Trust or Control? Mid-level bureaucratic dynamics and the quality of education provision in India's Districts

Summary (no more than 500 words) How can public service quality be improved in developing countries? Is tighter control or more autonomy and trust conducive to service quality?
Difficult-to-monitor tasks, i.e. thick tasks that are "implementation intensive" (Pritchett, 2014) and not solely logistical, pose a challenge to efforts preventing bureaucrats from shirking and require different accountability systems than thin tasks (Andrews, Pritchett and Woolcock, 2017). Those following the New Public Management (NPM) paradigm and its variations argue for quantitative targets, monetary incentives, close surveillance and easier exit by creating private options (Fukuyama, 2013). Those questioning this paradigm point out the successes of bureaucracies where bureaucrats are using informal channels in the public interest, do not shirk even when not watched and show a high esprit de corps despite a lack of monetary incentives.
In the field of education, there has been a shift in recent years away from the quantity of schools built towards how much children learn. Thus, the focus of analysis has moved from thin tasks that are easy to observe and monitor (e.g. has a school been built?) to thick tasks that are difficult to monitor (e.g. do one hundred thousand teachers distributed across twenty thousand primary schools engage children in learning?).
Preliminary research questions include:
1. Is local embeddedness of District-level bureaucrats conducive to the quality of school education?
2. Does exposure to high-quality District bureaucracies influence the performance of bureaucrats in later postings?
3. Where positive outliers exist, which approach better explains the processes by which they came to be or operate?
The following working hypotheses will be tested:
1. Local embeddedness is conducive to bureaucratic performance and even more crucial for thick tasks than thin tasks
2. Exposure to high-quality bureaucracies changes bureaucrats' preferences for a limited amount of time
3. Hidden positive outliers can better be explained by autonomy and trust rather than tighter control
The proposed PhD aims to answer whether control or trust and autonomy are conducive to the quality of education by analyzing panel data on several hundred of India's Districts. It aims to contribute to the larger debate on reforming public services to ensure that citizens get their rights fulfilled. To do so, a mixed methods approach will be used that includes large-N statistical analyses as well as in-depth qualitative work.