Env. & Dev. Econ. Research SeminarVoting and information: Evidence from a randomized controlled trialStefano Carattini
18. Dezember 2019
Abstract: Information asymmetries affect many real-world decisions, including voting behavior. Important welfare losses may occur if people's voting behavior is biased because of information asymmetries and misleading political campaigns. One crucial example is environmental policy. Growing evidence suggests that information asymmetries affect people's perceptions of Pigouvian policies. In particular, people tend to underestimate the incentive effect of such policies and overestimate their drawbacks. In combination with a ballot initiative on carbon taxes, taking place in Washington state, we realize a large randomized controlled trial providing information on carbon taxes to over 285,000 individuals. We collect data at the precinct level and measure variation in voting behavior across treatments and with respect to the counterfactual. We find that our interventions affect voting behavior, especially for voters located in precincts relatively opposed to the initiative and more exposed to information countering the arguments for carbon taxes. Additional evidence is provided at the individual level with a survey of about 1,000 individuals, using the same treatments as in the randomized controlled trial. The survey confirms that the informational treatments change voting preferences through beliefs’ updating.
Zeit: 12:15 - 13.45 Uhr
Raum: 0079 VMP 5