Procurement auctions with entry and uncertain number of actual bidders: Theory, structural inference, and an application.

详细信息   

• 作者：Zheng ; Xiaoyong.
• 学历：Doctor
• 年：2005
• 导师：Li, Tong
• 毕业院校：Indiana University
• 专业：Economics, General.
• ISBN：0542269244
• CBH：3185404
• Country：USA
• 语种：English
• FileSize：8176756
• Pages：198

文摘

Motivated by several interesting features of the highway mowing auction data from Texas Department of Transportation (TDoT), this dissertation proposes a two-stage procurement auction model with endogenous entry and uncertain number of actual bidders. Our entry and bidding models provide several interesting implications. Most importantly, for the first time, we show that even within an independent private value paradigm, as the number of potential bidders increases, bidders' equilibrium bidding behavior may become less aggressive because the "entry effect" is always positive and may dominate the negative "competition effect." We also show that it is possible that the relationship between the expected winning bid and the number of potential bidders is non-monotone decreasing as well. We then propose a semiparametric Bayesian method to estimate entry and bidding models jointly taking into account unobserved auction heterogeneity using a fully structural approach. Our results indicate significant unobserved auction heterogeneity. Also, the entry costs are a significant part of bidders' submitted bids. Furthermore, we quantify the "entry effect" and the "competition effect" with regard to the expected winning bid, and find that as the number of potential bidders increases, the procurement cost increases as well in our data. Hence policies that encourage more potential bidders might cause the government's procurement costs to increase. Beyond estimation, we further adapt the semiparametric Bayesian model selection criteria developed by Chib and Jeliazkov (2001) and Basu and Chib (2003), and then apply them to contrast the model estimated in this dissertation with other competing models such as the auction model without entry. The results show strong evidence that prefer the model we developed in this dissertation to the standard auction model.