The 3:1 Rule, the Adaptive Dynamic Model, and the Future of Security Studies

The 3:l Rule, the Adaptive Dynamic Model, and the Future o f SecurityStudies Joshua M . Epstein I T h i s article is part of a continuing dialogue on the methodology of conventional balance assessment . The immediate exchange began in the Spring 1988issue of this journal, when I raised questions concerning a variety of methods, notably the 3:1 rule employed by John Mearsheimer.’ It is essential to be specific. My Spring 1988 article responded to the 3:l rule as Mearsheimer stated it in 1982: “An attack requires more than a 3:l advantage on each main axis to succeed.”2 In his present article, Mearsheimer seeks to rebut my criticisms of the rule, and to establish that my own, very different, methodology is seriously flawed. Each attempt is unsuccessful. This relatively narrow debate has broader implications: If conventional balance assessment is ever to approach the status of a science, certain minimum standards will have to be met. Terms must be defined with sufficient clarity that hypotheses are testable. And the hypotheses themselves must take as bedrock the indisputable datum that warfare is a dynamic process. I do not believe that the 3:l rule, as articulated by Mearsheimer, meets either standard. And, for that reason, among many others, I do not believe it offers a promising approach to the practical problem of conventional balance assessment, or a promising avenue of scientific inquiry to the field of security studies. The 3:l Rule: In What Units and Under What Circumstances? Mearsheimer writes, “Epstein’s case against the rule rests on misuse of a historical data base which is itself f l a ~ e d . “ ~ Whether the data base is flawed, While the author bears sole responsibility for all views expressed here, he is grateful to John D. Steinbruner, Bruce G. Blair, Ethan Gutmann, and Shimon Avish for their contributions. Ioshua M . Epstein is a Senior Fellow at the Brookings Institution and Visiting Lecturer in Public and International Affairs at the Woodrow Wilson School, Princeton University. He is currently writing a book on the European conventional balance and conventional arms control. 1. Joshua M. Epstein, “DynamicAnalysis and the Conventional Balancein Europe,” International Security, Vol. 12, No. 4 (Spring 1988). 2. John J. Mearsheimer, ”Why the Soviets Can’t Win Quickly in Central Europe,” International Security, Vol. 7, No. 1 (Summer 1982),pp. 16-17. See also p. 15 and p. 19 n. 33. 3. John J. Mearsheimer, ”Assessing the Conventional Balance: The 3:l Rule and its Critics,” International Security, Spring 1989 (Vol. 13, No. 4) 0 1989 by the President and Fellows of Harvard College and of the Massachusetts Institute of Technology. 90 The Future of Securify Studies I 91 whether I misused the data base, are matters to which we will turn below. Here, it is critical to emphasize that my central objection-which is not even mentioned by Mearsheimer-did not ”rest on” historical data at all.4It was instead that, as Mearsheimer uses it, the 3:1 rule is too ill-defined to permit an empirical test. That is different from saying it is well-defined and wrong. Here is what I wrote: The first question is, simply, 3:1 of what? In what units is this rule supposed to apply? . . . To make the claim clear and establish the historical regularity that is asserted, the rule’s proponents must: (a) specify the units in which this (or any other) magic ratio is to be measured; (b) go back and explicitly rescore a respectable sample of historical breakthrough operations in those units; and (c) demonstrate the statistical relationship they claim.5 Mearsheimer seems to believe he has addressed this problem of units by saying that “the force ratio should be expressed in a measure that captures the combat power inherent in the forces of the opposing sides,” which “is a function of the firepower, self-protection, and mobility of the forces in question ” (p. 63, emphasis mine). But what function? In a particular historical case, does one calculate firepower (F), self-protection (P), and mobility (M) and add them up? Or is ”combat power” the product: F x P x M? Or is it some other...