Do Town-Hall Debates Increase Support for Democracy?

Citizens often overestimate the amount of corruption in their government. In May 2018, researchers ran an online survey asking Tunisian citizens about their perceptions of corruption. The average respondent estimated that more than 67% of Tunisian households have bribed police officers. Yet, a representative poll conducted by the International Republican Institute shows that the true number is only 12%. Despite existing literature on corruption and beliefs, two fundamental questions remain unanswered: 1) what psychological mechanisms cause people to have biased beliefs about bribery?; and 2) does having biased beliefs about bribery affect people’s real-world economic decisions? To answer these questions, researchers are running an experiment in Tunisia. They begin with a Tunisian news website and hire individuals to translate articles about corruption from French to Arabic. The treatments will vary the content of the articles, in an effort to vary the translators’ beliefs about bribery. The outcomes will measure a variety of the translators’ economic behaviors, including their financial investments, job applications, and political contributions. The policy implications of this research are substantial. If overestimating corruption does in fact distort economic behavior, then a large-scale information intervention that debiases people’s beliefs could, by itself, significantly accelerate economic growth.