Storable Votes

National Bureau Of Economic Research Inc, Cambridge MA

Investigators

Abstract

The inability of majority voting to account for the intensity of voters' preferences leads to many practical problems. Where majority voting is the rule, minorities may be inadequately protected. In addition, the obvious opportunity for mutual gain existing whenever two voters have strong preferences over different issues inevitably results in vote trading, compromising transparency and penalizing less experienced outsiders. In international organizations, in particular, the recognition that sovereign countries should not be forced to accept decisions to which they strongly object, typically leads to national vetoes, risking paralysis and increasing bargaining. Ideally these vetoes should be costly, allowing member countries to express the intensity of their preferences but only at some cost. This project studies a simple scheme that appears potentially able to reconcile the advantages of majority voting with the protection of more intensely felt interests. Consider a committee meeting regularly over time to take decisions that affect all of its members. Decisions are taken by majority vote. However, each committee member is given a stock of votes at the beginning of the year, with the proviso that these are all the votes at his disposal for the year, but he is free to cast any number of them at any one meeting. Committee members can then choose to increase their weight over more strongly felt decisions, at the cost of reducing their influence over other choices. The idea of using more resources, here more votes, when a decision is valued more is very natural, but storable votes have no precedent in the literature. Initial results suggest that votes that can be stored over time have better welfare properties than non-storable votes; perform well even if voters are not fully rational, are fairer to minorities, and are both more transparent and welfare superior to tradable votes. In addition, the mechanism is very simple, does not violate any ethical prior and could realistically be implemented. The theoretical results have been confirmed in initial experiments. The research project will analyze the theoretical properties of storable votes, moving from simple if plausible examples to more general situations, and using experiments as check on the accuracy and practical relevance of the theoretical analysis.