Myths of Empire: Domestic Politics and International Ambition.

Is realism dead? Has it finally succumbed to the theoretical and empirical onslaught to which it has been subjected? If the answer is yes, which theories have taken its place? If the answer is no, what explains its durability?(1)

I argue in this essay that structural realism, qua theory, must be viewed as deeply and perhaps fatally flawed. Yet at the same time, qua paradigm or worldview, it continues to inform the community of international relations scholars. The reason for this apparent paradox is no less sociological than epistemological. Borrowing from Thomas Kuhn, I argue that structural realism will not die as the cornerstone of international relations theory until an alternative is developed that takes its place.(2) In the absence of that alternative, students of world politics will continue to use it as their cornerstone; in an important sense, structural realism continues to define the discipline.

The books under review mark a continuation of past efforts to attack structural realism by emphasizing the domestic sources of international relations. In so doing, each offers unique methodological approaches and an array of fascinating case studies. They raise insightful questions about the conditions under which domestic political and ideological factors can shape foreign policies that lead to such outcomes as overexpansion and war.

But the central questions to be raised in this essay are: to what extent do the works under review contribute to the crucial task of theory building in international relations? Do they offer alternative theories of world politics that promise to supplant realism? If not, to what extent are the modifications provided generalizable beyond the specific cases analyzed? Or, instead, have these and other works simply led us into a period of theoretical "crisis," in which the discipline finds itself dissatisfied with existing theories but as yet unable to construct new ones?(3)

That the books under review seek to challenge realism on its "home ground" of national security cannot be doubted. Bruce Bueno de Mesquita and David Lalman, for example, state that "Although we set out with no preconceived notions about the relative merits of ... realpolitik and domestic interpretations [of state behavior] ... the logic and evidence developed in this book provide a foundation for the claim that a perspective that is attentive to the domestic origins of foreign policy demands gives a richer and empirically more reliable representation of foreign affairs than a realist emphasis."(4) Richard Rosecrance and Arthur Stein write: "The study of grand strategy, which deals with what influences and determines nations' policy choices for war and peace, is an ideal arena in which to examine 'realist' approaches. It is, after all, the realm in which countries should be most expected to follow realist imperatives.... And yet the findings of this volume suggest that this ... response to security stimuli does not always and may not even usually occur."(5) And for Jack Snyder, "recent exponents of Realism in international relations have been wrong in looking exclusively to states as the irreducible atoms whose power and interests are to be assessed ... domestic pressures often outweigh international ones in the calculations of national leaders."(6)

In mounting their attack, the authors highlight not only the logical flaws contained in realism but also its inability to explain (much less predict) many of the outcomes that are of greatest interest to contemporary scholars, including the end of the cold war and the repeated failures of states to balance against threatening powers.(7) That these books, which are all within the field of security studies, should point to realism's weaknesses is especially significant, for it is in that domain of international relations where its legacy probably remains the strongest.

Yet I conclude that these works are unlikely to produce a "paradigm change" in international relations scholarship, for two reasons. First, none of them goes beyond its case study material to produce a generalizable alternative theory; indeed, as we will see, Snyder explicitly omits whole types of states from his analysis. Second, none of these works provides even a decisive modification of structural realism, if by such a modification we mean the generation of alternative hypotheses from existing realist assumptions (i.e., states as primary actors; anarchy as the international condition; state behavior as rational) or from a changed set of assumptions.

Because of this failure, those scholars who believe that realism is dead should prepare themselves for a shock. Indeed, a notable example is Bruce Russett, who ironically borrowed Mark Twain's famous remark about the premature report of death in an article about American decline.(8) Reviewing War and Reason among other books, Russett stated that "these broadsides leave a sinking hulk" where realism "once ruled the theoretical seas."(9) And he quotes Zeev Maoz, who has written that system theories are "useless theoretically," "empirically meaningless," and "normatively objectionable."(10) What these critiques overlook is that scholars will continue to cling to realist planks until they are rescued.

In the following section, I briefly discuss what we might reasonably expect from a theory of world politics, whether it be domestically or systemically oriented. I then examine each book, highlighting both the insights and problems associated with the argumentation. I conclude by examining the status of structural realism in light of these critiques and providing some suggestions for how scholars might proceed if they seek to develop a progressive research program.

In search of theory

If theories of the domestic sources of world politics are to replace or significantly modify structural realism as the cornerstone of international relations theory, they must be able not only to falsify it but also to articulate an explicit model of how a given set of domestic factors can produce particular international outcomes, the most important being war and peace. In order to take that next crucial step, they must be able to conceptualize the causal mechanisms at work that lead from the domestic factors that have been identified as critical (democratic regime type, for example) to foreign policies (for example, free trade) and to specific international outcomes (peace, for example). In other words, they must offer explanations of international relations that either work from the "inside out" or, if their more modest task is to modify structural realism, that specify the domestic process by which uncertain systemic pressures are translated into particular policy responses. Further, the theories must be generalizable beyond the case studies treated if they are to have any hope of exercising a decisive influence on international relations theory.

The need for theories that incorporate the systemic and unit levels in an operational way has, of course, become an old theme within the discipline. For instance, writing in these pages some fifteen years ago, Peter Gourevitch asked: "Is the traditional distinction between international relations and domestic politics dead? Perhaps. Asking the question presupposes that it once fit reality, which is dubious."(11) He urged scholars to examine domestic and international politics "as a whole," and this continues to be a major, if elusive, goal.(12)

In the interim, the theoretical issue at stake concerns the explanatory mileage we get from adopting one particular model over another, no matter how closely its assumptions fit reality. Gourevitch noted that political scientists had adopted two major approaches for exploring the problem of how states behave with respect to their external environment. The first, most powerfully associated with but not limited to structural realism, privileges the autonomous nature of the anarchic international system and focuses on the pressures that it places on every nation-state. In this view, the foreign policies of states are best explained as a rational response to these external pressures. To be sure, the reactions of states to the international system are processed through some form of domestic intermediation. Political structures, bureaucracies, and ideas and beliefs can all play a role in shaping policy, and structural explanations often provide no more than a starting point for analyzing state behavior.(13) Still, for those scholars who accept the power of a structural paradigm, the most parsimonious explanation of a country's foreign policy behavior is not found in the psychology of its leaders, its regime type, or its political ideology; instead, it is located in the relative position of the state in the anarchic international system, as measured by its capabilities - its capacity for independent action.

The second approach rejects or downplays the utility of system-level theorizing. It argues either that there is no objective international system with an independent existence or that systemic pressures are so weak and uncertain that they are indeterminate with respect to the foreign policy choices that states make and the outcomes of their international interactions. In order to understand state behavior, therefore, scholars must reject the "billiard ball" model of structural realism and begin their exploration inside the "black box" of domestic politics. The causal logic of this explanation thus begins with what is happening inside a particular unit.

Two brief examples, one theoretical, the other historical, seemingly provide strong support to the inside-out framework. From a theoretical perspective, the strongest inside-out alternative to structural realism would seem to be the theory of the democratic peace. A cottage industry has emerged in recent years that seeks to explain why states with democratic regimes do not go to war with one another. The word "theory," however, is a misnomer in this context, for proponents of this school have accumulated observations rather than any sustained causal logic.(14) Further, even if the causal logic were established, democratic peace theory would still have a difficult time explaining the outcomes of conflicts among democratic states in such issue-areas as trade and finance. In economic relations among democracies, threats and coercion - the stuff of realist theory - are ever-present as part of international negotiations, while the value of international institutions in mitigating these conflicts may be overrated.(15)

But these criticisms aside, we could hypothesize with some confidence that a world composed solely of liberal states would not confront the same type of security dilemma that exists when these same liberal states must coexist with authoritarian regimes, because liberal states view one another …