Article Review 5 on “Is Anybody Not an (International Relations) Liberal?”13 min read

Security Studies coverBrian Rathbun asks an arresting question, and a fair one.  Several years ago Jeffrey Legro and Andrew Moravcsik hurled down the gauntlet by asking “Is Anybody Still a Realist?”  Their answer was:  not really.[1]  If Legro and Moravcsik are correct that nearly every IR scholar today considers domestic factors causal in some fashion, then we must ask whether that makes everyone a liberal.  And if the answer turns out affirmative, is that not a problem for liberalism?  If we are all liberals now, then does liberalism have any meaning in IR research?

H-Diplo | ISSF Partnership
Diane Labrosse and Thomas Maddux, H-Diplo/ISSF Editors
George Fujii, H-Diplo/ISSF Web and Production Editor
Commissioned by Christopher Ball
H-Diplo | ISSF Article Review 5
Brian C. Rathbun.  “Is Anybody Not an (International Relations) Liberal?”  Security Studies 19:1 (2010).  DOI: 10.1080/09636410903546558.
Author: John M. Owen IV, University of Virginia
Published by H-Diplo | ISSF on 28 July 2010

Brian C. Rathbun.  “Is Anybody Not an (International Relations) Liberal?”  Security Studies 19:1 (2010).  DOI: 10.1080/09636410903546558.

Reviewed by John M. Owen IV, University of Virginia

Brian Rathbun asks an arresting question, and a fair one.  Several years ago Jeffrey Legro and Andrew Moravcsik hurled down the gauntlet by asking “Is Anybody Still a Realist?”  Their answer was:  not really.[1]  If Legro and Moravcsik are correct that nearly every IR scholar today considers domestic factors causal in some fashion, then we must ask whether that makes everyone a liberal.  And if the answer turns out affirmative, is that not a problem for liberalism?  If we are all liberals now, then does liberalism have any meaning in IR research?

Now a ubiquitous liberalism might not be a problem if IR were what Thomas Kuhn calls a “normal science” and liberalism its dominant paradigm.[2]  Liberalism could be for IR what Darwinism is for biology:  pervasive, assumed, universally defended against any anomalies, indispensable to research.

But if normal sciences exist, IR is certainly not one; and, for Rathbun, liberalism can never make IR into a normal science, because liberalism is no paradigm.  A paradigm, he insists, must be internally consistent and integrated; it must specify entities and how those entities relate.  Here Rathbun makes a crucial move, deliberately conflating Kuhn’s paradigms with the research programs of Imre Lakatos[3] and research traditions of Larry Laudan.[4]

We shall revisit Rathbun’s move below, but for now let us follow it.  Liberalism, he argues, is no paradigm because it lacks a core logic.  Liberalism is typically depicted in one of two ways:  as constituted by a particular value on a dependent variable, where any theory predicting international cooperation qualifies as liberal; or as constituted by a particular category of independent variable, where any theory using domestic-level factors (Waltz’s second image)[5] counts.  Either version of liberalism – claiming a dependent variable or an independent variable – falls short.  Labeling any theory liberal that predicts international cooperation allows a host of contradictory, or at least unconnected, mechanisms.  Institutionalism is often associated with liberalism, but there are rationalists who say that institutions foster cooperation by lowering transaction costs, and constructivists who say that institutions foster cooperation by promoting common identities.  Labeling any theory liberal that looks to states’ domestic properties causes the same problem.  Some explanations for the democratic peace appeal to domestic institutions, as constraints on executive decision making or revealers of preferences and capabilities.  Other explanations appeal to liberal-democratic norms or identities.  These mechanisms do not belong together, argues Rathbun, because they appeal to different logics.

Rathbun acknowledges (15-17) that many scholars typically classified as liberal are untroubled by liberalism’s lack of a core logic.  Among the most influential, Robert Keohane shies away from the liberal label, and Bruce Russett is more interested in explaining empirical puzzles than in vindicating a theory or putative paradigm.  And both, significantly, want to challenge the structuralism and pessimism of Waltz’s neorealism.  Both are more against neorealism than for liberalism, whatever the latter may be.

But that leaves the many college textbooks that persist in presenting liberalism as one of IR’s competing paradigms, and it also leaves what seems to be Rathbun’s chief target, the reconstituted liberalism of Moravcsik’s influential 1997 article.[6]  Moravcsik’s liberalism is the outstanding attempt to make liberalism an internally consistent, cohesive research program.  What the various types of liberal mechanisms – commercial, republican, ideational – have in common is a focus on the origins of state preferences, which are exogenous to states’ strategic interactions.

Rathbun recognizes value in Moravcsik’s clarifying move, but insists that this effort, too, comes up short.  It repeats the error of identifying a level of analysis with a paradigm; it shoehorns together mechanisms – material interest, ideas about the good society – that have no theoretical connection.  For Rathbun, liberalism appropriates mechanisms from these two legitimate paradigms.  It is a poacher.

This is a serious and useful challenge.  As Rathbun notes, scholars will go on using domestic-level mechanisms, and explaining international cooperation, whether or not they are utilizing a common causal logic.  But insofar as these scholars see themselves, or are seen, as participating in a common research program, comparable to quantum mechanics in physics or monetarism in macroeconomics, they must consider this challenge.  In other words, if they see themselves as making progress, as most philosophers of science understand progress, they must answer Rathbun.

As someone whose own work makes causal claims for domestic and transnational groups and ideologies, but has little interest in vindicating liberal IR theory or even identifying as a liberal, I nonetheless offer a twofold reply to Rathbun.

First, I think that he begs the question by presupposing that rationalism and constructivism are real paradigms and can thus sit in judgment of liberalism.  Rathbun regards the divide between rationalism and constructivism as real and meaningful, such that when liberalism crosses that divide it becomes ipso facto incoherent.  But it is never clear why rationalism and constructivism merit paradigmatic status if liberalism does not.  In what sense, for example, do theories that operate on different levels of analysis hang together, while theories that incorporate material and ideational variables do not?  Some constructivists – Alexander Wendt[7] and Martha Finnemore[8] among the most prominent – focus on the international system or society, while others – Christian Reus-Smit,[9] Richard Price,[10] and many others – focus on the domestic level.  The same is true of rationalists:  some – John Mearsheimer,[11] Dale Copeland,[12] Andrew Kydd[13] – are systemic, while others – Kenneth Schultz,[14] James Fearon,[15] Bruce Bueno de Mesquita and his collaborators[16] – are domestic.  Indeed, Rathbun notes with approval the rationalist framework of David Lake and Robert Powell that is agnostic as to level of analysis.[17]  Why are these approaches not “appropriating” explanatory variables from neorealism and liberalism?  For a half century, IR scholars have taken care to separate their levels of analysis, and many have regarded a theory as confused if it appealed to more than one.[18]

Take the democratic or liberal peace.  Rathbun accuses liberalism of appropriating the study of this phenomenon from rationalism and constructivism; in this case, liberalism is true to its tendency to claim any theory that predicts international cooperation.  Set aside the historical point that early scholars of the democratic peace – Rummel, Doyle, Russett – were self-consciously operating within a liberal (anti-realist) approach and not a rationalist or constructivist one.  Why not reverse the charges and accuse constructivists who claim Doyle’s ideational explanation of poaching from liberalism?  Why not accuse rationalists who claim Schultz’s transparency mechanism of doing the same?  Why does coherence require sticking with either ideational or material variables, but not both, yet not require sticking with one level of analysis?

In truth, Rathbun’s problem here points to the difficulty of specifying what “super-theories” or families of theories are – a difficulty he recognizes in his discussion of philosophers of science (10-12).  Rathbun asserts that a paradigm must specify entities and how these entities relate; in Laudan’s words, it must “[outline] the different modes by which these entities can interact” (11).  Now, a paradigm’s entities we can surely recognize:  states, institutions, individual persons, neurons, and so on.  But “outlining the different modes by which these entities can interact” is vague.  Liberalism takes individuals and the groups they construct as primary entities.  On Laudan’s grounds, why does liberalism lack a logic if it allows that individuals interact according to either a logic of consequences (rationalism) or a logic of appropriateness (constructivism)?  Why is that any less coherent than rationalism’s openness to entities’ being individuals or states?

To my mind, it is not.  Liberalism is no less (or more) a paradigm than is rationalism or constructivism.  To be more precise and argue my second point:  the most helpful conception of a “super-theory” is that of Laudan, which he calls a research tradition,[19] and realism, liberalism, rationalism, and constructivism are all legitimate research traditions.  Here we return to Rathbun’s crucial move on10-12, where he collapses the views of Kuhn and Lakatos into those of Laudan.  I am not sure that he can do this, however, or that he is fair to Laudan.  Laudan was writing against Kuhn and Lakatos, and his research traditions differ in important ways from those scholars’ “super-theories.”  Research traditions do not have fixed elements, but evolve over time as the theories they comprise contest one another.  Nor do research traditions entail the theories they comprise; nor do the theories entail the research traditions.  What this means is that, in Laudan’s words, “there are a number of mutually inconsistent theories which can claim allegiance to the same research tradition, and there are a number of different research traditions which can, in principle, provide the suppositional basis for any given theory.”[20]

If we adopt Laudan’s research traditions, then it appears that liberalism is indeed one.  It also appears that certain IR theories could in principle belong to liberalism and also to one or more alternative traditions.  Liberalism is a research tradition owing to its specifications that individuals (rather than states, classes, ethnic groups, and so on – although individuals may form such groups) are the primitive entities; that individuals may relate to one another morally or instrumentally, depending on conditions; and, if instrumentally, through coercion or cooperation, depending upon conditions.  A theory such as Keohane’s institutionalism may be both liberal and rationalist:  liberal owing to its emphasis on conditions under which cooperation emerges, and rationalist owing to its positing of a logic of consequences.[21]

A social science with internally inconsistent, overlapping research traditions may be messier than we would like.  But it appears to be the best depiction of how IR research actually works over time.  Insisting that a super-theory cohere and remain fixed, like an individual theory, is asking too much.

But if Laudan is right, we are not alone; this is the way science works.  For physicists, chemists, geologists, and our other colleagues in the natural sciences, theories belong to research traditions that are qualitatively different from the theories they comprise.  The good news is that these traditions are not simply undifferentiated holding pens for theories.  Research traditions do help clarify matters for scholars.  Not all theories belong in all traditions.  For instance, traditions help researchers identify which problems are important and which are not.[22]  Liberalism directs researchers toward problems of individual well-being; realism, toward problems of international stability.  Thus liberalism is generally more interested than realism in explaining great-power interventions in small states, because realists tend to doubt that such interventions affect the structure of the international system.[23]

Is anyone not an (international relations) liberal?  Yes indeed.  Scholars who regard states or socioeconomic classes as the primary units of analysis are not liberals.  Scholars who expect no sustainable progress, understood as increasing individual security and prosperity, are not liberals.  It is not a problem that many scholars who are liberals, so defined, disagree about theories and methods.  The problem would come if we cast liberalism aside – or realism, for that matter – simply because we mistakenly thought it was less scientific than its competitors.


John Owen is an associate professor and director of the Honors Program in the Politics Department at the University of Virginia.  He studies the effects of ideas and cultures on international relations and vice versa.  His is author of The Clash of Ideas in World Politics: Transnational Networks, States, and Regime Change, 1510-2010 (forthcoming in 2010), and of Liberal Peace, Liberal War: American Politics and International Security (1997).  He also is co-editor of Religion, the Enlightenment, and the New Global Order (forthcoming in 2010), which was produced under the auspices of the Institute for Advanced Studies in Culture.  He has published in Foreign Affairs, International Organization, Internationale Politik und Gesellschaft, International Security, International Studies Quarterly, National Interest, Perspectives on Politics, and several edited volumes, most recently A New Realism, ed. Ernest May, Richard Rosecrance, and Zara Steiner (forthcoming in 2010).  He holds an A.B. from Duke University, an M.P.A. from Princeton University, and a Ph.D. from Harvard University.  He has held fellowships at Harvard, Stanford, Princeton, and Oxford universities.  His research has been supported by grants from the Mellon, MacArthur, Donchian, and Earhart foundations.  He is a member of the editorial board of International Security.


© Copyright 2010-2015 The Authors.

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 United States License.


[1] Jeffrey Legro and Andrew Moravcsik, “Is Anybody Still a Realist?” International Security 24:2 (1999), 5-55.

[2] Thomas Kuhn, The Structure of Scientific Revolutions (Chicago: University of Chicago Press, 1960).

[3] Imre Lakatos, “Falsification and the Methodology of Scientific Research Programmes,” in Criticism and the Growth of Knowledge, ed. Imre Lakatos and Alan Musgrave (New York: Cambridge University Press, 1970), 91-196.

[4] Larry Laudan, Progress and Its Problems: Toward a Theory of Scientific Growth (Berkeley: University of California Press, 1977).

[5] Kenneth N. Waltz, Man, the State, and War (New York: Columbia University Press, 1959).

[6] Andrew Moravcsik, “Taking Preferences Seriously: A Liberal Theory of International Politics.” International Organization 51 (October 1997), 513-53.

[7] Alexander Wendt, Social Theory of International Politics (New York: Cambridge University Press, 1999).

[8] Martha Finnemore, National Interests in International Society (Ithaca: Cornell University Press, 1996).

[9] Christian Reus-Smit, The Moral Purpose of the State (Princeton: Princeton University Press, 1999).

[10] Richard Price, The Chemical Weapons Taboo (Ithaca: Cornell University Press, 1997).

[11] John J. Mearsheimer, The Tragedy of Great Power Politics (New York: W. W. Norton, 2001).

[12] Dale C. Copeland, The Origins of Major War (Ithaca: Cornell University Press, 2000).

[13] Andrew Kydd, Trust and Mistrust in International Relations (Princeton: Princeton University Press, 2005).

[14] Kenneth Schultz, Democracy and Coercive Diplomacy (New York: Cambridge University Press, 2001).

[15] James D. Fearon, “Domestic Audience Costs and the Escalation of Disputes,” American Political Science Review 88:3 (1994), 577-92.

[16] E.g., Bruce Bueno de Mesquita et al., The Logic of Political Survival (Cambridge: MIT Press, 2003).

[17] David A. Lake and Robert Powell, eds., Strategic Choice in International Relations (Princeton: Princeton University Press, 1999).

[18] Waltz, Man, the State, and War; J. David Singer, “The Levels of Analysis Problem in International Relations,” World Politics 14:1 (October 1961), 77-92.

[19] Larry Laudan, Progess and Its Problems: Toward a Theory of Scientific Growth (Berkeley: University of California Press, 1977).

[20] Laudan, Progress, 85 (italics in original).

[21] Robert O. Keohane, After Hegemony: Cooperation and Discord in the World Political Economy (Princeton: Princeton University Press, 1984).  Note that Keohane considers the possibility of altruism as well, which might place his institutionalism within the constructivist research tradition as well.

[22] Laudan, Progress, 81-93.

[23] Michael W. Doyle, Ways of War and Peace (New York: W. W. Norton, 1997); John M. Owen IV, “The Canon and the Cannon: A Review Essay,” International Security 23:3 (1998/99), 147-78.