Commission on Physical Sciences, Mathematics, and Applications
|
Commission on Physical Sciences, Mathematics, and Applications |
Paul K. Davis, Rand
Deterring the invasion or coercion of weak or medium-strength states that are important but not vital interests of major states is a key strategic challenge of the new era. This paper describes strategies for doing so. It begins by using decision-modeling methods to identify factors that would influence the decisions of would-be aggressors, including factors idiosyncratic to individual leaders. It then discusses how both immediate and general deterrence might be strengthened by a variety of political, economic, and military measures. The measures discussed include reasonably capable defensive forces that cannot easily be bypassed, operational arms control to make surprise attack more difficult, forward-deployed protector forces, and formal arrangements through regional security structures that would assure the long-term punishment of aggressors through political and economic isolation and, perhaps, military measures. The paper also encourages identifying and rooting out "dangerous ideas" that increase regional tensions and hatreds, and that could encourage aggression during a crisis. The following pages document the methods described here and include extensive references to relevant literature in political science, psychology, history, and strategy.
This paper was developed for an international conference dealing with long-term stability and security in a multipolar world. Rather than discussing stability and security in the broad, however, it focuses on the challenges that follow from my central premise that a principal strategic issue for the developed world is how to deter invasion or coercion of weak and medium-strong states when the security of the threatened states is important but is not a "vital" national interest of the powers that might be the protectors.
This premise is provocative, primarily because of the reluctance of democracies to face up to challenges that do not clearly affect their truly vital interests. To some, it conjures up images of entangling alliances, world policeman functions, strategic overextension, and quagmires. To others like myself, it seems to be a sober expression of reality. If accepted, it has a considerable impact on how one thinks about foreign policy and defense planning.
In what follows, I start by illustrating how this deterrent challenge may arise and why it is so difficult. I then describe how deterrence issues can be examined with the aid of an analytic approach that focuses on influencing the decisions of human beings. This includes actually modeling the decisions of such leaders. I next abstract from this discussion a way to summarize deterrence factors in the form of a "success tree" that can help guide the development of strategies. Finally, I draw on insights from the decision-modeling approach to describe potential deterrent strategies that might be recommended to weak or medium-strong states, on the one hand, and strategies that might be recommended to the United States and its partners of the developed world, on the other. Many features of the strategies are familiar from other approaches, but some reflect more uniquely the decision-modeling's emphasis on the perceptions and reasoning of adversaries.
Let us begin by considering the challenge of deterrence in rather general terms. Who is to be deterred from doing what, what kinds of deterrence are worth distinguishing, why is deterrence sometimes difficult, and why are there some reasons for believing it is feasible to do better in the future than in the past?
The major states of the developed world want to deter international aggression as part of maintaining regional stability. Usually, however, the objective is discussed in abstract terms. To be more concrete, consider the following range of threats that might arise in the next 20 years as viewed from one American perspective.
None of these is implausible in the long run. Some, however, are more difficult to contemplate than a repeat of Saddam Hussein's invasion of Kuwait. For example, the threats involving Russia or China are uncomfortable because neither state is behaving aggressively today toward its neighbors and there is no interest in labeling either of them as a future "enemy." If matters go well, Russia and China will develop, liberalize, prosper, and interact continuously with other states as partners in developing a better world. On the other hand, that is not guaranteed and the extreme nationalist movement in Russia is certainly a matter of concern, as is the degree of bitterness expressed by some Russian military officers about the state of affairs in Russia and what amounts to the loss of empire. Although Russia's army is currently in disarray, it will remain huge and may pull itself together. It is also unclear whether, in the years ahead, China will view the world in classic balance-of-power terms or take the more liberal perspective reflected in the U.N. charter and the actual behavior of nearly all developed states.
The other complication in thinking about future threats is that many of the threats are to particular weak and medium-strong nations whose security is desirable but is not necessarily a "vital" national interest of the United States or other major states. As a result, it is difficult for governments even to discuss such threats within the context of national defense planning. Nonetheless, any of the aggressions indicated could be a serious affront to broad interests, even if not vital interests. But how do we deal with such threats, especially when they seem so remote and less than vitally important?
In this regard, consider that one of the paramount blunders of the last decade was the judgment by national leaders and strategists as they observed the disintegration of Yugoslavia that a war among the various emerging factions, although highly regrettable, would not strongly affect their own national interests. This view changed grudgingly with CNN's broadcasts of ethnic cleansing, émigrés flowing into neighboring countries, and the partial dashing of hopes for a new world order, but it seems clear that current world leaders do not yet know how to deal with threats to less than vital interests. Even if they had personal concepts on the matter, there is great public reluctance to get involved in unnecessary conflicts in foreign lands.
Given, then, the existence of potential challenges, especially to weak and medium-strong states that are not obviously vital interests of the United States or other major states, let us next consider deterrence and what we mean by the term, since it has many variants. In this paper:
By and large, the security challenges facing the United States and its NATO allies involve extended conventional deterrence. America's ally South Korea, of course, has a direct threat today from North Korea. In the distant future, Korea may have a virtual threat from China. Ukraine, the Baltic states, and Poland see direct threats. In what follows, I shall consider challenges of both direct and extended deterrence, of both the general and immediate varieties.
There has been so much said about deterrence that one might think that the issues and necessary strategies are well understood. Nuclear deterrence, to be sure, has succeeded for many decades, and the leaders of major states fully appreciate the reasons for avoiding nuclear warfare. The reality is much less happy, however, when one looks at direct and extended conventional deterrence. Although it seems to have worked for NATO's Central Region, Huth and Russett have demonstrated that immediate deterrence has failed more often than it has succeeded over a large set of crises in the 19th and 20th centurieseven though the aggressor ultimately failed roughly two-thirds of the time, which suggests that deterrence "should" have had a better track record. That it has failed so often under such circumstances is sobering and even alarming.
Some of the myriad reasons for deterrence having failed are as follows:
Even this list is not long enough. Consider that aggressive personalities such as Saddam Hussein and Slobodan Milosevic still seem to ascend to power all too frequently. Consider also that expectations have changed because of the alleged lesson taught by Bosnia about ethnic and religious differences being enduring and fundamental. And, finally, we should also face up to the sober reality that the United Nations is thoroughly ineffective in dealing with security threats requiring prompt and decisive actions.
In light of these discouraging observations (see also Watman and Wilkening, 1995, and Arquilla, 1995), is deterrence even feasible in difficult cases? There are in fact several reasons for optimism:
To put it differently, despite the Bosnian debacle, one can argue that overall trends are still favorable. We should not focus unduly on exceptional cases.
Against this background let us now move to a discussion of deterrence theory. Although much has been written on the subject, the usual tendency has been to treat only some aspects of the subject while ignoring or giving short shrift to others. There is nonetheless a substantial volume of serious thought on which to draw in considering deterrent challenges and potential strategies.
Over the last decade I have taken a rather different approach to the study of deterrence than has been customary. It focuses on the decision making of leaders and on "natural" variables. The approach is motivated by several observations. First, the incentives and perceptions of aggressors are often intensely personal, as one can appreciate by thinking of Saddam Hussein's 1981 reaction to the threats of Iran's Khomeni, of current-era North Korean leaders who must worry about their personal survival, of Saddam Hussein in 1990 as he compared trends to his self-image, or to Slobodan Milosevic with his dreams of a Greater Serbia. Historically, we might think of Hitler in this century or, for example, Alexander the Great of antiquity.
A second observation here is that "great men of history," whether appropriately identified as such or merely self-proclaimed, are "special." Many do not reason in the same way we think normal political leaders do. Their values are different, their attitudes toward risk are different, and their interpretation of information is different. So also is the reasoning of states dominated by ideological and ethnic-hatred considerations "special."
All of this suggests an approach to deterrence that focuses on influencing the decisions of human leaders or groups of leaders. That is, instead of everything being a matter of abstract power balances, successful deterrence depends on one or more human beings reaching certain conclusions after thinking about the situation and alternatives. The decision makers are attempting to be rational, but an observer might think the reasoning or actions to be "irrational" or "crazy." It is preferable to avoid that terminology because it is misleading and generates the notion that worrying about how to deter will be fruitless.
With such motivations in mind, my colleagues and I have developed an approach for modeling the decision making of adversaries. Consider first a view of the proximate issues at the time of a decision. It can be used in group discussions about decision makers, by decision makers themselves, or by analysts reasoning about what opposing leaders are up to.
As mentioned above, potential aggressors attempt to make rational decisions. The approach represents this in a simple but unusual way by having the modeled adversary consider options and examine likely and possible consequences of those options, as suggested in Table G.3.1. The format here is that for each option the reasoner estimates the likely outcome, most favorable outcome, and worst-case outcome. He then makes an overall assessment of the options for action based on these estimates. Each outcome is characterized by one of the values Very Bad, Bad, Marginal, Good, or Very Good.
| Table G.3.1 Generic Decision-Table Format for Assessing Options | ||||
|
Likely |
Most |
Worst-Case |
Assessment |
|
|
Option 1 |
||||
|
Option N |
||||
Table G.3.2 illustrates how a table might be filled in for two different models of the same leader viewing a particular situation (not defined here). In the example the two models see the same facts differently. Model 1 is perhaps more pragmatic, risk-averse, and pragmatically incremental. He chooses the incremental option, which has low risks. Model 2 is perhaps more ambitious, more risk taking, and quite unhappy with the status quo and mere marginal improvements. He chooses the aggressive option despite the substantial risks, primarily because he sees great upside potential and also assesses the likely outcome to be at least Good.
Table G.3.2. Illustrative Judgments for Two Models Considering Options
MOST LIKELY BEST-CASE WORST-CASE NET OUTCOME OUTCOME OUTCOME ASSESSMENT Option Model Model Model Model Model Model Model Model 1 2 1 2 1 2 1 2 Bargain- Marginal Bad Good Marginal Bad Very Marginal Bad ing and Bad compromise Security Bad Bad Good Good Bad Bad Bad Bad threat- ening coercion Limited Bad Bad Good Good Very Very Bad Bad attacks Bad Bad Full- Bad Good Very Very Very Very Bad Good scale Good Good Bad Bad invasion
This simple representation of the decision can be very useful in thinking about someone else's reasoning or one's own reasoning. In its highlighting of likely outcome and both upside opportunities and downside risks, it is a "natural" representation of what we do every day. It is arguably much more natural than expressions in terms of utilities, for example. At the same time, there is much that is implicit, just as there is much implicit when we make our own decisions.
To understand how a potential opponent might reach individual judgments about, for example, the worst-case outcome (would it be Very Bad, Bad, Marginal, Good, or Very Good?), we need:
Developing alternative images is a crucial antidote to the normal focus on so-called best-estimate thinking. To develop alternative "images" of the opponent's reasoning, one can use a combination of essay writing, attribute lists, influence diagrams, and cognitive maps. As in the example of Table G.3.2, in one image the opponent may be pragmatic and incrementalist; in another he may be exceedingly ambitious and frustrated. Perhaps he will also feel cornered, surrounded by enemies, and desperate. These images may incorporate (Davis and Arquilla, 1991a) a variety of well-known psychological phenomena such as those discussed in the literature under "prospect theory," which may encourage greater or lesser risk taking than deemed rational by students of decision analysis.
To illustrate some of these concepts, Figure G.3.1 shows contrasting cognitive maps or influence diagrams used in a study of Saddam Hussein (Davis and Arquilla, 1991b). They represent different images of Saddam's perceptions about the economic situation in mid-1990. Figure G.3.1a represents the cause- effect relationships emphasized in the intelligence community's "best-estimate" understanding of Saddam prior to the invasion. Figure G.3.1b represents an alternative image that could readily have been formulated and disseminated at the time, except for the pressures to focus on a single best estimate. It includes additional factors such as Saddam's perception that his problems were the direct result of Iraq being squeezed deliberately by his enemies (the United States, Kuwait, and Saudi Arabia among them). It also highlights the connection between his economic travails and his grandiose ambitions.
Figure G.3.1 Saddam's image of the 1990 economic situation: two models.
Although nearly all experts would have agreed on the factors in either diagram being "significant," the dominant mental image (see Figure G.3.1a) gave some of the factors little emotional weight. The diagrams highlighted differences of perspective about how Saddam might be viewing the world. We used a number of such diagram pairs in depicting our two images or models of Saddam Hussein. Although we started our work after the invasion and therefore had no trouble constructing a model to explain it, our work proved both insightful and predictive for Saddam's subsequent behavior through February 1991 (i.e., his failure to pull out of Kuwait in the kind of compromise American strategists feared).
As a side note, the decision-modeling approach can be applied not only to crisis decision making but also to peacetime decisions. For example, a recent study (Arquilla and Davis, 1994) applied the methods to understanding the decisions of potential proliferators. Figure G.3.2 is a composite cognitive map developed in that study to indicate the factors potentially affecting the reasoning of states considering development of nuclear weapons. Note that in our work the most important factor is security. Other factors may include the desire to keep superpowers (read "United States") out of the region or the desire to coerce neighboring states. More recently, my colleague Zalmay Khalilzad and I applied the methods to assessing strategies for dealing with North Korea (unpublished).
Figure G.3.2 A generic proliferator's cognitive map.
The next step in the approach is to identify the practical real-world factors that dictate judgments about things such as risks (i.e., in the terms of Table G.3.1, about worst-case outcome). For the case of Saddam Hussein before the invasion decision, the factors affecting perceived risks might have been as indicated in Figure G.3.3.
Figure G.3.3 Possible map of Saddam's assessment of risk before invading.
By merely "eyeballing" Figure G.3.3, one can reason about what judgments Saddam would have made given the information available on the various factors. It seems easy to understand why he considered the risks acceptably low. However, if one of the items is not yet sufficiently explicit (e.g., "warnings from U.S." in the bottom center), the hierarchical decomposition can be continued to greater depth. The warnings from the United States included a range of diplomatic communications of varied "firmness," a very small military exercise in the Gulf, and no preparation for large-scale military operations. On balance, Saddam saw the warnings as unimpressive.
One further item deserves mention. In considering what factors affect decisions, it is important to recognize that "everything" can matter—everything from, say, knowledge about the military balance in heavy armor to whether the target of potential aggression has somehow personally insulted the decision maker or his state. Some factors may be moral or cultural, whereas others may be what I call "dangerous ideas," ideas that have a much greater effect on encouraging military action than they "should" have by virtue of logic and reality. Some of these dangerous ideas include deep-seated hatreds and paranoia, as when the target is felt to be the cause of all sorts of troubles.
The decision-modeling approach can be quite rich, building in highly specific information about a particular leader or group of leaders and about the context in which the decisions are being made. For the purposes of this paper, however, let us instead skip the decision modeling itself and leap to a more abstract representation of what emerges as a view about how to affect a decision about invasion.
Figure G.3.4 provides an overview representation in the form of a "success tree" showing the determinants of a "good" decision (not to invade) as a hierarchy of variables, the highest-level factors of which are as follows:
I have used the success-tree approach (or its cousin, the fault-tree approach) successfully in a number of strategic studies and models over the last decade. One of its primary virtues is that it highlights visually the various components of the problem on which one may wish to focus while developing strategy, i.e., while identifying ways to influence decisions and actions by one's opponent. A second virtue is that it encourages comprehensiveness and integrativeness (although, in practice, something is usually omitted through inadvertence or misjudgment).
The hierarchical structuring is also important because it demonstrates how one can deal with the analyst's chronic nightmare, the curse of dimensionality. The cosmic issues often depend on a vast number of variables, which makes analysis and convergent reasoning very difficult. However, by representing the variables as appearing at different levels of a hierarchy, one is essentially specifying a top-down way to analyze the problem: one starts at a high (relatively abstract) level, works at that level until one runs into difficulties, and then goes to the next level of detail as necessary to understand issues and circumstances. This continues recursively to whatever depth is necessary. Such an approach is useful for simplifying discussion. It is also a natural basis for formal decision modeling.
Figure G.3.4 Success tree for deterrence seen as a decision.
To summarize, there are some useful methods for thinking about deterrence (and other issues such as proliferation) in terms of the decision making of human beings. They encourage us to identify key variables, to order them hierarchically, to identify options, and to assess how, under different mindsets, foreign leaders might evaluate the most likely, best-case, and worst-case outcomes of those options. This can be a systematic way of addressing the issues. Figure G.3.4 is a more abstract representation of the determinants of deterrence, but it can be used as a kind of checklist in thinking about more specific decisions and decision variables. Against this background, let us now turn to what might constitute the principles of good strategy.
If one takes a decision-modeling perspective, what kinds of insights emerge about how weak or medium-strong states might deter strong neighbors? What kinds of advice might be given to states such as Kuwait, Ukraine, Poland, Taiwan, or a unified Korea in the shadow of China? What advice might be given to even weaker states such as Lithuania, Estonia, and Latvia? In what follows I list the insights that appear to be most important. Many of these could have been derived from a more standard political science/international relations perspective, and I make no claims about the results being unique. At the same time, my experience is that by taking a decision-modeling perspective and attempting to be realistic about how real human beings make decisions, one finds oneself taking much more seriously than otherwise a number of "soft" factors that are consistently ignored or brushed aside in most discussions of deterrence. These include factors such as the "feelings" that the nations at issue have toward one another and the particular objectives and values of individual leaders or groups of leaders, which may have little to do with the objectives and values of their publics.
In any case, Figure G.3.4 summarizes high-level variables important in a decision-modeling perspective, and if we start from the left and move rightward, the following insights come to mind. They are provided here in the form of advice to weak or medium-strong states.
Perhaps the most obvious suggestion is to avoid unnnecessarily making an enemy of a strong neighbor and, instead, to take measures to eliminate sources of difficulty. Thus, the suggestions here would be as follows:
The issue of "dangerous ideas" merits special discussion because it is seldom discussed. What I have in mind here ranges from formal religious and otherwise ideological teachings that encourage and perpetuate prejudice and hatred (e.g., teachings about Israel and Jews that can be found in Arab schoolbooks) to misconceptions that are important in particular crises. As an example here, when Yugoslavia began to fall apart, one factor influencing Western Europeans and Americans to stand aside was the widespread and fatalistic notion that the people of the Balkans were backward, tribally oriented, highly disputatious, and still fired with the same ethnic hatreds as in the early part of the century. Civil war might be unfortunate, but it was allegedly inevitable. Would history have been different had Western political leaders and citizenries seen the Balkan people as "real people" who had in fact been living together for many years with substantial suppression of the ancient hatreds? It is impossible to say, but the Western misimpressions were not helpful. They were dangerous ideas that might have been defused by a public relations campaign.
Eliminating sore points is important, but building positive feelings is at least equally so. Interdependence and continual close contact guarantee nothing (in principle they could increase hatreds) but by and large, they help create good relations. The obvious suggestions here are as follows:
Eliminating incentives for aggression and improving relations are fundamental to improving general deterrence (to the point at which aggression is so "unthinkable" that it is not thought of as something to be deterred). For countries that have a choice, however, there is no substitute for defense, because circumstances and intentions change. The suggestions here go beyond the obvious by specifically highlighting the need to avoid "holes" in the defense. Defense should not be thought of as a "political" issue, but rather a military issue. It is not enough to have an army; a nation also needs to have a sensible strategy, properly prepared forces, and preparations that anticipate clever attacks by the adversary. The admonitions, then, are the following:
Encouraging forward deployment may be a bitter pill politically, but there is no adequate substitute for assuring that one's allies will be perceived as committed.
One of the most fruitful classes of measures generically appears to be arms control focused on how forces are located and postured, rather than on their precise size and configuration. Many proposals for arms control can be counterproductive, but others can substantially improve stability. In particular:
More controversial is the idea of nonoffensive defense:
It is difficult to argue from some high moral position that a weak state faced with a large and worrisome neighbor should not have nuclear weapons. Indeed, thoughtful American presidents have for decades chosen to quietly tolerate nuclear activities by Israel. It may have been hypocritical at one level, but the ultimate judgment was sound. The question is how far should this go? Again, what advice would an honest and objective strategist give to a weak, or even a medium-strong, state?
I admit ambivalence and, on bad days, some fatalism about proliferation. Mearsheimer and others arguing the case for the stabilizing role of nuclear weapons have a point (Mearsheimer, 1990). However, the following arguments in the form of advice appear to me persuasive on balance:
None of these arguments is ultimately compelling, but they seem persuasive in most cases of practical interest. Israel still appears to be the obvious exception, primarily because it is so small and its neighbors remain strongly and implacably hostile, despite the continuing peace process, which may change this in time. In a situation where religious or ethnic-hatred issues reign, we should not expect the normal rules of conventional deterrence to apply readily. Ideologues are willing to take greater risks, greater casualties, and even losses in pursuit of their goals.
Let us next turn to what major states can do to extend deterrence to weak or medium-strong states. The challenges are great, but there are nonetheless some principles.
The recurring problem here has been that nations have been ambivalent in peacetime about whether to get involved in events elsewhere, especially in the absence of an immediate threat, and especially when "getting involved" could antagonize another nation with which better rather than poorer relations are desired. This was the problem with the United States deterring Iraqi invasion. We see the same kinds of issues arising today in debates about NATO expansion. NATO expansion could, on the one hand, fill vacuums and establish the interests of the West in the continued security of various eastern and Central European states. On the other hand, it could antagonize Russia and provide fuel for the dangerous Russian nationalist movement.
Interestingly, even the "aggressive" proponents of NATO expansion have so far limited their goals to countries such as Poland. But what about Ukraine and the Baltic states? On the one hand, it is difficult to imagine NATO defending these states in a traditional manner. It is also clear that NATO's interests in Ukraine and the Baltic states are less than "vital." I would argue, however, that aggression against either of them would be altogether unacceptable in the modern world and that their security is very much a matter of NATO interest. Strategy, then, would include expressing those interests frequently.
If there is a single fatal flaw in extended deterrent strategies based on decisive military moves in a crisis, it is that democracies have a great deal of trouble being decisive in ambiguous circumstances. Even "obviously" prudent military measures such as prepositioning military forces in the region and enhancing states of readiness for deployment are often politically difficult because of concerns about provocation or escalation of tensions.
Such difficulties could perhaps be greatly mitigated by facing up to them in peacetime and developing much of the necessary political consensus, both domestically and politically, by including appropriate people in seriously conducted crisis games. Then, upon receiving strategic warning of a real crisis, the key people (including legislators and major allied leaders) could be brought into such gaming early so that they could themselves work through the logic for acting rather than dissembling. If this were successful, leaders such as the U.S. president could take appropriate hedging measures without being savagely attacked on the political front.
Many of those who would support early intervention to deter invasion of weak states or debacles such as in the Balkans tend to assume that a clear show of force would suffice. Often, however, their "clear show of force" would be long on show and short on capability. This is inherently dangerous when the object of attention is a strong and aggressive personality willing to take risks. Such figures tend to be impressed by power, not empty threats. And, indeed, shows of strength by Western European nations or NATO might well be empty because there might not be the political support for going further. An important distinction here is shows of force that do and do not put that force in harm's way, with the latter being far more effective than the former because they reduce the room for dithering if war begins. To put it differently, there is still a role for trip wires. However, when dealing with risk-taking aggressors, wise leaders will not deploy trip wires without starting the process of providing massive followup. Military conservatism on this score is well justified.
Continuing the theme of the importance of communicating credibly the willingness to fight, it seems important to increase rather than decrease forward deployments, preferably in forms that cannot be readily bypassed.
Important alternatives to permanent stationing of trip wires include (1) prepositioning equipment in the country at issue to permit rapid reinforcement in crisis; (2) creating other infrastructure to facilitate rapid reinforcement; (3) conducting frequent joint exercises in the country to remind everyone of security ties, even if informal; and (4) maintaining naval and air forces in the region.
If we turn from abstractions to specifics, considering the real or virtual threat to a particular weak or middle-strength state, it is usually the case that quick substantial enhancements of defense capability are possible if merely the right basis is laid in advance. This, however, may involve extensive coordination in the realm of command and control, logistics, and combined operations. Further, it may involve deploying tailored capabilities, some of them in short supply, rather than mere masses of equipment. Often, "smart" intervention is likely to mean providing air forces with precision-strike capability and superb theater-level reconnaissance and intelligence capabilities, along with the necessary command and control to exploit it.
Another form of "smart" intervention might be to supplement the defender's forces with high-quality indirect fire weapons that would greatly increase the vulnerability of attacker tanks and permit a kind of defense in depth (see also Kelley, Fox, and Wilson, 1994).
An important element of extended deterrence is avoiding self-deterrence, as well as coercion of regional allies. The problem, again, is weapons of mass destruction (WMD). Although defense against the WMD threat is essential, the preferred strategy here is to deter use of WMD, essentially by credibly threatening massive response (see, e.g., Gompert, Watman, and Wilkening, 1995). With today's precision weapons, such a response could be conventional. Further, it could have a "countervalue" or "counterforce" character, depending on needs. Countervalue attacks could be quite discriminating.
Here there is a complete commonality with the advice offered to countries concerned about direct deterrence. Operational arms control in particular (e.g., limits on the deployment locations and states of readiness) of forces can drastically alter the quality of strategic warning, even to the point of making justified preemptive attacks plausible. This, in turn, is becoming increasingly important as the result of the proliferation of weapons of mass destruction and missiles: were the United States to intervene in a regional crisis 5 or 10 years from now, there might be a high premium on early and decisive counterforce attacks on the aggressor's means for delivering nuclear, chemical, and biological weapons.
WMD issues are becoming so important that it seems clear that defenses against WMD are now essential. In this context defense includes counterforce, post-boost intercept, terminal intercept, and passive measures such as dispersal and hardening. Without such defenses, the option for intervention and, therefore, the credibility of extended deterrence may be severely undercut.
With very few exceptions, it seems exceedingly unlikely that the United States or its allies will be willing to intervene in regional conflicts without clear legitimacy in the international community. Unfortunately, the United Nations currently is incompetent in dealing with military crises, especially when competence includes speediness and decisiveness in circumstances of ambiguity. Further, the prospect of depending on U.N. military operations, as distinct from U.N.-sanctioned operations led by the United States or some other major power, should be sobering for anyone thinking about the challenges of successful immediate deterrence. The major nations need to develop alternative ways of legitimizing and conducting the necessary actions. Ideally, this would mean changes within the U.N. structure and decision making, but that may not prove feasible.
A key element of deterrence planning should be recognition that immediate deterrence, however important, is a slender reed on which to base security. Immediate deterrence has failed too many times in the past, and the reasons for it having failed are still salient. It follows that in addition to plans for military and other measures in a crisis, an overall strategy of extended deterrence should:
Because immediate deterrence may fail, especially with respect to attacks on weak or medium-strong states, defense of which does not represent vital interests of potential protectors, the United States and the civilized and forward-looking world community as a whole should worry more about developing and advertising credible options for severely punishing aggressor states—not just in the immediate aftermath of an attack, but for many years thereafter. Perhaps the metaphor should be of "putting aggressor states in jail" for terms of, say, 5 to 10 years. In other instances, an appropriate response might include military attacks to destroy substantial portions of the aggressor's military forces or infrastructure (e.g., its navy) or appropriate elements of the civilian value structure, all with conventional weapons. With sufficiently high accuracy and targeting, such attacks could be relatively discriminative. The attacks could be one-time events, "punishment," but not the start of a continuing war. There need be no quagmire.
The principal issue here is that of credibility. Would the world community or leader states punish militarily a successful aggressor that also possessed nuclear capabilities and the means to delivery nuclear weapons against their own countries (either by missiles or by terrorists smuggling devices into them)? The initial reaction of many observers is "decidedly not," unless there were vital interests at stake. Although the argument is plain enough, its implications seem puzzling in instances in which the potential punisher states have escalation dominance in every dimension and the aggressor is rational, however unpleasant. Certainly, military punishment options would be risky, but the long-run stakes could be high.
The argument is unlikely to be resolved, but a few observations appear to be objectively valid. In particular, general deterrence by threat of punishment options could be much enhanced by (1) missile defenses; (2) well-exercised and advertised military options for selective but severe punishment, coupled into long-term isolation activities politically and militarily; and (3) pooling of risk by international cooperation (e.g., a punishment option by NATO might be better than a punishment option only by the United States).
This enhancement of general deterrence seems to be a good investment. Enhancement of immediate deterrence through threat of punishment will be a risky proposition, but competition in risk taking is hardly a new issue.
What, then, can be said in summary about deterrence in defense of weak states, especially when one takes the perspective that deterrence is ultimately about influencing decisions? The principal conclusions of this paper are as follows:
Although nuclear capability could enhance deterrence, it is also likely in most instances to excacerbate tensions and assure that careful military plans will be laid for attack. Nuclear capabilities are likely to be vulnerable and therefore might be destabilizing in a crisis.
Allan, Charles T. (1994), "Post-Cold War Deterrence", a research survey, Washington Quarterly, Vol. 17, No. 3 (Summer).
Arquilla, John (1992), Dubious Battles: Aggression, Defeat, and the International System, Crane-Russak, Washington, D.C.
Arquilla, John (1995), "Bound to Fail: Regional Deterrence After the Cold War," Comparative Strategy, Vol. 14, No. 2 (April-June).
Arquilla, John, and Paul K. Davis (1994), Modeling Decisionmaking of Potential Proliferators as Part of Developing Counterproliferation Strategies, MR-467, Rand, Santa Monica, California.
Asmus, Ronald, Richard Kugler, and F. Stephen Larabee (1995), NATO Expansion: The Next Steps, Rand, Santa Monica, California.
Bennett, Bruce, Sam Gardiner, and Daniel Fox (1994), "Not Merely Preparing for the Last War," Chapter 15 of Davis (1994b).
Blainey, Geoffrey (1973), The Causes of War, The Free Press, Macmillan Publishing Co., New York.
Cimbala, Stephen (1992), Force and Diplomacy in the Future, Praeger, New York.
Cole, Paul (1994), The Conduct of a Nuclear-Capable Nation Without Nuclear Weapons: Sweden Without the Bomb, MR-460, Rand, Santa Monica, California.
Davis, Paul K. (1987), "A New Analytic Technique for the Study of Deterrence, Escalation Control, and War Termination," in Stephen Cimbala (ed.), Artificial Intelligence and National Security, Lexington Books, D.C. Heath and Co., Lexington, Massachusetts.
Davis, Paul K. (1988), Toward a Conceptual Framework for Operational Arms Control in Europe's Central Region, R-3704-USDP, Rand, Santa Monica, California.
Davis, Paul K. (1992), "Behavioral Factors in Terminating Superpower War," in Stephen Cimbala and Sidney R. Waldman (eds.), Controlling and Ending Conflict: Issues Before and After the Cold War, Greenwood Press, New York.
Davis, Paul K. (1994a), "Improving Deterrence in the Post Cold-War Era: Some Theory and Implications for Defense Planning," Chapter 8 in Davis (1994b).
Davis, Paul K. (ed.) (1994b), New Challenges in Defense Planning: Rethinking How Much Is Enough, Rand, Santa Monica, California.
Davis, Paul K. (1994c), "Protecting the Great Transition," Chapter 6 in Davis (1994b).
Davis, Paul K. (1994d), "Planning for Adaptiveness," Chapter 4 in Davis (1994b).
Davis, Paul K. and John Arquilla (1991a), Thinking About Opponent Behavior in Crisis and Conflict: A Generic Model for Analysis and Group Discussion, N-3322-JS, Rand, Santa Monica, California.
Davis, Paul K. and John Arquilla (1991b), Deterring and Coercing Opponents in Crisis: Lessons from the War with Saddam Hussein, R-4111-JS, Rand, Santa Monica, California.
Davis, Paul K. and Lou Finch (1993), Defense Planning in the Post Cold-War Era: Giving Meaning to Flexibility, Adaptiveness, and Robustness of Capability, Rand, Santa Monica, California.
Doyle, Michael (1995), Correspondence, International Security, Vol. 19, No. 4 (Spring).
Dror, Yehezkel (1971), Crazy States: A Counterconventional Strategic Problem, Heath Lexington Books, D.C. Heath and Co., Lexington, Massachusetts.
Gompert, David (1994), "How to Defeat Serbia", Foreign Affairs, Vol. 73, No. 1.
Gompert, David, Kenneth Watman, and Dean Wilkening (1995), U.S. Nuclear Declaratory Policy, MR-596, Rand, Santa Monica, California.
Helling, Klaus and Björn Niemeyer (1995), Risikoverhalten in Krisensituationen (Risk-Taking Behaviour in Crisis Situations), Institute für Angewandte Systemforschung und Operations Research, Universität der Bundeswehr, Munich, Germany.
Howard, Michael (1984), The Causes of Wars, Harvard University Press, Cambridge, Massachusetts.
Huber, Reiner K. (ed.) (1990), Military Stability: Prerequisites and Analytic Requirements for Conventional Stability in Europe, NOMOS Verlagsgesellschaft, Baden-Baden.
Huber, Reiner K. and Rudolf Avenhaus (eds.) (1993), International Stability in a Multipolar World, NOMOS Verlagsgesellschaft, Baden-Baden.
Huber, Reiner (1994), "Deterrence in Bosnia-Herzegovina: A Retrospective Analysis of Military Requirements Before the War," European Security, Vol. 3, No. 3 (Autumn).
Huth, Paul (1988), Extended Deterrence and the Prevention of War, Yale University Press, New Haven, Connecticut.
Huth, Paul and Bruce Russett (1988), "Deterrence Failure and Crisis Escalation," International Studies Quarterly, Vol. 32, pp. 29-45.
Jablonsky, David (1991), Strategic Rationality Is Not Enough: Hitler and the Concept of Crazy States, Strategic Studies Institute, U.S. Army War College, Carlise Barracks, Pennsylvania.
Kelley, Charles, Daniel Fox, and Barry Wilson (1994), "A First Look at Defense Options for Poland," in Davis (1994b).
Kugler, Richard (1995), Toward a Dangerous World: U.S. National Security Strategy for the Coming Turbulence, Rand, Santa Monica, California.
Layne, Christopher (1994), "Kant or Cant: The Myth of Democratic Peace," International Security, Vol. 19, No. 2 (Fall).
Mearsheimer, John (1990), "Back to the Future: Instability in Europe after the Cold War," International Security, Vol. 15, No. 1.
Mearsheimer, John (1983), Conventional Deterrence, Cornell University Press, Ithaca, New York.
Møller, Bjørn (1995), "Common Security and Non-Offensive Defense: Are They Relevant for the Korean Peninsula?" presented at the International Conference of the Korean International Studies, June 16-17, Seoul. See also his book Common Security and Non-Offensive Defence, a Neorealist Perspective, Lynne Rienner, Boulder Colorado, 1992.
NATO (1995), "Symposium on Coping with Uncertainty in Defense Decision Making," The Hague, January 16-19, 1995, Technical Proceedings, two volumes, Panel 7 on the Defense Applications of Operational Research, NATO Defense Research Group.
Olmstead, Thomas (1994), "Psychology and the CIA: Leaders on the Couch," Foreign Policy, No. 95 (Summer).
Ronfeldt, David (1994), Beware the Hubris-Nemesis Complex: A Concept for Leadership Analysis, Rand, Santa Monica, California.
Russett, Bruce (1995), "And Yet It Moves," a correspondence item, International Security, Vol. 19, No. 4 (Spring).
Spiro, David (1995), International Security, Vol. 19, No. 4 (Spring).
Watman, Kenneth and Dean Wilkening, with John Arquilla and Brian Nichiporuk (1995), U.S. Regional Deterrence Strategies, Rand, Santa Monica, California.
Wilkening, Dean and Kenneth Watman (1995), Deterring Nuclear Threats from Regional Adversaries, Rand, Santa Monica, California.
Wolf, Barry (1991), When the Weak Attack the Strong: Failures of Deterrence, N-261-A, Rand, Santa Monica, California.
Woodward, Bob (1991), The Commanders, Simon and Schuster, New York.
Zimmerman, Warren (1995), "Origins of a Catastrophe: Memoirs of the Last American Ambassador to Yugoslavia," Foreign Affairs, (March/April).
[NAP Home Page] — [Report Home Page] — [NSB Homepage]
[Table of Contents] — [Next Section]
Copyright 1997 National Academy Press