Campaign Contests

Abstract

I develop a formal model of political campaigns in which candidates choose how to distribute their resources over two different policy issues. I assume that campaigning on an issue has two simultaneous effects, both rooted in social and cognitive psychology: It increases the perceived quality of the advertising candidate in that issue and it makes the issue more salient, thereby increasing the issue's perceived importance to the voters. Whether a candidate can increase his vote share during the contest depends on the interplay of strategic issue selection, which depends on candidates' comparative advantages, and the aggregate resource allocation to the issues. The aggregate resource allocation---or campaign agenda---depends on an issue's importance, the firmness of voters' conviction regarding candidates' relative quality, and the divisiveness of this issue. A candidate increases his vote share during the campaign contest if he has a comparative advantage on the issue that receives more aggregate spending. Consequently, the contest may be biased in one candidate's favor and an a priori less popular candidate might be the actual odds on favorite. I show that a relatively unimportant issue might receive most aggregate spending and thus could decide the election.