Nobel Laureate Thomas Schelling described the tens of thousands dead in the Korean War as “undoubtedly worth it” because the United States needed to defend its reputation for resolve.[1] In 2013, President Barack Obama implicitly disagreed with Schelling, when he refused to carry out military operations based on his ‘red line,’ after Syrian President Bashar al-Assad used chemical weapons. Of all the hot-button debates about the use of military force, few arouse such recurring, intractable debates as whether, and when, it is worth fighting for reputation.
Keren Yarhi-Milo. Who Fights for Reputation: The Psychology of Leaders in International Conflict. Princeton: Princeton University Press, 2018. ISBN: 978-0-691-18034-2 (hardcover, $95.00); 978-0-691-18128-8 (paperback, $32.95).
31 January 2019 | https://issforum.org/to/ir11-10
Editors: Jeff D. Colgan and Diane Labrosse | Production Editor: George Fujii
PDF Version
Contents
- Introduction by Jeff D. Colgan, Brown University. 2
- Review by Carla Martínez Machain, Kansas State University. 5
- Review by Michael Poznansky, University of Pittsburgh. 8
- Review by Kenneth A. Schultz, Stanford University. 11
- Review by Rachel Elizabeth Whitlark, Georgia Institute of Technology. 14
- Response by Keren Yarhi-Milo, Columbia University. 19
Introduction by Jeff D. Colgan, Brown University
Nobel Laureate Thomas Schelling described the tens of thousands dead in the Korean War as “undoubtedly worth it” because the United States needed to defend its reputation for resolve.[1] In 2013, President Barack Obama implicitly disagreed with Schelling, when he refused to carry out military operations based on his ‘red line,’ after Syrian President Bashar al-Assad used chemical weapons. Of all the hot-button debates about the use of military force, few arouse such recurring, intractable debates as whether, and when, it is worth fighting for reputation.
It is into this debate that Keren Yarhi-Milo wades. She deftly sidesteps some of the swampiest terrain around the value of fighting for reputation. Instead, she investigates the question of which individual leaders are most likely to believe that reputation matters, and use their military capabilities accordingly. Her answer focuses primarily on a single psychological attribute: self-monitoring. High self-monitors are more likely to want to defend their reputation, and believe that they are capable of doing so with their actions, than low self-monitors are. To understand how self-monitoring affects leaders’ propensity to use military force, she considers a second crucial factor: the strength of an individual’s belief in the efficacy of military force. She calls ‘hawks’ and ‘doves’ those who believe military force is typically effective, or ineffective, respectively.
Four scholars offer their reactions to Yarhi-Milo’s second book in this roundtable.[2] The reviewers are unanimous that Yarhi-Milo’s book is important, timely, and well-argued. It approaches the question systematically, and offers multiple types of evidence in support of its claim.
Just as importantly, and perhaps more so, the reviewers offer constructive criticism – both for Yarhi-Milo’s project, and for the scholarly community doing research on individual leaders in the field of international relations (IR). Michael Poznansky and Carla Martínez Machain give us reason to pause before too hastily accepting the book’s claims. Poznansky points out that Yarhi-Milo’s theory expects different behavior for certain types of leaders during high- or low-stakes crises, and points out that the book’s empirical tests do not do much to differentiate between those types of crises. He also raises some methodological concerns about the book’s investigation of the role of individual advisors like Cabinet officials, which could be influenced by selection bias. Martínez Machain raises some similar concerns about the causal inferences. In particular, she wonders about the book’s assumption that high self-monitoring Presidents “take threats to US reputation for resolve personally” (29), and asks whether there might be circumstances under which even those types of individual are able to separate their own reputations from that of the state.
Kenneth Schultz offers a much broader critique of the research agenda regarding leaders in IR. He points out that we now have, as a field, a significant body of mid-level theory and empirical findings that connect individual attributes with the use of military force. Those include a leader’s pre-political beliefs, their experience before gaining office, their values, and their risk tolerance, as indicated by revolutionary or rebel activity.[3] The question Schultz raises is whether and how the various results cumulate. The plethora of factors that “matter” empirically leaves a great deal of indeterminacy. As he points out, “we could conclude that Lyndon Johnson escalated the Vietnam War because he was Southerner, because he believed that most threats derive from external aggression, because he served in the military but had only limited combat experience, or because he was a high self-monitor.” What is missing is anything close to a parsimonious theory of how leader psychology affects military conflict in IR.
Similarly, Rachel Whitlark wonders how a leader’s concern about a reputation for resolve weighs against his or her desire for a reputation for other attributes, like intelligence, prestige, etc. She uses the interactions between North Korea’s Kim Jong-un and President Donald Trump to illustrate this concern. She points out, “we can easily imagine that a leader’s level of narcissism might influence his or her decision-making during crises. Indeed, it is difficult to conclude from President Trump’s approach to these nuclear negotiations that a reputation for resolve is a central concern, even if he might be a high self-monitor …”
Whitlark also points to a second concern about heterogeneity in the type of cases that Yarhi-Milo raises. Many of the cases in the book involve a response to an attack on US troops, as occurred with Lebanon in 1983 or Haiti in 1993. Whitlark avers, “These types of attacks seem different in kind from the original theoretical explanation offered by the author, who notes that some leaders are likely proactively seeking to demonstrate resolve (183). In other words, are instances where leaders seek to show resolve substantively the same as situations where a leader responds to an attack?” Prospect theory might imply different behaviors in the realm of losses (for example, when the U.S. is attacked) than in the realm of losses.[4]
Yarhi-Milo thoughtfully responds to these criticisms, and considers new avenues for research. All told, the exchange leads to a deeper appreciation of the book’s strengths as well as its limitations.
Participants:
Keren Yarhi-Milo is a Professor of Political Science and International and Public Affairs and Associate Director of the Arnold A. Saltzman Institute of War and Peace Studies at Columbia University. In addition to the book reviewed here, she is also the author of Knowing The Adversary: Leaders, Intelligence, and Assessments of Intentions in International Relations (Princeton University Press, 2014).
Jeff D. Colgan is the Richard Holbrooke Associate Professor in the Department of Political Science and Watson Institute for Public and International Affairs at Brown University. His research focuses on international order, especially as related to energy and the environment. His book, Petro-Aggression: When Oil Causes War, was published in 2013 by Cambridge University Press. He has published work in International Organization, Foreign Affairs, World Politics, International Security and elsewhere. He also occasionally blogs at the Monkey Cage and Foreign Affairs.
Carla Martínez Machain is an Associate Professor of Political Science at Kansas State University. She received her Ph.D. in Political Science from Rice University in May 2012. She also received her B.A. from Rice University, in Economics and Political Science in 2007. Martínez Machain’s research focuses on foreign policy analysis, with a focus on military policy and international conflict. In particular, she has been engaged in work on conflict outcomes, defense cooperation, and repression. She has done much work on the various effects of U.S. troop deployments abroad on host states’ policy. Her work has appeared in various journals, including the Journal of Conflict Resolution, International Studies Quarterly, the Journal of Peace Research, and Conflict Management and Peace Science. Martínez Machain has also served as Section Program Chair for the SSIP section of the International Studies Association and region President for ISA Midwest. She currently serves as part of the editorial team for the ISA journal International Interactions.
Michael Poznansky is Assistant Professor of International Affairs and Intelligence Studies in the Graduate School of Public and International Affairs at the University of Pittsburgh. His research has been published or is forthcoming in the American Journal of Political Science, the European Journal of International Relations, International Studies Quarterly, the Journal of Global Security Studies, and the Journal of Peace Research. His current book project, Covert Action in the Shadow of International Law, examines the role of international law in a state’s decision to pursue regime change using covert or overt means.
Kenneth A. Schultz is Professor of political science at Stanford University. His research examines international conflict and conflict resolution, with a particular focus on the domestic political influences on foreign policy choices. He is the author of Democracy and Coercive Diplomacy and World Politics: Interests, Interactions, and Institutions (with David Lake and Jeffry Frieden), as well as numerous articles in peer-reviewed scholarly journals. He was the recipient the 2003 Karl Deutsch Award, given by the International Studies Association, and a 2011 Dean’s Award for Distinguished Teaching, awarded by Stanford’s School of Humanities and Sciences. He received his Ph.D. in political science from Stanford University.
Rachel Elizabeth Whitlark is an assistant professor in the Sam Nunn School of International Affairs at the Georgia Institute of Technology. Her research explores nuclear proliferation, counter-proliferation, and presidential decision-making. Her Security Studies article “Nuclear Beliefs: A Leader-Focused Theory of Counterproliferation,” won the 2019 International Studies Association’s International Security Studies Section, Best Security Article Award. She is currently completing a book that explores how U.S. and Israeli leaders decide to use preventive military force against adversarial nuclear programs. Her articles and commentary have appeared in such outlets as Security Studies, International Studies Quarterly, The Washington Quarterly, The Monkey Cage, and The Week, among others.
Review by Carla Martínez Machain, Kansas State University
In her book Who Fights for Reputation: The Psychology of Leaders in International Conflict, Keren Yarhi-Milo explores a much-studied but still widely debated question in foreign policy analysis: How much do the characteristics of state leaders matter in determining a state’s foreign policy actions? Quite a bit, argues Yarhi-Milo. This book is a detailed and careful, yet also powerful, statement of the role of the individual in foreign policy. Its mixed-method approach is representative of the work of a new generation of scholars who move beyond the qualitative/quantitative divide to produce robust and relevant work.
Yarhi-Milo’s work speaks to a broader question on the role of the individual in foreign policy. Her specific research question is about the conditions under which leaders will engage in belligerent behavior in order to build (or maintain) reputations for being resolute and not backing down in crises. It also addresses the question of how we can conceptualize and measure the abstract concept of resolve.
Yarhi-Milo argues that the leaders most likely to “fight for reputation” and to use force in order to shape perceptions of themselves and their country are those identified as “high self-monitors.” Self-monitoring is a trait defined in psychology as the individual “strategically cultivat[ing] their public appearances” (20). High self-monitors place importance on social status and are willing to modify their behavior (sometimes in ways that are incongruent with each other) in order to improve others’ perceptions of them. Thus, controlling for other factors, high-monitors are more likely to use force, and more specifically, they are more likely to use force to signal resolve. To ensure that leader hawkishness is not driving her observations, Yarhi-Milo distinguishes between high and low self-monitoring hawks and doves, categorizing leaders into four types based on these two dimensions. Though she finds that low self-monitor hawks are a rarity (this is one of the argument’s weaknesses, which I discuss later), she does identify a greater propensity for conflict involvement among high self-monitoring doves (“reputation believers,” 37) as compared to low self-monitoring doves (“reputation critics,” 35).
This book contributes to the foreign policy literature that focuses on the role of the individual. Foreign policy analysis has considered the role of individuals and their psychology for years. For example, Robert Jervis’s “Hypotheses on Misperception” continues to be highly relevant to foreign policy articles.[5] In addition, newer work by authors such as Rose McDermott, and Maryann E. Gallagher and Susan H. Allen explores how it is that the individual psychological and biological characteristics of leaders influence foreign policy outcomes.[6] This book adds to this literature by focusing on a specific, measurable characteristic in a leader’s psychology and applying it to a particular foreign policy behavior. In this sense this work is narrow, but it fits into a wider literature that builds an argument for psychology altering foreign-policy outcomes. It is important to note that Yarhi-Milo does not argue that system or state-level variables do not matter in determining a leader’s willingness to use force, or in their willingness to use force to signal a strong reputation. Rather, the point she makes is that even after we control for all of these other variables (which she does in both her quantitative and qualitative analysis), this individual characteristic can make a difference in a leader’s willingness to use force.
The book also contributes to a growing literature that attempts to either measure or explain the intangible measure of ‘resolve.’ For example, Jessica Weeks and Roseanne McManus study how leaders can effectively send signals of resolve to their opponents.[7] Though resolve is something that scholars have consistently argued matters in determining foreign policy choices, finding appropriate measures of resolve (that are not post-hoc ones) remains a challenge. Yarhi-Milo does not necessarily measure resolve per se, but she identifies a characteristic of leaders that have a preference for it and uses it as an independent variable to explain the willingness to use force in order to signal resolve.
One of the strengths of this book is its mixed-method approach. Yarhi-Milo proposes a theory that depends on measuring an unobservable characteristic—resolve. This means that it can be hard to determine causality and to argue that force is actually used to demonstrate resolve as opposed to other, perhaps more pragmatic, reasons. The mixed-method approach, in which the weaknesses of each method are addressed by another one, is the best way to deal with this problem. Yarhi-Milo tests the theory using three methods: she ran survey experiments in the United States and Israel, undertook a statistical analysis (in which militarized interstate dispute initiation and involvement by the different Cold War Presidents is the dependent variable explained by the Presidents’ self-monitoring), and offers a series of cases studies (Presidents Ronald Reagan, Jimmy Carter, and Bill Clinton) that use process-tracing to determine how leaders made their decisions during crises and how much importance they gave to reputation in determining whether to use force.
This book is a complete and robust one. Of course, all books have potential points of weakness. One point that Yarhi-Milo makes is that presidents who are high self-monitors will “take threats to US reputation for resolve personally” (29). This is an assumption set up in the theoretical development of the book, and, as an assumption, need not be tested. Still, something that I believe could be further explored is to what extent this is true. Yarhi-Milo provides three detailed case studies of presidents, and there are examples of presidents identifying the reputation of the country with their own (Clinton in Somalia is a clear example, 230). That said, I imagine that there is variation in the extent to which this happens, and I would have liked to have seen the author further explain whether identification of the President with the country’s reputation is higher in high self-monitors than in low ones. For example, are there circumstances under which individual leaders are able to separate their own reputations from that of the state? Is this dependent on the individuals’ personality or on the circumstances they find themselves in? Some illustrative examples of cases in which leaders did not identify their own reputation with the reputation of the state would have been helpful in allowing readers a better understanding of this connection and when it is strongest.
More broadly, this study’s specificity is also the book’s major shortcoming in that it makes its contribution somewhat narrow. The study focuses on the effect of a very particular personality trait and how it interacts with another characteristic, the level of hawkishness, to contribute to the use of force as a signal resolve. In addition, because of the rarity of low self-monitor hawks (the author notes that none of the U.S. presidents in her sample meet her selection criteria for low self-monitor hawks), the distinction comes down to only low and high self-monitor doves. Though Yarhi-Milo does find such an effect, the book could have more persuasively argued that these distinctions critically affected foreign policy outcomes. In other words, a greater discussion of what the counterfactuals could have looked like would be an effective way to persuade the reader that this personality trait matters.
Dealing with counterfactuals is of course always difficult. From a methodological perspective, an alternate way to present the analysis would be to take a cue from labor economics and use a decomposition model to simulate the counterfactual.[8] This would allow the author to analyze how much of a leader’s conflict propensity is due to their behavior (low versus high self-monitoring) as opposed to other observable characteristics such as the international environment they found themselves in, or even their hawkishness level.[9] In this case, even though we do not observe as many actual low self-monitor hawks, we could simulate what one would look like in term of its conflict behavior and thus emphasize the effect of the leader’s personality.
The book itself discusses a variety of avenues for future research. I particularly find a lot of promise in exploring how it is that resolve-signaling behavior can change, even if self-monitoring remains constant. Yarhi-Milo notes that leaders may learn from experience, and thus that their levels of hawkishness may evolve during their tenures as president. In addition, the book deals largely with the interaction between hawkishness and self-monitoring; what other factors can interact with self-monitoring to influence foreign policy behavior? Yarhi-Milo’s identification of the individual effect of self-monitoring allows other scholars to study its interactions with other variables, such as gender. For example, work by Valerie Hudson and Patricia Leidl asks questions about the foreign policy behavior of female leaders.[10] Would self-monitoring female leaders behave the same way as male ones when signaling resolve? There is also room to explore the role of self-monitoring in the behavior of non-state leaders, such as the leaders of rebel or terrorist groups. For example, Oliver Kaplan discusses the role that cognitive dissonance plays in determining the willingness of rebel leaders to victimize civilian populations.[11] Future work could apply Yarhi-Milo’s framework to non-state leaders, and study whether the same principles hold. This approach would have the added advantage of having a large pool of observations.
Overall, this book is a methodologically sound and valuable contribution to both the foreign policy analysis and political psychology literatures. The question of how leaders’ self-monitoring levels lead them to be more willing to engage in conflict is one that is relevant for both policy and academic communities. This book is a welcome addition to a growing field of foreign policy analysis research that considers variables at various levels of analysis and does not shy away from thinking about how they interact with each other. It is an engaging read that will likely have a great impact on the study of foreign policy.
Review by Michael Poznansky, University of Pittsburgh
It’s been seven years and we still can’t stop talking about U.S. President Barack Obama’s ‘red line’ incident. The story goes like this. During the early stages of the Syrian Civil War in 2012, Obama issued a warning to Syrian President Bashar al-Assad, stating that “[w]e have been very clear to the Assad regime that a red line for us is we start seeing a whole bunch of chemical weapons moving around or being utilized. That would change my calculus.”[12] A year later Assad used sarin gas against civilians, testing Obama’s resolve. Rather than following through on his threat to use force, Obama settled for a deal in which the U.S. and Russia would oversee the destruction of Syria’s chemical weapons. The reaction was severe. Just last year, President Donald Trump tweeted: “If President Obama had crossed his stated Red Line In The Sand, the Syrian disaster would have ended long ago! Animal Assad would have been history!”[13] Regardless of whether Obama’s critics are right, this episode raises broader questions about reputation in international politics. In particular, why are some leaders willing to fight to project a reputation for resolve while others are seemingly unwilling to do so?
Who Fights for Reputation: The Psychology of Leaders in International Conflict provides an answer. Keren Yarhi-Milo argues that the interaction of two factors explains variation in leaders’ willingness to fight for reputation. First, whether leaders are high or low self-monitors matters. High self-monitors are able, and believe in their capacity, to shape how others see them. Low self-monitors exhibit the opposite tendency. High self-monitors are thus more willing to fight to demonstrate resolve than low self-monitors. Beliefs about the efficacy of military force matter as well. While both high self-monitor hawks and doves are likely to fight for face, the latter may attempt to use non-military means before resorting to force. Low self-monitor hawks may be inclined to use force, but typically not for reputational reasons. Low self-monitor doves are reluctant to use force in most cases, especially to save face and show resolve.
Anyone who reads this book will be struck by the methodological firepower Yarhi-Milo brings to bear. Who Fights for Reputation employs three distinct kinds of tests. In Chapter 3, Yarhi-Milo teams up with Joshua Kertzer to validate the argument’s microfoundations using a survey experiment on both American and Israeli populations. For the quantitative analysis in Chapter 4, Yarhi-Milo uses an original survey of 68 presidential historians to identify whether U.S. presidents from Harry Truman to George W. Bush were high or low self-monitors. Chapters 6 through 8 contain detailed case studies of how presidents Jimmy Carter (a low self-monitor dove), Ronald Reagan (a high self-monitor hawk), and Bill Clinton (a high self-monitor dove) reacted to crises bearing on America’s reputation for resolve.
This book makes several important contributions. First, it joins a growing body of literature in international relations that takes leaders and their individual differences seriously.[14] Second, it moves beyond the literature’s emphasis on whether reputations matter (an important question in its own right) and seeks instead to explain under what conditions leaders believe reputation matters.[15] Third, it serves as a model for how to do innovative, multi-method research on a question of enduring importance.
As with all good books, I came away from reading Who Fights for Reputation with several questions about the argument and evidence. To begin with, the survey experiment chapter does not explicitly distinguish between high and low-stakes crises, which is central to the book’s theoretical argument. Yarhi-Milo and Kertzer argue that “the effect of self-monitoring on the hawks’ willingness to fight when treated with reputational considerations should be much less pronounced [than for doves] … because … both low and high self-monitor hawks are already predisposed to fight to achieve other, nonreputational benefits” (45). That is technically true, but only when the material stakes involved are high. When the stakes are negligible, low self-monitor hawks (reputation skeptics) should “be less motivated to fight…” (37). This dynamic plays out in Chapter 7 when Defense Secretary Weinberger, a low self-monitor hawk, advocated for restraint in Lebanon and Grenada since “vital issues [were] not at stake” (206).
Without knowing whether the scenarios provided to the American and Israeli samples are high or low-stakes crises, it is hard to say whether the experimental results fully validate the theory. If the crisis scenarios were perceived by respondents as involving high stakes, then Yarhi-Milo and Kertzer’s predictions would indeed be supported by the evidence, which finds no real differences in the hawks’ willingness to use force across the various treatments. If, however, the respondents perceived the crisis scenarios as involving low stakes, it would change the interpretation. This leaves the door open for future researchers to conduct survey experiments in which the stakes of a crisis vary specifically to assess whether the hawk hypotheses hold. If Yarhi-Milo is right, we would only see a null effect in high-stakes scenarios; low-stakes scenarios should reveal separation between high and low self-monitor hawks.
This important issue—distinguishing between high- and low-stakes crises—also rears its head in the remaining chapters. In Chapter 4, the quantitative analysis, Yarhi-Milo utilizes the Militarized Interstate Disputes dataset as the main dependent variable. The MIDs data, however, do not distinguish between crises that were vital to the disputants and those that were not. This, in turn, limits the inferences we are able to draw from the analyses for the reasons stated above. A similar issue arises in the case study chapters. To be sure, coding crises as vital or nonvital is undoubtedly challenging, especially in observational contexts. But further discussion is warranted given its importance to the theory. Verifying experimentally, quantitatively, and qualitatively that individuals who exhibit high self-monitoring tendencies do indeed fight on nonvital issues while low self-monitors do not is among the most promising contributions.
A second question turns on the role and relevance of advisors. Generally speaking, the impact of advisors on foreign policymaking is an understudied phenomenon that is only just beginning to receive serious attention in scholarly outlets.[16] Yarhi-Milo’s efforts to incorporate them into the case studies is therefore refreshing. But there is little discussion of why some advisors were selected as being more influential than others. This is especially notable given that the omitted ones seem relevant in key instances. In Chapter 6, Carter’s National Security Advisor, Zbigniew Brzezinski, and his Secretary of State, Cyrus Vance, are identified as the two key players while Secretary of Defense Harold Brown, who is described as being involved in some of the decisions under study, is discussed only in passing. Similar comments could be made of Chapter 7, wherein Reagan’s CIA Director (William Casey), National Security Advisor (William McFarlane), and Vice President (George H.W. Bush) make notable appearances even though Secretary of State George Shultz and Secretary of Defense Casper Weinberger are identified as the two most influential advisors. Offering a defense of these decisions and some ex ante coding criteria for identifying advisor relevance would have been useful.
Related to this is the lingering question about whether and under what conditions advisors matter. In the Carter cases, the President routinely ignored National Security Advisor Brzezinski, a high self-monitor who often advocated fighting for face, and sided instead with Secretary of State Vance, a low self-monitor who reflected his inclinations not to. In most of the Reagan cases, the President listened to advisors like Secretary Shultz who shared his general proclivity for action and desire to show resolve. Indeed, Yarhi-Milo observes that “[t]he only time [Reagan] favored Weinberger was in the last phase of his fight to keep [U.S.] forces in Lebanon, where it became clear Congress would not allow him to continue the mission” (220). If presidents simply listen to advisors who share their own inclinations, this would seem to cut against the notion that studying them is intrinsically important. The really interesting cases would be those in which presidents differed from advisors in their preferred policy but were persuaded to change course. Figuring out when this is most likely to happen and when presidents will simply lean on those who affirm them is a valuable area for future research.
Finally, one issue that comes up in various places that would be useful to explore in greater depth is covert action’s relationship to reputation. In various places, Yarhi-Milo invokes presidential preferences toward covert action to code leaders’ military assertiveness. Carter’s and Clinton’s apparent aversion to the so-called quiet option is used as evidence that they were doves (129-130; 228-229). Conversely, Reagan’s attraction to covert action is used to classify him as a hawk (177-178). A broader question related to this tool of statecraft, however, is what a leader’s use of it says about their willingness to show resolve. For instance, does Reagan’s secret support for the Afghan mujahideen actually serve as a signal of resolve since American complicity was widely suspected, or does it betray an unwillingness to devote the resources necessary to achieve outright victory?[17] Although Yarhi-Milo argues that “covert action could still improve America’s reputation for resolve vis-à-vis the Soviet Union because Moscow and regional American allies would draw inferences about US resolve from the ever-greater resources expended on Operation Cyclone” (186), what does it imply about Reagan’s commitment to show resolve given that escalation concerns prevented a more overt response? In short, even if the decision to pursue covert action sends more of a signal of resolve than doing nothing, an unwillingness to escalate by going overt may have watered down the message.
These various questions aside, Who Fights for Reputation is an impressive accomplishment that will be required reading for anyone seeking to understand leaders, reputation, and crisis behavior. The breadth and depth of the analysis set a high bar for scholars interested in entering these debates.
Review by Kenneth A. Schultz, Stanford University
Given that academic political science often lags behind the real world by many years, it should be noted that the recent renaissance of scholarship on how leaders’ personal traits affect their foreign policy choices started before the presidency of Donald Trump gave us a stark reminder that ‘individuals matter.’ Works like Leaders at War, by Elizabeth Saunders, and Why Leaders Fight, by Michael Horowitz, Allan Stam, and Cali Ellis, helped reignite interest in how the psychological and intellectual baggage that leaders bring with them into office affects when and how they use military force.[18] Keren Yarhi-Milo’s Who Fights for Reputation? is a strong addition this research agenda, providing a model for how to rigorously demonstrate the effect of an individual trait on an important phenomenon: the willingness to fight, not simply for the issue at hand, but in order to establish a reputation for toughness that extends to other issues and future interactions.
International relations scholarship has had no shortage of things to say about reputation, and the fact that Yarhi-Milo found a new angle on old problem is no small feat. Most recent work focuses on the question of whether adversaries take past actions into account when assessing a country’s resolve,[19] and, if so, whether reputations adhere to states or to individual leaders.[20] There is comparatively less work on when it makes sense to invest in reputation in the first place. Classic models assume that the value of a reputation depends on the strategic context, including how likely a future challenge is and how similar the issues are from one interaction to the next.[21] Who Fights for Reputation? argues that there is also unappreciated heterogeneity across individuals. In particular, leaders vary in terms of how much they “self-monitor” (11), a personality trait that influences how much people think about how others regard them and how willing they are to engage in deceptive behavior to shape that image.
An exciting feature of the newest generation of work on leaders is that it genuinely locates the main source of variation in individual traits. Not all research on leaders, even those rooted in psychological approaches, does this. For example, in Leaders and International Conflict, Goemans and Chiozza assume that leaders are motivated by their political and personal survival, but the main source of variation is regime type, which determines how conflict outcomes affect leader tenure. Similarly, a good deal of the earlier work on psychological and cognitive influences focused on biases and limitations that were effectively universal: traps and mistakes that all people are prone to, such as selective attention, confirmation bias, or the framing effects identified in prospect theory.[22] While important, the fact that these qualities are universal means that they are not themselves the drivers of variation in outcomes. The latest wave of research, by contrast, is more focused on qualities that vary across individuals: i.e., personality traits, preexisting beliefs, or prior experiences. The advantage of these theories is that they are well-placed in principle to explain and predict variation in outcomes across individuals.
Nevertheless, the task is not trivial, and the evidentiary challenges are significant. What is particularly impressive about Who Fights for Reputation? is how effectively it meets these challenges. The variety of evidence brought to bear in support of this claim is very impressive. The book combines survey experiments on American and Israeli public samples, a quantitative analysis on the use of force by U.S. presidents, and careful historical analysis of decision making by three presidents. In addition to providing a rich understanding of some important episodes in U.S. foreign policy, the statistical evidence points to strong effects: Even after controlling for a number of domestic and international variables, high self-monitoring presidents are involved in nearly twice as many militarized interstate disputes (MIDs) and initiate nearly twice as many MIDs as low self-monitoring presidents (100). The cumulative effect of these tests is impressive, and many potential concerns are anticipated and addressed along the way.
Let me mention two hurdles in particular that Yarhi-Milo clears quite ably. One major challenge facing this kind of analysis is to identify the presence or absence of a trait in a way that is independent of the outcome the trait is hypothesized to influence. Recent work has taken a number of strategies to address this problem. Saunders infers presidents’ beliefs about the desirability of intervention by analyzing speeches and writings from before they became president; Horowitz, Stam, and Ellis collect data on leader background and experiences from before they ascended to that role; Dafoe and Caughey proxy variation in honor concerns by differentiating Southerners and non-Southerners.[23] For Yarhi-Milo, placing survey respondents on a self-monitoring scale presents no problem, since they can be asked standard questions designed for this purpose. Doing the same for U.S. presidents, most of whom are dead and none of whom could be asked to take a personality test, presents a more difficult problem. To deal with this, Yarhi-Milo uses an innovative technique: a survey of presidential historians that asked them to answer the standard questions based on their knowledge of the individual. To try to ensure that these answers were not influenced by the foreign policy choices she examines, the historians were asked to think about the president’s personality from the time before he was first elected, and the instructions were careful not to tip off respondents into thinking about the foreign policy outcomes that self-monitoring is thought to influence. There are any number of concerns one might have about this exercise; Yarhi-Milo, to her credit, is aware of them, and takes great pains to assess possible weaknesses (see, in particular, 76-78).
A second concern is that there are large number of ways that people differ from one another in terms of their values, personality traits, beliefs, and cognitive abilities. How do you choose which one(s) to focus on? How do you show that it is your trait of interest that explains the variance, not some other individual factor? With respect to first question, Yarhi-Milo makes a convincing case that self-monitoring is theoretically relevant to reputation building. In this context, the most likely alternative, and potential confounder, is a president’s general beliefs about the efficacy of military force, a feature that is connected political ideology, personal values, and prior experiences.[24] It is quite plausible that hawkishness derives in part from a belief that force is useful because of its reputational effects. To address this, Yarhi-Milo explicitly introduces the hawkishness into her theory as a separate dimension, arguing that this trait interacts with self-monitoring to create four types of individuals depending on whether they score high or low on these two dimensions. Her basic expectation is that high self-monitoring has a larger effect on doves than on hawks because reputational concerns can override the former’s general skepticism about the use of force. She also independently measures the hawkishness of both her survey respondents and the U.S. presidents and includes the measures in her quantitative and qualitative analyses to assess the way in which these two dimensions work together.
As a self-contained exercise, the book is very well executed, persuasive, and interesting to read. It is a model for doing research into personality traits and foreign policy, and other scholars in this area should take inspiration from the multi-method approach. To the extent that I have concerns, they revolve around the implications of these results for the broader research agendas on the use of force and individual-level characteristics.
First, how should these results change the way scholars empirically analyze the use of force? One clear message is that there is leader-level heterogeneity arising from variation in self-monitoring and that studies of, say, U.S. participation in MIDs need to control for this factor. However, most such studies already include president-specific fixed effects—effectively assuming that each leader has a different average conflict propensity—which in most applications provide a more efficient way to control for this trait, along with any other fixed attributes that vary across presidents. Moreover, given the costs of trying to measure self-monitoring in a wider set of leaders, both across countries and across time, as well as the danger of measurement error in instruments like a historian survey, introducing leader-specific fixed effects is likely to be the best strategy for most researchers. But doing so means side-stepping the specific argument about self-monitoring, rather than engaging it.
A more fundamental question for this entire research agenda is whether and how the various results cumulate. Individuals vary from one another in a myriad different ways. Saunders points to preexisting beliefs about sources of instability; Horowitz, Stam, and Ellis point to conditions of upbringing and prior experiences; Dafoe and Caughy highlight exposure to cultures of honor; now Yarhi-Milo focuses on self-monitoring. On their own, each work makes a case for the importance of its preferred variable or set of variables. Collectively, the emerging image is that ‘individuals matter’ but that they differ from one another in multiple, likely non-overlapping ways. From this set of writing alone, we could conclude that President Lyndon Johnson escalated the Vietnam War because he was Southerner, because he believed that most threats derive from external aggression, because he served in the military but had only limited combat experience, or because he was a high self-monitor.
How should we sort this out? Are these independent factors that work additively, or are some causally prior to others? Yarhi-Milo tells us that self-monitoring has a large genetic component and that this disposition becomes relatively stable after childhood (23). This means that it is unlikely to be a product of later experiences. At the same time, she is reluctant to see it as a reflection of other underlying personality traits, such as the Big Five (neuroticism, extraversion, openness, agreeableness, and conscientiousness, 76). Thus, the possible ways in which individual leaders differ from one another is only starting to be tapped. As this research agenda progresses, it will no longer be enough to show that ‘individuals matter’ when controlling for domestic and international factors; we also need some guidance on which of the many ways people differ from one another matter most.
Review by Rachel Elizabeth Whitlark, Georgia Institute of Technology
What explains variation in concern over reputation for resolve? When are leaders willing to use military tools to demonstrate such a reputation? These are the questions that Keren Yarhi-Milo sets out to answer in Who Fight for Reputation? The Psychology of Leaders in International Conflict. Whereas the existing literature exploring the role of reputation in conflict generally focuses on the strategic environment or offers crisis-specific explanations, Yarhi-Milo argues that variation in state behavior and willingness to fight for reputation stem from variation in national leaders’ psychological dispositions. Specifically, she explains how individuals vary in terms of their propensity for self-monitoring—the degree to which they regulate their behavior in social situations. This self-monitoring behavior is a stable trait stemming from both genetic and early childhood environmental influences (2). Accordingly, those leaders with the greatest willingness to fight for “face” should be those who believe they can manipulate their own state’s reputation and who want to appear resolute, steadfast, or strong in the eyes of others during international crises (11). Those individuals with these characteristics are Yarhi-Milo’s high self-monitors: They are inclined to modify their behavior strategically in order to cultivate status-enhancing images, will be particularly concerned about reputation for resolve, and will seek to act primarily to enhance their social status. By contrast, leaders who do not possess these views should be markedly less likely to fight for face (11). If we control for leaders’ general beliefs about the use of force, such low self-monitor leaders are less likely to use military force to fight for reputation than their high self-monitor counterparts (14). Indeed, Yarhi-Milo demonstrates that high self-monitor leaders are more than twice as likely to engage in or to initiate militarized interstate disputes than low self-monitors.
More specifically “self-monitoring concerns the extent to which individuals strategically cultivate their public appearances” (20). High and low self-monitors differ in their interest in and skill at managing such impressions. High self-monitors are sensitive to external cues about their behavior and modify it as necessary to achieve some end. By contrast, low self-monitors lack the interest and ability to manipulate their behavior strategically, caring more for honesty and authenticity (22). Following Yarhi-Milo’s coding scheme, Presidents John F. Kennedy, Lyndon Johnson, Ronald Reagan, and Bill Clinton were high self-monitors, while Gerald Ford, Jimmy Carter, and George H.W. Bush were low self-monitors. Harry S. Truman, Dwight Eisenhower, Richard Nixon, and George W. Bush cannot be classified (74). In addition to their self-monitoring dispositions, leaders also vary in terms of their beliefs about military assertiveness. These views (hawkish or dovish) regarding the efficacy of force have a moderating effect on the self-monitoring status, and when combined, shape how willing leaders are to use force to secure reputation (35). The author then selects cases for analysis within the full population to assess as much of this variation as is possible, yielding treatments of Reagan as a high self-monitor hawk, Clinton as a high self-monitor dove, and Carter as a low self-monitor dove (117).[25]
Yarhi-Milo examines these propositions through an analysis of American presidents both during and after the Cold War. She deploys a mixed methodological analysis to confirm that her argument holds—that so called high self-monitors will be particularly concerned about reputation for resolve and will be willing to fight in order to protect or preserve it. As she demonstrates, this desire to appear resolute stems less from tangible or instrumental benefits that reputation for resolve might offer, and more from an intrinsic psychological benefit image confers in their mind (12).
The contributions of Who Fights for Reputation? are manifold. First, despite engaging complex psychological phenomena, the author offers an intuitive central story drawing on a commonly held notion that reputation should somehow matter in the complicated world of international politics. Second, Yarhi-Milo expands the ongoing debate over the nature and implications of audience costs (the penalty a leader might incur from his or her constituency if seen backing down in a crisis),[26] which is currently well-represented in the literature, to include how reputational concerns are specifically relevant to such cost considerations. Likewise, she adds to the burgeoning literature on the role of individual leaders and how they are influential in international relations, international security, and foreign policy (4).[27] More specifically, Yarhi-Milo demonstrates how psychological traits have a causal role in foreign policy decision-making processes and outcomes, as distinct from the more well-known role of beliefs,[28] experiences,[29] and biases.[30]
Methodologically, Yarhi-Milo sets an incredibly high standard for social science research. Yarhi-Milo calls her approach both layered and stacked—designed to test the theory from its microfoundations through to its performance in a set of historical cases (15, 269). Each next step complements previous choices and accounts for the weaknesses or vulnerabilities of the other tools employed. To begin, a cross-national experiment (designed with Joshua Kertzer) tests the microfoundations of the psychological argument that individuals differ in their degree of concern for reputation and documents the existence of variation in this so-called self-monitoring status across the general population (as distinct from within the population of leaders). Next, to measure the study’s independent variables—leaders’ self-monitoring dispositions and beliefs regarding the efficacy of force—Yarhi-Milo leverages an original survey of 68 presidential historians who offer their expert opinions regarding the concerns over reputation by the presidents they know best. The survey results complement the author’s own coding decisions and lend support for the presidential categorization the author presents. Third, statistical analysis demonstrates correlation between the aforementioned explanatory variables and the outcome variable, the application of military instruments to project an international reputation for resolve (16), as well as the substantive significance of high self-monitor behavior (17). Finally, case analysis augments the author’s classification of American leaders by their self-monitoring type and allows her to process-trace the theorized causal mechanisms of the psychological characteristics from the pre-presidential period through to their role during the crises of interest.
The collective impact of the analysis is impressive. Yarhi-Milo answers her original question, under what conditions leaders will be willing to use military instruments to project a reputation for resolve, by offering a probabilistic argument where high self-monitor hawks are more likely to use force to demonstrate resolve and enhance reputation (13). By contrast, high self-monitor doves will be more reluctant to use force and escalate in crisis situations (14). The historians’ survey and inclusion of advisors’ views are especially creative approaches in this effort and are the kinds of tools that should be included more in the literature.
Nevertheless, as with all scholarship, challenges persist, and issues remain unexplored. Below, I comment on some of these challenges as they affect the research design, the theory, and the case analyses.
While the research design described above is extensive, it does raise questions. In particular, a key exclusion criterion appears problematic. The author describes that she excludes cases in which U.S. national interests were at stake, in as much as those cases would make it difficult to disentangle the reputational versus interest-driven motivations of the presidents (118). While the logic behind this decision is understandable, including such situations might make a useful addition. This is the case for two reasons. First, the distinction between cases where national interests are at stake and where they are not appears murky given that reputational concerns seem part and parcel to national interests. In other words, reputation for resolve does not exist in a vacuum and is often if not always engaged towards some larger strategic goal, even if indirectly. While high self-monitors may care about reputation intrinsically, reputation’s utility is still in service of larger national considerations. Moreover, the national interest is inherently a subjective consideration, so it is problematic to attempt to sort cases for analysis accordingly. Second, it is difficult for this book to demonstrate policy relevance if the book’s analysis only centers on cases without national interest considerations, as historically these are the episodes of central importance to U.S. policy. While I return to other policy considerations below, note that the deliberate inclusion of such cases would go a long way in demonstrating the theory’s relevance and substantive impact. Thus, if Yarhi-Milo could demonstrate not only that U.S. national interests were at stake and that the leader took military action explicitly for reputational concerns (in addition to or in place of national interest), such evidence would be compelling. Ideally, a case would also exist in the historical record where interests were threatened and called for military action to secure, but reputational concerns convinced the leader to abstain. Such a case, where interests would require one action, but where reputation would require an opposite or different action, could usefully serve to highlight the work reputation was actually doing. Admittedly, this is methodologically difficult and potentially historically impossible, but even some kind of approximation or hypothetical example would be helpful to think through.
Thinking theoretically, it is not obvious that reputation for resolve is the only relevant status concern for state leaders as the author asserts (21). While certainly reputation for resolve is one such possibility, another potential reputational concern could be one’s ego and attendant considerations. Indeed, the recent summit meetings between North Korea’s Kim Jong-un and President Donald Trump seem to suggest that both leaders are concerned not only with the substantive interests at stake with the nuclear negotiations, but also with how they are perceived as individuals—intelligent, prestigious, savvy, etc.[31] These and other characteristics that are more often associated with individuals may be relevant for analysis, and we can easily imagine that a leader’s level of narcissism might influence his or her decision-making during crises. Indeed, it is difficult to conclude from President Trump’s approach to these nuclear negotiations that a reputation for resolve is a central concern, even if he might be a high self-monitor, an issue with which the author unfortunately does not engage.
Perhaps most importantly, Who Fights for Reputation? leaves the reader to wonder whether backing down in a crisis or retreating following an attack is substantively the same as seeking out opportunities to demonstrate one’s resolve. Many of the cases the author explores are instances where the leader at the time is forced to respond to an attack on troops or other American assets, as happened with Lebanon in 1983 (198) or Haiti in 1993 (241-250). These types of attacks seem to be different in kind from the original theoretical explanation offered by the author, who notes that some leaders are likely to proactively seek to demonstrate resolve (183). In other words, are instances where leaders proactively seek to show resolve substantively the same as situations where a leader responds to an attack? The latter type of event recalls prospect theory and how actors bargain or fight harder for assets that are already in their possession but are less committed (resolved) for items not yet obtained.[32] Indeed, prospect theory suggests that not wanting to cut and run should be substantively different from seeking to fight proactively for reputation. It would have been useful for the author to engage prospect theory directly, even by distancing her own theory from it, and also discuss potentially important substantive differences between the cases under exploration. A discussion on this point is therefore warranted given that it would help clarify how reputation interacts with actors’ identification of domains of loss or gain, as well as shed light on the very nature of the episodes under analysis, as they may not all be created equally.
Last, on the topic of the cases, there are instances in which the material presented provides substantial evidence for complementary explanations, which might have been explored for a more nuanced analysis. For example, with President Carter, national interests appear to obscure the reputation-based argument Yarhi-Milo seeks to describe. This manifests during the author’s discussion of the Horn of Africa case, which appears to rest not on Carter’s low self-monitoring dove status as Yarhi-Milo intends, but rather on the fact that the President had different priorities than his advisors (134-147). Specifically, it seems that Carter prioritized arms control and the ongoing-SALT II negotiations above all else and was unwilling to jeopardize the delicate diplomacy through interventions elsewhere. It is this prioritization of interests that appear to have been driving Carter’s decisions, as opposed to any reputational concerns (or lack thereof).
Similarly, Yarhi-Milo uses the Soviet invasion of Afghanistan to highlight Carter’s low self-monitoring dove status and argue that because of this characteristic, he should not have been persuaded to use force by arguments concerning U.S. reputational damage. Indeed, while Carter may well be a low self-monitoring dove, the bulk of the change in behavior again appears to have been interest-driven. Specifically, only as time passed did Carter perceive U.S. national interests to have been at stake because of growing Soviet expansionist aims (123, 158). This revelation seems somewhat problematic, not only because the research design would rule out such cases for analysis, but also because it is not necessary to understand anything about Carter’s psychology to understand his actions. In other words, his low self-monitor status drops out of the explanation required for the change over time—that expanded Soviet activity in Afghanistan required a U.S. response. While it is understandable that this case is included in as much as it confirms the theory’s expectations—that Carter was not swayed by the reputational element of National Security Advisor Zbigniew Brzezinski’s argument for action—Carter’s behavior does appear to have been driven by concerns over interests and changing views thereof. This conclusion challenges the argument the author makes.
A similar difficulty plagues the Reagan episodes, as much argument is implied through the evidence rather than demonstrated, which again leaves much to the reader’s interpretation. In numerous instances, the author asserts that reputation was driving Reagan’s actions; unfortunately, given the evidence marshalled, this is but one of many possibilities. For example, Yarhi-Milo describes at length how in the 1980s Reagan demonstrated U.S. resolve by increasing assistance to the mujahedeen in Afghanistan (185-190). While this is factually accurate, it is not obvious that reputation was doing the causal work. A likely competing contender is that U.S. behavior can be explained by material concerns or mission-specific factors. For example, quoting from National Security Decision Directive 75, the author highlights that “[t]he U.S. must rebuild the credibility of its commitment to resist Soviet encroachment on U.S. interest and those of its Allies and friends, and to support effectively those Third World states that are willing to resist Soviet pressures…” to claim that this “establishe[s] the restoration of American credibility as a policy priority” (188). The section continues, [Washington must provide] “security assistance and [conduct] foreign military sales,” to demonstrate that the United States would not cede the so-called periphery…” (188). While certainly credibility and by extension reputation are salient here, the questions as to why or to what end remain ambiguous. While it is a plausible inference that assistance was offered specifically to improve U.S. reputation for resolve or to demonstrate credibility of commitments, the move could also have been indicative of steps taken to balance Soviet actions or the desire to defeat rather than contain the adversary, as Reagan preferred. Noting that reputation is a salient feature of a case is not the same as assigning it a causal role for the behavior undertaken. In the end, it is not clear that we can know definitively; the author’s recognition of the uncertainty and ambiguity that exists with all historical work would go a long way to assuage such concerns.
Beyond the above considerations, it is worth pondering the wider applicability of Who Fights for Reputation? Two issues come to mind. First, at the conclusion of Yarhi-Milo’s book, the policy implications seem potentially overstated. As the author notes, “high self-monitors engage in, initiate, and prevail in more crises that involve coercive military instruments compared to their low self-monitor counterparts” (100). In this way, high self-monitor status is a more important predictor of crisis behavior than party affiliation (100). Indeed, the author believes that this should be important information for how voters select presidents. While this is plausible conceptually, one wonders how voters can practically assess the self-monitoring status of a candidate. Yarhi-Milo undertook an extensive research effort to categorize her subjects; the notion that the average voter could reasonably assess these characteristics among presidential candidates seems somewhat outlandish. Given that the average American voter is often disengaged from politics, especially in matters of foreign policy, evaluating such significant information may not be attainable. Yarhi-Milo suggests that voters are attracted to high self-monitors because they are relatable and personable and that they therefore vote them into office accordingly (102, 266). A different research design is necessary to assess this possibility.
Second, given the very small variation that exists within the population of American presidents—Carter is the only low self-monitor dove, and the full spectrum of psychological views is rather narrow (116-117, 266)—how policy relevant is this analysis? The author herself acknowledges this limitation: “In the United States, most see value in actually fighting for reputation for resolve; but occasionally we encounter those reluctant to fight even when their advisors think they should” (7, 56). If this suggests that most people are hawks and dovishness is rare, is this not a sufficient indicator useful for concerned citizens and analysts alike? While Yarhi-Milo diligently challenges the marginal improvement her theory offers beyond simple considerations of leaders’ beliefs regarding the effectiveness of force (88), as a short-hand or heuristic, a skeptic will not be convinced that much else is necessary. The theory on offer is without a doubt interesting, but given the complexity and narrow substantive applicability, to what end?
Last, it is striking that the best evidence for the reputational theory comes from the presidents’ advisors. In both the Carter and Reagan chapters, the discussions between the presidents and their respective teams, as well as deliberations between the advisors themselves, offers some of the most compelling evidence for the range of self-monitoring status and its connection to policy preferences. This is true both of Secretary of State Cyrus Vance and Brzezinski with respect to Jimmy Carter and Secretary of Defense Casper Weinberger and Secretary of State George Shultz with respect to Ronald Reagan. The evidence is so compelling in fact that at times it causes the reader to wonder whether or not the story at hand was actually one of policy advocacy or of leaders’ susceptibility to persuasion by their talented advisors (for example, 103, 209). Indeed, Yarhi-Milo’s own words raise this concern: “The Reagan presidency thus consisted of a series of public performances, with Reagan playing the role that his advisors and speechwriters crafted” (174). It is difficult not to take this sentence as indicative of the fact that advisors have strategic craftsmanship abilities on serious matters of decision-making. While Reagan may be an extreme case, the impressive discussion of the role of advisors and their divergent self-monitoring dispositions should encourage us to take a closer look at their causal role. Indeed, new literature is already heading in this direction.[33]
In conclusion, with Who Fights for Reputation? Yarhi-Milo offers an important book in its aims, its expansive research design, and the contribution it makes to our understanding of the role of reputation in international politics. This book will likely serve as a model that other scholars will seek to emulate as they pick up the mantle of investigating psychological factors in international relations and design comprehensive analyses to guide their efforts. Indeed, as Yarhi-Milo herself notes, “I hope this book will infuse new ideas into the discussion of reputation in international relations” (278). It is difficult to imagine events transpiring any other way.
Response by Keren Yarhi-Milo, Columbia University
I would like to thank the editors of H-Diplo/ISSF and Jeff Colgan for organizing this roundtable for Who Fights for Reputation. I am grateful to the reviewers for their thoughtful comments and their engagement with the book’s argument and findings. They provide a number of useful insights, some that I had wrestled with while writing the book, and others that require further reflection and suggest fruitful areas for future research. In what follows, I try to address the major points raised by several reviewers.
First, several reviewers raise questions about the theory pertaining to the relationship between leaders’ reputation concerns and other factors that might be influencing their decision-making. For example, Rachel Whitlark raises questions about whether reputation for resolve is the only relevant status concern for high self-monitor leaders. Indeed, as the discussion in the book notes, the positive relationship between status concerns and reputation for resolve is not one that we should observe across all different domains. In other words, reputation for resolve is certainly not the only possible status concern of high self-monitors. The theory I develop in this book makes it clear that status concerns will look different across various domains. This is why I limit the scope conditions of the theory to cases of military crisis situations, where I argue that reputation for resolve is a salient and relevant status concern. But I also argue that reputation for resolve might not be a salient status concern in other issue areas that do not involve military crises (29). Put differently, while the status concerns may not be identical or equally salient across domains, the theory should still apply in other issue areas as long as the behavior of high self-monitor can be shown to be linked to concerns about image and social status.
Experimentally, Josh D. Kertzer and I are able to show that in hypothetical military crisis situations, reputation for resolve is an important status concern: high-self monitor doves given a neutral treatment reminding them that the eyes of the international community were upon them were significantly more likely to fight for reputation compared to low self-monitor doves. These results, coupled with causal mediation tests and text analysis of open-ended responses, reveal that high self-monitors’ behavior in a crisis is driven specifically by concerns about reputation for resolve, rather than any other status or material concerns (56). At the same time, Whitlark’s question is important because the theory allows for the possibility that other status concerns might inform leaders’ behavior in other issue areas. Indeed, leaders might seek to cultivate their reputation for protecting the environment in their behavior at climate change negotiations, or try to cultivate their reputation as a guarantor of human rights in humanitarian intervention decisions. This is not at odds with the theory, which explicitly posits that the status concerns could mean different things in other domains. Finally, as a broader point, despite the substantial work scholars have done on this topic, we still do not have good ways of assessing whether resolve, or reputation for resolve, has unique qualities that do not translate to other types of traits or reputations. I invite future scholarship on this question because I believe it will substantially help us understand the dynamics of different reputations in international relations (IR) more generally.
Another set of questions pertains to the relationship between the President and the interests of the state. For instance, Carla Martínez Machain asks whether there are circumstances under which individual leaders are able to separate their own reputation from that of the state. This is a good empirical question. While the cases in the book do present presidents who often hitch their reputations to that of the state, analyzing the variation in this linkage is an avenue for future research. On pages 29-30 of the book, I note several theoretical conditions under which presidents might be more likely to separate their own reputations from that of the state. Again, however, we need to subject these to careful empirical analysis. To the extent that there is a difference between the two, the theory of the book might be less useful to explaining the behavior of states. Moreover, it should be noted that since reputation is really in the eye of the beholder, whether or not the leader separates her reputation from that of the state could be independent of what observers think which can lead to interesting misperceptions.
Relatedly, Whitlark and Michael Poznansky both raise questions about the relationship between leaders’ reputation concerns and variation in U.S. interests in a crisis. This deserves further discussion, but addressing this would require a substantial (and quite problematic) data collection effort and a somewhat different research design, which is why it currently lies outside the scope of the book.
Whitlark suggests consideration of cases where U.S. interests were not at stake, and Poznansky raises similar methodological concerns about distinguishing between high and low-stakes crises. Again, the issue of disentangling U.S. interests from other concerns is a tricky one; in the quantitative analysis, the Military Interstate Disputes (MIDs) dataset used does not indicate whether each crisis is highly relevant to U.S. interests or not. One can imagine that contemporary commentators might also differ on their assessments and opinions about U.S. interests in each crisis. To address this issue in the case studies, I tried to place each crisis in context in order to assess perceptions at the time of whether it was a high stakes or low stakes crisis. In the large-N analysis, I assume that high self-monitors should have an overall greater level of motivation for any given level of interest. High self-monitors are always more prone to engage in status-seeking behavior regardless of the interests involved, and I discuss how this plays out when controlling for hawkishness in the book (69). We can assume a similar difference in behaviors will still play out across all levels of U.S. interest and across high and low stakes crises. I further control for the capabilities of the adversary and the geographical region of the crisis as proxies for stakes (84). Nevertheless, a more fine-grained analysis to show that we see this difference between high and low self-monitors even in crises where U.S. interests are low would be a welcome addition.
Next, Poznansky asks about the implications of covert action for the theory. He wonders what high self-monitor leaders’ use of covert action says about their concerns about reputation for resolve, and whether resorting to covert action sends a more “watered down” signal of resolve. It is important to consider these demonstrations of resolve not only in the context of international pressures from adversaries, but in the context of domestic constraints as well. Even high self-monitor Presidents are constrained by domestic considerations––they cannot just maximize their reputation for resolve at the expense of other factors, such as domestic public opinion or the preferences of their own party. Given situations of strong domestic political opposition, high self-monitor leaders choose covert action as the best means to signal their resolve to the adversary while keeping the domestic opposition in the dark, or what Erving Goffman calls “audience segregation” (31). While the variation in clarity of message across covert or overt actions remains a question for future research, covert messages can still signal resolve by allowing leaders to engage in riskier activity and introduce escalation risks should the adversary choose to retaliate.[34]
Finally, several reviewers suggest promising directions for future research. Poznansky correctly points out that future work on individual leaders should study the dynamics between individuals in small group settings. He asks about the role of advisors in this theory, probing why some advisors were selected for study and the conditions under which advisors matter. In my research design, I selected the two advisors that played the most dominant role in the decision-making process across the crises selected. But it is true that in some crises there are additional advisors who played an active role in the decision-making. I do not analyze their self-monitoring inclinations because there was not enough available material about their personalities that allowed for such an in-depth analysis, and also because doing so would have resulted in a considerably denser and longer manuscript. Future work should try to more explicitly theorize about the interaction between the personalities of the president and those of his main foreign policy advisors.
I also explicitly leave out a discussion of why presidents select certain advisors, and in particular, whether hawkishness or self-monitoring of advisors shape their selection in a systematic way. Addressing this question requires a different kind of a research design. Based on the work here, however, we can appreciate the complexity of this task. For example, President Ronald Reagan and Secretary of State, George Schultz are both high self-monitors and both shared similar views of the Soviet Union. Both cared deeply about reputation for resolve in Lebanon, although Reagan felt that he was forced to withdraw due to domestic political pressure. Carter and Brzezinski differed in their self-monitoring and often disagreed on the Soviet Union. Although Carter’s behavior may have shifted toward Brzezinski’s policy preferences over time, he did not accept Brzezinski’s rationale for caring about reputation for resolve as the reason for a more hawkish turn against the Soviet Union after their invasion into Afghanistan. Both presidents, then, had advisors who varied across both their self-monitoring and beliefs, and this is probably true of most Presidents. Moreover, attributing the President’s behavior to the influence of his advisor(s) requires evidence that we rarely have even in high-quality primary documents. Nevertheless, formulating hypotheses about how leaders select advisors and testing them, especially when taking into account the role of personalities and not just beliefs, is a fruitful area of future research. The literature on individual leaders has made substantial progress in showing the independent causal effects of leaders’ personalities and beliefs, which should allow the next generation of researchers in this area to take these insights up one level to explore group dynamics.
Indeed, this book adds to a growing literature on leaders in IR. At the same time, as more people work in this space, it is important that we seek to also be clear about whether our results can be explained by other leader-specific factors. My goal in this book was to demonstrate how and why self-monitoring, a particular personality trait from the psychology literature, could and should be incorporated into the international relations lexicon. In this spirit, Martínez Machain and Kenneth Schultz both ask questions about the external validity of theories focused on one personality trait, and highlight the need for scholars in this research program to synthesize the various findings about the role of individual leaders. Schultz takes stock of the program and asks how scholars should think about accumulating the findings and generalizing them, if individuals differ from each other in countless ways. This is certainly a pressing question, and I can only begin to answer it here. By showing this personality trait’s relevance in discussions of leader reputation for resolve, I aimed to lay out a template for future scholars of leader personalities to show the causal effects of particular traits and beliefs. While these types of studies of particular personality traits raise concerns about external validity, this particular trait seems especially salient for this question of reputation. While we know from existing literature that self-monitoring is a distinct factor, as it is uncorrelated with these other personality traits (76), I also probe whether the results can be replicated by substituting self-monitoring with another personality trait. Indeed, in the large-N analysis, I tested for the significance of the other “Big Five” personality traits, and none of them generated the same effects as self-monitoring. In the case studies, I am also in conversation with the work of Allen Dafoe and David Caughey, and probe whether Southern culture of the presidents can explain the pattern that I see (which it cannot) (114).[35]
While this book takes explicit steps to take into account the role of alternative leader-specific explanations, future scholars should focus on this question of accumulation––how the different personality traits, beliefs, experiences, and other factors identified at the individual level interact within leaders as well as across leaders’ inner circles. This book and the other works that Schultz cites have shown that certain individual factors have independent causal effects on leaders’ behavior,[36] so the next step is to think about how the interaction of these traits influences behavior of both individuals and small groups. It could be the case that some traits in combination provide positive feedback and enhance certain patterns of behavior, tending Presidents to the extremes. It could also be the case that other combinations of traits might push behavior in opposite directions, canceling each other out or producing indeterminate outcomes.
Overall, I was delighted to see that despite the unique challenges involved in studying the causal role of personality traits on the behavior of leaders in crises, the reviewers found the book convincing and methodologically rigorous in making a valuable contribution to the literature showing how individual psychological characteristics matter. Beyond the particular argument I make about the psychological sources of concern about reputation, I wrote this book with an explicit aim to develop a template of how to study leaders systematically, and the benefits different methods could offer. I am especially excited about all the different ways the next generation of scholars can build on this book, and push the research on individual leaders in new and promising directions.
Notes
[1] Thomas C. Schelling, Arms and Influence (New Haven: Yale University Press, 1966): 124.
[2] Her first book was: Keren Yarhi-Milo, Knowing the Adversary: Leaders, Intelligence, and Assessment of Intentions in International Relations (Princeton: Princeton University Press, 2014).
[3] On leader beliefs see Elizabeth N. Saunders, Leaders at War (Ithaca: Cornell University Press, 2011), and Rachel Elizabeth Whitlark, “Nuclear Beliefs: A Leader-Focused Theory of Counter-Proliferation” Security Studies 26:4 (2017): 545-574. On their prior experience see Valerie Bunce, Do New Leaders Make a Difference? Executive Succession and Public Policy Under Capitalism and Socialism (Princeton: Princeton University Press, 1981), and Michael C. Horowitz, Allan C. Stam, and Cali M. Ellis, Why Leaders Fight (Cambridge: Cambridge University Press, 2015). On leader values see Brian C. Rathbun, Joshua D. Kertzer, Jason Reifler, Paul Goren, and Thomas J. Scotto, “Taking Foreign Policy Personally: Personal Values and Foreign Policy Attitudes,” International Studies Quarterly 60:1 (March 2016): 124-137. On risk tolerance see Jeff D. Colgan, “Domestic Revolutionary Leaders and International Conflict,” World Politics 65:4 (2013): 656-690; and Matthew Fuhrmann and Michael C. Horowitz, “When Leaders Matter: Rebel Experience and Nuclear Proliferation,” The Journal of Politics 77:1 (2015): 72-87.
[4] Rose McDermott, Political Psychology in International Relations (Ann Arbor: University of Michigan Press, 2004).
[5] Robert Jervis, “Hypotheses on Misperception,” World Politics 20:3 (1968): 454-479.
[6] Rose McDermott, “The Biological Bases for Aggressiveness and Nonaggressiveness in Presidents,” Foreign Policy Analysis 10:1 (2014): 313-327; and Maryann E. Gallagher and Susan H. Allen, “Presidential Personality: Not Just a Nuisance,” Foreign Policy Analysis 10:1 (2014): 1-21.
[7] Jessica L. Weeks, “Autocratic Audience Costs: Regime Type and Signaling Resolve, International Organization 62:1 (2008): 35-64; and Roseanne W. McManus, “The Impact of Context on the Ability of Leaders to Signal Resolve,” International Interactions 43:3 (2017): 453-447.
[8] Ronald Oaxaca, “Male-Female Wage Differentials in Urban Labor Markets,” International Economic Review 14:3 (1973): 693-709 and Alan S. Blinder, “Wage Discrimination: Reduced Form and Structural Estimates,” Journal of Human Resources, 8:4 (1973): 436-455.
[9] Daina Chiba, Carla Martínez Machain, and William Reed, “Major Powers and Militarized Conflict,” Journal of Conflict Resolution 58:6 (2014), 976-1002.
[10] Valerie M. Hudson and Patricia Leidl, The Hillary Doctrine: Sex and American Foreign Policy (New York: Columbia University Press, 2015).
[11] Oliver Kaplan, “The Art of Rhetorical Traps in Civilian Self-Protection,” Journal of Peacebuilding & Development 12:3 (2017): 111-116.
[12] Ben Rhodes, “Inside the White House during the Syria ‘Red Line’ Crisis,” The Atlantic, 3 June 2018.
[13] Donald J. Trump (@realDonaldTrump), Twitter post, 8 April 2018, 6:12 AM, https://twitter.com/realDonaldTrump/status/982969547283161090
[14] Michael C. Horowitz, Allan C. Stam, and Cali M. Ellis, Why Leaders Fight (New York: Cambridge University Press, 2015); Elizabeth N. Saunders, Leaders at War: How Presidents Shape Military Interventions (Ithaca: Cornell University Press, 2011).
[15] For a summary of the former, see Daryl G. Press, Calculating Credibility: How Leaders Assess Military Threats (Ithaca: Cornell University Press, 2005); Alex Weisiger and Keren Yarhi-Milo, “Revisiting Reputation: How Past Actions Matter in International Politics,” International Organization 69:2 (2015), 473-495. For a recent article on the latter, see Todd S. Sechser, “Reputations, Resolve, and Coercive Bargaining,” Journal of Conflict Resolution 62:2 (2016): 318-345.
[16] Elizabeth N. Saunders, “No Substitute for Experience: Presidents, Advisers, and Information in Group Decision Making,” International Organization 71:1 (2017): 219-247.
[17] On the various types of signals covert action can send, see Austin Carson, Secret Wars: Covert Conflict in International Politics (Princeton, NJ: Princeton University Press, 2018); Austin Carson and Keren Yarhi-Milo, “Covert Communication: The Intelligibility and Credibility of Signaling in Secret,” Security Studies 26:1 (2017): 124–56.
[18] Elizabeth N Saunders, Leaders at War: How Presidents Shape Military Interventions (Ithaca: Cornell University Press, 2011); Michael Horowitz, Allan C. Stam, and Cali M. Ellis, Why Leaders Fight (New York: Cambridge University Press, 2015); for a recent review, see Michael C. Horowitz and Matthew Fuhrmann, “Studying Leaders and Military Conflict: Conceptual Framework and Research Agenda,” Journal of Conflict Resolution 62:10 (2018): 2072-2086.
[19] See, for example, Daryl Grayson Press, Calculating Credibility: How Leaders Assess Military Threats (Ithaca: Cornell Univ. Press, 2005); Joe Clare and Vesna Danilovic, “Reputation for Resolve, Interests, and Conflict,” Conflict Management and Peace Science 29:1 (2012): 3-27; Alex Weisiger and Keren Yarhi-Milo, “Revisiting Reputation: How Past Actions Matter in International Politics,” International Organization 69:2 (2015): 473-495.
[20] See, for example, Cathy Xuanxuan Wu and Scott Wolford, “Leaders, States, and Reputations,” Journal of Conflict Resolution 62:10 (2018): 2087-2117; Jonathan Renshon, Allan Dafoe, and Paul Huth, “Leader Influence and Reputation Formation in World Politics,” American Journal of Political Science 62:2 (2018): 325-339.
[21] See, e.g., Thomas C. Schelling, Arms and Influence (Westport: Greenwood Press, 1976), chap. 2; James E. Alt, Randall L. Calvert, and Brian D. Humes, “Reputation and Hegemonic Stability: A Game-Theoretic Analysis,” The American Political Science Review 82:2 (1988): 445; Krista E Wiegand, “Militarized Territorial Disputes: States’ Attempts to Transfer Reputation for Resolve,” Journal of Peace Research 48:1 (2011): 101-113.
[22] This category includes Yarhi-Milo’s first book; Keren Yarhi-Milo, Knowing the Adversary: Leaders, Intelligence, and Assessment of Intentions in International Relations (Princeton: Princeton University Press, 2014).
[23] Saunders, Leaders at War; Horowitz, Stam, and Ellis, Why Leaders Fight; Allan Dafoe and Devin Caughey, “Honor and War,” World Politics 68:2 (2016): 341-381.
[24] See, e.g., Brian C. Rathbun et al., “Taking Foreign Policy Personally: Personal Values and Foreign Policy Attitudes,” International Studies Quarterly 60:1 (2016): 124-137; Horowitz, Stam, and Ellis, Why Leaders Fight.
[25] For a discussion of the case selection strategy, see pages 116-117.
[26] One canonical work on this topic is James D. Fearon, “Domestic Political Audiences and the Escalation of International Disputes,” The American Political Science Review 88:3 (September 1994): 577-592.
[27] Alexander L. George, “The ‘Operational Code’: A Neglected Approach to the Study of Political Leaders and Decision-Making,” International Studies Quarterly 13:2 (1969): 190-222; Robert Jervis, Perception and Misperception in International Politics (Princeton, NJ: Princeton University Press, 1976); Margaret G. Hermann, “Explaining Foreign Policy Behavior Using the Personal Characteristics of Political Leaders,” International Studies Quarterly 24:1 (1980): 7-46; Daniel L. Byman and Kenneth M. Pollack, “Let Us Now Praise Great Men: Bringing the Statesman Back In,” International Security 25:4 (2001): 107-146; Hein E. Goemans, Kristian Skrede Gleditsch, and Giacomo Chiozza, “Introducing Archigos: A Dataset of Political Leaders,” Journal of Peace Research 46:2 (2009): 269-283; Giacomo Chiozza and H. E. Goemans, Leaders and International Conflict (Cambridge Cambridge University Press, 2011); Stephen Peter Rosen, War and Human Nature (Princeton, NJ: Princeton University Press, 2009); Sarah E. Croco, “The Decider’s Dilemma: Leader Culpability, War Outcomes, and Domestic Punishment,” American Political Science Review 105:3 (2011): 457-477; and Jeff D. Colgan, “Domestic Revolutionary Leaders and International Conflict,” World Politics 65:4 (2013): 656-690.
[28] Elizabeth N. Saunders, Leaders at War: How Presidents Shape Military Interventions (Ithaca: Cornell University Press, 2011); Brendan Rittenhouse Green, “Two Concepts of Liberty: U.S. Cold War Grand Strategies and the Liberal Tradition,” International Security 37:2 (2012): 9-43; and Rachel Elizabeth Whitlark, “Nuclear Beliefs: A Leader-Focused Theory of Counter-Proliferation,” Security Studies 26:4 (2017): 545-574.
[29] See for example, Michael C. Horowitz and Allan C. Stam, “How Prior Military Experience Influences the Future Militarized Behavior of Leaders,” International Organization 68:3 (June 2014): 527-559; and Matthew Fuhrmann and Michael C. Horowitz, “When Leaders Matter: Rebel Experience and Nuclear Proliferation,” The Journal of Politics 77:1 (2015): 72-87.
[30] Jervis, Perception and Misperception in International Politics.
[31] Simon Tisdall, “Trump believes the North Korea summit is all about him. But Kim has a plan, too,” The Guardian, 9 June 2018, https://www.theguardian.com/us-news/2018/jun/10/north-korea-summit-trump-believes-all-about-him-kim-has-plan-too.
[32] Daniel Kahneman and Amos Tversky, “Prospect Theory: An Analysis of Decision under Risk,” Econometrica 47:2 (March 1979): 263-291.
[33] Elizabeth N. Saunders, “No Substitute for Experience: Presidents, Advisers, and Information in Group Decision Making.” International Organization 71:S1 (2017): S219-S247.
[34] On this, see also Austin Carson and Keren Yarhi-Milo, “Covert Communication: The Intelligibility and Credibility of Secret Signaling,” Security Studies 26:1 (2016): 124-156.
[35] Allen Dafoe and David Caughey, “Honor and War: Southern US Presidents and the Effects of Concern for Reputation,” World Politics 68:2 (2016): 341-381.
[36] Elizabeth N Saunders, Leaders at War: How Presidents Shape Military Interventions (Ithaca: Cornell University Press, 2011); Michael Horowitz, Allan C. Stam, and Cali M. Ellis, Why Leaders Fight (New York: Cambridge University Press, 2015); and Dafoe and Caughey, “Honor and War.”