PI: Orit Kedar
Project coordinator: Ms. Liran Harsgor
An electoral system is an essential and principal component of representative democracy. It translates preferences of citizens to a legislative body and thus mediates votes, aggregates ballots, and inevitably distorts preferences. No electoral system can count all voices; no electoral system can voice all preferences. Even in a hypothetical world in which social equality exists, turnout is perfect, and all citizens have equal access to decision makers, any electoral system tilts the playing ground in different ways. Theorizing and empirically analyzing how the electoral system affects political outcomes is the aim of this study.
Perhaps the most important factor defining an electoral system is the number of seats allotted to a district-the district magnitude. It is long established that the district magnitude affects key features of the political landscape in a country, from representation (Rae 1967), through the number of parties (Duverger 1954, Cox 1997, Clark and Golder 2006), whether the government consists of a single-party or a multi-party coalition (Lijphart 1999), parties' strategy (Dow 2001, Merrill and Adams 2007), voters' considerations (Kedar 2005), and even redistribution policy (Iversen and Soskice 2006). In fact, Duverger's Law (1954), stating that single-member district tends to encourage a two-party system is one of the most established relationships in political science. Combined with Duverger's hypothesis that proportional representation results in multipartism and insights of later works (Rae 1967, Riker 1982, Taagepera and Shugart 1989, Cox 1997), differences between the district and the national level notwithstanding (Chhibber and Kollman 2004), the general expectation is that the smaller the district magnitude, the smaller the number of parties in a country.
Most countries, however, have many district magnitudes, not one. The Norwegian Storting (2009 elections) has a district of as few as four representatives but also a district of seventeen, and nine other magnitudes in between those two. Spain's Congress of Deputies (2008) has a district of three representatives but also a district of thirty five. Districts of the Portuguese Assembly of the Republic (2009) range in their magnitude between two and forty-seven representatives, Switzerland's districts vary between one and thirty-five, and so on. The distribution of districts within a country, the range of magnitudes, the variation in magnitude, their skewness, the ratio of largest to smallest, the proportion of seats in parliament elected via large districts compared to those via small districts, and how this heterogeneity correlates with political cleavages, all vary greatly from one country to another. In spite of the prevalence of and differences in this within-country heterogeneity in district magnitude, with a single exception (Monroe and Rose 2002), all studies in the voluminous literature on electoral systems and its effects sidestep it by focusing solely on a magnitude of a middle district in each country.
The proposed study is the first large-scale comparative project that conceptualizes, theorizes and empirically analyzes how district heterogeneity affects representation.
The study develops approaches to thinking about district heterogeneity in a politically relevant manner. It theoretically and empirically addresses questions such as: how does the structure of electoral districts shape representation? How does the design of district heterogeneity affect the national party system? How does district heterogeneity affect party coordination?
In general, the study consists of three parts:
(i) Theoretical analysis aimed at understanding the political implications of different properties of district heterogeneity and the distribution of districts in a country. In this analysis I will borrow techniques from other fields in the social sciences, theorize, conceptualize, and develop new measures for analyzing the effect of electoral districts on key aspects of the political system.
(ii) Cross-national empirical analysis informed by the theoretical part. This analysis will rely on district- as well as national-level data collection in a broad cross-section of countries.
(iii) Controlled in-depth comparison of representation and the party system in two pairings of countries that vary in their district structure.
My (terrific) team of research assistants:
- Ms. Liran Harsgor, PhD student, Department of Political Science, Hebrew University
- Ms. Maayan Mor, Masters student, Department of Political Science, Hebrew University
- Mr. Raz Sheinerman, PhD student, Department of Political Science, Hebrew University