Publication Type: Paper (prepared oral presentation)
Abstract: What would happen if voters had to vote according to another electoral system? For example, if the French presidential election were held according to the FPTP system instead of the traditional two-round system, would voters change their vote and how?

To answer to these questions and to measure the psychological effects associated to two electoral systems, here the one and the two-rounds ones, we used a multi-design approach based on individual data.

First, a survey conducted just before the first round of the 2007 French Presidential election including questions about vote intentions under the actual rule (two-round ballot) and some hypothetical questions about vote intentions under the one-round system. This investigation, which is similar to an in vivo experimentation, makes it possible to vary a parameter and only one (the electoral rule) and to examine the variations of votes, all other things being equal (national or local election, issues, number of competitors and electoral campaign). Realized in a given country, at one precise time, this kind of simulation is obviously contextualised.

Second, a series of experiments in laboratory (led in France and Canada) aiming at comparing the effects of the two rules. This kind of simulation, not very widespread in political science, is the in vitro experimentation, which aims at controlling the whole determinants of the vote, and thus to fix the context in which the survey data are collected. This "abstract" character confers an essential virtue on experimentation results: they can be generalized.

The results of these two methodologies shows that the multi-design is the sole method capable of taking jointly into account both candidates' size and position and thus measuring the "multi dimensions" of the psychological effects of electoral rules.