zbMATH — the first resource for mathematics

Voting behavior and information aggregation in elections with private information. (English) Zbl 0898.90050
Summary: We analyze two-candidate elections in which voters are uncertain about the realization of a state variable that affects the utility of all voters. Each voter has noisy private information about the state variable. We show that the fraction of voters whose vote depends on their private information goes to zero as the size of the electorate goes to infinity. Nevertheless, elections fully aggregate information in the sense that the chosen candidate would not change if all private information were common knowledge. Equilibrium voting behavior is to a large extent determined by the electoral rule, i.e., if a candidate is required to get at least \(x\) percent of the vote in order to win the election, then in equilibrium this candidate gets very close to \(x\) percent of the vote with probability close to one. Finally, if the distribution from which preferences are drawn is uncertain, then elections will generally not satisfy full information equivalence and the fraction of voters who take informative action does not converge to zero.
Our approach is related in some respects to the approach taken by S. Lohmann (1993) and D. Austen-Smith (1990). Lohmann uses a similar framework to analyze the effects of private information on costly participation in political protest movements while Austen-Smith examines the incentives for strategic voting in small two-alternative elections. Neither Lohmann nor Austen-Smith considered the asymptotic properties of their models. Our results are also related to the literature on information aggregation in auctions: P. Milgrom (1979), Wilson (1977), and W. Pesendorfer and J. Swinkels (1995). In another related paper T. R. Palfrey (1985) analyzes information aggregation in a Cournot model.

91B12 Voting theory
Full Text: DOI