My friends Sebastian Rosato and John Schuessler (R+S) recently penned an interesting article on applying realism as a prescriptive foreign-policy theory. Some have attacked their analysis on nonrealist grounds, positing that liberalism is a productive part of U.S. security, disputing their historical case studies, contending that realism plays a greater role in public policy than (whinging) realists imagine and arguing that realists have no special claim to policy rectitude.
This debate is essential to improving both IR theory and American foreign policy. And since I mainly agree with R+S’s arguments on the points just mentioned, I would like to propose an “internal” critique. R+S’s prescriptive project will impoverish realism as a predictive theory of world politics. In fact, structural-realist premises cannot produce a single strategic prescription for all great powers. But don’t stop reading yet—these theoretical complaints yield an important policy conclusion: America should follow a different set of prescriptions than what R+S suggest.
R+S lay out three prescriptions for realist strategy: balance against great powers and strategically important minor powers through military build-ups and alliances; ignore minor powers in unimportant areas; and encourage local coalitions against strategically important minor powers. Three structural-realist assumptions inform these policies: the international system is anarchic; the present and future intentions of states are inherently uncertain; and war outcomes are unpredictable. In such a world, weakness can be deadly, so states must be highly attentive to the balance of power. But because war is costly and risky, security is best obtained by deterring aggressors and avoiding conflict where possible. R+S defend this logic against the alternatives of appeasement, cooperative reassurance, buck passing, bandwagoning and preventive war.
This analysis suffers from theoretical, prescriptive and logical problems. The root cause of these problems is that states often have good reasons to pursue “opportunistic” strategies, primarily buck passing and war.
Theoretically, the structural-realist project is in trouble if R+S are right: theirs is a world without a geopolitical explanation for war. If the rational response to anarchy, uncertainty and risk is to balance and deter, then there is no rational cause for war at the level of the international system—even the rationalist trinity of commitment problems, private information and indivisible spoils require the willful pursuit of aggressive behavior in order to cause war. As those who have pursued R+S’s project in the past concede, this makes war a result of domestic political pathologies. The structural realist dream of using the balance of power to “explain why some historical periods were more conflictual than others” (pp. xxii) falls by the wayside.