University of California at Los Angeles
November 18, 2013
Why do countries ever go to war with each other? Why can’t rival powers just work out an arrangement that would be better for both of them than an armed conflict? It is commonly assumed nowadays that if states were able to understand how far their adversaries were prepared to go to achieve their aims, bargains could be struck relatively easily and wars could be avoided. The problem, the argument runs, is that governments have “incentives to misrepresent” how tough they are in order to improve their bargaining positions, and for this reason even rational states can easily misjudge how strongly their rivals feel about a particular issue. Whether war can be avoided might therefore depend on how well states are able to deal with that problem—that is, on how well they are able to make their real preferences clear to their opponents.1
And a number of theorists have emphasized one particular way this can be done. The “audience costs” mechanism, they argue, allows states to make their real preferences known. If a government would pay a political price with its “audience” at home if it backed down in a crisis, its rival would be more inclined to take what it says seriously than if it could bluff with impunity. By taking a tough line in public, the leadership would in effect be “tying its hands”: the prospect of having to pay a price at home would tend to keep it from backing down, and the adversary, knowing this, would see that tough public statements could not be dismissed as “cheap talk.” It is often taken for granted, moreover, that this audience costs effect is stronger in democracies than in other kinds of regimes, and that this gives democratic states a real advantage in international bargaining.
The audience costs argument is of particular interest because the phenomenon it focuses on is something of a double-edged sword. On the one hand, by enabling the adversary to see more clearly how far a state is willing to go—that is, by enabling it to distinguish between serious threats and empty posturing—the audience costs mechanism can play a real role in the process through which a bargain is reached and war is avoided. On the other hand, states might try to exploit the audience costs mechanism by making public threats and locking themselves into hard-line policies in the hope this would lead their adversaries to back down; the upshot might be a war that would not have occurred if they had not used that tactic.2 Thus the audience costs theory might help us understand why crises get resolved and wars are averted, but it also might help us see why crises escalate and wars break out. This, however, creates a certain problem of indeterminacy: the ability of the theory to generate unambiguous predictions is more limited than one might think. One could argue, for example, that democracies are better able than authoritarian regimes to avoid war because it is easier for them to make their threats credible. But one could also argue that they are more likely to go to war, because their ability to use the audience costs mechanism to gain a bargaining advantage gives them an incentive to dig in their heels, making it harder for them to make the concessions needed to settle a crisis.
How then should one go about seeing if there is any substance to the theory? A number of scholars have tried to get at the issue by using statistical methods, but since audience costs are not directly measurable, this has not been easy to do. (They, in fact, would not be observed at all if, as the theory suggests, a government’s ability to tie its hands in this way would lead the adversary to back down.3) One therefore had to find something that could serve as a proxy, and the most common assumption here was that audience costs are a function of how democratic a state is.4 That assumption, however, is somewhat problematic and has by no means been universally accepted.5 But even if it were valid, a statistical analysis based on the assumption that a country’s democracy score can serve as a proxy for audience costs might not tell us much about the issue scholars are really concerned with. Such an analysis might suggest that democratic political structures are important, but it cannot show they are important “because of their ability to generate domestic political audience costs,” and not for some other reason.6
Given these problems, a number of scholars have concluded that there is really only one way to proceed: one has to study particular cases.7 But how exactly are those cases to be selected? They obviously cannot be chosen in a purely arbitrary or random way. There has to be a compelling reason for thinking a particular case will tell you something important about the general issue. And for nearly forty years the standard assumption has been that two sorts of cases are of particular interest in this context: “easy” or “most-likely” cases, where one would expect the theory to apply if it had any validity to it at all, and “tough” or “least-likely” cases, where one would be surprised if the sort of thing the theory emphasized turned out to be important. If the theory is not supported even by the “easy cases,” then you would have to wonder whether it really helps you understand anything; but if a theory passed a tough test, then that would be important evidence of its power.8 That kind of guidance, however, can only take you so far. It does not tell you much about how, in practice, you would go about selecting the particular cases to be studied.
There are a number of ways to do this, but in this article I want to talk about one that has not been widely used, but which can be quite effective in certain contexts. This particular method is based on work historians have already done. One can scour the historical literature and look for historical interpretations that have a certain resonance in terms of the theory you are interested in (and the audience costs theory is of course what we are interested in here). One can ask historians whether they can think of any examples of the sort of mechanism one has in mind; one can even post a query on one of their email discussion groups, H-Diplo for example. Political scientists may argue about theoretical issues among themselves, but historians, by and large, have no dog in those fights—as a rule they are scarcely aware of them—and it can generally be taken for granted that in developing their interpretations they have no interest in supporting any particular political science theory. So if, for example, they come up independently with arguments about specific historical cases that have a certain audience costs flavor, those cases should be of particular interest to people looking for some way to test the audience costs theory. If one is trying to see how much the mechanism emphasized by that theory actually counts for in the real world, these are cases one would especially want to examine.
So suppose you identify an interpretation of this sort. What do you do next? Your goal is to assess a particular historical claim, but doing that is not quite as easy as it might seem. Relatively narrow questions are generally bound up with much broader issues, so to assess a specific claim, you often need to go into those broader issues of historical interpretation in some depth. And to get to the bottom of those interpretive issues, you cannot simply rely on what particular historians say; the simple fact that historians often disagree among themselves on those issues means that no particular interpretation can be accepted on faith. But few political scientists have the time needed to go into the sources on their own and work out an interpretation that makes sense to them. In such circumstances, the best way to proceed is to analyze the historiographical debate on the issue—that is, to assess historical arguments in terms both of their internal logic and of the adequacy of the evidence put forth to support them. And in analyzing those historical debates, you can often reach some fairly solid conclusions not just about how the episodes in question are to be interpreted, but also about how specific claims linked to those basic interpretations—claims that have a particular importance in the context of some theory—are to be assessed.
All this is very general and my main goal here is to show how this method works in practice by looking at how it can be applied to the case of the audience costs theory. One key claim associated with that theory is the idea that political leaders are to some extent locked into a particular policy by the tough public statements they make. Can one think of any historical case where this was so—or, more precisely, can one think of any historical case where a scholar claimed this was so?
There are not many such cases one can point to, but there is one very important argument of this sort—an argument that supports a more general interpretation of U.S. policy on Vietnam during the early 1960s. A number of scholars have claimed that President John F. Kennedy was locked into a policy of doing whatever was necessary to prevent South Vietnam from falling to the Communists because of the tough public statements to that effect which he had made. If true, this would be strong evidence supporting one major part of the audience costs argument—or at least strong evidence showing that the sort of mechanism audience costs theorists have in mind can play an important role in the real world. Other scholars, however, have denied that one can infer from the fact that strong statements were made that Kennedy was determined to “pay any price” to prevent the loss of South Vietnam. Some of them go so far as to argue that the president had actually decided to withdraw from Vietnam by 1965 even if that would result in a Communist takeover of that country. If true, that would mean that Kennedy did not believe that tough public statements limited his freedom of action to anything like the extent that the audience costs theory would lead you to think—and no one was in a better position to judge how much freedom of action he had than the president himself. That in turn would suggest that one key element in the audience costs theory—the idea that by issuing strong public statements, a political leader is burning his bridges—is weaker than many people believe.
What this means is that one can get at the issue not directly, by just looking at the historical evidence and treating historical works as repositories of facts, but rather indirectly, by analyzing historical debates. One can look at those historical arguments about Kennedy and Vietnam—arguments in which an audience costs-related claim often plays a key role—and try to see how well they stand up in the light of the evidence, and especially the evidence the scholars who make those arguments put forward to support their claims. What does this analysis show, and what in particular does it tell us about the audience costs theory?
Locked into a Policy?
It used to be taken for granted that American policy during much of the Cold War period, and in particular U.S. policy on southeast Asia in the early 1960s, was to be understood in mainly ideological terms. The “containment of communism,” as two distinguished analysts wrote in a well-known work on America’s Vietnam policy, was the “core consensual goal of postwar foreign policy”; the key decision, from which all else essentially followed, was that South Vietnam could not be allowed to fall into Communist hands.9 And for years it was commonly argued that this basic point applied in particular to the Kennedy period (1961-63). Only after the war had turned sour in the late 1960s, the argument ran, were the basic assumptions underpinning America’s postwar policy seriously challenged; in the early 1960s those assumptions were still strong enough to essentially determine policy.
In the last couple of decades, however, many historians have come to take a rather different view. Most scholars today no longer view Kennedy as a simple Cold Warrior. People used to quote the famous passage from Kennedy’s inaugural address about how America would “pay any price, bear any burden, meet any hardship, support any friend, oppose any foe to assure the survival and the success of liberty” as though this told us something fundamental about the sort of policy he was to pursue. But the prevailing view among historians nowadays is that this was not the real Kennedy at all—that the real Kennedy was much cooler, less ideological, and more power politically oriented than that sort of rhetoric might suggest. Even Kennedy’s Vietnam policy is often interpreted in those terms; one writer, John Newman, went so far as to argue twenty years ago that Kennedy was determined to withdraw from that country “come what may” after the 1964 election.10
Newman’s thesis received a good deal of attention when his book JFK and Vietnam came out in 1992, in large part because Oliver Stone used some of Newman’s arguments at that time in his well-known film JFK; Newman had, in fact, worked with Stone on the movie.11 Indeed, as Stone himself noted, the movie suggested “that it was Vietnam that led to the assassination of John Kennedy”—that people within the government conspired to murder him because he was determined to change the course of American policy and actually withdraw from that country.12That theme struck a chord with certain sections of the left, but some prominent left-wing intellectuals strongly objected to the picture Stone had painted. Noam Chomsky in particular was so disturbed by the fact that so many people accepted the view of Kennedy as a “shining knight promising peace, foiled only by assassins bent on stopping this lone hero who would have unilaterally withdrawn from Vietnam had he lived” that he wrote a book that sought to refute the idea that Kennedy wanted to withdraw from Vietnam unconditionally.13 In that book, Rethinking Camelot, Chomsky maintained that for Kennedy there could “be no withdrawal without victory.”14 The country could not be allowed to fall to the Communists; the war simply had to be won; America had to “stay the course.”15
And one of the key arguments Chomsky made to support that interpretation had a certain audience costs flavor. He quoted a number of the public statements Kennedy made, especially in 1963, insisting that America was not going to withdraw from South Vietnam and allow the country to be taken over by her enemies. “If the United States were to falter,” the president said at one point, “the whole world, in my opinion, would begin to move toward the Communist bloc.”16 This sort of “inflammatory rhetoric,” Chomsky writes, “could only serve to undermine withdrawal.” A president who really wanted to pull out would never have used that kind of language. The argument that some people make that Kennedy intended to withdraw after the 1964 elections—that it would be easier to deal with a right-wing backlash then, after he was reelected, than in his first term, when a withdrawal might compromise his ability to remain in office—was “hardly credible.” “Nothing would have been better calculated to fan right-wing hysteria,” he writes, “than inflammatory rhetoric about the cosmic issues at stake, public commitment to stay the course combined with withdrawal from that commitment as the client regime collapsed in 1964, election on the solemn promise to stand firm come what may, and then completion of the withdrawal and betrayal. That plan would have been sheer stupidity.”17
This book was taken quite seriously when it came out, even by writers who did not share Chomsky’s political views. Tom Wicker, who for many years had worked for the New York Times first as a political reporter and then as a columnist, reviewed the book for Diplomatic History, the main journal published by the Society for Historians of American Foreign Relations (SHAFR). So exhaustive was Chomsky’s study of the sources, Wicker wrote, “that his conclusion can hardly be disputed by a fair-minded reader: ‘President Kennedy was firmly committed to the policy of victory [in Vietnam] that he inherited and transmitted to his successor, and to the doctrinal framework that assigned enormous significance to that outcome; he had no plan or intention to withdraw without victory.’” That conclusion was based in large part on a study of what Kennedy actually said: his public remarks, in particular, expressed “nothing but determination to win.” Wicker was clearly convinced by the inference Chomsky drew from his survey of what Kennedy was saying in public. It could be argued, Wicker wrote, that “these public statements were a smokescreen to conceal Kennedy’s real plans; but if so, Chomsky asks, why would such a ‘political animal’ as JFK make a sudden withdrawal more difficult for the public to accept by repeated claims for the importance of victory?”18
The distinguished historian Robert Dallek also referred approvingly to the Chomsky book in his SHAFR presidential address in 1996, and he too emphasized what we would now call this point about audience costs. There was good reason to think, Dallek wrote, that had he lived, Kennedy would have escalated the war just as his successor Lyndon Johnson did. He thought that Chomsky’s book made a “convincing case” that Kennedy did not intend to withdraw from South Vietnam “without a greater test of the Communist drive for control.” And the use that book made of Kennedy’s public utterances was particularly worth noting. “Chomsky points out,” Dallek said, “that had Kennedy intended to withdraw, it is hard to understand why he so consistently spoke publicly about holding the line in Vietnam. JFK was too astute a politician to have created a public expectation that he intended to abandon after reelection in 1964.”19 And indeed many other scholars have made arguments of this sort over the years.20
What then is to be made of the Chomsky argument? If there was a significant gap between Kennedy’s real policy and the sort of rhetoric he used—that is, if he was much less committed to avoiding the loss of Vietnam than his public statements might lead one to think—that would shed some light on the issue we are concerned with here. It would suggest that he was not as locked in by his public pronouncements as the audience costs theory might lead one to suppose. So how convincing is the Chomsky argument? Is it really impossible for a “fair-minded reader” to dispute his conclusions?
The answers turn, of course, on what the evidence shows. Chomsky thinks the sources are absolutely unambiguous. “There is no hint in the record,” he says, that Kennedy intended to withdraw without victory.21 The president, one official wrote, was determined to win the war, and there was “not a phrase in the internal record to suggest that this judgment” by one of Kennedy’s closest advisers “should be qualified in any way.”22 Arguments to the contrary, in Chomsky’s view, are to be understood essentially as exercises in myth-making. After the Tet offensive in January 1968 many people turned against the war, and since Kennedy was the great hero of the liberal intellectuals, he now had to be portrayed as a kind of secret dove.23 But that interpretation, as he sees it, was baseless. Arguments to that effect, as laid out in the Newman book and in Arthur Schlesinger’s 1978 biography of Robert Kennedy—were “concocted without a shred of evidence.”24
So Chomsky certainly gave the impression that there is not much to back up the idea that Kennedy’s support for the policy of keeping South Vietnam out of Communist hands was anything less than whole-hearted. To be sure, he was aware of the fact that a few stories suggested that Kennedy, in the final analysis, would have accepted the “loss” of Vietnam. In a 1972 memoir, Kennedy’s aide Kenneth O’Donnell said the president had told Senator Mike Mansfield that he agreed with him on the need for a withdrawal, but for domestic political reasons could not pursue that policy until after the 1964 elections. O’Donnell also claimed that after Mansfield had left, the president said that after he was reelected, he intended to bite the bullet and withdraw from Vietnam, no matter how much of a right-wing backlash there was at home. But O’Donnell’s testimony, Chomsky feels—and most historians agree with him on this point—should not carry much weight.25 It is not just that accounts published by Kennedy’s acolytes after people had soured on the war are inherently suspect, or that Mansfield’s accounts of his meeting with Kennedy varied.26 Even if the story about Mansfield was essentially correct, it would not tell us much about what Kennedy really felt. As Chomsky points out, the president might have just been telling Mansfield what he thought the Senator wanted to hear.27
The problem is that whereas Chomsky leads the reader to think that the idea that Kennedy was not fully committed to winning the war in Vietnam is supported only by very questionable “post-Tet reconstructions” of this sort, the evidence is much more abundant than he suggests.28 To be sure, some of it can be dismissed with the same sort of argument Chomsky used to dismiss the Mansfield story. Senator Wayne Morse, for example, from the start a leading dove on Vietnam, later said that Kennedy told him shortly before he was assassinated that he fully agreed with Morse’s criticism of the administration’s Vietnam policy.29 But this can certainly be written off with the argument that Kennedy was just telling Morse what he wanted to hear. And a number of accounts that took the line that the president intended to withdraw without victory could easily be written off as post-Tet myth-making, if they were all we had. The accounts by Robert McNamara, Kennedy’s Secretary of Defense, by Theodore Sorensen, one of Kennedy’s closest advisers, and by the historian Arthur Schlesinger, who worked in the White House during that period, can certainly be put in that category.30 But to dismiss the stories to that effect told by a whole series of other former officials would be more of a stretch. The later accounts given by McNamara’s deputy Roswell Gilpatric, by Roger Hilsman, a State Department official who played a key role in making policy in this area at the time, and by Michael Forrestal, who had been responsible for Vietnam policy at the Kennedy White House, fall into this category.31 Hilsman’s claim, in particular, was quite categorical: “What Kennedy told Hilsman in private was that he did not expect victory, and that he intended to withdraw anyway.”32 As for Forrestal, he told an interviewer in 1971 about a meeting he had with the president the day before he was shot. The two men should meet again, Kennedy said, when Forrestal got back from a trip to Indochina he was about to make:
because we have to start to plan for what we are going to do now in South Vietnam. He said, “I want to start a complete and very profound review of how we got into this country, what we thought we were doing and what we now think we can do.” He said, “I even want to think about whether or not we should be there.” He said, because this was of course in the context of an election campaign, that he didn’t think we could consider drastic changes of policy quickly. But that what he wanted to consider when I returned and when people were ready to think about this more clearly was how could some kind of a gradual shift in our presence in South Vietnam occur.”33