GAME THEORY AS A PART OF EMPIRICAL ECONOMICS

There is something slightly madcap in agreeing to make a hundred year prophecy about a field of study less than fifty years old, particularly a field that has undergone considerable evolution in that time. Yet this is the situation of game theory. Although it has antecedents going back much further (e.g. in the work of Cournot, Edgeworth and Zeuthen), game theory did not become a coherent field until the publication in 1944 of von Neumann and Morgenstern's Theory of Games and Economic Behavior. And many of the extensions and reformulations that shaped modern game theory came only in the I950S and 6os, in the work of Aumann, Harsanyi, Nash, Shapley, Selten, and others. I will also speculate about the future of experimental economics, which is one of the tools but by no means the only one that I anticipate will play an important role in helping game theory bridge the gap between the study of ideally rational behaviour and the study of actual behaviour. Although it too has older antecedents, experimental economics is also a fairly new line of work, having originated more or less contemporaneously with game theory. Indeed, many of the earliest experimental economists are today known primarily as distinguished game theorists, and were drawn to experimentation by the chance to test game theoretic predictions, and observe unpredicted behaviour, in a controlled environment (see e.g. the experimental work in the I950S and 6os of Maschler, Nash, Schelling, Shubik, and Selten).1 Since the safest part of a long term forecast is the far future, let me state at the outset that I am cautiously optimistic that, a hundred years from now, game theory will have become the backbone of a kind of micro-economic engineering that will have roughly the relation to the economic theory and laboratory experimentation of the time that chemical engineering has to chemical theory and bench chemistry. Game theory is, after all, the part of economic theory that focuses not merely on the strategic behaviour of individuals in economic environments, but also on other issues that will be critical in the design of economic institutions, such as how information is distributed (e.g. Harsanyi, I967-68; Aumann, 1976), the influence of players' expectations and beliefs (e.g. Kreps and Wilson, I982), and the tension between equilibrium and efficiency (e.g. Myerson and Satterthwaite, i983). And game theory has already achieved important insights into issues such as the design of contracts and allocation mechanisms which take into account the sometimes counterintuitive ways in which individual incentives operate in environments having decision makers with different information and objectives.