Charlie Plott's letter to The Economist

The Economist

The issues about laboratory experimental economics methods, addressed in your May 8 issue editorial, are not new. They have been examined extensively. The rapid growth in use of these methods is precisely because the implicit questions raised in your editorial have been answered in many ways. The issues are (1) whether or not there are principles of economics that can be observed operating under laboratory conditions and (2) what might be the value of using laboratory methods to study them, if such principles do exist. There is a third issue, which seems to grate against the editor's sense of what constitutes proper science, and that is how the profession as a whole deals with theories that are rejected by laboratory experiments.

Have experimentalists uncovered basic principles of economics? The law of supply and demand might be the best example. This principle, which predicts the price and volume of ultimate market equilibration, works with amazing accuracy under appropriate conditions. The price that is discovered in competitive markets is exactly the one predicted by an application of the competitive model. The nature of price stability and the information content of prices are both actively studied. The findings are robust, not depending on income, culture, education and perhaps not even age (since it seems to work with children). Furthermore, not only do multiple market systems follow the same principles, markets respond to institutional changes in predictable (but not always understood) ways. I think that no other branch of science can claim success in identifying principles that govern something as complex as a multiple market system with human participants. And, there is no need to stop with markets. Game theory and Nash equilibria are demonstrating similar powers of prediction. Circumstances have been identified where the models are stunningly inaccurate, yet even when the inaccuracies occur, experiments have been able to supply explanations from closely related disciplines, such as psychology.

Have the principles been put to valuable use? Experiments have played a central role in several major instances. These include the allocation of the rights to land at major U.S. airports, regulations governing pricing in natural gas pipelines, the Ethyl case in antitrust, the design of the auction mechanism used by the Federal Communications Commission, the architecture of the Regional Clean Air Markets in Southern California, the electric power markets in operation in Southern California, decisions regarding access to public railroad tracks, methods of allocating resources on Space Station Freedom, etc. Many other applications are underway. The editorial carried the implication that laboratory experimental work has no applications and that impression is seriously wrong.

Has the profession acted responsibly when the basic theories are rejected? The preference reversal phenomenon, used as an example in the editorial, was introduced to the economics literature by David Grether and me. The editorial reported accurately. We concluded that the phenomenon is inconsistent with all theories of economic decision making. Has the profession abandon the theories as a result? Of course not and it would be silly to do so. While the theories do have problems, paradoxes do exist, they nevertheless continue to help us understand phenomena of staggering complexity. They continue to help us design and implement very successful policies. Critics of the theories, scholars who relish in pointing out that the theories have been falsified, have produced absolutely nothing that will do a similar job. Much like the editorial, these scholars suppose that the sole purpose of economics is to study individual choice. They ignore the fact that much of economics is about markets.

The editorial did not provide many references for those who would want to pursue the issues. In addition to the excellent reference contained in the editorial, readers should be directed to sources such as Davis and Holt. There are also several other good books and collections of papers. (Douglas D. Davis and Charles A. Holt, Experimental Economics, Princeton University of Press, Princeton New Jersey (1993). Vernon L. Smith (editor), Experimental Economics, Edward Elgar, 1990. John D. Hey and Graham Loomes (editors), Recent Developments in Experimental Economics, Volumes 1 and 2, Edward Elgar, 1993. Daniel Friedman and Shyam Sunder, Experimental Methods: A Primer for Economists, Cambridge University Press, 1994. Elinor Ostrom, Roy Gardner and James Walker, Rules, Games and Common-Pool Resources,The University of Michigan Press, 1994. A new Journal Experimental Economics, Special Issues of Economic Theory.)

It is nice to know that the use of experimental methods in economics caught the eye of The Economist. However, whether or not economics or any other science, is a "proper science" according to the criteria used in the editorial, is probably not an answerable question. Of more interest to the readers are the existence of a laboratory experimental methodology in economics, its limitations, and the areas of potentially successful applications.

Professor Charles R. Plott
Edward S. Harkness Professor of Economics and Political Science
California Institute of Technology
Pasadena, California 91125

626 395-4209 (voice)
626 793-8580 (fax)