The dismal science enters the lab

I took part in the small conference held in Pittsburgh in 1990 where the authors of the various chapters in this handbook presented their first drafts. Copies of these proved invaluable sources for teaching and research during the following two or three years, and I waited impatiently for the final versions to appear. However, as I guess most editors of multiauthored volumes will testify, such a project can only progress at the speed of its slowest contributor. Thus it has taken five years to bring this book to market, during which time several other texts about experimental economics have been published. So is this handbook past its sell-by date? Or has it been worth the wait?

The answer is: a bit of both - although the care that has gone into the book, and the consequent quality of the various contributions, more than compensates for the delay in making the finished product available to the students, teachers and researchers who will surely learn much from it.

It is true that some chapters seem a little dated, with rather few detailed references to work after 1992. But then, as with any other rapidly developing branch of a discipline, no hard copy can hope to be really up to date for long. What is more important is that, first, it should be accessible to those nonspecialists who want to discover what role(s) experiments might play in the understanding of economic behaviour, and who can learn from others' experiments even though they may never conduct an experiment of their own.

It should also be helpful to those who wish to begin to explore for themselves the use of experiments in their own areas of research, and who would welcome guidance about the longest ladders and the slipperiest snakes they are likely to encounter.

And it should be stimulating and provocative for those who are already undertaking experimental work, but who can still benefit from being reminded to think critically about basic principles, reflect on the nature of good and bad practice, and remain alert to the questions and doubts that many researchers have about the "ecological validity" of the whole experimental enterprise.

On the first two counts, this handbook succeeds admirably. On the third, the performance is patchier. But given that the number of practising experimental economists, though growing fast, is still small, the target audience at present must mostly consist of people for whom the first two counts are more important. For such people, this handbook has a very great deal to offer.

Consider an example. Suppose that a student has been attending a standard course examining the issues and problems associated with the provision of public goods. The student is perplexed (or intrigued, or possibly disheartened) by the apparent mismatch between what standard theory predicts about some of these issues and what casual observation and/or intuition suggests is happening in the world around him. But it is hard to test those predictions rigorously with the kind of data available in the nonexperimental world; nor is it easy to explore the relative effectiveness - or uselessness - of different "mechanisms" proposed by theorists to overcome certain problems such as the underprovision of public goods.

However, experimental methods offer the possibility of creating an environment where these predictions and mechanisms can be studied in a more systematic way, and chapter two of the handbook leads the student (and the general reader) through the basic principles involved, in tandem with a critical review of some of the landmark experiments conducted in this area. The concluding paragraphs of the chapter summarise some of the main questions and puzzles that remain, offer some conjectures about these and make some suggestions for future research.

Much the same broad format is used in subsequent chapters, which deal with bargaining and coordination problems, the functioning or malfunctioning of different types of market, and individual decision making. In all cases, the authors try (and generally succeed) to set out clearly the main issues being investigated, describe the key features of experiments intended to address those issues, and offer some comments about the strengths and weaknesses of those designs.

If I have one reservation about this handbook, it concerns the uneven quality of some of the commentary, particularly the general questions about the (undoubted) limitations of experimental methods in economics. But this is a book written principally to demonstrate the considerable scope and potential of economics experiments, and it achieves that objective very well.

Graham Loomes is professor of economics, University of York.