Consequences of Electoral System Change in Modern Democracies

Research on electoral systems has generated deep insights into the effects of electoral systems in the last decades. Nonetheless, four methodological problems remain. First, the classical cross- sectional design does not allow to control for all possible variables; second, we confront a omitted variable bias; third, time trends are often neglected; and fourth, the so called endogeneity problem remains unsolved. The quasi-experimental longitudinal design of our project allows us to develop solutions for these problems. For this purpose, we focus on the consequences of electoral system change rather than electoral systems. Concretly, the project aims to answer three questions: first, whether change in electoral systems influences relevant dimensions such as party systems, political representation or quality of democracy. Second, we test whether these effects can be explained by existing theories about the effects of electoral systems. Last, we approach a question relevant from a perspective of political theory as well as political practise: to which extent influence and structure political institutions the behavior of political actors? We use the extensive set of literature on the effects of electoral system and formulate hypotheses about the consequences of change. For this purpose, we identify three dimensions which might be influenced by electoral system change. First, change in electoral systems might lead to modifications in party systems. The second dimension, quality of representation, looks at disproportionality between votes and seats, diversity in political programming as well as congruence between interests of voters and representatives. Third, electoral system change should as well have consequences for the quality of democracy, which is characterized via electoral turnout and satisfaction with democracy. We analyse the relationship between electoral system change and the three dimensions at the national macro level, the meso level of the constituency and the individual micro level. We include all OECD and/or EU-member states since the introduction of the universal male suffrage. To ensure comparability, we include only those elections categorized as free and fair. We employ time-series cross-section and matching procedures. To allow for the extensive selection and evaluation of data which is necessary for this project, it is conducted in cooperation of Prof. Fortin-Rittberger (University of Salzburg) and JProf. Harfst (University of Greifswald). The workload is equally divided between the two universities over five steps of procedure. In a first step, all participating researchers will familiarize with the topic and a coding scheme for electoral system change will be developed. Afterwards, all changes in electoral systems will be identified and during the third step, we will gather data on the dependent variable. The last two steps provide time to analyse the relationship and test hypotheses as well as prepare publications and the final report. Afterwards, all data will be accessable online to the public for further research.

What happens when electoral rules change? The objective of our project was to examine the consequences of electoral system change on the size and composition of party systems, amongst a myriad of other outcomes. A central aspect of the project revolved around the collection of data depicting electoral institutions and their consequences at the level of electoral districts, which is the level at which the immediate effects of electoral systems should theoretically occur. This innovative data collection allows researchers to account for the entire electoral chain in more than 9000 electoral districts in twelve modern democracies and capture the strategic entry of parties, strategic voting decisions of voters, and the electoral rules that structure the conversion of votes into seats. By integrating the different moving pieces located between institutions and party systems, we overcome some of the limitations and challenges previous empirical studies have faced. Our findings show that the number of available mandates in a district impacts party system size. And, that this relationship is conditioned by the sociological context. While part of the electorate reacts to the electoral system by strategically defecting from their preferred party to a more viable party, the effect of strategic voting is blurred by a comparatively large amount of non- strategic behaviour by voters. In the end, the tactical considerations of voters prove elusive, perhaps even more so than previously believed. The mechanical effect, which refers to the transformation of votes into seats, turns out to be the most decisive link of the electoral chain. These findings contribute to the further development of theories of political institutions and their capability to effectively structure the behaviour of political actors. Empirically, the results of the project indicate that targeted analyses capturing the temporal dimension and that include insights from the field of electoral behavior offer promising avenues to tackle the elusiveness of the psychological effect. In addition to these insights surrounding electoral rules about party systems, the Austrian sub-part of the project answered a series of related inquiries about the role of electoral rules as a conditional variable. We demonstrate the limited conditional effects of electoral institutions when analyzing gendered differences in political knowledge, the quality of representation of womens preferences and the impact of electoral fraud on satisfaction with democracy.