Click here to access a PDF of this report.
Mar 27, 2019
AUTHOR: Henry Sokolski
Improving the Role of Intelligence in Counterproliferation Policymaking: Report of the “Speaking Truth to Nonproliferation Project,” 2018
By Henry Sokolski
March 2019
This report is the culmination of a two-year project sponsored by the Nonproliferation Policy Education Center, which engaged more than 50 senior, retired and serving policymakers, intelligence officers, and top academic national security analysts. Its findings are based on hours of group discussions and private conversations that helped develop new primary histories of eight nuclear proliferation cases: India, Pakistan, Israel, Taiwan, South Korea, Libya, and an Argentine and a separate South African nuclear rocket case.
Each history was prepared by an academic historian and was based on open sources. Former officials who had direct roles in these cases then critiqued these accounts. Additional private interviews were conducted with participants to fill in historical gaps. The purpose of the case studies was to identify when and how intelligence shaped or prompted nonproliferation policy actions and, if it did not, why. This set of historical conclusions prompted a more general discussion of how policy and intelligence officials might improve their collaboration to prevent and curb further nuclear proliferation and how academics might contribute by enhancing their treatment of such issues.
The project addressed three broad, related questions:
- How can the role of intelligence in the making of nonproliferation policy be improved?
- How can the nonproliferation agenda get the priority it deserves?
- How can the nonproliferation community be sustained and strengthened?
Improving the Role of Intelligence in Policymaking
The first step in answering this question is to identify shortcomings in the nonproliferation intelligence process. One of the major findings of this study is that when effective nuclear nonproliferation policy actions are absent, it is rarely due to a lack of timely intelligence. A key reason actions are not taken is that the relationship between intelligence and policy is not simple, automatic or linear. Conventional wisdom has it that “actionable” intelligence is the critical ingredient most lacking and most needed to enable sound, timely nonproliferation action. In fact, the cases this project examined suggest otherwise.
At least as often as not, even when timely, repeated, and persuasive intelligence on nuclear proliferation was available, policymakers deflected, ignored, or downplayed it in favor of pursuing what they viewed as more urgent political, military, economic, or diplomatic objectives. This was most evident in the Israel case where the Intelligence Community IC had considerable evidence
- that Israelis had stolen 100 to 300 pounds of weapons-grade uranium from a nuclear facility in Pennsylvania;
- that Israel had tested a nuclear device in violation of the Limited Nuclear Threshold Test Ban Treaty, which Israel had signed; and
- that Israel had repeatedly deceived US nuclear inspectors.
Pakistan’s nuclear weapons progress was also well understood by US intelligence, yet, again, US policymakers were reluctant to act. The problem in these cases was not inadequate intelligence but the existence of other policy priorities that made policymakers reluctant or unwilling to act decisively against proliferation.
When policymakers were unwilling to act in the face of persuasive intelligence that a nuclear proliferation concern existed, the IC was reticent to push the policy community to act. There are understandable reasons for this. The IC wishes to preserve its independence from the policy process. It also is appropriately concerned about preserving its sources of intelligence and amassing more information, whereas policymakers are more inclined to use and share intelligence selectively to achieve specific policy goals, or to avoid upsetting policy goals. Again, the historical case of Pakistan is relevant. In this case, the Intelligence Community’s preservation of its sources regarding A. Q. Khan certainly complicated early public discussion of this case.
Ideally, there is a natural give and take in discussions of when to use perishable intelligence that would benefit from close intelligence and policy interaction and understanding of each other’s concerns and priorities but in practice there is tension between the two communities about the protection and uses of intelligence. The IC’s diffidence in advocating forcefully on behalf of the facts makes it easier for policymakers to ignore or override evidence when there are reasons to do so—the IC rarely plays an effective corrective role.
In some cases, such as North Korea, Israel, India, and Pakistan (once their nuclear weapons programs were fully underway), policymakers concluded, rightly or wrongly, that they did not have any strong policy levers with which to block these nuclear weapons efforts. In these cases, it was not a lack of good intelligence that inhibited action, but the policymakers conclusion that no attractive action was possible.
Thus, the historical record suggests that inadequate intelligence is not the principal cause of ineffective nonproliferation policies.Rather, when policymakers pursue other priorities that they value more highly or when they believe they possess no good policy options, persuasive intelligence will not prompt vigorous nonproliferation actions. If the policymaker does not believe taking clear, short-term risks is warranted to avoid larger, long-term risks, inaction is most likely. The reticence of the IC to push the policymaker to act or to clarify the long-term risks facilitates such outcomes.
The historical cases examined in this project also cast some light on when intelligence does cause policymakers to act decisively. A key factor is the wider political context within which a case is unfolding. In the case of South Korea in the 1970s, the State Department was still embarrassed by Congress’s discovery that, despite State’s denial, US heavy water had been used by India to make its first bomb in 1974. This revelation led, over State’s objections, to the passage of the Nuclear Nonproliferation Act of 1978.
With the new law in force, the executive branch didn’t want to get caught flat-footed again. When evidence emerged in the late 1970s of South Korea’s intent to make its own nuclear weapons, the White House acted on the intelligence almost immediately and pressured South Korea to abandon the project. In the case of Taiwan, President Nixon’s recognition of the Peoples’ Republic of China and the passage of the Taiwan Relations Act made Washington policymakers particularly sensitive to any indication Taiwan might go nuclear. Again, the White House acted early, upon the very first indications, and with good results.
In the case of Libya and the South African and Argentine nuclear-capable rocket cases, which this project examined, several factors disposed policymakers to act. These included the failure to prevent Iraq’s acquisition of rockettechnology, congressional hearings and passage of the Missile Technology Control Act, and the transition of the governments in Argentina and South Africa to popular self-rule. In the Argentine and South African rocket cases, moreover, there were long-term trade and military assistance bans in place and several key US government departments acting in close operational cooperation with the IC. In contrast, more than once, intelligence relating to Israel’s, Pakistan’s, and North Korea’s nuclear weapons-related activities was downplayed, deflected, or ignored by US policymakers as untimely, unconvincing, or unwanted, as the context and overriding policy concerns then were quite different from the South Korea, Taiwan, Libya, and rocket cases.
More and better intelligence is always desirable but the findings of this project suggest that the main deficiencies in the intelligence-nonproliferation nexus lie elsewhere. The fundamental problem has to do with how nonproliferation intelligence is used (or not used) by policymakers and that, in turn, raises issues about the relationship between the IC and policymaking and the need to focus on long-term threat assessments and policy planning.
Operationally, these findings recommend a more conscious effort to increase collaboration between intelligence and policy making officials by:
- Reviewing intelligence tradecraft guidelines regarding how and when intelligence officers and policymakers should interact.
The goal would be to identify worrisome proliferation trends at the earliest possible point and to formulate and execute practical counteractions. Government officials should be encouraged to make judgments about a proliferation-related circumstance at a relatively inchoate stage so they can take modest, early actions—before it becomes “too hard” to act (i.e., when only radical, high-risk actions are feasible). This will require both the intelligence and policy communities to tolerate ambiguities, express uncertainties, and, when appropriate, to take modest risks. The costs of failure are high enough to encourage such early efforts.
- Expanding long-term, operational nonproliferation collaboration and planning between the intelligence and policy making communities.
Useful lessons can be learned about assessing and neutralizing proliferation from current ongoing efforts to interdict proliferation activities and to lengthen nuclear proliferation timelines. Could the activities of the Strategic Interdiction Group of the National Counterproliferation Center and other integrated targeting and joint policy-analyst and operator cells be scaled up to deal with the broad array of global proliferation concerns? Could this be done without squelching healthy, competing, different interpretations of the intelligence? What new analytical, collection, and operational tools would they need? Also, what can be learned from earlier, failed efforts at such integration, e.g., the Strategic Assessments Group that had existed in CIA’s Directorate of Intelligence? These questions deserve serious answers.
- Encouraging increased joint instruction of policymakers and intelligence officers.
The instruction and training of intelligence and policymaking officials in the field of nuclear nonproliferation would benefit from an analysis of case histories; both those where intelligence was used effectively to encourage and shape nonproliferation policy actions and those where it was not so used. It would be useful to determine how current government-sponsored education and training might better inform both intelligence and policy making staffs in the subject matter. Ideally, joint training and education of intelligence and policymaking staff should occur at entry and mid-level so that relationships are built before crises unfold.
Gaining the Priority the Nonproliferation Agenda Deserves?
The use of intelligence will not be improved so long as policymakers are inclined to sacrifice nonproliferation concerns to advance other priorities. If policymakers valued achieving nonproliferation more highly, they would be less likely to resist or ignore the implications of proliferation intelligence. The historical record suggests, however, that willingness to prevent nuclear proliferation and, therefore to take prudent early action, has rarely been very high.
A Declining Concern?
One of the broad conclusions of this project is that the priority attached to nonproliferation has recently declined at a point when longterm planning and modest early actions would be most useful. During the 1950, 60s, 70s, 80s, and 90s, policymakers and the public feared the further spread of nuclear weapons to additional states far more than they do today. Policymakers also were much more willing to take risks to prevent such proliferation. They viewed acquisition of nuclear weapons by France, Israel, China, India, Pakistan, and North Korea, and the attempted acquisition of nuclear weapons by Iraq, with considerable apprehension.
More recently, though, these concerns appear to have relaxed. Novel, counterintuitive academic notions of nuclear stability—that more nuclear powers may be better, that nuclear proliferation has been inconsequential and its risks overblown —have enjoyed a certain following. Most recently, President Trump ruminated that Japanese and South Korean nuclear proliferation might not just be inevitable, but beneficial and that America’s interests might be best be served by allowing or encouraging their acquisition of nuclear arms. The British Foreign minister Boris Johnson, meanwhile, has argued that peace in the Middle East and Gulf would be best served by letting Iran acquire nuclear weapons.
Policy Fatigue
In addition, policy fatigue has set in for reasons that cannot be easily dismissed. Two to three decades ago, there were a number of worrisome nuclear proliferation cases—Pakistan, India, North Korea, Iraq, Iran, Libya. Now, with the fall of Qadhafi and Saddam, the acceptance of India and Pakistan’s nuclear status, the only clear-cut cases of nuclear worry are Iran and North Korea. Unfortunately, their nuclear behavior cannot be easily managed. More important, for most news watchers, there do not seem now to be many other states on the cusp of acquiring nuclear arms. Egypt, Turkey, Algeria, and the United Arab Emirates are uncertain cases that, at best, seem years away from getting nuclear arms (unless they change course and engage in crash programs that would be difficult to hide). Saudi Arabia might get bombs from Pakistan, but most pundits discount this possibility as an immediate worry. As for Japan and South Korea, they both may have the way but still seem distant from exercising the will to acquire their own nuclear weapons, unless their faith in US security guarantees dramatically falters.
Emerging Proliferation Worries
Are there, then, no nuclear proliferation problems worth working? Actually, just the opposite. The proliferation problems we face and can most readily address, however, are not so much (as it was in the past) specific trouble states, but emerging trends, which if not slowed or blocked will produce a bow wave of future nuclear-armed states and possible use within five to 15 years. Among the trends this project identified were:
The growing interest in Pakistan and Russia, and possibly in China and India as well, in the early use of nuclear arms against a variety of conventional threats. Russia and Pakistan both contend that a single or a few nuclear weapons can and should be used as a firebreak against a variety of conventional military attacks.If China and India should adopt similar views, the need to acquire not just hundreds, but possibly thousands of new, additional warheads would push new production and prompt possible emulation globally.
An increasing number of new scenarios for actual or threatened nuclear use by Russia, China, India, Pakistan, North Korea, and Israel.
The likelihood that if the next use of nuclear weapons is seen as having militarily benefited the first user, many other countries might want nuclear weapons and be willing to use them first even if this risked major nuclear exchanges that would be disastrous.
The possible ramp up and stockpiling of fissile materials production in China, Japan, and South Korea that could enable these states to break out with large numbers of nuclear weapons.
The possible emulation of Iran’s enrichment efforts by other states once Iran is clear of the restrictions of the Iran nuclear deal.
The possible emulation of Brazil’s naval reactor and enrichment program by South Korea, Iran, Pakistan and others.
The increasing dissemination of nuclear weapons design and production- elated technologies, including both new (3-D printing and CAD CAM production techniques) and old (weapons code and design information).
None of these trends has caught the immediate attention of the public or much of the US government.
Reflecting this lack of urgency, congressional oversight is much reduced from what it once was. The Senate has disbanded the Governmental Affairs Subcommittee on Nuclear Proliferation, which had been an important watchdog in the area. Nor are there any longer annual IC or routine congressional staff reports on nuclear proliferation developments.
What nuclear nonproliferation oversight there is is far less focusedon identifying and slowing worrisome longterm trends than on reacting to all too advanced nuclear proliferation crises, for example:
- What to do about North Korea’s latest nuclear or nuclear-capable missile tests;
- How best to constrain Iran’s nuclear and long-range missile programs;
- How to get Syria to honor its International Atomic Energy Agency inspection obligations after Israel bombed a suspect production reactor, etc.
All of this is symptomatic of policy fatigue as well as a growing disinclination to plan. No success against nuclear proliferation, however, is likely, or even possible, unless government officials are willing and able to identify futures that have clear proper nouns; outcomes they wish to avoid; and happy endings they prefer to secure and against which they can plan. Each of these specific futures should be seen as master narratives guiding US policy. As a negative but immediately accessible point of departure, these narratives might begin with the notion that we should, at a minimum, avoid future Irans, Iraqs, Syrias, and North Koreas. The narratives should also tackle identifying worrisome long-term proliferation trends and how best to stem them.
Operationally, this suggests the need to:
- Develop a consensus on the nonproliferation futures the United States desires and nuclear proliferation futures we wish to avoid in the development of key US policy and intelligence requirements.
The starting point here should be to call for a clearer description of these security futures in US Defense Department review and guidance efforts. These futures narratives should, in turn, drive more of the IC’s development of its National Intelligence Priorities and a structured, routine liaison with mid- and senior-level policymakers. This effort should be more normative in character, aimed at identifying where Washington wants to get to rather than merely passive analysis. The fruits of and progress in institutionalizing this collaboration (perhaps in the National Counterproliferation Center, a revitalized Strategic Assessment Group, or similar body) should, in turn, be a topic for congressional oversight by the intelligence, foreign affairs, military and nuclear proliferation-related committees.
- Revitalize Congressional oversight.
The Senate Governmental Affairs Committee might reestablish its Subcommittee on Energy and Nuclear Proliferation. Unlike the House, which has a subcommittee of the Foreign Affairs Committee dedicated in part to strategic weapons proliferation concerns, there is no Senate committee of any sort dedicated to overseeing the management of US nonproliferation efforts. These dedicated “proliferation” committees should seek to hold joint hearings with the relevant intelligence and armed services committees to clarify the executive branch’s own views of what proliferation threats deserve the most attention, why and how. Special attention should be paid not just to specific countries and crises, but to troubling longterm trends, such as growing interest in “peaceful” production and stockpiling of nuclear explosive plutonium and the buildup of enrichment capacity in trouble regions like East Asia and negative trends in foreign countries’ nuclear weapons use doctrines. Congress also should request and receive routine classified briefs from the executive and produce routine assessments of their own. These, in fact, are required by law.
- Ask each new administration to report publicly what the security risks of further nuclear proliferation in relation to possible use might be and how nuclear use by others might jeopardize the interests of the United States, international security, and the average American citizen.
- Game the most worrisome nuclear proliferation and nuclear weapon use scenarios to expand the consciousness of senior policymakers and intelligence officers at the start of any new administration.
This was done at the most senior levels at the start of the Reagan administration with useful results. Congress, which has a continuity of government responsibility if wars eliminate much of our country’s executive branch, has a particular responsibility to understand how likely such wars might be and how they might best be avoided. Gaming of such crises for both senior executive officials and congressional members needs to be encouraged again.
Sustaining and Strengthening the Nonproliferation Community
If a key to more effective use of intelligence is the elevation of nonproliferation on the policy agenda so that it competes more successfully against other priorities, then it is essential to have a robust and well trained nonproliferation community able to make the case for the nonproliferation cause. In order to ensure that the requisite expertise is sustained, it is essential that new generations of nonproliferation experts are recruited and are well trained. It is not clear from the deliberations of this project that this essential task is being undertaken. On the contrary, it appears that efforts to expand academic instruction and analysis of nuclear proliferation cases have yet to significantly improve US nonproliferation policy and students’ preparation for possible public service.
The problem is that the cultures and norms of academia diverge from policy needs, and it is difficult for academics to step outside of the mold. Certainly, the demands of peer review placed on academics are difficult to overestimate, especially for political science and the study of public policy. These fields place a premium on the development of novel and general conceptual frameworks. Development of such frameworks is imperative to secure academic prominence, but they leave government actors mystified.
Similarly, political science scholarship prizes “proven” sources of information—primary documents and citations from prestigious published works. This reduces the incentive to develop new sources or to examine the existing sources critically. Whenever possible, general statistical treatment of aggregate data is also encouraged (although significant nuclear proliferation issues tend not to produce large aggregates). All of this is done in the name of promoting a more “scientific” approach to political analysis.
In the profession’s zeal to be more scientific, there has been a not-sosubtle deemphasis of narratives or political stories—the lifeblood of political discourse, insight, and instruction. First-hand accounts from prime actors often get downplayed or passed by political scientists as being merely “anecdotal.” Academics tend to place more confidence in citations from newspapers, journals, and especially articles by other academics. This preference, of course, easily leads academics to rely only on what they believe to be “safe” sources, which further limits the scope of their investigation of the real-life political world.
As a result, academics have yet to fully exploit some of the richest and most powerful sources informing public policy—the narratives of those who have acted in the very cases of greatest interest. This undermines the quality and power of the analysis political scientists can tap. Graham Allison’s classic work, The Essence of Decision, became powerful political science in no small part because of his willingness to speak through narratives on the Cuban missile crisis, narratives that were fresh and new at the time. As was noted by the former editor of the prestigious journal International Security at one of the project plenary meetings, there is a clear need now for more, not less oral history, as more and more documents relating to nuclear policy are declassified. These documents can only be made meaningful by interviewing those involved in their development or implementation.
Deemphasizing political narratives also affects the character of policy instruction. Thus, education in the field of nuclear proliferation has become increasingly stylized with a heavy emphasis on novel theories and concepts to explain, for example, why nations proliferate and the extent to which proliferation exerts a “stabilizing effect.” But it is less focused on how best to prevent nuclear proliferation and what has or has not worked and why. The latter, it goes without saying, is essential practical preparation for public service.
Joint Professional Military Education. (JPME) may be a key source of demand or at least receptivity for academic programs (a) for expanded strategic thinking about proliferation and (b) for mixing heterogeneous audiences of academics and operators.a JPME emphasizes “not what to think but how to think,” and critical thinking. Therefore, JPME, especially at the most senior military service schools, is nominally open to issues (such as nuclear proliferation) of the highest importance to grand strategy and that have been out of fashion since the end of the Cold War and the rise of global terrorism as the leading threat. Unfortunately, today, the challenges that nuclear proliferation could pose to US national security interests and that of its allies over the next five to 15 years and what might be done to cope or reverse these threats has yet to be a featured, prominent topic. This is a mistake.
How can nonproliferation training be improved and made more relevant? The experience of this project suggests several steps:
- Bring more of “town” to “gown” by encouraging visits to academic centers from retired intelligence and policy making officials who can talk in the first-person about cases identified with proper nouns (e.g., Israel, Pakistan, North Korea, etc.).
Their “war stories” may not be the complete analysis but can vividly expose the student and professor to the power of various narratives, thereby enriching the resources from which both can ponder how sound existing policy conceptual frameworks might be. These stories also can help determine how more amendable nuclear proliferation problems might be to policy action more generally. Departing senior and mid-level officials should be encouraged by their agencies to share which two or three cases they were most involved with and to keep a tally on which departing official is most conversant about which case and making use of this information in securing oral histories both classified and unclassified.
- Increase the number of year-long internships for promising graduate students and faculty in policy making, legislative and intelligence staff positions to familiarize the student and teacher with line-staff operations in preventing nuclear proliferation.
Additionally fellowships for graduate students (including students with all but dissertation status) interested in public service (rather than teaching) should be sponsored by nongovernmental organizations to help them find full-time work on Capitol Hill and support them in such posts for the first year.
More also needs to be done to reform the existing security system, which increasingly is a barrier both to students wishing to enter public service and to understanding of the past.
- Create military/civilian mixed seminars, conferences and briefings, contributions to individual classes, development of full two to three week courses, and summer institutes when senior military service and intelligence schools focus on longterm national security problems that should include nuclear proliferation.
The National Defense University currently works with some 600 O-5 and O-6 military officers, almost one fifth of whom are internationals. The goal of their instruction is to think broadly and outside their comfort zone. The officers attend these schools because their service has marked them for advancement, and many of them are receptive to “charging their intellectual batteries” and figuring out what they want to do in their futures. It would be useful to present nuclear proliferation to them as a field of high national need and intellectual and operational challenge.
Yet another model worth emulating is the two-week summer seminars series on national security held at several campuses in the 1970s. Each summer the seminar enlisted promising new PhDs. In the past, this included students that later went on to become deputy secretary of defense, undersecretary of defense for policy, chairman of the National Intelligence Council, senior members of the National Security Council, directors of prestigious graduate school national security programs think tank scholars and senior industry officials. The National Defense University, the Naval War College, the Air War College, the Army War College, and the Naval Postgraduate School all could and should serve as venues for such educational efforts. Finally, the Sherman Kent School for Intelligence Analysis in CIA, the National Intelligence University, and the Center for the Study of Intelligence should be tapped to create similar educational opportunities.