Extended deterrence in the Taiwan Strait: learning from rationalist explanations in international relations
Extended deterrence characterizes situations where one state (for convenience, the defender) seeks to prevent another state (the challenger) from attacking a third state (the protege). The defender does so by threatening to take actions, including, but not limited to, military response, that will supposedly deny the fruits of conquest to the challenger or punish it to such an extent that its prospective costs for committing this unwanted act exceed whatever gains it may hope to acquire. Compared to direct deterrence, which aims at preventing an attack on the defender’s home territory, the threat of extended deterrence is inherently more doubtful. As Schelling (1966, 36) emphasized, “the difference between the national homeland and everything ‘abroad’ is the difference between threats that are inherently credible, even if unspoken, and threats that have to be made credible.” The challenger is less sure of whether the defender will actually carry out its threat–that is, whether the defender is just bluffing. Whereas it is easy to imagine a state’s resistance and even retaliation after it has suffered a direct attack, it is far less certain that a defender will come to the aid of an ally under attack. To paraphrase Charles de Gaulle on France’s rationale for pursuing its own force de frappe, would Washington retaliate against Moscow in the wake of a Soviet nuclear strike against Paris, knowing that this action would in turn almost certainly entail the destruction of New York in a Soviet counter-strike? Expressing similar doubts, Adolf Hitler questioned the United Kingdom’s commitment to Poland, remarking, “Why should Britain fight? You don’t let yourself get killed for any ally” (quoted in Huth 1988a, 5). Such skepticism is not unwarranted, because only 27 percent of treaty commitments were honored by states in the defense of their allies in past wars (Sabrosky 1980).
Although recent U.S. policy has alternated between deliberate strategic ambiguity and assertive declaratory support for Taiwan, there is little doubt that it wishes to forestall a Chinese invasion of this island and, as such, represents a case of extended deterrence. Despite the obvious importance of this triangular relationship to regional peace and stability, international relations scholars have not applied their considerable repertoire of theories and methods to its analysis. Area specialists, who have conversely been generally detached from mainstream international relations research, have dominated the research on this case. I try to bridge this evident gap in this article, focusing particularly on the possible contributions of rationalist interpretations in international relations scholarship in advancing our understanding of this case. Truth in advertising, however, compels me to acknowledge at the outset that this article is not about describing or explaining the current state of triangular relationship among the parties, nor is it concerned with predicting or prescribing this relationship’s future evolution. I seek instead to present some ideas, arguments, and propositions that have heretofore not been recognized or addressed systematically in discussions on extended deterrence as applied to Taiwan.
Rationalist interpretations do not imply that people are omnipotent in their ability to procure and process information. We know all too well that people are subject to a variety of cognitive and perceptual errors (for example, Jervis 1976; Levy 1997; Kahneman and Tversky 2000; Tversky and Kahneman 1977). This recognition of limits to rationality, however, hardly warrants general attributions of naivete, even stupidity, to government leaders. On the contrary, it seems sensible to start from the premise that officials know their counterparts far better than scholars may wish to acknowledge. Washington, Beijing, and Taipei, for instance, invest enormous time, effort, and resources in trying to gain an accurate understanding of each other. Academics have a hard time claiming any special insight or unique source of wisdom, whether it is based on mastery of the other side’s language, intimate familiarity with its culture, or access to timely and sensitive information with restricted distribution. If anything, they are usually at a considerable disadvantage on these scores when compared to diplomats, intelligence analysts, and even journalists and business people. Indeed, academics in fields such as history and political science typically operate in the realm of common knowledge, outdated information, and mundane data. This confession in turn implies that at least for some of us, our individual and collective forte lies with the analysis of persistent empirical patterns and the formulation of general models of foreign policy conduct.
Rationalist interpretations belong to such endeavors. They assume that people engage in strategic behavior. They attribute to Americans, Chinese, and Nigerians alike the capacity to learn from history and to adapt subsequent behavior accordingly. They also expect people to anticipate the actions of their counterparts and to make appropriate ex ante adjustments based on this anticipation. They finally treat people as purposeful agents; that is, at least when it comes to issues of major importance to them, people try to be deliberate even if not all of their choices eventually turn out to fulfill their hopes and expectations. As generic attributions that are supposed to apply to all people, rationalist interpretations suggest general tendencies about people’s behavior in particular classes of situations. They thus offer baseline expectations about this behavior. Case-specific and context-specific knowledge can then presumably be brought to bear in further refining these expectations. Accordingly, far from implying any incompatibility between generalist and specialist knowledge, there should be synergism in the work done by members of both communities. China specialists should therefore be able to show where and how generic rationalist attributions can be revised and improved in the particular case of extended deterrence in the Taiwan Strait. They should be able to contribute an important “value-added” component to general, baseline expectations. At the same time, however, and as I hope to show, rationalist interpretations can present issues and concerns that may have heretofore been inadequately addressed by country specialists.
EXPLAINING WAR OCCURRENCE
Wars are costly and risky. Waging war means the expenditure of blood, sweat, tears, and dollars. Leaders of the warring states endanger their own popularity, power, and even life and liberty (for example, Lyndon Johnson, Benito Mussolini, Slobodan Milosevic). So why would leaders want to bear these costs and run such risks? Why could they not reach a settlement without going to war, thereby sparing themselves all of the associated costs and risks of war? That wars are inefficient in the sense just described should not be difficult to grasp. That leaders have an incentive to settle without paying a heavy price for waging war should also be evident. So what prevents them from reaching a compromise without resorting to arms?
Mark Twain reportedly remarked that a difference of opinion makes a horse race. Wars evidently happen because the leaders of the opposing states have inconsistent ex ante expectations about the benefits and costs of war. Presumably, both sides must be optimistic about how they will fare in the coming conflict. Their respective subjective estimates of success sum to be more than unity (that is, the sum of the probability assigned by each belligerent to a successful war outcome is greater than one). To ask why wars happen is then tantamount to asking why the two sides are jointly optimistic or, in other words, why they have inconsistent ex ante expectations. As already implied, if the two sides have consistent expectations, they could have and would have negotiated a settlement on the basis of their anticipation of the eventual outcome of the military contest–and in doing so, they would have avoided the inefficiency of war.
One could point to misperception and emotion standing in the way of officials coming to their senses and to their pulling back from the precipice of confrontation. Rationalist arguments, however, suggest that one need not resort to such factors in explaining the belligerents’ inability to reconcile their ex ante expectations, even though such factors surely played a role in sustaining and even exacerbating mutual hostilities in some past crisis situations. Of course, misperception (underestimation of an enemy’s intention or overestimation of its capability) and emotion (fear, worry) can also discourage war. Among history’s more infamous cases was Neville Chamberlain’s appeasement policy. As will be explained, rationalist interpretations lead to the conclusion that even in the absence of misperception and emotion, there are good reasons to be weary of the chances that the opposing sides will be able to reach an agreement that averts war. This remark does not mean that wars are pervasive or inevitable; rather, the structure of the relationship between the two would-be opponents is such that it is difficult to arrive at an ex ante settlement even if officials are able to keep their cool during crises and make sensible, reasoned judgments.
Why? Part of the answer lies with deliberate misrepresentation. Anyone who has played the game of poker or has haggled over prices at a flea market or automobile dealership knows this ploy. In poker, one disguises one’s hand to make it appear either weaker (to keep the other players in the game to enrich the pot) or stronger (to pressure the other players to withdraw from the game). In bargaining over sale prices, one often pretends to be less interested in the merchandise or to be less willing or able to afford it than is actually the case. Likewise, government leaders are known to exaggerate their interest (or stake) in an issue and their motivation (resolve) and capability to protect this interest. In our ordinary lives and the world of foreign policy, bluffing is not an uncommon phenomenon. Indeed, one can go so far as to argue that as in the game of poker, the point of statecraft is to achieve “cheap” victories–that is, to get the other side to comply with one’s wishes without having to exert oneself and pay the necessary price. It involves, in other words, precisely efforts to beguile, intimidate, withhold relevant information, and, if you will, foster misperception (Gartzke 1999, 570-71). To outlaw bluffing or, more broadly, attempts to disguise, is tantamount to changing entirely the character of poker. As in this game, officials are not about to show their “hole card” by revealing their state’s true intentions or capabilities, or their “bottom line” demands (minimally acceptable deals).
As in poker and at flea markets, those who engage in statecraft expect others to have an incentive to undertake deliberate misrepresentation. Officials therefore are inclined to discount verbal statements (rhetoric) and even standard posturing behavior from the other side, dismissing them as not credible unless backed up by more tangible evidence. That it is taken for granted that all sides engage in deliberate misrepresentation helps to explain in part why it is difficult to reconcile the ex ante expectations of the would-be belligerents. The officials of the opposing sides know that sometimes states issue serious threats, but sometimes they also bluff. How can these officials tell the difference between a defender who is genuinely committed to the defense of its protege and one who makes empty threats? As demonstrated in research on economic sanctions, the target of such coercive statecraft has to decide whether the sanctions are meant to be a precursor to even more damaging deprivations to follow if it does not alter its policy, or whether the sanctions are intended as symbolic gestures to appease public opinion and are used in lieu of (that is, as substitutes for) more tangible and forceful actions (Schwebach 2000). An inability to penetrate the other side’s strategic behavior and an unwillingness to disclose one’s own true intent and capability contribute to the difficulty of reaching an ex ante settlement even when the parties agree that a conflict will be costly.
Deliberate misrepresentation is compounded by the officials’ felt need to address different audiences. On the eve of World War I, Chancellor Bethmann Hollweg sought to simultaneously convey Berlin’s firm support for Vienna’s cause and play down its belligerence to neutralize London’s support for Paris and St. Petersburg and to secure the support of the German Social Democrats for his policy. In the offshore-islands confrontations of the 1950s, President Dwight Eisenhower threatened U.S. intervention to deter a Chinese attack while trying to “leash” Chiang Kai-shek lest he tried to escalate the crises in an attempt to entangle the United States in an unwanted reversed invasion to retake the mainland. More generally, it should hardly be surprising if a liberal Democratic administration wants to appear “tough” on China in its dealings with a conservative Republican Congress while presenting a more conciliatory posture when actually talking with Chinese officials. Similarly, Beijing’s leaders may well misrepresent their policies depending on the institutional or political constituent in question, such as whether they are speaking to military cadres charged with national defense or coastal industrialists with a stake in maintaining export access to the U.S. market. Scholars who study cross-national patterns with a broad stroke are poorly trained and equipped to grasp such variations in the officials’ presentation of different faces to different audiences. Country specialists, however, should be much better positioned to take on this analytic challenge.
That leaders can present different faces to different audiences is common knowledge. This common knowledge in turn underscores the reason why it is difficult to reconcile the disputants’ expectations. Moreover, this knowledge means that leaders can sometimes try to act “tough” in foreign policy for domestic reasons. They may, for example, undertake economic sanctions against an adversary not because they really expect to extract concessions from this adversary, but rather with the view of making partisan political and economic gains at home (such as to defuse political criticisms and appease protectionist interests). Foreign leaders are not so dense as to overlook this motivation, which seems to drive such efforts to politicize trade and investment (Kaempfer and Lowenberg 2000, 1992, 1988). This being the case, it is not surprising that economic sanctions fail in most cases and that, even knowing this fact, leaders continue to employ them (Pape 1998a, 1998b, 1997).
But why could the parties not just be reasonable? Why could they not reveal the bases for their inconsistent expectations in the hope of working out a solution to avoid war? As Fearon (1995) has shown persuasively, inconsistent expectations can also be attributed to the possession of private information. Private information is by definition privileged data known only to one side. Such information may incline officials to think, and often reasonably so, that their troops are better trained and equipped, have a higher morale, and are provided with a more viable battle plan than they are given credit for by the opposing side. Such considerations led the Japanese leaders to assess their probability of success in the 1904 war to be roughly even odds, whereas the Russian leaders were supremely confident that military superiority lay with their side. Even though almost all historians agree that the Japanese would have preferred a negotiated settlement with the Russians instead of going to war, how could Tokyo have gotten St. Petersburg to revise its (unrealistic) ex ante expectations? Statements about the quality of Japanese armed forces would not be believed by the Russians. Why should the Russians believe such statements in view of the known tendency for officials to make deliberate misrepresentation? To disclose Japan’s military plan or to show more tangible evidence of its military disposition and preparations would in effect be to forfeit any strategic advantage, including the element of surprise. These are, after all, state secrets, and there are good reasons why states want to keep such information secret. Finally, to announce that Japan truly prefers a compromise settlement (in this case, a division of spheres of influence between Russia and Japan over Manchuria and Korea, respectively) to war is tantamount to undercutting Tokyo’s own bargaining position and, concomitantly, abetting St. Petersburg’s overconfidence and intransigence.
Hence, a reluctance to disclose private information impedes efforts by states to reconcile their ex ante expectations in the hope of striking a deal to avoid war. In the context of extended deterrence in the Taiwan Strait, Beijing is unlikely to volunteer any private information regarding, say, an effective way to provide air cover for its amphibious forces and naval convoy in a possible invasion, nor can Washington be expected to publicize, again hypothetically, its intention to actually refrain from direct military intervention in such an event. Any such disclosure would be self-defeating for obvious reasons. Parenthetically, academics–whether they be generalists or specialists–are unlikely to have access to officials’ private information. At best, they can only undertake retrospective analysis after states decide to de-classify highly sensitive material (such as that on the Cuban missile crisis, Allison and Zelikow 1999).
Rationalist accounts point to yet a third factor that hinders settlement (Fearon 1995; Gartzke 1999; Wagner 2000). This factor has to do with the commitment problem. Even when the two sides have consistent expectations, a deal may still elude them. More likely than not, the proposed deal re-allocates resources–money, territory, power, authority–between the parties so that there will be a relative gain for one and a relative loss for the other. Officials representing the side suffering relative loss are especially reluctant to enter into such an agreement if the gains and losses tend to be cumulative. What is there to prevent the side making the relative gain from reopening the same issue or opening new issues with further demands to redistribute resources? That is, gains made in this round can be used to seek additional concessions in future encounters. This situation presents a problem for commitment because the agreement lacks enforceability. Promises made by the party with relative gain are not credible because nothing prevents it from reneging on them in the future, taking advantage of its increased strength and therefore bargaining power. Nor can the party suffering relative loss commit itself credibly to vigorous resistance if it is presented with further demands for concessions (Neville Chamberlain learned this lesson too late and paid a heavy price for it). It is subject to the challenger’s salami tactics to extracting additional concessions. Both sides know that in the next round and the ones after, this party’s strength and hence bargaining position would deteriorate progressively.
This situation applies to cases of power transition. The ascending power has an incentive to avoid becoming the target of a pre-emptive strike from the current hegemon. It therefore has reason to promise that it will not use its power (current and prospective) in ways that are detrimental to the current hegemon. The latter understands its declining status and should also like to avoid the inefficiencies of a confrontation. Both parties, however, are caught in the predicament of an inability to make a binding commitment. The one experiencing upward mobility cannot commit to not use its increasing power to its advantage in the future, no matter what it says, and the one experiencing downward mobility cannot credibly commit to resist any future encroachments when it would be weaker (if it will not fight now, why should it be expected to fight then?). Note that misperception does not play a role in this scenario; in fact, both sides are in agreement about the unfolding power trajectories.
Commitment problems are especially severe in civil wars. In situations of ethnic rivalry and hostility, the majority group can offer guarantees to respect and honor the rights of the minority. But should the minority trust such promises? What is there to restrain the majority once it has gained exclusive control of the security forces, education institutions, and mass media? In the absence of such perceived restraint, the minority would rather fight to secede than to be incorporated in a majority-controlled state. The commitment problem therefore tells, at least in part, the story of recent conflicts in the former Yugoslavia (Fearon 1998). It also explains why some civil wars are protracted even when the contestants would have preferred to end them. Conclusion of the civil war would entail the disarmament of the combatants and the formation of a national army. After disarmament, each side becomes vulnerable to an attack from the other, especially if the other is to somehow gain control of the now integrated armed force. One common and striking pattern emerging from past episodes of civil strife is that settlement is facilitated when there is an interested, powerful, and impartial outside party to enforce the terms of the agreement. Conversely, in the absence of such a foreign presence providing security guarantees against possible defection by either party, a civil war is likely to drag on (Walter 1997).
It does not require much of a leap of inference to understand how the same commitment problem affects the prospects for settlement in cross-Strait relations. Notwithstanding all of the pledges made by Beijing to respect various spheres of autonomy for Taiwan, those pledges lack a binding quality once national unification is established. President Jiang Zemin’s “Eight Points,” enunciated in 1995, sought to alleviate Taiwan’s concerns by reassuring its people that on reunification with China, they would still be able to maintain their judicial, legislative, and financial autonomy (including tariffs and tax authority) and, moreover, keep their own armed forces and retain their independent international ties. Reunification, however, would force Taipei to forfeit its sovereignty claim, a huge source of protection and hence bargaining lever if recognized by the international community (notwithstanding the fact that sovereignty claims sometimes are insufficient to prevent international intervention in civil wars, most recently in Kosovo). Still, these verbal reassurances did not impress Taipei because of their lack of perceived enforceability. Even if Taipei is willing to give the current leaders of Beijing the benefit of the doubt, it may rightly be skeptical about whether future Chinese leaders can still be relied on to live up to the pledges made by their predecessors.
By the same logic, Washington may seek to reassure Beijing that its plan for ballistic missile defense is intended strictly for rogue nations such as North Korea. Despite such verbal reassurances, however, Washington faces the same commitment problem. Why should Beijing believe in U.S. promises because, once built, such missile defense systems can easily be redeployed against China’s meager strategic forces? What will be the penalty for Washington if it fails to live up to its promises? Who will be powerful enough to impose this penalty on the United States? As additional examples of the commitment problem, what guarantees does Israel have that the Golan Heights will not be used to attack it if this territory is returned to Syria, what guarantees does North Korea have for U.S. assistance in the provision of energy if it dismantles its nuclear reactors, and what guarantees could Saddam Hussein extract from the United States to cease its efforts to topple his regime in return for his compliance with UN disarmament programs? Would it be reasonable for each of these parties to expect more rather than less pressure in the wake of its concessions? For instance, could Iraq’s disarmament make it easier and more probable for the United States to attack? If the target believes so, it is more likely to resist than yield.
Contracts are self-enforcing only if a prospective aggrieved party can unilaterally withdraw from a relationship and this withdrawal is in itself the penalty for the other party’s failure to live up to its obligations. Self-enforcing agreements are only applicable to continuing sequences of exchanges, that is, situations that do not involve a one-shot deal with the possibility of serious and irrevocable damage to one of the contracting parties in case of opportunistic behavior by the other (the examples cited in the previous paragraph tend to lack this important qualification of repeat plays, therefore further compounding the commitment problem). Bilateralism and reciprocity are also central features of self-enforcing contracts (Yarbrough and Yarbrough 1992). “Reciprocity (even in the negative sense of threatened retaliation) is an essential feature of self-enforcing agreements because it forms the enforcement mechanism in the place of third-party intervention” (Yarbrough and Yarbrough 1986, 16). The point about self-enforcing contracts is, of course, that even though a settlement may be mutually attractive if both parties comply, its outcome would be detrimental to one party in case of non-performance by the other. This realization in turn impedes an agreement from being reached absent a viable enforcement mechanism. Self-interest demands that the greater the hazard of opportunism, the greater would be the contracting parties’ ex ante insistence on guarantees and safeguards that would minimize their vulnerability to ex post cheating and defection.
To ensure compliance and discourage opportunistic behavior (such as failing to perform according to the terms of an agreement), it helps to have a powerful and impartial external enforcer, or bilateral hostage-taking and issue linkages (if you fail to perform, you will lose the equivalent of a bond or escrow deposit, or if you cheat on this issue, I will get back at you by retaliating on that issue). Naturally, the “hostage” in question must be of sufficient value to its provider, and the issues involved must be of equivalent importance for such mutual reassurance to be credible. At the risk of being repetitive, the theoretical point to be underscored is that reassurance must be costly to its provider to be credible. The other side will not believe in any reassurance unless its provider will really have to forfeit something valuable in case of nonperformance or opportunistic behavior. With respect to this issue (and others to be discussed below), country specialists can make a valuable contribution. They are, for example, in a more qualified position to offer advice on what is truly important to Beijing, which can be offered to Taipei as a “hostage” or bond to alleviate the latter’s concerns about the former’s commitment credibility. Likewise, what should Washington be willing to offer to Beijing to demonstrate its sincerity when it professes that its planned missile defense system does not have Beijing’s nuclear forces in mind as a possible future target? Note that an offer of a token side payment is insufficient because whether it be a missile defense system or an integrated national army, the alteration of the status quo in question presents the so-called “first-mover advantage” such that the initial beneficiary can use its recently gained advantages to press for further concessions, whereas the other party bears all the risks of being further exploited.
Time sensitivity also enters strategic calculation in that the commitment problem, as already mentioned, applies well into the future. How can current political incumbents bind their successors to their pledges? Can the promises made by a Democratic president be disavowed by a subsequent Republican president? Can commitments made by the executive branch be contradicted by the legislative branch? This commitment problem–sometimes in the form of a particular administration pledging not to engage in protectionist policies or violate basic human rights–is often addressed by entering into multilateral agreements or joining international organizations (for example, Pevehouse 2002; Mansfield et al 2002). This is done so that the current administration may be able to use international treaty obligations to bind the hands of future leaders (such as by forcing them to balance the budget and accept multiparty competition), and to do so by increasing the costs of reneging in terms of their international ramifications (that is, more states will be upset because of a failure to live up to the agreement’s terms). Naturally, the severity of the commitment problem depends on whether a country has a fragmented power structure with many veto groups and stakeholders, whether it faces recurrent electoral cycles involving a high rate of turnover among officials, and whether it cares much about international opinion or is prone to act unilaterally.
Even though everyone can agree that wars are expost inefficient, states are often unable to reach a settlement to avoid them. This inability to settle is because of incompatible ex ante expectations about how a potential war will turn out. Such discrepant expectations can be explained without recourse to factors such as misperception or emotion. They are quite natural because of the tendency to engage in deliberate misrepresentation and the possession of private information. Even when the parties have consistent expectations, a settlement may still prove elusive. They are understandably reluctant to settle when the stakes are immensely high and the temptation for opportunism is equally great. The commitment problem points to the difficulties of enforcing contracts in a world of structural anarchy and self-help, even when such agreements may be beneficial to the parties. A key insight of the metaphor of the prisoners’ dilemma is, of course, that because of their fear of being taken as the “sucker,” both parties eschew cooperation that would have brought them greater benefit. They would settle instead for a suboptimal arrangement to avoid their worst-case scenario (whereby the sucker receives a long jail sentence, but his accomplice is released from prison).
DETERMINING DETERRENCE SUCCESS
How does one know when a deterrence policy has succeeded? By definition, deterrence succeeds if nothing happens. That is, the nonoccurrence of the unwanted event (an attack by the challenger) is a necessary condition for demonstrating deterrence success. However, this nonoccurrence is hardly sufficient proof. After all, the alleged challenger may not have any intention to attack in the first place. Alternatively, even if it has such an intention, it may have been deterred by some consideration other than the defender’s attempt at deterrence. Was the U.S. deterrence against the USSR effective? That Moscow did not attack Western Europe might have been because of the effectiveness of this U.S. deterrence. It could also be that Moscow never had any intention of launching such an attack, or that it might have been dissuaded from carrying out such an attack by its economic weakness, dissension among its leaders, concern for a challenge from China, or some other reason. Thus, one needs to guard against spurious inferences, giving credit to deterrence for keeping peace and stability when this policy does not quite deserve such credit. To avoid this error of false positives, one would have to engage in counterfactual reasoning, imagining how history would have turned out differently in the absence of an attempt to deter. Naturally, this approach treats history as a matter of contingency, recognizing that alternative outcomes are possible given different policies or circumstances. This is another assignment that country specialists, with their intimate knowledge of the supposed challenger’s domestic politics and strategic culture, are much better equipped to undertake.
Conventional wisdom has it that deterrence success depends on the balance of capabilities and the balance of interests between the defender and the challenger. The party with greater capabilities and more important interests is less likely to back down. Therefore, if the challenger has weaker capabilities and less at stake, it will not mount a challenge in the first place–it has been effectively deterred by the defender. Alternatively, when faced with a challenger with stronger capabilities and more at stake, the defender will concede by not resisting the former’s demand to alter the status quo. A slight variation to this account adds resolve (or determination) as a third variable. A deterrence situation thereby also becomes a contest of will. Assuming equal capability and matched interests, the side that is more highly motivated or committed to its cause is expected to prevail. A related and critical argument contends that to be seen to have the means to carry out the deterrence threat is not enough; the challenger must also believe that the defender has the will to carry it out for deterrence to succeed.
In reality, a challenger can be completely convinced that the defender has both the necessary will and ability to execute its threat and still launch an attack against a protege and even against the defender itself. On the eve of the Pearl Harbor attack, Japanese leaders had no illusion that the United States was much stronger than Japan, estimating that the balance of capabilities favored Washington by as much as 800 or 900 percent. They also believed that Washington was quite serious in communicating its resolve to oppose Japanese invasion against the European colonies in Southeast Asia (the protEgEs in this U.S. attempt at extended deterrence). Tokyo chose war because it thought it could not bypass the U.S. opposition to its seizure of Southeast Asia, and because it felt imperative to acquire the resources from this region in order to sustain its economy and war machine. Japan’s leaders would have preferred to take over Southeast Asia without provoking a war with the United States. However, precisely because they believed that the United States was not bluffing when it issued its deterrence threat, their options were reduced to having the United States impose war on them after their initial assault on Southeast Asia and seizing the strategic initiative by striking the U.S. Pacific Fleet pre-emptively. Neither superior force, nor credible resolve, nor acknowledged interest, helped extended deterrence in this case.
The critical missing point is that all of the plausible options available to the Japanese were unattractive. Concession to the U.S. demand to withdraw from Southeast Asia and China was unpalatable. Inaction was unacceptable because after eighteen months, Japan’s stockpile of military fuel would have been exhausted. Playing for time would accordingly have meant strangulation by slow motion. After depletion of its strategic reserve, Japan would have lost the war without having fired a shot, and the United States could then dictate the terms of Japanese surrender. Attacking Pearl Harbor therefore became a matter of choosing among the least of several evils. Faced with a choice of slow strangulation in the face of the allies’ embargo, fighting an enemy with superior forces, or giving up Japan’s aspirations to be a great power, “the only escape from the dilemma was by blunting one of its horns–to accept war with the U.S., but to attempt it under circumstances where the chances of victory were higher” (Russett 1969, 134).
This account shows that the defender’s ability and will to impose costs on the challenger is not the key to understanding deterrence success or failure. Neither the balance of capabilities nor the balance of interests tells the critical part of the story. Rather, the challenger’s relative valuation of the opportunity costs of compliance versus noncompliance with the defender’s demands holds the key. The challenger would surely prefer to not have to fight a war with the defender, but it sometimes has an even stronger distaste for giving up other aims and objectives that the defender demands. The literature on economic sanctions provides a parallel argument (Wagner 1988). Surely, the targets of such trade interruption or investment strike would rather not suffer this deprivation. However, they may still decline to comply with the sanctioner’s wishes because they find the price of compliance to be even costlier to the extent that this action entails sacrifices in national sovereignty or honor, political integrity, regime or even personal survival. What matters to the state being targeted for deterrence or sanction is its comparative disutility for compliance versus noncompliance, not the severity of the deterrence or sanction threat. Aggregate evidence from many past episodes of sanctions shows that the success of these influence attempts is not determined by either the severity of costs imposed on the target or the amount of concessions demanded from it (Drezner 1999, 1998; Pape 1997).
Thus, the target’s likely costs of challenging the deterrence or suffering the sanction are not meaningful in isolation. To that extent, deterrence and sanction research that focuses only on the defender’s ability and willingness (credibility) to impose costs misses the point. The point is that compliance will only be forthcoming if its costs are seen by the challenger to be less substantial than the costs of noncompliance. Similarly, the significance of concessions being sought from the target can only be evaluated relative to its other values and objectives. It must care less about giving up what the defender is asking than what the defender is offering in order for deterrence (or sanctions) to work. Surely, Saddam Hussein would have wanted to stop the economic embargo against his country, but he would rather not pay the price for lifting this embargo if it meant his personal or political demise. Scholars with generalist training are not very good at determining the challenger’s comparative expected disutilities, whereas researchers with special insight or access to this state’s decision logic or process may be better able to shed light on this important matter.
So what is a defender or sanctioner offering to the challenger to encourage the latter’s good behavior? In the case of extended deterrence, the defender offers to not execute its threat if the challenger does not attack the protEgE. It therefore offers to maintain the status quo. In the case of sanctions, the sanctioning state promises to restore the status quo ante, such as to resume trade or investment, if the target state complies. The difference between these two cases, of course, is that in the former the defender is seeking to forestall an undesired action by the target, whereas in the latter it is trying to compel the target to change an existing behavior or policy. That more is required of an influence attempt to get a target to give up something than is required to dissuade it from doing something in the first place is well understood in traditional international relations scholarship. More recently, prospect theory (Kahneman and Tversky 2000), focusing on psychological phenomena such as the tendencies for loss aversion and endowment effect, reinforces this understanding. Rationalist explanations, however, offer additional insight on this topic.
Think of deterrence and sanction efforts in terms of an exchange model. One state is offering to do something or to refrain from doing something in return for the other state to behave (or not behave) in a particular way. The first state is therefore engaged in linkage politics with an explicit expectation of bilateral reciprocity. In making the deterrence or sanction threat, this state is in effect declaring that it cares about certain things (such as Taiwan’s autonomy, human rights in China) more than other things (such as military confrontation with Beijing, trade with China). It has already been suggested that such a declaration is subject to deliberate misrepresentation. The main point, however, is about the clarity of the qui pro quo being presented and the likely attractiveness of this proposed deal to the challenger. What exactly is being threatened if the latter attacks the protege? Why should the challenger believe that it would be better off making a joint deal over, say, trade and human rights, rather than negotiating separately on each issue?
As Wagner (1988) noted, the sanctioning state has a bargaining leverage over its target to the extent that it has not extracted fully its advantages in its previous dealings with the target. That is, unless the former has not received terms as favorable as it could have received in a fair exchange, it would not have anything to offer the latter to compensate for its changed behavior. The would-be influencer needs to apply the foregone benefits on one issue to seek reciprocity from the other side on a different issue. Why? Because the target may not be able to get nearly as good a deal from other commercial partners should trade or investment be disrupted. A key determinant of influence success therefore revolves around the question of exactly in what ways and to what extent the sanctioner has settled for a less advantageous deal in one case so that it can effectively apply this concession to demand a more advantageous deal in another case. According to the logic presented earlier, that the target will suffer serious deprivation and that it will be subject to heavier costs than the sanctioner in case of a rupture in relations are not the most critical factors in assessing the prospects of influence success. Note moreover that the substitutability of trade (or investment) is critical (that is, the question of how easily the target can replace the lost trade or investment), a consideration that in turn underscores the importance of multilateralism already touched upon.
The discussion so far suggests that a preoccupation with the defender’s capacity and inclination to impose costs on a prospective challenger-the search, if you will, for the latter’s break point, such as in the Vietnam War (Mueller 1980)–is partial at best and misleading at worst. It directs attention to quantitative accounts (such as body counts, bombing sorties, hamlets pacified) without getting at another crucial consideration. Whereas A may be able to inflict more pain on B, B may be able to offset this asymmetry by being able to accept and sustain a higher level of tolerance for pain (Rosen 1972). Morale, resolve, perseverance, and national purpose are intangible qualities that are nevertheless critical in determining the contestants’ capacity to endure privation. Any reader of Mao Tse-tung’s strategic writings would not have missed his emphasis on the human spirit prevailing over material conditions (“nuclear weapons are paper tigers”). States develop a reputation for their relative sensitivity to such privation, such as the alleged U.S. allergy to body bags. They can also actively mobilize and prepare the material and psychological conditions for enduring privation (for example, digging shelters, storing grain, waging people’s war). Note that the relevant privation may include, especially for democracies, the price of declining popularity and electoral defeat.
Note further that a challenger can expect serious setback on the battlefield and still launch an attack. As will be discussed later, military action is itself a signal intended to reveal private information that cannot be credibly disclosed otherwise. The purpose of such action is not outright battlefield victory (the Viet Cong never achieved this feat). Rather, it is to force the defender to revise its prior expectations. The challenger just has to do “better” than what the defender had expected it to do, such as by fighting harder and persevering longer. Consequently, one can lose on the battlefield and still “win” in the sense of forcing the other side to rethink the probability of its success, the costs (including domestic political costs) likely to be associated with the outcome, and the amount of time and other investments probably required to accomplish it. By demonstrating that it can actually perform better on the battlefield than its opponent had expected, one in effect is saying by action that one deserves a better deal than that to which the other side has heretofore been willing to agree. The Viet Cong’s Tet offensive and Anwar Sadat’s decision to launch the Yom Kippur War demonstrate the rationalist logic just discussed. They do not have to defeat the Americans and Israelis; they just have to prove that they can do (hopefully for them, substantially) better than what their adversaries had expected.
According to the rationalist perspective, wars are indeed an extension of negotiation. In fact, by clarifying, validating, or refuting the disputants’ prior expectations, battlefield developments enable them to reach an agreement that has thus far eluded them because of deliberate misrepresentation and private information. Fighting forces the parties to back up their words with deeds. They now have to prove that they are in fact as strong as they said and as resolved as they professed. Such information is much more credible and hence useful to the two sides because it is much less subject to manipulation and misrepresentation. Naturally, the more discrepant or inconsistent the disputants’ prior expectations, the more fighting is required to reconcile these expectations, that is, to convince both that they are better off by settling than by continuing with the conflict. Seen in this light, deterrence success or failure is no longer a matter of the mechanical balance of capabilities and interests between the two sides, as seen by academics. To return to the poker analogy, the “game” involves multiple rounds of betting (which signals the strength of one’s hand and one’s resolve to stay around) whereby the players’ expectations become increasingly more convergent. This is not to imply that bluffing and deception are banished, only that one is more inclined to believe that another player really has a full house if he or she makes the maximum bet, rather than if he or she throws only a dollar in the pot. Moreover, to say that the players’ expectations converge does not imply that these expectations are necessarily correct (after all, the one holding the stronger hand may actually back out of the game).
Finally, by casting deterrence in this way, I argue that this is an iterative policy and interactive process. That is, deterrence is not a matter of simply preventing X from happening so that the policy’s success can simply be judged by whether or not X comes to pass. Rather than treating this undertaking in a binary way or as a once-for-all unilateral commitment, both the threat to defend and the threat to challenge can be increased or decreased, that is, adjusted over time by the parties on the basis of their evaluation of previous encounters. What is being deterred can be a question of more or less (how much X rather than whether X, such as the level of terrorist attack against Israel, the extent of speculation against the Japanese yen, or the amount of drugs smuggled across the Mexican border). If this view is accepted, the discussion about nonevents at the beginning of this section needs to be revised. Deterrence success, then, also becomes a matter of degree (for example, the number of terrorists, speculators, smugglers, or, more familiar to most of us in our daily experiences, speeding motorists who were dissuaded from undertaking the undesired act because of the deterrence threat). With this changed perspective, a defender’s threat can be seen to have restrained the scope, severity, and duration of the challenger’s attack on a protege even if it did not prevent this attack from taking place. China’s 1979 “pedagogical war” against Vietnam presents such an example (the USSR was the defender in this case). Note also that in this case, China was more interested in forcing a policy change on Hanoi, rather than making any territorial gain at the latter’s expense.
EXPLORING EXTENDED IMMEDIATE DETERRENCE
Extended deterrence can be general or immediate. When the United States announces that it will oppose any Soviet attack against a NATO partner without having any particular protege in mind and in the absence of any specific move by Moscow to prepare for such an attack, Washington is engaging in extended general deterrence. When China warned the United States against crossing the 38th parallel after Washington had made specific moves to invade North Korea, China was undertaking extended immediate deterrence. Likewise, when the United States threatened to resist Chinese moves against the offshore islands during the 1950s, it was applying extended immediate deterrence (Morgan 1983).
What determines the success or failure of extended immediate deterrence? Would the amount of trade between the defender and protege, the level of arms imported by the latter from the former, and the existence of a formal alliance between them make a difference? Would the possession of nuclear weapons by the defender make a difference? Research on questions such as these has been carded out by international relations scholars seeking systematic patterns across all known historical cases of such deterrence in the past century or so (for example, Huth 1988a, 1988b; Huth and Russett 1993, 1988, 1984; Huth et al 1993). Significantly, the factors just mentioned (trade, arms transfers, alliances, and nuclear weapons) are common knowledge. Presumably, leaders in Beijing are not ignorant of such facts, and they will draw some inferences from them.
For instance, a formal treaty of mutual defense offers a meaningful clue to a defender’s intentions. Why? Because alliance ties carry costs and, precisely because of these costs, alliance ties offer a credible signal to the challenger about the defender’s seriousness (even though, as mentioned earlier, states often fail to honor their treaty obligations). What kind of costs? Members of an alliance engage in peacetime coordination, share their defense plans, and, above all, announce to the rest of world that they forfeit their discretion should war come (Morrow 1994). That is, they commit themselves to supporting each other in case of war, thus denying themselves the option of reconsidering their allegiance after a war has occurred. This is a huge opportunity cost, especially because this binding commitment increases the risk of being dragged into a war provoked by a protege that a defender would otherwise rather not fight, or vice versa. Finally, an alliance commitment is costly because should a state renege, it will have to pay a price in its international reputation for unreliability. Other states will be less inclined to trust its pledges in the future. When a state takes on an ally, therefore, it knowingly and deliberately sends a costly signal in the hope that this communication would be credible to potential aggressors. The absence of a formal alliance, such as in the case of Washington’s relationship with Israel and Taiwan, also of course becomes a signal. What makes Washington want to treat, say, Taiwan and South Korea differently in terms of their status as a formal ally? As a signal of its commitment, Washington stations its own troops on the front-line in Korea to serve as a trip-wire. Systematic analysis of past cases of extended immediate deterrence led Huth (1988a, 215) to conclude that “the deterrent value of an alliance commitment cannot be separated from the analysis of the immediate balance of forces and the importance of a trip-wire force in position for the defender [on the protege’s territory].”
Although common alliance membership and other factors such as trade ties and armament sales should make extended general deterrence more credible (because they tend to show that the defender has a significant interest in maintaining ongoing relations with the protege), they should predict that extended immediate deterrence is less likely to succeed (Fearon 2002, 1994b). Recall that in the latter situation, the challenger makes specific moves against a protege with the full knowledge that the defender has already made a general announcement of intention to protect targets that include this protege. The challenger’s behavior is understood by the rationalist perspective to suggest that it sees itself to be more resolved than the defender, or else it would not have mounted the challenge in view of the defender’s prior pledge to resist its aggression. By the same logic, the challenger must also feel that it has a greater stake in the issue being contested. Moreover, the more the defender enjoys a favorable balance of capabilities, the more resolved the challenger must be for it to proceed with its challenge. A less committed or motivated challenger would not have run this risk of initiating a confrontation with a stronger adversary. A corollary implication is that as the bilateral balance of forces becomes more symmetric, it becomes more difficult to infer the challenger’s resolve from its behavior.
An analyst may, however, seek clues of motive and resolve from prior encounters between the defender and challenger. The United States and China had confronted each other on a number of occasions in the Taiwan Strait, especially in 1954 and again in 1958 over the offshore islands of Quemoy and Matsu. Accounts of these prior encounters overwhelmingly agree that China (the challenger) backed down, and the status quo was preserved. A minority and unconventional interpretation of these episodes argues that China (more specifically, Mao Tse-tung) was actually trying to prevent a change in the status quo by locking the Nationalists (and the United States by extension) to the defense of the offshore islands to avoid their detachment from Talwan, and hence to minimize the danger of Talwanese independence, as such a detachment would have enhanced this prospect both politically and geographically (Whiting 1975). It has also been suggested that China escalated these confrontations to prevent Taiwan from joining Washington’s multilateral anticommunist alliance of the 1950s (Stolper 1985) and more recently to influence Taiwan’s presidential election in 1996. These anecdotes suggest that there can be considerable ambiguity about the identity of the deterrer and deterree and about which status quo each party is supposed to protect (see also Lebow and Stein 1990, 353-55). They are introduced as a caution against the matter-of-fact assertions that often cast China in the role of a challenger to the status quo. From Beijing’s perspective, the threats of Taiwanese independence and U.S. missile defense undermine the status quo.
If the conventional wisdom is correct, Beijing was forced to back down in previous confrontations. What does this say about possible future confrontations? To answer this question from a rationalist perspective, one would consider the reasons that would incline the side “losing” the previous round to accept another confrontation. From this perspective, one would argue that each time a state engages in a public confrontation, it commits its prestige and reputation. Therefore, backing down involves not only the cost of giving up (even if only for the time being) one’s demands on the issues being contested but also the damage done to one’s prestige and reputation. Unless one accepts the implausible proposition that leaders do not care about their prestige or reputation (what Sinologists would call “face”), one would expect them to carefully calculate this possible cost of confrontation (another reason why wars and even public disputes that fall below this level of antagonism are inefficient). This is especially so if one has already been forced to back down in previous encounters–and especially when the defender has by its prior action demonstrated its resolve to stand by its deterrence threat.
The rationalist logic argues that having taken into account this common knowledge about the past and having anticipated the defender’s likely response in the next confrontation, a challenger must be even more resolved or highly motivated than before if it decides to mount another challenge. Given the challenger’s decision to challenge, one must assume that it must have already neutralized that which is known publicly (such as to academics); in other words, considerations such as the balance of forces and the defender’s ties with the protege must have already been taken into account (that is, discounted) by the challenger. The question then becomes what private information the challenger must possess to offset the publicly known facts about the defender’s prior resolve and success at deterring it. This reasoning leads to the expectation that the more often previous confrontations have resulted in disappointing and even humiliating outcomes for the challenger, the more likely it is that the next round of extended immediate deterrence will fail. Indeed, these encounters do not even have to produce negative outcomes for the challenger to have this hypothesized effect. Just the fact that the parties have had past clashes indicates that the challenger must be more highly motivated in a subsequent encounter. This expectation receives statistical support from the aggregate evidence of past deterrence episodes (Fearon 1994b).
One example comes from the deliberations of the Russian officials on the eve of World War I. Because Germany had successfully forced Russia to back down in its support for Serbia on previous occasions in the Balkan conflict, St. Petersburg was determined not to suffer the same indignity again (even though it was willing to pressure Belgrade to make substantial concessions to Vienna in the hope of avoiding a large conflict). Interestingly, while Berlin was applying extended immediate deterrence against St. Petersburg in the hope of forcing the latter to call off military mobilization against its ally Austria-Hungary, St. Petersburg was also seeking to deter Vienna from humiliating its ally Serbia. Therefore, the events of July and August 1914 showed that multiple sets of extended immediate deterrence occurred in Europe. More recently, in 1978 and 1979, we saw the same situation when China attempted extended immediate deterrence in connection with Vietnam’s attack on Cambodia, and the USSR in turn tried to protect its ally Vietnam from a Chinese attack.
A defender’s military superiority, particularly its possession of nuclear weapons, should not affect the success of extended immediate deterrence. A challenger presumably has already considered this publicly known factor in its decision to mount an attack. That it has chosen to proceed despite this factor (for example, North Korea, North Vietnam, and Iraq since 1945, and Japan in World War II when facing a nonnuclear but vastly stronger adversary in the United States) can only mean that it is highly resolved. This resolve in turn implies that the ensuing struggle can be expected to be especially bloody and protracted. Argentina’s war against the United Kingdom over the Falklands/Malvinas may be a partial exception to this generalization. Although the United Kingdom is a nuclear power, Leopoldo Galtieri and his colleagues apparently believed that Argentina commanded local military advantage and could achieve a quick and decisive victory before the United Kingdom could do anything to reverse this fait accompli. Rationalist explanations of war would argue that only highly motivated and resolved states would select themselves into a conflict with an adversary that is known to have a significant military edge. The less motivated and resolved ones would not have chosen to initiate a conflict, producing cases where the dog did not bark (that is, the nonoccurrence of the expected attack, representing successful deterrence).
Earlier, I mentioned that trip-wires can enhance the credibility of a deterrence threat. The deployment of the defender’s military forces in the immediate vicinity of the contested territory and especially in forward positions enables it also to act expeditiously and effectively to counter a challenger’s attempt to create a fait accompli. Paul Huth (1988a, 74) argued that “the effective military deterrent is the capacity of the defender to repulse an attack and deny the adversary its military objectives at the outset and early stages of an armed confrontation.” In this view, the challenger typically seeks quick and decisive victory over the protege at a relatively low cost. In 1938 and 1939, German military superiority over Czechoslovakia and then Poland rendered irrelevant Britain’s and France’s declarations of support for their proteges. Even if it was their intention to come to the aid of their allies (which Hitler doubted), Britain and France lacked the capacity to influence the course of events in the immediate areas of contest. In any case, slow and limited French and British mobilization encouraged Hitler’s belief that these countries were simply bluffing. Confident in the belief that the United States had vastly superior forces to deliver a quick and decisive knockout blow to the North Koreans, General Douglas MacArthur was inclined to ignore Chinese warnings of intervention. The Chinese threat of deterrence did not appear credible because Beijing did not have enough forces ready and in the immediate vicinity in time to head off a U.S. decision to cross the 38th parallel (Whiting 1960). The earlier discussion on private information would suggest that even if it did, Beijing would have had a tough time convincing Washington of the seriousness of its intent without giving away the element of strategic surprise.
The timeliness of warning is important. It is much harder to get a challenger to call off a campaign after it has been launched. The U.S. invasion of North Korea is one example. Argentina’s seizure of the Falklands/Malvinas offers another. How could early steps deter a challenger before it escalates the conflict to the point that the defender is faced with a choice of either direct military intervention or humiliating concession? Given the known tendency for officials to engage in deliberate misrepresentation, a defender must take actions that a less committed or resolved state would not have been willing to undertake. James Fearon (1994a) convincingly argued that domestic audience costs provide a way for leaders to demonstrate that they are serious and not just bluffing. Domestic audience costs come as a result of leaders staking out a public position on a policy in such a way that they will suffer serious damage to their reputation and career if they fail to subsequently carry out their pledges. In a democracy, opposition parties, legislative committees, media critics, and ultimately the voters can censure officials for reneging on their policy commitments. The more often, specific, and public are pledges of support extended to an ally, and the higher the level of officials making these pledges, the greater the audience costs (both international and domestic) in the event of nonperformance.
These audience costs, described by Fearon (1997) as representative of a policy of “tying hands,” are only paid ex post if the defender backs down. In contrast, the alternative policy of “sunk costs” entails costly ex ante steps, such as deploying troops on the protege’s territory and establishing a joint command, to signal credible deterrence. The “tying hands” approach is more tempting because one avoids paying the costs if deterrence is successful, and for this reason its practitioners tend to include both the more resolved and less resolved states (the latter seeking to achieve a “cheap” victory without actually intending to “walk the walk” if challenged). Although seemingly more efficient than the “sunk costs” approach, it has the offsetting liability of tending to “lock in” the adversaries and is therefore more likely to provoke escalation.
Democratic leaders pay a heavy political price when they are unable to honor their commitment. That autocratic leaders, facing fewer internal constraints and weaker political opposition, are more immune to domestic audience costs makes their policy commitments less credible. Autocratic leaders can avoid or minimize such costs because they are less accountable to their constituents and watchdogs. This difference in domestic audience costs makes foreigners less inclined to believe in the declared policy intentions of autocratic leaders than their democratic counterparts (with the corollary implication that the autocratic leaders will have to overcome this handicap by resorting more to “sunk costs” to demonstrate their resolve). Significantly, both autocratic and democratic leaders pay the same price in their reputation for unreliability in international circles if they do not honor their commitments. The difference in the credibility of their commitments lies in their different susceptibilities to domestic censure. Therefore, other things being equal, threats of extended deterrence are more believable if they are made by democratic leaders. As a corollary, however, when democratic leaders equivocate in declaring their policy intentions, it means that they wish to avoid domestic audience costs. This inference in turn erodes the credibility of their (vague) deterrence threats. Adolf Hitler was unimpressed by the vacillating statements coming from London in 1938 and 1939. His skepticism was reinforced by his judgment of the inherent stakes for the parties involved (as suggested by the quote in the opening paragraph). Recent research argues that a defender’s inherent stake in a protege’s security (as judged by the challenger) is more important than communicative signals in determining the success or failure of extended deterrence (Danilovic 2001 a, 2001b). Some deterrence attempts, such as the doctrine of massive retaliation in response to aggression against seemingly inconsequential allies or neutrals, are deficient because they fail the test of inherent stake.
There is a further implication from this discussion. There will be a transition period when China leaves behind the legacy of totalitarianism of the first three decades of the People’s Republic but before it reaches the point of full, consolidated democracy. During this interim period, its political process will become more competitive, there will be more pluralist voices, and public opinion will matter more for foreign policy, if for no other reason than the fact that rival politicians will be more inclined to mobilize mass support for their own political gains. Beijing’s leaders, in other words, may become more subject to domestic audience costs. In the context of Robert Putnam’s (1988) famous metaphor of two-level games, foreign policy settlements will increasingly require domestic ratification by multiple interest groups. It is not clear that for such transitional polities international accommodation will necessarily prevail over assertive nationalism (Mansfield and Snyder 1995). Although they may very well question the legitimacy of their leaders and the effectiveness and appropriateness of their government’s policies for a variety of reasons, national reunification is one issue that is likely to enjoy fundamental consensus among Chinese of all stripes.
It was mentioned earlier that officials often join international organizations to enhance the credibility of their commitment to particular courses of action (such as not to inflate their economy or abuse human rights). They undertake international treaty commitments with a view not only of signaling their commitment to domestic and foreign audiences (for example, international investors) but also as a way of binding future leaders. This logic also implies that when a state mobilizes international support for its policies and seeks the approval and endorsement of international organizations for these policies, its declared objectives and intentions are inherently more credible than when it engages in unilateralism. Indeed, one critical factor in determining the effectiveness of international sanctions is whether such efforts are undertaken as a matter of collective action. Multilateral coordination and enforcement provide prima facie evidence that the leading state is serious given the transaction costs involved and the likely costs of reneging on one’s commitment to multiple collaborators (Martin 1993, 1992). By extension, extended deterrence–whether general or immediate–is more compelling when it is presented as an outcome of multilateral diplomacy. Note that this greater credibility is being attributed to the signals conveyed by the collective action and not necessarily to the enhanced capability accrued by the defender’s coalition.
The coalition partners of the leading defender are often in a predicament: they want to show the challenger that they are serious in joining the defender in opposing aggression, but they also want to restrain the defender from escalating a crisis to an unwanted point. In 1990-91, those states joining the United States in threatening Iraq with military action in the hope that such action might not be necessary found themselves ironically tied to a U.S. policy that escalated the confrontation with Baghdad whereby military action became more unavoidable (Nossal 1994). Also, as noted in the earlier reference to “leashing” Chiang Kai-shek, the defender may be rightly concerned about being “entrapped” by its own protege in an escalating conflict. These concerns point to what economists describe as the problem of moral hazard–that commitment begets opportunism.
Contrary to the standard practice of Sinology, I have not relied on official statements, media reports, or announced state actions (such as China’s military exercises in 1996) to describe or explain the relations among China, Taiwan, and the United States. Because such material is public information, readers have ready access to it. It is not clear that the story told by academics is necessarily credible, except perhaps after the passage of time has produced the disclosure of critical private information. After all, academics do not have the contemporaneous private information used by officials in deciding foreign policy, and they are not any less immune than these officials to succumbing to the deliberate misrepresentation to which such public information may be subjected. Although this acknowledgment may be disappointing and indeed ego-deflating, I hope that it does not diminish the relevance of rationalist interpretations offered in this article. These interpretations offer a reasonably coherent perspective that can account reasonably well for the aggregate historical evidence on war, deterrence, and sanctions. Moreover, they offer some insights that contradict conventional wisdom.
It is well-known that people tend to resort to different explanations when accounting for their own behavior and the behavior of others, especially adversaries (Mercer 1996). They are inclined to point to circumstances beyond their control when explaining (justifying) their own behavior that contradicts others’ interests. Conversely, they rely on others’ inherent dispositions (“bad faith”) when explaining the latter’s undesired behavior. After receiving repeated assurances from U.S. diplomats that Lee Tenghui would not be issued a visa to visit the United States, Chinese officials were shocked to learn within a matter of days of the latest such assurances that Washington had reversed its decision in May 1995 (Garver 1997). They were disposed to see this reversal as a sign of duplicity and betrayal of prior commitment, rather than as an outcome of bureaucratic incompetence or disagreement. This event followed earlier attempts by some elements of the Beijing leadership to press for a more assertive policy toward Washington in view of incidents such as President George Bush’s decision in September 1992 to sell F-16 fighter jets to Taiwan. This decision was perceived to renege on the U.S. pledge in the 1982 communique to decrease the value of such arms sales over time. It took active intervention by Deng Xiao-ping, who pointed to Bush’s electoral motivations for this sale, to overrule the dissident leaders (Garver 1997). Such a decision, based on a situational explanation of the other’s unwanted behavior, tends to be rare and to contradict the central tenets of attribution theory (a nonrationalist perspective).
Still, that they do happen is a cause for hope. Social conventions, compatible institutions, and dense webs of intense and long-standing interaction buttress relations among friendly countries despite their occasional, even serious, disputes. These ties foster political trust and a concern for the “shadow of future” (Axelrod 1984), which in turn offer reasonable assurance that promises will be kept and opportunism will be counterproductive. Moreover, democratization should provide additional impetus for transparent policy and public disclosure, reducing (although not eliminating) the role of private information. Given the commitment problem and the problematic reliability of third-party intervention and adjudication, self-enforcing agreements enable cooperation to occur even in the face of mutual awareness of possible opportunism. In their absence, the ex ante mutual benefits of cooperation would not be sufficient to reach a settlement because both parties are aware that breach is sometimes advantageous ex post. Self-enforcing agreements try to avoid such policy predicaments by “making the expected future benefits of continued compliance to serve as the bond” (Yarbrough and Yarbrough 1986, 18). That is, relations governed by these agreements are such that the present value of the expected future benefits of cooperation exceeds the expected value of defection or cheating to each party. The challenges to wise and prudent statecraft discussed in this article present the most compelling reasons for a policy of sustained and comprehensive engagement between the United States and China because there is no other way to develop those conventions, institutions, and webs of interaction necessary to enhance trust, promote transparency, and alleviate concerns for opportunism. Although it may be insufficient to bring about the spirit of Camelot, it should go a long way to banish the world of Hobbes.
Allison, Graham T., and Phillip Zelikow. 1999. Essence of decision: Explaining the Cuban missile crisis. 2d ed. New York: Addison-Wesley.
Axelrod, Robert. 1984. The evolution of cooperation. New York: Basic Books.
Danilovic, Vesna. 2001a. The sources of threat credibility in extended deterrence. Journal of Conflict Resolution 45 (3): 341-69.
–. 2001b. Conceptual and selection biases in deterrence. Journal of Conflict Resolution 45 (1): 97-125.
Drezner, Daniel W. 1999. The sanctions paradox: Economic statecraft and international relations. Cambridge: Cambridge University Press.
–. 1998. Conflict expectations and the paradox of economic coercion. International Studies Quarterly 42 (4): 709-31.
Fearon, James D. 2002. Selection effects and deterrence. International Interactions 28: 5-29.
–. 1998. Commitment problems and the spread of ethnic conflict. In The international spread of ethnic conflict: Fear, diffusion, and escalation, edited by David A. Lake and Donald Rothchild. Princeton: Princeton University Press.
–. 1997. Signalling foreign policy interests: Tying hands versus sinking costs. Journal of Conflict Resolution 41 (1): 68-90.
–. 1995. Rationalist explanations for war. International Organization 49 (3): 379-414.
–. 1994a. Domestic political audiences and the escalation of international disputes. American Political Science Review 88 (3): 577-92.
–. 1994b. Signaling versus the balance of power and interests: An empirical test of a crisis bargaining model. Journal of Conflict Resolution 38 (2): 236-69.
Gartzke, Erik. 1999. War is in the error term. International Organization 53 (3): 567-87.
Garver, John W. 1997. Face off.” China, the United States, and Taiwan’s democratization. Seattle: University of Washington Press.
Huth, Paul K. 1988a. Extended deterrence and the prevention of war. New Haven: Yale University Press.
–. 1988b. Extended deterrence and the outbreak of war. American Political Science Review 82 (2): 423-43.
Huth, Paul K., Christopher Gelpi, and D. Scott Bennett. 1993. The escalation of great power militarized disputes: Testing rational deterrence theory and structural realism. American Political Science Review 87 (3): 609-23.
Huth, Paul K., and Brace M. Russett. 1993. General deterrence between enduring rivals: Testing three competing models. American Political Science Review 87 (1): 61-73.
–. 1988. Deterrence failure and crisis escalation. International Studies Quarterly 32 (1): 29-45.
–. 1984. What makes deterrence work? Cases from 1900 to 1980. World Politics 36 (4): 496-526.
Jervis, Robert. 1976. Perception and misperception in international politics. Princeton: Princeton University Press.
Kaempfer, William H., and Anton D. Lowenberg. 2000. A public choice analysis of the political economy of international sanctions. In Sanctions as economic statecraft: Theory and practice, edited by Steve Chan and A. Cooper Drury. London: Macmillan.
–. 1992. International economic sanctions: A public choice perspective. Boulder, Colo.: Westview.
–. 1988. The theory of international economic sanctions: A public choice approach. American Economic Review 78 (4): 786-94.
Kahneman, Daniel, and Amos Tversky, eds. 2000. Choices, values, and frames. Cambridge: Cambridge University Press.
Lebow, Richard N., and Janice G. Stein. 1990. Deterrence: The elusive dependent variable. World Politics 42 (3): 336-69.
Levy, Jack. 1997. Prospect theory, rational choice, and international relations. International Studies Quarterly 41 (1): 87-112.
Mansfield, Edward D., Helen V. Milner, and B. Peter Rosendorff. 2002. Why democracies cooperate more: Electoral control and international trade agreements. International Organization 56 (3): 477-513.
Mansfield, Edward D., and Jack Snyder. 1995. Democratization and the danger of war. International Security 20 (1): 5-38.
Martin, Lisa L. 1993. Credibility, costs, and institutions: Cooperation on economic sanctions. World Politics 45 (3): 406-32.
–. 1992. Coercive cooperation: Explaining multilateral economic sanctions. Princeton: Princeton University Press.
Mercer, Jonathan. 1996. Reputation and international politics. Ithaca, N.Y.: Cornell University Press.
Morgan, Patrick. 1983. Deterrence: A conceptual analysis. Beverly Hills, Calif.: Sage.
Morrow, James D. 1994. Alliances, credibility, and peacetime costs. Journal of Conflict Resolution 38 (2): 270-97.
Mueller, John E. 1980. The search for the ‘breaking point’ in Vietnam: The statistics of a deadly quarrel. International Studies Quarterly 24 (4): 497-519.
Nossal, Kim R. 1994. Rain dancing: Sanctions in Canadian and Australian foreign policy. Toronto: University of Toronto Press.
Pape, Robert A. 1998a. Evaluating economic sanctions. International Security 23 (2): 195-98.
–. 1998b. Why economic sanctions still do not work. International Security 23 (1): 66-77.
–. 1997. Why economic sanctions do not work. International Security 22 (2): 90-136.
Pevehouse, Jonathan C. 2002. With a little help from my friends? Regional organizations and the consolidation of democracy. American Journal of Political Science 46 (3): 611-26.
Putnam, Robert. 1988. Diplomacy and domestic politics: The logic of two-level games. International Organization 42 (1): 427-60.
Rosen, Steven. 1972. War power and the willingness to suffer. In Peace, war, and numbers, edited by Bruce M. Russett. New York: Sage.
Russett, Bruce M. 1969. Refining deterrence theory: The Japanese attack on Pearl Harbor. In Theory and research on the causes of war, edited by Dean G. Pruitt and Richard C. Snyder. Englewood Cliffs, N.J.: Prentice Hall.
Sabrosky, Alan N. 1980. Interstate alliances: Their reliability and the expansion of war. In Correlates of war II, edited by J. David Singer. New York: Free Press.
Schelling, Thomas C. 1966. Arms and influence. New Haven: Yale University Press.
Schwebach, Valerie L. 2000. Sanctions as signals: A line in the sand or a lack of resolve? In Sanctions as economic statecraft: Theory and practice, edited by Steve Chan and A. Cooper Drury. London: Macmillan.
Stolper, Thomas E. 1985. China, Taiwan, and the offshore islands: Together with an implication for outer Mongolia and Sino-Soviet relations. Armonk, N.Y.: M. E. Sharpe.
Tversky, Amos, and Daniel Kahneman. 1977. Judgment under uncertainty. Science 185 (27 September): 1124-31.
Wagner, R. Harrison. 2000. Bargaining and war. American Journal of Political Science 44 (3): 469-84.
–. 1988. Economic interdependence, bargaining power, and political influence. International Organization 42 (3): 461-83.
Walter, Barbara E 1997. The critical barrier to civil war settlement. International Organization 51 (3): 335-64.
Whiting, Allen S. 1975. New light on Mao; Quemoy 1958: Mao’s miscalculations. China Quarterly 62 (June): 263-70.
–. 1960. China crosses the Yalu: The decision to enter the Korean War. New York: Macmillan.
Yarbrough, Beth V., and Robert M. Yarbrough. 1992. Cooperation and governance in international trade. Princeton: Princeton University Press.
–. 1986. Reciprocity, bilateralism, and economic ‘hostages’: Self-enforcing agreements in international trade. International Studies Quarterly 30 (1): 7-21.
Steve Chan is a professor of political science at the University of Colorado, Boulder
COPYRIGHT 2003 Heldref Publications
COPYRIGHT 2003 Gale Group