Recent experimental evidence appears to reject simple Nash Equilibrium models of bidding behavior in First-Price auctions. We present a methodological critique of this evidence. Existing tests have concentrated on deviations of subjects from predictions in the message space of the auction: bid deviations. We suggest that it is more natural to evaluate subject behavior in expected payoff space. We conclude that the evidence against the simple models is not significant enough to warrant their rejection.

In the supply of public goods, far less free-riding actually occurs than traditional theory predicts. As one explanation, the social composition function(SCF), which aggregates individual contributions into an available social total,may not always take the standard summation form. Theoretical considerations indicate that free-riding should be least for an SCF of the weakest-link type but greatest for the best-shot type. Using a sequential protocol, our experiments strongly confirmed theoretical anticipations under all three types of SCF. Even under the more onerous sealed-bid (simultaneous play) protocol, the experimental subjects were able to make some partial progress toward the theoretical ideal.

Non-cooperative bidding theory for sealed-bid auctions generally implies testable predictions that are conditioned on the risk attitudes of agents. Received laboratory experiments that purport to test this theory do not generally control for the risk attitudes of subjects. Those experiments exhibit behavior inconsistent with popular bidding models that assume that agents have the same aversion to risk or are all risk neutral. This paper constructs an explicit Bayesian prior distribution for the risk attitudes of experimental subjects and reconsiders the experimental results. It finds that observed bidding behavior is still inconsistent with the Nash predictions when explicit prior weights are attached to alternative assumptions about subject risk attitudes. Thus one cannot account for observed bidding anomalies by appealing to uncontrolled nuisance variables such as risk attitudes. Non-cooperative bidding theory for sealed-bid auctions generally implies testable predictions that are conditioned on the risk attitudes of agents. Archetypical of this result is the Nash Equilibrium prediction for First Price auctions for an object that is valued by agents in an independent and private manner. Received laboratory experiments that purport to test this theory do not generally control for the risk attitudes of subjects. Those experiments exhibit behavior inconsistent with popular bidding models that assume that agents have the same aversion to riskor are all risk neutral. In this paper we construct an explicit prior distrlbution for the risk attitudes of experimental subjects and reconsider the experimental results. We find that observed bidding behavior is indeed consistent with the Nash predictions when explicit prior weights are attached to alternative assumptions about subject risk aversion. However, when one allows for risk loving subjects as well, observed behavior is inconsistent with Nash predictions. Thus one cannot account for observed bidding anomalies by appealing to uncontrolled nuisance variables such as risk attitudes. In section I we consider a specific Nash Equilibrium (NE) bidding model due to Cox, Roberson and Smith (1982) and Cox, Smith and Walker (1988) that clearly illustrates the risk-sensitivity of the theoretical predictions. In section II we provide independent evidence of the risk attitudes of experimental subjects in a test for risk attitudes developed by Harrison (1986a). This evidence allows us to construct an explicit prior probability density function over the coefficient of (constant relative) risk attitudes employed in the specific bidding model of section I. In section III we reconsider the evidence from the First Price (FP) experiments reported in Cox, Roberson and Smith (1982) and Cox, Smith and Walker (1983a, 1983b). I. A Specific Bidding Model Cox, Roberson and Smith (1982), hereafter CRS, present a model based on a power function utility specification for agent i: Ui(y) y= (1) Received for publication August 31, 1987. Revision accepted for publication December 18, 1989. * University of South Carolina. I am grateful to two anonymous referees for helpful comments, although they are not responsible for my conclusions.

Economic theory been under severe attack in recent years. The source of this attack been the observation of apparently robust behavioral in decisions that experimental subjects make in controlled environments. The implication of these observations drawn by some is that many of the fundamental tenets of economic theory are systematically misleading as a descriptive model of human behavior. Many alternative models of individual and group behavior have been proposed which can account for some or all of the apparent anomalies. In the Theory and Misbehavior of FirstPrice Auctions (Harrison, 1989), I argued that the effort to extend or generalize received auction theory as a response to such was misdirected. Specifically, I argued that the observed in the experiments in question may simply reflect the failure of the experiment to meet widely accepted sufficient conditions for a valid controlled experiment proposed by Vernon Smith (1982 pp. 930-9). The result of this failure is simply that the opportunity cost of in these experiments is, by any reasonable standard, minuscule. Observed anomalies may then not be at all: they reflect theoretically consistent behavior under conditions where misbehavior is virtually costless. My critique is quite general in going well beyond auction theory and sealed-bid experiments. Perhaps for this reason there is some truth in the assessment of John D. Hey (1991 p. 195) that it has stirred the passions of the experimental community throughout America. The sad corollary of that assessment, however, is that experimentalists must be a pretty dull lot if such modest concerns as mine stir their passions. Section I restates the payoff-dominance critique in general terms to introduce the nonexperimentalist to the main issues. To address some of the issues raised by my critics, Section II contains a detailed numerical example of an important experimental procedure for eliciting the certainty-equivalent of any lottery that uses the G. M. Becker et al. (1964) procedure. In Section III, the generality of the problem is briefly catalogued, so as to emphasize that this is a debate over much broader methodological matters than sealed-bid auction experiments. Section IV addresses directly some of the specific comments of my critics. Section V identifies a number of qualifications to my critique of existing experimental practice.

A fundamental difficulty in devising any test of general efficient market hypotheses is the specification of the relevant public and private information sets of traders. Without a consensus as to reasonable empirical specifications, tests of the hypotheses remain minimal or controversial. In this study we examine the efficiency of experimental asset markets in which one may unambiguously identify the public and private information sets of traders. We focus on how market efficiency is affected by the presence of futures markets in settings that incorporate different types of uncertainty and inside information. Our experimental results support four conclusions: (1) market outcomes tend to evolve toward strong-form informationally efficient equilibria, whether or not futures markets and/or event uncertainty are present; (2) the presence of futures markets clearly stabilizes spot prices; (3) the presence of futures markets tends to speed the evolution of asset markets to more efficient equilibria where there is event uncertainty; and (4) futures markets promote the "leakage" of inside information, with strong-form predictions outperforming semi-strong-form predictions.

In the supply of public goods, far less free-riding actually occurs than traditional theory predicts. As one explanation, the social composition function(SCF), which aggregates individual contributions into an available social total,may not always take the standard summation form. Theoretical considerations indicate that free-riding should be least for an SCF of the weakest-link type but greatest for the best-shot type. Using a sequential protocol, our experiments strongly confirmed theoretical anticipations under all three types of SCF. Even under the more onerous sealed-bid (simultaneous play) protocol, the experimental subjects were able to make some partial progress toward the theoretical ideal.

The use of dichotomous choice (DC) methods has become increasingly common in applications of the contingent-valuation method (CVM)1 to elicit the that an individual might have for nonmarket environmental goods.2 This hypothetical DC method involves a subject responding yes or no to a hypothetical question that asks whether or not he would be willing to make a commitment to pay some stated amount contingent upon the provision of an environmental good. The growing use of this method is primarily based on the assumption that the method yields incentive-compatible results. This implies that subjects will answer the CVM's hypothetical question in the same way as they would answer an identical question asking for a real economic commitment and that, therefore, the hypothetical DC method will result in accurate estimates of true willingness to pay. Explicit or implicit acceptance of this assumption is seen in a number of recent studies. For example, the use of the DC method in CVM studies is strongly recommended by a panel3 convened by the National Oceanic and Atmospheric Administration (NOAA) of the United States Department of Commerce to examine the use of hypothetical CVM survey questions (see NOAA, 1993 pp. 4608, 4608, 4612). The hypothetical DC method has been used by the Attorney General of the State of Alaska in a major application of the CVM to assess damages caused by the Exxon Valdez oil spill of 1989 (see Richard T. Carson et al., 1992). A major CVM study of potential environmental damages due to proposed mining activity in the Kakadu Conservation Zone of Australia employed the DC method with a similar rationale (see David Imber et al., 1991 p. vi). It is clear that if a subject perceives that his expected utility is affected by the possibility of the good actually being provided he has no incentive to misrepresent. We can presume that in an application of a real DC method, where payment and provision of * Cummings: Policy Research Center, College of Business Administration, Georgia State University, Atlanta, GA 30303-3083; Harrison and Rutstr6m: Department of Economics, College of Business Administration, University of South Carolina, Columbia, SC 29208. We are grateful to Peter Bohm, Bengt Kristr6m, and three referees for helpful comments. Ashley Abbott, Lloyd Brown, Colin Day, Tanga McDaniel, Helen Neill, and Melonie Williams provided excellent research assistance. We acknowledge financial support provided by the State of New Mexico's Waste Management Education and Research Consortium and Resources for the Future. We retain responsibility for all errors. 'For a critical review of the debate over the CVM, see Cummings and Harrison (1994) 2Homegrown is a term primarily used in experimental economics. It refers to a subject's value that is independent of the value which an experimenter might induce for the good (see Vernon L. Smith, 1976). The idea is that homegrown values are those that the subject brings to an experiment. 3Consisting of Kenneth Arrow (Co-chair), Robert Solow (Co-chair), Paul R. Portney, Edward E. Leamer, Roy Radner, and Howard Schuman.