Experimenting in economics

This year's economics laureates — economist Vernon Smith of George Mason University and psychologist Daniel Kahneman of Princeton University — have both been pioneers in the field of experimental economics. But the similarity between their contributions ends there.

MONDAY, OCTOBER 14, 2002 

This year's economics laureates — economist Vernon Smith of George Mason University and psychologist Daniel Kahneman of Princeton University — have both been pioneers in the field of experimental economics. But the similarity between their contributions ends there. Vernon Smith has used experiments to test and essentially validate existing theories of market. Daniel Kahneman, on the other hand, has applied experimental data to question the assumption of rationality in the traditional theory of decision-making under uncertainty and constructed an alternative theory.

 To be sure, several economists including his teacher at Harvard, late Edward Chamberlin, preceded Smith in applying the experimental approach to testing theories. But having made the most important early contributions and trained a large number of young researchers, Smith remains the undisputed central figure in the field.

 Smith's most celebrated contribution, made in 1962, tested the validity of the standard perfect competition model. In the simplest form, this model says that in a competitive market, exchanges take place at a price that equates demand and supply. Smith randomly assigned the roles of buyer and seller to his subjects in a so-called double-oral auction for a fictitious good. He then gave each buyer and seller a different reservation price. Each buyer was to buy one unit of the good at a price not to exceed his reservation price and each seller was to sell one unit at a price not below his reservation price. The difference between the subject’s reservation and transaction prices was to be his profit.

 Smith assigned reservation prices $32, 30, 28, 26.... to buyers and $8, 10, 12, 14... to sellers. The sequences implied seven buyers with reservation prices of $20 or more and an equal number of sellers with reservation prices of $20 or less. Thus, if we draw demand and supply curves based on the sequences, they would intersect at the price of $20 and quantity of seven units.

 The subjects knew only their own reservation prices and were forbidden from colluding. They were informed of all bids, transactions and prices, however. Smith, who had all the information, hypothesised that if the exchanges took place at or near $20, perfect competition model would be validated.

 Much to his surprise and contrary to the expectation based on Chamberlin’s (flawed) experiments, Smith found that most transactions were at prices close to $20. Moreover, over successive time periods, the variance of the price declined and it rapidly converged towards the theoretical equilibrium of $20!

 Smith’s subsequent work has focused on testing the role of alternative institutional mechanisms, particularly in the context of the auction theory. He has tested the predictions of alternative auctions such as the English, Dutch, and first- and second-price sealed-bid auctions. He has also initiated the use of the laboratory as a “wind tunnel” (a laboratory set-up used to test prototypes for aircraft) to study alternative institutional mechanisms for deregulation, privatisation, and the provision of public goods.

 Turning to Kahneman, he is credited with both a compelling critique of the traditional theory of decision-making under uncertainty and construction of an alternative theory (with Amos Tversky, deceased in 1996). Based on survey and experimental research, he argued persuasively that in complex decision situations under uncertainty, individuals do not make rational calculations as assumed by the traditional theory. Instead, they rely on heuristic shortcuts or rules of thumb. For example, they would overstate the probability of violent crimes in a city if they personally knew someone who had been assaulted even if they had access to more relevant aggregate statistics.

 Kahneman and Tversky have gone on to offer what they called the “prospect theory” of decision-making under uncertainty. Relying on experimental data, this theory departs from the traditional theory in two key respects. First, individuals are assumed to be sensitive to the way an outcome deviates from the status quo rather than to the absolute level of the outcome. Second, individuals are more averse to losses relative to the status quo than they are partial to gains of the same size.

 Prospect theory is able to explain several regularities that appear anomalous in the traditional theory. For example, it explains the propensity for people to take out expensive small-scale insurance when buying appliances. It also explains the willingness of individuals to drive to a distant store to save a few dollars on a small purchase but not for an equally large discount on an expensive item.