Electoral Systems and Turnout: Evidence From a Regression Discontinuity Design

Abstract

This case study will show you how to carry out a Regression Discontinuity Design by way of an example about voter turnout in elections. In 2002, municipalities in Poland were assigned to either a majoritarian or a proportional electoral system based on a population threshold of 20,000 inhabitants. This assignment by use of a population threshold introduces a discontinuity that splits practically identical municipalities into two groups, one group with a proportional electoral system and another group with a majoritarian electoral system. We will show you in this case study how to use this discontinuity to reliably and accurately estimate the effect of a change from a majoritarian to a proportional electoral system on voter turnout.