Policy Series: Trump and International Relations Theory: A Response to Robert Jervis’s “President Trump and IR Theory”

Poppy field, Helmand Province, AfghanistanCan IR theory help us understand what is about to happen? Can it help get us through the Age of Trump? Or, will Trump destroy IR theory in the same way that he eviscerated most accepted theories of electoral politics? In a cage match between Trump and Theory, the smart bet might be on Trump, but perhaps this says more about the fragility of IR theory than it does about Trump.


H-Diplo | ISSF POLICY Series
America and the World – 2017 and Beyond

“Trump and International Relations Theory: A Response to Robert Jervis’s “President Trump and IR Theory”[1]
Essay by Michael N. Barnett, George Washington University

Published on 7 February 2017 | issforum.org

Editors: Robert Jervis, Francis Gavin, Joshua Rovner, and Diane Labrosse
Web and Production Editor: George Fujii
Commissioned by Diane Labrosse

Shortlink:  tiny.cc/PR-1-5L
Permalink:  http://issforum.org/roundtables/policy/1-5L-Trump-IR
PDF URL:  http://issforum.org/ISSF/PDF/Policy-Roundtable-1-5L.pdf

Can IR theory help us understand what is about to happen? Can it help get us through the Age of Trump? Or, will Trump destroy IR theory in the same way that he eviscerated most accepted theories of electoral politics? In a cage match between Trump and Theory, the smart bet might be on Trump, but perhaps this says more about the fragility of IR theory than it does about Trump.

What is IR theory good for? Lots of things, including prediction, as emphasized by Robert Jervis’s introduction to the roundtable.[2] Perhaps prediction is too strong a word, too closely associated with logical positivism, the search for invariant laws, and if/then statements. Instead of prediction, we might be content with probabilistic thinking and the search for historically-bounded generalizations that can be cautiously applied to the imminent. My sense is that this is what most IR scholars do most of the time. This is the spirit of Jervis’s speculations about world politics in the Age of Trump. Institutionalism uses its insights to not only explain the past but also to offer suggestions for social engineering the future. Constructivists might insist that they do not engage in forms of prediction, but I think that they protest too much. And in my experience even poststructuralists will engage in prognostication of one kind or another. Birds do it, bees do it, even overeducated fleas do it. But we do so cautiously, and without the same bravado as we do when using our theories to explain the past. Our theories are backward-looking, not forward-looking; what looks deterministic from the vantage point of history looks completely stochastic from the perspective of the here and now gazing into the future. I do not mean to suggest that IR theorists should abandon any effort to think about what their theories have to say about how President Trump’s foreign policy and impact on world affairs, but we should recognize the inherent limits of the enterprise.

Most IR theories are based on the existence of identifiable, durable constraints. Jervis uses the venerable levels of analysis (or “images”) to work through the array of forces that might block Trump from doing what he tweets he will do. One of the striking features of his analysis, though, is that these constraints are fairly permissive, or at least they have lots of hidden doors that Trump can walk through. There is human nature and rationality, but Jervis hesitates from assuming that Trump’s nature is like all others or that he even remotely resembles the tenets of rational choice (more on this in a moment). Nor is Jervis confident that President Trump will be easily socialized. Even Barack Obama seems to have abandoned his initial assessment that Trump will come to realize the enormity and complexity of the job, and, therefore, be much more cautious and risk-averse. Apparently 70 year olds who have succeeded while listening to their base instincts and no one else might not feel (or even recognize?) the forces that pinned President Obama to the ground.

What about the domestic level of analysis or second image? Both political parties made their efforts to block his rise, and there he sits on top of the American political system. Professor Jervis suggests that the bureaucracy might be our firewall. Perhaps. But it is a sign of our desperate times that we are now putting our faith in bureaucracies, which are often treated as the enemies of rational decision making. We are no longer attempting to maximize utility but minimize disaster.

And then there is the systemic level of analysis or third image. The classic argument is that states, in pursuit of their national interest of security and prosperity, will be saved from their worst instincts and driven to conform with the requirements of system stability. Trump will not start a trade war or alienate China because doing so would wreck the American economy. Trump will not really backtrack on America’s security commitments because it would undermine the Western order that has been in place since World War Two, which would ultimately harm U.S. security. In other words, the “national interest,” defined in terms of security and wealth, will discipline even the most unruly politicians. So, Trump will be forced to conform to the national interest.

Which is what, exactly? As we know, the national interest is like a lot of our concepts—we are certain that it exists but struggle to fill in the details. Indeed, IR theorists often concede that the national interests is whatever the President decides it is. So, we can mock the slogan of “Make America Great Again,” and agree with Jon Stewart that Trump does not have the right to decide what it means to be great, but most of our theories are equally vague. And if an interest is what someone in power says it is, then it is difficult to see how constraining it can be.

If Trump is able to redefine interests any way he wants, surely he will be stopped in his tracks by American values such as liberty, equality, freedom, democracy, fair play, and so on. Maybe. Although he probably cannot redefine American values, he can certainly ignore them or decide that they are standing in the way of American interests (just as realists would predict or recommend?). So far Trump has praised the authoritarians, and said little in support of human rights, the rule of law, and democracy. He seems to suggest that the U.S. has no stake in Crimea, Georgia, or the Ukraine, or anywhere else the Russians seem to covet. Trump has argued that getting rid of the Islamic State of Iraq and Syria (ISIS) is more important than getting rid of Bashar al-Assad, and so he will work with Assad to get rid of ISIS. We might not agree with his choices, but Trump is not the first to insist that they are consistent with the national interest.

In any event, the systemic level of analysis does not seem to hold much promise, or at least is too permissive to identify with confidence the constraints that might contain Trump. Where should we turn, then? Jervis recommends three classics of international relations and comparative politics–and the one thing Peter Gourevitch, Karl Polanyi, and Samuel Huntington have in common is that international politics and processes are in the background and all the work is being done on the home front. [3]

There are other structural theories of international politics that might come to the rescue, including constructivism. There are as many different flavors of constructivism as there are of realism, but they operate with a thick conception of society, a society dense with social norms, rules, scripts, and discourses that do more than just operate as constraints and provide mechanisms of reproduction of the social order—they also constitute the actors. Indeed, these social rules can give the appearance of being so powerful that constructivists often have to add the disclaimer that agency still matters and that actors are not cultural dupes. If they didn’t think that agency played a role, then the “agent-structure problem” wouldn’t be much of a problem. But, in my reading of constructivism, most of the emphasis is on structure. So what happens when Trump (agent) meets the world (social reality)? I suspect that constructivists will emphasize the cultural constraints of norms, laws, and institutions, making clear that while Trump might be all logic of consequences, the world still operates according to the logic of appropriateness. Trump might really think that waterboarding is an appropriate way to treat enemies of the state, but most of the world does not, and will probably impose significant costs on him (and his American accomplices) if he follows through. Consequently, the military, assuming that they even believed that waterboarding was appropriate, would recommend alternative, more efficient and less costly ways of getting the job done. And even if Trump defied international law, one violation does not shatter a norm. Indeed, the response to the violation might be to rally around the norm and strengthen it. We will see.

Thinking about how the international might constrain or channel Trump’s foreign policy underscores something that both realist and constructivist theories have in common: their structures might be constraining, but they also are quite permissive. Neorealists have tied themselves into knots trying to identify how the (material) international system might affect the probability of great power war, alliance behavior, and assorted other features of high politics.

Constructivists draw from sociological theories that presume a fairly thick society, but nevertheless debate how thick is thick. But at what point does not very thick become thin? We don’t have to answer that question, we only ask ourselves: is international society thinner or thicker than domestic society? If we can agree that domestic society is thicker, and Trump just ripped through it, then international society might be nothing more than a Maginot Line. After all, Trump shattered nearly every conventional wisdom of electoral politics, making fools of the experts. According to them: President-elect Trump’s electoral possibilities should have ended the very moment he showed up on that Mexican border to announce his candidacy and declared that he would protect the American people from the know-nothing American political class and Mexican rapists. His candidacy should have ended when he continued to think the unthinkable, say the unsayable, and tweet whatever his impulsive fingers typed. It should have ended when he mocked minorities and the disabled and showed disdain for the rule of law. It should have ended when he boasted that the secret of his success with women owed to tic-tacs and assault. He shredded the American electoral playbook. And now we have President Trump. American politics is what Trump makes of it. Anarchy is what Trump makes of it. Or maybe, Trump makes anarchy.

But perhaps we are looking at this all wrong. Rather than imagining how Trump has remade American politics and is potentially about to do the same to international politics, why not see American and international politics as having made Trump. Trump is not a disrupter of the times–he is a sign of the times. He not a cause–he is an effect. We should be thinking of Trump as an accelerant—a chemical that turns the fire into an inferno. It is not as if Trump has intruded on a world of stability. Initially IR scholars interpreted the post-Cold War moment as one of hope and promise, but over the last decade the themes are darkness and uncertainty. The West is in decline. China is on the rise. The old international institutions are too sclerotic to handle the change and the new ones seem like placeholders rather than fit for purpose. Global governance might be good enough, but seems to be barely keeping its head above a rising sea level. Three decades ago IR scholars celebrated a third wave of democracy and now they are watching populism take revenge. Trump is entering at a moment when the rules are already in flux and the trade winds are blowing in his direction.

Because so many IR scholars are looking to the interwar period and referencing Weimer Germany to make sense of these times, perhaps they might do well to go back and revisit the theories that emerged in response to the crisis. Namely, critical theory in general and the Frankfurt School in particular. Deeply schooled in Marxist theory and influenced by modernist traditions, theorists such as Theodor Adorno, Max Horkeimer, Herbert Marcuse, Erich Fromm, and Ernst Bloch attempted to wrestle with how seemingly rational, modern, and civilized societies might go mad.[4] And they tried to understand such madness not as an alien life force but rather as a product of the very same forces that were also responsible for enlightenment and visions of progress. If IR theorists turn to critical theorists, it won’t be the first time. Hans Morgenthau, the grand master of realism, engaged these and other scholars as he himself attempted to understand a world that descended into self-destruction.[5]

There is a sense in which our IR theories are trying to make sense of the nonsensical. Our most beloved IR theories assume that actors are reasonably rational. Rational choice theory props up much of mainstream IR. I am well aware that rational choice theorists insist that the model’s underlying tenets are treated as “as if” assumptions for the purposes of modeling rather than statements about how real humans really think and process information. Constructivists are not as hard on rationality assumptions as they might appear, because many will work with the classic Weberian distinction between value and technical rationality. Rational choice and rationality appear to be acceptable baselines, and without them prediction would become nearly impossible.

But at what point do conveniences become contrivances or even delusions? What are Trump’s interests? Besides self-promotion, the mother of all self-interest, no one is really quite sure. Can Trump rank order his preferences in a consistent and reliable way? Perhaps, but it is hard to know what to make of a President-elect who appears to spend more time tweeting about the play “Hamilton” than he does thinking about the future of the Western alliance. Does he seem to search for better information? This is the man who says that daily intelligence briefings should be weekly and that the intelligence community mistook a “400 pound guy” for a sophisticated corps of Russian hackers. And what about information processing? Does he even process information? He gives the appearance of having the impulse control of a 13 year-old boy on stimulants. Does he weigh alternatives? Does he assign probabilities to outcomes? Does he look for the most efficient means? Does he relate means to ends? Does he try to surround himself with those who will give him the best advice based on the best information? Do we think that he will even listen to dissent? I have a difficult time imagining him even designating a devil’s advocate to subsequently ignore. If the IR community is fearful of life with Trump, it is precisely because he falls far below the bar of rationality. Maybe someday presidential historians will discover that it was all an act, in the same way that they now believe that President Eisenhower tried to project the image that he was aloof from daily policymaking when in fact he was quite involved. In the meantime, let’s take a pledge: we agree that rational actor models should never be applied to the Trump administration–not by us and not by the students we teach and advise.

Maybe social psychologists and those who are open to the presence of “animal spirits” can provide a better insight into Trump’s decision making. According to these theories of information processing, many associated with the pioneering work of Daniel Kahneman and Amos Tversky all of us are susceptible to forms of bias–motivational, confirmation, and on and on.[6] But at what point does the concept of bias cease to be a description and become a euphemism? During the campaign various political elites and pundits declared that Trump was “unfit for office.” What did they mean? They were doing more than simply warning about “System I” thinking. Nor were they worrying about an inflated ego or narcissism, which probably affects every politician (and probably quite a few academics). Instead, they were suggesting that he had deep psychological, indeed psychiatric, problems. Psychiatrists were suggesting that he had more than his fair share of disorders according to the Diagnostic and Statistical Manual of Mental Disorders (DSM–5), but rather that he might be mentally ill. Intelligent, yes. Savvy, yes. High functioning, yes. A sociopath, quite possibly. This doesn’t mean that we cannot predict his behavior. After all, forensic psychiatrists often try to predict the behavior of sociopaths. But they don’t tend to rely on rational choice or prospect theory to do so.

So, what is IR theory good for? It might be helpful at predicting what happens starting January 21. But it might be a greater help at keeping IR theorists warm in the chilly nights ahead. IR theory as the opium for the academics. It is reassuring. It suggests that the world still makes sense, or we can still make sense of the world. But is this a trick we play on ourselves, an act of self-delusion in order to maintain the belief that the world and our choices still make sense? Do we really think that IR theories were built to understand Trump and his world? And what is the point of doing so? At this moment I am reminded of an old joke retold by Woody Allen at the end of his classic Annie Hall. In the same way that times like these make me question IR theory, Allen is trying to make sense of why we seek relationships. “A guy walks into a psychiatrist’s office and says, hey doc, my brother’s crazy! He thinks he’s a chicken. Then the doc says, why don’t you turn him in? Then the guy says, I would but I need the eggs. I guess that’s how I feel about relationships. They’re totally crazy, irrational, and absurd, but we keep going through it because we need the eggs.”


Michael Barnett is University Professor of International Affairs and Political Science at George Washington University. Among his books are Empire of Humanity: A History of Humanitarianism (Cornell University Press, 2011), and Paternalism Beyond Borders (edited, Cambridge University Press, 2017).

© Copyright 2017 The Authors



[1] This essay was written in early January 2017. Addendum, 30 January 2017: I thought I might have been too hard, but it turns out that I was too sanguine.

[2] Robert Jervis, “President Trump and IR Theory,” ISSF Policy Series “America and the World—2017 and Beyond,” Introductory Essay, https://issforum.org/ISSF/PDF/Policy-Roundtable-1-5B.pdf.

[3] Karl Polanyi, The Great Transformation (Boston: Beacon Press, 1957), Samuel Huntington, Changing Order and Political Societies (New Haven: Yale University Press, 1973), and Peter Gourevitch, “The Second Image Reversed: The International Sources of Domestic Politics,” International Organization 32:4 (Autumn 1978): 881-912.

[4] Stuart Jeffries, The Grand Hotel Abyss: The Lives of the Frankfurt School (London: Verso, 2016); Theodor Adorno and Max Horkheimer, The Dialectic of Enlightenment (Stanford: Stanford University Press, 2007).

[5] Richard Ned Lebow, “German Jews and American Realism,” Constellations 18:4 (December 2011): 545-566. DOI: https://doi.org/10.1111/j.1467-8675.2011.00658.x.

[6] Amos Tversky and Daniel Kahneman, Tversky, and D. Kahneman, “The Framing of Decisions and the Psychology of Choice,” Science 211:4481 (30 January 1981): 453–458; Buster Benson, “Cognitive Bias Cheat Sheet.” https://betterhumans.coach.me/cognitive-bias-cheat-sheet-55a472476b18#.77qmmqvy9.