A Model of Candidate Location When One Candidate Has a Valence Advantage

that, under some fairly weak assumptions, for all levels of the valence advantage, the advantaged candidate chooses a more moderate position than the disadvantaged candidate. Empirical studies of congressional elections by Fiorina (1973), and Ansolabehere, Snyder, and Stewart (2001) support this result. D~ owns' (1957) seminal work presents a puzzle for scholars of U.S. elections: Why do candidates not converge to the median voter? One of the best explanations is that the candidates themselves have ideological preferences. As Calvert (1985) and Wittman (1977, 1983) have shown, when candidates have such preferences and they are sufficiently uncertain about the preferences of the median voter, then they do not converge, although they may moderate their positions from their ideal point. The basic version of the Calvert-Wittman model, however, does not account for another basic fact of U.S. elections: often one of the candidates has an advantage due to a nonpolicy factor, such as incumbency, greater campaign funds, better name recognition, superior charisma, superior intelligence, and so on. Stokes has dubbed such factors valence factors, and he has noted their importance in U.S. elections (Stokes 1963;1992).1 In this article I extend the basic Calvert-Wittman model by giving one candidate a valence advantage over the other. I show that if the valence characteristics of the candidates are approximately equal, then when a candidate's valence advantage increases, this causes her to move toward the