Structural Analysis of Auction Models

Open Access
Gabrielli, María Florencia
Graduate Program:
Doctor of Philosophy
Document Type:
Date of Defense:
June 16, 2009
Committee Members:
  • Coenraad Arnout P Pinkse, Dissertation Advisor
  • Coenraad Arnout P Pinkse, Committee Chair
  • Herman J Bierens, Committee Member
  • Sung Jae Jun, Committee Member
  • Runze Li, Committee Member
  • first--price autions
  • econometrics
  • collusion
  • structural approach
The analysis of auctions is an active area of research for both theoretical and empirical economists. Concentrating on a particular auction format, namely the First--Price Sealed--Bid auction, this dissertation contributes to the analysis of auction both from a methodological viewpoint and from a more applied perspective. The first chapter of this dissertation provides and overview of the literature relevant to this thesis. There, we review the theoretical background to auction models, collusion and the econometric literature on nonparametric estimators. Chapter 2 proposes a semiparametric estimator for the distribution of private values within the class of indirect methods that have been analyzed using the so--called structural approach to estimate auction models. The proposed estimator is shown to have desirable statistical properties namely, it is consistent and has an asymptotic normal distribution. Moreover, the estimator attains the parametric rate of convergence. Chapter 3 concentrates on the study of collusion in auctions. The main objective of chapter 3 is twofold. First, to provide a methodology to detect collusion using a structural approach, and second to apply the methodology to field data on procurement auctions for highway construction in California.