A Different World:
Relative Causal Inference and the Study of Mixed Electoral Systems
Abstract
Mixed electoral systems are often viewed as social laboratories where one can perform
“controlled comparisons” and hence evaluate the relative causal effects of majoritarian
and proportional rules while holding constant potential intervening variables. Of course,
the inferences made by such studies rely for their validity on the assumption that the
majoritarian and proportional components of the election are independent. Recent
studies, however, have demonstrated that interaction, or “contamination,” is generally
observed. In this paper, I employ the logic of relative causal inference to illustrate the
consequences that the violation of the assumption of independence has for the
establishment of the causal effects of SMD and PR. Then, I describe the idea of
“contamination” and outline its implications for the empirical analysis of mixed
electoral systems. I submit that mixed systems constitute a distinct class of electoral
institutions, one that offers psephologists the opportunity to investigate a wide range of
new electoral and legislative outcomes.
Convention