Can Relaxation of Beliefs Rationalize the Winner’s Curse?: An Experimental Study

We use a second-price common-value auction, the maximal game, to experimentally study whether the Winner’s Curse (WC) can be explained by models which retain best-response behavior but allow for inconsistent beliefs. In the maximal game, the WC can be rationalized only by a belief that others use weakly-dominated strategies. Yet, we find that the WC is widespread. In addition, we create environments where, regardless of beliefs, there should be a correction of the WC. We find little evidence of such a correction. Overall, our study suggests that the WC, at least in initial periods of play in the laboratory, represents a more fundamental departure from NE than a mere inconsistency of beliefs.