UCLA

In this dissertation, I investigate challenges to the delivery of public services to citizens in young democracies. Focusing on the elite actors whose daily job it is to implement development programs, I seek to better understand why politicians and civil servants behave in the ways that they do, and in particular, why each actor may be motivated to engage in misconduct. Regarding development, I focus on the provision and construction of local public goods, such as classrooms, health clinics, and water wells. To build new infrastructure, politicians and bureaucrats must plan projects, select beneficiary communities and award contracts to firms as part of a (theoretically) competitive procurement process. Despite considerable investments in new infrastructure, the results are often disappointing, and corrupt procurement practices are often to blame. It is puzzling to understand why civil servants administer corrupt procurement deals, when it is politicians who gain from this misconduct, and when such behavior conflicts with norms of bureaucratic integrity.

My central argument is that bureaucrats in Ghana engage in corruption because politicians retain discretionary control over their careers. Therefore, the introduction of meritocratic hiring is not enough to protect against widespread corruption because unchecked politicians hold strong incentives to distort administrative processes. These incentives, I argue, stem from politicians' need to generate funds to finance political parties and election campaigns. In short, experienced and professional bureaucrats engage in corruption, even when they do not personally benefit from it, because refusing to do so can jeopardize their livelihoods and interfere in their careers.

My theory calls into question the traditional view in the literature of bureaucratic delegation that giving politicians greater control over bureaucrats improves policy outcomes. I argue that while granting politicians oversight tools -- such as control over bureaucratic transfers, salaries, or promotions -- may align bureaucratic incentives with politicians' goals, it can weaken overall government accountability to the citizenry by allowing politicians to suborn bureaucrats to enable corrupt activities.

Methodologically, I adopt a mixed-methods approach. I combine data from an original survey of local government bureaucrats (N=864), with qualitative data from interviews with bureaucrats, politicians, development practitioners and local governance experts, as well as observational data from a new database of local government bureaucrats (N=40,000). In my survey with bureaucrats, I use experimental and sensitive measurement techniques to measure partisan bias in the allocation of projects across communities and the awarding of contracts to companies.

Overall, my results suggest that meritocracy, increasing the pay of civil servants, and increased levels of bureaucratic monitoring may do little to improve the delivery of public services in environments where bureaucrats continue to be controlled by politicians who hold strong incentives to capture state funds.