This paper argues that campaign finance policy, in the form of contribution limits and matching public financing, can be Pareto improving even under very optimistic assumptions concerning the role of campaign advertising and the rationality of voters. The optimistic assumptions are that candidates use campaign contributions to convey truthful information to voters about their qualifications for office and that voters update their beliefs rationally on the basis of the information they have seen.The argument also assumes that campaign contributions are provided by interest groups and that candidates can offer to provide policy favors to attract higher contributions.

This paper explores a group rule–utilitarian approach to understanding voter turnout, inspired by the theoretical work of John C. Harsanyi (1980) and Timothy J. Feddersen and Alvaro Sandroni (2002). It develops a model based on this approach and studies its performance in explaining turnout in Texas liquor referenda. The results are encouraging: the comparative static predictions of the model are broadly consistent with the data, and a structurally estimated version of the model yields reasonable coefficient estimates and fits the data well. The structurally estimated model also outperforms a simple expressive voting model.