In some respects, nuclear deterrence and assurance are now more complex than during the Cold War. Deterring a new set of nuclear-armed or potentially nuclear-armed adversaries while also assuring threatened allies requires continuous and informed balancing of both objectives. As current nuclear nonpeers emerge as near-peers or peers, they may not act as we expect. The nonpeer states that currently possess nuclear weapons and who are developing them are often ruled by regimes difficult to penetrate and about which decision-making dynamics are difficult to divine. Planning for the future must accommodate the uncertainty associated with the transformation of regimes from those that are staunch adversaries to those that may be tomorrow’s allies. The situation is further complicated by the need to address the possibility of surprise in areas of technology or unforeseen changes in equipment effectiveness. Finally, because research addressing deterrence and assurance has declined since the end of the Cold War, the conceptual basis for developing and improving U.S. strategy and for equipping forces may not be as robust as it once was.
In Chapter 3, the committee reviewed and assessed tools, methods, and approaches (collectively referred to henceforth as “methods”) that might be used to improve our understanding of how nuclear deterrence and assurance may work or fail in the 21st century and the extent to which such failures might be averted by proper choice of nuclear systems, technological capabilities, postures, and concepts of operation for U.S. nuclear forces. The committee had background in and was briefed on current analytics efforts. It concluded that while methods are important, the key to high-quality analysis in support of nuclear deterrence and assurance is
qualified people who have extensive experience in the nuclear deterrence and assurance domain as well as in the relationships of nuclear options to general deterrence and assurance. The committee identified two types of methods that should be emphasized because of their relevance to the added complexity and uncertainty inherent in a deterrence and assurance environment that contains a more and more diverse set of nuclear adversaries. These are methods for (1) gaining insight into different styles, modes, and motives of an actor’s decision making (discussed in Chapters 3, Appendix D, and Appendix E) and (2) dealing with “deep uncertainty” (discussed in Chapters 2 and 3). The committee also points out the need for analysts to be conversant in and use a suite of analytic methods, as well as the promise of hybrid methods in which different tools and methods, or the results of different approaches, are integrated—for example, using human gaming to inform quantitative modeling, as discussed in Chapter 3.
In considering how the Air Force should best approach deterrence and assurance analyses, the committee developed a top-level framework, primarily as a basis for categorizing these tasks and associated requirements for methods. Conceptually, the framework is straightforward. The reality is that the Air Force analytic community is not resourced to perform the analyses identified in this framework, many if not most of which require a whole-of-government perspective.
At a conceptual level, deterrence and assurance proceed through a sequence of steps, beginning with characterization of the situation or scenario involving potential actions adverse to the interests of the United States. That characterization leads to the identification of alternative U.S. objectives and then a characterization of the players in terms of their objectives, constraints, and values. Because of uncertainties associated with this characterization, alternative characterizations would ideally be constructed. The next step consists of determining feasible response options in the context of available capability and legal and political constraints, followed by the construction and assessment of a set of integrated, well-hedged, whole-of-government options, the choice of initial actions and the execution of a strategy, observing and adapting as the situation unfolds. It is within this context that the Air Force fulfills it deterrence and assurance mission. The subject of this study was analytic methods used to support Air Force decisions as it organizes, equips, and trains to meet its responsibilities in deterring adversaries and assuring allies.
The committee developed and applied criteria for evaluating methods. No candidate stands out alone. For example, methods related to actor-specific modeling and deep uncertainty have matured over the last two decades and are particularly relevant. Qualified analysts will, based on broad knowledge and expertise with the spectrum of available methods, select and apply those which are most appropriate. For many analysis tasks, a suite of methods will be the sensible and preferred tactic. As noted above in examining the current analysis efforts in nuclear deterrence and assurance the committee observed that analysts were doing a remarkable job given
the resources available. The community makes excellent use of classic analytical methods. It has begun to address the challenge of actor-specific knowledge, but it is not currently prepared to fully exploit developments in this domain. This community has taken preliminary steps to address deep uncertainty. A significant impediment to improved analysis in the deterrence and assurance domain is the limited number of analysts assigned to the deterrence and assurance mission and the organizational barriers that separate military and nonmilitary analytical agencies addressing deterrence and assurance in a whole-of-government context.1
In broad terms a responsive analysis will include the tasks illustrated in Figure 4-1. Initially, potential adversaries and allies must be identified, together with the deterrence and assurance goals associated with each—those viewed in the larger context of influence, to include combinations of carrots and sticks. Since strategies should be tailored to specific adversary/ally combinations, separate analyses are required for each combination. As displayed in Figure 4-1, the first and most important task in a specific crisis is to understand both adversaries and allies, which can be aided with leadership profiles. These profiles, addressed in Chapter 3 and Appendix E, are designed to identify an adversary’s or ally’s valued assets, help identify the range of behaviors that might be seen in crisis, assess the barriers to reception of deterrence messages, and estimate responses to perceived messages. They should describe likely changes in these factors as situations change. Given this information, and recognizing that peer/near-peer, regional and nonstate actors pose significantly different challenges, sets of capabilities can be generated and evaluated in terms of effectiveness of actions (“messages”) in producing a desired change in adversary behavior. This information can then be used to construct alternative organizations, equipment, and training, assuming different but explicit contributions from other services and government agencies.2
Alternatives should be analyzed and evaluated for flexibility, adaptability, and robustness, primarily in the context of uncertainty. Leadership profiles will be subject to varying degrees of uncertainty and error. Accordingly, provision must be made for undesirable, unexpected, and surprising behavior by adversaries and allies. Similarly, alternatives must be examined and evaluated from the perspective of technological surprise and unexpected changes in equipment effectiveness.
1 Hunter Hustus, Technical Advisor, Office of the Assistant Chief of Staff of the Air Force for Strategic Deterrence and Nuclear Integration, personal communication to the committee on December 19, 2013.
2 With respect to the Air Force, a broad spectrum of contributions could be brought into play, including, intelligence, surveillance, and reconnaissance and cyber operations. Also, there are second-order contributions such as the use of Air Force assets to deploy missile defense systems.
FIGURE 4-1 Notional tasks involved in deterrence analysis.
Sensitivity to the roles of other services, other government agencies, and possible actions by allies should also be analyzed. Keeping in mind the presence of deep uncertainty, assessments must consider the risk of being unable to deliver a particular capability at the time it is needed or the risk that the capability does not produce the desired effect. Such assessments can characterize the alternative under consideration taking into account actor-specific and situation-specific knowledge.
Based on the information developed for the set of adversary/ally combinations, recommended alternatives can be synthesized, integrating the information from the separate analyses to produce one or more options for consideration by the leadership of the Air Force. It is clear that the Air Force is a major but not the only member of the deterrence team. Each of the Services and many other departments and agencies have roles and responsibilities and should be considered in developing understanding and conducting analysis. However, to meet its Title 10 responsibility to organize, train, and equip the Air Force contribution, the Air Force should undertake a series of tasks related to analysis in support of deterrence and assurance.
Finding 2-2. Analytic Framework. Because the U.S. approach to strategic deterrence and assurance needs to be continually adapted, a management plan is required that defines comprehensively the set of continuing analytic foci, which includes nuclear command and control; air and missile defense; cyber, space, geostrategic, and technological changes; and the challenges of tailoring deterrence and assurance to adversaries and allies. This analytic management plan is in addition to tasks related to weapons, forces, personnel, and the nuclear enterprise in general.
Recommendation 1. In support of senior Air Force leadership guidance, including the Flight Plan for the Air Force Nuclear Enterprise, the Air Force should develop and maintain a comprehensive strategic deterrence analysis plan to identify the tasks that produce information required to organize, equip, and train Air Force nuclear deterrence and assurance forces and support combatant commanders (Air Force, 2013).
Rationale. Organizing, equipping, and training Air Force elements to be used in conventional and nuclear deterrence and assurance is a critical and complex challenge. The Air Force should develop sound and defensible strategies for developing and fielding its force. Given the current state of scholarship generally and expertise in the Air Force in particular, a multiyear plan for study and analysis of the widening range of deterrence and assurance issues is a key requirement. Furthermore, once in place, an Air Force deterrence and assurance analysis program (DAAP) would provide a means of ensuring that sufficient attention is paid to generating flexible, adaptive, robust strategies, which the committee believes are essential in the nuclear deterrence and assurance domain in the 21st century.
The envisioned DAAP would rely on input from the Department of Defense and other U.S. government agencies. Based on its deliberations, the committee believes that tools, methods, and approaches are available but that an institutionalized means of cross-agency collaboration and coordination does not exist. The analysis plan would provide the basis for establishing such an organization and defining its responsibilities.
Implementation. With respect to implementation, the Assistant Chief of Staff for Strategic Deterrence and Nuclear Integration should be assigned responsibility, with contributions from the Deputy Chief of Staff for Intelligence, Surveillance, and Reconnaissance; the Deputy Chief of Staff for Operations, Plans and Requirements; the Deputy Chief of Staff for Logistics, Installations and Mission Support; the Deputy Chief of Staff for Strategic Plans and Programs; and the Director for Studies and Analyses, Assessments and Lessons Learned, as well as the Air Force Global Strike Command and the Air Force Materiel Command, for developing and recommending to the Chief of Staff of the Air Force an outline of the DAAP.
Finding 3-2. Psychological Framework. Deterrence and assurance are largely a psychological concept. Thus, a proper evaluation of proposals for them will rely not only on the balance of military forces but also, whenever possible, on an understanding of the mindset and decision making of the adversary or ally.
Finding 3-3. Tailoring Key Messages. To elicit the intended response, it is important for the sender to have methods and tools that can detect opportunities and send messages tailored to a recipient that is open (willing and able) to make a response based on available information rather than on motivational, affective, or cognitive biases in a deterrence or assurance situation.
Recommendation 2. The Air Force should focus analytic enhancements in support of deterrence and assurance assessment on the human and human organizational factors at the heart of deterrence and assurance.
Rationale. In identifying and assessing analytic “issues and factors that must be considered in seeking nuclear deterrence of adversaries and assurance of allies in the 21st century,” the committee noted that deterrence is largely a psychological concept and that sophisticated evaluation of the requisites for deterrence and assurance does not rest solely in the balance of military forces but must include insight into the mindset and decision making of the adversary or ally. An understanding of the impact of any action taken, including unintended consequences, must be central to the design of strategies for deterrence and assurance. Thus, as
the number of possible adversaries has grown, so has the need for actor/situation knowledge. The adoption of tailored deterrence results in a set of unique cases that must be considered.
Many analytic methods exist for exploring the nature and content of an individual’s or a collective’s decision making. These include various content analysis approaches, leadership profiling, qualitative and quantitative cognitive decision modeling, and representing an actor’s decision making in agent-based and simulation models. However, it is important to note that because the Air Force is not the only consumer of these analyses it should work to coordinate its needs with the U.S. government agencies that produce information about international leaders in the course of executing their assigned missions. To be skilled users and to generate comprehensive and feasible requirements the Air Force must develop and maintain expertise in this domain. The Air Force will not be solely responsible for production but should make use of Air Force capacity and joint assignments to augment efforts carried out by the primary agencies—that is, the Central Intelligence Agency and Defense Intelligence Agency within the intelligence community and U.S. Strategic Command and other military commands—and ensure that those efforts meet Air Force requirements. Actor and multiactor modeling support both planning and operations. Performed on a continuous basis, this modeling will provide the Air Force with analytic input appropriate to specific deterrence and assurance needs and better estimation of the likelihood of the success of an action based on the decision and risk propensities of adversaries and allies.
Implementation. With respect to implementation, the Air Force Research Laboratory, with input from the Air Force Global Strike Command and the Director for Studies and Analyses, Assessments and Lessons Learned, should be tasked to provide to the Chief of Staff of the Air Force a description of the current state of the Air Force’s analytic capabilities in actor and multiactor modeling and a recommended way ahead.
Finding 3-4. Tailored Deterrence. The methods of content analysis and leadership profiling in conjunction with other methods have the potential to help meet requirements of actor-specific knowledge for a strategy of tailored deterrence. An alliance among content analysis, leadership profiling, abstract modeling, and gaming and simulations as a suite of methods is possible in order to solve the complex problems associated with studying the decision-making dynamics of single groups and multiple autonomous actors as decision units.
Recommendation 3. The Air Force, working with its Service partners and the Department of Defense more generally, should pursue research on deterrence
and assurance with a coherent approach that involves content analysis, leadership profiling, abstract modeling, and gaming and simulations as a suite of methods. It should organize its investments in analytic and other activities accordingly.
Rationale. While a variety of methods to generate actor- and situation-specific knowledge are available to support Air Force planning for deterrence and assurance, the problem of looking ahead over a planning horizon of 20 years or more places additional demands on the need to understand potential adversaries and allies, being cognizant of the fact that today’s adversary may be tomorrow’s ally and that regional political–military situations may change, sometimes quickly. One approach to uncertainties such as these is to base analyses on a set of generic decision makers similar to but larger than the four categories described in Chapter 3 in order to explore the degree to which adversaries or allies are willing and able to receive different types of deterrence or assurance messages.
A multimethod approach to this research is necessary and should include many or all of the following methods in addition to others: game theory, human gaming, simulation, qualitative cognitive modeling, agent-based modeling, leadership profiling, and content analysis. An understanding of the variation across decision-making units and contexts will lead to more robust plans by allowing analysts and Air Force leadership to consider a range of motivations, behaviors, consequences, and situations. It would also provide the Air Force with a better appreciation of the implications of leadership changes in state and nonstate adversaries and allies.
Implementation. With respect to implementation, the recommended research deals with an interagency issue. The Office of the Secretary of Defense (OSD) should take the interagency lead, in collaboration with the Joint Chiefs of Staff and U.S. Strategic Command. The Assistant Chief of Staff for Strategic Deterrence and Nuclear Integration should be the focal point for the Air Force and should prepare an Air Force advocacy briefing for approval by the Chief of Staff of the Air Force. The briefing should identify relevant agencies inside and outside the Department of Defense. Once approved, it should then be taken to the Joint Chiefs of Staff, OSD, and the U.S. Strategic Command as a basis for OSD action in an interagency initiative.
Finding 2-1. Deep Uncertainty. Planning to support deterrence and assurance with both current operations and longer-term programs to organize, equip, and train is characterized by deep uncertainty, described more fully in Chapter 3. Nonetheless, methods exist for dealing with such uncertainties effectively, primarily by hedging and capabilities for adaptation (Hallegate et al., 2012).
Recommendation 4. The Air Force analytic community should pursue methods of understanding and incorporating the concept of deep uncertainty.
Rationale. Among the factors that contribute to deep uncertainty in deterrence and assurance planning are the lack of actor-specific/situation-specific knowledge, limited capacity to predict how messages will be interpreted, random events that may occur during crises or periods of tension, technological surprise, and the impact of fleet-wide capability degradation. Substantial progress has been made on how to plan under deep uncertainty, in which a set of techniques is employed including, for example, alternative cognitive models, test cases, and portfolio management.3 Use of such techniques is consistent with the analytic approach referred to as hedging, with an emphasis on developing strategies and plans that are flexible, adaptive, and robust.
Implementation. With respect to implementation, the Air Force Research Laboratory, coordinating with the Director for Studies and Analyses, Assessments and Lessons Learned and the Assistant Chief of Staff for Strategic Deterrence and Nuclear Integration, should identify current and anticipated analysis issues in which the concept of deep uncertainty is important and then recommend a program to develop and reinforce relevant knowledge and expertise in the analysis workforce.
Finding 3-4. Tailored Deterrence. The methods of content analysis and leadership profiling, in conjunction with other methods, have the potential to help meet requirements of actor-specific knowledge for a strategy of tailored deterrence. An alliance among content analysis, leadership profiling, abstract modeling, and gaming and simulations as a suite of methods is possible in order to solve the complex problems associated with studying the decision-making dynamics of single groups and multiple autonomous actors as decision units.
Recommendation 5. Air Force analysis supporting nuclear deterrence and assurance issues should draw from a suite of appropriate methods, including hybrid methods that combine and integrate different methods.
Rationale. In examining the need to solve and understand the decision-related dynamics of effective deterrence and assurance, the committee recognized the potential value of conducting analyses on the basis of a combined approach. Indi-
3 Davis (2012) is a broad review of RAND’s work on dealing with uncertainty. For further discussion of methods to support “robust decision making,” see Lempert et al. (2006).
vidual methods that might be included are content analysis, leadership profiling, abstract modeling, and gaming simulation. In many respects this is consistent with current and past practices for conducting deterrence analyses in which a wide range of methods have been used.
The notion of tailoring deterrence poses a set of analytic challenges in which certain attributes and factors will differ, perhaps significantly, across the range of adversaries, allies and regions. The committee believes that methods must be selected, adapted when necessary, and applied by analysts with two types of expertise: (1) sufficient facility with a variety of analytic methods to be able to distinguish appropriate use of each and (2) knowledge of the deterrence and assurance actors and processes relevant to the analysis task.
Hybrid methods involving the integration of expertise drawn from multiple disciplines, and the application of the analytic approaches of those disciplines in an integrated and novel way, were evident in the committee’s investigation and assessment of theory, applications, and research addressing decision-making units. In this domain and across the extent of nuclear deterrence and assurance analysis, hybrid methods offer greater breadth and accuracy because of the multiple disciplines involved. They may contribute to developing a wider range of insights.
Implementation. With respect to implementation, Assistant Chief of Staff for Strategic Deterrence and Nuclear Integration should coordinate with the Director for Studies and Analyses, Assessments and Lessons Learned to describe the unique attributes of deterrence and assurance analysis and the value of integrated hybrid approaches. Based on that description, the Director for Studies and Analyses, Assessments and Lessons Learned and the Air Force Education and Training Command should recommend a program to ensure that analysts have the knowledge and expertise required to bring appropriate hybrid approaches to bear on the analyses of deterrence and assurance issues.
Finding 3-1. Long-Term Career Development. Education and nurturing of experts in deterrence and assurance will not happen without a management plan to do so in the Air Force (and other services, particularly the Navy), partly in coordination with joint assignments but also bearing in mind longer-term career development and assuring adequate expertise (a Service responsibility).
Recommendations 6. The Air Force should maintain its cadre of career analytic professionals (both civilian and military) with expertise in nuclear deterrence and assurance strategy to improve Air Force support to Combatant Commanders’ planning and operations, since methods can inform, but never replace, the judgment of
expert analysts. This could be facilitated by specific treatment of analysts in Vector 5 of the Flight Plan for the Air Force Nuclear Enterprise (Air Force, 2013).
Rationale. In the course of its efforts the committee was briefed on current analyses related to deterrence and assurance and on various methods. These briefings led to a critical finding—namely, that the primary element in improving and sustaining high-quality analysis of deterrence and assurance is the continued education and nurturing of people, which should include frameworks, theory, and critical reasoning. The nation currently has a small pool of such analysts, who are very capable, but the pool is not large enough.
While the qualifications required of an analyst in the deterrence and assurance domain include a thorough understanding of the methods widely used throughout the military analysis enterprise, deterrence and assurance have attributes that require specialized expertise. Unfortunately the number of deterrence and assurance “experts” appears to be declining as personnel with experience dating back to the Cold War retire. It is possible for people to gain and retain knowledge necessary to conduct sophisticated deterrence and assurance analysis and planning without becoming career specialists. Such knowledge can be acquired through academic courses and experiential learning tailored to the 21st century security environment, yet deterrence and assurance analysis is currently underresourced. If the Air Force is to develop analytically based strategies and perspectives that are credible in the joint arena, and if Air Force leaders are to be prepared with reliable, informed reviews of alternative options considered in that arena, then the relevant analytic community must be adequately resourced.
Implementation. With respect to implementation, the Air Force Education and Training Command should be tasked, in coordination with Deputy Chief of Staff for Intelligence, Surveillance, and Reconnaissance; the Director for Studies and Analyses, Assessments and Lessons Learned; and the Assistant Chief of Staff for Strategic Deterrence and Nuclear Integration to provide a way-ahead briefing for the Chief of Staff of the Air Force.
In the process of preparing for this report the committee was given the opportunity to interact with analysts currently engaged in planning for deterrence and assurance and in supporting deterrence and assurance missions. Given the resources available, these people are doing a remarkable job. The current community makes excellent use of classical analytic methods. It has begun to address the challenge of actor-specific and situation-specific knowledge but is not resourced to exploit advances in these disciplines. It has taken only preliminary steps to address
deep uncertainty and has limited capacity for the research necessary to develop new deterrence and assurance concepts, theories, and strategies.
The report’s recommendations respond to observed shortfalls and identified opportunities. The development of an comprehensive analysis plan will provide a framework in which to develop strategies for organizing, equipping, and training Air Force personnel. It will allow the Air Force to avoid overreliance on tools that are most appropriate for physics or engineering questions and contribute to the adoption of well-hedged, robust, and adaptive strategies. Increasing the Air Force analytic capacity to understand and utilize human and human organization factors will inform the region by region contributions the Air Force must make to tailored deterrence, facilitate earlier recognition of potential failure, expand understanding of the risk-taking behavior of adversaries and allies as well as allowing more specific tailoring of the Air Force response to potential deterrence or assurance needs. Advocacy of research to develop a generalized understanding of leadership, decision making, and behavior dynamics related to deterrence and assurance will improve the robustness of longer-term planning, provide a region by region baseline deterrence environment and assist in responding to leadership changes in adversaries or allies. Incorporating deep uncertainty into Air Force analyses supporting strategic planning will reduce the risk of being unprepared for unforeseen situations, increase awareness of the value of hedging in the face of uncertainty, and provide an approach to identifying and dealing with unintended consequences. All of these recommendations rely on the cadre of Air Force career analytic professionals. These professionals ensure that the Air Force has credible and analytically based perspectives in the joint arena, and that Air Force leadership is provided with informed and reliable information to support selection of an Air Force strategy, plans, and materiel.
Air Force. 2013. Flight Plan for the Air Force Nuclear Enterprise. Washington, D.C. http://www.af.mil/Portals/1/documents/news/FlightPlanfortheAirForceNuclearEnterprise.pdf.
Davis, P.K. 2012. Lessons from RAND’s Work on Planning under Uncertainty for National Security. RAND Corp., Santa Monica, Calif.
Hallegatte, S., A. Shah, R. Lempert, C. Brown, and S. Gill. 2012. “Investment Decision Making Under Deep Uncertainty: Application to Climate Change.” World Bank Policy Research Paper 6193. http://econ.worldbank.org.
Lempert, R.J., D G. Groves, S.W. Popper, and S.C. Bankes. 2006. A general, analytic method for generating robust strategies and narrative scenarios. Management Science 52(4):514-528.
This page intentionally left blank.