Bad And Good News About The Sealed-Bid Mechanism: Some Experimental Results
This paper surveys a set of experiments conducted by Roy Radner and myself (1989) and Peter Linhart, Radner, and myself (1989a; b), all of which were conducted to investigate the properties of the sealed-bid mechanism, a mechanism used to structure bargaining under situations of incomplete information. What we find is both bad news and good news for the mechanism. The bad news is that the behavior of subjects seems quite sensitive to the parameters of the prior distribution of types used and the number of rounds over which the experiment is run. As such prior distributions become more skewed (in a sense to be defined below), and the number of rounds in the experiment are increased to 75, bidder behavior becomes less linear. Such movements have a regular pattern, however, and this phenomenon is remarkably consistent. The good news is that such behavioral shifts do not seem to interfere with the efficiency of the mechanism. This is of course important for the acceptance of the mechanism in the real world, since if efficiency were to change dramatically as we change the parameters of the environment, adoption would have to proceed on a case-by-case basis. The fact that our results imply robustness of efficiencies indicates that, at least on a practical level, the sealed-bid mechanism may be a viable way to structure bargaining in many large-scale economic organizations. I. The Sealed-Bid Mechanism
- Export
- BibTeX
- Sources
- openalex