Governance Innovations in Sierra Leone
Principal Investigator
Abstract
Candidate debates have rich historical roots, offer a unique communication platform, and have become integral to contemporary campaign strategy. There is, however, no definitive evidence of whether they affect actual voting behavior. The relative scarcity of political information in the developing world offers an attractive testing ground, where the effects of debates could be more pronounced, persistent and directly linked to electoral outcomes. We experimentally manipulate citizen exposure to debates in Sierra Leone to measure their impacts on, and the interconnections between, voter behavior, campaign spending, and the performance of elected politicians. We find positive impacts on citizen political knowledge, policy alignment and votes cast on Election Day. We then document an endogenous response by participating candidates, who increased campaign expenditure in communities where debate videos were screened in large public gatherings. Over the longer term, we find that debate participation enhanced the subsequent accountability of elected Parliamentarians, who demonstrated greater constituency engagement and development expenditure over their first year in office. To unpack causal mechanisms, individual treatments disentangle the effects of general political knowledge from the information conveyed about candidate persona, and find that both matter. Overall, the experiments speak to the central question in political economy of whether elections effectively discipline politicians, and show how information provision can trigger a chain of events that ultimately influences policy.