DMCA
Identification and Estimation in Highway Procurement Auctions under Unobserved Auction Heterogeneity (2004)
| Citations: | 48 - 1 self |
BibTeX
@TECHREPORT{Krasnokutskaya04identificationand,
author = {Elena Krasnokutskaya},
title = {Identification and Estimation in Highway Procurement Auctions under Unobserved Auction Heterogeneity},
institution = {},
year = {2004}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
The accurate assessment of participants’ private information may critically affect policy recommendations in auction markets. In many auction environments estimation of the private information distribution may be complicated by the presence of unobserved heterogeneity. This problem arises when some of the information available to all bidders at the time of the auction is subsequently not observed by the researcher. This paper develops a semi-parametric method that allows a researcher to uncover the distribution of bidders’ private information in a standard First-Price procurement auction when unobserved auction heterogeneity is present. Sufficient identification conditions are derived and a two-stage estimation procedure to recover bidders’ private information is developed. The procedure is applied to data from Michigan highway procurement auctions and compared to the estimation procedures traditionally used in the context of highway procurement auctions. The estimation results suggest that ignoring unobserved auction heterogeneity is likely to result in substantially biased estimates and may lead to erroneous policy recommendations.
Keyphrases
unobserved auction heterogeneity highway procurement auction bidder private information standard first-price procurement auction sufficient identification condition michigan highway procurement auction unobserved heterogeneity semi-parametric method auction market policy recommendation private information distribution two-stage estimation procedure participant private information estimation result erroneous policy recommendation accurate assessment many auction environment estimation
