Informed Entry in First-Price Auctions
Vadim Marmer
Professor of Economics
Vancouver School of Economics
University of British Columbia
We study the role of information in first-price procurement auctions with endogenous entry. At the entry stage, bidders do not know their private costs but receive imperfect signals and decide to pay the entry cost and enter based on the signals’ strength. While, under more informative signals, entering bidders tend to be more efficient (can complete the contract at a lower cost), more informative signals may also reduce entry. Therefore, the overall effect of more informative signals on procurement costs is ambiguous. We estimate the model semiparametrically using data on highway maintenance auctions from the Texas Department of Transportation. We find that adjusting the informativeness of signals to the optimal level can reduce the cost of procurement by up to 38% of the engineer’s estimate. Our analysis also takes into account the possibility of auction failure due to insufficient entry. We find that imposing a binding reserve price is a more efficient mechanism to address insufficient entry than canceling contracts when there is only one active bidder. The former mechanism can reduce the cost of procurement by up to 32% of the engineer’s estimate relative to the latter.