Mechanical effects of electoral systems on proportionality and parliament fragmentation

Abstract

The following master’s thesis studies the mechanical effects of electoral systems on two electoral outcomes – proportionality and parliament fragmentation. The aim of the study is to investigate whether there is a precise universal relationship between proportionality and fragmentation across different electoral systems. The thesis places itself into the general framework of new institutionalism, saying that institutions including electoral systems matter, but their precise effects depend on the context in which they operate. We propose that if context is taken into account and held under control while analysing pure mechanical effects of electoral systems, a clear universal pattern emerges between proportionality and fragmentation. A computational experiment is carried out using constituency level data of 5 countries and 10 elections from the CLEA database. The results show a squared relationship between proportionality and parliament fragmentation, not a linear one that has been a tacit assumption in the debate between the proponents of majoritarian and proportional electoral systems. Testing this squared relationship on broader data proves its validity, especially for PR electoral systems.