Does the Electoral Rule Matter for Political Polarization? The Case of Brazilian Legislative Chambers

Rodolpho Bernabel
2015 Brazilian Political Science Review  
This study explores the effects of electoral rules on political polarization in the legislative branch of government. Since in Brazil the districts are also the states, and senators are chosen according to the plurality-majority rule while representatives are determined by a proportional rule, the comparison between legislative chambers enables one to test whether the plurality-majority rule induces politicians to behave less moderately, and whether the proportional rule has the opposite
more » ... To estimate these effects, roll call data from 1988 to 2010 was analyzed and legislators' ideal points were estimated using WNOMINATE. Evidence in favor of the hypothesis was found, although not in every circumstance. three anonymous reviewers for helpful comments. I also thank Fernando Limongi, Danilo Medeiros, and the Centro Brasileiro de Análise e Planejamento for generously sharing the database for this study. Last but not least, I thank Umberto Mignozzetti for sharing the R code to convert the data into a format suited to run the WNOMINATE function. All remaining flaws are my own. Does the Electoral Rule Matter for Political Polarization? The Case of Brazilian Legislative Chambers 82 (2015) 9 (2) 81 -108 his article explores the influence that particular electoral rules may have on political polarization. More specifically, the polarization in the legislative chambers, measured by the way legislators vote in roll calls. In this paper, I will explore how two different vote-counting rules affect the relationship between legislators and their parties. This relationship is understood by the concept of polarization. When the members of one party vote in one way and those of the other party vote in another way, and if this behavior is observed most of the time, then these two parties can be described as polarized. In a roll call vote, a legislator can either cast a "yea" or a "nay" vote. As this procedure is done repeatedly in one legislature, we can measure how similar, aggregated, or even polarized, are the members of two or more distinct parties. There can be many causes of polarization. In McCarty, Poole and Rosenthal (2006) , the authors argue that income, immigration, and campaign finance affect polarization. Here I am going to argue that the method of counting the votes by which a legislator is elected can also have an effect on polarization. The relevance of such a study becomes noticeable when there is a redesign or reformation of an electoral system, to identify the practical consequences of different electoral rules. The seminal work on the topic is Duverger (1957) , in which the author shows that a pluralitymajority rule leads to political systems with two parties, whereas a proportional rule enables multi-party systems to continue their existence. However, the relationship between electoral rule and legislative behavior has not yet been fully scrutinized. If a plurality-majority rule leads to a less fragmented party system, it is reasonable to expect that the legislators of a party would behave in a more loyal way. This hypothesis is investigated in this paper. Electoral rules are frequently analyzed with regard to their normative features, such as representativity and rationality. However, it is important to understand the practical consequences that a rule may have such as political polarization. The Brazilian case studied here is helpful to test the effect of two electoral rules on political polarization; plurality-majority and proportional rules. This paper will show that the members of Congress elected through the pluralitymajority rule behave more extremely than those elected via the proportional rule, i.e., they vote more often with their party. As will be explained presently, the T Rodolpho Bernabel 83 (2015) 9 (2) 81 -108
doi:10.1590/1981-38212014000200012 fatcat:lvzoejd77zh2doefzooka5hrdy