Research project proposal

REGIONAL MANIFESTOS PROJECT: EXTENDING THE MANIFESTO DATASET TO SUB-NATIONAL ELECTIONS

(AN APPLICATION TO THE SPANISH REGIONAL ELECTIONS)

Sonia Alonso (WZB) Braulio Gómez (IESAA)

October 2007

I. Background information: the already existing Manifesto dataset.

1.1. Introduction

The Manifesto research dates back to 1979, when the Manifesto Research Group of the European Consortium for Political Research was founded by an international group of political scientists. The MRG’s aim was to “measure party policy positions over time and between countries by means of content analysis of elections programs” (Volkens, 2002: 462). The work of the MRG was expanded in 1989 to the Comparative Manifestos Project, directed by Hans-Dieter Klingemann at the Wissenschaftszentrum Berlin für Sozialforschung (WZB). Since then, the WZB provided resources for updating and expanding the manifesto data collection, first to all OECD countries and then to 24 Central and Eastern European countries.

The major task of the Manifesto research group was to develop a classification scheme that would parsimoniously cover the whole content of election programs in comparative as well as longitudinal perspective.

1.2. The Manifesto dataset: a brief description

Why election programs? According to Klingemann, “election programs assess the importance of current political problems, specify the party’s position on them, and inform the electorate about the course of action the party will pursue when elected. With a few exceptions all election programs (…) have been legitimized by party conventions. Thus, they represent the party’s official point of view which is, of course, not only meant to unite the party internally, but –with the help of the mass media– to attract voters and win elections” (1987:300).

1 The units of analysis of the Manifesto dataset, until now, are the parliamentary parties at national elections. The Manifesto data cover all the relevant parties within a system. The relevance of parties is defined as the coalition or blackmail potential of a party in a given party system (Sartori 1976:121-125). The data sources are the content analyzed election programs or their nearest equivalents, as well as publicly available election statistics. There are 48 countries in the dataset, including 24 OECD countries and 24 Central and East European countries. The number of parties in the dataset amounts to 623. The period covered is 1990 to 2003.

Until recently, the Comparative Manifestos Project has only included national elections. However, a new Manifesto project for European elections has already started. In 2000, Hermann Schmitt started to develop the idea to apply the Comparative Manifestos Project on elections to the European parliament. The “Euromanifestos Project” at the Mannheim Centre for European Social Research “intends to collect and code all European Parliament election programs of all parties ever represented in that body” (Wüst & Volkens, 2003). Following on the steps of the Euromanifestos Project, the Regional Manifestos Project that we are presenting here covers the gap left at the sub-national level of regional elections.

A classification scheme with 56 broad categories, grouped into seven policy domains, was developed for the Comparative Manifestos Project. Each category of the coding frame is specified by a set of typical issues and political ideas. The classification scheme comprised 26 bipolar positions, such as “Welfare State Expansion” as opposed to “Welfare State Limitation”, 29 unipolar positions (or valence issues) such as “Environmental Protection”, and one general thematic concern for which no direction could be identified, namely “Economic Goals”. Each sentence is grouped into one, and only one of the 56 categories. Because of the different length of elections programs, percentages of the total number of coding units devoted to each category are calculated. This percentage is in turn interpreted as an indicator for the importance of a position policy area to the respective party (Volkens, 2002: 464).

The validity and reliability of the Manifesto dataset and its estimates has been successfully tested on numerous occasions (Gabel and Huber 2000; McDonald and Mendez, 2001a, 2001b).

1.3. The applications of the Manifesto dataset in political science

There are four main topics and/or fields where the dataset from the Manifesto research project has been used widely and successfully: a) “The crowning achievement of the Manifesto Research Project has been to measure party policy change in a variety of countries over an extended time period along the Left- Right dimension. Being able to do so is important because of the centrality of this dimension in political discussion and analysis… It is probably the single most important indicator of party policy, and a pointer to underlying ideology which meshes with

2 membership of a generic ‘family’ and other distinguishing party characteristics” (Budge and Klingemann 2001:20). b) The analysis of the party manifestos has served to challenge widely held assumptions about the operation of party systems. As Volkens points out, “[t]here is no overall trend of convergence to the center of the party systems. Parties do not become more similar. Thus, neither the famous thesis on the ‘end of ideology’ (Bell 1962) nor the ‘catch-all parties’ thesis (Kirchheimer 1966) get any empirical support. Contrary to these hypotheses, most parties fluctuate a lot on the left-right scale, but even the American parties stick to their ideological space” (2002: 466). c) The Manifesto dataset has allowed political scientists to investigate the relationship between parties and voters, central to the operation of representative democracy, and to provide answers to questions such as these: “To what extent have parties’ election programs moved or been synchronized with movements in the preferences of their own supporters? Are there differences in the degree to which particular types of parties cling to a particular set of ideological and policy positions, in contrast to other parties which are more willing to relax such positions in an effort to court broader electoral support?” (Klingemann 1995: 183) d) The Manifesto dataset has facilitated the analysis of the relationship between parties, governments and policy outputs. A core question of accountable government is “whether promises parties made in their election programs and governments made in their public policy statements were congruent, or incongruent, with actual government policies or ‘policy products’ of governments” (Hofferbert, Klingemann and Volkens 1995: 322). The dataset has been used specifically to analyze the logic of coalition governments, by comparing election programs of parties, put forward before the election, to the government declarations held by incoming coalition governments after the election (Laver and Budge 1992).

Up to now, “the manifesto research group is by far the biggest show on the road as a source data on party manifestos” (Laver and Garry 2000: 620) and is “…probably the most important political science research project in this field…” (Laver 2001: 7).

II. The Regional Manifestos Project: filling the gap of regional elections.

After the Comparative Manifestos Project, which covers national elections, and the Euromanifestos Project, which covers European elections, there is one more fundamental level that still needs to be done: the regional level. We propose to elaborate a dataset of the electoral programs presented by parties in sub-national elections, beginning with the case of .

3 2.1. Why is it necessary a Regional Manifestos dataset?

Among the many research directions opened up by the Manifesto dataset, the study of party positions along the left-right dimension over time has captured the most attention. However, the Manifesto dataset offers the possibility of analyzing another political dimension which cross-cuts the left-right divide and which is present in many –or most– democracies, namely the center-periphery dimension, which includes the nationalist and/or cultural cleavage, but also, and most importantly, given the widespread trend towards administrative and political decentralization in Western Europe, issues of multilevel governance.

The analysis of the electoral programs of parties in sub-national elections will open up new directions of research. Firstly, four types of comparisons will be possible: (1) comparisons of changes in policy positions or in emphases over time within specific parties; (2) differences in policy positions or in emphases across parties; (3) differences across regions in Spain; (4) differences between elections at the national and sub-national level. Secondly, by examining the differences across regions, and between the regions and the state, we will be able to study the evolution of intergovernmental relationships and the implications for multilevel government. Thirdly, we will be able to gather priceless information about internal party functioning. Do national parties defend the same policy programs at the regional level? How big are the discrepancies between the regional and national level? Which parties are more divided and which are less divided in their policy positions across the territorial levels? The analysis of national parties at the regional level is highly under-developed in political science, and the elaboration of the Regional Manifestos dataset would be the first step in the direction of filling in this research gap.

2.2. The Coding Procedure

The coding procedure comprises a quantification (how many statements do parties make?) and a classification (what kind of statements do parties make?). The classification scheme contains “general categories that are used to cover the total content of elections programs by identifying the statements of preference expressed in the programs” (Wüst & Volkens, 2003: 64). This classification has already been developed by the Comparative Manifestos Project. However, in order to, on the one hand, guarantee the comparability of the Regional Manifestos Project with the Comparative Manifestos Project and, on the other hand, make in-depth regional-specific analyses possible, the standard Comparative Manifestos Project coding scheme needs to be modified. This will be done by extending the existing classification scheme through the incorporation of regional-specific sub- categories.

2.2.1. Quantification: the coding unit.

The coding unit in a given program is the “quasi-sentence”, defined as an argument. An argument is the verbal expression of one political idea or issue. In its simplest form, a sentence is the basic unit of meaning. Therefore, punctuation can be used as a guideline

4 for identifying arguments. If different issues –however sort- are dealt with in the same sentence they constitute different quasi-sentences even if they apply to the same policy field. On the other hand, the same argument may be very long and may occupy a lot of space, but still be only one quasi-sentence (Volkens 2002:3).

2.2.2 Classification: the Regional Manifestos coding scheme.

At this stage we have not developed the classification scheme fully. However, here we discuss some general guidelines of how we need to proceed in order to elaborate this classification scheme.

The Regional Manifestos classification scheme is based on the existing scheme elaborated for the Comparative Manifestos, but extends it by creating new sub-categories. The existing classification scheme is made up of seven policy domains and 56 broad categories distributed among the policy domains. Not all the policy domains and policy categories are equally applicable to all the Autonomous Communities of Spain. Moreover, it is immediately obvious that this classification scheme was thought of for the national-state level, and not for the regional one. For example, there are considerably fewer categories for the domain 3 “Political System” than for the domain 4 “The Economy”. The analysis of the regional level, however, will require that we develop a significant number of sub-categories in the domain 3, which is so important for multilevel political systems and for the relationships between the central state and the regional administrations. Similarly, concerning external relations, most Autonomous Communities in Spain do not have prerogatives concerning foreign policy and, therefore, some of the categories in the domain 1 will not be filled for many electoral programs (for example, “foreign special relationships”), although other categories will (for example, “European union: positive and negative”). To give one last example, the domain 7 “Social Groups” will be particularly relevant for the regional level, especially in the Spanish case for those Autonomous Communities that represent historical nationalities. New sub-categories will have to be created to account for the position of the parties towards those groups (minority groups) coming from regions of Spain different from their own.

We have used bold letters to highlight those categories that in our opinion will require further extension into sub-categories in order to adapt the classification scheme to the regional level.

DOMAIN 1: External Relations Categories: 101 Foreign Special Relationships: positive. 102 Foreign Special Relationships: negative. 103 Anti-Imperialism 104 Military: positive. 105 Military: negative. 106 Peace 107 Internationalism: positive 108 European Union: positive

5 109 Internationalism: negative 110 European Union: negative

DOMAIN 2: Freedom and Democracy Categories: 201 Freedom and Human Rights 202 Democracy 203 Constitutionalism: positive 204 Constitutionalism: negative

DOMAIN 3: Political System Categories: 301 Decentralisation 302 Centralisation 303 Governmental and Administrative Efficiency 304 Political Corruption 305 Political Authority

DOMAIN 4: Economy Categories: 401 Free Enterprise 402 Incentives 403 Market Regulation 404 Economic Planning 405 Corporativism 406 Protectionism: positive 407 Protectionism: negative 408 Economic Goals 409 Keynesian Demand Management 410 Productivity 411 Technology and Infraestructure 412 Controlled Economy 413 Nationalisation 414 Economic Orthodoxy 415 Marxist Analysis 416 Anti-growth Economy

DOMAIN 5: Welfare and Quality of Life Categories: 501 Environmental Protection 502 Culture 503 Social Justice 504 Welfare State Expansion 505 Welfare State Limitation 506 Education Expansion 507 Education Limitation

6

DOMAIN 6: Fabric of Society Categories: 601 National Way of Life: positive 602 National Way of Life: negative 603 Traditional morality: positive 604 Traditional morality: negative 605 Law and Order 606 Social Harmony 607Multiculturalism: positive 608 Multiculturalism: negative

DOMAIN 7: Social Groups Categories: 701 Labour groups: positive 702 Labour groups: negative 703 Agriculture 704 Middle Class and Professional Groups 705 Minority Groups 706 Non-economic Demographic Groups

(Source: Volkens 2002)

2.3. Organization and Fieldwork

This project is a collaboration between the Instituto de Estudios Sociales Avanzados de Andalucía and the Research Unit “Democracy: Structure, Performance, Challenges” of the Wissenschaftszentrum Berlin für Sozialforschung (WZB). The coordinators of the project are Braulio Gómez (IESAA) and Sonia Alonso (WZB). They will work under the supervision of Andrea Volkens (WZB), responsible of the Comparative Manifestos Project, who has the greatest technical expertise on the dataset, after being in charge of it since 1989.

The fieldwork will be carried out by coders who are in charge of collecting the electoral programs and codifying them following the instructions that will be provided by the coordinators. The coders should preferably be PhD candidates who can use the data they are collecting for their own research and who, therefore, have an additional incentive and/or interest in the data collection. There should be between one and two coders for each Autonomous Community (region), depending on availability of candidates and on the number of electoral programs to be coded. The coordinators are the ones responsible for finding the coders and making sure that they pass the coding reliability test.

7 2.4. Parties to be included in the Regional Manifestos Project for Spain

Region Parties Andalucía Partido Comunista de Andalucía-PCE Alianza Popular-AP, Partido Popular Partido Socialista Obrero Español de Andalucía-PSOE Partido Andalucista Izquierda Unida-Los Verdes-Convocatoria por Andalucía

Aragón Partido Comunista de España Partido Aragonés Regionalista Alianza Popular Partido Socialista Obrero Español Izquierda Unida Partido Popular Chunta Aragonesista

Asturias Partido Comunista de Asturias Partido Socialista Obrero Español Izquierda Unida Federación de Partidos de Alianza Popular, Partido Popular Partiú Asturianista

Canarias Alianza Popular-PDP-UL, Partido Popular Partido Socialista Obrero Español Agrupación Herreña Independiente Coalición Canaria Coalición PIL-FNC Coalición Canaria-PNC

Cantabria Partido Regionalista de Cantabria Partido Socialista Obrero Español Alianza Popular-PDP-UL, Partido Popular

Castilla-La Mancha Alianza Popular-PDP-UL, Partido Popular Partido Socialista Obrero Español

Castilla y León Alianza Popular-PDP-UL, Partido Popular Partido Socialista Obrero Español Izquierda Unida Unión del Pueblo Leonés

Cataluña Esquerra Republicana de Catalunya Partit dels Socialistes de Catalunya (PSOE-PSC) Convergencia i Unió Alianza Popular-PDP-UL, Partido Popular Iniciativa per Catalunya Iniciativa per Catalunya-Verds-Esquerra Alternativa

8 Comunidad de Madrid Partido Comunista de España Alianza Popular-PDP-UL, Partido Popular Partido Socialista Obrero Español Izquierda Unida

Comunitat Valenciana Alianza Popular-PDP-UL, Partido Popular Partido Socialista Obrero Español Esquerra Unida Unión Valenciana

Extremadura Alianza Popular-PDP-UL, Partido Popular Partido Socialista Obrero Español Izquierda Unida

Galicia Partido Socialista Obrero Español Alianza Popular-PDP-UL, Partido Popular Bloque Nacionalista Galego Partido Socialista Galego-Esquerda Galega Partido Coalición Galega

Illes Balears Unió Mallorquina Alianza Popular-PDP-UL, Partido Popular Partido Socialista Obrero Español Entesa PSM-Nacionalistes de Mallorca Pacte Progresista (PSOE, Verds, EU, ENE, ERC) Esquerra Unida-Els Verds

La Rioja Partido Riojano Progresista Alianza Popular-PDP-UL, Partido Popular Partido Socialista Obrero Español

Navarra Herri Unión del Pueblo Navarro Partido Socialista Obrero Español Partido Nacionalista Vasco Izquierda Unida Convergencia de Demócratas de Navarra Euskal Herritarrok Aralar Nafarroa Bai

País Vasco Alianza Popular-PDP-UL, Partido Popular Euskadiko Ezquerra Partido Socialista de Euskadi- Partido Socialista Obrero Español Partido Nacionalista Vasco

9 Eusko Alkartasuna Izquierda Unida Euskal Herritarrok Aralar Partido Comunista de las Tierras Vascas-Euskal Herrialdeetako Alderdi Komunista

Región de Murcia Partido Comunista de España Alianza Popular-PDP-UL, Partido Popular Partido Socialista Obrero Español Izquierda Unida

2.5. Budget

Calculating the budgets for contracts awarded to third persons on collecting and coding the electoral programs:

First step: We start by calculating an average amount of 200 Euro to be paid for collecting and coding each program. Second step: Numbers of relevant parties differ between Autonomous Communities as well as between legislative periods. Therefore, we calculate all parties above 5% of seats in parliament to be included in the data collection. Number of party programs to be collected and coded, by four-year periods (see Annex for a yearly disaggregated calculation):

2004-2007: 55 party programs……….. 200*55=11.000 Euro 2000-2003: 61 party programs………...200*61=12.200 Euro 1996-1999: 65 party programs………...200*65=13.000 Euro 1992-1995: 69 party programs………...200*69=13.800 Euro 1988-1991: 60 party programs...……… 200*60=12.000 Euro 1984-1987: 63 party programs………...200*63=12.600 Euro 1979-1983: 54 party programs………...200*54=10.800 Euro Total (1979-2007) ……………………………...85.400 Euro

Third step: We add another 20% to the amount of money calculated in step two. By adding 20%, we ensure the inclusion of 1. small relevant parties, in particular small parties participating in government coalitions, 2. electoral coalitions in which participant parties issue separate programs and 3. exceptionally long programs, which is quite often the case in Spain.

Total amount applied for: 11.000 Euro + 20% (2.200 Euro) = 13.200 Euro for the period between 2004 and 2007.

10 ANNEX

Overview for Calculating Budgets: Covered Programs and Number of Parties in Regional Elections between 1980 and 2006.

Region 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 Andalucía 4 (5) 4 4 Aragón 4 (5) 4 (5) 4 Asturias 3 3 (4) 3 (5) Baleares 3 (6) 4 (6) 4 (6) Canarias 4 (7) 4 (8) 3 (7) Cantabria 3 3 (4) 3 (4) Castilla/Mancha 2 2 (3) 3 Castilla/León 2 (4) 2 (5) 3 (4) Cataluña 3 (6) 4 (5) 5 (6) 5 Extremadura 2 (4) 2 (5) 3 (4) Galicia 3 (6) 3 (5) 3 (5) 3 La Rioja 3 3 (4) 3 Madrid 2 (3) 3 (4) 3 Murcia 2 (3) 3 (4) 3 Navarra 3 (7) 5 5 (8) 5 Pais Vasco 5 (7) 5 6 (7) 6 (7) Valencia 2 (3) 3 (5) 4 Total 3 8 3 4 36 9 3 10 41 5 3 10 42 5 3

Region 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Andalucía 3 (4) 4 4 4 Aragón 5 5 5 4 (5) Asturias 3 (4) 3 (4) 3 3 Baleares 4 (7) 4 (6) 5 (7) 4 (6) Canarias 4 (5) 4 3 (4) 3 (4) Cantabria 3 (5) 3 3 3 Castilla/Mancha 3 2 2 2 Castilla/León 4 4 (5) 3 2 (4) Cataluña 5 5 5 5 (6) Extremadura 3 (4) 3 3 2 Galicia 3 3 3 La Rioja 3 (4) 3 3 3 Madrid 3 3 3 3 Murcia 3 3 3 3 Navarra 5 (6) 5 (6) 5 (6) 3 (5) Pais Vasco 6 (7) 6 (7) 5 5 (6) Valencia 4 3 3 3 Total 9 52 4 3 6 52 4 8 49 4 8 5 38

Footnotes: (n) – number of parties in parliament. n – number of relevant parties in parliament, that will be included in the dataset.

11