Theory and Misbehavior of First-price Auctions: Comment

Abstract

Glenn Harrison's recent (1989) article in this Review has provoked the loudest debate yet heard among experimental economists. His critique of the James Cox, Vernon Smith, and James Walker (1983a,b, 1985a,b, 1988; henceforth CSW) laboratory studies of first-price sealed-bid auctions with independent and private values, in my view, raises a fundamental methodological issue that is well worth debating, and the CSW research program deserves further discussion. In this short comment I confine myself to what I see as the most important issues and do not try to address all evidence relevant to the ongoing debate. First, I summarize Harrison's central argument very briefly. After pointing out two shortcomings in Harrison's specific analysis of CSW, I propose a methodological reconciliation. Finally (and with some trepidation since my own experience with the auction data is relatively slight), I suggest an alternative approach to understanding the anomalies discovered by CSW.