Replication Data for: Campaign Messages, Polling, and Elections: Theory and Experimental Evidence

How do politicians' track records and campaign messaging interact and affect voters' welfare? We analyze this question, theoretically and experimentally. In the theoretical model, which we implement in the experiment, politicians choose how much of an economy's resources to allocate to the citizenry - keeping the remainder for themselves - then face re-election against a challenger. Both incumbents and challengers have private information about their own quality which determines the economy's level of resources. We vary whether candidates can send campaign messages, and the level of variability in candidates' quality. We observe that both higher quality variability and allowing campaigning benefit citizens by allowing them to better select and hold accountable higher-quality officials. Also, when incumbents have performed poorly or when quality variability is high, challengers' negative campaigning (criticizing the incumbent) increases, and incumbents' positive campaigning (emphasizing their own strengths) decreases.