Key points are not available for this paper at this time.
Direct experimental tests of expected utility theory (EUT), in which subjects are asked to choose among alternative gambles, or to make judgements as to their willingness to pay (WTP), and/or willingness to accept (WTA) payment for a gamble, have not been kind to EUT. As noted in the survey by Paul Slovic and Sarah Lichtenstein (1983), the results of these interrogations are remarkably consistent in a wide variety of contexts and are robust under examinations designed to determine the effect of monetary incentives, experience, and other factors that might have accounted for the discrepency between subject responses and the predictions of EUT. On the other hand, experimental studies of individual and market behavior based upon EUT models of market decision making have yielded results showing high consistency with the predictions of these models (see the references in Smith, 1985). Are individual revealed preferences in some market contexts more likely to be rational (consistent with EUT) than individual responses to choices among alternative prospects? Several studies designed to solicit WTP and WTA responses for a variety of goods have found a wide disparity between the buying price and selling price measures of individual value (see the study and the citations therein by Jack Knetsch and J. A. Sinden, 1984; hereafter K-S). Values of WTA obtained in this way are frequently an order of magnitude greater than values of WTP although theoretically WTP and WTA differ by no more than a presumed small income effect. These results should not be dismissed by economists on the grounds of poor subject motivation because the experiments include some (see those in K-S) that have carefully introduced actual monetary payments and cash compensations and have not relied on hypothetical choices. These results are explained by K-S (and by most of the authors of such studies) in terms of the Daniel Kahneman and Amos Tversky (1982) framing paradigm in which people reveal that they are less willing to spend wealth which they consider to be part of their endowment than wealth not so considered. We would emphasize that such behavior is irrational only in the narrow sense of EUT as a behavioral hypothesis which may not only be a poor predictor of individual choice, but may not be a satisfactory guide to action. For example, the differential treatment of wealth that has become part of one's endowment may have important survival value which is imprinted in decision behavior. However, K-S (p. 508, fn. 3) report one, perhaps significant, experiment in which they do not get the usual WTA-WTP disparity. In this experiment, cash payments and offers for a lunch made to respondents entering an office cafeteria showed a much smaller, statistically insignificant, disparity. In this case, the respondents were about to participate in a familiar market for the commodity being evaluated. On this same theme an important new study by Don Coursey et al. (1984) uses a repetitive series of Vickrey (second price) auctions to determine market WTA and WTP prices for entitlements to an unfamiliar item, which they compare with hypothetical measures of WTA and WTP. They find that although individual bids in the first of a sequence of Vickrey auctions show a large WTA-WTP disparity, ending bids, after a *Graduate Student in Finance, University of Pennsylvania, Philadelphia, PA 19104; Professor of Economics, University of Arizona, Tucson, AZ 85721; Assistant Professor of Economics, Indiana.University, Bloomington, IN 47401. Research support from the National Science Foundation is gratefully acknowledged.
Knez et al. (Mon,) studied this question.