Single-Target Implicit Association Tests (ST-IAT) Predict Voting Behavior of Decided and Undecided Voters in Swiss Referendums

Abstract

Undecided voters represent a major challenge to political pollsters. Recently, political psychologists have proposed the use of implicit association tests (IAT) to measure implicit attitudes toward political parties and candidates and predict voting behavior of undecided voters. A number of studies have shown that both implicit and explicit (i.e., self-reported) attitudes contribute to the prediction of voting behavior. More importantly, recent research suggests that implicit attitudes may be more useful for predicting the vote of undecided voters in the case of specific political issues rather than elections. Due to its direct-democratic political system, Switzerland represents an ideal place to investigate the predictive validity of IATs in the context of political votes. In this article, I present evidence from three studies in which both explicit and implicit measures were used ahead of the vote on four different referendums. Explicit measures predicted voting better than implicit attitudes for decided voters while implicit and explicit attitudes were equally good predictors among undecided voters. In addition, implicit attitudes predicted voting behavior descriptively, but not significantly better for undecided voters while, also from a descriptive point of view, explicit attitudes predicted voting better for decided respondents. In sum, results suggest that, as argued in previous research, the predictive value of implicit attitudes may be higher in the context of issue-related votes but still not as high as initially hoped-for.