Holding Individual Representatives Accountable : The Role of Electoral Systems

Abstract

Voters are reluctant to sanction representatives for individual misconduct if they have to balance candidate-level and party-level factors in their choice, but this trade-off is affected by the electoral system. Our general theoretical model explains why individual accountability can empirically occur in single-member district (SMD) systems but is expected under less restrictive conditions using open-list proportional representation (OLPR). The latter not only decouples party and candidate choice but also makes seat allocation more vote elastic. For a thorough empirical test of our argument, we draw on real-world evidence from state-level elections in Bavaria, Germany, which are held under an unusual mixed-member system. Exploiting a recent public scandal involving one-third of representatives, we examine how electoral punishment of the same candidates by the same voters differs across electoral rules. Drawing on difference-in-differences as well as matching/regression estimators, we show that electoral punishment is substantially larger under OLPR than under SMD systems.