U.S. Conventional Prompt Global Strike: Issues for 2008 and Beyond (2008)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

3 Political, International, Policy, and Doctrinal Issues The basic policy conclusion of the Committee on Conventional Prompt Global Strike Capability, based on its collective knowledge and judgment, is that there would be important political and strategic advantages to the United States in being able to strike high-value targets having time-sensitive urgency that could not be effectively engaged by currently available conventional strike systems. Existing conventional capabilities are extensive (and should be improved, where possible, to enhance capabilities in the interim before any conventional prompt global strike [CPGS] system can be operational), and no CPGS system could fill all the gaps in the U.S. ability to strike potentially important targets. However, CPGS would fill some gaps and add more options for dealing with threats to the United States and to its allies and to others whose interests are important to its own. Having those options available could be valuable in the campaigns against terrorism and against the proliferation of weapons of mass destruction (WMD) capabilities, as well as for larger-scale military efforts. (The nature of these targets and the proposed capabilities and limitations of the Conventional Trident Modification [CTM] and various other CPGS concepts are addressed in detail in Chapter 2.) CPGS systems, however, present challenges and questions as well as opportu- nities. The essential policy judgment that must be made in deciding on acquiring a system—and on the doctrine and principles for its potential use—is whether the additional capabilities that CPGS would provide outweigh the costs and concerns that CPGS may present. The answer may depend on the type of CPGS system under consideration. On the one hand, for CTM, some challenges are smaller than would be the case with other CPGS systems because of the limited scale of the CTM program, CTM’s relatively low cost and technological risk, and the opportunity to take advantage of the mature Trident system. On the other hand, 61 

62 U.S. Conventional Prompt Global Strike some alternative concepts could be designed and deployed—with years of delay, perhaps many years—so as to mitigate some concerns that CTM would raise simply because it does build so directly on an existing strategic-nuclear system and has some limitations on its effectiveness. The need to consider the full range of issues presented by CPGS concepts With the introduction of any significant new military capability, the doctrine for use of the system must be clearly defined and agreed on by policy makers and military commanders and planners. This is especially the case with a system such as CTM (or any other CPGS system): these systems represent a unique military capability, and their use (and perhaps even possession) has implications across the spectrum of diplomatic, policy, and military considerations that go far beyond even the complex and important technical and operational military questions presented. Setting the frames of reference for thinking through and planning for these issues is a critical step in deciding whether to develop and acquire the capa- bility—and in preparing for its wise and effective addition to the nation’s instru- ments of military power. Developing doctrine is an essential part of acquiring a CPGS capability, and it requires thinking through not just potential scenarios for crisis and conflict in which CPGS might be used, but also all of the implications of possessing such a capability. The Primary Issues The primary issues that have been raised with respect to CPGS are these: (1) some CPGS systems might raise ambiguity problems because of using stra- tegic-nuclear delivery systems (notably, the Trident submarine-launched ballistic missile [SLBM]); (2) the advertised promptness of the CPGS capability will not be achievable for years because it requires exquisite intelligence, detailed surveil- lance, rapid decision making, and so on; and (3) the CTM concept would degrade the quality and certainty of the nuclear deterrent force. The committee analyzed all of these issues in considerable detail. Additional matters addressed in this chapter include command and control and the requirement for presidential authorization, the potential for inappropriate or mistaken use, the prevention of an accidental launch of a nuclear weapon when a conventional strike has been ordered, over- flight and debris, access to forward basing, proliferation, arms control and treaty issues, and strategic considerations. The consensus of the committee that emerged after study was that (1) the ambiguity problems have been overstated and can be substantially mitigated; (2) there are likely to be cases where intelligence could be available and decisions could be made promptly enough to make a short execution time highly valuable, even though minimizing other parts of the end-to-end planning time will take 

POLITICAL, INTERNATIONAL, POLICY, AND DOCTRINAL ISSUES 63 years to achieve (as discussed in Chapter 2 and not repeated in this chapter); and (3) any negative effects on the nuclear deterrent force would be quite small. The remainder of the chapter addresses the broader range of issues systemati- cally, but the complexity of that discussion should not obscure these bottom-line conclusions. A Broad View of the Issues to Be Considered Decisions on CPGS entail consideration not only of technological possi- bilities and military missions, but also of the challenges and opportunities that having (or lacking) such a system would present for national policy and national interests. Moreover, in deciding whether to acquire such a system, it is essential to take into account the doctrine and procedures that would be developed and applied to determine how and when the options afforded by CPGS should be used (and withheld). The possession of CTM, or of any other CPGS system, would add to the nation’s options—that is, to what the military is technically capable of doing. However, it is essential to have for CPGS, or any other system, a structure for operating it and making judgments about what it is to do—and consideration of those issues must be part of the decision on whether to acquire CPGS. The United States has for many years had long-range strike capability, which it has also exercised, as illustrated by the 1986 strike on Libya and antiterrorist-related strikes into Afghanistan and Sudan in 1998. Both involved aircraft. Strikes might be conducted tomorrow with such systems if the location of terrorist leaders were known. It might seem, then, that most of the policy-level issues have long since been resolved and precedents established. Arguably, however, conventional prompt global strike raises some new issues. The specific questions that need to be considered in developing policy and procedures in the CPGS context include the following: • What are classes of high-value targets that the United States may want to strike? • What intelligence will be needed, and how fast can it be provided and exploited? • What information can be gathered and processed in advance (e.g., early warning to focus preparations, precise geolocation of plausible targets, and pre- liminary or contingent decision making)? • What is the decision-making process? What is the chain of command? • What will have to be done to plan the mission? What preplanning can be done in advance of facing a decision on actual execution? When, for example, can a contingent decision be made—to use the system if certain information is avail- able about an identified type of target, subject only to a rapid, final confirmation by the President? 

64 U.S. Conventional Prompt Global Strike • Given sufficient warning-and-preparation time, how effective and reliable will the CPGS be? • Apart from the nature of the target, what information will be required as essential elements for making a decision—for example, probability of success (taking into account possible countermeasures and weapon system constraints), likely collateral damage (including the possibility of fallout from destruction of a WMD facility), overflight and debris impact, subsequent opportunities for attack, time to execute the strike after the order is given, and so on? • What are the alternative ways of attacking the target? • What consultations or notifications should be made in the United States (e.g., Congress) and in other countries? • What are the possible and likely reactions of the target country and other countries? What is the next move for the United States in the aftermath of success or of failure? • Is there any chance that an attack with a CPGS system will be misinter- preted as a nuclear attack, particularly in the midst of a crisis? • What are the longer-term reactions of other countries to the existence of the capability (and the use) of a CPGS system? What would be the effects on proliferation? • What are the vulnerabilities and limitations of the system to be deployed, and what should be done to correct or mitigate them? These issues are not fundamentally different from those that would arise with the use of any military force, but CPGS gives the capability to act very quickly, and therefore it is all the more critical that these questions—and others—be thought through and prepared for in advance. The alternative is errors arising from failure to consider all the relevant issues, or decision-making and process- ing delays that lose the advantage of having the technical capability to act with sufficient speed. Faster execution time will by definition make it possible to strike faster. Whether this gain will be critical is entirely dependent on the particular situation and its circumstances. However, the interval from launch to impact on the target is far from the only factor in determining whether the United States has a capa- bility for prompt global strike. Promptness of strike is measured not only by the time interval between launch and impact but by the whole end-to-end time line. There must still be a careful and informed decision to use the capability, and the attack must be properly planned. Intelligence must be gathered, distributed, and evaluated; decision makers must be informed of the intelligence and the options and consider the pros and cons of ordering an attack; detailed operational plans, including precise geolocation of the target, must be prepared, approved, and transmitted; and necessary strike preparations, including bringing weapons to ready-to-launch status, must be made. Some of these necessary antecedent steps can be done in parallel and some 

POLITICAL, INTERNATIONAL, POLICY, AND DOCTRINAL ISSUES 65 can be done well in advance. This is particularly true if the need or opportunity arises out of an ongoing confrontation in which contingent decisions can be made to use (or not to use) the CPGS capability if the opportunity presents itself, subject only to a final confirmation by the President. However, it will always be the case that these steps take time and that they require enabling capabilities, for example, for intelligence gathering, target location, and mission planning. If the United States is going to deploy a CPGS system, all of these issues must be addressed. This is so because being able to gain the full benefit of an invest- ment in a prompt execution capability requires both that enablers supporting that capability be available and that steps be taken to ensure that each element in the end-to-end time line—not just the time from launch to impact—is made as short as it can be. If it is valuable to invest in a system that has the advantage of a very short time of flight, then it is also valuable to minimize the time required to use the system effectively, bearing in mind that there must be time for deliberation and for ensuring that the strike is properly planned and that hair-trigger decisions are likely to lead to errors. Moreover, shortening the execution time once an order reaches the launcher does not necessarily make the overall time to decide on and plan an attack shorter; nor should it always do so. There may be cases in which the principal benefit of a shorter execution time is that it gives more time for other steps in the process to be carried out properly, while still not missing the opportunity if the decision is made to strike. Planning for making effective and prudent use of the execution capability also requires addressing doctrinal issues and decision procedures as well as technical parameters. Accordingly, a key part of the decision to proceed with CTM, or any other CPGS system, should be a commitment to a major comprehensive effort to develop theory, doctrine, and commonly accepted understandings (analogous to that for nuclear theory, including extended deterrence) as well as to taking other steps necessary to minimize the total end-to-end time line. Command and control and the requirement for presidential authorization Although CTM’s absolute firepower is very limited compared with that of many other modern military systems, its use would almost certainly occur in highly sensitive political contexts, because of either the nature or the location of the target. Almost by definition, potential targets of CTM would be highly impor- tant to national interests (e.g., terrorist leaders, WMD supplies or facilities), but executing the strikes could have ramifications well beyond the immediate opera- tion. Opportunities to use CPGS would often raise policy, legal, and diplomatic issues regarding the use of American military force against targets located in the sovereign territory of nations where the United States was not simultaneously conducting other military operations. These issues have arisen often in the past, 

66 U.S. Conventional Prompt Global Strike but the fact that CPGS systems offer the opportunity to reach such targets rapidly will not reduce the significance of the issues. CPGS would in some situations provide a military option that otherwise would not exist, but the availability of that option should not blind decision makers to the existence of alternatives, both military and otherwise. These aspects of an attack using a CPGS system mean, in the committee’s judgment, that a basic element of the doctrine for CTM, or any other CPGS sys- tem, is that the authority for the execution of these very limited, high-value strike options should rest exclusively with the national command authority, that is, the President acting through the Secretary of Defense in transmitting orders to the military, even in the midst of a conventional conflict (such as in Iraq or Afghani- stan currently). Such a reservation of execution authority provides maximum assurance that all of the relevant political and strategic considerations relating to the strike being proposed (e.g., ambiguity, likely immediate and longer-term responses, collateral damage, consultations, alternative strike options, and so on) are taken into account. Presidential authority could, in appropriate circumstances, include instructions to the relevant military commander to employ the system if conditions defined by the President were met. This “escape clause” should be used with extreme caution, however, and except in circumstances that are both truly extraordinary and very clear-cut, should always be subject to last-minute, prearranged confirmation by the President. In setting up a system for the authorization of strikes using CTM, it will be essential to establish a clear chain of command below the President. Presumably, the relevant regional combatant commander would have operational control of strikes against targets in his area of responsibility (AOR), as is the case today when long-range strike aircraft based outside a regional commander’s AOR execute strikes at targets in “his” region. However, whatever the details of the chain of command, CPGS systems (like nuclear weapons) should not be consid- ered simply new capabilities to be available to field commanders to employ on their own authority. Potential for Inappropriate or Mistaken Use A concern has sometimes been raised that CPGS capability should not be acquired because it could be misused. More specifically, those who hold this view argue that CPGS would give the President a non-nuclear option that might be particularly susceptible to misuse, because decisions to use the system could seem “too easy”: that is, the capability would facilitate military attacks deep in the territory of other countries, not just without using nuclear weapons but without deep and extensive commitment of other elements of the American military and other resources, and with little immediate risk to American personnel. 

POLITICAL, INTERNATIONAL, POLICY, AND DOCTRINAL ISSUES 67 The defect in this argument is that it applies too broadly. Any military system could theoretically be used incautiously, but the United States cannot, for reasons that go far beyond CPGS, make defense-acquisition decisions on the presump- tion that presidents will act foolishly. Developing a sound procedure for decision making will help ensure that the mere capability for very rapid action does not become a pretext for ill-considered strikes. The potential for inappropriate deci- sions—which exists with any military capability—should be dealt with by care- ful planning and procedures for using a new system, not by denying the nation the possibility of using the military capability wisely. The committee believes that in the emerging doctrine, the use of CPGS systems should be so much an exceptional, rather than a routine, operation as to make that casual, ill-considered decision making very unlikely. Moreover, it is untenable to say that the country should forgo having the means to strike effectively against targets threatening its national interests. There is some chance that capabilities could be used unwisely; the relevant question is how to minimize the risk of “unwise use.” There is a somewhat separate concern that the fleeting nature of the target will produce pressure for rapid decisions and result in ill-considered action that might be avoided if more time had been taken to reach a decision. However, the pressure for very rapid decision on CPGS operations would be driven by the nature of the target, not the speed with which the attack, once ordered, can be carried out. Potential CPGS targets are of sufficient importance that if CPGS were unavailable, those targets might well be struck, albeit less effectively, with other conventional systems. In that case the time for decision would be even more limited than with CPGS, and the effectiveness would, almost by definition, be less. Indeed, there are cases in which the greater speed of execution would allow more time for deliberation and reduce the perceived need for early delegation of authority to local or even tactical commanders. The committee’s judgment is not that these arguments concerning misuse or error are entirely irrelevant or absolutely wrong. The fact that a CPGS system could by its inherent nature be used rapidly, simply makes it all the more important that the decision to use it be made carefully including, where possible, deciding in advance for (or against) its use in the event that an opportunity arises. The committee believes that the increased possibility of successful uses of CPGS in cases where that is the right decision justify having the capability and outweigh the theoretical possibility of inappropriate uses. Moreover, while it is obviously important to do all that can reasonably be done to avoid mistakes, the fact that CTM attacks would necessarily involve limited numbers and limited effects (compared with options involving existing conventional forces that might require defense suppression, a larger attacking force, and risk to personnel, not to mention comparison with nuclear attacks) would reduce but not eliminate the potential ramifications of mistakes. 

68 U.S. Conventional Prompt Global Strike Impact on nuclear deterrence and stability At a minimum, it must be a basic premise of CTM, or any other CPGS system, that it should not weaken nuclear deterrence. In particular, introducing CTM as a conventional system on the Trident nuclear-powered ballistic missile submarines (SSBNs) should not be permitted to interfere with core nuclear deterrence or with the readiness, deployments, training, discipline, day-to-day operations, research, and in general the focus and resources necessary to maintain that core deterrence. The concern is sometimes expressed that because CTM adds a conventional mis- sion to a platform and support system hitherto exclusively devoted to the nuclear deterrence mission, it will risk a lack of focus and diversion of attention and priority (including funding for technological work needed to keep the force fully capable of meeting its mission) from strictly nuclear issues. As for a potential lack of sufficient weapons on SSBNs in appropriate patrol areas to maintain necessary levels of available forces to support nuclear war plans, the very modest reduction in on-station nuclear loads and the adjustment of patrol areas necessary for the CTM mission would not compromise the nuclear mission. (Because the United States has reduced the number of reentry vehicles [RVs] on Tridents to below their design potential, it is the committee’s judgment that the off-loads for the two CTM missiles per on-patrol SSBN could be offset by increas- ing the number of RVs on other missiles on the same submarines.) For reasons explained below, the dangers flowing from a possible identification of the location of the submarine resulting from the launch of a CTM are minimal. Two concerns that have been expressed are that the patrol areas would no lon- ger be optimized for survival and that the launch of a conventional Trident missile would reveal the SSBN’s location, thereby endangering its survivability for the nuclear mission. Although it is true that the patrol areas would not be “optimal,” they would still be extremely large and beyond what is considered needed by any current analysis. No other nation possesses the assets (nuclear-powered attack submarines, patrol aircraft, and so on) to conduct large-area search-and-destroy operations against U.S. SSBNs. As for launch revealing location, that problem has been studied and dealt with over the years. The problem would be no worse than with a nuclear warhead, and the Navy long ago developed techniques to pro- tect SSBNs after missile launch. These techniques include deceptive trajectories to complicate backtracking, and rapid SSBN movement from the launch point. Within a short period of time, an SSBN’s location would be unknown within an area large enough to deny plausible effective attack. Finally, it should be noted that in the event that circumstances changed, with the emergence of a serious threat to U.S. deterrent forces, SSBNs could readily revert to patrol areas and practices fully optimal for nuclear deterrence. With respect to the “softer” elements of maintaining the nuclear mission, the committee concludes that the Navy has developed the plan for CTM in a way that will hold the conventional role to the same high standards as for nuclear missions 

POLITICAL, INTERNATIONAL, POLICY, AND DOCTRINAL ISSUES 69 and will maintain attention and priority on the nuclear mission. The very fact that—unlike other systems with a potential nuclear but a primary conventional role—CTM would necessarily be employed only infrequently and on very special missions means that there is little potential for a routinized conventional mission to drive the nuclear role into the background; there is little risk that nuclear mis- sions would come to be seen as secondary to the “real” conventional mission, with an attendant reduction in attention to all the no-doubt tiresome and petty-seeming rules and procedures that (properly) go with nuclear weapons. Some argue that having an extensive CPGS capability would make it possible for the United States to declare a policy of no-first-use of nuclear weapons—that is, limiting U.S. use of nuclear weapons to a response to an actual nuclear attack. Apart from the more general arguments about whether a policy of no-first-use is wise, it is difficult to believe that the capacity for a very limited number of quick, highly accurate conventional attacks is significant to the debate. For example, an effective non-nuclear response to a massive chemical or biological attack, what- ever its wisdom or feasibility, would hardly rely greatly on the capacity of CPGS systems for rapid and precise attacks. Rather, the response would rely primarily on overwhelming force. CPGS is no easy—or tempting—road to no first use. From a very different perspective, it is sometimes argued that a reliance on CPGS weakens the deterrence of WMD attack (or other equally grave provoca- tions short of a nuclear attack), because investment in CTM would imply that the United States does not intend to use nuclear weapons in such circumstances: that is, that CTM betrays a lack of will in a context in which the nation would very much want adversaries to believe that it would use nuclear weapons. This concern seems overblown. As noted above, neither CTM (nor even other, more extensive CPGS systems) would likely be critical one way or the other to a non-nuclear response to a major WMD attack. More important, having a conventional option for certain limited missions does not in any way reduce the credibility of threats to respond with nuclear weapons to grave provocations. Quite apart from relatively technical concerns with ensuring that CTM does not reduce the effectiveness of the nuclear deterrent (and the important “ambigu- ity” question, discussed below), issues have been raised concerning the potential impact of CTM on nuclear stability. It could be argued that CTM or other CPGS systems could make nuclear war more likely because they would enhance U.S. capabilities in ways that render more likely U.S. attacks on high-value targets in nuclear-armed countries, to which those countries might respond by using nuclear weapons. Of course, any attack on a nuclear-armed country is a very serious mat- ter, simply because of the possibility of escalation, but the probability of a nuclear response to such a conventional attack is surely lower than the probability of a nuclear response to a nuclear attack. CTM and other CPGS systems are sometimes argued to have a more general effect of raising the nuclear threshold by allowing conventional attack where nuclear weapons would otherwise have been used. The degree to which the avail- 

70 U.S. Conventional Prompt Global Strike ability of an additional conventional option raises the nuclear threshold for the United States depends on the degree to which a strike with CTM would actually substitute for a nuclear strike, as contrasted to (merely!) giving the President an effective option in place of a nuclear alternative or inaction in a situation where there would otherwise have been no effective U.S. action. Opinions differ on how often, if ever, a U.S. President would actually use a nuclear weapon in the sort of scenarios for which a CPGS system would be designed. It is certainly possible to conceive of cases where a threat is so great that there is a very real possibility that the United States would use a nuclear weapon (e.g., in a case of absolute confirmation of an impending nuclear attack on a U.S. city), and a CPGS capability might allow the President to substitute its use for a nuclear strike and thereby act effectively while avoiding a nuclear strike. In this very limited sense, CPGS would raise the nuclear threshold. (There is a separate question about whether in a case where the President would actually have ordered a nuclear attack, any CPGS option would appear sufficiently effective—and dra- matic—to be an adequate substitute.) More important and certainly more likely, CPGS would be a valuable new instrument of national policy because in extremely serious cases it would avoid the dilemma of having to choose between using a nuclear weapon or making no response at all. The committee believes that in the face of such a dilemma, the disadvantages of using nuclear weapons are such that there are very few cases in which the President would actually choose to use a nuclear weapon. But without CTM or some other CPGS system, the lack of an effective and sufficiently prompt non-nuclear response and the appropriately extreme reluctance to use nuclear weapons for any but extraordinarily grave threats would leave the United States in the position of being unable to take any effective military action. On balance, it is the view of the committee that CPGS would have little effect on nuclear stability or deterrence because in most cases it is essentially an improvement—perhaps a decisive improvement—over existing but less satis- factory conventional options (or for no response at all). In virtually all cases in which CPGS might be used, even in those cases where other conventional options would be ineffective, there is little, if any, real possibility of the United States actually using a nuclear weapon. However, providing an option to fill part of the gap between what current conventional systems can do and the point at which a nuclear attack would be a wise action is an important advantage of CTM, whether that gap is large or small. Preventing AN accidental launch of a nuclear weapon when a conventional strike has been ordered Any CPGS concept must ensure that there is no possibility that an order to carry out an attack with the CPGS system would accidentally result in a nuclear attack, or that such an order would accidentally result in the launch of a missile 

POLITICAL, INTERNATIONAL, POLICY, AND DOCTRINAL ISSUES 71 with a nuclear warhead even if no nuclear detonation occurred because the weapon did not arm. This is a particular issue in the case of CTM, because the system involves having conventionally armed missiles on a submarine that also carries nuclear weapons (see Chapter 2 for a discussion of nuclear safety issues). There are multiple measures—effectively “fail-safe” procedures and mecha- nisms—that can reduce, and in principle eliminate, any risk of an accidental launch of a nuclear weapon when a conventional strike has been ordered. These include taking procedural and physical steps to prevent (and if possible make physically impossible) (1) the launch of a nuclear missile in response to a conven- tional launch order, (2) the loading of nuclear-armed missiles into launch tubes for conventionally armed missiles, or (3) the transmission of a nuclear launch order when a conventional launch order is intended. The Navy’s plans for CTM are designed to meet these objectives by, among other things, including distinct command-and-control systems for the two types of missiles on the submarine (details are provided in Chapter 2). In addition, CTM would exploit an overall command and operational management system that has been repeatedly used to launch without error training missiles from an SLBM also armed with nuclear missiles. These “fail-safe” arrangements should combine procedural steps and physical controls—that is, procedural steps relying on the following: the careful selection of personnel, application of a screening program to personnel involved with CTM, well-understood and well-designed processes (e.g., command-and-control protocols), continual training and exercises, and simi- lar personnel-oriented measures; and physical controls, such as air gaps between conventional and nuclear launch systems, and required external inputs to arm missiles and warheads. These measures would mean that errors, if they occur, do not result in an accidental launch or detonation of nuclear weapons. NUCLEAR Ambiguity The most prominent objection on policy grounds to the conventionally armed Trident missile, and to some degree to all CPGS proposals, has been the “nuclear ambiguity” concern: the possibility, with potentially catastrophic consequences, that a launch could be misinterpreted by foreign observers as a nuclear attack. The concern most commonly raised is that a major nuclear power that regards itself as a potential adversary of the United States, that is, Russia or China—having detected the launch of a CTM from a deployed Trident SSBN and realizing that the missile was coming out of a ship that carries similar missiles with nuclear warheads—might misinterpret the nature of the event that it was observing. Before going through the issue more systematically, it is noted that a key ele- ment of the concern as most frequently articulated is that in the past, SSBNs and Tridents have been associated exclusively with nuclear weapons. If today a Trident were launched in the direction of Russia or China from an ocean area where the United States did not routinely do missile tests, and if that launch were detected, 

72 U.S. Conventional Prompt Global Strike there would be little alternative to interpreting the missile as being nuclear-armed. (Even in that situation, the apparently targeted country would have every reason to refrain from an immediate nuclear response and every reason to doubt that the United States was actually starting a nuclear war with a single missile.) One thing, and perhaps only one, is absolutely clear about the ambiguity problem: Simply using something other than Trident missiles (or another legacy ballistic system) as the delivery platform does not avoid the problem entirely. Indeed, ballistic missiles, whatever their nominal maximum range, have the virtue that in the event of an actual launch, well-understood and extensively practiced surveillance systems can quickly acquire sufficient data to determine their trajectory and hence the observer can be highly confident of their target. There have been hundreds of launches of long-range ballistic missile rockets for testing, training, space launch, research, and other purposes over the past several decades, and these have served to familiarize warning system operators with the identification of the nonthreatening nature of such events. If the maximum range of a new CPGS ballistic missile system (e.g., a Submarine-Launched Global Strike Missile [SLGSM] or a two-stage Trident based in the Indian Ocean) is too short to pose a threat to critical targets inside Russia or China, this would decrease the ambiguity problem. However, detection systems may not be able to identify what particular type of missile has been launched, and there may be uncertainties about the launched missile’s maximum potential range. Nor is prior association with nuclear weapons—that is, being a legacy sys- tem—a major factor, much less a decisive factor contributing to the misinterpreta- tion of a conventional attack. There simply is no “bright line” between nuclear and conventional systems when relatively long-range platforms are being considered. While long-range ballistic missiles are now associated exclusively with nuclear weapons, that was once equally true of many systems that have since come to have extensive conventional roles (e.g., the B-52, B-1, B-2, fighter bombers, sea- launched cruise missiles [SLCMs], air-launched cruise missiles [ALCMs]). If the United States develops and deploys CTM (or any CPGS system derived from an intercontinental ballistic missile [ICBM] or SLBM), even the psychological association of the system with an exclusively nuclear role will erode. If CTM itself were developed and deployed and then finally used—after years of discus- sion, testing, exercises, exchanges, and so on—first impressions and intuition would be different: A conventional attack would probably be the first and domi- nant “explanation” of what was going on (as it is today when the United States uses in conventional combat systems what were once exclusively or primarily nuclear—notably, long-range bombers or cruise missiles). In the case both of other CPGS systems that might be developed in the future to have operational advantages over CTM and of other ballistic-missile-derived systems of far greater maneuverability (e.g., boost-glide missiles, aircraft), it will be impossible for other countries to be certain of the target until after it has been hit. Merely avoiding a declared nuclear role for a new and highly capable CPGS 

POLITICAL, INTERNATIONAL, POLICY, AND DOCTRINAL ISSUES 73 system (or relying on a system with no nuclear legacy) would not eliminate the ambiguity problem, or even necessarily reduce it, relative to CTM or other legacy- based CPGS systems. Any system—old or new—that the United States might des- ignate as conventional-only (e.g., a submarine with an exclusively CTM load, or a new missile or hypersonic glide vehicle) could be armed with nuclear weapons. Other countries could never be confident that there is not a nuclear warhead on a new U.S. system simply because the United States says so. It is true that the United States has never put nuclear weapons on a delivery platform without testing and exercising it and its attendant nuclear security and command-and-control system in that role (though without an actual nuclear weapon); however, other countries would not necessarily regard that as decisive, particularly insofar as the issue is the use of such a platform in a limited but highly effective role as a precursor to a larger attack. Even with an entirely new CPGS system, it would be necessary to take measures to address the ambiguity problem. Having started with this observation in the belief that it is telling, the com- mittee now addresses the issue more systematically and, admittedly, with some complexity. There are two, logically and practically distinct, aspects of the ambi- guity issue. The first is the possible misinterpretation by an observing nation of a conventional CTM strike on a third party as a nuclear strike on its own territory. The second is a mistaken conclusion by an observing nation that a conventional CTM attack headed toward its territory is a nuclear attack. The issues to be considered in analyzing the ambiguity question are as follows: • Who would be able to detect the launch? • If a foreign nation were to detect the launch, would it be able to identify the missile type correctly and estimate the trajectory of the missile or of the reentry vehicles (RVs?) • If the launch were detected by a foreign nation, what would happen? • Would that nation’s nuclear forces and surveillance systems be alerted and, if so, what would be the consequences? • Would a “retaliatory” strike be ordered? • Even if there were no immediate adverse effects, what long-term reactions might be triggered? It is certainly appropriate to give serious attention to the ambiguity problem. The stakes are extremely high. In some sense the issue is not whether misinterpre- tation is unlikely, but rather, how unlikely is unlikely enough? In this connection, context matters greatly: reactions in time of general calm could be very different from those during a general crisis. Today and for the foreseeable future, U.S. relationships with Russia and China are less tense than in the Cold War era. In some important sense, there is less danger of accidental war, ambiguous launch interpretation, and so on. This is not to say that the technical issues to which this 

74 U.S. Conventional Prompt Global Strike chapter now turns are unimportant, but only that the larger context against which they will be seen has changed in a way that lowers the greatest risks. If a proposal were made for the use of CPGS against a third country at a time of increased U.S.-Russian tension, decision makers would need to take into account that in such circumstances there is a higher probability of overreaction or mistaken interpretations of events—of an inclination to put the worst possible face on events. Moreover, quality of information is an important factor. If a launch cannot be detected, there is no risk of overreaction; if detection and tracking capabilities are good, the same is true. Only detection with incomplete ability to understand the event (e.g., satellite detection of a launch without good radar-based capacity to predict trajectories) would be potentially dangerous. The dangerous condition is one in which the observer has imperfect information, as opposed to either no information at all or very accurate information. The problem of “misinterpretation” of a CPGS attack on a nuclear-weapon state (i.e., the target country knows it is being attacked and may be unsure of the nature of the attack) is potentially of particular seriousness. If the target country has nuclear weapons of its own, an attack that is (correctly) perceived as directed at critical targets deep in its territory could trigger a nuclear response. It is far less likely, but not impossible, that a nuclear power, knowing that the United States had the capacity to mount CPGS attacks, might be motivated to adopt a policy of responding with a nuclear counterattack to any U.S. conventional attacks on critical targets. The existence of these risks means that any use of CTM, or any other CPGS system, would have to be carefully considered in the broader context, particularly if the question were one of an attack on a nuclear power. Nonetheless, there are many reasons to regard the risks as sufficiently low and manageable that, while their existence needs to be taken into account—and the use of CPGS even forgone in certain cases because of concerns about escalation—they do not constitute a reason to forgo acquiring the capability. The risk of escalation makes any military action against a nuclear power a grave matter indeed, but these risks are no greater just because a CPGS system is involved. Consider first the most likely case—a conventionally armed ballistic missile launch directed at a country other than Russia or China. Currently, there is no reason to believe that any foreign country other than Russia has a warning system that could detect such a launch. Assuming that Russia would detect the launch, the warning system would also likely have sufficient tracking to conclude in a few minutes that the target was not in Russian territory. There is strong historical evidence to support this conclusion. Because of flight tests and space launches, Russia, like the United States, is accustomed to detecting and monitoring missiles and rockets after launch to establish their trajectories and confirm that whatever they are, they are not attacks directed at their own country. Even if the Russians were unsure of the trajectories, they would certainly discern the necessarily very limited scale of a CPGS attack and would be unlikely to conclude that the United 

POLITICAL, INTERNATIONAL, POLICY, AND DOCTRINAL ISSUES 75 States was starting a nuclear war with Russia in a “bolt-from-the-blue” attack with so few missiles. (Indeed, in discussing the much-publicized “Norwegian sound- ing rocket” incident, Russian spokespersons acknowledged recognizing that the event, though puzzling and even worrying, was not a U.S. attack for precisely this reason—its small scale.) Correct interpretation would be overwhelmingly likely if, for a period of years, Russia had been reading about, talking to U.S. experts about, and observing exercises in the use of CTM or some other CPGS system. By the time the system was deployed and operational, the natural interpretation of its launch would be that, as had long been anticipated, the United States was exercising its CPGS capability. It is likely that when their warning system detected any other than the most routine foreign missile launch, Russian commanders would increase the readiness level of their surveillance and command systems. That step alone does not increase the danger of a nuclear response, because it does not affect the alert state of the forces themselves—only of the command and surveillance systems. Indeed it is certainly intended to cause the situation to be monitored more closely and would cause greater attention to be paid to actions that might inadvertently cause esca- lation. Even if the Russian civilian leadership did mistakenly conclude initially that Russia was the target of a small nuclear attack, it would have every reason not to order an immediate counterattack because the few missiles in flight could not significantly degrade the country’s ability to respond after the situation was clarified. This would be true even if Russia had a policy of “launch under attack” for major attacks or was concerned about the possibility of a precursor nuclear attack disabling its warning system. Today other nuclear-weapons states (including China) appear to have very little capacity to detect a ballistic missile launch (and therefore very little potential for misinterpreting what they do not know). When China develops early-warn- ing capabilities, as it presumably will in time, the observations above about the Russian capacity to analyze a ballistic missile attack correctly—and to refrain from nuclear response until it has—would apply. However, in the Chinese case, because China’s nuclear forces are so much smaller than Russia’s, an erroneous perception that a small nuclear attack was underway would perhaps be somewhat more likely to be regarded as a serious threat than would be the case for Russia. The word “perhaps” is used here because the possibility seems very small, and if China were concerned about bolt-from-the-blue attacks, it would have many mechanisms by which to improve the survivability of its forces. What is at issue is the relative risk of an accidental nuclear war and of the consequences of not using a CPGS weapon. If CTM or another type of CPGS sys- tem is deployed, everything reasonably possible to further reduce the risks should be done, if only because there is always some risk that an unexpected event—the   eoffrey G Forden, Pavel Podvig, and Theodore A. Postol. 2000. “False Alarm, Nuclear Danger,” IEEE Spectrum, Vol. 37, No. 3, March. 

76 U.S. Conventional Prompt Global Strike sudden appearance of a unscheduled missile launch—might somehow set off a chain of unintended and supposedly impossible reactions. The stakes are so high that it is important, if any form of CPGS is deployed, to take additional measures to further increase the chances that Russia (and China) would not misread its use. These could include cooperative measures to increase information about the system and its operation (notification, transparency arrangements, a joint warning center, data exchanges, participation in and observation of exercises, inspection regimes) as well as extensive and candid discussions of the nature of the system and the U.S. doctrine for its use. There are also technical devices that could be installed on a mutually agreed basis that would provide additional confidence that a CPGS system declared as armed exclusively with conventional weapons had not been covertly nuclearized. An example of such a device would be the continuous visual monitoring system outlined in Appendix H of this report. All the foregoing said, it is impossible to avoid the conclusion that whatever the risk of misinterpretation might be, it would be higher in cases where the issue was not a launch against a third country, but an actual attack on targets in Russia or China (albeit conventional and limited). Such an attack could, by definition, occur only at a time of very serious tensions and, quite possibly, active military engagement, between the United States and the target country. In such a context, the chances of overreaction, while still extremely small, would be higher than in other situations. While both sides would have an interest in controlling escalation, the combination of tension, suspicion, and uncertainly would be dangerous, and the risk of events getting out of hand would be increased. This is not to say that potential use of CPGS even against a nuclear-weapons state should be ruled out in all circumstances. The problem of misinterpretation of a prompt precision attack on a nuclear-armed state is a special case of the general and very serious problem of the control of escalation in a conflict between nuclear- armed states. It is obvious that any such conflict carries great risks, but it could only come about if great interests were at stake, so that refraining from military action for fear of escalation could carry great costs. The challenge of using mili- tary force in a conflict with a nuclear power without unintended escalation is by no means limited to CTM (or any other form of CPGS capability), but it does mean that the use of a CPGS capability to attack highly critical targets deep within the territory of Russia or China would require great justification and consideration. (It is difficult to conceive of situations in which the benefits of using CPGS against targets in Russia or China would outweigh the risks.) In sum, concerns about nuclear ambiguity will always be a potential issue, one whose significance depends not primarily on the technical characteristics of the CPGS system but on the context, scale, and target of the attack and on the degree to which transparency and confidence-building measures have been employed. There will be circumstances where ambiguity concerns might and should make a President less likely to use CPGS, especially in a crisis or for attacks against a nuclear power, but that is no reason to deny the possibility of its use in other 

POLITICAL, INTERNATIONAL, POLICY, AND DOCTRINAL ISSUES 77 situations where the risk of misinterpretation is not present. Moreover, the risk arising from the ambiguity of CPGS systems is one of the lesser risks—militarily and politically—associated with attacking another country, especial a nuclear- armed country. Overflight and debris As discussed in the section above on ambiguity, the overflight of third coun- tries situated between the attacking and the targeted countries (or flight paths that generate concern about impending overflight) may lead decision makers to assume mistakenly that their own country is under attack. Such concerns are particularly important when the third party has both the ability to detect overflight and the ability to respond militarily. For CPGS systems that rely on air-breathing systems, overflight may also render the missile vulnerable to the air defenses of third parties. In addition to issues of ambiguity and air defenses, overflight can pose politi- cal problems. Many countries would protest alleged violations of their airspace, as Japan did when North Korea conducted a test of a ballistic missile on a trajectory over Japan. Such protests could occur even if the third party had no capability to detect the overflight, because in most cases the fact that overflight had occurred would be obvious after the fact. Such protests could have significant political ramifications and would have to be taken into account in the decision to order a strike. Potential political problems could be mitigated by asking for overflight permission in cases where such permission is likely to be granted, or by notifying the government about the planned overflight in cases where permission would likely be denied, but in both cases unacceptable security compromises or delays in execution would be possible. In most cases, the political consequences of overflight, apart from ambigu- ity concerns, would not be a decisive or even an important reason not to conduct a strike. The longstanding position of the United States is that transit through space on ballistic flight paths or orbits is not a violation of national sovereignty or airspace. Moreover, the United States has conducted nonconsensual overflight with aircraft when this was deemed necessary for very high priority operations. One example is the 1998 U.S. cruise-missile strikes on al-Qaeda training camps in Afghanistan, which involved overflight of Pakistan. CPGS strikes presumably would be rare, and limited to the most important and time-urgent targets, in which case possible post-strike protests of overflight of most third countries would not be a major factor in a decision to strike (although it might play some role in mis- sion planning). Thus, the ability of a particular CGPS system to avoid overflight of selected territory should be considered an advantage, but not a decisive one. Submarine- based systems, such as CTM and SLGSM, have the advantage that submarines can in some cases be positioned to avoid overflight of selected countries. The 

78 U.S. Conventional Prompt Global Strike most important overflight concerns—those related to potential ambiguity prob- lems involving Russia—can be addressed by designating routine patrol areas that avoid overflight. But the repositioning of submarines to avoid overflight of other countries could take many hours or even days, which might not be compatible with the prompt nature of the CGPS mission (and, in the case of CTM, would interfere with the core nuclear-deterrent mission of the submarines). Boost-glide and air- breathing systems would have substantial maneuver capability on their path to the target, which could be used to avoid overflight of some countries or air defenses. All else being equal, this is an advantage, but it is not dispositive. Political concerns regarding overflight would be amplified by debris falling on a third country, which could result from the normal operation of the system (e.g., third-stage rocket motor) or from a system failure (e.g., the premature impact of a warhead or the crash of a cruise missile). Such concerns would be amplified further if the debris resulted in casualties or property damage. As in any use of military force, the potential for civilian casualties or other collateral dam- age would be a factor in the decision to order a strike. In most cases, the risk of collateral damage from overflight (as opposed to warhead impact in the intended target area) would be minimal, even for ballistic systems in which there can be no choice about the flight path from a particular launcher to the chosen target. In the case of CTM (and presumably SLGSM), the final stage of the missile contains enough propellant to allow mission planners to divert it hundreds of kilometers from its normal path, toward a body of water or a sparsely inhabited area, in order to minimize collateral damage. Boost-glide or air-breathing systems can take flight paths that avoid densely populated areas, but as with ballistic missiles, they are vulnerable to failures that could deposit debris anywhere along their flight path, which could diverge substantially from the planned flight path. CPGS sys- tems could be equipped with systems to self-destruct automatically or to impact in uninhabited areas in the event of deviations from planned flight paths or of other failures that would prevent the warhead from reaching its intended target. Although CTM as currently planned would not have this capability, the committee sees no technical reason why it could not, if this was considered a serious prob- lem. (The objection to self-destruction mechanisms in nuclear missiles—that the target country might be able to exploit them—does not apply nearly so strongly to limited conventional strikes.) The limited risk associated with debris can be illustrated with a simple cal- culation. For most systems, the debris resulting from normal operation or system   elf-destruction would not eliminate debris; it would turn the large warhead into many small frag- S ments. If the self-destruction occurred after the boost phase was completed, these fragments would continue along the same path as that which the intact warhead would have traveled. Although the total mass of debris that impacted the ground would be the same, the impacts would be spread over a much larger area and would occur at much lower velocities. This could reduce the potential for property damage and a high-casualty event such as the impact of an intact warhead on a school or hospital, but a complete assessment would require further study. 

POLITICAL, INTERNATIONAL, POLICY, AND DOCTRINAL ISSUES 79 failure would have the potential to cause deaths out to a distance of perhaps 10 meters from the point of impact, corresponding to a lethal area of a few hundred square meters. The average population density of Earth is about 40 persons per square kilometer, so the expected number of deaths from a random impact is on the order of 0.01, or 1 death per 100 impacts. Given diversion capabilities that caused debris to land in less-populated areas, the expected damage would become much smaller. Thus, the risks associated with debris are unlikely to be a signifi- cant concern in situations serious enough to trigger a CPGS strike. Most uses of military force carry some risk to civilians, sometimes in third countries, as in the 1998 strike against al-Qaeda training camps in Afghanistan, which resulted in the crash of cruise missiles in Pakistan (though no casualties resulted from these failures). ACCESS TO FORWARD BASING Conventional Trident Modification and most other proposed CPGS concepts need not be deployed on forward U.S. military bases. This fact has led some proponents to suggest that CPGS systems could substantially compensate for or hedge against a loss of access to bases and other facilities in foreign countries, or even that they could provide a basis for curtailing forward deployments. The problems for overall U.S. military capability that would arise from such reduced access are much greater than could be addressed by CPGS systems. In addition, some CPGS concepts depend to one degree or another on access to for- ward bases, as do many of the enablers (e.g., intelligence derived from unmanned aerial vehicles [UAVs] or from Special Operations Forces). Thus, the United States will retain strong interests in forward basing regardless of the deployment of CPGS systems. Some CPGS options, such as CTM, are often said to be truly global in a way that forward-based options cannot be. But forward basing in regions such as the Middle East, where targets for CPGS are most likely to be located, will remain important not because such basing makes rapid attack possible, although that will sometimes be the case, but because the very interests that make CPGS potentially useful in those regions will also make it appropriate to maintain other military capabilities there. It would be a mistake to believe that CTM or other CPGS sys- tems would make possible a meaningful reduction in forward basing. Moreover, even if the United States goes forward with CTM, the most mature CPGS option, it will be several years until this capability is operational. Until that time, the United States will be entirely dependent on forward-based systems (fighter aircraft, armed UAVs, submarines armed with SLCMs, B-52s armed with ALCMs, tanker support for long-range aircraft based in the United States) to conduct strikes on distant targets. If the need for a CPGS capability is truly urgent, then it is no less urgent to maintain and, where possible, improve 

80 U.S. Conventional Prompt Global Strike forward-based capabilities to carry out strikes against potential CPGS targets as rapidly as possible. Proliferation Although there have long been dual-capable aircraft, the long-range ballis- tic missiles deployed by the nuclear-weapons states have been armed only with nuclear warheads. Attempts by other countries, such as North Korea and Iran, to acquire long-range ballistic missiles generally have been viewed as an integral component of an effort to acquire a nuclear-weapon capability. Such attempts have been opposed by the United States mainly on these grounds, sometimes with the argument that for reasons of cost and accuracy, ballistic missiles are inher- ently unsuited to conventional roles. The presumed connection between missiles and nuclear weapons is highlighted by the Missile Technology Control Regime (MTCR) guidelines, which begin as follows: “The purpose of these Guidelines is to limit the risks of proliferation of weapons of mass destruction (i.e., nuclear, chemical and biological weapons), by controlling transfers that could make a contribution to delivery systems (other than manned aircraft) for such weapons.”  Although the text also refers to chemical and biological weapons, the 500-kilo- gram payload limit specified in the guidelines is based on the estimated mass of a nuclear warhead. Some have argued that the deployment of CTM or other CPGS systems will break the presumed link between ballistic missiles and nuclear weapons, result- ing in the spread of long-range missiles to additional countries. In this view, the deployment of a ballistic CPGS system by the United States would allow other countries to claim more credibly that their long-range missile programs were intended for conventional, not nuclear, delivery and are just as legitimate as the U.S. CPGS system. Similarly, it is argued that a weakening of the presumed coupling between missiles and nuclear weapons would weaken the MTCR and related efforts to limit the spread of missile technology. Although the Nuclear Non-Proliferation Treaty recognizes the right of only five countries to possess nuclear weapons, many countries argue that this division between “haves” and “have nots” should not be extended into other realms. In this view, the MTCR derives much of its legitimacy from its connection to nuclear nonproliferation—a connection that would be weakened by CPGS. There is also a related argument that the U.S. acquisition of CTM or other CPGS capability would encourage or legitimate efforts by other countries to develop equivalent systems. There are, for example, reports of Russian officers who, after criticizing U.S. plans, note that such a system would be useful for Russia as well. The systems that others might develop would not necessarily mimic U.S. CPGS systems in being designed to deliver small payloads with high   See <http://www.mtcr.info/english/guidetext.htm>. Last accessed on March 27, 2008. 

POLITICAL, INTERNATIONAL, POLICY, AND DOCTRINAL ISSUES 81 accuracy. Indeed, they might be intended for very different sorts of targets, such as satellites. Finally, it is argued that to the extent that U.S. CPGS systems threaten impor- tant targets in potential adversary countries, these systems might also stimulate other countries’ efforts to develop and deploy capabilities that might deter such attacks, such as nuclear and other weapons of mass destruction and associated delivery systems. This is sometimes countered by arguments that CPGS, by dem- onstrating the U.S. ability to destroy adversary missile and WMD sites rapidly, would dissuade countries from even attempting to field such capabilities. On balance, the committee judges that the deployment or use of CTM or other CPGS systems is by itself unlikely to have a substantial effect on the proliferation of ballistic missiles or nuclear weapons or on the development of similar systems by other countries. In general, countries will do what is in their own national interest and within their technological capability and financial capacity, regard- less of what the United States does—or does not do—about CPGS. CPGS should not serve as a significant additional stimulant for the acquisition of systems that could attack high-value targets in the United States (e.g., in order to deter U.S. use of a CPGS system), because CPGS would not add meaningfully to the already very substantial U.S. strike capabilities that potential adversaries must take into account. Nor is CPGS likely to have a substantial effect on efforts to impede the spread of ballistic missile or nuclear technology. Countries that possess the tech- nological capacity and have the strategic need for these capabilities have powerful incentives and self-interests in limiting the spread of these technologies. It is dif- ficult to see how either developing CPGS or forbearing to do so could alter this to any significant degree. Arms control and treaty issues Two arms control agreements that are currently in force impose restrictions on the deployment of intermediate-range and long-range delivery systems: the Intermediate-Range Nuclear Forces (INF) Treaty of 1987 and the Strategic Arms Reduction Treaty of 1991 (START). The Moscow Treaty of 2002 (Strategic Offensive Reductions Treaty, or SORT) also requires reduction and limitation of strategic nuclear warheads. Below is a discussion of the impact of these treaty restrictions on CTM and other possible CPGS systems. Intermediate-Range Nuclear Forces Treaty of 1987 The Intermediate-Range Nuclear Forces Treaty was signed by the United States and the Soviet Union on December 8, 1987. After the breakup of the Soviet Union, Russia, Belarus, Kazakhstan, and Ukraine became signatories and assumed the obligations of the former Soviet Union. The INF Treaty prohibits flight-testing, production, and deployment of all ground-based ballistic missiles 

82 U.S. Conventional Prompt Global Strike and ground-launched cruise missiles with ranges between 500 and 5,500 kilo- meters, regardless of warhead type. The prohibition applies to missiles deployed anywhere in the world; thus, the United States could not deploy a conventionally armed intermediate-range ballistic missile or ground-launched cruise missile at a forward base, such as Diego Garcia in the Indian Ocean. The INF Treaty is of unlimited duration. Accordingly, its restrictions remain in force unless the United States or Russia determines that extraordinary events have jeopardized its supreme national interests and exercises the right to withdraw from the treaty. The INF Treaty places no restrictions on manned aircraft, air-launched or sea-launched systems, or ground-launched systems with ranges less than 500 kilometers or greater than 5,500 kilometers. Strategic Arms Reduction Treaty of 1991 The Strategic Arms Reduction Treaty, which was signed on July 31, 1991, requires the United States and Russia to limit their deployed strategic arsenals to no more than 6,000 warheads, with no more than 4,900 on ICBMs and SLBMs, in accordance with agreed counting rules. The agreement limits deployed warheads by limiting the number of delivery vehicles and associated launchers to 1,600. After the breakup of the Soviet Union, a protocol to the treaty made Russia, Ukraine, Kazakhstan, and Belarus parties to the treaty and required the latter three to transfer their nuclear weapons to Russia. The required START reductions were completed by all parties in December 2001. START covers all ICBMs and SLBMs and their associated launchers, includ- ing new types of ballistic missiles, and no distinction is made between nuclear- armed and non-nuclear-armed ballistic missiles. Thus, the submarine-launched CTM and SLGSM missiles and launchers would be subject to treaty limits. For example, each CTM would be counted as one SLBM launcher and eight warheads toward the aggregate limit on launchers and warheads. This should not present any practical problem, however, because under the Moscow Treaty of 2002 (see below), the United States plans to reduce the number of launchers and attributed warheads well below the START limits. The CTM missiles would, however, be subject to all START rules, including reentry vehicle on-site inspections to con- firm that they are not armed with more than the permitted number of RVs. SLGSM would be handled differently from CTM. Presumably it would have to be either exempted on the basis of an agreed verification scheme or treated as a strategic weapon system (with an attributed number of RVs and an agreed definition of launchers, and so on). Similarly, deployments of conventionally armed ICBMs would be covered by START. In addition to the limits on launchers and warheads noted above, START requires that ICBMs be deployed only in silo, road-mobile, or rail-mobile launchers; soft launchers are prohibited except at test ranges and space-launch 

POLITICAL, INTERNATIONAL, POLICY, AND DOCTRINAL ISSUES 83 facilities. ICBM bases must be located more than 100 kilometers from test ranges and space-launch facilities. Test ranges are limited to no more than 25 missiles and 20 fixed, soft launchers, which would restrict the number of conventional ICBMs that could be deployed at Vandenberg Air Force Base, California. Finally, START prohibits deployments of ICBMs outside national territory, as well as air-launched and surface-ship-launched ballistic missiles. These restrictions and prohibitions make it difficult for the United States to deploy conventional ICBMs while avoiding overflight and ambiguity problems that would arise from deploy- ments at existing ICBM bases. START defines a ballistic missile as a “vehicle that has a ballistic trajectory over most of its flight path.” Therefore, boost-glide systems (even if derived from existing ICBMs or SLBMs) may not be subject to START if the glide portion of their flight is more than half of the total flight path. As noted in Appendix G of this report, however, the glide range generally may be much less than half of the total flight path, unless the initial speed is comparable with ICBM or orbital speed. However, at some small sacrifice in payload, glide range could be increased to exceed the ballistic range considerably by boosting the weapon to near-orbital speed. Conventionally armed, long-range, air-launched cruise missiles are not restricted by START so long as they are distinguishable from nuclear-armed, long-range, air-launched cruise missiles. Moscow Treaty of 2002 On May 24, 2002, the United States and Russia signed the Strategic ­Offensive Reductions Treaty (SORT) in Moscow. Under SORT, they agreed to reduce the number of operationally deployed strategic offensive warheads to 2,200 or fewer by December 31, 2012. The warhead limit takes effect and expires on the same day. The treaty contains no verification provisions or agreed counting rules. Because SORT applies only to nuclear warheads, it imposes no restrictions on ballistic missiles or other delivery vehicles armed with conventional warheads. Thus, SORT should have no effect on CTM or any other CPGS system. Future Agreements START is scheduled to remain in force until December 5, 2009. The parties are committed to meet at least 1 year before this date to consider whether the treaty should be extended. Although START contains provisions for an indefinite series of 5-year extensions, neither the United States nor Russia has indicated that it wishes to extend START in its current form. This has led some observers to conclude that the United States should simply ignore the restrictions that START would impose on CPGS systems, because START will expire before CPGS sys- tems could be deployed. 

84 U.S. Conventional Prompt Global Strike The START restrictions are unlikely to disappear entirely, however; the United States and Russia are likely to attempt to extend some of its provisions or to replace it with another agreement. One issue in any such negotiation would be the degree to which the United States would insist on modifications or provisions designed to protect certain CPGS systems or deployment options. As in any nego- tiation, this would have to be weighed against other goals for both parties. A key issue in the negotiation of a follow-on agreement would be verification and transparency measures. Because the central goal of such an agreement would be to limit nuclear weapons, there might be a temptation to try to exclude con- ventionally armed ballistic missiles from any such measures. Such an exclusion would lead to another kind of ambiguity problem and seriously complicate any effort to verify limits on nuclear systems. It will be desirable to reassure Russia (and others) that CPGS systems are not armed with nuclear weapons by using transparency measures that could be more intrusive and revealing than would be acceptable with nuclear systems. Looking to the longer-term future, the existence of substantial numbers of CPGS systems might be a concern because of the possibility that they could be rearmed with nuclear weapons. This would not be a concern at the relatively high limits on nuclear warheads established by START or even the Moscow Treaty, but it could arise if the United States and Russia move to significantly lower levels— for example, less than four times the number of long-range conventional-missile warheads. Such concerns could, however, likely be resolved with transparency measures to reassure other parties that conventional missiles are not armed with nuclear warheads (and, perhaps, that additional nuclear warheads are not available to arm the conventional missiles). Thus, assuming that the United States is willing to accept far-reaching trans- parency measures—which it ought to be, for reasons related to resolving ambigu- ity concerns—CPGS should neither be a barrier to future arms control agreements nor a reason not to pursue future arms control agreements. Strategic considerations A significant fraction of Russian strategic nuclear forces—submarines in port and long-range bombers in soft shelters—could be vulnerable to precision attacks with conventional weapons. If penetrating warheads capable of destroying hard- ened surface targets are available, some command, control, and communications facilities and, with very high accuracy, missile silos, could be vulnerable. Even if the initial deployment is very limited, any CPGS system could in principle be substantially expanded after its development and testing have been successfully completed. Such expansion would raise at least the theoretical prospect of a conventional preemptive strike capability against Russian nuclear forces. These concerns would be more acute for China, because China’s nuclear forces are much smaller and less diverse than those of Russia. However, even for China, 

POLITICAL, INTERNATIONAL, POLICY, AND DOCTRINAL ISSUES 85 large numbers—at least many hundreds—of CPGS weapons would be needed to mount an even marginally plausible conventional threat to a sophisticated nuclear force. Support for and deployment of any of the CPGS systems considered in this report would not imply that such a massive “strategic” CPGS system should be built, and the committee would not endorse such a project. If the United States were perceived by Russia or China to have embarked on such a course, they could respond to the development of this capability in many ways. In an attempt to strengthen deterrence against such attacks, they might, for example, declare that they would respond to a conventional attack on their nuclear forces and other strategic assets as if it were a nuclear attack. For China, this would represent a major change from its current declaratory policy of no-first- use of nuclear weapons, but it would be a change that China might well make as a response to a clear threat to its deterrent. Russian or Chinese leaders might believe, however, that threats to order nuclear retaliation to conventional attacks would be less than credible in the eyes of U.S. leaders and that CPGS might therefore reduce their ability to deter U.S. attacks. It would be reasonable to expect that Russia and China would respond to the prospect of a large-scale CPGS system by moving, as rapidly as technology and resources would permit, to decrease the vulnerability of their nuclear forces to precision conventional attacks. Such efforts might include the following actions: • Increasing the number or readiness of mobile launchers—mobile ICBMs and submarines at sea—which are difficult or impossible to target with CPGS systems; • Expanding the total number of deployed launchers or warheads (or refus- ing to implement already-agreed reductions) in order to ensure the survival of a given number of warheads after a conventional attack; • Hardening facilities to make them less vulnerable to conventional attack; • Developing close-in defensive systems to protect point targets, such as ICBM silos, from conventional attack; • Enhancing the ability to launch vulnerable nuclear forces rapidly on warn- ing of, or in the midst of, a conventional attack; and • Using cover and deception to disguise or misrepresent the locations of their systems. Some of these actions, such as increasing readiness or the ability to launch on warning, if taken by Russia or China, could increase the probability of mistaken attacks on the United States. It should be emphasized that there would be serious practical obstacles to expanding a CPGS system to the point where it might be viewed by Russia as having a significant capability against its nuclear forces, even if the capacity to destroy hard targets had already been developed and demonstrated. The largest barrier is deploying the many hundreds or even thousands of missiles that would 

86 U.S. Conventional Prompt Global Strike be required to mount a significant conventional threat against Russian nuclear forces. This could be accomplished in less than a decade only if many, if not most, of the existing nuclear-armed Minuteman III and Trident II missiles were converted to the CPGS mission. But such conversions would simultaneously draw down U.S. nuclear forces, with the result that the overall threat perceived by Rus- sia would not increase and might even decrease. Moreover, Russia is unlikely to consider plausible a scenario in which the United States would sacrifice a large fraction of its nuclear deterrent force to the CPGS mission. Thus, as a practical matter, deploying a CPGS system that could threaten Russian nuclear forces would likely involve building at least several hundred additional intercontinental-range missiles and their associated launchers.  The additional cost of the missiles alone would be several tens of billions of dol- lars; launchers would add billions more. This would be a highly visible project that would take more than a decade to carry out, giving Russia plenty of time to consider its response. Thus, even conservative Russian planners would have no reason to react to the limited CPGS deployments considered here by modifying their nuclear forces or associated policies. They could instead monitor the situa- tion, while making clear to the United States that larger deployments would be a matter of serious concern. As discussed elsewhere, the United States can take steps intended to assure Russia that it is not the intended target of a CPGS system and that the system has little capacity to destroy certain types of targets. In addition to keeping the number of deployed CPGS systems small, the United States can limit CPGS deployments to areas that minimize the potential for attacks deep in Russian territory, and it can adopt transparency measures, such as launch notification, joint early warning, technical briefings, inspections, and devices integrated into the payload that signal a conventional missile launch. At the same time, U.S. decision makers should be aware of the possible reac- tions of Russia and China—and the potential impact of these reactions on U.S. security—to a decision to move from a limited CPGS capability, such as that of CTM, to one capable of high-volume attacks that could possibly threaten Russian or Chinese nuclear forces. For this, and many other reasons, the United States should, in deciding to proceed with any CPGS program, make clear both in its declaratory policy and in its procurement actions and doctrinal development that it is not embarking on any such far-reaching effort.   possible exception would be the redeployment in a CPGS mode of some or all of the Minute- A man II missiles that were withdrawn from deployment and their silos launchers that were destroyed under START. These missiles are currently in storage in the Rocket Systems Launch Program and could be redeployed, but this would require building new silo launchers, which would require many years and billions of dollars to complete. In addition, it would be necessary to identify and clarify any START restrictions on redeploying a missile (such as Minuteman II) which has been designated under START as a “retired missile.”