Dr. Gideon Akavia
Center for Military Analyses, Israel

February 10, 1999

Offense-defense theory, essentially based on the idea that war is more likely when conquest is easy, holds a prominent position in international relations theory. The Spring 1998 edition of International Security was led by two articles on offense-defense theory, "Offense, Defense, and the Causes of War" by Stephen Van Evera and "What Is the Offense-Defense Balance and How Can We Measure It?" by Charles L. Glaser and Chaim Kaufmann. Despite the prominence of both these articles and offense-defense theory, there are problems with the theory which have not been adequately addressed. Three main points highlight the weaknesses of the theory. First, Van Evera outlines the "effects" of offense dominance, yet several of these effects do not directly or logically follow solely from offense dominance. Second, attempts to improve the definition of the offense-defense balance impoverish the phenomena upon which the theory is based. Finally, one must ask why offense-defense theory is so popular if it is so problematic. These three points will be outlined further below.

In his article, Van Evera lists ten "war-causing effects" of the offense-defense balance. Three of Van Evera’s effects are discussed below. The first of these effects is that when conquest is easy, aggressors win more decisively. Therefore, there is greater incentive to strike first. However, this conclusion does not follow either logically or historically. Thus this is not an effect of the offense-defense balance. Another proposed effect of the offense-defense balance that Van Evera lists is that when conquest is easy, resources are more cumulative. Aggressors are therefore tempted because small gains rapidly turn into large gains through conquest. This consequence also does not follow directly from the offense-defense balance. Finally, Van Evera claims that when conquest is easy, empires are easier to conquer. However, this statement is too simple. The attacker must at some point defend his gains. This means that even if the initial aggressive moves toward conquest are easy, the overall conquest of an empire may not be as simple as Van Evera implies. Rather than there being a single taproot for all of these effects as Van Evera believes, there are actually independent causes for these phenomena. There is a difference between estimating the specific balance and the existence of a general theory of conflict, yet this difference is not recognized by either Van Evera or Glaser and Kaufman. It is impossible to characterize the smaller pieces of the balance with the general theory. For instance, at what level is the offense-defense balance measured? Is it measured through the effectiveness of tank rounds or land warfare?

The second problem with offense-defense theory is that efforts to improve its definition are unacceptable. Glaser and Kaufman try to evaluate the balance by taking the first move advantage out of the equation and waiting for the defender’s response to begin the analysis. However, removing the first moves of the conflict from evaluation is impossible. These are critical moves which the theory should be able to explain. Improving the definition in this case impoverishes the phenomena. Glaser and Kaufman also assume that states act optimally for the purposes of analysis, yet states do not always act optimally. Using assumptions of non-optimal behavior and decision making complicate the analysis, but it is impossible to accurately analyze the balance under the assumption of optimality. Furthermore, as Goddard points out, it is difficult (if not impossible) for the analyst to know what is or was optimal in a given situation. In many situations, the debate over what the optimal decision would have been cannot be resolved. For World War One, for instance, it is only possible to determine the offense-defense balance in certain periods and places. It may be impossible to come to an overall consensus for the war as a whole.

Measuring the offense-defense balance is difficult to do effectively. Although there are certain factors that are said to favor the offense or the defense, the balance itself is seldom so clear. For instance, many analysts agree that mobility favors the offense. Although this may be true in the first stages of the war, the defender also needs mobility in order to effectively counterattack. In many cases, offense is chosen over defense because offense better suits the goals of the state. Germany chose to take the offensive in World War Two because that was the only way for them to achieve the additions of territory that it felt to be crucial to its survival. Thus, an understanding of the offense-defense balance was largely irrelevant in this case. Furthermore, the effectiveness of a combination of capabilities depends on the scenario in which those capabilities will be employed. This leads one to the question of how to choose the standard scenario according to which the offense-defense balance will be determined.

Finally, given the above critiques of offense-defense theory, one must ask why the theory is so popular. There are three main reasons for the affinity towards the theory. First, the theory has mnemonic power. However, the desire for mnemonic power has left out important trends such as the learning which takes place during a war and may cause a continual change in the offense-defense balance. Second, the theory has heuristic power. However, in order to create a ballpark figure for the offense-defense balance, it is necessary to evaluate the creative powers that take place in war as well as to consider more mathematical analytical tools. Third, the theory has rhetorical power. Analytic measures such as the three-to-one rule and analytic procedures are popular because they can be easily employed and discussed.

In summary, offense-defense theory is not a good theory of large-scale war. The offense-defense balance can only be evaluated at a given time and place, not for an entire conflict. Furthermore, the theory cannot describe the creative process that goes on during a war or take account of opponents with initiative. These are serious failings that must be accounted for in order to improve the utility of offense-defense theory as a whole.

Andrea Gabbitas - Rapporteur

Back to Seminar Schedule, Spring 1999