Electing Legislatures

Abstract

A "legislature" is defined to be an assembly of at least two elected officials which selects final policy outcomes. Legislative elections therefore concern the electoral choice of such an assembly. The classical two-candidate, single-district, model of electoral competition is not a legislative election in the sense of this essay. In the classical model the legislature comprises the winning candidate: this agent has monopolistic control of the legislative decision-making machinery, and implements his winning policy. With this system, voters have a straightforward "best" voting rule for any pair of candidate positions offered in the election: vote sincerely. In the multi-stage legislative electoral system, final outcomes depend on the entire composition of the legislature and the specifics of legislative decision-making. With such a system, voters' decisions are considerably less straightforward, which in turn complicates candidates' strategic choices. This paper presents a fairly technical review of the spatial-theoretic literature on legislative elections. The paper was commissioned by Norman Schofield for the conference on Coalition Theory and Public Choice (Fiesole, Italy: May 1987). On the one hand, the task was easy: the literature is small and much of it involves my own work. On the other hand, the task was difficult: the literature is small and much of it involves my own work. In any event, I am grateful to Professor Schofield for giving me the opportunity and incentive to raise some issues with which I have long been concerned. He is in no way responsible for any errors or omissions the paper might contain. I feel perfectly free, however, to blame him for the appearance of self-indulgence that the essay surely has.