Gabriel Madson (Duke University)
Abstract: Issue voting, where citizens select candidates primarily for their positions on political issues, is a normatively appealing theory of voting. A public whose political behavior is driven by preferences on issues, rather than non-issue considerations, is better equipped to hold representatives accountable for their policy actions. While there is evidence that voters engage in issue voting to some extent, we know quite little about how voters engage in issue voting. That is, how many, and which, issues are voters considering and how are these considerations combined to form a vote intention? Often, issue voting research has focused on whether it occurs or not, how prevalent it is, or, given that it occurs, how utility is calculated over single issue dimensions. But with few exceptions, this work fails to consider how voters process issues available for consideration in any given election and how these issues are aggregated by the voter to reach a decision. The most influential framework in the study of issue voting has been spatial theory, which assumes the public utilizes issue information in a highly motivated and sophisticated manner. From this, a number of critics have argued that issue voting is not prevalent because of well-documented limitations and biases of human psychology in the political sphere. In turn, several alternative voting models have been proposed including identity-based voting, retrospective voting, and voting based on simple cues. These alternative forms of voting share two characteristics: (1) they rely on heuristics that allow voters to make a choice without considering all available information about the candidates, and (2) the information considered by voters is not issue-based. In contrast, I argue that psychological constraints do not preclude issue voting. That is, voters use issue-based heuristic strategies when choosing candidates for public office. These heuristic issue voters cut a middle-path between the extremes of spatial theory and ignoring issues outright – one that places value on both issues and efficiency. While political behavior research has previously considered issue voting under psychological constraints, it has focused on the meta-choice of issue voting versus other decision rules, like partisan voting — that is, how citizens “decide how to decide”. This research departs from previous work on issue voting by instead looking at how the public can engage in different forms of issue voting. I adapt a series of heuristic decision rules from the judgement and decision-making literature, as well as one full-information normative benchmark, to the issue voting context and study the prevalence of these rules among the public. Empirically distinguishing these rules is difficult, however, because the polarized political landscape typically leads voters to the same decisions regardless of how they use issue information. Using two different nationally sampled experimental surveys, I examine respondents’ tendencies to engage in different forms of heuristic issue voting in randomly (Study 1) and algorithmically (Study 2) generated electoral environments where different rules are able to reach divergent decisions. I then employ a Bayesian Cognitive Model, which is intended for analyzing structured cognitive models like these systematic decision rules, in order to estimate the proportion of voters expected to be engaging in each type of issue voting. Results across the two studies find evidence of voters using a variety of issue-based heuristic decision rules, allowing them to meaningfully engage with issue information in a simplified and efficient manner. Two heuristic strategies in particular, Equal Weights (an issue agreement tally model) and Take-the-Best (similar in concept to the ‘issue public’), are estimated to be used by a high proportion of voters. Additionally, use of these strategies is found even when party cues, the most powerful of non-issue cues, are readily available. Together, this suggests that heuristic issue voting is a plausible framework for modeling voting behavior. Voters appear to have a variety of heuristic strategies at their disposal that allow them to meaningfully engage with issue information despite cognitive and motivational constraints. Understanding how voters engage with issue information is a fundamental area of political behavior that needs more exploration before we conclude that voters cast aside issue information altogether in favor of social identities and non-issue shortcuts.