The Mechanical and Psychological Effects of Electoral Systems: An Appraisal with Experimental Data

The paper proposes a way to measure the mechanical and psychological effects of majority runoff versus plurality. Building on a series of laboratory experiments on presidential-type elections, we evaluate these effects with respect to the probability of election of a centrist candidate. In our experiment, the runoff system slightly favours the centrist candidate, but this total effect is small because the mechanical and psychological effects tend to cancel each other. The mechanical effect of runoffs is to systematically advantage the centrist, as usually assumed; but our study detects an opposite psychological impact, to the disadvantage of this candidate.