October 25th, 2012

Bringing Strategy Back Into IR?

Infinity Journal recently released a brief that notes that war, strategy, and policy are not well-taught in IR courses:

"[M]ost IR courses talk around war, as if it were ancillary, something avoidable, an aberration — so long as war isn’t studied, it isn’t a threat. Unfortunately, most IR degrees do not offer the student a full appreciation for how strategy works or how war extends the reach of policy. Many IR courses do not actually give us a functional understanding of policy."

The current state of separation between strategy and international relations does not have to be terminal. I am not suggesting that IR should consume strategy, but rather that a bridge be built between the two fields of inquiry to facilitate fruitful analysis and scholarship. The study of ends, ways, and means and the conduct of war is useful for anyone studying or practicing international relations, and strategy has much in common with certain aspects of IR.

First, it should be acknowledged that there are certainly many elements of IR and strategy that inherently cohere. There are functional areas of IR such as the study of civil war, deterrence, strategic bombing, or military doctrines that directly take on aspects of strategy. One’s perspective on IR also impacts how you view strategy’s core assumptions. Certainly realists may be more kind to the idea that the nature of war remains consistent across time than those who place stock in the power of norms, institutions, and ideas. Likewise, whether or not you believe in balancing, hegemonic wars, offensive vs. defensive realism, or Graham Allison’s organizational and/or bureaucratic politics model does matter for how you look at the origin and conduct of wars.

At the same time strategy is also cross-paradigm. Social culture, political ideology, domestic politics, and bureaucratic preference are all part of the analysis of strategy. One does not have to be a neorealist to study strategic theory, and in fact some strategists have harshly criticized IR concepts such as offense/defense theory. Colin S. Gray advances a synthesis of liberal and realist ideas in making the case that the Soviet Union and the United States had legitimate security concerns as well as mutually incompatible domestic regimes. Strategic literature often focuses on states but also recognizes that sub-state and/or cross-state actors are meaningful within the international system.

The short answer for why IR and strategy have diverged lies in questions of method. I took a course in Net Assessment from CSBA’s Barry Watts in the fall of 2010 that introduced me to an fascinating rebuttal Watts wrote to Robert Pape’s work on strategic bombing. Using insights from complexity science as well as post-1980s advances in quantitative military analysis, Watts observed that outputs such as the decision of a government to bow to the demands of a military coercer will rarely have easily discernible, linear, and correlate inputs on which a predictive theory could be constructed.

Echoing J.C. Wylie’s dichotomy between cumulative and sequential modes of strategy, Watts observes that Pape’s view of strategic bombing ignored the way effects in one area of operation could have subtle effects on others. For example, the massive diversion of German war production to air defense efforts interfaced with pressures on the Eastern front. Watts also points out that Pape’s idea that Japanese surrender was predicated on Tokyo’s realization that it could not defend the home islands defies historical accounts of the panic inflicted by both nonnuclear incendiary bombing and nuclear attacks. Ignoring these effects leaves one without a clear idea of how Germany and Japan ultimately were defeated, and the important role air power played in those outcomes.

Analysts of strategic theory are generally uncomfortable with “either-or” explanations and seek holistic ideas of causation. I refer to Colin S. Gray’s paper on battle outcomes in the Battle of Britain as an example. Certainly Gray puts more weight on some factors than others but he emphasizes their complex interaction. Williamson Murray writes about the contingency of the German victory in 1940, looking at how a better led and trained French force might have prevented German breakthrough in the South. 

The contextual and holistic approach also tends to appeal to practitioners from the policy and operational worlds because it feels intuitively right to them. That is why we see a stream of papers, blogs, and magazine articles targeted towards policy audiences that insist that war is complex. It is, and that can sometimes be lost in IR approaches to studying conflict. We still can’t really agree on what precisely led to Serbia’s capitulation in the Kosovo conflict, and Daniel Byman and Matthew Waxman’s frustration with the state of the debate led to them echoing Watts in calling for a synthetic look at airpower’s interaction with other factors.

Thankfully, political methodology has since caught up with some of strategy’s methodological demands. There are a number of methods of quantitative and qualitative analysis that can help an analyst capture some of the factors that have traditionally eluded political science approaches to war. They are too numerous to elucidate here, and those interested in exploring further should order some of the Oxford handbooks in the previous two links for a look at where the field is currently sitting.

Complexity-based approaches, including some methods that incorporate agent-based modeling, are also now beginning to be explored.  In particular I am very interested in exploring the idea of co-evolution as a means of interpreting strategic history. One can also note that Clausewitz himself was a prototypical social scientist. Just because strategy and the conduct of war is complex does not necessarily mean that social science is not methodologically well-equipped to examine either.

At the same time there are also very concrete differences between IR theorists, who are  looking to create broadly scientific knowledge about the political world, and military historians trying to capture the intricacies and complexities of the past. It may be true that the dividing line between such cross-purposes has been significantly blurred but it has not completely eroded. Strategic theory—the study of ends, ways, and means—straddles those boundaries somewhat uneasily. Even so, strategy need not be alien to IR.

As M.L.R Smith noted, “[strategic theory] advances a set of propositions that if true can be held to explain certain facts or phenomena. In this regard, strategic theory reveals itself less as a set of hard and fast rules, but more as a series of purposive assumptions that guide analysis.” Those assumptions are probabilistic. Clausewitz notes that sometimes “war by algebra” occurs and the enemy backs down without a fight after assessing the correlation of forces. But it’s significantly more likely that combat is needed to compel strategic decision.

Highly abstract large-N theories that depend on regression methods, game theory and formal modeling all have gotten a bad rap in the past but have their place in assessing certain kinds of strategic questions as long as their limitations are acknowledged. Quantitative approaches in general have not been treated fairly—certainly they have been misused but qualitative models are just as imperfect in their representation of highly complex realities. Putting common assumptions to formal rigor can also illuminate strategic problems, as Phil Arena demonstrates in his look at two-person attrition games.

At the end of the day, no one method or model will be able to capture all of the complexity of war. But that’s OK. Strategy is conceptually very simple but also very difficult to practice. So where can we innovate? In his rebuttal to Pape, Watts suggested applying complexity-based approaches and favorably cited Robert Jervis’ System Effects. George Mason University’s Aaron Frank is doing precisely that, looking at the applicability of evolutionary theories to the Military Revolution debate and doing agent-based modeling of intelligence processes. Certainly those approaches have weaknesses of their own but also suggest promising ways forward. There are also useful new methods being advanced in the field of military innovation/diffusion by Michael Horowitz, Dima Adamsky, and Evan Laksmana.

Ultimately strategic theory should not be subsumed into IR. There are historical traditions in strategic theory that will always mark it as a distinct—and interdisciplinary—entity. Edward Mead Earle created the American strategy field by bringing together historians, political scientists, sociologists, and hard scientists into a single forum to consider the challenges posed by the Axis. Strategy is eminently practical and as such will always be rooted in such interdisciplinary assemblages.

Instead of swallowing strategy whole, IR should simply borrow from it, just as it borrows from economics, anthropology, law, philosophy, and sociology. Likewise strategy has borrowed a great deal of underlying assumptions from IR’s various strains to explain state and non-state strategy and conduct of war. More fruitful exchange can only improve both fields and make them more useful to those seeking to better understand the dynamics of human conflict.

May 1st, 2012

Walt and Unreal Realism

Stephen Walt asserts that, had realists been in charge of US foreign policy, a host of maladies could have been avoided. There are, however, big problems with this piece as well as Walt’s general approach to realism. Much as I would like to believe that, say, putting Hans Morgenthau in charge of the NSC would have led to exceptionally better policy and strategy, a view that emphasizes commonsense “realism” does not reveal much foundation for his assertion.

First, international relations realism is such a diverse tradition that putting “realists” exclusively in charge of US foreign policy would likely produce a cacophony rather than the harmonious symphony he suggests. When Walt writes about China he neglects the significant difference between how an offensive realist and a defensive realist like himself might perceive the nature of US-China relations, the Asian-Pacific security system, and whether or not it is possible for the US or China to really overcome the security dilemma. This is not an academic dispute of how many Thucydides citations can dance on the head of a pin—it has real policy implications. And this is without going into the longstanding challenge that scholars of Asian IR such as David Kang have made against structural realism’s applicability in East Asia. 

It is worth noting that despite, in policy circles, the rhetorical and practical persistence of liberal approaches to viewing the world since Wilson, American foreign policy takes on a significantly schizophrenic character. The United States cheers on the Arab Spring, but not in Bahrain—where it might threaten favorable US military arrangements. A cottage industry of critics often see deliberate hypocrisy, but a more charitable way of viewing this imbalance is that the United States is torn between sincere idealism and the day-to-day “real” demands of international politics. If “idealists”—which Walt claims have exclusively monopolized American policy—cannot use their power to impose their preferences what hope do realists have?

Walt also stacks the deck when constructing his list of counterproductive things realists would not have done if they had been in charge. Given that the a sizable chunk of US IR scholars—constructivist, liberal, and realist (to say nothing of theories outside the “big three”)—opposed the invasion of Iraq, it is also not very clear why realists should get a pat on the back for coming to the conclusion that going to Baghdad was a bad idea. One might demonstrate “realism” in the sense of pragmatism for opposing the war, but you don’t need to read Waltz’s Theory of International Politics to do it. 

Second, Walt misunderstands the relationship of theory to policy. IR scholarship is scholarship—it seeks to generate knowledge about the world around us. That knowledge can help aid action, but in and of itself does not constitute an operational approach for action. Carl von Clausewitz wrote in On War that theory prepares the mind for action, but cannot dictate action, and Alexander George has written extensively about the interaction between IR scholars and policymakers, and argues that general knowledge of international relations can only be an input to, not a substitute for, policy analysis within the government. General knowledge will aid the decisionmaker, but cannot tell him or her what the right answer to a given individual situation may be. IR theory is systemic in nature—it is about (within the framework of dominant big three IR) causal inferences. It was never optimized for the purpose of telling policymakers how to handle individual cases.

Aside from this, policymaking is distinguished by domestic political considerations, international policy linkages, and bureaucratic foodfights. Even if Walt’s realist policymakers understood the right solution, boning up on Offense-Defense Theory tells you nothing about how to operationalize it within the American political system. The invasion of Iraq stands as a paramount example of this problem. The confused termination of the first Gulf War, US sanctions policy, a bipartisan commitment to containing Iraq without thought of how to sustain such a policy in the long term, and the way domestic politics produced bipartisan support for regime change in the 1990s all most likely led to the 2003 war. And lest Walt blame it all on neoconservatives, I would point him to Candidate Bush’s 2000 speeches eschewing nation-building and Condoleeza Rice’s pre-9/11 writings about returning to realism. As John Lennon sang, life happens when you make other plans. A theoretical commitment to realism does not ensure more policies in line with realism.

Finally, while Walt’s blog heading identifies him as a “realist in an ideological age,” it can sometimes be difficult to distinguish sound realist theory in his writings. This critique is not constructed for the purpose of excommunicating Walt from the realist “church” as his theoretical commitments are substantial. Moreover, while I also generally am sympathetic to IR realism any readers of this blog can see that I hardly limit myself to any mode of analysis in looking at security, war, and warfare—I am the last one to tell Walt he cannot use a useful idea here and there. But he advertises himself prominently as a realist and urges that realists should run American policy, so the following critique is appropriate.

First, Walt mangles the difference between a strategy of restraint and offshore balancing:

Since the end of the Cold War, prominent realists have called for the United States to adopt a more restrained grand strategy that focuses on maintaining the balance of power in key areas (e.g., Europe, East Asia, and the Persian Gulf) but reduces America’s global footprint and keeps the U.S. out of unnecessary trouble elsewhere. Such a strategy would also force U.S. allies to shoulder more of the burden and discourage them from either “free-riding” or “reckless driving” (i.e., adventurism encouraged by overconfidence in U.S. support). For instance, realists would never have adopted the Clinton administration’s foolish strategy of “dual containment” in the Persian Gulf, or the Bush administration’s even more reckless effort at “regional transformation.” Instead, realists would have maintained a robust intervention capability but kept it offshore and over-the-horizon, bringing it to bear only when the balance of power broke down (as it did when Iraq invaded Kuwait in 1990). Had we followed this approach from 1992 onward, it is even possible that al Qaeda would never have gotten rolling in a big way or never tried to attack the United States directly.

Walt significantly misrepresents the Offshore Balancing literature. OSB, as Dan Trombly sometimes calls it, is not simply offloading existing security commitments to allies and floating around waiting to pounce if something goes wrong. Whether onshore or offshore, the overall goal is still ensuring US dominance over the Persian Gulf security system.  Cost, who bears the burden, and how much control the US has over allies’ behavior all concerns the ways US aims are achieved rather than the ends. Moreover, burden-sharing and offloading commitments onto allies puts the United States inside a security architecture and proscribes certain political commitments—even if they are minimized. This may describe a strategy of Restraint (a more accurate term used by Barry Posen that Walt conflates with OSB), but it is not OSB.

The entire point of Offshore Balancing is to avoid diplomatic entanglements altogether, not maintain the same entanglements on the cheap. Arguing that an offshore balancer uses alliances to maintain existing security commitment goes against the spirit of the original literature and the historical record from which it is based. As Trombly has written, offshore balancers in fact foster security competition to prevent the emergence of any one power. Offshore balancers do so by remaining deliberately aloof from onshore alliance structures, refusing to commit. Sometimes this idea is taken to an extreme, as Truman’s infamous quote about World War II illustrates:

“If we see that Germany is winning we ought to help Russia and if Russia is winning we ought to help Germany, and that way let them kill as many as possible, although I don’t want to see Hitler victorious under any circumstances. Neither of them thinks anything of their pledged word.”

As noted above, Walt and other realist critics conflate this with a strategy of restraint. America and Splendid Isolation-era Britain balanced by pitting their rivals against each other, preventing the rise of a hegemonic power with the ability to threaten the British or US mainland. OSB policies, of course, are compatible with other measures in the policy toolkit, and they may or may not be the right policy for the United States in the 21st century. But if Walt’s selling point is that realists would make better policymakers, he should make better use of realist theory to do so.

Some of Walt’s arguments are also more characteristic of liberalism than realism. In one bullet, Walt argues that realists would have launched a “focused effort to destroy al-Qaeda,” but in the next breath he claims that realists would have also “treated al-Qaeda like criminals.” But an approach that aimed to manage al-Qaeda with law enforcement tools and a law enforcement framework, presumably in cooperation with other states, is more characteristic of a liberal approach to international relations and security. It is, after all, IR liberals who tend to minimize military interventions as “police actions” and argue about the necessity of processing terrorists through existing domestic and international institutions. This critique was, in fact, advanced as the standard antidote to American unilateralism by liberal critics of neoconservatism during the height of the Bush years.

A focused effort to “destroy” al-Qaeda, by necessity, would not be an effort to arrest them and put them on trial. It involves, given that many reside outside the US in fortified sanctuaries, the use of coercive tools of power projection to break up enemy networks. But Walt has opposed the drone war by appealing to arguments about international norms of legitimizing targeted killing and the risks of inflaming domestic opinion in the Middle East. This is not a very “realist” argument to make, and especially sounds off-key coming from a structural realist. Walt might have solid grounds, in say, wondering about tools of power projection might do for shifting states or maybe some strategically significant sub-state actors’ perception of the offense-defense balance or causing them to balance or soft-balance against the United States. But international norms?

A structural realist would not worry about states imitating the United States in guaranteeing their own security—since the point of the realist critique of institutions and norms is that they do not meaningfully constrain actors from acting to make themselves more secure. In fact, Walt’s co-author wrote a very useful critique of institutions. Unless Walt wants us to suggest that China or Russia will suddenly start flying drones to zap dissidents living in California or New York, what’s the point of the critique? And even before the US began flying drones, the KGB and Latin American dictators were flagrantly killing dissidents in Western states. Is Walt going to make a causal argument that American targeted killings led to Putin knocking off people in British sushi joints?

The historical record of how states have handled irregular opponents, especially the 19th century British Empire (which Walt, Layne, and others suggest the US should emulate for its grand strategy) is also instructive. The British frequently waged wars to crush irregular threats to the crown, and used the legal system not as a means of processing enemies through the proper set of institutions but making an example out of those who dared oppose the Crown. It also frequently violated the sovereignty of other states with its naval forces in order to hunt down and eliminate pirates who violated the sovereignty of the Empire by waylaying Her Majesty’s servants. Britain during the height of offshore balancing was at war for nearly all of Queen Victoria’s reign in “dark defiles” from modern Pakistan to Rorke’s Drift in wars of choice that make even Iraq look, at times, like a work of Clausewitzian strategic genius. Sure, Britain was offshore in Europe. But if the point of Walt and Layne’s OSB critique is to make the US avoid landpower commitments in the Third World, it’s hard to find support for such a policy in the historical record.

Walt’s own significant analytical confusion about what a “realist” approach to US national security would constitute is strong evidence of the fact that policy analysis in practice tends to be a grab bag of multiple theoretical traditions and theory’s primary role is to increase the analytical tools available to a policymaker.

Loading tweets...


A blog on states, communities, and organizations in conflict by Adam Elkus.

Portrait photo: Marshal Liu "One-Eyed Dragon" Bocheng