The effect of information on voting behavior and electoral outcomes: An experimental study of direct legislation

The extent to which an electoral outcome is responsive to a voter 's preferences over outcomes depends on the voter 's ability to express her preferences in the act of voting. For instance, if a voter knows everything there is to know about the consequences of her voting decision, then her vote will be a relatively accurate measure of her preferences. In contrast, if a voter is uncertain about the personal consequences of her voting behavior, then the correspondence between the vote she casts and her preferences over possible electoral outcomes will be less clear. In this paper, I use a series of experiments to examine the relationships between the observed voting behavior of relatively uninformed voters and the voting behavior that these same voters would have exhibited had they been completely informed about the consequences of their actions. My goal in running these experiments is to show what we can infer about individual and collective preferences from observed voting behavior. The experiments are based on observations of, and prior theoretical research on, direct legislation elections (i.e. the referendum and initiative). In direct legislation elections, voters, as opposed to elected representatives, are given an opportunity to accept or reject specific resolutions to particular collective choice problems. I use the experiments to identify conditions under which relatively uninformed voters can use simple cues to cast the same votes they would have cast had they been better or completely informed. That these conditons resemble regular characteristics of actual direct legislation elections suggests that voters who economize on information costs are often able to govern effectively. These