Modeling Negative Campaigning

Abstract

Negative campaigning is an important aspect of campaign competition but plays little or no role in existing models of campaigns. Within the context of plurality elections for a single office we model the incentives that affect the use of negative campaigning. Under simplifying but still quite general assumptions we show a number of results, including the following key conclusions: (1) for two-candidate competition the front-runner will engage in more positive and less negative campaigning than the opponent; (2) in a three-candidate contest with one candidate clearly trailing by a large margin and playing mainly a spoiler role, that candidate will only engage in positive campaigning; and (3) in any three-candidate contest, no candidate engages in negative campaigning against the weaker of his two opponents, so that to the extent there is negative campaigning, it will be directed against the front-runner or it will come from the front-runner. These results have direct empirical applications to multicandidate primaries and nonpartisan contests and can provide insight into recent general elections as well.