André Walter

Proportional representation (PR) electoral systems have become widespread because they are expected to ensure the representation of interests with small or geographically inefficiently distributed voter bases. Yet in reality, most PR systems consist of a large number of districts that vary substantially in size, and some have surprisingly low magnitude. Existing research shows that such differences matter greatly for political outcomes but offers no explanation for their origins. We argue that the design of electoral districts in newly adopted PR systems is systematically linked to electoral geography and partisan politics. If parties with concentrated voter bases can influence the design of the new electoral system, they will create a significant number of low-magnitude districts. In general, parties involved in designing districts benefit from electoral disproportionalities under the new PR rules. Empirically, we use newly collected district-level data for several Western European countries in the early twentieth century.

Proportional representation (PR) electoral systems have become widespread because they are expected to ensure the representation of interests with small or geographically inefficiently distributed voter bases. Yet in reality, most PR systems consist of a large number of districts that vary substantially in size, and some have surprisingly low magnitude. Existing research shows that such differences matter greatly for political outcomes but offers no explanation for their origins. We argue that the design of electoral districts in newly adopted PR systems is systematically linked to electoral geography and partisan politics. If parties with concentrated voter bases can influence the design of the new electoral system, they will create a significant number of low-magnitude districts. In general, parties involved in designing districts benefit from electoral disproportionalities under the new PR rules. Empirically, we use newly collected district-level data for several Western European countries in the early twentieth century.

Analyzing the voting behavior of Swiss members of parliament (MP) using newly collected individual, district, and cantonal level data, we show that both electoral disproportionalities and the insurgent parties’ electoral potential are important determinants of MP voting behavior on the adoption of proportional representation (PR). However, in contrast to the prominent electoral threat thesis, the insurgent party’s high electoral potential decreases the probability that MPs of established parties support PR. The reason for this relationship is partisan redistricting, whose relevance has so far been largely ignored in the literature. We demonstrate that adapting electoral district boundaries for political reasons, if possible in a given institutional context, can be a powerful alternative to the adoption of PR, because it allows established parties to retain parliamentary majorities even as an insurgent party’s electoral potential increases.

As the first country to introduce proportional representation (PR), Belgium has attracted considerable attention. Yet, we find the existing explanations for the 1899 breakthrough lacking. At the time of reform, the Catholic Party was politically dominant, advantaged by the electoral system, and facing reformist Socialists. Nevertheless, they single-handedly changed the electoral system and lost 26 seats in the first election under PR. We argue that the Catholics had good reasons to adopt PR. Majoritarian rules tend to create high levels of uncertainty because they provide incentives for non-dominant parties to cooperate. Such electoral coalitions are facilitated by multidimensional policy spaces that make electoral coalitions other than between nonsocialist parties possible. PR reduces the effectiveness of cooperation between non-dominant parties, but such certainty comes at a price. In addition, in the presence of dominant parties, divisions over electoral system reform often result in intra-party conflicts that may be more decisive than inter-party conflicts.

Previous research on welfare state development in federal states has claimed that interstate competition induces a race to the bottom with regard to social expenditure. While existing research has focused on the impact of migration inflows of poor households on states with high provision of social expenditure, it has remained silent with regard to migration outflows. In this article, I argue that migration outflows are associated with increased social expenditure if eligibility restrictions for social benefits between states are relaxed. Facing outflows of workers and the unemployed, low-spending states respond with increases in social expenditure to retain a sufficient workforce and counteract rising wage levels. As a result, states with higher levels of outward migration are catching up with more generous states instead of engaging in a race to the bottom. To support my claim, I use a time-series cross-sectional data set of Swiss cantons covering the period 1930–75.