next up previous index
Next: asa.business.04 Up: ASA Business and Economic Previous: asa.business.02

asa.business.03


Sponsoring Section/Society: ASA B&E

Session Slot: 2:00- 3:50 Wednesday

Estimated Audience Size: xx-xxx

AudioVisual Request: xxx


Session Title: Empirical Auction Models

Theme Session: Yes/No

Applied Session: Yes/No


Session Organizer: Marshall, Robert C. Pennsylvania State University


Address:

Phone:

Fax:

Email:


Session Timing: 110 minutes total (Sorry about format):

110 minutes total Opening Remarks by Chair - 0 or 5 minutes First Speaker - 30 minutes (or 25) Second Speaker - 30 minutes Third Speaker - 30 minutes Discussant - 10 minutes (or none) Floor Discussion - 10 minutes (or 5 or 15)


Session Chair: Marshall, Robert C. Pennsylvania State University


Address: Department of Economics, 613 Kern Graduate Building, Penn State University, University Park, PA 16802

Fax:

Email: rcm10@psu.edu


1. Econometrics of the First Price Auction with Asymmetric Bidders

Bajari, Pat,   Havard University


Address: 200 Littauer Center, Department of economics, Harvard University, Cambridge, Ma 02138

Phone: 617-495-8040

Fax: 617-495-8570

Email: pbajari@arrow.fas.harvard.edu

Abstract: This research develops an approach to econometric modeling of the first price sealed bid auction with independent private values. Previous econometric work imposes the assumption that the distribution of bidder valuations is identical for all bidders. This research allows bidders to draw valuations from potentially different distributions. Thus heterogeneity of bidders is allowed. The approach to inference in this research is Bayesian. Difficulties in classical inference, such as nonstandard asymptotic theory are therefore circumvented. Existence and uniqueness of equilibrium for the theoretical model is established. The general econometric approach is then applied to the specific problem of asphalt firms bidding for government contracts in the vicinity of Minneapolis, Minnesota. This research finds that both location and differing levels of efficiency across firms are important factors in determining the price level. Estimates for firm profitability are produced. The structural model is compared to a linear specification. Non-nested hypothesis testing strongly favors the structural model. A test for bid rigging in the market is proposed. The bid rigging test differs from previous approaches in that an equilibrium model of competition and an equilibrium model of collusion are directly compared. The results strongly favor the hypothesis of competition. Finally, this research compares expected price levels under both a first price and second price auction.


2. The Impact of Synergies on Bidding in the Georgia School Milk Market

Marshall, Robert C.,   Pennsylvania State University


Address: Department of Economics, 613 Kern Graduate Building, Penn State University, University Park, PA 16802

Phone:

Fax:

Email: rcm10@psu.edu

Raiff, Matt E., Pennsylvania State University

Abstract: Each summer milk processors around the country participate in sealed bid procurements for the right to provide public schools with milk throughout the subsequent academic year. School district contracts are an important part of a complicated vehicle routing problem that milk processors solve on an ongoing basis. There are allegedly substantial cost savings for a milk processor from servicing a district that is directly adjacent to one they already service. In this paper, following the work of Krishna and Rosenthal (1996), we construct a procurement model allowing for cost synergies. The equilibrium bid function, which reproduces the single object first price procurement bid when synergies are not present, maps directly into an empirical specification. Using data from a time period when bidders were allegedly acting non-cooperatively we find significant and substantial support for the presence of synergies in the bidding. This implies that models of collusive bidder behavior must take into account the presence of synergies in bidder valuations.


3. Identification and Estimation of a Class of Game Theoretic Models

Florens, Jean-Pierre,   Univ. des Sciences Sociales, Toulouse


Address:

Phone:

Fax:

Email:

Protopopescu, C., Univ. des Sciences Sociales, Toulouse

Richard, Jean-Francois, University of Pittsburgh

Abstract: A large class of game theoretic models is represented by the following statistical model: $\xi_i \sim F$ i.i.d. and $x_i = \varphi(\xi_i,F).$ The xi's are observable but the $\xi_i's$ are latent variables. The function $\varphi$ is known and represents the strategy of players. The strategical component is expressed by the dependence of $\xi$ upon F. More sophisticated versions of the model are obtained by introducing exogenous variables in the distribution of $\xi_i$ and/or in the function $\varphi.$ The paper considers parametric and nonparametric estimation of the distribution function F. Issue of particular interest are the identification properties of the model and, in the nonparametric case, the speed of convergence of the estimator of F. The latter depends upon the properties of the game and leads to define a concept of (statistical) order of game.

List of speakers who are nonmembers: None???


next up previous index
Next: asa.business.04 Up: ASA Business and Economic Previous: asa.business.02
David Scott
6/1/1998