Date of Award
5-2015
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Legacy Department
Economics
Committee Chair/Advisor
Paul Wilson
Committee Member
Matthew Lewis
Committee Member
Raymond Sauer
Abstract
The first two chapters of my dissertation are essays on estimation of procurement auctions with secret reserve prices. In the last chapter, I develop a test on strong disposability versus weak disposability for non-parametric data envelopment analysis (DEA) estimators. In auctions with secret reserve prices, the possibility that the item goes unsold in the first round generates a particular form of multi-round auction in which information on bids is revealed after each round. If bidders have an imperfect estimate of project costs, the information revealed by observing the bids of others in the previous round can mitigate the classic winner's curse problem that arises when firms' costs share a common component. Using data from the Indiana Department of Transportation (INDOT), I test for the existence of a common value component and analyze how releasing bid information helps to cure the winner's curse using reduced form analysis and structural estimation. I find that the common value component in bidders' costs have a more important role in round 1 than later rounds. The released bid information cures the winner's curse by providing bidders with more accurate cost estimates. Counterfactural studies indicate that using secret reserve prices benefits a government under a common value paradigm as opposed to public reserve prices. Previous studies on multi-round procurement auctions assume bidders are risk neutral and myopic. I make the extension by allowing bidders being risk averse and forward looking. Using the data from INDOT, I estimate the structural model with myopic bidders and detect risk aversion of bidders. I then estimate the model with forward looking bidders. Using results of structural estimation, I conduct a series of counterfactual studies to address the following question: which of the following is the best selling mechanism in terms of the government expenditure and the probability of no sell, auctions with public reserve prices, secret reserve prices, no reserve price or some other format? Non-parametric data envelopment analysis (DEA) estimators have been widely applied in analyses of productive efficiency. However, most existing empirical studies have assumed strong disposability of inputs and outputs. This constitutes a restriction which should be test with data. Using methods developed in Kneip et.al (1998) and Kneip et.al (2008), this paper derives the rate of convergence and the asymptotic distribution of DEA estimator under the weak disposability assumption. With the information about the rate of convergence and the asymptotic distribution of DEA estimator under both strong disposability and weak disposability, I can test the weak disposability assumption against the strong disposability assumption applying the central limit theorem results of Kneip et.al (2014) and using the similar method as in Kneip et.al (2013). Monte Carlo results illustrating the performance of the test in terms of size and power are also presented.
Recommended Citation
Wu, Di, "Estimation of Procurement Auctions with Secret Reserve Prices" (2015). All Dissertations. 1529.
https://open.clemson.edu/all_dissertations/1529