In the larger debate about how political institutions affect fiscal and economic performance, attention has returned recently to electoral systems. It now appears that proportional systems minimize distortions in one class of democracies (wealthy, small, open economies with weak political competition), while majoritarian systems do so in another class (larger, poorer, or more contentious democracies); but we lack an explanation for this observed regularity. In two recent papers, the principal investigator develops and begins to test a model of distortionary vote buying in which majoritarian voting systems minimize social cost where votes are cheap (easily bought), but proportional representation (PR) does so where votes are dear (hard to buy). This accounts plausibly for the observed regularity; but other testable implications arise from the model (and were not anticipated in its construction); for example, (a) total social cost rises monotonically as votes grow cheaper under PR, but first rises and then declines under majoritarian systems, and (b) the economically "worst" outcome under PR is considerably worse than the worst under any majoritarian system. The current model is a very simple one of interaction between incumbents and voters, and tests of it to date have been cross-sectional and have involved only a few indicators. In future research, it is proposed: (a) to model party competition explicitly; and (b) to test on a much broader array of indicators (some of which are constructed laboriously) and on time-series data for all democratic regimes from 1960 to the present.
