A Note on 'Experimental Auction Markets and the Walrasian Hypothesis'

This note serves to correct an erroneous inference regarding price dynamics and to graphically illustrate the importance of model specification in the context of a very simple and fascinating structure. In an earlier JPE article, Vernon Smith concluded that excess supply by itself was an unreliable predictor of the speed of price adjustment. On the basis of regression procedures applied to experimental data he found that the potential rent to be captured exerts the dominant influence. Two alternative statistical procedures, a Tobit specification and a nonparametric test, dramatically deny this inference. Excess supply dominates excess rent as a predictor of the rate of adjustment, but in fact neither Hypothesis adequately captures the random behavior of price movement.