The Threat of Abstention: Candidate Strategies and Policy Representation in U.S. Presidential Elections

Abstract

Introduction The unified theory of voting that we have studied in the multicandidate elections of France, Britain, and Norway – a theory that does not consider turnout effects – does not imply divergence of party strategies in two-party elections, such as those typical of presidential elections in the United States (see Chapter 4, section 4.3). Indeed, Erikson and Romero (1990) have shown that for the unified voting model, a convergent equilibrium is likely to exist in which the two candidates locate at the weighted mean of the voter distribution, with each voter's policy preferences weighted according to the elasticity or marginality of her vote choice probability. How can we explain the persistent tendency – observed empirically – for the two main American parties to adopt positions that, although they may not be far apart by European standards, are nevertheless distinctly different? To this end, we include in the unified model not only the voter's choice of candidate or party, but also the voter's decision whether to vote at all. The high rate of abstention in the United States, relative to that of most European countries, is well documented (Lijphart 1997; Dalton 2002: Chapter 6). Therefore, if abstention can affect party strategies, we might expect the effect to be particularly pronounced in the American context. We consider two possible motivations for abstention: (a) alienation, in which no candidate's utility reaches the voter's threshold for voting, and (b) indifference, in which the voter finds insufficient difference between the candidates to motivate casting a ballot.We label the resulting specification the unified turnout model . Later, in Chapter 9, we will explore policy strategies in Britain under the unified turnout model.