Introduction

Introduction

How Voters Compensate for Diffusion of Power: Empirical Evidence Orit Kedar Department of Political Science University of Michigan ISR 426 Thompson St. Ann Arbor, MI 48106 [email protected] www.umich.edu/~oritk 1. Introduction How does voter choice depend on institutional environment? In Chapter 2 I presented a model of voter choice in parliamentary democracies. I showed that, other things equal, in political systems that facilitate bargaining and compromise, a party ideologically similar to the voter is less effective in representing her interests than a party placed away from her and away from the center of gravity of the party system. In systems in which power is concentrated, on the other hand, policy formation involves little bargaining, and thus voter motivation to turn to extreme parties is diminished. Generally, I argued, the more institutional mechanisms, parliamentary norms, and practices facilitate diffusion of power and compromise, the more voters will compensate for the dilution of their vote by endorsing parties whose positions differ from their own views. In this chapter, I test this implication of the model. The argument above integrates national- level factors into the analysis of individual choice, and the empirical analysis will have to do so as well. To this purpose, I first turn to micro-level data about voter choice in a variety of parliamentary democracies and estimate an empirical model of voter choice consistent with the model presented above. I then utilize macro-level data about institutional diffusion of power and further investigate how variation in voter decision rule across polities estimated in the first step is explained by institutional mechanisms. The first step of my analysis, then, will focus on voters, and the second on institutions in which voter behavior is embedded. Empirically, a cross-national individual-level analysis walks a fine line. On the one hand, I am in search of a cross-country generalization about the principles that guide voter decisions. The institutional implication of the compensatory vote model aims exactly at that. On the other hand, the context in which voters make their decisions differs across countries. Not only do the parties on the ballot among which voters choose differ across countries, but so do the social and political cleavages that lead voters to endorse different parties. Any cross-national empirical 1 analysis of voter choice will have to acknowledge these differences and allow for causal heterogeneity across countries. My estimation approach allows me to establish a general relationship between institutional diffusion of power and voter decision rule across polities, while allowing the micro-foundational model of voter choice to vary across countries, depending on the political context. This chapter is designed to walk this fine line. It begins with individual-level analysis, addressing issues of measurement and estimation. It then turns to presenting evidence regarding voter decision rule in various parliamentary systems, focusing on distances between voters and parties in general and compensatory vote in particular. Building on these results, it then shifts the discussion to the national arena, examining different mechanisms that facilitate compromise and diffusion of power. Finally, it analyzes how voter decision rule varies by institutional environments. 2. Selection of Cases and Data The institutional hypothesis implied by the compensatory vote model guides my selection of cases. To review, the argument predicts that the more institutional mechanisms encouraging compromise are built into policy formation, the more voters will compensate for the expected dilution of their vote and thus will endorse parties whose positions are more extreme than their own views. The empirical analysis in this chapter draws on fourteen western parliamentary democracies that vary greatly in their institutional mechanisms: Australia, Belgium (Flanders), Canada, Denmark, Iceland, Ireland, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, and the United Kingdom. Lijphart famously lays out a two-dimensional map of power sharing in democratic polities. The first, the executive-parties dimension, summarizes the arrangement of executive power, the party system, the electoral system, and interest groups. The second, the federal-unitary dimension, summarizes the centralization level, cameral structure, rigidity of constitutional 2 amending, judicial review, and central bank independence. Averaging polity values on these dimensions between 1945-1996, as well as during a subset of this period, 1971-1996, he then places each polity on the two-dimensional map. Table 1 presents the value of each of the fourteen polities on Lijphart’s two dimensions (measured in the 1971-1996 interval), along with the 25th and 75th percentiles. As we can see from the table, Canada, Britain, Australia, and Spain have a relatively high concentration of power on the executive-parties dimension, with the former two scoring among the lowest, while Switzerland, Belgium, and the Netherlands with particularly high level of fragmentation, are in the upper quartile. Britain, Iceland, and Sweden represent systems with particularly low levels of federal decentralization, while Australia, Canada, and Spain are in the upper quartile. Overall, then, the cases represent a wide range of institutional arrangements. In the analysis below I draw on Lijphart’s measures of power diffusion and modify them in accordance with the theoretical argument. I elaborate on my measures below, but let me mention here a few details about institutional arrangements in these polities. First, the cases vary greatly in their district magnitude. In fact, of all parliamentary democracies, my sample includes three with the lowest possible district magnitude (Britain, Canada, and Australia), and the one with the highest, the Netherlands (150). Second, two of the cases, Switzerland and Belgium, scoring among the highest on the executive-parties dimension, represent an extreme arrangement of power sharing: consociational democracy. Finally, Belgium, Canada, Switzerland, and to a certain degree Spain represent federal systems, while Iceland is among the most centralized polities on this dimension. (The process of devolution in the UK is not much reflected in Lijphart’s measure which averages values between the 70s and the mid 90s. This is not a concern here since the election under study is the 1987 British elections.) Finally, Belgium represents an extreme case of party-system fragmentation in that it has two regional party systems and a national party system is wanting. 3 -- Table 1 here – In addition to variation in institutional features, the combination of cases represents variation on an additional aspect relevant to voter behavior. Issue voting has become an increasingly important predictor of voter choice during the 90s (e.g., Barnes 1997). An analysis drawing on data from this particular period in time bears the risk of producing results specific to cases in which issues were of particular importance to voters. Utilizing data from the late 80s the 90s, and the aughts of the 21st century allows me to mitigate this potential concern. 2.1 Measurement For most cases, individual-level data are drawn from the Comparative Study of Electoral Systems (Modules I and II). Britain and Norway are exceptions; results for Britain draw on data collected by the 1987 British Election Study, and results for Norway draw on the 1989 Norwegian Election Study. Measurement of all micro-level concepts relies on self-reported data from the surveys above. I elaborate on these measures below. 2.1.1 Voter position As the theoretical model, the empirical analysis, too, is unidimensional. In all cases but Britain, the surveys offer a general left-right placement scale. The British Elections Study offers seven issue items for self-placement instead. However, since in the 1987 British elections the three main parties aligned in the same order across all seven items, a single summary dimension can provide a good representation to the policy space. (For a detailed description of party placements on the issues in the British case, see Appendix 1.) 2.1.2 Party position Studies of electoral politics often grapple with measurement of party positions. Two different approaches are found in the literature. The first warns against potential projection by voters whereby respondents are likely to view the parties they endorse as ideologically similar to them. If this is the case, according to this approach, party positions should be measured independently 4 of voter choice. Thus, this approach uses “exogenous” measures such as the average of individually perceived party position (Rabinowitz and Macdonald, 1989), party position as perceived by experts (Laver and Benoit, 2005), or party position measured from the party manifestos (Gabel and Huber, 2000). An additional consideration mentioned by Rabinowitz and Macdonald in favor of this approach is that our theories demand that each party would have a single position on every issue. However, as Westholm explains, while the researcher ought to be able to identify a single position for each party, it is not essential that all voters have an identical perception of parties’ positions (1997, 870). The second approach calls for a measurement choice that depends on the particular research question. Thus, according to this approach, a study whose focus is voter choice should measure party positions as they are

View Full Text

Details

  • File Type
    pdf
  • Upload Time
    -
  • Content Languages
    English
  • Upload User
    Anonymous/Not logged-in
  • File Pages
    43 Page
  • File Size
    -

Download

Channel Download Status
Express Download Enable

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

  • Not to be reproduced or distributed without explicit permission.
  • Not used for commercial purposes outside of approved use cases.
  • Not used to infringe on the rights of the original creators.
  • If you believe any content infringes your copyright, please contact us immediately.

Support

For help with questions, suggestions, or problems, please contact us