Research Program on Econometrics of Auctions and Procurements
Program Director: Isabelle Perrigne.
Members: Herman Bierens, Robert C. Marshall, Joris Pinkse, Quang Vuong.
Over the past twenty years, economists have developed an increasing interest in confronting auction models to auction data. Researchers have focused first on testable implications involving observed bids and possibly other information such as firms' ex post revenue to detect whether bidders behave according to some particular auction models. A more recent approach goes further by assuming that observed bids are the outcome of a specific auction model. As such, the auction model provides the structure on which the econometric model relies. This approach allows researchers to recover the key elements of bidders' behavior that can be used for policy evaluations and recommendations. Given their complexity, auction models lead to non standard econometric techniques and challenging questions regarding whether observations are sufficient to recover the key elements of bidding behavior. This problem is known as identification. Our faculty have been key players in the development of quantitative methods and in addressing such questions for analyzing auction data in various settings. The development of such tools allows business leaders and policy makers to assess the potential gain or loss that can result if the design of the auction and the value of the reserve price are changed.
In addition, our faculty have been interested in developing econometric methods that minimize parametric assumptions on the key elements of the model to avoid misspecification and potential bias in policy evaluations. These methods are easily implementable and computationally simple making them usable beyond the academic world. They can be adapted to a large set of situations and thereby have greatly contributed to the development of the empirical analysis of auction and procurement data. Our faculty continue to work in this area as many important problems remain to be studied such as how uncertainty affects bidding behavior. Moreover, additional econometric tools need to be developed to assess empirically the many auction models provided by auction theory.
Auctions can be viewed as relatively simple models of imperfect information in which bidders have some private information such as their valuations for the auctioned object. Other models involving asymmetric information can be found in contracts, regulation and nonlinear pricing in which some agents such as firms or consumers have some information about their efficiency or tastes. Contracts, regulation and nonlinear pricing affect many sectors of the economy such as electricity and telecommunications to name a few. An important set of empirical questions could be addressed in these areas. The experience acquired by our faculty in the econometrics of auctions and procurements is a stepping stone for the future development of econometric methods for regulation and contracts .