The Political Consequences of Electoral Laws in the 1990 Hungarian Elections

Abstract

Over the past thirty years a well-developed body of theory has emerged describing how different electoral laws interact with voters' preferences to promote particular party systems.' In this study, I test several empirical predictions generated from this theory. Specifically, I examine hypotheses relating the electoral formula, which describes how votes are translated into seats, and the district magnitude, which indicates the number of seats to be distributed, to the number and representation of parties in a party system. The immediate influence of these electoral laws on a party system is straightforward.2 The electoral formula shapes the party system by limiting the representation of small parties. The majority formula, for example, allows only the party with the majority of the votes to gain office, while proportional representation (PR) formulas distribute multiple seats to parties according to their vote share. The district magnitude influences the party system by limiting the number of parties that can be represented. For these reasons, electoral systems with large district magnitudes and proportional representation (PR) formulas are expected to be more equitable in their distribution of seats and provide representation for more parties than systems with small district magnitudes and majority electoral formulas.3 In designing a test of these hypotheses, one must recognize that the influence of electoral laws on party systems depends crucially on the political context in which they operate.4 In a polity with several strong local parties, for example, the majority formula may lead to as large a party system as a PR system. In addition, the extent to which district magnitude can influence the number of parties gaining representation depends on the number of parties competing for office. In a polity where three parties compete, district magnitudes greater than three will not constrain the number of parties represented from these districts. Thus, electoral laws merely promote, within their political context, particular party systems.5 As a consequence, a rigorous test of the aforementioned hypotheses must control for contextual factors that might influence the electoral results independently of the electoral laws. This test requires data from elections with two characteristics: identical voter preferences over a constant set of parties and variation in electoral formula and district magnitude. Actual elections, unfortunately, have never completely satisfied these requirements. Faced with this constraint, scholars have examined the existing data from a variety of angles, each with its own advantages and disadvantages. Most often, scholars have used cross-national studies which allow for comparisons of party systems from nations with a broad range of electoral formulas and district magnitudes.6 These studies enhance our appreciation of a wide variety of electoral systems but are unable to control for voter