Sanctioning Regimes and Chief Quality: Evidence from Liberia

We investigate how different sanctioning regimes and the quality of local leaders affects public goods provision in Liberian villages. We conduct a public goods experiment where leaders act as third-party punishers under one of two exogenously imposed sanctioning regimes. Under the first "flat fee" regime leaders receive a flat fee as compensation but do not receive any monetary gains from punishment. Under the second, "incentivised" regime leaders receive the punishment (tokens taken from the game participants) as compensation. We use villagers' perceived measures of corruption of their leader as our preferred measure for leadership quality. To empirically distinguish between sanctioning itself and the identity of the person sanctioning we have a treatment variation where a random villager acts as the third party punisher. We find that real village leaders elicit higher contributions than random villagers or groups without sanctioning. We also report that the effectiveness of sanctioning is attenuated by chiefs that are perceived to be of low-quality, especially when the sanctioner has no material incentive to punish. This suggests that low-quality chiefs are less likely to exert effort for public goods if they do not also privately benefit from it. Finally we find that people's preferred regime choice seems to depend on their real-life experiences in the village rather than their individual characteristics. Current development programmes heavily rely on community self-management and local institutions. Our paper supports the idea that a programme's success is likely to depend on whether villagers deem their leader to be credible norm enforcers.

Keywords: Corruption, Public goods, Monitoring, Sanctioning, Field Experiment

JEL Classification: C9, K42, O17, O12