UC San Diego

This dissertation examines China's attempts to incorporate new, seemingly democratic f̀ire alarm' oversight institutions into their authoritarian political system. China's market economic reforms created new principal- agent problems that their traditional top-down oversight institutions cannot rectify, and these problems are forcing Chinese leaders to look for new solutions. Democratic leaders solve principal-agent problems by transferring monitoring authority to the citizenry, and Chinese leaders are borrowing that strategy by giving their own citizens new fire alarm institutions for holding local-level officials accountable to national-level laws and policies. However, Chinese leaders do not deploy these institutions uniformly across all policy sectors. Instead, they vary their fire alarm institutions on a sector-by- sector basis, and this dissertation attempts to explain that variation. I argue that Chinese leaders divide their fire alarm institutions into two broad categories : governmental and non-governmental. Governmental fire alarms operate through state agencies; non-governmental fire alarms operate through non-state communication and organization networks. Governmental fire alarms are relatively safe, but they do not work well in China's politically controlled administrative environment. Non- governmental fire alarms are relatively effective despite China's political controls, but they also pose additional risks. I argue that the risks associated with non- governmental fire alarms vary across policy sectors. I hypothesize that Chinese leaders distinguish between high- and low-risk policy sectors and enable the more dangerous (and also more useful) non-governmental fire alarms in the low-risk sectors only. This strategy maximizes the benefits from fire alarm oversight while also minimizing the potential political risks from non-governmental fire alarms. Unfortunately, this strategy also limits fire alarm effectiveness - especially in the high-risk sectors with the most restrictive non-governmental fire alarm policies - so China's fire alarm institutions will not be as effective as their (relatively unlimited) democratic counterparts. I test my hypothesis by comparing Chinese fire alarm oversight strategies across three policy sectors: rural land expropriation (high risk), environmental protection (medium risk) and food and drug safety (low risk). The case studies from these three sectors support my argument that Chinese leaders differentiate between governmental and non-governmental institutions and only employ non-governmental fire alarms when the sector-specific political risks are relatively low