There is a large body of research on the effect of electoral rules of citizen-elite ideological congruence (e.g., Blais and Bodet 2006; Golder and Stranski 2010; Lupu et al. 2017; Powell 2009). Less attention has been paid to a corollary of this phenomenon: the partisan effects of electoral systems. Across established democracies after the Second World War, it is widely assumed that right-wing parties are more likely to reach political office in countries with majoritarian electoral rules (Iversen and Soskice 2006; Doring and Manow 2015). However, a systematic analysis of the mechanisms that provoke this pattern is almost completely missing in the literature.
Do voters under proportional representation (PR) behave in a systematic different way than under majoritarian electoral rules? What if the one person, one vote principle comparatively holds less in non-proportional countries generating as a result ideological bias?
In this paper, we answer these questions by making three main contributions. First, we describe the variation in the partisan effects of electoral systems across countries over the entire post-war period in 23 OECD countries. Second, we conduct a time series cross-sectional analysis to identify some of the correlates of the conservative bias of majoritarian electoral rules. Within this framework, we supplement this aggregate-level analysis by employing individual-level data from the Comparative Study of Electoral Systems (CSES) project. Third, we provide causal evidence of the partisan effects of electoral systems at the cross-national level by combining the use of the synthetic control method and an instrumental variables approach.
Our goal is not to provide a novel theoretical answer of the partisan effects of electoral systems. Rather, by employing a new dataset that includes the available information of surveys and elections for 22 countries over 75 years, we offer evidence of a time-consistent relationship between three main factors, voting behaviour of the middle class and geographical concentration of left-wing parties, and type of electoral system.
Our findings suggest that at least part of the conservative bias of majoritarian systems is due to the lower propensity of middle class voters to vote for left-wing parties under these electoral rules as well as the lack of electoral nationalization of left-wing parties in non-proportional systems. By doing this, we offer novel insights on the causes of variation in levels of economic redistribution across democracies.