Incorporating Emotion Into Cue-Based Political Judgment Modeling

Although the influence of partisanship on American voter behavior is well documented, there are political decisions when party elites are either silent or not in agreement where parties might hold less sway. The present research theorized that in such cases, voters’ emotional projections about the policy rather than a review of available information about the law predict their decisions. This study took a policy capturing approach to modeling political judgment by using participant responses to a proposed policy to create a series of regression models in which partisanship, forecasted emotions, and utilization of policy-relevant cues predicted support for the First Step Act, a bipartisan policy which did not receive united support from either political party. Results showed that affective forecasts regarding the Act were the most robust predictors of voter support. Further, voters were inconsistent in identifying which policy relevant cue they were most responsive to when deciding to support the policy. Finally, results showed that a fast and frugal model which predicted support using the two cues most important to each voter was just as accurate as a model which predicted support using voter response to all cues. These findings suggest that when voters do not have clear direction from partisan affiliation, they turn to their emotional reaction to the policy as a source of guidance in their political judgment.