The long peace, the end of the cold war, and the failure of realism. (Symposium: The End of the Cold War and Theories of International Relations)

Nation-states are engaged in a never-ending struggle to improve or preserve their relative power positions.

--Robert Gilpin

The greatness of the idea of European integration on democratic foundations consists in its capacity to overcome the old Herderian idea of the nation state as the highest expression of national life.

--Vaclav Havel

The dramatic events of 1989-91 are widely recognized to have usered in a new era in international relations. Prominent realists maintain that a shift is under way in the international system from bi-to multipolarity. Some of them predict that a multipolar world will be more conflictual and urge states to acquire nuclear weapons.(1) Realists and neorealists alike argue that superpower behavior since 1945 is consistent with their theories. I contend that it sharply contradicts these theories.

I develop my argument by looking at realist explanations for three of the more important international developments of the last half century: the "long peace" between the superpowers, the Soviet Union's renunciation of its empire and leading role as a superpower, and the post-cold war transformation of the international system. Realist theories at the international level address the first and third of these developments, and realist theories at the unit level have made an ex post facto attempt to account for the second. The weakness of these explanations raises serious problems for the realist paradigm.

Evaluating realism

The realist paradigm is based on the core assumption that anarchy is the defining characteristic of the international system. Anarchy comples states to make security their paramount concern and to seek to increase power as against other values. Power is defined as capability relative to other states. Drawing on the core assumption of anarchy and the "self-help" system it allegedly engenders, realists have advanced a variety of sometimes contradictory propositions about international relations. Realists disagree, among other things, about the relative war-proneness and stability of multipolar versus bipolar international systems, the importance and consequences of nuclear weapons, and more fundamentally about the weight of power as an explanation of state behavior. The competing predictions of realist theories make realism difficult to falsify. Almost any outcome can be made consistent with some variant of realist theory.

Operationalization

Testable theories require careful conceptual and operational definitions of their dependent and independent variables. These definitions must be conceptually precise and stipulate how the variables are to be measured or their presence determined. Realist theories do not meet these conditions. They do not share common definitions of the core concepts they use to construct variables. Individual definitions of national interest, power, balance of power, and polarity allow for an unacceptably wide range of conceptual and operational meaning and make it difficult to test realist propositions against evidence drawn from specific cases. Neorealism, the most scientifically self-conscious of realist theories, is particularly inadequate in this regard, as my critique of its explanation for the long peace will demonstrate.

Specification

Theories must stipulate the conditions associated with predicted outcomes. It these conditions are met but repeatedly fail to produce the predicted outcomes, the theories can be rejected. If predicted and unpredicted outcomes occur, the theories are inadequately specified.

Power transition theories comprise the branch of realism that analyzes great power responses to decline. These theories failed to envisage the possibility of a peaceful accommodation between the two poles of a bipolar system or that one of them would voluntarily relinquish its core sphere of influence to bring about that accommodation. Such an anomalous outcome constitutes strong grounds for rejecting power transition theories. Realists have sought to save their core insights by treating the end of the cold war as a special case and reformulating their propositions to take it into account.(2) Anomalous cases often serve as catalysts for better theory. But as the second part of my critique will show, realist attempts ex post facto to explain Mikhail Gorbachev's reorientation of Soviet foreign policy are neither logically consistent nor empirically persuasive.

Utility

Good theory is based on good assumptions. Realists maintain that their core assumption of anarchy accurately captures the dynamics of the international system and generates powerful explanations of interstate behavior. Some recent literature contends that the assumption of anarchy has no theoretical content and cannot generate useful or testable propositions.(3) I contend that international structure is not determining. Fear of anarchy and its consequences encouraged key international actors to modify their behavior with the goal of changing that structure. The pluralist security community that has developed among the democratic industrial powers is in part the result of this process. This community and the end of the cold war provide evidence that states can escape from the security dilemma.

A critical case?

At the final session of a 1991 conference on international relations theory and the end of the cold war, a prominent participant expressed his dissatisfaction with the proceedings.(4) The end of the cold war, he insisted, was a "mere data point" that could not be used to test or develop theory. However, neorealism drew on a single case of bipolarity to construct its theory. If that case does not fit the theory, it raises serious doubts about the validity of the theory. Other realist theories have cast their empirical nets more widely. The end of the cold war and the ongoing transformation of international relations also raise serious problems for these theories. This essay does not test in a formal sense any of these theories; such tests are precluded by the lack of specification as well as by my own reliance on only a few cases. Rather, it attempts to demonstrate that historical evidence since 1945 contradicts many realist claims and expectations and suggests the need for alternative approaches to the study of international relations.

Realism and the long peace

Security specialists consider it remarkable that the superpowers did not go to war as did rival hegemons of the past. Many realist theories attribute the absence of war to the bipolar nature of the postwar international system, which they consider less war-prone than the multipolar world it replaced. All of them have poorly specified definitions of bipolarity. None of the measures of bipolarity derived from these theories sustains a characterization of the international system as bipolar before the mid-1950s at the earliest.

For the sake of brevity, I will discuss only two realist theories that emphasize the restraining effects of bipolarity, those of Hans Morgenthau and Kenneth Waltz. They are arguably the most influential international relations theories of the cold war era.

Measures of power and polarity

The first edition of Morgenthau's Politics Among Nations coincided with the beginning of the cold war; in that and subsequent editions, Morgenthau worried that the United States and the Soviet Union would stumble into a nuclear war despite their mutual recognition of its destructiveness. For Morgenthau, the long peace was not an analytical puzzle but a desperate hope.(5) Morgenthau believed that postwar international relations was shaped by bipolarity and nuclear weapons. Both were double-edged swords. Bipolarity was "a mechanism that contains in itself the potentialities for unheard-of good as well as for unprecedented evil." It "made the hostile opposition of two gigantic power blocs possible" but also held out the hope of regulating that opposition through an equilibrium of power maintained by moderate competition. Nuclear weapons made leaders more cautious and more insecure. The nuclear arms race reduced international politics to a "primitive spectacle of two giants eying each other with watchful suspicion." Human survival depended on mutual restraint. This was not a function of the polarity of the international system but of the skill and commitment of leaders.(6)

Drawing on Morgenthau's insight that bipolarity had the potential to promote a more stable international order, Waltz built a formal deductive theory of international relations.(7) In an effort to create a parsimonious theory at the system level, he gave explanatory weight to the nature of the system, the number of actors, and the distribution of their capabilities. He downplayed the explanatory power of state attributes, including leadership.

Writing in the late 1970s, Waltz was struck by the seeming stability of the postwar order and the success of the superpowers in defying earlier predictions that the cold war would sooner or later turn hot. He attributed the absence of war to bipolarity, which, he maintained, was less war-prone than multipolarity. Waltz argued that war arose primarily because of miscalculation; states misjudged the relative power or the power and cohesion of opposing coalitions. The latter error was more common because of the difficulty of estimating accurately the power and cohesion of shifting and often unstable coalitions. In a bipolar world, where hegemons rely on their own vastly superior power for their security, coalitions are less important and "uncertainty lessens and calculations are easier to make."(8)

Waltz regarded military technology as a unit attribute and outside his theory. He sought to minimize its consequences and insisted that the "perennial forces of politics" were more important than nuclear weapons in shaping the behavior of nations. Nuclear adversaries "may have stronger incentives to avoid war" than conventionally armed states, but then the United States and the Soviet Union also found it more difficult to learn to live with each other "than more experienced and less ideological adversaries would have."(9)

Waltz's Theory of International Politics has one major dependent variable, the war-proneness of international systems, that is explained by one independent variable, the polarity of the system. The theory resides entirely at the system level: war-proneness is a system property, and polarity is a structural characteristic of the system. Waltz is unyielding in his contention that a theory of international relations should not incorporate variables at the unit level or use system-level properties to predict the behavior of individual units. Bipolarity affects state behavior only indirectly by structuring constraints and incentives for leaders.(10)

Many international relations scholars and historians contend that nuclear weapons have played a far more important role in preserving the peace than Waltz's theory acknowledged. Waltz has come to accept the contention of his critics. In 1981, he upgraded the role of nuclear weapons, arguing that they "have been the second force working for peace in the postwar world."(11) In 1986, he conceded that the introduction of nuclear weapons, a unit-level change, had a system-level effect.(12) In a 1990 essay Waltz went further and argued that "The longest peace yet known has rested on two pillars: bipolarity and nuclear weapons." Nuclear weapons deterred attacks on states' "vital interests"; and "because strategic weapons serve that end and no other, peace has held at the center through almost five postwar decades, while war has frequently raged at the periphery."(13) Waltz reaffirmed this argument in 1993.(14)

Waltz's 1990 essay argued that the international system was undergoing a peaceful transition from bipolarity to multipolarity. Neorealism recognized the possibility of system change--although not peaceful system change--but maintained that multipolar systems were more war-prone. While not rejecting this core proposition of neorealism, Waltz's essay indicated that it was no longer relevant. The long peace would endure because the superpowers possessed nuclear weapons. Waltz was arguing that nuclear weapons, by his definition a unit-level capability, can explain war-proneness, the most important system-level property. Such a "reductionist" argument vitiates the need for a theory of international relations whose principal purpose is to explain war-proneness. This may be why Waltz has subsequently backed away from his characterization of the international system as moving from bipolar to multipolar.

Waltz now insists that the international system remains bipolar even after the breakup of the Soviet Union.(15) His depiction of the post-cold war world as bipolar is strikingly at odds with the views of other prominent realists. More to the point, it cannot be derived from the definition of power in Waltz's Theory of International Politics. Even before the collapse of the Soviet Union, U.S. Defense Department studies showed that Japan, the United States, and Western Europe were steadily increasing their lead over the Soviet Union in the development and application of almost all the technologies critical to military power and performance.(16) Post-Soviet Russia is in a demonstrably weaker position.

What distribution? What capabilities?

Realist definitions of power are imprecise, making it difficult to develop measures of polarity. The most thoughtful treatment of capabilities remains that of Morgenthau. In his chapter on the elements …