Incumbency Advantage, Voter Loyalty and the Benefit of the Doubt

Abstract

The standard Downsian model of two-party competition cannot account for the well-known fact that incumbents rarely lose elections. This paper modifies the model to take account of the tendency for voters to give the incumbent the `benefit of the doubt'; i.e. voters may vote for an incumbent even when they prefer the policies of the challenger. We show that if all of the voters give the incumbent a very small benefit of the doubt, or if only a few voters give a larger benefit of the doubt, the incumbent may become invulnerable to defeat by any challenger, especially if his position is a relatively central one relative to voter ideal points. We also show that taking account of voter willingness to give some form of benefit of the doubt to incumbents also helps us to understand why politics deviates from the standard Downsian model in two other ways: (1) tweedledum-tweedledee politics is uncommon, and (2) incumbents often win by much larger than bare majorities. Data on attitudes toward government jobs policy and on US involvement in Central America, as reported in the 1984 NES survey, are used to illustrate that the more centrally located the incumbent, the smaller the benefits of the doubt that are required to ensure re-election. In an Appendix we present a number of new and useful theorems about the stability properties of situations in which voters provide incumbents with some form of benefit of the doubt. Our results show how voter willingness to give incumbents benefit of the doubt can make incumbents, especially centrally located ones, virtually impossible to defeat. Thus, `benefits of the doubt' serve to reduce competition and deter entry.