The Effects of Probabilistic Forecasts and Stakes on Electoral Participation in an Exterminal Setting

Abstract

This study looks at the effect of probabilistic forecasts and policy stakes on vote intention in an experimental setting. This study examined stakes and forecasts through exposing subjects to mock newspaper articles in two experiments: a fictitious mayoral election and the 2018 U.S. Senate race in Arizona. The study found that varying the policy stakes of an election had a significant impact on intention to vote, while varying the probabilistic forecasts did not have a significant effect on intention to vote. In the mayoral experiment, when the mock article said that the candidates disagreed on a major policy (high stakes treatment), subjects were significantly more likely to indicate that they would vote in that election in comparison to the article where subjects were told that the two candidates agreed on that same policy issue (low stakes treatment). In the Arizona experiment, I found that low stakes made subjects significantly less likely to say they would vote in comparison to the control treatment. Both stakes and probability had some effects on candidate favorability and candidate preference in my mayoral experiment. My results lead me to conclude that the media has the potential to influence political participation levels through the way they cover elections.