war on the rocks

Doubling Down on Reputation: Defining State Resolve and Why it Matters

“You shouldn’t think of Putin as such a primitive guy. It’s totally clear that the Syrian and Ukrainian crises had nothing to do with one another.”  This is what Fyodor Lukyanov, editor of Russia in Global Affairs, had to say when Julia Ioffe asked him whether President Obama’s decision to not bomb Syria in 2013 emboldened Putin to invade Ukraine. “Russia is a nuclear superpower, and this kind of rationale vis-a-vis Russia is senseless.”

In a recent article for War on the Rocks, I argued that the conventional discussion of the relationship between credibility and state reputation is misguided.  I marshaled evidence from my study of international crises to demonstrate that the record of U.S. compellence (the use of threats to induce a target state to change its behavior) is inconsistent with the argument that the United States must intervene in country A in order to effectively coerce country B.  This simple interpretation of the role of state reputation, often advanced by the commentariat and critics of this president in particular, does not fit with reality:  States do not behave as we would expect if reputation operates in such a straightforward manner.

Alex Weisiger and Keren Yarhi-Milo criticize this approach in a recent response.  I welcome this opportunity to consider a topic of critical importance to both academic and policy debates about the use of power and influence in international politics.  If we define “reputation” as “a state’s record of past behavior,” then I would agree that the United States’ “reputation” probably influences the way potential adversaries and allies assess American threats and promises.  We might call this the “general reputation” theory of state behavior.  If the United States never followed through on any commitment anywhere, target states would be unlikely to take the United States’ demands very seriously.  Nor would I disagree that the United States’ record of behavior in alliances or in past military conflicts may influence the way other states perceive the United States and how they respond to its demands.

I did not set out to evaluate these much broader claims about how a state’s past behavior influences the way other states respond to its future threats and promises.  Instead, I considered a much narrower definition of state reputation and how it affects the behavior of other states: What I term the “take-action reputation” theory of state behavior holds that the failure to follow through on past commitments will render tomorrow’s threats ineffective, and by extension, that the United States must follow through on all commitments today to render its threats effective tomorrow.  When critics argue that the failure to bomb Syria emboldened Putin to invade Ukraine, they are citing this “take-action reputation” theory of the role of credibility and reputation.

Weisiger and Yarhi-Milo are concerned that my study’s narrow focus provides a misleading picture of the role of reputation in international politics.  As I stated in the piece, I was specifically examining whether we could find evidence consistent with the “take-action reputation” explanation for U.S. coercion (although perhaps I could have labeled it more clearly as such).  I restricted my focus to interstate coercion not because there are no other realms in which we can observe the role of reputation, but because this is an area of international politics in which it should be particularly easy to observe the effects of “take-action reputation,” if it operates in the way that critics of the U.S. decision not to bomb Syria in 2013 argue.  Critically, I demonstrated that during the period I studied (1945 to 2007) the United States always followed through on compellent threats in crises. As such, there should be no doubt in the minds of target states about Washington’s “take-action reputation,” and yet in the post-Cold War period we still observe these threats fail more often than not.  Clearly, there is something else going on here.

In my article, I also examined two of the highest-profile cases of U.S. coercion from the last few decades: Iraq and Libya. I did not claim that past U.S. behavior played no role in how these states decided whether to concede to U.S. demands in 2002/2003 and 2011, respectively.  Instead, I demonstrated that the act of following through on past U.S. threats was not enough to convince those states to concede, as we would expect if the “take-action reputation” theory was correct.

As I discuss in my book, Cheap Threats: Why the United States Struggles to Coerce Weak States, before the 2003 invasion Saddam did consider America’s performance in past conflicts.  He did not, however, look at the perfect U.S. record of following through on past compellent threats in crises and then decide to concede, as the “take-action reputation” theory would lead us to expect.  Instead, Saddam believed that the United States would take action against him and chose to resist anyway.  The threat, in other words, was entirely credible, and yet that was not enough to convince him to concede.

Weisiger and Yarhi-Milo cite their recent article in International Organization to argue that reputation plays a much more important role in international politics than I acknowledge.  They set out to examine how a state’s reputation for resolve influences the likelihood that it will become involved in future disputes.  Taking the results of their models at face value, it is not clear that Weisiger and Yarhi-Milo are actually measuring what they claim to be in their statistical tests.  They borrow Mercer’s definition of resolve: “the extent to which a state will risk war to achieve its objectives.”  A highly resolved state is one willing to run a high risk of military conflict to prevail in a crisis.  Weisiger and Yarhi-Milo then define reputation for resolve as “others’ perception of that state’s willingness to risk war.”  Thus a state has a reputation for resolve when other states perceive that the state is willing to risk war to get its way in disputes.

Weisiger and Yarhi-Milo claim that their data show how a state’s reputation affects the likelihood that it will be involved in future disputes.  (In straightforward cause and effect terms, “reputation” is the cause and “future dispute involvement” is the effect.)  What they actually measure, however, is whether winning or losing a previous dispute (the cause) affects whether a state is likely to find itself in a future dispute (effect).  Weisiger and Yarhi-Milo label this “win or loss” variable as “reputation” in the tables reporting their statistical results, but they don’t actually measure and test whether the “perception of [a]state’s willingness to risk war” (reputation, as Weisiger and Yarhi-Milo define it) influences the likelihood that a state will be challenged in the future.  In other words, Weisiger and Yarhi-Milo claim to measure how the perception of a state’s willingness to risk the onset of war affects the likelihood that a state will find itself in a future conflict, but they actually measure how the outcome of a previous dispute affects the likelihood that a state will be involved in a future conflict.

Weisiger and Yarhi-Milo cite these results as evidence that reputation matters.  In fact, their research demonstrates nothing more than the fact that a state that is good at winning disputes is less likely to find itself fighting in the future.  This is an interesting finding, but it does not necessarily have anything to do with resolve or reputation as they define these concepts. To buy their argument about reputation, we would have to believe that all states would look at a given dispute (anything from a war to an argument about fishing rights), derive the same conclusions from it, and then apply those conclusions in the same way to future (potential) conflicts with the states in question years or decades later.  This seems like a large number of assumptions to have to make to conclude that reputation matters in the way Weisiger and Yarhi-Milo claim.  Do we think that everyone derives the same conclusions from the 2003 U.S. intervention in Iraq and will apply those conclusions consistently to future conflicts with the United States?  In many international conflicts, the involved parties cannot agree about who prevailed, let alone the conclusions that should be drawn from them.

This discussion illustrates an important issue that hamstrings current discussions of the role of credibility and reputation.  Both the academic and policy literature suffer from a failure to clearly define and employ the concept of state “resolve.”  In fact, as I demonstrate in my book, there are three types of state resolve: the willingness to execute military action, the willingness to sustain costs in an ongoing conflict, and the willingness to employ violence against the opponent.  The failure to accurately identify and discuss these different components of state motivation hampers our ability to evaluate claims about state resolve, and by extension it hampers our ability to evaluate claims about the role of state reputation.

Consider the following example that illustrates why these distinctions are so important:  Imagine that the United States has become increasingly concerned about a civil war in a small, “rogue” state in the Middle East.  The president is afraid that civilians may soon come under direct attack from their government.  He does not want a massacre on his watch, but he also knows that the American public will not tolerate a prolonged ground campaign to over-awe the combatants and to enforce an ensuing peace agreement.  He decides to authorize air strikes against government forces to degrade their strength and to prevent them from marching into the heavily populated northeastern zone of the country.  He declares from the outset of the operation that the United States would not be sending ground troops into the country and that the air strikes would be limited to a period of four weeks.  All care would be taken to minimize the risks to civilians on the ground from the bombing.

What is critical for our purposes is that high “resolve” in one dimension of state motivation does not necessarily correspond to high resolve in the other dimensions of state motivation—especially for a country like the United States where the use of limited military force is relatively low-cost.  You may have already realized that this is a not-even-thinly veiled reference to the 2011 U.S. intervention in Libya.  In this situation, we would characterize the United States as highly willing to initiate military action but unwilling to pay high costs in the ensuing conflict or to inflict high levels of violence on the target.  In other words, the United States’ willingness to launch military action against the rogue state—indeed, its willingness to carry out a threat—does not mean that it will be resolved to pay high costs to secure its preferred outcome or to escalate the violence to a level necessary to secure a brute-force victory against a stalwart opponent.

When we fail to distinguish among these different types of state resolve, we make it much more difficult to understand the root of America’s struggle to coerce weak states and to understand the role of reputation in international politics.  Weisiger and Yarhi-Milo’s arguments about reputation focus on the resolve to initiate military action, but their models measure performance in war (possibly the resolve to sustain costs, but they do not make this point).  The two types of resolve may be related, but it is inappropriate to simply equate one with the other, and doing so generates misleading findings about the roles of resolve and reputation in international conflict.

It is precisely because the United States is perceived as suffering from a resolve mismatch that weak states choose to resist in the face of a U.S. compellent threat.  The United States is perceived as willing to initiate conflict but not necessarily resolved to pay and inflict high costs in an ensuing fight.  Because the United States is so powerful and the use of force is relatively cheap, it is extremely difficult for the United States to accurately signal high resolve to suffer or inflict costs in conflict, despite the fact that it may have a reputation for following through on its threats.  This is why Saddam Hussein would be willing to resist a superpower that he believed was highly resolved to initiate military action against Iraq but lacking the resolve to pay or inflict the costs necessary to dethrone him.

My original article examined one pathway through which a state’s past behavior might influence the outcome of future events. Weisiger and Yarhi-Milo examined another, but marshaled no direct evidence to contradict my claims about the (dis)utility of the “take-action reputation” theory of state behavior.  I stand by my assertion that it is foolish for the United States to undertake interventions for the primary purpose of reinforcing its reputation.  Targets take more into account than America’s “take-action” record when deciding whether to concede to a U.S. demand, which is why the historical record of U.S. compellent threats is not consistent with the logic of this theory.  As Weisiger and Yarhi-Milo would no doubt agree, there is certainly room for more work on the roles of resolve and reputation in international politics.  I am glad to have sparked such a lively debate about such an important topic.


Dianne Pfundstein Chamberlain, PhD, is an Associate Research Fellow with the Saltzman Institute of War and Peace Studies at Columbia University.  She was formerly an Assistant Professor in the Department of Political Science at the University of Massachusetts, Amherst, and a Postdoctoral Research Fellow in the International Security Program at the Belfer Center for International Affairs, Harvard Kennedy School.  She is the author of Cheap Threats:  Why the United States Struggles to Coerce Weak States (Georgetown University Press, 2016). 

Image: Kremlin.ru