This dissertation empirically studies bidder behavior in procurement auction environments. I apply auction theory to develop models explaining market outcomes in two distinct settings. The empirical analysis uses detailed bid-level data and takes advantage of the particular features of each market to estimate the models. I then use the estimated parameters to conduct counterfactual analyses and identify specific drivers of bidder behavior.In the first setting, I study government procurement auctions for road repair and construction projects in California. This market is characterized by significant turnover of participants and there is a number of government programs in place designed to prolongrm tenure. The first aspect of this market which I investigate in Chapter 2 is learning-by-doing among inexperienced firms that have recently entered this market. I extend previous methods for estimation of dynamic auction games by developing a dynamic model of experience accumulation that allows for intertemporally-linked costs and endogenous auction participation. I document the presence of learning-by-doing in the data, and use the theoretical model to show that bidders with a higher level of experience have lower average costs than inexperienced bidders. Consistent with dynamic incentives, my empirical results indicate that inexperienced firms charge lower markups and bid more aggressively in order to leverage learning-by-doing and achieve cost reductions in the future.Next, in Chapter 3, I address the subcontracting side of the highway procurement auction market in California. I empirically investigate whether experience in the subcontracting market can improve future prospects of new entrants. The evidence on learning-by-doing from Chapter 2 suggests that experience obtained in the primary market plays an important role in the future success of such firms. However, many recent entrants participate as subcontractors before actively submitting bids as primary contractors. I conclude that despite the fact that experience gained as a subcontractor might be more limited in scope and task-specific, it can be a valuable stepping stone for firms that have only recently entered this market.Finally, in Chapter 4, which is joint work with Dr. Elena Krasnokutskaya and Dr. Christian Terwiesch, I turn my attention to a very different auction market. In this case, we analyze data from an online market for programming services to study bilateral trading patterns that arise between buyers and sellers from distinct world regions. Our methodology allows us to control for both the country-specific differences in seller quality and the country-specific differences in buyers;; preference for particular bidder attributes relative to price, as well as for differences in their outside options. We decompose the clustering in trading into effects generated by various components of the model. Wend that differences in cost distributions across seller countries drive most of the clustering pattern.
PDF
download
878987797
028-85220240
