Essays on Informal Governance: Enforcement and Accountability under Weak Institutions

This dissertation deals with two interrelated problems of village-level governance in low-income countries: mobilizing contributions to local public goods and ensuring that such contributions are not embezzled or misappropriated. The motivating question is: How are contributions to village-level public goods sustained, if the enforcement capacity of village governments is limited and village leaders are widely perceived as corrupt?.;In village communities, formal institutions of enforcement and accountability are often dominated by informal social relations, especially in countries where state capacity in the rural areas is limited. When formal institutions of enforcement and accountability are imperfect, challenges to local revenue mobilization include not only the potential inability of local authorities to control free riding, but also the risk that contributions are misappropriated by local leaders. Misappropriation of contributions can affect both the aggregate welfare effects of local revenue mobilization and its distributional consequences, with potential repercussions for citizens' willingness to contribute. Nevertheless, many villages are able mobilize hard-to-enforce contributions to a respectable extent.;To shed light on this puzzle, this dissertation relies on a combination of behavioral experiments, surveys, and open-ended interviews with village leaders and randomly sampled adult village residents in forty-eight villages in rural Tanzania. The research is designed to discriminate between alternative causes of cooperation between village leaders and their constituents, as well as among village residents themselves. The results provide novel insights into three distinct aspects of village-level public goods provision.;One experiment concerns the sensitivity of local revenue mobilization to the possibility of misappropriation. Results from this experiment show that village residents' disposition to make confidential, voluntary contributions to a public good is only minimally reduced by the possibility that their contributions can be embezzled. Rather than being indifferent towards the risk of embezzlement, it appears that village residents anticipate voluntary moderation. A simultaneous experiment reveals that individuals voluntarily refrain from maximal rent extraction, especially if they are entrusted with resources that benefit members of their own community or that have been voluntarily contributed by their peers. Out-of-sample tests confirm that greater voluntary restraint in a village is associated with a lower sensitivity of voluntary contributions to the possibility of embezzlement.;Next, the dissertation examines the preferences and incentives of village leaders in situations where they have the opportunity to misappropriate public resources. In a behavioral experiment with actual village leaders, which is validated through observational comparisons with their rent extraction behavior in the real world, most elected village chairpersons are intrinsically less willing to misappropriate their co-villagers' contributions to a public good, compared to the appointed village executive officers and randomly sampled residents from their own villages. However, village leaders respond ambiguously to the need to anticipate the consequences of their rent extraction decisions for their reputations or future social interactions in their village. On average, the resulting incentives cause externally appointed village executive officers, but not elected chairpersons, to engage in less rent extraction. This suggests that at least some elected village leaders do not anticipate to be held to high standards of public-spiritedness.;Lastly, the dissertation explores the role of peer sanctioning and reputational mechanisms in local revenue mobilization. Contrary to existing laboratory findings, social incentives in the real world -- caused by the need to anticipate the consequences of free riding for their reputations and future social interactions in their village -- encourage some village residents to contribute less to a public good than they would otherwise be willing to. This could be a consequence of distributive norms in village communities, which promote selective tolerance for free riding and potentially create incentives for individuals to hide their true willingness to contribute, in order to not be perceived as wealthy, exploitable, or suspiciously generous. Moreover, the extent to which social incentives in a village discourage contributions to public goods is negatively correlated with real-world revenue mobilization capacity in the village and positively correlated with the prevalence of an informal coercive institution.