Stickers and Hunters: Perceptions of Candidate Integrity in a Personalised Election

In their latest research, Laver and Sergenti (2012) explain the dynamics of electoral competition between candidates/parties and voters by assuming that political actors adapt their policy positions according to known patterns of voter support for them. The subject of our interest is the voters’ response to changes in policy positions, which is a variable that Laver and Sergenti generally overlook. This may undermine the voters’ trust in the candidate’s future moves as he becomes less predictable (Enelow and Munger 1993). A politician who changes his positions may come across to voters as dishonest, inconsistent, unreliable, and indecisive, and thus loses authority (Sigelman a Sigelman 1986). We assume that a number of other intervening variables, as policy recall, issue salience and framing will alter the effect of issue repositioning by political candidates. To further address the issue, we run laboratory experiments with unidimensional highly salient policy issue, multiple rounds of voting and dynamically updated positions of certain candidates. Issue salience, pragmatic or moral framing of the policy, or degree of political knowledge; none was shown to have a correlation with the perception of candidates’ integrity. Nor did there prove to be any significant correlation between a preference for integrity (as a candidate’s character trait) and selection of the candidate. In accordance with the directional spatial model of electoral competition (Merrill, Grofman 1999), we have however concluded that voters construct their own beliefs about the positions of the "Hunter" (candidate who changed the position), and the credibility of the Hunter’s change of positions is influenced by the degree of the change. This finding presents a challenge to further electoral models of party competition based on multi-agent modeling. NOTE: we submitted this paper into two panels, our first preference being the Experiments in Political Science but not by a wide margin :).