Playing the games

James Morrow's superb book provides the best account of ideas from game theory tailored to the interests of political scientists, which is currently available. It is ideally suited to form the basis of a serious graduate course in positive political economy, such as are increasingly being offered in American PhD programmes in political science.

Morrow's book proceeds from the ground up, assuming very little prior knowledge of the literature. It would, however, certainly speed things along if a prospective reader had taken a masters level course in micro-economics. Notwithstanding, all basic concepts are explained in the book, including basic ideas such as Bayes rule and utility theory. There is also a terse mathematical appendix. The applications developed in the book are explicitly geared towards political scientists. They are, however, somewhat oriented towards examples from the US - a reflection of where most of the work in this mode has been done to date.

An important example which runs through Morrow's book concerns the study of deterrence in international politics. One country (the challenger) is deciding whether to invade another (the defender). The challenger does not know the resolve of the defender (and possibly vice versa). The defender may also choose how vehemently to resist. Then how can we determine when and whether a challenger will proceed with invasion and how much resistance he will face? Formulating this problem as a game and determining the conditions under which different things might happen is a challenging intellectual exercise. Whether this is the way in which we would choose to model Saddam Hussein's decision to invade Kuwait is moot, but game theory provides the grammar for the argument.

A second useful illustration concerns the application of bargaining theory to legislatures. Suppose that a proposer in a legislature can suggest a piece of legislation. If it does not pass, then another individual may be given the power to propose instead. Again the legislature votes yes or no. With a no vote, the process again proceeds to a different proposal and so forth over some specified period of time. What is the outcome likely to be? Again, setting up and studying the problem is challenging. Morrow gives an account of what we understand about such problems and a sense of how the rules and institutions in legislative decision making may affect the policies that are passed. Questions of modifying procedures frequently emerge: having an analytical base from which to study these issues allows us profitably to consider some interesting "what if" questions.

Morrow's book is basically a textbook which focuses on technique. It avoids philosophising about methodology in political science. This may be a source of disappointment, especially to sceptics requiring the author to justify why they should read a book which is fairly hard going for those without a mathematical background. The models used in much game theory, when applied to economic situations yield very little in the way of immediately testable results. That, however, may be seen as a reminder that invoking rationality in individuals does not of itself produce anything very determinate. It is necessary, at a minimum, to supplement models with specific institutions and protocols.

Given the current state of political science in the UK, I suspect that it will be some while before Morrow's book figures prominently on graduate reading lists. It seems more likely that economists who become interested in political science will be the ones to promote use of the book. This would follow the trend set in the US. However, for political scientists who are keen to know how game theory might advance their understanding of political phenomena, there is no better place to begin than by reading this book.

Timothy Besley, currently at Princeton University, has been appointed professor of economics, London School of Economics.