Why do people vote? An experiment in rationality

The study presents the findings of an experiment conducted during the 1993 Canadian fedeal election campaign. Students in two universities were exposed to a ten-minute presentation about the rational model of voting and the 'paradox' that so many people vote when it is apparently irrational on a cost-benefit basis. Our data indicate that exposure to the presentation decreased turnout in the election by seven percentage points. This result contributes to the debate abut the effect of rational-choice models on real political behavior. More important, the experimental panel data permit the presentation's effect to be decomposed, and this helps explain why people do vote. In this study, turnout was reduced mainly because the presentation diminished the respondents' sense of duty, an effect that was indirect, because there was no reference in the presentation to such motives. Framing the voting act in rational-choice terms induced some students to reconsider whether they should feel obliged to vote.