Contamination effects in mixed-member electoral systems: a dissemination of measurement techniques

Abstract

There is widespread agreement that both tiers of a mixed-member electoral system do not operate independently of one another and that instead they interact, producing contamination effects. These tend to raise the number of parties away from what Duverger’s Law or the M + 1 rule suggest. However, the field has not yet determined an undisputed methodological approach to answer the following question: raised by how much, exactly? This paper identifies and differentiates amongst three major measurement techniques, herein titled the Difference Approach, the Likeness Approach, and the Simulation Approach. By doing so, it provides a more concise map of each of their logics, their varied implementations, their drawbacks, as well as possible ways forward.