The electoral college vs. a direct vote: Policy bias, reversals, and indeterminate outcomes†

Abstract

In this essay we address three issues from the debate over substituting a direct vote for the Electoral College: (1) the extent to which the Electoral College induces candidates to advocate policies that they would not advocate under a direct vote; (2) the probability that the winner under a direct vote is not the winner under the Electoral College; and, (3) the probability that third parties—both national and regional—force elections into Congress. We conclude, first, that while candidate platforms are sensitive to election procedure, the likelihood of politically significant differences has diminished over time. Second, computing the probability of reversals in 2‐candidate contests, we find that, even in close elections, this probability does not exceed 0.2. Finally, and contrary to our initial supposition, we conclude that regional as against national third parties are not necessarily more likely to induce indeterminate outcomes, ceteris paribus. We also explore situations in which the Electoral Colleg...