Norman Schofi eld · Gonzalo Caballero · Daniel Kselman Editors Kselman · CaballeroSchofield Advances in Political Economy Institutions, Modelling and Empirical Analysis Eds.

Th is book presents latest research in the fi eld of Political Economy, dealing with the integration of economics and politics and the way institutions aff ect social decisions. Th e focus is on innovative topics such as an institutional analysis based on case studies; Norman Schofi eld the infl uence of activists on political decisions; new techniques for analyzing elections, involving game theory and empirical methods. Gonzalo Caballero Daniel Kselman Editors 1 Advances in Advances in Political Economy Political Economy Institutions, Modelling and Empirical Analysis

Social Sciences / Political Science

ISBN 978-3-642-35238-6

9 783642 352386 Advances in Political Economy

Book ID: , Date: 2013-02-21, Proof No: 2 PDF-OUTPUT Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF Norman Schofield r Gonzalo Caballero r Daniel Kselman Editors

Advances in Political Economy

Institutions, Modelling and Empirical Analysis

Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF Editors Prof. Norman Schofield Daniel Kselman, PhD. Center in Political Economy Centro de Estudios Avanzados Washington University in St. Louis en Ciencias Sociales Saint Louis, MO, USA Juan March Institute Madrid, Spain Assoc. Prof. Gonzalo Caballero Faculty of Economics University of Vigo Vigo, Spain

ISBN 978-3-642-35238-6 ISBN 978-3-642-35239-3 (eBook) DOI 10.1007/978-3-642-35239-3 Springer Heidelberg New York Dordrecht London

Library of Congress Control Number: 2013933596

© Springer-Verlag Berlin Heidelberg 2013 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s location, in its current version, and permission for use must always be obtained from Springer. Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. While the advice and information in this book are believed to be true and accurate at the date of pub- lication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF Introduction

Political Economy is both a growing field and a moving target. The concept ‘politi- cal economy’ remains something of an open signifier, alternatively used to describe a methodological approach in political analysis, grounded in the application of for- mal and quantitative methods to the study of politics; or one of any number sub- stantive areas in the contemporary social sciences. In economics, new institutional economics (Williamson 1985, North 1990) has established the fundamental impor- tance of history- and polity-specific governance structures in sustaining economic markets. Comparative research has investigated the effect of democratic institutions and processes on economic policy and outcomes, research given perhaps its most comprehensive statement in Persson and Tabellini (2000) and Drazen (2001), which have constituted the so-called “macroeconomics side” of political economy (Merlo 2006). Development economists increasingly recognize that, absent sound gover- nance institutions, standard macroeconomic prescriptions for economic growth and stability often fail to bear fruit (Rodrik 2007). Economists have also recently joined political scientists in examining the role of economic factors in explaining demo- cratic transitions and the evolution of political regimes (Acemoglu and Robinson 2000, 2006). Dewan and Shepsle (2008) have emphasized that in recent years some of the best theoretical work on the political economy of political institutions and processes has begun surfacing in the political science mainstream, and they con- sider that this is a result of economists coming more firmly to the conclusion that modeling governments and politicians is central to their own enterprise. Moving to political science, work on the modernization hypothesis, motivated by the consistently high cross-national correlation between democratic consolida- tion and economic development, has also recognized the role of economic factors in determining the evolution of political regimes (Moore 1965; Przeworski et al. 2000). Furthermore, comparative political science in many ways beat economics to the punch in recognizing the role that political institutions play in determining the economic trajectories of developing and still industrializing economies (Haggard and Kaufmann 1990). Economic class structures, and their embodiment in labor unions and professional organizations, have occupied an important place in compar- ative politics research on the economic institutions of advanced industrial societies

v

Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF vi Introduction

(Hall and Soskice 2001). Studies of voter behavior have identified both the role that conjuntural economic factors play in informing voter choice and the relationship between voters’ professional context and their preferences for redistribution. As al- ready mentioned, the label political-economy also refers more loosely to the appli- cation of formal and game theoretic methods first developed by economists to the study of political phenomena, including legislative bargaining (Shepsle 1979; Kre- hbiel 1998), government coalition formation (Laver and Schofield 1990; Laver and Shepsle 1996), and campaign position-taking (Cox 1987, 1990; Schofield 2006). In this sense, the effect of economics has been felt more strongly in contemporary political science than any other social science (Miller 1997). As evidenced by this brief, and necessarily incomplete, literature review, polit- ical economy is a concept with fairly flexible boundaries, encompassing research from a wide variety of fields and approaches. For example, Weingast and Wittman (2008) viewed political economy as the methodology of economics applied to the analysis of political behavior and institutions, but they assumed that it is not a sin- gle approach because it consists of a family of approaches. Previously, two views had been distinguished in the new political economy, and both have contributed to the advance of the understanding of modern political economy: on the one hand, Hamiltonian political economy has been interested in economic patterns and perfor- mance, but it considers that political institutions and political choices are relevant explaining factors; on the other hand, Madisonian political economy has assumed that the economic approach is central in political analysis, quite apart from eco- nomic content (Shepsle 1999). Rather than an explicit “field” or “discipline” in and of itself, the notion of political economy represents rather a growing awareness in both political science and economics that their respective contributions to our un- derstanding of society are intelligible only in mutual conversation. It is one thing for scholars in both disciplines to recognize the interdependence of their subject matters; it is another to create professional fora in which practitioners of these two disciplines come together. The current volume results from the latest in a series of conferences designed to engender a closer collaboration between economists and political scientists. Its contributions represent a broad spectrum of research, and its contributors a diverse group of scholars from diverse academic traditions in political economy. Nonetheless, as a group we share a commitment to mutually beneficial in- terdisciplinary collaboration, such it has been shown in previous efforts (Schofield and Caballero 2011). These conferences took place in April and May of 2012. The first was held at the Juan March Institute in Madrid, Spain, and was entitled Contemporary Applications of the Spatial Model. Ever since Downs’ seminal work (1957), the spatial model has been a workhorse in formal political theory. While its core content addresses how parties choose the relative extremism or moderation of campaign positions, its re- sults have also been used in studies of economic policy and redistribution (Meltzer and Richard 1978; Persson and Tabellini 2000). The Madrid conference brought together a group of leading scholars working on contemporary applications of the spatial paradigm, including theoretical contributions on spatial consequences of pri- mary elections and the spatial consequences of vote buying; and empirical contri- butions on the measurement of parties actual policy positions, the extent to which

Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF Introduction vii

voters accurately perceive such positions, and how these perceptions are moulded by voters’ ideological predispositions. The second conference was held in Baiona, Spain, and supported by the Erenea Research Group at the University of Vigo, and the Center in Political Economy at Washington University in Saint Louis. This conference was in fact the second installment of the International Conference on Political Economy and Institutions (ICOPEAI); and like the first, which was held in June 2010, it brought together political scientists and economists from many countries. The spatial model featured prominently in Baiona as well; but to this agenda was added a variety of papers on political transitions, democratic performance and human capital formation, social networks, and new institutional economics, and voting. There was substantial overlap in the participants at both conferences, allowing for a fruitful extended dialogue that, along with an internal peer-review process, has improved the content of the volume’s contributions. The editors thank the University of Vigo, the Juan March Institute, and the Cen- ter in Political Economy, Washington University in Saint Louis for the support they provided. In addition, an earlier version of Chap. 4 was presented at the conference on the Political Economy of Democratic Institutions, organised by Lauretta Frederk- ing at the University of Portland, June 2009. We thank Lauretta and the University of Portland for organising this earlier conference. We have decided to structure the volume in three sections, each dealing with a particular emphasis in political economic research: Institutions, Modelling, and Empirical Analysis. Each chapter in this book went through a review process before publication. These chapters deal with theoretical and empirical issues over the behavior of in- stitutions and the operation of democratic elections. Saint Louis, USA Norman Schofield Vigo, Spain Gonzalo Caballero Madrid, Spain Daniel Kselman

Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF Further Reading on Political Economy

Acemoglu D (2008) Oligarchic versus democratic societies. J Eur Econ Assoc 6:1Ð44 Acemoglu D, Robinson J (2000) Why did the West extend the franchise? Growth, inequality and democracy in historical perspective. Q J Econ 115:1167Ð1199 Acemoglu D, Robinson J (2006) Economic origins of dictatorship and democ- racy. Cambridge University Press, Cambridge Acemoglu D, Robinson J (2012) Why nations fail. Profile, New York Acemoglu D, Johnson S, Robinson J (2001) The colonial origins of comparative development. Am Econ Rev 91:1369Ð1401 Acemoglu D, Johnson S, Robinson J (2002) Reversal of fortune. Q J Econ 118: 1231Ð1294 Acemoglu D, Johnson S, Robinson J (2004) Institutions as the fundamental cause of long run growth. NBER, Washington Acemoglu D, Johnson S, Robinson J (2005) The rise of Europe: Atlantic trade, institutional change, and economic growth. Am Econ Rev 95:546Ð579 Acemoglu D, Johnson S, Robinson J, Yared P (2008) Income and democracy. Am Econ Rev 98:808Ð842 Acemoglu D, Johnson S, Robinson J, Yared P (2009) Reevaluating the modern- ization hypothesis. J Monet Econ 56:1043Ð1058 Boix C (2003) Democracy and redistribution. Cambridge University Press, Cam- bridge Buchanan JM, Tullock G (1962) The calculus of consent: logical foundations of constitutional democracy. University of Michigan Press, Ann Arbor Bunce V, Wolchik S (2010) Democracy and authoritarianism in the post commu- nist world. Cambridge University Press, Cambridge Coase RH (1984) The new institutional economics. J Inst Theor Econ 140:229Ð 231 Clark G (2007) A farewell to alms. Princeton University Press, Princeton Collier P (2007) The bottom billion. Oxford University Press, Oxford Cox GW (1987) Electoral equilibria under alternative voting institutions. Am J Polit Sci 31:82Ð108

ix

Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF x Further Reading on Political Economy

Cox GW (1990) Centripetal and centrifugal incentives in electoral systems. Am J Polit Sci 34:903Ð935 Dewan T, Shepsle K (2008) Recent economic perspectives on political economy, Part I. Br J Polit Sci 38:363Ð382. Downs A (1957) An economic theory of democracy. Harper Collins, New York Drazen A (2001) Political economy in macroeconomics. Princeton University Press, Princeton Easterly W (2007) Globalization, poverty and all that: factor endowment versus productivity views. In: Harrison A (ed) Globalization and poverty. Chicago Univer- sity Press, Chicago Eggertsson T (1990) Economic behaviour and institutions. Cambridge University Press, Cambridge Epstein D, Bates R, Goldstone J, Kristensen I, O’Halloran S (2006) Democratic transitions. Am J Polit Sci 50:551Ð568 Greif A (2006) Institutions and the path to modern economy. Cambridge Univer- sity Press, Cambridge Haggard SM, Kaufmann RR (1992) The politics of adjustment. Princeton Uni- versity Press, Princeton Hall PA, Soskice D (eds) (2001) Varieties of capitalism. Oxford University Press, Oxford Jones CI, Romer P (2010) The Kaldor facts: ideas, institutions, population and human capital. Am Econ J Macroecon 2:224Ð245 Kingston C, Caballero G (2009) Comparing theories of institutional change. J Inst Econ 5(2):151Ð180 Kitschelt H, Mansfeldova Z, Markowski R, Tóka G et al (1999) Post-communist party system. Cambridge University Press, Cambridge Kornai J, Eggleston K (2001). Welfare, choice, and solidarity in transition: re- forming the health sector in Eastern Europe. Cambridge University Press, Cam- bridge Krehbiel (1998) Pivotal politics. A theory of U.S. lawmaking. University of Chicago Press, Chicago. Laver M, Schofield N (1990) Multiparty government. The politics of coalition in Europe. University of Michigan Press, Ann Arbor Laver M, Shepsle KA (1996) Making and breaking governments: cabinets and legislatures in parliamentary democracies. Cambridge University Press, Cambridge Libecap GD (1989) Contracting for property rights. Cambridge University Press, Cambridge Meltzer AH, Richard SF (1978) Why government grows (and grows) in a democ- racy. Public Interest 52:111Ð18 Menard C, Shirley MM (eds) (2005) Handbook of new institutional economics. Springer, Berlin Merlo A (2006) Wither political economy? Theories, facts and issues. In: Blun- dell R, Newey W, Persson T (eds) Advances in economics and econometrics. Theory and applications. Cambridge University Press, Cambridge, pp 381Ð421 Miller GJ (1997) The impact of economics on contemporary political science, J Econ Lit XXXV:1173Ð1204

Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF Further Reading on Political Economy xi

Mokyr J (2010) The enlightened economy: an economic history of Britain 1700Ð 1850. Yale University Press, New Haven Moore B (1965) Social origins of dictatorship and democracy: lord and peasant in the making of the modern world. Beacon Press, Boston North DC (1990) Institutions, institutional change and economic performance. Cambridge University Press, Cambridge North DC (2005) Understanding the process of economic change. Princeton Uni- versity Press, Princeton North DC, Wallis JJ, Weingast BR (2009) Violence and social orders: a con- ceptual framework for interpreting recorded human history. Cambridge University Press, Cambridge Ostrom E (1990) Governing the commons. Cambridge University Press, Cam- bridge Ostrom E (2005) Understanding institutional diversity. Princeton University Press, Princeton Persson T, Tabellini G (2000) Political economics: explaining economic policy. MIT Press, Cambridge Persson T, Tabellini G (2003) The economic effect of constitutions. MIT Press, Cambridge Przeworski A (1991) Democracy and the market: political and economic reforms in Eastern Europe and Latin America. Cambridge University Press, Cambridge Przeworski A (2006) Democracy and economic development. In: Mansfield E, Sisson R (eds) Political science and the public interest. Ohio State University Press, Columbus Przeworski A, Alvarez ME, Cheibub JA, Limongi F (2000) Democracy and de- velopment: political institutions and well-being in the world, 1950Ð1990. Cam- bridge University Press, Cambridge Rawls J (1971) A theory of justice. Harvard University Press, Cambridge Rodrik D (2007) One economics, many recipes. Globalization, institutions and economic growth. Princeton University Press, Princeton Schofield N (2006) Architects of political change: constitutional quandaries and social choice theory. Cambridge University Press, Cambridge Schofield N (2007) The mean voter theorem: necessary and sufficient conditions for convergent equilibrium. Rev Econ Stud 74:965Ð980 Schofield N (2009) The political economy of democracy and tyranny. Olden- bourg, Munich. Schofield N, Caballero G (eds) (2011) Political economy of institutions, democ- racy and voting. Springer, Berlin. Schofield N, Gallego M (2011) Leadership or chaos. Springer, Berlin Schofield N, Gallego M, Jeon J (2011a) Leaders, voters and activists in elections in the Great Britain 2005 and 2010. Elect Stud 30(3):484Ð496 Schofield N, Gallego M, Ozdemir U, Zakharov A (2011b) Competition for pop- ular support: a valence model of elections in Turkey. Soc Choice Welf 36(3Ð4):451Ð 482 Serra G (2011) Why primaries? The party’s trade-off between policy and valence. J Theor Polit 23(1):21Ð51

Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF xii Further Reading on Political Economy

Shepsle KA (1979) Institutional arrangements and equilibrium in multidimen- sional voting models. Am J Polit Sci 23:23Ð57 Shepsle K (1999) The political economy of state reform. Political to the core. Braz J Pol Econ 19:39Ð58 Sokoloff KL, Engerman SL (2000) Institutions, factor endowments and the paths of development in the new world. J Econ Perspect 14:217Ð232 Weingast BR, Wittman DA (2008) The reach of political economy. In: Weingast BR, Wittman DA (eds) The Oxford handbook of political economy. Oxford Univer- sity Press, Oxford, pp 3Ð25 Williamson OE (1985) The economic institutions of capitalism: firms, markets, relational contracting. The Free Press, New York Williamson OE (2000) The new institutional economics: taking stock, looking ahead. J Econ Lit 38:595Ð613

Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF Contents

Part I Institutions Transaction Cost Politics in the Map of the New Institutionalism ..... 3 Gonzalo Caballero and Xosé Carlos Arias Political Transitions in Ancient Greece and Medieval Italy: An Analytic Narrative ...... 31 Leandro De Magalhães A Collective-Action Theory of Fiscal-Military State Building ...... 47 LuzMarinaArias Stable Constitutions in Political Transition ...... 67 Katja Michalak and Gerald Pech Quandaries of Gridlock and Leadership in US Electoral Politics ..... 91 Evan Schnidman and Norman Schofield Sub-central Governments and Debt Crisis in Spain over the Period 2000–2011 ...... 129 Fernando Toboso Deciding How to Choose the Healthcare System ...... 145 Olga Shvetsova and Katri K. Sieberg

Part II Modelling Challenges to the Standard Euclidean Spatial Model ...... 169 Jon X. Eguia A Non-existence Theorem for Clientelism in Spatial Models ...... 181 Daniel Kselman Nonseparable Preferences and Issue Packaging in Elections ...... 203 Dean Lacy and Emerson M.S. Niou

xiii

Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF xiv Contents

When Will Incumbents Avoid a Primary Challenge? Aggregation of Partial Information About Candidates’ Valence ...... 217 Gilles Serra Measuring the Latent Quality of Precedent: Scoring Vertices in a Network ...... 249 John W. Patty, Elizabeth Maggie Penn, and Keith E. Schnakenberg

Part III Empirical Analysis The Politics of Austerity: Modeling British Attitudes Towards Public Spending Cuts ...... 265 Harold D. Clarke, Walter Borges, Marianne C. Stewart, David Sanders, and Paul Whiteley Modeling Elections with Varying Party Bundles: Applications to the 2004 Canadian Election ...... 289 Kevin McAlister, Jee Seon Jeon, and Norman Schofield Spatial Model of Elections in Turkey: Tracing Changes in the Party System in the 2000s ...... 317 Norman Schofield and Betul Demirkaya Do Competitive Districts Necessarily Produce Centrist Politicians? ....331 James Adams, Thomas L. Brunell, Bernard Grofman, and Samuel Merrill III A Heteroscedastic Spatial Model of the Vote: A Model with Application to the United States ...... 351 Ernesto Calvo, Timothy Hellwig, and Kiyoung Chang Inferring Ideological Ambiguity from Survey Data ...... 369 Arturas Rozenas Biographies ...... 383

Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF Contributors

James Adams Department of Political Science, University of California, Davis, Davis, CA, USA Luz Marina Arias CEACS, Juan March Institute, Madrid, Spain Xosé Carlos Arias Faculty of Economics, University of Vigo, Vigo, Spain Walter Borges Division of Liberal Arts and Life Sciences, University of North Texas-Dallas, Dallas, TX, USA Thomas L. Brunell School of Economic, Political and Policy Sciences, University of Texas at Dallas, Richardson, TX, USA Gonzalo Caballero Faculty of Economics, University of Vigo, Vigo, Spain Ernesto Calvo Government and Politics, University of Maryland, College Park, MD, USA Kiyoung Chang Government and Politics, University of Maryland, College Park, MD, USA Harold D. Clarke School of Economic, Political and Policy Sciences, University of Texas at Dallas, Richardson, TX, USA; Department of Government, University of Essex, Colchester, UK Leandro De Magalhães Department of Economics, University of Bristol, Bris- tol, UK Betul Demirkaya Center in Political Economy, Washington University in Saint Louis, Saint Louis, MO, USA Jon X. Eguia Department of Politics, New York University, New York, NY, USA Bernard Grofman Department of Political Science and Center for the Study of Democracy, University of California, Irvine, Irvine, CA, USA

xv

Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF xvi Contributors

Timothy Hellwig Department of Political Science, Indiana University, Blooming- ton, IN, USA Jee Seon Jeon Center in Political Economy, Washington University in Saint Louis, Saint Louis, MO, USA Daniel Kselman Center for Advanced Studies in the Social Sciences, Juan March Institute, Madrid, Spain Dean Lacy Department of Government, Dartmouth College, Hanover, NH, USA Kevin McAlister Center in Political Economy, Washington University in Saint Louis, Saint Louis, MO, USA Samuel Merrill Department of Mathematics and Computer Science, Wilkes Uni- versity, Wilkes-Barre, PA, USA Katja Michalak Department of Public Management and Governance, Zeppelin University, Friedrichshafen, Germany Emerson M.S. Niou Department of Political Science, Duke University, Durham, NC, USA John W. Patty Center in Political Economy, Washington University in Saint Louis, Saint Louis, MO, USA Gerald Pech Department of Economics, KIMEP University, Almaty, Republic of Kazakhstan Elizabeth Maggie Penn Center in Political Economy, Washington University in Saint Louis, Saint Louis, MO, USA Arturas Rozenas ISM University of Management and Economics, Vilnius, Lithua- nia David Sanders Department of Government, University of Essex, Colchester, UK Keith E. Schnakenberg Center in Political Economy, Washington University in Saint Louis, Saint Louis, MO, USA Evan Schnidman Cambridge, MA, USA Norman Schofield Weidenbaum Center, Washington University in St. Louis, St. Louis, MO, USA Gilles Serra Department of Political Science, Center for Economics Research and Teaching (CIDE), Mexico City, Mexico Olga Shvetsova Department of Political Science, Binghamton University, Bing- hamton, NY, USA Katri K. Sieberg Department of North American Studies, University of Tampere, Tampereen Yliopisto, Finland

Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF Contributors xvii

Marianne C. Stewart School of Economic, Political and Policy Sciences, Univer- sity of Texas at Dallas, Richardson, TX, USA Fernando Toboso Faculty of Economics, Departamento de Economía Aplicada, University of Valencia, Valencia, Spain Paul Whiteley Department of Government, University of Essex, Colchester, UK

Book ID: , Date: 2013-02-21, Proof No: 2, UNCORRECTED PROOF Metadata of the book and chapters that will be visualized online

Please note: Images will appear in color online but will be printed in black and white.

Book title Advances in Political Economy

Book subtitle Institutions, Modelling and Empirical Analysis

Book copyright year 2013

Book copyright holder Springer-Verlag Berlin Heidelberg

Chapter title Transaction Cost Politics in the Map of the New Institutionalism

Corresponding Author Family name Caballero

Particle

Given Name Gonzalo

Suffix

Division Faculty of Economics

Organization University of Vigo

Address Campus As Lagoas-Marcosende, 36310, Vigo, Spain

E-mail [email protected]

Author Family name Arias

Particle

Given Name Xosé

Given Name Carlos

Suffix

Division Faculty of Economics

Organization University of Vigo

Address Campus As Lagoas-Marcosende, 36310, Vigo, Spain

E-mail [email protected]

Abstract In recent decades, the new institutionalism has strongly emerged in social sciences. Institutions have come back to the main research agenda in economics, politics and sociology. This paper presents and analyzes the program of Transaction Cost Politics within the map of the new institutionalism. Transaction Cost Politics constitutes an extension of the New Institutional Economics towards the analysis of politics, and it points out the relevance of institutions in political markets that are characterized by incomplete political rights, imperfect enforcement of agreements, bounded rationality, imperfect information, subjective mental models on the part of the individuals and high transaction costs. The paper reviews the main contributions of Transaction Cost Politics and we study the relationships of Transaction Cost Politics with Rational-Choice Institutionalism, Constitutional Political Economy and the New Institutional Economics.

Chapter title Political Transitions in Ancient Greece and Medieval Italy: An Analytic Narrative

Corresponding Author Family name Magalhães

Particle De

Given Name Leandro

Suffix

Division Department of Economics

Organization University of Bristol

Address 8 Woodland Road, Bristol, BS8 1TN, UK

E-mail [email protected]

Abstract Models of political transitions have mostly focused on the 19th and 20th centuries. Their setup tends to be specific to the contemporary period. This chapter reviews the events that led to democracy in ancient Athens and to rule by council in medieval Venice. We confront the available models of political transition with these events. We find evidence that war and economic conditions played a key role. The political economy models that incorporate these features do well in explaining the transitions in both ancient Athens and medieval Venice.

Keywords Political transitions – Wars – Ancient Greece – Athens – Venice – Genoa – Democracy – Republic

Chapter title A Collective-Action Theory of Fiscal-Military State Building

Corresponding Author Family name Arias

Particle

Given Name Luz

Given Name Marina

Suffix

Division CEACS

Organization Juan March Institute

Address C/Castello 77, Madrid, 28006, Spain

E-mail [email protected]

Abstract Prior to the emergence of the fiscal-military state, many monarchs depended on economic and local elites for the collection of tax revenue and defense. Why did these powerful elites allow the ruler to increase fiscal centralization and build-up militarily? Building on historical accounts of colonial Mexico and 17th century England, this chapter develops a game-theoretic analysis that explains why increases in fiscal centralization are more likely when the probability of a threat of internal unrest or external invasion increases. Elites free ride on fiscal contributions under fragmented fiscal capacity. Centralized fiscal collection and enforcement serves as an institutional devise for the elites to overcome free riding and ensure the provision of military protection. The analysis shows that an increase in the probability of a threat is more likely to result in centralization when the alignment between the elites’ and the ruler’s vulnerability to the threat is high, and in the presence of economic growth. The analysis also suggests that institutions that allow rulers to commit, such as representative assemblies, may not be necessary for fiscal centralization to transpire. Examples from European and colonial history provide support for the implications of the theoretical analysis.

Chapter title Stable Constitutions in Political Transition

Corresponding Author Family name Michalak

Particle

Given Name Katja

Suffix

Division Department of Public Management and Governance

Organization Zeppelin University

Address Am Seemooser Horn 20, 88045, Friedrichshafen, Germany

E-mail [email protected]

Author Family name Pech

Particle

Given Name Gerald

Suffix

Division Department of Economics

Organization KIMEP University

Address Abay 2, 050010, Almaty, Republic of Kazakhstan

E-mail [email protected]

Abstract This chapter develops a spatial model where an autocrat selects a status quo constitution which a succeeding elected constitutional assembly may or may not accept as a blue print for negotiations on constitutional reform. If the autocrat expects that the future constitutional assembly is dominated by parties which favor redistribution, he does not want to bind himself by the constitution. If the middle-class opposes redistribution or the middle class and the right dominate the constitutional assembly, stable constitutions exist which are in the interest of the autocrat. This framework is applied to transition processes in Chile and Egypt.

Chapter title Quandaries of Gridlock and Leadership in US Electoral Politics Corresponding Author Family name Schnidman

Particle

Given Name Evan

Suffix

Division

Organization

Address 11 Vandine Street #2, Cambridge, MA, 02141, USA

E-mail [email protected]

Author Family name Schofield

Particle

Given Name Norman

Suffix

Division Weidenbaum Center

Organization Washington University in St. Louis

Address Seigle Hall, Campus Box 1027, One Brookings Drive, St. Louis, MO, 63130-4899, USA

E-mail [email protected]

Abstract In 1964 President Johnston was able to overcome Southern Democrat opposition to the Civil Rights legislation. Recent opposition by Republicans in Congress has induced a form of legislative gridlock, similar to the situation facing Johnston. This paper argues that the current gridlock is more pernicious than in 1964 for two reasons. The pivot line in the two dimensional policy space has shifted slightly so that voters are more clearly separated by different preferences on civil rights. Secondly the era of deregulation since the election of Reagan has brought money into the political equation, especially since Citizen’s United decision of the Supreme Court. The argument is based on a formal model of the 2008 election and shows that excluding money, both candidates in 2008 would have adopted centrist positions. We argue that it was money that pulled the candidates into opposite quadrants of the policy space. We suggest that the same argument holds for members of Congress leading to the current gridlock. Before discussing the current gridlock between the executive and legislative arms of government we draw some parallels with earlier episodes in US political history, particularly the early years of the Roosvelt presidency and the lead-up to the passage of the Civil Rights legislation in 1964. We also suggest that in fragmented or multiparty systems, based on proportional representation, such as in the euro area, small parties will adopt radical policies far from the electoral center, thus inducing coalition instability. This phenomenon coupled with a fragile fiscal system based on the euro also has created difficulties in dealing effectively with the fall-out from the recession of 2008–2009.

Chapter title Sub-central Governments and Debt Crisis in Spain over the Period 2000–2011

Corresponding Author Family name Toboso

Particle

Given Name Fernando

Suffix

Division Faculty of Economics, Departamento de Economía Aplicada

Organization University of Valencia

Address Av. Tarongers s/n, 46022, Valencia, Spain

E-mail [email protected]

Abstract This chapter studies the quantitative evolution of sub-central sovereign debt in Spain over the period 2000–2011 and compares it with the evolution of central debt. As an intense process of political and fiscal decentralization has taken place since the mid eighties, the paper examines whether this drive to decentralization has been paralleled by any fiscally undisciplined behavior on the part of Spanish sub-central governments over the period considered. Some key formal legal rules and informal behavioral norms present at sub-central politics in Spain are examined, including legal controls on borrowing by sub-central governments. The empirical analysis will be based on the internationally comparable public finance figures provided by sources such as the OECD, the Eurostat and the Bank of Spain. The paper concludes that economic performance seem to be the key factor for explaining the evolution of sub-central, as well as central, public debt before and after the world financial crash. The analysis shows that in terms of the Spanish GDP the debt burden generated by sub-central governments in Spain decreased over the 2000–2007 period. However, this debt has soared from 8.5 per cent of Spanish GDP in 2007 to 16.4 per cent in 2011, adding 85 thousand millions euros (about 106 billions US dollars) to the stock of total public debt in Spain in just four years. Central government added 267 thousand millions euros (about 334 billions US dollars).

Chapter title Deciding How to Choose the Healthcare System

Corresponding Author Family name Shvetsova

Particle

Given Name Olga

Suffix

Division Department of Political Science

Organization Binghamton University

Address P.O. Box 6000, Binghamton, NY, 13902, USA

E-mail [email protected]

Author Family name Sieberg Particle

Given Name Katri

Given Name K.

Suffix

Division Department of North American Studies

Organization University of Tampere

Address 33104, Tampereen Yliopisto, Finland

E-mail [email protected]

Abstract The continuing debate in the United States over the form of health care provision is illustrative as to how difficult that choice can be. The choice is further complicated by political activity—lobbyists with a vested interest in various formats—and a noticeable effect from path dependence—people are used to what they have and are afraid of change, and some groups actually stand to lose from change, at least in the short run. What might the decision have been in the absence of these effects? This chapter creates a model to explore this question. In particular, we appeal to insights from Buchanan and Tullock (1962), Rawls (1971) and Kornai and Eggleston (2001) to ask what type of health care provision would a polity choose from behind the veil of ignorance, and what type of mechanism—unanimity (constitutional) or majority (legislative) would they prefer to use to select it?

Chapter title Challenges to the Standard Euclidean Spatial Model

Corresponding Author Family name Eguia

Particle

Given Name Jon

Given Name X.

Suffix

Division Department of Politics

Organization New York University

Address 19 West 4th, 2nd floor, New York, NY, 10012, USA

E-mail [email protected]

Abstract Spatial models of political competition over multiple issues typically assume that agents’ preferences are represented by utility functions that are decreasing in the Euclidean distance to the agent’s ideal point in a multidimensional policy space. I describe theoretical and empirical results that challenge the assumption that quasiconcave, differentiable or separable utility functions, and in particular linear, quadratic or exponential Euclidean functions, adequately represent multidimensional preferences, and I propose solutions to address each of these challenges.

Chapter title A Non-existence Theorem for Clientelism in Spatial Models

Corresponding Author Family name Kselman

Particle

Given Name Daniel

Suffix

Division Center for Advanced Studies in the Social Sciences

Organization Juan March Institute

Address C/Castelló 77, 28006, Madrid, Spain

E-mail [email protected]

Abstract This chapter proposes a spatial model that combines both programmatic as well as clientelistic modes of vote-seeking. In the model political parties strategically choose: (1) their programmatic policy position, (2) the effort they devote to clientelism as opposed to the promotion of their programmatic position, and (3) the set of voters who are targeted to receive clientelistic benefits. I present a theorem which demonstrates that, in its most general form, a spatial model with clientelism yields either Downsian convergence without clientelist targeting, or an inifinite cycle. Put otherwise, in its most general form the model never yields a Nash Equilibrium with positive levels of clientelism. I relate this result to past research on instability in coalition formation processes, and then identify additional restrictions, regarding voter turnout and the set of voters which parties can target, which serve to generate Nash equilibria with positive clientelist effort.

Chapter title Nonseparable Preferences and Issue Packaging in Elections

Corresponding Author Family name Lacy

Particle

Given Name Dean

Suffix

Division Department of Government

Organization Dartmouth College

Address HB 6108, Hanover, NH, 03755, USA

E-mail [email protected]

Author Family name Niou

Particle

Given Name Emerson

Given Name M.

Given Name S.

Suffix Division Department of Political Science

Organization Duke University

Address Durham, NC, 27708-0204, USA

E-mail [email protected]

Abstract In this chapter we develop a model in which candidates have fixed positions on a single issue dimension on which one candidate has an advantage by being closer to the median voter. The disadvantaged candidate can introduce a new issue to win the election. When all voters have separable preferences and the advantaged candidate moves last on the new issue, there is no way for the disadvantaged candidate to win. When some voters have nonseparable preferences over the issues, the disadvantaged can take a position that the advantaged candidate cannot beat. Candidates in an election can benefit from introducing new issues, but only when some voters have nonseparable preferences. Using data from a 2004 survey, we show that a substantial percentage of US voters have nonseparable preferences for many issues of public policy, creating incentives and opportunities for political candidates to package issues.

Chapter title When Will Incumbents Avoid a Primary Challenge? Aggregation of Partial Information About Candidates’ Valence

Corresponding Author Family name Serra

Particle

Given Name Gilles

Suffix

Division Department of Political Science

Organization Center for Economics Research and Teaching (CIDE)

Address Carretera México-Toluca 3655, Lomas de Santa Fe, Mexico City, 01210, Mexico

E-mail [email protected]

Abstract When can a party insider feel safe from an outside challenge for a future nomination? In most countries, parties can choose whether to hold a primary election where the rank-and-file members take a vote, or to allow party leaders to directly appoint an insider candidate of their liking. The cost of primaries forces candidates to drift away from the party leader’s policy preferences in order to cater to primary voters. This paper postulates a benefit: primary elections can reveal information about the electability of potential candidates. I refine the formal model in Serra (2011) by making the realistic assumption that such information is revealed partially rather than fully. A signaling mechanism is introduced whereby candidates send noisy information that is used by primary voters to update their beliefs. This leads to surprising insights about the behavior of primary voters: under some circumstances they will use the information provided by primary campaigns, but under other circumstances, they will choose to completely ignore such information. In addition, the results predict that popular incumbents will not be challenged in a primary election, which is consistent with empirical observation. Finally, a prescription for parties is to allow their primaries to be tough given that stiff competition will improve the expected ability of the nominee.

Chapter title Measuring the Latent Quality of Precedent: Scoring Vertices in a Network

Corresponding Author Family name Patty

Particle

Given Name John

Given Name W.

Suffix

Division Center in Political Economy

Organization Washington University in Saint Louis

Address 1 Brookings Drive, Saint Louis, MO, 63130, USA

E-mail [email protected]

Author Family name Penn

Particle

Given Name Elizabeth

Given Name Maggie

Suffix

Division Center in Political Economy

Organization Washington University in Saint Louis

Address 1 Brookings Drive, Saint Louis, MO, 63130, USA

E-mail [email protected]

Author Family name Schnakenberg

Particle

Given Name Keith

Given Name E.

Suffix

Division Center in Political Economy

Organization Washington University in Saint Louis

Address 1 Brookings Drive, Saint Louis, MO, 63130, USA

E-mail [email protected]

Abstract In this chapter, we consider the problem of estimating the latent influence of vertices of a network in which some edges are unobserved for known reasons. We present and employ a quantitative scoring method that incorporates differences in “potential influence” between vertices. As an example, we apply the method to rank Supreme Court majority opinions in terms of their “citability,” measured as the likelihood the opinion will be cited in future opinions. Our method incorporates the fact that future opinions cannot be cited in a present-day opinion. In addition, the method is consistent with the fact that a judicial opinion can cite multiple previous opinions.

Chapter title The Politics of Austerity: Modeling British Attitudes Towards Public Spending Cuts

Corresponding Author Family name Clarke

Particle

Given Name Harold

Given Name D.

Suffix

Division School of Economic, Political and Policy Sciences

Organization University of Texas at Dallas

Address PO Box 830688, Richardson, TX, 75083-0688, USA

Division Department of Government

Organization University of Essex

Address Colchester, C043SQ, UK

E-mail [email protected]

Author Family name Borges

Particle

Given Name Walter

Suffix

Division Division of Liberal Arts and Life Sciences

Organization University of North Texas-Dallas

Address 7300 University Blvd., Dallas, TX, 75241, USA

E-mail [email protected]

Author Family name Stewart

Particle

Given Name Marianne

Given Name C.

Suffix

Division School of Economic, Political and Policy Sciences

Organization University of Texas at Dallas

Address PO Box 830688, Richardson, TX, 75083-0688, USA E-mail [email protected]

Author Family name Sanders

Particle

Given Name David

Suffix

Division Department of Government

Organization University of Essex

Address Colchester, C043SQ, UK

E-mail [email protected]

Author Family name Whiteley

Particle

Given Name Paul

Suffix

Division Department of Government

Organization University of Essex

Address Colchester, C043SQ, UK

E-mail [email protected]

Abstract The fallout from the 2008 financial crises has prompted acrimonious national debates in many Western democracies over the need for substantial budget cuts. Among economic and political elites there is broad agreement that substantial public sector budget cuts are necessary to address unsustainable sovereign debt and to establish long-term fiscal integrity. Many ordinary citizens see things differently, since austerity measures threaten programs that challenge longstanding public commitments to education, health and personal security that constitute the foundation of the modern welfare state. We investigate the nature of public attitudes towards the budget cuts using surveys from the British Election Study. The results suggest that cuts currently are widely perceived by the public as essential for Britain’s long-term economic health. But an upward trending view that slashing public services will cause serious difficulties for families may lead many people eventually to say enough is enough. It is likely that support for the cuts will be undermined by a lack of visible results in the real economy.

Chapter title Modeling Elections with Varying Party Bundles: Applications to the 2004 Canadian Election

Corresponding Author Family name McAlister

Particle

Given Name Kevin

Suffix Division Center in Political Economy

Organization Washington University in Saint Louis

Address 1 Brookings Drive, Saint Louis, MO, 63130, USA

E-mail [email protected]

Author Family name Jeon

Particle

Given Name Jee

Given Name Seon

Suffix

Division Center in Political Economy

Organization Washington University in Saint Louis

Address 1 Brookings Drive, Saint Louis, MO, 63130, USA

E-mail [email protected]

Author Family name Schofield

Particle

Given Name Norman

Suffix

Division Weidenbaum Center

Organization Washington University in St. Louis

Address Seigle Hall, Campus Box 1027, One Brookings Drive, St. Louis, MO, 63130-4899, USA

E-mail [email protected]

Abstract Previous models of elections have emphasized the convergence of parties to the center of the electorate in order to maximize votes received. More recent models of elections demonstrate that this need not be the case if asymmetry of party valences is assumed and a stochastic model of voting within elections is also assumed. This model seems able to reconcile the widely accepted median voter theorem and the instability theorems that apply when considering multidimensional policy spaces. However, these models have relied on there being a singular party bundle offered to all voters in the electorate. In this paper, we seek to extend these ideas to more complex electorates, particularly those where there are regional parties which run for office in a fraction of the electorate. We derive a convergence coefficient and out forth necessary and sufficient conditions for a generalized vector of party positions to be a local Nash equilibrium; when the necessary condition fails, parties have incentive to move away from these positions. For practical applications, we pair this finding with a microeconometric method for estimating parameters from an electorate with multiple regions which does not rely on independence of irrelevant alternatives but allows estimation of parameters at both aggregate and regional levels. We demonstrate the effectiveness of this model by analyzing the 2004 Canadian election.

Chapter title Spatial Model of Elections in Turkey: Tracing Changes in the Party System in the 2000s

Corresponding Author Family name Schofield

Particle

Given Name Norman

Suffix

Division Weidenbaum Center

Organization Washington University in St. Louis

Address Seigle Hall, Campus Box 1027, One Brookings Drive, St. Louis, MO, 63130-4899, USA

E-mail [email protected]

Author Family name Demirkaya

Particle

Given Name Betul

Suffix

Division Center in Political Economy

Organization Washington University in Saint Louis

Address 1 Brookings Drive, Saint Louis, MO, 63130, USA

E-mail [email protected]

Abstract The Turkish political party system underwent significant changes during the first decade of the 21st century. While secularism and nationalism remained the defining issues of electoral politics, both the number and the ideological positions of parties in the political system changed considerably. In the 2002 elections, none of the parties from the previous parliament were able to pass the electoral threshold. The new parliament was formed by the members of the Justice and Development Party (AKP) —a new conservative party founded by the former members of Islamist parties—and the Republican People’s Party (CHP)—a party with a strong emphasis on a secularist agenda. In the 2007 elections, AKP consolidated their power by receiving 46.6 % of the votes while CHP increased their share of the vote by only 1.5 percentage points to 20.9 %. In addition, the Nationalist Action Party (MHP) and independent candidates supported by the pro-Kurdish Democratic Society Party (DTP) were able to win seats in the 2007 elections. In order to explain these changes, this paper applies the spatial model to the 2007 elections and compares the results to previous analyses of the 1999 and 2002 elections (Schofield et al. 2011). First, we run a pure spatial model to estimate the relative role of the ideological position and the valence of political parties in determining their electoral success. Second, we supplement the spatial model with the demographic characteristics of voters. Finally, we use simulations to determine whether a Nash equilibrium exists for the position of political parties or candidates.

Chapter title Do Competitive Districts Necessarily Produce Centrist Politicians?

Corresponding Author Family name Adams

Particle

Given Name James

Suffix

Division Department of Political Science

Organization University of California, Davis

Address Davis, CA, 95616, USA

E-mail [email protected]

Author Family name Brunell

Particle

Given Name Thomas

Given Name L.

Suffix

Division School of Economic, Political and Policy Sciences

Organization University of Texas at Dallas

Address 800 W. Campbell Road, Richardson, TX, 75080, USA

Author Family name Grofman

Particle

Given Name Bernard

Suffix

Division Department of Political Science and Center for the Study of Democracy

Organization University of California, Irvine

Address Irvine, CA, 92697-5100, USA

E-mail [email protected]

Author Family name Merrill

Particle

Given Name Samuel

Suffix III

Division Department of Mathematics and Computer Science

Organization Wilkes University

Address Wilkes-Barre, PA, 18766, USA E-mail [email protected]

Abstract Using the first dimension of DW nominate scores for the U.S. House and Senate over the period 1956–2004, we analyze how the degree of ideological polarization between the parties varies as a function of district ideology, defined in terms of Democratic presidential support in the district. We find, as expected, that the more Democratic-leaning the district at the presidential level the more liberal are the representatives from the district, and that for any given level of Democratic presidential support, Democrats elected from such districts are, on average, considerably more liberal than Republicans elected from such districts. However, we also find that—consistent with theoretical expectations of spatial models that have recently been put forward—the ideological difference between the winners of the two parties is as great or greater in districts that, in presidential support terms, are the most competitive—a finding that contradicts the intuitive expectation that the pressure for policy convergence is greatest when the election is most competitive.

Chapter title A Heteroscedastic Spatial Model of the Vote: A Model with Application to the United States

Corresponding Author Family name Calvo

Particle

Given Name Ernesto

Suffix

Division Government and Politics

Organization University of Maryland

Address 3144F Tydings Hall, College Park, MD, 20742, USA

E-mail [email protected]

Author Family name Hellwig

Particle

Given Name Timothy

Suffix

Division Department of Political Science

Organization Indiana University

Address Woodburn Hall 210, 1100 E Seventh Street, Bloomington, IN, 47405-7110, USA

E-mail [email protected]

Author Family name Chang

Particle

Given Name Kiyoung

Suffix Division Government and Politics

Organization University of Maryland

Address 3144F Tydings Hall, College Park, MD, 20742, USA

E-mail [email protected]

Abstract How do candidate policy positions affect the citizen’s vote choice? From the Downsian tradition, a common response to this question is that voters identify where contending candidates are located on policy space and then select the candidate closest to them. A well-known finding in current models of political psychology, however, is that voters have biased perceptions of the ideological location of competing candidates in elections. In this chapter we offer a general approach to incorporate information effects into current spatial models of voting. The proposed heteroscedastic proximity model (HPM) of voting incorporates information effects in equilibrium models of voting to provide a solution to common attenuation biases observed in most equilibrium models of vote choice. We test the heteroscedastic proximity model of voting on three U.S. presidential elections in 1980, 1996, and 2008.

Chapter title Inferring Ideological Ambiguity from Survey Data

Corresponding Author Family name Rozenas

Particle

Given Name Arturas

Suffix

Division

Organization ISM University of Management and Economics

Address LT-01129, Vilnius, Lithuania

E-mail [email protected]

Abstract The chapter presents a Bayesian model for estimating ideological ambiguity of political parties from survey data. In the model, policy positions are defined as probability distributions over a policy space and survey-based party placements are treated as random draws from those distributions. A cross-classified random-effects model is employed to estimate ideological ambiguity, defined as the dispersion of the latent probability distribution. Furthermore, non-response patterns are incorporated as an additional source of information on ideological ambiguity. A Markov chain Monte Carlo algorithm is provided for parameter estimation. The usefulness of the model is demonstrated using cross-national expert survey data on party platforms.

Keywords Ideological placement – Ambiguity – Bayesian – Latent variables – Missing data

Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 Institutions atI Part Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 DOI (eds.), al. et Schofield N. e-mail: Arias X.C. Spain Vigo, 36310 Lagoas-Marcosende, As Campus Vigo, e-mail: of University Economics, of Faculty h e ntttoaimi oilg spr fti mrigprdg ntesocial the in paradigm emerging Weingast. this Barry of and part Shepsle is Kenneth sociology scientists in Hall, political institutionalism Peter new includes Olsen, The Johan institutionalism Peters, new Coase, Guy the science as Ronald of political such as as literature far such the as authors In concerned, Ostrom. of is Elinor contributions and Williamson the Oliver devel- North, on been Douglass role based has notion, (NIE) economics, of Economics re- subject in Institutional ongoing the New oped an on The levels been institutions. applied has “institutions of and change there that theoretical and therefore the assume and at to analyzed both effort be begun search can have they sciences that social and matter” different The then. since into incorporated became also Nee this institutionalism and science, new (Brinton political sociology the entirely than and from be responsive changing, it far less quickly can is been was neither “it but had legitimate; that sociology completely Although and not ignored”. science” is political it consistent; in ( or appeared Olsen coherent the and has March of institutionalism that, one new to become prior “a had years Two institutions economics. of in economics areas liveliest the the to that address Society presidential Economic his in Royal affirmed (1986) Matthews mid-eighties, the During Introduction 1 Arias Carlos Xosé and Caballero Gonzalo New the of Map Institutionalism the in Politics Cost Transaction .Cbleo( 2012). Caballero Spain, G. (Baiona, International ICOPEAI Second Institutions, the and European and Economy the 2011) Political Society at France, on Choice workshop (Cargese, Conference Public specialized Economics the a Institutional of in New Meeting presented for Annual was School the version at renewed presented This was 2009). paper (USA, this of version initial An hr a enacnieal n oal nraei eerho institutions on research in increase notable and considerable a been has There 10.1007/978-3-642-35239-3_1 [email protected] [email protected] B ) · ..Arias X.C. dacsi oiia Economy Political in Advances 1998 pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , ). , stated 1984) 3 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 esadhwisiuin fetdefiinyi cnm.Cae( Coase economy. in (1990a efficiency North affected institutions how derstand of institutions and interaction political studying for transaction step therefore governance. fundamental attempt, an a only is are analysis over frequently made which transactions transactions, political and ments of outcome Tommasi the and policy, as (Spiller and time policies institutions public between view relationships should the of we understanding deeper a have to Weingast and (North making situations political different of analysis the al. in et space (Schofield policy dimensional higher a of nise uha h feto xrciepltcladeooi ntttos(Ace- institutions economic and political extractive of Robinson effect and the moglu as such Caballero issues and in (Schofield sciences boost- social thereby to the 2011 foundations rise among theoretical given relations new has interdisciplinary on sciences based ing social economy political in (Nee to concept approaches impetus key institutionalist a new of as variety institutions a on focus the and Walter Dimaggio, others. Paul among as Nee, such Victor authors and of Powell contributions the includes it and sciences, 4 hssne C lo st aemr es u ftepltclmresw observe. we markets political the of out sense more make to us allow TCP sense, this markets political of it model” into the theory of cost power (North explanatory transaction the incorporating increase by substantially model can choice transaction (Caballero rational of relationships standard economic notion the of the limits furthermore, the and surpassed mainstream costs economic the for nizations Levinson and lofrtoei hc l atcpnsaepltcas sdatwt yWigs and Weingast by with dealt as politicians, are citizens ( participants Marshall all between which out in those carried for ( transactions also North political both political for which of only politicians, functioning not and the so for is greater even This is markets. eco- relevance affects decisively their costs then transaction exchange of presence nomic actors the If the costs. of transaction part high the bounded and on agreements, models mental of subjective character- enforcement information, are imperfect thereby imperfect rationality, which rights, contracts, markets, political and political incomplete by in transactions ized institutions as of condition relevance their the highlighting in relationships political of otPltc eerhporm(C)cridotoe h attet er (Wein- years twenty 1999 past Transaction the Marshall the over and through out carried processes gast (TCP) and program institutions research political Politics of Cost study the with dealt h oino rnato ot a h e ocp htteNEue oun- to used NIE the that concept key the was costs transaction of notion The policy- on effects systematic have arrangements institutional different The hs h rtr fisiuin”hsbcm nusinbei oilsciences, social in unquestionable become has institutions” of “return the Thus, C sspltcltascina h nto nlss n xlisteevolution the explains and analysis, of unit the as transaction political uses TCP .Ti oenpltcleooyo ntttoshsicue eeatadvances relevant included has institutions of economy political modern This ). Williamson ; 1990b 1988 nbe h utfiaino h motneo ntttosadorga- and institutions of importance the of justification the enabled ) .35.I hsmne,tenwtascinlisiuinls has institutionalism transactional new the manner, this In 355). p. , 2008 ,EsenadOHloa ( O’Halloran and Epstein ), 2011 1999 1988 ,tesuyo oilodr(Schofield order social of study the ), ,aogothers. among ), 2011 ple n Tommasi and Spiller ; ;North 2007 ,temdln fteatoiainrgms(Schofield regimes authoritarian the of modeling the ), 1989 .Pltcllf scaatrzdb xhne,agree- exchanges, by characterized is life Political ). 1990b agr n McCubbins and Haggard ; 1990b Dixit ; n ii (1996 Dixit and ) 1999 1996 2003 n ple n oms ( Tommasi and Spiller and ) 2005 , , 2007 2003 .Ti a rvddastrong a provided has This ). ). pti n O’Halloran and Epstein ; 2010 .CbleoadXC Arias X.C. and Caballero G. , 1998 2001 n h utilization the and ) 2001 mhszd but emphasized, ) .Bti ewant we if But ). 1937 .“Modifying ). , 1960 2007 and ) .In ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 oiia cnm.Tecnlsosaeotie tteedo h chapter. the of end the at outlined are conclusions The Economy. Political td rmTP Section TCP. from study Section constructed: (NIE). was Economics (TCP) Institutional Section New Politics the Cost and (RCI) Transaction Institutionalism which Rational-Choice on basis fundamental the way. Section comparative sciences. a in ap- and TCP other of analysis, of political with characteristics goal of analytical it main approach the relates positive shows The a and paper is TCP, arguments. this TCP of rivals. its and sources mention close theoretical both only the proaches, we contributions, searching main references, is the paper of of the rest contends relevant the review- most for When the day. and specify present the we to literature, up the TCP theoretical ing of the contributions studies main and the institutionalism and new foundations of places panorama paper current the the Moreover, within economy. TCP political in institutionalism transactional new rvdsacnrlrl otento fceil omtet hc utfisthe justifies which commitment, TCP credible (6) of transactions. notion political the inter-temporal to of of importance series role a central there- pub- of a of and result vision provides the progressive one a the in economic as witnessing complex are the policies more we in lic becomes times, recent structure than In institutional field (5) efficient world. political political an the of Transaction in design (4) the higher fore outcomes. be and to institutions tend between costs relations the explaining affect when institutions therefore and agents, policy the and public of game, structure political incentive the of or the rules determine (implicit trans- the they of of are outcome set Institutions the (2) a are policy-makers. as policies among public interaction actions sense, approach political this transactional consider In to relations. the contractual us of explicit) leads application field The political (1) the proposals: to following the by ized different the compares O’Halloran contracts, of mechanism enforcement governance the studies also unit, rnato otPltc nteMpo h e Institutionalism New the of Map the in Politics Cost Transaction e oszal rgeso o ntttosaeudrto.Nvrhls,teeis there Nevertheless, understood. are institutions new how an The on has theories. reopened progress and renewed have science, sizeable political on to institutions and based led sociology century, economics, sciences in 20th social emerged has the the institutionalism into of research decades for two agenda last the During Institutions of Definitions Sciences Social 2.1 the into Overview An Institutionalism: New 2 oiis Section Politics. Section a as Politics Cost Transaction of approach the analyzes and reviews chapter This rnato otPltc TP,bsdscnieigtecnrc sa analysis an as contract the considering besides (TCP), Politics Cost Transaction 4 6 tde h udmna ruet n otiuin fTascinCost Transaction of contributions and arguments fundamental the studies nlzstegvrac fpltcltascin nCnrs sacase- a as Congress in transactions political of governance the analyzes 2 1999 tutrsadaot h one ainlt upsto Esenand (Epstein supposition rationality bounded the adopts and structures rsnssvrlapoce fnwisiuinls ihntesocial the within institutionalism new of approaches several presents 3 raiainlsrcue fgvrac r ut relevant quite are governance of structures Organizational (3) outputs. euainlcapital reputational .Afis praht h hoeia ae fTPi character- is TCP of bases theoretical the to approach first A ). 5 3 rsnstetoapoce fnwisiuinls htformed that institutionalism new of approaches two the presents hw h rnato ot r ohg npltclmarkets. political in high so are costs transaction why shows 7 oprsteTPapoc ihta fConstitutional of that with approach TCP the compares n h raiainlfrua fteState. the of formulae organizational the and 5 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 lo o xhneoe pc n ie,adisiuin rdc netit by uncertainty (North “reduce exchange” institutions of and would structure time”, that stable and arrangements a institutional space creating of over set exchange a for of devise costs allow to the lower necessary to was order “In it theory. exchange, cost approach transaction This to economy. reference an specific of a enforcement structure assumes incentive and the rules so- provide informal they and rules, and economic formal mechanisms, political, of structure consist that Institutions constraints interaction. devised cial humanly the rules”. say, the to of is are “enforcement the society endogenizes a thereby in and game” motivation the of importance theory of the a “rules emphasizes of the approach how “institutions-as-equilibria” of the theory while selected, a on focuses approach rules (2011 Kingston “institutions-as-equilibria” and the Greif and ( approach. approach Caballero “institutions-as-rules” the and least incidental at Kingston the duce to and view According ideology view. the view, institutions conflict social the view, institutions cient epeetd o xml,Aeol n oisn( Robinson can institutions and of Acemoglu views example, different For several and presented. institutions, be of definition unique no 6 rdtosi cnmc,pltclsineadscooy lhuhteeaeseveral institutional are old there the although sociology, from and points. new distinguished connection science The be political economics, century—can last approaches. in the 20th several traditions during the using bases of theoretical out decades new carried two on developed be been can has institutions institutionalism—that of study The Approaches Institutional 2.2 ue yhwwl h aktapoiae eotascincs eut”(North results” cost transaction mea- zero “is a efficiency approximates of market level 1990b the the and well markets how political by of sured efficiency of level the mine t cin ntrso eain ewe oe n iutos.Ptr ( Peters situations”. and appropri- roles between define relations that of routines terms and in rules actions interrelated ate of “collections are institutions h utpedfiiin ftetrsisiuin sdi cnmc,pltclscience political economics, in of used sociology”. institutions many and terms “encompasses the is definition of which this definitions of multiple and each the factor”, organizations, social and nonphysical norms, man-made, beliefs, a rules, interrelated of (2006 systems Greif Avner as by analysis (Greif institutional in Analysis opments Institutional Comparative and al. et Historical Aoki game. among a meaning of and lutions values shared of individual sense affect institution. in- some must the An of be institution members (B) An should polity. (C) There and/or time. (D) An society over behavior. (A) the stability institution: some of political shows feature of stitution structural concept a the constitutes to characteristics institution key four adds further codn oteNrha prah ntttosaeterlso h ae that game, the of rules the are institutions approach, Northian the to According olwn h ntttosa-ue prah ac n le ( Olsen and March approach, institutions-as-rules the Following h ntttosa-qiiru prahdfie ntttosa qiiru so- equilibrium as institutions defines approach institutions-as-equilibrium The .360). p. , 2001 sue hsve fisiuin,atog eettertcldevel- theoretical recent although institutions, of view this assumed ) xeddta esetv:teinstitutions-as- the perspective: that extended ) 1990b .3)cnie “institutions consider 39) p. , .39.Isiuin deter- Institutions 359). p. , 2007 2009 .CbleoadXC Arias X.C. and Caballero G. itnus h effi- the distinguish ) ,w hudintro- should we ), 1989 1999 tt that state ) ,p.18) 1998 ; Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 iuls n h yohsso ainlt,gigbyn h one rationality. bounded the beyond going rationality, of hypothesis the and vidualism cnmsshsbe oie nrcn ie.Freape ot ( North example, For times. recent ( in Greif modified been has economists rnato otPltc nteMpo h e Institutionalism New the of Map the in Politics Cost Transaction niiul,woatwt nopeeifrain ncuaemna oesand models mental of inaccurate part information, (Nee the transactions incomplete on costly with action act purposive who presumes individuals, generally and tradition choice-theoretic of set a of up made was science political in tradition institutionalism old The (b) oee,i oie h supino ainlt n ute dsatm dimension theory, time a microeconomic adds of further of and tools (North rationality assumptions analytical of basic assumption the the the as modifies preserves it well however, it as and competence, theory, and neoclassical scarcity of modification a is (Eggertsson Indus- Systems, Economics New Economic and the Comparative Law History, Economics, and Economic ar- Cost New scientific Transaction the and Organization, different Analysis, institutional created trial Rights from of was Property arose relevance as but the contributions such institutionalism highlighted eas these old that and contributions the predeces- aspects, of immediate from organizational set their arise a in not to these did thanks for looking therefore be NIE not sors”. should we then less-interventionist antecedent, a shows it and models”, “economizing (Rutherford orientated character is and it choice reductionist, and rational individualistic of formalistic, to effects more the be to of institutional- appears “new study manner, ism” this the In efficiency. and on exchange frameworks institutional voluntary alternative of conceptualization the alism, c h ale oilgclisiuinls inee yTlotPros( Parsons Talcott by pioneered institutionalism sociological earlier The (c) Commons, John Veblen, (Thorstein economics in institutionalism original The (a) eetees h dao eiu itbtenteodadnwinstitutionalist new and old the between rift serious a of idea the Nevertheless, nteohrhn,NwIsiuinls ntesca cecsasmsthe assumes sciences social the in Institutionalism New hand, other the On neoois os (1984 Coase economics, In naporaefudto o h e ntttoa prahi oilg,where Society sociology, in and provides approach (1922—Economy sociology, Weber institutional of new the periods for modern foundation appropriate and an classical the in established analy- institutional analysis (1998 emphasize Nee not as did Just it but sis. institutions, of existence the sumed inlscooia praht oprtv ntttoa analysis. institutional comparative to approach sociological tional normative and historicism holism, structuralism, (Peters analysis legalism, as char- general such certain acteristics assumed and contributions heterogeneous multi-approach based analysis and theories power. its of devised conceptualization the and on structures institutional many the to of rejection economic their certain terms a as (Rutherford towards holistic interventionism tendency well their in as and argue criterion framework”, welfare to individualist behavioristic the tendency and a institutionalists “collectivist nature, old a anti-formalist such and an of contributions by The adopted marked analysis. and was holism analysis of neoclassical methods of the foundations the rejected Ayres) Clarence n srm( Ostrom and 2006) 1994 ). 1999 1994 1998 ). .5 onsottetaiino oprtv institutional comparative of tradition the out points 5) p. , ). 2007 .I ed omv oad ehdlgclindividu- methodological towards move to tends It ). utie ht“fmdr ntttoait a any had institutionalists modern “if that sustained ) upse h iiso h ehdlgclindi- methodological the of limits the surpassed ) 1994 .I a etrdo itiuieconsequences distributive on centered was It ). 1990 spoal h eteapeo h tradi- the of example best the probably is ) .Teaayia rmwr fteNIE the of framework analytical The ). 1994 1937)as- , 2005 ), 7 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 n alr( Taylor and 1 itrclIsiuinls eeoe nrsos otegoptere fpoli- of theories group the to response in developed perspectives, Institutionalism behavioral Historical the which (1) to of reaction each a sociology, as and being: institutionalism” these science “new political a in itself called approaches three existed there l mrcnisiuinls,sm set ftenwisiuinls r con- (Rutherford are years recent institutionalism in new institutionalism the old the of to the aspects back necting from some sharply concern institutionalism, themselves considerable American differentiate the old to of institutionalists spite economic In new institutionalism. among economic the with old convergence the possible a of towards thinking project institutionalist ( new Hodgson the of and evolution contributions, North’s the ( via al. institutionalism et Groenewegen sense, this In 8 3 oilgclisiuinls a endvlpdi oilg,epcal nor- in especially sociology, in developed been has institutionalism Sociological (3) American the of study the from arose (RCI) institutionalism choice Rational (2) iutnosywt h osldto fteNwIsiuinlEoois Hall Economics, Institutional New the of consolidation the with Simultaneously epae.()I mhszstehgl-neatv n mutually-constitutive and moral highly-interactive and the scripts cognitive emphasizes systems, It symbol as definition (B) such their templates. elements and of do, set scientists a political includes than define broadly institutionalists more Sociological much (A) institutions manner: following the in characterized h oreo oiisadcneunl faypseirpltcldcso (The- decision political posterior Steinmo any and of determine consequently len decisions and early political politics of of initial relevance course the the the history: emphasizes political approach in- throughout This polity. decisions and organiza- the formal the of in as structure embedded institutions conventions tional and defines norms it routines, procedures, and formal functionalism, structural and ties eue eentaotdt anefiiny u nta hudb considered and Hall be which should to (1996 pro- instead institutionalism, Taylor of and but efficiency, type forms This gain institutional culturally-specific-practices. to as the adopted of not many were that cedures considered It the theory. institutions. to the ganization reference by of affected by actors origin the institution the to the functions of to those of series respect by existence provided With a value the (D) as interaction explains outcomes. strategic politics RCI political of see institutions, role of to the determination tends emphasizes when the It rationality It in of (B) (C) model behavior. dilemmas. a human action incentives. employs collective explain and It (A) rules to follows: of tries as function it are a features is benefit its behavior their of that maximize Four sustains to and RCI try rules therefore individuals of which and system within a economics behavior as “new for institutions the incentives perceives from approach inputs This some organization”. received of it and ap- behavior this congressional to contributions main the of some provided proach. who Pierson and Thelen Skocpol, gov- ( Theda by Hall by made Peter coined has decisions was institu- initial term whose of an institutionalism, persistence generates Historical the which ernment. in dependence” results “path which a inertia, exists tional there that implies This 1996 noprt h otiuino ac n le ( Olsen and March of contribution the incorporate ) ttdta uigteegte n ieiso h 0hcentury, 20th the of nineties and eighties the during that stated ) 1986 1992 soeo t rnia rcros oee twsSteinmo, was it however precursors, principal its of one as ) Thelen ; 1995 1999 on oebigsbtennwadold and new between bridges some found ) ;Pierson 2000 iro n Skocpol and Pierson ; .CbleoadXC Arias X.C. and Caballero G. 1998 one u the out pointed ) 2001 ,cnbe can 1984), ). 2002 ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 h nlsso h cin fpltclprisaditrs rus nteohrhand, other the On groups. interest and parties political of actions the of analysis the etaie sbigls omladhgl ntttoaie,sc sKcete al. con- et Kickert are as that such politics institutionalized, in highly relations and ( formal many less are being there as that ceptualized between assuming Insti- relationships” society, Representation “institutionalized Interest and such hand, State of one structure the the On analyzes map. tutionalism the in incorporated been have rdtoaitapoc freape,tecnrbtoso odo isn,but (Peters old Wilson), analysis Woodrow empirical the advanced to of more nearer contributions a and imply the descriptive others examples, are (for contributions independence approach these of the traditionalist Some or such banks. of institutionalization issues, central case legislative of the applied the government, of government”, parliamentary group “divided and a the presidential of between study differences Empirical the the governments. on as of centered decisions been and has politics institutionalism the gov- that conditions indicates studies structure emphasizes furthermore empirically and ernment it approach effects, because their This and and issues. differences approach institutional institutional theoretical certain empirical of traditional lack of its set to a due map the in added as construed be and should sociology environment societies. in of its approach process institu- and new individualization The of an institution politics. sources the of irrational insti- between role molding the new relations the on a of based and conception sociology, old the in an tions, thought between of or distinguish Weber school can and as tutional sociology we such historical and like authors areas sociology classical in organizational maintained of tradition been time analysis has tradition the institutional Such exo- from strong Durkheim. the a right renounces been research has approach sociological There this in (B) therefore preferences. and of politics, geneity in participating those for sqieueu o nesadn h iest,puaimadcmlxt ftenew map the extended of this complexity adapted, and sciences. social pluralism or diversity, in discussed the institutionalism understanding be for could useful and quite clear is not classifica- are the criterions of some tion Although Institutionalism. International and Representation Institutionalism Interest Institutionalism, Sociological Institutionalism, Economics, Empirical Institutional New Institutionalism, Historical Institutionalism, (Peters Choice approaches nal eight using sented in ucinadteriflec ntebhvo fidvdas(ac n Olsen and (March individuals of institu- behavior how the on understanding 1984 influence for their organizations and within function values tions role and central norms the to highlights soci- assigned institutionalism truly Normative a and (A) institutionalism institutionalism. normative ological a namely, approaches two into divided is rnato otPltc nteMpo h e Institutionalism New the of Map the in Politics Cost Transaction 1997 ial,pitn u h i ftesuy w te ntttoaitapproaches institutionalist other two study, the of aim the out pointing Finally, oevr nte prah miia ntttoaimi oiis a been has politics, in institutionalism empirical approach, another Moreover, nta a,tescooia ntttoaimidctdb aladTyo ( Taylor and Hall by indicated institutionalism sociological the map, that In pre- been has sciences social in institutionalism new of map complete more A e ntttoaitpatc eas tehne h oillgtmc fthe of legitimacy social adopt the can participants. enhances its institutions it and concerned, because organization is practice institutions institutionalist as of In new change (C) a and actions. origin individual the and as institutions far between relationship the of nature , hw h neetrpeetto ntttoaimi seilycnee on centered especially is institutionalism representation interest The show. ) 1989 .Isiuin odteronpriiat n upymaigsystems meaning supply and participants own their mold Institutions ). 1999 :NraieIsiuinls,Ratio- Institutionalism, Normative ): 1999 ). 1996) 9 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 clatvt samaso otiigteucranyo cinadpltclresults. political and action of uncertainty the polit- in containing institutions of of integrated means role are a the actions as out activity individual sets how RCI ical manner, and this sequence, In what decisions. collective mea- in into what and structured, taken is be information can the sures how situations, decision-making in cluded fisiuin,isiuinlcoc n ogtr uaiiyo ntttos(Wein- institutions of durability long-term and gast choice institutional institutions, (Peters of RCI the to by individuals institutionalism” to “actor-centered generating bestowed expression role up important the end the used did indicate have this it In authors institutions. to stages, some political attend early of sense, role its not the incorporated during did which developments manner Choice theoretical relevant Rational a though model in explicit even institutions and However, clear behavior. a individual presenting of human for for characterized theory specific is a rational-choice provide behavior, not human approaches, do other which on against institutionalism, normative activity As as individuals. political such of nature basing the of explain that importance theories the behavior inherits it and individualism, rne ihnsc nisiuinlfaeok codn oKsradOstrom and Kiser to According framework. pref- institutional their maximize that an ( to incentives seek such and who rules within agents, of political erences set of a possibilities ra- choice as from the conceived restrict analysis are institutional Institutions theory. of choice micro-foundations tional the of importation the through hieter.W a hrfr s h ocp fRI(Shepsle RCI of concept the Taylor use rational and of therefore agenda institutions can research of We the importance theory. into the institutions choice assumed political that included tasks and life of political set in a of to tradition rise academic for gave later, choice choice Sometime War. rational rational World of Second development the principal after the politics was studying Choice Public of program The Institutionalism Choice Rational 3.1 Institutional New and Choice-Institutionalism Rational 3 NIE. the and RCI of analysis institutional- detailed new a the perform within first TCP with to integrate overview to need the order we complete In ism, NIE. should the we of and incorporation interna- the compatible, in but different lines are research tionalism relevant ( the Nye and of Keohoane One by led of values. been behavior of has the institutionalism set tional shape a predictability, promote and and level behav- stability States international the generate as explaining they regimes when since perceives structure institutionalism institutions, International of States. role of the ior highlights along and politics lines international conceives institutional Institutionalism International of approach the 10 1982 ainlcoc hoyhspoie itntv e fapoce otestudy the to approaches of set distinctive a provided has theory choice Rational methodological assumed that approaches choice rational the from emerged RCI nti es,teveso aladTyo ( Taylor and Hall of views the sense, this In Economics 1996 ,isiuin r ue htidvdasuet eemn htadwoi in- is who and what determine to use individuals that rules are institutions ), .17.Ti prahpoie ytmtcteteto institutions of treatment systematic a provides approach This 167). p. , 1996 Weingast ; 1996 , 2002 Peters ; 1999 n ees( Peters and 1996) ). .CbleoadXC Arias X.C. and Caballero G. 1977 ). 1986 1999 1999 , ninstitu- on ) 2006 ). ;Hall Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 a eiyta ir ee eal aeagetiflec nresults. on we influence great and a results have influence details institutions level how micro on that verify centered and constitu- can is justice as of analysis such courts The institutions the elections. bureaucracy, democratic body, the executive all the almost body, legislative on the done tions, been has work that out 2006 election. critical and micro-foundations revolutions the as provides such approach phenomena across The macro-political about for (D) institutions theory survival. distinctive similar and a of form yields stability, effects institutions their endogenous the of of study The and (C) country related countries. under given outcomes a More- and change. behavior within the conditions institutions of underlying comparisons the affords as approach this change over, outcomes analysis the and through behavior and how outcomes of explic- and behavior is with in methodology constraints implications institutional The corresponding distinct their (B) compare action. that models on institu- through constraints comparative, of itly as effects modeled the studying are for which methodology tions, systematic and explicit an provides ntttosa noeosvrals hti osy h ntttostk particular take studies institutions level why other say, the to (b) is (Weingast institutions; forms that of variables, effects endogenous the as study institutions that analyzes i.e., nous, r oa n a nuecodnto rudte (Shepsle them around coordination induce arrangements may Institutional by rules. and provided simple are focal be game to are are the tend institutions of these why rules and The themselves; and players others. another, the not of but instead circumstances form some in one altered take institutions why as such neato ewe cosadi eemnto fpltclrsls(aladTaylor and (Hall results political of determination in and actors between interaction institutions) it (self-enforcing and so do actors to specific incentives no by (Weingast have altered actors be these why to show institutions must allow must stability institutional et ofcltt oprto tutrs scamdb enat( Weingast by claimed as structures, agree- ex-ante cooperation as facilitate appear institutions to manner, ments this In level. institutional the at cilitated endwt h rbe fsaiiyo eut n h rbe fcnrlo public of (Peters control rasa of tabula problem a the con- on been and formed has results are RCI of Institutions (2) stability (3) bureaucracy. process. of political problem the the in with actors cerned central the are utility personal cooperation. from gains obtain to institutions need we that affirms opttr,adraieta ue eetteetr ru fidvdas Shepsle individuals. of group entire the benefit rules ( that realize and competitors, benefits. mutual personal for obtain or looking to way formal individuals gains in opportunistic the found permits be this Therefore, and can concerned. institutions, cooperation informal through all problems benefit action collective and resolve to problems action collective resolve rnato otPltc nteMpo h e Institutionalism New the of Map the in Politics Cost Transaction 1986 na a sWigs’ ( Weingast’s as far as In ihrsett enats( Weingast’s to respect With enat( Weingast w eaaelvl faayi a edsigihdi h C (Shepsle RCI the in distinguished be can analysis of levels separate Two ntttoait frtoa hiehglgtterl fisiuin nstrategic in institutions of role the highlight choice rational of Institutionalists C sue h olwn he etrs 1 ainlidvdasta maximize that individuals Rational (1) features: three following the assumes RCI niiul bev htisiuinlrlsas ii h hiepsiiiisof possibilities choice the limit also rules institutional that observe Individuals that cooperation voluntary of structures as institutions political considers RCI Weingast ; ttsta n oprto hti o otya h niiulaetlvli fa- is level agent individual the at costly too is that cooperation any that states ) 1996 1996 ). 1996 1996 onsotfu hrceitcfaue fRI A hsapproach This (A) RCI: of features characteristic four out points ) ,nml;()Alvlcniesisiuin sfie n exoge- and fixed as institutions considers level A (a) namely; ), ). rtlvlo nlssi ocre,w aet point to have we concerned, is analysis of level first 1996) 1996 eodlvlo nlss tcvr questions covers it analysis, of level second ) 1999 2006 2002 ). .Amdlof model A ). ,we he when ), 1986 11 , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 pti nastigta ilmk tvsl oefutu”(Coase fruitful” more vastly to it tries make but will theory that that price others setting replace to a to not try in but not it does matters “put NIE economic body. some theoretical richer to a respond require to us enables theory Price Economics Institutional New 3.2 institutions. those (Moe of institutions (Peters creation of content “goodness” the functionalist of a in sense has a interested case, is are any institutions in who of approach, actors creation This of the that action possibility rational the a recognizes it although created, are 1996 12 coasean passage the where world a (North assumes matters rights, and time property ra- of institutions, incomplete and instrumental with costs and framework transaction theoretical information positive a perfect considers of it assumption and tionality, neoclassical the rejects it framed not were traditionally that mainstream. questions economic new in NIE the answering (1987), of Arrow by consists indicated movement As issues. institutional of incorporation the plies r hs hc r eie rmtesigning the costs transaction from derived that of are such out which problem, carrying those contracting The are a function”. as “transaction understood the be perform can to transactions required opportunistic costs against of sum resources the their (Eggertsson relationships” project exchange assets and in their behavior theft, defend rights, and ownership new transgressions acquire against to try individuals when arise costs rdc rnato ot.W a en rnatoscssa h eore used resources (Allen the rights as property costs transfer will transactions which and transfers, define maintain rights can to property We out costs. carry transaction will produce they i.e., transactions, out carry I tde ntttosadhwisiuin neatwt raiainlarrange- organizational Shirley with and say, interact (Menard to economy institutions is within how that ments and inter-related, institutions mutually studies are which NIE approaches, (North both processes incorporates change institutional NIE as institutions well as between performance, relations economic the and studied that approach macroanalytical a eco- generated cost “transaction ( to Coase (Williamson rise gave hand, nomics” which one organizations the of greater approach a On obtaining microanalytical performance. and costs economic transaction in reducing for efficiency medium a are institutions that h rnatosta ol euti oilefiinygis oee,a gis this against as However, gains. efficiency social in result would that transactions the oto n opine(Eggertsson compliance and control h hoeia rmwr fteNwIsiuinlEooiscmie the combines Economics Institutional New the of framework theoretical The but competition, and scarcity of assumptions neoclassical orthodox accepts NIE rpryrgt r n’ blt oeecs hie vrago.Idvdaswill Individuals good. a over choices exercise to ability one’s are rights Property nawrdwt eotascincss h ate ocre ol ar u all out carry would concerned parties the costs, transaction zero with world a In .Hwvr hsisiuinls osntepantedtiso o institutions how of details the explain not does institutionalism this However, ). oino rnato ot ihthe with costs transaction of notion 1975 1994 , ). 1985 , 1996 1990 ;wieo h te ad os ( Coase hand, other the on while ); ). 2005 2005 ex-ante northian ). 1991 2005 ;Ostrom facnrc n fits of and contract a of ,ta st a,“transaction say, to is that ), .2) rnato ot are costs Transaction 27). p. , oino ntttos such institutions, of notion .CbleoadXC Arias X.C. and Caballero G. 1990 1999 , 1999b eeae a generated 1937) 2007 n concludes and ) ). ,wihim- which ), 1990a ex-post 1960) ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 a u oteicesn eun fisiuinlcag:()Isiuinlcag is change Institutional (A) change: institutional of returns increasing the always to not due tal to agents NIE efficiency. however, the greater change, provokes a for institutional costs search transaction of the towards of source coincide existence the accord- the are institutions, that prices of sustains relative evolution analyzing which when to view situation. ing efficiency future our an affect adopt that opportunities can of We set the delimits also but results nomic n a fbign h hiegpadbnigteeouino oit vrtime over society a of evolution the binding and gap choice (North the bridging of way one h ntttoa arx ytepoesso oillann n yteceto of creation the by and learning social of of processes externalities the the the by by costs. reinforced matrix, transaction is institutional dependence high the path with framework, markets theoretical char- imperfect this is and In change returns Institutional institutions. increasing create by to acterized time of passage the of importance environment” the external order the North and structure and are to individual (Denzau institutions create linked individuals that environment; mechanisms closely mind) the representations the be interpret internal (to will to the create models are systems mental models cognitive Such “Mental factors. institutions. key with as mental individuals in subjective individuals the of of considers behavior models NIE the context, understand uncertainty to an order within decision-making In necessary”. the of is understanding evolve an they and way matter, prejudices and dogmas, myths, ideologies, ideas, one ainlt,b ociigteidvda sitninlyrtoa u only but rational (Williamson intentionally way limited as a individual of in the model conceiving the assume by and rationality, deduced, subjectively bounded reality been between have that distinguish models to and formation need is there (North them: perception and around self world incomplete individual’s process the and subjectively of altruism individuals information utility: (b) motiva- or behavior; individual influence wealth also maximizing (a) limitations to because: ra- limited defective orthodox not is the are behavior that tions human considers of NIE decision-making. approach human tionality of study the requires ftegm” hc ildtrietecs fcryn u rnatos(North transactions out carrying of cost the “rules own determine its have will will 1990a which society every game”, sense, the the this which In of in out. environment institutional carried the being of is nature the transaction on as well as transaction cific the of producing value in the implied costs whenever the ahead than (Coase go greater same is only no transactions will such such therefore from from rights gains and production of expected costs, the readjustment surpass transaction The costs positive marketstransaction. such whenever of economic out carried presence anything, is the cost transaction by not characterized does negotiation are where world hypothetical Institutionalism New the of Map the in Politics Cost Transaction h I rusta h rcse fisiuinlcag r omlyincremen- normally are change institutional of processes the that argues NIE The eco- current the determines only not framework institutional the manner, this In oehrwt h td fmna oesadhmnbhvo,NEasmsthe assumes NIE behavior, human and models mental of study the with Together ( North lines, these Along nesadn h eainhpbtenisiuin n cnmcperformance economic and institutions between relationship the Understanding h ee ftascincsswl eedo h hrceitctat fec spe- each of traits characteristic the on depend will costs transaction of level The hrdmna models mental shared ). 1990a 1960 ). ). 1994 1990a ,p.4). nwihidvdasmk eiin.Pt eedneis dependence Path decisions. make individuals which on .NEdfnsta niiul c ihicmlt in- incomplete with act individuals that defends NIE ). 2000 1994 ). .32 ttsta hsoydmntae that demonstrates “history that states 362) p. , 13 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 A ainlCoc ntttoaimwsitrse npltclmresadinsti- and markets political in interested was Institutionalism Rational-Choice (A) Insti- New the to especially, Economics: and, tutional Institutionalism Rational-Choice to look a implies (Shepsle economy a from political analysis approach in political theoretical view to the Economics of Institutional application New an the as of emerged has Politics Cost Transaction Politics Cost Transaction 4 on questions interesting (2005 and Eggertsson advances institutions. relevant incorporate contributions tutional (Coase in” are his- you every society and the on country depends every for answer and different models torical a mental get costs, transaction “you situation...there positive changes: with institutional economies of case the in another is no one way better (North history find of will course but the history reverse change to can and difficult institutions Individual it informal (C) and limitations. formal char- informal in and is changes formal change specific of Institutional evolution broadly slow (B) are a framework. that by institutional policies acterized of basic favor the in with weighted heavily consistent is that process incremental an 14 in o nesadn ilneadsca re nhmnhistory. human in order de social and and jure violence (2009 de understanding al. of for et exercise tions North the Recently, of power. result political is a idea facto are main institutions the where economic institutions equilibrium in that ( persistence and Robinson change and simultaneous Acemoglu of model intermediation. appear financial institutions for contracting matter while investment, to and growth economic long-run on ( the Acemoglu Johnson change. explores and economic (2005) and institutions North science, policy. cognitive institutional among relationships and technology social failure, tional h I sarsac rga htcniulyeovs n eetnwinsti- new recent and evolves, continually that program research a is NIE The to economy one from exported be cannot conclusions analysis economic Positive hieIsiuinls ogttemi oeo rnato ot n history, and NIE. costs the to transaction look of should role we therefore main and Rational- the and forgot Institutionalism, Institutionalism Rational-Choice in Choice ra- found TCP not the However, is approaches. model traditions old tionality institutionalists the or other institutionalism from normative difference as such economic big for a model implies rationality a which of behavior, assumption the defends it when tutionalism a ntttoscnttt xat ooeainareet mn politicians” among agreements co-operation “politi- (North ante that ex indicates constitute and institutions institutions, In- cal Rational-Choice political to on According focuses behavior. TCP political stitutionalism, for rationality assumed of and model structure cooperation a a as institutions political understood tutions, 1990b 2005 .39.Frhroe C onie ihRtoa-hieInsti- Rational-Choice with coincides TCP Furthermore, 359). p. , ocueta rpryrgt ntttoshv rtodreffect first-order a have institutions rights property that conclude ) rsnsagnrlfaeokt eeto institu- on reflect to framework general a presents ) 1999 .Udrtnigtefudtoso TCP of foundations the Understanding ). 1999a cnmcsse eas everything because system economic ,p.5). 1990a rps h hoeia founda- theoretical the propose ) , 1990b .CbleoadXC Arias X.C. and Caballero G. ). madisonian 2008 osrc a construct ) on of point Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 a hieapoc.O h n ad hr sn ietrltosi ewe CPE between relationship direct no is ratio- Fig. there (instrumental) in hand, the TCP one surpassed and the NIE the On theoretical that approach. the way choice NIE—surpassed same nal the the of in extension RCI an continuation, appearance of TCP—as the framework main allowed sense, its analysis this political as In towards TCP. CPE NIE of with the of Choice, extension Public the of whereas emergence the allowed matters). time ysis of passage approach, transactional Figure a rationality, (bounded NIE the of transaction positive by characterized is world economic the that out points NIE (B) 1 Fig. Institutionalism New the of Map the in Politics Cost Transaction oeie mlctareet(rtascin mn oiymkr”(ple and (Spiller explicit, makers” sometimes policy on a among focuses is transaction) TCP policy (or of “public agreement approach that implicit the considers sometimes he literature, and of transactions tradition political relevant a by nowadays.interactions TCP TCP. with of dialogue background fruitful the a in not incorporated have been institutionalisms not these empirical, have Even emergence references as the their on such and influence and TCP, institutionalisms, had NIE of other not via have that indirect institutionalism, is sociological out influence or point (the normative and TCP should NIE over we of influence Finally, antecedent direct an a RCI). as not shown has it is but institutionalism RCI historical hand, other the on C sdfeetfo C eas C sue he hrceitcfoundations characteristic three assumes TCP because RCI from different is TCP hl rnatoa nlsshdbe ple oeooi n organizational and economic to applied been had analysis transactional While omneo oiisoe ie n hrfr lbrtsatertclframework theoretical matters. a history elaborates where therefore and time, over polities (North of formance agreements” of ex- to enforcement actors imperfect the assump- 1990b of of the part and on the environment, on built their models is plain subjective politics of information, of costly theory of cost tions transactions matters. transaction political time “A studying of institutions. when passage foundations and NIE the three that these considers assumes impli- and TCP the rationality assuming bounded by rationality of instrumental cations rejects It institutions. and costs rnato otPolitics Cost Transaction 1 hw o h xeso fRtoa hieter oad oiia anal- political towards theory Choice Rational of extension the how shows .35.Mroe,TPi neetdi xliigtedfeeta per- differential the explaining in interested is TCP Moreover, 355). p. , 1 eas hi hoeia onain aedfeetoiis and origins, different have foundations theoretical their because 15 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 eetu sue yTP(North and TCP analysis by political into for assumed too verified integrated thus also been are were has features dependence These path studies. organizational of the notion the and analysis, (1990a economic North the matter. determines history incentives, and structures that output. rules political of biases and set institutional costs The a transaction costs. of as transaction volume acts high and polity actors of the of structure part subjective the information, on models imperfect mental rationality, bounded imperfect agreements, rights, political of incomplete enforcement in- by of characterized relevance markets the political highlights in It stitutions contracts. and transactions as relationships (Dixit political achieved be always process would political plan the efficient world, an ideal since an as matter, such country not in would a that conclude of and government process of reasoning (Eggertsson type the zero to are the according costs by transaction theorem affected Coase’s as long not of version is macro that growth a economic those derive which to to less us (North unexecuted) allows ef- them stay situation value permit would This improvement” who would “Pareto owners that (no more from rights them rights value of of assignment transfer simple through a ficiency to correspond would activity imperfect (Dixit history and in incomplete time”, is “real which in evolution, place in takes polit- which process construes analyzed and dynamic likewise be a TCP costs. can as transaction activity they economize ical that to is matter, political effect institutions positive their that tackle political and rigorously that to sustains tries TCP that analysis. perspective microanalytical a from takes costs. that transaction policy transaction positive economic of of of effects politics application the direct whereas a account them, solutions be into with governance would deal sense the original to its order and processes in in costs costs political with diverse transaction up of positive come actors analysis of that an existence is the TCP on costs: based transaction of politics and Tommasi 16 uin hp netvsadpeeecso actors. of insti- preferences way ap- and the incentives has institutional regarding shape program especially both tutions institutionalism, TCP historic the of with RCI, are overlap agendas with that there connection the points its that between through detect and overlap and Furthermore, and proaches. common in intersection aspects of many points have RCI and institutionalism choice ( Weingast rational as and of such Katznelson approaches institutionalism moreover the and appreciated historic who of Longstreth, authors and are Thelen Steinmo, There historic RCI. between of the relationship foundations and the institutionalism indicated the recently through has TCP furthermore turn on Literature NIE. in influence exercised which his- has Really, path, institutionalism. institutionalism and historical torical of history arguments with of contact importance greater a the facilitates assumes also approach actional h I a oue oto t fot ndmntaigta asg ftime of passage that demonstrating in efforts its of most focused has NIE The of evolution the explains and analysis of unit the as transaction political uses TCP political zero, are costs transaction where world neoclassical pre-coasean the In transactions political studies and individualism methodological assumes TCP 2007 .3.I hssne esol on u h itnto ewe TCP between distinction the out point should we sense, this In 3). p. , eeddterlvneo ahdpnec in dependence path of relevance the defended ) 1990b 2005 Dixit ; 1990 aercnl niae hthistoric that indicated recently have ) .Btw a oase ute in further step a go can we But ). 1996 1996 .Teeoe uhatrans- a such Therefore, ). .CbleoadXC Arias X.C. and Caballero G. , 1998 ). 1996 1990b ). ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

690 689 688 687 686 685 684 683 682 681 680 679 678 677 676 675 674 673 672 671 670 669 668 667 666 665 664 663 662 661 EDITOR’S660 659 658 657 656 655 654 653 652 651 PROOF650 649 648 647 646 645 hti oesrdc rnato ot yrligo at inl oiiin aean have politicians signal, party on relying by costs transaction reduce voters if that ( Hudson and Jones parties. political 2001 of theory cost transaction a towards social vance and opportunism rationality, Bogt economic bounded (Ter as strictly institutions factors organizations, than government relevant less of are autonomization aspects the of trans- process political the reducing the of (Majone means like costs a institutions action as supranational understood or best banks Commission—is central European the as agents—such dependent through justified foundations: approach. is two rational following it and the individualism that assumed methodological strictly is initially perspective had this that of institutionalism an novelty The politics. understand to sur- institutionalism transactional manner, this RCI. In of RCI. suppositions ( of the Weingast challenges passes by the presented been of already one had rationality.as which bounded models, subjective assume on and based rationality revise instrumental to ( reject North’s them TCP (North lead and the theories” will NIE interpret incorrect that the to initially feedback which their information by correct models receive approach and correct or rationality have them instrumental either around actors the such world way, “the factors that this cognitive meant In and RCI myths. cultural of of and importance ideology the beliefs, highlight as to order in models rnato otPltc nteMpo h e Institutionalism New the of Map the in Politics Cost Transaction ac tutrsadisiuinldsg ntepbi etr(sah n Martimort and others (Estache sector to public the suited 1999 in poorly design institutional and and structures example, nance “sovereign an the is as exam- affairs such For (Williamson transactions, foreign possible. some which as to of suited much transactions” well so is costs bureaucracy public transaction ple, minimizes public that transac- regulation, structure of firms, type tional hybrids, each (markets, governance of insti- of characteristics bureaucracy, adequate models the the all on that Then, depend tions. assumes will TCP governance which Firstly, of on contributed. tutions issues relevant significantly three are has parties political TCP and agents independent to gation transactions. political O’Halloran in and problem political Epstein hold-up firm, and TCP, the the economic of of analyzed theory theory between the from the similarities lessons of and some Taking review differences interaction. a several included showed It it powers. and of delegation the to perspective ( Dixit and omdtetorlvn rcdns nipratcnrbto oTPfo politi- ( from O’Halloran TCP and to Epstein contribution been has important science An cal precedents. relevant ( two Marshall the and Weingast formed while program, the for bases retical eadn h ancnrbtoso C,w hudpitotta ot ( North that out point should we TCP, of contributions main the Regarding order in matter ideology and history that indicating of possibility the opens This rationality strict from escape to been has NIE by made effort principal other The h praho C sueu o raiainsuis ulcbracay dele- bureaucracy, Public studies. organization for useful is TCP of approach The xlrdhwpltclprisrdc oes nomto ot n hyargue they and costs information voters’ reduce parties political how explored ) Gallego-Calderón ; 1990b 1996 .)s ...) 1999 , , 1998 1994 .I hswy C noprtssvrlefrst td gover- study to efforts several incorporates TCP way, this In ). ol ecniee fw att eemn h etorganiza- best the determine to want we if considered be hould r h w udmna otiuoswopoie h theo- the provided who contributors fundamental two the are ) rpslicue h data niiul aedecisions make individuals that idea the includes proposal ) 2001 2003 1999 .I at hr r miia tde htso htin that show that studies empirical are there fact, In ). .Tidy oecnrbtoso C aetidt ad- to tried have TCP of contributions some Thirdly, ). Ruiter ; 2005 .Scnl,dlgto fpwrt in- to power of delegation Secondly, ). 1999 ,wihapidtetransactional the applied which ), 1990b .36.Nevertheless, 356). p. , 1988 n ot (1989) North and ) 1990b 1996) 1998, 17 ) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

736 735 734 733 732 731 730 729 728 727 726 725 724 723 722 721 720 719 718 717 716 715 714 713 712 711 710 709 708 707 EDITOR’S706 705 704 703 702 701 700 699 698 697 PROOF696 695 694 693 692 691 einrligo eaaino oesaogseilzdaet LfotadMar- and (Laffont agents specialized among (Weber timort powers politics of separation international on relying in design hierarchy the and choices cltascin n h ujciemdl fteatr nraeteaon of af- amount ideologies the different increase (therefore actors markets exchange). political the political in fect of more models polit- even subjective costs in relevant transaction the particularly and are transactions O’Halloran un- information ical and an asymmetric strong (Epstein example, of exert for arrangements situations may when, Moreover, contractual rights key ex-ante control is over ex-post control influence Moreover, of exceedingly: occurs. rights matters eventuality residual relations inter- contracted the power and ex-post of vague the contracts possession containing that political the prototype, implies explicit, contract This were terms. incomplete they pretable an if to Even respond markets. clearly the economic measure in and system observe transaction price political to of difficult objects the is as (Pierson it such performance, and political unclear of factors opaque, different is politics of world terri- of governance power. the of “electorate-parliament- be distribution can chain torial example the another by be yet can configured and example is government-bureaucracy”, an that complex: agency-relation especially vertical be to the rela- tends actors agency” political “common amongst a relation of (Dixit terms principals in multiple interpre- agents with enable tionship that political effects transactions between spillover political interaction of Moreover, presence of wide agreements. tation the tacit to even due for- agents and nor many verbal affect a explicit on is neither rest parties are the and contracts of mal political one many when furthermore, especially subject; happens multiple This transactions. rights. political of of cases change rules. means safe less this of and set polit- a authority safe, around for While revolves normally manner. politics fight are Therefore, unlimited the that an includes rights in property competition them on ical possess place agents takes do competence nor economic safe not are they tions: ones important most the of of and understanding some relevance better emphasized. and be the a costs, must understanding get transaction for political to of essential try characteristics are we considerations when Several concept politics. irreplaceable convert an transactions political into in costs them transaction of intensity and nature peculiar The Exchange Political in Costs Transaction High 5 Zelner and (Henisz governments design the been have TCP in topics (Patashnik institutions Other budgeting reputation. of party maintain to incentive 18 hrl,teeaehg nomtoa rbesi oiia rnatos The transactions. political in problems informational huge are there Thirdly, many in identified perfectly be cannot and many are parties contracting Secondly, interac- political within constraints strong to subject are rights property Firstly, 1998 2000 n h oenneo h eainhpbtenpiaeivsosand investors private between relationship the of governance the and ) .I hssne oiia akt akamaueetfrualk the like formula measurement a lack markets political sense, this In ). 2004 1996 ). 1996 ,tecutis nentoa institutional international countries’ the ), .Frhroe h tutr fagency- of structure the Furthermore, ). .CbleoadXC Arias X.C. and Caballero G. 1997 ,teinstitutional the ), 1999 ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

782 781 780 779 778 777 776 775 774 773 772 771 770 769 768 767 766 765 764 763 762 761 760 759 758 757 756 755 754 753 EDITOR’S752 751 750 749 748 747 746 745 744 743 PROOF742 741 740 739 738 737 ceie yapt ihicesn eun.Pltclisiuin edt sals a establish to tend institutions Political returns. towards increasing bias char- with scenarios path political lower a in (Dixit effective a by less organizations acterized to are of mechanisms leading learning choice markets, and corrective economic intense the in less than a weaker and and efficiency slower are markets in ical effective equally are that mechanisms analogous (Pierson no and politics rights are property there as possesses (such markets), marketplace horizons economic capital time the im- lengthening While whose for run. mechanisms decisions long strong political the some those in of out nature play only the plications with contrasts (Pierson and logic, electoral effort augmented the are between problems relationship informational the and that 2000 unclear such quite others, becomes action of my effect actions of consequences on the highly makes politics depend of nature collective The transactions. rnato otPltc nteMpo h e Institutionalism New the of Map the in Politics Cost Transaction eaial ihri oiia akt hni cnmcoe (North ones economic in than markets political in higher tematically paper”. on words but are enforcement without them “rules complex make cluded, more Tommasi to and with safeguards (Spiller policies time institutional over public flows, more effective Moreover, benefit require show. non-contemporaneous will will with characteristics section those transaction next of the case as the and such vote, Tommasi a and for (Spiller sidered costs transaction intertempo- increases The exchanges 2007 time. political over of enforced be nature should ral that not agreements are is, political policies that intertemporal enforcing agreements, public and Since striking requires process. cooperation typically transactions, political spot are the constrain promises possibilities ( such mechanism mitment but compliance a contracts to political subject not in unit exchange fundamental (Vanhuysse markets economic in those influence than high (Dixit weaker a process significantly achieve are political that politics the institutions above, in designing the incentives To of difficulties costs. of the transaction reduce add (Pierson to must exchange adjustments we hamper institutional can of power out of authority. carrying structure on the based sense, constraints incorporates this that In density institutional an exists there 1996 esnlojcie.I hswy oiial eeattascincssaeas to government also of are use self-interested costs through transaction determined relevant endogenously politically extent way, great a this achieve In to strategically objectives. them personal manipulate actors to political gravitate intentionally; increased to (North are tend economy issues the costs from transaction polity high Moreover, Norway. such in costs, transaction solidations ( political high Sorensen of example impact for the as to due complicated very are tions itl,rgrigtepsaeo ie h hieadeouinfre npolit- in forces evolution and choice the time, of passage the regarding Fifthly, political of range wide a characterizes action collective of problem the Fourthly, ae nteecaatrsis mn tes rnato ot edt esys- be to tend costs transaction others, among characteristics, these on Based a are promises action political mechanisms, enforcement the regarding Sixthly, .Mroe,tesottr oio fpltclatr,woaeitrse in interested are who actors, political of horizon short-term the Moreover, ). .Ti stecs ftoecnrcswoeblsaentsmlaeul con- simultaneously not are bills whose contracts those of case the is This ). aalr n Arias and Caballero ; ttsquo status 2000 ). hc apr hneadaatto onwstain,and situations, new to adaptation and change hampers which vlae hnh tde oa oenet con- governments local studied he when evaluated 2006) 2003 .Svrlcs tde hwta oiia transac- political that show studies case Several ). 1990b n oiia rnato ot sometimes costs transaction political and ) hr at enforcement party third 2007 1996 ,bcuea srm(2004 Ostrom as because ), ,adteicniesrcue in structures incentive the and ), 1996 n iie com- limited and ) .Ta st say, to is That ). 2000 1990b a con- has ) ,adthe and ), Dixit ; 2002 19 ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

828 827 826 825 824 823 822 821 820 819 818 817 816 815 814 813 812 811 810 809 808 807 806 805 804 803 802 801 800 799 EDITOR’S798 797 796 795 794 793 792 791 790 789 PROOF788 787 786 785 784 783 edt h nerltdpltclbhvosta hrceieteplc-aigpro- policy-making the characterize Tommasi inheritances and that historical (Spiller behaviors and cess political institutions interrelated political We the how fundamental. to scenario is lead each society in each know in to processes need political (Acemoglu structure prevalent that be governance of can institution and cies ucs fplc eomwl eedo h blt oatro dp ntttosin institutions rent- adapt without or policies (Murshed good alter difficult to to is commitment seeking ability credibly where the and the on direction where depend desired countries, the will less-developed reform in policy higher of are success costs transactions that expounds a enetne oepantewrig fdemo- al. of et workings (Scarstacini the actors explain political to and extended institutions been cratic has (2007) Tommasi and Spiller of iia rnato ot ndsgigU diitaieaece Wo n Bohte and (Wood po- agencies manipulates administrative coalition US enacting 2004 designing the in that costs shows transaction that litical evidence empirical is there fec onr stekyfco omk oiia rnatosdfcl reasier, or of difficult some (Scartascini are transactions determinants technologies political institutional enforcement and key make horizons the to time players, factor of key number the the and is framework country institutional The each an moves”. of and unilateral responsibility, for fiscal leeway for excessive a incentives executive, with little executive the provinces with grants aligned been with that often system bureaucracy has federal a that activities, judiciary in a legislative transactions objectives, in long-term political uninterested no intertemporal legislature promote “a iden- not country: they do and the that TCP features from key Argentina some in policy tified public of foundations institutional the iinws“o ipet ov udmna rbesi eiltv exchange” legislative in tra- problems logrolling fundamental the solve But Tullock. to and simple (North Buchanan “too by in was initiated logrolling dition are or was they vote-trading that which on tradition in focused projects the was those pass analysis to Pre-transactional cooperation interested. and exchange on for TPC look of islators approach TCP. of the analysis of legislative case of a possibilities case as the the governance introduces and show section transaction to This institutions. useful and be transactions political can study case A Transactions Political of Governance The Case-Study: A 6 ma- via (Twight act costs to transaction predicted political are officeholders of there which and nipulation under enforcement), conditions or specific agreement some negotiation, are (us- of manipulation costs transaction-cost costs, political informational of ing ways several are There mechanisms. 20 otmoaeu eetflw n o-iutnosecags tipisthat, implies It exchanges. non-simultaneous and flows benefit contemporaneous hrfr,hg rnato ot npltclmresipyta nfcetpoli- inefficient that imply markets political in costs transaction high Therefore, navr eeatapidwr,SilradTmai( Tommasi and Spiller work, applied relevant very a In oiia gemn mn eiltr sncsayt asblsi oges Leg- Congress. in bills pass to necessary is legislators among agreement Political nCongress in ). 1990b .I at eiltv xhnehshg rnato ot u onon- to due costs transaction high has exchange legislative fact, In ). 2007 2001 ). ). 2007 1994 2003 .I hssne h framework the sense, this In ). .I hssne o example, for sense, this In ). 2010 .Suyn h institutions the Studying ). 2003, .CbleoadXC Arias X.C. and Caballero G. .Fnly ii (2003) Dixit Finally, ). 2007 aestudied have ) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

874 873 872 871 870 869 868 867 866 865 864 863 862 861 860 859 858 857 856 855 854 853 852 851 850 849 848 847 846 845 EDITOR’S844 843 842 841 840 839 838 837 836 835 PROOF834 833 832 831 830 829 ewe ogese nodrt civ aoiista emtcagn the changing permit that majorities transactions achieve allow quo. to that tus order committees in of congressmen system US a between the the that establishes regarding being Congress exchange, rights legislative States property of is United costs of of Congress transaction system majority high the the the the of hand, reduces role and committees other executive the the preferences, the On political reduced. as same clearly long the As represent model. legislature Spanish the the in system chical committees. parliamentary on each rights of property head to the the rights and has property seats, group committee grant their not for does deputies Congress individual Spanish the indus- the The of transactions. model political organizational facilitate trial to presented are (Caballero governance party institutional national political of power each the of and committees leaders weak charac- politics, been electoral has party-based Congress by Spanish property terized the of transaction individual organization legislative industrial with the the committees recently rights), powerful facilitate politics, that electoral rights (candidate-based property strong have congressmen of po- structure of the organization. determinants determine legislative they institutional of and main governance organization, legislative the in Electoral of rights congressmen. two property individual are litical of systems rights “party- committee property and between committees of rules of structure distinguish the structure for institutional to for relevant the key is Moreover, important is congressmen. it is of since incentives rules”, It electoral the “candidate-centered rules. and rules” electoral electoral centered and political by trans- reduces of that regardless mechanism congressmen affiliation. right among party property negotiations their of independent favors type and a mecha- costs Having is ex-post. action a committee them, agreements a constitutes the for in agenda declining valued position avoid the can a highly to committee more each access which are restrictive by which The nism votes. committees Leg- trading mechanism. those of exchange on rightsinstead market seat to of explicit a not “institutionalization the seek as The substitutes so islators A. control” right of their agenda waive jurisdiction the may the on B of committee of proposals com- of legislator the of members jurisdiction influence the a of return selection system, the In influence this B. to mittee intention Under his costs. cede can transaction A committee low relatively had Congress ex-ante constitute (North politicians” institutions among “political cooperation sense, make about to agreements this try In agreement should easier. the Congress exchanges allows of organization legislative that industrial governance of the of and problems structure congressmen these institutional among resolve an establish not simultane- to Marshall does vote sary and (Weingast exchange a transactions for of legislative up of form come enforceability market not explicit do many The that secondly, ously. and, bills trading concern inhibit trades can legislative flows potential benefit of patterns differential firstly, Institutionalism New the of Map the in Politics Cost Transaction nti a,lgsaietascin n gemnsaecridotvaahierar- a via out carried are agreements and transactions legislative way, this In which in Congress of model prototype a represents Congress American the While affected are institutions legislative of organization the and behavior Legislative ( Marshall and Weingast hrfr,pltclpris(irrh)i h pns aeadcommittees and case Spanish the in (hierarchy) parties political Therefore, 1988 nlzdhwteCmiteSse fteUS the of System Committee the how analyzed ) 2011 .I hssne ifrn oesof models different sense, this In ). 1990b ). 1988 .I sneces- is It ). sta- 21 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

920 919 918 917 916 915 914 913 912 911 910 909 908 907 906 905 904 903 902 901 900 899 898 897 896 895 894 893 892 891 EDITOR’S890 889 888 887 886 885 884 883 882 881 PROOF880 879 878 877 876 875 wihhsbe neie rmnolsia cnm) hl C noprtsthe incorporates TCP rationality while substantive economy), of neoclassical model from the inherited adopts been has CPE (which orthodox The assume. they converted which TCP. is in Transaction analysis theory. political for political unit approach excellence to such par economics and a into from analysis, economic expanded for later contrac- initially the was assumed approach, TCP transactional hand, or other tual the On centered agreements). was constitutional and around reaching short example was hand, (for CPE issues other the specific of the on studies on as political expand and in to mainstream; analysis unable was economics transactional it the the hand, in one fundament the on methodological limited: a be contrac- to this out of turned extension approach analytical tarian and application the However, theory. economic improvements possible democratic offer and of order order such same. of constitutional the effects to studies the methodolog- into CPE same research the Choice. to on societies Public founded was as which CPE, postulates the ical of (Buchanan start politics the constitutional meant of book level This the the from distinguishing ordinary-politics categorically of by level action, collective of analyzes on framework level ocue h ovnec fteuaiiyrl o aigti yeo decisions of type this making for rules rule over “rules unanimity and (the the consensus, of towards convenience attitude the cooperative concludes a generates decisions constitutional in from Starting CPE. dimension. Choice: this tive Public of within study development The principal game. the sum to positive rise a gave as choice configured and of identifiable is type clearly political politics” not of constitutional are game of groups great and level “the individuals the of at interests the ( useful when Buchanan especially choices, is to approach According exchange politics-as-exchange. individ- such methodological and presuppositions: choice three by rational summarized ualism, be can core hard Its new the of characteristics the clarity. of more some with show approach to will institutional approach Public TCP rational and a economics. CPE implied Comparing from (CPE) politics. analysis Economy the Political political constituted Constitutional madisonian approach and Choice rationality on instrumental program the research TCP, main and NIE to Previously Politics Cost Transaction and Economy Political Constitutional 7 (Caballero policy-making different for institu- the implications Each different in transacting. has factors and structure making key tional decision as facilitates appear that governance case of American models the in system) (decentralized 22 h ancnrbto fBcaa n ulc ( Tullock and Buchanan of contribution main The oal ifrnebtenCEadTPrsdsi h ua eairmodel behavior human the in resides TCP and CPE between difference notable A for as well as analysis political for both framework, contractarian a defends CPE P tde h fcec fcntttoa ue nterpstv n norma- and positive their in rules constitutional of efficiency the studies CPE century. a half over developed been has program research Choice Public The ytmi studied). is system ” tt quo, statu logrolling, P niae o h elo uncertainty of veil the how indicates CPE raon h td ftecs of cost the of study the around or a oips two- a impose to was 1962) .CbleoadXC Arias X.C. and Caballero G. 2011 ). 2003 1966 ). ), Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

966 965 964 963 962 961 960 959 958 957 956 955 954 953 952 951 950 949 948 947 946 945 944 943 942 941 940 939 938 937 EDITOR’S936 935 934 933 932 931 930 929 928 927 PROOF926 925 924 923 922 921 esetv npstv nlss n srifre nnraiepioohcltheo- normative-philosophical in reinforced non-ideological developments. is under- and retical and to non-historic analysis, order a positive into in assumes in delving matter perspective CPE However, of ideologies actions. and challenge political history the manner, stand assumes this and In studies matters. cognitive political in dimension historic that transaction political of intertemporality. view as analysis simplistic such political and elements other for static incorporate a costs the not assumed transaction did On it of efficiency. case, role any optimal central in the to and, stress leads not that does CPE solution hand, entailed cooperative difficulties the a explain achieving to us in permitting thereby exchanges, political in costs ic hn h C eerhpormhsidctdteiprac ftransaction of importance the indicated has program research TCP the then, Since is evolution (Dixit institutional acts political of past path and the a rules that this more constitutional furthermore is of In and up acts effects. made political type long-lasting and than rules level have between of them distinction matter of the weight that some the defends gauge when TCP can manner, especially one acts, and specification political the future of to leave for to constitutions open due due Thus, terms and contingencies. contractual contingencies, contingencies verify many future foreseen and observe for all objectively even foresee to rules difficulty to specifying incapacity of incom- their complexity as the to perceived due are constitutions contracts TCP, formal plete to integrates According framework the institutions. this within informal and society, element and a more of one framework just institutional is complex also constitution agents Therefore, Furthermore, and ignorance. groups influence. several their of of but exert interests criterion” veil the “justice and “rawlsian” structure a power a is negotiation everything where behind not wherein not elaborated-rules are are constitutions agents such ( that Dixit problems. states information despite strategically act will decisions stitutional level. decision constitutional is the it whenever in to de- irrelevant out constitution constitution action carried political (“the not to any politics role framework making and key This thereby economy a everything”), game. of confers termines sum results and and positive rule operation a of the of notion understand that the of as analysis with such the constitutions permitting politics” links thereby of uncertainty, game of veil great a “the behind out carried are sions more not on logics is political that and CPE economic approaches. from behavioral integrates human emanates TCP flexible because economicism TCP suppositions greater in These a patent NIE). why quite of understand characteristic to is key (which are rationality bounded of model Institutionalism New the of Map the in Politics Cost Transaction ot ( North 8Conclusion C sue hoeia esetv hticroae h motneo the of importance the incorporates that perspective theoretical a assumes TCP transaction of relevance the highlight program TCP the within out carried Works codn oteTPtertclapoc,teaet novdfrmkn con- making for involved agents the approach, theoretical TCP the to According deci- constitutional that indicates constitutionalists of framework theoretical The 1990b n ii (1996 Dixit and ) rvddtetofudn otiuin oTCP. to contributions founding two the provided ) 1996 ). 1996) 23 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 1012 1011 1010 1009 1008 1007 1006 1005 1004 1003 1002 1001 1000 999 998 997 996 995 994 993 992 991 990 989 988 987 986 985 984 983 EDITOR’S982 981 980 979 978 977 976 975 974 973 PROOF972 971 970 969 968 967 rsn o nteNIE. the in too present iebtTPhsntaeutl noprtdterl fceco npltcltrans- political in coercion of political role the in incorporated (Nye factor adequately important actions not very has beyond are TCP goes coercion but models and life power mental are (c) approach shared the North’s rationality); of example, bounded limits (for the TCP and well-defined (b) diffuse always exist; slightly not not are does contents theory whose general approach the- this an general possibly is a and lacks TCP institutions, (a) political considered: of be ory should issues Among can three analysis. TCP, the structure of costs into weakness governance rationality transactions the bounded the political incorporates therefore approach (b) this institutions, (c) analysis; inefficient matters; trans- of of political existence unit (a) the the ones: explain following as the considered are are TCP of actions strengths relevant Three TCP. of transactions political on institutions different the of outcomes. role and the analysis comparative understand of to importance order the is in concludes program and This sciences. analysis social positive of on map centered institutionalism new the in niche its occupies RCI. with the contact with in overlap was an TCP was when there content sense, historical institutionalist this between in historical overlap and Like- and institutionalism, TCP. choice intersection by rational of choices. analysis and points political individual considerable to were of exported there fundament was wise, NIE the of on perspective and historical based doses This theory determinism institutional historic of any importance an eliminated the established but understood analysis which institutional NIE, for through history was influence main The proach. (Shepsle spective tttsa xeso fNEtwrsa nlsso oiisfo a from con- politics TCP of sense, analysis this an towards In NIE matters). of time extension ratio- of an (bounded stitutes passage NIE approach, from of different transactional foundations is a characteristic TCP nality, three However, assumes behavior. TCP political because for as- RCI and rationality structure of cooperative model a a as sume institutions political understand both institutions, institu- historical and NIE RCI, of based programs possible the was tionalism. of TCP advances of institutionalist development the and on content On appearance, institutionalism. the contrary, international the and institutionalism institutional- interest-representation sociological ism, institutionalism, approaches institutional empirical the institutionalism, with normative elements of common few its very has of and one approach, own map as the TCP and incorporate view must of sciences point social transaction insti- in approaches. a political institutionalism from manner, new study this the of the In of object determine outcome. the institutions political become political and tutions how costs studied transaction has of and volume markets political in costs 24 sacnlso,w hudpitotsm tegh,wanse n challenges and weaknesses strengths, some out point should we conclusion, a As that program research institutional intrinsically and true a as appears thus TCP ap- TCP the on influence indirect important an had has institutionalism Historical C onie ihRIbcuebt r neetdi oiia akt and markets political in interested are both because RCI with coincides TCP its with institutions political studies that institutionalism transactional a is TCP 1997 1999 ;Moe2005 ). .I n ae hs he ekpit fTPare TCP of points weak three these case, any In ). .CbleoadXC Arias X.C. and Caballero G. madisonian per- Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1058 1057 1056 1055 1054 1053 1052 1051 1050 1049 1048 1047 1046 1045 1044 1043 1042 1041 1040 1039 1038 1037 1036 1035 1034 1033 1032 1031 1030 1029 EDITOR’S1028 1027 1026 1025 1024 1023 1022 1021 1020 1019 PROOF1018 1017 1016 1015 1014 1013 td fpltc hog C.Tesac o hoyo ntttosbsdon based institutions (North of sciences theory social a different for the among 1990a search reconciliation The favors TCP. choice through individual politics of study in- several social pure, The in too vocation prototype. multidisciplinary sciences. remain subject a that hybrid assume simultaneously disciplines a should scientific as stitutionalisms analysis the transactional against proposed as and science in fertile” ingly ( Coase ( Bates and sense, this In institutions. in interested are we ( and when Coase in assess convenient profile is multidisciplinary to a sciences need that social we too Tran- implies institutions, (c) institutionalism approaches. of disciplinary institutional scending different change the and from order coming role In contributions past. integrate notion, the the in the between than understand fewer exchange are to institutionalisms and of dialogue types (2006 some Shepsle more between sense, ferences this and In more institutionalisms. plu- of be methodological types assumes several should that There TCP a (b) in workralism. useful econometrical are techniques analysis, experimental comparative and institutional case-studies, work: pirical ae 21)Tenwisiuinls.I:Telgc n oko ogasNrh conference, North, Douglass the of work of and foundations legacy the The In: and institutionalism. Arrow new (ed) The (2010) G R Feiwel Bates In: essays. the Cambridge on Press, Reflections MIT (1987) analysis. KJ institutional 14:1Ð18 Comparative Arrow Econ (2001) poverty. Law al Res et and costs? M prosperity transaction Aoki power, are of What origins (1991) DW Rev the Allen fail: Econ nations Am Why institutions. (2011) and J elites Robinson power, D, Acemoglu of Persistence (2008) JA Robinson develop- D, and institutions Acemoglu of economy political for notes 113(5):949Ð995 Lecture Econ (2007) Polit JA J Robinson institutions. D, Unbundling Acemoglu (2005) S politics. Johnson and D, commitment conflict, Acemoglu Social theorem? Coase political a not Why (2003) D Acemoglu References Institutionalism New the of Map the in Politics Cost Transaction hoisotnhv togreprclbs hni sa neoois.North economics”. in usual is their than correct, base is impression empirical my stronger if ( a and effects have etc., the system, often of political account theories the take to system, have legal “We science: the hard of into it convert to subjects other neter ngnrl o utpltclter reooi hoy.I hssense, this Sciences. Social In Institutional theory”. sci- New economic the social or to restructuring theory road for political a foundation is just a there not being general, to in way theory its ence on well is approach, cost 1999 NS,Wsigo nvriy t Louis St. University, Washington 727Ð734 CNISS, pp York, New Press, University York New policy. economic of theory York New Publishers, Crown 98(1):267Ð293 Cambridge Press, MIT ment. 31(4):620Ð652 Econ Comp J h rnatoa prahbr neooi nlssmngdt akethe tackle to managed analysis economic in born approach transactional The em- more needs TCP (a) presented: are future the for challenges three Finally, ok ntehptei ftemrig fpltcladeooi theory economic and political of marriage the of hypothesis the on works ) .Acrigt ot (1999 North to According ). 1999b 1999b onsotterlvneo oiisfrtenwinstitutionalism. new the for politics of relevance the out points 2010) .5 ieiehglgtdhw“yrdsbet r fe astonish- often are subjects “hybrid how highlighted likewise 5) p. , .4 eeddtecneineo ikn cnmcsinewith science economic linking of convenience the defended 4) p. , .35,“htCaesatdwt transaction with started Coase “What 315), p. , utista h dif- the that sustains ) 25 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1104 1103 1102 1101 1100 1099 1098 1097 1096 1095 1094 1093 1092 1091 1090 1089 1088 1087 1086 1085 1084 1083 1082 1081 1080 1079 1078 1077 1076 1075 EDITOR’S1074 1073 1072 1071 1070 1069 1068 1067 1066 1065 PROOF1064 1063 1062 1061 1060 1059 ogo M(98 h praho ntttoa cnmc.JEo i XXXVI:166Ð192 Lit Stud cost Econ transaction J Polit economics. a institutional institutionalism. of safeguards: approach and The new (1998) hazards GM three political Hodgson the Explicating (2004) and BA science Zelner WJ, Political Henisz (1996) France. and R Britain Taylor in intervention PA, State of Hall politics the economy: the Governing (1986) PA Hall In: policy. public of determinants the and institutions Political (2001) build- MD North McCubbins C. (eds) S, Douglass institutionalism: Haggard G old Caballero and new N, integrating Schofield On (1995) In: al equilibria. et J or Press, Groenewegen rules University Institutions: Cambridge (2011) economy. C modern Kingston the A, to Greif path the 88(2):80Ð84 and Rev Institutions Econ Am (2006) analysis. A institutional comparative Greif and Historical (1998) transaction A political Greif of role the sector: public the in design institutions. Institutional regulatory (1999) of R Gallego-Calderón design the and costs transaction Politics, policy (1999) D to Martimort approach A, politics Estache cost transaction A powers. Delegating Arbor Ann (1999) Press, S Michigan O’Halloran Cam- of D, Press, University Epstein institutions. University Imperfect (2005) Cambridge T institutions. Eggertsson and behaviour Economic Econ countries. (1990) less-developed T for Eggertsson politics transaction-cost from lessons Some (2003) AK In: Dixit study. case a and framework a policy: economic and politics cost Transaction Press, (1998) MIT AK perspective. Dixit politics transaction-cost a 47:3Ð31 policy: Kyklos economic institutions. of making and The ideologies (1996) models: A mental 2(2):1Ð6 Dixit Shared Newslett (1994) ISNIE DC society. North the AT, 2(1):3Ð10 of Denzau Newslett task ISNIE The Coase. (1999b) 140(1):229Ð231 Ronald RH Econ with Coase Theor interview Inst An J (1999a) 3(1):1Ð44 economics. RH Econ institutional Coase Law new J The cost. (1984) social RH of Coase 4:386Ð405 problem Economica The firm. (1960) the RH of Coase nature The (1937) político análisis RH el Coase política: economía la de teórica reorientación Una (2003) XC Arias G, Caballero the in legislators amateur and system committee foundations, Institutional (2011) 156:59Ð86 constitutional Sistema G institucional. of economía Caballero foundations nueva La Logical (2001) consent. G of Caballero calculus The (1962) G for Tullock Center JM, program. research Buchanan a of development Eco- and (1987): origins the J choice: Buchanan Public (2003) In: JM process. Buchanan political of theory individualistic Thou- An Foundation, (1966) Sage Russell JM sociology. Buchanan in institutionalism new The (1998) V Nee M, Brinton 26 oiisapoc.IdCr hne13(6):901Ð915 Change Corp Ind approach. politics XLIV:936Ð957 London Press, University University Oxford Cambridge policy. and parliaments Presidents, (eds) Cambridge Press, MD McCubbins S, Haggard 29(2):467Ð476 Issues Econ J bridges. ing Heidelberg Springer, voting. and democracy institutions, of economy Political Cambridge Mannheim Session, Joint ECPR costs. Research Policy Bank World Cambridge Press, University Cambridge powers. separate under making bridge 15(2):107Ð133 Polit Mar- St. economies. advanced the in York organization New economic tin’s, and Institutions (ed) M Baldassarri Cambridge 8:131Ð164 Política Cienc Esp Rev and transacción. de democracy costes institutions, de of economy Political (eds) G Heidelberg Caballero Springer, voting. N, Argentina, (USA, Schofield perspective comparative In: institutional Spain). an Congress: Spanish the of governance Arbor Ann Press, Michigan of University democracy. University Mason George choice, public of study the Austin Press, University Texas philosophy. moral and theory predictive Between nomics. Oaks sand .CbleoadXC Arias X.C. and Caballero G. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1150 1149 1148 1147 1146 1145 1144 1143 1142 1141 1140 1139 1138 1137 1136 1135 1134 1133 1132 1131 1130 1129 1128 1127 1126 1125 1124 1123 1122 1121 EDITOR’S1120 1119 1118 1117 1116 1115 1114 1113 1112 1111 PROOF1110 1109 1108 1107 1106 1105 srmE(07 hlegsadgot:tedvlpeto h nedsilnr edo institu- of field interdisciplinary the of 2:8Ð10 development Update the IHDP growth: paper. and on Challenges words but (2007) are E enforcement Ostrom without Rules Cam- (2004) action. E collective Ostrom of institutions of evolution The commons. the Governing (1990) E Ostrom arrange- contractual changing and trade, rights, property state: the about for Thinking framework (1997) conceptual JV a Nye orders: social and Violence (2009) BR Weingast gov- JJ, institutions Wallis of DC, evolution North the Press, commitment: University and Princeton Constitutions (1989) change. BR Weingast economic DC, of North Levi process J, the Alt In: Understanding theory. (2005) economic and DC political North of marriage the of 84(3):359Ð368 anticipation Rev In Econ 2(4):355Ð367 Am (1999) Polit time. DC Theor through North J performance politics. Economic of (1994) and theory DC cost North polities transaction A Uni- of (1990b) Cambridge DC development performance. North economic historical and the change institutional to Institutions, (1990a) approach DC cost North transaction R Swedberg A N, (1989) Smelser In: DC sociology. and North economics in institutionalism institution- new new The The (2005) (eds) V V Nee Nee M, Brinton In: institutionalism. new the of Sources Dis- (1998) rent-seeking, V and Nee commitment for institutions politics, cost 3(2):215Ð233 Transaction Politics Perspect (2001) Dordrecht institutions. SM Springer, political Murshed economics. and institutional Power new (2005) of TM Handbook Moe J (2005) Econ M growth. Shirley economic C, of Menard sources the and institutions York of New economics Press, The Free (1986) Am institutions. RCO life. Rediscovering Matthews (1989) political JP in Olsen factors JG, organizational March institutionalism: new The (1984) JP Olsen political JG, a governance: March democratic of limits the and institutions Nonmajoritarian (2001) G powers. of Majone separation the institutional and design of institutional synthesis costs, Transaction metatheoretical (1998) a D Martimort action: JJ, of Laffont worlds three The (1982) E 5(2):151Ð Ostrom Econ L, Inst Kiser J change. institutional of theories the Comparing (2009) for G strategies Caballero networks: C, Little, complex Kingston transition, Managing (1997) in JFM politics Koppenjan EH, world Klijn interdependence: WJM, and Kickert Power (1977) JS Nye institu- RO, choice Keohoane rational and historical between Intersections (2005) BR Weingast costs costs. transactions I, a transactions policy: Katznelson on public and based integrity political analysis parties, an Political (2001) parties: J political Hudson P, of Jones role The (1998) J Hudson P, Jones Institutionalism New the of Map the in Politics Cost Transaction inlaayi.JIs ho cn3(3):239Ð264 Econ Theor Inst J analysis. tional institutional Cambridge new Press, the University of bridge frontiers The (eds) J Nye Diego J, San Drobak Press, In: Academic coercion. economics. with world a in ments Cambridge Press, University Cambridge history. human recorded interpreting 49(4):803Ð832 Hist Econ J England. seventeenth-century in choice public erning York New Princeton Foundation, Sage Russell cooperation. and Competition (eds) E Ostrom M, Cambridge Press, versity 145:661Ð668 Econ Theor Inst J economies. Princeton Press, University Princeton sociology. economic for Handbook (eds) Oaks Thousand Foundation, Sage Russell sociology. in alism University Nations United WIDER. 125. paper cussion 96:903Ð918 78:738Ð749 Rev Sci Polit 157:57Ð78 Econ Theor Inst J approach. transaction-cost 42:673Ð684 Rev Econ Oaks Eur Thousand Sage, inquiry. political of Strategies (ed) E Ostrom approaches. 180 London Sage, sector. public Boston Brown, York New Sage, Russell institutionalism. intersection of choice points rational situations: and and historical Preferences between (eds) BR Weingast I, Katznelson In: cionalism. 49:70Ð88 Stud Polit approach. 94:175Ð189 Choice Public 27 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1196 1195 1194 1193 1192 1191 1190 1189 1188 1187 1186 1185 1184 1183 1182 1181 1180 1179 1178 1177 1176 1175 1174 1173 1172 1171 1170 1169 1168 1167 EDITOR’S1166 1165 1164 1163 1162 1161 1160 1159 1158 1157 PROOF1156 1155 1154 1153 1152 1151 enatB 20)Rtoa hieisiuinls.I:Ktnlo ,Mle V(d)Political (eds) HV Milner I, Katznelson In: Klinge- institutionalism. RE, choice Rational Goodin (2002) In: BR perspectives. Weingast choice rational institutions: security Political international (1996) to BR approach Weingast costs transaction a anarchy: 50:136Ð amidst Stud Hierarchy (1997) Polit K approaches. Weber Polit economic competing Theor politics: J in theory. Efficiency integrating (2002) P An Vanhuysse manipulation. transaction-cost Political al (1994) et C S Twight Steinmo politics. comparative in institutionalism 2:369Ð404 Historical Sci (1992) Pol S Rev Annu Steinmo politics. K, comparative in Thelen institutionalism organiza- Historical government (1999) K of Thelen autonomization the to approach cost transaction A (2003) H trans- Bogt A Argentina. Ter in policy public ap- of foundations transactions institutional a The (2007) policy: M public Tommasi PT, of Spiller foundations institutional costs. The transaction (2003) M political Tommasi of PT, impact Spiller the (eds) consolidations: B government Rockman Local R, (2006) Rhodes RJ S, Sorensen Binder In: institutionalism. choice Polit-Braz Rational Econ (2006) Rev core. KA the Shepsle to Political reform. state (ed) of economy HF political Weisberg The (1999) In: K institutions. Shepsle equilibrium and equilibrium 7(3):243Ð283 Econ Institutional Springer, Philos politics. (1986) Polit of regimes. KA authoritarian soul Shepsle Modeling and (2008) heart M the Levinson N, chaos: Schofield or Leadership (2011) al et M Gallego N Schofield Springer, voting. and democracy institutions, of economy Political (2011) 34:503Ð536 G Welf Caballero Choice N, Soc Schofield orders. Social (2010) N Schofield Amer- Latin in arenas Inter- and actors, 580, institutions, W.P. Political (2010) transactions. M Tommasi political E, Stein of C, Scartascini determinants institutional 15(3):173Ð194 The Perspect Econ (2007) J now. C and then Scartascini Law Cambridge economics: J Press, Institutional University (2001) Eur Cambridge M governance? economics. Rutherford in public Institutions to (1994) M applicable Rutherford In: economics science. cost political transaction Is contemporary (2005) in DWP institutionalism Ruiter Historical Rev (2002) Sci T Polit Am Skocpol politics. P, of study Pierson the and returns increasing dependence, Path Lon- (2000) Routledge, P the institutionalism. Pierson new on the perspective science: political cost in transaction theory Institutional a (1999) budgeting: G Peters of natural contractual York The New McGraw-Hill, (1996) action. E social of Patashnik structure The (1937) T Parsons 28 cec.Tesaeo h icpie otn e York New Norton, discipline. the of London state Press, The University science. Oxford science. political of handbook new A (eds) HD mann 41:321Ð340 Q Stud Int cooperation. 149 6(2):189Ð216 Univer- Cambridge Cambridge analysis. comparative Press, in sity institutionalism historical politics: Structural (eds) the in autonomization of 16:149Ð186 cases Econ six Law with J confronted Eur framework Netherlands. cost transaction political a tions: Cambridge Press, University Cambridge approach. cost action 19(2):281Ð306 Organ Econ Law J Argentina. to applications with proach 127:75Ð95 Choice Public London Press, University Oxford institutions. political of handbook Oxford 19:39Ð58 Econ Polit J York New Agathon, politics. of science the science: Political Heidelberg Berlin Inter- works. democracy How (eds) M Washington Tommasi Bank, E, Development Stein American C, Scartascini In: policymaking. ican Washington Bank, Development American York New Norton, 20:287Ð303 discipline. Econ the of state the science: Political (eds) Milner I, Katznelson 94:251Ð267 don 29:189Ð212 Sci Policy institutions. budgeting of design .CbleoadXC Arias X.C. and Caballero G. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1242 1241 1240 1239 1238 1237 1236 1235 1234 1233 1232 1231 1230 1229 1228 1227 1226 1225 1224 1223 1222 1221 1220 1219 1218 1217 1216 1215 1214 1213 EDITOR’S1212 1211 1210 1209 1208 1207 1206 1205 1204 1203 PROOF1202 1201 1200 1199 1198 1197 odB,BheJ(04 oiia rnato ot n h oiiso diitaiedsg.J design. administrative of politics the and costs transaction Political Lit (2004) Econ J J ahead. Bohte looking BD, Wood stock, taking economics: institutional new The (2000) OE perspective. Williamson economics cost transaction a bureaucracies: private and Public (1999) perspective. OE Williamson organization and economics, law, the con- realism: relational legal Revisiting markets, (1996) firms, OE capitalism: Williamson of institutions Press, economics Free The implications. (1985) antitrust OE and Williamson analysis hierarchies: legislatures, and why Markets or, (1975) congress; OE of Williamson organization industrial The (1988) WJ Marshall BR, Weingast Institutionalism New the of Map the in Politics Cost Transaction oi 66(1):176Ð202 Polit 38:595Ð613 15(1):306Ð342 Organ Econ Law J 5(2):383Ð420 Change Corp Ind York New Press, Free tracting. York 96(1):132Ð163 New Econ Polit J markets. as organized not are firms, like 29 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 Democracy nrae t agiigpwr hyso httems lsi etrwl etaxed be will sector elastic most Bates the emerge. that show to They parliament power. bargaining by its rule increases for easier ( it Lien made and that changes economic the oeso oiia rniin odmcayo nteetnino h suffrage the Robinson of and extension (Acemoglu centuries the Persico 20th and and on 19th Lizzeri or the on democracy focus to to tended have transitions political of Models Introduction 1 Keywords Magalhães De Leandro Narrative Analytic An Italy: Medieval and Greece Ancient in Transitions Political .D aahe ( Magalhães De L. 1 in back fight. go to we wars seem which as issues and policy war But contention to papers. the go to and these whether defence, in be is to intended good is- public These period defining infrastructure. the the as history, for such goods, relevant public are economic sues of provision the over and .Shfil ta.(eds.), al. et Schofield N. UK 1TN, BS8 Bristol Road, Woodland 8 Bristol, e-mail: of University Economics, of Department DOI hsppri opoiea nltcnraiet etwehrtemdli eMaga- Genoa. De and Venice, in Medieval model Athens, the Ancient whether of in test objective to The ( narrative 1688. Giovannoni analytic and of an lhães Revolution provide hap- Glorious to ( (what the Giovannoni is and points and paper case Magalhães threat this English De their In the and lost). on is fight), is war to focus players a wars when the com- players which to (the the available and citizens to policy war pens of the to assembly both go an determine to Wars to (whether paper). may power the that over in game hand elite bargaining to mercial the ruler model absolutist They an transitions. bring political explaining in role o ealddsrpino h ehdo nltcnraieseAis( Arias see narrative analytic of method the of description detailed a For ieaueo h itrcleegneo nlsv ntttoshsfcsdon focused has institutions inclusive of emergence historical the on Literature eMglãsadGoann ( Giovannoni and Magalhães De 10.1007/978-3-642-35239-3_2 [email protected] 1985 oiia transitions Political · Republic ,freape omlz h data h a lsiiyo sector a of elasticity tax the that idea the formalize example, for ), B ) 2004 dacsi oiia Economy Political in Advances srlvn oteudrtnigo oiia transitions political of understanding the to relevant is 2012) lvdradOxoby and Llavador ; pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , · Wars 2012 · nin Greece Ancient rps oe hr aspa key a play wars where model a propose ) 2005 , 1 ,dsue vrredistribution, over disputes ), · Athens 2012 · ). Venice 2012)the · Genoa 2001 31 · ; Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 osrit nteeeuie(ewl alsc eieadmcayo ueby rule or democracy a regime a such call considerable parliament). and will or institutional assembly (we council, by The assembly, executive rule bureaucracy. to the transition centralizing a on a is constraints here on financial in interested in or to are investments system, is we states, change papers judicial the these of a size of on the objective capacity as The such good. changes public institutional common-interest explain a is country the ( Arias and oiia oesmyb eesr opoietergteooi netvswhen incentives economic right the provide observable. not to is necessary effort be may state. powers the political to ( contribution Hanssen financial (1988 the quasi-voluntary and Levy of to Fleck form in power a made of include is and transfer must argument trade a stitutions similar observe on should dependent A we more classes. position agriculture, becomes commercial policy on economy preferred less the the and As to sector. manufacture closest elastic be most will the policy of equilibrium the that and less 32 opeettepo rmatn ntertra frvlto.Cnrr oAcemoglu to Contrary revolution. of threat ( their necessary Robinson on is acting and redistribution from High poor future. the the prevent in to poor the for redistribution higher sure ( Robinson ( Tilly (see capacity of build ( states role helped the and at state the look to need we themselves Medieval and transitions and wealth the Greece war. commercial ancient understand of both to for rise but true joint we holds Italy, the As parliament) occur. by of rule to predictions (or transition broad democracy a their for below, easier see it will makes environment economic particular a asta eetterlradhskn u o h omrileie xmlsof Examples elite. commercial elite’s commercial the the not for dynastic markets but the costly products. expand kin, that are his wars commercial wars and are misaligned wars there ruler aligned but of the won, example benefit is the key war that to aligned A wars an involved. and if ego-rents ruler returns no economic the high are both receive ruler to commercial the the return and both available: elite economic also are ego-rent little wars an aligned Alternatively, brings receives elite. ruler commercial this the but bias: intrinsic winning, an from have misaligned, called wars, Some at is ruler finan- the the when war, buys commercial defensive and a the during problem (where elites commitment commercial risk. commercial assembly the this the an solves by of to assistance role) preferred cial power wars leading over the the Handing to plays going elite future. to the commit in to elites unable be will ruler n iies eMglãsadGoann ( Giovannoni and Magalhães De citizens. and ( Morelli and Jackson 2000 h oe nD aahe n ivnoi( Giovannoni and Magalhães De in model The ae n in( Lien and Bates xesv ieauehsfcsdo o h heto a rv h omto of formation the drove war of threat the how on focused has literature Extensive asaeitoue nD aahe n ivnoi( Giovannoni and Magalhães De in introduced are Wars ,Bse n eso ( Persson and Besley ), ,weetehnigoe fpwri omtetdvc oen- to device commitment a is power of handing-over the where 2001), 2012 2001 ) nteeppr,awri omntra n h eec of defence the and threat common a is war a papers, these In )). ,Lv (1988 Levy 1985), ,D aahe n ivnoi( Giovannoni and Magalhães De ), 2006 ,weewr aedfeetrs-eadrto o rulers for ratios risk-reward different have wars where 2007), ou nacetGec oso htteetninof extension the that show to Greece ancient on focus ) 2009 ,GnaoiadVt ( Voth and Gennaioli ), ,adFekadHnsn( Hanssen and Fleck and ), 2012 lo o ifrn ye fwar. of types different for allow ) ulso cmguand Acemoglu on builds 2012) 1990 ou nwr.The wars. on focus 2012) ,HfmnadRosenthal and Hoffman ), 2011 2012 ,Bi ta.( al. et Boix ), 2006 ,weesal in- stable where ), ybidn on building by ) .D Magalhães De L. eciehow describe ) 2011 ), Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 nbevbeefr opatadmiti lv re nteAhna il htonly that hills Athenian input the in must trees owners olive maintain property and and plant Workers to problem. effort unobservable inconsistency time a mitigate h oiia rniin nacetGec n eivlItaly. medieval ( and Greece Giovannoni ancient and in the Magalhães gauge transitions to De political able the in be should model we merchants the questions, to of these the ruler answer relevance foreign that to a trying policy; preferred By merchants ruler? foreign the sitting that the on or ruler; aristocracy opinions their the from diverging that resources take evidence withheld had transitions there the Is merchants Did threat? the foreign wars. strong by and of played period was a any, during if place role, the what been check have and to assembly seems constraint main the but Venice fleet. role, For the citizens. a of the played financing by also manned was this Navy Genoa their the in and see, later government representative and will Army of we the steps As where first Greece, the wars. for during relevant effort particularly exerting seems model de- for who return citizen-soldiers, in the redistribution democratization—to mand of form the commitment—in ble ( Vindigni and Ticchi commercial the of who assistance ruler the a without by wars characterized defensive elite. is and which wars rule, dynastic absolutist on unstable Gio- goes of and period Magalhães a De by ruler). ceded that the is to were threat model they ( credible the vannoni (if no strength of be in military would predictions power intermediate there hand-over the of hegemonic to countries of willing in one be occur stronger should Therefore, may the transitions ruler assistance. under sitting rule their the either they for as to If return rule contender. transition self stronger a to the of or probability against contender, war the defensive increase assis- a they financial on so, withhold do ruler other to sitting with prefer the alliances then if to of alterna- may example, tance because elite the for (maybe commercial prefer wars satisfied, The ruler winning must is powers). sitting at elite foreign condition better the This commercial is ruler. that The ruler sitting invading threat lost. the their credible is to war a ruler invading the be tive if must replaced there be assembly, will an to power over oiia rniin nAcetGreece Ancient in Transitions Political 2 Narrative Analytic An Italy: Medieval and Greece Ancient in Transitions Political hr em ob la ikbteneooi ciiy npriua rd,and trade, Lien particular and Bates in as activity, such economic models ( of between predictions Greece. the link supports ancient clear evidence This in a democracy. government be to democratic the seems discuss towards briefly There let’s trends detail, general in Athens from of evidence reforms political the into looking Before 1985 pcfi oacetGec,FekadHnsn( Hanssen and Fleck Greece, ancient to Specific nsmay ewl otruhhsoia xmlso rniin orl by rule to transitions of examples historical through go will we summary, In with Italy medieval and Greece ancient in evidence the confront also will We ( Giovannoni and Magalhães De ,Lv (1988 Levy ), 2012 loso httastost ueb alaetaelkl ob pre- be to likely are parliament by rule to transitions that show also ) ,adD aahe n ivnoi( Giovannoni and Magalhães De and ), 2009 ,weetetra fwrhlsteeiemk credi- a make elite the helps war of threat the where ), 2012 hwta o naslts ue ohand to ruler absolutist an for that show ) hwhwdmcaycan democracy how show 2006) 2012). 2012 nunderstanding in ) 33 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 nossec so mle cl.Teewsltl cnmcgi o h Spartan the for gain producers. economic grain little to power was over There time handing scale. the in also smaller elite but a observable, of easily more is production inconsistency grain in workers the by exerted ig aetesle eopwr‘fteasml pk crookedly’. spoke assembly the two ‘if the and power Eventually, villages veto easier. into themselves phalanxes two gave population into should kings the organization Assembly the divided military full established also made a It that It which matters. and tribes, century. state Elders, on 9th 28 decision of the final council have in (Raaflaub a Sparta, (see sometime of Greece down Kings of hereditary laid rest Rhetra, the Great in the and Sparta farms) their in Wallace from equality off and of time the sense and some weapons created the afford could who (those landowners Vin- and Ticchi war. of during effort model their the of context the (2009 in digni understood better are changes These eopwr h iiaiaino l set flf,ptnilyas oig suggests Democracy. voting, fledged also fully potentially a life, not of was aspects Sparta all that of militarization the power, veto steeoemr ieyt rs h rae h an rmsligti ieincon- time ( this solving from gains Democracy the problem. greater this the are. mitigate arise problem to sistency to way likely The a investment. more is laborers’ therefore producers the is of the fruits to the power expropriate of aristocracy to hand-over the not because ante arises ex problem commit inconsistency cannot time The later. years fruit bear 34 3 2 Hanssen and Fleck of models the Neither production. surrounded grain and ( for coast soil the quality from away high Sparta, by in even appeared representativeness invariably of have would They that Greece. note trade. from to on off-shots dependent interesting and near highly is cities been Izmir), (coast It coastal Megara facing (Sicily). are Turkey), Syracuse island these of and all (coastal island), (coast Chios (Greek Pontica Naxos Heraclea Greece), Athens), Lybia), mainland of Greece), on (coast mainland be- coast Cyrene of period to (coast the Ambracia (next mainland cities—in (Sicily), of Argos archaic Acragas (coast (Sicily), Achaea some are: Croton that These Greece), evidence constitutions. is ( democratic Wallace there 480BC—had and fore how Raaflaub discuss land). 43) (versus wealth p. commercial of importance tive Thebes, and achieved Corinth Argos, as as such such soils, grains, oligarchies. richer were for with unsuitable cities soil whereas democracy, dry moderate with Cities model. their for 2006 2006 e aflu n alc ( Wallace and Raaflaub See ( Wallace and Raaflaub See lc n ase ( Hanssen and Fleck h dacsi afr rcieta e oHpiergmnsmne ysmall by manned regiments Hoplite to led that practice warfare in advances The h teineapecnrsswt praacrigt lc n Hanssen and Fleck to according Sparta with contrasts example Athenian The h te motn osdrto sta omo oenetwt oedegree some with government of form a that is consideration important other The ( Giovannoni and Magalhães De In .Sat’ atpan eeielfrgoiggan o nyi h effort the is only Not grain. growing for ideal were plains vast Sparta’s ). rD aahe n ivnoi( Giovannoni and Magalhães De or ) ,weepwri rne otectznslir nodrt guarantee to order in citizen-soldiers the to granted is power where ), 2007 .3).I prati oktefr ftesto ascalled laws of set the of form the took this Sparta In 37)). p. , 2006 2007 2007 .34). sources. p. primary , and details more for 39) p. , xedteraayi oohrcte n n support find and cities other to analysis their extend ) 2012 2012 n ftekyvralsi h rela- the is variables key the of one ) 3 xli hs ntttoa changes. institutional these explain ) .D Magalhães De L. 2 eie this Besides 2007, Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 l oi rbsi tia(h einsronigAhn)it e e rbsand, tribes new ten into four Athens) the surrounding reorganized region 500, (the Attica to in assembly tribes the Ionic extended old 508 in reforms Kleisthenes’s 508BC Kleisthenes, 2.1.2 lsl eae otepltcleooymdlpooe yFekadHanssen and Fleck by proposed seem model reforms political-economy maximize institutional the to ( incentive Solon’s to an of related created interpretation trade closely as The poor, production. the and agricultural elite the between possibly rdbetrast h ueo h pra lt.Wtotsc hetteewsno was there threat a few such power. been Without over have hand elite. to Spartan to elite the seems the of for also There incentive rule but land. trade, the of on to lack threats least the credible hegemonic—at to only becoming not Sparta due be to could Sparta in institutions democratic ae yafrintra,btaedsrbda h euto oi-cnmcstrife socio-economic of result the as described ( are Osborne but Athens. threat, within foreign a by also vated Solon offices). pow- executive (the limited land-owning Archons with law. nine top criminal tribes) the the and Ionic of to civil offices codified four restricted the probably the were was of as Assembly classes, each the from in (100 Participation ers. 400 of Assembly an infiepteyadapoa ue otasotoieoladwn)hdbeen had wine) Osborne and to oil According onwards. olive 700 Athe- transport around time. from to ( that Mediterranean (used the during over amphorae rapidly all changing and found was pottery Athens fine of ( nian economy Osborne the Moreover, how elite”. scribe the between conflict ter t oete epahre pno xn rneither. or oxen, of span capac- a the no horse, on was a and keep output, Status agricultural either of to life). amount ity to the public instead by linked in measured was been was but place aristocracy, which had hereditary wealth, a status a to how therefore belonging change by (and to determined longer society was 594 Athenian in in reform defined Solon’s of innovation main The 594BC Solon, 2.1.1 three Ephialte-Perikles’ of and consisted 508, in accounts, reforms most 462Ð450. Kleisthene’s in by 594, reforms has, in Athens reforms Solon’s in steps: democracy to transition The Athens 2.1 Narrative Analytic An Italy: Medieval and Greece Ancient in Transitions Political 4 2006 2009 .3) o ute eal n rmr sources. primary and details further for 30). p. (1991, Hansen See ihnteitrrtto fD aahe n ivnoi( Giovannoni and Magalhães De of interpretation the Within h eom fSln(nieteltrrfrs ontse ietymoti- directly seem not do reforms) later the (unlike Solon of reforms The ). hsnwtaerltdwat eeae optto ihnteeie and elite, the within competition generated wealth related trade new this ) 2009 .21 ecie oo’ ol s“ ol fbit- of world “a as world Solon’s describes 211) p. , 4 oo’ eom loincluded also reforms Solon’s 2009 .23 oso ode- to on goes 213) p. , 2012 h akof lack the ) 35 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 sebyo h 0 aeacersyi oeg oiyt h olt lse all classes hoplite Persia. the with to himself alliance Kleisthenes policy an with foreign use supporting into in for put immediately say ostracized was clear power new a This gave Attica. over 500 the of Assembly Spartan the and Isagoras defeat implemented. to subsequently were able reforms were Kleisthenes’s and forces. rioted into Kleisthenes people forcing described Athenian and reforms The Isagoras political exile. support to the again proposing Athens by invaded de- Sparta support was above. who popular Kleisthenes, gather Archon. to elected tried was feated, Sparta, with alliance an favored who ue.TeeGnrl omne h ryjitywt h oeac oeo the of (one Polemarch the Archons). with jointly nine Army the commanded Generals These duced. bet atr ipa’cide;i xhnefrtehsae ipa eti exile in use went Hippias made hostages was (who the and for Sigeion. Hippias exchange again in in to attacked children; Sparta lost Hippias’ Athens). capture and to defend 511 able to in support mercenary Athens Thessalian attacked of Sparta 510. until 561 transition. this ( understand Giovannoni and Magalhães De in would els Perikles and century. century 5th the 5th in early later the post repeatedly in General the Archons elected hold be nine would the Cimon within and great Polemarch Themistocle yielded of Both Generals policy. The Athenian offices. over public influence other most to repeatedly—contrary held .27 o oedtisand details more for 277) p. (2009, Osborne see Delphi, of oracle the sources. primary to due was intervention 5 period certain property. a or allowed for title city which losing the leave ostracism, without to time of figures of law political important the confined for introduced longer vote no to also were Athenians Kleisthenes and Attica rights itself. of Political (villages) Athens regiment). demes one to the supply a all to of to had part extended tribes be also ten were to the afford of could (each who regiment those hoplite to rights political extended importantly, most 36 11 10 9 8 7 6 teswudb nipratadto oSat’ ewr fale gis Argos against of allies size of and network Sparta’s wealth to city-state). The rival addition Peloponnese. (a important the an be beyond would influence Athens its increases to policy hi noua tyrants. unpopular their teinsoisaottefl fHpisete mtSatnitreto rmninta the that mention or intervention Spartan omit either Hippias of fall the about stories Athenian ( Osborne See sources. primary ( and Osborne details See more for 34Ð35) sources. pp. primary (1991, and Hansen details See further for 35) p. (1991, Hansen See e onr n aos(1991 Samons and Fornara See ( Osborne See litee’ eom a motn iiaycneune.Tenwyformed newly The consequences. military important had reforms Kleisthenes’s Isagoras, power. for fought factions Athenian exile in Hippias tyrant the With teswsudrterl ftetrn essrtsadhssnHpisfrom Hippias son his and Peisistratos tyrant the of rule the under was Athens mod- both and threats foreign to linked directly were Kleisthenes of reforms The 7 son ( Osborne .27 o oedtisadpiaysucs e loHne ( Hansen also See 275). sources. p. primary 2009, and details more for 277) p. 2009, 2009 6 h eeaswr lce ypplrvt n h otcudbe could post the and vote popular by elected were Generals The .28 o oedtisadpiaysucs e loOe ( Ober also See sources. primary and details more for 278) p. , 8 nteps,Sat a eeae lisb eieigcte from cities delivering by allies generated had Sparta past, the In 2009 9 ugssta prasmtvswr ato deliberate a of part were motives Sparta’s that suggests ) .5)frmr eal n rmr sources. primary and details more for 56) p. , 2012 5 n51 or ftnGnrl a intro- was Generals ten of board a 501, In n iciadVnin ( Vindigni and Ticchi and ) 11 h raiaino tiain Attica of organization The 10 .D Magalhães De L. epus help 2009) 2007). .36). p. 1991, Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 urn ree u h xldtrn ipa emdt ei linewt the with alliance in be to seemed Hippias tyrant exiled 6th Persians. the late but the Greece, in quering rights political of extension war. the a understand during to effort century. way exert another citizens is the and This redistribution) guarantees (which power htpltclpwr ol eetne oteTee n hti hteventually what is that and Thetes the Perikles. and to Ephialtes extended under be happened would powers political Magalhães De that in models Both Thetes. ( the Giovannoni class, and property-less the from come to have nta iie-odesms xr nbevbeefr nwr nTch n Vindigni and Ticchi In war. in effort ( unobservable exert must citizen-soldiers that in which transition, political a trigger to to was not ( place. aim chose took Giovannoni their eventually model) Instead, and the Athens ruler. in Magalhães on their elite attack defend De commercial help (the Spartan of Athenians The the context rule. where the war their the defensive Sparta, in defend that against interpreted, to noteworthy fight mercenaries be to is on can themselves It rely Athenians regiments. to the had Hoplite summon but to the able man not and were tyrants finance both to the needed in generals of board elected an of creation 501. the year with further even increase would e rbsad19dmssrntee n oenzdteAhna ry reducing mercenaries. army, Athenian on the modernized dependence and their strengthened demes 139 and tribes ten Narrative Analytic An Italy: Medieval and Greece Ancient in Transitions Political 16 15 14 13 12 itr nanvlbtl gis esaa aai.I 7,teDainlau was league the Dealian led the Athens 478, Aegean. 480, In the In Salamis. in Navy. strike supremacy at Athenian naval silver Persia the Athenian recent against solidifying to a battle created added from naval be revenue a the to in use triremes victory to 100 harbor later for the and pay for fortified, pay to be to Assembly to the Peiraieus persuaded of Archon as Themistocles Persia. with hetpritda iia( rseosiln iaigAhn ncommerce) in Athens rivaling island prosperous (a Aigina as persisted threat mle raiigams 000mnt o them. row to men 20,000 almost de- organizing a foreign implied was of Navy choice deliberate the a on and to investment power been naval The have Athenian policy. windfall transfer. would strength silver sum alternative to the lump move The invest Assembly. liberate a to the Athenian decision by each the approved that pay was notice Navy to re- the important democratic on Athenian also the is of developments It further forms. the understand to important is 2009 e son ( Osborne See (1991 Hansen See ( Osborne See (1991 Hansen See ( Osborne See h hett tesrmie ih o nywr h esasitn ncon- on intent Persians the were only not high, remained Athens to threat The h teinam a aall ihtems riso h al 0hcentury, 20th early the of armies mass the with parallels had army Athenian The citizens Athenian the to elite the from power over handed reforms Kleisthenes’s h oiia osqecso hstr otesawr la s10trimeres 100 as clear were sea the to turn this of consequences political The ,etra het aea qiiru osbe hr h lt ad over hands elite the where possible, equilibrium an make threats external ), 13 h e teinam eetdtePrin tMrto n40 The 490. in Marathon at Persians the defeated army Athenian new The 2009 2009 2009 .36). p. , sources. primary and details more for ) n i atclr iciadVnin ( Vindigni and Ticchi particular) (in and 2012) .310). p. , 308). p. , 279). p. , 12 h oe ftepol vrfrinaffairs foreign over people the of power The 16 oto hs e would men these of Most 15 ol predict would 2009) hstr otesea the to turn This ,a a as 2012), 14 sided 37 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 h atbsino h rsorc neto linewt pra h rnfrof transfer goes The Athens and Sparta. Navy, the with finance alliance who those on of intent weight relative Aristocracy the the increases power of Giovan- bastion and last Magalhães the De interpret of can model the we ( within noni that Ephialtes suggests of transition evidence political This the Kleophon. Hy- lamp-manufacturer lyre-maker Kleon, and tannery-owner perbolos, examples: ( following Hansen the wealthy. After gives considerably Aristocracy. still 39) the but includ- p. of birth, and lower part of were to often Athens up were in they Indeed, Perikles leaders com- policy. political the foreign main gave over the system control Perikles, ing political of degree Athenian some the elite that mercial expect should we policy, foreign solutions. an n omn o ieaohrt omn)atieefroeya i some (in year one purpose). for this trireme main- for for, a together pay command) pool to would to individuals citizens another rich wealthiest hire cases the (or required command that and tax) Navy tain, rotating the financing (a in liturgy component a important An was regiment. Hoplite a financing of that en wy h sebywstle oad h ors citizens. poorest the Hoplites towards of regiment tilted to large was a Cimon to Assembly of due the that willingness is away, the element being key against other also The those were Spartans. issue: the reforms key the assist Hoplite a political which was large policy the under foreign a proposed conditions over with dispute that The a revolt. away that Helot shows was approved a was Cimon suppress reform Athe- while Sparta the help passed of to General was contingent and reform Archon these The an of Navy. Cimon, Opposers nian and Assembly). Aristocracy the institu- the (mostly other included Demos aristocracy—to reforms the owning Areopagos—the land of the the representative from of more power bastion tions transfer a and to body reform judiciary a main proposed Ephialtes 461/2 In 462BC Ephialtes, 2.1.3 38 19 18 17 eswr rn ofe-iigpolm,adtxaodnewscmo.Christ common. was avoidance tax and problems, free-riding ( to prone were tems rd.Wti tescmec a nne ymrtm on n togcom- strong a and loans maritime emerged. by elite financed mercial was commerce Athens Within paid trade. be to started they as life, public in part who active it. more and for a Navy, take and the to after re- for Thetes soon a followed the essential reforms allowed as were Perikles’s Empire. who role from naval Thetes, directly the Athenian benefited therefore of prominent empowerment a the for of support sult the and reforms mocratizing eutdi itrdsuewt io srczda etidt ees h reforms, the reverse to ( tried Raaflaub he assassinated. as eventually ostracized Ephialtes Cimon with and dispute bitter in resulted 1990 e asn(1991 Hansen See ( Raaflaub See ( Raaflaub See o h elhetidvdasi oit oqaivlnaiyfiac Athenian finance quasi-voluntarily to society in individuals wealthiest the For ihEpr,Ahn eaetecne falrentoko Mediterranean of network large a of center the became Athens Empire, With ecie ndti h xeto h a viac rbe n h attempted the and problem avoidance tax the of extent the detail in describes ) 2012 .Ehats(isl naitca)rdcstepwro h Areopagos, the of power the reduces aristocrat) an (himself Ephialtes ). 2007 2007 .10 o oedtisadpiaysources. primary and details more for 110) p. , .18 n ilt( Millet and 118) p. sources. , primary and details for 113) p. , 18 h nnigo h aywscnieal ifrn from different considerably was Navy the of financing The 1983 o eal n rmr sources. primary and details for ) 2007 .12 xlisteede- these explains 122) p. , 19 fcus,sc sys- such course, Of 17 .D Magalhães De L. h reforms The 1991, Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 yatn meo ihrwtaigpiiee rne oVnc htfre the King formed The wealth. that 1024. Adriatic. Venice Venetian to in the of granted Constantinople along privileges backbone cities in trading ap- Venetians withdrew exile political conquered Emperor to and Byzantine and religious sent attacked contentious and swiftly Grado, broth- further ousted Stephen of mar- two to Patriarch was and placed Due as Otto Venice, Torcello. 16 Otto in pointments, of at 1017, positions Bishop dodge In religious as Hungary. important a most and of made third Em- two Stephen was his the Byzantine King Otto raised in the ers Pietro of retired. of 1005, daughter and niece in the dodgeship the son ried the to eldest success- son to his very eldest of Otto, a his death son, was premature marry the II to came With able Orsoleo Venice perors. was Pietro family, monarchy. and Orsoleo ruler hereditary the ful a With becoming succession. to for close them setting fathers, their families. nteae eeBznieofiil pone yteEprri Constantinople in (the Emperor assembly the general a by in appointed officials ( Byzantine (Lane were area 1172, council. his the place of in took decisions the reform by political abide key to were constrained The was there 1032. dodge but illustrate the monarchy, in 1032 when hereditary changes around a institutional events becoming clear checks The to no considerable 1172. close came and with monarchy dodgeship 1032 system the elected executive: Republican how the from a on transition to Venetian balances heredity) and the of in degree dates some (with important two are There Venice 3.1 Genoa and Venice Medieval in Transitions Political 3 and Kyriaziz Sparta. with conflict and hegemony maritime ( of Zouboulakis policy a pursue to on Narrative Analytic An Italy: Medieval and Greece Ancient in Transitions Political 20 interests. commercial their were further Sparta) to power and yield (Persia and threats financ- gain foreign in to constant able indispensable under stability was and its elite defences, and commercial Athenian democracy ing The conquest. Athenian Macedonian of more the consolidation and ( until the more Giovannoni and understand depend Magalhães to to De propriate starts of wealth model the Athens’ commerce, and on de- instrument as military war (2009 main into Vindigni effort the and put Ticchi and participate in both scribed to individuals for incentives creating Navy. Athenian the of needs financial the to linked e ozlsd aa(2011). Lara de Gonzáles see o ecito ftesl nocn ntttosta rmtdtaei eiei hsperiod this in Venice in trade promoted that institutions enforcing self the of description a For ihtm,pwru ogswr bet as hi ost uetgte with together rule to sons their raise to able were dodges powerful time, With eiebgnishsoyudrcnrlo h yatn mie h rtrulers first The Empire. Byzantine the of control under history its began Venice nsmay h oiia rniint eorc nAhn a la oein role clear a had Athens in democracy to transition political the summary, In ) yte8hcnuy eiewseetn hi ruler their electing was Venice century, 8th the By (2003)). Norwich 1973), 2004 loageta h ieo nuneo h omrilcasis class commercial the of influence of rise the that argue also ) concio 20 or ihsc xenlpesr,teitrmdodge interim the pressure, external such With Arengo .Hwvr steAhna aybecomes Navy Athenian the as However, ). otlkl oiae ytepowerful the by dominated likely most ) 2012 em oeap- more seems ) 39 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 osrit nteeeuiecm n17 n ol elne oetra threats external to linked Navy. the be Venetian consolidate the would would of and financing that 1172 the law sixty-two in to the for and came in family executive change Michiel the The the on years. of constraints seventy-six members last by the held of been out had dodgeship the that elected an as rule the to of (2003) continued power executive dodge Norwich The monarch. the sons. onwards, and their intact, 1032 to remained dodgeship however, From the dodgeship, monarchy. passed fathers that hereditary no signal a that clear notes a favor sending in not were Venetians change did the a pretensions they choosing dynastic was By no There nepotism. the with regarding dodge. dodge specially described ruler, a the a and to for elections behavior counterbalance acceptable for was a what called as already councillors reform law of clear Existing positions no then. was next law there old Venetian (2003) their the to Norwich in as link According no Venice. chose with of Venetians silk-merchant families wealthy The powerful a Venice. 1032, in in Flabanico rule Domenico dodge hereditary establish to attempt i f40wst enmntdb h egbrod fVnc ohl office hold to Venice of neighborhoods the by nominated be uncon- Coun- to whose Great A dodge, was in. the herself 480 found of Venice of position power institutions the cil the Roman for The blamed difficulties. western constrain was in and to power Navy strained eastern designed a the were had both followed and financially, with that straits war dire at in now Empires, were They in reforms. murdered the and litical ousted to was burnt) He in be city. Venice to the to had to return Plague (that streets. to the ships the brought forced and also was men the he Vitale During loose but time. Vitale and Plague, gain did fleet; to only the Emperor Not the in Vi- humilation. by spread Dodge ploy plan. Plague a peace be the a to wait, out out asked work turned Emperor and what Byzantine Constantinople ex- accepted the to were tale sea, go men at to able already embassy fleet all Venetian the and a With contribute, Navy. a to the ordered man had Vitale to means Dodge pected with debts. citizen previous every for pledged loan: been forced already had decade a least ( Norwich ditions. and Constantinople in confiscated property Empire. Constantino- the and opposite of citizens settlement ports Venetian other Genoese all (the had Galata for and at Venetians the ple) Genoese blamed the Emperor on The break-down. attack of an point the at were Venice and nosuemme fteOsoofml re osieteddehpbtwas but dodgeship the seize to Venice. tried to family return Orseolo his the before ousted. of died but member back, obscure called An was Otto abdicated. Centranico 40 21 o oedtisseNrih(2003 Norwich see details more For owc ( Norwich ( Lane 1172, by monarchy hereditary a to aversion this with Even family’s Orseolo the to response a as came Venice in reform important first The eoeimdaeyeetn e og,teVntasdcddt moepo- impose to decided Venetians the dodge, new a electing immediately Before con- financial strenuous under preparation war the led Michiel II Vitale Dodge 21 2003 hp )dsrbshw n17,rltosbtenByzamtium between relations 1171, in how, describes 8) Chap. , 2003 hp )mnin htaltervne ftesaefrat for state the of revenues the all that mentions 8) Chap. , hp 5). Chap. , .D Magalhães De L. 1973 remarks ) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 nuneo h oyRmnEprr ob ue yaPdsa oege who foreigner a Podesta, year. a one by of ruled mandate be a with to the Genoa (under Emperor) rule 1190 Roman by would agreed Holy families power the the Since of clans, (clans). key influence the families of noble any different with between consolidated never war civil into escalate ularly of None aristocracy. In the stable. 1339. and proved in rulers, government Genoa foreign of of forms podestas, Dodge these by first ruled the of or was as Captain council Boccanegra Genoa as by Simone Boccanegra between and rule Gugliemo 1257 of to in rise transition People the a the are for candidates date best accepted The historically parliament. clear no has Genoa Genoa 3.2 done a as people the to choice their nominated present Council ( and of Norwich Great dodge (see citizens the the deal all directly, choose the comprising to dodge including and electors the state, Arengo, 11 nominating the the of of by officials Instead elected chief Venice). officially the then nominate (until thereafter dodge and year one for Narrative Analytic An Italy: Medieval and Greece Ancient in Transitions Political 22 objective essential the powers state. with foreign the the not of antagonizing wealth and of the or mind, monarchy, for in hereditary a interests up would their they setting power, with of in policy Once state. foreign the of over of defence hands decide the the finance in to and entrenched afford threat was could policy external who foreign those high over of power period the a straights, from financial During away dire populace. and merchants, the wealthy and the quasi-nobility, state, old the the of financiers the to power transfer ( to Giovannoni and Magalhães De in model Venice when point a at threat weakened. consolidation foreign militarily the considerable was that under done so was terms, regime the new accept the sued to of immediately refused also who Venice Byzantium, dodge. with (Venetians new peace creditors the there for under the seems State of It the willingness finance Navy). to a the themselves) demonstrates finance which to resentment, loans little forced was men the wealthiest (from the bonds of government new one ( Norwich was to Ziani According Sebastiano Venice. Dodge in Republic. its the strengthening of financiers while pyramid administrative the in of power base gained the Senate center’. and The apex dodge. the coun- the The both (2003 six. restrain Norwich to to two affairs. power from foreign given councillors of also number were the cillors increase to was reform other podestaria. e pti (1996 Epstein See hogotishsoy eo swl nw o nenlsrf htwudreg- would that strife internal for known well is Genoa history, its Throughout h oiia eomi eieo 12cnb etudrto nlgto the of light in understood best be can 1172 of Venice in reform political The the towards power in change a reflected clearly dodge next the of choice The o aetertcaayi fthe of analysis theoretic game a for (2006) Greif and details more for 88) p. , 2003 hp )frdtisadLn (1973 Lane and details for 9) Chap. , nepesteefc fteerfrst ‘weaken to reforms these of effect the interprets ) 2003 hp ) in upne amn nthe on payment suspended Ziani 9), Chap. , .Teerfrsse ob designed be to seem reforms These 2012). 22 p 511) The 95Ð101)). pp. , 41 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 ti ( ( stein Epstein what particular in the people, called the and families noble the between eca lt rfrafrinrt hi itn ue.Ti em ohv entecase com- the the been that have be to seems must This it Genoa. ruler. credible sitting in aristocratic be their repeatedly to to (or foreigner threat monarch a the prefer current For elite Genoa). mercial the of replace case necessary to the A in ready families rule. someone foreign ( threat, support Giovannoni outside and to Magalhães ible classes De merchant in transitions the political sec- of for The condition will war. (commercial) the aligned is an aspect versus ond war (dynastic) misaligned a ob- ( of commercial Giovannoni choice following and and Magalhães finances De stable abroad. classes with jectives merchant government the stable versus a rule in aristocratic interests impose to attempts and concerns nastic ( Giovannoni and hães France later and 1331Ð1335, example, from key Naples the of King is Milan. the then podesta 1311, and The in choice. ruled by VII but Henry but conquered, were they because Not the or poor the not artisans, master and people traders ‘the new (Epstein middling towards elite’ the a geared traditional office, and were in policies People him regime’s the put new who of The Anziani. Captain 32 as of Boccanegra council Gugliemo elected revolt popular n eido ntberl yteodnblt,tepol eotdadcetda created and revolted Gugliemo). people of the (grandnephew nobility, Boccanegra old Simone electing the Dodge by of rule position new unstable of period (probably a fell and he nobles). before the years of five some by for coup office a in 147). to (p. stay due state to Genoese the creation was finance the Boccanegra to to securities’ led Gugliemo public and for debt markets state ‘precautions on a default of the preventing at aimed was Boccanegra a lot alb 34udrteimnn tako nalac aeu fnobles of up made alliance an power. Boccanegra of from attack Simone excluded imminent commercial money’. been the making for had under who of 1344 Pera by task fall to congenial to out more also was the was to fleet war new civil from a without debt 1340, ( public By Epstein a consolidate ventures. debt. is to of old none had master any governor councillors: again a new repudiating The and and ducal profession. maker, defences 22 by shield themselves Genoese the identified a strengthened that of butchers, those three of 16 guild’ drapers, for wool ‘two the details are have there ( Epstein we and nobles. noble; the that of not notes and classes 205) merchant p. the of rule the was this Again, 42 24 23 e pti (1996 Epstein See (1996 Epstein See eie h oflc ewe ifrn ol aiis hr a loaconflict a also was there families, noble different between conflict the Besides h vnsi eo ihih w motn set ftemdli eMagal- De in model the of aspects important two highlight Genoa in events The foreigners. by ruled repeatedly was it that was Genoa of characteristic Another n13,atrapro ne oeg ueb oetAju igo Naples, of King Anjou, Robert by rule foreign under period a after 1339, In 1996 ooograsso popolo .17 ecie h vnso 27 fe rs nteeooya economy the in crash a after 1257: of events the describes 137) p. , .208). p. , 204). p. , 1996 .Tefis stecercnitbtentenblt’ dy- nobility’s the between conflict clear the is first The 2012). .27 oe htteeplce eetda‘unn away ‘turning a reflected policies these that notes 207) p. , h ihmrhnswowr o ato h oiiy Ep- nobility. the of part not were who merchants rich the , 1996 .18.Itrsigy n ftefiaca eom of reforms financial the of one Interestingly, 138). p. , 24 2012 oe hscnitwt the with conflict this model ) 2012 .D Magalhães De L. .206) p. 1996, sacred- a is ) 1996, 23 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 rniint ueb alaetsol epeee ya ntbepro hr the where period unstable an by ( preceded be Giovannoni should and parliament Magalhães by rule De to of transition model the to specific Ma- De is transition of political of merchants ( aspect Giovannoni wealthy this and were explains galhães best who in that dodges and model chose The 1172 1172) birth. of lower and changes (1032 institutional moments the key proposed both above. families (Kleis- discussed powerful aristocracy reforms the the state institutional Venice, of the the In proposed members that plausi- Perikles) which a so and in seems Ephialtes, elite, This Athens, thenes, threat. in commercial foreign events the a of against to interpretation itself ble or defend to people funds the enough raise to can either power over hand etvspae e oei rvnigasal omo oenetfo appearing from government of ob- form dynastic stable competing a Genoa. with preventing in aristocracy in role divided key a a that played clear jectives seems It the ruler. in single ( shortcoming a Giovannoni a and recognize Magalhães to De (as is in difficulties explanation model financial alternative and An threat 1172). foreign in hostile did Venice a faced country ( the Giovannoni when and moment Magalhães De of logic nVnc n ae nGna h oe htbs ep sudrtn hs driving these understand us helps (2009 Vindigni best both and that dodges, Ticchi their model is chose The must to forces Genoa. forces voice in a Similar given later Athens. was and the Venice populace in of the in Thetes) how inclusion in (the political role class the a landless for played have the also and of for Sparta, later key in been and have Rhetra Hoplites to Great seems the motive of This citizens. creation own the state’s the by manned army an also was this production; grain Genoa. for and Venice soil for (2006 inadequate true Hanssen and and insufficient Fleck little had As had Athens sea. have like the to to seem they turn Moreover, to times. but those choice of that cities worlds executive trading medieval the leading evidence and on the ancient constraints were The the considerable flourished. with in trade institutions cities representative The where developed prediction. arise this to support to likely seems more with are ( sectors and governments Lien economic Venice, tative and the Bates Athens, of of ( importance in papers Hanssen in The trade growth elasticity. international relative tax economy— high and a the to Athens, in problem due in inconsistency Genoa—or time production condi- a economic solve oil to help response olive to driving a order One as in transitions. institutions either representative political tions: of for creation forces the driving is different force of emerges picture A 4FinalRemarks Narrative Analytic An Italy: Medieval and Greece Ancient in Transitions Political eas on vdnet upr w set fpltcltastosta are that transitions political of aspects two support to evidence found also We to chooses elite aristocratic established an as about come also may transition A h a hr osal rniint ueb sebyi eo?Wti the Within Genoa? in assembly by rule to transition stable no there was Why oiia rniinmyas oea ouint rbe fhwt motivate to how of problem a to solution a as come also may transition Political ,adD aahe n ivnoi(2012 Giovannoni and Magalhães De and 2006), 2012 ). ). 2012 2012 hr h rsorc smdldas modeled is aristocracy the where ) ,teesest aebe oclear no been have to seems there ), ,Fekand Fleck (1988), Levy 1985), oe,acetcity-states ancient notes, ) 2012 rdc htrepresen- that predict ) .Tefis sta a that is first The ). 43 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 rlzdpwrmyhv ob salse eoei a ehanded-over. be can it Cen- before parliament. established be by to rule have through may executive power tralized the constraining towards step intuitive enil n oh(2011 Voth and Gennaioli onr W aosL I(91 tesfo litnst eils nvriyo California CREI-Barcelona paper, Working of conflict. University military and Pericles. capacity State to (2011) ancient H-J Voth Cleisthnes N, to from Gennaioli application Athens with (1991) model II LJ a Hill, Samons democracy: Chapel CW, of Press, Fornara origins Carolina The North (2006) FA of Hanssen University RK, genoese. Fleck the and Genoa by (1996) rule to SA absolutism from Epstein Assoc transition the Philol and financing Am War Trans (2012) F Athens. Giovannoni Political L, classical In: Magalhães in De states. antidosis of and formation avoidance the Liturgy and (1990) wealth MR War, Christ politics. (2011) and R taxation, Giovanni rights, B, property Codenotti capacity: C, state Boix of origins The (2009) Social T government. Persson representative T, and development Besley taxation, on note A (1985) D Lien RH, centralization. Bates to fragmentation 91(4):938Ð963 from Rev Mexico: Econ colonial in Am capacity transitions. fiscal political Building of (2012) theory LM A Arias (2001) JA Robinson D, Acemoglu References Byzantium. against Navy the to fund had to dodge order The jeopardy. in loans in forced empire Western to in- the resort commercial and eventually Venice, key Venice’s Byzantine In put the that supporters. both policies in his followed terests con- and Michiel a Ephialtes II was Vitale by revolt Dodge opposed Helot the move a against containing city policy the in foreign defend Sparta tentious to to of mercenaries Tyrant assistance The on Cimon’s merchants. rely Later, the to Sparta. or had citizens example, the for of Hippias, support Athens, the without war to goes ruler 44 fVnc eeal o hscudsgetta h etaiaino oe a de- (as power of centralization the that suggest (1990 dodges Tilly could the in This or scribed Athens to. of able tyrants were the Genoese Venice that interests. way of dynastic the in with power clans consolidated different never leaders between strife internal was system ( Giovannoni and Magalhães De constrained was dodge the of Byzantium power both the from when 1172. lose Empire in eventually threat Roman would direct western it under the whom was from Athens Venice and (to threat. and Sparta external war), and credible Peloponnesian Persia a the both intermediate be from of must threats states there clear in because faced only is transitions This political strength. observe military should we that predicts rs,Berkeley Press, 49(1):115Ð146 Econ Law J Greece. 958Ð1528 pp University Bristol Mimeo, parliament. 120:147Ð169 45Ð74 pp Heidelberg, Springer, voting. and democracy institutions, of economy 99(4):1218Ð1244 Rev Econ Am Caltech 567, paper working science Madrid Institute, March Juan paper, Working h eodapc sta h oe nD aahe n ivnoi( Giovannoni and Magalhães De in model the that is aspect second The ial,Gnapoie neapeta hwdtelmttoso h oe in model the of limitations the showed that example an provided Genoa Finally, ,HfmnadRsnhl( Rosenthal and Hoffman ), ,adAis(2012 Arias and ), .A motn seto h eos political Genoese the of aspect important An 2012). )myb nipratadcounter- and important an be may )) ,Bse n eso ( Persson and Besley 2000), .D Magalhães De L. 2009 2012) ), Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

690 689 688 687 686 685 684 683 682 681 680 679 678 677 676 675 674 673 672 671 670 669 668 667 666 665 664 663 662 661 EDITOR’S660 659 658 657 656 655 654 653 652 651 PROOF650 649 648 647 646 645 il 19)Ceco,cptl n uoensae.Bakel Oxford Blackwell, states. European Lucca and IMT paper, capital, Working Coercion, democracy. (1990) endogenous C and Tilly War (2009) A Vindigni D, In: Ticchi Greece. archaic in trends egalitarian and power” “People’s of (2007) Origins RW In: Wallace Athens. KA, mid-fifth-century Raaflaub in London demokratia Routledge, of of edn. breakthrough University 2nd The Greece. BC, (2007) 1200Ð479 ancient making KA in the Raaflaub democracy in Greece of (2009) Origins R In: Osborne man’. that besieged Econ ‘I in Trade London (2007) J Penguin, In: J Venice. Athens. Q of Ober history fourth-century A franchise. in (2003) credit JJ the of Norwich structure and the and growth, loans Maritime competition, (1983) Partisan P Millet (2005) of scope RJ the and Oxoby Democracy suffrage? H, the extend Llavador elites the did Berkeley Why Press, (2004) California N of Persico University Baltimore revenue. A, Press, and University Lizzeri rule Hopkins Of Johns (1988) The M republic. Levy maritime a Venice (1973) FC Lane economic an 97(4):1353Ð1372 change: institutional Rev and Econ power Am sea Democracy, war. (2004) and MS Zouboulakis bias NC, Political Kyriaziz (2007) taxation. M and Morelli warfare M, of Jackson economy political the fall: we Divided (2000) and J-L principles Rosenthal structure, PT, Demosthenes: Hoffman of age the in democracy Athenian The New (1991) Press, MH University Hansen Cambridge economy. modern the to path the and Institutions (2006) A Greif lit- enforcement: contract for institutions public-order Self-enforcing, (2011) Y Lara de Gonzáles Narrative Analytic An Italy: Medieval and Greece Ancient in Transitions Political rgn fdmcayi nin ree nvriyo aionaPes odn p22Ð48. pp London, Press, California of University Greece. ancient 2 Chapter in democracy 5 of Chapter 105Ð154. Origins pp London, Press, California of University Greece. ancient in democracy 4 Chapter 83Ð104. pp London, Press, California 36Ð52 pp London, Press, Hogarth The Ð Windus & Chatoo economy. ancient the 119(2):707Ð765 Econ 120(3):1155Ð1189 J Q reform”. of “Age Britain’s to application an with government, 17:117Ð132 Econ Law J Eur law. naval Athenian the of analysis paper working Caltech Oxford Blackwell, ideology. of economy York Political In: 95Ð115 pp 1050Ð1350. Heidelberg, Venice, Springer, voting. in and government democracy institutions, limited and regulation, igation, 45 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 .Shfil ta.(eds.), al. et Schofield N. Spain 28006, Madrid 77, e-mail: C/Castello Institute, March Juan CEACS, DOI onigsse.Ma rwhfrteMxc iyTesr vrgdams 60 1800. almost and averaged 1770 between Treasury decades City the Mexico of each the in ac- for percent proficient growth a at Mean and succeeded bureaucrats system. officials fiscal counting public with success Mexico administration limited central colonial with the in strengthening century, met nineteenth Spain early in the modernization until fiscal at attempts parallel though a n cnmc,Safr a col n h etro eorc,Dvlpet n the and University. Development, Democracy, Stanford in on at Program Center (CDDRL) Olin the Law M. and of John School, Rule the Law Research, Stanford Policy financial Economics, Economic the and for without Law Institute possible Stanford been the not have from on would support Center Berkeley’s work UC This the Governance. at and University, and Conflict Stanford Institutions Political at and Workshop Institutions and History Endogenous Economy and on Political Science Workshop on Social Conference the International Baiona, II in Gavin the Institutions Rabushka, at Alvin participants Pancs, the Romans and Marichal, de Wright, Carlos Gonzalez Alberto Yadira Levi, David, Jáuregui, Margaret Paul Luis Haber, Bernheim, Steve Greif, Douglas Avner Lara, B. Dobado, Baron, Rafael David Dincecco, referee, Marc anonymous Díaz-Cayeros, an to grateful am I 2 1 (O’Brien reputation bureaucracy, world-renown Brewer fiscal acquired efficient navy and and the new and created, a in- was of army was hands an collection the in fiscal put England, and In cen- centralized the roles. creasingly granted unprecedented century and eighteenth new the government in tral state fiscal-military the of emergence The Arias Marina Luz Building State Fiscal-Military of Theory Collective-Action A ..Ais( Arias L.M. century. eighteenth the of end the by percent 12 over to 1670s the in percent n oa ltsfrtecleto ftxrvneaddfne iclcpct was capacity Fiscal defense. and revenue tax of collection the for elites local and centuries. nineteenth early and eighteenth late the in lowed li (1985 Klein (1988,Table2).O’Brien ro oti rniin otmnrh eeddt ag xeto economic on extent large a to depended monarchs most transition, this to Prior 10.1007/978-3-642-35239-3_3 [email protected] B 566Ð574). , 1989 ) .Txpoed sapretg fntoa noers rm3.5 from rose income national of percentage a as proceeds Tax ). dacsi oiia Economy Political in Advances sawy oeflmtvto ntxpolicy. tax in motivation powerful pay’ a should always else is ‘somebody earmarked, not are benefits Where pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , , 2 te uoensae fol- states European Other l (1983 Alt .194) p. , 1 Even 1988 47 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 aieisiuinbidn eaiet h ne fwr htepan h aito nficlcapacity fiscal in variation the explains what wars of centralization. onset the to relative institution-building tative tal. et Weingast a expendi- and public (North like on credible representation, commitments tures making of by institutions cooperation fiscal higher political increases them of parliament, granting establishment upon The rulers revenues. constrain to fiscal need the underscore approach second Persson and Besley zd nvra,eutbe n oetal oepoutv ytm ftxto and taxation of systems productive more finance.” central- potentially modern, and more [ equitable, of families universal, imposition ized, noble the against as defences well and privileges as domains, maintained necessary. ecclesiastical corporations, was estates, and elites “provinces, the cities, fis- Europe of privileged strengthen modern fraction to early a across or least all army at Indeed, an of provide compliance lacked to the ruler revenue administration, a increase cal that unilaterally extent to the in- To ability to militarily? ruler the build-up the allow and elites centralization powerful fiscal politically crease and fiscally did Why elites. different paaue.Teeaetoapoce nteltrtr.Oeapoc ihihsthe conflicts. highlights approach military One of literature. administrative the role states’ in European approaches two of are growth There the apparatuses. for impetus major the as tablishments h rvnilEtts h sebyo h lry n a amr a a author- tax had farmers Crown. tax the and to clergy, funds the transferred of and assembly ity the Estates, provincial the tsvg ( Stasavage 3 corporations, return. and in elites privileges other these and with exemptions them loans granted and and taxes. of transfers dispatch monetary and aristocrats, negotiated collection as Rulers the for corporations—such level responsible in ecclesiasts—were national organized or Provin- actors merchants the elites. economic local at the or by institutions authorities collected monies cial fiscal the of enforcement and recipients and as Treasuries primarily monitoring taxes. functioned in collect minimally to invested capacities ruler central The fragmented. 48 8 7 6 5 4 nf ies ouain(Huntington 2001 population diverse a unify omo ovy,adhdacnrc ihte o h olcino the of collection the for them the with Mexico. in contract colonial defense in a with tax had them (sales) and provided convoys, colonies, of the in form competition foreign from guild n eluh( Hellmuth and ubro tde aeepne hsagmn n nlddqaictos e o instance for See qualifications. included and argument this expanded have studies of number A ( Weber instance for See ( Bonney also See 432). (2011, O’Brien ( Marichal ( Rosenthal (2011 O’Brien instance, for See, h sa-iiaysae hn mle osso sa n te rvlgsfor privileges other and fiscal of losses implied then, state, fiscal-military The xlntosfrficlcnrlzto mhsz h edt nraemltr es- military increase to need the emphasize centralization fiscal for Explanations ,o h eet ftxto eoemr ain octzn (Levi citizens to salient more become taxation of benefits the or ), 2011 6 2002, 2007 ). 68). 1998, 8 .Frasnhss e apel( Campbell see synthesis, a For 1999). 67)adSih( Smith and 76Ð77) , 2003 ,adCx( Cox and ), 2009 1922 7 rcptlivsos(Mann investors capital or ) ,Sopl(1979 Skocpol ), iiaycnit aiiaeficlrfr eas wars because reform fiscal facilitate conflicts Military ,adLv (1988 Levi and (1989), Brewer 421Ð423), , 2011 5 1948 .Eta (1997 Ertman ). n trs( Storrs and 1999) 1989 ). 4 ofa n Norberg and Hoffman ; h pns rw rtce h merchant the protected crown Spanish The ,Tly(1990, Tilly ), 1968 n ’re ( O’Brien and 1993) Herbst ; anan hti stetmn frepresen- of timing the is it that maintains ) 2009). 1993 2000 1988 ,Bne (1995 Bonney ), 3 o ntne nFrance in instance, For n ie n Kane and Kiser and ; .Epaain nthe in Explanations ). 2011). .]and city...] states, 1994 ). n Dincecco and ; ,adBrewer and ), 1988 ..Arias L.M. alcabala and ; Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 agie eaaeywt ahgop sa rgetto e ofe iigand riding free to led fragmentation fiscal revenues. fiscal group, lower each with grants separately similar (2008 bargained upon Summerhill conditional France. grants in towns” subsidy (1971 other make Henneman from to from towns quote led 57) jealousies (1985, “fiscal Lien and Bates rights.” among group cooperation narrow inhibit to tended privileges, distinct the own their with each groups lt oprto nsaigteitra raiaino h state. the of organization internal the shaping in cooperation elite we h ltsadterlr ntepoeso sa-iiaysaeformation. state fiscal-military of be- process than the empha- rather in chapter elites, ruler, the powerful the such, fiscally and As the elites territory. of the among the contribution tween problem of the commitment with defense a elites the the sizes for of resources cooperation and the men impeded that regimes century. eighteenth fiscal the in regions some in reforms imple- military successfully and officials fiscal Spanish mentated yet assemblies representative lacked ( America Marichal parliaments. of formation the ihmr sa n oriepwr a aeicnie oeporaeeie or elites expropriate to incentives ruler have A debts. may its problems. powers on commitment coercive renege of and because fiscal taxation cooperation more of fiscal with benefits taxation. in the of translate in benefits increases not the that may however, salient highlights, making approach by second govern- centralization The and fiscal monarchs negotiate for window to a ments provide conflicts Military state. fiscal-military tesrcueo h sebis iie ste eeit separate into were they as divided assemblies, the of structure “the ( Ertman documented. well been nisiuinlfaeokta loe ltst omtt otiuet military to contribute well. to as contributing commit were to others ensuring elites and by low-contribution allowed protection provided a that centralization framework in fiscal institutional stuck that an were to argue ruler amount I the optimal con- equilibrium. socially and the low-public-good-provision the elites on a than The ride to less protection. free led contributing to military then thereby incentives fragmentation others, had fiscal group of conflict, to elite tributions military benefits each of earmarking problem: threat action a on collective of rested face income the fiscal In ruler’s elites. the regimes, fragmented In Building State Fiscal-Military of Theory Collective-Action A 12 11 10 9 n h opineo coswt efr-nacn oprto.Se o ntne lo (1993 Olson instance, for See, cooperation. welfare-enhancing (2006 with Greif actors of compliance the ing atsa.Gef(1998 Greif Baltistan. rlzto ihn ntttoso ersnaini lc.Teeiec nDincecco in evidence The place. in representation ( of institutions no with tralization 2011 ute,i ie fwr h ue a icutteftr oeta te iies(Levi citizens other than more future the discount may ruler the war, of times in Further, 56Ð57). (1988, Levi also See ( Emerson prevent- in problems free-rider and action collective of role the emphasized have scholars Many nti hpe,Iudrcr h olcieato rbe rsn nfragmented in present problem action collective the underscore I chapter, this In oho hs prahscnrbt oorudrtnigo h rwho the of growth the of understanding our to contribute approaches these of Both htcletv cinpolm r neett sal rgetdsae has states fragmented fiscally to inherent are problems action collective That curiae 7 hw hti ayErpa ttsficlcnrlzto aebefore came centralization fiscal states European many in that shows 27) , n ri ta.( al. et Greif and ) n edtenblt,ceg,adtetwst ou ntedfneo their of defense the on focus to towns the and clergy, nobility, the lead and rvdsasmlrisgtrgrigsaefraina nerirsaein stage earlier an at formation state regarding insight similar a provides 1983) , 9 2006 oemnrh,hwvr ucee ticesn sa cen- fiscal increasing at succeeded however, monarchs, Some 12 lohglgt h motneo iiaytrasadtene for need the and threats military of importance the highlights also ) 1994 ). 0 oe bu h sae nGraythat: Germany in Estates the about notes 50) 1999, 2007 1 ihihsta ooilSpanish colonial that highlights 51) , 2Ð2)ntsta eas rulers because that notes 224Ð225) , 11 curiae félite of 1988 that ) ). 49 10 ), Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 ocue ihadsuso fteipiain n ute vne o research. for avenues further and implications section the final of The discussion history. Mexican a colonial with and concludes English from evidence with gument oepantastosbtenficlrgmsbtol h mato eie ntenme fwars of number the on seek regimes not of do impact They the only decentralized. but is fought. regimes power fiscal fiscal France, between when and transitions elite explain Britain and to in crown powerful finance between a war state for of of preferences study role comparative the ( their Rosenthal incorporate In and not Hoffman changes. does fiscal however, for blocking good analysis, in public Their elite a capacity. of salience fiscal the in emphasize increases also They institutions. market-supporting and capacity eio n h ako tnigame nbt ein,i lgigtebnfi from ruler. benefit the the and aligning colonial in elites regions, for the War both between Years’ in protection Seven armies highlights military the standing of evidence and lack England The the for and argument. War Mexico, Civil theoretical the the of importance of the implications provides Mexico the colonial for century eighteenth support and England century seventeenth in up power their government central a to taxes. surrendering levy by to gain elites local and corporate lentv ast omtt oprt o ees,teeie cuec ofiscal to acquiesce elites lack the elites from defense, the for stakes and cooperate ruler, the centralization. the to that capacity and commit extent elites to the the ways To between alternative higher. aligned when are are and centralization activity protection rents, fiscal economic military economic future cause for of to ruler prospects the likely on the more dependent more is are elites unrest the internal when or invasion external of 13 by ruler, indexed a as corporations, to referred henceforth government, central A Game The 1.1 in- government. central invasion have a elites or to local taxes unrest and levy of corporate to which power threat under their a surrender conditions to the of incentives on probability state is the fiscal-military focus The why when creases. explain likely to more framework is theoretical building a provides section This Model Formal 1 50 15 14 members. their all on binding agreements make to able are that cnmccroain rgid eg rsort,mrhnso clsat) ayshlr have scholars Many ecclesiasts). or merchants aristocrats, (e.g. guilds or corporations economic War. Years’ Seven the after Mexico colonial in reorganization military ( rvdsadtie itrclaayi ftescesu nraei sa etaiainand centralization fiscal in increase successful the of analysis historical detailed a provides 2012) itrcly h oprtoswr ersne ylclatoiis(..mjr)o ed of heads or majors) (e.g. authorities local by represented were corporations the Historically, ( Persson and Besley Arias Mexico. colonial eighteenth-century of history the on built is framework theoretical This rsn h omlagmn nSect. in argument formal the present I build military and centralization fiscal in increases the from evidence Historical which under conditions the analyze to framework game-theoretic a provide I 13 h nlsssosta nices ntepoaiiyo threat a of probability the in increase an that shows analysis The n elyadPrsn( Persson and Besley and 2009) 1997 14 n oeta ( Rosenthal and ) i = 1,...,n hs oprtosaecmoe fagents of composed are corporations These . lutaeteiprac ftedfeec in difference the of importance the illustrate 1998) Section 1. 2011 td h on eeomn ffiscal of development joint the study ) 2 R ofot h hoeia ar- theoretical the confronts neat with interacts , 15 n economic ..Arias L.M. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 rpssaplc profile policy a proposes n tleast or at accepts and corporation ruler. Each the good. by proposed public profile the policy the to pro- rejects contributions ruler and the centralization, goods, proposing private not profile If policy good. “fragmented” public a the poses of level a and goods, htterlrpoie ocorporation to provides ruler the that i good, lic corporation. olcieAto hoyo iclMltr tt Building State Fiscal-Military of Theory Collective-Action A 17 Strayer instance, For 16 systems. tax of development the in ( forms corporate of importance the stressed x τ) (x, (g yial lyda motn oei oriaigeooi ltsadriigteeiesawareness elite’s building. the fiscal-military raising negotiate and to elites need economic the coordinating about in role important specific an the played of corporations. typically or collection lending coins), the The and silver miners). rulers mint for between loans fleets (e.g. guaranteed merchants, also commodities corporations for to by (convoys taxes added defense value local regions), of provision specific between trade to nfr a payments tax increase The uniform taxes. levy to power the ruler costs the to surrender corporations the whereby fragmented under them enforce to unable is capacity. ruler fiscal the because voluntary are butions hr r w eid.I h rtpro,alpaesobserve players all period, first the In periods. two are There Timing 1.1.1 hoe hte opooea nraei sa etaiaino oke fiscal keep to or centralization fiscal in increase an fragmented. propose capacity to whether chooses i od( good lic a rcest h ue.Ec iaea otati bevdol ytepristo parties the by fis- only corporation their observed with of is contract contract part A bilateral transfer contract. Each corporations ruler. the the the contracts to private proceeds these cal of means by ruler. the and to The proceeds monies. the de- fiscal of of ruler some collection transfer the and and the is, taxes enforcement levy of the That corporations for contributions fragmented. corporations the is the on on capacity depends pends fiscal ruler initially the defense, because military corporations provide to order In ,Petih(1972 Prestwich (1971), Henneman 1970), itrcleiec uprsgvn h ue gnastigpwr oacsadpbi officials public and Monarchs power. agenda-setting ruler the giving supports evidence Historical access exclusive (e.g. monopolies royal include goods private provided publicly of Examples 1 ntescn eid fterlrpooe nices nficlcentralization fiscal in increase an proposed ruler the if period, second the In h ue straee iha nain(rurs)wt probability with unrest) (or invasion an with threatened is ruler The ne rgetdcpct,tecroain aecnrbtost h pub- the to contributions make corporations the capacity, fragmented Under h ue a rps otecroain nices nficlcentralization, fiscal in increase an corporations the to propose can ruler The h ue sal oefrebltrlcnrcswt h niiulcorporations, individual the with contracts bilateral enforce to able is ruler The ,...,g F = oterlr ne etaiain h ue ulcyanucsadenforces and announces publicly ruler the centralization, Under ruler. the to ((x G g n ) i —iiaydfnei eesr odfa h naes(rteunruly). the (or invaders the defeat to necessary defense—is )—military 1 , ≥ n ,τ ¯ f ≤ 1 ,ta euti ee ftepbi good public the of level a in result that 0, s(tity nraigin increasing (strictly) is ),...,(x n oprtosacp,terlrivssi sa-iiaystate fiscal-military a in invests ruler the accept, corporations 17 t ≥ fpooiga nraei sa etaiain h ruler the centralization, fiscal in increase an proposing If n ,τ {t,x,G rmeach from 0 n )) etepol fterlrsuiaea fest each to offers unilateral ruler’s the of profile the be } ossigo a amns etro private of vector a payments, tax of consisting {τ,x,g ,BtsadLe ( Lien and Bates ), i i . necag o payment a for exchange in g i } and pcfisteamount the specifies hc nldsavco fpayments, of vector a includes which , f g i g j ,adLv (1988 Levi and 1985), > 0for G i = = θ f(g) x h ue then ruler The . j i hs contri- These . ≥ τ where , i fgood of 0 ). ≥ θ pub- A . 0. 16 g Let 51 = i Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 aofo ahcorporation. each of payoff G ingly, v epce)pyf fec oprto we sa aaiyi rgetd is: fragmented) is capacity fiscal (when corporation each of payoff (expected) 18 period. first the in to agreed policy the implements and 52 21 20 19 le rae eedneo h ue n hrfr ihrbnfi from benefit higher a therefore and ruler the on dependence greater plies h ulcgood. public the ( against defense own their have may they or say, territory, colonial uprisings. Spanish internal of takeover British a h lt a o eceil,hwvr fe h ue a nrae sa etaiain discuss I centralization. conclusion. fiscal the increased in from corporations has fragmentation and ruler ruler to the between reversion after problem however, of commitment credible, threat the be A not contracts. may exchanges private because elite is on ruler the and ruler and rely corporations 1 the capacity Weingast both and period fragmented from and corporations commitment under in North the ensures agreement between e.g. Reputation problem issue. (see the an commitment to on not this agreed renege fragmented, those is could than capacity ruler higher fiscal the payments tax centralization, collect in forcibly invests ruler a Once eret hc corporation which to degree cet sa aaiyrmisfamne,terlrpooe famne”policy “fragmented” a proposes ruler profile the fragmented, remains capacity fiscal accept, oprto’ aofdpnso h mut ftepiaegood, private the of amounts the on depends payoff corporation’s A Payoffs 1.1.2 the implements ruler period. first the the centralization, in fiscal to in agreed profile increase policy an “fragmented” propose not did ruler h het(fa naino net oteetn httecroaindpnson depends Let corporation protection. the and that rents extent future to the for vulnerable ruler to the is unrest) of corporation or survival invasion A the an others. (of than threat protection the military from more benefit good, lic where ett h ue,and ruler, the to ment oaindpnso h ue o cnmcrns( rents economic for ruler the on depends poration if ruler the with exchange individual θ> its of function a is payoff ee feooi activity, economic of level θ h ue n h oprtoscudas ifri hi ecpino h rbblt fathreat a of probability the of perception their in differ also locally. could taxes corporations dispatch the and and enforce, ruler collect, The assess, of to effort case the the captures in cost even This rents economic their keep to able be may corporations some instance, For regarding problem commitment ruler’s the incorporate explicitly not does framework This .Ti a eicroae nteparameter the in incorporated be can This ). sicesn n ocv in concave and increasing is = ,tecroainspyf sas ucino h ulcgo.Temr cor- a more The good. public the of function a also is payoff corporation’s the 0, f(g y v {τ,x,g sicesn n ocv in concave and increasing is and i G ,g n n(xgnu)oealeooi activity, economic overall (exogenous) on and , u −i y i F r h auso h rvt n ulcgos epciey tagiven a at respectively, goods, public and private the of values the are ) } (x where , n h oprtosacp rrjc h ue’ rpsl fthe If proposal. ruler’s the reject or accept corporations the and , 21 i ,G) ial,frany for Finally, e = i g > v(x i scroainiscnrbto to contribution i’s corporation is θ scorporation is 0 i stepoaiiyo threat, a of probability the is i , y) ¯ x eet rmtepbi od( good public the from benefits and + θ θα nices neooi ciiyicessthe increases activity economic in increase an , G y ¯ i , y and α v i x . f(g i y ¯ y i ¯ > scs oclettaxes. collect to cost ’s , i y ,g o all for 0 G y −i ¯ > ), α i ,and 0, y ,tehge h eetfrom benefit the higher the ), ¯ 18 i − and , fls than less If τ g i y( i α G scorporation is y − ¯ i θ 0, v( G oecorporations Some . ahcorporation’s Each . ≥ = τ 0, y) .Ahigher A ). i ¯ aaerz the parametrize 0 − .I,b contrast, by If, 0. y) ¯ = x 20 e n i ¯ n h pub- the and , i = .Rcl that Recall 0. h function The . corporations .Accord- 0. ..Arias L.M. G i 1989 . spay- ’s 19 α i The im- .If ). t . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 ainis: ration Assume c(x,G) r,respectively: are, all n nyif: only and ration eetadcno eecue rmtemltr protection. military the from excluded be cannot and benefit otiuin falohrcroain ie htcorporation that given corporations other all of contributions ov o h uesrtg ugm-efc aheulbi SN)o h game the of (SPNE) equilibria ruler. Nash the subgame-perfect by preferred pure-strategy the for solve I Equilibrium 1.2 Building State Fiscal-Military of Theory Collective-Action A 23 22 n rvt n ulcgospooe yterlrudrficlfragmentation. fiscal under ruler the by proposed goods public and private and etaiigtxcleto. eas h ue olcsadefre ae ne fis- under taxes enforces and collects centralization, ruler cal the Because collection.) tax centralizing etaiainmr ieya oga h oprtosrciigalresaeo h upu benefit surplus the of share large a (high to receiving transition good corporations a surplus public makes the the the corporations as from the of long to all as surplus extracts likely the so of more share and centralization higher offer a take-it-or-leave-it Giving a corporations. the makes from ruler the Also, equilibria. trivial promised and be defensive can between war distinction a the of on (1983 spoils more Emerson The For see excluded. good. warfare private are predatory a others is while groups, latter some the only whereas to good, public pure a is mer where hr r oiieetraiiso hs htd o otiuet uliga army an building to contribute not do that those because on externalities positive are There suetecroain cetterlrspooa hnidfeet hsalw et ueout rule to me allows This indifferent. when proposal ruler’s the accept corporations the assume I for- The warfare. predatory from warfare defensive distinguishes assumption non-exclusion This hnficlcnrlzto sipeetd h epce)pyf fec corpo- each of payoff (expected) the implemented, is centralization fiscal When corporation If h ue’ aofi h eeu bandfo oprtos amns Let payments. corporations’ from obtained revenue the is payoff ruler’s The fficlcpct sfamne,corporation fragmented, is capacity fiscal If i hn h ue’ aof ne rgetdadcnrlzdficlcapacities fiscal centralized and fragmented under payoffs ruler’s the Then, . i t v(x eevspayoff receives etecs fpoiigpiaeadpbi od nbt sa regimes. fiscal both in goods public and private providing of cost the be scorporation is f c i sicesn n ovxin convex and increasing is sicesn in increasing is , y) ¯ + i θα e eet h ue’ rgetdplc proposal, policy fragmented ruler’s the rejects i = i 23 u y i C u u o all for 0 i α f(g ybcwr nuto,Ifis td h hieo payments of choice the study first I induction, backward By u stxpyet Rcl h ue sets ruler the (Recall payment. tax ’s (x R F i R C ). 0 x G) (x, G) (x, i = ,G) g i ,g −i v( −i ). i 0, = o all for . = = ), y) v(x ¯ y nt i ¯ =1 n + i − − , (τ i θα y) x .Thatis,if ¯ c(x,G) g i i + and i + y(f( − g θα τ i i ) i G i − 0,g − cet oiyprofile policy accepts y(G, − and , e c(x,G), F. i −i θ> ≥ y) ), ¯ c θα x y) ¯ − ,algop with groups all 0, i x i where , j y t, 22 = f( i t i and 0 sntcontributing. not is 0,g = x i t g = −i −i o all for c n corpo- and 0 ), {x x eoe the denotes i y i G ¯ ,τ . = i i α ,G when i 0for > } (1) 53 if 0 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 stesm o l oprtosbcuetecoc of choice the because corporations all goods private for of amount same equilibrium the the that is Notice capacity. fiscal fragmented der aaiyi oe hntescal optimal socially the than lower is capacity 1 Lemma 1,...,n nrae hntegop neatol ihterlradaeual oslethe solve to unable are and ruler the themselves. among with problem only action interact collective groups the when increases Proof betv ucin h ue’ e fpotmxmzn oiisis: policies profit-maximizing of set ruler’s the function, objective ore oterlrol oteetn hti receives it that extent the to only ruler the to sources x h grgt au rmpbi odpoiini h oprtoswr bet com- to able were pay: corporations the to if mit provision good public from value aggregate the en- of means of corporation, no choice each has ruler’s of ruler the contribution the lower and the good, Internalizing contributions. public these the forcing to contributions their in others on and ( in constraint participation the SPNE, the In 54 h qiiru payment equilibrium the ol en otfo h rerdn rbe.Tedifference The there pay, problem. to riding commit free to the themselves from police cost and no coordinate be to would able were groups the If f/∂g ∂f h rtodrconditions: order first The terize eas fteasmtoson assumptions the of because ol eal oices i aofb nraigtepyetfrsm corpora- some for payment the increasing for by solving payoff By his tions. increase to able be would ∗ ( x, ne rgetdficlcpct,ec oprto a netvst rnfrre- transfer to incentives has corporation each capacity, fiscal fragmented Under Lemma ovn eoti h unique the obtain we Solving ˆ ∈ g r max arg G) ˆ h proposal The . i G h oilyotmllvlo ulcgo rvso solves: provision good public of level optimal socially The ˆ − characterize , h eutflosbecause follows result The . ∈ θ r max arg 1 h qiiru ee fpbi odpoiinudrfamne fiscal fragmented under provision good public of level equilibrium The Y(G x losu odfietesca otdet rerdn steices in increase the as riding free to due cost social the define to us allows − ∈ x,g j R = G n ∗ i ∈ i =1 n , ∗ R α y) ¯ n j ∈ e G i n ∂y/∂G =1 i − x r max arg ˆ τ i − G =1 slwrta h oilyoptimal. socially the than lower is i n sas qa otescal pia amount optimal socially the to equal also is v(x Y( ∗ rmeach from c(x,G). τ g ˆ .From( v(x i ∈ G, ˆ i R o ahcroain usiuigeach Substituting corporation. each for , · n θ y) i ¯ )> y) f/∂g ∂f ¯ , y y) i ¯ − =1 n and x ˆ i + α c(x,G) ,tefis re conditions: order first the 2), and θα ,where 0, i i i ∂y/∂G i sbnigpriiaincntan,w obtain we constraint, participation binding ’s c f =1 = n i .) y g ˆ sicesn in increasing is ∂c/∂G θα f(g h ue rpsst h oprtosun- corporations the to proposes ruler the : . i G

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 rps centralization. propose rgetto.I h ue’ aofgvntetxpyetncsayt bancom- obtain to necessary payment tax the from given pliance frag- under payoff fiscal than ruler’s centralization preserve the under to If higher but fragmentation. be threat must payoff credible ruler’s no the Second, has mentation. ruler the and centralization refuse epficlcpct rgetd umn over Summing fragmented. capacity fiscal keep uthl o tleast at for hold must (4) in constraint tion tain θα amnsadpoie h oilyotmlaon fpiaegoods private of amount optimal socially the provides and payments oain cetadterlricessficlcnrlzto ftepoaiiyo a of probability the if that: centralization such fiscal is increases threat ruler the and accept porations estemxmmuprbudo e sa rnfr ychoosing by transfers fiscal net on bound upper maximum the sets oigpriiaincntan hnterlrpooe centralization: proposes ruler the when constraint participation lowing n rpsto ie h odto ne hc oiyprofile policy which under condition the gives proposition ing n( in amn htec oprto swligt a necag o h pia ee of level optimal the for exchange in pay good: to public willing the is corporation each that payment h ih adsd f( of side file hand right The esb hoigtescal pia amount optimal trans- fiscal socially maximizes the ruler choosing the centralized by is fers capacity fiscal when that notice First, centralization fiscal in Investment 1.3 Building State Fiscal-Military of Theory Collective-Action A rpsto 1 Proposition ycmestn ihpiaegoods private with compensating by (5) in maximum higher the (a than higher payment tax a tansi ( in straints ban ihrmxmmpyet rmtoecroain h eetmr from more benefit ( who good corporations public those the from payments maximum higher obtains u qiiru for equilibrium aofb nraigtetxpyetfrsm oprtos Let corporations. some for payment tax the increasing by payoff R C i x G) (x, usiuigi ( in Substituting w odtosms odfrficlcnrlzto oocr is,teparticipa- the First, occur. to centralization fiscal for hold must conditions Two rtoti h PEasmn h ue a olc oprto-pcfi tax corporation-specific collect can ruler the assuming SPNE the obtain first I opermrsaotti aiu a amn r nodr is,teruler the First, order. in are payment tax maximum this about remarks couple A 3). [ {ˆ y(G nt τ, t v(x i ≤ M x, ˆ ∗ x , ≤ ≤ i g ˆ i y) , ¯ ri h rset feooi ciiyincrease. activity economic of prospects the if or ) } idfrall for bind 5) y) v(x ¯ h oprtoscnrfs etaiainadfreterlrto ruler the force and centralization refuse can corporations The . i n [ ¯ − v(x + i oprtosi esta h aoffrom payoff the than less is corporations , SPNE the At n i [ ¯ y(f( , v(x α θα = i y) ¯ i , .Scn,terlrcnoti opinefo corporation from compliance obtain can ruler the Second, ). 4)for y) n i ¯ i θ + y(G, , 0,g where , ≥ y) ¯ + θα F −i + scntn n ie yteeulbimplc pro- policy equilibrium the by given and constant is 4) θα ( i y) ¯ τ ˆ ), + [ y θα i i i Y(G , y(G, − tews h ue ol eal oices his increase to able be would ruler the otherwise , h ue rpssplc profile policy proposes ruler the y) t c(x ¯ x G ˆ ∗ i i y(G, t ] = , ∗ ≥ ∗ g) ∗ , etebnigcntan n( in constraint binding the be ˆ , ,G y) (t y ¯ ¯ y) v( ¯ 1 ∗ n ovn for solving and ]− y) ¯ − ,...,t ∗ x ˆ − ) i ]− y , −[ Y( C y) ¯ f( where , C G, n ∗ + ˆ 0,g ) n ¯ . − i θα y) G ¯ e oprtos h oprtoscan corporations The corporations. i c(x,G) −i ∗ ] i + y ahcroainfcstefol- the faces corporation Each . ), C c(x n f( t i y ¯ α sacntn.I olw that follows It constant. a is eoti h aiu tax maximum the obtain we n ovn for solving and g), i ˆ ∗ {ˆ /n , , − τ, G) y ˆ ¯ F x, ˆ for ] hrfr,teruler the Therefore, . −ˆ . g {t ˆ 5)for } i g ∗ h ue osnot does ruler the , = i {t ,x −ˆ ∗ t 1,...,n. ∗ i ,x ∗ G τ i ,G h follow- The . i ≡ x ∗ ∗ − ∗ ,G ∗ h con- The . t v(x sdefined as } eob- we , e , i ∗ l cor- all . i } , san is y) ¯ i (4) (6) (5) 55 + at Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 ento 1 Definition ihrta hi aiu a ae eieblwteSN hnterlrst a sets ruler and the centralization when under SPNE corporations the all below for derive payment I tax rate. uniform tax maximum their remaining than the higher oblige to capacity enforcing and exactly fatra.Ntc httelwrteaeaecroainsdpnec nteruler the on dependence corporation’s average probability the the lower the the to on ( that relative and Notice good riding, threat. public free a of overcoming of amount from optimal gain the overall providing corporations’ from ruler the to increase o given a for osnthl,terlrpooe rgetdpolicy fragmented a proposes fragmented. ruler remains the capacity hold, fiscal not and does profile (6) of condition values If smaller increase. ties for ruler’s occurs the and centralization corporations’ Also, θ the protection. between divergence military the from smaller benefit the probabil- the threat, of a values of smaller for ity occurs centralization fiscal is, That centralization. nt 25 y) ¯ Y( ∗ } h olwn eutgvstecondition the gives result following The i ∗ ∗ nrae eas h tkso l par- all of stakes the because increase, G, } ilscnrlzto neulbim I equilibrium. in centralization yields , ˆ i y) ¯ stenme fcroain whose corporations of number the is e y) n ¯ i − −¯ + [ Y( θ n c(x i i ie odto ( condition gives oprtost a a rate tax a pay to corporations G, α ˆ t i ∗ ∗ i /n , )> y) − ¯ G) ˆ t ] p c(x ∗ ] etepvtlcorpo- pivotal the be , . i α 0. ∗ i is ,G = n

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 nfor in hnteaeae h rniint etaiainocr o oe rbblt fthe of probability higher lower Proposition a vulnerability the for in of occurs of than centralization degree threat that to a to transition has the relative average, corporation corporation the pivotal each than the of threat If a corporation. to pivotal vulnerability the by weighted ohPropositions Both Implications 1.4 Proof Building State Fiscal-Military of Theory Collective-Action A 26 or high. small, too is is protection centralization military in from investing between elites of alignment corporate cost the ruler’s the and the because ruler fragmented the to remain riding. benefits free therefore the pay overcoming may from fa- to regime gain willing activity social fiscal is economic the A corporation increasing of by a level and maximum centralization higher the under op- a transfer increasing the to equal, by of willing centralization else provision are cilitates the all corporations from the Also, benefit that centralization. the more under is the higher and the protection, military rents, fiscal dependence timal future in corporations’ for the increase sur- higher ruler The an the the future. on cause on economic stakes their to for corporations’ likely ruler the the more higher of vival is the build-up, threat military a and of centralization probability the in increase vntog ti oilyotml hsocr eas h ue nue osin good. public loss across the a discriminating from endures than benefit rather ruler their transfer to the according uniform because corporations a occurs requesting This condi- from optimal. revenue the socially fiscal corporation, is average it the though than even good public Proposition the in from tion benefit lower a has twt h eoiao ncniin(6 condition in denominator the with it h ulcgo) nices nficlcnrlzto eed ntesz fte‘ac- the of size the on from depends benefits centralization ( corporations’ fiscal coalition the in cepting’ all increase extract an to good), unable public the thus is (and payments specific nlssepaie htuls h ue a h upr fsm ftecorpora- The feasible. the not of are some up build of military support and centralization the fiscal has in ruler increase the the tions, unless that emphasizes analysis good public the from benefit lower corporation. a average has the corporation than pivotal the if optimal, socially hscnb enceryb ertn h eoiao ncniin( condition in denominator the rewriting by clearly seen be can This h anfo rvdn h pia muto iiaypoeto snow is protection military of amount optimal the providing from gain The Proposition ial,acul eak bu h hoeia rmwr r nodr is,the First, order. in are framework theoretical the about remarks couple a Finally, α p Y t p h ue rpssatasto fadol if only and if transition a proposes ruler The ∗ G and ∗ , y ¯ τ ˆ i − n ovn for solving and , 2 Y( hw nadto hti h ue sual ocletcorporation- collect to unable is ruler the if that addition in shows G, 1 ˆ n ¯ 2 .I atclr emyntosrecnrlzto hni is it when centralization observe not may we particular, In ). and y) ¯ osnthl n h ue osntpooecentralization propose not does ruler the and hold not does i 2 l leconstant. else all 1, α n ihih h anipiainfo h nlss An analysis. the from implication main the highlight i = θ ): Y ie odto ( condition gives G ∗ , y ¯ − Y( 26 G, ˆ oee,i h ioa corporation pivotal the if However, y) ¯ 7). nt i p ∗ α ≥ n i u − sflosadcomparing and follows as 7) R F Y ( τ, ˆ G ∗ x, ˆ , y ¯ G) ˆ i Substituting . α i − n α p . 57 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 esi uligaficlmltr tt.I ooilMxc,tewr findependence of wars the Mexico, colonial suc- In their state. for fiscal-military justice a building of in administration cess the in infrastructure the and implement authority to recognized ability (1989 and Brewer authority reforms. credible the fiscal-military with government no central threat. or of a ruler periods of long probability the relatively in enjoyed increase had the and to prior armies involvement lacked military below Eng- both evidence seventeenth-century (at they historical in objectives The that state defensive. shows The fiscal-military were nature. a Mexico military eighteenth-century of good groups, and up public land build specific a the to of and assigned initially) private exclusively least a preda- war with both of created confounds is spoils army protection the an and When protection. goals military tory of nature good public the negotiate Section to compliance. rulers their for obtain need to the elites (2) corporate and and the protection, elites’ with military re- corporate from the fiscal-military between benefit a ruler’s alignment to the the and leading vulnerability analysis military theoretical (1) the form: by highlighted factors main o nln hl Sect. while England for dnefo h rniint sa-iiaysaei eetet-etr England seventeenth-century in Mexico. state colonial fiscal-military eighteenth-century a ev- and to historical transition discusses the section this from argument, idence theoretical the for support provide To Evidence Historical 2 centralization. external fiscal or in internal increase credible of an the probability to the or lead in not authority increase will legal an threat with sanction, and actor monitor and central to monitor have a ability credibly must Lacking can ruler ruler collection. the the tax believe hold, enforce must to corporations the results and the authority For legal centralization. implement and propose role. the important external with an or negotiation such centralization, play alternative impose not except has or may army threat ruler corporations an credible finance the to however, no revenue If, of has sources fragmentation. ruler fiscal the preserve and to centralization refuse can corporations 58 27 in e clse ( McAlister see (1791 tion, ( Urrutia Stein and and Fonseca instance Stein fis- Mexico’s for colonial see of half For reforms, ( respectively. second financial century, the and in mid-seventeenth administrative, underwent the England cal, and and century Mexico colonial eighteenth that the capacity, military in and ment, ’re (1988 O’Brien ( 1973, ,Keh ( Kuethe 1994), ayshlr aedcmne h motn hne nficlamnsrto n enforce- and administration fiscal in changes important the documented have scholars Many lo h ae fEgadadclna eiocrooaeteiprac fa of importance the corroborate Mexico colonial and England of cases the Also, isolate to us allow they because pertinent are Mexico and England of cases The eod h eu mlctyasmsterlrhsteatoiyadteaiiyto ability the and authority the has ruler the assumes implicitly setup the Second, 1987 ,Ken(1998 Klein (1987), Elliott ), , 2011 1986 2003 1953 ,Bee (1989 Brewer ), ,Ace (1981 Archer ), .Rgrigclna eiosmltr reorganiza- military Mexico’s colonial Regarding (2007). Marichal and ), ,GtérzSno ( Gutiérrez-Santos ), 2.2 icse h vdnefrclna Mexico. colonial for evidence the discusses n rwradHlmt ( Hellmuth and Brewer and ) , 1978 ,Jueu ( Jáuregui ), n lit ( Elliott and ) oe h motneo h rts crown’s British the of importance the notes ) ,Mrca n ot Mantecón Souto and Marichal (1982), Fisher 1961), 27 raieteeiec rudtetwo the around evidence the organize I 1999 2006 ,Casot (1990 Coatsworth ), .Temi ore o nln are England for sources main The ). 1999 ). 2.1 rsnsevidence presents ,Kih (2002 Knight ), ..Arias L.M. ,Brading ), ), Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 ogopwsspro oterest. the to superior was group no n olcin n a oto vrtesaedprmnsi hreo implement- of charge in assessment departments their state for rules. rules the de- those the over ing taxes, Parliament control of had elites. types and and wealthy collection, levels the the and of from selection cooperation the on obtain cided and negotiate to need from navy. benefit the elite’s strengthening the and and executive’s army the standing pro- of protection a alignment military creating an of for lack conditions the the with together vided for stability protection political (O’Brien greater internal overseas” afford of interests to threat also commercial but growing stability, economy’s political the and above order state, centralized maintain among more military to and consensus all and stronger political altogether fiscal an a for England’s “forged elites events wealthy in England’s of watershed to succession a challenges the for Importantly, and history. warfare conditions civil the of created interregnum hierarchy An rivalries. domestic against state viigdrc ae nln.I a o ni 79ta itmngdt introduce to managed Pitt that 1799 until not was It land. on taxes direct avoiding novmn ihmjritrainlcnit uigta iepro.According time-period. that military ( during Brewer no to conflicts had international aristocrats major ef- five with also involvement of (Brewer were out the all” aristocrats “four from at landed 1640s army experience Its the standing centuries. by a seventeenth demilitarized; lacked late fectively England the to nations, fifteenth European late other to contrast In England 2.1 Centeno state. minted newly its stabilize to centralization. region to the different lead for in ( not years did elites fifty Spain the almost from by took independence It faced of threats aftermath external the and in regions internal The importance authority. the of legal example of an provide instability political resulting and (1810Ð1821) Building State Fiscal-Military of Theory Collective-Action A 29 28 armed and support private on heavily prior ships. Further, depended merchantmen war. navy Years managed the Hundred fleets century, the seventeenth French during the ports and to English Castilian various century. sack seventeenth and the seize established English to be of an to half began and second only complicated, the power naval England in English of difficult. num- invasion continent and the an sizes of army made invasion in deployed increase large troops The of Europe.” modern ber early in war of scale the 2002 n ’re ( O’Brien and (1977) Horowitz (1989 Brewer summarizes paragraph This h motn oepae yPrimn nficlmtesgvseiec fthe of evidence gives matters fiscal in Parliament by played role important The the secure to need the for point turning a marked (1642Ð1651) War Civil The rusta hswsarsl fteatoiyvi etb h pns crown: Spanish the by left void authority the of result a was this that argues ) 1989 29 2,“nln a hlee o utb e nua oiinbtby but position insular her by just not sheltered was “England 12), , nfc,telne ltssttetrsfrcoeainb initially by cooperation for terms the set elites landed the fact, In 28 1989 2011). 2.Ti a atyarsl fEgadsnon- England’s of result a partly was This 12). , 8Ð13). , 2011 2) The 426). , 59 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 73 dniyawtrhdi ooilMxcnhsoy h ee er’War Years’ Seven The history. Mexican colonial in watershed a identify 1763) 32 31 30 port the 1760, men. In 1,000 Mexico. Mexico. around colonial of by and in Spain, guarded Gulf New locations were the with guarded fortress trade of most European and coast all the for the of exit on and one Veracruz, a entry therefore garner of of point to port the efforts the was the port in This by 1762 evidenced in is army Mexico field colonial of position vulnerable The rule. colonial of Ameri- years its 200 in first challenges its external during or territories internal can no practically faced crown Spanish The Mexico Colonial 2.2 tax. income first Britain’s 60 36 35 34 33 uhlre oc a eesr oesr h euiyo h kingdom. a the Viceroy, of the security the to ensure according to Notwithstanding, necessary force. was force previous larger the much of size the times n oiieafreo ,0 e nadaon Veracruz. around and fortifications in the men ready 8,500 to of funds force Marqués Viceroy a extraordinary The mobilize of War. Years’ and pesos Seven 3,398,471 the spent of Cruillas end the de before just attack to nerable ocsa h ot n odr o h rtcino h pns mrcnempire. American Spanish minimal the on of only protection relied the for and borders territory and colonial ports its the at defend forces to expenses major in incur teswr oprtn n otiuigwt h ogn faficlmltr state. fiscal-military a that of group forging elite the every with contributing ensure the and to to cooperating policies and were elites fiscal others the public with and negotiate to centralized need of the importance to Also, tran- attests possible. parliament a was by monarch, played low-public-good-provision and role and elites the low-contribution wealthy a the of of out majority bene- a sition the for aligning protection By military centralization. fiscal from increasing fit in war) Civil the after taxa- (evident [direct accept to classes political the prompted (O’Brien credit tion]” public of collapse able etr,tomr ieoate eecetd e rnd n í el lt.Teviceregal Madrid. so, in The Even government America. the Plata. in with subjects la directly its 18th negotiated de and individuals Spain the Río powerful in and and In crown corporations the respectively. Granada colonial between New link New Peru, a viceroyalties, created: and as two functioned were Mexico governments in viceroyalties contemporary territory more roughly the two divided comprising century, crown Peru, Spanish and the Spain conquest, the After America. here. presented argument cal clse (1953 McAlister 315). (1981, Archer (1953 McAlister Spanish colonial in viceroyalties Spanish the of one Spain, New of part was Mexico Colonial theoreti- (1953 the McAlister supporting Mexico colonial of analysis historical thorough a provides (2012) Arias (1988 O’Brien h ako iiaypoeto ncnucinwt h ee er’Wr(1756Ð War Years’ Seven the with conjunction in protection military of lack The nsm nls sa itr hw h motneo heto nenlunrest internal of threat a of importance the shows history fiscal English sum, In ). ,7). 2Ð3). , ,2). 1988 22). , 30 Ol h riso eouinr rneadteprob- the and France Revolutionary of armies the “Only 34 h oto earzwsvul- was Veracruz of port The 31 h rw i o edto need not did crown The 35 hswsafreeight force a was This 36 ..Arias L.M. 32 33 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

690 689 688 687 686 685 684 683 682 681 680 679 678 677 676 675 674 673 672 671 670 669 668 667 666 665 664 663 662 661 EDITOR’S660 659 658 657 656 655 654 653 652 651 PROOF650 649 648 647 646 645 ae tepe sa eomtruhteceto fteUino rswt the with Arms of Oli- Union of the Count-Duke of the creation 1626 the through In War. reform Years’ fiscal Seven a the attempted of vares importance the to regions. attest other bureaucracy, in fiscal than efficient extent larger more in- a and apparatus, to larger establishments structured officials a military crown’s of highly renovated and means the more by Colombia a revenues Ve- and into fiscal in creased Panama administration and state of Peru, and the regions more Mexico transformed coastal were capitals and and imperial monarch the Cuba Spanish In racruz, the threat. of British existence the the by affected on relied patronage and lege l o rmc ewe rti n rne o h rttm nErpa history, European in time first territory. the colonial For strug- in a France. occurred was and war battles Britain The between colonists. primacy the to for known gle was conflict the as War, French-Indian n h aneie n oa authorities. local and elites officials main crown the between and contestation political and compromise, bargaining, through century. eighteenth the of of- half royal second by undertaken the reorganization in military ficials and centralization fiscal the marked hne h aac fpwrwti ooiltrioyaogteErpa pow- European the among territory colonial within power ers. of balance the changed Building State Fiscal-Military of Theory Collective-Action A 42 41 40 39 38 37 oiiefo h mrcst pi,teficlmltr rniincudnthave not could transition fiscal-military monies. the continental always Spain, with were financed to transfers been net Americas because the at Furthermore, imposition, reform. from top-down of a positive stages for upon initial rely the could in crown least the that colonies Spanish the in omlotra fwri Europe. in war of outbreak formal as ftedmgahcrcvr fteIda ouaini h rthl fthe of half increase unrest first internal levels. the be- saw new in Mexico Also, to population colonial attack. in Indian provinces British the many of against century, recovery eighteenth possessions demographic colonial the Spanish of cause secure to who need Britain, the Great for victory a power. with European 1763, dominant 10, the February as follow- on emerged ended Royal Paris war the British of The the Treaty Philippines. France, the 1762, the and in of ing Manila Spain August and of In Cuba, war. Havana, kings agree- captured the Navy Bourbon An into the hands. Spain brought British between Compact, in 1761 Family was August America in mainland autumn made the in By ment territory Europe. French and all Africa, 1760, coastal and of Philippines, the isles, Caribbean the aihl( Marichal ( Grafe and Irigoin See ( Inglis and Kuethe ( Fisher (2006 Elliott ( Bonney instance, for See, aldatmt oipeetficlrfr ale nteclna eidalso period colonial the in earlier reform fiscal implement to attempts Failed privi- of network elites’ the where regions the in successful more was Reform rwn oyo coasi eosrtsta etaiainwspursued was centralization that demonstrates scholarship of body growing A h iiaydfassfee ySandrn h ee er’Wrhighlighted War Years’ Seven the during Spain by suffered defeats military The 37 n15,Fac elrdwro ra rti.Ti a h einn fthe of beginning the was This Britain. Great on war declared France 1756, In 219). 1982, 2007 9) h oflc nNrhAeia olbgni 74 w er eoethe before years two 1754, in began soil American North in conflict The 292). , 39 88) uteadIgi ( Inglis and Kuethe 48Ð80), , h ee er’Wr oehrwt h nrae ninunrest, Indian increased the with together War, Years’ Seven The 1985 2008 2Ð2) e loPqet ( Paquette also See 122Ð123). , n lit ( Elliott and 2004) n aihladSuoMneó ( Mantecón Souto and Marichal and ) 38 41 hr eebtlsi ni,NrhAmerica, North India, in battles were There 1985 40 2006). hr eefwmltr rplc forces police or military few were There ,Baig(1971 Brading ), 2007 ). 1994). , 1987 ). 42 61 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

736 735 734 733 732 731 730 729 728 727 726 725 724 723 722 721 720 719 718 717 716 715 714 713 712 711 710 709 708 707 EDITOR’S706 705 704 703 702 701 700 699 698 697 PROOF696 695 694 693 692 691 ilMxc akn nainetbtenteft fteconadta ftelocal protection. the military of of that provision and crown the the regarding of elites fate corporate colo- the and in between alignment reform an implement lacking to Mexico inability nial the for evidence provide centralization, fiscal hszn nte omtetpolmta ewe h ltsrsligfo free from resulting elites the between problem—that commitment another phasizing sa etaiain h ue ol eeeo h oiyareetadunilater- debts. and its in agreement invests on policy default ruler the or a taxes on Once future renege centralization. increase could in ally ruler increase pol- the the the centralization, implement before fiscal to to commit agreed can ruler profile the icy assumes argument theoretical the First, capac- fiscal of centralization the protection to military agree from can they benefit survival that their ity. the ruler) so the on (and of stakes future that have with powerful economic pro- aligns fiscally their military the for provide of ruler to majority the need a of the until on not The agreement is contribute. is it also there tection, if others a even ensuring of that, by shows face protection analysis to military the elites to fis- in allows contribute in that riding to framework increases free institutional commit to an to provides lead leads centralization can Fiscal capacity unrest threat. fiscal internal Fragmented or invasion centralization. external cal mechanism of complementary a threats providing why by for literature the to contributes chapter This 3Conclusion tempo- Only effect. reforms. took fiscal rates the tax resisted trade subject elites to not colonial increases was rary the however, territory, and Spanish Colonial threats de- the empire. territorial was the between to contribution of annual Europe part fixed every mainland A from in colonies. manded American defense Spanish of the burdens and kingdoms the sharing of goal 62 45 44 43 oecssrptto a o eeog olmtrlr,adta oeelaborate more that and representation). rulers, of institutions limit in (e.g. to that required however, enough be argued, may impor- be have arrangements sufficiently not Scholars institutional is may maintain. reputation future to cases reputation the a some If has agreements. ruler his the on tant, renege to incentives ruler’s a naon oteConi xhnefrtergtt aeueo h eeusrsligfo a from resulting revenues the of to use promised make ( to latter Ortiz ( right the Domínguez the which See for tax. exchange through royal in individual Crown specific or the to corporation the amount private legal were an major a colonies pay to The its and able repayment. and Crown were and Spain the rulers in interest between transaction of taxes, this service certain for the collect used guarantee directly form and to advance in right payments the obtain corporation/debtor the granting By .Frtecs fEgad e rwr( Brewer see England, of case the For 2011). e o xml ot n enat( Weingast and North example for See debts. its pay to ruler the for devise commitment a as serves fragmentation fiscal that Notice (1986 Elliott h vdneo eoito ihteeie,aderirfie tepst increase to attempts failed earlier and elites, the with negotiation of evidence The yaayi opeet hsltrtr ntene ocntanterlrb em- by ruler the constrain to need the on literature this complements analysis My h nlsshr ihihstoise htsgetaeusfrftr research. future for avenues suggest that issues two highlights here analysis The 246Ð274). , ,Gefe l ( al. et Greif 1989), 1960 93). 1989, ,adAvrzNgladChamley and Alvarez-Nogal and (1998), Conklin ), 43 44 ,adBlo n oof(1989 Rogoff and Bullow and 1994), euain oee,lmt the limits however, Reputation, asientos. Asientos eecontracts were ..Arias L.M. 45 ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

782 781 780 779 778 777 776 775 774 773 772 771 770 769 768 767 766 765 764 763 762 761 760 759 758 757 756 755 754 753 EDITOR’S752 751 750 749 748 747 746 745 744 743 PROOF742 741 740 739 738 737 elyT eso 21)Plaso rseiy h oiia cnmc fdvlpetclusters. development of economics political politics. the and prosperity: of taxation, Pillars rights, (2011) property T capacity: Persson T, state Besley of origins The government. (2009) representative T and development, Persson T, taxation, Besley on note A (1985) D-HD Lien Americas RH, centralization. Bates to 1759Ð1812. fragmentation from Spain, Mexico: colonial New in capacity in fiscal Building policy (2012) LM military Arias Press, and Mexico finances New Bourbon of Univeristy (1981) 1760Ð1810. CI Mexico, Archer bourbon in army Cortes The the (1978) II, Phillip CI between Archer constraints under policy Debt 41:181Ð222 (2011) Choice C Public structures. Chamley tax C, of Alvarez-Nogal evolution The (1983) JE Alt References contributions. of allows enforcement the regime the believe fiscal and new policies a actors The of those creates collection. coordination tax that shock the enforce for and the and actors, monitor that credibly key key: can the ruler are among rel- conditions problem the Two and action necessary. state government collective is central a the actors For both fiscal of sufficiently places. income a evant and the centralization, times affecting fiscal jointly increase other shock to some- at large capacity us capacity fiscal tell fragmented fiscal may on here of relying analysis evolution the the doing, so about to In thing leading centralization. in fiscal protection in military increase of an aspect public-good defensive, the emphasize ico in- disappears, the threat the of constant. after irreversibility nature else to all sunk-cost lead the can in centralization addition, increase fiscal admin- In in an fiscal vestment reverse. make a to could difficult and capabilities, centralization army enforcement fiscal standing and a Incorporating monitoring of disappears? with creation threat istration ob- the the not once between we centralization complementarity should fiscal the Why in irreversible. reduction are a capacity serve fiscal in investments the whether mentioned. problems the commitment and two representation and the centralization between needs between research links timing Future the centralization. carefully fiscal representation to more of agree study to a to institutions elites to that the corporate response for corroborates the necessary in Mexico not that aligned are colonial extent are of the protection case military to The from centralization, threat. benefits fiscal ruler’s of the stages and ca- initial necessary,elites’ fragmented not the prefer are in representation corporations least of the institutions at that and riding suggests free ruler also overcoming the results from The benefit both pacity. the large, If sufficiently enough. not be cred- not is guaranteeing may that ruler suggest the results from The ibility capacity. fiscal fragmented under riding Building State Fiscal-Military of Theory Collective-Action A rneo nvriyPes Princeton Press, University Princeton 99(4):1218Ð1244 Rev Econ Am 14(53):53Ð70 Soc Polit paper Working 37(3):315Ð350 11-06 Madrid Albuquerque de III Carlos Univ. history, economic in WP bankers. Genoese and ohtetertclagmn n h vdnefo nln n ooilMex- colonial and England from evidence the and argument theoretical the Both explain to dynamics incorporate not does argument theoretical the Second, 63 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

828 827 826 825 824 823 822 821 820 819 818 817 816 815 814 813 812 811 810 809 808 807 806 805 804 803 802 801 800 799 EDITOR’S798 797 796 795 794 793 792 791 790 789 PROOF788 787 786 785 784 783 uire-atsD(91 itraMltrd éio 3511.EiinsAee,México Ateneo, Ediciones 1325Ð1810. México, de case Militar the Historia enforcement: (1961) trade. D and medieval Gutiérrez-Santos commitment, from Coordination, (1994) lessons BR economy: Weingast modern P, Milgrom the A, to Greif path the and Institutions (2006) In: A por Genoa. Greif medieval Impr. late edn. growth: 1845Ð1853 economic and hacienda, systems real political la Self-enforcing de (1998) A general Greif Historia Rev (1791) Res Am C Lat Urrutia 1750Ð1821. F, America, Spanish Fonseca in politics and society, Soldiers, (1982) J Fisher Rethink- (eds) E Hellmuth J, Brewer In: structure. state modern modern early early and in medieval Variation (1999) in T regimes Ertman and states building leviathan: Baltistan. the in of authority Birth and (1997) T state-formation Ertman of study Yale A 1492Ð1830. kinship: America, Charismatic in (1983) Spain RM and Britain Emerson world: 1700. Atlantic before the America of and Empires Spain (2006) (ed) JH L Elliott Bethell University In: Yale America. decline. Spanish of Colonial age (1987) an JH in Elliott statesman the Madrid olivares: Financiero, of Derecho count-duke IV. The Felipe (1986) de Hacienda JH y Elliott Política (1960) A Ortiz from Domínguez evidence change: political and taxation, Warfare, (2011) A Vindigni G, Federico M, Cambridge Dincecco 1650Ð1913. Europe, 71(1):133Ð161 finances. public Hist and Econ transformations J Political responsibility. (2011) ministerial M 106(3):483Ð and Dincecco Econ hazard, Polit moral War, J (2011) II. G Philip Cox under Spain and debt sovereign Pennsylvania of México America. theory Mexicana, The Latin Alianza (1998) atraso. in J del Conklin nation-state orígenes Los the (1990) and J Coatsworth war debt: 19:163Ð185 and Sociol Blood Rev Annu (2002) sociology. MA fiscal and Centeno state 97:155Ð The Econ (1993) Polit JL J Campbell debt. sovereign Britain of model in recontracting state constant A century (1989) eighteenth K Rogoff the J, leviathan: Bullow Rethinking (1999) (eds) E London Hellmuth Hutchinson, J, power. of Brewer sinews Spanish The Colonial (1989) (ed) J L Brewer Bethell In: empire. American 53(3):389Ð its Rev and Hist Mexico Am Bourbon (1987) Hisp DA Mexico. Brading colonial late in elite and Government University (1973) Cambridge DA 1763Ð1810. Brading Mexico bourbon in merchants long and the Miners during (1971) DA France Brading and Britain of history fiscal comparative the Press, Towards University Oxford (2004) 1200Ð1800. R c. Europe Bonney in state fiscal the of Oxford rise Clarendon, The finance. (1999) state (ed) and R systems Bonney Economic (1995) (ed) R Bonney 64 ftemrhn ul.JPltEo 102(4):745Ð776 Econ Polit J guild. merchant the of Princeton narratives. Cambridge Analytic Press, (eds) University Cambridge BR Weingast JL, Rosenthal M, Princeton Levi Press, A, University Greif R, Bates Mexico Torres, G. V. Press, University 17(1):217Ð222 Oxford Germany. and Britain in state century Oxford eighteenth the leviathan: ing Cambridge Press, University Cambridge Europe. 12(4):413Ð444 Soc Polit Cambridge Press, University Cambridge Press, University Cambridge Haven New Press, 71(4):887Ð914 Hist Econ J Risorgimento. Italian the Cambridge Press, University 513 Pennsylvania Press, University 178 Oxford Press, University Oxford Germany. and Cambridge Press, University Cambridge America. 414 Cambridge Press, Eu- its and Cambridge Britain Press, industrialization: University and Cambridge Exceptionalism 1688Ð1815. rivals, (ed) ropean L Prados In: century. eighteenth Oxford ..Arias L.M. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

874 873 872 871 870 869 868 867 866 865 864 863 862 861 860 859 858 857 856 855 854 853 852 851 850 849 848 847 846 845 EDITOR’S844 843 842 841 840 839 838 837 836 835 PROOF834 833 832 831 830 829 tsvg 20)Ceil omteti al oenErp:NrhadWigs revisited. Weingast and North Europe: 28:2Ð37 modern Rev early Hist Am in Hisp commitment 1575Ð1770. Credible Spain, (2002) New D in Cambridge Stasavage taxes Press, University Sales Cambridge (1948) revolutions. RS social Smith and States (1979) Levi A, T Greif Skocpol R, Bates Totowa In: Littlefield, reconsidered. absolutism & of Rowman economy I. political Edward The under (1998) JqL finance Rosenthal and politics, War, (1972) M Prestwich auteG 20)Saecvlsceycoeainadcniti h pns mie h intellec- the empire: Spanish the in conflict 87(3):567Ð576 and Rev cooperation Sci society Polit State-civil Am (2007) development. GB Paquette and democracy, Dictatorship, (1993) M Olson possible its and state fiscal exceptional an of evolution historical and nature The 41(1):1Ð (2011) Rev PK Hist O’Brien Econ 1660Ð1815. taxation, British of economy political gov- The institutions (1988) of PK evolution O’Brien Hist the Am commitment: Hisp and Constitutions 1763Ð1766. (1989) Spain, BR New Weingast of DC, the North army of the financing of the reorganization and The Spain (1953) New LN situados: McAlister and Silver (1994) M Mantecón and Britain Souto Spain, C, between wars Marichal the M and silver Mann Mexican In: empire: finances. of state Bankruptcy (2007) English C of Marichal analysis an 1130Ð1815: Berkeley society, Press, and California State of (1988) University revenue. M and Mann rule Of of (1988) establishment M the Levi III, Charles reform: enlightened and Absolutism Press, (1985) Tennessee GD of Inglis University AJ, Kuethe society. and military York crown, New 1783Ð1815: Press, Cuba, University (1986) Cambridge era. AJ in colonial Kuethe expenditures the cajas and Mexico: las income (2002) royal A de Knight empire: partir Spanish a the of análisis finances un American The 1680Ð1809: (1998) españa, HS nueva Klein la de economía administration La tax (1985) of HS bureaucratization the Klein structure: state and Revolution inten- los (2001) de J época Kane la E, en Kiser administración Su España. Nueva de Empire Hacienda and Real nation-state La to (1999) L path Jáuregui Spanish a absolutism: Haven for New Bargaining Press, (2008) University Yale R societies. Grafe changing MA, in Irigoin Univer- order Manchester Political III. (1968) William S of Huntington reign the in politics and policy Parliament, (1977) H modern Horowitz early in taxation and warfare of economy political 1450Ð The government (1997) representative JqL and Rosenthal liberty, P, crises, Hoffman Fiscal (1994) (eds) K Norberg Princeton control. P, and Hoffman authority in lessons comparative Africa: in power and financ- States war (2000) of J development Herbst the France: century fourteenth in taxation Royal (1971) JB Henneman Building State Fiscal-Military of Theory Collective-Action A a cnOgn18(1):155Ð186 Organ Econ Law J Princeton Press, University Princeton narratives. Analytic (eds) B Weingast JL, Rosenthal M, 39:263Ð298 Stud Am Lat J uladpltclatvte fteultramarine the of activities political and tual 64(2):408Ð446 Rev economy Hist British Econ the Nelson. of to industrialization Cromwell and from commercialization precocious the for significance 32 34:72Ð96 Hist Econ J England. seventeenth-century in choice public erning 74:587Ð613 Rev Hist 33(1):1Ð32 Am Rev Hisp century. eighteenth the in Caribbean the in empire Spanish Cambridge Press, University Cambridge 1760Ð1810. France, 73Ð123 pp York, New Blackwell, Basil capitalism. and war, States, (ed) 109:118Ð143 Present Past Cuba. in reorganization commercial and alcabala, the Knoxville Albuquerque Press, Mexico New of University 1680Ð1809. Bolivia, and Peru, Mexico, colonial 34(4):561Ð609 Mex Hist reales. 107(1):183Ð223 Sociol J Am France. and England modern early in México UNAM, 1786Ð1821. dentes: 88(2):173Ð209 Rev Hist Am Hisp Building. Manchester Press, sity Diego San Press, frontiers Academic The (eds) economics. JV institutional Nye new JN, the Drobak of In: development. economic for lessons historical Europe: Stanford Press, University Stanford 1789. Princeton Press, University Princeton Press, University Princeton 1322Ð1356. ing Consulados n cnmcsceis .1780Ð1810. c. societies, economic and 65 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

920 919 918 917 916 915 914 913 912 911 910 909 908 907 906 905 904 903 902 901 900 899 898 897 896 895 894 893 892 891 EDITOR’S890 889 888 887 886 885 884 883 882 881 PROOF880 879 878 877 876 875 ee 12)Eooyadscey 98en emntr e York New Bedminster, edn. 1968 society, and Economy Oxford (1922) Cambridge Blackwell, M Blackwell, Weber Basic 990Ð1992. revolutions. AD European states, (1993) European C Hisp and Tilly capital performance. Coercion economic (1990) C and Tilly institutions, political bargains, Fiscal Prince- (2008) Press, University WR Princeton of Summerhill state. modern honour the III, in of essays origins Charles medieval Europe: of the On century age (1970) eighteenth JR the in Strayer in state Spain fiscal-military New The and (2009) Spain (ed) C empire. Britain, Storrs of Great and Apogee France (2003) state: BH democratic Stein the SJ, of Stein birth the and debt Public (2003) D Stasavage 66 mHs e 88(2):219Ð233 Rev Hist Am ton Farnham Ashgate, Dickson. P.G.M. Baltimore Press, University Hopkins John 1759Ð1789. Cambridge Press, University Cambridge 1688Ð1789. ..Arias L.M. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 DOI (eds.), al. et Schofield N. Kazakhstan of Republic Almaty, 050010 2, e-mail: Abay University, KIMEP Economics, of Department Pech G. Germany Friedrichshafen, e-mail: 88045 20, University, Horn Zeppelin Seemooser Governance, Am and Management Public of Department ongtaeatasto ewe h l eieadteopsto seBrown (see opposition the and suspended regime be to old had the constitution between the 2011 transition whether a or laid negotiate as constitution apply to down, Egyptian would step succession the to of in Egypt, rules of the out whether president arose then immediately Mubarak, question Husni the for called When governments. House elected White freely the by assembly succeeded the were regimes constitution, autocratic where prior tries the of constitution. rejection new absence a a the on after In negotiations policy. or free quo enters status constitution the prior changing it a of constitu- assembly, form over of succeeding the negotiations the take enters by which assembly reform accepted the tional If when point. outcome model quo default We autocrat status the the modified. mover, becomes the be first selecting the can in As point game. free quo dynamic is a status as the reform how policy and of design with constitutional rule space rights policy policy property a in guarantees provides it reform. point and Moreover, quo policy. constitutional elected social status on succeeding and a redistribution negotiations a con- legitimate dimensions for by as quo print defines accepted status constitution blue be a A a not selects as may autocrat assembly or an constitutional may where constitution model This spatial stitution. a develops paper This Introduction 1 Pech Gerald and Michalak Katja Transition Political in Constitutions Stable ews otakKihShaebr n omnShfil o epu comments. helpful ( for Michalak Schofield K. Norman and Schnakenberg Keith thank to wish We oercnl,cntttoa ucsinhsbcm nisei ayAa coun- Arab many in issue an become has succession constitutional recently, More 10.1007/978-3-642-35239-3_4 .AtrMbrkeetal eind h nei iiaygvrmn,ie the i.e. government, military interim the resigned, eventually Mubarak After ). [email protected] [email protected] B ) dacsi oiia Economy Political in Advances pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , , 67 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 a infiatbrann oe ntetasto rcs,wswdl ent be interests. to property seen and privileges widely their was to process, on which transition hold military, the the to existing in the of able power of leadership discontinuity bargaining the of significant hand, one process. other had is transition the case the Egyptian On in the involved constitution. it parties authoritarian of all face to the acceptable on not Therefore, be to out turned tion constitution. new a over negotiations wards osiuinlsuccession. constitutional h ieo nlzn hsppri perda fteam a eoigissup- by its dominated revoking legislature was a army of the face at the if project, in as process constitution parties. appeared Islamist reform this it constitutional of paper the outcome for the this con- port judge finalizing different to of slightly early time this too the in still is choice it constitutional Whilst analyzing text. to itself lend should e foecniestemltr spr fti re,oecudeulyageta h transition the that argue equally could one order, this pact. of a part as as classified military be can the considers one if Yet ( Leff and Munck otie infiataedet n ie tpvn h a oparliamentary to way the which paving constitution at provisional aimed a elections. and adopted amendments Forces, significant Armed contained the of Council Supreme 68 5 4 3 2 1 oni fteAmdFre ae ahrsmlrcoc rbe hndraft- when problem choice constitution. similar provisional rather the a of Supreme faced amendments the the Forces successors, its ing Armed by print the blue of a as Council accepted is which constitution a order select property con- the place. formal in yet kept the perseverance was lacked where constitution the autocrat case under the ambivalent established by an down is handed transition stitution Egyptian the that suggest a ogadrcn itr fcntttoaim h hla osiuinwsa was constitution Chilean The constitutionalism. of history recent and long today. place a in has remains national 1980 mod- the of on a constitution military and the the left, parts, versus large the civilians In of of council. parties power security of voting relative admittance the constitution, chang- of the for amendments ification of quorum The parts the 1989. vital of lowering in non were the plebiscite ing which powers, by presidential amendments constitution on constitutional reformed restrictions negotiated a included of center part the as par- and Chilean adopted right the defeat, the electoral thereby Following of election, constitution. ties for the stood of tran- Pinochet letter the 1988, the which im- by In by and sticking place: rules the took junta set finally the which democracy but to of power, sition workings of exercise internal its on the constraints governed posed only not subsequently which n otsadVa ( Vial and Montes and (2002) Barros see overview June an 15 For Times, Financial dissolved, parliament orders court Egypt of framework the In case: Egyptian the of ambiguity the on agree schemes classification Other May 20 Times, ( Financial Dunnes freedom, and Egypt’s Brown see process the of details For fteeeet,tofcssadot nteoehn,teMbrkconstitu- Mubarak the hand, one the On out: stand facts two events, these Of hl,a h eodeapewihw oka,i la xml fsuccessful of example clear a is at, look we which example second the as Chile, oevr hltti ae ok notepsiiiyfra uortcrgm to regime autocratic an for possibility the into looks paper this whilst Moreover, hr r la ifrne u losmlrte ewe gp n hl:Chile Chile: and Egypt between similarities also but differences clear are There 1 h lce alaetstoto htpoe ob up odto- road bumpy a be to proved what on out set parliament elected The h gpintasto a ecasfida n fdfa fteodorder. old the of defeat of one as classified be can transition Egyptian the 1997) 4 5 n18,teCienmltr ut dpe constitution a adopted junta military Chilean the 1980, In 2012 . 2011 ). 2005). 3 npicpe u framework our principle, In 2 2012 hs w observations two These .Mcaa n .Pech G. and Michalak K. . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 hni a adycntandteatca n ssgicnl mne ntere- the in amended significantly is and even autocrat process the process. reform form constrained the hardly in cases has accepted Hereby, con- is it says. prior constitution when it preexisting the what a elect of where will irrespective possible are outcome, assembly bargaining succeeding default a as property. theoretically, stitution stationarity that, the the by on argue abide compromise also to will have We he will time, he some that for with- expects himself assembly autocrat constitu- constitution constitutional the a succeeding if i.e. a Only constitution, amendment. by further accepted stationary out be a to write predicts to he which choose tion may he demise, his before anann rpryrgt sapeeust o osiuinlstability. constitutional for in interested prerequisite is a which is theoretical middle-class a rights a provides property that model maintaining idea our stated Here, frequently autocrat. constitutions the stable the for dominate, of underpinning right interest the the with in single coalition are a a which forge exist not to expected If redistribu- is opposes constitution. it class the or middle tion the by and himself assembly autocrat constitutional favor bind the the which to dominates parties party want If by not constitution reform. dominated does a constitutional is he assembly over down redistribution, constitutional bargain hand future to to the that expects prefers expects he strictly who autocrat on the depends not or Whether behalf. process. their on on constitution influence a sufficient exert write can him class make assume owning to we property autocrat paper the this the of In group demise. a interest his writing the after in that play autocrat, power the the which influence through avenue can an constitution, creates it way, that in in constitution outcome constitution. new default a negotiating a in involved as satisfied. otherwise constitution risks not the preexisting against is the insure the to changing players on order in agree the interested in are yet of end which may one parties may constitution successor of elected reform the constraint other constitutional reasons, participation the those of from For the attempt concessions if secure The to conflict process. tries open the each where in attrition may involved of parties parties game negotiating newly a The to in uncertainties. attempts point up many society reversion end creates a a compact which as social in also its process but arrange reform reference The a reform. as constitutional both for serve can which process transition the the in military. during bargained the which broker by right power projected the was a of which as parties power the served of were shadow which these military Chile by in the exercised whilst mainly influence revolution was period. the transition this the brought is during Egypt was regime countries old In power the both of with to associated transition organizations Common and the parties protests. hand, other of street letter the by the on about obeyed Egypt, hesitation, In some after constitution. regime, the the because Barros (see place junta took the transition of dealings the on constraint binding Transition Political in Constitutions Stable 6 e ..Odsok( Ordeshook e.g. See oevr eso hti h uortcnhn onacnttto immediately constitution a down hand can autocrat the if that show we Moreover, eso,fis fal htcntttoseitwihaesal ntetransition the in stable are which exist constitutions that all, of first show, We preexisting a use to assembly constitutional succeeding a expects autocrat the If the in point focal natural a as constitution preexisting a see we paper, this In 1997 ,Esel (2001). Easterly ), 2002 .Ms significantly, Most ). 6 69 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 uleulbimaayi hc xed n ple h rsn rmwr othe to framework present the applies and extends which analysis equilibrium full a osiuinlpoesswe aigteatraieo ecnigit conflict. into descending by of ( abide alternative Pech agents the where facing properties—i.e. processes—when self-enforcing constitutional with constitutions design Weingastto and which (North by rights devices son property commitment uphold as to pledge constitutions 1989 credibly of can ( role government Weingast the a by as on introduced accept focus been agents coordi- approaches action—has which as of coordinated constitutions crossing for understanding the ab- triggers of “red-lines” the way as in different devices—understood achieved A nation outcomes constitution. a to of compared sence outcomes Pareto-better on coordinate ok tcntttosi em ftepoete n eiaiiyo h oigrlsit rules voting literature the of of strand desirability and Another properties mechanism. the provides. a of terms as in law constitutions of at looks rule the contain which tutions aeo hl n gp,oecnageta hie vrpltclisiuin were outcomes. institutions policy political particular the over promoting In choices or preventing outcomes. that at argue policy aimed can for clearly one often implications Egypt, has and the Chile policy with of quo together case status institutions of a selection of than the legitimization rather directly, selection more policy choices for policy rules making general provide typically norms constitutional u a anyamda rvnigprimn rmslcigplce hc eeaantthe against were which policies selecting from Muslim rulers. parliament military the preventing Egypt the by post-revolutionary at of dominated of interests aimed set-up parliament institutional mainly a the dissolve was with to interfered but Forces which move Armed decision a the The was antisystem. of brotherhood not Council were Supreme they provided the constitution, of reform the under admitted were they Barbera See game. political the from emerge stages ( later Jackson at and which rules of selection the determines constitution. American the a constitution. can 7 by advanced was terms outcomes (1986 in desirable particular Riker achieve in to constitution situations the strategic and manipulating general in of rules of selection the at Looking Literature Related 1.1 70 10 9 8 lyr ihnasailplc rmwr n hssaeplc outcomes. policy institu- shape of thus selection and framework veto the of policy or set spatial the setter a determine within agenda institutions how players an shows on example, of predicated for strand move (2002), a are Tsebelis the in which tions. model game as our political such places the choices This of previous model. equilibria spatial derives a which in literature point of quo status a of selection ntecs fCie ate ftelf eentamte ne h iohtcnttto but constitution Pinochet the under admitted not were left the of parties Chile, rules of of case choice the constitutional In and/or historical the how ask may one setting general more a In (2002 Schofield (1996). Riker also See e,freape esah( Gersbach example, for See, norfaeok osiuinpoie oa on hc nbe gnsto agents enables which point focal a provides constitution a framework, our In 2008 ra uortt ielglyesrndgaate ohsfloes(Myer- followers his to guarantees enshrined legally give to autocrat an or ) 2009 .Mroe,Gosa (2002 Grossman Moreover, ). 10 n av ta.( al. et Naqvi and ) n auof( Lagunoff and 2004) ial,i nacmayn ae,Mcaa n eh( Pech and Michalak paper, accompanying an in Finally, ihhsaayi fteeet edn oteaoto fteAmeri- the of adoption the to leading events the of analysis his with ) 7 u ae oescntttoa hiei em ftestrategic the of terms in choice constitutional models paper Our n abr n ako ( Jackson and Barbera and 2004) 2007 ou nsl-nocn rpriso consti- of properties self-enforcing on focus 2012) ). lbrtso hslgcadapisi oteeouinof evolution the to it applies and logic this on elaborates ) ie odtosudrwihi spossible is it which under conditions gives ) 2006 ). .Mcaa n .Pech G. and Michalak K. 2012 1997 provide ) 9 8 .Other ). Whilst Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 inlcmuiaindimension. tional/communitarian T bet fe efc netv otatt h uort ywihi fessupport offers it which be by must autocrat, clientele the the to either contract incentive reasonable, perfect be a to offer assumption to this able For clientele. his be. conflict of of case the the military, in the redistribution powerful achieving more of ne- possibility the free the but of will everyone, cost limited to more the high expect negotiating especially we freely be case, clientele. of to this autocrat’s gotiations option In conflict. the the into when as degenerates player modelled constitution powerful the we particularly a what harmony is some and military see military can The we the be cases between model, both our In interest of autocrat. of purposes the the of for clientele can, the with and autocratic leadership associated class Egypt, military possessing of the the In with aspect but clientele. interests lingering separate autocrat’s shared organizationally a the were as with military transition it and of associate government aftermath closely Therefore, the military. and the in junta within military the from the emerged identify junta the can case one Chilean the In player. a class, dle inascae ihatxrate tax a with tion—associated rbe fteatca.Section tion autocrat. the of problem Section rejection. its after or constitution prior osiuini pair a is constitution A Model The 2 Section Paper the of Outline 1.2 of significance stability. the constitutional on for focuses wealth detail, class more middle in paper, That process. transition Chilean Transition Political in Constitutions Stable 12 11 ue.Saiiypoete feetrlrlsaedsusd o xml,i abr n Jackson and Barbera in example, for discussed, are rules electoral ( of properties Stability rules. rae oefrrlgo npbi ieo h eaieiprac fteso- the of importance relative the or principle. life market public free the in versus versus religion principle secularism solidarity for authoritarianism, cial versus role liberalism greater representing a scale a along ie ocuin rmorframework. our from conclusions rives Section constitution. Section prior autocrat. a the of for presence constitutions the in reform tional 2004). ed o ics nti oe ue oenn otcntttoa hie uha electoral as ( such Kitschelt choices constitutional post governing rules model this in discuss not do We × hr r he oi-cnmcgop,tecineeo h autocrat, the of clientele the groups, socio-economic three are There utemr,w sueta h uortpretyitraie h preferences the internalizes perfectly autocrat the that assume we Furthermore, 5 X icse plctost ifrn xeine fpltcltasto n de- and transition political of experiences different to applications discusses =[ 2 M esu h oe.Section model. the up sets 0, 1996 n h okn class, working the and , 1]× nsta h aoiyo oiycocscnb usmdudradistribu- a under subsumed be can choices policy of majority the that finds ) . t x) (t, ersnigacutysbscchoices basic country’s a representing , t 4 adsca policy social —and xed u eut oadnmcstig Sec- setting. dynamic a to results our extends L 2.1 ed o xlctymdltemltr as military the model explicitly not do We . 3 rsnsngtain nteasneo a of absence the in negotiations presents nlsstesai osiuinlchoice constitutional static the analyses 2.2 eal agiigo constitu- on bargaining details x 12 hc a emeasured be may which h oiyspace policy The 2.3 11 eie optimal derives nredistribu- on R h mid- the ,  71 is Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 iie.Ti bevto evstepltclters tugigfra explanation an for struggling theorist political the leaves observation This citizen. oeaueo oaincpue h e noepr only. part income net the notation—captures of abuse some is hc seel itiue mn ebr ftetregroups. three the of members among distributed evenly is which rate tax a at tribution group oilplc cl where scale policy social ihswihwr endudrta osiuin netepoet re sac- general is through order achieved be property taxes. only general the can i.e. Once property rules, of constitution. redistribution property that principal, the in under of cepted defined acceptance implies were negotiations which for rights template restricting a over In as bargains choices. constitution assembly admissible of prior choices the of to set where space appro- the case bargaining or specify the the defend must for to Yet we able conflict. be reform, to of constitutional be expect situation agents may a which scenario in wealth this of priate with level con- prob- associate the the less as over we of is negotiations which thought free level assumption of wealth This case reversion class. the The class for other stitution: socio-economic outcomes some one construct of we of when expense possibility lematic the the out at rule itself and enriching transfers in equality pose 13 each of representatives of are incomes Gross group policies. redistributive different of effects support. “sells” for exchange autocrat in the clientele alternatively, his or, to rules advantages constitutional those favorable for exchange in 72 15 14 n where and suethat assume u ol aet iutnosyafc h rpryo h idecasa well. alone as property class their middle of affect the beneficiaries of not of property will majority the regime vast affect new the simultaneously the to to by have assurances an would measures provides but expropriative such it that but a of system court on thinking old in clearly the of tried is are way choice regime proper old a The the Such zero. country. a of of rate order tax economic a the or reduces of level. choice 1 election constitutional the of the constitution, rate is the out choice tax bargaining turns extreme a freely it However, either as laws. such selecting tax scenarios simple to by bargaining determined some normally for are that policies tax because sight, first soewemn e o w loain hr h noeraiaini h same, the is realization income the where allocations two R for yet overwhelming is agiigotoe when outcomes bargaining uhacpac osnti eea ueotta niiulcsso ufi”ercmn under enrichment “unfair” of cases individual that out rule general in not does acceptance Such at counterfactual be to appears stage constitutional the to policy tax a of choice the Assigning that assume effectively We titypeestealcto where allocation the prefers strictly u nalsceisw nwo,aeaeicm xed h noeo h median the of income the exceeds income average of, know we societies all In h e noedsrbto sotie rmtxn noeaalbefrredis- for available income taxing from obtained is distribution income net The netn u supino esbetxplce noteuiiyfnto for function utility the into policies tax feasible on assumption our Inserting o ipiiy easm htalgop aetesm iewe acltn the calculating when size same the have groups all that assume we simplicity, For i = i α n eoigaeaeicm o redistribution for income average denoting and i v w i v (x) R i x M =−| >w + w t , where ∈[ u 15 x x i R t x) (t, i x R L 0, | = i 2 R h tlt ucino iie eogn oclass to belonging citizen a of function utility The . , a eiorpi rfrne hr h tlt function—with utility the where preferences lexicographic has 1] atrstels soitdwt elztoso the on realizations with associated loss the captures i w x and rcesfo h a nnealm u transfer sum lump a finance tax the from Proceeds . = M i n = scitizen is L,M,R αv and M t and i ge on agree (x) x x L x scoe to closer is + eefcieyasm htacpigthe accepting that assume effectively we , yet ersnstebispito group of point bliss the represents i ( snticm fe ae n transfers and taxes after income net ’s 1 α t − easm that assume we , R t)w = x .I re ouiul assign uniquely to order In 0. i R + . 13 w t w. eobtain we , .Mcaa n .Pech G. and Michalak K. R 14 sicm motive income ’s hrb eim- we Thereby i .We i Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 ainpyofo ahplayer each of off pay uation 1 ito rcs ae h omo pncnitadsc oflc a eprecipitated be can conflict define nego- such We and free party. conflict the open any where of by or form referendum the a takes process in requires, tiation support constitution widespread reform factually, draft or the where formally scenarios with compatible is sumption ntrso noeraiainadteicnie o rtn osiuinwould constitution a writing for case incentives The the and minimal. realization be income of terms in befrdsrbto as distribution for able l lyr ofel eoitn h osiuinwt soitdlottery associated with constitution the negotiating freely I to players all 1 Definition negotiating freely of lottery the values to expected players the all where by point preferred policy the particular, in constitution of contain, the and negotiating empty freely over non acceptable is generally are which outcomes of all rcs nteasneo eal osiuinis constitution default a of absence the in process on fve h xettosoe h nlotoetk h omo lottery a of form the take outcome ante final ex the an over from expectations unfolds,  con- the process in- the view is, constitutional of of That the outcome uncertainty. point precisely some the with how constitution, predicted of default be dependently a only can of process absence reform the stitutional in that assume We Constitution New a Negotiating Freely 2.1 a only that assumption the taxation make expropriating we non distributions with share to equilibrium income order relevant political In empirically a practice. of for in expro- possibility of policies favor the tax in for expropriating majorities allow observe have hardly should we we societies when democratic priation in that fact the of Transition Political in Constitutions Stable osiuin nuigfrhmefadfutotoeof a outcome negotiating default con- freely a the to of himself revert outcome and for the outcome ensuring with this constitution, dissatisfied reject is can who process player reform a stitutional point any at that maintain aia sdpoe.I h aia smsl netdi h aua eore sector, resources natural the in invested mostly is islikelytobelow.Weassume capital the If re- deployed. to is likely capital are practice, In nationalization value. its its in duce involved problems agency capital, productive rmcnaiyof concavity From redistribution. ={ ={ >t x oethat Note i eutwihw s ntepofo Lemma of proof the in use we which result a , and (x,t,π(x,t)) x,t 0 ( 1 > − | t .Tecs where case The 0. x t) (x, γ) r fee.Mr omly edfieteset the define we formally, More offered. are I of I a lsdgah nwa olw,w ou ntecs where case the on focus we follows, what In graph. closed a has stestotfail ucmswihaewal rfre by preferred weakly are which outcomes feasible out set the is w i R } v  ihpoaiiyweights probability with satal vial o eitiuin If redistribution. for available actually is and n iert of linearity and t w 0 (x x t) (x, = γ = 0 orsod oastainwhere situation a to corresponds 1 ,t slkl odpn ntekn fidsr nwihthe which in industry of kind the on depend to likely is t (1 0 i 0 −γ)w ( = ∈} = ) 1 steepce au of value expected the as − R,M,L strivial: is 0 R , γ)w 3 +w i u = in M R L,M,R +w t >w hnetrn h osiuinlreform constitutional the entering when tflosthat follows it L (,)< π(x,t) .As R M a noc t rfre outcome preferred its enforce can Eu . n en vrg noeavail- income average define and w>w i 2 () hr eso htteset the that show we where o all for 1 hogottepprwe paper the Throughout . u I L i 0 x x Eu h etawy favors always left the , fotoe hc are which outcomes of = and w and Eu i R  u < () t x t) (x, osssmil of mainly consists L , i t ()  o hslottery. this for a noc its enforce can : uha as- an Such . h contin- The . i (x 0 ,t  0 ,i.e. ) for 73 γ Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 itsoe osiuinlrfr rd-oodsg ftecnttto.Ti e of set This constitution. the of design nego- de-novo players or of reform set pre-determined constitutional a over that tiates suppose Moreover, autocrat. the by down loconvex. also eto htepcain vrteotoefo reyngtaigteconstitution the lottery negotiating a freely of from form the outcome take the over expectations that sertion st sueta ahpryi ie hnet mlmn t rfre outcome preferred its implement to chance a given lottery is probability a party a such each with provides that which assume game bargaining to a is modelling consistently of way ble Proof realization ae httepryipssispeerdplc on.Tu,if Thus, point. policy preferred its imposes party the that tates em 2 Lemma following the range, intermediate the holds: In minimal. lemma be would constitution a writing for rfre ucm ntetasto and transition the in outcome preferred 74 16 constitution a Suppose of Presence the in Constitution a Negotiating 2.2 and and party u oe osntatmtt xli oflc u nta ssacnitseai to scenario conflict. conflict of a shadow uses the instead in but settlement conflict a explain rationalize to attempt not priori- does a model form our players example, where case For of the against conflict. cover results entering to the model of before the any coalitions extend changing may Without we conflict. paper open this of form the take negotiations get concave, o te pcfiain,seMcaa n eh( Pech and Michalak see specifications, other For ecnmdf a fsb ditn oflc cost conflict a admitting by offs pay modify can We goigtetiilcase trivial the Ignoring (x R ( 0,x i 0 lopee hspit tms ein be must it point, this prefer also xetduiiyfo reyngtaigtecnttto is constitution the negotiating freely from utility expected , ycnaiyof concavity By ,t V V R M i i 0 I 0 0 if if n xett elz on ntercnrc uv fte i.Nt that Note win. they if curve contract their on point a realize to expect and − ) t x) (t, = = For sntvnsigysal ..teeis there i.e. small, vanishingly not is w w for ε) P P M M 1 R R w ihcranyoe otr ihepce outcome expected with lottery a over certainty with ≤ > is v v >t M i i P w, w, ( x x 1,x 0 > j R R . > 16 c w v L for 0, + + tlattepoint the least at , pcfigatxplc combination tax/policy a specifying ntecs hr hsopruiyaie,rtoaiydic- rationality arises, opportunity this where case the In ) n if and  ,if P P h set the hc scmo nweg oalpaes n possi- One players. all to knowledge common is which i t M M 0 = M = v v L,M,R. i i R is h oiyraiainis realization policy the wins, ,tersl fLemma of result the 0, x x I is h oiyraiainis realization policy the wins, M M snnepyadconvex and empty non is + + R I P P a ontigaoti.Aan incentives Again, it. about nothing do can ycneiyo rfrne and preferences of convexity By . x 2012). L L 0 v v ,t i i 0 ε> x x utb in be must L L L uhthat such 0 + + and P K 1 R i − M 2 w hc sicre ffree if incurred is which P i nyhne nteas- the on hinges only I . t x) (t, a omacoalition a form may + .Mcaa n .Pech G. and Michalak K. Because . c L ( 1,x L L w 1 ( titypeesto prefers strictly 0,x i − a enhanded been has is h policy the wins, M + P ) x P R for R 0 ) v L ,t hs for Thus, . w w sstrictly is w 0 .As M  , ≤ I (2) (1) M w is Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 tlatb h lyr novdi osiuinlbrann.If bargaining. constitutional in involved players the by least at Proof srjce.Hne ahplayer each Hence, rejected. is agiesi eemndeoeosyt h oe.I htflosw ou nthe on focus Sections we bargain. follows parties what two In where model. case the to exogenously determined is bargainers Transition Political in Constitutions Stable 17 Section game. bargaining party two the of applications rmbrann ntecntttoa eompoeshv netvst ceteven accept to incentives of have process outside reform constitutions constitutional the in bargaining from eodb tlatoepae.Hne eeto fapooa hntedefault the when proposal a conflict, of into rejection descending from a off Hence, player. is one constitution least at outcome by an implementing vetoed in results proposal a rejecting rpsl umte by submitted proposals spsil ucmsfrtebrann aebetween game bargaining the for outcomes possible as la rdcinwihcaiinfrs neednl ftedfutcnttto.SeMichalak See of constitution. formation. default set coalition the the endogenous of for given the independently is (2012) effective—there Moreover, forms, Pech in is assembly. and coalition always rule which coalition the voting prediction grand in clear the and because a coalitions representation empty effective non of is of strength coalitions—which set effective Relative the clientele. determine their assembly of support secure can oecuetecs hr gnsaepeitdt aen agiigpwra all at power bargaining reform. no constitutional over have bargains to which requires predicted coalition only strictly the it are enter as is agents they assumption proposal where when mild a case a is making the which exclude of coalition hold to a probability results in our ante player All ex each application. for the our positive that in singular assumption be the to under out turn they although valued P admpooe aeweetepooe ae aei-rlaei fe othe to offer take-it-or-leave-it a Let makes proposer player. the the other to where time game any proposer at random revert to option Recall, the draft. has constitutional still outcome. reform society cooperative the non in on group agreement any an that find however, to unable are gainers constitution, the gotiating P osiuino l he ate agi vrcntttoa eomaestraightfor- are reform the model. bargaining constitutional impose two-party to over the position bargain of a extensions parties in ward three is party all one or where constitution cases different The cases. remaining tion 3 Lemma I. in in constitutions constitution over a preference choose strict will a has autocrat he the bargaining, players two with j i ed o xlctymdleetosbtrte sueta h ersnaie fec group each of representatives the that assume rather but elections model explicitly not do We → → netepedtrie agiesaccept bargainers pre-determined the Once If etiko h agiigpoeuea aigtesmls omo w player two a of form simplest the taking as procedure bargaining the of think We j c i Naturally, . c ∗ pair a ycntuto of construction By ∈ ∈ I I , , fteeaetobranr n h uortuiul rfr constitu- a prefers uniquely autocrat the and bargainers two are there If x t) (x, hscnttto ssrcl rfre vraycnttto o in not constitution any over preferred strictly is constitution this u j (c) c/ ,j i, ∈ ∈ Γ stedfutuiiywihplayer which utility default the is I ij I ∈{ (c) hc maximizes which eut neeyaetraiighso e otnainpay continuation her or his realizing agent every in results R,M,L I i eacrepnec hc sin oec hieof choice each to assigns which correspondence a be Γ(c) e,a h olwn em hw,i h ttcmodel static the in shows, lemma following the as Yet, . saproposer, a as c evsa h eal ucm hc rvisi h bar- the if prevails which outcome default the as serves ,any i hnmkn eompooa to proposal reform a making when , } and V c 0 yti eie lyr h tn obenefit to stand who players device, this By . ∈ i = u I 3.1 i P ssrcl rfre otedfutoutcome default the to preferred strictly is x t) (x, j i oethat Note . → and j n umte by submitted and , c ujc to subject 3.2 ahrta eetn ofel ne- freely to reverting than rather 17 ics ngetrdti special detail greater in discuss P 3.3 j i → c elzswe proposal a when realizes u i hc ilutmtl be ultimately will which ie noeve fthe of overview an gives j j and x t) (x, and c/ ∈ j P I h equilibrium the , ≥ j j rpslcan- proposal a , → u j saproposer, a as i j hoe for chooses , ih eset be might (c) I whenever .If I c/ . ∈ 75 I c , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 Γ in r losrcl rfre vrcntttoswihaentin not are which constitutions over preferred strictly also are tions ae fPropositions of out- cases default their all where over unanimity constitution of a case prefer the to outcome weakly extends default least also its at It over come. parties constitution the other prefers the strictly and party dominant The cess. rfre yalprisoe h eal ucm.I h eane ftepprwe paper the of remainder the In outcome. default the consider over parties all by preferred n,hne fascn re magnitude. order second a of of hence, and, preference unique a has autocrat ihteotoepooe ybt proposers. in both point by ideal proposed proposer’s outcome the the not with select to proposer one of choice the titydmnt h lentv fraiigtedfutotoefo oflc with accept conflict they from that outcome commands default rationality the and realizing certainty of alternative the dominate strictly osdr aewhere case a considers osiuinlcoc xssin exists choice constitutional elz on ntercnrc uv r fti violates this if or, curve contract their on point given up, a set realize is game bargaining the way The Constitutions Optimal 2.3 against rejected be not 76 19 18 o osri rpsl,ie n osiuinntin not constitution any i.e. proposals, constrain not and hr nqeyeit constitution a exists uniquely there in proposal a choose owiebth sas nifrn ewe rtn n o rtn constitution. a writing not constitution and which writing to between respect indifferent also with is indifferent he only but not write is to autocrat the hence come, c/ igacost, a ring 4 Proposition which process: constitutions bargaining i.e. the constitutions, in stationary amended as not are game static the of ofconstitutions boundary the on constitution hr xssasainr osiuinwihi tlata odfrteautocrat the for good as least at is which constitution stationary as a exists there ij ∈ ntecs fProposition of case the In that Recall c hslmaetnst h aeweeol n at oiae h eompro- reform the dominates party one only where case the to extends lemma This (c) I . j e o hscs efidta h uortawy nsu ihhsdfutout- default his with up ends always autocrat the that find we case this for Yet . sielpit in points ideal ’s , c/ I ∈ stecoc e fteatca n banuiu pia hie nthe in choices optimal unique obtain and autocrat the of set choice the as x I o n constitution any for c R osntcntantepooe te hnb eurn i rhrto her or him requiring by than other proposer the constrain not does ∗ c = ∈ hnteatca a ietypooeacnttto ihu incur- without constitution a propose directly can autocrat the When srsrce ob in be to restricted is x I M ,soevenif hc does. which I 5 I h olwn rpsto hrceie srcl)optimal (strictly) characterizes proposition following The . I c and e tcntan h epne nrjcigapooa.If proposal. a rejecting in responder the constrains it Yet . .For ihu rcpttn oflc.With conflict. precipitating without L 6 where sn h em,w a ocueta hs constitu- these that conclude can we lemma, the Using 6. oiae h osiuinlasml n ooptimal no and assembly constitutional the dominates M i I and c nti ae h uortmycos constitution a choose may autocrat the case, this In . and ∈ c L I hc sntstationary not is which j agiswt an with bargains R u h rfrnei nyi em fplc realization policy of terms in only is preference the but c otr on lottery a I ∗ agi n ge on agree and bargain h uortms iht osri tleast at constrain to wish must autocrat the , ∈ I hc speerdb h uortwhen autocrat the by preferred is which c h w agieshv netvsto incentives have bargainers two the 18 Γ ec,acnttto hc does which constitution a Hence, M ij (c) I c at nfvro eitiuinthe redistribution of favor in party oeta ycntuto of construction by that Note . ssrcl oiae ythe by dominated strictly is , t ihnnzr egt must weights zero non with x t) (x, I hysildsge over disagree still they , c , eas ecno agree cannot he because i. in c/ e ∈ . .Mcaa n .Pech G. and Michalak K. I ∈ o which for I r tlatweakly least at are I , I Γ oraieapoint a realize to . 19 ij (c) Proposition assigns Γ(c) x . = i c ’s 7 , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 o oniewith coincide not otwe omtigt osiuinad ssoni h ro fProposition of proof the in shown as and, a constitution incurs a autocrat constitutional to the static committing problem, when choice the cost constitutional discussing dynamic when the In autocrat problem. the choice for constitutions optimal for with coincides case This binding. are constraints Both b). c) case outcome. default the is in ec,teotmlcntttoa hiecicdswt hspoint. this with coincides choice constitutional optimal the Hence, that h osriton constraint the pia rpsli nqeydfie.I hspiti eetda eal,teconsti- the the default, as and selected convex is strictly point If are stationary. this is If maker tution defined. proposal uniquely the is proposal of optimal preferences maker, proposal is ahpooe a opropose P to has proposer proposal this each selecting by off better is stitution Proposition nqepooa hc,i eetda eal eut nasainr constitution stationary a in results default as selected if which, proposal unique nti eto edrv h pia osiuinlcoc o h uorti ebe- Lemma he from know if we autocrat As the imminent. tution for is choice demise constitutional his optimal that the lieves derive we section this In Choice Constitutional Static 3 point In a) possibilities: three with left are we and given Proof Transition Political in Constitutions Stable 21 20 ntecnrc uv,i.e. curve, contract the on and ssrcl ovxuls tcicdswt the with coincides it unless convex strictly is n and ing either tlt ie hti sin is it that given in utility be to has posal ili culyipc ntescesrsdcso te hnb eurn hmto them requiring by than other decision in successor’s only the amendments on propose impact actually it will tttoa hiea h rbe fpcigacnttto rmwti h set the within from constitution a picking of problem the as choice stitutional i oseta h point the that see To that see To → I hspooiinalw st ou nsainr osiuin hnlooking when constitutions stationary on focus to us allows proposition This etspoethat suppose Next R I fteatca selects autocrat the If . j speeecsaelexicographic, are preferences ’s u snt hscicdswt h aewhere case the with coincides This not. is and P c en h Pareto-set the Define i i ilb cetdb h agies e nyi h osiuini nteset the in is constitution the if only Yet bargainers. the by accepted be will (c) B → c is ups that suppose First . ij P j foeproposal one If . R (c) j 4 or → spooa suiu hnthe when unique is proposal ’s osntapply. not does R i P c snt nta ae with case, that In not. is . spooa,osrethat observe proposal, ’s nld h aevleof value same the include j → I c i = a ebnig e proposal a Yet binding. be may nta of instead R B ( ij 0,x ae rpsltebnigsgeto h onayof boundary the of segment binding the proposal a makes B (c) P ij R i P (c) → ) B c ∩ B I ssainr hnslce sdfuti h aewhere case the in default as selected when stationary is ij j nldsalwrvleof value lower a includes = c ij ec eaegigt ou nteatca’ con- autocrat’s the on focus to going are we Hence . I ∩ (c) c maximizes (c) hnti stedfutotoe fteproposals the If outcome. default the is this when R . P I speerdpito the on point preferred ’s ihrpooe utpooepoint propose must proposer either , oethat Note . o h bargainers the for B ⊂ ij R P I (c) nta ae rpsl oniewt points with coincide proposals case, that In . t sarsodrwl eetaypooa hc does which proposal any reject will responder a as nta of instead P = nta ae h osritta h pro- the that constraint the case, that In . t iei idn,rcl htb u assumption our by that recall binding, is line 0 u h rpsrraie h ihs utility highest the realizes proposer the h uorti swl f fh selects he if off well as is autocrat the , i given B t B = ij ij (c) 0-line. c (c) P Setting . u ∩ i j P t ⊂ (c) constraint and t I = hnteohr h autocrat the other, the than aiie h proposer’s the maximizes 20 I scne.When convex. is -iei nqeydetermined. uniquely is 0-line and )Constraint b) . j nete case, either In c ie h eal con- default the given n eal consti- default any 3, P = j P → B ij i urnesthat guarantees maximizes (c) sbinding is c I R L hnit when sbind- is t or a a has = 0is u M 77 21 I I 9, I j , . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 t uinlcmat fw dniyteMsi rteho ihiswlaegasas goals welfare its consti- with party, future brotherhood the Muslim negotiate the to identify the brotherhood we Muslim If the compact. and tutional protesters street the to scenario that consequences. on its focuses against insurance autocrat provide the to constitution, order writes the in he writing when when case that, autocrat the plausible the clientele, autocrat’s to the of known perspective is where the From this constitution. that quo status particular the in suppose and form rbblt,i sbte oselect to better is it probability, eetaypito h oiotlpr ftecnrc uv n ewl hoeto choose will he and curve contract the of part horizontal the on point any select e oiotlpr ftecnrc uv ie where (i.e. curve contract the of part horizontal per curve. rpss h ilpooetepitwhere point the propose will she proposes, .. Case 3.1.1 the which outcomes the obtained. have on could had groups have two military—would other the or Mubarak it setter—be hr h otatcreitret with intersects curve contract the where ups ti nw htatrtransition after that known is it Suppose Reform Constitutional on Negotiate L and M 3.1 Proposition from Finally, 78 22 amendment. no with accept predecessors the which constitutions i.e. eetetxrt stesm swith as same the is rate tax the Here c 0 set between the curve of contract boundary the lower of the that suppose we Initially nifrnecurve indifference oiotlpr ftecnrc uv ie where (i.e. curve contract the of part horizontal oalwrtax lower a to as such point a at constitution Fig. in curve contract 0 ∗ e Sect. See sdmntae nteapni.I htcs,teatca at ochoose to wants autocrat the case, that In appendix. the in demonstrated is 1 w m z .If 1 osetelte on,spoeta h uortpcsa picks autocrat the that suppose point, latter the see To . 4 L eko htw a ou nsainr constitutions, stationary on focus can we that know we rpss h ilpooeahge a ttepoint the at tax higher a propose will she proposes, L z c hc loi ntebudr of boundary the on is also which and ∗ ntepitwhere point the in M c ∗ I ec,a ogas long as Hence, . httecnrc uv savria iefor line vertical a is curve contract the That . m onie with coincides M z hscmswt ihrtxrt.If rate. tax higher a with comes This . and l 0 t nescswt h otatcurve. contract the with intersects = L ) nta ae h uortmay autocrat the case, that In 0). eoit vrcntttoa re- constitutional over negotiate t = l 0 )bcueti ol imply would this because 1) I l nescswt h contract the with intersects 0 I nescstevria part vertical the intersects , L I nescswt h lower the with intersects M nescswt h up- the with intersects rpsswt positive with proposes .Mcaa n .Pech G. and Michalak K. utraiealower a realize must I n corresponds and I n the and M M Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 itosbetween tiations and eoit with negotiate .. Case 3.1.2 Proof positive strictly is constitution a constitution down a ing down handing prefers strictly autocrat ia osiuininvolves constitution timal rpsto 5 Proposition n osiuini otyi em fwat,teatca rfr o owiea write to not 6 prefers Proposition autocrat the wealth, of terms in costly constitution. is constitution a ing with that suppose Next osntviolate not does select when 1 Fig. Transition Political in Constitutions Stable etyo h ttsqocnttto.To constitution. quo status the of dently Proof zero. is autocrat the to constitution wealth a transition writing post affect of value not monetary does but policy only affects constitution au oteatca because autocrat the to value rps h eal outcome default the propose osiuini re ostsyhsplc rfrnewt da point write ideal to with like preference still policy would his he satisfy yet to zero order is in constitution constitution a a writing of value monetary the w L L M x piu constitution Optimum e icsinabove. discussion See e icsinabove. discussion See sepce to expected is > nteieltxrt of rate tax ideal the on ∈[ w x w M M M ,x vrreform over ntesai oe with model static the In ntesai oe with model static the In ≤ L L R ] w M spriiaincntan,ie h a aeis rate tax the i.e. constraint, participation ’s and scoea osbeto possible as close as a esta vrg fetv elhad hrfr,agrees therefore, and, wealth effective average than less has M ilrsl ntemxmmlvlo eitiuinwhich redistribution of level maximum the in result will t ∗ c t = t oec te.Ntrly loi h aeweeteop- the where case the in also Naturally, other. each to 0 = > ,wiigtecnttto fespstv monetary positive offers constitution the writing 0, .I htcs hc silsrtdi Fig. in illustrated is which case that In 1. 0. L L R . x and h eiorpial rfr wealth, prefers lexicographically who , and R h w eoitr ilnecessarily will negotiators two The . M M sbranr and bargainers as sbranr and bargainers as . h oeayvleo hand- of value monetary The t = w t w x M 0 . indepen- , R ec the Hence M > fwrit- If . 2 ≤ nego- , w w , the , 79 a

B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 and with negotiate when exist nld the include autocrat. the for optimal not is it but constitution h oeayvleo rtn osiuini titypstv,a h reversion the as positive, involves strictly constitution is a of constitution absence a the in writing outcome of value monetary The hr h a aei oetgvnthat given lowest is rate tax the there When constitution: quo nti eto easm that assume we section this In 3.2 to value monetary positive a guarantees 2 Fig. 80 equilibrium nti aeteei amn between harmony is there case this In When ucm o isl:If himself: for outcome .. Case 3.2.1 include “Pinochet” to the going of not was authors left. government the the successor of that parties any argue that the believed to from had possible emerged constitution therefore, which reform is, constitutional It of negotiations. element The one transition. was and negotiating participation admission transition concertación political their from Chilean moderate parties the left-wing the banned and in had right constitution up “Pinochet” the set of bargaining PN the the effectively with was This reform. tional utpropose must on ntecnrc uv with curve stitution contract the constraints, on in further point point Without ideal satisfied. her be choose to has constraint participation oeta if that Note ti ayt e nFig. in see to easy is It w M R osiuinwhich constitution A M ≤ and c w rpss h ed oofrtepoint the offer to needs she proposes, L ∗ t sepce to expected is w ocicd ihpitin point with coincide to c M = ∗ c M M L ∗ sasainr constitution. stationary a is vrreform over xs nta case, that In axis. 0 R spwrt noc ucmsi h oflc cnroi weak, is scenario conflict the in outcomes enforce to power ’s > tef n if And itself. eoit nCntttoa Reform Constitutional on Negotiate osnot does w R M 3 I rpssh ed ooffer to needs he proposes htapitsc as such point a that rpssagainst proposes hc snttepitwt h oettxrt u a but rate tax lowest the with point the not is which R rpss ewnst propose to wants he proposes, R L R and ysrtgclycosn h ttsqocon- quo status the choosing strategically By . and M I M M where and M utobtain must r rdce ongtaeo constitu- on negotiate to predicted are ilawy ge natxrt fzero. of rate tax a on agree always will R c z ∗ z t nterrdsrbtv ol.Yet goals. redistributive their on h a ooffer to has she , snta pia hiefrastatus a for choice optimal an not is swl.Thus, well. as sminimal, is t 0 > M 0. m h point the z and R .Mcaa n .Pech G. and Michalak K. c ∗ z a nueabetter a ensure can swl.Therefore, well. as sas stationary a also is L t z ≤ utobtain must gi because again t ∗ to M R would I ,soit may L l 0 ’s .

B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

690 689 688 687 686 685 684 683 682 681 680 679 678 677 676 675 674 673 672 671 670 669 668 667 666 665 664 663 662 661 EDITOR’S660 659 658 657 656 655 654 653 652 651 PROOF650 649 648 647 646 645 wealth, itswith where tiates case the consider Finally rvlo l he lyr.I htcs,ayslcinof selection any case, that In players. three all of proval Fig. hr h a aegt iia seFig. (see minimal gets rate tax the where . te Cases Other 3.3 Case 3.2.2 and with negotiate when 4 Fig. and with negotiate when 3 Fig. Transition Political in Constitutions Stable ostesm atr si h aeo eoitosbetween negotiations of case the in as pattern between same negotiations the lows of case the For r.Mroe,tecnttto a oeayvlet h uortbecause autocrat the to value monetary has constitution the Moreover, ary. Fig. in illustrated is aihn yLemma by vanishing stitution l ls a oeta vrg fetv wealth, effective average than more has class dle h aehlsi rpsli h osiuinlbrann aenesap- needs game bargaining constitutional the in proposal a if holds same The w w 4). M M M M pia constitution Optimal constitution Optimal ≤ > sepce to expected is to expected is c c ∗ w w ∗ saancoe ntepitin point the in chosen again is R w ycosn h on in point the choosing by R R nta ae ti tagtowr that straightforward is it case, that In . M vrreform over reform over ≤ w gi,i ses ocekta hscnttto sstation- is constitution this that check to easy is it Again, 4. 2 n,hence, and, M t ∗ a eo vrg fetv elhadnego- and wealth effective average below has I

B&W IN PRINT B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

736 735 734 733 732 731 730 729 728 727 726 725 724 723 722 721 720 719 718 717 716 715 714 713 712 711 710 709 708 707 EDITOR’S706 705 704 703 702 701 700 699 698 697 PROOF696 695 694 693 692 691 rpsto 7 Proposition results: our l he players three all nfvro eitiuini rdce ob dominant, constitution be a writing to not predicted and writing is between redistribution of favor where in cases the in an autocrat or the dominant for is value monetary positive has only constitution a autocrat. netvsfrwiigacnttto ol opeeyvns.Tesm ple to applies same Hence, The conflict. vanish. open completely where with case would than the constitution place a better writing no for in incentives them puts the which for value value reversion monetary if positive hand, creates other the constitution On a autocrat. writing cases those in Hence, either h uorto hte owieacnttto o eue owehrtedominant the whether to reduces now select constitution will a party write to whether of autocrat the eicr yntraiighspeerdplc outcome policy which preferred is cost his the it to realizing think weights not we zero by Therefore, non successfully attach incurs another time. of will he long for autocrat consequences a the post the for that his assume hand, felt to in be other reasonable might be the constitution On will a zero. he down the than handing implement that greater and probability write still the to is know when want day not time will may therefore, a autocrat and, at the demise constitution addition, his to In of adhered been date demise. to has regime’s precise likely actually the the is it before it that acceptable time practice, be In some to himself. for constitution constitution a the for condition by incurring a bound without be being where successors as his choice such for constitutional cost constitution of default any model the choose static can a autocrat introduced the has section previous The Choice Constitutional Intertemporal of Model A 4 in point preferred its the propose from will within gain constitution it reform to process, a stands constitutional selects who party the player dominant all the the to by if template Moreover, preferred a process. strictly as reform be acceptable would weakly sce- it be conflict and least the at players in constitution would constitution point a a ideal of such her conflict, preexistence realizing the from when If even her position prevent nario. a can such in parties party other one the put may rule Majority reform. constitutional proposing than possibility another with 82 on in point elzsdrn h iewe i ucsosdlvrapeerdplc outcome. policy preferred a constituency deliver his successors gains the his to when and time himself constitution the the during by realizes abide to has he which ocmlt u xoiin ups htoepryi ufiin ocrythrough carry to sufficient is party one that suppose exposition, our complete To M I fteei n oiatpryi h osiuinlrfr process reform constitutional the in party dominant one is there If speitdt edmnn n fulfills and dominant be to predicted is si h aewhere case the in as , M ihngtain between negotiations With t w L R sdmnn.Tefloigpooiinsummarizes proposition following The dominant. is speitdt edmnn,i offers it dominant, be to predicted is and M c ec,teatca eet i preferred his selects autocrat the Hence, . eoit ihec other. each with negotiate M c . sncsayt rvn ecn into descent prevent to necessary is and w R M rbetween or < ( w h uorti indifferent is autocrat the 0,x rwhen or R .Mcaa n .Pech G. and Michalak K. ) . I uigtetm for time the during If h usinfor question The . L L R and M ran or sdominant. is and R , M rwith or writing R party their R Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

782 781 780 779 778 777 776 775 774 773 772 771 770 769 768 767 766 765 764 763 762 761 760 759 758 757 756 755 754 753 EDITOR’S752 751 750 749 748 747 746 745 744 743 PROOF742 741 740 739 738 737 i emo fc.Ti cnrocmrsstecsswhere cases the during comprises binding scenario non This is office. which of constitution term a his writing consider only would autocrat the constitu- the writing to value monetary positive a is there claimed, tion. have we as if, 0 aeweetesaial pia osiuinde o nov hoigtepoint the choosing involve not does constitution those in optimal in least statically At the constitution. optimal where statically case stationary, the choose to want sarily assembly. succeeding the in dominant is redistribution in o htcnttto,tecs fcommitment of cost the constitution, that For the tion. writing to value Proof monetary positive a is there model. static where the cases in all constitution in down hand to great sufficiently 8 great: Proposition sufficiently strictly is autocrat future the the of which weight constitution the a if exist writing must prefers there positive, strictly is model fteatca ad onacnttto,h hoeteconstitution the choose he constitution, endogeneity. a such down of hands possibility autocrat the ignore the we If the constitution, autocrat’s and the states values of both choice the in assume time weights these of though autocrat, length the expected of the rate on discount depending that assume, We Transition Political in Constitutions Stable 23 n h eeto omteti titypstv,i.e. positive, strictly is commitment of benefit the and h uortcnraietesm a f swt ocnttto nteiiilperiod initial the in constitution no with off as pay off pay a same and the realize can autocrat the eas h ucsoswn ocos eomconstitution reform a choose to want successors the because fh osnthn onacnttto,hsttlpyofis off pay total his constitution, a down hand not does he If n h osritdrn i emi fc n raigsrne netvsfralow a for incentives stronger demise: creating his and after office regime in tax term his during constraint the ing an nn rbran with bargains or inant ie i h ihs oa eet ..h maximizes he i.e. benefit, total highest the him gives e h ro fLemma of proof the See I nalcssweeteei opstv oeayvlet rtn constitution a writing to value monetary positive no is there where cases all In yslcigaconstitution a selecting By ial,ee hncosn idn osiuin h uortmyntneces- not may autocrat the constitution, binding a choosing when even Finally, nalcss hr h oeayvlefo rtn osiuini h static the in constitution a writing from value monetary the where cases, all In M where at poe ordsrbto,teatca ae rd f ewe loosen- between off trade a faces autocrat the redistribution, to opposed party h uortcnawy hoet addw h ttclyotmlconstitu- optimal statically the down hand to choose always can autocrat The ( 1 − t δ) esmnml ..i h aewhere case the in i.e. minimal, gets u hc nrae ihtetm nofie a tefdpn nthe on depend itself may office, in time the with increases which , R ftewih ftefuture the of weight the If , Γ(c (Γ hr sabnigcnttto hc h uortsrcl prefers strictly autocrat the which constitution binding a is there V R  )) t x) (t, 3 M . V ≥ R agi n h aewhere case the and bargain u ( ∅) = R 0 c ( ntescn eid h atrrltosi follows relationship latter The period. second the in =  1 ∈ / − ( 1 I δ)u − hc sntbnigdrn i emi office, in term his during binding not is which δ)u R t x) (t, , R δ ( , 0,x nteatca’ betv ucinis function objective autocrat’s the in + R δu ) L + u R R u sepce ongtaewith negotiate to expected is δu (x Γ(t,x) R L (x R 0 ∗ ran or ,t . ∗ ,t w ∗ t x) (t, ) M ∗ − ) . ( M ≤ 1 − u − u w R ∈ at nfvrof favor in party R ( δ) I 0,x and Γ(c (Γ . 23 t x) (t, and R L )  ∈ δ sdom- is sfinite is / .Even which I))> 83 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

828 827 826 825 824 823 822 821 820 819 818 817 816 815 814 813 812 811 810 809 808 807 806 805 804 803 802 801 800 799 EDITOR’S798 797 796 795 794 793 792 791 790 789 PROOF788 787 786 785 784 783 h idecasudrasfcetysrn ecie hetb h ethv been left—have the by threat perceived strong sufficiently and a right under the class between middle 1—negotiations hypothesis the in writing constitution for ditions n ti egtdwith weighted is it and yaial pia osiuinlchoice constitutional optimal dynamically Proof nescinof intersection nahge au of value higher a in h uortcosstesaial pia constitution optimal statically the chooses autocrat the t rpsto 10 Proposition constitution: a writing of desirability the on insight .I natca xet hthsoncineewl aen nuneo succeed- a on influence no have will clientele own his that expects autocrat an If 2. succeeding a on influence have will clientele own his that expects autocrat an If 1. emerge. hypotheses two analysis our From Transition Political of Experiences Different to Application 5 9 Proposition 84 ortraie etrotoewt weight a with outcome better a realizes tocrat h mle h eet Finally, benefit. the smaller the as fwiigaconstitution, a writing of mle a aesc as such rate tax smaller uortgets autocrat ortt rt osiuingiven constitution a write to tocrat proposes rtclweight critical Proof constitution optimal statically the on in hrfr,teciia weight critical the Therefore, tion. h smaller the 0 t n h ttclyotmlconstitution optimal statically the and ntecs fteCiencntttoa rjc,i em luil httecon- the that plausible seems it project, constitutional Chilean the of case the In biul,frvr small very for Obviously, fh xet httemdl ls rfr o eitiuinpolicy. redistribution low a prefers constitution class a middle write the to that incentives expects strong he has if only object. he to assembly, able constitutional being ing left the without policy redistribution preferred opposing their parties impose al- to that constitution, able expects be a he will if write vanish to incentives incentives those has though generally he assembly, constitutional 0 → nalcssweeteei oeayvleo rtn h osiuin the constitution, the writing of value monetary a is there where cases all In Let 0. c ∗ p t o ups htteatca hoe osiuinwt slightly a with constitution a chooses autocrat the that suppose Now . 0 M t , ∗ m h esvlewiigacnttto has constitution a writing value less the δ ntecs where case the In etepoaiiythat probability the be ihawih of weight a with ∗ h rae h oe of power the greater The z uhta if that such n h otatcurve. contract the and V R ( n for and z 1 u nFig. in − R Γ(c)) (Γ p δ h uortmyntwn ohn onaconstitu- a down hand to want not may autocrat the , M δ δ u )δ al eo htweight, that below falls → p 1 R L However, . δ nta case, that In . M Γ(c)) (Γ ∗ − δ 1, eoitswith negotiates . δ ∗ nybcmsrlvn ftedsac between distance the if relevant becomes only .If u h olwn rpsto eeaie this generalizes proposition following The . z R M eut nahge au of value higher a in results (t L t c c ≤ R  0 ∗ rpssi h agiigpoes If process. bargaining the in proposes >t sidpnetof independent is rpss h proposes she proposes, ) u ntetasto cnroand, scenario transition the in ssfcetylret nueteau- the induce to large sufficiently is sicesn in increasing is t ( R z 1 ∗ (t ec h ucm swrefor worse is outcome the hence , δ w → t V  | R hr xssa exists there (t . 0,  R ,x c c c z sbenefit ’s ∗ ∗ ∗ h au- the ,  ) ,sothe results hence and nthe in t R 0 L , , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

874 873 872 871 870 869 868 867 866 865 864 863 862 861 860 859 858 857 856 855 854 853 852 851 850 849 848 847 846 845 EDITOR’S844 843 842 841 840 839 838 837 836 835 PROOF834 833 832 831 830 829 h osiuin rts oeet uha h n e ycpe iesi 1983 of in miners authors copper the by of led mind one the Sater the and on as Collier ( been (see such Vial have movements and must Protest Montes threat was constitution. by left-wing project the suggested a constitutional as of the communism, possibility on against the bulwark embarking a for build regime to Pinochet the of rationale eutn ocoeacnttto hc id i w cin.Fnly ti unclear is it Finally, been actions. have own would he his cases binds these which In process constitution rule. his transition a of any chose horizon to in time player reluctant outcome long strong a default predict a high may be a he to or enforce military to the power is predict significant expectation the may with when autocrat even low The be bargaining: may multiparty writing might why constitution is which reasons of are down value bargaining there monetary handed Moreover, multiparty the process. be bargaining will where the constitution in case amended a significantly the predict be we In that place, down. predict take we handed to dominant, expected be be to will expected In constitution is reform. brotherhood no constitutional Muslim negotiating the where in case players the different the constitution. of a writing strength to bargaining all at in value role no dominant is a there plays process, brotherhood the reform Muslim on constitutional the that If, the that bargaining. expects is give multiparty he expectation would of the that form hand, This provided the other redistribution. takes constitution the to process a fits reform opposed write constitutional best class to the Egypt incentives middle that for strong a view right autocrat with the the the country supports of further a candidate class” presidential of “middle the case this of wooing of voters the the coefficient and secular Gini Chile’s Egypt’s often than Moreover, lower the wealth. opportunities of existing is hand, improvement of on redistribution other emphasis the an the than compati- with more rather set On are mind model. which middle-class aspirations a descrip- our voiced with the protests ble of street fits context the best started the which probably groups in goals leftist welfare of social tion its with Pech brotherhood, and Muslim (Michalak depth greater in wealth case class Chilean middle the between in relationship transition the stable represented. it discuss and which sug- we class process paper middle accompanying the transition of an the mind In the in on showed not was concertación redistribution mod- the that The gests of transition. parties constitutional the successful Chile’s which of eration case the in been has Schmitter wealth and O’Donnell (see class working and 1986 middle privileged, of members oiia rcs.A ttre u,asbtnilpr ftelf loojce oac- to the objected outside also it reform. left keep political the towards to of vehicle utmost part a its substantial as did a constitution it the out, cept because turned formal regime it a from old As in plausible process. the not was political of would constitution view left successor of the a point That over autocrat. the negotiations the in by involved met be be sense to believed were or met Transition Political in Constitutions Stable 24 o isnigve e ai (1987 Tapia see view dissenting a For hrfr,tepeito formdlciial eed ntepiraotthe about prior the on depends critically model our of prediction the Therefore, The corresponds. transition Egyptian the case which to see to difficult more is It class middle of influence the which is remains which question interesting An uthv eiddtejnao uhalneigthreat. lingering a such of junta the reminded have must ) 1996 n h ocle poet”mvmn hc involved which movement “protesta” so-called the and ) ). 24 foeacpsta one that accepts one If 2005 2012 85 ), ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

920 919 918 917 916 915 914 913 912 911 910 909 908 907 906 905 904 903 902 901 900 899 898 897 896 895 894 893 892 891 EDITOR’S890 889 888 887 886 885 884 883 882 881 PROOF880 879 878 877 876 875 ae yatcai osiuinlcoc hntetastosdsusdi hspa- this in manipu- discussed be transitions to the open than less choice were per. constitutional players where Africa autocratic main South template. as by in constitutional emerged lated or Poland ANC inherited in the any as and for such Solidarnocz transitions play that the to suggests in point role compromise latter a reach The there Fur- to different necessary is process. are it process there transition find transition that which the such interests during serves heterogeneous diverging sufficiently with it clientele groups is redistribution, autocrat’s society oppose the if to for only wealthy thermore, proxy sufficiently natural is are a class there transi- as middle that constitutional suggests the transi- successful This If during facilitate outcome. which if tion. preferred conditions benefit its political such impose and to no economic able is is is There rel- party class classes. clientele one middle the other tion of the of position representatives as insurance bargaining to provides long the ative improves it as and successors, bargaining redistribution its transition to transi- by opposed from the accepted excluded during is expected being constitution is against a bargaining clien- autocrat’s is If multiparty the where process: discussion to cases tion preceding benefits all the offers almost and compact in model constitutional tele the a down from handing emerges that which lesson main The Discussion Further 6 constitutional a within rights property rights those property securing emerging about the compact. think members in can owners—often share they their property secure before emerging paper. nomenclature—to for this former important in the discussed more of cases be the will from different it quite Indeed, are stakes Leff observe the and we order, Munck Poland property (see power of incoming case and the old not in the 1997 least was stable between at pact of This through because failure countries. transition trying the communist a of on former lack light the of a for case shed completely the may in This transition protect. to im- constitutional arrangements the wants constitutional is autocrat of though, element the highlights, an which as model that rights our predicts property What established model observe. of to our portance rare hand, be other would the cases On such redistribution. supporting class middle of case amendments. the major to with correspond under but would established transition observation interpretation order constitutional This property stable second place. the in a that kept observation suggests was the constitution model with the our line However, in out acceptability. more turned is to autocracy which the bar during a a practice have be constitutional to compliant would to fact expected non this that they of say that interpretation those to orthodox extent by An be the reform. even over to revoked bargaining benefit future formally in to was share stood constitution have the would that who fact players the interpret to how 86 elc bevtosweecntttoa ucsinwstidi h aeo a of face the in tried was succession constitutional where observations lack We .Hwvr ntecs hr e osiuinhst eiiiea emerging an legitimize to has constitution new a where case the in However, ). .Mcaa n .Pech G. and Michalak K. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

966 965 964 963 962 961 960 959 958 957 956 955 954 953 952 951 950 949 948 947 946 945 944 943 942 941 940 939 938 937 EDITOR’S936 935 934 933 932 931 930 929 928 927 PROOF926 925 924 923 922 921 lution nyb civdtruhtelna a system: tax linear the through achieved be for only off pay joint the maximize

nti pedxw hwta ntecs where case the in that show we appendix this In Appendix Transition Political in Constitutions Stable 26 25 ls ons e ..BrnadDemir( Diermeier and Baron e.g. see points, bliss n h a aei eemnda h eiulsatisfying residual the as determined is rate tax the and aepolicy same hsproblem, this yconstruction, By tandotmzto problem optimization strained h otatcrei etclfr0 for vertical is curve contract the optimum, which of unit a half costs hsplc realization policy this exceeds ny odmntaeti on,nt that note point, this demonstrate To only. a fec ru eaiet vrg vial noeas income available average to relative group each of gap a rate tax rae than greater into converted is ∂v If than greater is ratio the that show can One L M ∂x o 0 For h pia rpslcnb nepee sfollows: as interpreted be can proposal optimal The At Writing max

(x = M  ) x w x M / L M =−  t L swat ymr hnaeaewat,weetelte scluae okn at looking calculated is latter the where wealth, average than more by wealth ’s 1,

(x and n at and 0 ti ayt hwthat show to easy is It .   ) R ) M ≤ = t ∂v  + ,tefis re odtosfra neirslto of solution interior an for conditions order first the 0, = ( L ∂x L 1 (x P t v − w  and x ) M L x .Frcneine edfieteincome the define we convenience, For 1.

→ , nrae e nto tlt eraeby decrease utility of unit per increases 2001 ( M L x μ(x −1) M ∗ L 2|x M s.t. )  a ewitnas written be can R = − −1if scs fcompensating of cost ’s ie httasesbetween transfers that given ). =  M − v (t − v L → L L L

 w>

(x) ,x x (x sicm yoeunit. one by income ’s L M L L w )  |  L

L

) + ) , →−∞ +w 0. 2 M L hnpooigto proposing when , given x + ( L  M 1 stert twhich at rate the is t nydpnso h ratio the on depends only ∗ − and <(w w w . x M L c t)w e R ycniut of continuity By . −(w +(w = steplc hc would which policy the is . M M 25 L M (t M − −w ftertois ratio the If agi and bargain −w

− ∗ w ,x x M L L t L ) L w ) afwybtenthe between way half −w x ∗ −w ) = ,i.e.if )  ta admissible an at ≥ R ∈[ R L ovstecon- the solves μ w . u M M M and x stert at rate the is M t M eet the selects and ∗ M − sincome ’s ,x w ,x ,it −1/2, M μ swealth ’s L M w M ∗ L ,aso- .Inan L .Call and ) ≶ > .

.As

can (4) (3) M L 87 w 0, L . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 1012 1011 1010 1009 1008 1007 1006 1005 1004 1003 1002 1001 1000 999 998 997 996 995 994 993 992 991 990 989 988 987 986 985 984 983 EDITOR’S982 981 980 979 978 977 976 975 974 973 PROOF972 971 970 969 968 967 ai 18)Sbee s el lgldd el ittr.Cue132 June):22Ð29 113(22 Cauce dictatura. la de ‘legalidad’ la 32:1Ð20 de Sci uso Polit el J Sobre Br (1987) constitution. J the Tapia Newhaven of Press, Evolution University Yale (2002) rhetoric. N Newhaven of Schofield Press, strategy University Yale The Sci manipulation. (1996) Polit political W Am of Riker art capital. The on (1986) state W the Riker of dependence rational Structural Economics, (eds) (1988) R M Gekker Wallerstein T, A, Boylan Przeworski In: order. legal the and Rationality (2009) G Pech about conclusions Tentative rule. authoritarian from Transitions (1986) stabil- PC of Schmitter agents G, O’Donnell or turmoil of Reflections democracies. new for Constitutions (1997) P govern- Ordeshook institutions constitution. of evolution the the of commitment: sustainability and Constitutions and (1989) law B Weingast of DC, rule North The (2012) G Pech state. KJB, constitutional Neumärker the N, of Naqvi foundations and problem credibility autocrat’s Eastern The and (2008) America R South Myerson democratization: and transition of Modes (1997) CS Leff GL, Munck roots authoritarian the Chile: in process building constitution The (2005) constitution T 1980 Vial the Esteban, autocracy: J BehavMontes an in Econ choice Games constitutional Strategic institutions. (2012) political G Pech of M, Michalak reform and stability Dynamic (2007) R Lagunoff Press, University Cambridge democracy. social European 8733 of paper transformation The working (1996) NBER 23:449Ð464 H Welf conflict? Kitschelt Choice or Soc Constitution vote? (2002) one HJ person Grossman one 2012 Why June (2004) 15 Times, H Financial Gersbach dissolved. parliament orders court 6:317Ð335 2012 Growth Egypt May Econ 20 J Times, development. Financial Cam- economic freedom. and Press, Egypt’s consensus University class Cambridge Middle (2001) 1808Ð1994. W Chile Easterly of history The (1996) WF and Sater questions S, some Collier answer amendments constitutional draft Egypt’s (2011) M Dunnes N, Brown policy. foreign change, for rulebook constitution’s Egyptian The (2011) N constitution. 1980 Brown the and Junta, the Pinochet, dictatorship. and Constitutionalism (2002) represen- R proportional Barros in parliaments and governments, heteroge- Elections, a (2001) in D weights Diermeier voting DP, assigning Baron nations: of weights constitu- the On and (2006) rules MO majority Jackson self-stable S, choose: Barbera to how Choosing (2004) MO Jackson S, Barbera References 88 e 82:12Ð29 Rev economy political of frontiers series Routledge philosophy. normative and choice Baltimore Press, University Hopkins John democracies. uncertain 90:55Ð72 Choice Public 49:803Ð832 ity? Hist Econ J England. century seventeenth in choice public ing Freiburg 02-2012, paper working CEN 102:125Ð139 Rev Sci Polit Am 29:343Ð362 Polit Comp perspective. comparative in Europe Stock- Assistance, Electoral and Democracy for Institute holm International democracy. stable a of Chile in 67:569Ð583 Cambridge bridge 2011 March 4 Peace, International for Endowment Carnegie others. raise 2011 February 3 cessed foreignpolicy.com/posts/2011/02/03/the_egyptian_constitution_s_rulebook_for_change Cambridge Press, University Cambridge 932Ð967 Econ J Q systems. tation 114:317Ð339 Econ Polit J union. neous 119:1011Ð1048 Econ J Q tions. .Mcaa n .Pech G. and Michalak K. http://mideast. .Ac- Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1058 1057 1056 1055 1054 1053 1052 1051 1050 1049 1048 1047 1046 1045 1044 1043 1042 1041 1040 1039 1038 1037 1036 1035 1034 1033 1032 1031 1030 1029 EDITOR’S1028 1027 1026 1025 1024 1023 1022 1021 1020 1019 PROOF1018 1017 1016 1015 1014 1013 enatB(97 h oiia onain fdmcayadterl flw mPltSiRev Sci Polit Am law. of rule the and democracy of foundations political The (1997) B Weingast Prince- Press, University Princeton work. institutions political how players: Veto (2002) G Tsebelis Transition Political in Constitutions Stable 91:245Ð263 ton 89 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 ( h ags utinbn,Keiasat n13 httigrdtesqec fbank of sequence the of triggered failure that the 1931 was in it Kreditanstalt, 1929, bank, of Austrian collapse largest market the the with started have may Depression” ( (2010 Reich 1 with deal to how fear about oil. the concerns for is and appetite “weirding” background US or the the warming In global uncertainties markets. of resulting effects stock The the account global debt. of in current public swings US US violent of the level induced the to have facilitated contributed as has well save, as currency to deficit, manipulated propensity China’s its and distri- rapidly, with wealth growing and coupled are income has Brazil on and and and India advantage possibilities China, employment comparative bution. on of effect structure dramatic rev- a ratio international internet had this the the increase with changed Entitlements, coupled likelihood, %. has economy all 100 global olution in the about to will, to to transformation generation” risen debt The since boom public more. has even “baby US but aging the % the 2001, 33 to to to % due 1993 49 From from debt. fell of ratio GDP quandary quandaries. economic economic political the severe is of number First a faces currently States United The Introduction 1 Schofield Norman and Schnidman Evan Politics Electoral US in Leadership and Gridlock of Quandaries .Shimn( Schnidman E. .Shfil ta.(eds.), al. et Schofield N. USA 63130-4899, 1027, MO e-mail: Box Louis, Campus St. Hall, Drive, Seigle Brookings Louis, One St. in University Washington Center, Weidenbaum Schofield N. USA 02141, e-mail: MA Cambridge, #2, Street Vandine 11 DOI ,Rcmn( Rachman 2011), ,Sih(2012 Smith 2012), o aybost aehv drse hs unais u ecnmninGlrih( Galbraith mention can we but quandaries, these addressed have name to books many Too aypol o erta efc eeiino h 90s hl h “Great the While 1930’s. the of repetition a face we that fear now people Many 10.1007/978-3-642-35239-3_5 schofi[email protected] [email protected] ,Mlnvc( Milanovic ), B ) 2011 ). ,Scs(2011 Sachs ), dacsi oiia Economy Political in Advances 1 2010 ,Lessig (2011), Harvey (2011), Mandelbaum and Friedman ), pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , ,Na (2012 Noah ( 2011), Buchanan (2011), Steyn ), , ,Stiglitz ), 2008 91 ), Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 adofi infiatices nra SGPa eldvlpet nnew in developments well as GDP US real in increase significant a in off paid h 90s osvl lofcdopsto rmsuhr eort n a to had ( of and Kennedy presentiment Democrats As a bill. southern In anti-lynching from legislation. an just on Deal opposition 1938, way New faced 25, give Senate the also June the of Roosevelt of in last 1960’s, (FLSA) seats the Act 6 the was and Standards election, House Labor the the Fair to in The prior seats election. hous- 81 1938 in of the for gain later in Republican subsidies a him to and led with budget, railroads met the own- and of public system” recommending nationalization social 1938 utilities, in the ing. Roosevelt the of to of wrote “trustee ership later a ideas Keynes earlier 1934. as Keynes’s May him of pre- praise some moves with policy to these consistent which Although were of Reserve accounts. (Keynes they some Federal savings book, the reforms, Keynes’s in allowed rates dated banking Q the interest Regulation introduced regulate Corporation speculation. 1933, and to control Insurance 16, States to Deposit June designed United Federal were In the to the works. in attempt established public (FDIC) to had through created Act partly all GlassÐSteagall unemployment, were the with and Indus- Administration) Administration) deal (National Adjustment Works NIRA (Public (Agricultural 27, PWA the Emergency May AAA the on (Federal Authority), Administration), Court Valley FERA Recovery Supreme (Tennessee the trial the TVA Corps), by the Conservation Ti- unconstitutional (Civilian Authority), (although deemed dol- CCC Act was the Recovery The Act Industrial took 1935). the National essentially of the that and I amendment standard), tle an deal gold provid- (to with the and Act off also lar Adjustment but Prohibition, Agricultural production, (finishing the over Act, revenue), Act with Banking government Revenue Emergency needed the much Beer-Wine including ing and Deal, New Economy the the of beginnings enhanced. the and preserved through be could economy world the of efficiency and liberty the ue13,ta e oteeomu otato fwrdtaeaddeeigo the of deepening and trade world in of earlier contraction enormous Act Tariff crisis. the Smoot-Hawley to the led with that coupled 1930, US, June the and Europe in failures 92 2 e lotewr yFse ( Fisher by work the also See oee,tevrosefrs ro o13,t euaeteeooyeventually economy the regulate to 1938, to prior efforts, white various the the Congress. However, alienate the deadlock irretrievably indefinitely would and repair system] beyond establishment racial southern South’s the on assault h eeeeooi onuni 97 asdprl yatmt oblnethe balance to attempts by partly caused 1937, in downturn economic severe The pushed he 1933, 16, June until 3 March on inauguration Roosevelt’s in- From the and consume, to propensity the another, invest one to to ducement adjusting of task the clear made 380) (1936: Keynes As 2 osvl ugdadtesxwe lbse ofimd[htafrontal a [that confirmed filibuster week six the and judged Roosevelt ttecs fteelreeto h ucin fgvrmn t involve] [to government of functions the of enlargement the of cost the at 1930a , b, 1933 .Kye isl a rte otePeieti 1933 in President the to written had himself Keynes ). ndb eaina udmna as ftedepression. the of cause fundamental a as deflation debt on 1933) 1999 4)notes, 343) : .ShimnadN Schofield N. and Schnidman E. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 eadadspl fitrainlliquidity. international of supply and demand providing be Plan. by might Marshall that the recovery strategy through the European a liquidity After triggering pursued terms. US internationalism,” current the in “hegemonic however, billion 1945, termed $500 April of in order Roosevelt the of of death assets with Union, clearing a today. bated corpora- large of ability ( Livingston the input. labor to reducing due when even been output increase have to tions may increases productivity These fWrdWrI e omjrwrso oiia cnm yVnMss( Mises Von by economy political of works major to Reces- ( led Schumpeter Great II current War the World of of consequences the cor- of is out argument increasing way this of a (If pattern sion.) suggests growth. earlier it led the then reversed consumer rect, had reanimated Deal and New inequality the income that argument good a nteOC onre eas ftes-ald“oiia uiescce n the and cycle” business “political deficit. so-called payments the of US pace because continuing gathering countries dis- was inflation OECD was that system the suggested Woods Smith- in report Bretton the McCracken war the by post 1977, but In the rights mantled. 1971, drawing December special of of agreement creation sonian the through stability maintain h SadU n cnmcgot ea.Fo 92t 98 n h election % the 3 and about 1988, at to grew 1982 GDP From from US these began. Bush, growth stripped For G. economic eventually 1979. of and was in UK Inflation Thatcher and problem. Margaret US the the under was UK government the election leaders, the in two and Party 1980 Conservative November in imple- the Reagan changes of dramatic of the election to presidential background the $1.80 the after forms about mented 1970’s from the of rise chaos price The a 1974. and OPEC of formation ehooisadlreicessi atrpoutvt (Field productivity factor in increases large and technologies Politics Electoral US in Leadership and Gridlock of Quandaries 5 4 3 h euaoyrgm hthdbe u npaedrn h e Deal. New the during place in put been had that regime regulatory the hs events. these R ea) i ec R oa,adToa .Bie,J.(,Virginia). (R, Jr. Act, Bliley, J. Modernization Thomas Gramm and Phil Services Iowa), co-sponsors (R, Financial its Leach Jim after Texas), the named (R, Act, by GrammÐLeachÐBliley the 1999, called 12, were also companies) November financial other on owning from repealed company holding bank a at billion. (prohibiting grew $376 GDP/capita to grew US deficit 2000, trade to the while 1992 % became in 3.5 US election about the Clinton’s 1989, capitalization From in and speed. democratization Union gathered with Soviet coupled the Globalization, of hegemon. world collapse the the With billion. $115 about n17 h Shdataesrlso 22blinbtb 97ti a ecto 2 billion. $27 of deficit a was this 1977 by ( but Rosner billion $2.2 and of Morgenson surplus See trade a had US the 1970 In (1960 Triffin See y16,hwvr thdbcm biu htteewsa maac nthe in imbalance an was there that obvious become had it however, 1960, By for arguing was Keynes II, War World of close the by collapse, another of Fearful h eidfo h olpeo eorc nErp nte13’ oteend the to 1930’s the in Europe in democracy of collapse the from period The uigCitnsscn diitain h rvsoso h ls taalAct Steagall Glass the of provisions the administration, second Clinton’s During 1942 ). ,VnHyk(1944 Hayek Von ), 2011 o icsin swl steacutb lno ( Clinton by account the as well as discussion, for ) 4 h omdt omta olwdldt the to led followed that boom commodity The n opr( Popper and ) / nu,bttetaedfii loge,to grew, also deficit trade the but annum, 3 fot eemd n16Ð98to 1964Ð1968 in made were Efforts 1945 / arlt 1.5i January in $11.65 to barrel htaesilbigde- being still are that ) 2003 2011 Allen ; 5 hsended This provides ) 2011 1994 1940 )of 93 ). ), Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 Shueodntwrhfl rmaot$0tilo n20 oaot$0trillion $50 trillion. about $15 to about 2007 of in debt trillion public $70 a about with from 2009. US fell in worth the net left household has US crash market The burst. h osn akt h rsmto httemre ol euaeisl proved feared. itself Keynes regulate that instability could of (1986 market Minsky on the as based that just derivatives presumption unfounded, particularly The assets, market. risky housing in the speculating profits, of for % competed 65 2009. about in from GDP rising of US, % the 100 in to debt 1995 household in in GDP increase significant a to led oeg xhnersre f$. rlinadhlsaot$ rlini STreasury US in trillion bonds. $1 agency has about currently holds government China and result, and trillion a $3.2 As China US of 2011. with reserves the of deficit exchange months 2006 six trade foreign first In its the unstable. while for billion billion, became $130 $750 market reached reached global invest- deficit and the payments trade that of global meant balance of also expansion inequality has the global to increased. but that contributed has ment, has meant economies developed US has all the shock in in labor inequality Deregulation income this that that but surprise decreased no has is It 1989. since 0 ,Prua 0 ,Blim9 ,Gray8 ,Fac 7% rti 0% pi 6%. 56 Spain %, 80 Britain %, 87 France %, 83 Germany %, 96 Belgium %, 106 Portugal %, 109 %. 35 about only ree tl,Iead ree pi n Portugal. and Spain Greece, Ireland, Italy, Greece, o pto European the up together for of put package to rescue forced (EFSF) was Fund EU Stability The Financial default. of risk at themselves found are homes their than more owing themselves worth. find homeowners million 15 about and 94 8 7 6 ik odn vr$ rlini ik oeeg debt. sovereign risky in trillion $2 over holding risk, h uoenlvl fttlpbi etGPcretyae ree16% tl 2 ,Ireland %, 121 Italy %, 166 Greece are: currently debt/GDP public total of levels European The (2011 al. et Alpert (2011 al. et Alpert emnBohr i l o akutyo etme 5 08 n h bubble the and 2008, 15, September on bankruptcy for file did Brothers Lehman (Keynes when speculation. serious of is whirlpool position a the 158Ð159) on But bubble enterprise. the harm of becomes no stream enterprise do steady may a Speculators on predomi- bubbles . . the as increase. however, of does, enterprise. risk speculation the over of improves, predominates markets nance is speculation investment it of that life, organization case whole the the their As ac- over always the assets means for of no enterprise yield by prospective term the the forecasting and of market, tivity the of psychology the forecasting nadrgltdwrd n nacneto oa aad nnilinstitutions financial hazard, moral of context a in and world, deregulated a In lblzto a en ht2blinpol aejie h ol’ ao force labor world’s the joined have people billion 2 that meant has Globalization h otgo pedt uoe hr h etoehn en htmn states many that meant overhang debt the where Europe, to spread contagion The fImyb loe oaporaetetr pclto o h ciiyof activity the for speculation term the appropriate to allowed be may I If 7 e vni h erfo ue21 o21 os ausfl y$ trillion, $1 by fell values house 2011 to 2010 June from year the in Even 4 ilo o upr oEr ebrsae ndfcly including difficulty, in states member Euro to support for billion 440 ). oeta hn ae bu 0%o D,ivssaot1 n consumes and % 15 about invests GDP, of % 50 about saves China that note ) 6 a rud hsiblnecnla otekind the to lead can imbalance This argued. has ) aa lohsaot$0 ilo.Cepmoney Cheap billion. $800 about has also Japan e 8 5 ilo,al oisebonds issue to able billion, 750 uoenbnswr loat also were banks European .ShimnadN Schofield N. and Schnidman E. 1936 : Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 n 08gnrleetos nteeeto fNvme 0 01 h conser- 2004 the the 2011, in 20, Spain November of of Minister election Prime the In as elections. terms general for 2008 elected and was a (PSOE), form Party to Commissioner, European former a Italy’s Monti, deficit. government. Mario the reduce asked to measures then of president number a approved Parliament after signed Francois candidate socialist the by 6 of May the round on second by Hollande. the also imposed in election, Sarkozy measures, Presidential Nicolas austerity vote. of French the defeat the the the of the of of was % % importance rejection 18 greater 14 a Of with with as EU. seats seats seen both 108 took were 41 2012, Democracy reverses Lucas won 6 New These only 10, right, May party, center November of the left on while election center and vote, the the government, PASOK, In unity parties, minister. a major prime for interim way became make Papademos to down step An government. to her 2012. of March, fall 10 on the held to be leading will 2011, election October, early a 11Ð12 of on head parliament the as the Radi 2010 government. July 8 coalition from center-right Slovakia four-party of Minister Prime successor, and Party, his Democratic until government 2011. caretaker October a 3 of on appointed elec- head was parliamentary as Thorning-Schmidt, 2011 office Helle September in the remained in He Denmark tion. of for Minister Prime % as 12 position and Socialists crisis. promised for the and Party, % with Popular deal 28 the to to with measures coalition under vote austerity a Democrats, further the lead Social will of right Coelho Party. % center Popular the 39 the 5, took June Coelho, of election Passos billion the Pedro $116 In of 2011. bailout a 3, obtained May government caretaker on Coalition, the and 23, (National March on Christian parties signed and six Party of People’s coalition Swedish dif- Parlia- a Greens, and Finnish Alliance, leading Democrats). long the Left 2011, After by Democrats, time. under June Minister Social first Party, Prime 22 True the elected Coalition on the for was while National ment party held, Katainen center-right largest had negotiations, The they the ficult that seats. became seats Katainen, 34 51 Jyrki gained the of party 16 Finns lost Kiviniemi, Mari ister, (Prime Taoiseach became Gael, March. took Fine 9 only party, on Fail, opposition Ireland Fianna the of party, crisis. Minister) of governing held debt Kenny the election 2007, an Enda the in and and with seats 2011, 25, 78 deal 1, From February to 26. on February order dissolved on was in Parliament Union, Irish European the First, the by imposed measures unaiso rdokadLaesi nU lcoa Politics Electoral US in Leadership and Gridlock of Quandaries oéLi orge aaeo h edro h pns oils Workers’ Socialist Spanish the of leader the Zapatero, Rodríguez Luis José re- had Berlusconi, Silvio Minister, Prime Italian the 2011 12, November On agreed Papandreou, George Minister, Prime Greek the 2011, 5, November On Radi Iveta his lost Venstre, party, liberal center-right the of leader Rasmussen, L¿kke Lars re- Party, Socialist the of Socrates, Jose Portugal, of Minister Prime the Then Min- Prime by led Party, Center the Finland, in election general 2011 the In austerity the to opposition of because fallen have governments EU of number A oáwstelae fteSoa eortcadCrsinUnion— Christian and Democratic Slovak the of leader the was cová ˇ oáls oeo ofiec in confidence of vote a lost cová ˇ 95 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 n sarsl,gvrmn eurscaiinareet nedtefra model formal the Indeed agreement. representation, coalition proportional requires (Schofield government on result, based a systems as electoral and The have strife. polities partisan and EU volatility indebted political increased in resulted has inequality wealth unrest market continued created has which ambivalence. Spain ris- austerity political and the fiscal and Italy is the in further threat on matters in default emphasis Muddling debt defeats Merkel. the ing election Angela changed the Chancellor However, have German EU. may of the May strategy of sus- in deepening be France a can and to the states Greece eurozone member agreed and twenty-seven UK) the Europe the the of whether both but twenty-six debate of (all 2011, 9, much systems December is political on the though there tained, in Europe as In well States. bond as United destabilize market to the continues in about certainty crisis to debt limited euro been the has why intervention been is has EU which Reserve the GDP, markets. Federal contrast by EU the GDP In of GDP. of by % % intervention of 2 13 the % of US 11 order the Republican order the In of of the intervention England. the of an of in by defeat government Bank problem Conservative the the debt The its and 2008. with November 2010 dealt in UK May US in the Kingdom in administration United the in ernment Di Elio leader, Party reach Socialist to the under parties government six Rupo. a of form the coalition to 2011, agreement a tentative 1, forced a December debt on sovereign because Belgium’s Eventually 2010, Wallonia. of June and downgrading of Flanders election between the mea- after conflicts austerity form in of to further able seats been out had 350 carry government coalition the to mandate of a 186 and won vote Rajoy % Mariano sures. 44 by a led with parliament, (PP) party People’s vative 96 ek oepanti etiua ednyi h 08ad21 lcincce in cycles election 2010 and 2008 the in States. tendency paper United this centrifugal the of this remainder highlighted The explain has politics. trillion to partisan $14 US seeks over in of convergence debt of federal lack extreme with the deal to need the to from resulting able legislative been by have characterized activists now space, is policy elections. the recent of in a dimensionality exert role two important poli- the increasingly centrist of an adopt not Because played do has candidates money that Instead, argue cies. and election, presidential to 2008 similar the candidates, the political of on system effect electoral plurality electoral the convergent or post” strong the a past generates “first US that suggests here the- use political the we contrast This ory In center. instability. the governmental and from fragmentation far sustains positions polarization adopt generally will smallparties tation, nteUie tts hsucranyculdwt eae frsn noeand income rising of decades with coupled uncertainty this States, United the In un- of deal great a created has crisis economic global the of web complex The gov- Labor the of fall the were crisis debt the of effects political first the fact In No %, 96 of ratio GDP to debt a with difficulty, in itself found Belgium Even oninmda oe result voter median Downsian centrifugal 2007 neligti ae ugssta,udrpootoa represen- proportional under that, suggests paper this underlying ) oc nteplc oiin ftepris sarsl Spolitics US result a As parties. the of positions policy the on force nedteices npria rancor partisan in increase the Indeed gridlock. (Downs 1957 .W ics eeteet since events recent discuss We ). .ShimnadN Schofield N. and Schnidman E. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 n16 gis otenDmca poiini oges h rdoki ogesi 1964 in Congress in gridlock The Congress. in opposition Democrat Southern against 1964 in clrainet htivle h oildmnino ii ihsa ela the as well as rights like. civil the and of taxes dimension involving dimension social the economic the usual to involved components that realignment” of ical number a are There economy. the quandary: interrelationship current and the involves polity and constitutional the essentially between is that quandary a faces but acceptance, competition Hamiltonian and a capitalism. in liberty industrial believed of in Roosevelt regulation, while belief etc.) Jeffersonian trade a free espoused (through Wilson (Gould that resulted was had Wilson that inequality com- economic of power increasing 2008 the the to and opposed Progressives, interests as mercial election presidential 1912 the tested con ftelgco sn eore eeae ytrf rtcint nueinfrastructural induce to Northern protection the of tariff development industrial by the generated facilitating detailed canals, resources a states. and gives using railways Lind as economy. of such political logic improvements USA the the of of development account the of visions Hamiltonian and Spltc,a icse eo,sget htti oflc sa motn sever. as important as is in conflict dimension this social that the suggests below, of discussed importance op- as The politics, and “Moderate” US Burke. equality compromising and the and Hamilton and reason of hegemony, Enlightenment of hierarchical and support monarchy in to Boling- Paine, posed by and espoused Jefferson Enlightenment “Radical” Condorcet, a broke, between conflict belief, a of witnessed ( systems 1700 philosophical Israel rival by of context suggested general as more the in conflict this put ogtadwnteeeto f10 opeev h Epr fLiberty”. of “Empire the He preserve corrupt. to to 1800 capital of allow election also the would of won 1791 opening and in the fought that America believed of Jefferson Bank 1720s. noble First the cor- the Hamilton’s in and ascendency that commercialization Whig crass the Bolingbroke, the of Viscount by ruption destroyed John, been had Jefferson St. US. England Henry the of in constitution of occur could arguments feared Jefferson the that followed defect structural a to on due sions choices their make interests. simply would than voters rather that judgements suggesting of choice”, basis fit the a of Madi- “probability above. [1787] described a (Madison quandaries X the Federalist address in logic to son’s able be should government Wise Argument the of Logic The 2.1 Politics Activist 2 Politics Electoral US in Leadership and Gridlock of Quandaries 11 10 9 rmik(1990 Kramnick e ao(2012 Caro ( See Crick by books recent the also See i h lcino ..Jhsni 94wstebgnig fanw“polit- new a of beginnings the was 1964 in Johnson US L.B. the of 1932, election in The and (i) 1912 in 1800, in as just that is paper this of thesis The con- had Wilson Woodrow and Roosevelt Teddy both century 20th early the In nti ae eageta h Splt scretyual omk iedeci- wise make to unable currently is polity US the that argue we paper this In .Ide hc ( Chace Indeed ). o icsino o B a bet oc hog h ii ihslegislation rights civil the through force to able was LBJ how of discussion a for ) , 1992 o h otnigcnitbtenteJeffersonian the between conflict continuing the for (2012) Lind also See ). 2004 ,wohspitdotta h oenpro since period modern the that out pointed has who 2012), ugssta h ifrnebtenRoeetand Roosevelt between difference the that suggests ) 1995 ,Hthn ( Hitchens ), 1999 2007 a htaRpbi ol exhibit could Republic a that was ) ,Dwis(2011 Dawkins ), 11 euefco nlssto analysis factor use We ). 9 ecan We 10 97 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 eitc nsae httn ovt eorti otatt hs h edt vote to tend who those to contrast charac- in voter Democrat of vote pattern to the tend Republican. in that differences states dramatic in to teristics led Re- has the conservative This particular, and activists. economic In conservative social axes. of coalition both a combine from benefited that have publicans constructed be to coalitions winning thousands and buys media large activism. for pay electoral are to of time money man-hours volunteer of of sums and/or activists large cash provide that of to means able amounts now small This for providing elections. raised to deregulated could constrained essence that previous money in the and on campaign restrictions many political removed 2010, 21, January on ecpino h hrce riso addts uhpretoscnb influenced be can perceptions Such candidates. of traits al. character et the of (Clarke perception elections US of studies ut h eulcn logie ohHue fCnrs n1994. in Congress of re- Houses a both as South gained election the also Presidential of 1980 Republicans gain the The won the sult. Reagan been Indeed, has Party. Republican realignment the this by of term become consequence the has principal axis use second The we The axis. and axis. time, social over the important characterize that more beliefs the of many to o h aktoe oenetadable htwlaeepniuesol ede- be should expenditure preference welfare ( Campbell a that and belief care, Putnam a As health and creased. economic government government, over The of market equality. the size and for values services, traditional government gays, involves Amer- for axis African right to civil attitudes involves abortion, axis icans, social The dimensions. two these construct 98 14 13 current the to similarity some has paper. budget this 12 the in over discussed conflict Congress, in partisan gridlock with combined issue this over rwhbfr h usigo h alo a nrae h blt ftoewith those of ability the these term increased We has elections. balloon influence to the assets of economic bursting of the consequence before a is growth years 30 last the in occurred has realignment. this that regime regulatory the in hi rfrne,i oprsnt h itiuino lcoa rfrne,have preferences, electoral of that distribution mean we the extreme. activists this more to economic By become comparison preferences. to activist in accrued economic preferences, have of their that radicalization benefits a to significant power. these led the activist have of increased time, power also same has enhanced parties the the the At between to race contributed arms has an Indeed, inequality activists. wealth and income ing o icsino h hne nielg n lcoa upr o h eulcnpryi this in party Republican the for support electoral and ideology in changes the period. of discussion a for ilr( Miller ,Gla (2009 Gelman (2005), Saunders and ( Abramowitz Abramowitz and (2005) Saunders and Abramowitz See (2003 Schofield and Miller (2003), al. et Schofield See ii h xsec ftopltcldmnin a en hti spsil for possible is it that meant has dimensions political two of existence The (iii) decision, Court Supreme The i)Mdl feetosaetpclybsdo oe rfrne ln.Recent alone. preferences voter on based typically are elections of Models (iv) i)Atog h oilai a eoeeetrlymr motn,economic important, more electorally become has axis social the Although (ii) o icsino hsrainet e loMcltwieadWodig ( Wooldridge and Micklethwaite also See realignment. this of discussion a for 2007) 14 13 iiesUie .FdrlEeto Commission , Election Federal v. United Citizens 2010 2009 aeson eiist fvtr srelated is voters of religiosity shown, have ) ). , 2011 oilactivists social 2010 ,Shfil (2007 Schofield ), aeepaie h electoral the emphasized have ) ). cnmcactivists economic .ShimnadN Schofield N. and Schnidman E. o ciit nthis on activists for n coedand Schofield and ) 12 h change The Increas- . 2004) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 oaiaini h lcoaeadLya ta.( al. et Layman and electorate the in polarization ao ate a eoe rmoeanother. one from removed far parties major ogesoe h attreyasi re oguetevldt fteaoear- above and the of Obama validity between the contest gauge the to order of in details years the three gument. discuss last the then over and Congress 2008, to 2000 siaeo h itiuino oe rfrne o rfre oiin)i h US the in positions) preferred (or be Fig. 2004. preferences can example, of voter space election For of presidential policy surveys. distribution this voter the of the of from that nature estimate evidence elections The offer dimensional. recent first two for we least inferred this, at do is To space influence. activist political of terms in larization ttehato h oiia unayi edt eosdrtecntttoa sepa- constitutional the reconsider to US. need the a in is political powers quandary of a political ration of the context of heart the the in at occurs that seeking rent elite facto of de kind the a dilemma. the that of prisoners’ of suggests result nature propose a the we is of power that because influence model undue The exert machinery. to democratic labor, as such groups, interest h lcoa eni tcatcmodels. stochastic electoral in the mean (Downs to converge dimension electoral positions will one the parties adopt in that candidates voting inferred or deterministic has parties under and that win, have to assumption fo- competition order the to two-party in on tended of based has Models competition been judgements. electoral than typically on rather literature preferences on formal the cus noted, have we As Elections Modeling 2.2 in inefficiencies inducing thus ( rents Olson by economic work collect Earlier to economy. political order the in power facto de candidates. suggest between We race spend. arms candidates the that for money logic the by the provides therefore turn this in and media, the by Politics Electoral US in Leadership and Gridlock of Quandaries 18 17 16 15 08Peieta elections. Presidential 2008 2004 ( Duggan by work tlatipyta oiia hiela oamdrt rcnrs ucm.O the On outcome. centrist or moderate become has a politics to that evidence lead extensive choice is there political contrary, that imply least at hsfiuei ae nfco nlsso h mrcnNtoa lcinSuy(NS for (ANES) Study Election National American the of analysis factor on based is figure This ( al. et Fiorina by works the See ( Hinich and Enelow by work earlier the See (2011). Vermeule and Posner nters ftepprw osdrmdl fU rsdnileetosfor elections Presidential US of models consider we paper the of rest the In nti ae ecnie hoyo oiia hiewihacut o po- for accounts which choice political of theory a consider we paper this In v)Teiflec fmnyadteplrzto ihnCnrs ugssthat suggests Congress within polarization the and money of influence The (vi) v cmguadRbno ( Robinson and Acemoglu (v) ntenx eto egv oedtiso h atrmdlta eue o h 04and 2004 the for used we that model factor the on details more give we section next the In . 2006 ,adPtye l (2009). al. et Patty and ), 15 ,FoiaadArm ( Abrams and Fiorina 2005), 18 h rtlf ih ieso ersnspreferences represents dimension right first-left The aedsusdteaiiyo ltst exert to elites of ability the discussed have 2008) ,EisnadRmr ( 1990 Romero and Erikson 1989), 16 nplrzto mn activists. among polarization on 2010) hs oeso oiia convergence political of models These 17 1982 lofcsdo h blt of ability the on focused also ) 2009 1957 n cat ta.( al. et McCarty and ) Hotelling ; polarized n oerecent more and ) 1 ihtetwo the with 1929 rsnsan presents median, 2006 ,o to or ), )on 99 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 iejistepeerdpit fvtr h,acrigt h oi oe,would gives model model, logit logit The the half. one to of according probability who, equal voters with cleavage candidates of This the election. points choose Presidential preferred 2004 the the joins of model line logit Figure simple a interests. from obtained opposed of coalitions be Figure as interpreted be can and taxes and a expenditure government towards attitudes) (or 1 Fig. 100 20 19 hudb edl vial r oae otenrho this of north the to located are available readily be this should of right the on located are decreased be should programs welfare oilplc,priual ii ihs swl svtroiin bu brinetc. abortion about opinions voter as well as rights, civil particularly policy, social with hseuto lotpse hog h point the through passes almost equation This eortcniae er,hdasih datg vrteRpbia candidate, Republican the over advantage slight a had Kerry, candidate, Democrat economic h oilai sdfie ota oeswospotcvlrgt o asadbleeta abortion that believe and gays for rights civil support who on voters spending that that so and defined is market axis free social the The in believe who voters that so defined is axis economic The eas h oiia pc stodmninl ate nteUie ttsmust States United the in parties two-dimensional, is space political the Because (a,b,c) lcoa itiuinadcniaepstosi h ntdSae n2004 in States United the in positions candidate and distribution Electoral 1 lososetmtso h oiin ftetopeieta candidates. presidential two the of positions the of estimates shows also axis. = ( 19 −0.2, h eodnrhsuhor north-south second The 1.34, ρ dem −0.93) == y = 1 Setting . + exp 1.44x exp (a (a + − 1 + ρ bx 0.21 lososa shows also social dem bx i y ( + i 0, = -axis. + cy ieso eet tiue on attitudes reflects dimension −0.21) 2 1 cy i ) eoti h equation the obtain we i .ShimnadN Schofield N. and Schnidman E. ) x atsncevg line cleavage partisan -axis. n ugssta the that suggests and (1) (2) 20 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 rt n eeal r oae nteuprlf udat rmtoewotn to tend who those from quadrant, quadrant. left right upper lower the the in in Republican, located vote are generally and crat, oilcnevtvs(oe ih) h oilpyhlgclltrtr ensthe defines literature psychological social and The right) (upper right). Libertarians (lower left), Liberal conservatives (lower morality: left social to religious “quadrants” the four left), essence (upper in secularists are there that finds that psychology uh This Bush. positions) activist mean the for larger are bars (error 2004 2 Fig. Politics Electoral US in Leadership and Gridlock of Quandaries 21 activists. and estimates. ciit edt elctdi h pe eto h oiysae h enactivist mean The space. policy the of party left Democrat upper the the all are in opposition, estimates located In be space. policy to the tend of activists are right activists lower Republican the all in almost located groups, liberal socially and conservative socially h gr hw h tnaderrbr o hs siae,wt agrerrbr o activist for bars error larger with estimates, these for bars error standard the shows figure The Figure h w iesoaiyo h oiia pc scrooae ywr nsocial in work by corroborated is space political the of dimensionality two The oprsno enpria n ciitpstosfrDmca n eulcnvtr in voters Republican and Democrat for positions activist and partisan mean of Comparison 2 atsncevg line cleavage partisan hw h enpstoso eortcadRpbia at voters Party Republican and Democratic of positions mean the shows 21 Figure 2 ⎣ ⎡ ugssta huhteRpbia at otisboth contains party Republican the though that suggests Act : y x 04R Dem p Re 2004 eaae epnet h edt oeDemo- vote to tend who respondents separates −0.48 0.55 +0.75 −0.49 ⎦ ⎤ . 101 (3) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 ihapria laaeln ie by given line cleavage partisan a with ethn oani em fa mhsso utc hl h ih addmi is domain hand right the while justice authority. on of emphasis terms an in of defined terms in domain hand left 3 Fig. 102 22 rlicm a us hsaayi ssgetv facreainbtentetodmninltrait dimensional two the space. between policy correlation dimensional Fed- a two of to suggestive the opposition is and with analysis space associated This “conscien- weakly cuts. and tax and income abortion “openness” eral legalized is that with “openness” shows agreement Moreover, respectively. with he ideology associated liberal/conservative (CES) with Study correlated Election are tiousness” Congressional “openness”, 2006 as such choice. the traits political personality on Using of stability” effects “emotional the and explore “agreeableness” “extraversion”, to Mondak “conscientiousness”, analysis clusters. regression moral uses four the (2010) define al. These et “purity”. and “authority” “loyalty”, “fairness”, sion”, g ie rm20 o20 o20 a oae lgty nacokiedirection, clockwise a important. in increasingly slightly, become rotated had had axis 2008 social to cleav- the that 2004 the suggesting that to Notice 2000 McCain. from over lines Obama, age candidate, Democrat the of advantage the election, this For equation election. 2008 the for hc asstruhtepoint the through passes which s atraayi nfiemrltat,icuig“compas- including traits, moral five on analysis factor use (2009) al. et Graham precisely, More naayi o h 00cnetbtenGr n uhgvsasmlrresult similar a gives Bush and Gore between contest 2000 the for analysis An Figures itiuino oe da onsadcniaepsto n2008 in position candidate and points ideal voter of Distribution 3 and 4 hwtedsrbto fvtradatvs rfre positions preferred activist and voter of distribution the show 22 y ( y 0, = = −0.4) 1.87x 0.82x hscevg iesget h greater the suggests line cleavage This . − − atsncevg line cleavage partisan 0.34 0.4, .ShimnadN Schofield N. and Schnidman E. sgvnb the by given is (4) (5) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 ciit n h addtsi 2008. in candidates the and activists i.4 Fig. Politics Electoral US in Leadership and Gridlock of Quandaries al 1 Table hl Tables while poeUieslhat care health Universal Oppose services Government Less Question a opne Equally Companies Tax Equality about more Worry government Less spending Welfare Decrease Government to Market Prefer Government Bigger Oppose ii ih o gays for right Civil Immigration Decrease Abortion Support Eigenvalue Liberal v Conservative Amer Afr. Support access Gun values Traditional Disagree Table itiuino ciitielpit n addt oiin n2008 in positions candidate and points ideal activist of Distribution atrlaig o cnmcadsca policy social and economic for loadings Factor 1 ie h w iesoa atrmdlbsdo h NS20 Survey, 2008 ANES the on based model factor dimensional two the gives 2 and 3 ieterslso h siae fma oiin fvoters, of positions mean of estimates the of results the give 0 cnmcpolicy Economic 0. 0. 0. 0. 0. 0 0. 0 1. 0. 0 0 . . . . . 53 51 14 65 24 56 50 28 12 93 30 14 36 0. 0. 0. 0. 0 oilpolicy Social 0. 0. 0. 0. 1. 0. 0. . 12 22 37 14 55 10 25 53 60 83 60 45 103 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 oemdrt nteeooi xs n oerdclo h oilaxis. social the on radical more and axis, economic of become the distribution average, on average on moderate have the activists more to Party relative Democrat axes contrast, In both preferences. in electoral preferences their in radical more iin aebcm oeeteebten20 n 08 n a ocheck to way One 2008. ( and compare 2004 to between is extreme inference this more become have sitions in bandfo h atrmdl o h w elections. two the for models factor the from obtained tions ( al 3 Table 2 Table 104 23 siae oiino oilplc 0 policy social on position Estimated policy economic on position Estimated Question Democrats Activists eulcn 1. Republicans Democrats Non-activists eulcn 0. Republicans 0.76, eal fte20 atrmdli ie ntenx section. next the in given is model factor 2008 the of Details h orcinsget htRpbia ciit ae naeae eoemuch become average, on have, activists Republican that suggests correction The oprsno ( of comparison A h enatvs siae are estimates activist mean The ,(0.9, 0.76), ⎣ ⎡ bm n caneetrlyprevdpositions perceived electorally McCain and Obama election presidential 2008 the for data Descriptive Act y/sd x/sd : 04R Dem p Re 2004 −0. −0. ense 5%CIMa ..9 C.I % 95 s.e. Mean C.I % 95 s.e. Mean policy Economic 0.91) 00 20 10 41 70 17 20 72 −0.63 n ( and 3) 0.72 o 04ad20 epciey hscreto gives correction this respectively, 2008 and 2004 for ⎣ ⎡ . . . . Act 09 13 03 06 5 n ( and 3) +0.99 −0.64 : y x rvdssm vdneta ciitaeaepo- average activist that evidence some provides ) [−0. [ [−0. [ 08R Dem p Re 2008 1. 0. 66, 60, 38, 24, ⎦ ⎤ 1. 0. ntrso h lcoa tnaddevia- standard electoral the of terms in 5) 16]− −0. 84]− −0. , −0.82 1.41 02] 11] ⎣ ⎡ Act y/sd x/sd oilpolicy Social +1.14 −0.20 : 1. 0. 0. 0. 08R Dem p Re 2008 40 14 20 82 60 36 56 −0 Obama . . 75 20. 22 ⎦ ⎤ .ShimnadN Schofield N. and Schnidman E. 0 . −0.91 . . . . 05 11 09 04 1.56 23 [−0. [ [−0. [ 0. 0. Using 92, 29, +1.26 −0.22 65 99, 1. 0. , 37] − 44] −0. 0. (σ ⎦ ⎤ 46] 65] x . ,σ −0. McCain y n 788 219 449 37 59 ) (7) (6) 80 40 = Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

690 689 688 687 686 685 684 683 682 681 680 679 678 677 676 675 674 673 672 671 670 669 668 667 666 665 664 663 662 661 EDITOR’S660 659 658 657 656 655 654 653 652 651 PROOF650 649 648 647 646 645 nte“riigedge” “trailing the on f21 perdt efil rdcl ntelwrrgtqarn ftepolitical the of quadrant right lower the election in recent “radical” the fairly in be positions to Republican appeared Congressional 2010 in- the of The that and positions. Party, fact mean Republican the the activist in the indeed conservatives as social Party” much “Tea as of dominance not creasing but towards somewhat quadrants, shifted opposed therefore two have the parties two the for positions voter Average ecpin bu h optneadcaatrtat ftepltclcandidates. political electoral the in of shifts traits the character from and but competence the preferences, about electoral perceptions of distribution the the in in rotation changes clockwise a as interpreted be can This space. shift. political parties two the of tions a nacdteiflec fatvs groups. activist of influence the enhanced has candi- chosen their their from to receive available solve make they to must benefit date. resources activists the what the The estimating deciding modeling and activists. by contributions by supporting problem candidates optimization their that own to with is their available make model pattern resources can underlying shifting the they the the of contracts endogenize essence of to The consequence attempts candidates. the command. the it are candidates for perceptions support the activist in that of changes resources these the turn, by In influenced are perceptions These the of transformation Politics Electoral US in Leadership and Gridlock of Quandaries 26 25 24 n20 o$. ilo n20 o$. ilo n2008. in billion billion $5.3 $3.5 Pol- to about Responsive 2004 for from in Center went billion The $4.6 spending, terms. to candidate nominal 2000 including in in spending, 2004 election in that billion estimates 1 over itics to 1976 in dollars million optimistic. intelligent, honest, strong, knowledgeable, caring, moral, as such but conservative socially populist, quadrant). hand, left lower other the the (in on Democrats or leftist quadrant) economically right upper the (in Republicans pc,cue oepoietRpbiast osdracag fpryallegiance party of change a consider to to Republicans prominent some caused space, ned err ta.( al. et Herrera Indeed, traits character candidate to perceptions electoral links that model empirical an present we Below conservative economically but liberal socially cosmopolitan, be hand, one the on would, These the hspeoeo per ob udmna seto Spltc:a activists as politics: US of aspect fundamental a be to appears phenomenon This efrigtesm aclto o o-ciit o h ate efind: we parties the for non-activists for calculation same the Performing eageta h udmna hne nvtrcoc eutntol from only not result choice voter in changes fundamental the that argue We nrcn er,teiprac fatvs otiuin a nrae,adthis and increased, has contributions activist of importance the years, recent In Democrats. ⎣ ⎡ ⎣ ⎡ 04R Dem p Re 2004 y/sd x/sd 04R Dem p Re 2004 y x hfsi h ciitcaiin o h w ate hscuea cause thus parties two the for coalitions activist the in Shifts bev htsedn yprisi eea apin etfo 58 from went campaigns federal in parties by spending that observe 2008) −0.28 −0.37 24 atsncevg line. cleavage partisan 0.30 0.40 ftecevg iecag at leine hnteposi- the then allegiance, party change line cleavage the of +0.37 −0.33 −0.43 +0.49 ⎦ ⎤ ⎦ ⎤ , , ⎣ ⎡ ⎣ ⎡ 08R Dem p Re 2008 y/sd x/sd 08R Dem p Re 2008 26 y x h miia n omlmodels formal and empirical The −0.56 +0.72 −0.62 0.80 +0.36 −0.17 +0.40 −0.19 ⎦ ⎤ ⎦ ⎤ , . 105 (8) (9) 25 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

736 735 734 733 732 731 730 729 728 727 726 725 724 723 722 721 720 719 718 717 716 715 714 713 712 711 710 709 708 707 EDITOR’S706 705 704 703 702 701 700 699 698 697 PROOF696 695 694 693 692 691 h iblt ftecntttoa aac ewe xctv n eiltr nthe in legislature and executive between balance States. about constitutional United comments and the brief Presidency some of the makes viability between section the last conflict Sect. The ensuing by Congress. the in followed and groups election Republican 2010 2008 of election the midterm of the effect the the discuss to preferred we voter shocks of or distribution downturn the in economic change by occurs a points. line, caused inducing that economy, cleavage quandary, polarization political partisan “radial” political global the a of Secondly, of times allegiance. rotation at party slow switch a activists by as induced realignment ferential” the of polarization of degree the in change positions little candidate electorate. be candidates in may polarization resources, there and of though positions, even offer radical increases, de- more the to more of move become to Because they forced candidates. are and chosen victories, their election of to manding crucial become resources have the provide that elec- aware the they increasingly of become outcome activists important, the more As becomes economics. tion and politics between mechanism feedback a distribution wealth Pierson and and income (Hacker decade in Skocpol last inequality the over in occurred increase has the that for account well may nl edt oeRpbia hl ihsae edt oeDmca (Abramowitz Democrat vote to tend states rich while 2010 Republican vote to tend ingly ntesadr pta oe,ol candidate only model, spatial standard the In Model the of Outline An 3 “dis- the for polar- accounts so polarization activist become (Gelman of politics has in model politics center” This electoral appearing States. why United reason the a in provide ized here discuss we that 106 at ri ieyt eal ogvr eli h uue(Penn future 27 the in the well in govern effectively to govern able to be able to degree itself likely shown overall is have the or to perceived reflects past, is valence party general, the In which party’s model. to a the models, of empirical significance statistical In important. equally are valence electorate the by candidates ( Stokes as ense,Ws igna oiin,Mnaa ot aoia etcy lbm n North and Alabama Kentucky, strongholds. Republican Carolina, are South these All Montana, Carolina. Louisiana, Virginia, West Tennessee, h eet21 esssae httepoetsaewsMsispi olwdb Arkansas, by followed Mississippi, was state poorest the that stated census 2011 recent The ntenx eto epeeta uln ftemdlta eue nSect. In use. we that model the of outline an present we section next the In “circum- continuing the First, work. at forces two see we run long the Over to due resources these over candidates between race arms an is there Essentially and society in wealthy the benefited have to appears polarization this Moreover, ). 27 suulyasmdt eidpneto h at’ oiin n dst the to adds and position, party’s the of independent be to assumed usually is 2007 1963 ). , 1992 a mhszd h o-oiyeautos or evaluations, non-policy the emphasized, has ) 2009 n h aao htpo ttsseem- states poor that paradox the and ) positions atrt oes However, voters. to matter .ShimnadN Schofield N. and Schnidman E. 2006 2009 4 , hr ediscuss we where 2010 ). iro and Pierson ; valences ,of 3 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

782 781 780 779 778 777 776 775 774 773 772 771 770 769 768 767 766 765 764 763 762 761 760 759 758 757 756 755 754 753 EDITOR’S752 751 750 749 748 747 746 745 744 743 PROOF742 741 740 739 738 737 hsnpry n hs eore r eedn ntepryposition. party the on dependent are resources these and party, chosen etoso addt character candidate of ceptions hsi o rca assumption. crucial a not is this ae nmlioilcniinllgtmtos(N) nti oe,ec polit- each model, this In (MNL). candidate, methods ical logit conditional multinomial on based the of other or one towards bias or candidates. valence including models voting probabilistic or h at sdenoted is party the lence. 28 Politics Electoral US in Leadership and Gridlock of Quandaries 33 32 31 30 29 n rteohro h candidates. the of other the or one LNE. a be must NE any Thus share. vote increase to as so deviate to incentive ( ( oiin.Temdli ae nteasmto htteeaevrosadditional various as to are referred there is that date first assumption The the valence. candidate on of actual based kinds and is positions model equilibrium The of positions. estimates the between difference the center. electoral the to close positions adopt eta addtssol oecoet h lcoa rgn oee,teempirical the However, Figs. origin. in electoral resulting the work to and close Enelow move by should work candidates earlier dential the corroborated have ( paper Hinich this in mentioned are that oiin ihtepoet htn addt aeasaluiaea oeadyet and move unilateral small share). a vote make (or candidate candidate utility of no vector increase that a simply property is the LNE equilibrium” with stochas- A Nash model. positions same empirical the local MNL using the “a model It as formal of assumptions model. parallel tic existence a the for for of maximization conditions vote intercepts under the the (LNE) obtain from to obtained model, spatial were possible pure valence was a on of based estimates was it the since Downsian, where be to considered be can model fetn i vrl aec.I h miia oe ecnas siaetoad- two estimate thus also electorate, call can we the we which model before valence empirical of image aspects the his ditional In enhance valence. overall to his resources affecting these uses then didate oa origin. toral n cevyadPty(2006). Patty and McKelvey and (2001) Helpman and Grossman 2005), 1983 oidmgahcvlnerfr otepoest fmmeso aiu rust ihyregard highly to groups various of members of propensity the to refers valence Sociodemographic (2011 al. et Clarke See that points. assumed preferred voter is of it distribution convenience, the For unilateral of a mean the has is candidate origin no electoral that The so positions candidate of vector a is (NE) equilibrium Nash A ( Adams The model electoral stochastic formal and empirical an developed has work Recent deterministic considers that developed has literature new a decade last the Over hspprofr oegnrlmdlo lcin ht esget consfor accounts suggest, we that, elections of model general more a offers paper This j .I hsmdl ciit rvd rca eore ftm n oe otheir to money and time of resources crucial provide activists model, this In ). 33 dpsaplc position policy a adopts envtrtheorem voter mean n hw,b iuaino h ai fteMLmdl,ta presi- that models, MNL the of basis the on simulation by shown, and 1989) ,Aaoe n afe ( Palfrey and Aragones (2001), al. et Ansolabehere 2001), 28 30 miia nlsso h 04ad20 Speieta elections presidential US 2008 and 2004 the of analyses Empirical j a hrceie yan by characterized was , μ n adr ta.( al. et Sanders and ) j 1Ð4 (z j ) traits mlctyw dp oe rgnlydet Aldrich to due originally model a adopt we Implicitly . losget htpeieta addtsd o nfact in not do candidates presidential that suggests also set htalcniae hudcneg oteelec- the to converge should candidates all that asserts 29 z . j μ nteplc space, policy the in , j (z j ) sol eedn on dependent only is 2011 ri valence trait nrni reoeosvalence exogenous or intrinsic o miia nlssuigtevtr’per- voters’ the using analyses empirical for ) hnpry rcandi- or party, When valence. activist X 32 hnthe then , and z j n o on not and , ,BnsadDuggan and Banks 2002), oidmgahcva- sociodemographic ciitvalence activist 31 z k , ahcan- Each , k λ = j .This j ,but 107 of Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

828 827 826 825 824 823 822 821 820 819 818 817 816 815 814 813 812 811 810 809 808 807 806 805 804 803 802 801 800 799 EDITOR’S798 797 796 795 794 793 792 791 790 789 PROOF788 787 786 785 784 783 (θ nietyestimate indirectly oeso h lcoa epne 1 sapr pta,()i pta oe with with model model spatial full a a is is (4) (2) while socio-demographics spatial, with pure model a spatial is a is (1) (3) response: traits, electoral the of Table models class. and income education, ethnicity, Tablegender, in shown We as perceptions “honest”. these and of “strong” analysis “knowledgeable”, factor “caring”, a performed “moral”, of terms in dates, Here hrceitc ftevtro oecoc.Smlryteterms the Similarly choice. vote on voter the of characteristics addt rissoe 2008 scores traits candidate al 4 Table 108 nuneo oe hieo h oe’ ecpin ftecaatrtat ftecan- the of traits character the term of The perceptions voter’s didates. the of choice voter on influence iedsrbto steTp xrm au itiuin eoe by denoted distribution, value extreme 1 Type the is distribution tive vector candidate to contributions activist by generated valence endi em famti nue rmteEciennorm, Euclidean the from induced metric a of terms in defined the called constant, nrni reoeosvalence exogenous or intrinsic j NS20 logv oi-eorpi hrceitc frsodnsb the by respondents of characteristics socio-demographic gave also 2008 ANES candi- the of traits character the of perceptions individual gave 2008 ANES The easm oe tlt sgvnb h equation the by given is utility voter assume We η u i ij ∗ ) u atrlaig for loadings Factor (x ij = r niiulseicsaasgvn h nuneo sociodemographic of influence the giving scalars specific individual are (x i ( ,z i 1 ,z j ,..., ) j steosral opnn fuiiy h osatterm, constant The utility. of component observable the is ) = = μ μ j pta parameter spatial j λ u j (z (z ,..., j ij ∗ j + (x j ) ) ,is i μ ymdln h election. the modeling by ,z Eigenvalue Optimistic McCain canIntelligent McCain Honest McCain Strong McCain Knowledgeable McCain Cares McCain Moral McCain bm Optimistic Obama bm Intelligent Obama bm Honest Obama bm Strong Obama bm nwegal 0. Knowledgeable Obama bm Caring Obama bm Moral Obama Question j p (z j j ) ) satvs upr ucin esgetta ecan we that suggest We function. support activist ’s fparty of j stesohsi ro,woemliait cumula- multivariate whose error, stochastic the is + ) +  j (θ . j iigteiprac fplc difference policy of importance the giving , j h function The . η i ) + (α j τ 5 i ) −0. −0. −0. −0. −0. −0. bm risMCi traits McCain traits Obama hw h euto h logit the of result the shows − μ 3. 0 0 0 0 0. 0 0 j .ShimnadN Schofield N. and Schnidman E...... j 07 07 11 55 03 10 02 17 09 61 68 69 61 71 72 .Theterm β (z  j x ) i stecmoetof component the is − (α z · j 4. j  β τ 2 Ψ i ,on + sapositive a is −0. −0. −0. ) −0. −0. .Theterms 3. 0. 0. 0. 0. 0. 0. 0. 0. 0. oe the model  λ j 00 57 68 00 63 70 65 63 67 08 09 13 07 18 01 j X ,isthe .The Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

874 873 872 871 870 869 868 867 866 865 864 863 862 861 860 859 858 857 856 855 854 853 852 851 850 849 848 847 846 845 EDITOR’S844 843 842 841 840 839 838 837 836 835 PROOF834 833 832 831 830 829 yia oe.B hoigaplc oiint aiieatvs upr,teparty the support, activist maximize to position policy a choosing By voter. typical a * 5 Table Politics Electoral US in Leadership and Gridlock of Quandaries mzn qiirafrtemdladcmaeti otepstoso h candidates. party the As of tion. positions the to this compare and model the for equilibria imizing Table Using traits. and socio-demographics otepryi otatt h support the to contrast in party the to l oesbcm oelkl ospotparty support to likely more become voters all Education Observations South Hispanic o ieiod(L Ð299 BIC AIC (LL) likelihood log fia American African cantraits McCain canvalence McCain Variable okn Class Working Income edr(F) Gender Age traits Obama Spatial aeieObama Baseline prob h rbe o ahpryi htatvssaelkl ob oeeteeta the than extreme more be to likely are activists that is party each for problem The ntetertclmdljs rpsd ciitvlnei fetdb at posi- party by affected is valence activist proposed, just model theoretical the In < β 0. pta oi oesfrUA2008 USA for models logit Spatial 05 ** λ j prob satvs support, activist ’s < 0. 611 601 788 1 pta 2 p ris()S.&Dm 4 Full (4) Dem. & Sp. (3) traits & Sp. (2) Spatial (1) 1.)(10.1) (14.1) (7.6) −0. 0.85 01 84 *** *** *** prob < μ 0. j μ Ð243 513 494 1.30 (8.3) (6.8) −1. (7.6) 0.78 −1. a (z 001 k (z j 02 08 ) *** *** k nrae u oicesdcontributions increased to due increases , *** *** ) 5 eevdb party by received Mdl4 ecnetmt oemax- vote estimate can we 4) (Model j vrparty over Ð250 567 521 0.15 (12.3) (1.5) 0.36 (1.34) Ð0.55 −4. (2.8) 0.86 −2. (2.25) −0. (1.5) 0.03 (2.5) (3.78) (1.26) 0.29 (1.0) Ð0.01 16 60 54 k *** * . ** * *** k hn(ntemodel) the (in then , (10.3) (0.07) (0.51) 438 Ð207 (2.59) (0.50) (3.66) (3.08) Ð0.02 Ð0.23 494 (0.26) (6.44) (7.15) (3.4) 0.01 −1. 1.36 0.83 −3. −0. 0.22 −3. 0.44 (1.0) Ð0.01 16 58 70 79 *** *** *** 109 *** *** ** *** Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

920 919 918 917 916 915 914 913 912 911 910 909 908 907 906 905 904 903 902 901 900 899 898 897 896 895 894 893 892 891 EDITOR’S890 889 888 887 886 885 884 883 882 881 PROOF880 879 878 877 876 875 hna h ieo h lcinteefc fti upr sicroae nothe into incorporated is support this valences. of trait effect and socio-demographic the exogenous, various election and the the of Grossman estimates of empirical in time outlined the as at contracts Then feasible of set ( a Helpman to committed is of space. candidate policy each the for of balance gradients quadrants activist the opposite aggregate into define point the we candidates that Below two each suggests the groups. between that support contracts argue activist pre-election and of the condition, family analyze the to and used be candidate then can locus balance the determine ie w itntmtvsfritrs groups: interest for motives distinct two sider the supporters. by activist influenced their turn from in raise can are candidates judgements the These that candidates. judgements resources electoral the of of terms quality in the also about but policies, intended over preferences voters’ fatvs upr o candidate for support activist of oe ycnieigafml fptnilactivists, potential of family a considering by model gradients. activist of terms position in the influence. change activist thus of logic and the support, provides activist change and This electoral adopted. the balance candidate to for position motive electoral the to ieadmny ti esnbet ups htteatvs ucinwl exhibit equilibrium. will the Nash function then a concave, exhibit activist sufficiently will the model are that maximizing functions vote activist suppose these of to order) When terms reasonable (first returns. in is a decreasing denominated it as money, is this support and give activist time results because marginal Moreover, Theoretical “optimal condition. share. the balance vote determine therefore maximize must to party condition” The voters. centrist lose will 110 34 hr each where candidate activist that h aefralvtr.Wt utpeatvss the activists, multiple With voters. all for same the iet eit oa oices oeshare. vote increase to as so deviate to tive ae to cated didate aheulbimi etro addt oiin ota ocniaehsauiaea incen- unilateral a has candidate no that so positions candidate of vector a is equilibrium Nash A osdrnwtestainweeteecnrcshv enare,adeach and agreed, been have contracts these where situation the now Consider nteatvs oe h term the model activist the In pronouncements. policy politicians’ an the with influence those to [while] candidates, preferred of prospects ( Helpman and Grossman the of terms in only not candidates evaluate voters that model the to intrinsic is It agisbtenteatvssspotn candidate supporting activists the between Bargains eas ahcniaei upre ymlil ciit,w xedteactivist the extend we activists, multiple by supported is candidate each Because characterized be can candidates two the on activists of influence the that argue We otiuoswt an with Contributors j sposition ’s j j .Spoefrhrta h ciit aepoie hi resources. their provided have activists the that further Suppose 1996). by k stelna combination linear the is k ∈ k a nvtr’uiiy oeta h ciitvlnefnto for function valence activist the that Note utility. voters’ on has aac locus balance A r denoted are j z sedwdwt tlt function, utility a with endowed is j n h rfre oiino h ciit h eore allo- resources The activist. the of position preferred the and , R ,i hi aetertcmdlo ciit,con- activists, of model theoretic game their in 1996), rsto pia oiin o ahcniae This candidate. each for positions optimal of set or , lcoa motive electoral jk (U μ j j n hscnrc e a eue omlyto formally used be can set contract this and , k (z (z μ j j j ) j )) nadto,tecniaems hoea choose must candidate the addition, In . (z nune vr oe n hscontributes thus and voter every influences .Let j ) = nedt rmt h electoral the promote to intend μ jk k oa ciitvlnefunction valence activist total (R ∈ A jk j {A μ U (U 34 j .ShimnadN Schofield N. and Schnidman E. jk k j k hngvsa gives then hc eed ncan- on depends which , } (R nunemotive influence (z o ahcandidate, each for j jk ))) (U eoeteeffect the denote k (z j ))) otatset contract . aim j for j is , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

966 965 964 963 962 961 960 959 958 957 956 955 954 953 952 951 950 949 948 947 946 945 944 943 942 941 940 939 938 937 EDITOR’S936 935 934 933 932 931 930 929 928 927 PROOF926 925 924 923 922 921 erwt ( Meirowitz oi,adoesca.Teecnb on yfco nlsso uvydt as data survey of analysis factor by found be can These above. indicated social. one and endogenous an nomic, with deal explicitly not can- does of model terms his constraint. in but budget expenditure effort, of of influ- choice choice factors the didate what proxies or Meirowitz levels decision.” campaign can- these spending competition ence choose of candidates models Downsian how outcome. explain election not the influence would it that uttoatvs rusfrec addt.Ti dacnb eeaie omany to generalized be can idea This with candidate. case simplest each groups.) the for activist is groups (This curve. activist contract two similar just a by described be may groups at located Party, Republican the for supporters of groups activist two the qiiru ntevt hr aiiiggm sta h at oiin satisfy positions party the that is game maximizing share a vote the in equilibrium egtdeetrlma o party for mean electoral weighted uig u ntelegislature. the in but buying, t ocscniae osedmr,wieAhot n un eMesquita de Bueno and Ashworth while more, spend to ( candidates forces ity hs uvtr sdtrie yteecnrct fteuiiyfntoso the of functions utility the of the eccentricity this call a the We is curve by groups. contract activist determined this that is shown would be curvature groups can two whose It these adopt. that to policies candidate of their set prefer the represents This curve.” “contract a by ueo Seeto results. election US on the with ture influence associated valences will various resources the calcu- these of This which has to election. electorate model. degree the the the that contest perceptions of with to the estimate position-taking resources an candidate of activist requires maximizing consequences obtaining lation that of electoral of plausible result the necessity more the balance an the is are that of it candidates functions because the that utility defeat of argue We accept positions positions. to estimated policy the willing particular be to would candidates adherence that assume plicitly (1977 as Wittman regarded be then can activists. the candidates of the ac- preferences of aggregate policy positions the incorporating estimate estimated also The we influence. valences tivist these estimate we when Consequently, Politics Electoral US in Leadership and Gridlock of Quandaries 35 2009 e ot (2004 Coate See hsmasta,frec party, each for that, means This equation. balance sFig. As eco- one policy, of dimensions two are there suppose model, above the apply To h rtodrcniinfrtecniaepositions candidate the for condition order first The eetltrtr neetoshsfcsdo h fet fcmag expendi- campaign of effects the on focused has elections on literature recent A by proposed as positions, policy have candidates where models Electoral ups htcniae u aec oa oices hi lcinchances. election their increase to as so valence buy candidates that suppose ) 5 2008 niae,w a ersn h oflcigitrsso agisbetween bargains or interests conflicting the represent can we indicates, ,Clet(1985 Calvert ), okda vote at looked (1996) Snyder and Groseclose by paper earlier An example. for ) oe ht“addtsadprissedn hsmnythought money this spending parties and “candidates that notes ) 35 err ta.(2008 al. et Herrera ,Dga n e (2005 Fey and Duggan ), eulcncnrc curve contract Republican z j j el ie yteexpression the by given , = i  ij x i . ugs hteetrlvolatil- electoral that suggest ) j ,adDga ( Duggan and ), = (z h eortactivist Democrat The . dem ∗ dem ,z rep ∗ or ) rep tobeaNash hr sa is there , 2006)im- R catenary and 111 C , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 1012 1011 1010 1009 1008 1007 1006 1005 1004 1003 1002 1001 1000 999 998 997 996 995 994 993 992 991 990 989 988 987 986 985 984 983 EDITOR’S982 981 980 979 978 977 976 975 974 973 PROOF972 971 970 969 968 967 ueadfudta h qiiru o h oe 4 nTable in (4) model the for equilibrium vector the that found and dure ugs htti rtodrcniinwl urne httevector the that guarantee will condition order first this that suggest i.5 Fig. 112 o candidate for equation ciitrsucs h qiiru odto ol egvnb h condition the by given be would condition equilibrium the resources, activist test. a be then will ntemdlfrte20 lcingvni Table in given election 2008 the for model the In eew call we Here hsi eemndb h e fvtrpeerdpoints preferred voter of set the by determined is This h eodtr nti xrsini aldthe called is expression this in term second The pia eulcnposition Republican Optimal o each for j uesrtg aheulbimo h oemxmzn oiia con- political maximizing vote the of equilibrium Nash strategy pure ildpn ntepsto fteohrcandidate, other the of position the on depend will [ z j j el ste ie by: given then is − z j ∗ ] the z j el lcoa rdetfrpryj party for gradient electoral − z j ∗ z j el + − 2β 1 z j ∗

dμ dz = j 0. j

z  ciitgain o at j.We party for gradient activist = eue iuainproce- simulation a used we 5, 0. {x .ShimnadN Schofield N. and Schnidman E. i } ic nteasneof absence the in since , h coefficients The . 5 a ie ythe by given was z ∗ k .The = (z dem ∗ balance ,z { rep ∗ ij ) } Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1058 1057 1056 1055 1054 1053 1052 1051 1050 1049 1048 1047 1046 1045 1044 1043 1042 1041 1040 1039 1038 1037 1036 1035 1034 1033 1032 1031 1030 1029 EDITOR’S1028 1027 1026 1025 1024 1023 1022 1021 1020 1019 PROOF1018 1017 1016 1015 1014 1013 oiin eegvnby given were positions etosw ics h oflcsta nu ewe eortcPeietand President Democratic a between ensue that conflicts Congress. the discuss the we to sections candidate Democrat the pull space. policy activists the Democrat of left while upper space, policy the of h ciitgain of gradient activist the unaiso rdokadLaesi nU lcoa Politics Electoral US in Leadership and Gridlock of Quandaries ini h eae o xml,o aur 5 09 h eaevtd5 against 52 voted Senate the 2009, 15, January on example, For opposi- Senate. initiatives Republican the policy navigating Obama’s in after of only tion but some Congress administration through his passed in the successfully through Early filibuster. policies Democratic the blocked the of Senate on use out- the dependent in policy initially Republicans precise which still to the were degree administration Congress Obama’s of President houses from both comes in majorities Democratic Despite Election 2008 Post 4 h ciitma oiin are positions mean activist The oee,a icse bv,adsoni Table in shown and above, discussed as However, suigthat Assuming iia ocuin a emd bu ogesoa lcin.I h following the In elections. Congressional about made be can conclusions Similar quadrant right lower the to candidates Republican pull activists that suggests This 2β 1

dμ d z  = = = z ∗ ⎣ ⎡ z ⎣ ⎡ eet h nuneo ciit,te eoti netmt of estimate an obtain we then activists, of influence the reflects ∗ y x y x − z z z z el canObama McCain canObama McCain ∗ el act −0.25 +0.46 −0.37 = 0.59 = = ⎣ ⎡ ⎣ ⎡ ⎣ ⎡ y x y x y x canObama McCain canObama McCain +0.82 −0.32 +0.75 −0.22 −0.37 −0.12 +0.13 0.59 −0.82 1.41 epDem p Re ⎦ ⎤ ⎦ ⎤ . − +1.14 −0.20 +0.75 −0.22 −0.07 +0.10 ⎣ ⎡ y x ⎦ ⎤ ⎦ ⎤ ⎦ ⎤ canObama McCain h siae candidate estimated the 3, −0.12 +0.13 . . . −0.07 +0.10 ⎦ ⎤ 113 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1104 1103 1102 1101 1100 1099 1098 1097 1096 1095 1094 1093 1092 1091 1090 1089 1088 1087 1086 1085 1084 1083 1082 1081 1080 1079 1078 1077 1076 1075 EDITOR’S1074 1073 1072 1071 1070 1069 1068 1067 1066 1065 PROOF1064 1063 1062 1061 1060 1059 o h et etn rs adl) iial,o oebr1 09tecentrist the 2009 ultimately 1, (and November seat on Similarly, Senate handily). Florida Crist the Rubio, beating for Marco seat, darling, candidate Party the GOP Tea won the a and as Ultimately, Independent primary. nominated an Senate was become GOP to the forced in up was Crist lining Crist began Charlie they Governor that centrist much the so care against health on policies Obama’s opposed activists to 5 splits rulings. regularly right-leaning still fundamentally more which not of has court favor it current in 3, the 4 of 37. of high time to makeup num- all ideological 63 the an the increased of to changed Court has vote Supreme court a the the on by women to 2010 of justices ber 7, confirmed female August was left-leaning Obama on two she Similarly, She adding Sotomayor and Court. 31. Though after court the to year high on one 68 the serve exactly to of to almost Kagan, vote woman Elena a third woman, by the another 2009, and nominated 6, Hispanic August first on the Justice is Court take Supreme would as legislation tomayor healthcare broader Americans over all battle for the coverage that health 2009. of guarantee clear it most to not Obama, steps President was many by it of signed but first was the bill the as bill, When seen the health it. was for extending against voted bill voted a Republicans also Democrats to Forty 2 135, House children. to and The low-income 290 36. of of to vote millions a to 61 by insurance Senate 4, the February on passed approval discrimination final gave pay against bill a 2009 support- Democrats conservative dog” Republicans. “blue no on even but with House it, 196, the ing to passed 233 of proposal by trend 2009 budget the 2, Federal reverse April to trillion % and $3.6 12 warming (over The global trillion inequality. address $1.75 growing to at 2009 promises for included deficit It budget GDP). estimated of an with past, the of ties immediately was bill The votes. 60 just Obama. with vot- President it Republicans by passed 183 signed Senate and the favor the days; in few while voting a against Democrats within ing 245 Senate with and the House House of the the both discussion passed by including bill halt upon billion, to agreed $787 was of 61 cuts, bill of to tax compromise 9 some A majority bill filibuster. February required a stimulus On blocking the the thus senators. by bill, Republican of stimulus vote three size indeed of the did support 2009 Senate reduce the 6, the to for February return Senate On in the billion, program. in $780 recovery reached economic was Obama’s agreement of an support in 42 and 114 nOtbr 09 n ru dniyn sppls eulcn,te“e Party” “Tea the Republicans, populist as identifying group one 2009, October, In So- Sonia confirmed Senate the when victory important another gained Obama 22, January on success; of modicum a received even policies social Obama’s priori- the revised that budget year 10 a proposed thinking. Obama economic 2009 of 26, terms February a Great in On argument the over better since got a crisis we’ve have economic I where biggest think the situation I and the Depression, debt in of tested was worth us theory dollars’ landed their trillion that its and there, and getting thoroughly me to pretty prior time long pretty a for afterwards: commented Obama As o aet a htgvnta teRpbias eernigteshow the running were Republicans] [the that given that say to have I Now .ShimnadN Schofield N. and Schnidman E. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1150 1149 1148 1147 1146 1145 1144 1143 1142 1141 1140 1139 1138 1137 1136 1135 1134 1133 1132 1131 1130 1129 1128 1127 1126 1125 1124 1123 1122 1121 EDITOR’S1120 1119 1118 1117 1116 1115 1114 1113 1112 1111 PROOF1110 1109 1108 1107 1106 1105 eto c,adti a indit a yPeietOaao uy21. July on Obama President by law into Pro- required signed Consumer was 2010 and the this Reform 15, and obtain Street July Wall Act, On DoddÐFrank to tection the Senate. failed for the 60Ð39 passed bill voted eventually Senate Regulation it the Financial 2010, 19, the the May though on on supra-majority Even debate respectively programs. end 68Ð29 creation” and to 62Ð30 “jobs 22, vote voted multi-billion-dollar February Senate two On the 2010 opposition. implement 17, less to March far on with again passed and legislation 2010 of pieces major other otiue oti poiina el ntehat ilvt nteHuei early in against. Re- House voted 176 the Democrats while in Dog” bill, “Blue the vote for 39 bill voted and Democrats Republican health publicans 1 some the with Interestingly, Democrats in 219 well; Democrats. 2009, November as the opposition and this long Obama to as contributed coalition stymie blocking to their possible utilize continue as to intending 1872. were since Republican Republicans been that had that district a in Bill election candidate, the Democrat won the election endorse Owens special and Owens. district the congressional of 23rd out York’s New drop in to decided Scozzafava, Dede candidate, Republican Politics Electoral US in Leadership and Gridlock of Quandaries 39 38 37 36 rcniito”t oc huhahat iluigsml aoiyrule. majority simple as using known bill manoeuvre reach a health to using a attempt of though an talk force re- was 2010, to majority there 25, “reconciliation” and seat February failed, 60 On compromise the legislation. bipartisan the a of in through Democrats election opposed. push the Senate to Republicans special deprived quired 39 the 19 with in January votes Brown on 60 Scott Massachusetts Republican by by again victory Senate, the the However, in passed bill health sainms esont ebde eta vratnya pni codnewt h Byrd the 2012. 28, with June accordance on 4 in to span 5 ten-year a over neutral budget be to rule. shown be must islation 60. of super-majority a required has 1975 since but ity, bill. care health the on discussion fodbeCr c, nMrh2,21 n nudtdvrino h ilon bill the of version updated an and and Protection rec- “Patient 2010 under the required 23, 2010. bill, as 30, March the March 43, on of to Act,” draft 56 Care a of signed Affordable majority President simple the by and bill onciliation Democrats. Repub- the dissident President. 33 the passed by to Senate joined bill The legislation, care the health against a unanimously send voted Nancy to speaker, licans 220Ð207 House voted and Obama House President the by Pelosi, efforts strenuous after 25, March on rms il hsbogttesz ftecaiint h rtclsz f6 votes, 60 of size critical Senate. the the to in com- coalition decision a the a to of force agreed size to had the sufficient Brown, brought Sherrod as This and Democrats, bill. Nelson the promise Ben with Senators caucuses Democrat who as independent well an Vermont, of Sanders Bernie otayt xettosteSpeeCutuhl h osiuinlt ftehat aeatby act care health the of constitutionality the upheld Court Supreme the leg- expectations the to majority; Contrary simple a with Senate the pass can bill a whereby major- measure two-thirds a a required is originally Reconciliation It end. an to final debate bringing the at to aimed move motion to a lines, is partisan Cloture along 40, to 60 voted Senate the 21, November Saturday, On hl tsee ht“rdok nudoe h elhcr eilto,several legislation, care health the over ensued “gridlock” that seemed it While steHatcr eaehae poe umradfl f20 tbcm clear became it 2009 of fall and summer over up heated debate Healthcare the As 39 37 ial nCrsmsEe 09 the 2009, Eve, Christmas on Finally 36 yDcme 9 Senator 19, December By 38 Finally, 115 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1196 1195 1194 1193 1192 1191 1190 1189 1188 1187 1186 1185 1184 1183 1182 1181 1180 1179 1178 1177 1176 1175 1174 1173 1172 1171 1170 1169 1168 1167 EDITOR’S1166 1165 1164 1163 1162 1161 1160 1159 1158 1157 PROOF1156 1155 1154 1153 1152 1151 eindt aeavnaeo h ako pniglmt.Oecmitewsset was committee One limits. spending expressly of right, lack and the left of the advantage take from to each designed one committees, campaign “independent” nation.” the for consequences terrible with act political htrslsfo h oe fitrs ruslctdi h poe udat of quadrants opposed the in located polarization groups political of interest space. degree of policy the ef- power the be be the to can from appear government results would that that problem doubt the States of United Part the fective. in voters that surprising legislation. hardly passing these at because efficacy oppose If of simply enact 2010. lack or to of their party impossible to end either prove due blame the to incumbent could at any were electorate expire cuts, the to tax opposition, the Republican due of of Bush were resolution the which to the 2003, Republican regard and with and to bills, remained 2001 due also of collapsed problem cuts major Act A tax Change tax. the Climate carbon 2010, the a the 22, to 2009, July with opposition On 26, forward Senate. push the June in to Energy on action effort Clean reached House never American (ACES), the the Act Security called by and formally passage bill, Independence change Energy Despite climate an Act. Waxman-Markey Trade and Expanded Change Act, an Climate Immigration Act, Comprehensive and Reduction a Deficit Act, A Export Congress: and through put to bills major ilitrssicuigfrincroain—osedwtotlmti u elec- spe- our for floodgates in the limit open without will spend tions.” believe corporations—to I foreign that law interests—including of cial century a “reversed had court mrcnAto omte a pn 3. ilo,$66mlinad$7million $17 and million the $26.6 and million, respectively. Crossroads $32.8 American spent had Commerce, Repub- Committee PAC” of groups Action “super Chamber American these Three US that donors. the estimated undisclosed groups, from was activist % lican it to 60 2010, million, required 1, $280 not November spent are of had non- and As through donations donors. funneled unlimited their accept were disclose can money that of Novem- groups amounts the advocacy large In profit increase. election only mid-term will 2010 contributions is activist ber, it of contributions, importance campaign the on limits that removed clear which 2010, 21, January decision, on Court mission, Supreme the of result a As 2010 in Decision United’ ‘Citizens the of Impact 4.1 116 41 40 ildcrma tt fteUinsece osaehsha ndsgemn ihtePresident, the with disagreement in head his true.” shake not “that’s to words speeches the Union muttering the reportedly of State at decorum cial urm or utc aulAio pone yGog .Bs,boefo rdtoa judi- traditional from broke Bush, W. George by appointed Alito, Samuel million. Justice $7.1 Court raised Supreme Committee Action First Families America’s pro-Democrat The nJl,21,teFdrlEeto omsinhdapoe h raino two of creation the approved had Commission Election Federal the 2010, July, In ie hs netite urudn oiycocsi h eiltr,i is it legislature, the in choices policy surrounding uncertainties these Given four remained there 2010, November, in Congress 111th the of end the Nearing nhsSaeo h no drs nlt aur,PeietOaasi the said Obama President January, late in address Union the of State his In 41 wri ( Dworkin 40 ae aldteSpeeCutdcso a unprincipled “an decision Court Supreme the called later 2006) iiesUie .FdrlEeto Com- Election Federal v. United Citizens .ShimnadN Schofield N. and Schnidman E. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1242 1241 1240 1239 1238 1237 1236 1235 1234 1233 1232 1231 1230 1229 1228 1227 1226 1225 1224 1223 1222 1221 1220 1219 1218 1217 1216 1215 1214 1213 EDITOR’S1212 1211 1210 1209 1208 1207 1206 1205 1204 1203 PROOF1202 1201 1200 1199 1198 1197 mlctoso h 00Election 2010 the of Implications 5 to able be will the will Democrats, to groups the thanks Both amounts, to unions. unlimited and ties spend corporations close individuals, govern- with from less Ten, money and Commonsense raise taxes low called for other, advocate conservative The the ment. Growth, for Club the by up Politics Electoral US in Leadership and Gridlock of Quandaries ot6sasbtrtie aoiyo 1t 6(ih3Independents). 3 (with 46 to 42 51 of the Democrats majority in the a Senate seats the retained 63 In but 192. lost seats to 242 Democrats 6 of the lost majority Republican 2010, a November, to leading In House, woes. Democrats, particularly economic incumbents, their blamed electorate for the elections midterm the In occurred has that distribution wealth and account income Pierson well in and may al. inequality (Hacker and et the society (Layman increase in wealthy the years the axis for forty benefited social last have the to on the appears activists over inter- polarization of polarized particular positions of more policy the is become that election) have evidence midterm is a there extremely for because The est (especially spending. expenditure Democrat of total level than high higher somewhat spending the publican target of weak. to were element able Democrats key were incumbent a where bodies races activist Indeed, many Senate these candidates. to and that House Party million was Tea $1 campaign the about Republican funnel of successful to right-wing PAC most a the as of Fund Conservatives Senate the used News Murdoch’s Rupert from million $1 of Corporation. excess in pub- donation corporations highly singular These most a The was filings. licized Financial. Prudential tax Sachs, The the their Goldman to million. Chemical, in donate Dow $17 contributions include that spent these corporations GPS but disclose purposes. nonprofit, must 501(c)(6) political Chamber a for is used donors, Commerce be disclose of to to Chamber need supposed not as not does it (GPS) is means Strategies but This dis- Policy nonprofit. welfare Grassroots to social Crossroads required 501(c)(4) formed a was then but They committee, donors. independent-expenditure-only close 527 a as to roads 59 by failed bill 23. the September on on Senate cloture the at purchas- in when attempt 39 donors second on their A disclose vote advertisements. to a groups political blocked interest Republicans ing special Senate force congressional 41 would all November that when bill the 27 a July before on regulations defeated was finance election campaign new impose to russpotn eulcncniae pn bu 2mlin hywn12. won they million, $2 about spent candidates Republican supporting The groups pattern. spending the to .Hwvr h eortlse a edue be may losses Democrat the However, (2010). Bunch by predicted backlash the was This nte21 lcinccettlcmag pnigwsaot$ ilo,wt Re- with billion, $4 about was spending campaign total cycle election 2010 the In DeMint, Jim Senator, Carolina South groups, activist external the to addition In Cross- American formed first Gillespie, Ed and Rove Karl advisors, Bush Former 2006 e okTimes York New , 2010 iro n Skocpol and Pierson ; nlsssgetdta n2 os itit where districts House 21 in that suggested analysis iiesUnited Citizens 2007 eiin eorteffort Democrat A decision. Reich ; 2010 2010 42 ). ayof Many .This ). 117 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1288 1287 1286 1285 1284 1283 1282 1281 1280 1279 1278 1277 1276 1275 1274 1273 1272 1271 1270 1269 1268 1267 1266 1265 1264 1263 1262 1261 1260 1259 EDITOR’S1258 1257 1256 1255 1254 1253 1252 1251 1250 1249 PROOF1248 1247 1246 1245 1244 1243 ahntnwstems oaie ehdse ic onn h eaei 1977. in that remarked Senate Missouri, the from Republican joining in senator the since Republican environment former seen the defied the had that Danforth, he he C. said as John polarized has most election Lugar the 2010 Senator was demands. the Washington various after their Dodd over Senator Party as way same onciu eao hi odgv i aeitr peho h eaeflo on following: floor the Senate including the remarks on with speech 2010 valedictory 30, his November gave Dodd Chris Senator Connecticut funds responders. allocating first legislation 2001 even 11, September and to treaty healthcare arms provide immigra- nuclear to legislation, a Tell” Don’t legislation, included Ask, reform measure “Don’t Republican tion the This receive. repealing tax to on their set discussion on were blocking extension percent same 98 other the received the un- Americans that wealthy session cuts of duck the percent lame two for the wealthiest during up the legislation stood til all Party blocking on Republican insisting the before by Even session benefactors campaigns. entered candidates in- Congress Party wealthy 112th Tea and the corporations the op- large of that are many fact funded that the heavily independents despite dividuals conservative is This as business. themselves big to see posed activists Many Party activists Tea party the current and of activists disaffected mobilized of coalition winning the with itself striking Congress Representative wing. a on most for Party Tea had post Perhaps the leadership they from programs. special impact a other immediate creating leadership the cutting House is medicare, Republican by Party and deficit Tea security the the social about reduce of supportive to are want supported opposed generally are movement and supporters they populist party spending, tea a Although deficit West. of and to representation South the a in elites is by Party primarily Tea the that postulate tde fteTaPryidct htte r rdmnnl le,mdl class middle older, predominantly are Americans. they white that rural indicate and Party suburban Tea demographic Reserve, preliminary the Federal Further, of the taxes. and studies abolishing income Party even Tea like and the stances benefits, of policy unemployment backing extreme the to received subscribed Congress vocally of members elected newly the 118 43 at ebr r eyseicdmgahcsbgopwt rdtoa ouitcnen.See Tea concerns. that populist are traditional indications ( with all Schoen sub-group preliminary, and demographic Rasmussen only specific also very are a findings are their members although Party and emergence its since kco n ilasn(2010 Williamson and Skocpol nte le ttsa nteSnt,IdaasRcadLgr lal etthe felt clearly Lugar, Richard Indiana’s Senate, the the in on statesman elder rests Another debate productive, ultimately Senators. but between create passionate, relationships par- partisans personal a a determined between and when difference brawl the only end, tisan so the do In bipartisanship— bipartisanship. can with thus you debate consensus—and bipartisan- the And Political begin there. don’t nation. You arrive a process. you a as core not success the goal, and a at democracy, is is a ship it fact, as In strength weakness. our no That’s of debate. partisan pitched often and na fott ls i aerwt atn dieaotcmrms,retiring compromise, about advice parting with career his close to effort an In a build to have parties system, electoral US the of nature plurality the of Because rmtemmn forfudn,Aeiahsbe nae na eternal an in engaged been has America founding, our of moment the From 2010). aebe olcigsre n neve aao h e Party Tea the on data interview and survey collecting been have ) 43 hsdmgahcmk-plasoeto one leads make-up demographic This .ShimnadN Schofield N. and Schnidman E. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1334 1333 1332 1331 1330 1329 1328 1327 1326 1325 1324 1323 1322 1321 1320 1319 1318 1317 1316 1315 1314 1313 1312 1311 1310 1309 1308 1307 1306 1305 EDITOR’S1304 1303 1302 1301 1300 1299 1298 1297 1296 1295 PROOF1294 1293 1292 1291 1290 1289 ra tep yteamnsrto ohre akt h 90s(rerir when earlier) (or 1990’s the to desperation of back tactic. act political harken an acceptable to either an administration was as In- the compromise seen compromise. by be the can attempt of way an support this or in in conference President at- Bill press former President persistent the House volving former White of invited a wake Obama give the to Democrats, In Clinton liberal as values. prominent compromise several saw progressive they by Obama’s as tack President Party of Democratic more of betrayal compromise wings by This a extension. liberal down cut the unemployment tax the in bring of many lifespan will angered year wealthy two the the over not percent to unemploy- do 0.1 cuts forecasts of than economic tax extension positive these an most the that for tax even exchange predict that the (in fact allow the Americans to despite all benefits) deal for ment a extended struck be eventually to he cuts reach, within accomplishment islative election primarly a lose to Senator years. six-term 60 first in the was Lugar Mourdock. Richard adb eaeDmcas n rmsdt s hi aoiyi h e os to House new the for- in put majority their measure use spending to promised trillion and $1.2 Democrats, a Senate politics. by derailed driven ward activist Republicans DeMint in Senate compromise and to and Kyl roadblocks House Senators the of by reminder statements stark a These week on Year. provide the work during New work much and to so return Christmas to putting between need by might Senators holiday that docket Christmas Congressional the the “dis- and for Nevada) Car- institution from South the (Democrat from Reid respecting” Harry (Republican Senators Leader DeMint Majority irked Jim criticized productivity who and olina) and Arizona) from compromise (Republican toward Kyl step gov- Jon federal This the the 2011. funding of continue into passage to the measures ernment temporary on Tell,” on Don’t and treaty, Ask, arms “Don’t nuclear repealing on debate began day. Congress Republicans next 36 the law and into Democrats bill 139 112 the with signed and passed Obama favor measure President in the opposed. order 16 Republicans in December 138 ransom on king’s and Nancy midnight Speaker, at a Democrats class” House pay middle by “to the accusations Democrats help Despite forcing to 19. were to Republicans 81 that of Pelosi, in vote a invest bil- by on who Senate $858 15 for the landowners December passed the breaks overwhelmingly and bill of tax compromise expenses provisions grant The techniques. Other commuting and conservation benefits bill. transit insurance cut mass unemployment tax schoolteachers, extend the would on bill debate lion halt to voted Senate unaiso rdokadLaesi nU lcoa Politics Electoral US in Leadership and Gridlock of Quandaries 44 h raetsrntso h Spltclsystem. political US the of strengths greatest the ti ot oigta h onigFtesrpael ie h edfrcmrms soeof one as compromise for need the cited repeatedly Fathers Founding the that noting worth is It ept nrae oaiain rsdn bm otne opesfrayleg- any for press to continued candidate Obama Party President polarization, Tea increased to Despite election primary a lost Lugar Senator 2012 of May In far redemption. so beyond gone is have are policy, we we foreign Party, that Republican on overboard the authority in leading anybody the by challenged and seriously Senate the in person respected fe hsiiilcmrms a tuk h ojmsee ohv rknas broken have to seemed logjam the struck, was compromise initial this After The worked. ploy bargaining Republican the 2010 13, December Monday On fDc ua,hvn evdfietrsi h SSnt n en h most the being and Senate US the in terms five served having Lugar, Dick If 44 119 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1380 1379 1378 1377 1376 1375 1374 1373 1372 1371 1370 1369 1368 1367 1366 1365 1364 1363 1362 1361 1360 1359 1358 1357 1356 1355 1354 1353 1352 1351 EDITOR’S1350 1349 1348 1347 1346 1345 1344 1343 1342 1341 PROOF1340 1339 1338 1337 1336 1335 himno h ugtCmite h lern rpsdacti xedtr of expenditure and in Wisconsin decade. cut next a of blueprint the proposed fiscal over Republican blueprint trillion the Ryan, The $5.8 approve Committee. D. to Budget Paul the 193 Representative of to chairman by Republicans 235 many drafted voted with House 2012, again House the 19, for the 15, to of 81 April Senate, members On the the 59 against opposed. passed voting with also minority 167, bill Democratic the The to the half legislation. of 260 than rest was more the and vote House for majority House government Republican the federal final wrangling, the The much finance year. After to fiscal billion. 14, April $38 on by legislation budget passed the would cut and that Senate compromise the However, Jan- in 189. on majority to Obama. vote, Democrat 245 President the of to by pass margin signed was to a be Congress able not by obviously 112th be Bill not Care the was Health in repeal the House this repeal the to 2011, by 19, moves uary first the of One Congress 112th the in Gridlock 5.1 up increas- run and the Congress in divided conflict election. partisan highly 2012 escalating a the in that the to resulted surprise has of politics no driven party is activist results it ingly the the elections, Given longer ultra-wealthy. 2010 the no the and class of is middle disillusioned Party the Republican of but the wealthy while constituencies, poorest Constitution. the by required majority two-thirds the President exceeding of favor, achievement in vote treaty. policy cau- to the Democratic foreign cus unanimous for tangible a 26 joined most to Republicans the Thirteen 71 administration. voted as Obama’s Senate seen the was later treaty Rus- days and Two This States treaty. United the the the renegotiate required killed have to have would sia would text the amendment to The change any Russia. because with treaty Start, New 175. treaty, to control 250 arms by repeal repeal openly this serve to approved to 31 previously gays to had for 65 House possible vote The military. it did the making Senate in legislation, the Tell” but Don’t 55Ð41, Ask, of “Don’t vote the Senate a on failed citizens spending temporary a 2011. approve March did until Congress up 21 bill December On government. shrink 120 neryArl 01asudw ftegvrmn a nyjs vre ya by averted just only was government the of shutdown a 2011 April, early In the and richest the represent increasingly Congress not in capacity Democrats the together. the have However, progress we make that to We’ve but elections gridlock. progress, November from endless make the draw to to of doomed only to not wake lesson are the any we in there’s that shown If it’s weeks, productive generations. most few in the past of had these heels we’ve the that on comes years it two and decades, in had we’ve period said: Obama As new the to amendment an reject to 37 to 59 voted Senate the 20, December On become to students immigrant illegal allow to Act” “Dream the 18, December On hn tsfi osyta hshsbe h otpoutv post-election productive most the been has this that say to fair it’s think I .ShimnadN Schofield N. and Schnidman E. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1426 1425 1424 1423 1422 1421 1420 1419 1418 1417 1416 1415 1414 1413 1412 1411 1410 1409 1408 1407 1406 1405 1404 1403 1402 1401 1400 1399 1398 1397 EDITOR’S1396 1395 1394 1393 1392 1391 1390 1389 1388 1387 PROOF1386 1385 1384 1383 1382 1381 ontr,hsagrdmn nppls ice.Mawie nitneo closing on economic insistence the Meanwhile, address circles. to populist order in in many businesses angered other propos- has and budget down-turn, banks massive to his relief with and providing along secretary als steps, Commerce These former Electric Competi- Daley. was William and General staff executive Jobs of naming bank on Chief to Council second President’s addition President’s the the In of tiveness, Chairman crisis. as economic liberals Immelt Jeffrey the social CEO pro-business to retain response and attract his to with attempted economy.” has our Obama and contrary, the people to day. American way the the for a % for find 2 works can about that down and closed way work Dow a The committee’s in this issue on this build remain tackle “We can declared: but Congress agreement, closing any that the to to hopeful come reduce resistance not to could and Committee it Joint that wealthiest announced the the from however, coupled deficit 21, for reaction bill, November cuts the Initial On tax loopholes. of Congress. keeping corporate measures to on some insistence bill accept an jobs to with billion willingness European $447 a continuing indicated a and Republicans downgrade sending Poor by and crisis Standard debt debate, Ceiling Debt the increased. Bonds Treasury US for demand However, a t$43tilo n h urn STesr etws$42 trillion. $14.29 was debt Treasury US ceiling current debt The the debt. and public trillion US $14.3 the of at question was the over polarized completely ties nolw ept h vnulcmrms ntedb eln,o uut5 2011 5, AAA August from debt on Federal AA ceiling, US downgraded to agency, debt rating the credit bill the on Poor, the and compromise signed Standard eventual immediately the Obama Despite law. President Democrats 6 into against. and voting Republicans, trillion. 19 independent $1 with over 2 one August of and on cuts 74Ð26 future bill the determining commit- passed of Senate joint responsibility The designed the newly with through A vested billion billion. was $500 $900 tee additional over of an Democrats, caps with spending by billion, with $400 supported February, by House 269Ð161, ceiling the bill, debt 1 the August compromise increasing On a proposal. House passed the a game Representatives tabled The get of which Republicans. to Senate among Boehner, the fighting Republi- to A. intense 22 shifted of John then with week Speaker, a 218Ð210, House ended was by This vote efforts approved. The bill the spending. support in to cuts unwilling and cans ceiling debt the in population. programs the of social segments on certain spending on increases continued while tax cuts, opposite, and tax the of maintenance wanted the basically and Democrats spending in reduction a demanded publicans unaiso rdokadLaesi nU lcoa Politics Electoral US in Leadership and Gridlock of Quandaries 45 n ogKn od$. rlin te onre od$. rlin h eann 36tilo is trillion $3.6 China remaining Reserve. the trillion, Federal $3.2 the hold etc. in funds countries trillion pension other by $1.6 trillion, held and $1.3 hold agencies Kong government Hong other and in trillion $1.9 Fund, fti 62tilo shl yteU oenet 27tilo nteSca euiyTrust Security Social the in trillion $2.7 government, US the by held is trillion $6.2 this Of h eaeoe h osbl ihihstefc ht ept ei teto to attention media despite that, fact the highlights bill jobs the over debate The by off set turmoil economic the on acted Obama President 13, September On yJl,i emdta h oiia ytmwsaani rdokwt h par- the with gridlock in again was system political the that seemed it July, By h os nFia uy2,fial prvdapa o hr-emincrease short-term a for plan a approved finally 29, July Friday on House The + n h o nutilidxdopdaot2 ntefloigdays. following the in % 20 about dropped index industrial Dow the and , 45 Re- 121 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1472 1471 1470 1469 1468 1467 1466 1465 1464 1463 1462 1461 1460 1459 1458 1457 1456 1455 1454 1453 1452 1451 1450 1449 1448 1447 1446 1445 1444 1443 EDITOR’S1442 1441 1440 1439 1438 1437 1436 1435 1434 1433 PROOF1432 1431 1430 1429 1428 1427 h ogtr,tepria laaeln a oaefrhri lcws direction. clockwise a in further rotate may line cleavage partisan the term, long the raino addt ue As uha Mk sGetAan o Republican for Again” Great the Us to “Make eye as blind such a PACs, new turn Super while, to the candidate All Commission of election. Election creation general Federal the the be in allowed might voters precedent it moderate legal to that appeal primaries to the him Rom- in for positions Governor campaign, hard conservative that primary radically concern such adopt the taken declaring to had of publicly ney forced pro- end were was who strategists the Romney Adelson Republican By Gingrich, some Sheldon Newt positions. billionaire of policy and support conservative million in increasingly millions $55 of of tens supported tune which vided the Again” nomination, Great to the Us to “Make Perry PAC route the Rick as centrist such more money a activist facing seeking but to Governor admitted Newt Former initially House candidates, Romney the Republican Mitt other of the conservative Speaker to former seasoned himself and Comparing with Gingrich. Santorum along Rick Perry Senator Rep- Rick former Governor cycle Bachman, ideologues and Michele election Paul, Representative Presidential 2012 Ron darlings Republican The resentative Party Early persist. non-convergence. non- Tea and to included and money money frontrunners likely irrationality of and is this role falling center elections, the barriers highlights recent electoral in legal the role With to convergence large irrational. increasingly politics an US playing made has Money 2012 and Conclusion 6 Democrats. the some to while switch space, will policy quadrant the right of upper quadrant the left in lower the interests to business move Republican “cos- to the conservative is, begin economically That may Republicans. Party and populist liberal to realign- opposed socially party are the Democrats a as mopolitan” where in populism situation result old-style a will of to This ment era anti-Detroit, religion. an anti-banking, pro-evangelical to Street, and anti-Wall back anti-immigration, Bryan: those hearkening Jennings by be William expressed will by demands they expressed populist Party, Should traditional Tea collapse. the the economic the to the in from cater from votes party most gain Republican the will suffering the Party are Republican and result who the is voters that a it blue-collar possible Wall As “Occupy is and Party. the It spiked Tea group, protesters. populist has the Street” liberal-minded anger of distinctly second, populist demands a rates, populist spawned even unemployment Re- more conservative high the traditionally persistently to by of as attacks well to vulnerable as party publicans his and President the eroded has President. regulation, the financial for increased support of business spectre the and loopholes tax corporate 122 46 inefo h eulcnPryt h Democrats. the to Party Republican the from giance o xml,o pi 8 00AlnSetr h eao rmPnslai,sitdhsalle- his shifted Pennsylvania, from Senator the Specter, Arlen 2010 28, April on example, For hslc fspoti ohteppls n ompltnqarnsleaves quadrants cosmopolitan and populist the both in support of lack This .ShimnadN Schofield N. and Schnidman E. 46 Over Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1518 1517 1516 1515 1514 1513 1512 1511 1510 1509 1508 1507 1506 1505 1504 1503 1502 1501 1500 1499 1498 1497 1496 1495 1494 1493 1492 1491 1490 1489 EDITOR’S1488 1487 1486 1485 1484 1483 1482 1481 1480 1479 PROOF1478 1477 1476 1475 1474 1473 tl n pi Nvme 01.Temdlpeetdhr n eeoe further developed and here presented model Greece, Por- The 2011), 2011). (2011), (October (November Finland Slovakia Spain 2011) 2011), and (September, Italy (February, Denmark Ireland 2011), of (June systems tugal multiparty the in ments uncertainty. and gridlock such of presence the in powers of hyaotplce htwudb lcoal rainl eei o o money. for so not it politicians were influence irrational, activists electorally be put, would Simply that politicians policies bases. it for adopt Were activist irrational they be parties. these would between toward it line provide move groups cleavage to activist partisan the resources the the shift for potential and not the have politicians sectors Moreover, in populist persist. draw and to to cosmopolitan likely the is in groups center activist electoral irra- powerful electoral the this to elections, non-convergence “irrational”. recent and politics in In role tionality US large space. made increasingly policy has an the and playing that money important of so With more half rotated conservative become has socially has line money the pivot addition in Republi- electoral located moderate The are no coalition. Republicans now social-liberal all are the there because join 1964 social to in southern cans than the extreme executive choosing more and by is legislative civil the it branchs activists between resolved for gridlock Party but present support dilemma, The Republican Goldwater. party electoral conservative, Traditional Republican an Repub- liberals. in of a social thus fact that were disaffected the was attract of 1964 to of use rights election make the could in candidate Johnson lican for danger The overwhelmed. omk ogtr netet n oiydcsos ned onradVermeule and Posner attempting Indeed, leaders decisions. policy international policymaking ( and and in investments business uncertainty long-term for make increases introduction. situation to US the difficult the in a by mentioned creating choice briefly thus effective quandaries Al- of severe gridlock. the absence unable been with to The has grips Government tended voters, to recently the US come for has the choice to of of money appearance structure the of constitutional is influence they there the though because seen, the have systems with we electoral coupled As proportional polity, choice. to clear preferred a be voters to gave were post” the past ufiin oe nteSnt oefc coue,t lc h otenDemocratic southern the block to ‘cloture’, effect to Senate filibusters. with the Republicans, moderate in and votes Democrats sufficient northern support, liberal Humphrey’s of Hubert coalition with unstable created, an effectively Johnson (Branch 1964, South in of the through margin precisely on Act Congress, narrow effect to possible very Bill the Rights a of Civil (by a because sending 1960 delayed in he Nixon), President against elected victory was Kennedy After election. 2012 1960s. the for million President $100 Even raise Romney. to Mitt plans Republican USA” for “Priorities PAC Future” Obama’s Our “Restore and Perry Rick Politics Electoral US in Leadership and Gridlock of Quandaries 47 aswt 7Rpbiasad4 eort oigae h ilpse nJn 9b 3t 27. to Congress. against 73 Johnston by by 19 fight June long on a passed after again bill passed The 1965 aye. of voting Act Democrats Rights 44 voting The and Republicans 27 with days 2011 6)dsrbstedaao h ltr oeo u 0 94atrafiiutro 57 of filibuster a after 1964 10, Jun of vote cloture the of drama the describes 568) (2012: Caro nteohrhn,tercn uoendb rsshsldt h alo govern- of fall the to led has crisis debt European recent the hand, other the On opr( Popper tlzn h oe ehv rsne,w otattecretstainwt the with situation current the contrast we presented, have we model the Utilizing ru htteUie ttsnest eosdriscntttoa separation constitutional its reconsider to needs States United the that argue ) 47 2008 hswstefis iesneRcntuto htteSuhr eowas veto Southern the that Reconstruction since time first the was This rudta lrlt lcoa ytm,ohriekona “first as known otherwise systems, electoral plurality that argued ) 1998 .T uhteCvlRights Civil the push To ). 123 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1564 1563 1562 1561 1560 1559 1558 1557 1556 1555 1554 1553 1552 1551 1550 1549 1548 1547 1546 1545 1544 1543 1542 1541 1540 1539 1538 1537 1536 1535 EDITOR’S1534 1533 1532 1531 1530 1529 1528 1527 1526 1525 PROOF1524 1523 1522 1521 1520 1519 Acknowledgements will system economic political unstable climate generations. potentially of future this prevention burden of is such costs good severe goods public hege- the global public of important change, social pattern most of external the underprovision the Since and innovation. with mode as inequality consistent current both is the sustains US that and the adds mony in Stiglitz seeking the system. support rent dominate hegemonic political oligopoly to fund this of to came of able that maintenance provided was the speculation also and for of but globalization from form US, benefited extreme the Capital the to market. for costs basis necessary the (both financial disguised debt the econ- US mercantilism) unequal fund its to an willingness of such its become components and has China wage US by unskilled savings the artificial down that High reason forced trade this has increased for (Stiglitz it is While GDP, omy It and US. global US. years, the the in forty of in increase or rates capacity thirty an manufacturing last to the the contributed of of has deficits much trade of trade US increased loss massive such (and the the 1945, “globalization” to in resulting led back The has saw trade unstable. flows) Keynes institute inherently to As is rapidly. China system allow very to a grow attractive to seemed Af- it it dominance. allowing Union, in US Soviet mercantilism, standard of the be gold system of the this could fall sustain of the system to ter coming this pressure by exerted 1970’s mercantilism maintained. capital the was this Finance US 1973, of limit the to crisis Germany, in tried oil growth allies Nixon economic the Although its Indeed until cost. by years great strategies without 30 maintained trade first power mercantilistic hegemonic the its accept In maintained Japan. to US the willingness facil- Union its to Soviet system by the economic international oppose II and War World growth post itate a maintain to order in time that in the resolved. difficulties At been created 2008Ð2009. not has has of crisis also recession euro euro the this the from writing of fall-out on the based with system effectively fiscal far dealing cou- fragile phenomenon policies a This radical instability. with coalition adopt pled inducing will thus center, parties electoral systems, small the multiparty from or representation, fragmented proportional in on that suggests based (2012) Gallego and Schofield in 124 baoizA anesA(05 h a’ egtaog h elt fplrzdAmerica. polarized of 48 reality the along: get we can’t Why (2005) Haven A New Saunders Press, University A, Yale center. Abramowitz disappearing The (2010) A Abramowitz References qaiyadopsdt oacyadheacia eeoy n h opoiigModerate and compromising reason the of and hegemony, support Burke. hierarchical and in Hamilton and Paine, of monarchy and Enlightenment to Jefferson opposed Condorcet, and by equality espoused Enlightenment Radical a sal( Israel ou ()12.Atce1 Article 3(2):1Ð21. Forum eea esetv nrcn vnsi rvddb id( Lind by provided is events recent on perspective general A 2012 a one u httemdr eidsne10 inse oflc between conflict a witnessed 1700 since period modern the that out pointed has ) 2012 ,mkn h oflc ewe ao n aia oeintense. more capital and labor between conflict the making ), hsppri ae nwr upre yNFgat#0715929. grant NSF by supported work on based is paper This 48 .ShimnadN Schofield N. and Schnidman E. 2012 .H suggests He ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1610 1609 1608 1607 1606 1605 1604 1603 1602 1601 1600 1599 1598 1597 1596 1595 1594 1593 1592 1591 1590 1589 1588 1587 1586 1585 1584 1583 1582 1581 EDITOR’S1580 1579 1578 1577 1576 1575 1574 1573 1572 1571 PROOF1570 1569 1568 1567 1566 1565 letD okt ,RuiiN(01.Tewyfrad h e mrc foundation. America new The forward. way The (2011). N York Roubini New Borgo, R, America. Hockett in 1930’s D, 77(4):974Ð990 Rev the Alpert Sci yesterday: Polit Michigan Since Am activists. (1994) of party FL with University Allen model spatial government. Downsian Rev A party (1983) Econ JH responsible Aldrich Am and institutions. competition and Party elites, (2001) power, J of Adams Persistence (2008) J Robinson D, Acemoglu Politics Electoral US in Leadership and Gridlock of Quandaries emnA(09 e tt,bu tt.Rc tt,po tt.PictnUiest rs,Princeton Press, University Princeton state. poor state, Rich York state. New York blue Giroux, New state, Red and Press, (2009) Straus Free A Farrar, state. Gelman us. predator be The to (2008) used JK 1:337Ð357 That Galbraith Econometrica (2011) Pearson depressions. M great America. Mandelbaum of polarized T, theory Friedman a debt-deflation of The myth (1933) The I Fisher war?: Culture (2005) JC Pope politics. Rev SJ, American Econ Abrams in M, representation Am Fiorina of century. breakdown the the Disconnect: of (2009) SJ decade Abrams progressive M, Fiorina technologically most The Am (2003) vote. the AJ of model Field behavioral the Comput and Math equilibrium candidates. Candidate (1990) presidential DW American Romero RS, of Erikson location The (1989) Princeton M Press, University Hinich Princeton Econ JM, here? Games possible Enelow democracy candidates. Is policy-motivated (2006) with R Dworkin competition D Electoral Wittman (2005) BR, M Weingast Fey In: J, equilibrium. Duggan electoral and York objectives New Row, Candidate and Harper (2006) democracy. J of York theory Duggan New economic Press, An Free (1957) reality. A of Downs magic York The New (2011) Scribner, advertising. R hypothesis. Dawkins informative astonishing The and (1995) contributions F Crick campaign with competition Political York (2004) New S Knopf, work. can! Coate to we Yes Back (2011) (2011) P W Whiteley Clinton M, Stewart Press, D, Sanders University J, Toronto Reifler T, choices. Scotto political A, Kornberg Making HD, Clarke (2009) York TJ New Scotto Schuster, and A, Simon Kornberg Debs. and HD, Taft Roosevelt, Clarke York Wilson, New 1912: Knopf, (2004) motivations, power. J of candidates, Chace passage model: The (2012) voting RA multidimensional Caro the of Robustness (1985) RL York York New New Calvert Press, Harper, Matins Backlash. St. (2010) superpower. W York a Bunch New of Schuster, Suicide and (2011) Simon P fire. Buchanan of Pillar (1998) T office Branch of theory the elections: of model spatial the in voting Probabilistic (2005) J Duggan Games J, competition. Banks valence and platform with candidate. Elections favored (2009) a E Mesquita with de model Bueno S, Downsian Ashworth a in equilibrium Mixed (2002) T Palfrey Am E, elections. congressional Aragones in positions Candidate (2001) C Stewart JM, Snyder S, 2008 Ansolabehere 2004, studies. election national American rs,AnArbor Ann Press, 98:267Ð293 oga,NwYork New Longman, Norman Press, Oklahoma of University 93:1399Ð1414 84:1103Ð1126 Rev Sci Polit 12:417Ð435 Model 51:490Ð522 Oxford Behav Press, University Oxford economy. political of handbook Oxford The (eds) 2:772Ð804 Assoc Econ Eur J 30:450Ð461 Stud Elect 2008. America in choice electoral and politics Valence Toronto 29:69Ð85 Sci Polit J Am convergence. and uncertainty Heidelberg Springer, Banks. decisions: strategic Jeffrey and of choice honor Social in (eds) J essays Duggan D, Austen-Smith In: candidates. motivated 67:191Ð216 Behav Econ 103:131Ð161 Theory Econ J 45:136Ð159 Sci Polit J newamerica.net www. 125 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1656 1655 1654 1653 1652 1651 1650 1649 1648 1647 1646 1645 1644 1643 1642 1641 1640 1639 1638 1637 1636 1635 1634 1633 1632 1631 1630 1629 1628 1627 EDITOR’S1626 1625 1624 1623 1622 1621 1620 1619 1618 1617 PROOF1616 1615 1614 1613 1612 1611 enE(09 oe ffrsgtdvtn.A oi c 53:36Ð54 Sci Polit J Am and voting. far-sighted voting of strategic model equilibrium: A an (2009) three’s E company, Penn Two’s (2009) Haven New M Press, Ting University Yale JM, nations. Snyder of J, decline and Patty rise York The New (1982) Press, M Bloomsbury divergence. Olson York great New The Holt, (2012) endangerment. T Reckless Noah (2011) engagement: J civic Rosner G, and Morgenson Personality (2010) MA Seligson Rev D, Sci Canache Polit Haven New MV, Am Press, Hibbing University US. Yale JJ, economy. the unstable Mondak in an Stabilizing realignment (1986) partisan H Minsky and Activists (2003) York New N Basic, Schofield nots. have G, the and Miller York haves New Polit The Penguin, (2010) nation. J B right finance. Milanovic The (2004) campaign A Wooldridge and J, advantages Micklethwaite incumbency 57:155Ð contests, Behav Econ Electoral Games (2008) elections. large A in Meirowitz voting of theory Cambridge A Press, (2006) MIT J America. Polarized Patty (2006) RD, H McKelvey Rosenthal New K, Classics, Poole Library N, writings. McCarty Madison: (ed) J Rakove In: X. Federalist [1787]) York (1999 New J Basic, Madison thrift. York Against New (2011) Harper, J promise. Livingston of Land (2012) York M New Twelve, Lind lost. Republic, (2011) L Amer- Lessig among extension conflict and commitment Ithaca party Press, polarization, University Party (2010) Cornell al circle. et his GC Ithaca and Layman Bolingbroke Press, University [1968]) London Cornell (1992 Macmillan, radicalism. I money. bourgeois Kramnick and and Macmillan, interest Republicanism employment, (1990) 2. of I vol theory Kramnick general money, The of (1936) 26:755Ð769 JM theory Rev Keynes Yale self-sufficiency. applied National The (1933) JM In: Keynes money. on treatise A London. Macmillan, (1930b) 1. vol JM money, Keynes of theory pure The In: money. on Oxford treatise Press, A University (1930a) Oxford JM fear. Keynes from Oxford Freedom Press, (1999) University DM Oxford Kennedy enlightenment. 39:41Ð57 Democratic J (2012) Econ J competition. Israel in York Stability New Twelve, (1929) great. H not Hotelling partici- is voter God and (2007) Press, spending C University campaign Hitchens platforms, Oxford Policy capitalism. (2008) of C Martinelli crises D, the Levine H, and Herrera capital of back enigma its The turned (2011) richer—and rich D the Harvey made Washington how Politics: (2010) P democ- Pierson American JS, of Hacker erosion the and revolution republican Cambridge the Center: Press, (2006) MIT P groups. Pierson interest Econ JS, Special Hacker Rev (2001) politics. E interest Helpman GM, special Grossman and competition Electoral 90:303Ð315 (1996) Sci Polit E J Helpman Am GM, supermajorities. Buying Grossman (1996) J foundations. Snyder moral different T, on Groseclose rely conservatives and Lawrence Liberals Press, (2009) Kansas B of Nosek University J, ring. Haidt the J, in Graham hats Four (2008) LL Gould 126 utcniaeeetos oi c 4:251Ð278 Sci Polit J Q elections. multicandidate Rev Sci Polit Am behavior. political on effects trait of study 104:85Ð110 the for framework integrative an 97:245Ð260 70:681Ð699 180 York 104:324Ð346 Rev Sci Polit Am activists. party ican 6 vol Writings, Collected Keynes, in Reprinted London. 5 vol Writings, Collected Keynes, in Reprinted 92:501Ð513 Econ Public J pation. Oxford York New Schuster, and Simon class. middle the on Haven New Press, University Yale racy. 63:265Ð286 Stud 96:1029Ð1046 Psychol Soc Pers J .ShimnadN Schofield N. and Schnidman E. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1702 1701 1700 1699 1698 1697 1696 1695 1694 1693 1692 1691 1690 1689 1688 1687 1686 1685 1684 1683 1682 1681 1680 1679 1678 1677 1676 1675 1674 1673 EDITOR’S1672 1671 1670 1669 1668 1667 1666 1665 1664 1663 PROOF1662 1661 1660 1659 1658 1657 ita 17)Cniae ihplc rfrne:adnmcmdl cnTer 14:180Ð Theory Econ J model. dynamic a preferences: policy with Candidates (1977) D Wittman University Yale Wirtschaftens. und Handelns Des Theorie Nationalökonomie: (1940) University L The Mises Von [2007] London. Paul, Kegan & Routledge Haven serfdom. New to Press, road University The Yale (1944) Oxford crisis. F Press, Hayek dollar Von Clarendon and politics. Gold Electoral 57:368Ð377 (1960) (ed) Rev D R Sci Kavanagh Triffin Polit In: Am politics. Valence competition. (1992) party Norton, D and Stokes future. models our Spatial endangers (1963) society D Stokes divided today’s how inequality: of Price (2012) conservatism. York York JE New New republican Regnery, Stiglitz House, America. of Random After dream? making (2011) American M re the Steyn stole the Who and (2012) H party Smith tea The (2010) US: V York the New Williamson in Harper, T, democracy. realignments and Skocpol political socialism, Capitalism, and (1942) elections J Critical 51:518Ð Schumpeter (2003) Sci Polit A J Martin Am G, US. Miller the N, in Schofield coalitions activist and Elections (2007) Berlin G Springer, Miller chaos. convergent or N, Leadership for Schofield (2012) conditions M sufficient Gallego and N, Schofield necessary theorem: elec- voter of mean dynamics The the (2007) and N stokes Downs, Schofield (2011) P Whiteley MC, Stewart Random HD, prosperity. Clarke and D, virtue Sanders American reawakening York civilization: New of Knopf, price future. The America’s (2011) and economy JD next Sachs remak- the fundamentally Aftershock: is (2010) movement RB party Reich tea the how hell: as Mad York (2010) New D Schuster, Schoen and S, Simon Rasmussen and future. Simon Zero-sum us. (2011) unites G and Rachman divides religion how grace: Oxford American Press, (2010) University DE Oxford Campbell unbound. 307:19Ð22 RD, executive Economist Putnam The revisited. (2011) London enemies A Paul, its Vermeule Kegan and EA, and society Posner Routledge open enemies. The its (2008) and K government society Popper activist open politics: The (1945) American K of Popper transformation The (2007) (eds) T Skocpol P, Pierson Politics Electoral US in Leadership and Gridlock of Quandaries 189 Action Human as Trans. [1949] Haven. New Press, Chicago Press, Chicago of York New Oxford Press, University Oxford 51:217Ð240 Stud Polit 1860Ð2000. 531 74(3):965Ð980 Stud Econ Rev equilibrium. 41:287Ð314 Sci Polit J Br choice. toral York New House, York New Harper, system. two-party our ing York New Schuster, Princeton Press, University Princeton conservatism. of rise the and 127 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 DOI (eds.), al. et Schofield N. url: e-mail: Valencia, Spain Valencia, of University 46022 Aplicada, s/n, Tarongers Economía Av. de Departamento Economics, of Faculty hs fESA 95. E.S.A. of those 1 the regards As Spain. of criteria Bank Accounts fi- the National public the and comparable subsectors, Eurostat governmental internationally the by the breakdown OECD, the on by based provided be nance 2008Ð2011. will and 2000Ð2007 analysis periods empirical the behavior this over The undisciplined governments whether fiscally sub-central examines of any chapter part by the decentralization the paralleled on eighties, fiscal been mid has and the decentralization political since to of drive Spain process in place intense problems taken serious an has experiencing Because being respect. is gov- that this present central, country at as The borrowing a well countries. Spain, government as developed in sub-central, the borrowing western of ernment evolution 2008, all quantitative almost the of on in fall focuses turmoil chapter the in in been erupted has sector crisis financial the Since Introduction 1 Toboso Fernando 2000–2011 Period the over Spain in Crisis Debt and Governments Sub-central ofrne swl saoyosrfre,frhligm ocnetaeo htw l consider all ( we Toboso what F. on concentrate the to issues. in me key participants helping the thank for be I referees, to Spain. anonymous Political Baiona, as on 2012, well Conference 2Ð4, as International May conference, the in at held presented Institutions was and chapter Economy this of version previous A ae,a eaegigt e loi h aeo pi.O ore pi enjoys Spain course, Of Spain. most of in case debt the relations in total intergovernmental also of regarding see characteristics half organizational to than singular going more some are owing we by holders, are as debt usually cases, important governments most central and debt, the local public state, far central, Concerning sub-sectors: funds. four security into social down broken be to sector government ...9 a eetybe pae oSNA 09 hs rtraaeflycmail with compatible fully are criteria These 2009. S.N.A. to updated been recently has 93 S.N.A. www.uv.es/ftoboso 10.1007/978-3-642-35239-3_6 [email protected] B ) dacsi oiia Economy Political in Advances pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , , 1 eur the require 129 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 o eadesdete eas h mato hsapc ngoa ecnae is percentages global hand. on at aspect purpose this research will of present subsectors the impact the among the for in consolidated because irrelevant addressed internally either be figures addressed to are be issue how not an of not ear- issue is or grants this The conditional paper. But from example. financed government for be sub-central grants, may each marked expenditures of power these type regulatory of each of such Some regarding degree enjoy. aspects have exact qualitative governments the relevant discretion or of many expenditure, degree account the into example, take for not us, do they because mentioned. tion already reasons the public for their parliaments such with regional-state countries, as federal classified why in being are is figures states finance That as treated 2010. already cent are in per executives and 70 expenditures increased reached sub-central also already have expenditure total sub-central powers of in regulatory share autonomous their their and as substantially actors policy are key governments now regional-state Particularly, country. decentralized quite fiscally and eea ulcsrie hs aaeethsbe xenlzdt rvt rpbi enterprises, public or private to externalized been has management whose services public several ( ooosyejy o vnmr ulcplc epniiiista h ädrin Länder Germany. as the such than country au- responsibilities federal politi- government policy traditional own regional-state a public Each more their even 1982. in now to enjoys outset 1979 tonomously the from processes from electoral created and regional- significant were cal Seventeen executives of 1978. policy and in a parliaments and transition state 2007 democratic till since growth decentralization economic political rapid relatively a both controls enced legal including paper, the in governments. addressed sub-central be by to borrowing have on will that issues debt and 130 5 4 3 2 et n21.Asgicn eouino h oe otx o uttecentrally the since place just taken not also tax, has to government 1980s. power regional-state the each the to of govern- revenue, devolution regional-state tax for significant collected income A total 2011. of in cent per ments grants have 40 intergovernmental than reforms with less autonomy, several representing financial already their disposition, autonomous increase governments’ other to and place sub-central powers taken 2011. tax also at in Concerning cent income 2011. in per of cent 45 per sources total about 54 of about to cent to per 1983 13 1983 in from in went expenditures expenditures public total in of share governments’ cent Sub-central per 87 from went tt eeso oenet ntecs fsbetr,db sudb oilsecurity social by issued debt subsectors, of case the In government. of levels state uiiaiyo eae ulcetrrss si stpclydn hnelaborating when done typically is it statistics. as each Accounts and by enterprises, National municipal created public agencies by related governmental etc or other debt, by municipality revenue, as spending, well the as include governments, provincial used be will that figures n oooadSosn ( Scorsone and Toboso and 2006a) ifrn su htwl o eadesdi h ae sta nSana lehr hr are there elsewhere as Spain in that is paper the in addressed be not will (2010 that Timofeev issue and different Martinez-Vazquez A see problems measuring other and these On ( Toboso see this On ( Toboso see comparison this On fcus,alteefiue ersn og siaeo oiia decentraliza- political of estimate rough a represent figures these all course, Of ti otyt eto eeas htSanrpeet onr hthsexperi- has that country a represents Spain that also here mention to worthy is It 3 l hs grsarayidct htSanhsbcm politically a become has Spain that indicate already figures these All ,Tbs n croe( Scorsone and Toboso 2005), 5 ,(2006b 2006a), h aecmet pl otecnrladregional- and central the to apply comments same The 2010 ). )and( 2010 4 2 oa oenet’pbi finances public governments’ Local eta-aae ulcexpenditure public Central-managed 2010 ). ). ,Toboso ), .Toboso F. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 atr aeiflecdteeouino eta h ifrn ir fgovernment. of impact tiers these different of the at each debt which of in evolution magnitude the relative influenced the have the factors to research numbers econometric future assigning for of political leaves behavioral task also and chapter details The regulatory irrelevant. sector are public practices, other some as well as limits, eti pi hnfcoslne opltcladficldcnrlzto arrange- decentralization fiscal and political Sects. to sub-central in linked and stressed central factors As of ments. than evolution Spain the in explaining for debt relevant Stabil- exceed more European the factors to in seem Spain settled Section leading deficit Path. public factors Growth on and key limits ity as the sector, period second financial expenditure this the public during in in increase bailouts the for central counter-cyclical including needed initial governments, the the all at and by mainly recession economic adopted 2007, current measures since the debt to points and experienced also deficit has It level. Spain public that in show increase will considerable section sub-central a This by crash. debt financial issue said to the and followed over deficit place in Sect. in incur In details to governments. legal ability the including concerning Spain, or- period in multilevel the exists the that of section government aspects this key of In regarding crash. ganization singularities financial world some evolution the emphasizes the before paper show debt the central to as order well in as using sub-central figures case of finance Spanish the Sect. public on multilevel In 1996Ð2007 paper. Accounts period present National the the for in conducted investigated is analysis being empirical topic the to contributions eral Sect. in remarks, introductory Following Accounts. National in done is it as subsectors, among idated aaae ontpoiedsgeae grsfrsca euiyfnsdb o l countries. all OECD for The debt percentages. funds based security social IMF for with figures comparable desegregated provide strictly not not do are databases follow that percentages the fe tesdi h ieaue odnadothers and are Rodden we that literature. evolution debt the this central, in influencing stressed as factors often main well the as are which sub-central, investigate with briefly need happened has what examining Before Evolution the Influencing Factors Main the Are What 2 who government central the is it as Spain. in level debt central this the upon at decides included appear will funds 2000Ð2011 Period the over Spain in Crisis Debt and Governments Sub-central 7 rules. Account activ- government” National “general following the indirectly 6 of but the part funds) to as security belong accounted (central-regional-local-social directly organizations ities not these are When activities their etc. sector partnerships, business public-private to consortiums, mixed to e odn( Rodden See why is That IMF. the by provided statistics finance public multilevel regarding case the not is This hrfr,t copihisproe h rsn hpe sognzda follows. as organized is chapter present the purpose, its accomplish to Therefore, nteLiterature? Emphasized the Being in Are that Debt Sub-central of 2002 ,Rde (2006 Rodden ( 2003), al. et Rodden (2002), Wibbels and Rodden ), 3 h aei oea ead ote20Ð01pro that period 2008Ð2011 the to regards as done is same the 3 4 and ocue yepaiigta cnmcconditions economic that emphasizing by concludes 6 iue eern o“l oenet”aeconsol- are governments” “all to referring Figures h bv ttmn sntt li htdebt that claim to not is statement above the 4, 1 reyrve oeo hs e gen- key these of some review briefly I 7 aemd usadn contribu- outstanding made have ). 2 131 an Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 flclpbi od edt niiulcnrsmno eaost r oepn own-district expand to try spending. to aggregate financing senators restrain national to or try by they congressmen created that individual time incentives same to the the lead how at spending goods on public emphasize local All of Senate. and Congress USA eevdcnieal teto o nteltrtr,a ela h aacdbudget balanced the as well as literature, the in too attention considerable received countries. these all quality the for and factors precedent, relevant bailout seem any reporting of fiscal existence the of region, or state characteristics country, for institutional the circumstances of Specific all spending. under government superior sub-central seems disciplining arrangement institutional single no that ao n eeon(2007 Ferejohn and Baron hsi lorfre oa h omnpo problem. pool exists. common constituents the regional as over to taxes referred own also raising is to This are obligation sources financial no by external re-election and if ensure spending available to for politicians opportunities regional sub-central and of local part finding the on usually efforts incentive or strong said party in The the executive. results than and parliament coalition central or the party controlling political coalition different a to belong government central borrowing or borrowing enforced. to credibly access not unrestricted are have overspend limitations they to if incentive deficits the credibil- greater have lack incur may bailout and governments for sub-central demands at which agents reject in political to government ity, government to of central systems possibility by multi-tiered a commitment those is In the paribus. there ceteris pro- that out, government leaders bailed multilevel government be key for sub-central A setting for institutional literature. expectations the of vides whether strand be this to to seems according aspect governments sub-central of discipline sa eain r ieyt euti ectba ndcnrlzdplc-aigwt otbudgets soft with policy-making decentralized in bias deficit a in constrains. result to likely are relations fiscal h oeyugtfr“h tt-ein rmtecmo ol h better. the pool, ( state. common or the region from the state-region” in “the general for in get people you the more and The groups lobby parties, all of behavior n oilcrusacsa ela atcpnsdfe omc rmcs ocase? to case from much so differ participants as is: well however economic as question political, circumstances key if situation social The polity. each and and in the decisive in norms are as arrangements social well institutional informal as which economy as the This they well in debt. matter while as of values, levels arrangements, countries unsustainable institutional-legal some generating thus that in others means manner in financing conservative deficit fiscally gov- on sub-central rely a why in explaining in behave crucial ernments factors among these performance fac- in consider key differences They explaining are for federations. ones) systems only federal the necessarily characterizing recur- (not have details tors They institutional perspective. that institutional comparative stressed a rently from issues these on tions 132 10 9 8 2006 rmasml f3 onre,Ml ( Melo countries, 30 of sample a From (2005 North See elyadCae(2003 Coate and Besley mn hs ragmns h fetvns fdb n pniglmt has limits spending and debt of effectiveness the arrangements, these Among h netv a epriual eeati oiia gnscnrligasub- a controlling agents political if relevant particularly be may incentive The fiscal weaken necessarily not does se per decentralization fiscal and Political sn apeo 3cutisoe h eid18Ð00 lkao n Singh and Plekhanov 1982Ð2000, period the over countries 43 of sample a Using on osmlrapcsa e atr nmn ae.Teeatosconclude authors These cases. many in factors key as aspects similar to point ) )orOstron( ,Kih (2006, Knight ), , 2009 1990)and( drs hs omntxpo susmil eern othe to referring mainly issues tax-pool common these address ) 2000 2005). 9 2008 hw vdneidctn htintergovernmental that indicating evidence shows ) ,ImnadRbned( Rubinfeld and Inman ), 10 hsba a rvnthe driven may bias This 1997 ,Bqr(2002 Baqir ), .Toboso F. ,or ), 8 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 t r esal ohl eta oeneti h tep oipeetcounter- implement to attempt the in government central help policies. to cyclical able less are its years subsequent the over impact little find they since. again and ceilings such of adoption hnasnl oiia at otostegvrosi n h tt os hnwhen than house divided. state the was and control for governorship party faster the particularly controls made party relevant, were political were Adjustments single a years. seemed when subsequent factors in political adjustments the by deficits that deficits explaining found public affected also negatively He and significantly States. 1980s late the during US safis reaction. first a as sspnldt rmte16stl h 90 ihsc ocuin ose tal. et Kousser conclusion. a such ( with Shadbegian 1990s time. the till of 1960s ( periods the several from for data states panel have limitations between uses authors growth such several revenue without example, or and for expenditure states with in US difference not the significant is to no this regards detected But is with it as bodies. and case, management growth technically the independent of defined always by rate well non-compliance spending are detect the limits to if lowered easy particularly have periods, to boom seem However, the usual. limits during as these conclusion sole occasions a show many not in do data panel upon made Studies rule. 2000Ð2011 Period the over Spain in Crisis Debt and Governments Sub-central 15 14 13 12 11 erature. if so. rules do ceiling to said likely the are evade leaders to usu- way political debt government a this central represents Though it government. authorization, of central levels needs government” all ally “general at by called passed as so are the well budgets to of whose as out only entities enterprises, are and apply that public analysis agencies typically special commercial depth limits by public in debt some issued an that debt for forgotten excluding account be debt, into not guaranteed taken must be It to assessment. have sound that country to country ato ulcdfiisaddb ees lehie ta.( al. im- et negative Bloechliger a levels. produce debt that always and revenues recessions deficits tax that public doubt on on little impact pact be negative can the there and results, expenditures also other in increase matic nageaedmn n rwhta curda euto h ol nnilcrash. decline financial the world cushion the of to result attempt a an as in occurred 2008 that growth during and implemented demand were aggregate that on economies major the of r oiial oesniieo adtdad osqety oedfcl obe to difficult more consequently, expenditures and, current while mandated constraints, or changed. budget sensitive of more face the politically for- in the are curtail as to expenditures easier current is than mer more investment public affect often recessions 2008 e loAli-ur (2011). Allain-Dupré also See ( Rodden and Wibbels also See research. of line this on contribution ( Padoan cited, much seminal, a is (1979) Barro ( Levinson See fcus,sbcnrlgvrmnswt titblne ugtrlso etlim- debt or rules budget balanced strict with governments sub-central course, Of h osqecso cnmccce r lociia atr xmndi h lit- the in examined factors critical also are cycles economic of consequences The netgtscagswti ie tt,ntaogsae,floigthe following states, among not state, given a within changes investigates ) 12 14 2009 eesosuulyla odlbrt oneccia pnigmeasures spending countercyclical deliberate to lead usually Recessions oeb ( Poterba ,freape netgtstesz n opsto fteficlsiuu packages stimulus fiscal the of composition and size the investigates example, for ), ,FtsadMhv( Mihov and Fatás 1998), 11 13 huh gi,mn xetosadpriuaiiseitfrom exist particularities and exceptions many again, Though, fw locnie mato pnigdrvdfo h auto- the from derived spending on impact consider also we If 1994 15 ,freape hwdhwteeooi onun in downturns economic the how showed example, for ), 2006 ). 2006 n oe( Rose and ) 2006). 2010a , 2010b hwthat show ) 1996) 133 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 oa usadn eti eylwpooto vrtepro.Ruhyspeaking, Roughly period. the over proportion low very in debt outstanding total h alu rbe nteAgniacs,seJnse l ( al. et Jones ( see al. case, et Argentina Tommasi the in problem bailout the ino et o copihn hsrsac ups,tenx eto oue on focuses section next the evolu- period. purpose, the 1996Ð2007 regarding research the turn this for similar accomplishing statistics a For whether by debt. see paralleled and of is Spain tion conditions in economic debt investigate public in quantitatively central, turn to as indi- the well attempts as as paper different sub-central, This of quite evolution is section. the in paper introductory found present the attempts the similar in of other cated purpose with an The refuted literature. not or published is contradicted factors the burdens definitive often the are debt of find valid in to are role change attempting that the and exercises to econometric level Moreover, the compared task. determining as easy played in details cycle role organizational economic relative informal fiscal the the and with knowing dealing formal However, scholars specific governments. of by central attention sub the by attracting behavior are conditions economic to debt. greater in policies incurring budgetary then periods, active cantons, more Swiss these centralized implement enjoying In than to Cantons debt. and tend their incomes decentralization public political of than greater evolution greater regards the appear as affect needs share. spending Cantons to suddenly to among seem governments differences autonomy all recession political for economic to no revenues of has enough phases government are in of There However, organization debt. multilevel on growth impact economic relevant 1984 of between periods cantons in Swiss that the ( Vatter in and registered Freitag evolution 2000, the and on based and research engnrtdadn alu rbe a xse vrtepro.Mroe,Ta- Moreover, period. the over existed has problem bailout ble no and generated been ai-mrcncutisfrexample, for countries Latin-American esrda ecnaeo D,a Table as GDP, of properly percentage if considered a here as period 2000Ð2007 measured the over Spain in government debt? of public levels central cause or 2008 sub-central in to initiated regards in recession as evolved sever point debt the turning and public significant crash any has financial How world government. the of Did levels evolution Spain? whether negative sub-central investigating a at for by debt paralleled study public was case political of drive suitable decentralization quasi-federal a fiscal new and seems political it a this 1983, developing to in 1978 changes from system major experienced Spain As Spain in Debt Central and Sub-central of Evolution The 3 134 16 e az (2000 Tanzi See hr utb odutta ohognztoa atr swl sfcoslinked factors as well as factors organizational both that doubt no be must There of line this On direction. opposite the on work just growth economic of Periods ti vdn htpbi ect n usadn ulcdb eerdcda all at reduced were debt public outstanding and deficits public that evident is It eoeteWrdFnnilCrash Financial World the Before 1 lososta ohrgoa n oa oenet aecnrbtdt the to contributed have governments local and regional both that shows also 2001 o naayi fteptnilmcocnmcpolm.Fra nlssof analysis an For problems. macroeconomic potential the of analysis an for ) rSig n oms ( Tommasi and Saiegh or ) 2008 ,freape rvd miia eut showing results empirical provide example, for ), 16 1999 orlvn areooi itrinhas distortion macroeconomic relevant no 1 ). eel.Cnrr otecs nsome in case the to Contrary reveals. 2000 ,SnuntiadTmsi(2004) Tomassi and Sanguinetti ), .Toboso F. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 rcs fpltcldcnrlzto a eyintense, very was decentralization political of process a eosre nFig. in observed be can crash, been financial current always the has to Spain prior in period the debt over Table outstanding level as Total average one cent. EU that the per and than 3 GDP lower Spanish around the of ones by cent local per debt by 6 outstanding around 1990s, local remained mid and has the governments regional Since regional than debt. more outstanding times to did three governments about contributed government central the Spain. of Bank and domain Eurostat public OECD, Source: the Accounts). in National Figures Euros. of Millions and GDP of (% 1 2000Ð2011 Table Period the over Spain in Crisis Debt and Governments Sub-central 18 17 r beto o ul nosd h miia vdnetknfo h omrones former the rules from taken fiscal evidence these empirical where the countries endorsed, fully with not compared or If expen- absent total settled. are annual and/or been to deficit ceiling also top public have explicit the cases, ditures some to In limits debt. explicit partic- public outstanding operate, containing the policies rules fiscal fiscal their adopting which countries by developed in ularly many context decades, the two changed past significantly the have along that, well-known is effectively It period. to crash. financial able world be the will after in required system addressed strategies political consolidation be multilevel fiscal to new the to this issue respond different if is A problem section government. debt next of or the levels fiscal central relevant or any sub-central by at paralleled not was decentralization political rwhoe hs er fo 49prcn fSaihGPi 94t bu 0preton percent 40 about to 1974 in GDP rapid Spanish and significant of a cent experienced per has also 24.9 2000s. we Spain the (from If over in average years percent. expenditure 33.1 those public managed over total they growth 2006 that In account expenditures. into public take total of percent 2.9 managed government. of level central the at included is Funds Security u-eta osDb s%o oa et1.41.62.523.65 22.15 90424 19.56 86639 15.84 76148 Debt Total of % 59267 as Debt Govs Sub-Central Euros Millions in Debt Gov Central Euros Millions in Debt Govs Sub-central zone) (Euro AVERAGE EU Spain in GOVERNMENTS ALL Government Central Governments Local Governments Regional Governments Sub-central he er fe h prvlo h 98dmcai osiuin einlgvrmnsonly governments regional Constitution, democratic 1978 the of approval the after years Social by Three issued Debt protocol. excessive-deficit European the to according made are figures All rpial,ti vlto fpbi etboe onb eeso government of levels by down broken debt public of evolution this Graphically, eea xlntosmyhl on oad htocre nSandrn this during Spain in occurred what towards point help may explanations Several eea oenetdb nSanbfr h ol nnilcahboe onb levels by down broken crash financial world the before Spain in debt government General 1 loindicates. also 1. 17 fw aeit osdrto htoe hs er,the years, those over that consideration into take we If 176329 046291883 304416 312994 314766 114. 9736.1 27.7 39.7 31.0 46.2 37.1 61.1 51.5 0020 062007 2006 2004 2000 996. 8566.3 68.5 69.8 69.9 . . . 2.8 5.7 8.5 2.8 5.9 8.7 2.9 6.2 9.1 3.3 6.3 9.6 18 tcnb ttdta this that stated be can it 135 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 iln,btmil uigproswt nuheooi growth. economic enough dis- with budgetary periods maintain to during helping mainly in mechanisms but useful cipline, as rules these show to tend domain public the in Figures Spain. of Bank and Eurostat Source: Accounts). National Euros. of Millions and the GDP of (% levels by down broken crash financial world the before Spain in debt 1 Fig. 136 21 20 19 eortctransition. democratic aspects. settled crucial procedures also specific seem deviations of the potential effectiveness or with the aspects dealing enforcement determines for and greatly Monitoring controls rules. and these limitations of design specific tansadrqieet eestldb a o bann uhawitnautho- written a such con- obtaining specific for and for Law debt, required by issue been settled rization. and were always credit requirements has term and Minister long straints Finance access Central to governments the regional of approval written government. a of levels GDP, across burdens cycle of debt economic terms relative the in restrain and to debt Controls act central comparison. to not economic sub seemed relevant did of most this percentage the period, relative is boom the which the in public over increase in registered any was increase cause Euros an millions Thought financial in started. the measured recession once debt economic controls pe- associated sufficient the its been over and governments not crash have local rules by legal as these well However, as riod. regional by displayed discipline fiscal rnil,tog oeflxblt a nrdcdi h plcto fti rnil n20,and 2006, in principle this of application the 2009. in in on introduced latter was flexibility some though principle, have laws These finances. public since. local occasions on several 38/1988 in Law reformed been and 7/1985 Law Finances, Governments in w aswr asdi h eta alaetsnetevr einn of beginning very the since Parliament disposi- Central As governments’ the income. in local current passed at annual were sources government Laws financial regional two framing tion, each rules be of legal must % the credit 25 regarding all must than interests) First, higher and place. (capital be repayment in not annual been second, always And investment. have to requirement dedicated two year), a than n20,broigatvte yRgoa oenet eeas ikdt h aacdbudget balanced the to linked also were governments Regional by Regional activities on borrowing 2001, 8/1980 In Law Organic were: parliament central the in passed initially Law These ( Kumar and Debrun See oevr oecnrllglrlsfrbroighv laseitdsince existed always have borrowing for rules legal central some Moreover, huhteerlsadcnrl aeas encagdo eea occasions, several on changed been also have controls and rules these Though eea government General 21 npriua,frrgoa oenet ogtln emcei (longer credit term long get to governments regional for particular, In 20 2008 hrfr,te a aepae oei anann the maintaining in role a played have may they Therefore, ). 19 oee,the However, .Toboso F.

B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 fGPmr ahya nnttrsa ulcicm rmteErpa Union. European the from income public as terms net in percent year 1Ð1.2 each about figures). average more finance on GDP of 2001, ministry of until the receiving, to been according also 2006 has in 18.4 Spain 36.7 (from Moreover, to in years 1975 thirty revenue in about tax GDP in of Spanish points cent eighteen the per about in of increase registered an increase GDP: to the relation this from of come provided Most has resources. period income GDP financial the of public Spanish amount over the increasing registered period, an with this growth governments gov- over economic regional that of The levels fact significantly. the all grew to at also paid policies be public must financing out- Attention for the percent- ernment. available consequently, a Table been and, previous has as deficits in income and shown public public euros as reducing levels total in debt help in public also standing (both did period GDP), growth of the age over average on income hr emadln emdb a edne yteCnrlFnneMnse if Minister Finance Central the by denied issue be implemented. and fully may not credit debt are access plans term to these long deficit authorizations and cases, public all term eliminate reces- In short progressively required. of to also case Plans are in measures. deviations considered legislative explicitly such are in exceptions sions and governments, flexibility regional affecting some similar those though to as stability subject budgetary been to have regards governments in local obligations packages, regulatory these all of sult 22 as GDP, to from of 1996 went in percentage also deficit in public in significantly Sub-central ended show. reduced that figures period OECD were the levels over debt non-disciplined governments Sub-central sub-central next a of 2007. by the part political paralleled in the to been on emphasize drive not behavior to significant fiscal has going the decentralization are that fiscal show and we provided federalism as figures issue the key Therefore, a section. seems cycle economic The ( as known Finances is Public second Local Locales—LRHL initial The of 38/1988. two Law Law These Regulatory and effect. the 7/1985 in Law was were Constitution packages 1978 legislation the once transition, democratic 2000Ð2011 Period the over Spain in Crisis Debt and Governments Sub-central 24 23 ute euto o h eid20Ð04a h wlenwcutista nee h Uin EU period. the new entered the that for funds countries EU new the twelve of the most obtaining as are 2007Ð2014 2007 period and the 2004 for reduction further a previous the in incomes current of percent 30 exceed not must obtained year. 2013). amount in total is the that year): maximum, the years six ap- in exceptionally repaid and be to paid have 2009 be will the not that of authorizations could approval credit that extra the obligations for after obligations these ply months documenting some three for if had Law providers have and Decree government their urgent Royal Local pay contained 2008. Bud- to (which in for governments unpaid 5/2009 Law local remained Law General facilitate Decree to the Royal measures in and extraordinary settled mentioned) principles already basic Stability develops getary further (which 1463/2007 Decree ic 02 hssuc ficm sbcmn essgicn ntrso D,adhssuffered has and GDP, of terms in significant less becoming is income of source this 2002, Since along (repaid credit term short access freely to governments local for exists requirement extra An Royal through implemented were issues borrowing regarding reforms significant these of Two oevr tseseietta h pcaua nraergsee nttlpublic total in registered increase spectacular the that evident seems it Moreover, n2006. in −0.4 .Svrlrfrshv ae lc ic hs years. those since place taken have reforms Several ). 23 e euaoad a Haciendas las de Reguladora Ley nicesn mutof amount increasing An 1. ecn fGDP of percent −0.6 22 sare- a As 137 24 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 h ih ogtapbi eso.Testainlosee os fw osdrthe consider we if worse even looks Fig. situation that The unemployment pension. in increase having public huge populations a elderly get also of and to levels building banks, right expected saving the the and and current in banks regarding bubble affected needs seriously many the simi- are and the 2008, no Spain burst had of since finally has has case growth that and the economic sector GDP in regarding of issues record Key cent bad markets. per the international for 200 at Japan, near problems 2010. of financial burden of lar beginning debt the a since got particularly has example, situation, critical a such issued. previously enced bonds back interna- pay debt of to new even part issue and successfully the rates to interest on governments reasonable these uncertainty at of considerable ability the caused public over of investors has terms fi- tional in This severe the more debt. in been have and bailouts events, deficit few and a bankruptcies but recession, mention to of sector, nancial consequences the boom, the measures over those several and with reform banks, savings sector and financial bailed-out. banks already the recapitalize banks of and more case reorganize imply to the necessarily implemented is Growth taking being and as are Stability spending that financial European public measures the the other to in Several prior established (ESGP). registered limits Pact they the levels exceeding deficit then public crash, the particular, in triplicate countries, or European all double to with to generated lead together has recession governments, factors, economic national all singular by the other adopted some of measures severity counter-cyclical The initial 2008. events the dramatic of and some fall countries, the developed in around bubble erupted financial decade fis- major a and a than to political more up regarding after leading However, singularities irrelevant. the are that arrangements decentralization mean cal not does analysis above The Central as, well as Sub-central, with Happened Has What 4 138 noteErpa no raya ae18s hs hrdcneune aealso have limits consequences shared top These debt 1980s. and late deficit at was Treaty public fact, Union well-known European in the the pro- This, into introducing zone. clearly for the is argument in countries early partners main some since other a in reveals on registered case impacts zone debt negative Euro public significant to the ducing in belonging As increase countries Spain. sharp of of the case case 2010, the the in is course, as unions, of monetary increases, controls debt and deficit expectations. regarding debt worse external the also but banks) debt, and external enterprises public near households, just the (not by in debt solved external be is not account) higher will financial As problems future. the that investor from international compensation to idea no register- the (with is transmit figures Spain the spectacular that OECD Regarding a deficit government. to trade result, of according external ing, tiers a account all current As the at substitute. payments, place of a taking also balance as is but revenue come needed, public has are in expenditures sector decrease public export more dynamic and no more that and dropped has demand u hseouino etlvl ln osntepanwySanhsexperi- has Spain why explain not does alone levels debt of evolution this But debt and deficits public diminish significantly to able not were that countries In h motneo aigo h akteefo fetv sa ue n public and rules fiscal effective of thereof lack the or having of importance The eti pi fe h ol iaca Crash? Financial World the After Spain in Debt 2 hw.Ti eel o nyta internal that only not reveals This shows. .Toboso F. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 uhcnetdaece ste aeAAt ea rtesi 2006, in Brothers Leman to AAA gave they as agencies contested much poses. eidi hw nFig. in shown is Euros period millions the over thousand Spain 267 in debt than public more outstanding years. total adding four the 2001, said to dollars) in GDP US Spanish cent billions of 334 per greatest cent (about 52.1 the per caused 27.7 to has from 2007 2008 soared in in debt place public took central dras- that Outstanding the cycles main where impact. economic government as of in governments level point regional central turning with the tic years, at been four the has just to it in However added ex- contributors. debt euros also public millions has thousand Spanish debt 68 of governments’ about stock Sub-central with cent). increase per significant 72.1 a to perienced GDP of cent per 36.1 njs oryas ntecs fSan h oa ulcdb a obe nterms in Table doubled as has compared, debt are public years total 2007 the and 2011 Spain, if of GDP case Spanish the levels the debt In of public years. outstanding four affecting in problems just increases economic spectacular in in specific resulted the has of country, particularities each the with together 2008, for also plan bailout used Spain). the been in and banks finally Portugal, saving have and Ireland, the They for (Greece, Stability bailouts place. European or taken mentioned interventions have already justifying that the (ESGP) of Pact amendments Growth several the to lead that OECD Statistics the of by Institute provided National those Spanish with the consistent from fully domain are public the in Figures Source: ployed). 2 Fig. 2000Ð2011 Period the over Spain in Crisis Debt and Governments Sub-central 25 ecti h onr omr hn3 e eto D.Rmme htteErpa tblt and Stability European the that Remember GDP. of cent per 30 than more to country the in deficit etydwgaetertnsfrcnrladsbcnrlgvrmn eti Spain, in debt government sub-central and central for ratings the downgrade rently swl st,freape h orbnsrsudi rln,wihaone h nulpublic annual the amounted which Ireland, in rescued banks four the example, for to, as well As hr utb osrrs htMoys ic n &,tog hyaevery are they though S&P, and Fitch Moody’s, that surprise no be must There rpial,ti vlto fdb ylvl fgvrmn vrte2007Ð2010 the over government of levels by debt of evolution this Graphically, nalteecutis h xrml rtcleooi iuto rae since created situation economic critical extremely the countries, these all In nmlyeti pi pretg fatv ouainadnme fpol unem- people of number and population active of (percentage Spain in Unemployment hc nldsas rvosyasfrcmaaiepur- comparative for years previous also includes which 3, 2 hw (from shows 25

recur- B&W IN PRINT 139 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 ure f21,a rga n ayohraayt eurnl aldattention last called the recurrently at analysts started other have many and to Krugman seemed as that 2010, recovery of economic quarter weak dam- already finally have particularly the that 2010, measures aged since policy contractive implemented represent that being countries, either measures EU surprise several austerity in no all be term must short There the deficits. in public negatively which reducing down- debt, of as this of purpose debt repaying the issuing in consequence affects default for main of paid risk be a estimated to Obviously, greater rates imply course. interest grades in of increase countries the is other downgrading several this in as domain well public as the in Figures Spain. of Bank and Eurostat Source: GDP). the of (% levels by 3 Fig. of Bank and Eurostat OECD, Source: Accounts). National domain Euros. public the of in Millions Figures Spain. and GDP of (% 2 Table 140 rwhPc eurdi ob ne he e eta eea ue e sd h xetosalso exceptions the aside let rule, general a as cent per three under be settled. to it required Pact Growth u-eta osDb s%o oa et2.52.72.32.823.87 24.08 559459 488245 22.23 439420 175502 26.17 322584 154891 291883 125662 23.65 114400 Debt Total of % as Debt Govs Sub-Central 90424 Euros Millions in Debt Gov Central Euros Millions in Debt Govs Sub-central zone) (Euro AVERAGE EU Spain in GOVERNMENTS ALL Government Central Governments Local Governments Regional Governments Sub-central eea oenetdb nSanbfr n fe h ol nnilcahboe down broken crash financial world the after and before Spain in debt government General eea oenetdb nSanatrtewrdfiaca rs rkndw ylevels by down broken crash financial world the after Spain in debt government General 614. 396. 72.1 52.1 61.2 46.4 53.9 16.4 41.9 40.2 14.8 30.6 36.1 12.0 27.7 9.6 8.5 0720 0921 2011 2010 2009 2008 2007 637. 958. 87.2 85.3 79.5 70.0 66.3 . . . 1413.1 11.4 8.7 6.7 5.7 . . . . 3.3 3.4 3.3 2.9 2.8 .Toboso F.

B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 ic 07 ih27tosn ilosers(bu 3 ilosU olr)being dollars) US billions 334 (about terms euros absolute millions in thousand more increased 267 has with debt 2007, the since where government of level central the controls. issuing decentraliza- debt fiscal as in and introduced well political was as characterizing in change tion rules relevant increase serious basic no the substantial if extremely regarding even a the country Spain the by that in levels evident paralleled debt is been and in deficits it has public cent Therefore, 2008 per 2011). since 36.1 in experienced (from years cent recession four doubled per just Spain 72.1 in in that to governments stressed all 2007 has by chapter EU debt most the outstanding by debt, achieved total public be to not could regards estab- Path As Growth limits countries. and The Stability countries. European European the other in many lished concerning case the also been has 07algvrmnshdrgsee upu f( of surplus a registered had governments all 2007 ect n oa etatrtewrdfiaca rs.Ttlpbi ecti Spain in deficit public Total crash. financial that world straightforward the ( seems public after reached in it debt registered 2009, increase total spectacular and and the 2008 deficits for charged during deficits be effect cannot public 2007, regulations in regarding till these controls also GDP and were of regulations debt terms that As in and show. surprise no reduction provided is figures greater it the also average as experienced EU the Spain than in Spain key levels in a debt higher also played were period rates the growth along As revenue in role. tax mentioned have in as existed registered period, increase always the The over have chapter. debt that the of debt evolution this and influenced deficits positively doubt on no limits formal detailed The 2007. regulations strict if paribus. otherwise also ceteris be country, would enforced, the than and of easily settled more were needs default macroeconomic of the bor- risk to the their tend entering governments regards and articles these as devices that published excessively own likely many borrow is their it in to to controlled, ground left centrally not are common are governments debt. a activities rowing sub-central public is if of it evolution that stresses, the state chapter for to the irrelevant to as not are is fact, controls This debt In started. and arrangements crash deficit decentralization financial public fiscal and world and political turn- the regarding a singularities when the that 2008 that show eco- in say provided of place here impacts took figures The the point factors. ing that key indicates the over investigation Spain seem in conditions present debt nomic the public central, 2000Ð2011 as period well the as sub-central, of evolution the Regarding Remarks Concluding 5 contraction reveal. in available remains already economy figures the preliminary that as surprise 2012, public and no reduce Span- 2011 be to the in must expenditures pull it cutting government 2010, to of since able tiers deficits consumption all with internal and nor economy demand ish external neither With to. 2000Ð2011 Period the over Spain in Crisis Debt and Governments Sub-central oevr h nlsspoie ntecatras niae hti a enat been has it that indicates also chapter the in provided analysis the Moreover, nteSaihcs hsudsilndficlbhvo a o ae lc till place taken not has behavior fiscal undisciplined this case Spanish the In − . e eto D n20 n ( and 2008 in GDP of cent per 4.5 ) − 12prcn n20,weesin whereas 2009, in cent per 11.2 ) + . e eto D.Adthis And GDP. of cent per 1.9 ) 141 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 ilndbhvosfo apnn nteftr.Ftr oiia far antbe undis- reactions. cannot these chemical affairs of excludes political result that the Future predict chapter future. we present the as the in predicted happening in crash from financial nothing the behaviors is till ciplined there not least 2008, at in governments, undisciplined sub-central started fiscally of by part political paralleled the been that on not indicates behaviors has chapter Spain evolution the in overall in decentralization their fiscal provided attempt but and I analysis government did the regional nor Though also future, specific debt. the any stating for regarding of prediction by case any conclude extract the to me examine the attempt to let I among did Finally, not way though chapter. population no aspects, in present elderly key that also the of are of levels pension, expected purposes public by a the research the get in as to debt well right private the as of bubble having stock a firms, The after and sector. population building banks active the the households, in of which also cent 2008, place per since taking is- 22 GDP bursting key reach in to A registered unemployment rates. evolution lead negative interest country has extremely reasonable the the at in again debt is experienced new sue issuing of problems evolution successfully serious for the the chapter, 2010 explain Spain the since cannot that in evident alone mentioned is debt it As Zone, public problems. Euro singular the from in crisis suffer overall must the to Italy, in and Ireland than tugal, more increased have debt paper. above. and the mentioned in deficits one mentioned public the as in than countries others, situation restructuring some other in and no course, bailouts spending, to Of public lead first could central for being sector, in drastically needed financial registered the were expenditures increase revenues public the tax recession those where of including counteract period result to a a order In as well in recession. as reduced policies the Spain, stabilization of in in forces government engaged central the countries, recession evident, other economic financial was many subsequent crisis worst in and the as the world, as consider western soon im- the most As higher in what generated. the century by with past finances the again of relates public crisis government central of on adding level amount caused then central higher the the pact 2011, out- for by total in reason of added cent main stock debt the per the stressed, of to 16.4 previously dollars) As to US debt. 2007 billions public 106 in standing (about GDP euros Spanish millions going thousand 2007, of 85 since cent debt in per in increase chapter 8.5 spectacular whereas the a from 2011, governments, GDP, registered sub-central in also of Regarding have GDP terms they cent. Spanish In that per shows of years. 27.7 four cent was just per figure in 52.1 this Spain 2007 reached in has debt debt public public total central of stock the to added 142 ao P eeonJ(07 agiigadaed omto nlgsaue.A cnRev Econ Am legislatures. in formation agenda and Bargaining (2007) 110(6)1318Ð1354 J Econ Ferejohn Polit J DP, overspending. government Baron and Districting (2002) R Baqir OECD crisis. the from lessons investment: public of governance Multi-level (2011) D Allain-Dupré References 77(2):303Ð309 OECD 5, No papers, working development regional sSanhsbe ihihe sami otiuo,tgte ihGec,Por- Greece, with together contributor, main a as highlighted been has Spain As .Toboso F. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

690 689 688 687 686 685 684 683 682 681 680 679 678 677 676 675 674 673 672 671 670 669 668 667 666 665 664 663 662 661 EDITOR’S660 659 658 657 656 655 654 653 652 651 PROOF650 649 648 647 646 645 aehS oms 19)Wyi retn’ sa eeaims nfcet neigthe Entering inefficient? so federalism fiscal Argentina’s is 128(3):407Ð431 Why Choice (1999) Public cycle? M business Tommasi political the S, dampen in Saiegh rules fiscal management Do macroeconomic (2006) federalism: S of Rose fiction the Beyond (2002) budget E hard Wibbels of JA, challenge Rodden the and Decentralization (2003) (eds) J Litvack GS, Uni- Eskenland the Cambridge J, federalism. around Rodden fiscal performance of peril fiscal and promise and the paradox: grants Hamilton’s (2006) federalism: JA Rodden fiscal of politics. dilemas and The institutions (2002) budgetary of JA effects Rodden the crises: fiscal to regulated? responses be State borrowing (1994) government J Poterba subnational should How implications. (2006) long-term RB and Singh sustainability, AA, impact, Plekhanov crisis: the in policy Handbook Fiscal (eds) M (2009) Shirley PC C, Padoan Ménard In: analysis. institutional for framework A (2005) E Cam- actions. Ostron collective Press, for institutions University of evolution Princeton The commons. change. the Governing economic (1990) E of Ostron process the cross-country Understanding a (2005) relations: DC fiscal North intergovernmental and decentralization Fiscal (2000) RL J Melo Tax Natl Manag Financ states. Public revisited. the measures Decentralization from (2010) effectively A evidence Timofeev limits J, cycles: Martinez-Vazquez expenditure business and and tax budgets state Balanced can (1998) tolls: A TEL Levinson the whom For (2008) al et T Kousser funds: federal of distribution the and power, bargaining representation, Legislative (2008) B B Swedenborg Knight R, Congleton In: systems. federal in problems pool tax Common anal- an (2006) performance: B fiscal Knight and institutions, Politics, (2000) 11(4):43Ð64 M Perspect Tomassi Econ P, governments: J Sanguinetti federalism. MP, sub-national Rethinking Jones (1997) in DL discipline Rubinfeld RP, fiscal Inman and Econ Decentralization Public (2008) J states. A US Vatter the M, in rules Freitag la fiscal of de effects macroeconomic fortalecimiento The el (2006) I en Mihov fiscales A, Fatás instituciones las de papel El (2008) policy and MS impact Kumar crisis: X, economic the Debrun and governments Sub-central WP (2010b) OECD al crisis. et of times C in Bloechliger government of levels po- across a policy Fiscal goods: (2010a) public al local et C of Bloechliger provision 87:940Ð947 decentralized Econ versus Polit Centralized J (2003) debt. S public 83(4):1181Ð1206 the Coate Rev of T, Sci determination Besley Polit the Am On legislatures. (1979) in R Bargaining Barro (2009) J Ferejohn DP, Baron 2000Ð2011 Period the over Spain in Crisis Debt and Governments Sub-central ayit.JAp cnII(1):169Ð209 Econ Appl J labyrinth. 54:494Ð531 Polit World systems. multitiered Cambridge Press, MIT constraints. Cambridge Press, versity 43:6 Sci Polit J Am world. 102:798Ð821 Econ Polit J 53(3):426Ð452 Pap Staff IMF evidence. empirical cross-country Some December 178, No paper, working ADBI OECD Cambridge Elgar, Edwar economics. institutional new of Cambridge Press, University bridge Princeton 28(2):365Ð380 Dev World analysis. 10(1):13Ð47 51(4):715Ð732 8(4):331Ð361 Q Policy Politics State spending? reduce 118(532):1785Ð1803 J Econ Congress. US the from Press, evidence MIT evidence. and analysis policy: public and Cambridge design constitutional Democratic (eds), 61:305Ð333 Dev Econ J provinces. Argentina the of ysis 38(2):272Ð294 Publius system. federal swiss the from evidence 90(1Ð2):101Ð117 reglas las de política Economía España (eds) Caja al fiscales. et instituciones JV Salinas e In: empírica. evidencia y Teoría disciplina. 752 No paper, working department economics OECD responses. 2010 12, 112:2611Ð2637 Econ Public J analysis. economy litical 143 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

736 735 734 733 732 731 730 729 728 727 726 725 724 723 722 721 720 719 718 717 716 715 714 713 712 711 710 709 708 707 EDITOR’S706 705 704 703 702 701 700 699 698 697 PROOF696 695 694 693 692 691 ibl ,Rde 20)Bsns ylsadtepltcleooyo eetaie finance: decentralized of economy political the and cycles Business (2006) and J politics, Rodden policies, E, Argentina: Wibbels in federalism Fiscal (2001) P Sanguinetti S, democratic Saiegh in M, enjoy Tommasi governments regional do tax to power much How (2010) E Scorsone F, Toboso perspecti- y evolución autonómicos: gobiernos los de fiscal institutional corresponsabilidad análisis La (2010) Una F Toboso económicos. resultados y política democratic Descentralización in (2006b) governments F regional Toboso financing for choices organizational Key (2006a) F Toboso transition democratic since Spain in growth economic and working change fund Institutional monetary (2005) International F about. Toboso worry to gov- issues state federalism: of fiscal growth On and (2000) size V the Tanzi affect limitations expenditure and tax Do (1996) RJ Shadbegian versus Insurance behavior. fiscal and transfers Intergovernmental (2004) M Tomassi P, Sanguinetti 144 esn o sa eeaimi h U n irsAe l(d)Fsa oiysrelac in surveillance policy Fiscal (eds) al et A Wierts In: EU. the in Europe federalism fiscal for lessons 1:157Ð201 Assoc Econ Caribb Am Lat J Econ reform. institutional 2:157Ð176 Stud Fed Reg Spain? homenaje en ensayos transformación: y Valencia Crisis PUV, (eds) Bono. al Emerit et profesor I Antuñano al In: crisis. de tiempos en vas 37(146):121Ð145 Econ Latinoam Rev Desarro Probl Español. caso del issues. 2:61Ð82 regulatory Econ N Anal and Rev Mercuro Spain. legal M, to Oppenheimer 16 approaches Chapter In: 293Ð427. economic factor. pp York, alternative key New Sharpe, economics: a M.E. and as governments Law regional (eds) regulating 1978: in DC Washington, IMF, paper. 14(1):22Ð35 Policy Econ Contemp ernment? 62:149Ð170 Econ Int J discipline. aggregate .Toboso F. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 ae rae oe oepoeti usin npriua,w pelt insights to appeal we Our particular, effects? In these question. ( of Tullock this absence and explore the Buchanan to from in the model been in least a have at creates decision are change, paper the from and might lose have to What they stand run. actually what short groups to some and used change, are no- of dependence—people afraid a formats—and complicated path various further from in is effect interest choice vested ticeable The a is be. provision with can activity—lobbyists care choice political health that by of difficult form how the over to States as United illustrative the in debate continuing The Introduction 1 Sieberg K. Katri and Shvetsova Olga System Healthcare the Choose to How Deciding .Svtoa( Shvetsova O. ( Pauly including what Many, and systems. costs, efficient which at with care health associated provide are best would tradeoffs system compli- of (Newhouse is type costly situation which more on the even However, improving it right. technology—while make in death, a lifetime, improvements outcomes—also and and their be expensive, life in is should between care once difference care Health least cated. the health at make that everyone can by argue care used many health be to to access expected because is and that service a is it? select to use majority veil to or the prefer (constitutional) behind they mechanism—unanimity from would of choose (legislative) type polity what a would and provision ignorance, care of health of type what ask to DOI (eds.), al. et Schofield N. e-mail: Yliopisto, Tampereen 33104, Tampere of University Finland Studies, American North of Department Sieberg K.K. e-mail: Binghamton, 6000, USA Box 13902, P.O. University, NY Binghamton Science, Political of Department h eeto fahat aesse sahgl hre ujc.Hat care Health subject. charged highly a is system care health a of selection The 10.1007/978-3-642-35239-3_7 katri.sieberg@uta.fi [email protected] B ) dacsi oiia Economy Political in Advances 1962 pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , n oniadEgetn( Eggleston and Kornai and (1971) Rawls ), , 1992 .Tu,dbtsfocus debates Thus, ). 1986 ,and ), 2001) 145 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 h rsu’wt eadt h ucm.I diint h eeto fhat care health of selection the FDIC), the to (and addition Act Glass-Steagall associ- In the costs outcome. as such added the arrangements of to systems, the benefits regard the bear enjoying with to nonetheless ‘rescue’ unlikely while to the policy is median outcome suboptimal herself the by suboptimal choosing voter the bolstered with allow median ated is to The self-interest unwilling 2. is this and polity that occur, The is twist 1. problems The that provision. awareness of voter suboptimal set for to particular support lead typical cohesive this can In under for self-interest problems. off individual actions better but collective be policy, of would one case society the certain Instead, problems, case. a action particular to collective that apply to limited to not extends is model it systems, care health over making polity cost overall a minimize ignorance, to of order in society. veil the mechanism to the selection rule, the Behind majority as provision. unanimity under private select and would for provision, opt care would health polity of same system Buchanan entitlement of an logic select the to prior Appealing the (1962 choice. selec- of Tullock that the and outside affect how so would explore do itself we perspective, mechanism to institutional tion opportunity an an from given Further, if context. less social chosen not be more, would be system what will coverage private then goal, a is care costly. universal if ( provision—among that Hsaio private coverage. argue in universal problems of inherent lack avoid these, to insurance private of form pelt h feto rcst euesrlsdmn—oigta ihu this without that demand—noting states, Klarman surplus costly. reduce overly They become to objective. will prices provision main health of the incentive, is effect control the cost to Here, appeal consumption. care health in ficiency (1969 Klarman 146 stoedbtn h eraiaino elhaei mrc r asn nboth on raising are America 1 in healthcare on reorganization the debating those as In Tullock and Buchanan 1.1 cinpolm nnmt stee-nepeerdmcaimt aedecisions. make to mechanism preferred ex-ante collective the the of is version unanimity through particular this problem, addressed involving action cases be in can that argue issues, We other analysis. this among control, pollution provision, ucation eaegaeu oarvee o hssuggestion. this for reviewer a to grateful are We h aclso Consent of Calculus The n oeotyapc formdli htatog ti oiae ydecision by motivated is it although that is model our of aspect noteworthy One consider to interesting is it scientists, social among debate the of range the Given is bills unwanted and ( unexpected, Enthoven large, (including of Others risk the permit as and 557) it services (1969, eliminated. health in for indulge taste to one’s enhance one may insurance health that fe osdrn eea osbeepaain,tehptei sadvanced is hypothesis the explanations, possible several considering After mn tes setta h akti h etwyt nueef- induce to way best the is market the that assert others, among ) n al ( Rawls and ) ( 1962 ,w hwta ne nnmt,aplt would polity a unanimity, under that show we 1971), 1993 ,Bcaa n ulc s h aequestion same the ask Tullock and Buchanan ), ;Fuchs ); 1994 n ibr n hesv ( Shvetsova and Sieberg and ) 1996 ru o ihyregulated highly a for argue ) .SvtoaadKK Sieberg K.K. and Shvetsova O. 1 nvra ed- universal 2012) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 a eetbtamnrt upr nteeetrt,i h xrm paigfr“0preto 50 of percent “50 for speaking extreme the in electorate, the in support percent.” minority a but reflect can aeu xsec ftoo oedcsv oltospougtn oflcigpoli- simul- conflicting for promulgating each potential coalitions for decisive a more coalition cies. has or decisive minority two one of specified-size existence just a taneous by generating deciding of whereas advantage decision, the has majority ple decision unique ini- the is the requirement. one, possesses that this of who issue satisfying coalition person that rule blocking the for a some of with rule Unanimity, consent or decision endowment. the society a tial with the such ad- only of setting must occur means matter constitution can the right expropriation a that the is endowment (e.g., protecting It that it So land). expropriate of dress. to some endowment wanting of initial subsets Indeed, her property say. its of of they of possession unanimity, or a by speech, in decided free way: is be of this individual only in an can Rights right, which of any issue Bill with specific policy the Of a model population.” would is the one right of how a fraction of “the explanation of actu- their format is in the process, interest are of constitutional provisions rules actual constitutional decision any various the model ality how they illustrate that Tullock decisive claiming and deemed be not Buchanan potentially While to issue. population the an of on fraction any allow they the and to eral, delegated be can implementation policy put (we of to policy particulars healthcare rule bureaucracy). the on decision decisions that which fundamental suppose choose making policy can unanimously 2) a for 1) as that constitution premise to: healthcare the a with is into with Faced approach unanimity. start deci- their to by which then, propose to they of area arrived and is delineation area, itself a policy constitution each as the to constitution apply a to rule define their sion they ills, components: all cure main to two prescription here. a so do not we and and and Buchanan answer applied debaters, be their can current as theory the theory unlike a Since, and offer 5). action Tullock collective (p. between drawn?” line be dividing action the private shall “How controversy: the of sides System Healthcare the Choose to How Deciding 2 tlte rmipeetto fplc eiin hc ol emd ne each under made be would which decisions policy expected compare of to implementation needs from stage policy utilities constitutional unanimous given the a at individual for an most, and rules decision of decisions. set of Fig. sequence feasible Tullock, the the to sketches and of issue way Buchanan extremes the of the all process just and constitutional takes super-majorities, the to majority, capture simple To from unanimity. range areas policy in minorities. specified with provisions, omy oehwvrta aoiaincaiin nrpeettv oiseetdb aoiyi districts in majority by elected bodies representative in coalitions majoritarian that however Note nte yeo eiinrl omni osiuin ssml aoiy Sim- majority. simple is constitutions in common rule decision a of type Another gen- is approach their rules, decision of set feasible a defining to comes it When of consists choice constitutional of theory Tullock’s and Buchanan Specifically, ybcwr nuto,i re oko hc eiinrl ol eether benefit would rule decision which know to order in induction, backward By deciding for options then, large, at polity the addresses it as constitution a In 2 elsial hnmnrt eiinrlsfl nactgr ffdrlo auton- or federal of category a in fall rules decision minority then Realistically majoritarian rcdrs u ntttdwti constitutionally within instituted but procedures, 147 1 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 o oewl o h osiuinlsaea nBcaa n ulc ( Tullock and Buchanan in as stage constitutional the active for an does well reaching which bode difficulty, to not insurmountable comes an it present when preferences But conflicting withheld. to consensus, entitlements or seems violated unanimity be something Indeed, to want not excluded impractical. we be account. rule when well decision into best as a taken work just as are might unanimity difficulties makes stage enforcement This making Promise contract policy contract. once a such consideration knowl- at any from this made entering and transfers from credible, losers (upon utility be expected then of cannot would the losers realis- policy prevent the any a would with in from edge benefits But winner the stage. future share bargaining the won) the that having at commitment promised the receiving was setting of as assured tic be just in can transfer only one works utility where however enforcement the approach perfect That of environment transfers. ideal those an of bar- amounts and the transfers and utility over on Buchanan gaining agreeing futile. through even difficulty and the endless resolving if be suggest and Tullock can coalition, bargaining blocking things, a different is want they individual becoming to each making inability unanimity, decision the of under cost to Indeed, the lead prohibitive. to easily terms, Tullock’s so and Buchanan can in it or, decide because problematic is course, of Unanimity, Rawls 1.2 con- expected their from induction. proceed backward we by con- where select Tullock: rules and and of sequences choice Buchanan unanimous of a theory is the stitution reflects directly This procedure. feasible ( Tullock and Buchanan to according choice institutional 1 Fig. 148 ueaBcaa-ulc tl osiuinscesul:i re o h individuals the for order in successfully: constitution style Buchanan-Tullock a duce 1962) al (1971 Rawls oitc of Logistics nrdcsa supinwihalw h nnmt uet pro- to rule unanimity the allows which assumption an introduces ) not ohpe,sc swe ewn right a want we when as such happen, to .SvtoaadKK Sieberg K.K. and Shvetsova O. 1962 ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 ite orsodt onispsuae odfeetdges eaotteetwo these adopt We so- degrees. different different that to possible postulates is Kornai’s It to assumptions. correspond these cieties to according aligned are erences 2 eped o eiv htasc esnms a oefrbscncsaycare necessary basic for more pay must person sick a person that rich believe a not which do from disease people a from (2) die to person poor a want “Yes.” not cautious do very people a (1) be may answer the welfare ( that social Eggleston long-term out and dis- a Kornai turns as we It viewed Can be healthcare? could outcomes? for what healthcare function for life—what regarding consensus and some demand health least social at of the cern safeguarding about interest—the say of prefer- the one issue about can the assumptions model’s on the actors ground of to ences which on basis the for Looking Eggleston and Kornai 1.3 it as below. assumption model Rawlsian the the the describe of of we provision discussion when the the healthcare to to of return become applies technology will individuals We the of question. and in types function good the utility would which policy the the including by from revealed, mechanisms losses or issue-specific ignorance” gains of the future veil is their include “the it to behind that far regard issues Things how with others. determining some remain on when individuals On than in type weighed area. be their to policy know a need not that for will do individuals rule on that decision much imagine the to very of easier depends theoretical choice this area the see policy for we approximation specific Whether appropriate the area. policy an what particular as exactly a construct know for would be should information, rule identifying decision of deprived properly unanimity when the for thus and options fruit. institutional bear over to preferences procedure individual same each have the for i.e., sufficient have is themselves, to risk-aversion, minimalist their about first, about his uninformed including be, beliefs, similarly might identical constitution are a we individuals what If that Rawls. to to regard assumption interest with of mind outcome needed open constitutional is specific an the assumption keep to that lead but to con- risk-averse), order solution in extremely think maximin only assumptions are to the individuals additional invoking easy his de-facto by (thus it individuals makes cept these Rawls find of alike. and risk-aversion fact the ignorance” in about of are meta-constitutional “veil they the because the at alike behind terms, processes distributive decide Rawlsian In future individuals regulate. the stage will who in individuals constitution positions of the their comprised which identical about be incorporate to uninformed and society similarly same decision- a are the need the exactly else’s—we Literally, everybody be one. as must as inputs person decide each must of they unanimously, process making decide to able be to System Healthcare the Choose to How Deciding ti,o ore liaeya miia usinwehro o niiulpref- individual not or whether question empirical an ultimately course, of is, It fRws rmwr a eacpe,te tcudb rudta n individual, any that argued be could it then accepted, be can framework Rawls’ If hnahatypro Kra n Eggleston and (Kornai person healthy a than and treatment, medical standard with die to have not would 2001 oi ht ttevr least, very the at that, posit ) 2001 .50). p. , 149 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 sin oee ayo e epniiiisfrognzn e w healthcare own her organizing for responsibilities few legislation financing. observe or the we whom many principal, to however agent the society’s assigns as the functions patient—becomes form individual—a some an that in Fig. society the in where depicted mechanism process the about Thinking “Patient” the and “Society” The Actors: 2 consequences. polity the the suffer yet individuals and that let rule, to majority preferences unwilling under these is provision suboptimal Individual is problems. to action It lead collective can of stage. self-interest class constitutional special the the into problem at our designate held unanimously are more thus the of care the for subsidy public the withhold to not apply prefers to principal pay The to prefer Ð will principal the emergency, life-threatening a to came it If characteristics: following Ð the with function welfare social a of favor ini- 2012. Fall the and and Tampere on 2010 of based Fall University the in and University at appeal Binghamton experiments theoretical classroom in their validation of empirical grounds tial the on here, assumptions 150 spraet hc cso h oit’ behalf. society’s the sub-principal on con- appointed acts the the of which of will functions (super-agent) outcomes utility “contract” implementation respective healthcare the the principal’s maximize the of The which comparison of tracts principal. the drafting on full the depend the with simply entrust then of principal, to full super-agent who the a over of on preferences sub-principal relying a as to alternatively, delegation or, as policy policy. healthcare a that under the produces least chooses principal the the suffer would that he tautology in agent that a an ignore as on because we argument policy but our certain lower, building even the avoid be three for to may the payoffs of order agent-patients any assign of by achieved we payoffs payoff when The of level Thus principals. minimal society. view the the assign of we of point principal, the member ex-ante from destitute also structures most were choice the they alternative of the If compare preferences. would these they share event, they or and (Rawls healthy poor, one players or as maximin rich are be individuals will all one whether sick, stylized Rawlsian yet the knowing identify At not and players. specific stage, separate more “constitutional” as be incarnations varying to its need in levels thus “society” We changing the unfolds. experience process and the under as junctions decisions information different of make at it aggregation of comprise rules who different individuals because framework, this within ik(hs aei oeepnie ytels ik(hs aei esexpensive). less is care (whose sick less the by expensive) more is care (whose sick and treatment, life-saving accepted nte bevto oda rmFig. from draw to observation Another and society a in citizen every of preferences common the are presumably These in principal the within consensus popular the sketch above assumptions two The hsapoc losu ove h hieo h eiinbd hc then which body decision the of choice the view to us allows approach This 1971 ,ads ogtt vi h os osbetr of turn possible worst the avoid to sought so and ), 1 sta scey stognrlaterm a general too is “society” that is 1 bv sacoc facontractual a of choice a as above .SvtoaadKK Sieberg K.K. and Shvetsova O. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 hr h ai rfrne spsuae yKra n glso ( Eggleston and Kornai above. by postulated discussed as preferences basic the share PP. enforcement dis- Principal, and the Ex-Post implementation and the the types is at level own there Finally, the population. stage, their by the about principal in have types ex-ante society of the the tribution of from members differs that but principal information people interim of same by-unanimity the the majori- that of either Notice comprised IPU). be is or An can (IPM stage. choice, unanimity constitutional by constitutional the or the at tarian on acts depending EAP, IP, Principal, principal, Ex-ante Interim An PP. and IP, EAP, as within interests conflicting these of conflicting structures. implications potentially institutional how- the different and personifications exposes preferences Multiple model distinctive provision. Our with welfare interests. actors of separate matters imply in ever actor same unilaterally the case as that in hospital a or if care doctor her a principal. deny and societal to for, the site calls saving represents on contract life provider authorized medical the a the what cover some to is to up that or assets is no it principal treatment, has at extending the patient enforcement life a or by contract example, the conducted for If, and also policy, thereof. representative is the stage of implementation implementation of the stage the also is ut stefis roiy hl otmnmzto sscnay esa wyfrom away ade- stay We as secondary. viewed is level cost-minimization some while at priority, healthcare first the of is Provision is quate care. outcome to acceptable as regard long with as achieved minimized for it indecent prefers EAP, is principal, it constitutional that it. and without be life to sustain society to in necessary people is lowest it the even that necessary—in a of tion eln oshlr fpltcleooy ae a ruda a aka 95that: 1945 as back far as argued ap- has historically Hayek economy. been political of have scholars to assumptions pealing Kornai–Eggleston’s and premise Rawls’s Principal Rawlsian Ex-ante The 2.1 System Healthcare the Choose to How Deciding npoiigfrtoecmo aad flf gis hc e a make can ( few Fuchs which against life Matthews of added, (emphasis hazards provision. common adequate those for providing in a organize to help not should state fwat ushs h rtkn fscrt hudntb urnedt all to guaranteed be shelter food, not of should minimum clothing, some security and is: of that kind freedom; general first endangering the without has, ours wealth of l or(onigbt P n P)atr ersnigtepicpl eclaim, we principal, the representing actors IPU) and IPM both (counting four All below labeled are principal societal the representing actors of types three The treated is colloquially what of personifications different three have we sense, a In there principal, policy-setting the and principal Constitutional the to addition In noa stettl(raeae oto h oiyi ocre,w sueta the that assume we concerned, is policy the of cost average) (or total the as Insofar hr sn esnwy nasceywihhsrahdtegnrllevel general the reached has which society a in why, reason no is There 1996 6 losae htmdclcr et dmSihs17 defini- 1776 Smith’s Adam meets care medical that states also 16) , ufiin opeev health preserve to sufficient opeesv ytmo oilinsurance social of system comprehensive o steeayrao h the why reason any there is Nor . 2010 ) 2001 hc we which ) 151 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 h ortpso principals of types four the noeost h oiycoc n hsapi:(pcfi oiycoc;specific body. decisive choice; the is interaction policy of the coalition (specific of rules outcome the decisive pair: equilibrium to an the a according be must of thus coalition) decisive and membership the choice of the make-up policy rules, the entirety other the to as all outset endogenous the Under at society. fixed is the policy for of coalition the decisive unanimity, the under of local only anywhere members the Importantly, of even place. place set or in chamber take rules legislative the a can on to depending contract/policy referendum government, a the or and body of constitutional size a choice from the The constitutes of policy. which agent-patient society healthcare the with of the the contract fundamentals the the in chooses on It decisive individuals policy. be healthcare to of the constitution the coalition by empowered a as is composition principal interim Our Body Policy-Setting Principal—The Interim 2.2 elaborate We actors-principals. the of below. functions differences utility these on the in distinctions the lights no that. make for and preferences adequate societal minimally and individual is include about what not assumptions beyond do additional care we surplus provision, any care consideration basic in for preference lexicographic of is assumption what ( of Attfield see notion wealth, the ( societal that the Burau with suggest point care to a of research to level vary, is a may (there adequate adequate afford minimally as cannot view society to the possible that is it whether of discussion the 1 Table 152 nywt h evcswihete o ryu elhisrnecnfiac and finance can insurance health your or provided means you be by either will You or which insurance. service, services health of the of point type with at and only amount providers cover- specific from health a directly much entitlement purchasing either as of the buy choose, can mean you “You effectively read: as might would age contract that taking a for like Or responsibility system. Something yours the single-payer health.” of assume portion your we a return of withhold in care automatically and to earnings, going else’s everyone are sort. and some “we of be: contract might a patient/agent contract unanimous the The it offers Be principal respectively). interim MIP the a or majoritarian, versus UIP or body (the (unanimous) rule constitutional majority a simple is with choice legislature policy for venue the where tion nFig. In Table 2006 tlt ucin of functions Utility 1 ecmaesd ysd h rcs fplc aigadimplementa- and making policy of process the side by side compare we 2, umrzstecmoiino h A’ tlt ucin n lohigh- also and function, utility EAP’s the of composition the summarizes ,Hwl n cagln( McLaughlin and Howell ), PYes Yes PP Yes MIP Yes UIP EAP dqaecare adequate Minimal 1989 ) lo ie h KornaiÐEggleston the given Also, )). a burden tax Personal Yes Yes .SvtoaadKK Sieberg K.K. and Shvetsova O. aeaetxburden) tax (average cost Societal Yes ,Bakand Blank 1990), Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 pcfi hnta fteEPi ead owihcssbcm h ato h cal- the of part funding the the a) become the costs include which must to necessarily contract/policy regards the in that EAP Notice, culation. the of that than contract. specific a such decision fun- adopt non-enforceable, to is choose be latter would to the types all it certain that of While knowing show bodies available. even we be yet below also non-enforceable, issues, might damentally options implementation in-between their of have range contracts full the practice in i.2 Fig. System Healthcare the Choose to How Deciding eida da yeo uemre rvt nuac ytm ntemdlin model the In system. insurance private market contract pure the be a would of This Fig. type needs.” ideal health your an of behind regardless that, beyond nothing non-enforcement non-compliance/public individual the of case in the costs the cover to taxation from exempt are coalition decisive the of Members ( non-enforcement. public the non-compliance/ of individual case the finance to have will coalition decisive the of members Some unanimity: b eiinb aoiyrule: majority by Decision ) h tlt ucino iie samme fa nei rnia smore is principal interim an of member a as citizen a of function utility The 2 ( elmtorevswt hs w xrm ye fplc hie,though choices, policy of types extreme two these with ourselves limit we eiinby Decision a) (b) (a) 153 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 sntuiesl n ih pl,freape hnteteteto mirnsnnctzn is immigrants/non-citizens of treatment the when example, for apply, might concerned. and universal, not is u ufiin rvt ud ocvrtecs ftetet Parameter treatment. of cost the with- cover and to coverage funds adequate private purchased sufficient without out those for policy “no-care” implied choices, These rc sefre i aain oteetemv yP htw hwi redundant is show we that PP by move the there so taxation, via enforced is tract state to to node access to cision resort sufficient last The is overrun. assumptions. MIP KornaiÐEggleston cost the or cover from to UIP logically budget follows by state budget move of redundant in use games the This in justify principal. move last the the of be could preference This well. that as do IP death— to the Figs. or not to by sickness offer prefers worse continued to even turn patient—her able considered in severely, the be is most for would PP outcome he alternative sanction which the to poor because position the a to in to care IP, deny preference rich. not a the to principal holds and the PP patient Specifically, that the premises. assumption treat Eggleston’s our and with Kornai consistent to is adheres from observation infusions through This the or by patients budget. covered paying state be the overcharging to by transferred eventually either large, are another, at or patients, society way these one from in recoup costs, cannot such they and which costs life-threatening incur to immediately thus not denied Providers for be conditions. treatment to including market-type carry rules, ought the the to which to stage failed treatment according this have them the receive at and they that is pay coverage, who claimed cannot medical patient long patients sufficient the been though treat has unimplemented: to It goes not contract. contract or This a whether doctor up. on of shows a decision breach a patient as in make uninsured such to an behalf, has society’s where then the room individual emergency on a an acts is in who principal administrator authority ex-post or of the contract/policy, position market-type a the in citizen enforcing of time the At Stage Implementation Contract the at Principal Ex-post 2.3 fails. will contract/policy that the overrun tax of cost the enforcement the of the of share if share them what the on about know allocated fall expectations who is form actors burden can our tax bear that they general burden From the process. where majoritarian taxation, the regular enforcement via from from arise up might made which of shortfall is framework any failure constitutional that assume broad general thus the the overrun. will within to We cost implied state. access the a are the they is through that place, there and in when budget, indeed do state also are to c) provisions what and, such necessary), in that is as claim We that provisions, all fallout or the covered stipulate (only must services of level the b) principle, 154 3 h hiet noc rntt noc h n ae rvso eper hr h entitlement the where reappears provision care” “no the enforce to not or enforce to choice The ntemdlstrs hn h xps rnia,P,hstecoc ttels de- last the at choice the has PP, principal, ex-post the then, terms, model’s the In and 2a ,btw ev tumdle o ti eudn ie h assumed the given redundant is it for unmodelled it leave we but 2b, enforce e versus or ∼e o enforce not pl under apply , otat(ntecs of case the (in contract a Insurance elhplc oefrigthe enforcing to policy health .SvtoaadKK Sieberg K.K. and Shvetsova O. h con- the Entitlement, −p ntePP the in 3 ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 yteP n aei rvdd hnn uhuiiyls oaetocr hl no while occurs agent to enforced loss agent-patient. not utility the is by such contract made no is the then care hand the provided, other financing is the to care contribution on the and in If enforced, PP decline dying. is a the from worse, contract by or a getting life, the to avoidably of If from care quality loss denying care. the utility when request catastrophic service, to a of sustains chose point agent who the patient at only uninsured enforced sick be enforced. can with otherwise or contract case punished The the be cannot not de-facto therefore is and thus however, non-compliance, collection, constitute same, tax The enforcing enforced. is of compliance collection problem tax the that from assuming away stay we erally, gn ic teul o-atcpto nepomn hsecpn nvra tax. universal Fig. escaping in thus indicated employment in is non-participation This equals it since agent for. pay can To covered. much has how to and tient/agent Thus, purchased requesting. has is one she coverage system, much care market-type how a components: In two that. of reflect consists payoffs is treatment her no and if workforce, enough the be may (zero insurance enough To buy sought). or tax social the pay to . xr oto eieigHat aea mrec Care Emergency as Care Health Delivering of Cost Extra 3.1 Actors to Costs and Technology Health-Related 3 between that choosing include contract the not by do abides we agent so the ( Here, and a model. argument, our of con- normative in that cost separate, possibility treatment—but the a such bear cover for would to calls that unwilling, insurance sideration reality, just an In contract/policy. of not or the unable, treatment to financially serious according is costs sometimes the bear agent to the like not does she patient. erally a to refused indeed They is is MIP. it and care but UIP, if EAP, principal, loss actors-principals, ex-post other sustain as the of all by and functions felt premises, Kornai’s utility is it captures in society, present It because the also insurance. in care carry individual needed every or to patient pay applies a to it deny failure to her or having his from of loss utility this captures payoff System Healthcare the Choose to How Deciding nFig. In tsmc etrrtrsta hti ihedlf-aigmdcn seHlo and Halfon (see medicine life-saving high-end gener- in measures that preventative than in returns investment better much that ates developed has sub- major a consensus Specifically, a greater others. stantial to with it compared reached “technology” consider be medical can healthcare some outcomes under of efficiency health similar Scholars whether care. determine to regular objective and preventative for care and comply) h eiint o opyi h nilmn aei titydmntdfrthe for dominated strictly is case Entitlement the in comply not to decision The h gn ptet aushrhat n at orciecr fsc.Btgen- But sick. if care receive to wants and health her values (patient) Agent The 2 , c> o comply not ne h aktbsdplc.Ntbyn nuac osntb itself by not does insurance buying Not policy. market-based the under ∼ atrstefiaca fcec osfo usiuigemergency substituting from loss efficiency financial the captures 0 c ( .To comply). not nasnl-ae ytmrqie htteaetsasotof out stays agent the that requires system single-payer a in 2 yteuiiyls of loss utility the by o comply not comply eedn nasbae en either means subgame, a on depending comply, en oakfrcr necs fwa she what of excess in care for ask to means en oakfrcr nteaon h pa- the amount the in care for ask to means −t u otels fwgs Gen- wages. of loss the to due o complying not 155 c Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 ucm sta l aepeettv n eua ae n hnhat sfinanced the is when health pool. minimized redistributive when other and field is or care, principal medical “insurance” regular the society-wide the and a to preventative in in have cost structure all overall cost that the the is that outcome and prediction function the welfare to about social have lead we the that information in the that aims conclude the can we principal, the to cost overall regular where outcomes applied. the consistently prefers is principal its care as preventative the care and that regular conclude than expensive must more we is substitute, care emergency that fact technological the twice. itself for paying thus surplus, net a generate hsi eas nfiacn elhaealne ol efiacn h “investment” the financing be strength would fails. lender the approach a this by healthcare financing health bolstered in financing because be with is credit, would This make such credit for would of demand markets individual’s price financial the the of the and good available, other resources any the with failure Where market first-order above. the described following as failure market second-order the for accounts pay. to care preventative Canning and and (Bloom regular society that the possibility with of a that resources emergencies even productive of of the is part share boost a There the may replace prevent. treating to than will cheaper willing it is be that pay care should to regular cost we willing since lower latter, are care, a the we “regular” at for since replaces plainly, paying) Put it are medicine. as (and emergency medicine eventual preventative the be cheaper should of the procedures some emergency for for pay pay to to willing willing is who principal the that dictates aigpoeue ie si mle ytetcnlgclcaatrsiso medical of characteristics treatment technological of the cost by the implied innovations, is as rise, procedures ( saving Eggleston and transfers. those fi- Kornai authorizing to the are with They according why consistent necessity. preferences that, consequently present-day has luck and a principal pure collective case, are is sick the it the is and to that necessary, healthy why the from elaborate Here transfers very to care. nancial critical of place and exception specialized right the for the need (with urgent is available we to be care, comes it to for when unlikely funds individuals) private complexity. are few with funds further paying such add that of agent/patient mentioned possibility to the earlier costs mentioned the Having on bearing aspects Technology Increasing Are Healthcare of Costs Marginal 3.2 lives save Medicine to of having (Institute of emergencies instances the in reduces care regular and preventative consistent Hochstein 156 umn ptedsuso fteapcso eia ehooyta fetthe affect that technology medical of aspects the of discussion the up Summing with emergency an in lives saving for preference of premise the combining So h itntv aueo elhaea od nte ehooyrltdaspect, technology-related another good, a as healthcare of nature distinctive The logic then efficient, as medicine preventative of favor in tradeoff the accept we If o lotayidvda rfml,a h ot fmdclinvtosadlife- and innovations medical of costs the as family, or individual any almost For 2002 mn tes.Ti feti oetal xlie ytefc that fact the by explained potentially is effect This others). among , if n culybcmsvr l xed h ability the exceeds ill very becomes actually one 2002 ). .SvtoaadKK Sieberg K.K. and Shvetsova O. ,the 2001), 2000 and ) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 etntsta ihlvl fc-asfrhat nuac ilidc epet use to people induce will insurance health for this co-pays of for levels high care that health notes ment of value overconsumption. true constitutes the insured while indicates consumed uninsured care twice the while care group—thus ambulatory care use (1986 of year uninsured. lack the while of ative than part insured only would while for person much insured as that are than who illness people of an that purchase ing treat the (1986 to when insurance. more the occurs without of spend it spent have to the illness—or person because an a risks of insured causes greater event the insurance take the causes to in insurance coverage care—i.e. preventative of moral health on certainty Here, of less insurance. presence spend medical the to in when person role occur strong either a care. can health plays overuse hazard hazard to moral incentives that create insurance argues health He to subsidies tax that argue o u oe st htetn n hni h eadfrhatcr lsi?Pauly elastic? healthcare for demand the functions is payoff when the and generate extent ( we what as to ourselves is ask model our to for need we that question next The Peculiar? Healthcare of Consumption Is 3.3 (patient). Agent the angle: by following non-compliance the constitutes from what what situation beyond is the contract anything at receive look or to can insurance attempt one much words, not asking other how does In she decide for. as to paid long she individual as an purchase, to Fig. to up in care is as category it contract/policy our representation, health in market-type abstract falls the treatment requiring with and non-compliance sick of getting and money, ready enough accumulate treatment. to serious ability a the fund from to demographics size debt—not sufficient the of massive of development, savings portion career of large early accumulation a and is excludes education economy which professional modern of precur- stage a necessary the in a savings—during near earnings be almost high can Finally, potential career to individuals. one’s as sor maintaining high-earning risk strikes, for illness high even an present once impossible earn- life, and their in economy later developing knowledge-based And in hires. the The society— difficulties a in good. additional in a capacity individuals encounter as ing sickest who health the those of by demanded peculiarities precisely is the by care for Moreover, health account thus age. of to and variety wage-earning fails costlier the reasons again than patients predispositions once younger sickest genetic view are the to this they of related when or number care expensive genetic large need of A because reasons. sick related are for fallacious the weaker similarly the is speak, to so and, greater individual individual—the the sick sicker a the “collateral.” of funding, ability for earning demand subsequent the the and survival the in System Healthcare the Choose to How Deciding 1986 h oa aadnto a ubro rtc.ARN oprto experi- corporation RAND A critics. of number a has notion hazard moral The obnto ffiuet ucaeaeut muto nuac,nthaving not insurance, of amount adequate purchase to failure of combination A care health financing of means a become can savings individual that view The eiisteapiaino h cnmcmdlo nuac ohat aeto care health to insurance of model economic the of application the revisits ) o ramn o hc o r o lgbeudrti omo h social the of form this under eligible not are you which for treatment for 4)A neape al ie aashow- data cites Pauly example, an As 640) , 3) easmsta h rel- the that assumes He 636). , nour In 2. 157 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 h s fpeettv aewihmyb esnlycsl ntrso ieand structures care time health incentives. of resources, provide care it terms must health When of in individuals. use of economical costly behavior encouraging desirable personally to reward comes be to choose may may the which principal encourage the select To care effort, way. alternatives. preventative to straightforward available a of fail in of use patients schemes cost out incentive the when by the treatments addressed is of and be can hazard providers some This efficient moral out more of wipe and instance to cheaper are more potential who the one those Yet has to compared gains. merely as possibility coverage have this unlikely they are uninsured, when by they treatment treatment since of seeking However, cost delay problems. the to acute up or run eventually diseases so advanced developing and care preventative regular in to effort access enough care. that emergency claim on the only extension with reliance than Medicaid consistent costly the is less of This is benefits costs. care than the the lower that was than life”—or saved higher human life were per a cost of the “value that signif- typical argue decreased the they mortality efficiency, beneficial—child of of was use terms use made the In this people icantly. following of that number note findings, larger also Pauly’s a They with services, them. health consistent use that, to opportunity note increased They people. of portion t,cssntihtnig n tmgtntee euetecss ureadGruber and Currie util- costs: principal’s the preferences, the reduce even principal’s decreases ( not reasons the might it these to And for notwithstanding. contrary demand costs is ity, in pay, decrease to a inability therefore agent’s and to due care care medical sary the of is misuse this regarding concerns If let efficiency. else. not assessing anything in to purchase inflated careful be to pur- be that must them because we for insurance then impossible health case, it purchase to make not would choose chase many care, health for ity ellf hie aylwricm epemk nhat aeconsumption: care health in make people ( income Gladwell lower article, many popular choices rather a real-life decreased, In have costs. could overall them increased, using than and necessary (Gladwell were way neglected efficient they an in services necessarily not but care, health less 158 inlydcdddcso ue h nilmn ehns ihatmtcflttax flat constitu- automatic their with using mechanism choosing entitlement policy be the stylized will rule: two who decision to makers decided mechanisms tionally policy provision our healthcare between of extremes field the simplify We Market Versus Entitlement Choice: Policy Health 4 1996 nte seto oa aadwt gnsptet rsswe hyd o put not do they when arises agents-patients with hazard moral of aspect Another hseatct,mnfse nrdcddmn eo oebsclvlo neces- of level basic some below demand reduced in manifested elasticity, This util- low revealing than rather that, show it. examples of on colorful instead bandage numerous Gladwell’s a because put It’s he hazard. fixed moral bone of broken scourge a getting the defeated has he because Isrne)ae Contracts (Insurance-)Based xlr h fet rmteetnino eiadsrie oalre pro- larger a to services Medicaid of extension the from effects the explore ) tv ssls elhcr hnh ol fh a nuac,btta’ not that’s but insurance, had he if would he than care health less uses Steve .SvtoaadKK Sieberg K.K. and Shvetsova O. hsprry the portrays thus 2005) 2005 .Mn fthe of Many ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

690 689 688 687 686 685 684 683 682 681 680 679 678 677 676 675 674 673 672 671 670 669 668 667 666 665 664 663 662 661 EDITOR’S660 659 658 657 656 655 654 653 652 651 PROOF650 649 648 647 646 645 o aet es hndsgae udn sdepleted. is does funding provision designated and soft when the is in cease area preferences to policy function have lexicographic health a the the not be within to not constraint Due may budget or funding. the guar- may polity, to the which contribution of (healthcare), agent’s and player. good the a scheme the of funding of of type a delivery particular of the each consists of by chooses antee UIP, born principal EAP, are the to they that payoffs as contract and costs A principal, those the reflect within PP costs and MIP, of ( allocation Eggleston the is and component Kornai to according benefits purchase, nFig. in to whether chooses PP the ( Finally, scheme. coverage either of requirements nteedt e h epepytepiefrterondecisions. own their for unwilling price is the principal pay the people because the taxation, private let by by to covered determined end must spending the that care in spending, health extra on the constraint and intended purchase the between distinguish to here ( Shvetsova and Sieberg see u oe nlzstecoc fplc oeaeuigbcwr nuto.In induction. backward using coverage policy of choice Fig. the analyzes model Our Model The 4.1 healthcare what of person. principal each the to provided by be decision must the circumstances our what address under in we and rely do services not nor solvency financial do system, improves We the pool here. risk of the wide framework of population analytical that aspects assuming our these on of conclusions of outside resource Both making are constraint. requires choice budget pop- it policy the entire hand the within over other fit expenditures the to resource on ulation the max- allow and it would pool, hand, that risk one decisions the the allocation On of population. size its uses for and the care population imizes health the universal from fund Secu- taxes to Social collects money and generally that Medicare system of payer funding single the a like The with system rity). funded (a provision workers care all basic on The guaranteed tax insurance). universal payer health single of the healthcare is policy either Entitlement (of purchase individualistic fully versus System Healthcare the Choose to How Deciding 5 4 an as care health between decide must h,aogohrtig,cudb se odcd hte rntt odtebudget the hold to not principal or the whether decide of to strategy asked equilibrium be an could of things, part other a among viewed who, be could it but assumption, case address. particular we the that identifies the problems that of action assumption nature collective this the of is and it PP short, the In of process. preferences enforcement presumed the from follows but sumption e sntdb eiwr h em sf osrit per ob noyoo.W s h term the use We oxymoron. an be to appears constraint’ ‘soft terms the reviewer, a by noted As systems, insurance private versus payer single of efficiencies relative the regarding arguments For )ornot( oignwt h tlzdmdlo osiuinladplc hie payoffs choice, policy and constitutional of model stylized the to now Moving nti sa ecos otettesf ugtcntan nrgr ohat san as health to regard in constraint budget soft the treat to choose we essay this In ,w eittecocsmd sn nnmt ue nti iuto,teUIP the situation, this In rule. unanimity using made choices the depict we 2a, 2 oaltreatr-rniasrflc hi rfrne o eieighealth delivering for preferences their reflect actors-principals three all to I ∼ et h ain/gn,A either A, Patient/agent, the Next, . e h ue ftegvncvrg ceea on fservice. of point at scheme coverage given the of rules the ) 2012 ). entitlement, 2001 complies E .Teohrmdlsnecessary model’s other The ). rtruhpiae( private through or , 5 hsi o na o as- hoc ad an not is This ( c )ornot( ∼c Entitlement 4 insurance) ihthe with ) enforce 159 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

736 735 734 733 732 731 730 729 728 727 726 725 724 723 722 721 720 719 718 717 716 715 714 713 712 711 710 709 708 707 EDITOR’S706 705 704 703 702 701 700 699 698 697 PROOF696 695 694 693 692 691 vi opiain emrl drs h su ftxto ofiac xr elhcr spending care health extra finance to taxation of issue the here. address merely we complication, avoid anyipidadms eacutdfri h aof fteitrmprincipals interim the of payoffs the in Fig. for in form accounted Figs. extensive in be the both must in and included implied not is tainly decision taxing the While Taxation Majoritarian and Theorem Voter Median The 4.2 by Fig. designed in is UIP contract of the case when the because in (as is unanimity This (ex-ante) majority). dimin- under coalition decisive (e.g., the made size when is in redistributive decision ishes more the up end if from will equitable and covered more unanimity, be overruns by will cost mechanism mechanism. plus funding financing taxation) budget, policy-specific general policy-designated efficient the (health an exploiting designing combined to of The leads their lieu procedure in into majoritarian budget factors the state this that show and the We taxation under payoffs. place general takes expected it majoritarian that by cognizant are generated decisions expectations their in preferences. actors tax But voter healthcare. median majority. of the by on decided results always the is on back assume, fall here can we we that purchased, Thus care the any not cover in to but addressed taxation provided taxation General general section. was next of the policy in nesting discussed a literature is extant there background, the in and spending other from redistributing or period could next one but the areas. funds firm against additional is borrowing of constraint budget source for the The allow where health. budget, or national lives the human is of presumably price a at firm as constraint 160 6 napplto iha noedsrbto hti kwdtwrstelf,the left, the towards skewed is that distribution minority. income wealthy the of an on body taxes with of voluminous population burden The chief a the enabled. distribution placing In income is be the to majority it of by predicts mean literature the taxation below redistributive way thus is and income excess median that this premise of the much how voter. on median policy directly the healthcare by depend borne over will is preference burden overruns majority budget then tax- rule, generates general in from majority that treatment covered by are deny expenses on not unfunded agreed How- will these ation, ex-post, treatment. principal If, ex-post get need. the not of case where should the contingency they a which is beyond there coverage, ever, of amount certain a every so overruns), cost its US, burden. the the of in part case a bear the to be agree to may have as will person (or disproportionately to policy obligated legally the be finance could which coalition decisive the outside ngnrl ae a eue ofn oto evcs rjcs eitiuinshms t.To etc. schemes, redistribution projects, services, of host a fund to used be can taxes general, In osiuinlcoc o oiypoeueta emdlapisol otearea the to only applies model we that procedure policy for choice Constitutional parallel, in healthcare, of area the in shortfall potential the for up make to Thus, coaso sa oiy(e,eg,MlzradRichard and Meltzer e.g., (see, policy fiscal of Scholars and 2a 2 .Udrapiaeisrnesse,idvdaswl purchase will individuals system, insurance private a Under b. 2 ) hr osnteitaminority a exist not does there a), .SvtoaadKK Sieberg K.K. and Shvetsova O. 1981 , 1983 ti cer- is it 2, )relyon 6 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

782 781 780 779 778 777 776 775 774 773 772 771 770 769 768 767 766 765 764 763 762 761 760 759 758 757 756 755 754 753 EDITOR’S752 751 750 749 748 747 746 745 744 743 PROOF742 741 740 739 738 737 tvl oecnee,yttecaiinwt rfrnefor preference with coalition the yet centered, more atively than scheme insurance an under this If less coverage. pay wealthier of would level and voter their with insurance, beyond median care own healthy need a her who then for those prefers holds, voter of only median care pays the the for case, she pay this which voters In in all. at system burden insurance tax an no bear can voter median the ouaindcdn nhatcr oiygvnta otoern r aeu from up voting Table made the in are in .2 as overruns distribution and becomes cost joint taxation wealth, that the general given of Then policy dimension health. healthcare of the on dimension on deciding the population type, type on wealth wealthy type their the sick as of the well of .2 as type are health probabilities their the comprise know who and stage individuals interim that the simple, at it principal keep the to Suppose, dimension. each on lation the of minority small and relatively care a specialized by costly demanded population. the are to population maintenance) “oppo- (due the substantively disability above costs is the severe healthcare sickness” in most of but that “level wealth meaning mean the of median, distribu- that that the assume health- to that we on similarly direction, premise median skewed site” the the is Adding what receive. health see to of to and tion is pay MIP to to expects payoff dimension the care identifying to step next The Policy Healthcare on Preferences Median 4.3 more has then, voter, This Eibner voter. and mean Rosenthal 1999 (see the taxation than redistributive demand income to lower incentive a has voter median System Healthcare the Choose to How Deciding electorate the in types of distribution 2 Table payoff baseline the not to Relatively are taxation. AND amount general healthcare for own pool their fiscal for from the anything in pay be to to need going not will electorate the otebsln aofo .Ti dai ossetwt te eerho h link Corcoran the example, on For research spending. other social with and consistent ( share Evans is and tax idea voter’s This median 0. the of between payoff baseline the to ntedmnino eea aaini eoproa otiuint aigfrthe for paying to contribution coalition personal healthcare. majority zero on is the overrun taxation of cost general expectation of the dimension Thus the education. on public on spending local nra icmtne,tedsrbtoso elhado fwat ih erel- be might wealth of and/or health of distributions the circumstances, real In oiei h lutaini Table in illustration the in Notice popu- the in types binary with example hypothetical a consider illustration, an As eas h er eso h udn ocmeadCuil( Caudill and Holcombe burden. the of less bears she because ) Entitlement Entitlement hypothetical A d srflce ntepyfst h I nFig. in MIP the to payoffs the in reflected as , 2010 n htardcini h einvtrstxsaeidcshigher induces share tax voter’s median the in reduction a that find ) hshrpyf for payoff her thus ; oiy ihisuiomtx hyaetu aigsm positive some saving thus are they tax, uniform its with policy, Healthy Sick 2 2. Insurance hti hsrte xrm ae6 ecn of percent 64 case extreme rather this in that is d hc sgetrta requal or than greater is which .16 Poor .64 3. Insurance 1985 2005 hwthat show ) ih still might ,Nelson .04 Wealthy .16 161 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

828 827 826 825 824 823 822 821 820 819 818 817 816 815 814 813 812 811 810 809 808 807 806 805 804 803 802 801 800 799 EDITOR’S798 797 796 795 794 793 792 791 790 789 PROOF788 787 786 785 784 783 h iuto ifr.Hr,telf adsd ftete sietclt hti h UIP the in that to identical is tree the of side hand left the Here, differs. situation the care, emergency covering and Entitlement. 0 payoff, baseline the eas l iiesaecvrdunder covered are citizens all because −t both under adotiigtebsln aofo hnby than 0 of payoff baseline the obtaining ing—and eas h osntrs h payoff the risk payoff safely positive not can does she she that knows because Agent comply, the not PP, of preference that Given coverage. sufficient of h bv icsin ntesbaesatn ihtemv yUPo h ethand left the on UIP by move the from with generated starting payoffs sideofthetreeinFig. subgame the the with In game discussion. the above to the induction backward apply now can We 5Analysis pol- adoption. the its in on participation vote from can exempt but are icy that groups those to majority—due exceed payoff) first the is (EAP) principal ex-ante the of payoff 3 Fig. 162 −p ntesbaeo h ih adsd trigwt h oeb I,however, MIP, by move the with starting side hand right between the choice on a subgame faces the In it that knows UIP the then, subgame, the of top the At fteplc is policy the If fsesasoto h okoc wihi hti ae to takes it what is (which workforce the of out stays she if hieo h eiinrl o elhaeplc ttecntttoa Rwsa)sae(the stage (Rawlsian) constitutional the at policy Healthcare for rule decision the of Choice fhe if enforce Enforces r . h Phsn ifrnei aof u ohschoice, his to due payoffs in difference no has PP the Entitlement, and h ue n osnttetaptetwohsntpurchased not has who patient a treat not does and rules the 3 ,ifInsurance o enforce not h gn,i hscs osbte by better does case this in Agent, The . steplc,teP ban eaiepayoff negative a obtains PP the policy, the is Entitlement r Ðs n nta h a banthe obtain can she instead and , n oh a opoiecare provide to has he so and o complying not −c .SvtoaadKK Sieberg K.K. and Shvetsova O. oteUPwl p for opt will UIP the so , o comply). not n obtaining and comply- Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

874 873 872 871 870 869 868 867 866 865 864 863 862 861 860 859 858 857 856 855 854 853 852 851 850 849 848 847 846 845 EDITOR’S844 843 842 841 840 839 838 837 836 835 PROOF834 833 832 831 830 829 oe esnlcs otemda oe eut nahge tlt hntebaseline the than utility higher for a opt the in will will from results MIP who exempt voter the those also median thus, of payoff, the is care to the she cover cost And median to personal insurance. used The lower own be move. will her different which for tax a only general makes pays he most, outcomes, at the voter, to regard with expectation aoiy hnteegop a omacaiinadaotEtteetee tthe at even Entitlement adopt and coalition a stage. form legislative can groups these then majority, a netit st n’ w elhsau,w aeteptnilfreeyn ootfrEntitlement. for opt to everyone for potential the have we status, health own one’s to as uncertainty surance prefer n hi epciecneune,wl p for subgames opt Majoritarian will and payoff, consequences, Unanimity respective the their in outcomes and the anticipating EAP, Our mrec aea hi oehat aeoto ste ol ewt cesto access with be would they the with as person satisfied option poor as care a from be If health that to care. likely sole than regular less their lower are more as is unhealthy (or, care and care care emergency poor regular emergency the of we from then that if utility care) than particular, regular less expected In is the systems. care coverage that emergency coalitions care generally, of other value health for the Intro- to potential that well. regards the assume as with suggest outcomes arise health to the could us in is allows that but care now cost, health assumption its emergency in that that just ducing possibility not the care regular considered to not inferior have we point, this To Coalitions Alternate 6 that knows similarly, Agent, opt the still PP, will by PP choice the the safely side, Given can hand rules. she right the the enforce on to And not complying. Agent with subgame, System Healthcare the Choose to How Deciding 7 Entitlement, then w elhpeim n te amnsunder payments other and premiums health own udno udn h mrec aeo h ikpoor, sick the of care emergency the funding of burden tlt rmeegnycare, emergency from utility f nadto odfeecsi auso mrec essrglrcr,w nld ihenough high include we care, regular versus emergency of values in differences to addition in If, ThemovebyEAPinFig. u h I’ rfrne r ifrn rmteUPsads ihtesame the with so and UIP’s the from different are preferences MIP’s the But ftecmie ouaini h w bv rusi ag nuht constitute to enough large is groups above two the in population combined the If iial,i elh person, wealthy a if Similarly, pi sln stefloighls(where holds following the as long as Entitlement, −2c ol rfrteEtteetoption. Entitlement the prefer would while rmpyn o mrec aeisedo eua care. regular of instead care emergency for paying from , R o comply. not R iu e nfr a htsewudpay, would she that tax uniform her minus , 7 shat aefrom care health is U rj (R E U − 3 ,i.e.if pi pi hw h eiina h alin“eld stage. “veiled” Rawlsian the at decision the shows T (R rj suiiyfo aeta h ol eev under receive would she that care from utility ’s Insurance. rj )>U − Entitlement aslwrtxsunder taxes lower pays , T pi rj )>U (I − pi ): d Insurance, Unanimity rj (E) − IT rj IT T I ). hsaodn lower a avoiding thus , pi rj d shat aefrom care health is rj hnsewudalso would she then , eehge hnher than higher were , Entitlement obndwt her with combined , o comply.Thisnot hnher than 163 In- Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

920 919 918 917 916 915 914 913 912 911 910 909 908 907 906 905 904 903 902 901 900 899 898 897 896 895 894 893 892 891 EDITOR’S890 889 888 887 886 885 884 883 882 881 PROOF880 879 878 877 876 875 rbto fpeeecsadtetcnlg ftego rvso.I h tradition the In dis- provision. good the ( the Tullock given of and “suboptimal”, technology Buchanan is the of resolution and attempted preferences of their tribution in applied rule making constitutional the at found be to are differences Welevel. institutional differences. those institutional the involve that that might suggest suggesting variation here others, the than explain issues to such seem story with some theoretical grappling regards, above. time other described easier many much so problems have in action to similar collective so sit- of democracies, these developed case of the for special All Among support mind. the to among societal come are and regulation uations banking policies, and pay control, and pollution childcare, leave similar simi- political for parental with possibly established issues and unaddressed the other or reasons—maternity unresolved are which linger There for which structure. qualities issues decision manifested system. “wrong” lar of a single-payer class be fair to a a seems of in process sort po- one supply—some democratic healthcare and to makes legitimate able general This the quite US, not apparent the is an in process least by litical at characterized that, is something for healthcare preference of financing and Organization 7Conclusion164 eet,snl ae ytm a alt eipeetdi h eiinprocedure decision the if implemented be to fail may and systems Shvetsova payer citizens and single (Sieberg for benefits, systems risks insurance care private health Hsaio than and reduces (Blumenthal it inequities that avoid assertions from ranging systems to unfit particularly are issues. such that mostly up environments evolve take political be which engender could constitutions may issue amendment, interpretation an to of by impervious “constitutionalization” and the Rigid to and benefit. access care evolve of where Health significantly losses. example efficiency one that with but fraught those is is emerge, including newly that issues, Secu- those all and Social transform address with to US the process in policy done the is of as outside much principle process, funding policy its the fixing rity. status of and the flow entitlement it the and an giving to or ebb in it right relegate manifest positive to could in- a than this that of rather practice, of (quasi-)constitutional, In technology as process. legislative issue of on-going the state treat modern socially to the the is that given dustry means healthcare This to pre- assumptions. approach socially our that preferred suggest given the tentatively unanimity that We majoritarian. the show not approximates we is it problems, policy such action producing collective for of rule ferred set this in to policy and any stage were constitutional time. it rules-choosing, a if at ex-ante, issue rule the one decision to for different rules revert a pick to preferred it have would for society possible the issue, that for ecamta hs hn”ise r opolmtcbcuetedecision- the because problematic so are issues “hung” these that claim We hr r ueosagmnsi ao fsnl-ae nilmn elhcare health entitlement single-payer of favor in arguments numerous are There the on reliance that argue to us lead findings our scale, grander a on Second, or policy, healthcare the to regard with First, two-fold. are here conclusions Our 1962 ,w hwta,gvntepeeec distribution, preference the given that, show we ), 2005 ota ti oescal efficient socially more is it that to ) 2012 .SvtoaadKK Sieberg K.K. and Shvetsova O. .Rgrls ftheir of Regardless ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

966 965 964 963 962 961 960 959 958 957 956 955 954 953 952 951 950 949 948 947 946 945 944 943 942 941 940 939 938 937 EDITOR’S936 935 934 933 932 931 930 929 928 927 PROOF926 925 924 923 922 921 oeta ,Ebe 20)Imgainadtemda oe’ netv ordsrbt income redistribute to incentive Voter’s median the and Cambridge Immigration Press, (2005) University C Harvard Eibner Econ justice. H, J of Rosenthal economy. theory A medical (1971) the J in Rawls failure 6(3):3Ð21 market Perspect and Econ insurance, J health loss? 98:187Ð194 Taxation, welfare Choice (1986) much Public how MV voter. costs: Pauly median care the Medical of (1992) income JP the Choice Newhouse Public and government. Redistribution of (1999) size P the Nelson of theory rational 89:914Ð a Econ of Polit Tests (1983) J government. SF Richard of AH, size Meltzer the of theory rational July A 9 (1981) Post, Washington SF Insur The Richard Risk insurance. AH, J social Meltzer on makes. Hayek party (2010) third D the Matthews differences health the the hospital reforming the transition: Reimbursing (1969) in HE solidarity Klarman and conse- choice, the Welfare, on (2001) committee K late, Eggleston too J, little, Kornai too coverage: without Care 3:351Ð357 (2002) Econ Medicine Health pill. of model. magic Institute voter illusory median ‘Marketization’—the a (1994) in WC Hsaio spending government and shares de- Tax for (1985) framework SB integrated Caudill an RG, development: Holcombe health course Life (2002) M Hochstein Care N, Med Halfon expenditures. care health 1917 in variation Regional (1989) The CG system. McLaughlin health-care JD, failed Howell our behind 86(1):1Ð24 idea Rev bad Econ the myth: Am moral-hazard reform. The care (2005) health M and Gladwell values, 1): Economics, 12(Suppl (1996) Aff VR Health Fuchs competition. managed of principles and health. history child The and (1993) care, AC public medical Enthoven for of utilization support eligibility, insurance the Health and (1996) voter, J Gruber median J, the Currie inequality, constitutional Income of (2010) foundations WN logical Evans consent: S, of Corcoran calculus The (1962) G Tullock JM, care Buchanan health Chinese evolving 287(5456):1207Ð1209 discontents—the Science its nations. and Privatization of (2005) wealth diver- W and than Hsaio health D, convergence The Blumenthal more (2000) D nations: Canning across DE, priorities Bloom 16(3):153Ð156 health Ethics Setting Med (2006) J resources. V care Burau health R, of distribution Blank global The (1990) R Attfield References as decided matters these have to opt prin- and ex-ante potential the constitutional. large. ignorance, this of at recognized veil have society the would are the Behind cipal coalition, overall. for particular neces- society their importance to not for of costly is best particularly issues understandably legislature ALL while the choices, decide that majority to Some shows venue model best Our the carefully. sarily selected not is itself System Healthcare the Choose to How Deciding nteUie tts ae rsne tteana etn ftesceyfrteavneetof advancement the for society the of meeting annual the at presented Paper States. United the in 24(2):629Ð675 Lit 41:403Ð418 927 36(5):553Ð566 Cambridge Press, University Cambridge Europe. Eastern Washington in Press, sector Academy National services. care health of board uninsurance, of quences 46:197Ð206 Choice Public 80(3):433Ð479 Q Milbank research. and policy, health, veloping 27(8):772Ð788 29 August Yorker, New 24Ð48 111.2:431Ð466 Econ J Q 16097 No paper working NBER education. Arbor Ann Press, Michigan of University democracy. 333(11):1165Ð1170 Med J Engl N system. 27(3):265Ð281 Policy Health Public J gence? 165 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 1012 1011 1010 1009 1008 1007 1006 1005 1004 1003 1002 1001 1000 999 998 997 996 995 994 993 992 991 990 989 988 987 986 985 984 983 EDITOR’S982 981 980 979 978 977 976 975 974 973 PROOF972 971 970 969 968 967 166 ibr K hesv 21)Ksanseokue aktaudnogla Yhdysvaltain ongelmat kattavuuden ja Kustannustehokkuuden (2012) O Shvetsova KK, Sieberg evyeholsa(fcec n oeaepolm nUie tts elhae Puheen- Healthcare) (forthcoming) States’ Studies United Administrative Tutkimus, in Hallinon problems vuoro. coverage and (Efficiency terveydenhuollossa Hungary. 2 Budapest, 30ÐJuly association,” June representation, Calculation, counts? “What socio-economics, .SvtoaadKK Sieberg K.K. and Shvetsova O. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 Modelling atII Part Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 edfae yayohrplc hnpeeecsaeageae ymjrt rule majority cannot by it aggregated and are odd preferences is when agents policy of other (Black number any the by is if defeated rule policy be other majority any simple defeats line, policies real ideal the policy over the peaked prefers (Black single transitive agent are are the preferences assumption policy, this agents’ satisfying ideal If Preferences agent’s ideal. agent’s the the policy, to ideal above agent’s closer policies the below agents two policies that two given assume given criminal to that or budget; standard and is higher policy It ideal or etc. unique lower a sentences; the % a have harsher 0 of receive or from can subset lighter vary specify item a can can policy rates by law budgeted tax policies any ordered: %; feasible easily indeed 100 of are to left/right set issues natural the policy a Many represent with If line. can endowed real issues. are we policy issue order, over policy low/high given competition or a political over represent policies feasible to the useful are models Spatial Introduction 1 Eguia X. Jon Spatial Model Euclidean Standard the to Challenges ..Eua( are Eguia paper J.X. volume. working the the of for to publication Tyson updates the Scott provide after thank Political to even I in or welcome “Advances Schofield. errors volume N. ammend the and to in Kselman Comments chapter D. suggestions. a Caballero, as G. published by be edited to Economy”, meant is paper working This candidates both convergence: (Downs policy policy median to the leads choose candidates two between competition DOI (eds.), al. et Schofield N. USA 10012, NY York, New floor, 2nd 4th, West 19 e-mail: University, York New Politics, of Department ,tesadr prahi oasm htaet aea have agents that assume to is approach standard the (1972), al. Start- issue. et given Davis a with to corresponds ing dimension represent each models bundles: spatial policy Multidimensional over issue. preferences per policy one with bundles policy (Wittman preferences policy diverging have candidates oiia optto sal novsmlil oiyise.Cniae propose Candidates issues. policy multiple involves usually competition Political 10.1007/978-3-642-35239-3_8 [email protected] 1958 B .Snetemda oiycno edfae yayohr electoral other, any by defeated be cannot policy median the Since ). ) 1948 dacsi oiia Economy Political in Advances ;frhroe h einielplc mn l h agents’ the all among policy ideal median the furthermore, ); pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , 1957 uligo oeln’ ( Hotelling’s on building , , 1983 ;Calvert ) vni the if even 1929)), single-peaked. 1985 ). 169 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 h nrdcin xoeta ucin r loue PoeadRosenthal and (Poole used also are functions exponential Introduction, the r h otcmol sd(cat n Meirowitz and (McCarty used commonly functions loss most quadratic the loss or are curves. Linear the curves. on indifference indifference assumption on these default with assumption or associated standard standard function a inconse- a on is exists is consensus point This similar ideal utility. no However, agent’s her the or identi- from his valued perturbation for are the point quential policy of ideal agent’s ‘direction’ two an multi- the that from i.e. in distances such cally, identical preferences are at curves are on which indifference assumption points Circular common models. a spatial dimensional are curves indifference Circular Function Loss the About Concerns 2 in decreasing (Kramer are linear a that with utilities typically point, and this space, to policy distance Euclidean the the in alternative preferred most 170 2 of model voting probabilistic a 1 in instance, For function. loss the in of theories concavity various the the in function simplicity. or utility convenience the by motivated of appears form literature functional the the of choice The 1977 dersen necre u anantersrcin htuiiyfntosb ifrnibe(Plott differentiable be functions utility that 1967 restrictions the in- maintain of but curves shape ence Euclidean circular the (McKelvey preserve curves they difference but functions, utility general more I.Cnen bu h hp fidfeec uvs eaaiiyars issues. across separability curves: indifference of shape the about Concerns III. classes: three with functions. alone utility let preferences Euclidean functions, weighted whether utility or question quasiconcave Euclidean or into differentiable call by multidimensional that represented be a results can in utility present Euclidean I by generally, exponential represented More or adequately space. quadratic be linear, can are issues that multiple functions over preferences that tion Duggan and (Banks cave ht(tnad ocv tlt ucin ontfi h aawell. data the fit not do functions utility concave (standard) that ( repnnil(ol n Rosenthal and (Poole exponential or rvdsa xoaiaino ulda tlte ihaqaiieradtv aec term. valence additive quasilinear a with utilities Euclidean of axiomatization an provides 2011) I ocrsaottesaeo nifrnecre:cneiy n different and convexity, curves: indifference of shape the about Concerns II. nspoto hi supino xoeta tlt ucin,PoeadRsnhl( 1997 Rosenthal and Poole functions, utility exponential of assumption their of support In ( Dardanoni and D’Agostino .Cnen bu h ocvt ftels ucin cetn h Euclidean the accepting function, loss the of concavity the about Concerns I. h hieo osfntosi osqeta:ipratrslsrl rcal on crucially rely results important consequential: is functions loss of choice The iieteetertcladeprclcalne osadr supin in assumptions standard to challenges empirical and theoretical these divide I assump- the challenge that results empirical and theoretical of series a present I egt o ifrn dimensions. different for weights curves. indifference the of shape at tal. et Patty ; Schofield ; 1992 lno tal. et Clinton ; 1978 2009 Duggan ; ea n Merlo and Degan ; 2008 rvd naimtzto fteEciendsac;Azrieli distance; Euclidean the of axiomatization an provide 2009) 2004 1976 ,o ifrnibeadqaiocv (Kramer quasiconcave and differentiable or ), 2007 coedadSened and Schofield ; ,o hyrlxteasmto fcrua indiffer- circular of assumption the relax they or ), rDga n Kalandrakis and Duggan or ; 1985 osfunction. loss ) 2009 rEguia or ; 2007 2006 1 te hoisalwfor allow theories Other et .) sntdin noted As 2.5). Sect. , rSchofield or ; 2012 ,qartc(Fed- quadratic ), 2012 1977 ,quasicon- ), Wittman ; ..Eguia J.X. 2007b 1985 1973 argue ) , ). a) ). 2 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 eiso n su ihotoe nohrise)rqie httels ucinbe function loss the that requires issues) other on (Eguia outcomes quadratic with issue lot- one over on preferences weighted function of teries in independence decrease utility (i.e. that separability Euclidean utilities additive With multi-dimensional linear distances, neutral. with Euclidean risk a voters not words, with are other utilities voter In is- Euclidean a issues. linear given other of a on on attitude outcomes on lotteries risk depend over the preferences hence the and but so neutral, sue separability, additive risk satisfy are not voters do weighted that linear evidence are as (2009 utilities finding Eguia voters’ this that interpret assume They to Euclidean. appropriate is it that suggesting aeta rvdsabs tfrvtrcocsi Speieta lcin is elections presidential US in choices voter for fit best a provides that mate hni sntpsil orsaetedmnin oa oueuwihe Euclidean unweighted Schofield weights use and different (Miller with to utilities voters as Euclidean different so weighted for use dimensions instead the dimensions, must rescale various we and to the dif- utilities, possible to If not weight. weights is its relative it to different then according assign dimension voters each of reinstated, of groups circles measure ferent Euclidean of and to units weights solved, the same trivially rescaling the by is assign indiffer- voters problem circular) all the If than dimensions. dimensions, two (rather hence these Eu- with and elliptical case the others, the generating on in than weighted, curves important depend ence be more that be to functions may ought issues utility utilities some of that assumption is the distance clidean about concern first A Preferences of Convexity About Concerns 3 are functions convex. utility sufficiently voters’ are functions if utility voters’ median if the diverge candidates to but ( concave, converge Kojima candidates and Kamada equilibrium candidates, in two with competition electoral Model Spatial Euclidean Standard the to Challenges rblt.Ti ed st h eodcaso ocrs ocrsaottesaeof shape the sep- about additive concerns with concerns: consistent of is class curves. that indifference second function the the obtain loss to indifference we the us Euclidean curves, leads for indifference This parameter of arability. the a assumption of with shapes the fit different best discard under a if to func- check is to loss fit) and linear curves, best a curves the indifference provides Berinski Euclidean with tion (with function) finding loss (2007) quadratic Lewis’s a and requires curves indifference Euclidean der ulda itneand distance Euclidean t ucin aeteform the take functions ity the of which clear not is it appropriate.” evidence, more empirical conclude is convincing I assumptions any here. of relevance absence any the has in decision-making that economic to in easier averse makes risk it are because ple [ However, second exists. and equilibrium aversion’ an that ‘risk show with associated is it because son ( Osborne ekn ots oes ikattd,Brnk n ei ( Lewis and Berinsky attitude, risk voters’ test to Seeking at xoai ob nti nepeain ierEcienutilities Euclidean linear interpretation: this on doubt axiomatic casts ) h supino ocvt sotnaotd first adopted, often is concavity of assumption “the that warns 1995) 2011b .Teol a orcnieadtv eaaiiy(hc un- (which separability additive reconcile to way only The ). α saprmtrt eetmtd hyfidta h esti- the that find They estimated. be to parameter a is u i x x (x, k ) =− 2003 d(x,x ...]itisnotclearthatevi ). i ∗ ) α where , 2007 d(x,x sueta util- that assume ) i ∗ hwthat show 2010) ) ec htpeo- that dence saweighted a is α ˆ ≈ 171 1, Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 iybokpeeecs nfc,tecr fsml aoiyrl sntepyunder empty not Taylor is Thorson and rule (Rae and preferences majority Wendell city-block simple have agents of if core conditions general the more fact, In preferences. block city 2009 ne (Shepard ences nNreineetos oee,abnr oprsnbtenct lc utilities choice block city vote between predict the comparison on to binary based a aiming However, when elections. Norwegian preferences, in preferences. Euclidean Euclidean quadratic linear with elections have model presidential voters US assumes in (1997 that choice Westholm vote model of alternative fit an better than a provides preferences block agents issues, separable function. block on city bundle the policy to implies a according which is distance attribute, consideration measure each under in object distance the the if up that adding by attributes separable with ae nthe on based hc aaeeie by parameterized which l hc en qaeidfeec uvs(ihsurstle ta4 ereangle are degree that 45 functions utility a the by representable at in are decreasing tilted and coordinates), squares of axes (with the to curves relative indifference square define which adequate not simply preferences. nor is the unweighted utilities represent Euclidean neither to by convex, functions, imposed curvature utility strictly The Euclidean not ones. by weighted are represented preferences policy be If to cannot functions. preferable they policies utility quasiconcave of set (strictly) the indiffer-by each if any by is, for defined convex, that convex; set is (strictly) contour is upper curve the preferences. ence if ellipti- convex convex not of (strictly) class are are particular that a Preferences shapes represent have utilities Euclidean may Weighted curves cal. indifference functions: utility clidean tityqaiocv,adi sntdfeetal.Cascrslso h instability the on not results (Plott is Classic rule representation differentiable. majority utility not simple their is of block, it city this and are to quasiconcave, according preferences strictly ideal If their distance. to of closer points notion prefer they and distance”), “Manhattan 172 nissue on ftaeigo rd(hti h the why as is dimension, (that by grid dimension a distance the on up traveling adding if by points two between distance 1 and upry n ae ( Laver and Humphreys rnvsiadCria ( Corrigan and Grynaviski natraieasmto oEcienpeeecsi iybokpreferences, city-block is preferences Euclidean to assumption alternative An Eu- weighted by representable be not may preferences that is concern deeper A ). l 2 k r pca ae fteMnosi(1886 Minkowski the of cases special are ∈{ l l 1 2 1,...,K metric metric 1987 nsta oe ihct lc rfrne uprom a outperforms preferences block city with model a that finds ) l 1 x ersnal srcl)cne rfrne r representable are preferences convex (strictly) Representable . distance Arabie ;   1974 } x x hti,aet ihct lc rfrne aclt the calculate preferences block city with agents is, That . δ   − − ie h itnebetween distance the gives , x 2009 x cevyadWendell and McKelvey ; x − ∗ ∗ 1991 2006    x 1967 x 1 2 noerslsfo scooyadcgiiesci- cognitive and psychology from results invoke ) ∗ = −   = δ oageta gnsmauedsac oobjects to distance measure agents that argue to ) n htamdlta sue oeshv city have voters assumes that model a that find ) x ( McKelvey ; = ∗ k K  l  =1 1 k K 1 =1 k = rct lc itnei oeie called sometimes is distance block city or K =1 | x (x k x k − k k K − =1 x − 1976 k x ∗ | x | k ∗ x k ) n h ierEcienutilities Euclidean linear the and ∗ k 2 ) δ − aiyo ercfunctions, metric of family ) ontapyi gnshave agents if apply not do ) 1976 x 2 1  x suncsaiyrestrictive: unnecessarily is and 1 δ k ∗ . | upry n Laver and Humphreys ; where , x ∗ as: x k stepolicy the is ..Eguia J.X. 1971 (1) x ; Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 a ersn h aeudryn rfrne sn mapping a using that preferences such underlying same the represent can htteaethsalna tlt ucinover function utility linear a has agent the that f between pc ed oidfeec uvso ifrn hps h pta ersnainof representation spatial The vector shapes. a different into to of alternatives wishes of curves who set indifference theorist the to the of leads for mapping choice space different of a object using an individual: is an function study utility the of shape the fudryn rfrne.I,frisac,teeaetrepolicies three representations are are there functions instance, utility for agent the If, preferences. as underlying just of convenience, for used resentation tlt ucinover function utility per oefutu oetmt parameter estimate to fruitful more appears comparing than Rather Model Spatial Euclidean Standard the to Challenges δ that finds and elections regional Spanish tityqaiocv n ifrnibeuiiyfntos while functions, utility differentiable and quasiconcave strictly ihlna iybokuiiiscrepnigto corresponding utilities δ block city linear with h hp fteuiiyfntos(ulda,ct lc,Mnosiwt parame- with Minkowski block, city (Euclidean, functions utility ter the of shape the over preferences have eaal unless separable functions. ssgicnl ifrn from different significantly is aie seoeosygvn hr sasubset a is there given: exogenously as natives is preferences convex of assumption standard the whether and justified. fit, better a provide most and, elections across greatly results vary candidates. their across that However, puzzlingly, elections. estimates Presidential obtaining 2008 inconclusive, and are 2004 2000, the to sponding ( function rfrne htaentcne,aduiiyfntosta r ete tityqua- strictly neither parameter are estimate (2011) that al. functions et Ye utility differentiable. and nor convex, siconcave, not are that preferences htaet aeon have agents that ierydcesn nthe in decreasing linearly = = :{ oee,tesailrpeetto ftesto esbeplce sisl rep- a itself is policies feasible of set the of representation spatial the However, r o additively not are (1) expression in decreasing linearly are that functions Utility oto h ieaue n l ftedsuso bv,cniestesto alter- of set the considers above, discussion the of all and literature, the of Most functions utility which establish to necessary appears work empirical Further δ .Teetssspotteueo ierct lc vrlna ulda utility Euclidean linear over block city linear of use the support tests These 2. x,y,z .Ntc htayparameter any that Notice 2. saqeto nwa rmtv rfrne vratraie ow believe we do alternatives over preferences primitive what on question a is ) i prefers x f(x) n oigdt rmteAeia ainlEeto tde corre- Studies Election National American the from data voting and 2) }→ and z = R x hnw a a h he oiist h elln sn mapping a using line real the to policies three the map can we then , δ 0, uhthat such to = X f(y) [ y .T aif diiesprblt,teuiiyfnto utbe must function utility the separability, additive satisfy To 1. 0, ⊆ X to 1] δ nti iw h usino h dqaeasmto on assumption adequate the on question the view, this In . R = z δ = K ihielpita .Udrti esetv,w e that see we perspective, this Under 0. at point ideal with n agent and , u f(x) oe of power  . lna iybok and block) city (linear 1 x,x 2 1 δ and = = ∗ δ> 0, ,adte r l infiatydfeetfrom different significantly all are they and 1, =− f(z)  f(y) x i eut nsrcl ovxpeeecsand preferences convex strictly in results 1 − sidfeetbetween indifferent is k = K =1 x δ = ˆ δ ∗ n a htteaethsaquadratic a has agent the that say and 1 ∈ .Rivero(  x and 0.5 δ k ( othat so , [ 0.92, − X 0, δ = x 1] ⊆ k ∗ ,adqartcEciento Euclidean quadratic and 1, f(z) 1.17) R ihielpita .Btwe But 0. at point ideal with δ 2011 , δ K = hti ie,adagents and given, is that = oeo h estimates the of none ; estimates ) lna ulda) it Euclidean), (linear 2 n hnw a say can we then and 1 y δ g n arlottery fair a and δ< sn h utility the using :{ x , x,y,z eut on results 1 δ y o several for and }→ z and 173 (2) R Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 h e fatraie into alternatives of set the ahr ems aesbetv aso h e ftesto lentvs n o ahagent. each for one alternatives, of representation; spatial set unique the a of use set cannot the we so, of If maps issue. subjective given have a must on another we of rather, left or right the to is ftesto alternatives all of for set Euclidean the are of that with functions profile utility preference by any dimension- sented result: great positive with a representation obtain spatial we a ality, accept we If functions? utility clidean eaefe oasm n ereo uvtr,icuigpeeecsta r not are that preferences including curvature, of degree choosing any by curvature. assume convex elliptical to or circular free have are curves We indifference that assumption the relax to (1)or(Expressions Preferences of Separability About Concerns 4 Euclidean by represented be policies can over we preferences preferences, functions all model utility to can wishes questions: who new theorist ask the by made choice agent. determine endogenous the jointly of space preferences this underlying in the use on we make function we utility assumptions the the and alternatives of set the 174 3 rfrne falaet a erpeetdi oempwt uscnaeutility quasiconcave with the question. map empirical some that open in such an represented is are be functions application can agents given all any of differ- in rep- preferences and/or utility profiles quasiconcavity the violate preference that agents Whether such other entiability. of be preferences may the this the of achieves on resentations that depending However, mapping map. any chosen profile, the preference over functions utility quasiconcave by oany to in naysailrpeetto nesteudryn rfrnepol satisfies profile preference underlying (Eguia the conditions restrictive unless very representation cannot spatial we any alternatives, in of tions all set of the over preferences of functions the utility representation represent Euclidean spatial quadratic prefer- chosen using separable appropriately individual single-peaked, an single any repre- a spatial represent our of can in relation we issue exogenously ence per while are dimension case, issues one this than policy In more the no sentation. Suppose use to result. want positive we a and given, such obtain not do we Laslier ria rfrne vratraie nagvnised o eedo h realized the on depend not do issues. issue other given on a outcome on alternatives over preferences ordinal oei ob loae ewe ulcsedn nplc n,pbi pnigon spending public one, in- policy national on of spending public unit between fixed allocated a be Suppose to separable. is be come possibly cannot spending public nlz nadtoa opiain gnsmyntareo hc alternative which on agree not may agents complication: additional an analyze (2012) al. et Calvo o n igepae rfrnepol ihsprbepeeecs ecnmap can we preferences, separable with profile preference single-peaked any For new eonz httesailrpeetto ftesto lentvsi an is alternatives of set the of representation spatial the that recognize we Once iy ( Milyo δ> 2007 2000b rsreteasmto htpeeecsaesprbears issues: across separable are preferences that assumption the preserve 0 .I ecr o h ubro iesosi u pta representation, spatial our in dimensions of number the for care we If ). nsm space? some in n ( and ) bv,o aitoswt egt o ahdmnin lo us allow dimension, each for weights with variations or above, 2) δ< 2000a .Teegnrlztoso h tnadmdlfrom model standard the of generalizations These 1. X R into K oe htpeeecsoe utpedmnin of dimensions multiple over preferences that notes ) N oa orpeettepeeecso ie agent given a of preferences the represent to as so R fnt htpeeecscnb ersne yEu- by represented be can preferences what not, if 2011a K niiul ihqartcEcienuiiyfunc- utility Euclidean quadratic with individuals contain ). 3 K ≥ N N iesos(oooni and (Bogomolnaia dimensions gnsi elttemapping the let we if agents N gnscnb repre- be can agents ..Eguia J.X. δ = 2 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 r w opeetr sus uhta o n agent any for that such there issues, that endogenous complementary new, Suppose two preferences. considering are separable by have addressed agents is which over problem dimensions the policy then agents, all for same issues. of pairs redistributive various so across lax ( non-separability Lacy is such immigrants. policy favor redistribu- immigration part in so if would restrictive than policies is natives, policy only immigration pre- benefit if policies net, net tive safety safety have social generous agents the more that actually a and possible issue ferring policy is given it immigration instance, a about For preferences on issues. non-separable other values on ideal outcomes agents’ the on because depend but concerns, budgetary to ne a elb eaal ne hsrpeetto ftesto sus n in and issues, of set the of Prefer- ( representation one. Milyo’s this escape issue they under to the case separable devoted any let be spending and public well spending, of may public fraction ences total the be have be dimension agents dimension that first easily second assume the is we problem let which over This over preferences: Preferences separable. dimensions separable one. policy be issue the cannot redefining on two by and spent solved one amount issues the on with spending expenditures decrease of public amount must ideal the two opportu- so the issue increases, increases, also on two one issue policy on on spending spending of public cost nity as that consumption means over goods utility public marginal of Decreasing consumption. private and two, policy of vectors basis new a using by 1 Fig. Model Spatial Euclidean Standard the to Challenges iesos o ntne eunn onnsprblt ewe mirto and immigration defined newly between our non-separability over to separable returning agents instance, all make For case, to dimensions. this dimensions In create agents. across cannot heterogeneous we is issues different on preferences between dimensions. endogenous Fig. in coordinates of axes vrtetoise.Hwvr fw s ifrn ai fvcos sdpce in depicted as vectors, of basis different a use we if Fig. However, Fig. issues. in two agents the arbitrary over two for depicted functions, utility These faet aennsprbepeeecs u h orlto ewe susi the is issues between correlation the but preferences, non-separable have agents If due not non-separable, truly are preferences if arises difficulty insidious more A hsslto al faet aennsprbepeeecsadtecorrelation the and preferences non-separable have agents if fails solution This 1 n osdrtenwtodmninlvco pc ie ytetotilted two the by given space vector dimensional two new the consider and , bann separability Obtaining u(x 1 ,x 2 ) =− x 1 hnaet aesprbepeeecsoe h new, the over preferences separable have agents then 1, − 2000b x 1 i 2 − n ( and ) x 2 − 2000a x 2 i 2001a 2 critique. ) + , x b, 1 i , 2012 − x 1 i noeseiec of evidence uncovers ) x r o separable not are 1, 2 − x 2 i . 175 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 fsml aoiyrl sln speeecsaesot (Plott smooth are preferences as long as rule majority simple of h tutrdidcdeulbimter SeseadWeingast and (Shepsle theory equilibrium structured-induced the 1979 snteog opeitvt choice. vote predict to enough not con- is example, Rosenthal and second (Poole a literature For 2004 estimation outcome. point policy dimen- ideal policy chosen the each the sider affects which considered in are is order preferences sion the because irrelevant, preferences, is non-separable legisla- With dimensions the separable. policy which In various in dimension. the order by considers the dimension theory, ture policy equilibrium choosing structured-induced by standard solved the is instability the that poses aro susadpraeto ru faet h iare o aiu of maximum a for disagree, who agents of group or possible agent per per parameter and correlation one issues add of correlation must the pair on we disagree issues, who between agents preferences of preferences in the represent to order In sions. e ieso e gn rgopo gnswodsge nteewihs o a for weights, these on disagree parameter who one agents add of to group need of on or we maximum disagree agent dimensions who per different agents of dimension the preferences to per the assign parameters represent they more accurately weights need to the we order because In weights it. damaging, relative fix more different to -but assign consider- who dimensions a agents various of is the intro- problem This to to the issues. to instead across similar need correlation We setback, map. the able whole accommodate the to tilting parameters by duce problem this correct cannot we return to map whole dimensions. the defined tilt newly can over we model tilted, standard equally doing a curves curves to indifference in have non- agents their but very all if In if separable, preferences tilted; non-separable. agents are non-separable remain have agents of of agents group group terms, redraw one other technical can the of we of preferences preferences then the the lax, so, make is not to policy to immigration axes net safety when the immigrants smaller immigrants needy a help on prefer agents to money other net spend while safety lax, larger is a policy prefer immigration agents when some if net, safety social 176 sayqaiyta l oesarei od n ae h addt h possesses who candidate the makes and good, Valence is candidates. agree candidate about voters voters a all of add that preferences must quality actual we any is the that argues capture literature to of term valence body increasing an instance, For general more preferences. the smooth but and/or Eu- models, convex separable, of spatial of assumption multidimensional assumptions standard in the functions only utility not clidean questions work empirical and Theoretical Discussion 5 K(K−1 hl iltoso eaaiiyd o fetcascrslso h instability the on results classic affect not do separability of violations While f ntecnrr,dfeetaet aepeeecstle ndfeetdirections, different in tilted preferences have agents different contrary, the on If, tnadsailmdl ufrfo iiain htIhv o osdrdhere. considered not have I that limitations from suffer models spatial Standard 2 ,te fethww a nepe n s omnsailmdl.Consider models. spatial common use and interpret can we how affect they ), :i rfrne r o eaal,etmtn h da on fec legislator each of point ideal the estimating separable, not are preferences if ): ) N e parameters. new (K − 1)(N − 1) e aaeesi hr are there if parameters new N gnsand agents 1985 1981 1967 lno tal. et Clinton ; ,wihpro- which ), McKelvey ; K ..Eguia J.X. dimen- Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 rbeo o vrsc iesos hc r o rcsl end eetees if Nevertheless, defined. precisely not sep- are which are dimensions, preferences such whether over determine not to or difficult arable left/progressive more (from is issues It to cultural right/conservative). left/pro-state all to all (from includes of issues them. that list economic another over the groups and collapse separable that right/pro-market) they one be rather, dimensions, issues; to two many onto assumed include issues not not do are models functions spatial Many utility issues, such include iso lcoae hudtk noacuteitneeiec nnon-separability on (Lacy poli- evidence issues world existence of account real pairs into specific various take across to should models electorates spatial or or of ties competition application electoral However, of models choice. on policy results convergence or existence equilibrium model. the for fit poorer a provides htaet aeuiiyfntoswt parameter with functions utility have agents that gnshv tlt ucin ihtecrauecrepnigt h etestimate best the to corresponding curvature the with of functions utility have agents uho h hoeia ieauewl evldtd h anipc fotiiga obtaining of impact main The validated. of be estimate will better literature theoretical the of much and (Humphreys preferences block city on rely that Laver results Whereas, notably apply. most not assumptions, (Plott differen- these rule majority on nor of rely quasiconcave, instability that neither the literature are the in functions results utility Standard tiable. hence and convex, not (2)fortheare form functional separability, parameter additive eter estimate satisfy functional must work to the empirical want Assuming functions, we utility issues. if policy or, multiple (1) with form bundles the policy about over chapter erences this in theory. the with posed in theory challenges preferences any spatial the but of theories, address formalization richer appropriate must generate to component model spatial spatial a basic be can the improvements can to other added that or be preferences preferences uncer- policies, other-regarding dynamics, rationality, Valence, feasible bounded functions. over of tainty, utility preferences set convenient have analytically the agents by represents that represented that assumption space the vector model: spatial a the of pillar basic 2010 oipoetefi ffrhreprclwr nielpitetmto oes(Clin- models estimation point ideal on work al. empirical et further ton of fit the improve to utility differentiable or quasiconcave by agents representable with functions. environments not in are quasiconcavity apply preferences and of robust assumptions whose es- are under to preferences obtained needed of literature differentiability be the or would in work results theoretical what further tablish and relevant, more become would and (Ashworth location spatial candidate’s endog- the to to seeks Mesquita valence relation de on its Bueno research analyze Current to voters. and all it to enize attractive more it of more Model Spatial Euclidean Standard the to Challenges δ ihrgr osprblt,voain fteasmto yial ontaffect not do typically assumption the of violations separability, to regard With ute miia oki eesr oetbihwehraet aecne pref- convex have agents whether establish to necessary is work empirical Further fteetmtdparameter estimated the If δ ˆ , ihnteprmtrzdfml fuiiyfntos( functions utility of family parameterized the within sls hn1 h osqecsfrtertclwr r rmtc Preferences dramatic: are work theoretical for consequences the 1, than less is 2009 2012 2004 ro o-ifrnibeuiiyfntos(aaaadKojima and (Kamada functions utility non-differentiable on or ) rShfil tal. et Schofield or ; ol n Rosenthal and Poole ; δ nuiiyfntoso h om( form the of functions utility in 2009 Zakharov ; δ ˆ scnitnl rae hn1 vni ti o er2, near not is it if even 1, than greater consistently is 2011 2001a 1985 2009 1967 .I hscatrIaayecnen bu a about concerns analyze I chapter this In ). ,o oecoc oes yasmn that assuming by models, choice vote or ), Serra ; McKelvey ; , b, 2012 δ 2010 = ,s hti h oesexplicitly models the if that so ), htis that 2) vntog parameter though even 2 1976 and δ 2012 Schofield ; fteetmtdparam- estimated the If . ,isedo assuming of instead 2), δ ˆ = rs n Polborn and Krasa ; 2but 1978 δ> ˆ ilbe will 1 ,would ), 2010 δ = 177 2 ) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 lno D aka D iesD(04 h ttsia nlsso olcl aa nfidap- unified a data: call roll of analysis statistical The (2004) D Rivers information SD, lensing, Jackman ideological JD, beholder: Clinton the of eye The (2012) K Chang T, Hellwig E, uncer- motivations, Calvo candidate model: voting multidimensional 43:87Ð98 the Econ of Math Robustness J Cambridge (1985) preferences. Press, R University Euclidean Calvert Cambridge (2007) Sci elections. J-F 56(1):23Ð34 and Laslier Polit Econ committees A, J Polit of Bogomolnaia J Q theory decision-making. The electorate. group (1958) U.S. of D rationale the Black the in On aversion (1948) D risk Black of estimate An (2007) policy JB multidimensional Lewis in elections AJ, democratic Berinsky of model dynamic A 47(4Ð (2008) Econ J Math Duggan J JS, dimension. Banks valence a with preferences Euclidean for Axioms Psychometrika (2011) Y Games competition. Azrieli psychologist? valence and platform sophisticated with Elections (2009) unnecessarily E Mesquita de and Bueno S, Ashworth Euclid Was (1991) P Arabie References empirical future these with consistent non- are (bet- that differentiated findings. Future assumptions assigned legislators. make of or weights must voters the theories evidence of ter) in groups different differences systematic across systematic dimensions, find or different to issues, can of pairs we over separability whether and understand to smooth, ability our in study. we impaired the- processes our are political and we the separable, are, predict nor they and smooth, that and convex assume convex model. are models not we oretical preferences are processes preferences that political If evidence smooth. the empirical and understand convincing to lack ability shortcut we simplifying our Nevertheless, a on only effect are preferences limited Euclidean smooth with and that convex conjecture are but to Euclidean us not allowed are preferences if the- robust several are that results show oretical that Generalizations policy-making. government legislative competition, and electoral formation explain can that models spatial agent multidimensional each for dimensions between correlation of agents. degree of a group only achieved, or also not be estimating but cannot preferences, point, non-separable this ideal for if allow an pref- to else, that necessary or way be dimensions, may a it new such then the in over issues) new separable the define weighing are to or erences seek bundling either of should ways prefer- models (new between issues, dimensions cultural correlation and systematic economic a across of ences evidence reveals work empirical future 178 rah mPltSiRv98:355Ð370 Rev Sci Polit Am proach. the of applications contemporary on conference March model Juan spatial Fundacion In: vote. the and effects, 29:69Ð95 Sci Polit J Am convergence. and tainty 2:139Ð154 3:269Ð299 Sci Polit J Q spaces. 5):545Ð553 67(1):191Ð216 Behav Econ 56(4):567Ð587 uueeprclwr hl sals hte rfrne r ovxand convex are preferences whether establish shall work empirical Future of development the in tool useful extremely an been have preferences Euclidean ..Eguia J.X. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 son 19)Sailmdl fpltclcmeiinudrpuaiyrl:asre fsome of survey a rule: plurality under competition political of Choice models Spatial Public (1995) critique. Leipzig M Teubner, constructive Zahlen. Osborne der a Geometrie models: (1886) H Lett spatial Minkowski Econ in politics. deficiencies of Logical models (2000b) spatial J in preferences Milyo Euclidean with problem Sci A Polit Am (2000a) States. J United Milyo the in realignment partisan Math and spaces. Activists (2003) choice N multidimensional Schofield models. G, in Miller voting equilibria Voting formal (1976) in RE intransitivities Wendell global RD, McKelvey for conditions General (1979) RD McKelvey for implications some and Cambridge models Press, voting University multidimensional Cambridge in Intransitivities theory. (1976) game RD Political McKelvey Fun- (2007) In: A elections. Meirowitz in N, packaging McCarty issue Sci and voting, Polit responses. issue survey preferences, J Nonseparable unstable Am (2012) and D responses. Lacy error, survey measurement preferences, in Nonseparable preferences (2001b) nonseparable D of Lacy theory A Rev (2001a) Econ Sci Games D candidates. Polit Lacy differentiated between Am competition candidates. Political (2012) specialized M Polborn between S, Competition Krasa (2010) 16(2):310Ð334 Theory M Econ J Polborn equilibrium. political S, of model Krasa dynamical 41(2):285Ð A Econometrica (1977) rule. GH majority Kramer for conditions economic of paper class Working a policies. equilibria. On electoral rule (1973) and GH polarization majority preferences, Kramer Voter and (2010) F metrics Kojima Y, cognitive Kamada models, Spatial (2009) M 39(153):41Ð57 Laver J M, Econ competition. Humphreys in Stability evalua- (1929) candidate H of Hotelling models proximity in issues Specification (2006) Sci BE Polit Corrigan J JD, Am Grynaviski turnout. positive and law 49(3):549Ð570 Duverger’s implying Theory model Econ voting formation. A party (1992) of TJ 47(2):200Ð205 theory Feddersen Econ spatial Math A J (2012) preferences. JX for spatial Eguia (accepted of Foundations Theory (2011b) Econ JX profiles. Eguia preference of representation spatial the On (2011a) Polit JX J Eguia Q dimensions. multiple in preferences Theory neutral Econ risk J of representations bargaining. Utility policy (2009) legislative JX Dynamic Eguia (2012) T Kalandrakis J, Games elections. Duggan of models spatial 65(2):135Ð and Econ games zero-sum Polit for J existence democracy. Equilibrium (2007) a J in Duggan action political 144(5):1868Ð1894 of Theory theory Econ economic J An ideologically? vote (1957) voters A Do Downs (2009) A Econo- Merlo rule. majority A, Degan and orderings preference Social (1972) MJ Hinich MH, Welf DeGroot Choice OA, Soc Davis distance? Euclidean about special so What’s (2009) V Dardanoni M, D’Agostino Model Spatial Euclidean Standard the to Challenges xlntoso h ubro addtsadtepstoste ae a cn28(2):261Ð Econ J Can take. they positions the and candidates of 301 number the of explanations 105(3Ð4):273Ð289 66(2):179Ð182 97(2):245Ð260 Rev 1(2):144Ð158 Res Oper 47(5):1085Ð1112 Econometrica 12:472Ð482 Theory Econ J control. agenda model spatial the of applications contemporary on conference March Juan dacion 9(2):1Ð21 Anal Polit 45(2):239Ð258 76(1):249Ð271 Behav 104(4):745Ð765 297 40(1):11Ð30 Sci Polit J Br 14:393Ð420 Anal Polit importance). issue (with tion 36(4):938Ð962 publication) 4(4):379Ð385 Sci 60:52Ð74 Behav Econ 150 40(1):147Ð157 metrica 33(2):211Ð233 179 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 ahrvA(09 oe fcniaelcto ihedgnu aec.Pbi Choice Public valence. endogenous with location candidate of model space. A non-Euclidean a (2009) in A proximity voter-candidate Zakharov Evaluating Theory (2011) KW Econ Leiker Q, J 77(1):142Ð Li Rev model. Sci M, Polit Ye Am dynamic alternatives. of a synthesis a preferences: motivation: Candidate (1983) policy D Wittman with Candidates elec- of (1977) theory proximity D the of Wittman rule. defeat majority illusory the under direction: decisions versus Distance social (1997) of A Westholm generalizations Some (1974) Public SJ choice. Thorson legislative RE, and Wendell equilibrium Structure-induced (1981) Science BR science. Weingast psychological KA, for generalization Shepsle of law universal a Polit Toward ICO- J valence. (1987) of valence. RN types Shepard two endogenous with with election an elections experience? discourage of charisma Does model (2012) G A Serra what? of Cambridge Polarization Press, (2010) University Cambridge G democracy. Multiparty Serra (2006) I Sened N, Schofield United the of models formal and Empirical (2011) U Ozdemir convergent M, Gallego for C, conditions Claassen sufficient N, Schofield and necessary theorem: 28:461Ð490 voter Welf mean Choice The Soc (2007b) uncertainty. 40(3):575Ð594 electoral N Stud with Schofield Econ equilibria Rev Political games. (2007a) dynamic N simple Schofield of Instability (1978) N Schofield na- and ideology models: 1:71Ð90 voting Sci multidimensional Polit in J separability Br and outcomes. Integrality policy (2011) and Oxford G rules voting. Rivero Decision call (1971) M roll Taylor 29:331Ð347 of DW, Sci history Rae Polit political-economic J Am Rev a analysis. Congress: Econ call (1997) Am roll H for rule. Rosenthal model majority spatial KT, A under Poole (1985) possibility H Rosenthal its K, and Poole equilibrium of voting notion strategic A equilibrium: (1967) an CR three’s company, Plott Two’s (2009) MM Ting JM, Snyder JW, Patty 180 138:347Ð366 4:497Ð521 Parties Opin Public Elect J 157 14(1):180Ð189 91(4):865Ð883 Rev Sci Polit Am choice. toral 42:893Ð912 Econometrica 37(3):503Ð519 Choice 237:1317Ð1323 paper conference II PEAI Political 72(2):426Ð437 (eds) G Caballero N, Schofield Heidelberg In: Springer, voting. 2004. and and democracy institutions, 2000 of economy in elections presidential States 74:965Ð980 Stud Econ Rev equilibrium. Fun- Sciences, Social the December in 2011/265, Study March, Advanced Juan for dacion Center elections. regional Spanish in tionalism London Press, University 57(4):331Ð347 4(3):251Ð278 Sci Polit J Q elections. multicandidate and ..Eguia J.X. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 oiydbts u ahri ietadcnign xhneo oe o other (or votes of exchange contingent and participation direct public political national-level a of in in not forms rather grounded but are debates, linkages Such policy officials. elected and zens issrtgclycos:()terpormai oiypsto,()teefr they effort the (2) par- in position, political work-horse policy which programmatic the in their model been (1) spatial decades choose: a strategically for develops ties paper has This which theory. political model, concerns formal literature spatial clientelistic the traditional introduces to the explicitly contributions none into valuable exchange, make electoral all contingent While setting. on theoretic game a in inducements. choose targeted voters provide issues, to programmatic ability more parties’ or on one evaluatingbased to on addition stances electric in policy contexts, the parties’ such to In political access etc. sector, fuel, public alcohol, the machines, washing in grid, jobs forms: many take inducements These investigates political-economy comparative and .Shfil ta.(eds.), al. et Schofield N. 28006 77, C/Castelló Institute, e-mail: March Juan Sciences, Social Spain the Madrid, in Studies Advanced for Center DOI ruddi apinadgvrac taeiso suso ainllvlpbi policy. public national-level of issues on strategies ( governance Rokkan and campaign and in Lipset grounded to least at dates isacrigt hi rfrne vrtesm su iesos u otherwise, Put dimensions. par- issue these same among analyze the from typically over choose models preferences spatial then their voters to dimensions; according issue posi- more ties announce or typically one parties political on competition, tions political of models spatial In Introduction 1 Kselman Daniel Models Spatial in Clientelism for Theorem Non-existence A .Kemn( Kselman D. 1 oe hieadeiebhvo scnumtdidrcl,vacletvl applica- collectively via indirectly, issues. consummated policy is ble behavior elite and choice voter iia conaiiymcaimudriste‘epnil at oenet oe,which model, Government’ Party ‘Responsible the underpins mechanism accountability similar A eiso eetppr,rvee nSect. in reviewed papers, recent of series A 10.1007/978-3-642-35239-3_9 [email protected] B ) 1 ncnrs,agoigbd frsac ncmaaiepolitics comparative in research of body growing a contrast, In dacsi oiia Economy Political in Advances pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , .)for t ...) ,adse isbtenpltclprisadvtr as voters and parties political between ties sees and 1967), programmatic nil aeilo rfsinlrewards. professional or material angible lcin nwihteln between link the which in elections , clientelistic 2 eo,hsaaye clientelism analyzed has below, ikgsbtenciti- between linkages 181 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 cnm n omlpltclisiuin ncniinn h i flnaestrate- linkage of mix the political- conditioning com- controlled in publicly basic gies. institutions a these political of complement formal role To and the economy highlight politicians linkage. also that of authors development the forms of statics, both parative levels in intermediate equitably at more competitive- is invest as it so Finally, increasingly increases. and ness programmatic economic heavily of is levels high politics At development, increases. competitiveness as so increasingly and telistic, n hr samre fucmitdvtr ufiinl ag otpteblnei ao foeor one of favor in balance the tip to large sufficiently 28) voters (p. uncommitted bloc.” of another market a is “...electionsareclosebetweenrivalblocsofparties... there which and in those as elections competitive fine hr aeasge h ocp ifrn miia eeet.KtcetadWlisn( Wilkinson and Kitschelt referents. empirical different concept the assigned have thors a aelas ttms oteaoto fpstv qiiru eeso clientelism. of levels equilibrium positive of adoption ideologi- the one’s in to of interest times, support one’s at the balance leads, maintaining base to in cal that need with the median voters, electoral of the face subsets courting parties distinct political target to that abilities is tial condition second A voters. abilities with a of differential Equilibria have subsets Nash parties distinct that of target is emergence to condition the One for clientelism. of conditions levels also necessary positive It processes. of formation either set coalition yields a in instability form discusses on general literature past most to Section its result instability. general in theoretical effort or game convergence campaign voter the their to median of Thus, candidates traditional % platform. all 100 devote this for to promoting be and to will point, ideal Equilibrium voter’s Nash median game’s the resources choose the clientelistic then to low, responsiveness voter sufficiently if is contrary, the On instability. plurality. by electoral ized an always win and can recipients candidates of opponent, set ‘narrower’ their slightly by a proposal clientelistic propose positive clientelistic with some Equilibria given Nash effort: exist never will there behavior, candidate rvn hsmxi h neato ewe cnmcdvlpetadelectoral and development economic between interaction the functions. competitiveness. is production mix vote politicians’ this in Driving appeals edited programmatic their and recent to clientelistic (2007 outline Introduction of Wilkinson the briefly In and I clientelism. Kitschelt here of volume, contributions, causes specific the on paper’s research this theoretical highlight to as So Clientelism of Theories 2 comprise which actions Section the Section benefits. and sets. clientelistic functions, choice receive utility to their their targeted position, actors, are programmatic model’s who their the voters of presents of promotion set the the to (3) and opposed as clientelism to devote 182 2 opttvns santbytik ocp opeieydfieadoeainlz.Dfeetau- Different operationalize. and define precisely to concept tricky notably a is Competitiveness hntroti o ie n ate aedifferen- have parties and given a not is turnout When constraint. turnout binding hsi o osyta h aei t otgnrlfr sawy character- always is form general most its in game the that say to not is This 2 tlwlvl feooi eeomn oiisi evl clien- heavily is politics development economic of levels low At 4 hndmntae ht betsrne etitoson restrictions stronger absent that, demonstrates then rsn nagmn oepantemix the explain to argument an present ) 5 eae this relates .Kselman D. 2007 )de- 3 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 srn’pryielgclspotr rssa nedgnu ucm fstrategic of outcome endogenous and an ‘moderate’ as of competition. arises identity supporters the that ideological the such party below explicit, derived ‘strong’ is model stances In identify programmatic supporters. positions party of fixed ‘strong’ choice these and ‘moderate’ from of and subsets fixed, the exogenously as positions programmatic spatial its ties’ (2005 of Stokes extremism by relative models the Indeed, relationship and stances. the strategies address linkage party’s not a does between it Furthermore, strategies?” campaign grammatic drs h usinakdb ishl n ikno ( Wilkinson and not Kitschelt does ( by it However, asked efforts. question clientelistic sub- their the or devote voters address parties’ particular which the to i.e. voters constituencies, of clientelistic sets parties’ of nature the on and incumbent ‘strong’ separating distance among ideological inducing the so effectively candidates. of do clientelism challenger function a to of longer likely likelihood no is the more turnout well, are politicians As turnout supporters, supporters. but increase ‘moderate’ choose ideological to targeting to clientelism than who not use Rather is incum- turnout. who decision an to basic between not whose played or voter is whether potential game a the and distinction: ( politician major Nichter bent one shrinks. with competitor model against her similar or and a for distance party ideological biased incumbent the heavily the when (c) not between and i.e. stances; policy incumbent, programmatic incumbent’s the the of supporters ‘moderate’ are oetreigaemr ieyt rs hn a h benefit the (a) when: of arise relationships to clientelistic likely more equilibrium, are In targeting candidate. vote challenger a or be- ‘ decides incumbent benefit incumbent the a where providing supporter, potential tween a and politician incumbent decisions. an par- competitors’ that their such of context function strategic explicit goods. a an in clientelistic are decision strategies with linkage electoral can- target the ties’ which to embed by not electorate do processes the they of the Finally, segments nor which stances; moder- choose programmatic or didates parties’ extremism relative political the of and ation investigate appeals of however, clientelistic mix not, between do optimal relationship They politicians’ the appeals. conditioning electoral programmatic in and profile clientelistic risk incumbents’ and aversion. of tiveness, risk uncertainty politicians’ relative (c) and the provision; (b) good proxy); public to reasonable returns a vote eco- in: be which increases (for should goods thus development policy effort targeted nomic sup- small-scale clientelistic for voter of preferences of relative allocation the voters’ blocs On optimal (a) smaller uncertain. The gain certainty. is to but proposals with politicians voters, good port allow of public goods number to clientelistic large response hand, a voters’ other target as to insofar clien- ability to risky the opposed are offer as goods public Public generate goods. to telistic decision politician’s incumbent an consider o-xsec hoe o letls nSailModels Spatial in Clientelism for Theorem Non-existence A 2007 hsfis e fgm hoei aeshsmd aubecnrbtost research to contributions valuable made has papers theoretic game of set first This tks( Stokes competi- electoral development, economic of role the emphasize papers These (2007 al. et Magaloni volume same the In ,nml Wa spltcas pia i ewe letlsi n pro- and clientelistic between mix optimal politicians’ is “What namely ), 2005 nlzsa nntl-eetdpioe’ iem lydbetween played dilemma prisoner’s infinitely-repeated an analyzes ) B n h oeta upre eie ovt o the for vote to decides supporter potential the and ’ n ihe ( Nichter and ) eeo eiintertcmdlto model decision-theoretic a develop ) tplt oiia par- political stipulate 2008) 2007 B n aaoie al. et Magaloni and ) slre b voters (b) large; is 2008 analyzes ) 183 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 iltn ooe h ort etseigb ulcofficials. public by rent-seeking to they that door and the governance; open economic to of tend matters national- will on where systems credibility in prevalent lack more Fi- politicians be level increases. will effort appeals such clientelistic clientelistic that as to argue segment ‘inclusive’ they effort nally, overall ‘broader’ more effec- more becomes a devote more targeting parties that i.e. are as find appeals, targeted voters also be which will They electorate in appeals. the policy districts of targeted those by i.e. swayed bias, tively ideological of levels low uhlclintermediaries. local such rgamtccocseeg iutnosyi equilibrium. the in In parties’ simultaneously biased. emerge and ideologically ‘inclusiveness’ choices relative programmatic ‘less’ coalitions’ or stip- clientelistic ‘more’ exogenous below are developed not an model which do implies districts model turn their electoral in in of which parties ulation research, positions, reviewed programmatic above explicit the choose with as said, That tions. n Tabellini and agtdgoswl ealctd n d etetato ees o nieStokes which unlike to Not districts levels. of extraction rent set (d) the and good (c) allocated; public provision; ( be of good will level targeted goods a of targeted (a) level choose: a model (b) their provision; in Politicians environments. ibility sa oiy u ahrntoa fcas ako rdblt.Ide,i ol ihu uhcred- such actually without world patrons a local in of Indeed, credibility. presence of the lack ibility officials’ national rather but policy, fiscal ag ltomte n otatatv.I h pta oe,cmag platforms campaign continuum model, policy label dimensional spatial will cam- the I whose In what attractive. of candidate consist most the find choose versa. they simply vice platform they and paign non-strategic: strategy are campaign 2’s voters candidate contrast on In contingent is action optimal 1’s that rgt oiyoto.Cniae’ato-e nsailmdl ossso platform a of consists models spatial choice in action-set Candidates’ option. policy policy the ‘right’ and candidates to available policy candidates Label voters. and candidates actors: of marker types the with two contains game The Functions Utility and Actors 3 184 3 n agtdgost niiul n ml oilgop.Mr atclryassume particularly More groups. social provid- small and and policy, individuals public to national-level goods targeted of ing issues on proposals their implementing divide must candidates both proposals. that policy candidates’ of evaluation their on based choose then voters forms, ( 2005 2008 oee hyas oeta ti o arncin istesle htgnrt esta ideal than less generate that themselves ties patron-client not is it that note also they However efradVac ( Vlaicu and Keefer oebdcineitclnaesrtge ntetaiinlsailmdl assume model, spatial traditional the in strategies linkage clientelistic embed To P ,teatosfidta letls ilb agtdt lcoa itit with districts electoral to targeted be will clientelism that find authors the ), oecoett drsigtesto usin ake ntepoedn sec- proceeding the in tackled questions of set the addressing to closest come ) ∈{ x P 1, oehr ntecontinuum the in somewhere 2} 2000 addts eiinpoessaeitreedn,ie candidate i.e. interdependent, are processes decision Candidates’ . P otecmaaiesuyo sa oiyudratraiecred- alternative under policy fiscal of study comparative the to ) n suetruhu htol w addtscmee such compete, candidates two only that throughout assume and 2008 programmatic dp atclrpltcleooi oe (Persson model political-economic particular a adapt ) x ∈[ xedbepltcleffort political expendable 0, 1] improves uhta h policy the that such x oiypooas osdrasml uni- simple a Consider proposals. policy ∈[ x 0, oe efr scmae ooewithout one to compared as welfare voter = 1] stepltclsetu’ most spectrum’s political the is 1 aigcoe apinplat- campaign chosen Having . ewe rmtn and promoting between x = 3 efradVlaicu and Keefer stems ‘left’ most the is 0 .Kselman D. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 x htbt the both that h agtset target the supin1 Assumption vote-seeking: clientelistic of nature the to as assumptions ntedsrbto fvtrpeeecs o xml,i rcue taeyi which in strategy a precludes it example, For P preferences. voter of distribution the in l oeso h oiia lf’ ..woems-rfre oiyis policy most-preferred whose i.e. ‘left’, political target the might on candidate a voters example, all For inducements. clientelistic with target will they rfre oiyo h continuum their the on by policy foremost preferred and first defined are Voters electorate. range lge ntergtfo h range the from right the on ologues eitcbnfis(labeled benefits telistic r itiue nfrl nteplc space points policy ideal the that in throughout uniformly assume distributed simplicity, are For (‘right’). ‘left’ political the on icy ustmy tlati h btat ag nweefo h nieeetrt l the all electorate entire voter. the single from a anywhere to range down abstract, way the in least at may, subset edt agtmr hnasnl iie u esta h niectznywt letlsi induce- clientelistic with citizenry entire the than less but citizen single a ments. than more target to tend oiia etri h range the in center political e e hc otisbt lf’ad‘ih’vtr,s oga hs oescome voters these as long so voters, ‘right’ and ‘left’ both contains which set get iecmag taeyi T hmsalItre ycineitcefr? nother clientelism on In expended effort?” be to clientelistic effort my of level target overall I the choosing shall beyond whom words, “To is strategy campaign sive will voters versa. strategies vice and campaign proposals, clientelistic below, policy in national-level see engage will their we discount candidates As that vote-seeking. extent of modes the distinct to two to devote will they which a have candidates both that Models Spatial in Clientelism for Theorem Non-existence A 4 (labeled stances programmatic their promoting between uttu hoentol pta position spatial a only not choose thus must addtsms locos h usto oeswowl eetfrom benefit will who voters of subset the choose also must candidates h ot‘ets’qatl fvtr nterange the in voters of quartile ‘leftist’ most the etlaig(ih-enn)vtrtree ycandidate by targeted voter (right-leaning) left-leaning supin2 Assumption nterange the in ht ogl paig voter a speaking, roughly that, hs xrms oee,aeulkl ob bevdi h miia ol,weepoliticians where world, empirical the in observed be to unlikely are however, extremes, These m targets h rtasmto rhbt addtsfo hoigatre e ih‘breaks’ with set target a choosing from candidates prohibits assumption first The lcoa addtsms hoefo hsdsrbto fvtr hs which those voters of distribution this from choose must candidates Electoral nadtoa usinwihcniae utase ndvsn comprehen- a devising in answer must candidates which question additional An = .5. x i ∈[ both 1 / x Θ 4 mean i , doouso h ih nterange the in right the on ideologues P ∈[ 1 h agtset target The letlsi effort Clientelistic / . 2 ] 0, nteohrhn,i osntprevent not does it hand, other the On . and 1 / 4 ] C median iial tpeldsasrtg nwhich in strategy a precludes it Similarly . P 4 .Ti mle h fotconstraint effort the implies This ). igeunit single x omk hsmr ocee osdrormdlo the of model our consider concrete, more this make To i Θ i ∈[ ihielpoint ideal with P ftevtrpeeec itiuinaelctdat located are distribution preference voter the of utbe must 1 x C / 4 ∈[ x P , i 3 / is ∈[ 0, fcmag fotwihte utdivide must they which effort campaign of 4 ] vnydistributed evenly n oo.Define on. so and ; 1] 3 continuous / x x 4 Define . P , ∈[ u loteefr levels effort the also but , 1] x x 0, i n mdrts ntelf nthe in left the on ‘moderates’ and i <. x ∈[ 1] i x ∈[ 5 (i.e. i 0, in (x svoter as P 3 ..termost- their i.e. point, ideal 1 x / / G i emk h following the make We . 4 x 4 >. ∈[ , ] P i P ralvtr rmthe from voters all or ; 1] mn l ebr of members all among ∼ n rvdn clien- providing and ) 0, 5) x rmcosn tar- a choosing from G n hs nteleft the on those and uniform P i 1] otpeesapol- a prefers most P sielpitsuch point ideal ’s ( . x + x P i P C ) <. P [ G stemost the as agt ide- targets 0, = ;o only or 5; P C 1] .They 1. P and ,such ), .This C C 185 P P , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 e e nldsal‘oeae’i h range the in ‘moderates’ all includes set get iswihmy tlati hoy eosre miial.I uueieain ileaiethe examine will I iterations future In empirically. assumptions. observed both be relaxing theory, of in consequences least at may, which gies hoeaplatform a choose ersnstevlecniae taht inn office. winning to attach candidates value the represents iiyo inn h lcin n ileeg noeosya ucino both of function a as construction endogenously (by emerge strategies will campaign and candidates’ election, the winning of bility inl rproa nueet.Bgnwt h aua supinta odn all holding point profes- that ideal assumption with material, voter natural targeted a the constant with and else Begin issues inducements. personal policy or programmatic sional, both over preferences as ih epiaiyofie n/rvt-ekn hl tesmgtb rmrl policy-seeking. primarily be might others while vote-seeking and/or office- primarily be might aeaotteplce hc r mlmne sarsl fdmcai lcin eg Wittman (e.g. elections democratic of also result candidates a which as in implemented situations are to which 1983 model policies spatial the traditional about the care extended has research recent That set. target their par- in which spectrum in political strategy the electoral of sides an both of Assumptions from possibility said, extremists the include eliminate to to attempt natural ties quite seems it plausible: h term The addt’ agtstaeasmdt eev neulaon ftebnfisre- benefits the of amount equal an receive within contained to themselves from assumed find sulting are who set types voter target All candidate’s others. a of to members than certain set to target goods their clientelistic more providing from candidates precludes 5 if candi- even example, that For extent strategies. the that linkage assume to clientelistic also in stances will engage We policy dates point. programmatic ideal candidates’ her from discount removed voters further policy a than rather distribution preference the of range continuous a from 186 6 tne codnl.T prtoaieti oin osdrtefloigspecifica- voter’s following a the these of consider tion discount notion, will this operationalize voters To stances, accordingly. stances policy programmatic their implementing and/or h npit ftetre e owihteremaining the which to set target the of endpoints the ..terol oli eiigcmagsi oiia nubny implying incumbency, function: political utility is campaigns following devising exclusively the in are goal candidates only that their assumption i.e. Downsian seeking, the the On candidates. adopt electoral I substan- to hand, available model strategies other this campaign of approach, set the spatial expands traditional tially the to Compared clientelistically. denoted set, o and low policy h ueyofieseigasmto stesmls falcniaepeeec oes More models. preference candidate all of simplest the is assumption office-seeking purely The are both well, As immensely. model the simplify and technical, primarily are assumptions Both x Let uta addtsmyepo ohfrso lcoa ikg,vtr o have too voters linkage, electoral of forms both employ may candidates as Just P ;Calvert1985 oe ute from further moves x v P P C abs = ={ P Θ x svr ih u ipy fcniae xr iteefr npromoting in effort little exert candidates if simply, Put high. very is C [·] x i P voter , P P eoe h bouevlefnto uhta,holding that, such function value absolute the denotes 1 rgamtcutility programmatic ,G . ∈[ .Srm(1990 Strom ). 5 and x en h e fvtr agtdby targeted voters of set the Define x P P P ,x i ,alevelofeffort 2 u , ilatiueltl rn au oti oiywhen policy this to value no or little attribute will i,P oeiiaefo addts cinst eiso apinstrate- campaign of series a sets action candidates’ from eliminate do x P P , ( x ] x prog) . P i voter } ersnsa al tep oepanwysm candidates some why explain to attempt early an represents ) U ersn a represent P = = i G spormai tlt for utility programmatic ’s x π for P i P G ol rfrthat prefer would · · P P 1 β strategy : eoe opooigti ltom and platform, this promoting to devoted h marker The . − x i abs ∈[ [ π 1 x o candidate for / 1 4 P , = C 3 − / P 1 4 x 6 P x ] P = − π .Tescn assumption second The ). ∈[ i steti party’s this the as ] P hoeapolicy a choose 1 π . 0, − represents 2 .Temarker The ). 1] G P P P a hntetar- the when (as addtsthus Candidates . erae.Simi- decreases. ilb targeted be will P G hoe the chooses P P G .Kselman D. constant, sproba- ’s P x P svery is office- β> target = (1) x 0 i Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 agtmr n oevtr h fotlevel effort the voters more and more target eev usata tlt rmcineitcbenefits, clientelistic from utility substantial receive h yeo rgamtcuiiyo neett hspaper. this to interest of utility programmatic of type the hmevswti addt’ agtst osdrtefloigfntoa form: find functional who following voters the about Consider What set? campaign. target 1’s candidate’s a from within ‘0’ themselves of utility clientelistic a receive n in ing e.Since set. le httepoiino agtdgosbcmsls fceti xrml large extremely in efficient less becomes goods targeted amounts. of provision the that plies (weakly effort clientelistic additional of unit a for utility marginal 7 holding larly, Models Spatial in Clientelism for Theorem Non-existence A letlsi tlt f‘’fo htcniaesplce.S,i addt chooses 1 candidate if a receive So, simply policies. candidate’s candidate set that particular target the from a ‘0’ by of targeted utility not clientelistic are who voters that assume itnus ewe oeswoaein are who voters between distinguish uet be to sume ( with Beginning is oes rgamtcuiiyfor utility programmatic voters’ cies, count parameter exogenous The the beneficiaries. reducing thus population, larger and e ilrcieltl tlt rmcineitcbenefits, When clientelistic targeted. from narrowly utility and extensive little receive will set utility at,icuigbtntlmtdt n’ noe rfsin n utrlenviron- cultural and profession, income, one’s to limited not but including nants, targeted. broadly and n oe ilhv o candidate for have will voter any onson points fpbi odpoiinotnasm htvtr r ikaes;i ewr oasm hthigher that assume to were we if averse; risk models are Formal voters environmental pollution. that of reduced to assume levels from often security benefit provision the national constant, good of prefer public held nature of else who the Voters all when nonetheless, most-preferred. even will their provision ser- protection good not social public to is increased access provided from free good benefit others voters and said, protection, That environmental vices. some security, national prefer may for ee oiybnfisi eraigwt h xeto agtn per frequently appears Vlaicu tar- targeting and for of Keefer extent utility (e.g. the marginal models with political-economic citizens’ decreasing in that is notion benefits The policy geted increases. effort clientelistic as h ucinlfr n( in form functional The ’ eoiao,w aearaydefined already have we denominator, (2)’s to Moving nepesn voter expressing In oes eatct’t letlsi pel a aypsil miia determi- empirical possible many has appeals clientelistic to ‘elasticity’ Voters’ P . G letlsi pel.We h icutrate discount the When appeals. clientelistic 7 u P G i,P sarsl fti ucinlfr,temxmmpormai tlt that utility programmatic maximum the form, functional this of result a As x ntecretmdl h dimension the model, current the In . P P Θ ( eueteucranysronigpris blt oipeetntoa-ee poli- national-level implement to ability parties’ surrounding uncertainty the reduce client η P ersn itntpeeecsa oteielntr fpbi od.Sm voters Some goods. public of nature ideal the to as preferences distinct represent Θ ≤ per ntednmntr holding denominator, the in appears x 1 P .Wievoter While 1. =[ ’ ueao,teparameter the numerator, 2)’s ) osat as constant, ilawy eraewt h ieof size the with decrease always will ∀[ 1 / mle htvtr’pormai tlt for utility programmatic voters’ that implies 1) i 4 , i : ’s 1 x / i 2 letlsi utility clientelistic ] ∈ hnalvtr ihielpoints ideal with voters all then , Θ G P i P P ] suiiywl lasices with increase always will utility ’s ,u P erae ode voter does so decreases ol gi nraewith increase again would s‘’ hsocr when occurs this ‘1’; is P i,P stre e n hs h r o.W will We not. are who those and set target ’s δ e capita per δ ( client ssal ebr of members small, is x ersnstert twihvoters which at rate the represents P C o candidate for sapbi odcnium ifrn ideal differing continuum; good public a is P η ) utb itiue mn larger a among distributed be must = sa xoetwihw ilas- will we which exponent an is C δ  vnif even P letlsi osmto fall of consumption clientelistic slre ebr of members large, is δ osatvoter constant P 2008 + C Θ vnif even stre e.A candidates As set. target ’s P η Θ G P P i h effort the P P G x spormai utility programmatic ’s .Oeainly tim- it Operationally, ). ti motn ofirst to important is it , scandidate as sarsl,(2 result, a As . i  P P < . ilawy eincreas- be always will = P 1 hs eet are benefits these / stre e may set target ’s and 1 4 or i C C ...) sclientelistic ’s P P x i rher or his , i sminimal is x P P > decreases i captures ) starget ’s = starget ’s 1 / 2 x P will dis- 187 (2) . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 sasmn ihvle mn oeaevtr hnbt ate hoeeteitplce in policies extremist envision choose might parties one both example, when For voters choices. moderate voter among x values and high strategies assuming campaign as candidate of set the on ilrnoiei nubae a ie hoeec addt ihaprobability a with candidate each choose (i.e. way unbiased of an in randomize will oeseaut hs cin n hoetecniaewoeplce maximize policies whose candidate the For choose utility. and their actions these evaluate voters i n rgamtctreig oseti oethat note this see To targeting. programmatic and tic xii oedge fdfeetainacrigt hi oi-cnmcadclua status. cultural and socio-economic their to according differentiation of degree some exhibit {x fvtr h choose who voters of et safis u,i hspprw ilasm that assume will we paper this in cut, first a As ment. 188 9 8 appeals. clientelistic to responsive similarly are electorate an in voters all that lrlt rule. plurality euneisl snt nafis tg ohcniae hoeasto actions of set a choose candidates both stage first a In game not. the model, is spatial itself traditional the sequence in found that than complex more substantially candidate’s a pro- from to away versa. devoted taken vice effort definition, and by effort, of clientelistic thus, increment is additional formation effort linkage Every the grammatic (i.e. binding). vote-seeking be on spent will be constraint from will only effort clien- utility all receive incumbency, to politicians political which allocated gaining in be rent-seeking will without model appeals a campaign In telism. programmatic on expended not effort td fdmcai conaiiywl eur aeu ramn of treatment careful the a for require framework will empirical accountability an democratic into of here study developed framework theoretical the ing u safis u eassume we cut first a as odn o( to cording ucpil omr cncl lcoa appeals. electoral ‘cynical’ more to susceptible a hsehutvl xrs voter a express exhaustively thus elec- can national between their differentiate on which based arguments torates static comparative of variation. exogenous set and subnational first invariant for an potential of its assumption as the well Nonetheless, as determinants, exogenous and nous h oe a vnulyb xeddt iutosi which in situations likely to will extended electorate be given eventually a may within model voters The case: the be to unlikely is this empirically course, Of i i, P ∈[ ∼P 1 Voter hl u praht oeigcmag taeisadvtrpeeecsis preferences voter and strategies campaign modeling to approach our While P / ,G 2 0, i hc case which (in .Bitit hsmdlo oe rfrne sataef ewe clientelis- between tradeoff a is preferences voter of model this into Built ). ( v 1] P ∼P ,x h leainwihaie rmpltcleteimmymk oeae particularly moderates make may extremism political from arises which alienation the : u i i,P ) P ilcos h addt hs oiisyedtehgetuiiyac- utility highest the yield policies whose candidate the choose will rfrwhom for or , .I addtsaotsrtge htyield that strategies adopt candidates If 3). , ( x v P P } ) P, = oa omxmz hi utility their maximize to as so α ∼P  P einvtrsrsosvns ocineitccampaigns.We clientelistic to responsiveness voter’s median P G G sa epce’vt hr) h lcini odce under conducted is election The share). vote ‘expected’ an is ..tepooto fvtr o hmeither whom for voters of proportion the i.e. , δ P P ∈{ ob xgnu otegm itself. game the to exogenous be to u · · i,P 1, ( ( 1 1 2} ( − − v ,given P abs abs ) = [ [ i x x suiiyfrparty for utility ’s u i i i, − − v ∼P P x x P P ( and v ] ] ∼P ) ) + v ) if ∼P  U u hs admcoc lands choice random whose but δ x P δ +Θ C i G define P sivratars oes i.e. voters, across invariant is = η δ ∈ / P P sedgnul eemndby determined endogenously is P i Θ π  = dnia aof,then payoffs, identical P P sfollows: as δ . ( · if losu oietf a identify to us allows 9 α 1 β P liaey translat- Ultimately, x − nascn stage second a In . i steproportion the as ∈ C P Θ ) P n n all and any : δ , u sendoge- ’s .Kselman D. i,P ( 8 v v P Also P )> (3) = δ i Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 osbt addtsago elo reo nchoosing in freedom of deal good a candidates both lows nraehrpoaiiyo winning. of probability her increase usto oes lhuhcntandb Assumptions by continuous any constrained target Although can clien- voters. candidates of to which subset in devoted situation effort a with any Begin without appeals. telistic position policy most-preferred voter’s (Downs model median original Downs’ in made choice equilibrium oe programmatic voter ohpriscos dnia ltom n rgamtcefr ees Trivially, levels. effort programmatic and implies platforms this identical choose parties both n agtset target and set aehsals than less a has date yLemma By saptniltre set, probability target potential a as 1 Lemma {x Define Instability Clientelistic 4 Models Spatial in Clientelism for Theorem Non-existence A agtstthan set target with satre set, target a as 1 Theorem environments. unconstrained these in telism sets, target choosing in unrestricted are candidates decided as is long election as Lemma and such, parties As two flip. coin the a between by indifferent are voters all case which ro fTheorem of Proof party aehsagetrthan greater a has date addt ilawy aea pia eito:te a mrv hi hne of chances their improve can they to deviation: optimal winning an have always will candidate ilidc eeto ywihvrpryi eslkl ownteelection. the win to likely less is party whichever by defection induce will ia oiyposition policy tical upr falvtr in voters all of support ol hoea dnia ee fcineitceffort clientelistic of level identical an choose could P o eosrt h mosblt fNs qiirawt oiielvl clien- levels positive with Equilibria Nash of impossibility the demonstrate now I h ro fLemma of proof The htaotastaini which in situation a about What Θ ,G G . ∼P P P v 1 ,x P ∗ < =[ obtains. P hncniae a hoeaycniuu ag fvtrielpoints ideal voter of range continuous any choose can candidates When sa as 1 1 twhich at 1 1 / / π hncniae a hoeaycniuu ag fvtrielpoints ideal voter of range continuous any choose can candidates When , x eko htaysrtg etrwihmakes which vector strategy any that know we , 2 2 Θ ∼P x P C ( ysml hoigasrtg dnia ota fteropnn,in opponent, their of that to identical strategy a choosing simply by P P ,( i. there ahEquilibrium Nash P } e > =[ x .A el l oesnti ihrtre e ilrnoie since randomize, will set target either in not voters all well, As ). . with P nayNs Equilibrium Nash any in 1 1 1 taey h atri taeywihesnilyrpiae the replicates essentially which strategy a is latter The strategy. / / x x − 2 never 2 osdrastaini which in situation a Consider ∼P P 1 u tews,anytime otherwise, Put . π Θ rbblt fwnig nsc ae h oe probability lower the case, a such In winning. of probability / G , ε) 1 P 2 ∼P x P = rbblt fwnig mligta h poigcandi- opposing the that implying winning, of probability nayNs qiiru ahcniaems i with win must candidate each Equilibrium Nash any in ssrih-owr.Cnie aei hc oecandi- some which in case a Consider straight-forward. is ] P = where < ] xssaNs qiiru nwhich in Equilibrium Nash a exists x (since n twhich at and , 1 P / sc that (such 1 2 n eryietclbtsihl arwrtarget narrower slightly but identical nearly a and , , ∼P ε P C taeyand strategy → ∼P chooses a choose can .I odoing, so In 0. ilb itiue vrasihl narrower slightly a over distributed be will π C π P P v 1 = > P = P v 1 ={ v )adtre set target and 0) m / π chooses ∼P 2 2 C nti case this In ? ={ x = ∼P P P ={ ,G 1 x 1 P / m = hoe strategy a chooses Θ 2 and x sopnn ilwnthe win will opponent ’s ). , P P v P 1957 1, C ,x P ,G . P 2 ∅, ={ P π = rmaoe hsal- this above, from P ,ie ocos the choose to i.e. ), , P ∅} C P ,x x 1 x P <. Θ P sopponent ’s P − P sthe as } P > ,G , 5or G with x =[ 0 P P P o either for niden- an , ,x − x median- π G P P P ε P , v } , x >. P x ∼P < and 189 P P = ] 1 5 } . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 hieof choice set. of target range potential continuous a any as choose points can ideal parties voter when clientelism of levels positive with agtstchoose set target ae nterpormai tlt o h epciepris If parties. respective the for utility programmatic v their on based any ic nyoevtri otie in contained is voter one only since tainty: x h eitn addt’ rbblt fwnig h rcs sdslydi Fig. in displayed is in process increase The an winning. to of leads deviations probability candidate’s these deviating of each the sets; target narrower slightly but ping h netv odvaeadwntermiigvtr’spoto programmatic on support voters’ remaining the win and grounds. deviate to incentive the u letls agtn costs targeting clientelist out instability 1 Fig. 190 P ..tetaiinlmda-oe ovrec ihu letls.A neapeI which example under an conditions As the clientelism. derive without now convergence median-voter traditional the i.e. hn5 rbblt fwnig ydfiiinaysc eito ol require would deviation such any definition greater a By candidate candidate winning. deviating deviating of the the probability yields it % if 50 optimal than be only will vector strategy this ic h hoe nietclpafr u eoe oeefr opooigand promoting to effort more devotes but (since platform platform that identical implementing an chooses she since vector v Θ oNs qiiru;js hti a ocineitcNs qiiru.Frsuffi- For Equilibrium. Nash clientelistic no has of levels it high that ciently just Equilibrium; Nash no ∼P ∼P i P sopponent ’s scnandi addts agtst.A hspit ihrcniaewl have will candidate either point, this At sets. target candidates’ in contained is nwrs hnbt addtscntre n otnossbe fvtr,any voters, of subset continuous any target can candidates both when words, In Equilibrium Nash no is there that demonstrate arguments these together, Taken htaotastrategy a about What uhjceigfree mle agtst a otneutlol h voter the only until continue may sets target smaller ever for jockeying Such Theorem = ={ = Clientelistic v x v 1 i x m P = C ie agtwt nyoevtrtp)a which at type) voter one only with target a (i.e. n eito hc novscosn ifrn oiypsto with- position policy different a choosing involves which deviation any , , P v 1, 1 2 > ∼ osntncsaiyipyta h aei t otgnrlfr has form general most its in game the that imply necessarily not does ∅, = nue tigo eitosi hc addtscos overlap- choose candidates which in deviations of string a induces 0 P ∼ ∅} v m P ol choose could hnalvtr ilhv ihrpormai tlt for utility programmatic higher a have will voters all then , ohcniae i ihpoaiiy5 ,s eito from deviation a so %, 50 probability with win candidates both . δ h aesuiu ahEulbimwl be will Equilibrium Nash unique game’s the P ochoose to v P P ={ h lcin(Downs election the v ∼P G x P P G ={ ,G P = v < P ) ssc,albttesnl oe in voter single the but all such, As 1). x 1 ∗ P ,x = :a oga e opponent her as long as 1: , Θ P 1, v 2 ∗ P , ∅, x l eann oeswl choose will voters remaining all , = P ∅} } v m 1957 with n i h lcinwt cer- with election the win and when ). G π P η P = < = .A h strategy the At 1. n agtset target and 1 1 / 2 v nti case this In . 1 ∗ ∼P ∼P = .Kselman D. v chooses chooses 2 ∗ = ∼P P v m 1. ’s , , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 h ads ownoe,bcuesereceives she because over, win to hardest the ai fcineituiiyaoe making alone, utility clientelist of basis ee of level htadvainto deviation a that i vr eas h receives she because over, win .6/.5 em 2 Lemma ε ltom n targeting and platform, π from o-xsec hoe o letls nSailModels Spatial in Clientelism for Theorem Non-existence A 1.2 x nuh(i.e. enough hc ol ersn the represent would which ilyield will and x nldsaybr lrlt fvoters of plurality bare any includes nwhich in exist ceive v locating eaotd ned h addt nqeto ilb nifrn ewe n set any between indifferent be will question in candidate the indeed, adopted; be Lemma certainty. note ee of level ayn eesr odto aaeesare parameters condition necessary panying iinlvtr ntetre set target the in voters ditional candidate deviating the for vote o n ee of level any for il h eitn candidate deviating the yield egy P ihcrany l oesi h aemjrt agtstwudreceive would set target majority bare the in voters allow all would certainty: strategies with condition necessary the adopting case gis nopnn at opponent an against Appendix n i h lcinwt etit.B hoigteplatform the choosing By certainty. with election the win and ˆ ˆ P ∼P P P → rbblt fwinning of probability a (G o h otflxbedvainfrom deviation flexible most the So, hs when Thus, oee,if However, oietf hte rntadvainfrom deviation a not or whether identify To (G > > x v ˆ Θ = 0). e facmayn hie hc yields which choices accompanying of set a = P P ˆ v P P u 0% eesr odto eitosaedfie eea olw:frany for follows: as here defined are deviations condition Necessary %. 50 ,temda oe n l oesi h agtstwudchoose would set target the in voters all and voter median the 1, 1 (G ) v ) =[ˆ i,P G ={ .O l oesi hstre e,temda oe ilb h ads to hardest the be will voter median the set, target this in voters all Of 1.2. G and , = m = C δ salse ht o n eito from deviation any for that, establishes ) P P P P .Temda oe receives voter median The ). ( x P x v client When .4,. = ) < P < δ m 2 au of value a = o l ausof values all for (G ,adi which in and 0, ssaleog) hr a emn eitosfrom deviations many be may there enough), small is = x ,i eitn otechoices the to deviating if 1, ˆ and 1 6, G δ δ P otetre set target the to .6 P hr osntexist not does there ) v (G = = P 0,. η 2 m ), = v any < = P os’ elu,i qiiru,wiho hs eitoswould deviations these of which equilibrium, in us, tell doesn’t then 0 ,frayvleof value any for 0, P x hc lo h eitn addt ownteeeto with election the win to candidate deviating the allow which 6} ˆ .O l oesi hstre e,temda oe ilbe will voter median the set, target this in voters all Of 1. P ) ={ ,teacmayn oiypafr n agtstdeviations set target and platform policy accompanying the 1, 1, aepuaiytre set target plurality bare oeaoae oeta sln svtr au clientelism value voters as long as that note elaborate, To . n ssc eevsttluiiy1 utility total receives such as and , v π (G ∼P C P o n eito from deviation any for .4,. P > oaybr lrlt fvtr.Lemma voters. of plurality bare any to P π = ) eesr condition necessary P ] 6, 0% dp h olwn rcdr:Ifis identify, first I procedure: following the adopt I %, 50 G P h olwn em establishes lemma following The . ol lodvaet h strategy the to deviate also could v > P Θ u m 0,. P rbblt fwinning of probability a P i, . p 0% hnfrta ee of level that for then %, 50 < ∼P hoe eito to deviation a chooses iia oprsndmntae htalad- all that demonstrates comparison similar A . 6} =[ 1: e fcocswihyields which choices of set a ( Θ oallows to prog) ( 0,. p any π G u v =[ P 6] i,P m P x ˆ Θ = P = ilas prefer also will ( culyivle maintaining involves actually 0,. hr ilb a be will there (G prog) osnot does 1. from 1 x uhthat such ˆ v P P 6] v P u m (G m i, l oesi h agtstre- set target the in voters all , ), ownteeeto ihcertainty with election the win to eitos eoe as denoted deviations, osome to ∼P v = π oavalue a to x P ˆ m P P ) ∼ ( fteacmayn choice accompanying the if , .4 il h eitn candidate deviating the yield prog) (G > = P × x G π 0% osdrtecase the Consider %. 50 P x P (since P .9 v m ) P + P − ag e fdeviations of set large and , P G = > P = n agtstthat set target a and = snwsrtg,such strategy, new ’s G P ={ .36 x from 1 0% hnfrthat for then %, 50 P x 6fo h strat- the from .36 ownteelection the win to .Cery nthis in Clearly, .4. P v < ˆ v P x P ˆ x x ∼P π > = < 1 P (G P P v 2 P ={ (G u ,G hr osnot does there m ,s h will she so 1, .5 1, = = pofi the in (proof P > i,P hc yield which .4,. P + ) h accom- the ∼P P v n al- and .4 0% De- %. 50 , ( ) m ,x P x x client) ˆ ε ˆ osnot does x P P 4, .Since ). , P m nthe on (G (G where (since 0,. , sa as x P P 191 P 6} = ) ) } , , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 osnteita aofipoigdvainfrta level that for deviation payoff-improving an exist not does xsswihgat h eitn party deviating the grants which exists aehsn ahequilibrium Nash no has game aepuaiytre e osntincrease not does set target plurality bare a qiiru.Wienmrclydfeet h aeqaiaieipiain obtain of implications value qualitative the same of Nash regardless the no has different, game numerically the While and strategies, Equilibrium. linkage competing the between infinitely cycle huh fa h e fvtr ealweetrlcniae otre ihclientelis- with target to candidates electoral allow Assumptions we goods. voters tic of set the as of thought oltosi eedn ntecmimn ehooiswt hc n endows one which allowable with of technologies set this commitment (Humphreys turn actors the strategic in on and dependent one coalitions; is assumptions ‘allowable’ the coalitions of on set crucially, the depend, coalitions regarding stable makes equilib- of stable properties of existence the the and Both ria possibility processes. the coalition uncovered in instability again in- theoretical once for work has recent strategies unique), bargaining not sequential were troducing equilibria Nash often greater (though in generated stability cycling initially theoretical and approaches instability non-cooperative (Nash theoretical While for theory processes. potential coalitional game the uncover cooperative to of tended things form other for- the coalition in of primarily models came non-cooperative in result Humphreys results This (see instability mation sets. general target to distinct related but overlapping is opponent’s gen- adopting their most usurp by continually the can supporters in candidates clientelistic clientelism that fact of the levels from positive arises model with eral Equilibria Nash of absence The Discussion 5 Lemma What of %. value 100 any to for winning that, of is probability tells her increases which deviations of 192 u nooeaohrstre es ni ohpriseetal n pbc tthe at back up end eventually parties both vector until strategy sets, programmatic target median-voter another’s one Theorem into in cut uncovered dynamic strategic the rpsto 1 Proposition result: following uhdvaineit.I eito oseit(..if (i.e. exist does deviation a If exists. deviation such eiso eitos n oon so and deviations, of series v etitoso h e falwbecineitccaiin hc a om However, form. can which coalitions clientelistic Theorem allowable of set the on restrictions 1 ∗ = The ihrgrst h oe bv,tento f‘loal’caiin a be can coalitions ‘allowable’ of notion the above, model the to regards With v 2 ∗ Appendix = 1 v bv eosrtsta,wtotadtoa etitos osto clien- of set no restrictions, additional without that, demonstrates above m n tlwrlvl h aewl aen ahEquilibrium. Nash no have will game the levels lower at and , When otistepof o n au of value any For proof. the contains η 1 = 2008 η and 1, :athighlevelsof G 2008 if o necletrve) al eerho h subject the on research Early review). excellent an for hc r rmrl ehia,srea preliminary as serve technical, primarily are which 2, P . ssc,when such, As infinitum. ad δ < ≥ .377). p. , ,i h eito from deviation the if 1, 1 / 2 then π P v P > v δ m 1 ∗ ywihbt ate continually parties both which by 1, spoaiiyo inn,te there then winning, of probability ’s h aesNs qiiru ilbe will Equilibrium Nash game’s the hsi unst nmto another motion in sets turn in This . 0% o n au of value any For %. 50 = v 2 ∗ = δ< δ< v m 1 v , / 1 m n if and 2 δ< / G 2 tlatoedeviation one least at to )thissetsinmotion P hslast the to leads This . 1953 x 1 ˆ / P 2 δ< (G h w parties two the ,adamong and ), P 1 ) / δ .Kselman D. 2 = ≥ hnthe then x 1 m / 2 and no 2 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 upr ic t effort its since support effort oetal nbt ate’tre esi h einvtr neetnl,i sim- Assumptions to a added in is Interestingly, voter. restriction voter additional median only this the the that which is such in on sets ‘left’, game voters target the target ple on parties’ only voters both target can only in candidate on can one potentially voters other that target the i.e. effectively and spectrum, ‘right’ only which political the results can the of candidate of series each side a that on one assume report I we here when re- research, I emerge future While additional clientelism. for of with extensions levels experimenting these positive now serve with equilibria am for I allow equilibrium. which constraints in stable is coalitions telistic Models Spatial in Clientelism for Theorem Non-existence A ineffort tion Θ noesd ftepltclsetu and spectrum alienate political may the interests of voters voter’s target side only median one can the candidates on If to uncertain. is closely participation whose too voters extremist cater which policies that uhta oeswoeuiiyfrbt addtsi esthan less is candidates both for utility whose voters μ that Define above clientelism. such of the utility, levels positive combining reservation with by equilibria a Nash that, with generate established to sets now target have allowable turnout on the on I restriction with emphasis voters. themselves greater ideological concern and more to greater of having concentrate without to desires, as median-voter’s so the strategy campaign their alter recommences. then cycle The eoe bnig,isfra oesbe fvtr nteielgcletee will extremes ideological when the election on the voters from of abstain subset some as insofar ‘binding’, becomes einvtri hi agtst,a hc on ihrprycndvaet h me- the to deviate can party either point the vector which strategy only at programmatic include voter sets, candidates dian target both their until me- proceeds this in jockeying which voter Such of median part. set a target the is voter shrinking continually dian by support voter’s median the sc that (such rmai ucm ed oa nnt yl fcmeiievt okyn o the for jockeying vote competitive pro- of loyalties. median-voter cycle clientelistic the voter’s infinite from median an deviation to any leads result: outcome instability grammatic an with end again once nrsos oti deviation this to response In oe’ qiiru ee fcineitctreigis of targeting size clientelistic the of low to level high equilibrium from model’s moving of consequences of static values comparative the examine to of begin levels equilibrium support positive the of maintaining adoption the in to that times clientelism. with of at leads median base electoral ideological the one’s courting in interest eit austa when than values mediate h taeyvector strategy the ∼P hoentt oei h lcin When election. the in vote to not choose hsisaiiyaie u otefc htcmeiiepriscncontinually can parties competitive that fact the to due arises instability This o xml,spoefragmn’ aethat sake argument’s for suppose example, For ae npeiiayrslswihepo hs diinlcntans ecan we constraints, additional these employ which results preliminary on Based =[ C P (.3 nturn, In . δ G δ ei ihahptei hc aisagano one-nuto:the counter-intuition: of grain a caries which hypothesis a with Begin . C + nfc,oeallvl fcineimaehge when higher are clientelism of levels overall fact, In . P ∼P ε),x = = ) oiyposition policy a .2), P m and .8 v 1 ] a epn iial,ads nsc htbt ate pursue parties both that such on so and similarly, respond can where , = C ∼P v 2 C δ sdsrbtdoe lgtynroe agtstthan set target narrower slightly a over distributed is = ∼P sue xrml o aus u tews,higher otherwise, Put values. low extremely assumes ε P v v → m sopponent ’s 1 = = hrceie ya fotalcto of allocation effort an by characterized ,aplc position policy a .2, .B on so, doing By 0. v 2 = v x v m ti possible is it constraint, turnout binding P m μ>. n i h lcinwt rbblt 1. probability with election the win and hssrce unu osritimplies constraint turnout stricter This . ∼P = μ>. ,adatre set target a and .7, ,te h edt aac one’s balance to need the then 5, ol hoea dnia alloca- identical an choose could P ∼ h aestrotconstraint turnout game’s the 5 P a notmldvainfrom deviation optimal an has not ilwntemda voter’s median the win will x ∼P oooial eae to related monotonically = ,adatre set target a and .3, δ Θ sue inter- assumes μ P 1 savoter’s a as =[ and G P .5,. 2 = ,we P 193 7] .8 ’s . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 utesr htsm usto oesnticue ntertre e nonetheless set choose target to of their candidates the forces levels in this near included equilibrium significant or In not support. at candidates voters electoral election supporters with of the them centrist subset win provides to some of order that set in ensure small small, must are a sets to target When only median. spectrum; cater electoral political parties the of former sides respective the their in on voters all to effort clientelist target agtst hntoewiheeg when emerge which those than sets target ml,temda oe’ ihrsosvns otreigicesshrpreference her increases targeting announce to candidates responsiveness that high voter’s median the small, elrmvdfo h lcoa ein When positions median. ideological occupy electoral parties it the both point; from which ideal in well-removed voter’s equilibrium median polarized the highly than a is other positions programmatic choose parties be votes. must their extremists gain sets, to target order parties’ in in placated included are centrists only since positions: fcineitcefr.Teitiinbhn hsrsl sa olw:when follows: as is levels overall result higher this to behind lead intuition always The effort. not clientelistic does of goods targeted to susceptibility voter 194 oe ilpee htcniae eptertre esnarrow, sets target their keep candidates that prefer will voter ihcineitgos(rfsin eiin oieooi ttsec swl as voters. well swing as or loyalists etc) target party parties to status distributed groups socioeconomic are interest goods religion, the targeted re- (profession, whether identify expert goods parties: to clientelistic asked clientelist of were with sets countries target all positions. the in about programmatic spondents data parties’ provides on political also data of expend it provides extremism parties Finally, that or it efforts Secondly, moderation of relative competition. intensity the programmatic the estimate vis-à-vis to of clientelism model’s us First above on allow democracies. the will world 88 of ESCPL of the number cross-section all, a contemporary on a data sci- in provides parameters political basic support, University Citizen-Politician Bank Duke World on by with Survey administered entists Expert and an developed (ESCPL), via Linkages collected Information results. which oretical sets target orientation. broad ideological have their parties of clientelism voters of all elec- levels to the cater near high concentrated at sets and target median; narrow have toral parties effort clientelist of levels rgamtcadcnrs,ie ate aen agtsets purely target is no policy have effort parties re- clientelist i.e. of U-shaped’ centrist, levels ‘quasi and low a programmatic very bear At effort. should clientelist effort. set clien- to clientelist target of lationship of levels parties’ levels high of intermediate very ‘inclusiveness’ at polarized and the Finally, more low be very should both and at effort, point telist ideal approx- voter’s should median political positions the a programmatic imate of parties’ extent efforts: the linkage of with clientelist levels relationship system’s low ‘hump-shaped’ a very display and should high larization very at intuitions. efforts prevalent linkage curvilinear most clientelist are system’s of with political series relationship a ‘hump-shaped’ a of a extent display are the logic noted, already pro- this as polarized Firstly, in more choosing Embedded and stances. targeting clientelistic grammatic to effort overall less voting ned h qiiru iheteeysmall extremely with equilibrium the Indeed, o nyde h qiiru when equilibrium the does only Not hs yohsscnttt,pras h ae’ oteprclyrlvn the- relevant empirically most paper’s the perhaps, constitute, hypotheses These G P tas ocste dp nraigyplrzdpolicy polarized increasingly adopt them forces also It . ml agtsets. target small δ ssalrpeettepprsfis nwhich in first paper’s the represent small is δ uhta rgamtcplc appeals policy programmatic that such , δ sitreit.I h atr parties latter, the In intermediate. is δ δ scaatrzdb uhsmaller much by characterized is ssfcetysaltemedian the small sufficiently is δ iial,ielgclpo- ideological Similarly, . (Θ P =∅ vnif even ) tintermediate at ; tmasde- means it .Kselman D. δ svery is Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

690 689 688 687 686 685 684 683 682 681 680 679 678 677 676 675 674 673 672 671 670 669 668 667 666 665 664 663 662 661 EDITOR’S660 659 658 657 656 655 654 653 652 651 PROOF650 649 648 647 646 645 ingraet h eaeo letls’ omtv status. informa- normative provide clientelism’s will on it debate well the as to capacity; germane model’s predictive tion this its of identify analysis to empirical consider of serve future levels many will predictions intermediate which only not ‘exclusiveness’, with Thus, democracy. systems and for hand, extremism perilous other generate the to On tend laud- themselves. clientelism consider of many and which in ide- values with able inclusiveness, associated political also ac- and are and moderation model governance ological this forms in palatable systems envi- normatively clientelistic exogenous Highly more the countability. to when conducive option not ‘second-best’ is a ronment as serve may linkage clientelistic clrn-ekn,proaitcacutblt s hl etil mefc,better imperfect, certainly polit- while total constrain is, the to than accountability insufficient are personalistic institutions rent-seeking, public ical where countries in way, other n efitrse oiiin,eetrlrlsascae ihproaitcpolitics personalistic with associated rules constrain- actually of electoral capable politicians, infrastructure self-interested bureaucratic and ing legal exogenous gov- an the of on (Kselman absence work institutions own electoral human my of Finally, and consequences literacy). countries economic ernance and of clientelist mortality dimensions expected; infant select have (e.g. on might development counterparts we their as outperform cut fact clear in as not elected are credible consequences without ( environments Pierskalla and in officials, Fernandez welfare officials. elected social and citizens aggregate average improves between often intermediaries clientelistic as serving of of appraisal capable normative ( nuanced Vlaicu current more and growing Keefer a a linkage. offers However, clientelism position. of this and studies to norms in much democratic undoubtedly of is consolidation There the polit- criticize practices. for and which consequences growth, circles perverse economic with policy governance, accountability and ical of academic form flawed both a in as arguments is clientelism It hear mechanism. to linkage unusual democratic a not as viability clientelism’s on debate is normative role voter’s research. median future to the left which question in a situations is multi-party reduced es- static comparative in to contains above role robust model pivotal the are the a not; Whether hypotheses with outcomes. Furthermore, is voter equilibrium etc. other median game’s not, the the the endows tablishing is while which other parties, clientelistic political the turn is two while in party only extreme which one is options, which party strategic in current one distinct equilibria the have to extend parties lead which will might in work face situations future and to Ideally, sets, model electorate. target equally-sized unbiased parties to ideologically both access situations: an have strategic constraints, their budget of identical symmetry face the from results arises form equilibrium decisions the the policy em- Firstly, in their ways. in come important limited all of are above number results above a the in that applicability admitted pirical be must it claims, main per’s o-xsec hoe o letls nSailModels Spatial in Clientelism for Theorem Non-existence A huhi ifrn otxs hs aessaeteudrurn hta times at that undercurrent the share papers these contexts, different in Though eodtepprseprclipiain,isrslscryipiain o the for implications carry results its implications, empirical paper’s the Beyond pa- the of testing empirical permit may set data emerging newly this Although improve absence oennewe oprdt esproaitcrls ttdan- Stated rules. personalistic less to compared when governance faccountability. of symmetric 2008 oeta h rsneo oa arn,woare who patrons, local of presence the that note ) 2009 taeypols h ymtyo parties’ of symmetry The profiles. strategy n htcineimspolitical-economic clientelism’s that find ) 2008 ugssta,i the in that, suggests ) 195 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

736 735 734 733 732 731 730 729 728 727 726 725 724 723 722 721 720 719 718 717 716 715 714 713 712 711 710 709 708 707 EDITOR’S706 705 704 703 702 701 700 699 698 697 PROOF696 695 694 693 692 691 eito hc osntivleabr lrlt agtst u tews,i the if otherwise, Put set. target plurality Lemma bare in a identified involve strategy not does which deviation Lemma set. aofehnigdvaino h oiia ih novsteplatform the involves right political the on deviation payoff-enhancing em 3 Lemma right. political the right; on political applies the Lemma condition on payoff-enhancing. symmetric deviation payoff-enhancing a is a for set strategy target condition necessary plurality bare a without vector u oe.Temda oe ilb h oe rmti agtstwoealgac will party allegiance opposing whose the at set since point target ideal gain, this from to voter difficult the most be be will voter median The voter. necessity). lrlt agtsto h oiia ih is right political the on set target plurality u esaeayvtr npormai rud.T e hsnt ht when that, note this see To grounds. programmatic on voters any persuade When Lemma 6.2 If Lemma of Proof 6.1 Appendix Theoretical 196 esaesm oeso rgamtcgons ntr,teemyeitpayoff- exist may there turn, In for grounds. deviations programmatic enhancing on voters some persuade otismr oesta eesr ownteeeto,adtu ilntrepresent not will thus and choices election, condition the necessary win to the necessary than voters more contains give [ ihti ltomaetoewt da points ideal with those are platform this with tlt fa least at of utility h upr fvtr h r nistre e.I un n eito rmteoutcome the from deviation any turn, In set. target its v in are who voters of support the 1 / x 1 i, i,P 2 m G ∼P ovtrwl aeaprl rgamtcuiiyfor utility programmatic purely a have will voter no , = hslma el sta o any for that us tells lemma, This yAssumption By sarsl,when result, a As ,( ( P P prog) v ( 3 / 2 prog) G ≤ v hneo inn.Frhroe n agtstgetrta aeplurality bare a than greater set target any Furthermore, winning. of chance a 2 m = P − 1 . / v > 2 ≤ G o any For m = 3 and P 1 v 1 ilne oivleatre e fa es afteeetrt nodrto order in electorate the half least at of set target a involve to need will / / salse h eesr odto taeyfrapayoff-enhancing a for strategy condition necessary the establishes m ) 1 2 3 2 / ] tflosthat follows It . 1 . and 2 o l oes.A el oeta l oeshv programmatic a have voters all that note well, As voters). all for / P n h dooia wn Voter Swing Ideological the and 2 o n party any for sopponent ’s 1 o n candidate any for G G P bv,ti aepuaiytre e ilicuetemedian the include will set target plurality bare this above, P sopponent ’s P ≤ > P 2 1 / 1 o h Case the for hc ontivlecosn aepuaiytarget plurality bare a choosing involve not do which / 2 3 2 and , ∼P ∼P ed osntla to lead not does leads x ˆ h eesr odto taeywtotabare a without strategy condition necessary the x ˆ P P P (G ∼ (G hc hoe h einvtrprogrammatic voter median the chooses which chooses G sopponent ’s P ∼P P P P ) chooses ) > = and , h chooses who 1 x G / x m x ˆ 2 v P i . P m h eesr odto taeyfor strategy condition necessary the x (G = ˆ twl eipsil ofor to impossible be will it , P v ≤ ∼P and 1 m (G P ∼ tmybe may it , ) 1 / P P = chooses 2 ) π v hoe h einvoter’s median the chooses x 3 P m rcl bv ento of definition above (recall / i 2 h oeslatsatisfied least voters the : = > P − ,adfrteevoters these for and 0, 1 / G rae than greater v 2 possible m hnn deviation no then , P ,P 3 and salse the establishes ilol gain only will Θ ofor to x ˆ ˆ .Kselman D. P P (G (G 1 / G 2 P P P P P (i.e. ) ) = = to to ≤ Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

782 781 780 779 778 777 776 775 774 773 772 771 770 769 768 767 766 765 764 763 762 761 760 759 758 757 756 755 754 753 EDITOR’S752 751 750 749 748 747 746 745 744 743 PROOF742 741 740 739 738 737 ovtr nterange the in voters to ic:()te r oae lsrto closer located are they (a) since: oefis htsigielgclvtr a xs ohi h range the in both exist may left. voters political the ideological on swing deviations that for applies first process Note identical An right. political the on x ewl o dniy o any for identify, now will We ro fSub-lemma of Proof [ e o oesi h plcberne uhta hr sn wn oe ideological voter swing no is there that such range, applicable voter the in voters for fied mte) o any for omitted), ewe h ate’rsetv rgamtcstances. programmatic respective parties’ the between ino rgamtcutility programmatic of tion ( Models Spatial in Clientelism for Theorem Non-existence A 10 hscnb erte as: rewritten be can This mlctydefines implicitly hr sn wn oe dooia voter ideological voter swing no is there shso e rgamtcfrparty for programmatic party her for or utility his programmatic as whose voter a voter, ideological swing u-em 1 Sub-lemma Basedon( htteidfeec odtosfrasigvtri h range the in voter swing a for conditions indifference the that h range the oia voter logical h range the chooses x taeyvector strategy ro fLemma of Proof ˆ u 1 3 P P / / oeswt da points ideal with Voters S,P 2 2 (.8) Define < ,x − ( 3 x P prog) G / = S ] 2 ssc,i ( if such, As . P v − nterange the in n the and .7 [ Θ m ) x x sals h olwn Sub-lemma: following the establish I A.3) ˆ G S n h agtset target the and P P = and x sasigielgclvtri h range the in voter ideological swing a as , P =[ S 1] u . o any For v S, nterange the in P ..bt oest h etadt h ih of right the to and left the to voters both i.e. , m .5,. G 3 ∼P x n e opponent her and chooses S P hnoeparty one When C ( 7] when [ 1 1 > prog) P eeae value a generates A.3) 1 [ x / 1 . − i / 2 G eaeloigfrsigielgclvtr nterange the in voters ideological swing for looking are We 1 = / 2 ,x < (x 2 ,x P nt fcineitcefr ilb agtdt oesin voters to targeted be will effort clientelistic of units .2 : 1 x P ∼P / x S > P [ 2 u S ] P Θ 1 − ntr,i ( if turn, In . ] / i,P ilalhv ihrpormai tlt for utility programmatic higher a have all will ∼P ⇒ = G ti hnsrihfradt sals ht(algebra that establish to straightforward then is It . 1 2 P > / ,x chooses 1 2 P ( / 3 =[ , prog) 1 / 2 nplc pc,ad(b) and space, policy in P / > when ) 2 2 G ] = −{ ..when i.e. , x ∼P oseti,nt ht( that note this, see To . P ∼P 1 m P / G 2 o any for , ,( G chooses · h wn dooia voter ideological swing the , 1 v : P ∼P 3 eeae value a generates A.3) m P + 1 hoe h einvtrprogrammatic median-voter the chooses / · x 2 · − and  G S − ( x 1 chooses 1 P abs S nterange the in − G − P P G >x x P (x P x [ hoe nielgcldeviation ideological an chooses P ) chooses x P P ] and P P o xml,if example, For . > ) [ − hnteei osigideo- swing no is there then , 10 } v − G 1 1 / . m / ∼P x 2 G x 2 ,x S and S h olwn expression following the P ) = ] x P bv ple only applies above A.2)  [ > P . ] 1 [ x 1 = ie u specifica- our Given . / 1 P P >G > / 2 1 / 2 1 ,x x a eindifferent be may 2 ,x chooses 1 − S define , / P P 2 P >x [ . : ] abs 1 ] / P G x o ausof values for ∼P 2 r o satis- not are S ,x P P stesame the is [ x eknow we , when P x = x hnfor than m P S ] − then .8 n in and > sthe as (A.1) (A.2) (A.3) x ∼P 1 S 197 / ] P 2 . , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

828 827 826 825 824 823 822 821 820 819 818 817 816 815 814 813 812 811 810 809 808 807 806 805 804 803 802 801 800 799 EDITOR’S798 797 796 795 794 793 792 791 790 789 PROOF788 787 786 785 784 783 ro fSub-lemma of Proof [ u-em 3 Sub-lemma dooia voter ideological xrsinipiil defines implicitly expression 198 ro fSub-lemma of Proof [ hscnb erte as: rewritten be can This dooia voter ideological hr sn wn oe dooia voter ideological voter swing no is there voter ideological voter swing no is there u-em 2 Sub-lemma Basedon( enwmv oietfigielgclsigvoters swing ideological identifying to move now We omitted), ( in condition indifference omitted): ( in condition indifference u pcfiaino rgamtcutility programmatic of specification our 3 1 / / x x 2 2 P P G ntr,frany for turn, In − , , P 1] 1] G . ydfiiin f( if definition, By . ( if definition, By . P x x S S . x ecnetbihtefloigSub-lemmas: following the establish can we A.6) = = S = 3 3 / / o any For o any For 2 2 3 x −{ −{ / x x 2 S S S G −{ = G G nterange the in range the in 1 1 P 1 3 2 P P − +  > G − 1 G G eaeloigfrsigielgclvtr nterange the in voters ideological swing for looking are We eaeloigfrsigielgclvtr nterange the in voters ideological swing for looking are We · · ∅ 3 G G P − / 1 ( ( ( P P / 2 eeae value a generates A.6) .I ste tagtowr oetbihta (algebra that establish to straightforward then is It A.6). (algebra that establish to straightforward then is It A.6). 1 1 x eeae value a generates A.6) P P −{ · 2 x G S ( > > − + S Sub-lemma G 1 x 1 P − +G = P x x S 1 1 + / / ·(1−x P P 1 [ [ 2 2 when / 3 x x x P , , ) ) / 2 } } P P P when when ) 2 ) , , >x ovtr nteapial ag aif the satisfy range applicable the in voters no : the satisfy range applicable the in voters no : G P P 1 u P 1 ∼P ∼P x x fadol if only and if 1 P − if if i,P S S lost express to allows · chooses · fadol if only and if fadol if only and if  x 1 nterange the in range the in G / ( ( 1 P chooses chooses prog) 2 1 P − + ( 3 x x x / P x o any for , P S 2 S v S ) m − < S . and m m G 3 x / nterange the in [ [ ,te hr sn swing no is there then 1, P P 2 P x x and and − ) hnteei oswing no is there then , P P . x x P  G x x , , P P . G P S 1] 1] P chooses P P < > P sfollows: as < > o ausof values for of values for , chooses chooses 3 3 / / 1 1 / 2 2 / 2 2 − − [ G x h following the P x G G P P . , .Kselman D. P P x x 1] > . . P P .Given 1 > > / (A.5) (A.6) (A.4) x x 2 P P : 1 1 / / > < 2 2 , , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

874 873 872 871 870 869 868 867 866 865 864 863 862 861 860 859 858 857 856 855 854 853 852 851 850 849 848 847 846 845 EDITOR’S844 843 842 841 840 839 838 837 836 835 PROOF834 833 832 831 830 829 ai tnefo h einvtrplc dpe by adopted of policy values voter diate median the from stance matic be eitcefrswl edt etree oa oscr aemajority. bare a secure to as so targeted be to need will efforts telistic dooia upres n ntr iiie h ieof size the minimizes turn in and supporters, ideological osm usto oeswt da points ideal with voters of subset some to r ihielpit nterange the in points ideal with ers [ ae oehr xrsin ( expressions together, Taken httencsaycniinsrtg ie some given strategy condition necessary the that x G o-xsec hoe o letls nSailModels Spatial in Clientelism for Theorem Non-existence A rgamtcuiiyfor utility programmatic hnfrparty for than dniytesigielgclvtrfrany for voter ideological swing the identify oe,ie h xsec odtossiuae nSub-lemmas in stipulated conditions existence the i.e. voter, httesigielgclvtri h oe hs da on sietclto identical is point ideal whose voter the is voter ideological swing the that rgamtcpsto.Alvtr ihielpoints ideal with voters All position. programmatic ial,a oeeteevle of values extreme more at Finally, oe ilbe will voter bv httegm’ wn dooia oe a da point ideal has voter ideological swing game’s the that above nprl rgamtcgons n ievrafrvtr ihielpoints ideal with voters for versa vice and grounds, programmatic purely on oeaevle of values moderate ihrpormai tlt for utility programmatic higher nwrs when words, In x ∼ P rgamtcposition programmatic iutnosystse o both for satisfied simultaneously rud.Frany For grounds. 3 ˆ / x P i 2 P S P >x hnfrparty for than (G When oefo h bv wn oe nlssta,frayvleof value any for that, analysis voter swing above the from Note Sub-lemmas x − , ) 1] S ] chooses ..ta agtstms odcv oscrn aemjrt itr,i that is victory, majority bare a securing to conducive most set target that i.e. , G P S ∈[ ..mxmzstenme fvtr h prefer who voters of number the maximizes i.e. , ) P a ecutdo ochoose to on counted be can = x ntr,gvnthat given turn, In . P P 3 / , chooses 2 1] v x P − m x x S n h usto xrms oesi h range the in voters extremist of subset the and , 1 eas h atrhsntsfcetydsigihdhrprogram- her distinguished sufficiently not has latter the because , 2 P / S G ∼P ∈[ and 2 ( and = G P 1 P ∅ ∅ x 3 / lo st express to us allow chooses (G P = P 2 < − )} − P 1 1 x P ) 1] / / P > P 2 2 = stepsto hc aiie h ag of range the maximizes which position the is 1 G G G / nprl rgamtcgons tfollows It grounds. programmatic purely on > 1 x ept h atthat fact the despite 2 elu ht o any for that, us tell A.7) if if if P P P 3 / / S 2 h aenvrhsmr hnoeswing one than more has never game the ∼ 3 3 ) ti tagtowr osefo ( from see to straightforward is it , P G ept h atthat fact the despite P sfollows: as G 1 G < 3 / P 2 P P G G [ 3 3 1 x . / / x / P P Θ ncnrs,a interme- at contrast, In . nprl programmatic purely on 2 2 S P 2 , , − − , G P nldsteplatform the includes > 1] 1 ∼P G G owhich to , x 1 [ ilhv higher a have will / and 2, S P P x Θ 2 = ˆ x , G = G : S prefer P P , ∼P P 1 1] 3 G =[ > / >G > 2 P 3 ilhv a have will = 1 − P / x 1 > r never are ∼P 2 / m 1 P sclien- ’s G G 2 ,( 1 when , >G P . / P (A.7) (A.8) 2 ,vot- 3 x to A.8) ,i.e. / ,the ∼P P P 2 i 199 P − > P ’s ’s . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

920 919 918 917 916 915 914 913 912 911 910 909 908 907 906 905 904 903 902 901 900 899 898 897 896 895 894 893 892 891 EDITOR’S890 889 888 887 886 885 884 883 882 881 PROOF880 879 878 877 876 875 hni htietfidi Lemma in identified that is than h eesr odto taeyfrpyf-nacn deviations. payoff-enhancing for strategy condition necessary the h taeyietfidLemma identified strategy the nacn eito hns ostesrtg nLemma in strategy Lemma the establishes does so then deviation enhancing mrvn eito from deviation improving from eevsauiiyo 1 rmtesto actions of set the from ‘1’ of utility a receives h e opponent her the adt e ht sln stemda oe rfr h eitn candidate deviating the prefers voter median the as long as that, see to ward When x When Proposition of Proof 6.4 actions of set the from ‘1’ of utility a receives voter median The Lemma of Proof 6.3 voter swing the and point ideal median the between voters all targets which 200 aoiyi agtstadtemda oiysac (Lemma stance (Lemma policy left median or right the political and set target is majority when ( be will lasrcieahge tlt rmtedvainsiuae nLemma in stipulated deviation Lemma the in from stipulated utility higher a receive always ad Lemma hand, h einvtrsuiiyfrtencsaycniinsrtg when strategy condition necessary the for utility voter’s median the odto eitoswhen deviations condition opoeLemma prove To ˆ P (G ie eito from deviation a Given v P η m G η ) = .I h atr h einvtrsuiiyfrlclyotmldeviations optimal locally for utility voter’s median the latter, the If A.9). = P oavalue a to = ilbe: will 1 > ,a ogas long as 1, 3 / 2 1 / u − 2 2 m,P party , el sta,when that, us tells G ∼ u 2 G P m,P 2 rtetbihta,frany for that, establish first I , P x(G . u ˆ P ntetx,ie htfrayvleof value any for that i.e. text, the in 3 hnd l te oesin voters other all do then , m,P :( < P x(G ˆ P δ> A.9)>( ,adcnesl slong as conversely and 1, a osdrbt oal pia eitoswt bare a with deviations optimal locally both consider can 2 v ,), x(G G ˆ v m P o h Case the for m P 1 .I h omr h einvtrsuiiywhen utility voter’s median the former, the If 3). Θ oavalue a to ,), / 2 ˆ otencsaycniinsrtg,i sstraightfor- is it strategy, condition necessary the to P < 2 1 P 3 smr ieyt il aofehnigdeviations payoff-enhancing yield to likely more is ,), there Θ (G ..i h taeyfo Lemma from strategy the if i.e. , agbaoitd.Ti ntr mle that implies turn in This omitted). (algebra A.10) 1 ˆ / P 2 Θ η P (G ˆ ilbe: will ) P = osnot does P G = ,temda oe’ tlt o necessary for utility voter’s median the 1, = ) P (G G = G < P P G P 1. v ) + xs aofipoigdeviation payoff-improving a exist m 2 P > + nteohrhn,when hand, other the On . P G + δ< 1 stre e.Temda voter median The set. target ’s P 1 δ / − + δ 2 > 1 1 δ G / + G 1 − 1 2 + u o ievra This versa. vice not but 3, 1 / P / − P 1 there 2 2 G 1 h einvtrwill voter median the , < − / ,o eitost the to deviations or 2), G P 2 . Lemma 1 P G does P . 2 v G m ilsapayoff- a yields . xs payoff- a exist nteother the On . P < 2 2 .Kselman D. ilbe: will 1 identifies hnthat than (A.11) (A.10) η η (A.9) x P = S = = to 1, 1 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

966 965 964 963 962 961 960 959 958 957 956 955 954 953 952 951 950 949 948 947 946 945 944 943 942 941 940 939 938 937 EDITOR’S936 935 934 933 932 931 930 929 928 927 PROOF926 925 924 923 922 921 ( ita 18)Cniaemtvtos ytei fatraietere.A oi c Rev Sci Polit Am theories. alternative of 34:565Ð598 Sci synthesis Polit a J motivations: Am parties. Candidate political (1983) competitive D of theory Wittman behavioral from A evidence (1990) with K politics Strom machine of model formal a accountability: Perverse Cam- (2005) Press, SC MIT Stokes policy. economic Explaining Polit economics. Am Political ballot. (2000) secret G Tabellini the T, and Persson politics Machine buying? turnout or buying Vote (2008) 21:128Ð140 S Econometrica Nichter games. cooperative Two-person (1953) of JF Nash model a diversification: portfolio and Clientelism (2007) al et A perspec- cross-national Diaz-Cayeros A B alignments. voter Magaloni and systems Party (1967) (eds) S Rokkan SM, University Lipset Duke thesis, masters Economics choice. H, intra-party and Kitschelt Governance (2008) In: D Kselman introduction. an linkages: Citizen-politician (2007) S 24(2):371Ð Wilkinson Organ Econ H, Law J Kitschelt democracy. and clientelism Credibility, (2008) R 11:351Ð386 Sci Vlaicu Pol P, Rev Keefer Annu Coalitions. (2008) M Do Humphreys performance. economic political and strategies Partisan (2009) York New J Row, and Pierskalla Harper democracy. M, of theory Fernandez economic An un- (1957) motivations, A candidate Downs model: voting multi-dimensional the of Robustness (1985) RL Calvert References of values for that, see to straightforward is it turn In Models Spatial in Clientelism for Theorem Non-existence A 1 δ −G 77:142Ð157 99(03):315Ð325 Rev Sci Polit Am Argentina. bridge 102(01):19Ð31 Rev Sci Patrons, (eds) S Cambridge competition. Wilkinson Cambridge political Press, and H, University accountability Kitschelt democratic of In: Patterns Mexico. policies. and to clients, applications with investment electoral Toronto Press, Free tive. Economics of Department po- and accountability democratic Cambridge of Press, Patterns University policies. Cambridge competition. and litical clients, Patrons, (eds) S Wilkinson 406 3rdÐ September Canada, Toronto, association, an- science 2009 political the American at 6th presented the Paper of fortunes? meeting economic nual affect accountability democratic of modes 29:69Ð95 Sci Polit J Am convergence. and certainty, + 1 / P 2 ) a nyb rae hn‘’if ‘1’ than greater be only can δ> 1 / 2 agbaomitted). (algebra G P < ,tefunction the 1, G P 201 + Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 h ainlSineFudto rvddfiaca upr o hsrsac hog Grant through research this for support SES-0242255. financial No. provided Foundation Science National The by work from know We voter. opponent median an the faces of election position (1929 rule policy Hotelling the plurality adopted two-candidate has a who in candidate a Suppose Introduction 1 Niou M.S. Emerson and Lacy Dean Elections in Packaging Issue and Preferences Nonseparable .Lc ( Lacy March D. Juan The 2012. Model, 27Ð28, Spatial the April of Spain, Applications Madrid, Contemporary Institute, on Conference the at Presented DOI (eds.), al. et Schofield N. USA 27708-0204, NC Durham, University, e-mail: Duke Science, Political of Department Niou E.M.S. USA 03755, NH Hanover, 6108, HB e-mail: College, Dartmouth Government, of Department ed ntesrcueo oe rfrne nteoiia oiysaeadtenew the and space policy original the on preferences voter of structure the on pends issues. new the introducing into by issues elections new win of bringing potentially new to scope can extends bringing candidates observation the Losing as his election. expanding strategy But conflict. from this the of benefit into conceived groups may originally conflict Schattschneider political conflict. ( Schattschneider the a As in space? policy losers current the observed, in unbeatable is no who opponent has space policy can- one-dimensional A a positions. in defeat. changing position accept for to losing but candidates a recourse to the reputations pinned penalizing party is voters to who due to didate perhaps or space, issue are policy candidates the position the the on policy in that the further freely adopt Suppose moving election. also from tied to restricted a is yielding do voter, median can the candidate of challenging the best the space hte h taeyo nrdcn e susit neeto ilscedde- succeed will election an into issues new introducing of strategy the Whether an facing candidate a to available are strategies what ask: we chapter this In 10.1007/978-3-642-35239-3_10 [email protected] [email protected] B ) ,Bak(1948 Black ), dacsi oiia Economy Political in Advances ,adDws(1957 Downs and ), pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , , ,ta naoedmninlpolicy dimensional one a in that ), 1960) 203 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 rdcsamda naldrcin ilteeb neulbim(Plott equilibrium points an ideal be voter there of will distribution directions the all not which in will in median candidates case a for rare produces position the equilibrium in an Only and exist. expands generally space issue the multi- of a ality is model voter median the Proba- on diverge. variant space. to policy realistic candidates dimensional and for interesting most incentives activists the create to bly appeal all to contributions candidates for campaign need for un- the How- and candidates, tie. voting, Policy-motivated probabilistic a positions. voters, certain distinct in adopt resulting candidates voter, the median elections, the of position this ever, the to converge candidates oe yseigtepsto ftemda oe hnteplc pc sone- is space freely policy move (Hotelling the can space when candidates policy voter and the median single-peaked, in for the are compete preferences of election voter position plurality dimensional, the single-winner seeking a by in votes candidates Two results. oretical Schofield oto h eerho lcoa optto a enasac o lcoa equi- electoral for search a been has competition (Black electoral libria on research the of Issues Most of Number the and Competition Spatial 2 on competition. preferences candidate public of of logic complexity the influence The profoundly issues. can of issues portion variety policy substantial a important a on by public voting held election the are 2004 of nonsepa- preferences a from that nonseparable Results show that curiosity. then demonstrate theoretical We survey a opponent. than her more beat are preferences disadvantaged to rable the the position then for a issues, preferences find the for nonseparable cannot preferences have candidate op- separable the voters her have voters some that that all position if show If a only issues. We take but advantage. and beat, an issue cannot new candi- has ponent a which candidate introduce on one can issue candidate and disadvantaged single fixed between on are competition competing positions spatial begin dates’ of candidates model The a candidates. present We two elections. in strategically issues. other sues on positions candidate’s candidate’s the a from evaluates separately issue issue preferences each each separable than on booth position with voting voter the in a weight Conversely, greater federal separately. carry pref- increase nonseparable issues to have of voters pledges packages When also children. erences, their candidate and the mothers can- also single if a candidate to prefer only assistance that may abortion voter if a opposes only Or who a taxes programs. didate example, cut spending government to For specific promises issues. cut who to one related pledges candidate other, on a on position prefer position candidate’s nonsep- may candidate’s a voter has the for voter on preference a depends her When issue issues, nonseparable. across are introducing preferences by preferences election arable voter an in which advantage over an issues gain can candidates particular, In issues. 204 svtr n addtstk oiin nmr hnoeise h dimension- the issue, one than more on positions take candidates and voters As osprbevtrpeeecsoe potnte o addtst akg is- package to candidates for opportunities open preferences voter Nonseparable addt convergence candidate 1978 1948 nlwadHinich and Enelow ; ;Downs 1957 1929 rdcinrrl t elt.I ottwo-candidate most In reality. fits rarely prediction Plott ; Black ; 1984 1967 .Ti oyo ieaueofr la the- clear offers literature of body This ). 1948 ai tal. et Davis ; ;Downs 1957 1970 .Lc n ...Niou E.M.S. and Lacy D. .Tersl sthat is result The ). McKelvey ; 1967 ;Davis 1976 ; Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 fnneaal rfrne a dnie ntepbi hieltrtr er ago years literature importance choice The public that policy. the assumes in public (Kadane competition identified of was electoral issues preferences and nonseparable across of behavior preferences voting separable in have voters research issues. the the across of nonseparable are Much preferences voter whether on depending strategy iisfrcniae opcaenwise ihodise o lcoa gain. electoral for issues old with opportu- issues create new preferences package voter to nonseparable candidates that for voting show nities the we in chapter preferences this nonseparable of In extent public. the or strategies candidate for ences 1984 h hg n eort,ple oesaa rmteWis n wp h Whig the (Riker swept landscape and both electoral Whigs, from American the the from them from away differentiated voters party slavery pulled restricting Democrats, and on Whigs position the ex- their mimicked the development halt but economic to Whigs’, on platform the position a Republicans’ of 1860s The slavery. and of 1850s pansion the in the adoption party’s split Republican to try the (Riker should opponents candidates that their been of has support space issue the expanding on wisdom tal. et Elections in Packaging Issue and Preferences Nonseparable we w addts ahcniae(rpry dpsavco fisepositions be- issue game of competition vector a spatial adopts a party) on (or based candidate packaging Each candidates. issue two of tween model a present We Packaging Issue of Strategy The 3 well- a defeat candidate a help can issues (Schattschneider or opponent issue new positioned a introducing adopt. election, can an they in positions policy the in restricted multidimensional are the candidates on variations which empirically explored in has model and date meaningful to research substantively little Yet are plausible. movement candidate on (Aldrich issue restrictions an on position a to candidate 1983 a confine (Petrocik affiliated may change leaders to their party difficult and and are Activists that Parties issues plausible. elections, on real reputations be develop In not candidates space. policy may the movement in candidate freely move unrestricted Both can competition. candidates electoral that on require research results of foundation the form dimensions Hinich multiple the and (Enelow splits direction it possible that opponent every condition her in rare half that the in under winning position exactly new except voters new find unbeatable to a is space position by policy no the defeated positions, around de- be dance to can then position candidates will new Although also that adopts. a can space adopts issue opponent who the her challenger in a feat position However, another position. given always any winning is a defeat there dislodging since exist, easy not is does (McKelvey winner candidate Condorcet cycle a endless when almost dimensions multiple an in opponents their sw ilso,teitouto fnwise na lcincnb successful a be can election an in issues new of introduction the show, will we As hncniae r osrie nteraiiyt hnepstoso h issues the on positions change to ability their in constrained are candidates When in result instability general the and dimension one in result voter median The .Ltl oksnete a xmndteipiain fnneaal prefer- nonseparable of implications the examined has then since work Little ). .Vtr a eaiecniae o flpflpig nise.Alo these of All issues. on “flip-flopping” for candidates penalize may Voters ). 1972 1972 .I h bec fa qiiru,cniae a dp oiin obeat to positions adopt can candidates equilibrium, an of absence the In ). ;Kramer 1972 McKelvey ; 1982 1960 .Acasceapei mrcnpltc is politics American in example classic A ). 1976 ;Riker Schwartz ; 1982 1982 1976 , ). 1977 1986 Schofield ; nlwadHinich and Enelow ; .Teconventional The ). 1984 1978 ). 1996 .In ). 205 ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 positions eeaiy ersrc teto otoissues, two to attention restrict we generality, line rbepeeecsocrwe,frisac,ise1i osprbefo hl sseparable is 2 while 2 from nonseparable is 1 Niou and issue (Lacy instance, 1 for from when, occur preferences arable rta esnspeeec noeisedpnso h hie vial rthe or available choices the on depends issue issue. one another on on issues, preference outcome among person’s interdependence a imply space. that preferences the or Nonseparable of space. axes the the of axes to parallel are axes preferences whose ellipses or circles h lentvsadpeesteoeta scoett e Eeo n Hinich and (Enelow her to closest of is voter each that to one a point the alternatives, ideal prefers more her and from alternatives or (GED) the two distance across Euclidean generalized choice the a compares with confronted When function. in 206 1 addt’ rpslt u udn o dcto nestecniaeas promises also may a candidate of voter taxes. the disapprove unless cut the may education to voter for mothers, funding a teenage cut Or, to abortion. unwed proposal on candidate’s for position candidate’s candidate support the the welfare of if disapprove end But abortion. to against promises initially position may also announced voter candidate’s a example, a For of issues. other approve on stance candidate’s the on depends the marriage, arbitrary. same-sex is allowing issues the or in Security be- complementarity clear of Social distinction a direction without privatizing the issues as spending, For education such meaningful. is or direction, complements taxes negative “directions” in and clear positive decreases with tween issues or For increases dimension. other as the such on more gets she as mension ahohr esnwnsmr noedmnina h eevsmr nanother on more receives she as dimension Newing one and on (Black more dimension wants to person related a positively other: are each that issues are complements Positive complements. negative ohrielpitmaue ngnrlzdEciendsac.Teeoe voter Therefore, distance. on Euclidean position whose generalized candidate in the for measured votes point ideal her to point, oesit hs lsrto closer those into voters oiinon position inn oiinsneamjrt fvtr r lsrt hnt .CniaeB issue, Candidate B. new to a than issue A on on to position position closer are a voters announces of majority then a since position winning a nue da point ideal induced n aro suscudb opeeynneaal rprilynneaal.Prilynonsep- Partially nonseparable. partially or nonseparable completely be could issues of pair Any n e of set A ttesato h election, the of start the At eaal rfrne r niae yidfeec otusta r concentric are that contours indifference by indicated are preferences Separable elblvtrielpit ytevtrnme,1 number, voter the by points ideal voter label We position candidate’s a of evaluation her preferences, nonseparable has voter a If di- one of out less wants person a which on issues are complements Negative dmninlEciensae o upsso lutainadwtotls of loss without and illustration of purposes For space. Euclidean -dimensional AB z i cutpoint, A . otiigtecniae’pstos oe oe o h addt closest candidate the for votes voter A positions. candidates’ the containing A Y ={ M r niae yidfeec otuswoeae r o aallto parallel not are axes whose contours indifference by indicated are . ≥ X A Y oesec a da point ideal has each voters 3 ,Y addtscno hneterpstoson positions their change cannot Candidates . 2000 z i A A+B } stepito agnyo e nifrnecnor nthe on contours indifference her of tangency of point the as ;Lacy 2 and A 1 h oefrA n hs lsrto closer those and A, for vote who , suscnb eae oec te sete oiieor positive either as other each to related be can Issues ttemdon between midpoint the at B {X 2001 1951 ={ A ,X ). X ). B B ,Y }∈ B } AB respectively. , 1 Y ≡ X scoett h oe’ nue ideal induced voter’s the to closest is addt a hnanuc a announce then can A Candidate . X θ and i and , ∈ Y A n X addtsAadBadopt B and A Candidates . , and n uscnaeutility quasiconcave a and 2,...,m A = B .Lc n ...Niou E.M.S. and Lacy D. X B on B addt sin is A Candidate . en voter Define . X ,whovoteforB. AB ste dp a adopt they as Nonseparable iie the divides , 1984 i ’s ). i Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 onson points Fig. In duce. ne r eaal rnneaal.Spoeta l oeshv eaal prefer- separable have voters issue, original all the that across Suppose ences nonseparable. or only separable the are space, ences issue issue. new dimensional if a one even introduce a tie to in is a candidate challenger position a than a winning confronting for better a When recourse any issue. out do the staked on assumption, can has position rigid candidate who any more challenging choose a can a voter, challenger way winning the median the no the If is fixed. of there are position then positions the candidate adopted and has voter candidate when candidate that beat to sition points, oiinda h da on idcdo issue on (induced point ideal the at positioned oeshv nue da points ideal induced have voters matches oe on voter na su htteohrcniaecno vroe vni addt a move can B candidate if Even advantage overcome. an cannot has candidate freelyonissue other candidate the one that which issue in an scenario on realistic a capture to is sumptions issue on advantage an has A, demands, activist positions or candidate way, reputation issue Either party positions. on changing as for such candidates constraints penalize voters to or due issue the on nously issue, initial the on positions their contour indifference the on point. falls ideal position her whose to candidate closest the chooses voter Each Elections in Packaging Issue and Preferences Nonseparable nyi e susaeitoue vrwihsm oeshv osprbeprefer- nonseparable have voters some which over ences introduced are fixed issues new are if positions only candidate which on issue nue da on a elsic oBssd ftecton,i hc aeB case which in cutpoint, on the position of B’s side “mimic” B’s can to A switch However, well wins. may point ideal induced ahdlnsacoe yterpstoson positions their by anchored on lines positions dashed their by constrained are dates ini addtscnmv reysnetedsrbto fvtrielpit does points ideal al. voter et of (Davis directions distribution X all the in since median a freely produce move not can candidates if tion i.Bt addt a obte yitouiganwissue. new in a Proposition up introducing end by and better position A’s do adopt can to B is candidate do But, can tie. B a candidate that best the voter, median steol su,cniaepstosaegvnby given are positions candidate issue, only the is h taeyo nrdcn e su igsciial nwehrvtrprefer- voter whether on critically hinges issue new a introducing of strategy The way no is there then space, issue dimensional one a in winning is candidate a If addt nrdcsissue introduces B Candidate h oe nldstoadtoa supin.Frt addtscno change cannot candidates First, assumptions. additional two includes model The . B z  i  h e addt pc sthen is space candidate new The . X X B X ∈  eanfie.Scn,w sueta n addt,abtaiylabeled arbitrarily candidate, one that assume we Second, fixed. remain rcoe otemda oe hncniaeB h ups fbt as- both of purpose The B. candidate than voter median the to closer or A aee,respectively, labeled, on oesaelbldb hi da ons ,2 n ,wt nue ideal induced with 3, and 2, 1, points, ideal their by labeled are voters 1, natocniaepuaiyelection, plurality candidate two a In  B X Y  r rhgnlpoetoso hi nue da points, ideal induced their of projections orthogonal are , ic l oeshv eaal rfrne,teridcdideal induced their preferences, separable have voters all Since . n ofot nopnn h a tkdottepsto fthe of position the out staked has who opponent an confronts and X addt ol ea h oiino h median the of position the at be could A Candidate . X Y z n n e issue, new any and , z X 1 1 n a aeaypsto.SpoeBtkspo- takes B Suppose position. any take can and , ihrtecniaepstosaegvnexoge- given are positions candidate the Either . and z 2 and , z 3 AB X ntecniaespace candidate the on X z , 3 . n a oeol ln h vertical the along only move can and X hnta addt a edefeated be can candidate that then hr sn qiiru nti elec- this in equilibrium no is There .  ihnwcutpoint new with , ftemda voter, median the of ) Y 1972 facniaei inn none on winning is candidate a if yaotn position a adopting by A Y .A h tr fteelection, the of start the At ). htacniaecnintro- can candidate a that , and B n addt is A candidate and , AB A+B 2 z h candi- The . 2  h other The . oe 2’s Voter . z i A ∈  AB that 207 , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 n aepeeecsta r ihrsprbeo osprbea oga she as long as nonseparable or separable either are that prefers preferences have and ne.Smlreape r osbewe oeaevtr2hssprbepref- separable has 2 voter moderate a when possible are examples Similar ences. win. n w oesoelp u hs ra r u frahfrAdet e position her to due A for reach of out are areas these on But overlap. voters two any B oes nifrnecnor htinclude that contours indifference voters’ n .Frisac,vtr n ohprefer both 2 and 1 at voters line instance, dotted For can- vertical A 2. the Candidate and on nonseparable. position are a preferences 2’s adopt voter not but issues, two the across scoe oamjrt fvtr on A voters Candidate of candidates. majority the a to relative to voters closer the of is positions the preserve thus and to moves B After B’s position. Candidate to closer is 3 voter position; A’s Candidate Voters1and2arecloserto preferences. separable 1 Fig. 208 yfidn e su vrwihvtr2hsnneaal rfrne.Candi- preferences. on position nonseparable any has adopt issue 2 cannot on and voter election preferences which the voter over loses of A issue date distribution new the a of finding by outside is B issue didate on position a with in o htcudbeat could that A for tions A position, another find can A fidcdielpoints, ideal induced of distribution the on advantage scoe ovtr n than 2 and 1 voters to closer is r lsrt addt ’ oiino issue on position A’s candidate to closer are htAcno etwt n oiino h e su.I Fig. In issue. new the on position any with beat cannot have voters A all that when issue. A over original victory the a on guarantee can had that she preferences. B separable advantage for electoral position no the is maintain There on and position B’s issue candidate new adopt always the can A candidate preferences, separable have   htmitisher maintains that , oes1ad2aecoe to closer are 2 and 1 voters , oe ol epstoe nweei h su pc otergto oe 2 voter of right the to space issue the in anywhere positioned be could 3 Voter h xml osntrqieta h einvtrhv osprbeprefer- nonseparable have voter median the that require not does example The hnsm oeshv osprbepeeecs hnBcnfidaposition a find can B then preferences, nonseparable have voters some When X . l oeshave voters All B  oaypito h otdln at line dotted the on point any to z i  B  , X B hti oeeteeta n oe’ oiin Can- position. voter’s any than extreme more is that  r nteaesi hc h nifrnecnor of contours indifference the which in areas the in are B B   hnto than A ngnrlzdEciendsac.Teposi- The distance. Euclidean generalized in  B  B uta h a on was she as just  A hr sn oiinAcnaotthat adopt can A position no is There . A X oe ’ rfrne r separable are preferences 1’s Voter . ti lontwrh htBbegins B that noteworthy also is It . hncniaeBaot position adopts B candidate When . B A  to htalw e ownvtr 1 voters win to her allows that A  since Y .Lc n ...Niou E.M.S. and Lacy D. AB htwl lo e to her allow will that A  2 hnalvoters all When . oes1ad2 and 1 voters , sotieo the of outside is X u wins but Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 lsrt h einvtrta .CniaeBde o aet dp h position the adopt to have not issue does on B voter Candidate median B. the than of voter median the to closer ierpcae fise htapa ovtr hnteise niiulymight individually issues the when voters to en- appeal to not. that preferences issues nonseparable of giving voters’ packages exploiting West, of gineer and possibility are strategies North the Candidates’ by the in- decades. richer for for in made majority support voters electoral national and of a slavery Republicans interests to the the opposition linked of to improvements combination promoting products The ternal while their parties East. slavery ship major the of farmers two in Western expansion the markets help Westward of would one the that to development restricting party infrastructure of minor the platform a during a from party work- rose on Republican aid quickly The to 1860s globalization. programs and to insurance 1850s due of social Way” disappeared “Third and jobs a training whose advocated job ers Blair Tony with and of combined Clinton issue trade Bill him any issues. free with on of agreed set him they whole then with the reduction, agreed on deficit they and if defense, that national spending, voters taxing, American told 1980 in Reagan piv- one a find The to hold. the opportunity an to in B result position. candidate voters gives winning the preferences three for nonseparable of order with who voter in voters one otal of preferences Only number nonseparable critical election. a have not the must nonsep- is There in are preferences. preferences issues nonseparable has voter original example which the on from issue an arable introducing by position winning to moves she after B candidate supporting to switch 2 and osnthv ob oae ttepsto ftemda oe on voter median A the of Candidate position preferences. the nonseparable at located has be 1 to have voter not extreme does more a while erences to than Voters1and2arecloserto preferences. nonseparable 2 Fig. Elections in Packaging Issue and Preferences Nonseparable i akbt oes1ad2 and givenherpositiononissue 1 voters both back win issue on position a find B Figure oiia addtsfeunl rsn oeswt akgso sus Ronald issues. of packages with voters present frequently candidates Political  addt cannot A Candidate . B Apivotalvoterhas nissue on 2 lutae htacniaecnmv rmalsn oiint a to position losing a from move can candidate a that illustrates X .Voters1 Y to X A Y . X sln sseis she as long as 209 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 cniinl ftegvrmn eue h muto oe tsed nedu- on spends it money of amount the reduces government the If (conditional) the to taxes in pay people that money of amount the want you Do questions (unconditional) The education. on spending whether from detect Later nonseparable to were: issues. are questions other taxes “conditional” opin- of on two respondent’s outcome the preferences answered a to respondents asks reference surveys. survey without it the existing isolation, since in in to “unconditional” issue similar an question question on of ion a type answered this the first label For We respondents paired. were taxes, spending of education nonsep- and uncover issue issue taxes not—to each instance, some For paired related, preferences. obviously questions arable issue—some manageable, other in survey one impossible the only been with make have nonsepa- To would survey. detecting issues be voters, 20-minute of could some a combinations issues for all twelve issues across the remaining preferences debates the rable of candidate of each in all Although from prominently election. nonseparable figured the about that commentary issues and different twelve about tions atcpt natrewv uvy ihWv odce pi 7My3 (N 31 interviews 27ÐMay April telephone conducted than 1 Wave responses with valid survey, surveys three-wave and complete a reliable in to participate more respondents in allows Krosnick results survey and often (Chang the and of leisure, format their computer at The WebTV. by istered 6preti ae1 5preti ae2 n 7preti ae3. were Wave in rates percent Completion 77 3. and Wave 2, Wave in in interviews percent 85 completed 1, also Wave in respondents percent new 76 211 of sample A ne r osprbears sus usin eindt eetnonseparable citizens. detect US to of prefer- survey designed voter 2004 a Questions whether on issues. measure appeared preferences across to nonseparable designed are questions pub- ences Few include reality? surveys empirical opinion or curiosity lic theoretical a preferences nonseparable Are Preferences? Nonseparable Have Voters Do 4 210 2 etme 7Otbr7(N 7 17ÐOctober September nweg ewrsrcutdoe 000sbet ainiet atcpt nsresadmin- surveys in participate to nationwide subjects 50,000 over recruited Networks Knowledge h muto oe htpol a ntxst h Sgovern- US the to taxes in want pay you people would to that then ment now, money spends of it amount than the less percent 25 to cation to government US .Arno apeo h nweg ewrspnlwscoe to chosen was panel Networks Knowledge the of sample random A 2009). ou ite a o1 oeta esednow spend we than more % now 10 spend to we say than little, now more a spend % up we 25 go than to more say % somewhat, 50 up to go say lot, a up go now spend we than less % now 50 spend to we say than lot, now less a spend down % we go 25 than to less say % somewhat, 10 down to go say little, now a spend down we go than more levels % current now 10 at spend to remain we say than little, now more a spend % up we 25 go than to more say % somewhat, 50 up to go say lot, a up go = 4) n ae3 oebr1Ðeebr3 04(N 2004 3, 19ÐDecember November 3, Wave and 947); 2 h uvycnandques- contained survey The .Lc n ...Niou E.M.S. and Lacy D. = 38;Wv 2, Wave 1308); = 717). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 osprbePeeecsadIsePcaigi Elections in Packaging Issue and Preferences Nonseparable n,tae n mgainalso infiatpretgso oeta voters with potential voters of of percentage percentages small a significant even show that nonseparable spend- Recall all have defense preferences. respondents nonseparable imigration Medicare, of with spending, and percent education 20 trade, taxes, least ing, as at For such preferences. from issues, Issues nonseparable the ordered preferences. of of are percentage more issues combined or smallest The The half survey. the separable. the to are in largest issues responses the twelve of all percentages for complements) remaining negative and positive (both na- a owners weapons. and gun assault for immigration checks on background adoption, ban only, a sex 2 wave and same Security in and, and Social plan, marriage insurance spending, health sex care tional same answered health trade, Respondents and free people. spending and some defense to paired nonseparable that are questions issues other or- many reverse sues, in questions questions In conditional the increase. two answered to the others taxes question, answered der. wants unconditional respondents de- person the some to a before taxes sample, decreases, wants split-half person spending a a nega- education increases, nonseparable as education spending indicate as crease; education diagonal decrease as the to di- complements: in- taxes below tive the to wants Responses taxes above but of decreases. wants increases Responses issue spending person spending spending. the education a on education as complements: change crease positive of not nonseparable level do indicate the con- diagonal agonal the on prefer- the to depending whether on responses of reveal responses taxes crosstabulation all questions a questions, In conditional on nonseparable. ditional two conditional or separable spending The are education taxes. ences for of survey levels the different on appeared questions Similar edu- on spends it money of amount the increases government the If (conditional) Table is- spending and taxing for expected be should preferences nonseparable While 1 hw h ecnaeo epnet ihnneaal preferences nonseparable with respondents of percentage the shows h muto oe htpol a ntxst h Sgovernment US the to taxes want in you to pay would people then that now, money spends of amount it the than more percent 25 to cation odw o,syt 0%ls hnw pn now spend we than less % 50 now to spend say we lot, than now a less down spend % go we 25 than to less say % somewhat, down 10 go to say little, now a spend down we go than more levels % current 10 at now to remain spend say we little, than now a more spend up % we go 25 than to more say % somewhat, 50 up to go say lot, a up go now spend we than less % 50 now to spend say we lot, than now a less down spend % go we 25 than to less say % somewhat, down 10 go to say little, a down go levels current at remain 211 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 l rfrne vrise a aeidcdielpit htaemr extreme. more are that points ideal induced have can issues rather over distributed preferences normally ble are preferences (Fiorina voter bimodal that than and issues elected most and on voters moderate among polarization in rise (Abramowitz the officials about written been issues.has Much of bundles itics. on extreme more them make in that But may points space. preferences ideal the nonseparable induced of have have middle who preferences, the voters in moderate separable space, remain have multidimensional always voters a voters issue all moderate of dimensional which points one in ideal a space the In multidimensional preferences. a nonseparable or have space who ones the are space odtoa ntxssnevtr a aeprilynneaal rfrne (Lacy spending preferences education nonseparable for partially preferences have nonseparable may voters with since percentage taxes the on match conditional to have not does tion osprbepeeecscet potnte o addtst akg susstra- issues package to tegically. candidates for opportunities create preferences nonseparable * preferences nonseparable 1 Table 212 3 oeswoaemr dooial xrm.Ti nigsget htteexample the than that issues suggests most finding on This preferences Fig. extreme. in nonseparable ideologically have more midpoint to are the likely who at voters more ideological moderates, are seven-point ideological scale, standard as the of a themselves described on who self-placement Voters voter’s scale. the by down broken akrudCek sal epn a 451 ban weapons Assault Checks Background Marriage sal epn a akrudcek 448 checks Background Immigration ban weapons Assault care health National dpinMrig 2 6.82 626 Marriage Adoption ees pnigMdcr pnig6212 18 622 621 spending Medicare spending Defense Immigration spending Defense spending Medicare reTrade Free rvtz oilScrt reTae676. 617 Trade Free Security Social Privatize dcto pnigTaxes spending Education Taxes Issue h ecnae frsodnswt osprbepeeecsfrtxscniinlo educa- on conditional taxes for preferences nonseparable with respondents of percentages The N 2, wave from Questions Table h eut a loepaneiec ftedspern etri lcoa pol- electoral in center disappearing the of evidence explain also may results The 2 ecnae frsodnswt osprbepeeecs ore 04pnlsre of survey panel 2004 Source: preferences. nonseparable with respondents of Percentages a o efroffo eleetosweevtr ntemdl fteissue the of middle the in voters where elections real from off far be not may 2 3 hw h ecnaeo epnet h aenneaal preferences nonseparable have who respondents of percentage the shows 2010 2005 = Adoption ainlhat ae688 628 care health National rvtz oilScrt 2 14 623 Security Social Privatize odtoa on Conditional dcto pnig6346.7% 623 spending Education 462 .Btohreiec ugssta otvtr remain voters most that suggests evidence other But ). .A Fig. As ). 2 hw,mdrt oeswt nonsepara- with voters moderate shows, 2 2 622 2 3. 621 2 42 620 a 04 N 2004, May Positive N * * complements 1. 4...... 80 31 22 51 61 88 33 28 61 29 = .Lc n ...Niou E.M.S. and Lacy D. 735 complements Negative 2001 5 2 7 6 7 ...... 8 1 8 7 6 2 3 6 2 6 5 2% ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 da on ie h osrit fteotosbfr i—addt positions him—candidate before options the of and constraints the given point ideal zdo etit r ae nitrrtn h itiuino oe da onsissue points ideal voter of distribution the (Fiorina interpreting issue on by based are centrist, or ized reteitvtr permdrt eedn ntecntansipsdb other by imposed positions. constraints candidates’ the the on or extremist depending issues moderate appear moderate are voters appear voters moderate voters extremist make whether or may about preferences conclusions Nonseparable extreme. draw or to issues across preferences inn neeto yitouignwise,sm oesms e h susas of issues hope the any see have must to voters candidate some disadvantaged issues, new a linked. introducing For by victory. election to issue an path new winning a a introducing candidate’s alone simply a But not be opponent. may is entrenched issues an new unseating space Introducing of issue election. hope multi-dimensional an only a in to choice strategic moving nature chapter, a the be this alters can in fundamentally show already issues We we multiple election. As to elections. close issue of a one of from balance moving the that tip know can issues the Changing contestants. on tion on xlso fcnetnsbcuei snvrtu htteblnermistesame the remains balance the and ( that changed” inclusion true is the never number with is the if it . . because deals. contestants strategy of “Political exclusion wrote, Schattschneider E.E. As 5Conclusion issue on point ideal an has instance, for 2, Voter preferences nonseparable 2 Table Elections in Packaging Issue and Preferences Nonseparable akrudCek sal epn a 3. ban weapons Assault Checks Background Marriage Adoption ainlhat aeImmigration care health National Immigration sal epn a akrudcek 13.01 Trade checks Free Background Security Social Privatize ban weapons Assault Trade Free ees spending Defense dcto pnigTaxes spending Medicare spending Education Taxes Issue X B u hnissue when But .  i xrm.Dbtsaotwehrvtr r xrm rmdrt,polar- moderate, or extreme are voters whether about Debates extreme. —is X ecnae frsodnswt osprbepeeecs ore 04pnlsre of survey panel 2004 Source: preferences. nonseparable with respondents of Percentages vntog oe ’ da on a emdrt on moderate be may point ideal 2’s voter though Even . 2005 Abramowitz ; Y sitoue,h uprscniaeBseteeposi- extreme B’s candidate supports he introduced, is Adoption eiaespending Medicare ainlhat ae24 .22 23 Security Social Privatize care health National Marriage ees spending Defense dcto pnig59 spending Education on Conditional 1957 4) h aemyb ado oiia susas issues political of said be may same The 941). , 2010 .W edmr nomto bu voter about information more need We ). X htmkshmtemda voter median the him makes that 37 40 19.11 11 54. moderates Ideological 7. 9...... 52 03 08 58 92 13 13 %48 84 5% X non-moderates Ideological i induced his , 3 2 6 6 4 1 7 ...... 4 3 6 0 0 8 7 0 2 7 1 5% 213 A Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 ald“eeteis”o h r fpltclmnplto (Riker Issue manipulation political Riker candidate. William of what disadvantaged art of the a part or politics, “heresthetics,” for electoral called of beneficial strategy prove fundamental strategy a the issue is will packaging new preferences a nonseparable “flip- have introducing voters penalize new of some include voters to when conflict if only of But strategy scope the issues. costly expanding by potentially compete changing can a of candidates floppers,” Instead issues, opponent. advantaged existing an beat on to positions order in issue new a on position and new structure produces realistic competition additional electoral some also of results. Imposing model prediction surprising elections. spatial This real multidimensional advantage. the electoral fit on an to positions for appear change quest not will never-ending does candidates a and in exist issues candi- generally two on not with will model equilibrium primarily multidimensional an reality, a fit dates, In not multidimensional. does result is theoretical politics This because voter. median the of position the lt R(97 oino qiiru n t osblt ne aoiyrl.A cnRev Econ Am rule. majority under possibility its and Polit equilibrium J of Am notion study. A case (1967) 1980 CR a with Plott elections, presidential in Sci ownership Issue Polit (1996) JR J Petrocik for Am implications 12(1):5Ð31 some and responses. Polit models Theor voting survey J multidimensional referendums. in in Intransitivities with (1976) problem RD preferences A McKelvey (2000) EMS nonseparable Niou D, of Sociol Lacy theory Math J A spaces. (2001) choice D multidimensional Lacy over voting Sophisticated 13:47Ð54 Choice (1972) Public question. GH the of Kramer division York 39:41Ð57 On New J (1972) Longman, Econ J America. competition. Kadane polarized in a Stability of (1929) myth H The Hotelling war? Culture (2005) University Cambridge MP introduction. Fiorina an voting: of York theory New spatial Row, The and (1984) Harper MJ democracy. Am Hinich of JM, rule. theory Enelow majority economic An and (1957) orderings A preference Downs Social (1970) PC Ordeshook MJ, Econo- Hinich rule. OA, majority and Davis orderings preference Social (1972) MJ Hinich MH, DeGroot OA, internet: Davis the versus interviewing telephone RDD via surveys National (2009) JA Krosnick Hodge L, William Chang valuation. complementary 56(1):23Ð24 with Econ decisions Polit Committee J (1951) decision-making. 77(4):974Ð RA group Newing Rev of D, Sci rationale Black Polit the On Am (1948) activism. D party Black with American model and spatial Downsian polarization, A citizens, (1983) JH engaged Aldrich center: disappearing The (2010) A Abramowitz References 214 57(4):787Ð806 40(3):825Ð850 Sci 12:472Ð482 Theory Econ J control. agenda 45(2):239Ð258 2:165Ð180 York New Press, 64(2):426Ð448 Rev Sci Polit 40(1):147Ð157 metrica 73(4):641Ð678 Q Opin Public quality. response and representativeness sample comparing London Company, and 990 Haven New Press, University Yale democracy. hncniae aefie oiin na su pc,acniaecntk a take can candidate a space, issue an in positions fixed have candidates When to converge will candidates competing two model, spatial dimensional one the In .Lc n ...Niou E.M.S. and Lacy D. 1986 ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 cwrzT(97 olciecoc,sprto fise n oetaig mPltSiRev Sci Polit Am trading. vote and issues of separation 45(3):575Ð594 choice, Stud Collective Econ Rev (1977) games. T dynamic York New simple Schwartz Winston, of and Instability Rinehart, (1978) Holt, N people. Schofield semisovereign The 51(4):933Ð (1960) Rev EE Sci Schattschneider Polit Am scope. Haven New and Press, direction, University visibility, Yale Intensity, manipulation. (1957) political EE of Schattschneider art democracy The of (1986) theory WH the Riker between confrontation a populism: against Liberalism (1982) WH Riker Elections in Packaging Issue and Preferences Nonseparable 71:999Ð1010 942 Heights Prospect Press, Waveland choice. social of theory the and 215 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 htpriscnue h otoe n eortcoei the is one democratic methods selection and the open all most Of the candidates. use, of can nomination parties the be that in could party method the the of hand, pathizers one On categories. candidate-selection two a (or in speaking, fall Broadly can election. (CSM) given a method at office for compete later nev- display may to candidate forward talented coming hopefuls most unknown talents. the of their find possibility to the wish down this shut who In ertheless parties used. and rational democratized, always that is not candidate-selection show is which I under it conditions though widely the is world, explore com- option I the open paper This around an skills. parties join campaigning political outsiders true to their fresh available display let can to they process where would nomination petition option One the candi- party. democratize possible the to in best standing be the win previous candidate’s selecting to that and wishes of identifying regardless of that date, for over party question mechanisms advantage ambitious find A and an to themselves. such imagine elections have com- prove would to fair to We insiders a them outsiders. well-known the be lesser-known allow for than even parties election not well-known why primary might is less interest there a be and as might with; They such they competing ballot. petition ways: the are on two they name grandees least their party getting dominant at time the in harder to disadvantaged much belong a necessarily are have not party do the who nomina- in Outsiders party’s faction their election. secure future to a par- advantage for their an tion have within frequently position they comfortable particular, a In enjoy ties. to tend insiders other and Incumbents Introduction 1 Serra Gilles Valence Candidates’ About Information Partial of Aggregation Challenge? Primary a Avoid Incumbents Will When .Sra( Serra G. DOI (eds.), al. et Schofield N. Mexico 01210, City Mexico Fe, (CIDE), e-mail: Santa Teaching de and Lomas Research 3655, Economics México-Toluca for Carretera Center Science, Political of Department ned oiia at a s ait fmtost oiaetoewowill who those nominate to methods of variety a use can party political a Indeed, yalwn h atcpto faltemmes ciit n sym- and activists members, the all of participation the allowing by democratic) 10.1007/978-3-642-35239-3_11 [email protected] B ) dacsi oiia Economy Political in Advances pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , , rmr election primary open .By 217 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 li htpiayeetoshv rcia datg vreiecnee nom- elite-centered over concrete, advantage be practical To a parties. have their elections inations: within primary elections that primary claim of I use the allow casionally calculations strategic the in lies ex- world the the leaders claim around party I primaries of Hence, of CSM. use the undemocratic to equivalent for an is planation and this democratic effect, prac- In a in uncompetitive. between or election, choosing competitive primary hopefuls be a several can holding primary among “officially” that withhold race are tice will competitive parties they while a Thus, them, that place. satisfies such takes party that endorsements if candidate their hand, insider divide other an or the identify well- On challenger. not a any do behind overwhelm and leaders rally to advice strategic funding to endorsements, of decide public amounts leaders with to large her party whether provide choosing If will by they not. insider, so or known do candidate They will favored exclusive. nomination and a their closed recurrent United whether endorse or choosing the in inclusive, in In CSMs and say choice. open strong other that be a and have of also nature uncommon primaries elites strategic not party between the States, is forth indicates it and clearly Actually, back which not. elections, go or process to debat- elites nomination parties party the for see open repeatedly to we whether America ing Latin Throughout law. by mandated rvdsa ifrainrtoae o h xsec fpiayelections. model primary my of existence sense, the that for In election. rationale” general “information the an provides in ef- how be reveal would can pre-candidates primaries election Thus the forth. fective “trial” so and a campaign, a as manage advertisements, serve can primary between A sure. for within known never called is often effectiveness candidate, to party’s seeking individuals a the surrounding become the uncertainty about much information is have valuable there reveal to Indeed, campaigns contenders. expected appointment. primary be elite the an can because through happens election nominated This candidate primary a a than skills through campaigning nominated higher candidate a that is often most is candidates elections which seri- primary discussing have of even to adoption before parties Their adopt political for to select. common CSM to is what it on fact, deliberations In ous government. the by them on addts nms rsdnilsses oiia ate aelea nchoosing usually in is leeway it have and parties CSM, political their systems, presidential offered. most are it they In but candidates candidates. election; of the quality winning the of of prospects terms the its in paper, citizens of this for terms throughout the matters in argued also example, party As the meeting. For private for party. by a matters chosen the at choice be bosses of could party level election of gubernatorial handful elite or a the presidential upcoming at an decision for nominee members. closed-door party be a all could of of method vote nomination sisting candidates democratic the aspiring hand, the other in among the culminates competition On that organized party the same the to within refer I election, primary 218 hspprpsuae eett at edr hthlsepanwyte oc- they why explain helps that leaders party to benefit a postulates paper This at edr r o h otpr epnil o h a hi ate nominate parties their way the for responsible part most the for are leaders Party at htsae ayo h etrso h usqetgnrlelection general subsequent the of features the of many shares that party a h ate.Pecniae utpriiaei eae,bodattelevision broadcast debates, in participate must Pre-candidates parties. the hyrva nomto bu addts pelt voters to appeal candidates’ about information reveal they not h aeta rmre r xgnul imposed exogenously are primaries that case the pre-candidates closed hi uuevote-getting future Their . (or voluntary ,con- undemocratic), .Mypremise ahrthan rather .Serra G. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 n rmre r oeapaigt h at hti otdsdatgdgvnits given disadvantaged most is that party; party the of the policies. membership to and likely mass valence appealing more the more and are are elite primaries primaries the that and between finds congruence also is paper there this when particular, In corroborated. are primary on literature tell. the can in I new as are far results as without Both elections election parties. subsequent explain enough their a help inside good for opposed would re-nominated a be being result have to might This able are altogether. insider incumbents election an many why primary that a find prevent I to candidates, belief reputation of prior that the skill view Regarding the cordial. oft-mentioned in and about an competition light held with stiff are primaries contrast from their benefit in ensure can stands should party parties result a This that election. find primary I its primaries, of quality the the ing call information, I valuable which reveal vote-getter, proficient to primaries the call of I ability which The precisely. more studied be hoeia ieaueta eae om oe.Section model. my to relates that literature theoretical yCnoct( Condorcet by ( ntdt opt gis nte at.I hssne h oe al ntetradi- the in falls model to the process a sense, nom- as this voting were In modeling they party. of “noisy if another tion performance as against their concrete, interpreted of compete be are forecasts To to as party full. inated per- interpreted in the to be known within ability can be that her performances not signals” or contenders’ still his the would but assume election contender, general I a the assess in to well needed form information the of of revelation part the analyzing is than research rather that to respect with outsider tion and insider of quality the primary and of election, ideology primary the candidates. the parties, of losing of intensity of ideology the voters, the cost depends answer variables: the the fundamental indicate, outweigh will several results primaries on the As of platforms? benefit candidates’ the informational of control the question main does The When nomination. elite-centered is: an and election primary competitive a leaders’ party the that reveal paper trivial this and not in costs is results the decision The between party’s election. trade-off the primary a of face a selection leaders of the party benefits that with They is trusted ideology. point be their main to The share moderate candidate. quite too Indeed, or not elections preferences. extremist may too primary leaders’ voters be the primary cost: might that from oft-mentioned know far one bosses policies party on adopt the focus to I candidates push paper might this In leaders. party hnWl nubnsAodaPiayChallenge? Primary a Avoid Incumbents Will When 1998 h eto h ae sdvlpda olw:Sect. follows: as developed is paper the of rest The (2011) Serra in results previous the of many results, new these to addition In can variables new Two choice. modeling this with found are results new Several hsmdli otnaino h eerhi er ( Serra in research the of continuation a is model This between choice party’s a includes that model voting spatial a build I basis, that On to costs several carry might primaries above, mentioned as hand, other the On ). full 1785 nomto,b hc enta rmr lcin nyreveal only elections primary that mean I which by information, quality ,FdesnadPesendorfer and Feddersen (1996), Banks and Austen-Smith ), fpiais n h euaino h nie addt as candidate insider the of reputation the and primaries; of ro belief prior aggregate nomto— rdto initiated tradition information—a bu h nie’ kl.Regard- skill. insider’s the about 3 2011 2 nrdcsasailvot- spatial a introduces reysmaie the summarizes briefly .Temi contribu- main The ). partial information 219 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 eif bu r-addtsbsdo hi efrac ntepiaycampaigns. primary the in performance their Section on based pre-candidates about beliefs iainpsil nldn nr-at optto.CoadKn ( Kang and Cho competition. intra-party ( including competition al. possibly spending et nization a and Castanheira election, In primary population candidates. a the leader, between party and a mean by ( appointment party trary al. the et between Jackson divergence mean. of degrees different that ( different with Grofman in comparing and papers those Owen especially Several (CSM). one, methods causes. this candidate-selection their with aspects than common rather share stud- primaries, literature have of authors Most consequences elections. primary the on ied literature formal the to adds paper Elections The Primary of Adoption the of Theories Previous 2 elec- general the before party a inside place takes that Section nomination tion. a the is study Sect. It I In corresponding election. and valence. dimension general additional candidates’ the an the study with to to model, voting serve Downsian will the that of parties variant two between model ing 220 oifrainlapcso rmre.I aladadTrl (2002 Tirole and Caillaud In primaries. of aspects informational to elections. primary closed and open iin o at ohl opttv rmr lcin hc steproeof purpose The the theory. is democratic to which relevance election, of tions primary competitive Sect. Sect. a In Finally, hold paper. party. to this the party in a mass for the ditions and elite the between h lentv S.I at osnthl rmr,Sye n igas- pre-can- Ting willing and the all Snyder among primary, random a is at hold difference chosen not that be main does will assume A nominee also party nominee. the Ting a the that and If com- sume of Snyder CSM. parties valence valence. model, alternative expected and my the in the ideology As of increase not. terms primaries or in election primary both a pete hold parties’ to the decision is party’s paper my of focus the while that those candidates. platforms, paper of of do, of their choice spite I of choice In focus candidates’ as do. the the find, because parties differences is authors strong important have than The models more nominee. our primaries similarities, high-quality from a benefit identify parties weak to party a allow where papers contenders. of primary set of valence a the is about there information Then reveal primaries preferences. voters’ about (2005 information Meirowitz acquire In party. the of trustworthiness and (2010 al. et heira eut nti paper. this in results nte e fppr htrlt om oe,aetoeta aepi attention paid have that those are model, my to relate that papers of set Another nte lsl eae ae sSye n ig( Ting and Snyder is paper related closely Another ( Merrill and Adams instance, For 6 nrdcsacs faotn rmre ae ntelc fcongruence of lack the on based primaries adopting of cost a introduces 5 eeosasgaigmcaimfrpiayvtr oudt their update to voters primary for mechanism signaling a develops ,teueo rmre rvdsifrainaottecredibility the about information provides primaries of use the ), 2007 8 icse h anrslsadsget oeinterpreta- some suggests and results main the discusses td he ifrn oiainpoess narbi- an processes: nomination different three study ) 4 aease aki h lcoa process, electoral the in back step a take I , 2008 2010 otlt htpiayeetosmay elections primary that postulate ) Appendix ,prisslc hi nenlorga- internal their select parties ), ,piaisalwcniae to candidates allow primaries ), eieanme fcon- of number a derive I 7, otisaltepof fthe of proofs the all contains 2011 2006 h losuisa studies also who ) opr primaries compare ) 2008 n Castan- and ) compare ) .Serra G. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 ae eeosasgaigmcaimt eelprilrte hnfl informa- full than rather partial reveal to this however, mechanism papers, tion. signaling those Unlike a candidates. develops their of paper valence the revealing by ties n nvlneadteoeo rmre.A nAasadMril( Merrill and Adams in As primaries. on ( Ting one and the der and valence on one points ( ( Schofield inocr nalf-ih oiysetu.Idnt by denote I spectrum. policy left-right with a be in will occurs This tion cycle. nomination refer- their any that completed without election” already a “general have parties the parties two to all corresponds between after this competition occurs essence, In the elections. on primary to focus ence I section this In Parties Two the Between Election General 3 benefit voters of subset Interestingly, a office-seeking. only that are sense that the type parties in it. a particularistic from for is enjoy bonus valence candidates of a model, type different as this his serves have In that nomination. can valence party’s of parties their use obtain model to to my election bound in are whereas in parties law, candidate both state insider that an by CSMs. is choose CSM will difference same elite Another the party room. the smoke-filled assume a I contrast, In didates. Challenge? Primary a Avoid Incumbents Will When enn htte aeaotteplc mlmne fe h election the after implemented policy the about party care (Wittman are labeled they parties that election, that meaning assume this motivated, I in tradition, Wittman-Calvert-Roemer competing the parties Following two are There Parties 3.1 valence the affect to ( models Mesquita their de in and agents Ashworth the are allowed also parameter have papers other oe rsn eei napiaino h oegnrlmdldvlpdi Serra in developed model general more the of application ( an is here present I model X cie ndti eo.Tevlnedmnini eoe by denoted is dimension valence The below. detail in scribed 2008 2010 L aec-oiymodel valence-policy h oe nti ae soeo h e htcmie ohltrtrs the literatures, both combines that few the of one is paper this in model The ial,ti ae srltdt h ieaueon literature the to related is paper this Finally, ( Kselman = x ,Shfil ta.(2008 al. et Schofield ), ). X X ∈ R L R omlz h da on ftemda oe ntegnrlelection general the in voter median the of point ideal the normalize I . 1973 and 2007 eod hr sadmnincrepnigt aec,wihi de- is which valence, to corresponding dimension a is there Second, . eeosamdlweeaprnsms opt naprimary a in compete must aspirants where model a develops 2012) 2011 ;Calvert X ,Cril n atner ( Castanheira and Carrillo ), R epciey h w ate aedsic dooiss that so ideologies distinct have parties two The respectively. , ,adSra( Serra and ), enn hti ilhv w iesos is,teelec- the First, dimensions. two have will it that meaning , 1985 ). Roemer ; 2011 ,tepeiehr sta rmre eppar- help primaries that is here premise the ), 2001 .Parties ). ,Shfil n ee ( Sened and Schofield 2009), ,Cladr(2008 Callander 2008), noeosvalence endogenous L x and h oiyimplemented, policy the v R with , aeielpolicy ideal have L n party and ,Meirowitz ), v ,Sny- 2008), ∈ R Some . policy- + 2005 .The 221 R ), . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 ald“aec aaees n a egvnmn nepeain fra overview an (for ( interpretations Schofield many given see be can and parameters” “valence called au omlzdt eocrepnigt o-kle addt,adahg value high a and candidate, low-skilled a of to corresponding zero to normalized value candidate’s the as interpreted best da on.Cnrtl,party Concretely, point. ideal o-oiyatiue r ovtr nta lcin aaeessc as such Parameters election. that in voters to are attributes non-policy ilanuc h ltomdsge yhrprydrn h apin,adshe and campaigns, election. the the She wins her. during she nominated case party that in her platform party by such the implement designed of will platform will function candidate the utility the announce exact exact, the will be To having of party. if their salience of as lower preferences behave policy a the of adopt analysis will footnotes, they fuller in a results for these (2011) report Serra I to trivial. reader valence. the become refer results I the and and election the in L os taeial nodrt aiieterepce tlt.Icl hs platforms those call I utility. expected their maximize to x order in strategically so do ucin of functions | v l addtsaecaatrzdb parameter a by characterized are candidates All Candidates 3.2 party’s a of point ideal generic the as assume I generality of loss much without and zero, to 222 2 1 nltrscin ilseiytosprt ruswti party within groups separate two specify will I sections later In nei h lcin ehial ilasm httevlneo high-skilled that a meaning of parties, valence both of the differ- extremism that a the assume make than will larger to I strictly salient is technically sufficiently candidate is election; valence the that in assume ence will I cases, teresting da points ideal X hsi qiaett suigthat assuming to equivalent is This that note course, Of R L seteimwl emaue by measured be will extremism ’s R V nti oe,cniae onthv oiypeeecso hi w.Rather, own. their of preferences policy have not do candidates model, this In ial,prisfruaeplc ltom ocmeei h lcin n they and election, the in compete to platforms policy formulate parties Finally, and h klso addtsi parties in candidates of skills the |

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 silsrtdi Fig. in illustrated is e ihrsoei h aec ieso oeta opnae e xrms in extremism her compensates than more dimension. dimension policy valence the the in score higher her n nrae ihtecniaesvlne nesne h aec parameter valence the of essence, function In utility platform, valence. the candidate’s candidate’s that up” and the “shifts point with ideal increases her between and distance ( the Serra in with elaborated decreases As model. of have kind this candidates two. both the If between skill. equally highest randomize the will she has skill, candidate same whose the one the for vote will If assumptions. indifference following aigamr xrms ltom Candidate platform. extremist more a candidate having figure, this in depicted h nlss ilasm htteei einvtr hc call I which voter, median a is simplify there To that election. assume the after will implemented I policy analysis, the the about cares electorate The Electorate General The 3.3 over for advantage valence 1 Fig. Challenge? Primary a Avoid Incumbents Will When 1. following: the is election this of timing The Concept Solution and Timing 3.4 zero. to point v ideal her normalize I election. the in decisive are preferences ftewnigcniae h tlt ucinof function utility The candidate. winning the of ti ot okn oecoeya o h einvtrmkshrdcso in decision her makes voter median the how at closely more looking worth is It nadto oteplc implemented policy the to addition In M tr apinn.Tecniae’cmagigskills campaigning candidates’ The skills campaigning. start candidates’ the of Assessment L ilvt o h at hs addt aiie e tlt.Imk the make I utility. her maximizes candidate whose party the for vote will h feto a of effect The R hr ti sue that assumed is it where 1, U M R x v) (x, M ssrcl rfre ocandidate to preferred strictly is neapeo how of example An . M =−| x sidfeetbtentetopris she parties, two the between indifferent is ate nonetercniae who candidates their announce Parties : h lcoaeas ae bu h skill the about cares also electorate the , 2010), R x v |+ L sal ownteeeto because election the win to able is

B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 that noacut nSae3 necniae’skills, platform candidates’ candidate’s once other 3, the candi- Stage each taking In where platform account. 2 her into Stage promote in and made announce is must decision date first The attributes. valence their with call I ie ausof values given 1 Theorem al 1 Table be should variables important some section, this call of I defined. results main the stating Before Election General the of Results 3.5 2. 224 3 3. eipratt ealta P eursta l taeisfr ahequilibrium Nash a form strategies all subgame. that every requires in will SPE (NE) It a strategies. that pure recall in to (SPE) important be equilibrium and induction subgame-perfect backward is by concept solved solution be the must game The knowledge. elects common voter is median mation the assessed, and observed been v niae ntemi hoe nti eto.I utb eebrta aec was as valence that election, remember that be general enough must salient the It be section. of to this assumed on results theorem the main the determine in indicated will parameters These platform. econcrete, be 0 −V V of Value h roso l h eut oei the in come results the all of proofs The tg osntivleaydcso:tecniae r eeldt oes along voters, to revealed are candidates the decision: any involve not does 1 Stage hr r eea omnst aeaotTable about make to comments several are There addtsanuc hi platforms their announce Candidates platforms: policy the of Assessment h eea-lcinvote general-election The and = −V

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 da on fhrpry n tl i h lcinbsdo e ueirsil So skill. superior her on based election toward the biased win the is toward still voter implemented and median policy party, the the from her diverge of to able point is ideal skill highest with candidate the oee,i em fteplc mlmne ytecniaeatrapiay Solv- primary. a after candidate the party by implemented ing policy the of terms in however, π nweg mn oesadparties. and voters among knowledge called ain nuh h at ihtehgetsildcniaei bet ulplc all policy pull point. to ideal able its is to candidate highest-skilled way the the with party the enough, salient < v nomto bu ate n hi oeta addts codn ota informa- that that to some probabilities According on the candidates. based potential tion, skills their these and about parties about beliefs are information prior skills some still campaigning exist of is candidates’ there nomination candidate the However, party’s the of uncertain. each study values of to exact identity process the the election Consequently, stage, unknown. the this in At party. back a step within a candidates take I section, this In Process Nomination The 4 when However, point. ideal voter’s median the to completely converge ties x 2 Fig. Challenge? Primary a Avoid Incumbents Will When 4 over for advantage valence da on ilb ie by given be will point ideal oto oiy nohrwrs htw enb pry ilvr codn oteCSM. the to according vary will “party” by mean we what words, other In policy. control hoigaCMoe nte.Ide,cosn ohl rmr lcincould election primary a hold to choosing Indeed, another. affect over CSM a choosing h aewee0 where case the π hsielpitdpnso hc ru otosplc ihntepry nti eto ehave we section this In party. the within policy controls group which on depends point ideal This L ∗ R L and h eto hspprsest td h blt fparty of ability the study to seeks paper this of rest The and ≡ L X P(v ,adubae when unbiased and 0, x qiiru platforms Equilibrium π L R ∗ π R and R hnteei a is there when R R oiieyudrcrusacsseie eo.Teecudb cost, a be could There below. specified circumstances under positively epciey with respectively, scs-eetaayi stefia olo hsrsac.Iece in eschew I research. this of goal final the is analysis cost-benefit ’s = X R V) h eei da onso parties of points ideal generic the v nltrscin,hwvr party however, sections, later In . ,bae toward biased 0, X π R R L = oincrease to ≡ X hc depicts which 2, RM P(v ftemembers the if L = L v V) π when R = and 225 R by ’s 0

B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 hi xetduiiytkn noacuttebhvo ftervlpryi h general the in maximizes party rival that the strategy of behavior a election. the design account will into taking they utility rather, their expected and but impose their leadership candidate, passively not the the would on both they points that that ideal implies Note This RAF. strategically. the think by RAF by the desired followed strategies be policy to the party strategies follow instead, policy the If, formulate candidate. leaders its its then selection, ership n a nielplc point policy ideal an has and of elite The file”). and paper, separate a in done words, other being party that is assume analysis simply Such I here election. primary a choose to hyws hi addtst olwi ahcrusac.I party platforms If circumstance. policy each the in indicating follow of to charge candidates their in wish are they parties precisely, More election. vergence hro)between thereof) whotovotefor. on,with point, | a aeaynnngtv value: non-negative any take can Party Leaders Party Versus Members Party 4.1 party of decision parallel the paper this 226 agigt i h lcin Thus election. the win to paigning of skills campaigning actual lt is elite oiypeeecso at’ edrhpadisRF twl eueu omea- to voters. useful primary be its define will and I establishment It purpose, party’s RAF. a that its between With any, and if divergence, leadership the party’s sure a of preferences policy ilasm htteRFhsamda ebrwoepeeecsaedcsv in decisive are preferences call whose member I median election. a primary has the RAF the that assume will I X RM ate r lorsosbefrfruaigplc ltom ocmeei the in compete to platforms policy formulating for responsible also are Parties h akadfie(A)of (RAF) file and rank The ngnrl ewl have will we general, In R − ossso n“lt”(r“edrhp)ada“ebrhp o rn and “rank (or “membership” a and “leadership”) (or “elite” an of consists X d RE R X π niaealwreiems oguneisd h at.Nt that Note party. the inside congruence elite-mass lower a indicate | RM nitrsigitrrtto of interpretation interesting An . L stkna neoeosprmtr naycs,rmme htthe that remember case, any In parameter. exogenous an as taken is > R R .Teuiiyfnto of function utility The 0. seieadms ebrhp ihrlvl fteitra di- internal the of levels Higher membership. mass and elite ’s ilb eerdt as to referred be will RM d L X R U U and R RE RM L sthe as RE h einmme of member median the R X d sas oiymtvtd osmlf h analysis, the simplify To policy-motivated. also is a led hsnacniaeb n ehd In method. any by candidate a chosen already has RE = R (x) (x) R ssapiayeeto,te t addt will candidate its then election, primary a uses v ≥ with , scniae r eeldwe hysatcam- start they when revealed are candidates ’s L X =−| =−| nenldivergence internal RM 0. and L oteewl eatninbtenthe between tension a be will there so , RE v X X X h ih lob odrn whether pondering be also might who R RM RM RE RE hslaesi spolicy-motivated is leadership This . r ul nw hnvtr decide voters when known fully are − − d is > R x x .Teuiiyfnto of function utility The 0. | | sa the as is R n call I and nparty in congruence R R X where , RM ssalead- a uses e ideal her o lack (or .Serra G. d R R d ≡ ’s R Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 addt.If candidate. notie eoe by denoted outsider an hoefo,wihIcl the call I which from, choose oiisltrtr,Iwl call will I literature, politics h eetrt steprysRF ntefre case, former the In RAF. party’s the is selectorate the ate aeefciewy fefrigtercniae’promises. candidates’ their having because commit- enforcing credible for concrete of are ways making commitments exchange effective Those by have In nomination. policy the parties them. on if charge yield selecting in to those those have to they of ments ticket, wishes the the on names to their cater to forced are ssmaie nTable them. endorse- in either elite summarized nominated of an preferences is or party the election inherit their primary will a nominee inside the is ment, CSM group the whichever whether on of depending Therefore, agents perfect as behave h olo addtsta tcncos from. choose can it that candidates of pool the party party oa a as to This assumptions. ignored. following the or unknown opening the been motivates is previously observation it have empirical election, might primary who contrast, competitive aspirants elite In truly outside nominee. a to party natural allow door the to a as decides emerged party When previously a there has when nomination. because who usually party’s insider is trusted who it their a consultation, aspirants is further getting of without someone of number endorse to the chance chooses is realistic rules a nomination have across difference important An Outsiders Versus Insiders 4.3 the choose call to I file. decision and the rank and party’s run, the to to chance delegated a is has nominee candidate elec- outsider primary or competitive an nominate a where pri- directly hold tion to a could leadership it Alternatively, or the candidate. endorsement for insider an be elite endorse would an CSM default methods: The two election. exist mary There (CSM). party method of selection leadership the candidate, a selecting Before Endorsement Elite Versus Election Primary 4.2 2 Table Challenge? Primary a Avoid Incumbents Will When X r hoe with choose, ers R h nepeaini hti tiigt i h oiain h pre-candidates the nomination, the win to striving in that is interpretation The addtsaotteplc rfrne fterslcoae nohrwrs they words, other In selectorate. their of preferences policy the adopt Candidates n niiulwoi fcal otsigteprysnmnto ilb referred be will nomination party’s the contesting officially is who individual Any = R R X scandidate ’s pre-candidate.If h betv of objective The a w r-addtst hoefo,wihcnito h insider, the of consist which from, choose to pre-candidates two has RM . m R = m R elite ∈{ 2. RO h eetrt stepryslaesi.If leadership. party’s the is selectorate the , elite m ec,b dpigapiay h at sexpanding is party the primary, a adopting by Hence, . fe neieslcin max selection: elite an After fe rmr lcin max election: primary a After R , selectorate primary = insider elite } hnparty then olwn tnadlnug nteparty- the in language standard Following . n eoeby denote I and h ru ncag fslcigaparty’s a selecting of charge in group the R R a nyoepecniaeto pre-candidate one only has X m RM ed ocos candidate- a choose to needs R R RI = h ehdthat method the or .If X x x R R RE RE m U U RM RE epciey This respectively. , ntelte case, latter the In . R (x) (x) = m R primary =−| =−| = R X primary X RI slead- ’s RE RM and , − then − 227 x x | | , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 π usdrcandidate outsider has e nihsaotteaoto fpiayeetos npriua h possibility the particular primaries. in such prevent to elections, might leads primary insider assumption high-skilled of realistic a adoption this that show, the will about I insights As new model election. this general in the contrast, in In formation election. only general is the information in the information additional no is is information there models those In didates. ainwt epc oAasadMril(2008 inno- Merrill main and the by Adams is processed to mechanism and respect informational candidates This with by rule. vation sent Bayes in- signals some using partial noisy acquire voters reveal of to primary Primaries system means aspirants. a a of as through elections skills formation primary campaigning of the model about a information develop I section, this In Elections Primary of Benefit The 5 2. 1. following: the is nomination the of timing The Timing 4.4 228 ( of stages three the knowledge. by common followed is are information nomination this All the election. of general stages the three the i.e. section, 3. v probability fc,fo ateetos rfo ol.Acrigt htinformation, that to According polls. in from performance or previous elections, from past come from could candidate, office, information insider That its low-skilled. that or probability skilled the has about party or information The ex-ante. 0 prior uncertain values, some are skills two campaigning take pre-candidates’ the can of skill values candidate’s a before, tioned 2011 R RO call I fe hsnmnto,tegm spae xcl sdsrbdi h previous the in described as exactly played is game the nomination, this After klsi eeld fteCMi neieedreet h r-addtscommit pre-candidates of the interests endorsement, policy elite the an pursuing is to CSM the If revealed. of is interests skills policy the pursuing to commit contest nomination method The process. candidate-selection nomination a the choose of selection The h oiaindecision:Party nomination The h apinn kl ftecniaewoi nlynmntdby nominated finally is who candidate the of skill campaigning the no = n te oespsuaigta rmre eelifrainaotcan- about information reveal primaries that postulating models other and ) 2 1 ro nomto bu h usdrcniae h at eivsta the that believes party The candidate. outsider the about information prior . v RI π RI and fbighg-kle,with high-skilled, being of v RO RO h apinn klsof skills campaigning the partially a rbblt foehl fbighg-kle,hence high-skilled, being of one-half of probability a has fteCMi rmr lcin h pre-candidates the election, primary a is CSM the If : eeldi h rmr,adteei diinlin- additional is there and primary, the in revealed R RE eet t candidate. its selects fully n oifraini revealed. is information no and π RI RM eeldi h rmr lcin and election, primary the in revealed ∈ RI ( ,Sra(2011 Serra ), 0, n oeifrainaottheir about information some and and 1) nteohrhn,teparty the hand, other the On . RO h edr fparty of leaders The : epciey n call I and respectively, V oee,teexact the However, . ,Sye n Ting and Snyder ), R RI sImen- I As . shigh- is , RI .Serra G. a a has R Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 nomto s“os. ec sueta h resil fcandidates of the skills because true however, the imperfect, that assume are I forecasts Hence “noisy.” Those is party. information low other or the against skill elec- paign high primary primary-election reflect the a in a itself is candidate as a can there campaign of performance primary if the interpret precise, the members more the Party in skill. be helps performance To candidate’s subsequently one. a information skilled tion, This most pre-candidates. the choose their party of skills actual the beliefs. prior the for except uncertain ex-ante are which candidates, eiigwehrt odapiayeeto rsikt neieselection. elite an to stick when or leaders election party primary by a out hold carried to analysis whether cost-benefit deciding the paper, the of section final iaetesle hywuddrcl nominate directly would they themselves didate kle candidate, skilled between decide will who RO aigahg-kle addt.Tevleta at edr r ekn omaximize maximizes to seeking are leaders party is that value The candidate. high-skilled a de- nating for rationale subsequent information the an for provides party. ground paper political testing this a sense a mocratizing that as In team serve in- election. a primaries give general manage and So advertisements funds, journalists. political raise to design they terviews candidates, how other primary debate on the supporters, tested during of are Indeed, pre-candidates resources. at- infor- and party’s the valuable assets campaigns the reveal pre-candidates’ to can the about come primaries about information Specifically, not mation useful revealed. would (2) is And but pre-candidates nomination. voters those elite out- to inside-track Those appeal an campaign. pre- through large primary as tention a the register have during can skills might who primaries their in siders contenders Concretely, display this non-mainstream expanded. to achieve or is hoping Primaries untested nomi- nominees candidates to bonus”. their potential door skill of of the “primary pool skill open the The campaigning increase (1) expected ways. that the two call party increase For will elections. to I primary is nee. adopt benefit to the incentive informational leaders, the formalize I Here Information Reveal to Mechanism a as Primaries 5.1 Challenge? Primary a Avoid Incumbents Will When r eeldonly revealed are j ihsildcniaewudsml be simply would candidate high-skilled a = π h rms nti ae sta rmre ilrva oeifrainabout information some reveal will primaries that is paper this in premise The ocalculate To ae eto ecie oto rmre.Ti ilalwsuyn,i the in studying, allow will This primaries. of cost a describes section later A f oee,te hoet odacmeiiepiayeeto,tecandidate the election, primary competitive a hold to choose they however, If, ob ocee eoeby denote I concrete, be To nomi- of probabilities different have will method each differences, these Given R ol onterc n h oiainwl edlgtdt h at’ RAF party’s the to delegated be will nomination the and race the join would RI ≡ , RO P(v a that say I . P(v R = forecast R P(v V) = partially P(v od o hycluaewihcniaeslcinmethod candidate-selection which calculate they so, do To . R V = | s R m j fhwwl h ol efr ntegnrleeto cam- general-election the in perform would she well how of V = R = RI ) | fteei rmr election. primary a is there if V with , elite high | and primary s j ) if h efrac fcandidate of performance the oeta fprylaescos oslc h can- the select to choose leaders party if that note m RO j R spromnesoe ihsil and skill, high showed performance ’s ec h rbblt fnmntn high- a nominating of probability the Hence . ∈{ ) ol eedo h culsil fthese of skills actual the on depend would , primary π RI Hence . , RI elite h rbblt fnominating of probability The . P(v } . R = j V ntepiay with primary, the in | elite) = v RI π s RI and j . = v low m 229 RO R Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 ueo h fetvns fpiayeetosa nifrainrvlto method. information-revelation an as elections primary of effectiveness the of sure kl,adw a interpret can we and skill, aei h eea lcin with election, general the in have probability a has primary hs lmnsalwsuyn h eairo rmr oes When voters. primary of behavior the studying allow elements These Beliefs Their Update Voters Primary 5.2 same the observe members RAF the all and particular, In knowledge. common are if 230 ihsildwslret fbt aetesm ro,sewl admz equally. randomize will she prior, between same the indifferent have is both member If largest. party was a high-skilled When who candidate skill. the highest for the vote RI will have RAF to the believed election, general is the winning in interest its itiue admvralswoedsrbto eedon depend distribution whose variables random distributed oigbsdo pae eif olwn os inlhsisrosi Condorcet in roots its has signal noisy a following ( beliefs updated on based voting about beliefs prior their fparty of ability otocntnswoevle are values whose constants two to oigway: lowing em 1 Lemma si un u,terbhvo ildpn rcal nterpirble bu the about belief prior their on crucially depend will valence, behavior candidate’s their insider out, turns it As party of member a per- that that her say meaning if I primary, irrelevant, hand, signals”. the is the other during primary sent the the signals On in meaning, the campaign. formance highest, on primary was depend signal the not whose does in strategy pre-candidate best the was for performance vote whose to is strategy party her of if member a that say 1785 j h addts efracsaepbi,adteeoetevle of values the therefore and public, are performances candidates’ The nsm h r-addts performances, pre-candidates’ the sum, In neteprymmesosretecniae’promne,te a update can they performances, candidates’ the observe members party the Once hs ocpscnb sdt ecieteRFsbhvo uigaprimary. a during behavior RAF’s the describe to used be can concepts These and spromnesoe o kl.Iasm htacniaespromnei the in performance candidate’s a that assume I skill. low showed performance ’s s RO ,Ase-mt n ak (1996 Banks and Austen-Smith ), q n ec hyudt hi eif ae ntesm nomto.Given information. same the on based beliefs their update they hence and , RO R f“en orc” ecninterpret can We correct”. “being of will suesewl oefrteoewoepirpoaiiyo being of probability prior whose one the for vote will she assume I , napiayelection primary a In P P s s j j = = high high RI q q | | sand ’s π v v R steqaiyo hssga.Mr broadly, More signal. this of quality the as facrtl oeatn h efrac h would she performance the forecasting accurately of RI j j srn n l il“oeacrigt h signals” the to according “vote will file and rank ’s = = nalterslsblw h symbols the below, results the all In . , q 0 1 o ahvleof value each for π j RO ∈ = = = ≡ ( ssil sn ae ue hsapoc to approach This rule. Bayes using skills ’s 2 1 P P RI ,adFdesnadPsnofr( Pesendorfer and Feddersen and ), 1 , −2 (1 1) , s s −q) q RO j j nohrwords, other In . +2 = = 2 q s low low j 2 s and saniysga fcandidate of signal noisy a as RI | | v v π j j RI and π = = , ≡ h akadfiemembers rank-and-file the 1 0 s 1 RO −2 = = v s RI RI r independently- are , q q 1 q +2 2 − and and q q 2 . s R v RO π RO il“ignore will s s and q RI aeprob- have RI ntefol- the in samea- a is 1998 = and π .Serra G. s RO refer ). s j s RO ,I RI ’s Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

690 689 688 687 686 685 684 683 682 681 680 679 678 677 676 675 674 673 672 671 670 669 668 667 666 665 664 663 662 661 EDITOR’S660 659 658 657 656 655 654 653 652 651 PROOF650 649 648 647 646 645 ifrnei ie by given is difference kle addt htapiaybig bv neieslcin ali the it call I selection. elite an above brings primary bonus a that candidate skilled • • • • Challenge? Primary a Avoid Incumbents Will When 5 at ebrwl hoet irgr h addts efracs nparticular, In of performances. values Bayesian candidates’ a enough the even high the disregard that for to combine compelling choose per- so will will their be member from she might coming party beliefs decision, information prior new her the the However, make with formance. candidates to the available about information beliefs of prior all member each use that to given and are: beliefs prior the ential • h xetdsilo h at’ oie.HnetevleIa okn ofidis find to looking am I value the Hence increase nominee. between to party’s difference is the the election primary of a skill using of expected bonus the the below, derive I As selection. ship the for vote insider contest. will the Hence and best. account, was campaigns into candidate primary the signals in the performance whose take candidate will voters primary direction: rmr oeswl ipynttutta uhpromne ilcrytruhto through carry will performances candidate such insider the that Hence skill. election. trust high general outsider’s not the the simply with will compared voters skill low Primary insider’s the of indications strong E(v aknw ahrta iigteeatvleof value exact the giving than Rather now. task uigtepiayeeto ilb sls:piayvtr ih oeacrigto revealed information. according information new vote the any might ignoring voters that completely primary while possibility useless: information the be preexisting will opens election it primary the as during significant is result This outsider the with test esol epi idta,ee huhteata au of value actual the though even that, mind in keep should We s if if if if s if n Oif RO and h motn au is value important The hr r eea oeotyfaue fti eut h rtoebighwinflu- how being one first the result, this of features noteworthy several are There u etts st uniytebnfi fhligapiayisedo leader- a of instead primary a holding of benefit the quantify to is task next Our nteohrhn,terslsfritreit ausof values intermediate for results the hand, other the On R RI RI π π π ) π π = = RI RI RI scontinuous. is RI RI fapiay Studying primary. a of s s ∈ ∈[ = ∈ ∈ RO RO I ( (π 2 1 ( π, 0,π ilide evleal obigbae yteoutsider the by beaten being to vulnerable be indeed will 2 1 , s RI , oeacrigt h inl if signals the to according vote , 1) π) 2 1 E(v ] = , , ) , , goetesgasadawy oefrRI for vote always and signals the ignore goetesgasadawy oefrRO for vote always and signals the ignore with s oeacrigt h inl if signals the to according vote oeacrigt h inl if signals the to according vote R RO | primary RO S E(v ≡ ewl enmntdrgrls fterperformances. their of regardless nominated be will he ; S π hc ersnsteetapoaiiyo aigahigh- a having of probability extra the represents which , R P(v RI | primary ) S h A will RAF the − R o ag ti n o tcags stemain the is changes, it how and is it large how , = E(v V ) R | primary) | and elite) E(v s always = RI R V − RI = S | elite hc oe nthe in comes which , · P(v s simn gis noe con- open an against immune is S RO s s oefor vote RI RI v ) , R R R . 5 = n admz ewe RI between randomize and = = sdsre,teepce value expected the discreet, is ti ayt e htsuch that see to easy is It sasmdt erational be to assumed is s . s V RO RO π | RI RI elite , , n oefrR if RO for vote and n oefrR if RI for vote and oi h expected the in go vni treceives it if even ) O Appendix na open an in skill 231 , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

736 735 734 733 732 731 730 729 728 727 726 725 724 723 722 721 720 719 718 717 716 715 714 713 712 711 710 709 708 707 EDITOR’S706 705 704 703 702 701 700 699 698 697 PROOF696 695 694 693 692 691 ae noabte oie ntrso aec.Hliga nenlcnetwill contest internal amount the an in Holding candidate valence. high-skilled of a nominating terms of in probability nominee the increase better a into lated htasalo ag eet nwrta usini h etsubsection. next the in question that answer I benefit? large a or small a that P(v kl.We h nie addt swa,maigthat meaning weak, is candidate insider the When skill. where iha usdrcniaede o nraeteepce kl ftenmneat nominee the of skill expected for the that is increase reason not The does all. candidate outsider an with for hoe 2 Theorem party ac ntepiay hsrsl snwwt epc oteppr bu primaries about papers the to respect of. with perfor- aware new candidate’s am is outsider I result the that This of primary. regardless the her in for mance vote will they that compelling ei yetbihn hte rmre aeabnfi opryleaders. 2 party Lemma to benefit a have primaries whether establishing by begin I Appealing? More Primaries Makes What 5.3 by start I paper. the of rest of the terms buttress in will considerations that previous properties the key the rephrasing on focus will I 232 and 6 rmr lcinmgtntbigaybnfi htovr hnteisdrcandi- insider the When that whatsoever. meaning benefit strong, any is bring date not might election primary a h rmr odcd hte h smr ovnigta h at nie.This ( insider. Serra party in the findings previous than with convincing in conforms more performance it candidate’s as is expected outsider she was the whether result at decide look to serious primary a the take will members party ftenmneb titypstv mut h esni htfor that is reason The amount. skill positive excepted strictly the a increases candidate by outsider nominee an the with of compete to her forcing old, oiainprocess nomination smnindbfr,tesymbols the before, mentioned As hsdmntae o h nomto eeldi rmr apin strans- is campaigns primary in revealed information the how demonstrates This rmre hrfr obigabnfi o ml nuhpir bu h insider’s the about priors enough small for benefit a bring do therefore Primaries h upiigrsl oe rmhg rosaotteisdrssil nsc case such in skill: insider’s the about priors high from comes result surprising The π π R RI ≡ R = S ∈[ 1 sseigt maximize. to seeking is V scle h rmr kl ou n sgvnby given is and bonus skill primary the called is −2 π, | q q elite h rmr kl bonus skill primary The +2 2 h rbblt that probability The 1) π q 2 R ) . . . ≡ P(v , m R R , sgvnby given is = π RI π V RI | π ∈[ m and saoeacrantrsod ocn e ocompete to her forcing threshold, certain a above is R π, R ) π = snmnewl ehigh-skilled, be will nominee ’s 1) S ee otocntnswoevle are values whose constants two to refer  ssrcl oiiefor positive strictly is at ebr n h nie addt so candidate insider the find members party π π RI RI + if S m R if = m π R R elite π hc stevral that variable the is which , RI = S ≡ sblwacranthresh- certain a below is primary P(v π RI R ∈ = 2011). ( 0, π V R π π) π | primary) , RI 6 ≡ given n zero and ∈ 1 .Serra G. −2 ( ( 1 −q) 0, q S +2 R .Is π) 2 q − ’s 2 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

782 781 780 779 778 777 776 775 774 773 772 771 770 769 768 767 766 765 764 763 762 761 760 759 758 757 756 755 754 753 EDITOR’S752 751 750 749 748 747 746 745 744 743 PROOF742 741 740 739 738 737 esapaigteisdrcniaei obgnwt.Ti sceryse nFig. in seen clearly is This with. begin to is candidate insider the appealing less em 4 Lemma shows, circumstances. result certain following under correct the only as would is Surprisingly, technology intuition attractive. this primaries’ more the primaries in those improvement make any that suggest would on Intuition larger debates a like better essence, challenges a paid In more is media on. included the they so because because and or or television longer, them, could improvement to became attention This campaigns more election. primary general the the because in occur performances future of forecasts erae with decreases oprtv ttc ihrsetto respect with statics comparative performances h ro bu h nie addt’ skill, candidate’s insider the about prior the em 3 Lemma xetdt ehge,i eoe eslkl htapiaywl mrv pnit. insider labeled the upon threshold once improve certain zero will a reaches exceeds primary advantage voters a electoral to appeal this that candidate’s before, likely is mentioned an candidate less as insider through becomes fact, the nominated it In of be higher, skill would be the that As to candidate. candidate expected insider the the of namely skill selection, the elite upon improve to is bonus 3 Fig. Challenge? Primary a Avoid Incumbents Will When n constant and hc eit h au of value the depicts which titypstv for positive strictly xc,a nraein increase an exact, q high-skilled being of probability insider’s sthe as h esg sta h lcoa datg ruh ypiaisi agrthe larger is primaries by brought advantage electoral the that is message The eea nihsabout insights Several ti o unt td how study to turn now is It a unnwt tdighow studying to now turn can I S h rmr skill primary The safnto fthe of function a as quality π forecast h rmr kl bonus skill primary The h feto h rmr kl bonus skill primary the on effect The RI ( q qa ozero) to equal ote aepiaismr rls trcie rtdsrb the describe first I attractive? less or more primaries make they Do . π fpiayeetosa nifrainrvlto ehd obe To method. information-revelation an as elections primary of RI π hsmksitiiesne eas h eeto primaries of benefit the because sense, intuitive makes This . ftecniaescmagigaiiyi h eea election. general the in ability campaigning candidate’s the of RI q ∈[ mrvsteacrc fteperformances the of accuracy the improves S π S , a oefo h em bv,ms oal htit that notably most above, lemma the from come can oayices in increase any to safunction a as π S ] hne ihacag nistomi determinants: main two its in change a with changes , u snl for null is but S π S hne with changes RI ssrcl eraigwith decreasing strictly is . π q π RI mle httepiayperformance primary the that implies RI . π n h cuayo h candidates’ the of accuracy the and , ( 0,π RI q S eebrta ecninterpret can we that Remember . for ) famria nraein increase marginal a of and π RI ( ∈[ π, π, 1) π π RI . 1) . . for s RI π RI and ∈ ( s 0, RO q π) 233 as 3, is ,

B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

828 827 826 825 824 823 822 821 820 819 818 817 816 815 814 813 812 811 810 809 808 807 806 805 804 803 802 801 800 799 EDITOR’S798 797 796 795 794 793 792 791 790 789 PROOF788 787 786 785 784 783 ed ocryotacs-eetaayi hncosn hte ohl primary a hold to whether is choosing leadership when cost party not. analysis That or the cost-benefit election consequence, adopt. a a to out As carry induced cost section. to are following a needs the candidates carry in I that might detail policy skill. in Primaries campaigning the described bonus. high of skill a terms primary with in the however, candidate probability a extra nominating that of called probability larger of a skill is Table campaigning in the summarized qualitatively about is beliefs finding prior This ac- candidate. the benefits insider those on the whether crucially out, depends turns campaigns it occur primary As the tually candidates. during both revealed between information discriminate new to using to believed (2) are and prospect whose superior; candidate be outsider an with candidate insider the place marginally by primaries of quality the Improving increasing occur. may that events campaign id,ete onominate to either minds, addt’ kl.Fritreit auso h prior the of values intermediate For skill. candidate’s htde o utf iighracac ntegnrleeto.I u,frex- for sum, In election. general the in of chance values fluke low a a extremely her or just giving high was tremely performance justify her not per- believe her does will if that voters even be to Primary candidate expected better. outsider is the was candidate to insider formance primary lose the the when will hand, in she other unqualified, performance the overwhelmingly the her On in luck. that performance bad trust to her will due if com- was voters even overwhelmingly Primary nominated be be appalling. will to is she expected primary election, is general candidate the insider in the petent When benefit. its on nraetepoaiiyo aigtergtnmnto hie uha increasing an Such Fig. choice. will in nomination revealed depicted right information can- is the the effect the making of nominate of quality to probability higher the campaigns A increase primary performance. the the better join to a will attention with that didate close outsider unknown pay the will to They compared race. candidate insider the of merits increase indeed will signals of quality the increasing 4 Fig. 234 bonus osmaieti eto,tebnfi,we hr soe fpiayelections primary of one, is there when benefit, the section, this summarize To re- to voters primary allowing (1) benefits: potential two have primaries sum, In h eutge nteepce ieto o oeaepir bu h insider the about priors moderate for direction expected the in goes result The oee,frohrpir,teqaiyo rmr lcin ilba oimpact no bear will elections primary a of quality the priors, other for However, S q h fetof effect The ntepiayskill primary the on q ilhv oefc nti decision. this on effect no have will S h esni htpiayvtr r nueaotterelative the about unsure are voters primary that is reason The . 4. RI o ueo onominate to or sure for π RI rmr oesqikymk ptheir up make quickly voters primary , RO π RI o ue eadeso any of regardless sure, for agnlicessin increases marginal , 3 . .Serra G. q

B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

874 873 872 871 870 869 868 867 866 865 864 863 862 861 860 859 858 857 856 855 854 853 852 851 850 849 848 847 846 845 EDITOR’S844 843 842 841 840 839 838 837 836 835 PROOF834 833 832 831 830 829 c omtet o n hsi ol eteRFspeeecsrte hnthe than rather preferences RAF’s the be would determine would it which thus preferences leadership’s and to; commitments icy be would it second, And ayeeto;adete a h ilwnteeeto.If election. the win will she way either and election; mary addt aetesm kl,te ilbt nonetepafr n they for and expressions following 0 the platform to the lead announce considerations These both election. EU will the they in skill, tie will same the have candidate aymkslsn eslkl u ae inn esattractive. less winning makes but likely less losing makes mary ie rmr lcinhstodfeecswt epc oa lt endorsement: elite pre- an be to To respect adopt. with to induced differences that are might probability two the candidates primaries has first, that However, election policy nominee. primary the their a of of cise, terms elec- skill in primary competitive cost expected a a the adopting carry of increase leaders to party is to benefit tion the saw, just we As Elections Primary of Cost The 6 3 Table Challenge? Primary a Avoid Incumbents Will When yparty by 4 Table neieapiteto h ilannounce will she or appointment elite an h value the Theorem from rived xetduiiyof utility expected oneteplatform the nounce kl datg,sewl noneteplatform the announce will she advantage, skill aedcess(u oteitra divergence internal the the to to (due (due decreases candidate pri- date high-skilled a a using bonus benefit, nominating skill a of primary As probability endorsement. the elite an increases over mary election primary a choosing in ycosn ihrapiayeeto ra lt eeto.Icall I selection. elite an or election primary a either choosing by High Intermediate Low high-skilled, is that Expectation ygacn tTable at glancing By h olnwi ofidepesosfrteepce tlt of utility expected the for expressions find to is now goal The RE (m R h w oeta eet fapiayeeto safnto of function a as election primary a of benefits potential two The h rd-f faced trade-off The R selite ’s ) v . π RI RI ≡ v R R S − .A ot h aoffo aigtehgetsildcandi- skilled highest the having from payoff the cost, a As ). slaesi rmadopting from leadership ’s hc ie h ucmso h lcindpnigon depending election the of outcomes the gives which 1, X e,poal e,tknit con Low account into taken Yes, probably Yes, o ee o nomto goe Zero ignored information No, never No, with Replacing primaries of Benefit e,frsr o nomto goe High ignored information No, sure for Yes, v L ecnraiysetetaeofthat trade-off the see readily can we 4, R L RM .If snmnei ihsildicessfrom increases high-skilled is nominee ’s n h ilwnteeeto.If election. the win will she and rmr election Primary lt selection Elite RO n not and L RI scniaehsasilavnae h ilan- will she advantage, skill a has candidate ’s RE eelddrn h primary the during revealed information the Using that X R that Probability π π R RI RI RM X isteelection the wins scniaewudhv aepol- made have would candidate ’s + R RM fsewsnmntdb pri- a by nominated was she if splc platform. policy ’s S X m − RE R X sisCM tcnb de- be can It CSM. its as RE fsewsnmntdby nominated was she if .Ptdifferently, Put ). L scniaeand candidate ’s R π R RI scniaehsa has candidate ’s slaesi faces leadership ’s 0 −| R tlt of Utility EU π R isteelection the wins X RI sleadership ’s RE primary a of bonus Skill RE to − (m π RE X RI RM R apri- if ) + | R S 235 the S ’s . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

920 919 918 917 916 915 914 913 912 911 910 909 908 907 906 905 904 903 902 901 900 899 898 897 896 895 894 893 892 891 EDITOR’S890 889 888 887 886 885 884 883 882 881 PROOF880 879 878 877 876 875 d hoe 3 Theorem in.Yttescn esn(htpiayvtr ih etomdrt)i equally is moderate) too elec- be primary might about voters comments primary scholarly (that reason in second another the or Yet way tions. some in found quently h ihs xetduiiy n fi sidfeet sueta twl choose will if it only and that if assume adopted I be indifferent, will is primary it A if selection. elite and utility, an expected highest the une fcosn n S vrteother. the over CSM one choosing of quences in hscudhpe o w esn.O n ad h A ol es a on far so be could RAF the hand, that one leadership On the of reasons. right two the for happen could This tion. ad h A ol es a ntelf ftelaesi that leadership the of left the on far so too be are could voters RAF primary the the hand, because election primary a adopt with EU h edrhpi party in leadership The CSM a of Selection Optimal The 7 5 Lemma 236 EU htcs h edrhpwl o dp rmr lcinbcuetepiayvoters primary the because election primary a too adopt are not will leadership the case that R ls opiayvtr edt et h at lt?I eed nacranthresh- certain a on depends It elite? party the inter- to words, be other to old, In How need own. congruence. voters its elite-mass primary enough to do from enough close ensue close only is will democratization ideology party RAF’s nal the if only and if EU R r oicnretthat incongruent so are si un u,tefis esn(htpiayvtr ih etoetee sfre- is extreme) too be might voters primary (that reason first the out, turns it As re ihteersls h edrhpi party in leadership the results, these with Armed h nuto eidti euti that is result this behind intuition The ≡| RE RE T RE EU T (m (m nrdcdi h hoe.I h rfrne fteeieadtems fparty of mass the and elite the of preferences the If theorem. the in introduced , X (m ≡ centrist. RE RM R R R S (m = = [X h xetduiiyof utility expected The = − (1 h edrhpo party of leadership The RE R elite primary primary X −π (1 = RE −π L ) elite )(π | . L and )−X RI ) ) ) +S) htcniinlast h olwn eut ealn that recalling result, following the to leads condition That . =− =− EU L π T L −| − − − R RE ] ≤ . (X (X (X (X (X (m X ilcos h pia rule optimal the choose will d X RE R RE RE RE RE RE RE R R hntelaesi ilntaotapiayelec- primary a adopt not will leadership the then − = − − − − − + slaesi o ahvleof value each for leadership ’s R X 0) X X 0) X primary d T R RM ilaotapiayeeto fadol if only and if election primary a adopt will L RE L R ≤ π π )π )π

966 965 964 963 962 961 960 959 958 957 956 955 954 953 952 951 950 949 948 947 946 945 944 943 942 941 940 939 938 937 EDITOR’S936 935 934 933 932 931 930 929 928 927 PROOF926 925 924 923 922 921 hoe 4 attractive. Theorem most are primaries where situation the to corresponds which date, a a hrfr eitrrtda the as interpreted be therefore can val edt iietocss h rtcs is case first The cases. two divide to need I π of value the on depend crucially will results is the primary out, a turns adopting it of likelihood the more by theorem, elections primary given previous of the adoption the to makes According what likely. on insight gain to like would We Statics Comparative 7.1 primary competitive a hold will of party values the high For voters”) election. primary “partisan label I (which can intuition same The literature. Fig. existing from the obtained in be mentioned seldom is but intuitive X voter, primary median the point of ideal the of function a as method candidate-selection 5 Fig. Challenge? Primary a Avoid Incumbents Will When ealthat Recall oes)teprywl nos nisdrcniae o nemdaevle of values intermediate For candidate. insider an endorse will party the voters”) nemdaepir,adtescn aeis case second the and priors, intermediate ainhpwt h da on ftemda rmr voter. primary median re- the non-monotonic of a point has ideal CSM the the with Consequently, lationship candidate. insider an endorse will π X ewe h at’ edrhpadtepiayvoters primary the and leadership party’s the between n A ilb edt rmr.Te a fprsn h rvostheorem previous the phrasing of way a Then primary. a to that lead is be will RAF and larger ayeetosae h interval The are. elections mary 2. 1. is: RI RM RM ≡ tr ihlwaditreit ro eif bu h kl fteisdrcandi- insider the of skill the about beliefs prior intermediate and low with start I rmterslsaoei scerta h threshold the that clear is it above results the From tityicesn with increasing Strictly positive Strictly ∈ 1 hudtk o h oiaint edlgtdt at ebr.Sc inter- Such members. party to delegated be to nomination the for take should −2 ( T The h ieiodo pnn h S erae ihteitra divergence internal the with decreases CSM the opening of likelihood the 0, q q ti oe“iey htteitra iegnebetween divergence internal the that “likely” more is it +2 2 T π) ec,Isuyhow study I Hence, . q π . ups h nta xetto htR shigh-skilled, is RI that expectation initial the Suppose 2 hntethreshold the Then . and π ee otocntnswoevle are values whose constants two to refer 5 o o ausof values low For . S X RM (X wihIlbl“xrms rmr oes)teparty the voters”) primary “extremist label I (which T T RE , hne ihteprmtr ntemdl As model. the in parameters the with changes hc eemnstelklho fprimaries of likelihood the determines which − ieiodthat likelihood T,X π RI X π RE RM RI ∈[ + ∈ π, wihIlbl“oeaeprimary “moderate label I (which T) ( 0, 1) R orsod otevle that values the to corresponds π) . T orsodn ohg priors. high to corresponding ilaotaprimary.Fora a adopt will eemnshwlkl pri- likely how determines orsodn olwand low to corresponding π π RI R ≡ ob specific, be To . sestablishment ’s π RI 1 −2 , (1 ssc that such is −q) q +2 2 q 2 X and 237 RM ,

B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 1012 1011 1010 1009 1008 1007 1006 1005 1004 1003 1002 1001 1000 999 998 997 996 995 994 993 992 991 990 989 988 987 986 985 984 983 EDITOR’S982 981 980 979 978 977 976 975 974 973 PROOF972 971 970 969 968 967 n that find I ogadcalnigpiais sti ol nraeteaon finformation of amount having the from increase benefit (namely would pre-candidates actually this about can as revealed primaries, party ( challenging A Ezra and perspective. example long different for a (see provides resources candidates avoid above and to smooth energy and their short be draining should elections primary that advised times increase. primaries of information quality reveal the when to Fig. increases in ability primary observed its a be adopting when can leaders effect This party larger. for is attractive more is election Lemma from know we equal) ihtepirble bu h nie.Tecmsfo Lemma be from comes of can The effect effect insider. negative the the This about are. belief prior primaries the attractive with identify less will primary the Fig. a hence in that observed and likely less candidate, the better is, a candidate insider the competent more high-skilled being of probability insider’s the of function a as primary a adopting 6 Fig. and 7. 6. 4. 3. 238 5. hstrsodicesswt h rmr kl ou.Telre h rmr skill primary the larger Second, room. The decision. smoke-filled bonus. nomination skill the primary it the bonus delegating with increases for threshold tolerate this can leaders party that ueo rmre.Fritreit ausof values intermediate For primaries. of ture in crease h rttorslso hstermcrooaetebnfi fpiais First, primaries. of benefit the corroborate theorem this of results two first The hsrsl otait eti iwo rmre nteltrtr.I ssome- is It literature. the in primaries of view certain a contradicts result This h hr n orhrslsdcmoeteefc of effect the decompose results fourth and third The h fetof effect The tityicesn with increasing Strictly with decreasing Strictly with increasing Strictly with decreasing Strictly tityicesn with increasing Strictly q h feto h xetdcmeec fteisdrcniaei nutv:the intuitive: is candidate insider the of competence expected the of effect The . h ieiodof likelihood The S h oelkl ti htteeiewl og ponigteisdri a in insider the appointing forgo will elite the that is it likely more the , q π T> a eitrrtda nipoeeti h nomto-eeainfea- information-revelation the in improvement an as interpreted be can RI althings (all .Hneteewl laseitacrandsac ihteRAF the with distance certain a exist always will there Hence 0. 6 q hc eit o h ieiodo dpigapiaydecreases primary a adopting of likelihood the how depicts which sas nutv huhmr ope.A etoe,a in- an mentioned, I As complex. more though intuitive also is π RI hc ntr ilincrease will turn in which 3, on X q π X π L RE S RI if L n ec on hence and , π . RI ∈[ π q , .Ti euti e nteltrtr about literature the in new is result This ). π) , π n nestv to insensitive and 7 T RI hc eit o h ieiodof likelihood the how depicts which . nices in increase an , T S nohrwrs primary a words, other In . nistocomponents, two its in ) h theorem The 2001)). q 3 q otherwise hc established which ilincrease will .Serra G. S π as RI

B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1058 1057 1056 1055 1054 1053 1052 1051 1050 1049 1048 1047 1046 1045 1044 1043 1042 1041 1040 1039 1038 1037 1036 1035 1034 1033 1032 1031 1030 1029 EDITOR’S1028 1027 1026 1025 1024 1023 1022 1021 1020 1019 PROOF1018 1017 1016 1015 1014 1013 oelkl oaotpiaista datgdpris hywr l previously all were They of importance parties. (2011 advantaged Serra than in primaries found adopt to less likely or more more primaries make not alter will not consequently will information attractive. and performance of eventual decision amount the her nomination decreasing of the or irrespective their increasing candidates So up primary. outsider made the an already in have of voters favor Primary in ignored. minds being actually would primary steei efc ogunebtenteeieadtemmesi ftepry and party; the of membership the even and party information; elite if of the even between amount an congruence maximum perfect election by a is primary reveal there example, a primaries is for Strikingly, if even election. enjoyed, eschewed previous be be a will could won already reputation has or who type incumbent This her. appoint orably parameters other the of value any under adopted be never will primaries and zero bv which above rmre,a tcudol eotie ymkn h elsi supinta pri- that assumption realistic the making information reveal by only obtained can be maries only could it as primaries, signals of quality the increasing 7 Fig. Challenge? Primary a Avoid Incumbents Will When primary a adopt to likelihood the q hoe 5 will Theorem under leaders out, outsiders party turns to that competition it the reputation As opening enough not challenge. by good primary her a a nominate have avoid might to insider candidate an insider an for conditions nubnsgtr-oiae nterpriswtotapiaycalne h rea- The challenge. primary a without parties their in re-nominated get incumbents X π RI RE ilnthv n feton effect any have not will h orh fhadsxhrslsbodyidct htdsdatgdprisare parties disadvantaged that indicate broadly results sixth and fifth fourth, The nohrwrs h nie’ euaincudb ogo htlaeswl inex- will leaders that good so be could reputation insider’s the words, other In h atpr ftersl smr upiig o o ausof values low For surprising. more is result the of part last The npriua,tefloigrsl eat rmpeiu eerha tprovides it as research previous from departs result following the particular, In ec hsrsl rvdsa xlnto o h miia bevto htmany that observation empirical the for explanation an provides result this Hence ∈[ , X q h fetof effect The π, L altig qa)on equal) things (all and 1) . ups h nta xetto htR shigh-skilled, is RI that expectation initial the Suppose R π hntethreshold the Then π π RI a motn ekesswt epc to respect with weaknesses important has L RI . ilpeetteueo rmr lcin o l ausof values all for elections primary of use the prevent will hc sanwcontribution. new a is which oId o lbrt nte ee ahrIfcso the on focus I Rather here. them on elaborate not do I so ) T h esni htcniae’promne nthe in performances candidates’ that is reason The . T , partially hc eemnstelklho fprimaries, of likelihood the determines which ahrthan rather fully. L hr xssathreshold a exists There . any π RI π circumstance. nices in increase an , RI , ssc that such is inevitably q , X RM 239 is . ,

B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1104 1103 1102 1101 1100 1099 1098 1097 1096 1095 1094 1093 1092 1091 1090 1089 1088 1087 1086 1085 1084 1083 1082 1081 1080 1079 1078 1077 1076 1075 EDITOR’S1074 1073 1072 1071 1070 1069 1068 1067 1066 1065 PROOF1064 1063 1062 1061 1060 1059 at nie teparameter (the elec- general insider the party in performance their of primaries forecast that imperfect pri- assumes and the tion. noisy in paper a performance this candidate’s are per- that contrast, mary their meaning In of partially, forecast information election. reveal perfect general only a are the primary in the formance in performance candidates’ that ing parties weakest (2008 the Merrill to and beneficial previ- Adams most that by are in found findings primaries as the that of example some (for corroborates literature paper ous this in findings the indeed and n au of value any o xml nAasadMril( Merrill and previously, Adams modeled been in has example benefit Such for nominee. party’s the of valence expected accordingly. them select can voters campaigning primary their cam- reveal the primary candidates and a the skills run campaign, to election general candidates the forcing before by paign candidates’ particular, the In about voters. information pri- to reveal that appeal to is mechanism personal paper a world this as the serve in around can premise elections parties The mary so. political increasingly by candidates—and used their party select method election? that to frequent primary confident competitive a be challenge a are she holding a elections or than Primary against he rather her can safe appoint When feel directly election? will future insider leaders a well-known of any nomination the or for incumbent an can When Discussion and Conclusions 8 party to deter sure to are enough is elite incongruence the elite-mass of and amount democratization. RAF any Lemma the benefit, a from both bring comes all: not This at candidate. pri- advantage same skill, the candidate’s any nominate insider bring the not about expectations do high maries sufficiently for that is son 240 fteisdrcniae f nteohrhn,teisdri eivdt eextremely be ability to the believed is about insider expectation the moderate hand, other for the so on do If, How- only candidate. performance. insider will the impressive they of most out primary turns a by it with provided as pre-candidate information ever, the the select use to may voters campaigns primary expected, As cisely. re-nominated get challenge. primary incumbents pro- a many thus without that paper parties fact The their empirical altogether. by the eschewed for be previous explanation a will an won election vides who reputation incumbent primary enough an a good being a then of has virtue election, insider insider by example the party for the better votes, if the winning fact, leaders for In party be. to to believed appealing is less candidate are Primaries primaries. of use aigti elsi supinldt e nihs h ro euaino the of reputation prior The insights. new to led assumption realistic this Making mean- fully, information reveal primaries that assume models those However nipiaino hs w etrsi htapiayeeto ilices the increase will election primary a that is features two those of implication An hsnwstpas loe tdigtebhvo fpiayvtr oepre- more voters primary of behavior the studying allowed also setup new This d R sitlrbefrpryleaders. party for intolerable is S seult eoadhence and zero to equal is π RI un u opa rca oei eern the deterring in role crucial a play to out turns ) 2008 ,Sra(2011 Serra ), ,Sra(2011 Serra ), T seult eo hc en that means which zero, to equal is )). ,Sye n ig( Ting and Snyder ), ie htpiaisdo primaries that Given 2. .Serra G. 2011 ), Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1150 1149 1148 1147 1146 1145 1144 1143 1142 1141 1140 1139 1138 1137 1136 1135 1134 1133 1132 1131 1130 1129 1128 1127 1126 1125 1124 1123 1122 1121 EDITOR’S1120 1119 1118 1117 1116 1115 1114 1113 1112 1111 PROOF1110 1109 1108 1107 1106 1105 . ro fTheorem of Proof A.1 Proofs the with Appendix curse a blessing. necessarily a not be are actually campaign could primary but party, the this the of to for intensity According wisely. and nominee length that their the acquire select paper, to them sympathizers campaign. help Democratic general-election to the beforehand the for information beneficial of was course it the consequence, information a in such As anyway could that revealed is however, which been premise, have Obama—information My would Republicans. and the by Clinton misused about be infor- later revealed much too being was that that was claimed benefit Observers a mation primary. to 2008 points the paper this during But mentioned costs. seldom possible potential other the among among existed: resentment base, and costs party’s party division such the include their that election to deny primary competitive not costs a do of possible I drawbacks the election. general about long the worried too in was Democrats prospects Obama Those and sup- severe. Clinton Democratic too between Several and competition and election. the Clinton that the primary Hillary complained Democratic about porters between the reveal competition during will the Obama they is Barack information example more recent A the the words, pre-candidates. are, other make primaries In debates, contenders. challenging among more more critiques include tough campaigning could allow pre-candidates’ and parties longer, the example, campaigns test For to enough. primaries thoroughly of skills design the during the information improving of from revelation fit the parameter improve to (the is cycle sugges- one primary make several direct to most provides how the paper is but This tions, then, prevalent. reformers, more par- for political elections question why primary and competitive A when democratic. about internally care become should ties we institution, representative any in the disregard them. to completely to will according voters provided exclusively primary information words, vote other con- and In the priors. campaigns preexisting ignore primary their actually the will in voters performance primary tenders’ incompetent, extremely or competent Challenge? Primary a Avoid Incumbents Will When ahcniaessildpn ntopee fifrain t ro eif,and beliefs, prior its information: of pieces two about holds on it depend beliefs skill The high-skilled. candidate’s being each Party of probability election, highest have primary to believes a is there If Lemma of Proof A.2 Table ns ihapecitv oe fw eiv htdmcaiainsol occur should democratization that believe we If note. prescriptive a with finish I 1 eei atclrcs fTbe1i hoe fSra( Serra of 1 Theorem in 1 Table of case particular a is here 1 1 q .Pltclprisadtegnrlpbi a bene- can public general the and parties Political ). R sRFwl oefrtecniaeta it that candidate the for vote will RAF ’s 2011 ). 241 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1196 1195 1194 1193 1192 1191 1190 1189 1188 1187 1186 1185 1184 1183 1182 1181 1180 1179 1178 1177 1176 1175 1174 1173 1172 1171 1170 1169 1168 1167 EDITOR’S1166 1165 1164 1163 1162 1161 1160 1159 1158 1157 PROOF1156 1155 1154 1153 1152 1151 h eiinta h A ae pnrciigtoeculso estimates. of couples those • receiving upon makes RAF the that decision the ac)if mance) hc seuvln atrsm ler)to algebra) some (after equivalent is which fadonly and if eevn ie siae t otro eif ilb ie by given be will beliefs posterior its estimate, given a receiving RO ler,adntn ht1 that noting and algebra, ebr r ainl hywl paeterpirblesbsdo h performances RAF the on the based that beliefs prior Given their s update campaign. will primary they rational, the are members throughout acquired information the 242 ilrnoieeulyif equally randomize will if equally randomize will • • supin h A ilvt for vote will RAF the assumption, for vote will RAF the assumption, • e oeagba n oigta 1 that noting and algebra, some ter nifrneasmto adntn that noting (and assumption indifference if hc are which hc seuvln atrsm ler)to algebra) some (after equivalent is which RI P(v fteRFobserves RAF the If fteRFobserves RAF the If observes RAF the If fteRFobserves RAF the If h A ilvt for vote will RAF The hr r orcul fperformances of couple four are There h A ilvt for vote will RAF The h A ilvt for vote will RAF The h A ilvt for vote will RAF The and r ihsild fteRFue ae uet paeispirblesafter beliefs prior its update to Rule Bayes uses RAF the If high-skilled. are RO s RO = P(v ( π 0, 1|s ofr opeo otro eif bu h rbblte that probabilities the about beliefs posterior of couple a form to ≤ 0) RO RO P(v π , P(v P(v P(v = RI ( = 1, RO with , 1|s RO high RI 1) RI = RO = , s s s s = = RI RI RI RI ( ) td aho hmi un ln with along turn, in them of each study I , RI = π π = = 2 1 q 1 2q qπ ( = = 1|s = )to 0) RI RI 1 (s = < 1|s − low high 2 1 2 1 2 high low − 1|s RI RI < < RO low q > 1 <π <π RO ,s −2 q)π + : RI 2 1 2 1 : : = )to 0) 1 ( : RO ilvt for vote will , for vote will , ) q q = −2 ( 1 RI RI = +2 2 1 RI hc seuvln atrsome (after equivalent is which high − ) qπ q q low hn ie yindifference my given Then, . indifference my given Then, . − high) q +2 2 + htteRFcudobserve, could RAF the that π q)π 2 RI q)( RI ,teRFwl oefor vote will RAF the ), )

1242 1241 1240 1239 1238 1237 1236 1235 1234 1233 1232 1231 1230 1229 1228 1227 1226 1225 1224 1223 1222 1221 1220 1219 1218 1217 1216 1215 1214 1213 EDITOR’S1212 1211 1210 1209 1208 1207 1206 1205 1204 1203 PROOF1202 1201 1200 1199 1198 1197 h cuayo h inl,wihis which signals, the of accuracy the RO are au.W a s h A’ eairdsrbdi h rvoslma o that, For lemma. previous the in calculate described to behavior need RAF’s first I the of use value can exact We the value. calculating by start I Lemma of Proof A.4 selection, elite appoint an directly With will (1) observations: party two the from directly comes conclusion This Theorem of Proof A.3 that noting (and assumption indifference my A.1 Table Challenge? Primary a Avoid Incumbents Will When hc sstedfiiino odtoa rbblt twice. probability conditional of definition the uses which optdi h al bv.Mliligadadn hs rbblte ses but easy is probabilities those adding just I and which Multiplying performances, above. candidates’ table the the given in vote computed will RAF the how on depends that probabilities prior the on only depends 2 ihapiayeeto h rbblt fnmntn ihsildcandidate high-skilled a nominating of probability by the increase election will primary a with (2) if if if if if π π π π π Table ahsmadi htepeso ssrihfradt calculate. to straightforward is expression that in summand Each RI RI RI RI RI π fadonly and if P(v RI ∈[ ∈ = ∈ ∈ ( (π ( 2 1 π, 0,π o h nie and insider the for 1 2 R A.1 , , h rmr oea ucino h signals the of function a as vote primary The 1) π) = 2 1 ] ) V umrzsteersls hc swa h em claims. lemma the what is Which results. these summarizes | π primary) S RI ydfiiin uhthat such definition, by ≤ oefor f Vote s admz admz oefor Vote Randomize Randomize s oefor f Vote oefor f Vote oefor f Vote RO RI π = P(v = with , = low low 2 2 1 RO RO RI RI R · v RI 2 P(s o h outsider. the for = π ,v RO V ≡ RI RI | s ,s primary) RI n thus and , 1 −2 ,s (1 RO q s s oefor f Vote oefor f Vote oefor f Vote oefor f Vote RO −q) .And q RO RI | +2 v P(v = S = RI 2 q P(v l t rprisaedrvdfo this from derived are properties its All . high 2 ,v high ecnd ob oigthat noting by so do can We . . R P(v RO RO RI RI 1 P(s RO v R P(v = −2 (1 RI ) = R −q) V q RI · +2 and R = V | P(v ,s primary,s 2 q = | V m 2 RO v RI | < s s oefor f Vote oefor f Vote oefor f Vote oefor f Vote V R primary RO | RO RI v ,v | = 2 1 m RI = = ,teRFwl oefor vote will RAF the ), RO r ihsild which high-skilled, are primary R ,v low high RI ) = RO RO RO RO RI RO ,s ; elite s ) RO RI eed nyon only depends ) ,s ; ) = v RO RI = P(v π ; ,v s s oefor f Vote oefor f Vote oefor f Vote oefor f Vote oefor f Vote RI π RO RI v RI RI RI RO + = = .And ,v ,v ) high S low RO RO . 243 RI RI RO RI RI ) ) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1288 1287 1286 1285 1284 1283 1282 1281 1280 1279 1278 1277 1276 1275 1274 1273 1272 1271 1270 1269 1268 1267 1266 1265 1264 1263 1262 1261 1260 1259 EDITOR’S1258 1257 1256 1255 1254 1253 1252 1251 1250 1249 PROOF1248 1247 1246 1245 1244 1243 P(v 0 S [ calculate tityngtv for negative strictly ssrcl eaiefor negative strictly is that 0 π . ro fLemma of Proof A.5 that find versions we previous algebra in appropriate reported the are With calculations paper). detailed this (the of here develop to long too 244 aclt h ifrnilof differential the calculate I n steeoedcesn utlk hi egbrn ons Hence points. neighboring their like just with decreasing therefore is and and hc r h ausw eeloigfor. looking were we values the are which at ∂π 2 1 π, ∂S RI ⇔ ⇔ < snndfeetal at non-differentiable is P(v π RI a o aclt h au finterest, of value the calculate now can I o ene oaayetesg of sign the analyze to need we Now 1) RI > R S π π 1 =− π π +2 = = RI = R RI 0 2 ehave we RI < q q ⇔ = −q 0for < V π when −2 < 2 1 hc ssrcl eaie If negative. strictly is which 1 S , .If | V S 2 π leadership 1 q π ≡ 1 −2 2 | RI = +2 RI primary) .If 2 π π q q q P(v π < RI ⎩ ⎪ ⎪ ⎪ ⎪ ⎪ ⎨ ⎪ ⎪ ⎪ ⎪ ⎪ ⎧ S q = +2 −q 2 RI RI −2 0for π −π = π 2 1 ∈[ 2 1 q 2 2 1 ∈[ RI R ∈ q RI 2 − u hti aifidbecause satisfied is that but , .S ehv indeed have we So 0. 2 q and RI π, q ( hc sstse because satisfied is which = π ∈[ π ) 0,π 2 ntn ht1 that (noting ∈ q q = RI , = 1) − V 2 ( π π ∈ 2 1 2 1 ⎩ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎨ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎧ + 2 1 stelmaclaims. lemma the as , π π | RI RI ] ) π , 3 primary ( , π π π π) RI RI ∪{ 2 1 2 1 RI ehv that have we for S π 1) 2 1 = ∈[ RI RI RI q , ihsm ler n oigtecniut of continuity the noting and algebra some With . RI q ihrsetto respect with = .So π 1) + q q ehv that have we π π if π, − q 2 }∪ − RI efidthat find we , .If π if 4 1 π + + 2 1 1) ) RI π ∈ S π and q (π q 2 1 − RI π RI 2 + if ( ∈ q sdcesn with decreasing is − 0,π RI − , q 2 P(v π ∈[ ( 2q π 2 2 1 0,π − 2 1 RI S ∈ 2 1 ) S RI π + q ∪{ S> π, .If ] π RI − ( 2 1 R = 2 = RI = 2 1 S 2 1 π ] − π 1) q , 2q = ∈ 2 1 RI 2 1 π π =− + RI π) 0for 2 2 1 }∪( π π RI S RI (π u scniuu tbt points, both at continuous is but , 2 V + π RI q n hc t in If sign. its check and h ausaoeaeue to used are above values The . , q = > RI | for ∈ , π q 2 leadership) 2 1 ∂π 2 1 2 1 ∂S RI π 1 π − ( )wihi aifidbecause satisfied is which 0) + for ≤ , ) RI −2 0,π RI π RI π) , q π π RI 2 1 2 q q =− ∂π π ∈ RI π π . +2 ∂S 2 + n aearaynoted already have I and ] RI ∈[ RI RI RI if q ( q ehv that have we 0,π π ∈[ 2 q π − = 2 1 naltoeintervals. those all in RI n nlyif finally And . 2 RI , eebrn that remembering , if π q q + 2 1 π ]∪[ ∈ q π 2 , ] + 2q 2 RI 2 1 − ( ] − 2 1 2 1 π ∈ , q − q π , π) S 2 1 2 (π RI − 2 1 hc is which 1 π − ]∪[ decreases RI S , ∈ 2 1 2 1 2 1 .Serra G. = ( + ) which π 0,π π 2 1 RI RI q> , 2 1 π) > − ) ∈ S , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1334 1333 1332 1331 1330 1329 1328 1327 1326 1325 1324 1323 1322 1321 1320 1319 1318 1317 1316 1315 1314 1313 1312 1311 1310 1309 1308 1307 1306 1305 EDITOR’S1304 1303 1302 1301 1300 1299 1298 1297 1296 1295 PROOF1294 1293 1292 1291 1290 1289 h hoe)that theorem) the nraein increase that ntoeintervals, those in e h ro fTerm2i er ( Serra in 2 Theorem of proof the See Theorem of Proof A.8 ( Serra in 1 Lemma of proof the See Lemma of Proof A.7 of values the that of differential the calculate I Lemma of Proof A.6 Challenge? Primary a Avoid Incumbents Will When o ons1 ,5 n ,setepofo ons1 ,6 n fTerm4in 4 Theorem of 8 and 7 6, 2, 1, points of proof the see 7, and (2011 6 Serra 5, 2, 1, points For Theorem of Proof A.9 ihamria nraein increase marginal a with rvdthat proved ausof values agnlices in increase marginal n ssrcl nraigwith increasing strictly is and oiieefc on effect positive on effect indirect oiie oi hs intervals, those in So positive. ∂S ∂q oee,if However, If osummarize, To osuyteefc of effect the study To oaayetecsswhere cases the analyze To = S π 2 1 sicesn with increasing is RI hc ssrcl oiie if positive; strictly is which S ∈[ π ,respectively. ), nTheorem in S RI π, sicesn with increasing is π ilntafc it. affect not will 1) ∈ T π , S S S S (π S hog t feton effect its through sursosv omria hne in changes marginal to unresponsive is sursosv omria hne in changes marginal to unresponsive is a titypstv feton effect positive strictly a has and hnvrfor whenever q , scntn o l ausof values all for constant is , 2 1 π 1,if ) q π , ean nteinterval the in remains . q 5 4 ∂S ∂q are q 3 4 pit3i h hoe) ent hti nyhsan has only it that note we theorem), the in 3 (point π , S q = S π ∈ π π ihrsetto respect with for q 2πq ssrcl nraigwt agnlicessin increases marginal with increasing strictly is ( ean nteinterval the in remains = = iial oethat note Similarly . 0,π π π π 2011 1 π RI − −2 ∈ 2011). (1 ∈{ ) and , ∈[ ( π −q) q S 2 1 +2 ∂S ∂q π ). + , rvdi Lemma in proved I . π }∪ 2 π) q π 1 = , 2 − π , = and ;smlryif similarly 0; (π ] q ∂S ∂q n aepoe i on of 2 point (in proved have I And . q π , hc ssrcl oiie if positive; strictly is which [ oethat note , =− n hc t in remembering sign, its check and 2 1 2 1 π , ) π T = ∪{ 1 RI −2 hrfr,cmiigboth combining Therefore, . 2πq ∂q ∂ 1 adeult eo,s an so zero), to equal (and q 2 1 q −2 [ ( +2 2 }∪ 1 1 q −2 q q + (1 −2 q q +2 2 π for −q) 2 q . ( π ∂q q q ∂ +2 ] 5 2 1 q ∈ +2 2 + 2 hr utproved just I where , ( π 2 that q codn othe to According . ( π) 1 q q 2 π, −2 ∈ 2 (1 , hc sstrictly is which ,s iha with so )>0, ∪{ −q) ( 2 1 q q 1) 0,π +2 ] a strictly a has , hr just I where π 2 q ∂S ∂q } 2 ) . )< ∪ = π ( π, .So 0. ,so 0, = 245 1) q 1 2 . , , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1380 1379 1378 1377 1376 1375 1374 1373 1372 1371 1370 1369 1368 1367 1366 1365 1364 1363 1362 1361 1360 1359 1358 1357 1356 1355 1354 1353 1352 1351 EDITOR’S1350 1349 1348 1347 1346 1345 1344 1343 1342 1341 PROOF1340 1339 1338 1337 1336 1335 T sla 21)Mda-ciit rmda-oes pta oe ihitaprycompeti- intra-party with model spatial A median-voters? Polit or J Median-activists Q outcomes. (2012) policy D and Kselman processes jury Nomination unanimous (2007) K of Mattes inferiority L, Mathevet the MO, innocent: Jackson the Convicting PF, (1998) Galderisi W In: Pesendorfer elections. T, Feddersen primary congressional of burdens decisions and des benefits probabilite The la (2001) a M l’analyse Ezra de l’application sur Essai [1785](1994) de national M annual the Condorcet at J presented Paper voting. crossover Econ competition. and primaries electoral Open polarisation. (2008) and I organization Kang political S-J, Party Cho (2010) strategic N Sahuguet and BSY, Information Crutzen M, (2008) Castanheira M Castanheira un- J, motivations, candidate Carrillo model: voting multidimensional the of Robustness 75:671Ð697 117(4):1453Ð1489 (1985) Stud Econ RL Econ J Rev Calvert Q motivations. intermediaries. Political political (2008) jury as S Condorcet Parties Callander (2002) the J and Tirole B, rationality, Caillaud Games aggregation, Information competition. valence (1996) JS and Banks platform D, with government: Austen-Smith Elections good value (2009) candidates EB When Mesquita (2009) de EN S, Simas Ashworth WJ, Stone III, S with beginning Merrill elections J, two-stage in Adams strategies party and Candidate (2008) III S Merrill J, Adams References Lemma from Note Theorem of Proof A.10 that π prove I derivatives, partial 246 fteohrparameters. other the of eaie nteohrhn utcluae that calculated just I hand other the On negative. nwfo Lemma from know n nidrc fettruhisefc on effect its through effect indirect an and S ∂T ∂S [X RI ssrcl eraigwith decreasing strictly is in aucit unMrhInstitute March Juan Manuscript, tion. 2(1):67Ð94 Sci 1998 92(1):23Ð35. Rev Sci Polit Am voting. strategic under verdicts & Rowman representation. of politics the and York primaries New Congressional Littlefield, (eds) M Paris Lyons (trans) M, F Ezra Hewitt and I McLean voix. des pluralite la a rendues Association Science Political Midwest the of conference 26(2):212Ð242 Organ Econ Law J 118(530):845Ð874 29:69Ð95 Sci Polit J Am convergence. and certainty, 1996 90(1):34Ð45. Rev Sci Polit Am theorem. 67(1):191Ð216 Behav Econ Manuscript elections. congressional to applications with model spatial a 52(2):344Ð359 Sci Polit J Am primary. a osuyteefc of effect the study To (1 ∂π RE ∈[ ∂S −π RI (1 π −π L ti ayt aclt that calculate to easy is It . )(π , L π )−X RI ] . +S) L π L ] ec,when Hence, . 4 2 that that π RI ∂π ∂S S ems oeta thstoefcson effects two has it that note must we RI π 5 RI = snnpstv.W hrfr aethat have therefore We non-positive. is . q when 0 a titypstv feton effect positive strictly a has π RI ∂π ∂T ∈[ RI π, = π S RI 1) nttl ehv that have we total, In . −S ∈[ [X ehv that have we (1 π, −π RE (1 ∂T ∂S L 1) −π )(π n eebrthat remember And . ssrcl oiie n we and positive, strictly is L RI )−X +S) L 2 T π L ] = T hc sstrictly is which T o n value any for 0 ieteffect, direct a : dπ dT hnvrfor whenever dπ dT RI RI = < .Serra G. ∂π ∂T and 0 T RI ≡ + Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1426 1425 1424 1423 1422 1421 1420 1419 1418 1417 1416 1415 1414 1413 1412 1411 1410 1409 1408 1407 1406 1405 1404 1403 1402 1401 1400 1399 1398 1397 EDITOR’S1396 1395 1394 1393 1392 1391 1390 1389 1388 1387 PROOF1386 1385 1384 1383 1382 1381 ndrJ r igM 21)Eetrlslcinwt ate n rmre.A oi Sci Polit 67:490Ð498 J Rev Sci Am Polit primaries. Am maximizers. and utility parties as Parties with (1973) selection D Wittman Electoral Polit (2011) Theor MM J Polit valence. Ting and Jr, J policy JM valence. between Snyder endogenous trade-off party’s with The elections primaries? of Why (2011) model G A Serra what? 35:635Ð of Sci Polarization Polit J (2010) Br G 1988Ð1996. Serra Israel, in competition Multiparty (2005) I Sened N, Schofield under positioning Party (2008) AV Zakharov E, Schnidman convergent U, for Ozdemir conditions C, sufficient Claassen N, and Schofield necessary theorem: voter mean The (2007) Cam- N Press, Schofield University Harvard applications. and theory competition: di- Political persistent (2001) contests: JE two-party Roemer in competition Polit Polit electoral J Two-stage Theor (2006) finance. B campaign Grofman J and G, ambiguity. Owen advantages, incumbency strategic contests, and Electoral (2008) primaries A party Meirowitz Informational (2005) A Meirowitz Challenge? Primary a Avoid Incumbents Will When 55(4):781Ð795 23(1):21Ð51 72(2):426Ð437 663 Uni- Washington Economy, Political in Center Typescript rule. versity. plurality and representation proportional 74:965Ð980 Stud Econ Rev equilibrium. bridge Heidelberg 26(3):547. Welf Choice Soc positions. party of vergence 70(03):681Ð699 17(1):107Ð136 247 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 DOI (eds.), al. et Schofield N. e-mail: Schnakenberg K.E. e-mail: Penn E.M. Drive, e-mail: Brookings USA 1 63130, Louis, MO Saint Louis, in Saint University Washington Economy, Political in Center ue htosre eeso nunerlt oa neligltn qaiy fthe of “quality” latent underlying an to social as- vertices. relate approach the influence Our in of In network. levels problem examples. a observed common in that these vertices sumes a of of solving influence each to the in estimating approach of analyst new sciences—that relationship an a a to present form interest we chapter to of this propensity often are its node, of another terms with in defined opinion), citizen, country, etn hslre su otesd,dfutt h s fara odt ee oteltn construct latent the to refer to word real a of estimate. and use detect the to and, to constraints attempting default space is side, and method scope the precise our of to be issues issue might traditional larger the one note this While we “citability,” setting estimating. as is such method term a our use that and characteristic the of parlance) common networks. in causality of inference with issues methodological discuss ( Lazer include mluil.W rsn crn ehdta saporaefrmauiginflu- measuring for assumption appropriate this is that make networks method that many scoring constraints vertex, a practical other present every We other implausible. influence or to spatial, potential temporal, the reflect has vertex each that 2 1 records and include relationships, and social citations. ubiquitous, countries, judicial between are alliances science co-sponsorship, political legislative in of data network of Examples Schnakenberg E. Keith and Penn, Maggie Elizabeth Patty, W. John Network a in Vertices Precedent: Scoring of Quality Latent the Measuring ..Pty( 1RC4LM010958-01. Patty J.W. # Grant NIH by supported was research This h od“ult”i ipyapaeodr huhoeta sruhydsrpie(tlatin least (at descriptive roughly is that one though placeholder, a simply is “quality” word The reviews comprehensive Recent growing. and large is science political in literature networks The 10.1007/978-3-642-35239-3_12 [email protected] [email protected] [email protected] 2 B lhuhcmo ehd o esrn nunei ewrsassume networks in influence measuring for methods common Although ) 2011 · ..Penn E.M. 1 .I diin olre l ( al. et Fowler addition, In (2011). al. et Ward and ) ueia siae fthe of estimates Numerical dacsi oiia Economy Political in Advances · ..Schnakenberg K.E. pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , influence , fec oe( node each of 2011 umrz and summarize ) e.g. legislator, 249 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 (C n a fine oe fti xo o u upssi h blt ognrt crsfralterna- for all scores locate generate scores scale. these to Substantively, common data. ability a the the on in opinions compared is directly purposes not our are that for tives ( axiom Luce this by of described irrele- as power model of axiom that independence theoretical choice presuming the (IIA) The with by alternatives accordance fail. consideration vant in under generated or are objects succeed successes the observed to the of potential quality world relative the (unobserved) the has an identifies in object quality every latent which opinion’s in each identifies that more model ( in retical explored Penn and and developed Schnakenberg method, by The detail incompleteness. of problem this with incomplete. necessarily aa erpeetteosre ewr aab rp eoe by denoted graph a by out data lay network network first general observed we to the method Accordingly the represent chapter. applies We this one data. how in discuss network then a and preliminaries as some data our of conceive We Data Network from Quality Inferring 1 a whether observe to us allow not do data the opinions, opinion earlier given by cited some be assess cannot to ( us ( “succeeded” enable “failed” decision data given de- network a future The when and structure. instances current network for a precedent as Supreme generates connectedness—the decisions deci- cisions their previous These to of decision. due future usage method a explicit our in Court’s by cited analyzed be readily will are decision sions the that likelihood the as vertices. fine influencing dif- the reveal of edges “quality” some latent that the may so about vertex vertices, information each other of (2) of amounts ferent number and different “quality” vertices a underlying certain by their with influenced to edge be unrelated an are form that cannot reasons vertices some for (1) where networks in ence 250 3 interactions vertex influence to potential the idt h oino ult,a ehv icse earlier. discussed have we as quality, of notion the to tied than” greater “is or where o any for influencing that ngnrlntokstig,w nepe oncinfrom connection a interpret we settings, network general In 1 h ehdw ecieadepo nti hpe sitne oda explicitly deal to intended is chapter this in employ and describe we method The sa xml,w aete“ult”o urm or eiin,wihw de- we which decisions, Court Supreme of “quality” the rate we example, an As neligormdli nasmto htec vertex each that assumption an is model our Underlying ,...,C j community ∈ V v,w C ={ a o ie naltroiin.Hwvr eas ae decisions later because However, opinion). later a in cited not was i.e., i n nteohrhn,o course, of hand, other the On . with , i ) 1, n oi skonta hyhdthe had they that known is it so and , ∈ en e fsbesof subsets of set a being ol aebe cited been have would 2,...,n V v , ob usto vertices, of subset a be to hti e o u upssi httento fiflec econceptually be influence of notion the that is purposes our for key is What . E v w) (v, ⊆ } E ˜ sastof set a is .If ∈ E i j) (i, niae htteei nedge an is there that indicates i odfieti omly let formally, this define To . ∈ n 2012 etcsand vertices E ya ale pno.Orntoksrcueis structure network Our opinion. earlier an by V hnw nwthat know we then ,i one nasml aimtc theo- (axiomatic) simple a on founded is ), and , i k) (i, C ⊆ C a ie naltroiin or opinion) later a in cited was i.e., i ∈ V en h omnt fvertex of community the being E E iha with , potential v sasto ietdegs where edges, directed of set a is edntipythat imply not need to w sipyn that implying as omnt structure community i j 1958 and oitrc:i sknown is it interact: to nacommunity a in E from ˜ eastof set a be j .I usel the nutshell, a In ). neatdwith interacted v to ..Ptye al. et Patty J.W. G w w i “influences” = . ol not could 3 potential V E) (V, ede- We C i C has i = . j , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 i to n teshdn oeta oiflec.Teecnet r ersne yteset the by represented are contests These influence. not, which were to but in potential S influential “contests” no be to had of potential others collection the and had a some influential, generating were as vertices some structure community and work G qual- latent higher that the Thus, vertex with. interactions interact influential) of to higher (i.e., potential ity with the successful had had vertices they have that that vertices to with assume likely We more quality. are vertex’s qualities latent the representing scale mon k ersnsteotoeo contest. a of outcome the represents escntan hti odtoa ntecmuiysrcue h osriti that is constraint The structure. community the on conditional is that constraint ness insc seg weights), edge as such tion i xetdiflec ncontest in influence expected aebe once to connected been have Network a in Vertices Scoring Precedent: of Quality Latent the Measuring k to connected been n thus and k influence to potential tlz u ewr n omnt tutr oestimate to structure community and network our utilize ino hscntan mle,gvnacretcmuiysrcue htn ver- no that structure, community Satisfac- correct score. a expected given total implies, their equals constraint score this actual of total tion vertex’s each that or nwrs h xetdsaeo nuneof influence of share expected the words, In tutr sdsge oacmoaeti at n npriua easm that assume we particular in and fact, this accommodate community k to Our factor(s)). designed independent other is some structure for or interact, to opportunity an okn ttettlsto etcsta culyinfluenced actually that vertices of set total the at looking influence potentially can that vertices the of influence to tial by ∈ ∈ a rmynthv ucee ( succeeded have not may or may oncigto connecting sobserved). is a o fhg nuhqaiyt influence to quality enough high of not was ={ k n nosre retmtdcmuiystructure, community estimated or observed an and u oli oetmt ahvre’ ltn ult”soesbett network a to subject score quality” “latent vertex’s each estimate to is goal Our com- a on placed be can vertex each that is model our of assumption second The Let iial,w a aclt h hr of share the calculate can we Similarly, G C .Otherwise, ). S j i) (j, hsstis set This . i s eltteepce nuneo vertex of influence expected the let we mle that implies x ∈ k = V ∈ a optnilt influence to potential no had (x E : i 1 h oelkl ht o n ie vertex given any for that, likely more the , s v) (s, . ,...,x i k W is ,wihw eoeby denote we which ), k i k) (i, i u h ikwsntcetdfrsm esn(osbybecause (possibly reason some for created not was link the but , k ∈ ={ i ssaeo aetiflec eaiet h oa aetinfluence latent total the to relative influence latent of share ’s n E ) ( s i.e., k w ∈ ∈ ∈ o some for ahr tmyb h aeta pno that opinion that case the be may it Rather, . S R : E E(s,i) k ˜ i i w) (i, n Thus, . ssaeis share ’s i a h potnt oln to link to opportunity the had ersn ahvre’ aetqaiy hnfreach for Then quality. latent vertex’s each represent szr eas nti pno easm that assume we opinion this in because zero is E(i,k) ∈ v eadeso hte neg between edge an whether of regardless i.e., = k E ∈ s en ncommunity in being ∈ }⊆ = V S | i W } 1 (,i) A(s, ( hs vr etxta a influenced was that vertex every Thus, . i hr sn hnethat chance no is there i.e., C E(i,k) | actual j i if k i n wtotayadtoa informa- additional any (without and , ∈ x . i nacneti which in contest a in C k k osbybecause possibly , i ∈ k x qa if 0 equal , j nuneof influence W ncontest in o all for . i n tews.W a now can We otherwise. 0 and C ecnev fornet- our of conceive We . i x j i, k C ntentokdescribed network the in ∈ ujc oa unbiased- an to subject ,rgrls fwhether of regardless ), i i V mle that implies i k) (i, k ( , rbblt of probability i.e., in j i) (j, k k i and ∈ ,or a h poten- the has i E i ˜ ilconnect will ∈ i Thus, . ol have could (,k) A(i, E ee had never ˜ k implies a the had k ∈ i ,by and C 251 k ’s i , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 xr nuneon). influence exert xssavector a exists iain r rvddweeaporaei u discussion. our in appropriate where ( provided II are Spriggs citations and Hansford include analyses book-length hti nqeu osaa multiplication. scalar to up unique is that nodes node by selection/influence aty eoi osdrto faloiin te hntemjrt pno.Both opinion. majority the than impor- other Most opinions network. all of citation/precedent consideration judicial omit the we tantly, of operationalization our in 02 hs iwdi h hoeia rmwr rsne bv nSect. in above presented framework to theoretical 1791 the from in opinions viewed majority Thus, Court 2002. Supreme to opinions vertex majority Court a Supreme influencing of incapable which is in vertex situations later a in it. preceded and arise that time example, by for indexed to could, are impossible networks vertices are of links certain types which These in observe. networks analyzing to well-suited ticularly nutshell, opin- future a in in opinion the citation: of revisitation of augur of determinants opinion study an ions? the of the on factors to or focused approaches factor various what have adopted most have but scholars topic, precedent, the of breadth the nqeyietfidu osaa utpiain h ai faytonds quali- nodes’ two any of ratio the multiplication, scalar ties, to up identified uniquely eal ntemethod. the on details u ncttosa nidcto fti sg—a trce utie attention years. sustained 60 attracted over usage—has for scholars this science fo- of political a indication and particular, an legal in from Justices—and, as Court citations Supreme on by cus precedent judicial of use The Precedent of Quality the Measuring 2 Schnakenberg was. actually it than influential less or ( more Penn be and to estimated is tex 252 5 4 et h eaieqaiisiflecso h ifrn oe.I atclr as particular, In nodes. different the of qualities/influences relative the sents suiul dnie.Ti ratio This identified. uniquely is ie forntokaeSpeeCutmjrt pnos n fmjrt opinion majority if and opinions, majority Court Supreme i are network our of tices o esn fsae erfrteitrse edrt cnknegadPn ( ( Penn Merryman and include Schnakenberg offerings to Seminal reader interested the refer we space, of reasons For ie aoiyopinion majority cites eoemvn n ti motn ont htw r xlctyasrcigfrom abstracting explicitly are we what note to important is it on, moving Before euiieadt e ossigo h olcino iain yUie States United by citations of collection the of consisting set data a utilize We par- is it objects, between relationships unobserved imputes model our Because i and 2012 j opt ( compete x rv ht ujc oamnmlcnetdescniin there condition, connectedness minimal a to subject that, prove ) ∗ = (x 1 o n uuend htboth that node future any for i.e., j ∗ ,...,x eicueteedge the include we , i essta ynode by that versus n ∗ ρ ) j i n adsadPse ( Posner and Landes and 1954) htsle h bv ytmo qain and equations of system above the solves that ersnstehpteia eaiefeunyof frequency relative hypothetical the represents ρ j i ≡ 4 x x iwdsbtniey hsvco repre- vector this substantively, Viewed j i , n ehrt( Gerhardt and 2006) i j) (i, j naftr ots nwihboth which in contest future a in ∈ E . i and 5 1976 nupiigy given Unsurprisingly, j 2008 ,wiemr recent, more while ), aeteaiiyto ability the have .Ohrrelevant Other ). ..Ptye al. et Patty J.W. 2012 1 )formore ,thever- x ∗ is Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 aknsta r i hoy dnia oternig htwudb siae from estimated be relative would recover that and rankings data the the to of identical theory) subset (in any are examine that can rankings we that implies approach a any is by opinion every cited approach, not this is In that ranked). opinion be (any contest cannot of once database quality the all least in the rank at opinion would cited of other been approach question have straightforward that the most opinions The considering the opinion. when each data of influence this or approach might one ways goe h iigoiinsteteto h ie pno ( opinion cited the of treatment opinion’s citing the ignores etcs( vertices Year iaindt Fwe ta.( al. et (Fowler data citation pre- strictly that Data opinions only and any by influenced be it. can date opinion an words, In httedt eepo eed o lo st xlr oefully. more explore to us allow not do here employ we data the that nte pno hti eie nta aeya.Gvntenme fyasta ecnie,this consider, we in that year years that of in number decided the Given is year. that same a opinion that affects one in approximation cite decided is cannot that Court opinion the another year, any in that, (unrealistically) opinion. and citing the Clark in see occurs examples, citation recent For opinions. earlier cite Justices how ( of Lauderdale aspect this consider to ( opinions and opinions between relationship the of conceive should 6 role the inferring and opinion majority ( precedent the of constructing quality in bargain-and work the both at understanding processes for relevant ing are opinions concurring and dissenting Network a in Vertices Scoring Precedent: of Quality Latent the Measuring 9 8 7 o iainb vr usqetyrnee opinion. rendered ( subsequently eligible every by is by citation opinion into cited for each be take cannot that we opinions presume chapter, earlier—later opinions—and this discussed earlier of bias purposes temporal be the the of to only For sense likely account another. the less one in are with data that compared objects the of directly up” communities “break observing) to perhaps, freedom (or, the estimating compared. analyst directly been an not offers have it that Accordingly, objects compare/score to us allows employ we Structure Community Cases: Differentiating rdistinguishing). or oeta,frsmlct,w prxmt hs“ae hn eaini h es htw presume we that sense the in relation than” “later the this times approximate of we number simplicity, the for by that, citations Note differentiated has that work recent any of employed aware and not developed are We have scholars that ways of number a are there one speaking, how Practically regarding questions empirical and theoretical interesting many are there addition, In rcia osrit rhbtu rmrnigalo h pnos otntl,our Fortunately, opinions. the of all ranking from us prohibit constraints Practical hsw osrc h community the construct we Thus (i) n ahoiinta sctda es nei a is once least at cited is that opinion each and , eapyormto oFwe n ensSpeeCutmjrt opinion majority Court Supreme Jeon’s and Fowler to method our apply We eteya nwihopinion which in year the be aoiyopinions), majority i.e., ,SrgsI ta.(2011 al. et II Spriggs 2010), 7 , 8 very elaeec fteefrftr work. future for these of each leave We Year ml rprino h ubro oeta iain econsider. we citations potential of number the of proportion small e.g. 2007 i > (i) arbae l ( al. et Carrubba , ) hr r ubrof number a are There (2008)). Jeon and Fowler ), i , Year ). j , i C (j) a er,w sueta o n arof pair any for that assume we heard, was i o ie opinion given a for ⇔ 2011)). sdsusderir h method the earlier, discussed As j 9 ∈ contestant C 6 i nadto,orapproach our addition, In e.g. . e.g. omrt ta.( al. et Bommarito , i competition”) “in i.e., i aoal,critical, favorable, , sflos Letting follows. as . 2009 253 )) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 usto bevdadsbeun aoiyopinions. majority subsequent and observed of subset bevdadetmtdcvrae itnusigtevrosoiin n majority and opinions of various inclusion the opinions. require distinguishing will covariates on research estimated conclusion Such typ- and a scores. a observed offering these in before of justices foundations necessary of micro-level majority is the a study by further supported Obviously, are opinion. that ical policies support opinion the of h 0 otctdoiin ic 80wt neetwr oprn h ranking the comparing toward eye an with 1800 since opinions most-cited 204 the enwpeetterslso he eae nlss efis rsn u eut for results 2002. and our 1946 present between first We rendered opinions analyses. most-cited related 100 three the of results the present now We 3Results opin- an the with examine both) we (or ending terms, or whose theoretic beginning ion graph edges In all frequently containing most 2002. 100 subgraph and the smallest 1946 to attention between our opinions restrict cited we Accordingly, set. data entire the 254 12 11 10 rsn h eut o h 0 otctdoiin ic 1800. since opinions most-cited 100 the for results the present h eret hc h pno sesl nesod h eret hc t conclu- its for which proxy to degree might the applicable, understood, It broadly easily course. are is sions opinion of the unambiguous, which appeal not to degree this is the What terms opinion. majority substantive subsequent in any represents in citation potential a as opinion aa(Spaeth data saadd0pit nta ots.Wt hsvco fsoe o ahcnetn in Schnaken- contestant by developed each method ( for the Penn apply scores and directly of berg to vector possible this then is With it contest. contest, that contestant each in the Otherwise, points (contest). 0 opinion awarded majority is ( that 1/| in in awarded cited cited (contestants) is was ions contestant contestant the the ( If contest, participant opinion/contest. that contestant’s been majority the having of that member as in one contestant citation) least the at count cited ( and we opinion opinion community, majority an each to for subsequent and, ( was turn opinions that in the contestants of the choose scores we influence First, the for solve then we now we which algorithm, detection are community edges, our incident than for their describe. more used and many opinions, then these are contains After they graph exact). selected, be This to 2002. (3674, opinions and 100 1946 between rendered opinions ion, hstm eidicue l ae nteFwe n endata. Jeon and Fowler the in cases all includes period descriptive time rich Spaeth’s This which for data Jeon and Fowler the in cases all includes period time This true is this that Note hs aetiflec crsrpeet nesne h pelo ahmajority each of appeal the essence, in represent, scores influence latent These sn h er fteoiin ocet h omnte sdsrbdearlier, described as communities the create to opinions the of years the Using x ˆ = ( x ˆ 1 2012 ndegree in ,..., r loavailable. also are ) ognrt h aetiflec crso ahmjrt opin- majority each of scores influence latent the generate to 2012) x ˆ n despite ) . nme ftmsctd ak mn h o 0 mn the among 100 top the among ranks cited) times of (number h rsmto hta pno ih aebe esbeol na in only feasible been have might opinion an that presumption the 10 rprastelklho htteplc implications policy the that likelihood the perhaps or W | ons where points, otsat)a follows. as contestants) i.e., 12 W 11 ial,w consider we Finally, olwn ht we that, Following stesto opin- of set the is vial for available i.e., ..Ptye al. et Patty J.W. contest) i.e., won) i.e., Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 ae ic 1946 since cases most-cited 100 Top Sample: scores. with correlations cr infiatyhge hnalo h others. the 100 of all the than for higher significantly score score quality greater estimated is average score quality the estimated construction the by which cases. is for which opinions 1, of set than the is This period. Table 1946 Since Opinions 100 Top 3.1 considered. are 100 these all as when times cases many those as least of at ranking cited the been with have 1946 that since opinions opinions most-cited 100 the of 1 Network Table a in Vertices Scoring Precedent: of Quality Latent the Measuring 14 13 h neetn seto h orlto sntta ti oiiebt ahr htit that rather, but, positive Table is of it inspection Indeed, that 1. not to is closer cited, correlation not been to the is responsive has of positively opinion aspect posi- obviously an interesting is is that the opinion opinion times an majority of of Court number score Supreme the the subsequent surprising: not a is by tive cited been opinions. has of sample opinion the expand we when below question this to return partially potentially n ftee10cssi rdce oct either cite to predicted is cases 100 these of one 1 over for nln ihpeiu oko h ercaino h rcdnilvle(r tleast, at (or, value opinions. precedential the judicial of depreciation of the usage) on work previous with line in opinion. opinions the subsequent of of age the number by number the divided the and opinion opinion, opinion, the the citing the of citing age opinions the subsequent respectively, of with, scores opinions’ the Table relating interest. of are correlations the course, individuals. Of detained of States. famous rights United the procedural the the is clarified in opinion penalty scoring death highest the third of judicial constitutionality the opinion, for the second-ranked implications fied The broad decision-making. with bureaucratic law of administrative review in case well-known a is e,freape lc n pig I( II Spriggs and Black example, for See, account opinions two these so 100, to sum opinions 100 top the for scores estimated the that Note hsrnigi neetn nanme fwy.Tetptomjrt opinions majority two top The ways. of number a in interesting is ranking This httecreainbtenteoiin’soe n h ubro ie each times of number the and scores opinions’ the between correlation the That h eaiecreainbtenteaeo noiinadissoei broadly is score its and opinion an of age the between correlation negative The simple few a but scores, the of treatment full-throated a from us prevents Space 2 rsnsteoiin ihtetp3 siae aetqaiysoe o this for scores quality latent estimated 36 top the with opinions the presents Descriptive t ftesmo h siae crs nohrwrs n pno htctsexactly cites that opinion any words, other In scores. estimated the of sum the of /8th tod ihteIAaimo hc h crn loih sbsd We based. is algorithm scoring the which on axiom IIA the with odds at orlto #Ct/er Score): Cite/Year, (# Correlation orlto #Cts Score): Cites, (# Correlation orlto Ae Score): (Age, Correlation 14 ti motn ont,hwvr htti fetis effect this that however, note, to important is It 1 2010 rsnstrePasncreaincoefficients correlation Pearson three presents ). Chevron Miranda 13 2 indicates, or h o-crn opinion, top-scoring The Gregg eei paribus ceteris eiini hc h Court the which in decision lot1 ftetime. the of % 13 almost fortiori a htterank- the that , Accordingly, . Gregg,clari- Chevron, −0. 0. 0. 787 496 461 255 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 al 2 Table 256 2Bvn .SxUkonNmdAet...17 1 1971 Pape v. Monroe . . Agents. Named Unknown Six v. Bivens Education of 23 Board v. Brown 22 21 6Wosnv ot Carolina North v. Woodson Powell v. Stone 36 35 0Lcetv Ohio v. Lockett 20 4Bno .Maryland v. Benton 34 9Mle .California v. Miller 19 3Rt .Uie States United v. Roth 33 8Gdo .Wainwright v. Gideon 18 2Kt .Uie States United v. Katz 32 7Bkrv Carr v. Baker 17 1Wiwih .Sykes v. Wainwright 31 6NwYr ie o .Sullivan v. Co. Times York New 16 88 1 80 1.08 1957 Mills Lincoln v. Union Workers Textile Wade v. Roe 30 29 5Ednsv Okla. v. Eddings Winship re In 15 14 8Wrhv Seldin v. Warth 28 3Bclyv Valeo v. Buckley 13 7Tryv Ohio v. Terry 27 2Mtesv Eldridge v. Mathews 12 5SD lg rdsCuclv amn15 1. 1959 Georgia v. Furman Garmon v. Council Trades Bldg. S.D. 26 25 1AlntnHihsv er.Hu.Dv op 971 1977 Corp. Dev. Hous. Metro. v. Heights Arlington Williams v. Dandridge 11 10 24 akName Rank oelv e’ fSc Servs. Soc. of Dep’t v. Monell Sims v. Reynolds 9 Jordan v. Edelman 8 Wash. v. Strickland 7 Harris v. Younger Chicago of 6 University v. Cannon 5 4 iad .Arizona v. Miranda 3 rg .Georgia v. Gregg 2 hvo,UA n.v RC Inc. NRDC, v. Inc. USA, Chevron, 1 Craig h 6hgetsoigoiin.Sml:Tp10ms-ie ae ic 1946 since cases most-cited 100 Top Sample: opinions. scoring highest 36 The .Boren v. 961 1976 961. 1976 991 1969 911 1961 1. 1954 931 1973 971 1957 971 1967 981 1978 931. 1963 77 2 71 1.07 1977 921 1962 941. 1964 951 1975 89 3 91 1.08 1973 921. 1. 1982 1970 961. 1976 981. 1968 921 1972 9103 100 1.49 1976 901 1970 961 1976 941 1984 2 1979 962 1966 947 1984 2143 144 1. 1.62 1978 1. 1964 1974 1. 1971 965. 1976 erSoe#CtsCites/Year Cites # Score Year ...... 19 3 96 18 21 19 3 3 97 80 01 01 5153 155 25 6144 104 26 47 2 75 04 7114 131 27 5153 155 05 8275 207 28 6173 127 06 4193 149 34 8114 161 38 7114 131 491447 1722 9103 100 49 1972 2183 2 118 89 12 15 3 4 132 57425 77 2 70 17 87 3 78 3 3 58 92 4 68 3 65 129 74 73 89 62 2256 225 72 4269 266 14 2196 129 52 3 3 134 ..Ptye al. et Patty J.W...... 6 2 2 2 4 4 2 5 6 7 1 7 3 6 7 8 7 8 8 6 1 7 2 6 1 9 2 8 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 ihs-crn oiin is,temjrt pno akdfut-ihs nthe in fourth-highest Table ranked in opinion fourth reported the majority analysis at interesting 1946Ð2002 the get identical Things First, are analyses. opinions position. two majority the highest-scoring three in top scores the similar particular, very In have are. and scores of sets greater two is the score 1946 quality estimated between the period which the for opinions for of 1. analysis set than previous the the is this with 2002, As and period. this for scores enwpeetorrslsfrtetp10ms-ie pnosrnee between rendered opinions most-cited 100 top Table the 2002. for and 1800 results our present now We 1800 Since Opinions 100 Top 3.2 would of opinion all hypothetical of the scores to the prior whereas scores opinion, average hypothetical decrease. above the with by increase citation opinions would the those scores average that of than sense result lower the had a in initially as opinions that opinions the cited all of that of scores opinion scores opinions the hypothetical the compress a other cited example, would of for opinion being number extreme, previous for the the every At score) to opinion. proportional (or that inversely “reward” by which is cited opinion’s on opinion an axiom difference. subsequent IIA that this a the implies of by that based origins fact is the the consider method the of to the by because useful generated arises is that distinction It from the year. Specifically, differs per scores the citations our an that of by cite note number generated cannot to opinions important opinion is the an it of that though, This fact ranking Again, positive. the subsequently. strongly rendered for is is control down that scores handed opinion the was that it fact since the cited highlights been has opinion the year commonly is mea- it are the as network. scores or, analysis, the counts network way: citation in another opinions’ and known Put received the have identical. than they not different citations something are of suring scores number their the to to respect respect with with opinions the of ings Network a in Vertices Scoring Precedent: of Quality Latent the Measuring 16 15 aca upr o edrdsrmnto ne il X hc al o edreut nhigher in equity gender for calls which IX, Title under discrimination education. gender for support nancial scores. relative inherently they oenet opaelmt noseiy hsoiini,o ore mn h top the among course, local of and is, state opinion of This power obscenity. the on is clarified limits and 1819 place affirmed to since Court governments opinions the majority which in most-cited California, 100 v. the among opinion scoring mn h o 0 otctdmjrt pnossne1819. since opinions majority most-cited 100 top the among nta ae h aoiyoiinafimda niiulsrgtt u eiinso eea fi- federal of recipients sue to right individual’s an affirmed opinion that majority implying the scalar, case, positive a that by In multiplication to up only identified are scores the that Recall oprn hs crswt hs nTable in those with scores these Comparing ial,tecreainbtentesoeadteaeaenme ftmsper times of number average the and score the between correlation the Finally, 15 3 rsnsteoiin ihtetp3 siae aetquality latent estimated 38 top the with opinions the presents erecentralities degree 2 , snot is Chicago , of University v. Cannon ti ehp upiighwsimilar how surprising perhaps is it 2, fteoiin ntecitation the in opinions the of 16 h orhhighest- fourth The Miller 257 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 al 3 Table 258 23 2Fra .Georgia v. Furman 22 8Pep og op .NLRB v. Corp. Dodge Phelps 38 9Bvn .SxUkonNmdAet...17 1. 1971 Ohio v. Lockett California v. Chapman . . Agents. Named Unknown Six 21 v. Bivens 20 19 7Ahadrv ense alyAtoiy13 1 1936 Authority Valley Tennessee v. Ashwander 37 8Ednsv Okla. v. Eddings 18 6NAPv lbm xrl atro 981 1958 Patterson rel. ex Alabama v. NAACP 36 7NwYr ie o .Sullivan v. Co. Times York New Wainwright v. Gideon Jordan 17 v. Edelman 16 15 5Jhsnv Zerbst v. Johnson 35 4Bonv or fEducation of Board v. Brown 14 4Wosnv ot Carolina North v. Woodson Kelly v. Goldberg Wade 34 v. Roe 33 32 3Mno .Pape v. Monroe 13 1RdLo racsigC.v FCC v. Co. Broadcasting Lion Red 31 2Bclyv Valeo v. Buckley Carr v. Baker Williams 12 v. Dandridge 11 10 0Uie ttsv ooyVcu i o 901 1940 Co. Oil Socony-Vacuum v. States United 30 Brewer v. Morrissey 24 akName Rank 9Tryv Ohio v. Terry States United 1 v. Katz States United 29 v. Roth 1959 Louisiana 28 v. Duncan Garmon v. Council Trades Bldg. 27 Diego San 26 25 nr Winship re In 9 ahw .Eldridge v. Mathews 8 enlsv Sims v. Reynolds 7 reRR .Tompkins v. R.R. Erie 6 one .Harris v. Younger 5 ilrv California v. Miller 4 iad .Arizona v. Miranda 3 rg .Georgia v. Gregg 2 hvo,UA n.v RC Inc. NRDC, v. Inc. USA, Chevron, 1 P h 8ms nunilcssaogtetp10ms-ie ae ic 1800 since cases most-cited 100 top the among cases influential most 38 The aris dl har .Slaton v. I Theatre Adult 921 1972 971 1967 911 1941 981. 1978 921. 1982 941. 1 1964 1963 961 1976 991 1969 941. 1974 941. 1954 901 1970 981. 1938 901 1970 981 1938 932. 1973 963 1966 947 1984 931. 1973 911. 1961 961. 1 1976 1962 901. 1970 961. 1976 941. 1964 912. 1971 965 1976 971 3 1967 3 94 103 1.32 1.33 1972 1973 981. 1968 1. 1957 981. 1968 erSoe#CtsCites/Year Cites # Score Year ...... 6183 118 36 8144 3 104 38 96 41 28 1 88 02 3102 180 03 9103 130 39 49 4 91 44 5133 153 05 8114 161 48 7192 159 07 0 5 3 207 92 5 51 4153 155 54 69 3 91 16 39 3 97 13 7143 134 57 49 2 97 14 78 2 87 17 8103 100 58 6124 4 132 131 76 78 4103 100 84 9143 144 89 2192 189 92 3194 129 03 4114 131 04 8256 225 28 1 1 113 2 3193 149 73 2269 266 82 1196 129 21 29 2 97 22 2 89 29 5173 3 127 3 155 25 107 25 26 ..Ptye al. et Patty J.W...... 8 2 4 7 6 3 4 1 4 2 2 2 6 2 9 6 7 7 7 9 4 1 8 6 9 8 8 4 5 4 1 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 hs eotdi Table in reported those scores. of “plateau” large relatively a is there 4, or 1946 since cases most-cited 100 Top Sample: scores. of correlations 1800 since cases most-cited 100 Top Sample: scores. with correlations crso hs ae ntetosmlsi Table the in and samples samples two two the the in in respec- cases opinions between, these post-1946 correlations of 100 scores two the following of between ranks the (relative) similarity are the strong tively, telling of a Most scoring indicates rankings. the Inspection two as does. the opinions, opinions additional most-cited including of 204 to ranking the relative invariant the be that should implies opinions method opinions the scoring most-cited the 204 underpinning the axiom of IIA The analysis the from score 1800 opinion’s since that identi- of are values rank ’46” the “Post Tableto after the in ranks: rendered reported two those opinion and to Each scores cal 1946. two by since contains accompanied rendered that is period opinions 1819. 1946 time most-cited since entire the 100 opinions for top majority opinions most-cited the most-cited of 204 set smallest top the the is This for scores the calculated We 1800 Since Opinions 204 Top IIA: Probing 3.3 subsequent of rate annual average the both as with well associated positively as are citation. citations Table scores subsequent and in of scores, reported lower number correlations have the to of tend discussion opinions the in drawn conclusions in reported scores the in 19th only ranks yet 1946, Table since rendered most-cited 100 5 Table 4 Network Table a in Vertices Scoring Precedent: of Quality Latent the Measuring fter“ult”(esrda nuneo iaiiy,uigntokctto data. citation network using citability), basis or the influence on as 1800 since (measured opinions “quality” majority their Court Supreme of all score we chapter this In 4Conclusion method. the identifies that This IIA 1946. of since supposition opinions the for most-cited support 100 provides top agreement the for respectively, scores, and ranks eodvsa npcin ti sflt eosdrtecreain nlgu to analogous correlations the reconsider to useful is it inspection, visual Beyond aho hs orltosidct a indicate correlations these of Each 2 hspithglgt etr ftesoe nbt als fe h o 3 top the after tables: both in scores the of feature a highlights point This . Intersample Descriptive eaiet h nlgu crsfrteoiin edrdatr1946 after rendered opinions the for scores analogous the to relative hs r ipae nTable in displayed are These 1. ero’ orlto Coefficient: Correlation Pearson’s orlto #Ct/er Score): Cite/Year, (# Correlation orlto #Cts Score): Cites, (# Correlation pamnsRn Correlation: Rank Spearman’s Score): (Age, Correlation h Fl”vle,peetdi Table in presented values, “Full” The 2. very togareetbtente(relative) the between agreement strong 5. 4 n lsl ofr othe to conform closely and 6 correspond , older 1: −0. 0. 0. 0. 0. 995 981 849 425 466 259 . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 ...... al 6 Table 260 13 ot aoiav Pearce v. Carolina North 39 31 24 nvra aeaCr.v NLRB v. Corp. Camera Universal 42 32 32 er .Ohio v. Terry 27 33 42 iesv i nnw ae gns 910 1971 . . Agents. Named Unknown Six v. Bivens 22 24 52 orev Pape v. Monroe 23 25 62 a ig lg rdsCuclv amn15 0 1959 Garmon v. Council Trades Bldg. Diego San 25 26 72 ri .Boren v. Craig 24 27 83 anrgtv Sykes v. Wainwright 31 28 93 odo .NrhCarolina North v. Woodson 36 29 03 ohv ntdStates United v. Roth 33 30 33 etnv Maryland v. Benton 34 23 34 53 azv ntdStates United v. Katz 32 35 Post’46 Full Rank 62 o .Wade v. Roe 29 36 12 rw .Bado Educ. of Board v. Brown 21 21 21 ienv Wainwright v. Gideon 18 22 81 ae .Carr v. Baker 17 18 02 umnv Georgia v. Furman Valeo v. Buckley 26 13 20 19 09Mnl .Dpto o.Servs. Soc. of Dep’t v. Monell 9 10 21 ahw .Eldridge v. Mathews Okla. v. Eddings 12 15 12 11 31 rigo egt .Mto os e.Cr.17 1 1977 Corp. Dev. Hous. Metro. v. Heights Arlington 11 13 67Eemnv Jordan v. Edelman California v. Miller Ohio v. Lockett 7 19 16 20 15 14 71 e okTmsC.v Sullivan v. Co. Times York New 16 17 hvo,UA n.v RC Inc. NRDC, v. Inc. USA, Chevron, 1 1 rg .Georgia v. Gregg 2 2 annv nvriyo Chicago of University v. Cannon 4 3 iad .Arizona v. Miranda 3 4 0Dnrdev Williams v. Dandridge Wash. v. Strickland 10 6 8 7 enlsv Sims v. Reynolds 8 5 4I eWinship re In 14 9 one .Harris v. Younger 5 6 30 oprn crso ot14 ae fl ape 0 otctdoiin ic 1800) since opinions most-cited 204 sample: (full cases 1946 post of scores Comparing 91.08 0.79 1957 Mills Lincoln v. Union Workers Textile Name reRR .Tompkins v. R.R. Erie 910 1951 970 1957 990 1969 980. 1968 960 1976 970 1977 960 1976 945. 1984 930 1963 990 1969 970. 1967 964 1976 992 1979 920. 1972 962 1966 901 1. 1970 1984 981 1978 961 1976 31.27 1.03 1973 920 1962 960. 1976 940. 1954 910 1961 erScore Year 930 1973 981. 1938 31.40 1.13 1982 941. 1.26 1974 1.04 1978 941 1964 941. 1964 911. 1971 901. 1970 ulPost’46 Full ..Ptye al. et Patty J.W...... 80 21.05 82 10.98 81 00.95 80 81.04 88 71.21 87 61.18 86 41.15 84 31.17 83 21.07 82 21.01 82 77.52 67 01.28 90 81.06 78 71.08 77 35.14 23 41.12 94 31.25 93 42.62 04 91.74 19 01.38 00 91.34 99 81.49 98 32.72 03 71.50 17 29 41.47 14 41.58 14 31.49 13 11.50 11 21.65 02 41.62 34 11.89 31 1.10 ... Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 o etaiy;isedorsoeuiie h crso h other the of scores the utilizes influenced score potentially our instead centrality); tor rlt n erecnrlt.Oedfeec sta u esr osntuiiethe utilize not does measure our that is of difference cen- score eigenvector One as centrality. such degree networks and connected trality partially for measures centrality other a is directly it been influence—specifically, never have that objects axiomatic compare an data. utilize to the we in designed necessarily compared data is is our that network of nature it—our method incomplete predates scoring the that with opinion deal different that To opinion a problem incomplete. an the by that with cited and faced opinions be are other cannot we of numbers scale heterogeneous common cite a opinions on majority opinions such all placing In 6 Network Table a in Vertices Scoring Precedent: of Quality Latent the Measuring omrt JI,Kt ,Zle 20)Lwa emeswb oprsno aiu network various of Comparison web? seamless a as Law (2009) J Zelner paper, Working D, Katz precedent. II, Court MJ Supreme Bommarito US of depreciation The (2010) II JF Spriggs RC, Black References n en h atta easm httecmuiyo case a of community the influence that potentially assume could we that that cases fact the being ing of influence expected the vertex. represent equal hypothetical scores community that our on a then vertex with vertices, each inter- vertex possible following all hypothetical the of a set to were the leads there to This if interact. scores: to our potential of the pretation have not between did relationships” that “influence vertices impute we data community and network observed sw eiv tpoie sflnwmaueo oecnrlt htgeneralizes that centrality node of measure centrality. new in-degree useful data of Addition- network a concept case. incomplete the provides of the types it of other believe topic to the we method by by scoring as only also our not but apply to considered determined hope was be we case ally, to a structure which community in allow year the to intend we work future 94 oe .BnfiilIdsra onCr.14 0 1949 Corp. Loan Industrial Beneficial v. Cohen Slaton v. I Theatre Adult Paris Brewer v. Morrissey 45 43 39 38 38 37 Post’46 Full Rank 13 tn .Powell v. Stone Seldin v. Warth 35 28 41 40 ersnain fteUie ttsSpeeCutcru 19Ð05.I:Poednso the of Proceedings In: (1791Ð2005). corpus Court Supreme States United the of representations Louis Saint in University Washington h crscluae yti ehdaeaaoost esrso network of measures to analogous are method this by calculated scores The h nlsspeetdi hscatri rlmnr,wt novosshortcom- obvious an with preliminary, is chapter this in presented analysis The (Continued) s ncmuigtecnrbto flink of contribution the computing in Name s ,or {w ssc,i udmnal ifr from differs fundamentally it such, As . metric vertex : s w) (s, i ossso l ftecssta rdt t In it. predate that cases the of all of consists , ∈ E ˜ } ngnrtn siae fthe of estimates generating In . s v) (s, to erScore Year v 950 1975 930. 0. 1973 1972 960. 1976 ssoe(si eigenvec- in (as score ’s w i rcleto of collection or , ulPost’46 Full htcudhave could that . . 60.93 0.98 76 77 50.91 75 41.01 1.10 74 75 x i using 261 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 262 adM,Soe ,ScsA(01 ewr nlssadpltclsine nuRvPlSci Pol Rev Annu science. political and decisis. analysis stare Network vertical (2011) of A dynamics Sacks K, information Stovel The MD, (2011) Ward A Stenger Research T, Empirical Hansford the II, for J Center Spriggs database. Court Supreme States United The (2012) University HJ Washington paper, Spaeth Working contests. from Scoring (2012) 1950. EM in Penn cited K, Court Schnakenberg Supreme California the what authority: 44(1):61 of Polit authority York Sci The New Polit Wiley, (1954) PS John JH future. behavior. Merryman the choice to Individual back (1958) science: RD political Luce Econ in Law Networks J (2011) analysis. D empirical Lazer and theoretical a Princeton precedent: Court. Legal Supreme (1976) US RA the Posner WM, on Landes precedent of politics The York (2006) New II Press, University JF Oxford Spriggs law: precedent. TG, the of Hansford power and The analysis Network (2008) (2007) MJ PJ Gerhardt Wahlbeck 30(1):16Ð30 S, Netw Jeon Soc II, precedent. JF Court Spriggs Supreme TR, net- of Johnson authority political JH, The Fowler in (2008) Causality S (2011) Jeon B JH, Fowler Sinclair Polit JF, J Padgett Am DW, space. Nickerson doctrine MT, in Heaney JH, opinions Supreme Fowler Court of Supreme content Locating the (2010) controls B Who Lauderdale (2011) TS, Clark G Vanberg AD, Martin B, Friedman C, Carrubba 14:245Ð264 Louis Saint in University Washington paper, Working Louis Saint in University Washington Law, the in Louis Saint in 6(4):613Ð673 Rev Law Stanford 19(2):249Ð307 Princeton Press, University 15(3):324Ð346 Anal Polit precedents. Court Supreme of importance legal the measuring 39(2):437Ð480 Res Polit Am works. 54(4):871Ð890 Sci 56(2):400Ð412 Sci Polit J Am opinions? York, Court New ACM, ICAIL’09. law, and intelligence artificial 234Ð235 pp on conference international 12th ..Ptye al. et Patty J.W. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 miia Analysis Empirical atIII Part Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 DOI (eds.), al. et Schofield N. e-mail: Whiteley P. e-mail: Sanders UK D. C043SQ, Colchester Essex, of University Government, of Department ..Clarke H.D. USA Texas-Dallas, 75241, e-mail: North TX of Dallas, University Blvd., Sciences, University Life 7300 and Arts Liberal of Division Borges W. e-mail: Stewart M.C. USA Dallas, 75083-0688, e-mail: at TX Texas Richardson, of University 830688, Sciences, Box Policy PO and Political Economic, of School ..Cak ( Clarke H.D. pro- the companies, manufacturing publicized and widely firms massive, investment of of banks, foundation heels major the the on of constitute close bailouts that Coming security state. commit- welfare personal public modern and the longstanding health challenging education, while to disadvantaged ments the pro- aid Many threaten measures integrity. that austerity fiscal grams differently—proposed long-term things see establish citizens and ordinary loads address to debt necessary are sovereign cuts there budget unsustainable elites sector political public substantial and that sub- economic agreement Among for broad is reductions. need debt the and over cuts democracies budget Western stantial many prompted in have turbulence debates economic national ensuing acrimonious and crises financial 2008 in Beginning Sanders, David Whiteley Stewart, Paul C. and Marianne Borges, Walter Clarke, D. Harold Cuts Spending Public Towards British Attitudes Modeling Austerity: of Politics The 10.1007/978-3-642-35239-3_13 [email protected] [email protected] [email protected] [email protected] [email protected] · B .Sanders D. ) · ..Stewart M.C. dacsi oiia Economy Political in Advances · .Whiteley P. bnzrSroet hrt olco,1851 Collector, Charity to Scrooge Ebenezer operation? in Aretherenoprisons?...Andtheunionwork pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , , oss r hystill they are houses, 265 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 oe eutosi ulcsco pnigtrae ooetr h distributional the overturn to democracies. mature threaten contemporary spending in sector consensus policy public in reductions posed 266 lsi nlsso ulcspotfrpltclprisadterlaes n here and of nature leaders, the their investigate first and We parties preferences. policy political for for models support similar develop public we of analyses in ples reforms. the excluded con- implement have recent to Lansley, from how midwives Andrew on and ferences Secretary, plans nurses Health his physicians, his representing to groups profes- and reactions professional negative medical Cameron such by NHS, of criticism the repercussions by for compounded political been the about has Fearing skepticism hospi- moves sionals. Public its such investors. private Service, of and benefits Health physicians the National to the facilities other of and ownership tals and management devolve to being widespread, increasingly surveys. is opinion fast” (Bloomberg public too media in far, news below, “too the are in deficit cuts both the the enunciated clear that to sense Osborne years A seven George lion. take Chancellor will cuts and sector projected (Werdigier public the than that less anticipates been now (Burns). percent has eight exceeded growth unemployment and Economic percent five nearly was rate tion advisors Society”. his Big and “the Cameron termed that follow- in- responsibility by the and services devolution-of-power mitigated public for bono—a provide be offered pro would would who raises volunteers workforce civic-minded government annual by the participation percent creased in one Reductions of years. two prospect for ing frozen the was with pay employee years, broad Public a two 2011). (BBC chopping programs and over “unnecessary” benefits, of spending curtailing range government jobs, government from 490,000 percent eliminating 20 by (Burns Os- of years George average four Exchequer, 2010, next an the May the cut of in to Chancellor power his proposed and assumed borne, Cameron gov- Democrats David coalition Liberal Minister his Prime and when unemployment Conservatives growing of and ernment debt public rising of bination citizens. its for education re- and and care government health effective small of of provision agenda the than ideological privatization neo-Thatcherite pri- higher a un- accords enacting civil government to Coalition prompted ority Conservative-led have the cuts that tur- public-sector charges political and proposed rest ongoing Kingdom, result and United The protests the crisis. In public debt moil. violent, sovereign oftentimes country’s widespread, dra- the been Greece dictated remedy has have is to Fund case Monetary policies known International financial best the conian the and Union Perhaps painful European benefits. necessitate the and where will implemented services that have public cuts Kingdom—either in budget large-scale reductions United contemplating the seriously and are States or United the Spain, gal, so hswiig eea countries— several writing, this of As oesicroaigdmgahc tiuia n vlaievralsaesta- are variables evaluative and attitudinal demographic, incorporating Models attempt Government’s Coalition the than resistance more prompted has Nothing infla- UK the 2011, of end the slow—by been has goals these towards Progress com- pernicious a with Confronted experience. British the on focuses study This 2011 .Tepoetdlvlo pnigrdcin snwfly£2 bil- £123 fully now is reductions spending of level projected The ). 2011 .Tepa a ordc h ugtb 8 billion £83 by budget the reduce to was plan The ). rln,Iay ree Portu- Greece, Italy, Ireland, alia, inter 2011 n,a ilb documented be will as and, ) ..Cak tal. et Clarke H.D. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 oenetsedn mn riayctzn.Sbeun tde oue o on not focused studies Subsequent for support citizens. gov- reduce ordinary may in among disconnect benefits; this spending and that government cost suggested between Downs they not. relationship what is direct because there and a ernments, outputs contribute is they and there what markets, inputs between In policy those receive. relationships may of sustain direct there evaluations see to that citizens’ cannot speculated assumed between people (1960) be gap Downs benefits large must ago, the a years that of be 50 burden weighing Over tax careful services. by the and benefits determined and is provided state services welfare and the for support of government to responses for belief-system context and cultural spending ideological intervention. political study, general government the this the provides in on In that cuts than framework response. towards benefits—rather that response—attitudes and of services specific specifics on a inter- the government on and general focus economy support we national that the attitudes in between vention connection weak a only vlain n rfrne.BreadVea ( those Viegas of and impacts Borre political preferences. the and and preferences af- evaluations per- factors policy investigate and and We evaluations state. formation welfare policy the modern fecting the analyzes undergird that that work values of previous sistence from study our distinguish We Perspectives Theoretical 1 British the from data cross-sectional using Study’s cuts Election budget the towards attitudes public Cuts Spending Public Towards Attitudes British Modeling rc h yaiso pnosaotsligtecii n atr htacutfor account to that used factors are and data crisis the survey solving Monthly dynamics. about these Brothers. opinions 2008, Lehman of October dynamics of in the failure begins track analysis the This after about crisis. opinion month financial public the the dynamics performance solving the overall investigate of the also likelihood on We heavily the leaders? more their focus and they parties to- do ex- of attitudes or an conditions, cuts, as economic spending on cuts the weight the wards more intention towards place vote attitudes voters rival of Do estimate factor. importance planatory we relative Prime analysis, the and this determine Conservatives to of the models part for As support proposed Cameron. affect the David attitudes towards Minister these these attitudes how answer public To assess examine support? and we electoral cuts context, of British losses the major and in face propose questions that cuts leaders such and implement parties governing to will try are “no”, reductions is spending answer public the that If assertions necessary? leaders’ Lehman their accept of states crisis. welfare failure financial ern the the of onset public since the recovery of dramatized 2008 economic dynamics September in of the Bank likelihood analyze Brothers the to about (CMS) Surveys monthly opinion the Monitoring from data Continuous use atti- also BES We demographics, evaluations. incorporates economic model and The beliefs attitudes. tudinal/policy these of model multivariate a ale eerhhsrie usin bu hte nidvda’ vrl level overall individual’s an whether about questions raised has research Earlier h rpsdbde usps rsigpltclqetos ilctzn nmod- in citizens Will questions. political pressing pose cuts budget proposed The 2011 lentv oe(V eeedmSre.Te,w specify we Then, Survey. Referendum (AV) Vote Alternative 1995 aeosre htteeis there that observed have ) 267 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 ilb rm eemnnso hs attitudes. those evaluations of economic determinants that plausible prime is be it will Government, Cameron’s Mr. by proposed has 1930s. that the settlement of since democracies political-economic Depression Western Great in the discourse the of political mainstream essence group-level of the contours well-being and the is personal defined individual- are This affect of needs chances. significantly impact life basic can the that and ensure mitigating conditions to economic time net in same safety variations the a provide at im- lives, while help citizens’ to met, expected of are quality democracies the mature prove in systems political and governments te,mr aia,vie ntergtbaea nu oeg okr n growing Brown. and workers Gordon foreign and influx an Blair blame Tony right of the on governments voices radical, Labour more New Other, prac- previous profligate partners, the the coalition to of attributable their is and tices problem Conservatives the that the are contend part, Democrats, and Liberal their its amassed the of For been many blame. has of to debt economies massive who the that unsure are exercised as are recession Citizens “double-dip” partners. a trading of verge the on is omy o oiisi prpit eas,a onegadCak ( Clarke and Kornberg as because, support appropriate Prime analyze is to of policies models Government oriented for cuts Coalition electorally Extending spending Conservative-led Cameron. policies—the the David government Minister by of 2010Ð2011 set in a concen- we towards instituted Kingdom, attitudes United on the in instead cuts trate budget to reactions public analyzing When aainotnso oe upr o pnig(itradMouritzen and and (Winter spending spending social for are the between support state “price” lower connection welfare show that the often the of questions taxation of respondents Survey burdens reminding underestimated. tax by are the benefits but that idea citizens, the by on recognized rather but disconnect, the 268 clrgm n t ntttos n o pcfi uhrte moida individual as embodied authorities specific (Easton governments for polit- incumbent and the or for officeholders institutions, whole, its a and as regime community political ical the for phenomena—support three on spending. public in cuts large-scale policy immediate, specific of a therapy” through “shock ad- measures approach—the The austerity study crisis-induced this austerity”. of but “permanent imposition condition, called the general he dresses long-term what this accentuate of may wel- period situation the a and current entering net were safety social and the state, to fare commitments previous fund to difficult ( it Pierson finding ago decade a Over context. crisis Newton and (Confalonieri same the 1995 largely remain preferences policy general while lree al. Lewis-Beck political et (e.g., on evaluations Clarke egocentric effects do than stronger behavior have voting economy, and attitudes national and in the literature, contemporaneous voting of retrospective, economic evaluations i.e., the prospective in evaluations, facts economic stylized sociotropic to behavior. counter which and runs attitudes this political that affect notes may He programs state welfare of consequences hndlnaigfcosta fetpbi tiue oad h pnigcuts spending the towards attitudes public affect that factors delineating When oeso oiia upr nmtr n mrigdmcaisuulyfocus usually democracies emerging and mature in support political of Models a in cuts spending for support studying are we that emphasis bears it Moreover, individual-level the to responses that suggests (2007) Kumlin review, recent a In ). 2004 ). one u htmn onre are countries many that out pointed 1993) Circa 1965 al 02 h rts econ- British the 2012, early onegadClarke and Kornberg ; 1992 aeobserved, have ) ..Cak tal. et Clarke H.D. 2001 ,even ), 1992 1988 ). ; Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 ostsyidvda neet,ntt hwsldrt ihasca ls owihthey which to class social a with solidarity show belong. issues to economic not interests, on individual focusing satisfy are to voters Increasingly, does. economics but elections, hieadoinain oad addtsadise eg,Cmbl tal. et Campbell (e.g., issues al. and et electoral candidates Clarke cuing factors towards powerful orientations most and the of choice one is identification party that testifying heuristics. image of leader models and undercut partisan effects emphasize instead demographic that their choice that considerations and voter argues parties be policy also among might of differences Cutler and who coalitions. basis similarities support voters the observable informed simple, on on studies best choices back In the fall electoral even models. make that support to finds party expected he in elections, variables Canadian demographic of other and class social this In government. Coalition the by economic (2006 pursued current Dalton being the regard, policies by austerity reinvigorated the be and will crisis divisions al. class et of (Clarke relevance established political well is Britain in port Stokes and Butler (e.g., politics significance, electoral 1969 British the in concerning class social arguments of longstanding dominance, indeed to back hearken shares” “fair become mature contemporary may in insecurity vote” democracies. “economic financial the enhanced of components and important equality increasingly income rising ( and that Stegmaier piece, tended and Lewis-Beck a Lewis-Beck performance e.g., of regard, economic (see, this not of support In is aspects political economy which for the about most course, debates matter Of protracted well. been as have trouble make there Voters in and incompetence. parties are incumbent of trouble leaders signal in their is clear economy a the when is and economy attributions responsibility perform- weak is a government deciding the whereas that for well, indicates The guide economy ing strong useful democracies. A extremely ballot. mature one’s simple, cast in a to how provides agenda it electoral fundamental; is the economy dominate deficits. typically the issues and debt lence the public worsened simmering that of banks are context British costs a of in care bailout made health the being over attendant are anger per- and arguments the these expanding when of era rapidly All an is exploding. in net people made safety elderly social commitments best comprehensive of the a its centage for afford despite ago can that, century government a argue half British others over century Still 21st Britons. sys- hardworking no benefit of intentions, the expense exploit the who at miscreants miscellaneous tem and immigrants of numbers Cuts Spending Public Towards Attitudes British Modeling el nti ead eea ikacpac/vrinoinain a erelevant be may as orientations work acceptance/aversion at risk be general may regard, heuristics Other this cuts. In budget well. explana- towards meaningful attitudes provide people’s heuristics of leader tions and partisan whether is question nent ulr(2002 Cutler h atragmn spolmtcsneteei neomu oueo research of volume enormous an is there since problematic is argument latter The and insecurity financial inequality, income of significance the about Conjectures va- related and economy the that argued have long voting economic of Students .Atog h rwn ekeso oilcasa rdco fprysup- party of predictor a as class social of weakness growing the Although ). 2004 saogtemr eetvie ttn h aefrincluding for case the stating voices recent more the among is ) , 2009b a rudta oilcasn ogrmtesmc nmost in much matters longer no class social that argued has ) ei-eke al. et Lewis-Beck ; 2000 2008 2004 e loBartels also see ; .I h rsn td,teperti- the study, present the In ). , 2009b 1988 ,i spsil htthe that possible is it ), lree al. et Clarke ; 2008 ,hv con- have ), 2004 1960 269 ). ; Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 eosrtdadepce efrac nsc susd uht rv electoral drive to much do issues such choice. on performance parties’ expected of and assessments Voter’s agenda. demonstrated political the dominate typically al. issues tional, et Adams (see inspired security. he personal whom and national sys- promote educational that policies Pace and and with care environment clean well, health a as effective tems, issues affordable, valence favoring important majorities other sustainable massive are vigorous, there with However, coupled growth. unemployment pre- and economic people inflation of of is- numbers rates valence overwhelming The low with point. opinion, ferring economy, ideal the public common is divide a issue share that valence consensus—voters canonical reform opinion system strong participating a electoral manifest of or sues desirability War the Iraq trade-offs, the spending in taxation-social as such issues oe rw nSoe’cneto aec sus( issues valence of concept Stokes’ on draws model for pain “short-term as justified being gain”. are that long-term cuts government towards attitudes to 270 vlain bt ontv n mtoa)a h otipratcmoet of components important most the Fiorina as (e.g., choice emotional) political and cognitive (both evaluations the Voter in school American al. Columbia The et the and (Lazarsfeld Berelson 1950s models Lazarsfeld, and socio-demographic of 1940s studies from voting draw towards the We attitudes in intentions. of rooted voting model and composite cuts a spending specify to the (2004, choice al. electoral of et models Clarke peting by demonstrated been has 2009b models such al. for major justification et in tistical Lewis-Beck used routinely (e.g., in- are studies models behavior election Composite political model. of one explanations of different researcher corporating the beware should also economists individuals averse wager. risk the take whereas to forward, unwilling going be effects will beneficial have will cuts the lree al. et Clarke edriae rvd oefladprioiu xlnto feetrlchoice. electoral of explanation parsimonious and and powerful partisanship a issues, provide valence images on leader performance party of assessments al. with et Together deliver to (Clarke ability outcomes their policy important, infor- more desired lack and, who preferences voters policy for devices parties’ heuristic about influential mation al. as serve et they (Sniderman that pro- in cue partisanship similar voting valence are accessible time and in powerful point any a at vides ancestor, Arbor Ann its like However, h aec oiismdlpriasi a yai rpris(lree al. et al. (Clarke et properties McCutcheon dynamic (Campbell and has identification Clarke partisanship party model Michigan of politics venerable valence stability the the the Unlike stressed images. that model leader party and variables—partisanship h aec oiismdla riuae yCak ta.( al. et Clarke by articulated as model politics valence The npriua,w r neetdi aec oiismdl fpryspot The support. party of models politics valence in interested are we particular, In ogaoS.Toa qia andt eaetemno n ok Political book. one of man the beware to warned Aquinas Thomas St. ago Long on (1957 Downs .Ti steapoc etk nti td,asmln aibe rmcom- from variables assembling study, this in take we approach the is This ). 2009a eei paribus Ceteris n h ayavctso pta oeso at competition party of models spatial of advocates many the and ) ei-eke al. et Lewis-Beck ; Cmbl tal. et (Campbell 2009 1981 e loFiorina also see ; ikacpatpol ilb iln obtthat bet to willing be will people acceptant risk , Lewis-Beck ; 1944 2005 1960 2004 eesne al. et Berelson ; ,Soe otne htvlne o posi- not valence, that contended Stokes ), 2011 2008 n rmmdl htpsteconomic posit that models from and ) , 2009a dstoohrmjrexplanatory major other two adds ) n nteBiihcnettesta- the context British the in and ) 1981 1988 ui n McCubbins and Lupia ; 1963 Achen ; ). 1954 , 1992 1992 1991 2004 ,fo h oesof models the from ), .Ulk positional Unlike ). Franklin ; .Lae images Leader ). , 2009b ..Cak tal. et Clarke H.D. 1960 e also see ; 1998 1992 2004 ,in ), ). ). ; Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 1 a responses. questions, measure budget the to of used the three was towards In scale agree-disagree questions. five-point analyses—attitudes five to the responses using for constructed variable cuts—was dependent principal The Cuts the for Support Public 2.1 Specification Model 2 of below. sources presented Both analyses Britons. the 1,000 for approximately used of are data opinion monthly and polit- regular beliefs the a attitudes, (CMS)—measuring ical conducts Survey 18,556 Monitoring also Continuous and BES survey—the wave The by internet wave. pre-referendum out post-referendum the carried the completing being completing internet respondents fieldwork national 22,124 representative with a with 2011, to panel, administered May were and waves survey April Two YouGov. in conducted was vey Sources Data Cuts Spending Public Towards Attitudes British Modeling h uswudsrnte rti cnmcly n n ttn httect would cuts the that stating one that and stating economically, one Britain cuts, strengthen the would about cuts statements the two between choose to respondents asked rv,Dsprv,Srnl iapoeo o’ know. Don’t or disapprove Strongly Disapprove, prove, debt. Britain’s of cause main the is financial spending serious public cause Excessive to likely UK are Ð proposes the Government of the that health expenditure long-term public in the cuts for The essential Ð are expenditure public in cuts Government’s The Ð h fhqeto a oddthus: worded was question fifth The h usinfra o h rttrecmoet ftedpnetvral a sfollows: as was variable dependent the of components three first the for format question The ifiute o eadm family. my and me for difficulties economy. odpbi evcscnb rvddol ytegovernment. the by only know. provided Don’t be encourage can Ð to services more public do Good and services Ð funded publicly provide to less do view? should your government to The closest comes Ð statements following the of Which reces- into further it know. pushing Don’t by economy Ð Britain’s damage will cuts expenditure public The com- international the Ð and of growth economic impact Britain’s overall strengthen will the cuts expenditure about public view The your Ð to closest cuts? expenditure come public statements proposed following the of Which h orhqeto stated: disap- question fourth nor The approve Neither Approve, approve, Strongly between choose could Respondents statements: following the of each with disagree or agree you far how indicate Please epet rvd evcsfrthemselves. for services provide to people sion. petitiveness. h rts lcinSuy(E)sA altRfrnu Sur- Referendum Ballot AV (BES)’s Study Election British The 1 h orhquestion fourth The 271 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 gcnrceautosbt ersetvl n rsetvl.Tefco nlssof and analysis sociotropic factor The elicit prospectively. questions and retrospectively The evaluations both scales. evaluations economic Likert egocentric five-point measures on routinely via questions cog- BES constructed four The measuring with evaluations, analysis. variable economic factor predictor personal exploratory a and an contains national of cuts evaluations the nitive towards attitudes of other model would than cuts the for support sick, less the people. manifest that predicted would spending we caregivers because Similarly and and seekers. disabled job reduced for pub- were opportunities the benefits fewer find because meant re- cuts both would caregiver harsh, unemployed long-term cuts were with spending who those lic those and that cate- ill predicted single and We a disabled sponsibilities. into permanently unemployed the long-term with and gory short- combining vulnerability, we and Wales. and for whole, prediction a such coefficient no as negative made UK a We produce analysis. the multivariate would is the Scotland of in than consid- resident proclivities be a ideological being to that its considered hypothesized in is left particular in more Scotland erably emerged. differences regional pro- if social funded publicly of might set also extensive an educated for highly need grams. more the the to prediction. trickier sympathetic but a more income, be be to with proved correlates anticipated education but we often increase, increased, opposed Education to income be cuts As the bands. to for 14 non-whites support in expected that measured we “white was group, Income and social cuts. 1 the vulnerable as to a scored As minorities capture 0. with to as identifications, British” effort British” race “white an and into in ethnicity dichotomized age, other was respondent’s Ethnicity and age. a also of We of effects them. square curvilinear oppose en- possible the and we age the variable, old age, favor new of For and a vulnerability conservative cuts. computed the more the recognize might favor be responsibility they to might or less likely people cuts, have more older be possibilities; traditionally would two who care, tertained men, family mea- was that and age child expected and for We variable dummy years. 0Ð1 in a was sured Gender bands. income and education with estimated was cuts the regression. towards squares of attitudes least nature of ordinary continuous model the our scores Given factor variable, variable. use dependent dependent we the the and indicated as questions, analysis analysis five this the factor by to produced answers exploratory cuts structures the economic components factor that single UK’s principal a believed the that A respondent solve harmful. the to that be scores needed indicated high would were scores produce cuts low to whereas rescaled that problems, were thought questions respondent op- five know” the the “don’t a to when with Responses government, well. the as whether from about tion services respondents fewer asked did or question they more fifth say favored A to they choose. respondents to allowed option category which middle know A not recession. into UK the push 272 seswee h cnm samjrcnenfrms iieso h K Our UK. the of citizens most for concern major a is economy the elsewhere, As status workforce of effects the evaluate to variable dummy a included also We determine to Wales and Scotland for variables dummy included also model The ethnicity, age, gender, for measures demographic included variables Predictor ..Cak tal. et Clarke H.D. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 npa,wihw eintda upr o lcoa eom,spotfrtraditional for support reforms, electoral for support as designated we which play, in oa oenet hudhv oeauthority. more representation. have proportional should to governments changed Local members. be 600 should Ð to system reduced electoral be The should Commons Ð of House The Ð ycmuigtenme fpstv nwr iu h eaieoe.Teresult- The ones. negative the from minus ranges was answers index positive that scored ing of eight were number of then the field Respondent computing a labels. negative by from and words positive four between in to equally economic divided question up general select a the about could using feelings Respondents their elicited, situation. describe also to asked were were responses. respondents economy structures which the factor evaluation towards economic reactions single Emotional a that indicates items these Cuts Spending Public Towards Attitudes British Modeling 2 oad eomwr esrduigsvnqetosi h Vrfrnu post- referendum AV pre-wave. Attitudes the the cuts. in in one questions the and seven support survey using wave to measured likely are were more membership reform EU be towards opposing would in those who membership that individuals expectation of conservative the disapproval included with or also Union, approval We em- respondent’s European conservatism. was the a ideological to this measured general and priority that accused, of variable giving the measure a contrasts of proxy rights that a the scale as protecting ployed 11-point to respondents opposed similar ask as surveys a crime BES fighting on the placement for regard, opt choose this incorporated to In to variable us ideology. led this which of but variables, measures anchors, attitude alternative opposing other with the intertwined taxation as preferences increased cuts policy scale. using tax the scale on and 11-point 4.3 an spending of indicated on mean and data measured a The is with adverse, often them. risk ideology taking disliked slightly Left-right liked are really average really person the person a on a Britons indicated that indicated 0 10 where and scale risks 11-point vari- risk taking an The on model. the measured in was included able were also reform political as towards attitudes such and initiatives policy major of about consequence feelings cuts. since and spending ef- model sector cause this identification public in both party images likely other leader are the party the the of leaders for include direction coefficient not the negative did We predict a fects. not and Liberal did cuts We and the Party. Conservative for Labour the support between and correlations identification the positive Democrat as enacted predicted and served we proposed Non-identifiers cuts, coalition variable. the the dummy Because category. single mi- identification a various party in with reference placed Identifiers Party. were variables Liberal Labour parties partner opposition identification nor minority principal coalition party the the dummy and Party, Conservative Democrats created leading We coalition the above. for described choice toral eco- both For associated economy. cuts. be the would the for scores of support positive greater view increasingly with that positive predicted very we a variables nomic representing 4 and response epnet eeakdt vlaesvnsaeeto v-on ietscales: Likert five-point on statement seven evaluate to asked were Respondents ikacpac/vrin etrgtielg,attdstwrsE membership EU towards attitudes ideology, left-right acceptance/aversion, Risk elec- of models valence from drawn variables several includes also model The −4to ,with +4, 2 atraayi niae he atr were factors three indicated analysis Factor ersnigavr eaieemotional negative very a representing −4 273 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 ftetremjrpris(ai aeo,E iiadadNc lg)a elas well as Clegg) Nick and Miliband Ed leaders Cameron, (David the parties about major feelings three the measuring of variables incorporates model politics valence fmr btatisepoiiyvralstpclyepoe nDownsian-type al. in et employed Adams typically (e.g., manifestation variables models concrete spatial issue-proximity a constitutes abstract it more quality, of pro-con voting its analysis the Given in regression model models. cuts separate intention a spending in variable the explanatory an in above—becomes variable variables described dependent Attitudes the membership. cuts—the as EU the well to towards as support/opposition and reform, political ideology political The left-right economy. towards measuring the attitudes to included reactions model emotional beliefs and of evaluations measur- cognitive variables comprised ing model vari- conditions vulnerability economic The and above. region described income, ables gender, model de- ethnicity, sociodemographic education to The age, was intentions. the here voting included purpose explain best Our models parties. which other termine and Democrats Liberal Labour, vatives, regression. OLS with The category. estimated or reference was Liberal the best model constituted Labour, affect was best” vari- Cameron was Conservatives, party” party This “no which the country. know saying “didn’t for Persons they the that parties. dummies facing other 0Ð1 issue miscellaneous four and important Democrats as most measured the was on of also evaluations able We best model, like”. was politics “really valence party 10 the to which from dislike” variable “really or predictor 0 Conservatives another from di- measured the included ranging were for scale was Cameron 11-point vote David variable an to Leader using Conservative dependent intention about respondent’s the Feelings party. a model, another of terms intentions in voting chotomized Conservative the For Cameron David About Feelings and Intentions Voting 3 cuts. the associated for for be support support would diminished whereas which orientation with cuts, “left” the progressive a for proxied support reforms power with and electoral correlate institutions government would traditional of power for devolution of support devolution that the anticipated for We Westminster. support from away general and institutions, British 274 hteeypryspretg fsasi alaeti h aea hi ecnaeo h vote. the of percentage their as own. same its the on is govern Parliament can know. in it Don’t seats so of vote Ð percentage the party’s half every than That more get Ð important: party more one was That statement seats. which Ð their designate for to again respondents run the and asked resign question should pre-wave manifesto The party issues. the important against decide vote to who referendums MPs more needs Ð status. Kingdom its United keep The should England Ð of Church The abolished. be Ð should Monarchy The Ð eas siae eiso ia oeso oigitninfrteConser- the for intention voting of models rival of series a estimated also We 2005 .Fnly sprtedsuso bv,the above, discussion the per as Finally, ). ..Cak tal. et Clarke H.D. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 xlntr oe fvroscmeigmdlseictosdsrbdaoe we the above, in McKelvey described interested and specifications were McFadden model we calculated competing Since various category. of base Conserva- power the with explanatory as models serving logit intentions multinomial Democrats voting using Liberal tive estimated Labour, were for parties” intentions “other Voting and procedures. logit binomial using country. best the deemed facing party issue and important most identification the party on measuring variables dummy several the 1 Fig. Cuts Spending Public Towards Attitudes British Modeling h odtosfcn h onr.Te loaedvddaottecueo h cri- the Fig. of as about Specifically, cause recovery. sanguine the to about not path divided are policy are the Britons also and many They sis country. that the shows facing data conditions survey the BES 2011 May The Cuts Budget the to Reactions Public 4 rmalfieidvda oeswl efr etrta n niiulmodel. individual variables any predictor than better the perform incorporating will models model previous individual composite on five based all a images addition, from that leader In rivals. expect and other we identifications and model’ research, partisan ‘cuts issues, pure per- valence a party outperform of will incorporating range Ac- model a choice. politics electoral on valence of formance the calculus that the in hypothesize cri- element we a secondary, cordingly, such but address in substantial, to even a designed are policies that, to valence sis is reactions the voters’ hypothesis whether crisis, Our economic or power. of voters, explanatory times of greater preferences provides political model the politics for account largely cuts model. each by predicted correctly intentions voting of percentage the etb h aorPrydrn t euei fc,wt 2pretdisagreeing percent 32 with office, in tenure its during Party mismanage- Labour to cuts the spending by for necessity ment the attribute percent) (49 respondents the h osraievru l te ate oigitninmdl eeestimated were models intention voting parties other all versus Conservative The fpriua neeti h oe oprsn swehrattdstwrsthe towards attitudes whether is comparisons model the in interest particular Of pnosaotct npbi xedtr.Suc:21 E-M Vrfrnu survey referendum AV BES-CMS 2010 Source: expenditure. public in cuts about Opinions R 2 s kieIfrainCiei AC and (AIC) Criteria Information Akaike ’s, 1 lutae,ams afof half almost illustrates,

B&W IN PRINT 275 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 as fBiansdb rdcsi otniu rpsto;pbi gemn and agreement the public was proposition; spending contentious public a excessive is disagree- Whether produces while percent. debt percent, Britain’s 19 of 54 to cause to 26 from 41 fallen from has risen ment has difficulties personal serious cause geigta h usaeesnilt rti’ cnmchat a alnfo 68 from fallen has Fig. health (see economic 50s Britain’s high to the essential to percentage percent are the cuts frame, time the this that During 2012. agreeing con- January surveys and 2010 Study June Monitoring between Continuous ducted monthly the in contained questions tical would who the percent only 29 services. said to government compared percent fewer for services, themselves—55 feweropt public for provide good fend provide should to respon- can government individuals of government the on majority that a rely proposition But and the it. services damage of the will skeptical strengthen cuts were will the dents cuts believe percent spending respon- 43 the personal and believe and economy Britons services of percent of Thirty-six provision sibility. government between balance sophical the followed that crunch institutions. credit financial the major executives of towards and bank meltdown investments for Britons speculative bonuses of over over anger controversies presided government the recent who the in about manifested by persist as provided stories banks, News bailouts percent their opin- banks. massive 36 in British over but division stabilize anger debt, this to for widespread the explanation be of possible cause may A the ion uncertain. was were 27 spending Specifically, and cause. public disagreed, the that was agreed spending excessive percent public 37 whether on agreement be. of would els cuts they the said of impact percent personal Thirty-five the disagreed. for what difficulties know percent not cause 20 did would only Government percent whereas Forty-five Cameron economy. households the the cuts their of that of policies agreed health the percent) long-term that (47 the believed many preserve as to Almost essential cuts. were the caused policies Labour that survey 2 Fig. 276 rcn h yaiso hs pnosoe iewsacmlse sn iden- using accomplished was time over opinions these of dynamics the Tracing Figure lev- lower were there debt, public Britain’s of cause the of assessments Regarding pnosaotbde usadpbi evcs ore 00BSCSA referendum AV BES-CMS 2010 Source: services. public and cuts budget about Opinions 2 umrzsdt nattdstwrsepniuect n h philo- the and cuts expenditure towards attitudes on data summarizes .I otat gemn httect r ieyto likely are cuts the that agreement contrast, In 3). ..Cak tal. et Clarke H.D.

B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 ctadadavne g erae upr o h us h eaiecoefficient negative in The residing cuts. caregiver), the for disabled, support unemployed, decreased age (i.e., advanced support vulnerability and positive Scotland but produced gender, cuts, male the and for education, and income Higher signed. sup- of modest more showed partner, port Conservative coalition balky the occasionally with Conservatives’ support, the of party expected, predictor As powerful variable. a identifiers dependent be the in to variation proved the identification of percent 51 for counts h utvraemdlo tiue oad h usseie 1peitrvari- predictor 21 significant specifies statistically cuts are the 18 towards which of attitudes ables, of model multivariate The Cuts Budget the Towards Attitudes Analyzing 5 disagree. than spending debt excessive British that the think of to cause continue main the people seven-point is a more and but agreement disagreement, for range for point range 10 a across fluctuated has disagreement surveys BES-CMS 2012 2006ÐJanuary June Source: 3 Fig. Cuts Spending Public Towards Attitudes British Modeling −0.52. otscodmgahcvralswr ttsial infiatadcorrectly and significant statistically were variables socio-demographic Most ( B yaiso ulcoiinaotct npbi xedtr coe 08Jnay2012. 2008ÐJanuary October expenditure public in cuts about opinion public of Dynamics = 0.17) ( B = aorietfir eesrnl gis h us ihacoefficient a with cuts, the against strongly were identifiers Labour . 0.58) uprigtect.Ietfir ihteLbrlDemocrats, Liberal the with Identifiers cuts. the supporting (p ≤ .05) (Table .Temdlac- model The 1). 277

B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 † *** Adjusted estimates) (OLS survey 1 Table 278 yohszdrsls h rae soesspotfrtaiinlinstitutions—the traditional for support one’s is greater the produced The fight- the also crime reform results. support prioritize institutional and to hypothesized towards ties attitudes likely EU of less its dimensions sever three were to The Britain accused ing. for membership the calling those EU of were rights support than the cuts who emphasize people who Thus, those expected. and as performed kinds various the from security their to threats perceive indeed do cuts. elderly the that suggests age for N dcto .018 Education at Identification: Party variable Predictor cnmcEautos.264 Evaluations Economic osat.4 .061 .045 Constant .074 Beliefs: Ideological/Policy Conditions Economic Reactions Emotional tnct 09.024 .022 .067 .009 9.334E-05 .012 Wales Scotland Income Gender Ethnicity Squared Age Age Status Vulnerability Orientation Risk nmlyd iald care-giver disabled, Unemployed, tiue oad Reform: Towards Attitudes UMembership EU Accused of Rights v. Crime = p iial,ms tiuia aibe apn etrgtpltcloinain of orientations political left-right tapping variables attitudinal most Similarly, Labour osraie.581 .174 System Electoral Party Other Democrat Liberal Conservative rdtoa ntttos.091 .063 Institutions Traditional Power of Devolution 18, ≤ . 001; 556 R utvraemdlo pnosaotct npbi xedtr,My21 BES-AV 2010 May expenditure, public in cuts about opinions of model Multivariate 2 = ** . 51 p ≤ † . 01; * p ≤ . 5 n-aldtest one-tailed 05, − − − − − − − − − B .061 .518 07.020 .986 .007 09.238 .009 .032 .004 .073 .023 *** *** *** *** *** * *** *** *** * ** *** *** *** *** *** *** *** ..Cak tal. et Clarke H.D. 2.406E-05 s.e. .004 .007 .021 .016 .015 .019 .002 .011 .004 .002 .019 .002 .005 .002 .006 .006 .006 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

690 689 688 687 686 685 684 683 682 681 680 679 678 677 676 675 674 673 672 671 670 669 668 667 666 665 664 663 662 661 EDITOR’S660 659 658 657 656 655 654 653 652 651 PROOF650 649 648 647 646 645 vr smaue ytevrossmaysaitc rsne ntetbe t per- its table, the in presented How- statistics models. summary specific various the the of any by than measured better as does Tableever, model research, model”, composite earlier cuts the Echoing “pure that rivals. the uments remaining here, the interest of of significantly model all does other outperforms model The opposition competitors. politics various its valence than for the better measure, and diagnostic Conservatives every for By intention parties. voting predicts best model Table in rized issue. that important judgments include ( predictors a Democrats strong produces Liberal Other reform Conservative. voting electoral identifica- of Party for probability Conservative Support and increase). Conservative) (.38 voting high- tion of their to probability lowest in their increase from Fig. shifted (see cuts variables. the value dummy towards would est of attitudes Conservative case voting as of the .44 probability by in predicted increase zero the or that variables mean showed their continuous procedure at This of predictors case other the holding significant in while statistically values range as their also over Conservative we varied voting variables, were of predictors independent probability various of in effects changes the estimated of strength the about intuition 34preto h oeitnin n eeae McKelvey a classifies generates correctly and intentions model vote composite the This of model. is- percent logit 93.4 important binomial most a the using handle intentions to able best are is as party statis- predictors, which are significant on sue. variables are judgments predictor terms the prime 24 of identification the of all party about 17 feelings All for in significant. The coefficients variance tically Cameron. of the for sup- estimates affect in issue, and of percent important predictor minister, 62 most strongest explains the the model is on cuts composite best spending are public Conservatives identification for a party port that (with After cuts. judgment proponent spending the chief public and the the Cameron, of Osborne) Minister George Prime Chancellor about feelings and intentions nlzn upr o h osraie n ai Cameron Table David and Conservatives the for Support Analyzing 6 cuts. is the there for Finally, support increased cuts. with heightened the correlated with relevant, being to are acceptance opposition characteristics risk personality with general associated more reform is that for evidence system support expected, electoral also the as Again, of component vision. key Society” a devolv- “Big individuals, Cameron’s support and Mr. who governments of local those the to for power supporting obtains government more relationship of ing same likelihood The the cuts. is spending greater public Parliament—the and monarchy church, Cuts Spending Public Towards Attitudes British Modeling h nlsso h efrac frvlmdl o oigitnin ssumma- is intentions voting for models rival of performance the of analysis The sntdaoe eetmt h fet ffcosafcigCnevtv vote Conservative affecting factors of effects the estimate we above, noted As 2 eot h eut fetmtn opst oeso osraievote Conservative of models composite estimating of results the reports mn h v optn pcfi oes h aec politics valence the models, specific competing five the Among 3. .Ohrpwru rdcosaefeig bu aeo (.79 Cameron about feelings are predictors powerful Other 4). 3dces)o aor( Labour or decrease) −.33 1dces)aebs ntemost the on best are decrease) −.31 R 5dces nthe in decrease −.45 2 f.7 oprovide To .87. of 3 lodoc- also 279 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

736 735 734 733 732 731 730 729 728 727 726 725 724 723 722 721 720 719 718 717 716 715 714 713 712 711 710 709 708 707 EDITOR’S706 705 704 703 702 701 700 699 698 697 PROOF696 695 694 693 692 691 aibenticue nmodel in included not Variable Ð McKelvey ecnaeCretyClassified Correctly Percentage ai aeo,My21 E-Vsurvey BES-AV 2010 May Cameron, David 2 Table 280 ‡ † *** Age N Gender elnsAbout: Feelings Constant Income Education Status Vulnerability Beliefs: Ideological/Policy Reactions Emotional Evaluations Economic Important Most Best Party Wales Scotland Ethnicity Identification: Party pnosAotPublic About Opinions variables Predictor L ersinmodel regression OLS model logit Binomial UMembership EU rm .Rgt fAccused of Rights v. Crime Reform: Towards Attitudes Issue: = p oEooi Conditions Economic to ai Cameron David lcoa System Electoral Labour rdtoa Institutions Traditional Labour Clegg Nick Miliband Ed xedtr us.604 Cuts Expenditure osraie2 Conservative eouino Power of Devolution Party Other Democrat Liberal Conservative Party Other Democrat Liberal 18, ≤ . 001; 556 utvraemdl fCnevtv oeitnin n elnsaotPieMinister Prime about feelings and intentions vote Conservative of models Multivariate R 2 = ** p . 87 ≤ . 01; * p ≤ = . −1. −1. −1. −2. −1. −1. 5 n-aldtest one-tailed 05, osraievt intentions vote Conservative Bs −. −. −. −. −. −. −. −. −. −. −. −. −. 93...... 012 091 358 479 516 968 027 080 004 020 079 115 685 704 235 140 152 065 053 887 867 022 776 088 098 628 774 4 *** ** *** * ** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** * *** ...... 974 .059 ...... 152 019 051 183 045 030 003 040 047 044 129 1. 412 Ð 099 Ð 255 Ð 129 1. 140 110 134 Ð Ð 022 Ð 021 025 1 4. 310 191 157 017 083 032 . e . † s B elnsaotDvdCameron David about Feelings −. −. −. −. −. −. −. −. −. −...... 6 .051 .007 067 003 166 3 .064 033 0 .015 005 206 006 027 309 089 287 216 204 709 222 356 351 573 1 .051 .005 .011 161 153 010 003 187 007 Adjusted *** *** *** *** ** *** *** * *** *** *** *** *** * *** *** *** *** ..Cak tal. et Clarke H.D. R 2 = . 62 .046 .021 .097 .065 .029 .001 .014 .016 .016 .011 .019 .052 .093 .046 .046 .057 .061 .050 . e . ‡ Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

782 781 780 779 778 777 776 775 774 773 772 771 770 769 768 767 766 765 764 763 762 761 760 759 758 757 756 755 754 753 EDITOR’S752 751 750 749 748 747 746 745 744 743 PROOF742 741 740 739 738 737 h ust M epnet eeqiebaih ihteraeaesoebig4.0 being Fig. score From in average illustrated ahead. their as year with Moreover, to scale. the bearish, likely” the over quite “very on resolved were meaning be respondents 10 CMS would and outset, crisis unlikely” the the to team “very that BES (CMS) meaning likelihood 0 the Surveys the with forecast 2008, Monitoring scale October Continuous 0Ð10 in monthly a Starting use in 1930s. respondents the asking of downturn began economic depression growing global serious great the most the symbolized the of 2008 since onset 15th the signaled September and on crisis Brothers financial Lehman of failure Britain? The in Bears Many so There Are Why 7 proximate principal the are intentions. considerations vote un- politics of finding drivers valence This model. that politics point valence the the derscores than better marginally only is formance model logit binomial in predictors icant 4 Fig. Cuts Spending Public Towards Attitudes British Modeling elt!pol r ers eas h cnm si erbesaeadsoslit- shows and shape terrible in is economy the Conservative- because the bearish are 3.0. Since reality!—people exceeded 2.1. never has only score is average the 2012, office, January took Coalition in Democrat score Liberal average the that such htaetesucso hspsiim ipease oti usini:“It’s is: question this to answer simple A pessimism? this of sources the are What hnei rbblt fitnigt oecnevtv soitdwt ttsial signif- statistically with associated conservative vote to intending of probability in Change 5 hi esms a rw vrtime, over grown has pessimism their ,

B&W IN PRINT 281 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

828 827 826 825 824 823 822 821 820 819 818 817 816 815 814 813 812 811 810 809 808 807 806 805 804 803 802 801 800 799 EDITOR’S798 797 796 795 794 793 792 791 790 789 PROOF788 787 786 785 784 783 nesn 2002) Anderson, ‡ model logit multinomial † for defined Not Ð 3 Table 282 aur 02 ore coe 08Jnay21 E-M uvy n N unemployment ONS and surveys BES-CMS 2012 2008ÐJanuary data October Source: 2012. January 5 Fig. Composite oi-eorpis.04 Socio-demographics model) logit (multinomial intentions voting party Model other and democrat liberal Labour, B. Panel .17 Politics Valence Attitudes-Cuts .10 Beliefs Political .04 Conditions Economic Socio-demographics model) Model logit (binomial intentions voting Conservative A. Panel cnmcCniin .07 Conditions Economic oiia eif .14 Politics Valence Attitudes-Cuts Beliefs Political Composite edriae,pryietfiain at eto otipratissue important most on best party identification, party images, and leader (Burnham performance model better indicate values smaller Criterion; Information Akaike ieiodo ovn nnilcii etya n nmlyetrt,Ags 2008— August rate, unemployment and year next crisis financial solving of Likelihood ia oeso oigitnin,My21 M survey CMS 2011 May intentions, voting of models Rival ‡ ‡ .38 .77 .76 McFadden .72 .70 McFadden .25 R R 2 2 .29 .06 .54 .87 .17 .86 McKelvey Ð Ð Ð Ð Ð Ð McKelvey R R 2 2 72.3 64.6 80.9 94.4 93.8 69.0 classified correctly Percent 53.8 44.7 51.6 88.2 87.2 66.7 classified correctly Percent ..Cak tal. et Clarke H.D. 19052 12309.00 16479 17708 AIC 36445 32513 35014 10634 11344 AIC 28238.60 4506 4829 † † ...... 15 82 42 06 93 06 82 98 B&W IN PRINT 37 86 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

874 873 872 871 870 869 868 867 866 865 864 863 862 861 860 859 858 857 856 855 854 853 852 851 850 849 848 847 846 845 EDITOR’S844 843 842 841 840 839 838 837 836 835 PROOF834 833 832 831 830 829 r aaeest eetmtd ie h tutr ftemdl t aaeesare parameters its model, the squares. of least structure nonlinear the using Given estimated estimated. be to parameters are eovdoe h otcmn eris: year forthcoming the over resolved is 1930s. Keynes the over rehearsing Hayek risks privileging ill-advised recipe view, an a their pursuing In on insist agenda. numbers and economic are opposite Miliband, neo-Thatcherite times their hard Ed because the be Leader, that need Labour claiming than reminder by part, worse countered a have their includes Balls, for news Ed Chancellor, And, political Shadow tough. daily his Govern- for the indeed Coalition passes are austerity what the times for of that by most need advanced result, attendant a proposal As an policy ment. every and George virtually economy Exchequer, of ailing Coalition the touchstone the of Democrat made Conservative-Liberal Chancellor have the and Osborne, of Cameron Minister 2010 Prime May Third Government. since economy. faltering presence the the un- the is start and jump massive is Second a to hikes. of program tax government easing and Labour quantitative former precedented cuts the spending by of announcement mixtures 2009 news bad varying March of for litany unrelieved need largely a the delivered about have speeches which budget 2011 annual and Osborne’s specify 2010 George we 2009, Chancellor Here, in is Centers. first Jobs The at factors. dynamic lines such in of evolve three length should the they in run, movements long explain- with the for over equilibrium relevance but forecasts, transient these have in may as variation factors rate ing jobless to Other the able process. and be correction crisis should error the we solving an then for economy, forecasts the between of relationship the prospects model future and state present the negativeing strongly is rate unemployment being the election and 2010 crisis (r the the between correlation to solving The run-up of 2011. percent of the expectations 8.4 much in to throughout movements 2008 decreases upward October by modest Fig. followed in with regard, percent 2012, this 5.8 January In from in are. risen effectively has times is unemployment hard rate—it UK unemployment how that the of as car- definition mind indicator public operational single the no an in perhaps weight regard, the much this that as In about soon. ries sanguine anytime than resolved less be are people will economic many crisis about that news surprising bad not of is barrage it continuing conditions, a Given reviving”. of sign any, if tle, Cuts Spending Public Towards Attitudes British Modeling etrt;BUDGET rate; ment RESOLVE where: ingovernment; tion ( =− 1 h eutn oe fpbi oeat o hte h cnmccii ilbe will crisis economic the whether for forecasts public of model resulting The assess- for heuristic” “big the as unemployment use typically people fact, in If, − .67) L) RESOLVE . ε = t (t) = = oeatfrrsligteeooi rss UN crisis; economic the resolving for forecast tcatcerrterm error stochastic nulbdes QE budgets; annual = b − + 0 + λ b 3 1 b ∗ ∗ 1 QE UN(t ∗ ( 1 (t − − − L) i) 1) UN(t = + 0,σ (N( + uniaieesn;COAL easing; quantitative b 4 b ∗ 2 − ∗ COAL i) BUDGET 2 ) − , t α (t 1 stm,and time, is ∗ − i) RESOLVE (t + − ε(t) i) = 5 α documents unemploy- , (t = b − Coali- and 1) 283 (1) λ Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

920 919 918 917 916 915 914 913 912 911 910 909 908 907 906 905 904 903 902 901 900 899 898 897 896 895 894 893 892 891 EDITOR’S890 889 888 887 886 885 884 883 882 881 PROOF880 879 878 877 876 875 eiulDiagnostics: Residual etg ftevraini ulcfrcssfrrsligteeooi rss(ad- crisis economic the resolving for forecasts public in justed variation the of Tablecentage in displayed are period u oiia cnmcsok,telretbigterpaeeto aorb the vari- by of Labour effects of the replacement Westminster. adjusted the at being has coalition largest Democrat process Conservative-Liberal the correction that shocks, error 2008, economic the powerful political of is a ous autumn unemployment defining the hypothesized, joblessness Since As and process. crisis. forecasts economic with heuristic, the key of course future the are easing quantitative by and forecasts statements budget economic annual smaller, reduce of to Effects of month. paribus) presence each (ceteris the points effect, worked permanent has far) government Coalition (thus Coalition a the the as of all presence Specified noteworthy. the be Coalition especially of would impact crisis is the The economic year. by the next that Labour the likelihood over the of resolved of replacement forecasts public the lower and to worked easing, quantitative of round suffi- point being full joblessness one in nearly increase by cent also forecasts per Unemployment lower one to rate. a cient unemployment between with effects, the relationship short-term and correction nearly large crisis error of has rate the the a by resolving at month for eroded subsequent is forecasts source, each whatever in from percent system, the 75 to forecasts shock a these is that between parameter nifies mechanism adjustment correction the unemployment, error monthly the and of power the of Indicative *** Adjusted 2012 2008ÐJanuary October ahead, year in solved 4 Table 284 N osat4. Constant Election General 2010 Easing Quantitative 2009 Statements Budget 2009Ð2011 Rate Unemployment Mechanism Correction Error hnei nmlyetRate Unemployment in Change variables Predictor = p Heteroskedasticity: omlt:Jarque-Bera Normality: RH B 7. = LBQ ARCH: uoorlto:LBQ Autocorrelation: eut faayigtemdluigdt o h coe 08Jnay2012 2008ÐJanuary October the for data using model the analyzing of Results vrl,temdlpoie asmnosacuto ulcfrcssabout forecasts public of account parsimonious a provides model the Overall, 2009 March the budgets, annual expected, As well. as play in are factors Other 39 ≤ . R 001; R re- be will crisis financial the whether opinions of dynamics of model correction Error 2 5and −.15 2 = = ** 4 n l aaeesaesaitclysgicn n rprysigned. properly and significant statistically are parameters all and .64) . 64 p ≤ . (t 3,df 339, 3 epciey u ttsial significant statistically but respectively, −.43, 01; χ − 2 1) = = * = p -ECM 9. = 5. ≤ . 6,df 967, 7,df 573, 1,df 119, 2 p 12, (t . 5 n-aldtest one-tailed 05, − sson h oe a con o ag per- large a for account can model the shown, As 4. 1) = = = = . 834 2 p 12, ,p 1, ,p 6, = = = . . 751 529 . 619 ( −.95) −. −. −. −. −. −. B 712 426 146 743 172 289 954 nte01 scale. 0Ð10 the on *** *** ** * *** *** *** −.74, p<. p<.001) < (p ..Cak tal. et Clarke H.D. 0.Ti sig- This 001. . −.71 .744 .102 .149 .087 .071 .103 .022 s.e. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

966 965 964 963 962 961 960 959 958 957 956 955 954 953 952 951 950 949 948 947 946 945 944 943 942 941 940 939 938 937 EDITOR’S936 935 934 933 932 931 930 929 928 927 PROOF926 925 924 923 922 921 ofiec ntento’ blt osleteeooi rsshsbe aln as falling been has Furthermore, crisis employees. economic fired the the solve by to served frustrated ability either be nation’s by to the kindly used in regarded who be confidence those to or unlikely seekers are pres- mounting and job Sustained put programs enough. cuts serious relief job is sector on cause enough public sure say by will propelled to services unemployment people of public many levels high lead slashing economic may that long-term families Britain’s view for for difficulties trending essential upward Although economy. as an real perceived the health, in widely results are visible of currently lack cuts a by undermined be may tually delivery. care health Thatcherite effective a of privileging expense the are at Lansley and agenda Cameron Andrew Minister ideological is Secretary Prime cuts if Health asking maligned spending are they much deep Now his of time. right pursuit the single-minded at policy Since a right the the whether Service. in asked commentators Health have power, rein National press to to British cherished came attempt Coalition the Democrat patients restructure Conservative-Liberal concerned the to and plans groups government interest in professionals, medical eco- as continuing and services and benefits disappearing of malaise. face nomic could which the concerns in equity-fairness raise traction others present gain Still the fast”. reject too of far, others Many but “too possibility. restraint, as latter fiscal program of the need to the important regard acknowledge re- most with responses Open-ended the suggestive these about election. are survey country prove general referendum the AV next will facing BES the and issue the to forward in run-up question going the the policy to in effective sponses liability in as political conceived seen a well be be however not to cuts, may the outcomes 2011, mean policy and environment of 2010 economic uncertainty in- changing the economy required a Second, the infrastructure effectively. and government damage services cripple could could public cuts cuts provide the to the that and that it, several fear curing on of skepticism substantial stead mounting is face there benches government First, the fronts. on friends his and on govern- shrink to schemes into first and head ment. education jump to of willingness lack the and income—mitigate aging evaluations vulnerability—unemployment, that economic factors economic But identification, self-sufficiency. suggest indicate party that to factors demographic tied and is reactions, Support and cuts problems. budget fiscal and Coalition’s endemic budgetary address the of to for need seriousness the the and difficulties acknowledges debt that sovereign opinion successfully has public Cameron British date, To tomorrow. tapped sidekick pay- hamburger (Popeye’s demands a minister promises anti-Wimpy prime and the the today tomorrow), ment As payment the tomorrow. for pain today of gain hamburger trade a tiger of that sought cuts the promise spending ride the public with to enact to today chosen attempting have while colleagues times economic Coalition poor his and Cameron David Politics Performance and Crisis Economic Conclusion: 8 Cuts Spending Public Towards Attitudes British Modeling nlsso M iesre aasgetta ulcspotfrtect even- cuts the for support public that suggest data series time CMS of Analyses enunciated been has recently skepticism of element pointed, more third, A Camer- Mr. cuts, the toward attitudes in divide partisan strong a is there Although 285 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 1012 1011 1010 1009 1008 1007 1006 1005 1004 1003 1002 1001 1000 999 998 997 996 995 994 993 992 991 990 989 988 987 986 985 984 983 EDITOR’S982 981 980 979 978 977 976 975 974 973 PROOF972 971 970 969 968 967 ulrF(02 h ipetsotu fal oidmgahccaatrsisadeetrlchoice. O, electoral and Borre characteristics sociodemographic In: all: protest. of shortcut tax simplest or The (2002) revolt F tax Cutler spending: and Taxing (1995) K Newton voter. MA, British Confalonieri the and politics Performance (2009b) P Whiteley MC, Univer- Stewart Oxford D, Britain. Sanders HD, in Clarke choice Political (2004) P Q Whiteley MC, Opin Stewart Public D, revisited. Sanders identification HD, Clarke party of dynamics The (2009) A McCutcheon States. HD, United Clarke the York and Canada New Wiley, choices: voter. political Making American (2009a) The TJ (1960) Scotto York A, DE Kornberg New Stokes HD, Martin’s, WE, Clarke St. Miller Britain. PE, in Converse change A, Political Times Campbell (1969) NY path. DE austerity Stokes in D, faith Butler and optimism’ ‘can-do for Colchester pushes Essex, leader of British University (2011) study. E JF referendum Scarbrough Burns Ballot AV O, (2011) Borre Study In: Election economy. British the in intervention Government (1995) JM Viegas Dec view. O, Borre quickly: so not perhaps but austerity, Chicago Princeton Cameron’s Press, age. Chicago needs of gilded U.K. University Voting. new (2011) (1954) WN the Bloomberg.com McPhee of PF, Lazarsfeld economy BR, political Berelson the democracy: Uni- Cambridge Unequal competition. (2008) party of L theory Bartels unified A (2005) B Grofman III, the S breaking Merrill regression: J, Adams linear and variables, demographic psychology, Social (1992) CH Achen References economic of era present unfolds. the policies ways as austerity democracies predictable elec- mature and in understanding other hardship for and model Britain important the in remains animate choice politics toral to Performance ahead. expected performance years be leader the and can in party parti- making of mutable are considerations. evaluations with voters general politics coupled that more valence economy the the of and of attachments force state san evolving the the negated to support, not reactions party Rather, have on effects attitudes attitudes substantial Differing these at- exerting are performance. partisan but measures leaders, party austerity party of harsh the of assessments about images global with di- more economic context respond and place broader than voters tachments in British Rather them, policy election. around and general conditions hardships next the the to reflexively the in and be support not rectly will party cuts of spending drivers the toward sole attitudes that indicates model intention vote as viewed increasingly is solutions. an that short-term problem seem to a may intractable to cuts response spending harsh public and point, unjust some inappropriate, At rise. to continues unemployment 286 crruhE(d)Tesoeo oenet xodUiest rs,Ofr,p 121Ð148 64:466Ð490 pp Polit Oxford, J Press, University Oxford government. of scope The (eds) E Scarbrough Cambridge Press, University Cambridge Oxford Press, sity 73:704Ð728 Toronto Press, Toronto of University 6:A10 Oct 234Ð280 pp Oxford, Press, University Oxford government. of scope The (eds) perhaps-not-so-quickly-view.html 11: Princeton Press, University Cambridge Press, versity 14(3):195Ð211 Behav Polit research. voting in triangle iron ial,tefc htvlnepltc aibe omc odietecomposite the drive to much do variables politics valence that fact the Finally, http://www.bloomberg.com/news/2011-12-07/u-k-needs-cameron-s-austerity-program-but- cesd2 e 2012 Feb 24 Accessed . ..Cak tal. et Clarke H.D. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1058 1057 1056 1055 1054 1053 1052 1051 1050 1049 1048 1047 1046 1045 1044 1043 1042 1041 1040 1039 1038 1037 1036 1035 1034 1033 1032 1031 1030 1029 EDITOR’S1028 1027 1026 1025 1024 1023 1022 1021 1020 1019 PROOF1018 1017 1016 1015 1014 1013 itrS orte E(01 h epewn oehn o ohn:terl fasymmetrical of role the nothing: for something 30:B8 want Nov people Times Why NY (2001) measures. PE austerity Oxford Mouritzen extends Clarendon, S, Britain Winter politics. slowing, growth Electoral Its (ed) 57(2):368Ð377 (2011) D Rev Kavanagh J Sci Werdigier In: Polit politics. Am Valence competition. (1992) party DE of Stokes psy- models political Spatial in (1963) explorations DE choice: Stokes Oxford and Press, Reasoning University (1991) Oxford P state. Tetlock Welfare to RA, the of Brody need politics PM, they new Sniderman what The learn (1993) (ed) citizens P can Pierson revisited. dilemma: voter democratic The American (1998) The MD (2008) McCubbins A, HF Weisberg Lupia H, Norpoth Rev WG, Annu Jacoby outcomes. MS, electoral Lewis-Beck of determinants Economic (2000) M Stegmaier MS, Macmillan, Lewis-Beck Palgrave elections. presidential French (2011) of E University Bélanger R, democracies. Nadeau western MS, Pearce, Lewis-Beck major and the Sloan elections: Duell, and choice. Economics people’s (1988) The MS In: Lewis-Beck (1944) H Gaudet In: B, evaluations. Berelson performance PF, representative and Lazarsfeld preferences, a policy in values, state: support Welfare political The (2007) community: S and Kumlin Citizens (1992) 14:297Ð309 Behav HD Polit identification. Clarke party of A, Press, dynamics University Kornberg the Yale and Measurement elections. (1992) C national Franklin American in voting York New Retrospective Wiley, life. (1981) political MP of analysis Fiorina 12(4):541Ð systems Polit A World (1965) democracy. D a Easton in small too York is New budget Row, government and the Harper Why democracy. (1960) of A theory Downs Washington economic Press, An CQ (1957) edn. A 4th Downs politics, Citizen (2006) RJ Dalton Cuts Spending Public Towards Attitudes British Modeling luin.ErJPltRs39:109Ð143 Res Polit J Eur illusions. Cambridge Press, University Cambridge chology. York New Press, University Cambridge know? Arbor Ann Press, Michigan of University 3:183Ð219 Sci Pol London Arbor Ann Press, Michigan York New Oxford Press, University Oxford behavior. political of handbook Oxford The Cambridge Press, University Cambridge democracy. Haven New 563 287 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 iigsm rdnet h einvtrtermi utpedmnin (Poole dimensions multiple in theorem voter median Rosenthal the and set to uncovered the credence called some space giving policy the of subset the (Kramer in lie which equilibria, Nash Nalebuff were and median (Caplin electoral systems electoral the actual at to 1991 equilibrium regards with Nash to unrealistic pure-strategy necessary and conditions a strong the is difference, there this that reconcile assure to exist. made than not were do more generally attempts with equilibria While spaces Nash into pure-strategy two-party extended Condorcet these when a dimension, However, voters exists one median. there that electoral that showed the meaning models at These choice, point party. a policy for vote non-stochastic certainly and would space policy dimensional .Shfil ta.(eds.), al. et Schofield N. USA 63130-4899, 1027, MO e-mail: Box Louis, Campus St. Hall, Drive, Seigle Brookings Louis, One St. in University Washington Center, Weidenbaum Schofield N. e-mail: Jeon J.S. Drive, e-mail: Brookings USA 1 63130, Louis, MO Saint Louis, in Saint University Washington Economy, Political in Center DOI .MAitr( McAlister K. these (Hotelling in median electoral that the showed to con- generally converge between to It incentive relationship 1929 systems. strong the had two-party on parties in focused cases, parties theory and political formal stituencies in work Early Introduction 1 Schofield Norman and Jeon, Seon Jee McAlister, Kevin Election Canadian 2004 the to Bundles: Applications Party Varying with Elections Modeling nta fpr-taeyNs qiira(N)teeotneitmxdstrategy mixed exist often there (PNE) equilibria Nash pure-strategy of Instead 10.1007/978-3-642-35239-3_14 ). ;Downs schofi[email protected] [email protected] [email protected] 1978 B .Mn ie,ti noee e nldsteeetrlma,thus mean, electoral the includes set uncovered this times, Many ). 1957 ) 1984 · ..Jeon J.S. ie n Ordeshook and Riker ; dm n Merrill and Adams ; dacsi oiia Economy Political in Advances pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , 1999 1973 erl n Grofman and Merrill ; , .Teemdl sue one- a assumed models These ). 1999 Adams ; 289 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 hwn htrtoa ate ilcneg oteeetrlma ie h simple the given of mean property electoral desirable the the line to in yields election converge framework. is will still same spatial model parties and the This rational behavior for time. that voter every polls showing vote of the same theories to the multiple make went with not the voter might in the available he choices times, if the multiple that to insinuates corresponding probabilities This of election. vector a has voter the htamdli hc h niiulvt sntdtriitci otappropriate most is deterministic not is vote individual al. the et which (Schofield in model a that spaces. policy dimensional 2001 290 ec em n h niiulsdsac rmtepryi h oiysae This space. policy the in va- party by the the by weighted from determined is distance is individual’s distance utility the individual and where terms choice, models lence individual of typical with Endersby are deal and models, that (Dow microeconometric with errors associated distributed commonly value errors, extreme Type-I of consider tion to be necessary can are sort. valence and this of elections African- of models kinds of among building outcomes Both valence when the higher is. determining of candidate in is Republican important candidate the Democratic Thus, than for the candidate. voters Republican that vote American the said to for be vote likely to can more are United it voter. they much in than a very example, candidate For of are Democratic individual. the voters characteristics to African-American the individual elections, upon from differs States depends valence valence of a of kind of This kind ability this governing (Penn where perceived an electorate in average an specific, of members the members all election, as all to for interpreted an common party be is in which can takes valence characteristic. and party of underlying electorate, measure other a a some that is on valence position rely Aggregate measures the valence to these exogenous that Both meaning are valence). valence sociodemographic (or of valence types individual and valence) character (or the in (LNE) equilibrium Nash meaning pure-strategy valence. local mean, with a the model is to stochastic mean well convergence electoral govern for demon- joint to condition and the models sufficient developed party that these been and a upon have extends necessary of measure Schofield a ability sort. valence strates this a predicted of with the studies models in or ability years, utilized revealed and past recent the In the to future. in his linked the govern of is in independent to valence is party general, that a has In candidates space. of a policy that a quality valence within of context, sorts location this any In mean for. to accounted taken are is asymmetries valence that given occur not eso ftedcooosrsos oi itiuin hsdsrbto assumes distribution ex- an This is distribution. which logit distribution, response logit dichotomous workhorse multinomial the The the parties. of is the tension models from choice distances individual respective their of to give individuals that sn hsfaeok coed( Schofield framework, this Using h otatbtenteisaiiyterm n h tblt hoessuggest theorems stability the and theorems instability the between contrast The eeteprclwr ntesohsi oemdlhsrle pnteassump- the upon relied has model vote stochastic the on work empirical Recent valence aggregate valence: of types two into divided be generally can Valence .Hwvr hssesa dswt h ho hoeswihapyt multi- to apply which theorems chaos the with odds at seems this However, ). 1998 un tal. et Quinn ; β osatta sdtrie yteaeaeweight average the by determined is that constant a , 1999 2007 .Ti ido tcatcmdlsae that states model stochastic of kind This ). hw htcnegnet h enneed mean the to convergence that shows ) 2003 .Idvda aec sabtmore bit a is valence Individual ). .MAitre al. et McAlister K. 2004 .These ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 rae si spoe nywe h oiinfralprisi qa ozr nall on zero to equal is parties of combinations all different for in run position which parties the are appro- when there in be only when individuals Similarly, longer proven dimensions. those no is of electorate, would it mean coefficient the as convergence electoral in priate, the parties the voters Thus, to all four region. converge of respective first would mean its the party ex- the that fifth above or the mean, mean the while electoral In would overall to. mean the responding electoral to are converge the parties to that convergence centers have ample, electoral parties As the all. account at into made. well be do must not decision will of of they choose kind variety know even this a they sometimes or for resources, where reasons limited regions regions responsive the select in or in run in reasons within to run historical share not to for vote choose run can may its mean They Parties maximize electoral reasons. runs. to Thus, the it wants region. to which this regards it in of with Rather, region voters itself electorate. the position entire to to the appeal it to for for at- trying unreasonable only thus be runs is are only would and parties and it regions these regions, of the both one of However, in regions. one both run in in parties voters to the appeal of to This parties Four tempting five to. regions. with responding country different are a two Imagine parties and complicated. that more mean become electoral quickly can one problem having center. of electoral case the at simplest LNE the a be to is there electoral likely the more of con- variance the the the is, lower distribution of the and definition smaller candidates, the the among are that Given differences is center. equilibrium, valence conclusion strategy the the pure general at at a the PNE exist be coefficient, a to vergence not point of does the non-existence LNE for implies exist a must this that LNE meaning a As equilibrium, mean. in electoral mean electoral the from .Tencsaycniinfrcnegnei fc if is convergence for condition necessary The 2. .We h ieso fteplc pc s2 hntesfcetcniinfor condition sufficient the then 2, is space policy the of dimension convergence sufficient the the a When to and 1. regards necessary with a mean out electoral lays the coefficient: also to He convergence for opposite. condition the indicate values High h lcoa eni oa uesrtg aheulbim(Patty equilibrium that Nash meaning pure-strategy mean, Low local electoral models. a the across electoral is to compared an attraction mean in be strong electoral mean can indicate the electoral value it the this thus to of unitless, values attraction is of coefficient measure This a system. is which c, efficient, osti itiuin coed( vote Schofield stochastic distribution, this the lows for ex- partner Type-I stochas- empirical the the natural follows of a party distribution creates model. assumed a This the model. for voting matching votes tic thus individual distribution, an value that treme probability the that Bundles Party Varying with Elections Modeling nodrt sescnegnet h lcoa eni hscs,oems take must one case, this in mean electoral the to convergence assess to order In in are equilibria Nash local the where question the answers only this However, away position a adopt will party one least at fails, condition necessary the When sn hssaitclfaeokadteasmto htidvda hiefol- choice individual that assumption the and framework statistical this Using iesoso h oiysaeo interest. of space policy the of dimensions ovrec oteeetrlma sc is mean electoral the to convergence 2007 nrdcdteie ftecnegneco- convergence the of idea the introduced ) < 1.

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 e fteeetrtssielpitand point ideal electorates’s the of ber easm that assume We odcaet h lcoaepirt h lcin Let election. the to prior electorate the to declare to h omlSohsi Model Stochastic Formal The 2 longer no is (IIA) alternatives” (Train irrelevant un- met of the because “independence appropriate of longer assumption no is derlying model logit multinomial typical the regions, 292 h aai h pta oe sdistributed is model spatial the in data The of number the maximize gained to be themselves organize can can insight parties some these received. which votes not analysis, in or way this whether the us From to tell country, results. as can the logical they methods, election give of new this methods region these As for new Parliament. one point the in testing in ideal runs seats an gained only is party which regional the party 2004, regional in however, a has Canada 2004. in model voting stochastic the of situations. analysis complex with allow more along will estimation that convergence of of method notions This the Ya- estimated. be see to (VCL: parameters logit regional set allowing choice varying This model. to mamoto necessary logit assumption multinomial IIA basic the the upon distribution rely use value not extreme Type-I will same but previously, the used utilize assumptions will exten- model, an logit model, the mixed empirical estimate the This of to to model. sion the used manner in be for equilibria can similar find conditions to that sufficient a necessary method parameters and a in necessary introduce convergence derived the will show the we be Secondly, also that convergence. can positions will show We electorate policy will previously. complex of used we that more this, vector From a a (LNE). for that first equilibrium coefficient show the Nash determine to local will necessary We a positions. is conditions party order of this vector second Schofield First, any and by bundle. handle defined to party first adjusted same coefficient be convergence the can the for that vote demonstrate all will not chapter do individuals where electorates we paper electorates. complex this structurally in more However, handling regions. for method specific a within convergence,propose adherence analyze only spatial can stochas- and researchers the framework, valence, in current tendencies the party Under about model. gained tic information useful we the coefficient, convergence without the left behind are the- reasoning underlying the the drives with that mechanism problems oretical and methodology empirical utilized currently the nt n orsod otenme fdmnin eetdt ersn h policy the represent to selected dimensions of number space. the to corresponds and finite fprypstos Given positions. party of ial,t lutaeteemtos ewl nlz h aaineetosin elections Canadian the analyze will we methods, these illustrate to Finally, in model vote stochastic the analyzing for methods introduce we chapter, this In aho h parties, the of Each 2011 2003 ilalwfrageaeetmto fprmtr oocrwiealso while occur to parameters of estimation aggregate for allow will ) .Gvnta hr r rbeswt siaino aaeesfrom parameters of estimation with problems are there that Given ). X sa pncne usto ulda space, Euclidian of subset convex open an is j ∈ z P ahvoter each , where , P ={ i n sdsrbdb vector: a by described is stenme fmmesi h sample. the in members of number the is 1,...,j,...,p x i ∈ X z where = (z 1 } ,z i hoe policy, a chooses ∈ 2 ,...,z N ersnsamem- a represents R .MAitre al. et McAlister K. p w ) where , etevector the be z j ∈ w X is , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 tlt fzr n h te tlte r oprdt hspry hs h tlt andby gained utility the Thus, party. this to compared are utilities other for the voting and zero of utility a given between transition the theoretical making the easier. allows model, two also empirical the it corresponding choices, the distri- individual match value in to extreme norm Type-I model the a only assume follows not is errors we This the particular, bution. of In distribution individuals. cumulative among the distributed that commonly be to assumed ouain(..pryquality). party (i.e. population agent a for term valence exogenous where: rpi ru.Hwvr ecnstu hs em orpeetidvdaswt niiullevel individual with individuals represent to terms these up effects. set random can we However, group. graphic identity. sociodemographic of regardless electorate, the of members aigparty having h xetdvt hr andb at ie h oiin fteohrparties. other the of positions the given maximizing party positions, on a of relies by vector equilibrium gained A share of vote definition expected This the the (LNE). of equilibria each Nash where local model the of share. equilibria vote seek maximize We to attempts parties. parties other the of positions 1 Bundles Party Varying with Elections Modeling hoyadbigstegpbtenortertclmdladtecrepnigeprclmodel. empirical corresponding This the and space. model microeconometric policy in theoretical used the our commonly between of 3 are gap that dimension the models each utility bridges latent for and the theory distances with line Euclidian in the model our of places is summation the where 2 aec em enn htti emcnb iwda h vrg seseto a of group. assessment specific average a the of as members viewed the be to can ability governing term party’s this that meaning term, valence itnebtenindividual between distance nti ae,w sueta hstr scmo mn l ebr faseicsociodemo- specific a of members all among common is term this that assume we paper, this In individual utility the paper, the in later all applications among empirical govern the to with quality up parties match the To of assessment average an as conceptualized be can This ntr,w sueta h xetdvt hr o agent for share vote expected the that assume we turn, In easm nti oe htagent that model this in assume We Here, ie h tcatcasmto ftemdl h rbblt that probability the model, the of assumption stochastic the Given o h upsso hsppr hnw akaota qiira erfrt a to refer we equilibria, an about talk we when paper, this of purposes the For z , ρ j u ij a lob enas seen be also can ij ∗ j where (z) (x nofiei oprdt aeparty, base a to compared is office in and i seulto: equal is ,z j u u ) ρ ij ∗ ij u ij steosral tlt for utility observable the is (x (x i ( (x z) z i i ,z ,z ∗ i , = ssi h eaLEif LNE a be the said is , j j z) u ) ) Pr ij ∗ i = = = (x n party and V i λ u u ,z j 1 u j ij ∗ ij ( i j z) β − (x 1 (x ) j (x = sapstv osatand constant positive a is i = β i hc scmo hogotalmmesof members all throughout common is which ,z i ,z λ  ,z j n 1 z j j j j − j 1 ) )>u ∀i . ,u ), chooses + 2 − β( ∈ α j N  x i ij ij = i 2 ρ il  m w (x ij .A snra,w sueti at has party this assume we normal, is As 1. sa xgnu sociodemographic exogenous an is (x i 2 =1 soitdwt party with associated , i ( + i ,z z) ((x z ,z ∀j j α 2 jm l ),...,u ij omaximize to ), , z − ∀l j ∗ x saciia on fthe of point critical a is im = 3 ) ip 2 j h ro term, error The −  (x j . (x i given ,z 1 steEuclidian the is V m p j − ) ( i z) x oe for votes z j i im ,is . ie the given an from gains ) λ 2 j )) V  san is + j ij i 293 ( α by z) is ij j Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 functions oesadte eaeayvco fprypositions, these party simulate of we vector coefficients any MNL relate estimated then the and Using models consider. we model the lodfieasrc oa aheulbim(LE ob etrta titylo- strictly that vector a be to (SLNE) equilibrium Nash maximizes local cally strict a define also oal aiie tteposition the at maximized locally htn at a hf oiinb ml mutt nraeisvt hr.More share. vote so its increase positions to party vector amount a of small is a z LNE by vector a position that formally shift is their may move party equilibrium marginally no Nash they that if local change The would par- position. shares relocate that vote policy and assume their We space how move. policy estimate to This the can choose function. ties would in vote party moves vote-maximizing respective make no its to platform, in its opportunity maximum the local given a that at means, itself locates party each if htteegnausaealnnpstv.Mr ipypt vector, a put, simply More positive. non all meaning non-positive, are is eigenvalues derivatives second the of that matrix Hessian the and function vote 294 rm au itiuin h probability the distribution, value treme ties. nregion in n a eetmtd o ahvoter each For estimated. be can and given ie htw aedfie h rosa uuaieycmn rmaTp- ex- Type-I a from coming cumulatively as errors the defined have we that Given Thus z sgvnby: given is k where V( dV ihpopulation, with , dρ ρ z) dz dz jk ij j ( = ( j where ( j z f z) z) V k k (V ) j = = = = =

( 1 z so z) −j ( 2β(z 2β(z k ,...,V , . . . z), w =1 p etpclydnt nLEby LNE an denote typically We . 1 =z z ij k p j + exp =1 u = = = j j z k il ∗ − − = exp p (u N (x = n 1 j k x x k ij ∗ p i exp (z ,ofsize k i i 2β (u ,z ρ i ( =1 (x ) ) ∈ z)) jk ij 1 × z il ∗ k N k ,...,z i (f ρ 1 j ) ,z k (x ( jk ij rdce ytepriua oe ihppar- p with model particular the by predicted , oaodpolm ihzr ievle we eigenvalues zero with problems avoid To . × i 1 k w ρ − j i ) n party and ,z − ij ij )) ( i ρ 1 ∈ n ( x −1 1 k ρ j ij u N z) k − i )) ,...,z jk ij + ij ∗ , k (z) h rtodrcniinbecomes condition order first the ρ ρ (x k ) 1 jk ij jk ij = p i a utnma oi specification logit multinomial a has − ,z j ) ( j p exp 1 . j ρ ) h rbblt that probability the ) − ij uhta each that such (f ( . ρ z) jk ij k z ) oavco fvt share vote of vector a to , )(z z(K) −2 j − k where x = p i ) j V .MAitre al. et McAlister K. = exp j ( z 0, i z) ∗ K (f oe for votes ,isaLNE sweakly is k eesto refers ) (2) (3) (1) j Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 xadtecnegnecefiin oint nld hs niiullvlvalences level individual these include to notion coefficient convergence the expand engvnteeidvda aec esrsa el hwn rosincluding proofs showing well, the as to measures was convergence variables. valence that assess these valence, individual to individual intend these We an given equation. valence, mean utility of in kind vec- are included second zero previously we a a not thus, than assume rather and, positions we pur- cancelations party Second, the those of space tor. vector For make general policy proofs. cannot a the in for we convergence of occur though, assessing paper, to origin this cancelation the convenient of to some poses corresponds for policy mean allowed the this electoral of and respec- electorate the scale their the an that adjusted to of such have converge regions papers space parties Previous all regional mean means. in and electoral electoral run mean regional the tive electoral that to joint parties Thus, converge the party. that system to that find electoral converge for should an vote we to in equilibrium, choice parties in the at that has locate find actually we to that ideal incentives should voter the on average have themselves the parties position of they these point that Rather, results, meaning mean. means, voter electoral electoral electorate, overall median respective an the their other of at parts locate with all to before, line in agents explained run these in As not for do incentive origin. that no electoral parties is aggregate regional there are joint not there the does where at interest cases of parties in vector the position the of in- where we all model departure, have a important in most convergence the assess certainly to and tend First, model. vote eigenval- stochastic this their compute and points these at matrices Hessian given ues. the function, show vote to the need of derivative we first the that show to httecnegnecefiin ol ecluae hnw sueacommon a assume we when calculated agent be for could valence coefficient (2007 convergence Schofield Previously, the valences. that level individual not the will on coefficient) convergence depend the (i.e. convergence assess that measures resulting ρ fcus,aLEhst ea rtclpit oaltesto osbeLEcnbe that can assume LNE not possible do of different we somewhat set as is the derivative works all this earlier that so from Note that point, 0. than critical to a equation this at setting be by to obtained has LNE a course, Of oae taciia on fterrsetv oefunctions, given vote that respective show their to of need point we critical LNE, a are at points located that show to order In Bundles Party Varying with Elections Modeling osntrl on rely not does mean. electoral the etof dent ij h rtdrvtv of derivative first The used have that papers previous from departures key two make we paper, this In hsrsl siprati opeo as is,w e httefis derivative first the that see we First, ways. of couple a in important is result This hta individual an that , i .Ti sdet h atta ed o sueta l ate r oae at located are parties all that assume not do we that fact the to due is This ). λ j j or cosalmmeso neetrt.Ti nigalw sto us allows finding This electorate. an of members all across dV i α oe o party for votes dz ij V j ( j j naywyaiefo h aclto fteprobability, the of calculation the from aside way any in z) (z) = ihrsett n ieso fteplc pc is: space policy the of dimension one to respect with 2β n i =1 n j (z hsi necuaigrsl eas any because result encouraging an is This . j − x i )ρ ij ( 1 ρ − ij ρ equals ij z seult eo Then zero. to equal is , ) V j ( ρ z) j hs eneed we Thus, . z bigindepen- (being l gnsare agents all , nyshowed only ) 295 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 swl nodrfrtepitt ealclmxmm ie h qainfrthe for equation the Given maximum. local on a relies be it the that to see LNE, negative can point be a we must the elements, thus Hessian diagonal for and the main order of maximum, trace in local the well a Thus, as negative. be be to all order to have In eigenvalues matrix. the the with from associated repelled to is attracted party is the party large, vote-optimizing becomes rational, coefficient substan- the a position. viewed As much be position. how can certain of coefficient a measure convergence a This derivative point. second as given the tively find quickly a to how Hessian at used of be the changing can measure Nash is of equation unitless local eigenvalues reduced a a a the coefficient, works, is convergence earlier test, point a in a order as However, not second found. or proper be whether must the assess to for order Thus, in equilibria. to reduced be can LNE. Hessian a is interest of position the that say can we • follows: as defined matrix h eoddrvtv a engtv si 2 if is negative be can derivative second the • ewe h niiulsielpito n ieso n h at’ oiinon position party’s the and dimension As one dimension. on same the point ideal individual’s the between • edascn re odto.TeHsinmti fscn eiaie sa is derivatives second of matrix Hessian The electoral condition. the order that second a idea need the This to point. support critical further LNE. gives a possible a it at always always because is is result, mean it good mean, electoral that a respective meaning its also at zero, is locates equals party sim- a always some when equation doing that after see the Second, to point. easy ideal is voter’s it a algebra, of ple exogenous are they as long as 296 hs hsi h au fitrs hntyn oass hte rntapiti a is point a probability the not fast or how of whether measure assess the voter to as viewed that trying be when can value interest This of maximum. local value the is this Thus, Let h f ignleeet aetefloigform: following the have elements diagonal off The clrpout with product, scalar where h rgnTe ignletiso h esa o candidate for Hessian form: the of entries diagonal origin.Then the fthe of ie hsmti,i all if matrix, this Given eko httetaeo h esa seult h u fteeigenvalues the of sum the to equal is Hessian the of trace the that know We the that matrix characteristic no is there sort, this of models previous Unlike ots faciia on salclmxmmi h oefnto,tu N,we LNE, a thus function, vote the in maximum local a is point critical a if test To v t n s = i = oesadlt Let let. and voters (x oe o party for votes t 1 and , t ,x n n 1 1 2 i i =1 =1 t n n s ,...,x = 2β(ρ 4β

1,...,w 0 β 2 (x =[ nt and ij w j is ) )(1 z hne steprymkssalmvs ereason We moves. small makes party the as changes ievle fteHsinaengtv given negative are Hessian the of eigenvalues etevco fthe of vector the be − v j ρ t = and , − z i − j sj (z ρ 0,v )(x r eesrl oiie h nywyi which in way only the positive, necessarily are ij 1 j ) t is s ,z = 2β(x − − 2 1,...,w j 0] ,...,z z β(x tj it )ρ h lcoa oainemti about matrix covariance electoral the − i ij − z tj ( tj . 1 t β z ) ) hcodntso h positions the of coordinates th i , − 2 ) and ρ ( 2 1 ( ρ i 1 j ij − v n h qae distance squared the and , − )(1 2ρ t 2ρ − ij − j ij z ) 2ρ tj ) aetefollowing the have − ,v sgetrta 1. than greater is ij .MAitre al. et McAlister K. 1 s ) − z sj  z w ethe be then , × w Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 niiullvlvlne n ate hc u nsnua ein,as regions, singular in run include which to model parties the and of valences lowest restructuring level the slight individual had that the which made with party However, be the at- valence. could to been exogenous belong assumption have would an coefficient parties population, convergence the highest same of the the all over that optimize given to Previously, tempting system. electoral the in party eaLE ahscn eiaiehst engtv.Tu,ec osiun atof part constituent each Thus, negative. be to has derivative c( second each LNE, a be ovrec otevco fitrs sthat is interest of vector the to convergence h osblt httepryi oae tasdl on.A point. saddle eliminates a This at 0. located than is less party are the dimension that possibility each the to respect with derivatives second httema f2β(x of mean the that Bundles Party Varying with Elections Modeling ersnsteeetrlbhvo ftesse.Tu,fra lcoa ytm the system, electoral an for Thus, coefficient system. convergence the is: coefficient highest of convergence the behavior analysis with electoral complete party the a conver- ensure represents the the to Then Thus, party behavior. each define. a convergence for region, to of calculated each difficult be for very should valences becomes coefficient these party gence of valence values lowest multiple of definitions and multiple notion equation are mean the there the in that from given valence away as- fact, move the of In to make behavior. first longer equilibrium the be no be that can will should we party valence valences, lowest of the difference that a sumption to down reduced be longer saLEi sfollows: as is LNE a is etro at oiin,asfcetcniinfrtevco en oa Nash local a being vector the for a Given condition that positions. sufficient is party a equilibrium of positions, vector party given of a vector to convergence for conditions sufficient hsdfiiino h ovrec ofceti qiaett h ento provided origin, definition electoral the the This to (2007 at equivalent Schofield locate space. is in do policy coefficient parties convergence the all the when of of definition that dimension fact this each the by for supported electorate is the notion in individuals all over nwrs h ovrec ofceti qa otesmo envle of values mean of sum the to equal is coefficient convergence the words, In lcoa rgn oee,ti sol o n ieso,s h uldfiiino the of the definition at full the positioned so being dimension, is: one parties coefficient convergence for on only rely is this not However, origin. does electoral that coefficient convergence the to z) ti motn ont htacnegnecefiin a ecluae o each for calculated be can coefficient convergence a that note to important is It nsmay h ehdfrassigwehro o etro at positions party of vector a not or whether assessing for method the summary, In ie hsdfiiino h ovrec ofcet ecndrv eesr and necessary derive can we coefficient, convergence the of definition this Given utb esta 1. than less be must ). c( c( .If z)<1. z) i − = z n 1 i ) 2 2β(x i =1 ( w 1 c( c( − z) i =1 i n z) − 2ρ sls hn1 hnw a urne htthe that guarantee can we then 1, than less is 2β(x = z ij i arg ) ) p 2 vralvtr sa qiaetconcept equivalent an is voters all over ( it 1 c c( − p − z)

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 .I w iesos h ute ufiin odto sta det that is condition sufficient further the dimensions, two In 6. .Cluaetecnegnecefiin o ahparty, each for coefficient convergence the Calculate 7. .Cekta ahprypsto et h rtodrcniingvnteother the given condition order first the meets space. position policy party the in each positions that party of Check vector the 2. or z*, Define 1. 298 .DfieteHessian, the Define 3. .Tencsaycniinta h ievle l engtv sta trace that is negative, negative be are all eigenvalues eigenvalues the that the condition of necessary The all If 5. Hessian. each for eigenvalues the Check 4. • • • • z at positions: party • h etro oiin salclNs equilibrium. Nash local a is positions of vector the .Since 0. h ovrec ofcet labelled coefficient, convergence The z hc seuvln otecniinthat condition the to equivalent is which iw tj h f ignleeet aetefloigform following the have elements diagonal off The h ufiin odto sstse,adtesse ovre otevco of vector the to converges system the If and interest. satisfied, is condition sufficient the If ocmaeti eea oe ihteoepeetdi coed( Schofield in presented one the with model general this compare To ups htalprisaottesm oiina h lcoa mean electoral the at position same the adopt Then parties all that suppose where ignletisare entries diagonal oeta ahprysrsetv lcoa eni oiinta sawy a always is that position function. vote a the is in mean point electoral critical respective party’s each that Note arxaotteoii.Then origin. the about matrix esta ,te h ytmcnegsto converges system the then 1, than less ) ) 2 2 c( w 1. β(ρ ρ t . ij = sidpnetof independent is ij 1,...,w c( )(1 hnw anthv ovrec.I,however If, convergence. have cannot we then , n 1 n 1 c z) j dV i i − ( =1 =1 n n ≤ z) dz ρ j C w 4β 2β(ρ (z) ij = j . j hc h opnnsof components the check , )> ( 2 n 1 z) (x = ij i hsrdcsto: reduces this 0 =1 is o ahprypsto sfollows: as position party each for w 2β )(1 n − i =1 i n − t z .Welet =1 w js 2β(ρ ρ )(x ij i =1 n ) it c( (x 2β(x ij − z) )(1

i − z ersnsteeetrlsystem. electoral the represents , 0 jt it z − z ethe be t w . )ρ =1 j − t w 2ρ )ρ =1 ij z ij ( ij tj c 1 j ( )(x i n ) w =1 1 (z) − 2 i n ( =1 − 1 itw by ρ 2β(ρ ndimension in ij − ρ 2β(ρ ij )(1 − w 2ρ ) ij z lcoa covariance electoral = ij − iw )(1 ij ) 0 )(1 ) 2ρ − 2 .MAitre al. et McAlister K. − c( ij 1 − (C 2ρ ) ,then z)<1, 2ρ w j ij (C ( ,ifallare ij )> z)) )(x )(x j 2007 z ( itw )< z)) itw = − 0. 0, ), − Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 esrso aec,bt grgt n niiul o ahpryi h system, the in have party to each of need for we estimation individual, above, and an proposed aggregate and model both election valence, stochastic of the measures utilize to order In Bundles Party Varying Given Strategies Estimation 3 Bundles Party Varying with Elections Modeling .We I svoae,temlioillgtseicto sicreti ewant we if incorrect is specification logit multinomial the violated, is IIA When group, 1. to group different. from are preserved sets be choice ratios odds the all put, if that even Simply requires alternatives. a that irrelevant statement only of a is independence for IIA of assumption ballot the the upon reliant on compli- thus more becomes are in quickly and parties situation the region, regional electorate, one are an cated. the in of there words, on members run alternatives other when of of In only beginning, fraction bundle region. which the same one country in the in a with shown exist vote voters as voters However, the How- all of ballot. analysis. when all equilibrium an works when to is only true model it this only logit model, is conditional the this to of ever, data assumptions of the given translation Typically, easy system. the within sitions sue bv odrv h ovrec ofcet htis: that coefficient, convergence same the the derive is to assumed, above are used errors value as extreme Type-I when model logit multinomial model logit aaoo( Yamamoto h esnta e ehdi eesr sta utnma oi oesare models logit multinomial that is necessary is method new a that reason The od n siainpoeue ihti data. this with procedures estimation any do to • w dniyteHsinwt h matrix the with Hessian the identify tcnas esonta h ufiin odto o ovrec,i w di- two in convergence, for condition sufficient by the given that is mensions, shown be also can It hsteegnausaedtrie ytencsaycniintrace(C condition necessary the by determined are eigenvalues the Thus where hc ecnwieas write can we which , VL.Ti oe,wihflostesm pcfiaina h typical the as specification same the follows which model, This (VCL). I sthe is 2011 C rpsda prpit oe,cle the called model, appropriate an proposed ) w j β ( by z) ln ihtedt nodrt nlz qiiru po- equilibrium analyze to order in data the with along ρ = w ij c c ( (ρ C = z) dniymti.Since matrix. identity = j ∗ j 2β( ( = )(1 2β( z) = 1 − 1 − k p − exp ρ =1 2β( j 2ρ 2ρ ) exp (u 4β j 1 j ) ij ∗ ) − 2 trace( (u trace ( (x 1 2ρ ik ∗ i − ,z (x j ( ) 2ρ j i ,z )) (ρ 0 j 0 0 )< k ) ) j − )) ≤ )(1 0 I 1. w ( 1 − − ρ j 2βI) )> )(2β) ayn hieset choice varying ,w can we 0, j ∗ ( z)) 299 ≤ Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 o u oe,w suetefloigseicto o h bevduiiygained utility model. observed the the within for parameter voter specification following by a the for assume estimate we model, aggregate we our best and For the distribution, own determine their to from estimates, that come these use parame- estimates using individual the Then, these set. estimate that choice we that assume electorate, within we an individuals the in for set interest choice of of ters type individual each for that effects. random utilizing done best a is generally estimation distribution, of probability method some estimated This from parameters distribution. come normal the to assumed that are are meaning region valence structure, us- each of by for hierarchical types types effects both both of random parameters that a accommodate assume to ing able we is hand, VCL the other specific; individual the On region. of regardless we hi da onsadtepris da ons eod yvru fteusage the of virtue be- by differences Second, the points. to ideal apply parties’ they the weight and much points how ideal they not their how and tween in parties differ the have only of individuals which that each papers view means previous simply from This departure assumption. this a utilized not is This group. sociodemographic μ niiulfo region from individual where ex- For specific. region are that that regions assume values we all have model, to parameters our common the in are ample, that to parameters offered or that parties electorate of assume bundle an can different choice in a we each has models, case, region these each this as In In region, voters. a electorate. make as entire to seen re- the estimates is type logistic for these set individual valence aggregating of be then estimate to type aggregate there set an allowing choice by each done for models is gression This assumption. IIA the on the analyzing al- to match, parameters. parameters the still given of model system estimations empirical the of empirical the equilibria and from model transition formal easy the lowing of framework the Thus 300 siainpoeuede.Frt ie h tutr fteuiiyeuto,w as- we equation, utility the of structure that the sume given First, does. procedure structure. estimation regional clear a when with choice electorate with attractive an up very from lines a parameters clearly VCL estimate model the to makes This attempting and party. before each specified for model voting formal for the gained utility the of sum party for votes voter j C oe.A ihtpcllgtmdl,tepoaiiyta voter that probability the models, logit typical with As model. VCL party ulda itnebtenvtradpryhstesm nepeaina ihnthe within as interpretation same the model. has formal party and voter between distance Euclidian jr olw h yia oi pcfiain hc ttsta h rbblt htthe that probability the that states which specification, logit typical the follows h C oe ssrno fet o h niiulcoc e ye,meaning types, set choice individual the for effects random uses model VCL The h C ifr rmtpcllgsi ersinmdl,tog,b o relying not by though, models, regression logistic typical from differs VCL The sn h C,hwvr lcsafwlgtasmtoso h oe,a any as model, the on assumptions light few a places however, VCL, the Using htteaeaemme rmscoeorpi group sociodemographic from member average the that j λ hsceryheacia pcfiaino aec ed tefvr elt the to well very itself lends valence of specification hierarchical clearly This . j i steageaeetmt fteeoeosvlneo party of valence exogenous the of estimate aggregate the is rmvtn o party for voting from β scmo vralmmeso h lcoae eadeso einor region of regardless electorate, the of members all over common is μ jr steadduiiyoe h grgt aec htteaverage the that valence aggregate the over utility added the is u j ij ∗ (x stertoo h xoetae tlt fvtn for voting of utility exponentiated the of ratio the is r i e o oigfrparty for voting for get ,z j ) = j λ : j + β β scmo oalmmeso h electorate the of members all to common is  z j − x i + j and μ jr ξ jrs + ξ steadduiiyover utility added the is jrs s esfo oigfor voting from gets i .MAitre al. et McAlister K. oe o party for votes j and j β othe to and Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 n itiuin fteprmtr fitrs.W suethat assume We interest. of parameters the of distributions can ing interest. and of up parameters set the easily of is estimates sampler garner Gibbs to a utilized family, be Gamma the distributions, within statistical used those commonly generally from a (both come interest in effects) of fixed easier parameters and much the effects of random is each procedure that estimation Assuming setting. the hierarchical of Bayesian implementation the VCL, the using l.Wierno fet a eetmtdi rqets anr si demon- is as manner, frequentist a in estimated (2011 Yamamoto’s be with can strated effects electorate. random the over While vary ily. that sets behavior the choice examining party when with method electorate VCL an the in researcher use parameter parties the to the of opt that we as information Thus, discarding interest useful. essentially much find mixed is may as the probit of with mixed as often the However, are so absorbed. values, values is regional itself absorbed individual IIA are the errors the the thus, logit, as and, assumption set, matrix es- IIA choice error individual to the the the researched in of on the level allow rely the not at not does parameters does timate model probit which multinomial model, alterna- the Another However, probit either. informative. multinomial specific and set efficient the choice more is model. estimate is tive VCL to logit individual, thus researcher mixed each parameters, the standard of for allow values not than the does rather over logit region) gain mixed (i.e. Similarly, efficiency can type same significant estimates set the a parameter achieve choice demonstrating VCL, to each of used for mixed structure be achieved of can the be kind model given specific However, logit a mixed as results. viewed the aggregate be that can meaning and VCL model, effects. modeling logit random hierarchical of of through notion use achieved the the best through through is parameters pooling of Partial these allowing estimation pooling. and aggregate partial group the each al- derive within by estimated to IIA be parameters relaxes to VCL parameters model. the the of from each while information lowing assumption most IIA the the with relaxes us it providing that also is regions multiple with electorates handling flexible a is VCL the interest of procedure. electorate estimation all the When of assume by appropriate. choice met we is are model when assumptions this Thus, these region, another. a of in three or run region to choose one par- historically to have parties parties that bound regional ballots historically with same are systems the which electoral on ties many remain However, in exogenous year. not troublesome is to do be year region frequently can from specific parties assumption where a This systems in region. VCL electoral that run the some within of to success usage decision perceived our party’s its of a independent of virtue are that by person assume Third, each space. we model. policy for model, the the effects sociodemo- in within random position covariates and these one’s other regional that of the assume the to we of orthogonal put, each are Simply that effects random assumes group model graphic this effects, random of Bundles Party Varying with Elections Modeling hs itiuin r needn foeaohr hsasmto olw from follows assumption This another. of one all of that independent assume we are Further, distributions distributions. these normal underlying have all effects dom o plctost hsmdl emk e supin bu h underly- the about assumptions few a make we model, this to applications For h tutr fteVLlnsisl oBysa siainmtosvr eas- very methods estimation Bayesian to itself lends VCL the of structure The when method superior the is (VCL) logit set choice varying the that reason The xetto-aiiainagrtmfrestimation for algorithm expectation-maximization ) β , λ j n h ran- the and , 301 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 plcto oCnda Elections Canadian to Application 4 concern. some for cause a certainly is it biased fixed. to large, easily lead not is will problem is violation this However, bias policy assumption the the This within Canada). While location in and estimators. region (as cases, correlated some assumption are in orthogonality space is example, the For yielding independence that met. and assume when perfectly converge to garnered is to unreasonable those sampler is from it Gibbs However, indiscernible the assumed. virtually takes it are time that of results computationally, taxing amount is the assumption this to that adding shown However, has itself. model re- within this (aggregate, with normal spent hierarchy multivariate all time the a are of from sampler, level comes Gibbs each sociodemographic) the that gion, in assume draws easily the could thus We and orthogonal. variables, the that assumptions our 302 at NP,albrlprywoespoticesdfo h 00eetos in elections, 2000 the from increased Democratic support New whose the party with coalition liberal legislature. the a a control formed (NDP), Parlia- to initially Party in order Liberals seats in the the government and of coalition Martin majority a Paul a form gain to to needed percentage failed largest and it the since however ment and percent), Canada seats) (36.7 (135 for vote seats government the Canada most of minority the in gained first election Party the Liberal 2004 The yielded the 1979. Fed- However, it the different. because to no significant parties was different was 2004 three least and at legislature elected eral have Canadians history, recent In discrete a use to necessary IIA. is on it information rely when of not choice does amount best which sheer the model thus, the choice is, method, VCL Gibbs consuming the the time sociodemo- from from derived a gained of are is number samples which it the at and Though rate sampler. parties the down of slow large number can a the groups party, given graphic effects slow each random quite for as Similarly, estimated be observations. individual are can and VCL types set the choice of of Convergence number model. at be the to within needs gence mean the but correla- diffuse, problem. own be the its fix still of to can care value case take specified this to the it in effect if priors random as in normal the discrepancy predictor The allows the a tion. and for as variable controls include researcher omitted to the an need this, is we doing that In mean effects. sociodemo- the random with is the correlated this be region, theoretically party. and the will base group that of the graphic covariate and mean the interest is respective of this the party that the Given at between situated differences sociode- spe- its Euclidian indeed a of of given are level mean difference case, the effects a and this random predictor effects In random level a level interest. regional mographic as of its interest effect of mean solved random of the is the covariate party, cific than problem the hierarchy the of the that mean in propose the lower They adding biased. by be simply to very estimates the causes which emne l ( al. et Gelman n rcia oei eesr eadn h iencsayt civ conver- achieve to necessary time the regarding necessary is note practical One 2008 tlz ehdt i admefcso h collinearity the of effects random rid to method a utilize ) .MAitre al. et McAlister K. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

690 689 688 687 686 685 684 683 682 681 680 679 678 677 676 675 674 673 672 671 670 669 668 667 666 665 664 663 662 661 EDITOR’S660 659 658 657 656 655 654 653 652 651 PROOF650 649 648 647 646 645 3 ne-uidcinljbmblt,()hligQee n 5 h nuneof influence the (5) living, and of Quebec standard helping their in- (2) (4) dimension mobility, state, decentralization The welfare job axis. the inter-jurisdictional programs this (1) (3) social along towards in scale attitudes the interest left-right coded voters’ We higher a cluded scale. have imply gun left-right to values (3) the as lower position women, so their that helping (5) such (2) and dimension rich, Iraq social and in war poor the between (4) gap control, voters’ the of (1) combination the weighted towards “social,” a one attitudes is dimensions: policy dimension or social factors The two “decentralization.” were other there that conclude to Quebec. us within led ran only Quebecois Bloc as Quebec, Table within electorate. voters the Canadian those that the for suggests of shares representative of sets fairly two is these sample between similarity The shares. vote overall oepercentages vote etdb li ta.( al. et Blais by seats lected any election receiving the from in votes it of kept percent). portion 4.3 significant base seats, a voter (0 gain initial did increased it in small However, party part parliament. its the within for a elections, support the played 2004 Though of undoubtedly Party. the Liberal which scale the in party the of Canadian small share on the vote another quite the in was reducing not player Party while significant Green Similarly, a the percent). BQ, BQ 12.4 the seats, made Party’s (54 Liberal and the Quebec parliament in region of dent the a percent quite within put fifty showing showing strong nearly their This However, party. with was the Canadians. overwhelming, thus, for voting of was and, voters percent region Quebec twenty the in approximately within ran for support only ballot BQ the The on (BQ). only Quebecois government. Bloc control to party, coalition regional a form to garner to to able led able not party not was was new and Party the Conservative majority the of seat and a inexperience elections lead- relative the weeks to the the prior However, in mistakes elections. key neck the and neck to Party, were up Liberal Party ing Liberal the the within and scandal Party control- of Conservative of exposure the hope 2000 Given Party government. the Conservative Canadian the in gave the support parties ling splitting two the After of share. merger sig- by vote the which elections, party, Liberal’s formed Conservative the Progressive party into the the chipped and Canada, Party nificantly Alliance op- of the main Party of Party’s Conservative merger Liberal formed the The newly percent). the 15.7 was seats, ponent (19 government control to order 1 Table Bundles Party Varying with Elections Modeling h atraayi efre ntevtr’rsossi h uvyquestions survey the in responses voters’ the on performed analysis factor The osuyte20 aaineeto eue h uvydt o aaacol- Canada for data survey the used we election Canadian 2004 the study To Quebec’s was elections Canadian 2004 the of aspect interesting most the Perhaps culadsample and Actual 2006 .Table ). D 15 BQ 29 Green 36 Conservative NDP Liberal 1 hw oesae ihntesml n the and sample the within shares vote shows 12 culSml—l Sample—Quebec Sample—All Actual 4 . . . . . 555. 11.95 3. . 25 42 31.55 29 18.45 66 34. 65 71 71 1 lohscolumns has also 9. 8. 2. 08 13 01 02 68 303 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

736 735 734 733 732 731 730 729 728 727 726 725 724 723 722 721 720 719 718 717 716 715 714 713 712 711 710 709 708 707 EDITOR’S706 705 704 703 702 701 700 699 698 697 PROOF696 695 694 693 692 691 at oiin stema fvtr o htpry h at oiin ntepolicy the in positions party vector: The the party. by that given for are voters space of mean the as positions could party and axis), (y axis after decentralization. D looking to the of opposed on job while as position better axis), regarded negative a (x be a axis does have Social government would the federal interests of the left peoples’ that the believes on who be voter to a tend would poor and rich between oes oiin ln hs w iesosi aaaaea h lcoa origin, electoral the at are of Canada values in median dimensions and ( two mean these The along dimensions. positions decentralization voters’ and social the along zto mle ihrvle nti xs h usin sdi h atranalysis factor the in used Table questions in The found axis. be decentral- this can for on desire values greater higher A implies lives. ization their in governments Provincial versus Federal 2 Table 304 Federal-provincial Region Cross Moving Quebec Living of Standard Welfare Left-Right War Iraq police/military Only Gun Women Inequality 0; h uvyakdvtr hc at hywudb oigfr ow estimated we so for, voting be would they party which voters asked survey The sn h atrlaig ie nTable in given loadings factor the Using 0) oilsrt,avtrwotik htmr hudb oet euetegap the reduce to done be should more that thinks who voter a illustrate, To . uvyitems Survey 2. 1 rvnil()n ifrne()federal (3) difference no (2) better? provincial interests (1) your after looks government which general, In agree strongly disagree—(4) strongly (1) are? they live, jobs they the where where region to the move in should work find can’t people If less much Quebec? more—(5) for much done (1) be should think you do much people How leave Don’t (2) behind people leave (1) decent a has everyone living. that of it standard to see should government The agree strongly disagree—(4) strongly (1) after look to willing less themselves. people makes state welfare The right below? left—(11) scale (0) the on yourself right. place and you left would of Where talk sometimes people politics, In decision decision? bad bad (2) a decision or good decision (1) good a was this think war you the Do in participate to Iraq. not against decided Canada know, may you As guns. disagree have strongly to agree—(4) allowed strongly be (1) should military the and police the Only less much women? more—(5) for much done (1) be should think you do much How less much more—(5) much Canada? gap (1) in the poor reduce the to and done rich be the should between think you to much How ecmue h au o ahvoter each for value the computed we 3, .MAitre al. et McAlister K. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

782 781 780 779 778 777 776 775 774 773 772 771 770 769 768 767 766 765 764 763 762 761 760 759 758 757 756 755 754 753 EDITOR’S752 751 750 749 748 747 746 745 744 743 PROOF742 741 740 739 738 737 iencsayt u h ib ape a eln ec oidmgahcgroup the sociodemographic and (each convergence long to be can time sampler the Gibbs as the model run previously, VCL to noted necessary the As time sociodemo- in made. model, be important can effects very groups random is that underlying parsimony so the categorical as and viewed VCL are graphics the Due of High Higher. or No structure Bachelors categories: the Bachelors, 18Ð29, three No to categories: but into Diploma divided four School was High into Diploma, Education divided School older. was and 65 Age 50Ð65, recorded. 30Ð49, were level education and two the that evidence are identities. plots, two regional the strong from along have seen This, regions easily state. strong are independent that a somewhat differences has in its the Quebec maintain with voters as to Similarly, government, wants Canada. and of identity in decentralization regional advocate voter more want Quebec average to in the tend voters than Quebec of policies majority social The liberal only. more Quebec most Figure for mean. distribution with electoral overall voter issues, the Fig. the of decentralization left In noticeably on is middle. which split the Quebec, within view evenly in leftist fairly right moderately a are lying have voters and voters most issues that social see the we on in this, Fig. lies From party in Canada. Conservative in shown the voters while as access quadrant, social right the on upper left the to locates Party hs at oiin orsodcoeywt hs siae yBni n Laver and Benoit by estimated those with ( closely correspond positions party These al 3 Table Bundles Party Varying with Elections Modeling D( SD Federal-provincial region cross Moving Quebec a 811 39 28 28 Var % Cumulative Var % tnado Living of Standard Welfare Left-Right War Iraq police/military only Gun Women Inequality Components 2006 h uvyas olce oidmgahcdt.Frec epnet e,age, sex, respondent, each For data. sociodemographic collected also survey The √ var ,otie sn xetoiin n20.A ihteeetmts h Liberal the estimates, these with As 2000. in opinions expert using obtained ), egtn ofcet o Canada for coefficients Weighting )1 z ∗ = ⎣ ⎡ D S 20 .05 0.32 −0.38 1.27 −0.17 Lib. Con. 1 h Q ersnsteeetrlmean electoral the represents “Q” the 1, Figure . −0. −0. Social −0.78 D Grn NDP 0. 0. 0 0. 0. 0. 0 0 . . . . 09 27 35 38 71. 67 37 38 30 20 35 36 1 −0.63 30.23 −0.13 losostedsrbto of distribution the shows also . −1.48 BQU ⎦ ⎤ −0. −0. −0. −0. −0. −0. Decentralization 0. 0. 0. 0. 65 48 00 05 07 17 06 03 20 52 07 2 shows 305 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

828 827 826 825 824 823 822 821 820 819 818 817 816 815 814 813 812 811 810 809 808 807 806 805 804 803 802 801 800 799 EDITOR’S798 797 796 795 794 793 792 791 790 789 PROOF788 787 786 785 784 783 hmali h oe.I hscs,atrtyn ihtemdlfrsm ie it variable. our time, a efficiently, as some run sex to and for include sampler age Gibbs not model by the does the allow spurious model and yielded with time was toying preserve vote to after and Thus, sex education. case, between keep this relationship to In the sense that makes model. idea seemed it the good if a in see always and all is variables it them the thus between considered), relationships being the region examine each to for effect random a has 2004 in Quebec for positions party and 2 Fig. 2004 in Canada for positions party and 1 Fig. 306 ape o rud1000ieain n eevdnc omldsrbtosfor Gibbs distributions each normal is ran nice We it received while. so a converge, and run to iterations model while 100,000 this a around for take for sampler does Gibbs sampler model variance the this the let for before, to hyperprior stated necessary gamma assist As diffuse To prior. value. a this create each at we effects of VCL, random the the to effects of priors random convergence normal respective in the region’s of each mean in the party setting Euclidian in- each by We for of earlier. difference proposed variable mean procedure independent the correction clude effects the random and the interest use correlation we some of difference, given effects model, random the For the sociodemographics. between with model a for lences sn h ayn hiestlgtpooe ale,w estimate we earlier, proposed logit set choice varying the Using itiuino voters of Distribution voters of Distribution .MAitre al. et McAlister K. β n h va- the and Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

874 873 872 871 870 869 868 867 866 865 864 863 862 861 860 859 858 857 856 855 854 853 852 851 850 849 848 847 846 845 EDITOR’S844 843 842 841 840 839 838 837 836 835 PROOF834 833 832 831 830 829 ogrdcsteefcso h neetatcreainta cusi h sampler. the in occurs this that autocorrelation run inherent to the sampler of Gibbs effects the the allowing Similarly, reduces interest. long of parameters the of each Bundles Party Varying with Elections Modeling pc nodrt aiietervt hr.Bcueteotoeo h election the of outcome the Because share. vote their maximize to order vote Party. in Green to the space choose also instead is that Quebec voters in party draw somewhat BQ valence to of lowest and fails The is voters it BQ. Party same for as Conservative the Quebec, The similar Quebec. within of in valence in many are lower presence for Party stronger compete Liberal a they has the space, simply and preference BQ the that strong of a given areas has however, also Quebec; Party Liberal in The measure. presence valence aggregate the have by to estimated fails that and party dimension one-issue of a The plenty electorate. of Quebec. makes the more to still) which of appeal Party, is outside mass (and Green was competitor the is significant is it Quebec as a sense of be outside preference to party the valence it in lowest allows its positioning as Canada and presence of valence same the its space have However, not counterpart. does of Liberal simply is NDP party larger The the valence. as equivalent valence almost lower with somewhat parties, valence highest the are Party members of would party. electorate valence they highest uniform then the a is groups, we BQ see sociodemographic if that Thus, and or estimate electorate. affiliation down the regional look party, of specific that valence percent without entities highest 25 around the as to as parties available party view only this is this specifies for of it vote accurately strengths though the actually it even of could that one exemplifies that is This people which did. them the model, of of percent that 50 party of nearly given valence party, outside sense, highest the valence the makes fact, in This in equivalent is, Canada. almost BQ in the are that two shows the model regional this that However, Non-Quebec Quebec. see the can to in valence we party aggregate effect, the valence in random equivalent adding highest almost simply the is By Party is level. Conservative valence Party the However, Liberal Quebec. of the outside model, density. Canada the posterior running the before of dicted log-likelihood the of well. measure data preferred. are the a a DIC fit as of with not values seen model does Lower be a that rewarding model can a quality DIC penalizing model The but of this parameters, in measure of is a number it as small (as normal functions multivariate DIC be BIC. the to or assumed AIC case), is to deviance distribution analogue the posterior report model also the hierarchical We a When values. is these which report as (DIC), not simply criterion do included information we are thus they case, random sometimes, this sociodemographic interest in the substantive controls of While always party. be is each may report valence values for effect their effect we regional thus model, the group, this the report base In then For the intervals. 0. as credible Party at percent Liberal restricted 95 the use corresponding we the example, and values parameter the h eut fteVLaesoni Table in shown are VCL the of results The ealta eaeitrse nfidn hr h ate illct ntepolicy the in locate will parties the where finding in interested are we that Recall than valence larger even an with party, valence highest the is BQ Quebec, Inside Liberal the and Party Conservative the before, mentioned as Quebec, of Outside pre- been have would as First, things. of number a see can we model, this From β n h grgt aecsfis.We first. valences aggregate the and eso h C siae of estimates VCL the show We 4. 307 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

920 919 918 917 916 915 914 913 912 911 910 909 908 907 906 905 904 903 902 901 900 899 898 897 896 895 894 893 892 891 EDITOR’S890 889 888 887 886 885 884 883 882 881 PROOF880 879 878 877 876 875 308

Table 4 2004 Canada VCL model given sociodemographics (LPC base) NDP Conservatives Greens BQ

β 0.2598 (0.2234, 0.2976)

λj −0.5883 −0.0864 −1.826 0.4525 (−1.678, 0.4625)(−1.057, 1.107)(−3.446, 0.2405)(−0.9229, 2.322) NQ Q NQ Q NQ Q NQ Q

μrj 0.0341 −0.6085 0.0655 −0.1827 −0.4038 −0.4915 Ð 0.1797 (−1.014, 1.127)(−1.778, 0.4704)(−1.135, 1.04)(−1.465, 0.8110)(−2.493, 1.217)(−2.720, 1.062) Ð (−1.67, 1.565) DIC 2029.291 .MAitre al. et McAlister K. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

966 965 964 963 962 961 960 959 958 957 956 955 954 953 952 951 950 949 948 947 946 945 944 943 942 941 940 939 938 937 EDITOR’S936 935 934 933 932 931 930 929 928 927 PROOF926 925 924 923 922 921 eigen o hmad hs o N.I sitrsigt oeta oho hs parties these of both that note to interesting is It LNE. a not thus, and, them for ohngtv,te h esa sngtv ent n h at oaini ta at is location party the and definite negative are is Given they maximum. if Hessian local party; the each for then Hessians negative, the both with associated eigenvalues the in interested enn that meaning estimates their locate. account will into parties this taking other use position then where preferred and of outcome most election their anticipated find the information to of other idea information and an polls form use to parties disposal that their assume at we shares, vote these on depends Bundles Party Varying with Elections Modeling ee,w a sueta fapryi on omv,i ilb h re at.We Party. Green the be will it aggregate move, the to at going and as is Hessians well party the as a examine regions, if now both that assume in the can party that we valence Given level, and. lowest equilibrium the at is space fail Party policy they the which Green in condition for elsewhere party necessary move the to the fail, choose conditions meet will these we they of Similarly, if any equilibrium. that see in Given convergence. to position for coefficients this convergence at the locate to check choose can not will parties the of iswl ieg rmtevco feetrlmas h D n h re Party Green the and NDP that The meaning means. eigenvalues, electoral positive of have vector both the from diverge will ties da ons ecncluaeteHsino h oefnto o ahpryas party each for function vote in voter the voters the of of coefficient, about Hessian mean convergence other information the the the the the as at calculate as and well is can position positions mean we regional same party points, its of the ideal so vector at Quebec, this locate in Given not runs Quebec. will only BQ it as that parties means this that Notice n osblt sta l ate illct tterrsetv lcoa means, electoral respective their at locate will parties all that is possibility One rmteHsinsadtercrepnigegnaus ecnseta w par- two that see can we eigenvalues, corresponding their and Hessian’s the From c H H j H H | Con z z BQ Lib ∗ ∗ = = = = = z ∗

⎣ ⎡ sa follows: as is z 140.0034 −0.1194 −0.0002 −0.0326 −0.0004 −0.0365 c( Eigen1 Eigen2 0.0034 ∗ z = ∗ ) ⎣ ⎡ z D ∗ S 7 1.945 1.071 1.518 1.031 Lib. fayo h esasaentngtv ent,te one then definite, negative not are Hessians the of any if , 350.0021 −0.0365 −0.0705 −0.0676 −0.0002 −0.0705 −0.0004 −0.1286 Lib. Lib. 000 0 00 0 0 00 c( D Con. NDP z ∗ Con. ) c( o ahparty. each for    −0.0361 z ; ; ∗ D Con . NDP ) o ahpry o h esas eare we Hessians, the For party. each for D Grn NDP H H GPC NDP z ∗ sntavt aiiigposition maximizing vote a not is = = 360.0085 −0.0326 −0.0676 Grn.

010.0012 0.0012 0.0021 050.0085 0.0085 0.0085 . −0.08 −1.11 −0.5921 BQU BQ −0.0092 −0.0362 −0.0091 Grn. ⎦ ⎤    −0.1183 −0.1297 BQ z ∗ 309 is ⎦ ⎤ Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 1012 1011 1010 1009 1008 1007 1006 1005 1004 1003 1002 1001 1000 999 998 997 996 995 994 993 992 991 990 989 988 987 986 985 984 983 EDITOR’S982 981 980 979 978 977 976 975 974 973 PROOF972 971 970 969 968 967 hs oesi ubc hs hssyeo piie sncsayfrfidn the finding for necessary is among optimizer share of vote its style Canada. optimize this in to positions Thus, trying optimizing respective Quebec. only their is in optimize BQ voters to Canada, electorate. attempting those of are the all parties of over four portion shares other that different vote the given by a while necessary over better case, is optimizing any this method be In do This potentially can party. can other party in party the no party each of where eventually each positions positions for should the of this parties given set moving do the equilibrium we values, the If starting its on parties. arbitrary converge optimizes other some party the at style each of beginning sampling which positions rotation Gibbs in the a method given In optimization share vector. an vote mean create the maximizing we on vote optimizer, the located that of not show are pro- to parties the methods to optimization for validity use positions give to to need order we in tests, Thus, posed parties. the of tendencies maximizing the vote as just vector, mean the given when point the did. saddle not that test a is indicate Hessian at values Party located These Green also position. the is social Party that mean Green the means at This share axis. vote social its maximizing the to corresponds 1.379 where c h aiiigpstosfrtelrettrepris oee,i codnewith accordance in However, parties. three on largest effect little the has for to relocation Given positions their incentive there. valence, maximizing stay low have the relatively to of want Party are parties parties Green other two the these the while that mean and electoral the NDP told from the away were move we coefficients: what convergence echo positions the optimizing by vote the measures, our for Fortunately h etro neet ee eseta l ftecnegnecefiins except than coefficients, less convergence but the one of than all greater that are see BQ’s, we Here, for interest. of mean vector the axis. the on that be represents eigenvalue still second will the it position, as better axis a decentralization choose the they of when Thus, point. saddle a 310 4 edt hc h ags n osei tidctscnegnet h envector. of mean examination the and to its Party convergence of Green portions indicates the constituent it to the if belongs see coefficient convergence to largest one The largest the check to need eatr rmti oiinwudrsl nalredces nvt share. vote in small decrease a large a that in meaning result maximum, would local position this changing from quickly departure a indicates coefficient convergence ative sfollows: as ti neetn ont httecnegnecefiin edntb oiie si h aewith case the is as positive, be not need coefficient convergence the that note to interesting is It BQ Figure the match actually tests two these if know not do we are, they as taken However, ecnas tlz h eto ovrec ofcet oass ovrec to convergence assess to coefficients convergence of test the utilize also can We (z ∗ ) hssml niae atclrysrn eiet tyi h ie oiin neg- A position. given the in stay to desire strong particularly a indicates simple This . z opt ∗ 3 hw h oeotmzn oiin o ahpryi aaa hc are which Canada, in party each for positions optimizing vote the shows = ⎣ ⎡ D S −0.0259 9 2. 1.099 0.0649 0.0524 Lib. c GPC c( z ∗ 260.2281 0.0266 −0.0264 z ) ∗ Con. shows: = 1.379 w D Grn NDP + wihi hscs s2), is case this in (which 0.5657 337 . −0.1290 −1.069 BQ .MAitre al. et McAlister K. ⎦ ⎤ 4 hswe thus Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1058 1057 1056 1055 1054 1053 1052 1051 1050 1049 1048 1047 1046 1045 1044 1043 1042 1041 1040 1039 1038 1037 1036 1035 1034 1033 1032 1031 1030 1029 EDITOR’S1028 1027 1026 1025 1024 1023 1022 1021 1020 1019 PROOF1018 1017 1016 1015 1014 1013 various iin hnte i ttercretpstos Table positions? current their at did they than sitions their to proportional means electoral their differences. from valence ( perceived away respective distances Schofield with by axis, proposed same the of theory equilibrium the 5 Table 2004 Canada in positions 3 Fig. Bundles Party Varying with Elections Modeling ate neetrlsseswt ate htd o u neeyrgo.We par- When region. every in run not do that of parties positions with maximizing vote systems the electoral examining in for method parties a proposed we paper, this In 5Conclusion parties the if would it than worse means. do position electoral and extreme respective adapt more their to a at forced located to dis- is move all the is NDP that to and The parties choose Party better. the can of do Green it one and the mean, is Party electoral when Green the better the with As do satisfied mean. all the Liberals from away the locate and NDP BQ, the valences Party, Green aggregate the (i.e. region effects). each random from regional valences the plus actual are vote shares the the vote at using These and routine. predicted mean, optimization electoral the the by at determined positions, positions current maximizing their at party each for sample hsbg h usin huh o uhbte a h ate oa hs po- these at do parties the can better much how though, question, the begs This hstbesrntesornto httevco fmasi o N sthe as LNE a not is means of vector the that notion our strengthens table This oemaximizing Vote z ∗ oesae given shares Vote s GPC BQ NDP CPC LPC 12. 15. 29 36. Current 4 . . 29 42 65 66 71 ,teprislct along locate parties the 2007), 5 hw h oesae nthe in shares vote the shows 11 17 33 33 Mean 3. . . . . 79 55 89 34 42 12 16 33 33 Optimal 3. . . . . 52 96 80 29 43 311 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1104 1103 1102 1101 1100 1099 1098 1097 1096 1095 1094 1093 1092 1091 1090 1089 1088 1087 1086 1085 1084 1083 1082 1081 1080 1079 1078 1077 1076 1075 EDITOR’S1074 1073 1072 1071 1070 1069 1068 1067 1066 1065 PROOF1064 1063 1062 1061 1060 1059 ohnl nee ie ait feetrt ye hnpreviously. than types electorate ability of its variety and wider coefficient convergence even the an demonstrates of handle definition further to new This the examined. of be nature can general differences the equi- and new computed have information, be parties this can voter’s Given where libria each point. know equilibria ideal parties and examine that in group, meaning Similarly, to sociodemographic voters, example). region, possible the for of is regions, each regions it about three many information of model, perfect of out this is combination two of case, a in uses this be runs further can in party numer- region a region, are when a A there case thus, electorate. where (the bundle; nation’s cases party a to a not in applied to is arise be equivalent model which can of bundles” type and “party this case First, region ous used. two be the can to methods limited these which in situations plex electoral the at positions taking polit- by that shares vote theories center. their accepted maximize widely the always of within can contrast positions par- actors direct extreme ical valence in more lower is taking finding the by This Rather, of shares space. mean. mean vote policy electoral the maximize join to at able the locating shares were at BQ vote ties not included respective and which their Quebec mean, maximize in electoral to voters find- joint able empirical the not these at were locating Using the parties by elections. was that 2004 Quebecois found the Bloc that we in region the ings, Canada a population, in Quebec, Canada’s party in of valence ran percent highest only 25 it around though up even makes that found we methods, when empirical run to model ideal hierarchical the clear information have it that of structures. makes electorates amount complex Logit within sheer Set tendencies assump- the voting Choice examining IIA exist, Varying the already the on so from rely doing gained not of does methods that Though coefficient given convergence tion. nec- the parameters the for the utilize equilibrium estimating Nash for to method local essary new a positions a are party proposed positions also of We conver- these system. vector not the electoral generalized or of any whether notion handle us tell to generalized and able more is a which the proposed coefficient valence at We estimating gence bundles appropriate. for party methods longer different empirical no have and valence are voters of different theories region, existing every and in polls run not do ties 312 [,]< lh[]+beta[1] + alpha[k] <- { v[i,k] 1:K) in { for(k 1:N) in for(i model{ sampling. Gibbs the for algorithm the gives appendix This Appendix ie h cuaeotoe fteemtos hr r ubro oecom- more of number a are there methods, these of outcomes accurate the Given new the Using elections. Canadian 2004 the examined we methods, these Using * (d[(N * k1)i-[] + (k-1))+i]-d[i]) .MAitre al. et McAlister K. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1150 1149 1148 1147 1146 1145 1144 1143 1142 1141 1140 1139 1138 1137 1136 1135 1134 1133 1132 1131 1130 1129 1128 1127 1126 1125 1124 1123 1122 1121 EDITOR’S1120 1119 1118 1117 1116 1115 1114 1113 1112 1111 PROOF1110 1109 1108 1107 1106 1105 }} dgamma(.01,.01)I(.01,10) ~ taued[1,f,z] dnorm(0,taued[1,f,z]) ~ ed[1,f,z] 1:4){ in for(z 1:e){ in for(f } -10000 <- ed[1,f,5] 1:e){ in for(f taum[2,5]~dgamma(.1,.1)I(.1,10) taum[2,4]~dgamma(.1,.1)I(.1,10) taum[2,3]~dgamma(.1,.1)I(.1,10) taum[2,2]~dgamma(.1,.1)I(.1,10) taum[1,4]~dgamma(.1,.1)I(.1,10) taum[1,3]~dgamma(.1,.1)I(.1,10) taum[1,2]~dgamma(.1,.1)I(.1,10) dgamma(.1,.1)I(.1,10) ~ taua[5] dgamma(.1,.1)I(.1,10) ~ taua[4] dgamma(.1,.1)I(.1,10) ~ taua[3] dgamma(.1,.1)I(.1,10) ~ taua[2] dgamma(.1,.1)I(.1,10) ~ taub[1] dnorm(0,taum[2,5]) ~ m[2,5] dnorm(0,taum[2,4]) ~ m[2,4] dnorm(0,taum[2,3]) ~ m[2,3] dnorm(0,taum[2,2]) ~ m[2,2] 0 <- m[2,1] -100000 <- m[1,5] dnorm(0,taum[1,4]) ~ m[1,4] dnorm(0,taum[1,3]) ~ m[1,3] dnorm(0,taum[1,2]) ~ m[1,2] 0 <- m[1,1] dnorm(0,taua[5]) ~ alpha[5] dnorm(0,taua[4]) ~ alpha[4] dnorm(0,taua[3]) ~ alpha[3] dnorm(0,taua[2]) ~ alpha[2] 0 <- alpha[1] dnorm(0,taub[1])I(-5,5) ~ beta[1] 1:K]) dcat(pv[i, }} ~ vote[i] expv[i,k]/sum(expv[i,1:K]) <- pv[i,k] exp(v[i,k]) <- expv[i,k] + k] education[i], ag[region[i],education[i],age[i],k] ed[region[i], + m[region[i],k] Bundles Party Varying with Elections Modeling 313 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1196 1195 1194 1193 1192 1191 1190 1189 1188 1187 1186 1185 1184 1183 1182 1181 1180 1179 1178 1177 1176 1175 1174 1173 1172 1171 1170 1169 1168 1167 EDITOR’S1166 1165 1164 1163 1162 1161 1160 1159 1158 1157 PROOF1156 1155 1154 1153 1152 1151 o[,,,]< lh[]+m2z d2fz ag[2,f,w,z] + ed[2,f,z] + } m[2,z] + }}} alpha[z] <- tot[2,f,w,z] 1:a){ in for(w 1:5){ in for(z 1:e){ ag[1,f,w,z] in + for(f ed[1,f,z] + m[1,z] + }}} alpha[z] <- tot[1,f,w,z] 1:a){ in for(w 1:4){ in for(z 1:e){ in for(f }}} dgamma(.01,.01)I(.01,10) ~ tauag[2,f,w,z] dnorm(0,tauag[2,f,w,z]) ~ ag[2,f,w,z] 1:a){ in for(w 1:5){ in for(z 1:e){ in for(f }}} dgamma(.01,.01)I(.01,10) ~ tauag[1,f,w,z] dnorm(0,tauag[1,f,w,z]) ~ ag[1,f,w,z] 1:a){ in for(w 1:4){ in for(z 1:e){ in for(f }} -10000 <- ag[1,f,w,5] 1:a){ in for(w 1:e){ in for(f }} dgamma(.01,.01)I(.01,10) ~ taued[2,f,z] dnorm(0,taued[2,f,z]) ~ ed[2,f,z] 1:5){ in for(z 1:e){ in for(f 314 li ,Fune ,Gdni ,NvteN vrt 20)Eeto 06 o i eethe were big how 2006: Election (2006) J Everitt N, Nevitte E, Gidengil London Routledge, P, democracies. multiparty Fournier modern in in A, policy representation Party Blais policy (2006) M and Laver strategies K, Benoit party Michigan Modeling (1999) of III University S government. Merrill party J, responsible Adams and competition Party (2001) J Adams References hne...ral?Wrigppr nvried Montreal de Universite paper, Working really? . . changes. 43:765Ð781 Sci Polit J Am extreme? so strategies are why elections: Arbor Ann Press, .MAitre al. et McAlister K. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

1242 1241 1240 1239 1238 1237 1236 1235 1234 1233 1232 1231 1230 1229 1228 1227 1226 1225 1224 1223 1222 1221 1220 1219 1218 1217 1216 1215 1214 1213 EDITOR’S1212 1211 1210 1209 1208 1207 1206 1205 1204 1203 PROOF1202 1201 1200 1199 1198 1197 aaooT(01 utnma epnemdlfrvrigcoc es ihapiainto application with sets, choice varying for model response Cambridge multinomial Press, University A Cambridge (2011) simulation. for T methods Yamamoto choice Discrete (2003) Nether- K the Train in competition electoral Multiparty convergent (1998) A for Whitford K, conditions Quinn sufficient A, Martin and N, Schofield necessary theorem: voter mean En- The Prentice-Hall, theory. (2007) political N positive Sci Schofield to Polit introduction J An Am (1973) democracies. PC multiparty Ordeshook in WH, choice Riker Voter 28:283Ð312 (1999) Sci A Polit J Whitford Am A, 1968Ð1980. Martin elections K, presidential University US Harvard Quinn paper, (1984) Unpublished H voting. Rosenthal far-sighted K, of Poole model A (2003) maximization E victory Penn of probability and share vote expected of difference Generic (2007) Behav JW Patty Econ Games Cambridge models. Press, voting University probabilistic Cambridge in equivalence voting. equilibrium of Local theory (2005) united JW A Patty polit- (1999) and B theory Grofman Game S, (ed) Merrill P Ordeshook In: equilibrium. electoral state: of poor 39:41Ð57 Existence state, J (1978) rich Econ G competition. state, Kramer in blue Stability state, (1929) Red H (2008) Hotelling J Cortina J, York Bafumi New Row, B, choice and Shor Harper D, of democracy. Park comparison of A, theory a Gelman economic probit: An multinomial (1957) A and Downs logit Multinomial (2004) J Econometrica Endersby theorem. voter JK, mean Dow a choice: social and Aggregation (1991) B Nalebuff A, Caplin Bundles Party Varying with Elections Modeling atal otse utpryeetos okn paper Working elections. multiparty contested partially 97:257Ð293 Choice Public probit. multinomial on based model a Germany: and lands 74:965Ð980 Stud Econ Rev equilibrium. Cliffs glewood 43:1231Ð1247 29:149Ð173 Welf Choice Soc voters. probabilistic with elections plurality simple in 51:523Ð536 375Ð389 pp York, New Press, University York New science. ical Princeton Press, University Princeton do. they way the vote Americans why 23:107Ð122 Stud Elect research. voting for models 59:1Ð23 315 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 oei re ownsasi h parliament. the in seats win to order in vote 2 1 4.8 and elections 1995 in 4.3 to rising fragmen- (Ozbudun of parties of elections positions of degree 1999 high ideological number in effective historically the a the and by with marked number tation underwent were the Turkey 1990s in both The politics parties. of electoral political terms century, in 21st changes the of significant decade first the During Introduction 1 Demirkaya Betul and Schofield Norman 2000s the in System Party the Tracing in Turkey: Changes in Elections of Model Spatial .Shfil ( Schofield N. DOI (eds.), al. et Schofield N. Drive, e-mail: Brookings USA 1 63130, Louis, MO Saint Louis, in Saint University Washington Economy, Political in Center Demirkaya B. USA 63130-4899, 1027, MO e-mail: Box Louis, Campus St. Hall, Drive, Seigle Brookings Louis, One St. in University Washington Center, Weidenbaum 02 nwihnn fteprisfo h rvosprimn eeal opass to able were parliament previous in the election from threshold. early parties an electoral the by the of followed none was which crisis Mother- in center-right financial 2002, the 2001 and The (MHP) (ANAP). Party Party Action land Nationalist the (DSP), Party Left omrmmeso saitprisadteRpbia epesPry(CHP)— Party People’s the Republican by the founded parties—and party Islamist conservative of new members (AKP)—a former Party Development and Justice eutdi alaetwt v ate,ec ihsa hrsrnigbetween ranging elections shares 1999 seat The with parties. each Islamist %. parties, and 25 five and nationalist with % concurrent 15 the a parliament and of a parties in share center-left resulted vote and center-right the the in of rise share vote the in crease codn oteeetrllwo 93 oiia at ed owna es 0%o h national the of % 10 least at win to needs party political a 1983, of law electoral the to According Tables See 10.1007/978-3-642-35239-3_15 [email protected] schofi[email protected] 1 B and ) 2 1 o oeadsa hrso ate ntels orelections. four last the in parties of shares seat and vote for olto oenetwsfre ytecne-etDemocratic center-left the by formed was government coalition A dacsi oiia Economy Political in Advances 2 h e alaetwsfre ytemmeso the of members the by formed was parliament new The 2000 pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , Kalaycioglu ; 2008 , .Ti a atydet de- a to due partly was This ). 317 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 h ucso oDEHAP to successor the oiia iw oasnl ieso a oca h iest fise htmay that issues of diversity the conceal may dimension of single reduction the a However, to systems. ex- views political and political we widely-used across been comparisons can has characterization making How continuum The left-right in voters? a helpful landscape? along for electoral voters relevant and new position parties are this political the that of in describe preferences issues we voters’ on can the parties How plain other debate? and political shape AKP that of issues main elec- the 2007 are the the in addition, seats win In to %. able 20.9 pro- were the to tions. (DTP) by points Party increased supported Society percentage candidates CHP Democratic independent 1.5 while Kurdish and only votes (MHP) by Party AKP the Action vote elections, of Nationalist 2007 the % the of 46.6 In share receiving agenda. their secularist by a power on their emphasis consolidated strong a with party a g f e d c b a 1 Table 318 osiuinlCourt Constitutional h ucso oDEHAP to successor the rePt at Y 20 .4Ð0.15 Ð Ð 9.54 5.13 Ð 12.01 13.22 1.22 ANAP 22.19 DYP DSP Party Motherland Party Path True Party Left Democratic ecPryG .530 Ð 3.04 7.25 Ð Ð Ð GP 4.75 DEHAP HADEP Party People Democratic Party Democracy People’s Party Genc Turnout Total 13.02 Others 25.98 14.27 49.80 20.88 8.36 46.58 SP 19.39 17.98 34.28 8.71 MHP Ð CHP Party Felicity Party Action AKP Nationalist Party People’s Republican Party Development and Justice name Party needns08 .05.24 1.00 0.87 Ð 5.42 Ð Ð Ð 15.41 Independents DP FP Party Democrat Party Virtue aoiyo needn addtsaespotdb eortcSceyPry(T) hc is which (DTP), Party Society Democratic by supported are candidates independent of Majority eortcPol at stescesrt epesDmcayPry hc a andb the by banned was which Party, Democracy People’s to successor the is Party People Democratic lists CHP the in elections the entered Court DSP Constitutional of the candidates by The banned was which Party, Virtue to successor the is Party Felicity aoiyo needn addtsaespotdb eortcSceyPry(T) hc is which (DTP), Party Society Democratic by supported are candidates independent of Majority DP for vote to supporters their asked and ANAP elections with from merge withdrew to ANAP attempt failed a in Party Democrat to name its changed DYP h hne neetrlpltc ruh bu eea motn usin:What questions: important several about brought politics electoral in changes The oesae %—9921.Source: (%)—1999–2011. shares Vote a e www.ysk.gov.tr; .651 .52.29 2.25 5.13 4.86 0.0100 0.0100.00 100.00 100.00 100.00 70 91 42 83.16 84.25 79.14 87.09 shares Vote Ð6.22ÐÐ 1.26 2.34 2.49 2011 Ð 2007 2002 1999 www.resmigazete.gov.tr .Shfil n .Demirkaya B. and Schofield N. c d b f 0.25 0.65 6.59 Ð g Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 efpaeeto oeso etrgtcniumi xlie yehi n sec- and (Carkoglu ethnic characteristics by (Onis socioeconomic explained 2000s than is rather in continuum Turkey differences left-right in a tarian on politics voters electoral of were the Self-placement democracies describing industrial advanced in in and sufficient space economic political not that discussed the been define has that It continuum issues time. left-right social over the and of countries content across substantive change may the Moreover, other. each across cut 2 Table Turkey in Elections of Model Spatial ac ewe h oiino h oesadtepltclprisadthe and parties political the model and (Schofield voters spatial valence the of pure position a the between with tance analysis our start We Elections of Model Spatial and 2 voters of position the in changes the trace to order in end parties. elections other 2007 the the Hinich is on Is- to and there MHP (Carkoglu the axis, nationalist between than nationalism Turkish in mean 2011 the and placed On electoral parties end left. the other one the to with the on closer on located albeit DTP is located axis pro-Kurdish CHP is religion AKP while elections. the parties 2002 on lamist and and right 1999 voters of separate the analyses that on spatial dimensions the primary in the parties as political emerge nationalism and ligion iia edr htaeidpnetfo hi oiypstos(Stokes positions policy their from independent are that leaders litical eortcLf at S 3 2.3 Ð (24.73) 136 DSP Party Path True Party Left Democratic ohradPryAA 6(56)Ð Total (15.64) 86 Others ANAP Independents Party Democracy People’s Party Motherland D Party Virtue Party People’s Republican Party Development and Justice name Party ainls cinPryMP19(34)Ð (23.45) 129 MHP F Party Action Nationalist e e l m i c o i t .I hsppr eapytesailmdldsrbdi h olwn section following the in described model spatial the apply we paper, this In ). c y r a P t et—9921.Source: Seats—1999–2011. a P r t a S y r t yD 2008 P11(01)Ð (20.18) 111 FP Y 5(54)Ð (15.45) 85 DYP HADEP K 6 6)31(95)37(59.45) 327 (59.56) 341 (66) 363 Ð CHP Ð AKP .Tevlnetr eest h oes ecpin fpo- of perceptions voters’ the to refers term valence The ). PÐÐÐÐ PÐÐÐÐ www.ysk.gov.tr; 9920 072011 2007 2002 1999 05)9(.4 6(.4 5(6.36) 35 (100.00) 550 (4.74) (100.00) 26 549 (100.00) 550 (100.00) 550 (1.64) 9 (0.55) 3 ÐÐÐÐ Ð 7 3.6 1 2.)15(24.55) 135 (20.4) 112 (32.36) 178 www.resmigazete.gov.tr M λ β) (λ, Ð Ð Ð 0(27)5 (9.64) 53 (12.75) 70 Ð hc nldstedis- the includes which 2006 coede al. et Schofield ; 1963 exogeneous Ð Ð Ð Ð Ð 2007 .I the In ). 2009 .Re- ). 319 ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 oe.I h on model joint the the In and model. coefficient spatial the given parties center small the the terms. whether from valence see relative diverge where and to LNE LNE, incentive a a any for have be search to we simulation, mean By electoral dimensions. for condition necessary a ss h oesae o eiiyPryadYugPrywr eo nte20 elections 2007 the in % 5 below were Party Young and Party Felicity Table a for (see have shares vote may The which ysis. results, election the DTP. to for term compared valence the sample on the influence deflating in underrepresented are voters (Schofield 3 a on conducted (WVS) 2007. Survey in Values World sample representative on nationally based elections 2007 analyze We Turkey in Elections 2007 3 where equals equation, the In voter that utility the model, 320 5 4 position on n h party the and point au itiuin h necp term intercept The distribution. value niae htte ol oefrapltclpryi h olwn lcin and elections following analysis. the factor in the in party used political questions a the for all vote answered would they that indicated em ntemdl coed( an valence Schofield relative be model. the would and the voter coefficient mean in spatial the terms the given of (LNE) position Equilibrium the Nash not Local or whether about coefficients. information the gives estimate to model logit characteristics multinomial the a by determined use not We voter. is it the that of sense the in exogeneous is valence The nlssbcuea h iesre a odce n ni h lcin,te eetodistinct two were they our elections, in the separately until parties and recom- profiles. two conducted leader voter these was different their include survey with but to parties time elections decided the the We at DYP. from because for withdrew analysis vote ANAP voters elections. their 2007 that the mended in DP to name its oito ( sociation eecue h oeso eiiyPry on at n eortcLf at rmoranal- our from Party Left Democratic and Party Young Party, Felicity DTP of voters, voters of the position excluded We the measure to used questions survey to response of levels low to Due ( v.20090901 aggregate official 1981Ð2008 Survey Values World ial,w noprt h eorpi hrceitc fvtr notespatial the into voters of characteristics demographic the incorporate we Finally, ecniu u nlsswt h aclto fcnegnecoefficient convergence of calculation the with analysis our continue We (θ z www.worldvaluessurvey.org .A xlie nTable in explained As 1). j j 2007 · equals η i ) ).  eest h oidmgahcvlneo voter of valence sociodemographic the to refers x u i ij − j (x splc position. policy ’s z i j u ,z  ij M eoe h ulda itnebtentevoter the between distance Euclidian the denotes j (x ) (λ,β,θ) i = i ,z ihposition with λ 2007 fe aldatmtt eg ihAA,DPchanged DYP ANAP, with merge to attempt failed a after 1, j j ) .AgeaeFl rdcr SPJS Madrid. ASEP/JDS, Producer: File Aggregate ). = + rvsthat proves ) h tlt fvoter of utility the , λ (θ j λ j − j  · j 3 η ie h xgnosvlneo party of valence exogeneous the gives β i sa ro etrwt yeIextreme I type a with vector error an is elmtoraayi otevtr who voters the to analysis our limit We )  x − x i i − β esfo oigfrparty for voting from gets  c< z x j i  − 2 sasfcetand sufficient a is 1 4 + , z .WrdVle uvyAs- Survey Values World 2009). 5 .Shfil n .Demirkaya B. and Schofield N. x j   i euefco nlssto analysis factor use We j 2 rmvtn o party for voting from +  ω j stenme of number the is i o party for c<ω i c sideal ’s j which with z j is j j . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 ifrfo h nlsso rvoselections. previous of analysis the from differ trace( with ls ieso scnieal mle.Temjrt fvtr r concentrated are Fig. voters of of part majority northern The the smaller. on considerably is dimension alism h w xsi .7.Tu h oe oainemti ste2 the is matrix covariance voter the Thus 0.073. the is on axes the two variance on the The north nationalism. to Turkish south with from x association movement increasing A life. an public indicates and private in religion 02eetosi Turkey. in elections 2002 fqetosaotteattdso oestwr eiinadntoaima well as nationalism and religion toward issues. list voters social long and of a economic attitudes with as start the We other. about each questions from voters of differentiate that issues the identify election 2007 the in positions party 1 Fig. Turkey in Elections of Model Spatial 8 7 6 eiinadntoaimaepicpldmnin htcaatrz h ideological the characterize voters. that Turkish dimensions of principal position are nationalism and religion h eiinadntoaimdmnin epciey h at oiin r given are positions matrix: party following The the respectively. by dimensions nationalism and religion the reli- toward attitudes the to atti- co- strongly the very the that related Second, implies DTP. gion. not which of pro-Kurdish are smaller, nationalism the part considerably toward of is southern tudes voters axes the two the on the are between concentrated whom variance voters of of most figure, group the another is There tionalism. rmlf orgto the on right to left from the and dimension e akguadHnc ( Hinich Appendix and in Carkoglu given Appendix See are in analysis listed the are of model loadings the factor and The analysis factor the in used questions The xsi .2 hl h aineo the on variance the while 0.729 is axis Figure h oiino ate scluae ytkn h enpsto fisvtr on voters its of position mean the taking by calculated is parties of position The oe itiuinand distribution Voter 1 ∇ hw h oiino oeswt the with voters of position the shows ) = 2.Tecvrac arxrvastoipratpit that points important two reveals matrix covariance The 1.227. y xscrepnigt h ainls ieso.Amovement A dimension. nationalism the to corresponding axis 2006 x 7 xsidctsave htfvr nicesn oeof role increasing an favors that view a indicates axis n coede l ( al. et Schofield and ) 1 6 ∇= ihhge eeso soito ihTrihna- Turkish with association of levels higher with iia opeiu tde,oraayi hw that shows analysis our studies, previous to Similar

7 0.498 0.073 0.073 0.729 y xsi .9.Tecvrac between covariance The 0.498. is axis 8 x 2011 is,tevrac ntenation- the on variance the First, xscrepnigt h religion the to corresponding axis 2  (Table o pta nlsso 99and 1999 of analysis spatial a for ) 6 ). × matrix: 2 1. y axis 321 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 h at ihtelws aec sDPwith DTP is valence lowest the with party The * 3 Table 322 oe,we l ate r oae tteeetrlma,tepoaiiyta voter a that probability the mean, electoral is the DTP at chooses located are parties all when model, infiat h aec em r acltdwt epc oMP h etrof vector The MHP. to respect with calculated is valences are relative terms valence The The significant. success. electoral Table their in summarized and are parties, results political of valence and position ideological our in and elections 2007 the in to population related the is be in may sample. MHP both this supporters however, nationalist DYP DYP; of than the number dimension small this years, on previous center the the to Turkish closer unlike with association and policies measure that to Interestingly, questions related used nationalism. questions we are of language, that lack as questions the such used issues to we on Due since however, analyses. parties cautious, previous of position are the We the from dimension. in different change this other a on parties as other of this all each position interpret and to to the hand, close above, very discussed one are As the DTP hand. on other than DTP the pro-Kurdish on nationalism the parties the other between On the mean. polarization in electoral a the parties is of pro-Islamist religion right there of the the axis, position to on the located AKP is to it compared of elections, center position previous minor the Although to relatively between. closer with with in is ends located dimension opposite elections the parties previous at other located the are the AKP all to and CHP similar axis, religion is the On differences. parties of position The ovrec Coefficient Convergence ohradPryANAP DYP Coefficient Spatial Party Motherland Party Path True n 1 AKP Party Development and Justice name Party ainls cinPryMPÐÐÐ Ð Ð 0. DTP MHP CHP Party Society Democratic Party Action Nationalist Party People’s Republican infiatwt probability with Significant = euetepr pta model spatial pure the use We 5;Lglikelihood Log 558; z ∗ uesailmdlfr20 lcin.Nraie ihrsett MHP to respect with Normalized elections. 2007 for model spatial Pure = ⎣ ⎡ y: x nationalism β : (λ religion at K H H T Y ANAP DYP DTP MHP CHP AKP Party AKP = =− ( ,λ 1.413, 603. <0. CHP 001 7 McFadden 57; ,λ 0.07 0.31 0.623, 3 MHP h pta coefficient spatial The . M λ β) (λ, ,λ 90.16 −0.09 0.03 −0.67 0, DTP −1.688, R ,λ 2 oetmt h eainhpbtenthe between relationship the estimate to = DYP 1.537 −1. −1. 0. λ −1. 0. . 413 623 658 114 −1.479, λ ,λ 676 479 688 DTP * * * ANAP * * * −1.40 −0.10 =− ) .Shfil n .Demirkaya B. and Schofield N. β −1.676) 8.Acrigt the to According 1.688. s068adstatistically and 0.658 is 0.061 0.302 4.138 0.269 10.93 0.36 0.151 0.129 t.error Std. .22 .04 −0.23 −0.46 ⎦ ⎤ −10. −5. −5. −4. | t-value| 551 507 684 758 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 ob nLEsneteuprbudi mle hn2 tde o,hwvr satisfy however, not, 1. does than It greater 2. is than bound smaller lower the is since bound condition upper mean sufficient the electoral the since the LNE for an condition be necessary to the satisfies coefficient convergence the ievle,gvn oa qiiru tteoii.I diinancsaycon- necessary a that addition is In convergence origin. this the for at equilibrium dition local a giving eigenvalues, ( Schofield evlfrtecnegnecefiin is coefficient convergence the for terval nevlfrtecnegnecefiin sn h pe on fthe of bound upper the using coefficient convergence the for interval epnet htidctdta hywudvt o T osiue25%o our of % 2.5 The constitute well. DTP as for DTP vote by would supported they sample. that not indicated were that that in respondents DTP candidates of includes supported share this vote but however, the elections assess 2007 to the difficult Table in 2007. is it participate therefore, not candidates; did independent DTP above, explained As ec ofceto h uesailmdl hc sgvnby: given is which model, spatial pure the of coefficient gence h 5%cndneitra for interval confidence % 95 the h tnaderrfor error standard The λ ρ Turkey in Elections of Model Spatial n h oe on of bound lower the and = =[ = = DTP DTP exp 0.024 h esa,o h hrceitcmti fDTP: of matrix characteristic the or Hessian, The coed( Schofield exp 22.225 is (1. ( [− 3.101) 413) 2.398, ute hw htif that shows further 2007) + 1 10.084 + 2007 hw htteidpnetcniae eevd52 ftevotes; the of % 5.24 received candidates independent the that shows + exp −0.978] exp hw htteHsino h T sgvre yteconver- the by governed is DTP the of Hessian the that shows ) (0. C c + ( DTP 2.311) 623) = = = λ 5.409 DTP ρ 1.537 2 2β( DTP = = = = n h 5%cndneitra for interval confidence % 95 the and × + 2 1.253

2β( 1 s03.Acrigy h 5%cndneitra for interval confidence % 95 the Accordingly, 0.36. is 0.658 + exp + 8 0 .091 −0.087 × n ievra h tnaderrfor error standard The versa. vice and − 0.091 exp 1 1 0.658 β 2ρ (0. − +

is ( × DTP c< 0) 7 0.498 0.073 0.073 0.729 1.688) 2ρ 1.232 exp [ ( + 0.538, × 1 DTP ) c< −0.376 .W aclt osraieconfidence conservative a calculate We 2. (−1. [ − trace( exp ( 1.188, 1 + ) 2 + ∇− − (−1. ,ta h esa ilhv negative have will Hessian the than 1, × 1.012] 688) 0.778] exp 2 ∇ 0.024) × 1.871]  ) I 688) (  0) 0.024) −1 hs h 5%cndnein- confidence % 95 the Thus, . − + + × I h ofiec nevlfor interval confidence The . exp exp 1.227 ∇− ( (−1. 0.209) I ρ 479) DTP + is + exp β β [ exp 0.01, 6 so 0.061 is coefficient ( 0.012) (−1. 0.05] 676) 323 −1 . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 The ecmaeti ovtssae norsample: our in shares votes to this compare We h ievle fteuprbudfrtecaatrsi arxae014wt the with 0.144 are matrix characteristic the for bound eigenvector upper the of eigenvalues The l h ate r oae tteeetrlma hi rdce oesae were shares vote predicted their When mean case. electoral our in the LNE at an located is as: calculated mean are electoral parties the the that all verify can we terms, valence one point. has saddle matrix determinant a negative gives characteristic bound a the upper and for the eigenvalue, bound negative upper one The elec- and bound LNE. the positive lower that be the implies should and which estimate mean eigenvalues, characteristic negative point toral the have the of matrix above, eigenvalues characteristic see the the we that for As is negative. LNE both be are to matrix mean electoral the for dition hscmaio siprata ttlsu bu hte h o aec parties valence low ( the Gallego and whether Schofield about mean. electoral us the to tells move it to incentive as any have important is comparison This ol aemr netv opl h at rmteeetrlma to mean electoral the activists from party the party than the attractor pull stable to a their incentive not than more higher is have are equilibrium would equilibrium an the If at shares. parties vote valence actual low of shares vote position predicted at equilibrium an call h ievle ftecaatrsi arxare matrix characteristic the ( of eigenvalues The 324 aeacndneitra o h esa sn h pe on fthe of bound upper of the bound lower using Hessian the the and for interval confidence a late h eigenvector the e ntevcos h qiiru tteeetrlma sntasal trco for attractor stable a not is mean electoral the DYP. at or equilibrium DTP the vectors, the in see −0.961, smnindaoe coed( Schofield above, mentioned As ysmlto ae ntepitetmtso h pta ofcet n the and coefficients spatial the of estimates point the on based simulation By C eigen DTP = = = = −0.278) v le ftelwrbudfrtecaatrsi arxare matrix characteristic the for bound lower the of alues 0.968

2 2β( ( −0.961, 9 0.071 −0.294 × 0.071 1 0.538 ( −0.96, −

ρ (s 7 0.498 0.073 0.073 0.729 and 2ρ z AKP 0 = −0.277) × = = DTP −0.518 ρ 0 ihteeigenvector the with −0.403 ( ,s −0.279) ( ( 0.556, DTP 1 0.543, ρ ) CHP ∇− AKP z − 0 n ieversa. vice and 2 ,s ,ρ and × z  I 0.231, 0.246, MHP , CHP z a  and 0.05) 0

tbeattractor stable − 2007 7 ihteeigenvector the with −0.273 1 0.111 0.111 0.112 ,s ,ρ I, 3 ihteeigenvector the with −0.539 0.134, DTP ∇− 0.132, MHP z 1.525 hw htancsayadsfcetcon- sufficient and necessary a that shows ) 0 ,s ,ρ I, −0 DYP 0.025, 0.024,

DTP 2 z .241 0 7 0.498 0.073 0.073 0.729 × ,s ,ρ ANAP 0.778 hntelwr9 on of bound % 95 lower the when  0.03, DYP z 0.03, 6wt h eigenvector the with −0.06 0 ( ) −0.278, ,ρ .Shfil n .Demirkaya B. and Schofield N. × 0.023) 0.025) ANAP z ( 0 1 −  − 2 0.961) ( −0.277, × I ( ( −0.15) −0.279, 0.01) 7 with −0.273 β ecalcu- We . 2011 coefficient z ∇− ∗ 0.961) Hence, . .Aswe 0.96) 190) , I . . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 qiiru oiino h eiinai n h atrtksapsto a othe to the far shares of vote position left predicted as: a The calculated the takes axis. were to nationalism equilibrium latter the the far the on at position position and equilibrium a axis the takes in of religion south former seen the The are on DTP. differences position and obvious equilibrium CHP most the from The of pulled activists. positions are party parties the the the direction by which point in and equilibrium much how shows matrix This ae h au o aaadKrih n o uks n l te languages. other all and Turkish for 0 that and variable Kurdish, dummy and a Zaza as for it construct 1 and value home the at takes lan- speak primary respondents the the by socio- ethnicity that and measure guage We ethnicity variables. education, independent as age, status include economic we studies, previous predicted Following higher voters. a provides ANAP. equilibrium and the DTP CHP, shares for vote share vote sample the to Compared h nta at oiin n h at oiin tteeulbimi ie ythe by given is equilibrium between the difference at The positions graph. party the the matrix: of following and southwest positions the party on initial is the DTP and mean Fig. electoral in seen be can As equilibrium 2 Fig. Turkey in Elections of Model Spatial 9 e Appendix See z ial,w upeettesailmdlwt h eorpi hrceitc of characteristics demographic the with model spatial the supplement we Finally, yuigsmlto,w on nte N ihtefloigprypositions: party following the with LNE another found we simulation, using By ∗ − z oa Nash Local z 1 1 = = ⎣ ⎡ ⎣ ⎡ y: 2 y: o h ito usin sdt esr eorpi characteristics. demographic measure to used questions of list the for x nationalism x : nationalism : religion at K H H T Y ANAP DYP DTP MHP CHP AKP Party ρ religion at K H H T Y ANAP DYP DTP MHP CHP AKP Party z 1 = = l h ate te hnDPaecnetae rudthe around concentrated are DTP than other parties the all 2, ( 0.539, ρ AKP z 1 40.05 0.04 0.02 50.08 0.05 0.02 ,ρ 0.05 0.29 0.245, CHP z 1 ,ρ 30.11 −0.13 −0.72 0.131, MHP z 1 ,ρ 0.03, DTP z 51.14 −0.05 1 ,ρ 0.03, 20.12 0.12 −1.24 DYP z −0.70 1 −0.70 ,ρ 0.025) ANAP z 1 50.05 .05 −0.08 .17 −0.29 −0.58 ⎦ ⎤ ⎦ ⎤ 325 9

B&W IN PRINT Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 * ** *** 4 Table 326 pcfial,Aeivtr eemr ieyt oefrCPcmae oohrpar- other to compared CHP for vote al. et to (Schofield likely ties More more choice. vote were and voters sect Alevi religious specifically, between relationship a to point studies Previous eas h usinwsntakdt h respondents. the to asked not was question the because h eorpi hrceitc,teol n hti ohsbtnieyadstatisti- and substantively both is that one However, Among significant. only significant. statistically the are statistically characteristics, DTP still for demographic one is the the except it terms but valence the model of spatial none pure the than smaller Kurdish oi-cnmcStatus Socio-economic Age Education n Coefficient Spatial Variable eaieValence Relative infiatwt probability with Significant = infiatwt probability with Significant infiatwt probability with Significant ntejitmdl hc ssmaie nTable in summarized is which model, joint the In 5;Lglikelihood Log 558; on oe o 07eetos omlzdwt epc oMHP to respect with Normalized elections. 2007 for model Joint λ k β 2011 =− 565. <0. .W eental oicuerlgossc noranalysis our in sect religious include to able not were We ). <0. <0. T 4 DTP CHP K 1 AKP NP0. ANAP ANAP AKP Y 0 DYP DYP 0 AKP DTP CHP ANAP 0 DYP 11 ANAP DTP DTP AKP at ofcetSd error Std. Coefficient Party AKP T 0 0 DTP 11 CHP DYP H 0 CHP DYP 0 0. CHP ANAP 05 ;McFadden 6; 01 001 R −14 −14 2 −0 −0 −0 −5 −1 −0 −0 −0 −0 = 0 0 ...... 0. 653 5 1 359 8 1 486 541 0 113 5 0 252 0 228 025 183 171 965 063 4329 583 314 60 36 0 285 227 9 1 694 0 0 004 3083 032 571 8 0 288 527 0 118 025 603 17 * *** * ** * * * *** h pta ofcetis coefficient spatial the 4, .Shfil n .Demirkaya B. and Schofield N. 4329 3083 0 1 0 0 0 2 0 0 1 0 0 ...... 441 4 1. 045 61. 36 209 0 0. 305 193 811 1 0. 811 484 288 625 228 3 0. 032 0. 355 354 7 1. 174 1. 0. 107 025 1 2. 2. 013 012 229 066 4 2. 145 3. 245 3. 095 019 −0. −0. −1. −0. −0. −0. −0. −0. −0. −2. −9. −2. | t-value 249 423 542 503 181 826 003 003 744 988 72 565 109 004 005 651 104 995 326 167 411 205 392 164 738 266 | Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 rvdsabte fit. better a provides al 5 Table baseline the decade. as last DYP Table the using in by during results model Turkey the the in rerun summarized politics we and electoral comparison, en- facilitate in elections to 2002 change order and the In 1999 trace of to analyses previous us with ables results our of comparison A Elections Previous with Comparison 4 compare we If MHP. nationalist to the likely is which McFadden more baseline, the are the to speakers compared Kurdish DTP surprisingly, for vote Not ethnicity. is significant cally Turkey in Elections of Model Spatial 11 10 a obt H n H.Ti a eepandb h odpromneo AKP’s of performance good the by explained be can policies. This MHP. economic culmi- compared and increased the AKP CHP to of valence points both The DYP to parties. to center-right compared of decline parliament the the of in nation parties three the of valence norsml euti ag tnaderr ntejitmodel. joint the in errors standard large in result sample our in rmr eemnn fbt at dnicto n at rfrnefrAPvoters. AKP for preference party and identification party both of determinant primary ec ofcetfo . n20 o15i 07 hc mle nincreas- an implies which 2007, in 1.5 to 2002 in 5.9 from coefficient organiza- gence and financial providing in also activists MHP—can party and role resources. CHP the leaders. tional parties—AKP, party by three the explained all of be competence for of partly the terms part as perceived as valence is thought positive it policy be extent The economic can the toward policy to Economic term attitudes valence voters. coherent the among any variance preferences. detect the voters’ explain not that effect did that analysis policy factor Our economic about considerations ideological ovrec Coefficient Convergence Party Left Democratic Party Motherland Party Society Democratic Party Action Nationalist Party People’s Republican Party Development and Justice name Party pta Coefficient Spatial h nre o 99ad20 r h eut fteaayi nShfil ta.( al. et Schofield in analysis the of results the are 2002 and 1999 for entries The na nlsso 07eetos aacol ( Kalaycioglu elections, 2007 of analysis supporters an DYP of In number small the because baseline the as MHP use we section, previous the In n fteciia nig forcmaio stedces nteconver- the in decrease the is comparison our of findings critical the of One oprsnwt rvosyears. previous with Comparison R β 2 ftepr oe otejitmdl eseta h on model joint the that see we model, joint the to model pure the of 11 ti motn ont,hwvr hti spatclrte than rather practical is it that however, note, to important is It a omlzdwt epc oDYP to respect with Normalized 5. 1999 0.336 1. −0. 0. 0 0 Ð 0. 10 . . 724 734 49 666 375 h osdrbeices nterelative the in increase considerable The 071 ∗ onsta cnmcstsato sthe is satisfaction economic that points 2010) ∗ ∗ ∗ ∗ 2002 −0. 0.43 5. 1. ÐÐ −0. 1. 0 . 78 94 52 33 31 21. 12 ∗ ∗ ∗ ∗ 2011 ) 2007 −0. −0. 1. 0. 2. 2. 479 54 659 102 893 327 197 209 ∗ ∗ ∗ ∗ ∗ Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 soito (www.worldvaluessurvey.org Association 1 o togyd o ge rdsge ihec ftefloigstatements? following the of each with disagree or agree you do strongly How (1) Secularism you would list this on party which for tomorrow, election national vote? a were there If Choice Vote Survey Values World on based is paper this in elections 2007 (WVS). of analysis The Questions equilib- Survey the 1: Appendix in par- axis the this of on None point. mean equilibrium electoral on DTP the the DTP elections. of from except 2002 south diverge parties to the rium. DTP compared to of other are except is positions AKP each that ties initial and to position The DTP closer a mean except equilibrium. got takes electoral parties axis the of all elec- nationalism right of of the the the position left that to initial and the argue point The We mean to equilibrium axis. space. electoral the religion ideological the pulls the an- the to on AKP found of close of also southwest aligned we strength elections.We the DTP toral simulation, 2007 except in using in parties located By LNE all DTP mean. an with gives electoral LNE mean other the electoral to that convergence verified of likelihood ing 328 12 you Would life. your in is it important how indicate following, the of each For (2) he—i otsbfr h 07eetos h usin sdi u analysis our in used questions The elections. following: 2007 the are the before months three—six ol ausSre 9120 fca grgt .0991( v.20090901 aggregate official 1981Ð2008 Survey Values World b twudb etrfrTre fmr epewt togrlgosbeliefs religious strong with people more if Turkey for better be would It (b) Dis- Strongly Disagree, Disagree, Nor Agree agree Neither Agree, Agree, Strongly a ti eyiprat ahriprat o eyiprat o talimpor- all at Not important, very Not Religion important, tant? Rather important, Very is it say a oiiin h ontbleei o r ntfrpbi office. public for unfit are God in believe not do who Politicians (a) 12 edpbi office. public held h uvywscnutdbtenJnayadMrh20,ta is that 2007, March and January between conducted was survey The .AgeaeFl rdcr SPJS Madrid. ASEP/JDS, Producer: File Aggregate ). .Shfil n .Demirkaya B. and Schofield N. 2009 .WrdVle Survey Values World ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 al 6 Table Loadings Factor 2: Appendix to belonging as themselves describe home? sometimes at People speak Status: normally Socio-economic you do (4) language What 1—No Language: attained? years have (3) you . . that . level are educational you highest means the is This What please? Education: birth, of (2) year your me tell you Can Age: (1) Characteristics Demographic world. the to relate they how and Proud, themselves Very about Not views Proud, different have Quite People Proud, Very (2) Turkish? be to you are proud How (1) Nationalism Turkey in Elections of Model Spatial akguA(07 h aueo etrgtielgclsl-lcmn nteTrihcnet Turk context. Turkish the in self-placement ideological left-right of nature The (2007) A Carkoglu References ato h nation the of Part nationality of Proud Variance uuaieVariance Cumulative eiiniprati life in important Religion office public in beliefs strong with People god in belief Politicians’ n = td8(2):253Ð271 Stud 8 eiinNationalism Religion 588 ecieyusl sblnigt h:1Uprcas pe ideclass, middle Upper you 2 Would class? class, Lower class. 5 Upper lower class, Working 1 or 4 the: class, upper middle to the Lower 3 belonging or as class, yourself middle describe the class, working the Degree Education—With Level 9—University Education, Formal old. of part Disagree as Strongly myself Disagree, see Agree, I Agree, yourself? Strongly see each nation. with you Turkish disagree how the or about agree statements you following strongly how the me of tell you would card, this Using Proud All At Not atrloadings Factor 0 0 0. 0.748 0 0. 0. . . . 270 071 478 106 270 738 0 0 0 0 0 0 0 ...... 403 656 246 405 133 092 064 329 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 ol ausSre 9120 fca grgt .0991(09 ol ausSre As- Survey Values World 57(2):368Ð377 (2009) Rev v.20090901 Sci aggregate Polit official Am va- 1981Ð2008 competition. Survey a party Values support: of World models popular Spatial for (1963) Competition DE (2011) Stokes A Berlin Zakharov Springer, politics. U, of Ozdemir soul and M, heart Gallego the N, chaos: York or Schofield New Leadership (2011) Routledge, M politics. Gallego convergent of N, model for Schofield spatial conditions The sufficient (2008) and N Schofield necessary theorem: voter mean The (2007) N Lynne Schofield consolidation. democratic to challenges politics: Turkish the Contemporary and (2000) party E development Ozbudun and justice the crossroads: the at globalism Conservative restitution (2009) or Z Islam Onis of resurgence Helm: Stud the Turk at Elect 2000s. party development the and preferences. Justice in (2010) Turkey party E in Kalaycioglu Turkish organizations party of to orientation analysis Attitudinal spatial (2008) E A Kalaycioglu (2006) MJ Hinich A, Carkoglu 330 sociation. 36(3):451Ð482 Welf Choice Soc Turkey. in elections of model lence 74(3):965Ð980 Stud Econ Rev equilibrium. Boulder Publishers, Rienner 14(1):21Ð40 Polit Mediterr Turkey. in consolidation democratic to path thorny 11(1):29Ð44 Stud Turk party? predominant right-of-center the of 9(2):297Ð316 Stud 25(2):369Ð392 www.worldvaluessurvey.org grgt iePoue:AE/D,Madrid ASEP/JDS, Producer: File Aggregate . .Shfil n .Demirkaya B. and Schofield N. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 .Aas( Peltason Adams W. J. Jack the Irvine. by California, and of research Ferris) University SSHRCC Stephen Chair, by Foundation) J. supported (Bren and Winer was Stanley author (co-PIs third-named is #410-2007-2153 the authors grant by of project listing Clover this The to on assistance. Work indebted bibliographic alphabetical. are We for data. Ludeman who POLIDATA previous election Sue of a aggregate and Bensen on quality Behrend-Gethard Clark comments high to helpful with thanks for us special Butler provides owe Dan routinely also to We and manuscript. interest, the DW-NOMINATEdata of of Poole-Rosenthal period version us the to for available House making the for for Poole Keith to indebted are We attributed com- commonly plurality two-party is of that portrait politics simplistic tweeedledum-tweedledee the in from resulting way petition long a come have We District of Effects the About Wisdom Conventional Evaluating 1 III Merrill Samuel and Grofman, Bernard Brunell, L. Thomas Adams, James Politicians? Centrist Produce Necessarily Districts Competitive Do DOI (eds.), al. et Schofield N. e-mail: Wilkes-Barre, University, USA Wilkes 18766, Science, PA Computer and Mathematics of Department III Merrill S. USA 92697-5100, Democracy, CA e-mail: of Irvine, Study Irvine, the California, for of Center University and Science Political of Department Grofman B. USA 75080, Dallas, TX at Texas Richardson, of Road, University Campbell Sciences, W. Policy 800 and Political Economic, of School Brunell T.L. USA 95616, CA Davis, Davis, California, e-mail: of University Science, Political of Department fteUS Congress U.S. Members the the of Among Convergence Party on Composition 10.1007/978-3-642-35239-3_16 [email protected] [email protected] [email protected] B ) dacsi oiia Economy Political in Advances pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , , 331 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 eaoso h aesaefo ia ate eg,PoeadRosenthal and Poole (e.g., al. parties rival et between from man differences state ideological same the the on of work Senators including States, own United their the in mobilizing “base”). petition party’s by the but as position of think voter we median what the (i.e., supporters to partisan appeal by general not in elections support maximize candidates circumstances, plausible empirically certain n ue fet at iegne(ebradMorton and (Gerber 2001 divergence party affects rules ing i elcinwe hi oigrcrscicddwt hi tt at’ platform party’s state their to with elec- likely coincided Senate more records U.S. voting were of their Senators when study incumbent reelection that a win conclude In they constituencies. 1962Ð1990, partisan from their tions to appealing from grounds elections policy ( general in al. on responsive benefits et electoral are Schmidt derive Congress example, candidates For of that constituents. pothesis members their which of to ideology degree the the to predecessors at their and resemble behavior congress voting of members in elected newly which to degree the ne fsaelvlRpbia osiunis n htti ifrnei o natfc fthe of do artifact than an views conservative country. not more the prefer- is of substantially rest hold policy difference the South this the from the Democrats that from are partisans and than Democratic constituencies, that heterogeneous fact Republican more state-level substantially of are ences constituencies partisan cratic elections. ity ( Grofman e.g., see, painted, ftetoparties. two the of 3 2 1 ( Downs to 332 6 5 4 erl ( Merrill oiiin’sneeplc motivations, policy sincere politicians’ logic. constituen- Downsian legislative standard multiple the of affects occurs existence competition the incentives which that competing in (2) cies strong and, are divergence; there party convergence, for party for pressures Downsian oedvretelctoso h einvtrars ifrn itit,()the modify their to (b) in voter leeway district, districts, median the own the different have accommodate to officials across behavior platforms/legislative candidates/elected the voter policy which (a) their median to larger the extent be of the to greater locations expected the are diverse convergence more party for competition constituency oes atsnlylis h heto btnindet leain n otof host a and alienation, to due abstention of factors. threat other the loyalties, partisan voters’ a ( man eotaaye ugsigta h oiypeeecso tt-ee Demo- state-level of preferences policy the that suggesting analyses report (1999) al. et Grofman ( al. et Winer ( Grofman See (1998 Morton and Gerber (1983 Wittman e.g., See (1957 own Downs’ hr sas xesv miia upr o at iegnei w at com- party two in divergence party for support empirical extensive also is There o xml,w o eonz h hoeia oeta o iegnedeto due divergence for potential theoretical the recognize now we example, For ;AasadMril(2008). Merrill and Adams 2006); ,adabd fwr aigbc slata a sFoa ( Froman as far as least as back dating work of body a and ), 2003 1990 1957 ;seas ndr(1994). Snyder also see 2008); 5 o eetrve ftetertclltrtr nprydvrec nplural- in divergence party on literature theoretical the of review recent a for 2004) 4 ,Bte (2009 Butler ), n c h rae h ifrnei ainei h upr bases support the in variance in difference the greater the (c) and hoeia eerhas ugssta h osqecso multi- of consequences the that suggests also research Theoretical ,wr httsshptee bu h xett hc rmr vot- primary which to extent the about hypotheses tests that work ), ). 1 iw fprycnegneae oee,frls ipitcta often than simplistic less far however, are, convergence party of views ) 6 w e oictosaetercgiinta 1 paralleling (1) that recognition the are modifications key Two e-oninmdl ftetp inee yAasand Adams by pioneered type the of models Neo-Downsian ;Goels (2001). Groseclose ); ;Bre (2001 Burden ); 2004). ,Mle n coed( Schofield and Miller ), , 2004 2 addt oiainrules, nomination candidate ;GomnadBuel(2001 Brunell and Grofman ); 2003 1998 eosrt ht under that, demonstrate ) rfa n Brunell and Grofman ; 1996 okn at looking 1963) 3 ;Oe n Grof- and Owen ); at activists, party .Aase al. et Adams J. ettehy- the test ) 1984 Grof- ; Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 osiuny ota h rae h upr o eortcpeieta nominees presidential Democratic for own support their the of greater views the expected the the that toward find so holders constituency, liberal office also pull we more Moreover, that considerably districts. effects average, support, such constituency-specific on presidential from elected Democratic are, Republicans of districts legislativethan level such different given from sharply any elected for display Democrats that, ideology such similar records, of voting districts from elected are ics eo,orfidnshv motn mlctosfrter-ae oesof we models as theory-based (although, for competition). theoretical implications candidate important than have rather of findings descriptive our dimension and below, discuss first empirical officehold- is the by focus taken use positions Our policy we ers. the under- and of the measure district, our of the as measure DW-NOMINATE scores in our predisposition as take We ideological nominees 1956Ð2004. presidential lying period the Democratic over for Senate records. levels and voting House support legislative U.S. the their Demo- for in data among reflected analyze We similarity as officeholders, ideological Republican of and degree surrogate cratic a the as ideology—for view we district/state of level—which implications for district/state the the at at look we patterns individ- particular, voting those In presidential of Congress. locations of ideological members the become of who record uals empirical the in reflected are ing districts. similar in positions Republican with positions Democratic su ra,ntwehrtecniae aesmlrpstoso hs sus(e,eg,Sigelman e.g., (see, issues these on positions policy similar same take the candidates Buell emphasize and the candidates whether rival not whether areas, around issue revolves which campaigns, political voter changing districts. to uncompetitive adjust than to rather likely competitive lop-sided more in for are time legislators than over districts districts ideology within (moderate) that competitive finds for further He steeper districts. is advantaged. vote electorally presidential be (normalized) to to appeared voter median sug- the to analyses to closer of closer sets were were other who who while nominees nominees elections, presidential the general the that in analyses gest better of fared sets position ideological some party’s in their that report they i.e. results, 7 more were constituencies party reelection. state fur- for their run and, to to position, likely voter’s appealed state who median senators the that reflected thermore, records voting their when than Politicians? Centrist Produce Necessarily Districts Competitive Do 9 8 htGifi hw,ised sta h vrg dooyo ersnaie (rather representatives of varies ideology parties) average between other. differential the the the that on than is ones instead, uncompetitive shows, and Griffin hand one What varies the districts on similar districts in differ- competitive office-holders policy between Republican the how and of Democratic question between main ential our address not does however, analysis, fin’s lg ais n htti eaini oepoone mn eeal moderate generally among districts. pronounced uncompetitive more among is than relation districts this that and varies, ology u vdnede o pa oacretlvl eaeoe su wesi n ilgein dialogue and ownership issue over debate lively current a to speak not does evidence Our ideology representative average relating districts over slope the that finds Griffin particular, In ( Lotfinia and Kenny however, elections, presidential postwar of study a In rfn( Griffin xcl sepce,w n htrpeettvsfo poiepriswho parties opposite from representatives that find we expected, as Exactly nti sa eso httertclyepce atrso addt position- candidate of patterns expected theoretically that show we essay this In 2004 2006 Petrocik ; rusta ititcmeiieespooe epniees Grif- responsiveness. promotes competitiveness district that argues ) 1996 9 alne al. et Kaplan ; 7 2006). cosdistricts across 8 rfn oee,de o compare not does however, Griffin, stemda oe ide- voter median the as eotmixed report 2005) 333 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 qal aacd lhuhteagmnso dm ta.adButler and al. et Adams of arguments are partisans the when ( Although greatest al. is balanced. divergence et equally ideological Adams i.e., in pattern, 1 expected the Figure picts candidates. policy Republican maximal and motivate Democratic will tween compositions absten- partisan that, and argument, balanced loyalties opposite partisan with the voters’ advance alienation, for to account due who tion (2010), partisans. al. Republican et and where Democratic Adams i.e., of Indeed, compositions, proportions partisan equal balanced approximately with are districts candidates there in Republican maximized be and not Democratic need vote-seeking between convergence policy h oeigo dm ta.( al. et Adams of modeling type. that the of districts ideological from elected the parties in different par- differences two underlying to the the lead of representatives district in closeness/competitiveness a of of separately, also degree characteristics party We the each tisan from district. at whether evidence looking the at offer than look in rather and particular, we voting In periods presidential Congress. time of of different Houses function both across a found as differences parties consider two the of ers officeholders Republican and Democratic both are districts. liberal such more from the district, the in 334 10 addtsi em ftelcto n ieo addts ae n rprino wn voters. swing of proportion and bases candidates’ of size and location the of terms in candidates supin,w vlaetehptei httedfeec nplc positioning policy in difference the Republican these that and Under Democratic hypothesis between outcomes). the election evaluate presidential ideolog- we same from assumptions, the inferred with (as districts compare from characteristics we winners charac- ical Republican candidate when candidate and idiosyncratic out Democratic a treat washing of We voter effectively set district. the as median the advantages the win incumbency to of and able location teristics our be the and still seats, and from uncompetitive be different in can how party minority about the is of interest candidate the of locations the as districts competitive in large as dif- least districts. from non-competitive at candidates in be winning should of differences these positions parties, ideological im- ferent the arguments authors’ comparing lose), these when must candidates, that whom winning ply of only (one consider candidates analyses rival our between while divergence policy of degree the h oe n ulct tby oapa ols dooial-retdvtr (see voters ideologically-oriented less to appeal using to and buys base settings, it interest-group competitive publicity and and activist such chances, an money election in from the their support because, improve financial to is closer gaining seek as This are to such district. options winners strategic the that various in have imply candidates voter necessarily median not the does to seat competitive potentially a presidential the where districts in non-competitive. as is competitive outcome is outcome presidential the where sn ititlvletmtso h oe itiuin ulr( Butler distribution, voter the of estimates district-level Using u rmr ups st netgt h hoeia xettosdrvdfrom derived expectations theoretical the investigate to is purpose primary Our officehold- among similarity ideological of degree the at closely more look We efcso inr eas ercgieta doycai atr a drive may factors idiosyncratic that recognize we because winners on focus We ntercn hoeia oes nietesadr oninmdl en in being model, Downsian standard the unlike models, theoretical recent the In 2005 ,Bte (2009 Butler ), winners hudb tlata ag ndistricts in large as least at be should ,that (2010), al. et Adams and ), 2009 xlisplrzto among polarization explains ) eei paribus ceteris divergence gap .Aase al. et Adams J. 10 de- 2010) districts , pl to apply between be- Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 esdwe h lcini otcmeiie efidta,cnrr oteintuition the to contrary that, find We competitive. most is election the when persed 11 the from activism distant and more turnout increasing being by for opponent compensate their than can voter Candidates median district. move that more potentially electorate—they in typically a voter general in are the activists—who candidates than potential When it. from polarized mobilize support one’s further seek to to district appealing trying competitive by and or opponent, base the party attacking own by likeability, or competence own Miller and Schofield e.g., Politicians? Centrist Produce Necessarily Districts Competitive Do 12 rae egtt h oiypeeecso h ebr fteronpria osiunyta to than constituency. constituency candidate’s partisan rival own the their of of members members the attach more the of candidates are of preferences candidate, voters preferences party’s the marginal partisan policy opposition most that the the Given is for to support. voter than weight the to candidate greater party’s likely hence their and most for 0.5, is vote nearer she to one that likely the candidate for be the voting must to of candidates respect probabilities the voter’s with voter’s of the the other of take the higher or will the one abstaining, candidate of the probabilities more nonzero the Romero have voters and candidate, (Erikson a account for into vote preferences to decision the voter’s affect a likely committed will and contest House Miller the and in Schofield vote (cf. to activists decision potential of the efforts affect may they turnout, actual nteeetrt) hl goigteplc rfrne ftervlpryspartisan party’s rival the of preferences policy the constituency. voters ignoring independent of while any electorate), preferences of the preferences policy the in the with to (along constituency weight partisan party’s give own rival her par- to the motivated own from is her support candidate only attract partisans—each of can members candidate the neither each by constituency—and scenario, tisan participate this in to Because, decisions ap- they position. influences also if policy candidate are only candidate’s participate preferred voters they their that partisan so of activism, where prove and/or can- but voting party’s from rival positions, abstain the candidates’ to to prone the candidate party’s of their regardless prefer didate, invariably they that Adams strong true, be so diver- might increased this (2010 why motivates al. intuitively which see et To constituency, positions. candidates’ partisan the own of gence his/her to part for large positions other. similar the to over converge party one to favor expected strongly be that can contexts partisan they extension, sce- uncompetitive by “perfectly” and, this nario, in Therefore— positions identical others. contribute—margin-maximizing to no converge to activists will are and candidates there vote will to since candidates citizens partisans, both courting these while then even to partisans, grounds the Democratic policy with are identify on citizens electorate appeal all the say, in If, citizens party. all which same in that namely context, election ini otcmeiie dm ta.( al. et Adams competitive, most is tion faithful. party own oegnrly sn odtoa oi oe,Aase l ( al. et Adams model, logit conditional a using generally, More affect greatly not do candidates House the of positions the if even elections, House competitive In u miia nlssspotti xetto htcniae a ems dis- most be may candidates that expectation this support analyses empirical Our ycnrs,i opttv itit,ec addt smtvtdt pelin appeal to motivated is candidate each districts, competitive in contrast, By ogi nuto bu h addtsmgtb otdsesdwe h elec- the when dispersed most be might candidates why about intuition gain To osdraohreteestainweevtr’pria issare biases partisan voters’ where situation extreme another consider ) 12 2007 .Sc napa a eutfo mhszn one’s emphasizing from result can appeal an Such ). 1990 2010 .10) natocniaeeeto where election two-candidate a In 1107). p. , 2007 rtcnie the consider first ) ). further 2010 wyfo h median the from away ru httemr un- more the that argue ) least 11 mn their among competitive 335 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 rsdnil ahrta os/eaecnetseicwy ee eas DW- because Here, way. contest-specific House/Senate a than rather presidential) losers. oeielgclycnrs hnmmeso htsm at lce rmsf seats. safe from ( elected al. party same et that Ansolabehere of members than centrist ideologically more tidctsta eort n eulcn lce ncmeiiedsrcsaeat are districts competitive in elected polarized Republicans as and least Rather Democrats Congress. of that members indicates extremist most it the elect districts opinion. competitive district most most the of the present center to their the maximize tend to presi- to relative candidates in incentives winning that, possible the strongest districts support, the in electoral candidates greater give or think might great competitive. one most as the typically are win- terms, is the support parties between dential two difference ideological the together the from district, closer ners that parties in opposite voter of median candidates the toward pull should districts competitive that 336 15 14 13 are factors other why empir- reasons and several theoretical developed subsequent moving However has by opinion. support research constituency their ical of enhance center parties and the candidates to equal, being factors other groups? appealing interest more special be to may also and that base positions their noncentrist of to presenting center the by towards or converging opinion, by district support ( their Butler maximize they of believe models candidates theoretical ( al. the et to Adams only and not relevant is districts competitive districts. lopsided in elected those ( noncentrist important. to is advantages finding electoral win- this believe suggest are the We there policies positioning. that believe radical themselves below more candidates present present the districts we that competitive findings more empirical in and candidates The activists, ning party groups. by to or interest partisans, electorate, special their district to to the appeal of that positions center noncentrist electoral the presenting their near enhance themselves candidates positioning elections, by real-world prospects in whether, obvious not is it snt“adrn”t oesi h district. the she in that voters demonstrate to positions “pandering” noncentrist not announced is her because character perior rust nnetecampaign; the finance to groups at ciit owr ntecniaescampaign; energizing candidate’s viewpoints; the noncentrist on hold work who to activists supporters party party motivat- include among These turnout positions. noncentrist ing presenting for candidates reward may 2001 e alne n ike( Wilkie and Callander See ( Baron See ( Sened and Schofield See, nieteeadms te uhr,w en optto nantoa (i.e., national a in competition define we authors, other most and these Unlike sw oe bv h ai oninmdlpoie togitiinta,all that, intuition strong a provides model Downsian basic the above noted we As contesting candidates of extremism policy the on research empirical believe We u nlssi locnitn ihteeprclfidnso noaeeee al. et Ansolabehere of findings empirical the with consistent also is analysis Our ,wofidltl upr o h li htwneso opttv ae are races competitive of winners that claim the for support little find who ), 1994 n on(2004). Moon and ) 2005 relative 2001 icse bv,btas otemr eea usin Do question: general more the to also but above, discussed ) 2006). 2007 oka h ereo iegnebtenwnesand winners between divergence of degree the at look ) oec te—u o eesrl oeextreme—than more necessarily not other—but each to ). 14 n,cnicn oesta h addt so su- of is candidate the that voters convincing and, oeta hsfidn osntipythat imply not does finding this that Note 15 ie hs hoeia considerations theoretical these Given ipypt neeto otxsthat contexts election in put, Simply 13 oiaigseilinterest special motivating aia policies radical not .Aase al. et Adams J. qa,and equal, 2009) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 16 data for 2004, through ideology 1956 period. member’s from 1956Ð2004 elections a House the of the over measure for pooled our districts district as House and scores U.S. DW-NOMINATE extremism Taking for candidate data between relationship using the competitiveness, analyzing by begin We Party by 1956–2004, House, U.S. the in Extremism Ideological 2 in analyze we which data divergence. of ideological kind for the pressures that about informative believe the is parameters. of to paper regression positions reasons this ideological the good the are of on there estimates dependent Thus, biases part in which is candidates, election proportion Congressional each Democratic in the vote because this the problem elections, of endogeneity House/Senate the an in pe- in shares introduces if votes time vote Specifically, regression substitute on shares. longer scores vote DW-NOMINATE we much presidential regress if a the we arise for over problems themselves have districts elections statistical we of House/Senate potential and number addition, districts larger In the compa- of a riod. more predispositions over provide In partisan we data underlying candidates. districts comparable the all and for on challengers data data both level rable data of presidential few a locations using only by ideological are contrast, there these the that for contain fact controls the that by partly sets exacerbated only is incumbency problem con- of This for effects). skills House controlling other campaign given (and and any candidates backgrounds hand, the two other as the the such On election features elected. the idiosyncratic involves of is competitiveness test officeholder the given at a directly which comple- look they in as seats, seen House be for ( should contests of they al. way—and et most. contest-specific Ansolabehere matter a When who compe- or mentary. winners defining national the approaches—i.e., a is both in it to tition disadvantages course, look and of we advantages But, losers, are winners. for There of not positions but the winners at for only available generally are scores NOMINATE Politicians? Centrist Produce Necessarily Districts Competitive Do 17 htaeesnilyflt etk hsa vdneo infiatedgniyefcs(aaomitted (data effects endogeneity significant of lines evidence give as done this reasons). have races. take space We House we for flat. district that essentially in vote-shares vote-share are House Democratic that against to scores of DW-NOMINATE position slope half of spatial positive center relate Regressions a bias that (the seriously to can data 0.5 leads endogeneity So from to This vote-share. inferences Democratic liberal). Ð0.5 against less from regressed average, is interval position on spatial the why when and, on see from—restricted scenario distributed To a scale uniformly consider Republicans). Left-Right are slopes, the for positive voters toward versa the biases vice which districts (and House in in shares shares vote vote would against positioning Democratic regressing candidate lower Democratic with be liberal would associated more party i.e., each be negative, for slope than voter—the rather median positive the decidedly to proximities relative candidates’ the namely AEcodnt o vr eiltri osrie olewti h nthprpee with hypersphere, unit the within lie to constrained is legislator every for coordinate NATE sepandi h website the in explained As factors— spatial by determined completely is vote-share which in case extreme an for fact In n yDmcai rsdnilVt nteDistrict the in Vote Presidential Democratic by and http://polisci.ucsd.edu/faculty/poole.htm 2001 n tesdfiecmeiini terms in competition define others and ) 17 ehv lte member plotted have we h vrg DW-NOMI- average the , 337 + 16 1 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 oemmesmyhv ag iertrss htfrsm ogesstercodntscnbe can coordinates their Congresses some for that so terms linear than greater large have may members some 18 and score conservative most the as interpretable was: specification our party each for Thus election, Presidential contemporaneous the in district the vote in Democratic share normalized (district-specific) the against scores DW-NOMINATE 338 19 where ln,SuhCrln,Txs n Virginia. and Texas, Carolina, South olina, is percent district normalized the then nationally, candi- percent presidential 60 a and if district, example, a For 65 in share. percent vote 65 presidential gets national date the minus share vote dential sa siaeo ititielg.W ae hsvral the variable this label We president, for ideology. proportion district vote Democratic of estimate an as eodterneof range the beyond emyitrrttezr on ftenraie eortcvt rprinfrpeietas president district. 50Ð50 for a proportion from deviation vote percent, zero Democratic 50 as normalized or exactly vote almost the presidential is of national %) mean point the (49.9 zero either study representing the national the the interpret of of may period the mean the we is the over district Because district vote the estimates. the presidential Centering parameter in Democratic percent. informative zero generate share footnote is to in vote vote below) explained presidential as district (described necessary, the normalized is the If zero then on district. vote, vote that national in the as average same national his of ahead rkndw ytm eidaesoni Fig. in shown are period time by down broken ititielg;w loicue um aibefrdsrcsfo h South. the from districts for variable dummy a included also we ideology; district oto,wihw aea esr fdsrc ideology, district of measure pro- a vote as Democratic take district we represen- normalized which the regressed portion, the we on which party, in each scores DW-NOMINATE party, for that one tatives’ curve, for Each district regression districts. the quadratic the Democratic while a represent vote, which represents it Democratic of in national right those the the to than i.e., areas less districts, was Republican vote presidential represent Democratic line vertical the of edfietesuha rass lbm,Foia eri,Lusaa issip,NrhCar- North Mississippi, Louisiana, Georgia, Florida, Alabama, Arkansas, as south the define We presi- district to equal is president for proportion vote Democratic normalized the Specifically, South WNMNT score DW-NOMINATE [ ititideology District − ititideology District lt o olddt vrtepro 9620 r rsne nFig. in presented are 1956Ð2004 period the over data pooled for Plots 60 =+ WNMNT score DW-NOMINATE = + ftedsrc a oae nteSuh n eootherwise. zero and South, the in located was district the if 1 ecn,rflcigtefc httepeieta addt a v ecnaepoints percentage five ran candidate presidential the that fact the reflecting percent, 5 1 / 1 nordt,teeae1 aapit o hc h WNMNT crsare DW-NOMINATE scores the which for points data 12 are there data, our In Ð1. − j 1or1. j = ] 2 omlzdpeieta oein vote presidential normalized footnote = h qaeo h omlzdpeieta oein vote presidential normalized the of square the j j sdistrict, ’s = representative rcdn h election, the preceding j slgsaievtn eodi h w years two the in record voting legislative ’s j 18, = + b 1 b + 3 − [ b ititideology District nepee stems iea cr.However, score. liberal most the as interpreted 1 or 2 20 [ j ititideology District ititideology district eo,i re o h udai regressions quadratic the for order in below, sD-OIAEsoe ae on based score, DW-NOMINATE ’s 2 ra ftefiuet h left the to figure the of Areas . j sdsrc,a endin defined as district, ’s and j ] 2 j + o short. for ] ntesur fthe of square the on omlzddistrict normalized b 4 18 [South whichweuse .Aase al. et Adams J. ] , 1 plots ; (1) 19 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 rudtelnsrpeet9 ecn ofiec intervals The confidence level. percent district 95 and represent national lines the the at around shares vote presidential Democratic identical nteohrhn,pae h eovleo h needn aibea ititwoepresiden- whose district competitive around a behavior at on variable attention independent focusing the vote, presidential electorates. of national value president, the zero for matches the vote proportion places tial vote hand, Democratic normalized other be- the the reflect Using would on variable estimates presidential region. independent parameterization, the unrealistic this this of an Under over percent interest. of of havior zero point range received actual the zero candidate the outside the Democratic is use which the then we vote, where ideology, If district district variable. a of independent to measure the corresponds our of as point vote zero district the around behavior reflect estimates hddrgosi h grsrpeette9 ecn ofiec ein o h re- the for regions confidence percent 95 the represent gressions. figures the in regions shaded 20 regression The the Democrats. for for 6,161 sizes and sample Republicans the for 1956Ð2004; 4,613 period are the models presidential over Democratic scores national DW-NOMINATE the bers’ minus share footnote vote (see which presidential share district, vote member’s Democratic House district the to in president equal for ver- is proportion DW-NOMINATEscores vote for Democratic curves normalized regression the quadratic sus presents plot The Notes: 1956Ð2004. 1 Fig. Politicians? Centrist Produce Necessarily Districts Competitive Do 21 nFigs. in r ae ntefl aastwtottebekonb region. by breakdown the without set data full the on based are ( in parameters informative estimate to order in ewe eulcnadDmcai ersnaie aiswt ititcompeti- district member’s with House dispersion varies ideological the representatives tiveness. of Democratic on degree and ideology the Republican how district between estimate of to effects and nonlinear score, DW-NOMINATE of possibility the at lpsdwwr pryrsosvns) ohfrte15Ð04pro sa as each period for 1956Ð2004 curve the fitted for the both that responsiveness), so district, (party by downward voter slopes median party the of position the to spond o ipiiy h at-pcfi ersincre n hi ofiec nevl ntefigures the in intervals confidence their and curves regression party-specific the simplicity, For zero on centered is that ideology district of measure a employ to necessary is it why see To nlso fteterm the of Inclusion os udai eaino WNMNT crsadpria itiuinb district: by distribution partisan and scores DW-NOMINATE of relation quadratic House 1Ð2 20 21 Table sepce,teprmtretmtsrpre nTable in reported estimates parameter the expected, As upr h xetto htrpeettvs dooia oiin re- positions ideological representatives’ that expectation the support 1 eot hs ersincefiinsfrteUS os,adthe and House, U.S. the for coefficients regression these reports .Teergeso ie eepotduigtefl e fHuemem- House of set full the using plotted were lines regression These 12). [ ititideology District 1 ,nt hti udai ersin parameter regression, quadratic a in that note ), j ] 2 n( in losu oinvestigate to us allows 1) etclline vertical 1 t00represents 0.0 at n illustrated and hddregions shaded B&W IN PRINT 339 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 differences nraeo e ecn nteDmcai rprino h ititvote. district the each of for proportion unit Democratic a the of in tenth percent one ten about of of increase score DW-NOMINATE the Table in the in trend period, downward reported time coefficient 1956Ð2004 and patterns linear entire representatives voting Democratic the the over presidential on pooled the parameters analyses of the estimated function For a district. the as in substantially mean lines vary suggesting regression substantial, does these are of ideology Republicans slopes and downward Democrats the both that at note for significant addition, statistically In party—are each level. for 0.001 and the subperiod) each for as well (as riod (Fig. 1996Ð2004 and 1986Ð1994, (Fig. whole the minus share vote footnote presidential (see the Democratic share for district in vote curves to presidential president regression equal Democratic for national is quadratic proportion which present vote district, plots Democratic member’s These House normalized Notes: the periods. versus time scores DW-NOMINATE by separated data with 2 Fig. 340 22 The figure. the in noted eras the by separated just dooyi erzr) nwihcs h au ftesurddsrc dooyvral n( in variable ideology district squared the district of of value measure normalized the the case that which (so vote in national zero), the to near similar is is ideology district the in share vote rsdnilvt hrsa h ainladdsrc ee.The level. district and intervals national confidence percent the 95 represent at shares vote presidential sngiil.I hsrneo austepeitdefc fdsrc dooyo representatives’ on ideology district of effect predicted the values linear. of approximately is range scores DW-NOMINATE this In negligible. is hs siae pl omria hne ndsrc rsdnilvt hnteDemocratic the when vote presidential district in changes marginal to apply estimates These nteohrhn,i elo ttegpbtentetocre,wihreflects which curves, two the between gap the at look we if hand, other the On udai ersinfrtepeieta oesaeadielg o ..Huemembers House U.S. for ideology and share vote presidential the for regression Quadratic cosprylines party across n o aho h iepros15Ð94 9617,1976Ð1984, 1966Ð1974, 1956Ð1964, periods time the of each for and 1) efidvr usata ifrne ewe h win- the between differences substantial very find we , 3frRpbia ersnaie,idctn a indicating representatives, Republican for −1.03 .Alo hs onadsoe—o h ulpe- full the slopes—for downward these of All 2). etclline vertical 12).ThedataarethesameasinFig.1 t00rpeet dnia Democratic identical represents 0.0 at hddregions shaded 1 are rudtelines the around .Aase al. et Adams J. within-party 22 5for −0.75

B&W IN PRINT 1) , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 oCmeiieDsrcsNcsaiyPoueCnrs Politicians? Centrist Produce Necessarily Districts Competitive Do

Table 1 Regression of DW-NOMINATE scores versus partisan distribution of the district: House of representatives Period Regression coefficients Democrats Republicans Intercept South Ideology [Ideology]2 Intercept South Ideology [Ideology]2

1956Ð2004 Ð0.32*** 0.19*** Ð0.75*** 0.13* 0.27*** 0.06*** Ð1.03*** Ð0.54* (full period) (0.01) (0.01) (0.02) (0.06) (0.01) (0.01) (0.04) (0.24) 1956Ð1964 Ð0.33*** 0.30*** Ð0.36*** 0.33* 0.24*** 0.06* Ð0.91*** Ð1.11** (0.01) (0.01) (0.03) (0.14) (0.01) (0.03) (0.09) (0.44) 1966Ð1974 Ð0.32*** 0.24*** Ð0.64*** Ð0.22 0.23*** 0.01 Ð0.76*** 0.12 (0.01) (0.01) (0.04) (0.13) (0.01) (0.02) (0.07) (0.46) 1976Ð1984 Ð0.32*** 0.21*** Ð0.86*** 0.34 0.20*** 0.10*** Ð1.02*** 1.21 (0.01) (0.01) (0.06) (0.19) (0.01) (0.01) (0.12) (0.68) 1986Ð1994 Ð0.30*** 0.07*** Ð1.17*** 0.83*** 0.29*** 0.03** Ð1.03*** Ð0.56 (0.01) (0.01) (0.05) (0.17) (0.01) (0.01) (0.10) (0.61) 1996Ð2004 Ð0.33*** 0.05*** Ð1.06*** 1.16*** 0.39*** 0.02 Ð1.03*** Ð1.30** (0.01) (0.01) (0.05) (0.17) (0.01) (0.01) (0.10) (0.45) *Significance at the 0.05 level; **Significance at the 0.01 level; ***Significance at the 0.001 level. Significance levels are 2-sided The definitions of the independent variables are given in the text. Curvature in the expected direction (i.e., convex for Democrats and concave for Republicans) is significant for both parties for the full period and for the earliest and latest subperiods. Thus, the data offer evidence that the curves either bow outward or may be straight, but the coefficients are never significant in the other direction, i.e., the data offer no significant evidence that any of the partisan curves bow inward. N’s for the analyses are as follows: Full period (1956Ð2004): Democrats 6158, Republicans 4616; 1956Ð1964 period: Democrats 1304, Republicans 843; 1966Ð 1974 period: Democrats 1254, Republicans 873; 1976Ð1984 period: Democrats 1323, Republicans 843; 1986Ð1994 period: Democrats 1243, Republicans 932; 1996Ð2004 period: Democrats 1034, Republicans 1125 341 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 o Republicans. for 15Ð04 o o n ftefiebekonprosi hr vdneta the that evidence there bowed is significantly periods is breakdown party five either the for of curve any for nor (1956Ð2004) 23 ideology. district House. of that the to relative in effect polarization huge most a increased the has party in the Clearly, units reflecting 0.72 1996Ð2004, to the subperiods subperiod three of recent first composition the Table partisan of in each the in reported when DW-NOMINATE units As parties 50Ð50. major two is the district scores DW-NOMINATE of the members between House difference of percent typical and 30 the Democrats a between indicates from intercepts Republicans member regression in Democratic difference a The than district. conservative Democratic can more district be Democratic percent to 70 expected Fig. a be even in from data Congressperson pooled Republican the a instance erage, for parties; two the from ners 342 25 24 httepria a nteSuhrne rm03Ð.3i h rttosbeid o06 nthe in 0.69 to subperiods two first the in 0.32Ð0.33 from so ranges combined, subperiod. South be recent must the South most in variable gap the partisan for the estimate parameter that and intercept estimated the South, ape htageaet vr1000vtr,fid ytmtcdfeecsi eulcnand Republican for accounted in entirely be differences cannot systematic preferences. that constituency (1999Ð2000) finds same-party House by 106th voters, the 100,000 in behavior over voting Democratic ( to Clinton Similarly, aggregate Congress. in that parties the samples of polarization increasing the observes and eortcrpeettvsi itit htfauelpie rsdnilvt mar- and vote Republican presidential between lopsided gap feature ideological gins. that districts largest in the im- representatives observe wisdom Democratic conventional should this voting, we likely presidential that are of ex- partisanship plies terms and unequal in median highly competitive with less district be districts or to that national Given the anyway. the win from either to candidate from the pect away when district move greatest lopsided can a be party in will dominant because divergence competitive, noted not partisan is As that election parties. is the different wisdom conventional from the members above, House to between related divergence is ideological ideology between findings district differences interesting how most ideological concern our However the representatives. with Republican and along Democratic ideology, district to sponsiveness voting presidential similar with districts to patterns. non-Southern compared from districts, elected Southern were from they elected leg- when were conservative they more when compile records voting to islative tended periods—representatives in time that—particularly earlier suggest variable the dummy South the on estimates coefficient itive o ofr oti atr:isedte o u lgtyaa rmec te in other each from scale. away distribution slightly partisan out the bow of they middle instead the pattern: this to conform not iue4i ulr( Butler in 4 Figure ( Ono that note We category the For non-South. category, reference the to apply above reported gaps partisan The oee h uvsi Fig. in curves the However re- members’ House about implies data our what considered have we far So 2005 2009 ban iia lt o w ogess(9917 n 2003Ð2004) and (1969Ð1970 Congresses two for plots similar obtains ) per osgetti aecneiyfrDmcasadconcavity and Democrats for convexity same this suggest to appears ) 1 hc r te otefl 9620 aa do data, 1956Ð2004 full the to fitted are which , hs ifrne ag rm0.52Ð0.57 from range differences these 1, .. h ereof degree the i.e., divergence, partisan 25 inward oeta ete o h ulperiod full the for neither that Note tte00 ee.B contrast, By level. 0.05 the at 1 ugssta,o av- on that, suggests 24 ial,tepos- the Finally, .Aase al. et Adams J. 2006 ,using ), 23 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 esr fielg a h ifrnebtentesaesDmcai rsdnilvt n the zero. and on vote centered presidential is that Democratic measure state’s a the vote, between presidential difference Democratic the national was ideology of measure corresponding the than the smaller between districts. is gap extreme districts be partisan more moderate the to for in that tend winners gap imply Democratic districts not and competitive does Republican from districts of lopsided representatives sets from that those fact ideologically than The moderate the gener- ideology. more of are district middle size of these the function but the Democrats,” in a assess moderate “Districts relatively not that or do results note Republicans regression authors authors moderate linear relatively The by patterns; 1976Ð1996. represented similar period ally obtaining the vote, for presidential reported against are 1980s the for dices o nad sw ol xeti h atsngpnroe ncmeiiedis- competitive not in do narrowed curves gap the partisan that the is if remarkable expect bow- is would tricts. outward what we The pronounced; as not ones. inward, un-competitive is bow curves in ideology the in than so of districts more ing competitive or different most as the are in winners House Democratic and Republican h ate,btti osbeefc sntnoted. not between is out effect bows that possible curvature this show but to parties, appear the party each Republicans for and Wright plots scatter and Democrats authors’ (Erikson between responsiveness separation party sharp reflecting ide- trends a constituency and the do against of we 1980s perception as the mean obtaining during the ology, members plot House authors incumbent these of axes, ideology both for scale seven-point NES id,ec tte00 ee rbte seTable (see better pe- (1996Ð2004) or latest level the 0.05 and the (1956Ð1964) at earliest each the riods, for and period overall the 27 26 as election it), particular follows that that for election (both midterm state the each as measure for well a support as the election partisan quadrennial for underlying each above, the for reported president of for Representatives, vote of the use House We Senate. the U.S. on analyses the replicate We Party by 1956–2004, Senate, U.S. the in Extremism Ideological 3 coeffi- negative and significant Democrats indicate the Republicans for the term for quadratic cients the on coefficients positive Politicians? Centrist Produce Necessarily Districts Competitive Do 28 l eaosa esr fsntra dooyfo ahcnrs.Teposfor plots The Fig. congress. and Fig. each period) in from 2004 depicted ideology are senatorial of regressions measure the a as senators all otoln o itit nteSuha ehv oesol eueti fetad nayeet it subperiod. event, recent any most in the and, in effect Democrats this But the reduce scale. for should the curvature done of convex races, have side pronounced uncontested we right the as the explain won South on cannot Democrats the up in turn Southern districts to conservative for Democrats of for controlling curves number regression a quadratic the that causing be may study under swt u nlsso os itit sefootnote (see districts House of analyses our with As ( Wright and Erikson period the of part earlier the in scores Democratic the of curvature convex for explanation One eae lt r bandb rko n rgt(2000 Wright and Erikson by obtained are plots Related n yPeieta oei h State the in Vote Presidential by and 2000 4 wihdpcsrslsfrtesm uprosue o the for used subperiods same the for results depicts (which i.81 lopo olcl dooybsdo h D/C in- ADA/ACA the on based ideology roll-call plot also 8.1) Fig. , 3 wihpeet eut o h nie1956Ð entire the for results presents (which 28 gap outward 27 n h WNMNT crsfor scores DW-NOMINATE the and ewe eulcn n eort as Democrats and Republicans between 18 1 ,frteSnt-ae nlssour analyses Senate-based the for ), n Fig. and oigfrbt ate for parties both for bowing .I atclr sn the using particular, In ). 2). 2000 26 nohrwords, other In ,Fig.8.6).The 343 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 dnilvt hrsa h ainladsaelvl The level. state intervals and confidence for national percent 1335 95 the are at models shares regression vote the idential for The Democrats. sizes for sample 1353 the and 1956Ð2004; Republicans share period vote the presidential over Democratic scores equal NATE national is the which minus state, footnote Senator’s share (see the vote in presidential president Democratic for state proportion ver- to vote DW-NOMINATEscores Democratic for normalized curves regression the quadratic sus presents plot The Notes: 1956Ð2004. tts h vdnefrotadbwn ssgicn tte00 ee o both for level 0.05 the at significant is bowing outward mirrors are for vote winners evidence presidential Senate The the Democratic states. where mid- and the states Republican in different in that other as i.e., indicating each vote, scale, from away national ideology states Fig. out the state in bow in the again than curves of data, terms, The 1956Ð2004 dle partisan the republican: in to or divided, fitted as evenly democratic are are are overwhelmingly that records are states voting that in senators’ greater Republican or and great Democratic between differences Table House). 3 Fig. 344 WNMNT nt ntefis upro 9616,t .0uisi h most the in units 0.66 Table 0.80 (see from to 1996Ð2004 increasing increase 1956Ð1964, subperiod the differences subperiod recent these first reflects time: the competitive, over in is Senate units DW-NOMINATE state the the in parties polarization of opposing partisan composition differ- of partisan typical members the Senate the of when indicates in- scores DW-NOMINATE regression which the in Republicans, 1956Ð1964 between difference subperiod ence and the the Furthermore, Democrats for level. between statistically those 0.001 party—are except tercepts the each all slopes—for at for fact, significant linear and in also downward, level; subperiod) are the 0.05 each of the for all at as data, well significant, House (as the period for full case the the was As each data. for intervals confidence percent 95 the represent again regression. figures the in regions ial,oretmtso h eaedt gi upr h rpsto htthe that proposition the support again data Senate the on estimates our Finally, h atrsw siaefrteSnt aaaesmlrt hs o h House the for those to similar are data Senate the for estimate we patterns The eaeqartcrlto fD-OIAEsoe n atsndsrbto ydistrict: by distribution partisan and scores DW-NOMINATE of relation quadratic Senate .Teergeso ie eepotduigtefl e fSntr’DW-NOMI- Senators’ of set full the using plotted were lines regression These 12). adi ntigmr ifrn)i dooyi h otcompetitive most the in ideology in different) more anything if (and 2 eot h ersincefiinsfrteSnt,adteshaded the and Senate, the for coefficients regression the reports etclline vertical 2). t00rpeet dnia eortcpres- Democratic identical represents 0.0 at hddregions shaded rudtelnsrepresent lines the around .Aase al. et Adams J. B&W IN PRINT 3 which , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

690 689 688 687 686 685 684 683 682 681 680 679 678 677 676 675 674 673 672 671 670 669 668 667 666 665 664 663 662 661 EDITOR’S660 659 658 657 656 655 654 653 652 651 PROOF650 649 648 647 646 645 ore elkon n aebe dnie sn lentv esrso legisla- of measures alternative of using are, identified ideology. views been concerning tive have constituency findings and to known, Our responsiveness well party periods. course, and and time gap House across ideological the partisan generalize across the largely consistent they are and conclusions Senate, substantive bal- party evenly our across convergence more Moreover, candidate that prompts lines. district claim a simplistic in any competition electoral on anced doubt considerable cast findings Our Discussion 4 significantly bows periods breakdown the of any peri- party for same either or for periods—the curve period latest no inward. full and whereas the House; earliest the either the in for bowing for outward and exhibited period that ods full the for parties intervals confidence Democratic national the minus share Senator’s vote the presidential in footnote Democratic (see president state share for vote to proportion presidential DW-NOM- equal vote for is Democratic curves which regression normalized state, quadratic the present versus plots These scores INATE Notes: periods. time by separated 4 Fig. Politicians? Centrist Produce Necessarily Districts Competitive Do 29 ednyfrRpbiast emr oeaei opttv itit.Frhroe h patterns the Furthermore, districts. competitive in moderate more be to Republicans for tendency eiltv oigrcrsrsodt ititielg,adta eortcrepre- Democratic that and ideology, district to respond records voting legislative hrsa h ainladsaelvl The level. state and national the at The figure. shares the in noted eras etito ftedt ooe-etrcscagsteptenol eymrial,wt slight a with marginally, very only pattern the changes races open-seat to data the of Restriction udai ersinfrtepeieta oesaeadielg o eaoswt data with Senators for ideology and share vote presidential the for regression Quadratic 29 npriua,w n h xetdeiec hteetdofficials’ elected that evidence expected the find we particular, In etclline vertical 12 ).ThedataarethesameasinFig. t00rpeet dnia eortcpeieta vote presidential Democratic identical represents 0.0 at hddregions shaded rudtelnsrpeet9 percent 95 represent lines the around utsprtdb the by separated just 3,

B&W IN PRINT 345 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

736 735 734 733 732 731 730 729 728 727 726 725 724 723 722 721 720 719 718 717 716 715 714 713 712 711 710 709 708 707 EDITOR’S706 705 704 703 702 701 700 699 698 697 PROOF696 695 694 693 692 691 346

Table 2 Regression of DW-NOMINATE scores versus partisan distribution of the State: results for the U.S. Senate Period Regression coefficients Democrats Republicans Intercept South Ideology [Ideology]2 Intercept South Ideology [Ideology]2

1956Ð2004 Ð0.39*** 0.26*** Ð0.66*** 4.29*** 0.26*** 0.16*** Ð1.57*** Ð2.30** (full period) (0.01) (0.01) (0.07) (0.54) (0.01) (0.02) (0.09) (0.81) 1956Ð1964 Ð0.39*** 0.39*** Ð0.49* 3.08* 0.27*** 0.26** Ð0.72** Ð5.02* (0.03) (0.03) (0.22) (1.32) (0.02) (0.10) (0.26) (2.50) 1966Ð1974 Ð0.38*** 0.23*** Ð0.62*** 2.18* 0.19*** 0.19*** Ð2.05*** 1.22 (0.01) (0.03) (0.15) (1.08) (0.02) (0.05) (0.22) (2.04) 1976Ð1984 Ð0.37*** 0.23*** Ð0.67*** 5.50*** 0.17*** 0.28*** Ð1.59*** 2.86 (0.01) (0.02) (0.11) (1.14) (0.02) (0.03) (0.21) (1.80) 1986Ð1994 Ð0.36*** 0.13*** Ð1.05*** 0.47 0.29*** 0.09** Ð1.94*** Ð5.56* (0.01) (0.02) (0.12) (1.77) (0.02) (0.03) (0.21) (2.61) 1996Ð2004 Ð0.42*** 0.16*** Ð0.95*** 3.45** 0.38*** 0.04 Ð1.63*** Ð7.48*** (0.01) (0.02) (0.12) (1.23) (0.01) (0.02) (0.15) (1.16) *Significance at the 0.05 level; **Significance at the 0.01 level; ***Significance at the 0.001 level. Significance levels are 2-sided The definitions of the independent variables are given in the text. Curvature in the expected direction is significant for both parties for the full period and for the earliest and latest subperiods. Thus, as for the House, the data offer evidence that the curves either bow outward or may be straight, but the coefficients are never

significant in the other direction, i.e., there is no evidence that any of the partisan curves bow inward. N’s for the analyses are as follows: Full period (1956Ð2004): al. et Adams J. Democrats 1353, Republicans 1135; 1956Ð1964 period: Democrats 314, Republicans 187; 1966Ð1974 period: Democrats 270, Republicans 205; 1976Ð1984 period: Democrats 263, Republicans 242; 1986Ð1994 period: Democrats 272, Republicans 234; 1996Ð2004 period: Democrats 234, Republicans 267 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

782 781 780 779 778 777 776 775 774 773 772 771 770 769 768 767 766 765 764 763 762 761 760 759 758 757 756 755 754 753 EDITOR’S752 751 750 749 748 747 746 745 744 743 PROOF742 741 740 739 738 737 oe,rl alvtn,adcniaepstoig(.. ode al. et Bond (e.g., positioning out- candidate election and influence voting, that call variables numerous roll of comes, impact the mediates diversity cal their in opinion public of the center surpass the to that supporters, appeal benefits constituency. who party electoral candidates to derive to accrue may appeal that groups benefits that interest policies special noncentrist and activists, present developed who arguments candidates ( theoretical why additional Sened above and noted Schofield have by we as And, didates. policy maximal motivate will tne alyadBay( Brady and Bailey stance, 1998 argu- These alienation. to due abstention that, and conclude loyalties ments partisan voters’ of ( account Butler of arguments modeling-based Figs. polarization. in overall increase, seen not be reduce, might elected can districts been as have competitive would produce districts, that those lopsided than more extremist less in each be from cases different most sharply in while would districts, other, competitive in Republi- Congress and necessarily Democratic of contrary, would members the do competitive can On more results Congress. be in our to polarization districts overall (or that of increase however, Democratic redesigning emphasize, lopsidedly the We that are imply level. that not each presidential districts to the from relative elected at polarized are Republican) ideologically they more terms when even in than be districts, other may competitive vote, presidential from Demo- the elected that of representatives suggest results Republican our Indeed, and whether will Congress. cratic House to on the moderation doubt of of members casts politics for the it districts bring competitive that more is draw to result redistricting polarization using competitive the of implication i.e., intense, most is districts. competition competitive this highly when in positions their moderate to politicians the on fifty label past we the which over finding, result that latter as pattern, This be opposite competitive. to most tended the has suggests polarization data partisan the years districts— anything, competitive if most the fact, in in smallest is representatives Republican ideology. and district cratic for controlling when Republicans But than liberal more are sentatives Politicians? Centrist Produce Necessarily Districts Competitive Do aantteDmcai oesaei h os lcinb itit.Teeposso internal show plots These district). Congress. in by voting election ideological of House liberal measures the the different in many from among share (LCV) ranging consistency Voters vote Conservative associations of Democratic League monitoring conservative the the fifteen (against to (ADA) of Action each Democratic ( for by al. American et and Lee scores used. NOMINATE have we by ideological that of (DW-NOMINATE scores) House measure the particular in the of voting result the likely not are observed ial,oraaye r eeatt h ieycretdbt vrhwpoliti- how over debate current lively the to relevant are analyses our Finally, hoeial,orcmeiieplrzto eutsurswt h eetspatial recent the with squares result polarization competitive our Theoretically, practical The implications. practical and empirical, theoretical, have findings Our efidn vdneta h ereo dooia oaiainbtenDemo- between polarization ideological of degree the that evidence no find we Jones ; scnrr oteitiinta oiia optto xrsmxmlpressures maximal exerts competition political that intuition the to contrary is , 2003 .Wt epc osntr’rl alvtso retae o in- for trade, free on votes call roll senators’ to respect With ). eei paribus ceteris 2006 n hti eorpial ooeeu states homogeneous demographically in that find 1998) divergence rsn reasons present (1994) Baron and (2004), Moon ), itit ihblne atsncompositions partisan balanced with districts , n dm ta.(2010 al. et Adams and 2009) ewe eortcadRpbia can- Republican and Democratic between 2004 ltlgsaievtn eod sassessed as records voting legislative plot ) ra rgreater or great 1 and opttv polarization competitive 3 hs eititn to redistricting Thus, . ndsrcsta are that districts in 2001 ,wihtake which ), Koetzle ; 347 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

828 827 826 825 824 823 822 821 820 819 818 817 816 815 814 813 812 811 810 809 808 807 806 805 804 803 802 801 800 799 EDITOR’S798 797 796 795 794 793 792 791 790 789 PROOF788 787 786 785 784 783 odJ,Cmbl ,Ctrl B(01 h uzeo osiunydvriyrvstd conditional revisited: diversity constituency of Public puzzle The elections. (2001) Senate JB Cottrill in K, diversity Campbell and JR, Bond representation Rev Issue Sci (2006) J Polit Adams Am JK, voters. Dow uninformed B, and Bishin informed with competition Polit Electoral J Am (1994) trade. D free and Baron Senate the representation: Am and elections. Heterogeneity House (1998) U.S. D Brady in cross-national M, positioning Bailey a Candidate competition: (2001) party C of Stewart J, theory Snyder unified S, A Ansolabehere (2005) B a Grofman with S, beginning Merrill elections two-stage J, in Adams Polit strategies party J and elections. Candidate American (2008) in S strategies Merrill J, candidate Adams and turnout Voter (2003) S Merrill J, Adams to expected be should divergence candidate Why (2010) S Merrill B, Grofman T, Brunell J, Adams References a in held was election off the faced when than candidates state, these heterogeneous when a state. similar in homogeneous can- more held senate election no of an were positions in ideological parties the rival that Bishin from analyses positioning, large didates empirical candidate as report to be respect (2006) to with al. tends And et polarization districts. partisan competitive in that larger finding or empirical states—a non-competitive our in fits than that Re- states, be pattern and competitive Democrats’ to in between divergence records tend greater roll-call findings or states publicans’ Brady equal heterogeneous observe and will Bailey that we the that level, extent imply presidential the the are at To preferences competitive influences. electorally constituent more several states heterogeneous of influ- in one significant while but statistically votes, exerts roll-call that on factor ences only the are preferences constituent 348 30 h w ate cusi itit ht npeieta upr em,aetemost the are terms, support of presidential winners in the that, between districts interest, difference in greatest ideological competitive. occurs greater of parties or believe two great we as the and that However, find party. also records their voting we and Con- roll-call representatives’ district district. that district their the find of in reflect we function support research, a presidential previous as Democratic with varies Senate sistent of the terms and in House defined the ideology, in parties the between tion a esrdi em fdmgahccaatrsis(sn h ulvnidx ri em of terms in or index) Sullivan the (using characteristics demographic diversity. of ideological terms in measured was ihne l n htti ocuinhlsrgrls fwehrtesaeeetrt’ diversity electorate’s state the whether of regardless holds conclusion this that find al. et Bishin fet fdsrc iest ncnrsinleetos aucit ea & University A&M Texas Manuscript, elections. congressional in diversity district of effects 129(2):201Ð215 Choice 88(1):33Ð47 42:524Ð544 Sci 45:136Ð159 Sci Polit J Cambridge Press, University Cambridge factors. behavioral and spatial integrating analysis 52(2):344Ð359 Sci Polit J Am primary. 65(1):161Ð189 Choice Public ones. non-competitive in as seats competitive in greater) even 145(3Ð4):417Ð433 (or great as just be nsm nti ae ehv nlzdhwtedge fielgclpolariza- ideological of degree the how analyzed have we paper this in sum, In 30 hsfidn saancnitn ihorresults. our with consistent again is finding This .Aase al. et Adams J. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

874 873 872 871 870 869 868 867 866 865 864 863 862 861 860 859 858 857 856 855 854 853 852 851 850 849 848 847 846 845 EDITOR’S844 843 842 841 840 839 838 837 836 835 PROOF834 833 832 831 830 829 cmd ,KnyL otnR(96 vdneo lcoa conaiiyi h ..Snt:are Senate: U.S. the in accountability electoral 46:1061Ð1079 of Polit Evidence (1996) J R politics. Polit Morton American J L, of Kenny Am polarization A, The study. Schmidt (1984) case H 1980 Rosenthal a KT, with Poole elections, presidential in ownership di- Issue persistent (1996) contests: JR two-party Petrocik in competition electoral Two-stage Exten- (2006) myth. B the Grofman debunking G, Congress: Owen in polarization partisan of tests, origins theory, Electoral taking: (2005) position K candidate Ono and resources, campaign Sci activists, Polit Party Am (2004) States. W U.S. United the Moon the in from realignment Evidence partisan and policies? Activists elect (2003) or N Schofield affect House G, voters Miller U.S. Do (2004) of MJ competitiveness Butler the E, on Moretti diversity DS, Lee constituency of impact nomina- The presidential (1998) in models W spatial Koetzle of importance the issue on of Evidence examination (2005) An B campaigns? Lotfinia Senate L, Kenny U.S. in Dialogue 65(3):851Ð863 (2006) Polit T J Senate. Ridout U.S. D, the Park in advantage. N, avoidance valence Kaplan position a and Position-taking has (2003) candidate DR one Jones when location candidate of of model disadvantages A electoral (2001) T potential Groseclose The (1999) W Koetzle TL, Brunell S, Merrill B, Grofman experiment natural a party: different geography, Identical (1990) A Glazer R, Griffin B, Grofman senators U.S. two (ed) the N between differences Polsby ideological 7:25Ð46. the Sci Explaining Pol (2001) Rev T Annu Brunell convergence. B, two-party Grofman and Downs (2004) marginal- B Organ the Grofman of Econ defense Law a responsiveness: J democratic and representation. competition and Electoral (2006) systems JD Griffin election Primary (1998) RB Chicago McNally, Morton Rand E, constituencies. their Gerber and Congressmen D, (1963) Brady LA In: Froman Congress. in ideology Am constituency vote. of the Representation of (2000) model G behavioral Wright the R, and Erikson equilibrium Candidate York New (1990) Row, D and Harper Romero democracy. R, of theory Polit Erikson economic J An House. Wilkie. (1957) 106th the A Simon in Downs calls with roll and campaigns, constituents political Congress: in and Representation (2006) lies, the J and damned Clinton voting Lies, roll-call (2007) incumbents’ on S blocs Wilkie voting S, of Callander size 34:211Ð227 the Sci of Polit J effect Br The elections. (2009) congressional US DM in Butler positioning Congres- Candidate (ed) (2004) PF B Galderisi Burden In: elections. congressional of effects polarizing The (2001) B Burden Politicians? Centrist Produce Necessarily Districts Competitive Do natflaet elypnse?Eo nur 34:545Ð567 Enquiry Econ punished? really agents unfaithful 40(3):825Ð850 Sci 26:547Ð569 Welf Choice Soc positions. party of vergence (Fall):15Ð20 sions 34(4):611Ð633 Sci Polit J Br applications. and 97:245Ð260 Rev 119(3):807Ð859 Econ J Q House. 23:561Ð573 Q Stud Legis 1962Ð1996. elections, 123(3):439Ð462 Choice Public elections. and tions 50(3):724Ð736 Sci Polit J Am advertising. television candidate in discussion 45(4):862Ð886 Sci Polit J Am states. U.S. 50 the in democrats and 5(2):199Ð210 republicans Polit among Party Shel- variance RJ, ideological party: Johnston Catch-All 207Ð217 In: a pp 1960Ð1984. London, Routledge, Senate, geography. U.S. electoral in the Developments (eds) in PJ differences Taylor FM, party ley of magnitude 132Ð142 the pp York, on Congressional New (ed) Littlefield, PF and Rowman Galderisi representation. In: of model. politics effects the institutional and an primaries state: same the from elected 68(4):911Ð921 Polit J hypothesis. ity 11(2):304Ð324 Press, University Stanford elections. House in change 149Ð177 and pp Continuity Stanford, (eds) M Fiorina J, Cogan 84:1103Ð1126 Rev Sci Polit 68(2):397Ð409 60(2):262Ð286 Behav Econ Games 34(3):297Ð318 Q Stud Legis Congress. of polarization asymmetric York New Littlefield, and Rowman representation. of politics the and primaries sional 349 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

920 919 918 917 916 915 914 913 912 911 910 909 908 907 906 905 904 903 902 901 900 899 898 897 896 895 894 893 892 891 EDITOR’S890 889 888 887 886 885 884 883 882 881 PROOF880 879 878 877 876 875 ita 18)Cniaemtvto:asnhsso lentvs mPltSiRv77:142Ð157 Rev Sci Polit Am alternatives. of synthesis a motivation: Candidate (1983) competition D Wittman electoral of degree the in variation Explaining differentia- (2008) platform B of Grofman logic L, Kenny the S, platforms: Winer party and seats, marginal seats, Safe (1994) presidential U.S. J in Snyder convergence Issue engagement? or Avoidance (2004) Jr Cambridge EH politics. Buell legislative L, and Sigelman elections democracy: Sci Polit Multiparty J (2006) Am I States. Sened United N, the Schofield in coalitions activist and Elections (2007) G Miller N, Schofield 350 namtr eorc:US eaeeetos 9220.Peetda h nulmeigof meeting annual the March at Antonio, Presented San 1922Ð2004. Society, elections, Choice Senate Public the U.S. democracy: mature a in 6:201Ð214 Polit Econ tion. 48(4):650Ð661 Sci Polit J Am 1960Ð2000. campaigns, Cambridge Press, University 51(3):518Ð531 .Aase al. et Adams J. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 DOI (eds.), al. et Schofield N. Street, Seventh e-mail: E 1100 USA 210, 47405-7110, Hall IN Woodburn University, Bloomington, Indiana Science, Political of Department Hellwig T. e-mail: Chang K. e-mail: Park, College Hall, USA Tydings 20742, 3144F MD Maryland, of University Politics, and Government .Clo( Science Calvo Social E. the and Institute (WZB). March Berlin Juan Bernhard Center the Murillo, Research at Vicky conferences Moore, at Bernie Richard participants Birnir, Merrill, and Johanna Sam We§els, Adams, Lupu, Jim Noam thank Kselman, authors Dan the Grofman, suggestions, and comments helpful For from inherited model spatial standard the on ( built Black years have 50 over science For political choice? vote in citizen’s scholars the affect positions policy candidate do How Introduction 1 Chang Kiyoung and Hellwig, Timothy Calvo, Ernesto States United the to Application with Vote: Model the A of Model Spatial Heteroscedastic A n oiiso h addts h oe h tlt h oe eie rmselecting from derives voter the voter utility polls. the the the lower of at the preferences them com- candidates, the by the between implemented of difference policies be the and to greater policies The expected candidates. the peting and policies preferred own their oass addt rfrne;te a aeipretifrainaotlikely about information imperfect have shortcuts may informational they use preferences; may candidate They assess candidates. voters to the Secondly, of candidates. the preferences of policy said each and compare vealed can of misguided and those are want to they preferences preferences what such know policy that voters But case outcomes. the suboptimal to be lead may It polices. preferred their h uligbok falsailmdl fvtn r iia:fisl,voters firstly, similar: are voting of models spatial all of blocks building The 10.1007/978-3-642-35239-3_17 [email protected] [email protected] [email protected] 1958 B ) · n on ( Downs and ) .Chang K. dacsi oiia Economy Political in Advances 1957 pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , ,weevtr sesterltv itnebetween distance relative the assess voters where ), , know h re- the know 351 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 n ate?I hscatrw ilrlxti udmna supino standard to of voters assumption allow fundamental and this voting relax will of we models differ- to chapter words, spatial distance other this relative In In their other? assessing parties? the when ent to metrics vot- different compared have space one policy voters in the if change in what left dramatic the more to a say, move, perceive parties both ers two for when if change What space? this policy measure to metric the same to say, the move, all use location for policy voters parties. identical same distance, are the given space in a parties policy two left the if in is, distances That of involved. with actors metric drawn the results. are that preferences choice single-peaked assumption social research, the most theoretical in single- sensible explicit made and produce not transitive to While be avail- models to the our assumed using allowing are candidates peaked, preferences the thirdly, to And dif- distance information. rational that perceived able make policies their nonetheless the comparing voters of by But views decisions implement. biased eventually very will have candidates even ferent may they and choices; policy 352 iswihte ilntvt o Gabr n Brent par- and and themselves (Granberg between for distance vote the not as 1977 parties will well to they as proximity which for reported vote ties the to al. overstate intend et either to (Merrill they bias” tend parties which systematic respondents or reflects shown particular, which candidates In has fashion of a 200). research in platforms Previous or policy ignorance vote. of the the out misestimate of may models “voters spatial that on research existing the altered systematically is parties to effects. voters information from by distance a par- perceived here different the propose of where we that voting, end, to this location To candidates. policy and preferred ties their from distance the measuring h feto oiia ntttos ugsigta etiea issaemoderated are biases centripetal that suggesting institutions, political of effect al. the et (Sanders appeals non-policy candidates’ (e.g., voting of models spatial extensive of Merrill more validity as and theoretical years, Adams recent the in falsifies scholars testing confronting empirical puzzles many voting of of one models but spatial are testing loca- when hence, “correct” and, the space under ascertain policy to in Attenuation trying candidates when candidates. of problems tion and theoretical parties to of rise give positions in biases centrist resulting attenuated, overestimate the- often that the attenuated are of models parties yield tenants for general and positions the equilibrium effects validate Furthermore, statistical analysis ory. proxim- most expected of if than even tests magnitudes, smaller Empirical parameter find models. spatial often of voting accuracy ity predictive the to adversely h feto rxmt.Ohr aeagetdeitn pta oe oinclude recalibrate to to model spatial assigned Romero existing is and augmented weight (Erikson have factors a Others behavioral which proximity. in of model effect the “discount” a propose ample, u hti oeshv ifrn ecpin ftemvmn fprisi the in parties of movement the of perceptions different have voters if what But u mhsso nomtoa issi ietda bevdiaeuce in inadequacies observed at directed is biases informational on emphasis Our tepshv enmd oadestepolm dm ta.( al. et Adams problem. the address to made been have Attempts Haddock ; misreported 2003 rxmt.Atnae rxmt siae n etiea biases centripetal and estimates proximity Attenuated proximity. 1999 .Teebae r o rva n nmn ae contribute cases many in and trivial not are biases These ). ;Iversen1994 aioizadMcDonald and Rabinowitz ; 1990 stretch 2011 n nomto nrgrst the to regards in information and ) eeocdsi pta oe of model spatial heteroscedastic .Shlr lohv okdto looked have also Scholars ). or 1980 compress rnegadJenks and Granberg ; h oiyspace policy the 1989 ,frex- for 2005), .Cloe al. et Calvo E. ). 2001 , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 Hinich n rfrcniae httk nmr xrm oiin Rbnwt n McDon- and (Rabinowitz positions extreme more ald on take that candidates prefer and presidential U.S. three from choice—with data Section vote magnification—using elections. on for proximity Section correcting ideological covariates. of without specific effect and candidate the esti- estimating and is of behavioral magnification results of ports where function voting a of as model mated proximity we or Finally, heteroscedastic candidate. lensing, the a ideological to propose of themselves from terms distance in ideological perceived biases voters’ these of from conceive draw and we magnification example, optics motivating of a illustrate field As We to biases. the elections space. information presidential policy these American in of 2008 candidates magnitude and the place 1996, voters 1992, the how from in data reflected use are they how and of ases function a as expand or contract covariates to of space variety policy a the allowing candidate, the and da oiyotoet htpooe yec addt (Downs their candidate from each distance by proposed the that minimizing to by outcome candidates policy ideal among select voters office. that for assume running and, candidates politics political of not—about body preferences large or the a ignorant—rationally point, hand, the are other to voters the more that On parties. shows and research candidates relative of by their models proposed measuring of policies by set the decisions rich make to a voters developed rational have how by researchers divided explain hand, remains to one decide the voters On how gulf. on conceptual literature a the research, of decades Despite Positions Candidate of Perceptions Biased with Voting 2 (Kedar Hellwig(Dahlberg actors and party (Calvo across parties power cal of distribution the 2009 of consideration the through Vote the of Model Spatial Heteroscedastic A issatrtesaeo h oiysaeue yvtr oass hi rxmt to proximity their assess to voters by used space policy model voter the candidates. of to contort shape analyst biases the the information alter allows how spatial biases propose is standard we emphasis solution the our The within Downs, perceptions. Working and candidates. Black of and model parties of locations reported models directional voting or discounting proximity Houweling than van that employed and show commonly (Tomz more to fact, designs in experimental are, (Macdon- rules used calculus have vote others the Yet enter models. positions policy al. candidate et and ald party how for els h hpe rcesa olw.Tenx eto lbrtso nomto bi- information on elaborates section next The follows. as proceeds chapter The otnigmdl fvtn ifri motn as pta rxmt models proximity Spatial ways. important in differ voting of models Contending nti hpe u oli ocnrn h bevdsseai issi the in biases systematic observed the confront to is goal our chapter this In 1989 .Eetrlrlshv lobe hw oatrteicnie aigpoliti- facing incentives the alter to shown been also have rules Electoral ). 1984 .Fnly hr ru fshlr ru htvtr lomk decisions make also voters that argue scholars of group third a Finally, ). 2001 2012 u model Our .Acmeigsho rusta oesaemtvtdb conviction by motivated are voters that argues school competing A ). epoieanïeetmt ftedge fmgicto nthe in magnification of degree the of estimate naïve a provide We . .Mn ru htteesltosipoeo rdtoa proximity traditional on improve solutions these that argue Many ). .Mr udmnal,ohr oi lentv o-rxmt mod- non-proximity alternative posit others fundamentally, More ). 5 concludes. . losu oatrteprevddsac ewe h voter the between distance perceived the alter to us allows 2008 2011 e loLc n Paolino and Lacy also see ; n h oe’ ecpino at locations party of perception voter’s the and ) 2010 ). 1957 o information how nlwand Enelow ; proximity 4 353 re- Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 hn2(.)i h nta lne ospotaohrcniae hscnb taken be can in- This she candidate. another if support scale to the less planned at on instead Obama she (5.8) places if 6 respondent conservative (1.7) around similarly 2 Obama A than Obama. places for scale) the vote Democratic to respon- the on the tends conservative” Obama on placing place “extremely 7 to of (scored self-identified in dent task a asked the scale, illus- When liberal-conservative candidate example Obama. graphs Barack 1Ð7 the as The 2008, Take where elections. in space. of candidate these policy Party’s view of in their each located affect in is self-placements candidates respondents’ party how Fig. major trate In two elections. the presidential of U.S. al. to 2008 et distance and (Adams ideological for the as vote well to as expect Respon- identify not 1976 space. not do do ideological they they the distance parties which their in overestimate with consistently parties parties leanings to of political different location predictable very and with in the parties dents differ about reporting respondents information Survey when use decisions. ways correctly their and making discern when to candidates ability voters’ of view in fterlvn ate rcniae apoiaey orcl fte r omk reasonable make to are they if correctly (approximately) ( see candidates al. to or decisions. and et parties it, relevant Abramson on the (generally list, stand of form candidates this tions some the To where in them. perceive issue between to the issue), difference the cognicize a on to opinion ability an the have as include interpreted These position. issue to cording al noqeto h vrg mrcnsaiiyt atavt ae ncandidate on issues. certainly based of view vote set a This a cast understand.” or to one do ability on they American’s positions average time the a question for also into events—and what calls political forget little how prominent they underestimate most quickly to the easy how even is about it know so, even Americans But typical it. about and something learn shallow to be affairs to tends (1992 knowledge Zaller such As politics, ephemeral. about something know public may on voter work seminar ( his on In Zaller voting scale. formation, of left-right) opinion criteria (or liberal-conservative the a meet of could basis people few the very that showed He innocent.” cally avt mn oes eciigvtr’aiiist siiaecniaepositions candidate ( vast assimilate Converse a terms, to summary abilities is in voters’ there Describing Indeed, political voters. choice. and among deficits vote naïveté information of documenting models literature comparative based and proximity American said of validity the competence/reputation, extremity, for thereof. policy candidates combination positions, some competing policy or of of terms characteristics in the per- it about office—be for something reputation know a voters (Stokes building that positions parties than or rather candidates formance with valence-issues, on based 354 1 h uhr of authors The sa xml fti hnmnn osdrvtrcoc nte18,1996, 1980, the in choice voter consider phenomenon, this of example an As pessimistic more this support to data survey existing in evidence much is There h eerho oiia nweg n oe hientrlyclsit question into calls naturally choice voter and knowledge political on research The ). h mrcnVoter American The 1992 6 usi,“ aoiypy nuhatnint public to attention enough pays majority “a it, puts 16) , 1964 agl codCnes’ iw hl h typical the While view. Converse’s echoed largely ) Cmbl tal. et (Campbell 1 ucntyage htAeiasae“ideologi- are Americans that argued succinctly ) 1963 .Ec fteeapoce assumes approaches these of Each ). 1960 1 2009 epo epnetplacements respondent plot we adotsc rtrafrvtn ac- voting for criteria such out laid ) d htvtr utseteposi- the see must voters that add ) 2005 Bartels ; .Cloe al. et Calvo E. 1988 Page ; Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 e loAase l’ ( al.’s et Adams also See ecieteTre svr iea—t22i h oe gis h at seCloe al. et will Calvo voter (see conservative party similarly the A against candidates. voted its she of 2.2—if one liberal—at for very voted as approximately she Tories lo- at conservative, voter if the very a perceive scale as scale, Party point left-right the 0Ð10 identify the (8.9) scale on left-right 9 Party the Conservative of the far-right place the to on asked cated when that show 2005 from at lsrt hi w da on,wiennspotr lc h at further party the place non-supporters while point, the ideal locate own away. systematically their supporters to party closer effects: party projection of evidence strong as 1 Fig. Vote the of Model Spatial Heteroscedastic A 2 addt lcmnsaognnspotr.Maswt 0o ee epnet o reported. not respondents fewer or Studies 10 Election with National American Means Source: non-supporters. among placements candidate lines hs issaentsrcl nAeia hnmnn o xml,Biiheeto tde data studies election British example, For phenomenon. American an strictly not are biases These eotma addt lcmnsaogcniaesupporters, candidate among placements candidate mean report 2 encniaepaeet esssl-lcmns ..18,19,20.Notes: 2008. 1996, 1980, U.S. self-placements, versus placements candidate Mean 2005 hp 0 nlsso uvydt rmFac,Nra,adBritain. and Norway, France, from data survey of analysis 10) Chap. , ahdlines dashed eotmean report 2012 Solid 355 ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 eto epooeanvlwyt noprt siiainadcnrsigbiases contrasting and choice. candidate assimilation of incorporate model to term spatial way a we novel into what a to contribute propose is that we factors chapter section the this assess in and contribution model magnification Our to means candidates. a the provide and and to voter candidate, individual, the that between implies distance picture this may section, factors next contextual the in show we As o oesmk s fcniae’paeet hnmkn hi eiin(Adams decision their making when al. placements et candidates’ of use make voters how ovret igepit hsi nw notc shprpao,mr omny sfarsighted- as commonly, more or, hyperopia as optics in known is ness. This point. single a to converge oain hl eepc l oest bev h at nasnl tu”loca- “true” actual single its a from in differs party aberration that spherical the space, observe party ideological to the the voters of in all location tion expect the we of While “image” location. an projects that ideological the in parties de- of we location bias, perceived information the on space. in literature effects extensive lensing a an similar here of observer from scribe observer—the Drawing position the voter. policy of well. the location perceived as the being the per- by politics shaped communications, to the is the elite-mass apply representative alters political about perceptions to lens altered information comes a Such revealing it space. through while When an in passing source towards matter light space background of the of distribution the across beam of travels a location it ceived of as source curving background The ex- a observer. an from developed of light has model of which spatial physics, beam of heteroscedastic on field our literature the for from tensive example comes motivating intuition The a voting. with begin us Let Describe to Example Motivating A 3 356 3 htteiaeo h at per lsro ute wyfo t relocation. true its from away further or closer behind appears party voting the When of image the that oia itne.Tecraueo hsln a eapoiae yalrenumber large ideo- a in by aberration approximated be optical can of lens degree this the of explains curvature which The near- distances. lens is logical this voter voter of the the graduation a where case is one second the as in cases distance); these a of at first focus the sighted. to of (unable think farsighted might is We should. it than the party, ovxln ufr rmshrclaerto hnlgttasitdtruhteln al to fails lens the through transmitted light when aberration spherical from suffers lens convex A emgtsrieta uhbae u oasmlto n otatefcsshape effects contrast and assimilation to due biases such that surmise might We e sasm htalvtr e h oaino at hog a through party a of location the see voters all that assume us Let n otat nPlc Distances Policy in Contrast) and uta niiul orc hi ysgtwt ess ecnseuaeta there that speculate can we lenses, with eyesight their correct individuals as Just 2005 h bet hc per lsrta tsol.We voting When should. it than closer appears which object, the focal rnegadBrent and Granberg ; h uvn fteplc pc nrsos oifrain ntenext the In information. to response in space policy the of curving the : for on per ha fteojc,wihi rjce ute away further projected is which object, the of ahead appears point at,the party, a rvttoa lensing gravitational stretch or oa point focal 1980 compress rnegadJenks and Granberg ; .. h fetta atreet na on exerts matter that effect the i.e., : fteojc pryo addt)falls candidate) or (party object the of h oiysae leigteperceived the altering space, policy the Magnification 3 hfsteve fosresso observers of view the shifts 1977 erl tal. et Merrill ; (Assimilation .Cloe al. et Calvo E. convex against 2001 lens the ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 hnvtn o h at,alrespondents all party, the for voting When ecncne h ovxln fteRpbia at tispoetdai;ta is, that axis; projected its at Party individual Republican an the exists of there where lens convex the center can We ..I h model the In U.S. mirror. parabolic a on projection parabola light the simple or of a lenses ray use of a can convexity of the we estimating example approximation) our quadratic of a sake (e.g. the for but functions, different of Vote the of Model Spatial Heteroscedastic A hn ftefis fteea rpre itne n h eoda tu distance.” “true as second the and M< distance” can “reported We as party. when the these of Thus, of location first “true” the the to of position position, think voter’s candidate’s the the of from perception distance her the and and distances: position voter’s (Euclidian) quadratic the two from of distance ratio the the as defined is magnification that Note a at,designated Party, can lsrt rfrhraa from away further or to closer rae.Ti losu oset to us allows This arrayed. the which on axis respondent n otati rxmt oeso voting. assimilation estimate of to models departure proximity models of in to point contrast physics a provides and from it lessons second, link And might purposes. choice. we voter two how of serves for example intuition previous the the provides it parties, First, of location perceived the in ases for loetmt h tu”rte hnterpre itnefo h oest the to voters the from distance reported could the we than data, rather reported in candidates. “true” magnification the of estimate degree the also explain to information had by ecndsrb this describe can We sa lutaigeape e suetecs fteRpbia at nthe in Party Republican the of case the use us let example, illustrating an As hl hr r aydfeetfntoa om htcnb sdt siaebi- estimate to used be can that forms functional different many are there While x x i i ∗ h udai prxmto sthus is approximation quadratic The . h feto h esi to is lens the of effect the 1 h ouinis solution The . i h efrpre dooia oiino h aersodn sgiven is respondent same the of position ideological self-reported The . M> L N ehv esthat lens a have we 1 L iR ∗ epnet—ahwt ifrn mg of image different a with respondents—each iR = L iR ∗ ecie h eotdlcto fteRpbia at by Party Republican the of location reported the describes x magnification rmapsto epniua otepicplideological principal the to perpendicular position a from , i ∗ =− x i L L L − M 1 2 iR ∗ x iR iR i −1 ∗ L iR = = = compress htosre h tu”lcto fteRepubli- the of location “true” the observes that iR ∗ = x L + a i ∗ 2 ( iD ∗ (x (x + M ihti qaiy ecnue(1)tosolve use can we equality, this With . b = i i + o every for bx stretches )ofthemirrorthat − − (x √ x i L L i dooia itne oevr fwe if Moreover, distance. ideological i + 1 − M iR ∗ iR = − cx ) ) c L iR x 2 2 2b i iR i 2 ∗ . x dooia,dsac n when and distance ideological, . ) i bev mgsta r either are that images observe 2 + = . b x 2 i ∗ − ..magnification. e.g. , 4ca i tahsto attaches . R sposition—are ’s R as: 357 (2) (1) (3) (4) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 h tnadpoiiymodel. proximity standard the otepoiiymdlfrom model proximity the Ra- to by proposed ( First positions. McDonald party and on binowitz information incorporate voters how turn. of in tion these of each consider Let’s va- discussed attributions. factors candidates’ voter and those components, characteristics, incorporating spatial lence for (directional) strategy non-proximal a introduction: model us spatial the gives the in into hence, attributes and, voter vote and the candidate different systemic, of provides in and representation bring this to individual particular, way both In novel clearly.” a factors, “see to which capacity voter’s about the arguments shape testing for means a we distance, to only applies variance a the this Since label magnification. the for covariates the component second model, the The the model, a measure. of heteroscedastic proximity as ideological a variance the from the rescaling different Yet model only covariates. is to variance of us set allowing a of of feature function desirable linear the has (5) in expression em hc ilb sdt sesteefc fvralsta nuemagnification: induce that variables of effect the assess to used be will which term, anfiaini dooia rxmt a lob tefafnto fohrcovari- parameter other placeholder of a function use a therefore het- itself can be as of We also level ates. such can the covariates that proximity assume ideological by can in explained we magnification etc.), is binomial, negative component models, variance probit eroscedastic the which in models covariates. with modeled explicitly be described can that as party. distances a assim- ideological explain against stretching will ( or covariates in and of for compressing number vote a contrast, that to and argue expect and ilation further that step respondents one go for can proxim- different We reported is that shown parties has to contrast ity and assimilation on literature existing The Model Voting Proximity Heteroscedastic A 4 358 anfiaineuto aen fet h exp the effect, no have equation magnification rfrtecniaewoavctstersd.I h otx fAeia politics, American of context the from point, In larger side. neutral the their select the voters advocates of who Specifically, sides candidate voter. the the opposite from prefer on distance candidate’s are the candidates from proxim- of minimizing function when utility by positive gained obtain a not is voters is but utility that ity This posits issues. the model on directional positions candidates’ the model, proximity sian ehv usiue h nua anfiainetmt ihteexponentiated the with estimate magnification angular parameter the substituted have we (5) In yepiil oeigtemgicto nteielgclsae ( scale, ideological the in magnification the modeling explicitly By si h aeo eeocdsi hiemdl(lae n Brehm and (Alvarez model choice heteroscedastic a of case the in As is,tk ietoa fet.Drcinlmdl rvd natraieconcep- alternative an provide models Directional effects. directional take First, si scmol oewe siaightrseatcdsrt oes(e.g., models discrete heteroscedastic estimating when done commonly is it As .Ide,ltu sueta anfiaini h euto nomto processes information of result the is magnification that assume us let Indeed, 4). θ eeocdsi rxmt model. proximity heteroscedastic iR BZ ota log(θ that so , savco fidvda-pcfi otoswihaeuafce by unaffected are which controls individual-specific of vector a is , U 1989 ( V (x iR within R ) i ,tedrcinlmdlhsln entecifrival chief the been long has model directional the ), ) ∼ − =− N)(L N(μ h pta oeigtaiin ieteDown- the Like tradition. modeling spatial the α (x θ iR exp ,σ i − − θ 2 (θ ) N) L oieta falcvrae o the for covariates all if that Notice . iR iR ( ) 0) ) and 2 = + θ (x iR ,ad( and 1, BZ. i nle formagnification our of lieu in − N)(L ilb eue to reduced be will 5) N ietoa voters directional , iD − N) .Cloe al. et Calvo E. provides 5) . 1995 ,the ), (5) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 eeec o( to reference than h oeta hywuda rdce oeyb pta considerations. spatial by solely predicted of shares as larger would attain they to know than parties We allow vote space. advantages ideological the valence in that positioned work bias is previous party evaluations from the valence incorporating where whether than of explore rather perceptions However, we voters’ additively, insight. advantages this valence on respond party build strategies We party preferences. how com- voter and more outcomes to yields election model understanding utility for random models the plete into components non-proximity of sion oe a enmc otse,adwt ie results. mixed with and contested, much been has model at oiin safnto ftevtrspreto fteprysvlneappeals. valence party’s the voter of a if perception voter’s that the maintain We of function a as positions party voters. across identical are advantages valence at su rxmt,tu oiigta pta fet hp tlt niety hog valence through indirectly, utility shape effects spatial that positing characteristics. thus proximity, issue party ( Hellwig and around. Calvo way other the than rather models, proximity from arising Houweling lensing van ideological and (Tomz view proximity Paolino the for and support stronger finds and problems eeocdsi pta oe fteVote the of Model Spatial Heteroscedastic A 6 5 4 epc ocmeec,itgiy hrsa n h ie(dm tal. et (Adams like with the al. reputations and et related charisma, integrity, non-positional competence, parties’ to of respect scholarship importance recent of the deal emphasized great has A appeals. non-policy candidates’ incorporating for than respon- rather magnification, where angular model. of choice of extremity the degree directly the the affecting on shaping on information as candidates uses place It dents different. weight. greater is as carry variables approach components between directional Our race of model. horse a proximity mean of whether a order discern of the analysts context on thus the are within models another rival between one Tests against pitted are often typically assumptions other are modeling effects the which or King on and or one Lewis utility of (see voter favor relaxed measure in analysts util- Conclusions how voter fashion. on affecting additive hinge component in and each directly with ity directly, them compared have however, oko oe hieoto h eairltaiinwihsosta oesdislike voters that shows which in tradition support behavioral finds the of claim out this choice familiarity, voter on on built work are considerations reputational that ec dsavnae nteohrhn,aetpclyasmdt eurltdto unrelated va- be its to and assumed hand, typically one are the hand, on another. other space, one the policy in on location (dis)advantage, party’s lence a of assessment ers’ oehn fa xeto sSnese l ( al. et Sanders is exception an of Something (2005 al.’s et Adams especially See measurement previous around get to designs experimental used has however, research, Recent h xlntr oe fdrcinlmdl eaiet h oninproximity Downsian the to relative models directional of power explanatory The erlxteeasmtos emdltedge fba nvtrassmnsof assessments voter in bias of degree the model We assumptions. these relax We et osdrvlne u oe fielgcllnigpoie e strategy new a provides lensing ideological of model Our valence. consider Next, R 2009 sata oain hntedge fmagnification, of degree the then location, actual ’s coedadSened and Schofield ; 2010 6 4 ute,tesailmdln ieauegnrlyasmsta parties’ that assumes generally literature modeling spatial the Further, ,ti ae tlkl that likely it makes this ), .W aeti sisrcieeiec o sn ieto xrmt omodify to extremity direction using for evidence instructive as this take We ). 2011). i iw h mg faparty a of image the views 1999 nfidmdl loseWtmn(1983 Wittman see also model; unified ) .Mxdfidnsaie ietoa n proximity and directional aside, findings Mixed ). 2006 .Teesuisdmntaeta h inclu- the that demonstrate studies These ). h oe aec safnto fvoter- of function a as valence model who 2011) (x i − L iR ∗ ) 2 >(x R 4 spoial lsrt her to closer proximally as M et ftetomodels, two the of Tests i huddces.With decrease. should , − L iR ,Goels (2001 Groseclose ), ) 2 oteextent the To . 2005 5 2008 u vot- But ;Clarke ;Lacy 359 ), Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 erwise. ilscnitn predictions. consistent yields rsdnilcniaesparty, candidate’s presidential ucino oiyeteim aec,adpltclifrain xrse as expressed information, political and valence, extremism, policy of function atr o oe hieb hreig r“lrfig”teiflec fideological of politics influence the about “clarifying,” information or sharpening, to distance. by exposure choice whether voter for examine matters we information model, which proximity through tic Wilson surpris- and pathways (Gomez not different present Perhaps are out outcomes. effects sought election have extension, by researchers and, ingly, choice voter for matter (Converse affairs knowledge objective 1996 and Americans’ institutions of generally political and dearth large about the a acquisi- exists highlighting individual’s there the above, literature noted by uncontested As altered politics. is about vote information of the voter tion of on effect location the how candidate model of to way perceptions a provides model proximity heteroscedastic the iaea oeeteebto h aesd ftenurlpita esl,and herself, as point neutral the of side same the Valence, on otherwise. but 0 extreme more as didate rswrh,caimtc n h ie hudrcieabae vlainb the by evaluation between distance biased the a is, receive (that terms should positional like, in voter the and charismatic, trustworthy, tmakn hte hr sayhn hydsieaottetomi ate.Wt epne to and responses dislike, With and parties. like an both main scored by or scale two followed three-point dislike, the a is nor construct about like This we dislike items, Parties. they choice Republican binary anything two and is these Democratic there the whether about asking like item they anything is there ob siae.Tedrcinleffect, directional The estimated. be to netit n eitspotn ate hyko iteaot(vni hyshare they if (even about little preferences). know policy they party’s parties the supporting resist and uncertainty 360 9 8 7 teto h epnetpy onw bu oenetadpolitics. and government about news to pays respondent the attention ences, rsdnileetosi 90 96 n 08uigteMro hi Monte Chain Markov the 2008—using and 1996, 1980, in elections presidential ee-on cl o at dnicto.Ti setrdit h pcfiaini ( in specification the into ANES entered standard is the This using identification. dispositions party for partisan scale respondent’s seven-point the for control we that n(6), In elo attention.” of deal ihdsrbefaueo loigu omdltevariance, the model to us allowing of feature desirable with tics, spr of part as h esr scdd1 coded is measure The whether identify to respondents ask surveys Studies Election National American the Specifically, (1997 Alvarez others, among See, ihti nomto,tehtrseatcpoiiymdli ssoni (5) in shown as is model proximity heteroscedastic the information, this With eetmt e fhtrseatcpoiiymdl—n ahfrU.S. for each models—one proximity heteroscedastic of set a estimate We I .Mr otse mn coasi hte uhifraindiscrepancies information such whether is scholars among contested More ). i T 8 D represents iR h oiia nomto variable, information political The iR BZ is ersnsvoter represents i h etro controls. of vector the , sassmn of assessment ’s i sepsr opltclifrain n the and information, political to exposure ’s = T dntpymc teto, 2 attention,” much pay “don’t iR scoded is , θ iR i )andBartels( − spreto fteeteiyof extremity the of perception ’s = 7 R fsedsie ntigaottepry n oth- 0 and party, the about anything dislikes she if 1 ate h oesve sbigmr competent, more being as view voters who Parties snnpstoa ulte,o aec characteris- valence or qualities, non-positional ’s γ 1 D + iR + ftersodn ie ntigaotthe about anything likes respondent the if 1 2001 D o ieonly. like for 1 + .Eeo n iihs( Hinich’s and Enelow 1996). iR γ ssoe ftevtrpae h can- the places voter the if 1 scored is , I 2 Zaller ; i T sasbetv esr fhwmuch how of measure subjective a is , iR + = pysm teto, 3 attention,” some “pay 1964 γ 2004 3 I i . x el apn n Keeter and Carpini Delli ; i .Uigorheteroscedas- our Using ). and − θ ilk ny o neither for 0 only, dislike 1 iR L pcfida linear a as specified , iR R γ 1981 ssal.Lastly, small). is r parameters are s splc prefer- policy ’s 9 omlmodel formal ) ial,note Finally, = .Cloe al. et Calvo E. pyagreat a “pay (6) 5) Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 bu h oe’ eiinwt epc oielgclpoiiy ecnie each consider We proximity. ideological to turn. respect for in with account to decision to voter’s pertaining specified covariates the are remaining about effects The information elections. and 2008 valence, holds and directional, finding 1996, This significance. 1980, statistical the attain across model variance variance, choice the the the specify in When do estimates we 5). these when 3, However, estimated. 1, imprecisely (Models are on not signed, estimates tively do parameter that unspecified, those left infor- is and and model 6) valence, extremity, models 4, ideological in 2, of proximity function (Models dis- positional a mation as ideological of variance effect for the the model results do in the that differences with observe lies we however, Here, interest, tance. Our esti- The case. precisely candidate. each and the in signed mated selects positively are respondent partisanship the for likelihood coefficients the choice-specific on distance ideological of hiemdli ute pi ewe h etro xgnu otos(at iden- (party controls exogenous of vector the The between component. tification), variance split the further for is other model the choice and al. model choice et the (Spiegelhalter for WinBUGS equations—one in engine (MCMC) Carlo *** 1 Table Vote the of Model Spatial Heteroscedastic A el eotcefiinsadsadr rosfo siaightrseatcpoiiymdlde- model proximity heteroscedastic text estimating the from in errors scribed standard and coefficients report Cells N LogLik News to Attention Effect Directional Model Variance Ideological Constant Identification Party Distance Ideological Model Choice at Valence Party p< Table 0. eeocdsi rxmt oes ore mrcnNtoa lcinStudies Election National American Source: models. proximity Heteroscedastic 001, 1 rsnstemdlrsls h hiemdlicue h siae effect estimated the includes model choice the results: model the presents BZ ** n h dooia itnecomponent, distance ideological the and , p< 0. 8813 5027 041418 3064 2076 2570 1736 1838 (1.080) Ð0.594 0.099 0.096 (4.656) Ð0.750 Ð0.039 0.094 Ð0.056 0.071 (10.973) Ð0.190 Ð0.290 0.040 (0.009) Ð0.065 (0.018) 0.029 Ð0.067 (0.746) Ð0.068 9018 9619 082008 2008 1996 1996 1980 1980 123456 10. 987Ð392Ð058Ð774Ð753.1 Ð1717.4 Ð1075.8 Ð1389.2 Ð998.7 Ð1102.1 01, * p< ** 0. 008008Ð0.210 Ð0.028 (0.046) 0.078 Ð0.398 (0.099) Ð0.088 (0.171) Ð0.811 .4 .9 1.252 0.698 (0.092) 0.747 (0.009) 05, + *** *** *** *** p< 0. (0.008) (0.033) (0.302) ,totie tests two-tailed 1, *** hl nega- while Distance, Ideological (0.009) (0.101) (0.118) (x *** *** + *** ** i 2003 − L (0.008) (0.010) (0.060) .W siaetwo estimate We ). iR ) 2 . *** variations (0.011) (0.067) (0.132) (0.198) 361 *** *** ** *** Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 pl,hwvr ote20 lcin nti case, this In election. 2008 the to however, apply, itneta oeswoddnt ntrso dooia esn,tedrcinlef- directional the lensing, ideological of ideological terms on candi- In weight the not. fect (negative) viewed did greater who who put voters voters than themselves cases, distance than these extreme In more elections. as 1996 date and 1980 the for case rmcmn cosa ihrtocnevtv rtolbrlo osraieteway the conservative suffer or not this liberal did too from Obama, did. or infer conservative predecessors Barack too their can and either We McCain as John voting. across 2008, coming proximity the from in on of candidates effect the side that adverse preferred penalty result the no this (on 2008, had by positions point) And extreme neutral coefficients. taking the disappeared: of in Carter sizes altogether and relative had Reagan the facing by that evinced than less as penalty” was 1980, “extremity 1996 for in the However, Clinton Clinton, and positions. Bill Dole im- extreme and confronting equation as Dole perceived variance Bob they candidates, the what the in taking punished term voters proximity directional that The the plies 1996. on to applies coefficient story signed same negatively The candidates. position-taking es- extreme precisely on and he penalty large extreme the how 1980 of In on function coefficient be. timated a to as ideology candidate’s president particular for the Carter viewed Jimmy or Reagan Ronald terms. port policy on penalty no receive candidates placed extremely that meaning . xliigteEfc fVlneo rxmt Voting Proximity on Valence of Effect the Explaining 4.2 while is, That candidate greater. to is penalty candidates a non-proximal attaches model the proximity to the attached and voter, penalty the the than that extreme A more space. is candidate ideological the dis- the onthesamesideof when ideological of that indicate center when would vot- less the term matters to to directional tance distance or the extreme on candidates’ the coefficient that signed to so positively move stretched they as is differs space specification ers ideological heteroscedastic The the effects. that directional implies of influence the consider First Proximity on Extremity Candidate of Effect the Explaining 4.1 362 o h aec aaeesaecnitn coseetos n18,19,ad2008, and 1996, 1980, in on elections: estimated estimate across coefficients consistent the effect, are parameters directional valence the the Unlike for effects. valence consider Next nteetee fvlnei ufiinl ih oe ilprev httecandidate the that smaller. perceive will becomes voter vote a the high, sufficiently on valence is distance party’s valence if ideological a extreme, of As the smaller. In effect disutility the higher the up, increases, and goes party advantage smaller the distance of the that and makes location valence policy preferred voter’s the between antd fta eat sgetrif greater is penalty that of magnitude em ftehtrseatcmdl hsmasta svlneicess h voter’s the increases, valence as that distance, means ideological this perceived model, heteroscedastic the of terms hs eut ugs hti 90 yia voter typical a 1980, in that suggest results These stretches Voting h itnebtentevtradtecniae.Ti tr osnot does story This candidates. the and voter the between distance the at Valence Party N ietoa Effect Directional hnohrie eaiesg,o h te ad means hand, other the on sign, negative A otherwise. than , (x spstvl indadsaitclysgicn.In significant. statistically and signed positively is i − L γ iR 1 niae htseasgsarltvl heavy relatively a assigns she that indicates ) < 2 hik.Ptdfeety stedistance the as differently, Put shrinks. , .Table 0. γ 1 i R sidsigihbefo zero, from indistinguishable is 1 a esadls ieyt sup- to likely less and less was hw htti si atthe fact in is this that shows when L iR sfrfrom far is .Cloe al. et Calvo E. x i ,the Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 ec ilapa ute eoe n uhmr dooia.Aan hstatre- trait this va- Again, low results. ideological. model with more all parties much in while constant and their pragmatic, mains removed to and further marks themselves appear valence to will high close lence attach them who see voters ac- will words, their party other of irrespective In pragmatic, location. as policy parties tual valence spatial high and of ideological effect as parties the declines, valence proximity as spatial contrast, out” inter- pronounced. By “drown more this becomes voting. to that proximity is of Notice valence determinant line). high (dashed a of valence as effect low the has that it shows if pretation than line) (solid valence from away prahsismxmm ebcmssailycoe oec n vr oe nthe in voter every and spatial each to of closer population. utility spatially becomes the he and maximum, proposed, its effect, approaches policy In constant. the remain of will irrespective voting her,” proximity to next “right is Voter voter probability the 2 Fig. Vote the of Model Spatial Heteroscedastic A 10 uprstecniae iuae rbblte r ae nprmtretmtsfo Table from estimates parameter election presidential on U.S. based 2008 the are for probabilities 6 Model by Simulated perceived candidate. as the position supports candidate’s the how indicates and figure The 5. position ftecniaei lolctda ,then 3, at located probability also is candidate the If Ð cl)uigprmtretmtsfo oe nTable in 6 Model from estimates parameter using scale) 1Ð7 nti lutain h te addt ntetocniaerc spae t5o h Ð scale. 1Ð7 the on 5 at placed is race two-candidate the in candidate other the illustration, this In h nuto ssrihfradadsrrsn:vtr ilprev o valence low perceive will voters surprising: and straightforward is intuition The Figure i oi n ahdlines dashed and (solid party candidate’s the of valence perceived ’s i h feto at aec ntehtrseatcpoiiymdl oe:Fgr displays Figure Notes: model. proximity heteroscedastic the in valence party of effect The slctda nte17ielg cl.Teohrcniae(o hw)i oae at located is shown) (not candidate other The scale. ideology 1Ð7 the on 3 at located is 2 i ( speerdlcto,i oe esuiiyi ti emdt aehigh have to deemed is it if utility less loses it location, preferred ’s lutae hsefc o oeaeylbrlvtr(oae t3on 3 at (located voter liberal moderately a for effect this illustrates ∼0.63) i ned ovt o addt stecniaemvsi oiyspace. policy in moves candidate the as candidate a for vote to intends eadeso t aec level. valence its of regardless i rfr h addt iheulyhigh equally with candidate the prefers sacniaesvlneadvantage valence candidate’s a as 10 u stecniaemoves candidate the as But 1 o h 08election. 2008 the for fetteprobability the affect ) i ( oiotlaxis) horizontal B&W IN PRINT 363 1 i Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 addt.Smltdpoaiiisaebsdo aaee siae rmTable from election estimates presidential parameter U.S. on 2008 based are probabilities Simulated candidate. oae tpsto .Tefiueidctshwtecniaespsto sprevdby perceived as position candidate’s the how indicates axis figure The 5. position at located h an rmpoial oae addtsaelwrbts r h ossin- losses Figure the continuum. are ideological so the but on lower away are further candidates moving located by curred proximally from gains the vrdcie ail mn h nomda h addt oe wyfo the from away moves how- candidate utility, the This as increased. informed as levels i.e., the voter, information among as rapidly declines greater ever was Obama for voting close proximally located those costetreeetos nte18 n 96pls teto onw a no had news to attention polls, 1996 and tiontoNews 1980 the on In effect biasing elections. three to the proximity across ideological use information measure, to information case, likely our this im- more vote—in hand, are other their informed the inform politically on the coefficient, im- negative that component A plies variance individuals. informed the is among in distance utility variable ideological above: information as that the applies plies logic on same coefficient The in- political positive any, lensing. of if a ideological effect, influence on po- what has the examine to acquisition We ascertain behavior. formation attention voting to individual’s as sought an models on have information our researchers in Many news. captured litical effects, information consider Finally, Distance Ideological and News to Attention 4.3 Voter space. voter policy probability the displays Figure 3 Fig. 364 )and h feto nomto atnint es ntehtrseatcpoiiymdl Notes: model. proximity heteroscedastic the in news) to (attention information of effect The i slvlo teto onw ( news to attention of level ’s (x speieyetmtdadngtvl ind hsmasthat means This signed. negatively and estimated precisely is i − i n20,hwvr h ofceton coefficient the however, 2008, In Distance. Ideological slctda nte17ielg cl.Teohrcniae(o hw)is shown) (not candidate other The scale. ideology 1Ð7 the on 3 at located is L iR ) 2 nrae.Aogtels nomdielg atr less: matters ideology informed less the Among increases. osnteettesm eea effect general same the exert not does News, to Attention oacniae(a aakOaa,teuiiyof utility the Obama), Barack (say candidate a to i fetteprobability the affect ) lines dashed and solid ned ovt o addt stecniaemvsin moves candidate the as candidate a for vote to intends rta dooymtesfrvoter for matters ideology that or compressed, stretches itne eut hwthat show Results distance. 1 oe o the for 6 Model .Cloe al. et Calvo E. 3 i i

uprsthe supports B&W IN PRINT illustrates ( horizontal among Atten- Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

690 689 688 687 686 685 684 683 682 681 680 679 678 677 676 675 674 673 672 671 670 669 668 667 666 665 664 663 662 661 EDITOR’S660 659 658 657 656 655 654 653 652 651 PROOF650 649 648 647 646 645 rvd oe en o hwn o aec lnstepoiiyeffect. we proximity elections—yet the blunts across valence matter how showing also for like, means novel the a and provide re- integrity, with competence, appeals, valence we to non-positional elections, spect ideology Candidates’ consistently. and 2008 so partisanship surprisingly, and do Not and 1996, change. matter, and 1980, continuity the of voter mix in a of calculus uncover model voter’s proximity heteroscedastic the the Comparing via utility. effects lensing the model we (Ta-when utility voter of predictor stronger on ble a positions becomes distance extreme ideological the effect, at lensing Fig. self-identify partic- (see at biased, scale liberal-conservative those highly the among is exclusively however, not positions, The but policy select considerations. ularly candidates’ (ideological) do the policy States of based United view president voter’s the of office in the Voters to politics. candidates presidential American into insights h supin negrigsailmdl fvtn r ynwfmla:1 vot- 1) familiar: now by are voting of ers models spatial undergirding assumptions The Remarks Concluding 5 set again We election. 2008 the from estimates x parameter using again dynamic, this Vote the of Model Spatial Heteroscedastic A adn ieto,ortrepro nlsssosta h eat fcandidates’ of penalty the that shows Re- analysis contests. three-period electoral our across vary direction, both garding component com- information the information and an component and component, this valence Using a magnification ponent. in of component, model magnification directional Our adjust- covariates. angular a make non-proximity includes of of for—or level trio correct a the to via voting model means proximity we a As aberration, States. to—ideological United ments the in elections a dential of location actual the than rather space ideological the the in party. observe located individuals elec- party that a in assume we of lensing optics, image to—ideological in models solutions sheds lens offer from research of—and Borrowing this tions. problems physics, the from on insights light on new Drawing voting. of models proximity proximity heteroscedastic to (b) and lensing ideological more it. measure a to for offers (a) correct work way our informed perceptions, theoretically voters’ and the cogent in bias perceptual rela- this their acknowledge measuring effects information when that metrics show compress we different Furthermore, use parties. to to proximity voters tive allow Employing we single-peaked. particular, and In transitive are novel preferences a voter 3) and candidates; i = ae oehr h eut fteehtrseatcpoiiymdl provide models proximity heteroscedastic these of results the together, Taken nti hpe,w ple h eeocdsi rxmt oe otrepresi- three to model proximity heteroscedastic the applied we chapter, this In our information, in changes to response in vary to distances spatial allowing By know .Ytpraso raetitrs osuet fAeia oiiscm from come politics American of students to interest greatest of perhaps Yet 1). 3. eeocdsi rxmt model proximity heteroscedastic hi rfre oie;2 voters 2) polices; preferred their h oiysaei ytmtcwy.Wiew aentbe h rtto first the been not have we While ways. systematic in space policy the heteroscedastic prahi bet xli teuto issi current in biases attenuation explain to able is approach rxmt oe,w hwta h directional the that show we model, proximity .Adoc emdlte“hp”o this of “shape” the model we once And 1). eaeal orlxteeassumptions. these relax to able are we , know h eeldplc rfrne of preferences policy revealed the stretch and 365 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

736 735 734 733 732 731 730 729 728 727 726 725 724 723 722 721 720 719 718 717 716 715 714 713 712 711 710 709 708 707 EDITOR’S706 705 704 703 702 701 700 699 698 697 PROOF696 695 694 693 692 691 en o epnn u nesadn bu o oesdecide. voters how turn, about in understanding and, than our elections deepening proximity across pro- for effects ideological has means these contribution a of comparing our systematically basis short, for tool the In a sources? on vided other vote from to information of receiving likely Fox those sources like more outlets different MSNBC news among charged or politically distinguish News more also of viewers might Are information. work political Future extreme. more param- is directional as voting the on proximity coefficient that positive applied a indicating model yield eter, proximity would heteroscedastic election from the U.S. extrapolating 2012 that the example, case to the For be provided. might it have trends, we current framework the upon United improve the in politics presidential of nature characterizations volatile common States. with increasingly comport changing, changes po- the more these of How- the of among positions. stronger Both was policy or informed. voting candidate proximity (stretching) litically election, and enhancing 2008 recent voter of the of terms in ever, in effects effect the no (compressing) had blunting information to (Mc- poli- access sustainable American periods, electorally is that al. polarization sentiment et such general that declined Carty and a have polarized with penalties become has such comports be tics issues, tendency to the This used on time. candidates positions with presidential extreme while taking that by suggests penalized trend This zero. essentially apelA ovreP ilrW tksD(90 h mrcnvtr ie,NwYork New Wiley, voter. American The (1960) D Stokes information W, Miller lensing, P, ideological Converse A, beholder: Campbell the of eye The (2012) K Chang T, Hellwig E, systems. Calvo electoral different York under New incentives Press, centrifugal University and Centripetal Cambridge (2011) elections. T and Sci Hellwig committees Polit E, of Calvo J theory Am The (1958) elections. D presidential Black in effects information University votes: Princeton Uninformed choice. (1996) public L of Bartels dynamics the and a primaries of Presidential development (1988) policy: L abortion Bartels towards ambivalence American Arbor (1995) Ann Press, J Michigan Brehm of Uni- University RM, Cambridge elections. Alvarez competition. and party Information of (1997) theory RM Alvarez unified A multiparty (2005) in B representation Grofman III, policy S and Merrill CQ strategies J, elections. party Adams 2008 Modeling (1999) the III in S continuity Merrill and J, Change Adams (2009) DW Rohde J, Aldrich PR, Abramson References coefficient the of size the Indeed, effect, time. directional over the declined on as positions extreme taking 366 in fteSailMdl.IsiuoJa ac,2 pi 2012 April 27 March, Applica- Juan “Contemporary Instituto conference Model”. the Spatial for the prepared of and tions presented Paper vote. the and effects, 55(1):27Ð41 Sci Polit J Am 40:194Ð220 Princeton 39(4):1055Ð1082 Press, Sci Polit J Am values. competing of model probit heteroskedastic York New Press, 43:765Ð791 versity Sci Polit J Am extreme? so strategies are why elections: Washington Press, uuewr neetosi h ..adesweesol ih xedand extend might should elsewhere and U.S. the in elections on work Future 2005 .A o oiia nomto,orrslsipyta nearlier in that imply results our information, political for As ). D iR shl sgeti 96a n18,adb 08is 2008 by and 1980, in as 1996 in great as half is , greater mn hs eciigcandidates perceiving those among .Cloe al. et Calvo E. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

782 781 780 779 778 777 776 775 774 773 772 771 770 769 768 767 766 765 764 763 762 761 760 759 758 757 756 755 754 753 EDITOR’S752 751 750 749 748 747 746 745 744 743 PROOF742 741 740 739 738 737 ita 18)Cniaemtvto:asnhss mPltSiRv77:142Ð157 Rev Rev Sci Sci Polit Am Polit synthesis. Am a motivation: choice. Candidate voter (1983) and D Wittman positioning Candidate 57:368Ð377 (2008) Rev RP Sci Polit Houweling Am van competition. party M, of Tomz UK. models Cambridge, Spatial 1.4, (1963) manual D user Stokes WinBUGS (2003) D Lunn N, Best Cambridge A, politics. Thomas legislative D, Spiegelhalter and elections democracy: Multiparty elec- (2006) of I dynamics Sened the and N, stokes, Schofield Downs, (2011) P Whiteley MC, Stewart 83:93Ð HD, Rev Clarke Sci D, Polit Sanders Am voting. issue 70(3):742Ð752 of Rev theory directional Sci A Polit (1989) Am SE ambiguity. McDonald political G, of Rabinowitz of theory projections The voter (1976) in BI effects Page contrast and Assimilation (2001) unequal J and Adams ideology B, Grofman of III, dance S the Merrill America: Polarized (2005) to H efforts Rosenthal further K, on Poole science: N, versus McCarty 8(1):21Ð33 Sophistry Anal (2001) Polit O voting. Listhaug proximity vs. G, directional Rabinowitz Elect on SE, evidence experiments. No Macdonald using (1999) G voting King directional JB, versus Lewis proximity Testing York (2010) New P Press, University Paolino Cambridge D, parties. not Lacy policy, test for Voting a (2009) democracies: O European on Kedar Western ambivalence in of representation impact and the leadership member: Political party (1994) a T of Iversen less leader party advantage. a valence Making a (2003) has G candidate one Haddock when location candidate of model A elec- American (2001) the T Relat in Groseclose voting Hum economic election. and sophistication 1972 Political the (2001) JM in Wilson effects BT, contrast Gomez and Assimilation (1977) R Jenks D, Granberg candidates candidates: presidential of Am positions vote. issue of the Perceptions of (1980) model E behavioral Brent the D, and Granberg equilibrium Candidate (1990) D Romero University R, Cambridge Erikson introduction. an voting: of theory spatial The (1984) model. MJ spatial Hinich Downsian JM, the Enelow in uncertainty voter to York approach New new Row, A and Yale (1981) Harper MJ matters. democracy. Hinich it of JM, theory Enelow why economic and An politics (1957) about A Downs know Americans What (1996) and S Keeter parties M, political Carpini systems, Delli electoral of and impact Ideology The (ed) matter? D context Apter In: Does publics. (2012) mass S in British Dahlberg systems the belief and of politics nature Performance The (1964) (2009) P PF Converse Whiteley MC, Stewart D, Sanders HD, Clarke Vote the of Model Spatial Heteroscedastic A 102(3):303Ð318 http://www.mrc-bsu.cam.ac.uk/bugs York New Press, University 41(3):287Ð314 Sci Polit J Br choice. toral 121 40(2):199Ð221 Res Polit J Eur USA. the and France Norway, from evidence locations: party Cambridge Press, MIT riches. 63(2):482Ð500 Polit J model. proximity the rehabilitate 29(3):460Ð471 Stud 38:45Ð74 Sci Polit J Am voting. of models three 24(4):769Ð780 of Psychol Polit attitudes. party political in effects contrast and assimilation 45(4):862Ð886 Sci Polit J 45(4):899Ð914 Am Sci Polit J Am attribution. heterogeneous of theory a torate: the to closer are 30(7):623Ð640 that 68(6):617Ð625 issues Sci the Am on are. positions really holding they as than views supporters supporters’ their by perceived often are 84:1103Ð1126 Rev Sci Polit York New Press, 25(3):483Ð493 Sci Polit J Am Haven New Press, University manuscript, Unpublished positions. party of Gothenburg of perceptions University voters’ on characteristics individual 206Ð261 pp York, New Press, Free discontent. Cambridge Press, University Cambridge voter. 367 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

828 827 826 825 824 823 822 821 820 819 818 817 816 815 814 813 812 811 810 809 808 807 806 805 804 803 802 801 800 799 EDITOR’S798 797 796 795 794 793 792 791 790 789 PROOF788 787 786 785 784 783 alrJ 20)Fotn oesi ..peieta lcin,14Ð00 n ai E Snider- WE, Saris In: 1948Ð2000. York elections, New presidential Press, U.S. University in Cambridge voters opinion. Floating mass of (2004) origins JR and Zaller nature The (1992) JR Zaller 368 a M(d)Suisi ulcoiin tiue,nntiue,maueeterr n change. and error, Princeton measurement nonattitudes, Press, attitudes, University opinion: Princeton public in Studies (eds) PM man .Cloe al. et Calvo E. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 dooia miut mrvspeitoso h tnadsailvtn models for voting accounting spatial that standard finding the behavior of (Alvarez voting predictions on improves mostly ambiguity focused ideological hand, other the on om o spit u spoaiiydsrbtosoe oiysae ned the Indeed, space. policy over distributions probability as but points as not forms Cukierman and (Alesina platform certain anyo h candidates(Glazer the of tainty 1976 (Shepsle voters the of .Shfil ta.(eds.), al. et Schofield N. Lithuania Vilnius, LT-01129, Economics, e-mail: and Management of University ISM DOI ilcniaei h ..21 lcin,hsbe osatyacsdo remaining of presiden- accused a issues. constantly policy Romney, been key has Mitt on elections, vague example, 2012 too U.S. For the altogether. in details issues candidate tial about about ambiguous talking remain avoid statements, or policy platforms, clear contradicting communicating make of instead politicians however, based representatives often, their Quite choosing positions. voters those par- and on of issues terms policy in on competition positions electoral taking of ties think to wisdom conventional become has It Introduction 1 Keywords Rozenas Arturas Data Survey from Ambiguity Ideological Inferring .Rzns( and Rozenas Seligson, A. Mitchell Li, Fan suggestions. Reiter, and Jeremy comments Desposato, for Scott Stimson Aldrich, James John to grateful am I 2 1 anpooe a oeabgospa ormi nIa o u o10years.” 100 to “up for Iraq in remain to Mc- plan John whereas ambiguous months more 16 far within proposed Iraq from Cain troops U.S. the withdraw to promised isn data Missing OaaFesPlotDbt ihRemarks’, With Debate Pullout Lehigh, Fuels Scott ‘Obama by details?”, Romney’s Mitt are “Where example, For hs xmlssgettertcladeprclraost ra oiyplat- policy treat to reasons empirical and theoretical suggest examples These oepanielgclabgiy pta hoit aerfre orisk-attitudes to referred have theorists spatial ambiguity, ideological explain To 10.1007/978-3-642-35239-3_18 ,cnetdpnec fvtn eiin CladradWilson and (Callander decisions voting of context-dependence ), [email protected] 1997 B dooia placement Ideological ) Bartels ; dacsi oiia Economy Political in Advances 1972 1986 ,tedsr fpltcast vi iiieise (Page issues divisive avoid to politicians of desire the ), Campbell ; 1 pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , 1990 nte20 rsdnilcmag,Brc Obama Barack campaign, presidential 2008 the In · Ambiguity ,o taei eet fntcmitn oa to committing not of benefits strategic or ), 1983a uy4 2008. 4, July Times, York New 1990 , ozadvnHouweling van and Tomz b; ;Aldrich · , Bayesian 1995 otnGlobe Boston · aetvariables Latent .Eprclresearch, Empirical ). 2008 ue2,2012. 27, June , ,uncer- ), 2009 · 2 369 ). Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 naplc cl.Hwvr av plcto fteeapoce swnig sI as parties wanting, political is approaches place these to of experts application now. naive political discuss a or However, citizens scale. policy ask a which on sets, data many to lc oiia coso cl nafr fa nevlrte hnapit(Tomz point a than rather interval an of form a in Houweling scale van a and (Alvarez on candidate actors given uncer- political their a report place of to position respondents the asking about by ambiguity. measures tainty ideological direct measuring uses for approach approaches first The two offers literature existing The Measurement Ambiguity and Data Survey 2 al. et (Ansolabehere points’ al. often ‘ideal et points, Laver as are positions to these referred that assumption of even measurement the the under on operates scholarship positions existing The policy lacking. are point quantity this a measure dispersion. as to small platform vanishingly a a with whereas distribution distribution a dispersed as platform defined highly position” posi- be a a “no can have a by to competition: described more not party be a candidate of provides can a study distribution or empirical probability party to a a approach for as general means position it policy what Conceptualizing clear tion. not is it points, as Whitefield and (Evans positions al. et ideological Kitschelt defined lack to known are parties va- (Aldrich a 2003 preferences involving policy parties cannot diverse political it with decentralized example, activists of For of riety reach. positions policy limited study very to a applied has be points as platforms policy of notion 370 l.Frt ihdge fdsgemn ewe h epnet adhnehigh hence (and respondents the between disagreement of degree high a First, els. stan- ( sample Campbell issue by the on suggested use as candidates to placements intuitive the political appears of place deviation It dard judgments. to in asked variation are generates scales respondents where survey Every Disagreement Respondent Interpreting 2.1 (Bartels data survey missing the (Campbell in respondents the among disagreement from especially ambiguity, ideological of study context. systematic cross-national a a in for approaches these use to and depth temporal of expert terms or in polls available widely opinion is width. (e.g., geographical that data data survey of from kind surveys)—the ambiguity ideological estimate to lhuhteeaemlil esn osuyielgclabgiy fcettools efficient ambiguity, ideological study to reasons multiple are there Although nte prahi oueidrc ehd hr miut sifre either inferred is ambiguity where methods indirect use to is approach Another .Aohrcs ocrsdvlpn eorce,wee o ayreasons, many for where, democracies, developing concerns case Another ). 2003 1999 atnadQuinn and Martin ; Mainwaring ; 2009 .Ufruaey uhsresaerr aigi difficult it making rare are surveys such Unfortunately, ). 1983a σ , ˆ 1995 .Hwvr h nuto sflwdo eea lev- several on flawed is intuition the However, b). sa siaeo at’ dooia ambiguity ideological party’s a of estimate an as 1986 2002 Scully ; .Teeidrc ehd a eapplied be can methods indirect These ). .Ti ae rsnsasaitclmodel statistical a presents paper This ). 1995 .I oiypstosaedefined are positions policy If ). 1983a 1997 2001 1983 , )orbyaskingthemto rfo h patterns the from or b) lno tal. et Clinton ; ilradSchofield and Miller ; .Rozenas A. 2000 2004 σ ˆ ) ; ; Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 cln fet a euti alyifrneaotielgclambiguity. ideological about inference faulty in the result measurement for can adjustments the effects proper scaling of without respondents interpretation among different disagreement might Treating to they scale. well-informed, due are respondents placements and conflicting ambiguous not provide is party a if even fect: respondent- disagreement. the respondent observed disentangle the to on effects have party-level ( we and Campbell level ambiguity, position al. policy ideological et party’s (Marks estimate a respondents correctly of of part ambiguity the intrinsic on an information or of lack the indicate may ( Laver and Benoit in deviation standard 1 Fig. Data Survey from Ambiguity Ideological Inferring ifrn praht miut esrmn sofrdb atl ( Bartels by offered is measurement ambiguity to approach different A Values Missing Interpreting 2.2 induces procedure an for measurement that demonstrated the universally easily scale, is mean, almost ordinal sample are the an respondents between on dependence the parties Since place scales. to required placement of nature ordinal ugssta epnet r oelkl o opaeapryo oiysaeif scale is policy uncertainty a of source on the party model, a Bartels’ In place platform. to its not about uncertain likely are more they are respondents that suggests iuu ipydet h ahmtclpoete fteestimators the of am- properties less mathematical as the evaluated to be due necessarily simply will biguous positions extreme with parties Therefore, fcniae n hi miut,o tcnmrl ea riato h measurement the of artifact an use be we merely if can it model; or ambiguity, their Fig. and in candidates positions of party on Fig. data in real the pattern in quadratic represented well is pattern this deed, stecase. the is h hr a of flaw third The ef- scale-heterogeneity the to due occur might disagreement respondent Second, apema and mean Sample 2006 xetdata expert ) σ ˆ soretmt fabgiy esml anteaut which evaluate cannot simply we ambiguity, of estimate our as σ ˆ steetmtro dooia miut tm rmthe from stems ambiguity ideological of estimator the as 1 ol ersn h tu’rltosi ewe positions between relationship ‘true’ the represent could M aeoymaueetscale, measurement category μ ˆ n h apesadr deviation, standard sample the and , 1983a σ ˆ ≤ μ ˆ , .Tu,to Thus, b). 1986 μ(M ˆ and ,who ), 1.The σ 2007 −ˆ ˆ σ .In- ˆ μ) 371 .It ) . Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 ontol ihtersodn-ee nweg u lowt h miut fthe of ambiguity the to evaluated. with have being also may is but platforms knowledge that respondent-level about platform the uncertainty with the only that can not argue One do and media. further to idea exposure this or extend education like characteristics personal respondents’ 1 Table 372 r eeae natose rcs.I h rtsae epnet eciea‘true’ a perceive respondents stage, first the In process. two-step a respondents in generated by are provided data that Suppose Model A 3 on information of source additional an as data missing and separate ambiguity. in disagreement, (2) ideological respondent patterns effects, observed the scale-heterogeneity on exploit effects the (3) party-level for and adjust respondent- (1) the should ambiguity ical greatly varied made non-response they statements not high the content. did or such their either policy in have the parties on policy—the PLL statements given public and the any PAD about make ambiguous that are knowledge- assume they least the because to be rates reasonable parties to appear two is 11 the It and are 6 able. experts PLL and and rates PAD non-response rates: high with non-response ( the in Laver effects and expert-specific Benoit in survey expert xetPryPDPDJP PDr PD PBDNJ PAD Expert/Party 16 15 14 13 12 NA11 10 12NANA18516 8 5N 4N 31 21 1 6N 9718 7N Table nsm h icsinsget htapoe ehdfretmtn ideolog- estimating for method proper a that suggests discussion the sum, In neapeo isn aaptenfo eotadLvr( Laver and Benoit from pattern data missing of example An 1 hw necrto xetdt nAbna oiia ate rmthe from parties political Albanian on expert-data of excerpt an shows A91 31 378 7 13 14 13 12 9 NA A51 5N 533 3 5 4 3 4 3 15 7 11 7 18 NA 9 6 NA 15 5 NA 16 NA 16 15 10 NA 5 15 16 5 6 5 15 6 NA 3 NA 12 NA NA NA A81 31 637 3 16 18 13 17 8 NA A1A222N AN A1 8181 41 A1 7 5 15 NA 12 14 3 8 A1 3191 41 091 01 2006 6141 41 2N 41 .Eiety hr r at-pcfi and party-specific are there Evidently, ). i = AN 5N7N 4N 6N 2 1,...,N PLL N A1A1 A222 A1 AN A1 A52N nparties on 2006 RP PSD PS PR xetsurvey expert ) A9N 93161 41 384 j = 11 41 .Rozenas A. 1,...,J A A 0 4 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 ee e fctofpit with points cut-off of set dered lno n ei (2007 ( Lewis Albert and and Clinton Johnson in presented as aamdl.I ersnscoscasfid(ahrta etd irrhclmodel hierarchical nested) than (rather cross-classified (Zaslavsky represents It models. data and small typically preferred. is is survey model any in parties of ber h oe n( in model the and Johnson introduces to approach similar tive point, cut-off idiosyncratic ( an Albert have to respondent each low httemaueetsaehas suppose scale Formally, measurement perception. the latent his/her that of transformation affine an is in ( placement frame- differ Mckelvey the follow and might I Aldrich respondents that, by for that account work To is scale. measurement here the of issue interpretation key their The scale. measurement ordinal te hnarsodn ihhigh with respondent a than other cevy( Mckelvey respectively. respondents and parties across vary to about Thacker and (Huyse ture aetpreto fparty’s of perception latent ial,when Finally, from draw random a as perception to such distribution. positions each probability party that treat define we we Since distributions, issue-scale. probability given be some on party each of position Data Survey from Ambiguity Ideological Inferring 4 3 where opaepriso h ih ieo h cl eaiet epnetwt low with respondent a to relative scale the of side right the on parties place to all h sa admefcslna oe ope ihodnldt.For data. ordinal with coupled model linear random-effects usual the cl eeoeet.Arsodn ihalow a with respondent A heterogeneity. scale r enda asinpoaiiydsrbto functions: distribution probability Gaussian as defined are o xml,sc nepeaino epnetoiin a enue nters nlsslitera- ( analysis Poole risk and the Palfrey in used been has opinions respondent of interpretation such example, For h oe n( in model The lentvl,oecudseiyacmo oainadsaeprmtradal- and parameter scale and location common a specify could one Alternatively, ntescn tg,tersodn a orpr i/e bevdvleo some on value observed his/her report to has respondent the stage, second the In ic h oiysaei endol pt nafietasomto,Adihand Aldrich transformation, affine an to up only defined is space policy the Since j h oe eut ntesaigmdlb lrc n cevy(1977 Mckelvey and Aldrich by model scaling the in results model the , μ τ u i o drs how address not did but i 1999 and ugs ocntanteetmtsof estimates the constrain to suggest 1977) 2003 n lno n ei ( Lewis and Clinton and ) a ny2N only has 5)Ð(6) ψ σ i j .31.When 341). p. , 2 nlzdhwasmto fhtrgnosvrac fet inference affects variance heterogeneous of assumption how analyzed 1987) r xetseiclcto n cl aaeesacutn for accounting parameters scale and location expert-specific are 1 = a ese sa xeso fsm ieyue ordinal used widely some of extension an as seen be can )Ð(2) o each for 1 2004 y z ij ij ∗ .I otatt hs lentvs eallow we alternatives, these to contrast In ). ). N(M = ∼ j σ 3 m N ltomb respondent by platform hudb estimated. be should M c − μ j 1977 1 ψ iff epnetlvlprmtr.Gvnta h num- the that Given parameters. respondent-level h oe eebe h utpertrmodel multiple-rater the resembles model the , j onsadlet and points =−∞ 1) i ψ ,σ = c i 1999 iial,arsodn ihahigh a with respondent a Similarly, . m epnetlvlprmtr.I comparison, In parameters. respondent-level n sueta ahrsodn’ reported respondent’s each that assume and ) j 2 and 1 <ψ 2007 , and hp )adapidt xetdt by data expert to applied and 5) Chap. , i z σ ψ .Fran For ). j ∗ j c i + M = ed opaepriscoe oeach to closer parties place to tends C τ M =∞ o all for 1 ={ i ≤ slre oeparsimonious more a large, is i μ c c h aetplc positions policy latent The . M m .Let m ohv eoma n unit and mean zero have to +1 : on cl,ti alterna- this scale, point m i z and = ij ∗ 1,...,M eoeunobserved denote j ewudhave would we , σ j σ sand ’s } j 2 ea or- an be = τ i σ τ tends ψ i . 373 (2) for (1) ). i ’s 4 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 aigmsigesdpnetbt on both dependent missingness making oa el.Sneplc pc sdfie nyu oafietasomto,these transformation, affine to up information. only of defined loss in is result space not do policy constraints Since between well). intervals as equal do at fixed are points off Here, ond, at sabgos(high ambiguous is party smsigand missing is a ( if let Rubin convenience, and/or Thus, For and answer. ambiguous non-ignorable. NA Little very re- an of be observe (NA terminology to to the likely ambiguity. data perceived more in is ideological is missing one party the knowledgeable, the a not about is if in respondent that information patterns assume of I the source particular, exploit In additional to an extended as be sponses) can model The Data Missing for Model 3.1 model the identify to insufficient be ( to in out turn constraints These deviation. standard 374 e sue,oecnseiyheacia rosweesm at-ee parameters party-level sur- some cross-national where priors a hierarchical If specify distributions. can prior one specifying used, is by vey completed is model The Distributions Prior 3.2 r eslkl ob eotda As sipidb h isnns oe n(7). in model missingness the by implied as NA’s, as reported be Here, to likely less are hr oeprisado oersodnstn ohv oemsigvle than values missing more have to Parameter tend respondents others. some and/or parties some where epnetlvlscale respondent-level oie rt hti epneti o ihykoldebe(high knowledgeable highly not is respondent a if that first, Notice, .Tefloigrsrcin r moe instead: imposed are restrictions following The 1)Ð(2). z δ ij  sahprprmtretmtdi h oe.Fnly easm httecut- the that assume we Finally, model. the in estimated hyper-parameter a is sta r rw rmdsrbtoswt o tnaddvain(low deviation standard low with distributions from drawn are that ’s sasadr omldsrbto ucin eutn napoi oe.By model. probit a in resulting function, distribution normal standard a is r Pr ij i (r =1 N = α ij μ 1 tews.Temdlfrteosre aacnb rte as written be can data observed the for model The otherwise. 0 τ j = i esrshwmc isnns ntedt eed nthe on depends data the in missingness much how measures ψ = ∈[ y z 1) ij ij i and 0 σ c n at-ee ambiguity party-level and = = ∼ j 1 ,o oh h nwr ilehbthg aito.Sec- variation. high exhibit will answers the both, or ), −   N m Aif NA δ,c α ψ 0 i M + μ z i −1 =1 N j ij if α + 1 = c r + ψ ,w sueta h isn aaare data missing the that assume we 1987), σ ij m σ τ j i ψ 2 δ i j =

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 ,tesaeadsaeo h nes am itiuini xds htthe that so fixed is distribution gamma inverse the of scale and E shape the (10), In h itnebtenaytoctofpit.W set We points. cut-off two any between distance the parameter pc tms yafco fto oieta as that Notice two. of factor a by most at space hikg:teetmtso dooia miut nacountry a in adaptive mean ambiguity common induce the ideological priors wards of hierarchical The estimates estimates. the the shrinkage: on effect negligible a having f10( ihrvrac eue h pe fcnegnewtotafcigthe affecting without convergence of speed the reduces variance results). higher (a 100 of losszbeiisnrtclcto hfs aty h eeto oe parameters model selection the Lastly, shifts. α location idiosyncratic sizable allows oflo rnae omldsrbtoss that so distributions normal truncated follow to eylre epnetpae l ate nteopst xrmso h scale. the of extremes constraint opposite we the common, not on are parties alternatives these all of places both Since respondent a large, very interval. e frsodnsadnme fprisi country in parties of number and respondents of ber erzr,arsodn ol lc l ate ntesm on.Smlry if Similarly, point. same the on parties all place at would mass respondent the a zero, with near distribution degenerate a to collapses Let hyper-parameters. country-level upon depend Data Survey from Ambiguity Ideological Inferring h oe setmtduigMro hi ot al MM)mtosusing methods (MCMC) Smith Carlo and Monte (Gelfand approach Chain sampling Markov Gibbs using estimated is model The Estimation Parameter 4 positions ideological mean The taken. is survey the which t httems xrm at stouis(n-fho h cl)aa rmthe from away scale) set the we of parameters, remaining (one-fifth the units For two point. cut-off is largest party or extreme smallest 0 most have we the specification, that this ity under model: flexible highly yet and eset We r eldcmne nteltrtr Gla tal. et (Gelman literature the in well-documented are ubr(.. .) ilsapiron prior a yields 0.1), (e.g., number 0 (σ ,ec epnetcnepn rsrn h perceptual the shrink or expand can respondent each (12), in priors under Further, eaiet h cl fctofpit,tepirdsrbto of distribution prior the points, cut-off of scale the to Relative and jk 2 ) = α η 1 μ b δ r sue oflo omldsrbtoswt enadvariance and mean 0 with distributions normal follow to assumed are k = Setting . is 0 eutn navgebtpoe ro itiuin h hyper- The distribution. prior proper but vague a in resulting 100 e ohv o-omldsrbto ihma qa to equal mean with apriori distribution log-normal a have to set μ ln jk (δ) | a δ σ ∼ ∼ = jk 2 b ψ N N | ilsapir aineof variance priori a yields 4 k τ b b h ttsia datgso h irrhclshrinkage hierarchical the of advantages statistical The . ik ik k k ( (c 0,η ∼ ∼ ∼ ∼ i +1 N U Gamma Inv-Gamma μ − ) ( 1 2 1 0, b c , k i μ 2 1). ,v ihlrevrac;tu,tepir n up end priors the thus, variance; large with jk ,  ), (, δ ). ∈ (c a,(a 1990 1 − k δ,c .Let ). − ψ = v 2003 k i δ 1)b M 1,...,K → epciey Let respectively. = τ −1 b i N k hsipista,for that, implies This . k ). 2 ,rsligi identifiable in resulting 1, ,tedsrbto of distribution the 0, μ .Letting /2. k , + jk and δ) r assumed are ψ eoeacutyin country a denote J k , 4pirprobabil- prior .44 i k st h specified the to ’s r huknto- shrunken are eoetenum- the denote τ i  n( in N beasmall jk 13)also apriori denote ψ (13) (12) (11) (10) 375 i (9) (8) z ψ is ij i Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 parameters o bevddt n n o isn data: missing for likelihoods—one two one of and product data a observed is for which likelihood, data complete the yields This uino data of bution 1993 n( in .Sml aetperceptions latent Sample 1. Gibbs blocks: following the Specifically, the form. between distributional iterates known sampler have model the in parameters the aai eie yitgaigottemsigdt rmtecmlt aamdl so model, data complete the from data missing the that out integrating by derived is data fltn perceptions latent of stenme fN nwr o party for answers NA of number the country is in respondents of set the 376 θ .Gvnteltn variables latent the Given 2. ifr rmslcinmdl fmsigdt hr h itiuinof distribution the where data missing of on models selection from differs y obs eoealprmtr ftemdl h on itiuinof distribution joint The model. the of parameters all denote hsfcoiain ilsaptenmxuemdlwt hrdprmtr (Little parameters shared with model pattern-mixture a yields factorizations This sn rvosyseie ro itiuin,tefl odtoasfrms of most for conditionals full the distributions, prior specified previously Using ria data ordinal aeeso h model: the of rameters y ,tedsrbto of distribution the 7), obs sosre and observed is .I hsmdl hr sasto omnprmtr fetn ohtedistri- the both affecting parameters common of set a is there model, this In ). π( and y μ obs σ jk y z (,z r z, L(y, ijk mis , | and z y z | y jk y , h en ftepafrsaesmlda follows: as sampled are platforms the of means The . u o ntedt oe aaees h oe o h observed the for model The parameters. model data the on not but n isnns atr in pattern missingness and ijk r , ψ | ·∼N , π( θ) y ·∼ z n h ofcetvector coefficient the and ; ijk mis y = = θ) , N n isn aaindicators data missing and z π( smsigdt epcieyadlet and respectively data missing is ∝ , S r r y jk μ π( k × | eed ntevcoso miut n uncertainty and ambiguity of vectors the on depends z K =1 obs θ) jk /D z h eann ulcniinl ontdpn nthe on depend not do conditionals full remaining the , k y ijk K =1 ψ i | obs = ∈ r jk k ik N , odtoa nteosre data observed the on conditional i π( N , =1 h aepae party placed have who μ, ,σ jk + y k y j mis jk j 2 σ τ J =1 J obs =1 k ik j k , /D , τ ,ψ π(y ncountry in z , π(r , | y jk r ψ ik mis 2 , ijk ijk r r )π( σ μ, 1(c ntemdlo isnns given missingness of model the In . , jk ,z | 2 σ z α σ | jk ijk 1 r = , | 1(c σ − ,θ)π( ,ψ | τ r μ k , (α , ijk δ<μ .Let ). jk ψ ψ y ik ijk 0 epciey o rvt,let brevity, For respectively. , ,τ , r )π( ,α α α r

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 each .ln(δ) 4. ec a emntrduigtesadr ol,a,freape eee( Geweke example, for ( as, algo- Welch tools, and MCMC standard Heidelberger the of using Casella convergence monitored and be the can (Robert improve gence models to identified known weakly in is rithms and “sweeping” by tering .Tecefiinsi h isn aamodel, data missing the in coefficients The 5. nern dooia miut rmSre Data Survey from Ambiguity Ideological Inferring a odce n4 onre ihaot8prisad3 xet e onr on country per experts ( 30 Laver and parties and 8 Benoit about by with data countries expert 48 the in conducted to was model our apply We Performance Electoral and Ambiguity Ideological Application: 5 onr vrg of average country .Fl odtoasfor conditionals Full 3. oipeetteietfiaincntansi ( in constraints identification the implement To aebten3 n 0%. 50 and 30 between rate itiuinfrtehyper-parameter the for distribution ntvrac n oe on qa oln to equal bound lower and variance unit tnaddt umnainmto yAbr n hb(1993). Chib and Albert by method augmentation data standard where where (a ainprmtr r ape sfollows: as sampled are parameters cation atnsagrtmi mlyd h o-otro of log-posterior The employed. is algorithm Hastings values o-otro for log-posterior where (λ ψ − − ln ik 1) 1)σ ssmldfo h ettuctdnra itiuinwt mean with distribution normal left-truncated the from sampled is π sdvddb h onr vrg of average country the by divided is S S p ln σ jk 1 ijk σ jk jk 2 2 π = j (t (t) 2 j = = | −1 σ ,μ ,ψ a ψ, τ, μ, z, τ .5 ψ j ik −2 (α r ape rmteivregmawt shape with gamma inverse the from sampled are ) ik 2 Here, . ·∼ |· ik | τ + ,μ ,τ σ, μ, z, ψ ik jk (z 0 1/ 2 si utatd hspoeuei iia oheacia cen- hierarchical to similar is procedure This subtracted. is ’s (z N + ijk a iia form: similar a has σ ijk  λ α − 1983 . and 1 k − stetnn aaee htstt civ nacceptance an achieve to set that parameter tuning the is φ τ ,α ik jk τ j ∝− diagnostics. ) ψ ik )/ψ (z σ ∝−(N ) jk ijk + onthv eonzbefr,tu Metropolis- thus form, recognizable a have not do 2 jk σ ik + ) − (J i jk σ ∈ −2 and and i N jk ik −2 ψ ∈ k N /2 and ik r jk k b + ijk S μ D r k /2 jk + ijk jk ψ ln jk S sgmawt cl 2J scale with gamma is 1/2) ik = + ln 2 )/σ 2 ( p = max{μ = ijk p 3) ψ N α jk 2 ijk + ln ik 0 ln i 4 jk ∈ , ,atrec lc fiterations, of block each after ), + s iial,fo each from similarly, ’s; jk and ψ N ( σ + 1 jk jk jk ik ( 2 2 (z − 1 jk (z }−c σ + ijk − 2004 α − jk σ 2 r ijk σ 1 ijk j S ψ 2 σ − jk /η r r ape sn the using sampled are , −2 1 − jk ijk ) −2 to proportional is ik M 2 /ψ τ .37.Teconver- The 397). p. , μ ln ) + ik −1 ψ h epnetlo- respondent The . ln(1 ( ik S 2 )μ ik 2006 1 ψ ) jk h conditional The . μ − − ik 2 jk /2ψ jk − k  S p σ .Tesurvey The ). 2 ijk jk + . p − −2 λ /ψ ik 2 ijk ),  τ Proposal . n scale and c + ik ), ik i 1992 n rate and +1 ) b τ 2 ik k .The − )or 377 the c i , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 h esrmn ro ntecvrae ih edt nai neec bu the about inference invalid to of lead ( presence might model the covariates in Ignoring, parameters the values. regression in fixed error not error, measurement with the measured are that estimates aeit con h atta h covariates the that fact the account into take irrhclmdlaeue npaeof place in used are model estimates hierarchical posterior mean the setting, where h olwn oe sestimated: is model following the apemean sample respectively. ambiguity) ideological the of inverse and follows: as defined end-points the its with to scale related is elections. services party’s last public a the of whether in provision vote-share investigate and and to missing extremism taxation is of ideological of goal % issue Our the 9 experts. on about 1493 ambiguity with and entries parties, 10,603 364 contains values, dataset analyzed The average. 378 lorltdt at’ lcoa efrac.I aetemdlpoie correct provides model observe the should case one In ambiguity, performance. ideological electoral of party’s estimates a to related also (Alvarez Houweling behavior van voting to and related Tomz is ambiguity that found has literature vious ambiguity ideological the about information uncertainty. additional respondent and provide do data missing ideologi- of in measure evidence incorrect the an as yield ambiguity would serves cal deviation this standard Together sample of experts. the ambiguity the (1) of to that uncertainty related the is data and the positions of missingness party the that indicating 0.014 of tion in and tion, miut a a ugse ale) h orlto between correlation the earlier), suggested was (as deviation standard ambiguity sample Fur- the parties. political if of thermore, vote-shares and ambiguity ideological of estimates posterior av apesadr eito,wt orlto fol 6pret h posterior The parameter percent. mechanism data 36 missing only the of of correlation mean with deviation, standard sample naive lcoa efrac hudb low. be should performance electoral [ [ 1] 20] h oe n( in model The ntesre,teeprswr se opaepltclpriso h 0point 20 the on parties political place to asked were experts the survey, the In sn ietmaue fielgclabgiyadvtr’ucrany h pre- the uncertainty, voters’ and ambiguity ideological of measures direct Using h otro siae of estimates posterior The fe optn h otro itiuin of distributions posterior the computing After β at rmtsriigtxst nraepbi services. public increase to taxes raising promotes Party at rmtsctigpbi evcst u taxes. cut to services public cutting promotes Party 2 v ersn h feto dooia xrms n dooia rcso (the precision ideological and extremism ideological of effect the represent jk μ savt-hr fparty of vote-share a is steetmtdeprclcne fprypafrs h coefficients The platforms. party of center empirical estimated the is μ fteasmddt eeaigmdli valid is model generating data assumed the if ˆ n tnaddeviation standard and T(v setmtdi he etns ntefis etn,Iuethe use I setting, first the In settings. three in estimated is 17) jk ) = 2009 β 0 σ + rmtepooe oe r eydfeetfo the from different very are model proposed the from .Teeoe dooia miut hudas be also should ambiguity ideological Therefore, ). 17). β 1 | j μ jk ncountry in E σ ˆ μ (μ − ˆ npaeof place in and μ | y) σ ˆ |+ ( μ, sntavldmaueo ideological of measure valid a not is ˆ and σ ˆ α β oho h bv oesd not do models above the of Both . σ) k ˆ 1 σ 2 , jk 1 s025wt h tnaddevia- standard the with 0.245 is E T( and + sfralprisi h dataset, the in parties all for ’s μ (σ 1 ·) σ and | y) jk a ( saBxCxtransforma- Box-Cox a is E eainhpbtenthe between relationship (μ + σ eie rmtelatent the from derived n ht()tepatterns the (2) that and  | n( in y), jk σ ˆ , .I h second the In 17). E n h parties’ the and (σ | y)) .Rozenas A. r only are 1997 (17) β 1 ; Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

506 505 504 503 502 501 500 499 498 497 496 495 494 493 492 491 490 489 488 487 486 485 484 483 482 481 480 479 478 477 EDITOR’S476 475 474 473 472 471 470 469 468 467 PROOF466 465 464 463 462 461 π(μ h oun of columns the ugs htielgcleteimi aeyrwre yvtr n htideological that and would voters evidence by and rewarded ambiguity rarely theories is existing extremism as ideological counter-intuitive that suggest are results these of Both s tt h aa hsi copihdesl yadn tpi h ib sampling Gibbs the in coefficients step the a over adding priors by easily uniform accomplished Assuming is algorithm. This data. the to fit d c b a * 2 Table Data Survey from Ambiguity Ideological Inferring inmti o oe n( in model for matrix sign n scale and shape with tribution vrl to h oe seteeypo,a niae ylow by indicated as poor, extremely the is and levels model conventional the at of significant ambiguity fit are ideological overall coefficients its the and of party None a extremism. of or performance electoral the average between tionship the differences: of stark estimates and are sample estimates there the Evidently, sample model. If naive proposed the the from estimates use posterior that regressions frequentists two parentheses ihmean with AIC | Extremism: Intercept F( 1/( dooia rcso:0 precision: Ideological RMSE N R μ rqets ersininrn h esrmn ro ntecvrae.Sadr rosi the mean in Posterior errors Standard covariates. the in error measurement the ignoring regression Frequentist 5%hgetpseirdniyinterval density posterior highest % 95 error measurement the for accounting priors flat with regression Bayesian p< 2 2 jk — 2, eut ftetreaaye r eotdi Table in reported are analyses three the of Results hrfr,i h hr etn,telna ersinwt esrmn ro is error measurement with regression linear the setting, third the in Therefore, − | + 361) π( 0. y) μ σ 1, β jk | and dooia miut n lcoa efrac fparties of performance electoral and ambiguity Ideological ,s )( ** T(v (T does ( 2 p< π(σ X ) ∝  X ) 0. X fetvtr’behavior. voters’ affect − 1/s | ) 05, y) −1 representing X 2 epciey Finally, respectively. X  J/ oecnsample can —one *** β  ) T( (where 2 p<  T(v (T μ 0.1-e4 ( error measurement No 1124 ( −0. −2. 0.062 1. 364 ˆ 0. 0. 0. v . and 17) , 039 126 μ ) 115)( 03)( 612)( σ ˆ 0. 01 32 n oainematrix covariance and and 001 ) ** − μ J X and σ v  steoealnme fprisi h analysis) the in parties of number overall the is r sd hr sn ttsial rcal rela- tractable statistically no is there used, are β stevco fvt-hrs tec iteration, each At vote-shares. of vector the is σ )/ β 2. r elcdwt rwfo h posterior the from draw a with replaced are s 0.12 E −3. 1078 −0. 2. 24.5 1. 364 rmtemliait omldistribution normal multivariate the from 2 0. 0. 0. (μ 875 057 ssmldfo h nes am dis- gamma inverse the from sampled is 156) 042) 413) a 125 073 | y) *** , * *** E (σ s | 2 y) ( X is,ltu opr the compare us let First, 2.  X ) β −1 E 3. −0. −3. error measurement With [−4. [−0. [ 2. n ersinerror regression and (μ where , R 860 486, 049 823 2 | 393, 137, y) and , c 5. E 126] −3. 0. (σ X F 030] | 275] y) stede- the is statistics. d 379 b Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

552 551 550 549 548 547 546 545 544 543 542 541 540 539 538 537 536 535 534 533 532 531 530 529 528 527 526 525 524 523 EDITOR’S522 521 520 519 518 517 516 515 514 513 PROOF512 511 510 509 508 507 dooia miut.Teetmtso h oe ilb isdt h xetthat extent the underlying to to biased related be are will model missingness the data of of estimates patterns heavily The relies ambiguity. the model ideological the that First, assumption studied. further the be on should that issues several are rather there distributions probability of terms in allows parties and points. data of than survey positions the policy sample of hierar- represent structure naive latent informational to on the rich that focusing the fact exploits by the model from offered chical results is accuracy mea- greater what The informative deviations. than more into standard ambiguity and synthesized ideological accurate be of more can yielding sures approaches framework two missing inferential of these single patterns how a the demonstrated from study ambiguity ideological This approach of data. another degree the while methodologists—one infer respondents to political the attempted among by synthe- disagreement on used on focused previously focused model approach estimate approaches to proposed distinct allows The that two model data. sizing a survey evaluate and from construct ambiguity to was ideological study this of goal The Discussion 6 380 siae fielgclabgiywud(ehp roeul)la st believe sample to the us lead that erroneously) be—and (perhaps to would otherwise. two them ambiguity these expect that ideological show of should to circumvent estimates we merely to was us associated—as exercise allow are would this that of quantities positions analysis nature our policy the of Instead, their design problem. that communicate the this case in to nothing the means is be there fewer well and have might parties It political performance. electoral smaller the on ambiguity approach ideological this use more to is advisable is error it measurement practice. and standard the problem a with the as model of nature the the Fitting 1 given account. appropriate about into by Second, taken however). now increases is zero, credible precision cover error % ideological not 95 does of the interval effect and credible the lower % now 90 is (though extremism zero covers ideological of effect the First, and estimator. this naive expectations, per- the theoretical upon electoral the to worse measure closer the with that is suggests associated pattern empirical am- statistically this ideological both Since increasing formance. are model, extremism this and In biguity (AIC). Criterion Information Aikaike’s etheacia oe,temdlfi nrae rmtclya niae ylower by indicated as higher dramatically (RMSE), error increases squared fit mean model root the model, hierarchical tent ncnrs,i n sstemaue of measures the uses one if contrast, In lhuhtepooe ehdo nern dooia miut spromising, is ambiguity ideological inferring of method proposed the Although ti motn ont htw do we that note to important is It in error measurement the account into takes which model third the Finally, σ hw ulttvl iia eut,abi,wt oeipratdeviations. important some with albeit, results, similar qualitatively shows , σ eie rmteltn irrhclmdlde improve does model hierarchical latent the from derived not R li ohv on n aslefc of effect causal any found have to claim 2 μ and and F σ ttsisadsbtnilylower substantially and statistics eie rmtepooe la- proposed the from derived hntemeasurement the when /3 .Rozenas A. μ Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

598 597 596 595 594 593 592 591 590 589 588 587 586 585 584 583 582 581 580 579 578 577 576 575 574 573 572 571 570 569 EDITOR’S568 567 566 565 564 563 562 561 560 559 PROOF558 557 556 555 554 553 ishl ,MnfloaZ akwk ,Tk 19)Ps-omns at ytm.Cam- systems. party Post-communist (1999) G Toka R, York Markowski New Z, Springer, opinion. Mansfeldova modeling. expert data H, of Ordinal Kitschelt (1999) basis JH the Albert on VE, uncertainty Johnson under making tran- Decision initial (2004) an BH of Thacker presence the L, in Huyse control length 84(1):237Ð241 Rev run Sci Simulation Polit (1983) Am PD ambiguity. Welch candidate P, of Heidelberger strategy The (1990) A Glazer posterior calculating to approaches sampling-based of accuracy and the Chapman Evaluating edn. 2nd (1992) analysis, J data Geweke Bayesian (2003) DB Rubin densities. HB, marginal Stern JB, calculating Carlin to A, Gelman approaches Sampling-based (1990) AFM Smith AE, Gelfand post-communist in cleavages electoral of formation the Explaining (2000) S Whitefield G, Polit preferences. Evans agency and Rev characteristics, Sci agency opinion, Polit Expert Am (2007) data. DE call Lewis roll JD, of Clinton analysis statistical The (2004) D Rivers presi- S, the Jackman of JD, examination Clinton an ambiguity: issue of consequences electoral The (1983b) analysis. causal JE a Campbell candidates: presidential of positions issue the Econ in Public Ambiguity (1983a) J JE ambiguity. Campbell political and voting London Context-dependent Routledge, 30:709Ð728 (2008) democracies. Sci CH modern Polit Wilson in J S, policy Am Party Callander test. (2006) empirical M an Laver uncertainty: K, under Benoit voting Issue elections. house (1986) US L in Bartels positioning Candidate (2001) III Arbor CS Ann Stewart Press, Jr, Michigan 105(4):829Ð850 JM of Snyder Econ University S, J elections. Ansolabehere Q and presi- ambiguity. Information 1972 of (1997) and politics MR The 1968 Alvarez (1990) the A to America. Cukierman applications A, with in Alesina scaling parties of political method A of (1977) transformation R Mckelvey and J, origin Aldrich The parties?: Public Why dynamics. (1995) electoral for J implications Aldrich activists: party with model spatial Stat A Am (1983) J data. J response Aldrich polychotomous and binary of analysis Bayesian (1993) S Chib JH, Albert References strengths potential elicit model. also proposed mea- could the of direct benchmarks weaknesses plau- using as and empirical analyses ambiguity validation the ideological external evaluate of Second, could sures assumption. scales this single-point of standard sibility the Houweling on van and them re- Tomz some (e.g. place parties, scales placing interval in the where, use experiment spondents survey A violated. is assumption this Data Survey from Ambiguity Ideological Inferring rdeUiest rs,Cambridge Press, University bridge manuscript In 31:1109Ð1144 Res Oper sient. edn. 4th statistics, Bayesian (eds) AFM Smith AP, Oxford Dawid Clarendon, JO, Berger JM, Bernado In: moments. London Hall, 85(410):398Ð409 Assoc Stat Am wave, J first The Europe. eastern and central Berlin in Sigma, Elections Edition (ed) HD Klingemann In: democracies. 16(1):3Ð20 Anal 5(3):277Ð291 98:355Ð370 Behav Polit 1980. to 1968 from positions issue candidates’ dential 27(2):284Ð293 Sci Polit J Am 92(3Ð4):565Ð581 45(1):136Ð159 Sci Polit J Am 71(1):111Ð130 Rev Sci Polit Am elections. dential Chicago Press, Chicago of University 41(1):63Ð100 Choice 88(442):669Ð679 Assoc 2009 hl others while ) 381 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

644 643 642 641 640 639 638 637 636 635 634 633 632 631 630 629 628 627 626 625 624 623 622 621 620 619 618 617 616 615 EDITOR’S614 613 612 611 610 609 608 607 606 605 PROOF604 603 602 601 600 599 alvk 20)Heacia aeinmdln.I:PesJ e)Sbetv n objective and Subjective (ed) JS Press In: modeling. bayesian Hierarchical Polit Am (2003) Ambiguity. A candidate Zaslavsky of implications electoral The (2009) RP Houweling Sci van Polit M, Am Tomz competition. electoral and uncertainty Building ambiguity: (eds) of T strategy The Scully (1972) S, KA Mainwaring Shepsle In: York Chile. New in Springer, politics ed. party 2nd Reconstituting methods, statistical (1995) Carlo T Monte Scully (2004) G Casella CP, behavior. Robert voting and ideology, 70(3):742Ð752 information, Rev between Sci relationship Polit The Am (1987) KT ambiguity. Poole political TR, of Sci Palfrey theory Polit The Am (1976) States. United BI Page the in realignment partisan and for Activists Carlo (2003) Monte N Schofield chain G, Markov Miller via estimation point ideal Dynamic (2002) positioning KM party Quinn on AD, data Martin Crossvalidating (2007) R Bakker MR, (eds) Steenbergen T L, Scully Hooghe S, G, Mainwaring Marks In: democracy. feckless York Assoc New parties, Wiley, Stat weak data. Brazil: missing Am (1995) with J S analysis Statistical Mainwaring data. (1987) D incomplete Rubin RJA, multivariate Little for models as Pattern-mixture words using (1993) texts political RJA from Little positions policy Extracting (2003) J Garry K, Benoit M, Laver 382 aeinsaitc.Picpe,mdl,adapiain.Wly e York New Wiley, applications. and models, Principles, statistics. bayesian 103(1):83Ð98 Rev Sci 66(2):555Ð568 Rev Stanford Press, University Stanford Institutions. Democratic 31(3):511Ð539 Rev Sci Polit Am 97(2):245Ð260 Rev 10(2):134Ð153 Anal Polit 1953Ð1999. court, supreme U.S. the 26(1):23Ð38 Stud Elect integration. European on Stanford Press, University Stanford institutions. democratic Building 88(421):125Ð134 97(2):311Ð331 Rev Sci Polit Am data. .Rozenas A. Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 EDITOR’S16 15 14 13 12 11 10 PROOF 9 8 7 6 5 4 3 2 1 fsca hie oiia cnm,addmcay ehswitnsvrlbooks, several written theory has the in He democracy. topics on and doctorat economy, working including a currently political is and choice, He University social eco- Caen. Liverpool of and of from government University the both LittD from in a Wil- d’etat Ph.D.’s University, the has Departments Essex currently He the from is Science. in nomics Professor Political He and and 2012. Economy, Economics until Political of of 1992 Professor from Taussig R. Louis liam Saint in University ington Schofield Norman Biographies Editors’ 1 Biographies DOI (eds.), al. et Schofield N. Democracy a ulse ySrne n20.I 06h published he 2006 In 2003. in Change Springer Political by published was Choice, cial book, His 2011). Voting Springer, and Caballero, Democracy Institutions, of Economy Political h nvriyo io n ercie Pi nCneprnosPolitical Contemporaneous in from MPhil Honors a with received Economics he in and B.Sc. Vigo, a and of Ph.D. University his the received He Spain. Vigo, to elected was and 2005. 2002, in in Sciences theory and Arts political of to Academy recipient contributions American was for He the Na- 2010. prize a and Riker In was 2009 summers Turin. the and in University of in Yale Institution Hoover at ICER at the Professor at at Leitner professor Fellow fellow visiting tional distinguished visiting the Fulbright was 2008, he the and pol- 2008 2007 was electoral Spring on in he one and, 2003 recently University In most Humboldt awards, change. NSF regime of and number itics a of book, recipient a the completed been just has He (Munich). Chaos Oldenbourg with 2009 in by followed 2008, in Routledge book, His Press. University bridge ozl Caballero Gonzalo schofi[email protected] 10.1007/978-3-642-35239-3 c-uhrdwt ai alg,pbihdb pigri 01.H has He 2011). in Springer by published Gallego, Maria with (co-authored utpryGovernment Multiparty Srne,18) n dtdo oeie e oue,including volumes, ten co-edited or edited and 1985), (Springer, swl as well as a h ietro h etri oiia cnm tWash- at Economy Political in Center the of Director the was dacsi oiia Economy Political in Advances sa soit rfso fEooisa h nvriyof University the at Economics of Professor Associate an is utpryDemocracy Multiparty pigrVra elnHiebr 2013 Heidelberg Berlin Springer-Verlag © , h oiia cnm fDmcayadTyranny and Democracy of Economy Political The wt ihe ae,i 1990), in Laver, Michael (with h pta oe fPolitics of Model Spatial The ahmtclMtosi cnmc n So- and Economics in Methods Mathematical , wt tiSnd ohwt Cam- with both Sened) Itai (with c-dtdwt Gonzalo with (co-edited , oilCoc and Choice Social a ulse by published was edrhpor Leadership rhtcsof Architects The 383 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 EDITOR’S62 61 60 59 58 57 56 55 54 53 PROOF52 51 50 49 48 47 nvriyo aionai a ig n h etro eorc,Development Democracy, on Center the the at and She Studies Diego America. U.S.-Mexican San the Latin for in in in Center California labor of capacity the University from fiscal colonial positions centralized of con- fellowship of origins received and origins indigenous has change the the on institutional on and behind centered Americas, mechanisms has develop- the work political on and Her economic tinuity. particular of in economy and political the ment, on focuses research Her Choice. policy Science parties’ Political of of analyses Journal empirical encompasses in and which appeared competition research, has party programmes, His of States. models United spatial the both and Europe Western represen- political in and strategies, tation election parties’ behavior, voting studies He Davis. Adams James Contributors of Biographies 2 in- Roma journals, University various in the analysis at institutional Univer- and on Charles papers Bratislava cluding many at in published Institutions” University in has He Comenius California of Tre. at of “Economics University Prague, on the in Madrid, course in and sity a Sciences Barbara Social taught Santa the He in in Berkeley. Univer- Study California Visiting Washington Advanced of a for Sciences, University Center been Social the the has Institutional Louis, He New St. in for Compostela. sity Center de the Santiago at of Scholar University the from Studies 384 Management, ora fPltclEconomy, Political of Journal conaiiy ei loasueto oenTrihpltc,adhswitnon written has and politics, organizations. Turkish party modern and of democratic institutions student political and Turkish a instability, also legislative is corruption, in- He political accountability. phenomena, limited political not contemporary of but and variety cluding institutions a game political on if both organizations impact employs party the political study research to His modelling (Madrid). econometric and Institute theoretic March Juan Sciences, Social Reces- Great the on volume a co-edited and books sion. several written has He Política. insSociológicas ciones u aiaArias Marina Luz [email protected] [email protected] Kselman Daniel [email protected] cnmcModelling Economic aiePolicy, Marine sPoesro oiia cec tteUiest fClfri at California of University the at Science Political of Professor is , saRsac elwa h etrfrAvne tde nthe in Studies Advanced for Center the at Fellow Research a is eit eEtdo Políticos Estudios de Revista eevdhrP..i cnmc rmSafr University. Stanford from Economics in Ph.D. her received , , h ora fPolitics, of Journal the h mrcnPltclSineReview, Science Political American the ora fIsiuinlEconomics Institutional of Journal Environmetrics, oíiayGobierno, y Política lTieteEconómico, Trimestre El and eit saoad Investiga- de Española Revista and Analysis, Political eit saoad Ciencia de Española Revista , ca Coastal & Ocean h American the Biographies Brazilian Public Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100

EDITOR’S 99 PROOF98 97 96 95 94 93 at upr nGetBian h ntdSae n aaahsbe upre by supported been has Canada and States United the Britain, Great in support party of editor Ad- is also and He Studies Dallas, Essex. Electoral at of University Texas Government, of of University Department Professor, Sciences, junct Policy and Political Economic, of voting. in U.N. bargaining and terrorism, costs, of audience diffusion domestic the war, include civil interests research His Maryland. of (Argentina), Studies c iyadcmltdhrudrrdaesuisi cnmc tIA,Mexico ITAM, at Economics in studies Mex- in undergraduate raised City. her and completed born was and Marina City Luz University. ico Stanford at Law of Rule the and Contributors of Biographies 2 tics cec soito,hsbe ulse nUS,Erpa,adLtnAmerican Latin and European, the U.S., as in such published Political journals American been the has of sections Association, from Politics awards Science Comparative received the has and which repre- Representation research, political the His of issues Congresses. on and focuses elections research His sentation, Maryland. of University the at ence gerrymandering. partisan about book a on working America for “ Bad Univer- book are the His Elections from 1997. Competitive Ph.D. in his Irvine received California, representa- He of Congress. elections, sity U.S. around the revolve and interests parties research political tion, and teaching His Dallas. at Política as such journals in Económico topics sarrollo different on papers several published has He Valencia). co-edited also He Costas. Antón with the program of book research the analysis the published cia. of the recently development He in the Politics. worked Cost and Transaction actively economy of world he the years, in recent crisis In financial (Spain). Vigo of versity aodD Clarke D. Harold [email protected] Kiyoung Chang Kiyoung [email protected] oéCro Arias Carlos Xosé [email protected] ret Calvo Ernesto [email protected] Brunell L. Thomas [email protected] , Ail acln,21) jointly 2011), Barcelona, (Ariel, tormenta, la de después mercados y Políticas ol Politics World , and lcoa Studies Electoral eit eEooí Institucional Economía de Revista pna Publica Opiniao lresrsac neetrlcoc n h oiia cnm of economy political the and choice electoral on research Clarke’s . , P..Nrhetr)i nAscaePoesro oiia Sci- Political of Professor Associate an is Northwestern) (Ph.D. , h rts ora fPltclScience, Political of Journal British The saP..Sueti oenetadPltc tteUniversity the at Politics and Government in Student Ph.D. a is eit eEooí Aplicada, Economía de Revista mrcnJunlo oiia Science Political of Journal American P..Dk nvriy sAhe mt rfso,School Professor, Smith Ashbel is University) Duke (Ph.D. sPoesro cnmc n cnmcPlc tteUni- the at Policy Economic and Economics of Professor is sapoesro oiia cec tteUiest fTexas of University the at Science Political of professor a is , oíiayGobierno y Política Bai) n the and (Brazil), raiaind oiro mercados y gobiernos de Organización a ulse n20.H scurrently is He 2008. in published was ” ersnainadRdsrcig Why Redistricting: and Representation mn others. among , eit eCecaPolitica Ciencia de Revista (Mexico), eit saoad Ciencia de Española Revista earloEconómico Desarrollo atred aarrogan- la de torre La oprtv Political Comparative , h ora fPoli- of Journal the (Chile). ,(PUV, De- 385 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166 165 164 163 162 161 160 159 158 157 156 155 EDITOR’S154 153 152 151 150 149 148 147 146 145 PROOF144 143 142 141 140 139 States 2009), Press, University are books recent Science,the Political of Journal iesre nlssa h se umrSho nDt olcinadAnalysis and Collection Data in School. School Spring University Summer Concordia Essex the advanced the and public teaches at regularly on analysis He focus methods. series research interests time and teaching elections Clarke’s and voting 2004). opinion, Press, University Oxford ford: the as such journals in Science Political published of articles Canada Journal numerous in ican of Support author Political the 2008 is the He and Study. Study, the the Germany and include in projects Study Support research Election Political current 2009 Other British Study. 2009/10 America in the Support He for Political (Canada). investigator 2008 Council principal Research a Humanities and currently Coun- Sciences is Research Social Social the and and Economics (U.K.), the cil (U.S.), Foundation Science National the 386 n neooi hoyjunl uhas such journals Behavior Economic theory the economic as in such Science journals and Political science of pref- political policy Journal leading of in ican model spatial published the been and has elections, His erences and economics. voting public formation, are and party specialization choice on of work social fields theory, His political 2007. formal in economy, Technology Ph.D. political of his obtained Institute He California University. York the New at from Economics of Department the to ated financing the war in on government working of currently is size Assistant and transitions. the an political politics, on and currently Brazilian worked is on has He states, He Pennsylvania. American University. of was Bristol University he at the Ph.D. Professor his at during student years two visiting For University. a Bocconi from Ph.D. his received Louis. Saint in versity ietro h C etrfrteSuyo eorc i eerhdaswith is deals He research rules. electoral His and Democracy representation of of theories Study and and Science rights the voting Political for as of Center such Professor topics UCI and the (UCI), of Irvine at Director California of University at others. among eu Demirkaya Betul [email protected] o .Eguia X. Jon [email protected] Magalhães De Leandro [email protected] enr Grofman Bernard [email protected] Trno nvriyo oot rs) and Press), Toronto of University (Toronto: efrac oiisadteBiihVoter British the and Politics Performance sa sitn rfso nteDprmn fPltc n affili- and Politics of Department the in professor assistant an is , cnmcTheory Economic sagaut tdn npltclsinea ahntnUni- Washington at science political in student graduate a is steJc .Plao Be onain noe Chair Endowed Foundation) (Bren Peltason W. Jack the is aigPltclCocs oigi aaaadteUnited the and Canada in Voting Choices: Political Making tde cnmc tteUiest fBaii and Brasilia of University the at Economics studied , ora fPltc n oiia Analysis Political and Politics of Journal h mrcnPltclSineReview , Science Political American the n the and n the and oilCoc n Welfare, and Choice Social urel ora fPltclScience Political of Journal Quarterly ora fMteaia Economics Mathematical of Journal oiia hiei Britain in Choice Political Cmrde Cambridge (Cambridge: ae and Games h British the Biographies .Hismost Amer- Amer- (Ox- Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

230 229 228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208 207 206 205 204 203 202 201 EDITOR’S200 199 198 197 196 195 194 193 192 191 PROOF190 189 188 187 186 185 98.H a ulse nanme fjunl,icuigthe including journals, of number Review,the a Science in published has He 1988). Louis. Saint in versity Univer- Duke from is Science Political in Ph.D. sity. His College. Dartmouth at Law and the in papers published has Politics and She Opinion Louis. Saint in sity Science, Political n lcoa oiis i oko lblzto,eooi oig n at pol- party and voting, economic the globalization, in on appears globalization work itics economic His on manuscript politics. book electoral industrial a and advanced completing of currently economy is political He the societies. on focus interests research His versity. in Doctorate Honorary Copenhagen. an of has University and Fellow the Sciences, a from and is Science He Arts Political Europe. of and Academy U.S. visiting American the a the in chap- been universities of has book dozen He a and Review. nearly Science articles at Political member co-authored American faculty and the of in single co-editor ten research and including 250 ters, Press), with University books; Cambridge other by 21 (published books 4 of co-author Contributors of Biographies 2 eateto ulcMngmn n oennea eplnUiest.She University. Zeppelin at Governance and Management Public of Department a and University Yale at Washington. Professor of Visiting University a the been at scholar has visiting He Science. Political of nal Grofman), Bernard and and Voting Adams of James Theory with fied 2005; Press of University author bridge the in is University. Yale particularly He modelling, from science. statistical political mathematics and mathematical in involves Ph.D. research current his His received He University. Wilkes at eSogJeon JeeSeong [email protected] aulMrilIII Merrill Samuel [email protected] McCalister Kevin [email protected] Lacy Dean [email protected] ioh Hellwig Timothy [email protected] aj Michalak Katja [email protected] aigMliaddt lcin oeDemocratic More Elections Multicandidate Making sPoesro oenetadDrco ftePormi Politics in Program the of Director and Government of Professor is h ora fPolitics of Journal The sagaut tdn npltclsine tWsigo Univer- Washington at science, political in student graduate a is mrcnJunlo oiia Science, Political of Journal American and , srae fpltclsineadpbi diitaina the at administration public and science political of reader is mrcnJunlo oiia Science Political of Journal American sa soit rfso fpltclsinea nin Uni- Indiana at science political of professor associate an is CmrdeUiest rs 05 ihBradGrofman), Bernard with 2005; Press University (Cambridge sagaut tdn npltclsine tWsigo Uni- Washington at science, political in student graduate a is sPoesrEeiu fMteaisadCmue Science Computer and Mathematics of Emeritus Professor is lcoa Studies Electoral nfidTer fPryCompetition Party of Theory Unified A mn te outlets. other among , . ora fEetos Public Elections, of Journal PictnUiest Press University (Princeton n the and , h rts ora of Journal British The mrcnPolitical American rts Jour- British AUni- (Cam- 387 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

276 275 274 273 272 271 270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250 249 248 247 EDITOR’S246 245 244 243 242 241 240 239 238 237 PROOF236 235 234 233 232 231 and Science, Political of Journal research political Her psychology. bureaucracy, political Europe’. ‘Civil governance, and Eastern democratization, political on in institutions, politics, thesis Governance comparative a of include Quality with interests University the field and State the Reform Ohio in Service psychol- from (Ph.D.) political doctorate Institutions’ on her ‘Comparative focus received a of She with psychology. psychology organizational East- in and in degree ogy democratization Master’s on a focus and a with Europe science ern political in degree Master’s a holds 388 havior o nvriyadAssatPoesro oeneta avr nvriy His University. Harvard at Government Mel- of Carnegie was Professor at book he Assistant Sciences University, Decision and Washington University and to Economy lon coming Political to of Prior Professor 2001. Assistant in Technology of of stitute boards editorial Politics the retical the on of serves Editor rently Associate an as ing Modelling, puter iia Science litical Univer- Duke at Studies Security Asian Pro- in systems. Program international the of anarchic sity. Director in is Brehm, politics Used Niou alliance John fessor Systems and with China Control China,” in Communist the government the Politics of and Theoretical Analysis Union of Japan Comparative Ex-Soviet from the A Evidence in Self-Policing: Parties: vs. Large Cox, trol publi- for G. Vote recent with Taiwan,” Non-Transferable his and for politics, Single Scheme Asian the Self-Assessment Tan, Yet-sen’s East der G. Sun of with Dr. Policy,” field of Land the Analysis “An In include: 1994. cations June Ordeshook, Quarterly, P.C. with ies coauthored coau- Systems,” Debate,” International rela- Realist-Neoliberal chic international The Ordeshook, of P.C. Great: field Tastes with of thored the Filling, in co-author “Less publications include: the His tions Economy, is 1989). Political Press, He University Relations, Politics. bridge International Asian Theory, East Formal and in specializes ence, fpltclisiuin,a ahntnUiest nSitLus i okhas work His Louis. models Saint formal in and University theory in Washington game published fed- on at been the courses institutions, and graduate political Congress as of on well courses as bureaucracy undergraduate eral teaches regularly He stitutions. mro Niou Emerson [email protected] onPatty John [email protected] erigWieGvrig xets n conaiiyi h Executive the in Accountability and Expertise Governing: While Learning , ora fPolitics of Journal , ercie i hD nSca cecsfo h aionaIn- California the from Sciences Social in Ph.D. his received He . safra oiia hoitwoersac oue npltclin- political on focuses research whose theorist political formal a is cnmc Politics, & Economics oiis hlspy Economics & Philosophy, Politics, P.. .o ea tAsi,18) rfso fPltclSci- Political of Professor 1987), Austin, at Texas of U. (Ph.D., mrcnJunlo oiia Science Political of Journal American 96 i urn rjcsicuesuiso oa self- local of studies include projects current His 1996. , , ora fTertclPolitics Theoretical of Journal oprtv Politics Comparative ulcChoice Public ol Politics World nadto oserv- to addition In Welfare. & Choice Social mrcnJunlo oiia Science Political of Journal American lcoa Studies Electoral oiia Analysis Political aur 94ad“line nAnar- in “Alliances and 1994 January , eebr19;“etBnssun- Bonuses “Seat 1993; December , aur 94 n Plc Pa- “Police and 1994; January , h aac fPower,(Cam- of Balance The , ulcChoice Public , ae cnmcBe- Economic & Games , , ahmtcl&Com- & Mathematical nulRve fPo- of Review Annual and nentoa Stud- International ora fTheo- of Journal , Biographies Quarterly ecur- he , Journal Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

322 321 320 319 318 317 316 315 314 313 312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 EDITOR’S292 291 290 289 288 287 286 285 284 283 PROOF282 281 280 279 278 277 n2013. in Branch Contributors of Biographies 2 a frsac r olciedcso aigadeooi n oiia institu- political and economic and in ar- making His Ireland decision Mainz. of collective of tions. University are University the and research National and of Bochum Bulgaria the eas School in Andrews, Graduate University American St the then the at of Galway, Since positions University 1995. research the in and Germany Dortmund, teaching in held Bochum has Ruhr-University he from Ph.D. his ceived ore npstv oiia hoy tWsigo nvriyi an oi.Her graduate Louis. as the Saint in Politics in well published University as Washington been modelling at has theory, work agent-based undergradu- political and positive teaches systems on regularly courses electoral She on institutions. courses political ate and theory choice social inPlc ic 1945 Since Policy was eign and Politics 2005, International since in Academy British compara- of the and author of U.K. the the Fellow is of of a He aspects editor 1975. been various joint since has on politics He articles taught politics. and has tive books he numerous where Essex, of of author University the at search) Department the joining be will University. he York 2013, Eco- New of at and fall Politics Management the of of Bayesian In University Lithuania. and ISM Vilnius, the elections in at nomics non-democratic teaching is and he democratic Currently, statistics. of game-theoretic models on focuses statistical research and His University. Duke from both Sciences, cision Univer- Harvard at (2005Ð2009). Government sity of Professor Mellon Carnegie Assistant was at and she Sciences and Previously, (2003Ð2005) Decision 1999. University 2003 and in Economy in Berkeley Political Technology at of California of Professor of Assistant Institute University California the from the B.A. from her Sciences Social in Ph.D. hiei Britain in Choice Choice, lic h rts lcinSuy swl spltclpriiain lcinforecasting; election participation; political as well as Study, Election British the eadPech Gerald [email protected] lzbt agePenn Maggie Elizabeth [email protected] ai Sanders David [email protected] Rozenas Arturas [email protected] , wt enGilad ilapa ihteUiest fCiaoPress Chicago of University the with appear will Gaillmard) Sean (with ora fTertclPolitics Theoretical of Journal oilCoc Welfare & Choice Social atrso oiia Instability Political of Patterns sascaepoesro cnmc tKMPUiest.H re- He University. KIMEP at economics of professor associate is (2004); rts ora fPltclScience Political of Journal British saPoesro oenetadPoVc-hnelr(Re- Pro-Vice-Chancellor and Government of Professor a is od hD nPltclSineadMS nSaitcladDe- and Statistical in M.S. and Science Political in Ph.D. holds 19) n oato of co-author and (1990), (1986); efrac Politics Performance safra oiia hoitwoewr oue on focuses work whose theorist political formal a is mrcnJunlo oiia Science Political of Journal American oiga mie idn oe rts For- British Role: a Finding Empire; an Losing and , , ahmtcl&Cmue Modelling, Computer & Mathematical rfso enrcie her received Penn Professor Complexity. (1981); 20) i urn eerhi on is research current His (2009). nMessage On amkn n Co-operation and Lawmaking rm19 o20.H is He 2008. to 1990 from (1998); , ora of Journal Political Pub- 389 Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

368 367 366 365 364 363 362 361 360 359 358 357 356 355 354 353 352 351 350 349 348 347 346 345 344 343 342 341 340 339 EDITOR’S338 337 336 335 334 333 332 331 330 329 PROOF328 327 326 325 324 323 est nSitLouis. Saint in versity citizen- European assessing and measuring sector; public U.K. ship. the of politics the 390 hsrsac a enpeetdi eea nvriisi h United the in universities several Canada. and in America presented Latin been Europe, has States, research jour- This academic Gobierno. respected y in appeared America the have Latin as publications such are nals His interest States. of United in regions cam- the elections, main (CIDE) parties, and His Teaching political reforms. study and congressional as- to Research and theory an paigns, formal Economics currently uses is for Serra at he Dr. Center City. and research Mexico the 2010, post-doctoral until at do Oxford professor to of sistant fellowship University the a at won College He Nuffield 2007. in University Harvard Department the in student graduate a University. currently Harvard is Government, He of Louis. Saint in University ton eho,CmrdeUiest rs,20)adi o okn on working now Worldwide Legislatures is in and Change 2004) System Party Press, University Cambridge deshook, Institutions sta- Federal system Self-Sustainable party in co-authored of of She publishes determinants democracies. She and in conflict. federalism change societal and of bility of political economy amelioration political to of to incentives fields lead on the can is those focus how research and shaped long-term elites are Her they choices. as institutional mechanisms political by studies She USA. University, Binghamton hois n fhrkyitrssi efr tt ntttos iharcn ou on focus recent a with underlying institutions, state the structures. welfare evaluate care is to health interests analysis key her experimental of and One uses assess decisions, theories. to research these analysis her policy of end, decisions, effect make that people To the how work. address don’t to Her work—or theory the Finland. choice things address social in why to Tampere and science how of political of and University issues economics the from at approaches combines Humanities research and Sciences Social of vnSchnidman Evan [email protected] Schnakenberg Keith [email protected] laShvetsova Olga [email protected] Serra Gilles [email protected] ar Sieberg Katri [email protected] katri.sieberg@uta.fi rdae ihaP..i oiia cnm n oenetfrom Government and Economy Political in Ph.D. a with graduated ora fPolitics of Journal steEkoPoesro ot mrcnSuisi h School the in Studies American North of Professor Erkko the is sa soit rfso fpltclsineadeooisat economics and science political of professor associate an is bandhsA n Mi oiia cnm rmWashing- from economy political in AM and AB his obtained saP..sueti oiia cec tWsigo Uni- Washington at science political in student Ph.D. a is ,the ora fTertclPolitics Theoretical of Journal wt ihi iipvadPtrC Or- C. Peter and Filippov Mikhail (with wt ao Mershon). Carol (with einn eeaim Theory A Federalism: Designing and , ate and Parties Biographies Política Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

414 413 412 411 410 409 408 407 406 405 404 403 402 401 400 399 398 397 396 395 394 393 392 391 390 389 388 387 386 385 EDITOR’S384 383 382 381 380 379 378 377 376 375 PROOF374 373 372 371 370 369 tl.H a losre sivtdeio n uhrfraseilisepublished issue special a for institutions. author and economics and editor invited Choice as the Public in served the also Bologna, at has of scholar University He the visiting Italy. and University, a Leeds been University, Mason has George He Center, Público. Gasto y Presupuesto Economics , disciplinary ences Macroeconomy the isi oiia cnm—aaa ra rti n h ntdStates United the and Perspective Britain ; Great in Economy—Canada, Campaign Political Election in General sies British 1987 The Change: ciimi Britain; in Activism Membership; Party vative Membership; Party of Politics Grassroots—The a otiue osvrleie oue nteetpc n a ulse nvar- in published has the and including topics journals these He on academic federalism. volumes Spanish ious edited new tasks several the policy to on of contributed emphasis organization an has the with provision, as services well public as and analysis epis- economic and methodological in in issues are temological interests research His Spain. Valencia, of University are books recent Britain Her in journals. Choice research major other and Analysis, Political Review, Science n eea Studies Federal and Voter British inadteEooi n oilRsac oni UK) e rilshv been have articles Her (U.K.). Founda- Council Science in Research National Social published and the Economic by the funded and and Elec- tion 2009Ð2010), British 2005, the of (2001, co-investigator Study been tion has She methodology. research and havior, Foundation. Science National the of Sciences tics of Editor iy h a enEio of Univer- Editor the Rutgers been on and has and University UTD, She State sity. at and School Institute Studies Polytechnic the Virginia Graduate in Texas of Director Dean, faculties of Program Acting University Vice-Dean, Science the Executive Political been at and Director, has Sciences in also Policy Elections, She and and Dallas. Political Research at Economic, Opinion for of Center School Survey the The of Co-Director and sor, igahe fContributors of Biographies 2 alWhitely Paul [email protected] ennoToboso Fernando [email protected] h odcsrsac n ece nteaeso lcoa oiis oiia be- political politics, electoral of areas the in teaches and research conducts She aineC Stewart C. Marianne n oiia cec rga ietri h iiino oiladEconomic and Social of Division the in Director Program Science Political and , ; n a eetyc-dtdabo on book a co-edited recently has and Económico , Análisis de Revista Chilean rsuefrtePo—h oet ob n Policy-Making and Lobby Poverty Poor—The the for Pressure nentoa tde Quarterly Studies International CmrdeUiest rs,2009). Press, University (Cambridge h mrcnJunlo oiia Science Political of Journal American The h rts ora fPltclScience Political of Journal British The Ofr nvriyPes 04 and 2004) Press, University (Oxford steato of author the is n oato of co-author and , e aorsGasot:TeTasomto fteLabour the of Transformation The Grassroots: Labour’s New uoenJunlo a n Economics and Law of Journal European sSno etrra h eateto ple Economics, Applied of Department the at Lecturer Senior is ainaPbiaEpñl,E rmsr Económico Trimestre El Española, Pública Hacienda ihItniyPriiainTeDnmc fteParty the of Dynamics Participation—The Intensity High P..PltclSine ueUiest)i Profes- a is University) Duke Science, Political (Ph.D. h mrcnJunlo oiia Science Political of Journal American The h aorPryi Crisis in Party Labour The dacdSaitclMtosi h oilSci- Social the in Methods Statistical Advanced abig ora fEconomics of Journal Cambridge sitn dtrof Editor Assistant , reBus h oiiso Conser- of Politics The Blues: True efrac oiisadthe and Politics Performance , , h ora fPolitics of Journal The h mrcnPolitical American The ; h ora fPoli- of Journal The oiia oto of Control Political , ora fInter- of Journal ; o Voters How Associate , Controver- ; , Labour’s Regional Political and , 391 , Book ID:306518_1_En, Date:2013-02-19, ProofNo: 1,UNCORRECTED PROOF

460 459 458 457 456 455 454 453 452 451 450 449 448 447 446 445 444 443 442 441 440 439 438 437 436 435 434 433 432 431 EDITOR’S430 429 428 427 426 425 424 423 422 421 PROOF420 419 418 417 416 415 iiainadDemocracy and ticipation Membership Party 392 nDmcayadPriiain 9820 n odrco fteER funded ESRC the of Programme co-director Study. Research and Election He Council 1998Ð2003 British science. Research Participation, Social political and and Democracy in Economic on methods the quantitative of and Director pol- was capital British social parties; political citizenship, behavior; electoral itics, economy; political on: articles fifty roots [email protected] ; efrac oiisadteBiihVoter British the and Politics Performance ; oiia hiei rti:Ctznhpi rti:Vle,Par- Values, Britain: in Citizenship Britain: in Choice Political ; hr oc oiis iea eort tteGrass- the at Democrats Liberal Politics: Force Third uhro oato fmr than more of co-author or Author . Biographies