Ensuring Pareto Optimality by Referendum Voting

摘要

We consider a society confronting the decision of accepting or rejecting a list of (at least two) proposals. Assuming separability of preferences, we show the impossibility of guaranteeing Pareto optimal outcomes through anonymous referendum voting, except in the case of an odd number of voters confronting precisely two proposals. In this special case, majority voting is the only anonymous social choice rule which guarantees Pareto optimal referendum outcomes.This paper has been presented at the Murat Sertel Memorial Conference on Economic Theory, 14–16 May 2004, Istanbul; at the Third Meeting of the Society for Economic Design, 30 June–2 July 2004, Mallorca and at the Seventh Meeting of the Society for Social Choice and Welfare, 21–25 July 2004, Osaka. We thank all the participants. We are particularly grateful to Steven Brams, Christopher Chambers, Franz Dietrich, Tarık Kara, Marc Kilgour, Semih Koray, Jean-François Laslier, Christian List, Thomas Ratliff, Maurice Salles, Tayfun S önmez, Harrie de Swart and two anonymous referees for their valuable comments. We thank the faculty of Economic and Administrative Sciences of Bilgi University for supporting our participation in these conferences. Our research is part of a project on Electoral System Design which is financed by Bilgi University Research Fund.