Welfare criteria for tax reforms: Efficiency aspects

This paper provides unexpected support to Nash's view of how bargaining determines the social choice. We first present a spatial model of electoral competition with probabilistic voting and a very close relation between voters' preferences and their choice probabilities. The outcome of the electoral competition game is then shown to be precisely the social alternative that maximizes a Nash-type social welfare function (theorem 1). This electoral outcome is also interpreted as a unanimity likelihood maximum (corollary 1). Within our framework, existence and uniqueness of such an outcome are guaranteed (corollaries 2 and 3).