These data were developed as part of the NSF Project #1530632 "Bounds Approaches to Empirical Market Design." They have been cleared for public release by the company and are available for research purposes. All personally identifying information has been removed.
Please cite the following paper in any written work that uses this data:
Larsen, Bradley J. (2020)
"The Efficiency of Real-World Bargaining: Evidence from Wholesale Used-Auto
Auctions,"
Review of
Economic Studies,
https://doi.org/10.1093/restud/rdaa007.
A free pdf of the paper and appendices are available on the author's website. The earlier NBER Working Paper version (No. 20431) is found here. A detailed description of the data and a descriptive summary of basic features can be found in the paper and appendices. What follows is a summary.
The data come from six different wholesale used-car auctions, where used-car dealers by cars from other dealers and from large institutions. At these auctions, for each car, the auction house runs a secret-reserve price ascending auction that, when the auction price falls below the reserve price, is followed by alternating-offer bargaining between the high bidder and the seller.
The data consist of several hundred thousand observations. A single observation in the data corresponds to a single run of the auction-plus-bargaining mechanism. An observation contains information about the car (make, model, year, trim, mileage, odometer reading, etc.), the characteristics of the sale (location, timing, etc.), the auction (reserve price, auction price, etc.), and the bargaining (all offers/counteroffers, the final outcome, etc.).
The zip folder above contains all of the data (in Stata and csv formats) and code (Matlab and Stata). The README file in the zip folder contains information on the code and contains a codebook describing all of the variables.
Questions may be sent to data@nber.org or to the author.