UC Berkeley

This dissertation comprises three studies on the incentives and motivations of public sector personnel, including bureaucrats and politicians. Chapter 1 tries to understand how public organizations can strengthen the mission motivation of public employees. Chapters 2 and 3 explore how the supply of politicians can be improved by either appealing to intrinsic motivation of potential candidates or removing the cost and information barriers that prevents them from becoming politicians.

In chapter 1, I investigate if public organizations can motivate workers by emphasizing the organizational mission to them. I implement a field experiment in partnership with the Department of Health in Pakistan, where I randomly emphasize the public health mission to community health workers, provide performance-linked financial incentives, or do both. The mission treatment improves worker performance across incentivized (home visits) and non-incentivized tasks, while financial incentives improve performance only on the incentivized task. Financial incentives also become less effective at increasing home visits when combined with the mission treatment. Finally, the mission treatment improves downstream child health outcomes---there is a lower prevalence of diarrhea and higher vaccination rates. These results highlight that promoting an organization's mission can be a powerful motivator for public workers, especially in weakly institutionalized environments.

Chapter 2 asks the question: how can we motivate `good' politicians -- those that will carry out a policy that is responsive to citizens' preferences -- to enter politics? In a field experiment in Pakistan, I vary how political office is portrayed to ordinary citizens. I find that emphasizing pro-social motives for holding political office instead of personal rewards -- such as the ability to help others versus enhancing one's own respect and status -- raises the likelihood that individuals run for office and that voters elect them. Emphasizing pro-social motives also better aligns subsequent policies with citizens' preferences. The candidacy decisions are explained by social influence, and not information salience -- I find that social versus personal messaging matters only when randomly delivered in a public setting but not in private. Results also show that changes in political supply, not citizen preferences or behavior, explain policy alignment. Taken together, the results demonstrate that non-financial motivations for political entry also shape how politicians perform in office.

In chapter 3, I focus on understanding what barriers may prevent citizens from becoming politicians. Local governments are said to be susceptible to elite capture in the developing world. Reforms that aim to improve political competition by encouraging non-elite candidates may help reduce elite capture. This chapter reports findings from a randomized control trial that encouraged non-elite candidacy by relaxing three barriers to entry: the cost of running for office, information on benefits from the office, and information on their chances to win the elections. The experiment was conducted before elections for village councils in rural Pakistan and has three primary results. First, it is possible to identify prospective non-elite politicians through the involvement of community members. Second, subsidizing the cost of candidacy, through the provision of services of a lawyer, has a positive effect on the decision of non-elite candidates to enter politics. Third, treatment effects are stronger among candidates who hold higher beliefs about their chance of winning and who expect higher benefits from office.