To the Editors (Michael Poznansky writes):
In “You Can't Always Get What You Want,” Alexander Downes and Lindsey O'Rourke investigate whether foreign-imposed regime change (FIRC) improves interstate relations.1 With some exceptions, their answer is a resounding no. Not only does regime change rarely enhance relations between the intervening state and the target state, but it may make matters worse by exace rbating conflict. Downes and O'Rourke's study marks a significant contribution to analysts' understanding of foreign-imposed regime change and its utility as a tool of statecraft. One problem, however, is that Downes and O'Rourke do not adequately define success or failure independently of their empirical measures. This, in turn, makes it difficult to truly know whether regime change improves or worsens relations between intervener and target. The remainder of this letter attempts to explain why this is so.
The first step in knowing whether Downes and O'Rourke are correct that “you can't always get what you want” in the world of regime change is to specify what it is that leaders are hoping to accomplish by toppling a foreign leader. While Downes and O'Rourke note that “[a]n intervener's primary reason for installing a new leader in another state is to get that state to behave in the intervener's interest,” they do little to specify what this would look like (p. 85). Does behaving in the intervener's interest simply mean that the foreign-imposed regime will be less likely to engage in militarized disputes (MIDs) with the state that put it there? This, at least, is what Downes and O'Rourke's quantitative analysis, which uses MIDs as the core dependent variable, implicitly suggests. In this view, regime change improves interstate relations to the extent that it reduces the probability of a MID.
What if, however, the aim of intervention has less to do with reducing the odds of direct conflict with a target state and more to do with denying a rival power the opportunity to bring an ideological protégé into its camp? Indeed, intervening to topple ideologically threatening regimes or preventing a hostile ideological alliance from forming in the first place has been one of the key causes of regime change over the last several centuries.2 When intervention is undertaken for these purposes and successfully prevents such an outcome, one might reasonably conclude that regime change worked even if the intervener and the target experienced some form of a militarized dispute afterward.3 None of this is meant to deny the value of knowing whether or not FIRC reduces the odds that a given dyad will experience militarized disputes. My point is simply that one cannot say for certain whether FIRC improves state relations without first understanding what it would mean for the target to act in accordance with, or contrary to, the intervener's interests. In some cases, a reduction in the likelihood of MIDs may qualify as the primary purpose of an intervention and thus would represent a fitting outcome to investigate. When FIRC is undertaken to achieve other ends, it will be less appropriate.
Downes and O'Rourke's failure to explicitly define what interveners are hoping to accomplish by conducting regime change not only creates problems for adjudicating the success or failure of an operation on its own terms; it also creates problems when it comes to specifying what the appropriate counterfactual should be in a given case. To say with any confidence whether FIRC improved or worsened relations between intervener and target, one would need to know whether relations would have been better or worse had regime change never transpired (which, of course, has a lot to do with the specific goals of an intervention). To illustrate this point, consider the U.S.-sponsored intervention against Mohammad Mossadegh in Iran in 1953, which Downes and O'Rourke describe as a failure (ibid.). Their rationale is that regime change against Mossadegh was a contributing factor to the 1979 Iranian Revolution and the virulent anti-Americanism that followed. This is certainly plausible.
An alternative way of looking at this case is that relations between the United States and Iran would have been far worse had Mossadegh remained in office. Recall that the primary concern of U.S. policymakers at the time—rightly or wrongly—was that the communist Tudeh Party would seize power, dismantle democratic institutions, and move the country into the Soviet camp.4 Of course, one cannot rewrite history and access this counterfactual directly to know whether this would have happened. Nevertheless, thinking through this issue, both conceptually and theoretically, would be a step in the right direction.
Before closing, it is worth noting that divining the true intentions behind a given intervention is fraught with challenges. To begin, the intervener's publicly stated reasons for regime change may diverge from what its true aims and intentions are.5 Moreover, the goals of a regime change operation can vary over time, making it even more difficult to know if an intervention was a success or a failure. These issues may help explain why Downes and O'Rourke chose to focus on readily observable outcomes such as MIDs to assess FIRC's ability to improve interstate relations, rather than relying on the rhetoric of leaders to match stated aims to observed outcomes. Even still, without clearly articulating the specific, and perhaps case-specific, reasons why states pursue regime change in a given instance, one runs the risk of exploring relationships—for example, the effect that FIRC has on the likelihood of disputes or the causes of FIRC—that bear little resemblance to what policymakers were trying to accomplish. The theoretical and policy implications of this disconnect may be significant.
Alexander B. Downes and Lindsey A. O'Rourke Reply:
We thank Michael Poznansky for his response to our article.1 Poznansky offers two critiques of our work. First, he maintains that despite the title of our article, we never say what states want when they enact foreign-imposed regime change (FIRC). Without knowing what interveners sought to achieve through regime change, which varies across cases, it is impossible to evaluate whether a particular FIRC succeeded or failed.
We agree with Poznansky that it is important to understand why states carry out regime change. Indeed, scholars have identified an array of potential ideological, economic, humanitarian, and military motives.2 For any given FIRC, each explanation will hold some level of sway, and many interventions have overlapping motives. Yet we argue that regardless of the policy goals animating a particular FIRC, there is an overarching logic of regime change: policymakers from the intervening state hope, at a minimum, to install leaders who will serve as reliable custodians of their state's interests in the target country and faithful allies on the international stage. If the new leader behaves as the intervener expects, relations between the two states should improve and the likelihood of military conflict should decrease. The logic for selecting militarized interstate disputes (MIDs) as our dependent variable was therefore simple: MIDs are a useful barometer of the overall quality of intervener-target relations, and thus for whether FIRC worked out as policymakers anticipated. Our analysis shows that, on average, FIRCs do not improve intervener-target relations, and certain types of FIRC make them worse.
Poznansky's second critique is that our decision not to specify the motivations for FIRC renders us unable to identify the proper counterfactual. As he puts it, “To say with any confidence whether FIRC improved or worsened relations between intervener and target, one would need to know whether relations would have been better or worse had regime change never transpired.”
On the one hand, Poznansky's point is true, because not even randomized clinical trials observe the effect of a treatment and a placebo in the same individual at the same time. On the other hand, it is a curious criticism given that it can be leveled against any study attempting to estimate causal effects. That is why Gary King, Robert Keohane and Sidney Verba's classic methodological study referred to it as “the Fundamental Problem” of causal inference.3 Yet researchers have developed tools to mitigate the problem that are well understood—and that we employ in our study. For instance, we controlled for ten variables identified by previous studies as important correlates of interstate conflict, many of which (e.g., relative capabilities, regime type, similarity of alliance portfolios, and distance) could also be correlated with FIRC. Likewise, we used genetic matching to identify control cases that were highly similar (on average) to cases that experienced FIRC to approximate the counterfactual as closely as possible. Neither of these procedures (or any of the others described in the supplementary materials) altered our basic findings.4
Poznansky's more specific point concerns the effect of the U.S. covert leadership FIRC in Iran in 1953. In our article, we noted that many historians have concluded that U.S. sponsorship of the coup against Mohammad Mosaddegh contributed to hostility in Iran against the United States and Washington's hand-picked successor, Mohammad Reza Shah. When Islamic revolutionaries toppled the shah's regime in 1979, memories of 1953 ensured that U.S.-Iranian relations would be marked by intense antagonism.5 Poznansky, by contrast, argues that “relations between the United States and Iran would have been far worse had Mossadegh remained in office” because U.S. leaders feared that “the communist Tudeh Party would seize power, dismantle democratic institutions, and move the country into the Soviet camp.”
As with many counterfactuals, it is impossible to rule this one out definitively, but let us make two observations. First, as we pointed out in our article, the “threat” from the Tudeh Party was vastly overblown.6 For instance, as Maziar Behrooz's analysis found, “[T]here is no evidence that the party had a plan for securing political power for itself in the foreseeable future. It is difficult to imagine how the party could have ousted Mossaddeq with no plan, no real base in the countryside, and with approximately five hundred army officers and between six and eight thousand members and supporters in Tehran.”7 Indeed, despite Tudeh having foreknowledge of the 1953 coup, Tudeh suffered from so much intra-party factionalism that it fell into a “state of paralysis” and was unable to oppose the coup or even defend itself afterward.8
Second, the outcome of the Iran case is consistent with our statistical results. The 1953 coup against Prime Minister Mossadegh is an example of a successful covert leadership FIRC. In our study, we found that FIRCs of this kind heightened the likelihood of intervener-target conflict over the long term. This is exactly the pattern in the Iranian case. The United States obtained a pro-Western (if repressive) ally in the Middle East for twenty-five years, but a mortal enemy for the next thirty-eight years (and counting). Farsighted policymakers should consider the possibility that even FIRCs that pay immediate dividends can eventually go sour, with devastating consequences.
—Alexander B. Downes
—Lindsey A. O'Rourke
Alexander B. Downes and Lindsey A. O'Rourke, “You Can't Always Get What You Want: Why Foreign-Imposed Regime Change Seldom Improves Interstate Relations,” International Security, Vol. 41, No. 2 (Fall 2016), pp. 43–89, doi:10:1162/ISEC_a_00256. Further references to this article appear parenthetically in the text.
See John M. Owen IV, “The Foreign Imposition of Domestic Institutions,” International Organization, Vol. 56, No. 2 (Spring 2002), pp. 375–409; and John M. Owen IV, The Clash of Ideas in World Politics: Transnational Networks, States, and Regime Change, 1510–2010 (Princeton, N.J.: Princeton University Press, 2010).
It is also worth noting that if, for example, the goal of regime change is to ensure that a target state does not become communist—a core objective of many U.S. interventions during the Cold War—it may be far easier for the new regime to stay in power despite having been installed by an external power. The reason is that, by virtue of simply being “anti-communist,” the new regime satisfies the broad wishes of the intervening state and is thus free to pursue policies on the domestic front that reduce the likelihood that it will be deposed.
Michael Poznansky, “Stasis or Decay? Reconciling Covert War and the Democratic Peace,” International Studies Quarterly, Vol. 59, No. 4 (December 2015), pp. 815–826, at pp. 820–821, doi: 10.1111/isqu.12193.
The use of declassified documents, when available, can partially ameliorate this concern, because leaders are often more candid about their true motivations in private settings, where the incentives to lie, spin, or deceive are lower.
Alexander B. Downes and Lindsey A. O'Rourke, “You Can't Always Get What You Want: Why Foreign-Imposed Regime Change Seldom Improves Interstate Relations,” International Security, Vol. 41, No. 2 (Fall 2016), pp. 43–89, doi:10:1162/ISEC_a_00256.
Examples include John M. Owen IV, The Clash of Ideas in World Politics: Transnational Networks, States, and Regime Change, 1510–2010 (Princeton, N.J.: Princeton University Press, 2010); Martha Finnemore, The Purpose of Intervention: Changing Beliefs about the Use of Force (Ithaca, N.Y.: Cornell University Press, 2004); and Lindsey A. O'Rourke, “Secrecy and Security: U.S.-Orchestrated Regime Change during the Cold War,” Ph.D. dissertation, University of Chicago, 2013.
Gary King, Robert O. Keohane, and Sidney Verba, Designing Social Inquiry: Scientific Inference in Qualitative Research (Princeton, N.J.: Princeton University Press, 1994), p. 200.
On our matching procedure, see Downes and O'Rourke, “You Can't Always Get What You Want,” pp. 73–74, and appendix M of our supplementary materials, available online at doi: 10.7910/DVN/7Y4TD8.
For historical works that take this view, see ibid., p. 58 n. 48.
Maziar Behrooz, “The 1953 Coup in Iran and the Legacy of the Tudeh,” in Mark J. Gasiorowski and Malcolm Byrne, eds., Mohammad Mosaddeq and the 1953 Coup in Iran (Syracuse, N.Y.: Syracuse University Press, 2004), p. 106.
Maziar Behrooz, “Tudeh Factionalism and the 1953 Coup in Iran,” International Journal of Middle East Studies, Vol. 33, No. 3 (August 2001), p. 364.