DWIGHT R. LEE, J. R. CLARK
Abstract: It is common to employ small-number voting models to show how majority voting can lead to outcomes that are often at variance to what most people expect. For example, when government projects are voted on separately, and logrolling and side payment agreements are ruled out, each project can capture a majority vote even though all voters would be better off if none of the projects passed. The outcomes are typically driven by the assumption that the decision of each voter is influenced entirely by effect of their vote on their personal financial interests. But this is misleading since the financial interests of voters have little, if any, influence on how people vote when the number of voters is large. The same outcomes may be realized when there are a large number of voters, but they are likely to differ even though the distribution of financial interests is the same for the large group as it is for the small group.