Snap Judgments: Forecasting Politician Productivity under Limited Information

Snap Judgments: Forecasting Politician Productivity under Limited Information

January 31,2017Working Paper No. 3360

Voting is fundamentally a forecasting problem: voters try to predict the future performance in office based on incomplete information about candidates. Forecast inputs combine observable resume criteria with more subjective assessments of confidence and trustworthiness. In developing countries, the amount of information available can be quite limited.  This paper explores the accuracy of voter forecasts about candidate productivity under varying degrees and types of information. To do so, I ran a series of lab-in-the-field experiments in a weak media environment where ballot photos are both the first and last visual impression many voters have of candidates.  I show first that inferences based on candidate photos alone predict who later wins actual elections with probability greater than chance. I then link these inferences to objective measures of productivity and show that they identify more trustworthy politicians, who divert fewer public resources to personal use, and do so more accurately than a suite of resume qualifications and observable characteristics. Neither snap judgments based on photos nor observable characteristics distinguish politicians along concrete measures of effort. Estimates further suggest that part of the observed returns to electable faces capture intangible persuasion skills, which operate through a direct favorable appearance channel and an indirect accumulation of oral communication skills channel.