Automobile growth has created severe problems such as traffic congestion, air pollution, and carbon emission worldwide. To address these problems, Chinese local governments implemented a series of automobile regulations to slow down auto growth. They set up a yearly quota of license plates and require potential car buyers to obtain a license plate before buying an automobile. Local governments end up with three modes of allocating the scarce license plates: only lotteries, only auctions, and a mix of lotteries and auctions. The main debate is the tradeoff between the efficiency represented by auctions and the equity represented by lotteries. By drawing on two survey results in Beijing and Guangzhou, this thesis analyzes the automobile regulations with respect to the normative aspect, the empirical aspect and the behavioral aspect of the debate. I structure this thesis in three separate essays.

The first essay focuses on the normative aspect. It provides different ways of interpreting fairness, which is a guiding principle of the lottery-based allocation scheme adopted in Beijing. The essay builds up a three-layer framework to evaluate the fairness of the lottery-based allocation scheme. I conclude that the lottery is superficially fair because it only narrowly conforms to the equality rule, which generates a positive public view. The findings suggest that fairness is an extraordinary broad concept, and thus any policy can at most only conform to certain aspects of fairness, as in the Beijing case.

The second essay focuses on the empirical aspect. It analyzes how people chose between lotteries, auctions, and non-participation in Guangzhou, where a mixed mode of lotteries and auctions is adopted. This essay focuses on the marginal effects of socio-demographics, especially income, on choice. It turns out that high-income people chose both lotteries and auctions, in contrast to the common belief that lotteries benefitted the poor and auctions benefitted the rich. In fact, people responded strongly to the fluctuation of the number of participants and winning rates of lotteries and auctions. People speculated in both lotteries and auctions, instead of making decisions based on their needs or economic status.

The third essay focuses on the in-depth behavioral aspect. I analyzed the impact of risk biases on the choice between lotteries, auctions and non-participation under three different policy scenarios. Consistent with theories, people over-estimated the winning rates of lotteries and under-estimated the winning rates of auctions. When faced with a choice between only lotteries and non-participation or between only auctions and non-participation, the risk biases dominated the decision-making process. On the contrary, the cognitive biases did not influence decisions much if people were faced with lotteries, auctions, and non-participation, as in the Guangzhou case.