UC Berkeley

Recent studies have listed resource misallocation as one of the key reasons of cross-country income gaps. This dissertation uses China as an example and asks whether political distortions can be a source of resource misallocation, and if so, what are the cost of those political distortions. Chapter 1 asks whether Chinese state-owned firms' preferential access to credit have any impact on the aggregate R&D activities. I find that the difference in access to credit creates a wedge in the factor market, which leads to considerable welfare loss by misallocating R&D inputs across Chinese firms. Specifically, worse access to credit market creates severe disadvantages to private innovators when they compete with state-owned incumbents. Such disadvantages distorts private innovators' incentives to invest in R&D. As a result, the wedge in access to credit between state-owned and private firms lowers aggregate innovation, leading to slower productivity growth. I develop and structurally estimate an endogenous growth model with a factor market wedge to capture this effect. I show that removing the wedge increases annual productivity growth by 1.2 percentage points and total welfare by 23%. Compared with the static loss from cross-sectional markup dispersion, the dynamic loss from misallocation in the R&D sector accounts for 90% of the total welfare loss. Compared with other mechanisms that lead to welfare loss, distortions to R&D incentives are the primary cause of misallocation in the R&D sector. These distortions are also quantitatively important in explaining the welfare loss.

State-owned firms enjoying preferential access to credit serves as an important foundation of my analysis in Chapter 1. Chapter 2 investigates this empirical relationship in depth by providing causal evidence on the wedge in access to credit between state-owned and private firms. Specifically, I estimate the effect of the 2004 Chinese banking reform on allocation of credits. Before the reform, connections to the government give unproductive state-owned firms preferential access to credit from state-owned banks. The banking reform changes the incentive structure of the banks, making them profit-oriented and thus limits the role government connections play in loan allocation. Using exogenous variations in the numbers of state-owned firms resulting from a high-profile national defense project, I find that the banking reform improves allocative efficiency more in cities with more state-owned firms. This improvement is driven by extending credit access to more productive private firms who lack access to credit prior to the reform. With the newly acquired credit, they invest more and grow faster. A back-of-the-envelope calculation shows that the banking reform leads to 4% gain in sectoral TFP among the most affected cities.

Chapter 3 switches gear and asks whether constraints and incentives of authoritarian regimes can lead to resource misallocation. To this end, I exploit the spatial discontinuity at the boundaries of Soviet Zones that separated regions controlled by the Chinese Communist Party (CCP) and the rivaling Kuomintang during the First Chinese Civil War. I find that counties controlled by the CCP in the past receive 20% more targeted fiscal transfers from the central government today. Since Soviet Zone boundaries are determined by uncertain results of military operations, I can attribute the difference in fiscal support from the central government to differences in counties' Soviet Zone status. I argue that supporting Communist-controlled counties helps the CCP to maintain its political legitimacy as the ruling party of China, while supporting Kuomintang-controlled counties do not. Therefore, favoritism towards Soviet Zones originates from constraints and incentives the CCP faces. Finally, I document the cost associated with such favoritism. Despite receiving extra resources from the central government, favored counties do not exhibit faster economic growth. The lackluster outcome is driven by misallocation of local fiscal resources: local government in the favored counties have higher per-capita administrative expenditures and have more people on government payrolls. On the other hand, they do not spend more to support infrastructure, education, or agriculture. My results suggest that political favoritism can be an important source of resource misallocation.