Fields:
Primary: Energy and Environmental Economics
Secondary: Industrial Organization, Public Economics

Expected Graduation Date:
June, 2019

Thesis Committee:
Frank Wolak (Primary)
wolak@stanford.edu

Peter Reiss
preiss@stanford.edu

Lawrence Goulder
goulder@stanford.edu

This paper presents a previously unrecognized but important consequence of subsidizing green technology in the context of California residential solar market. I show that this market is characterized by many small and local solar suppliers as well as a few large solar suppliers. Large firms advertise heavily and pursue primarily the environmentally focused consumers, whereas small and local suppliers have a limited ability to advertise and pursue primarily the economically focused consumers. Using rich micro-level data, I estimate how subsidies are apportioned between sellers and consumers among different seller-customer groups, and to what extent subsidies have stimulated adoption. The estimate is robust to a spatial discontinuity design. Contrary to many studies in the previous literature, I find very different subsidy pass-through among different groups. Large suppliers capture almost all the subsidy benefits whereas small and local suppliers concede the majority of the benefits to customers. Because customers have different motives to go solar and the solar market is opaque, endogenous matching of consumers and sellers causes this large variation in subsidy pass-through. Over half of the subsidies fail to reduce prices to consumers and encourage adoption. Instead, a large fraction of subsidies mainly yields rents to large solar suppliers to sustain their advertising-supported business model.

China's air pollution has been consistently severe for the past several decades. It is largely due to the fact that for a long time, the evaluation of local governors for promotion has been solely based on local GDP growth. To boost short-term local economy, the local governors sacrifice environment qualities such as air quality. Starting from year 2013, the central government of China incorporated air quality into the evaluation system with the aim to mitigate air pollution. This paper examines whether this policy change has successfully incentivized local governors to reduce air pollution. The stylized facts of the fossil energy industry in China suggest that governors can easily move air pollution sources toward the provincial border so that they can maintain a good rank in the pollution-adjusted-GDP Tournaments. I manually collect daily air quality data for 190 major cities in China from 2000 to 2016, and map them with local weather information for over 2000 stations across China. By utilizing a DID strategy, I find that the new policy causes cities on the border to be 30% more air-polluted in terms AQI than their counterpart cities that are far away from the border in the post-policy periods. This finding is re-confirmed when using wind as an instrument. These results are largely due to the "tragedy of the commons" effect. In extreme scenarios, the total volume of air pollutants could even be higher after the introduction of the new policy.

Low social security participation is a big issue in China. This paper investigates the extent to which the fragmented social security system in China affects social security participation, especially among the migrant population. By exploiting the fact that China is in the process of a gradual shift from a fragmented social security system to a consolidated nationwide system, I manually collect data on the progress of such reform at the provincial level, which hasn't been digitalized before. Using a Diff-in-Diff strategy, I estimate the response of social security participation to the centralization reform, which relies on the assumption that the reform roll-out is orthogonal to other factors that affect participation. I find evidence that in those provinces where reform has finished at the provincial level, residents are 2.7 per cent more likely to enroll in social security programs. In addition, richer, older and more educated people are more likely to participate in social security. By controlling for demographics, I find that migrants are 4.5 per cent more likely to participate in the social security system if reform has happened compared with nonmigrants. Provision of relevant information about the reform also matters, though the effect is not statistically significant.

This project builds a multi-period structural model in a durable goods market with a monopolistic seller and a pool of forward-looking consumers who are heterogeneous in their search costs and price elasticities. By applying it to data, this model can be used to examine how possibly a discrete subsidy scheme can affect the seller's intertemporal pricing strategy and consumers' purchasing time. The implications are that under certain conditions, a time-varying subsidy can assist the seller in exercising intertemporal price discrimination, resulting in an increase in firm surplus and a decrease in consumer surplus.

This study aims to examine whether and how consumers' political ideology towards the environment may have affected the pass-through of green subsidies.