Policy deliberation and voter persuasion: estimating intrinsic causal effects of town hall meetings

Institutions are a vital part of the way we make decisions about policies and implement them, but most randomised experiments only focus on the policy itself. In this lecture, I call for more rigorous evaluations of political institutions and policymaking and discuss how this can work in practice with reference to work my colleagues and I did in Benin and the Philippines. We randomly assigned districts in both countries to use either standard, clientilistic campaign messaging or broad-based, policy-oriented campaign messaging with deliberation at town hall meetings. In the both countries, we see a greater vote share for the party that participated in policy-focused deliberation. In Benin, deliberation also increased voter turnout. People demand better politics, and if politicians change their behaviour, they are rewarded. I hope to see more experiments like these, which separate institutional effects from policy effects and help overcome issues like clientelism that plague politics in the developing world.