Zhao, Naibao (2018-04). Essays in Empirical Microeconomics. Doctoral Dissertation. Thesis uri icon


  • This dissertation introduces two essays with the focus on alternative approaches to empirical research in economics. The first essay uses a reduced-form approach to address a "macro-phenomenon:" the long-standing puzzle of China's high household saving rate. The second essay employs a structural model to evaluate the social welfare of a procurement mechanism: the A+B auction. In the first essay, we study the role of income inequality interacting with liquidity constraints in explaining the high household saving rate in China. The predictions implied by a simple lifecycle heterogeneous agent model are consistent with data facts. Using three large nationally representative data sets, China Household Finance Survey (CHFS), China Family Panel Studies (CFPS), and Chinese Household Income Project (CHIP), we find robust evidence that (1) the rich save more; (2) the poor are more likely to face liquidity constraints, and the effect of liquidity constraints on household saving rate is significantly positive; (3) income inequality has a significant positive effect on aggregate household saving rate; and (4) the marginal propensity to consume out of transitory income for poor households is significantly higher than for rich households. Our study provides a policy implication that economic policy of reducing income inequality would lower the aggregate saving rate and thus become a policy of economic transition and growth. In the second essay, we investigate an innovative and widely used contracting mechanism, the A+B auction, in highway procurement projects. We introduce a structural model with time incentives/disincentives and construction uncertainties under which contractors' actual working days may deviate from the bidding days in the construction phase. This may make the A+B mechanism neither ex-ante nor ex-post efficient. We demonstrate that the model primitives including the contractor's cost function, distribution of private types, and incentive/disincentive parameters are all nonparametrically identified. Using the data from highway procurement auctions in California, we provide strong empirical evidence that considering the existence of uncertainty in the structural analysis would lead to a significant efficiency loss. On average, the ex-ante (ex-post) welfare loss is 38% (37%) of the contract value. Counterfactual time incentive contracts under A design would decrease the ex-ante (ex-post) total costs by 26% (36%) of the contract value, with an average reduction of $2.6 million ($3.7 million).

publication date

  • May 2018