The Current Situation
Despite the deep reforms of the U.S. intelligence community (IC) carried out after 9/11, including the creation of the Director for National Intelligence (DNI) and the National Combatting Terrorism Center (NCTC), there is widespread agreement that more remains to be done. This is not a new thought. Before the ink was dry on the 2004 Intelligence Reform and Terrorism Prevention Act (IRTPA), there were warnings from insiders as well as outside experts that the law had not fully dealt with the challenges facing the IC. The critics pointed to the anomalous position of the DNI, a neglect of strategic analysis, accusations of the politicization of intelligence, and the difficulties that the IC has with failure, learning, and adaptation, as signs that all was not well within the IC.
A series of stumbles over the past seven years have given credence to the red flags raised by these experts. In quick succession, the IC failed to predict the so-called Arab Spring, the resurgence of al-Qaeda, the adventurism of Putin, the aggressiveness of China, and a number of terrorist attacks on the U.S., from the Detroit “underwear bomber” to the San Bernardino massacre. The intelligence failures surrounding the Arab Spring were especially important, since the IC had not understood the implications of an entire series of seismic shifts in the strategic landscape, suggesting that there are serious problems with the analytical side of the community. It is noteworthy that the community’s intelligence collection—clandestine and open source—appears not to have focused on the deeper questions of regime stability and the underlying causes for the Arab Spring. In this Backgrounder, the focus will remain on analysis and improvements to the analytic aspects of the IC.
The need for reform is made more urgent by the increasingly complex national security environment that the United States is facing, one that is dominated by violent non-state Islamic extremists; anti-status-quo states China, Russia, and Iran; and a weakened or alienated set of allies and partners around the globe on which the U.S. must depend for support in dealing with these threats and challenges. There are also well-grounded fears that this situation will be the “new normal” for at least the next decade. The IC must, therefore, become the kind of federated enterprise—organizationally, analytically, and culturally—that can constantly learn from, and adapt to, this highly volatile environment in order to better support decision makers.
Defining the Problem Set
Before presenting potential solutions to the serious challenges that exist within the intelligence community, it is important to carefully describe the specific structural, analytical, and cultural issues that must be reformed if the community is to be made ready to deal with the new security environment.
Structural Challenges. There are two main structural challenges confronting the IC, each of which seems to contradict the other.
A Powerless DNI and Bureaucratic Bloat: Too Much Centralization or Too Little? The current organizational structure of the IC is dominated by 17 diverse agencies, with the DNI as the nominal head. There are also a few mission-organized units, such as the NCTC, that bring together intelligence professionals from all the agencies with their own organically grown workforce to address specific areas of concern. Criticisms of this structure have come from two completely different directions, with some experts arguing that the DNI has too little power, while others argue that there is already too much centralization and bureaucracy in the IC. These two criticisms are not mutually exclusive.
A number of observers and experts have noted that the Director of National Intelligence lacks any real control over the IC. Unlike the Secretary of State or Secretary of Defense, the DNI does not oversee his entire domain, nor do the various agencies, other than the CIA, report directly to him. The DNI also cannot dictate to the heads of the CIA or Defense Intelligence Agency (DIA) in the way that the Secretary of Defense, for instance, can issue orders to combatant commanders. This is an authorities issue, bureaucratic problem, and personality challenge combined. The IRTPA sets out the position of DNI as one that would coordinate and encourage cooperation, rather than a “command” position like the Secretary of Defense. And while the Director of Central Intelligence should report directly to the DNI, the powerful and independent-minded leadership and bureaucracy of the CIA reportedly resented the intrusion of another layer of administration into their affairs and have fought against DNI attempts to assert his legal authority. In addition, the personalities of the DNI and President matter: A determined leader might be able to use his authorities to enforce more unification of effort but only if supported by the President.
The problems created by the anomalous position of the DNI are legion: There is no central hub that can enforce change throughout the IC, make the entire community more adaptable, or root out and fire bad managers and leadership. The DNI must be prepared to exercise the firing authority often in tandem with the head of the Department where the poor performance is observed. If the President wants to enquire about issues handled by multiple agencies, the President can turn to the DNI, but obtaining an answer requires strong White House backing for the DNI as opposed to the DNI resting on relatively weak authorities in law or policy to push IC elements for answers.
For some observers, this is a feature, not a bug, since they worry about too much centralization within the IC rather than too little. The intelligence community’s bureaucracy, in their view, presents a serious challenge to sound and independent analysis and the creation of the DNI position has only compounded this problem. These experts point out that there are too many layers of management within the CIA in particular, creating unnecessary bureaucratic hurdles for intelligence professionals and separating the writers of intelligence products from their customers by as many as eight distinct managerial levels. In their opinion, adaptation and learning must come from multiple directions, and the relatively loose control of the DNI allows each agency to find the best way forward for dealing with the myriad challenges confronting the U.S.
Organized for Individual Missions, Not Cooperation. Despite the relative weakness of the DNI, there is no doubt that the creation of this office has facilitated better, though not perfect, collaboration and cooperation between the other 16 agencies that make up the IC. At the same time, the new bureaucratic layer did not deal with the disparate organizational structures within each of the intelligence community’s constituent parts. Every agency creates subdivisions according to what its leadership feels best fits its own mission, rather than working together to create an overall structure that would facilitate cooperation throughout the community. Even the recent reorganization by John Brennan of the CIA, while integrating the analytical and operational sides of that agency, does not seem to take into consideration the need to cooperate with other institutions. The result is that there are only a few issues, such as counterterrorism, where cross-agency cooperation is fully supported by organizational design and structure.
Analytical Challenges. Perhaps the most serious problems confronting the IC are those related to analysis—the core function for most of the community. The challenges come from many different directions and include pressures to maintain consensus, a failure to encourage the development of real subject-matter expertise, a neglect of open-source materials, and a strategic analysis gap.
Lack of Diversity in Analysis. The widely studied problems of groupthink, confirmation biases, and cultural norms have all had their role in limiting how intelligence analysts have perceived and described problem sets. The group-style of writing all products makes these problems a critical threat to dissenting opinions and alternative viewpoints from the lowest to the highest levels of the IC. There are also a few additional issues, the result of the specific circumstances under which intelligence professionals must work, that might impose greater uniformity on analysis than the data supports. The ongoing investigation into allegations of intelligence-analysis manipulation at Central Command’s intelligence shop, for instance, has exposed the problem of a preferred “narrative” that analysts were pressured to maintain. It has even been alleged that managers altered finished products, and that analysts’ jobs were threatened if the analysts did not toe the party line. The problem of potential politicization, then, whether directly by an Administration or by managers eager to please their political leadership, remains a serious one. Another IC-specific challenge is created by the need for consensus when writing the intelligence community’s premier products, such as a National Intelligence Estimate (NIE). This might lead to a less than full exposure and discussion of objections to received wisdom or the downplaying of alternative viewpoints.
Failure to Encourage Expertise. Personnel and promotion incentives within the IC encourage generalization rather than specialization and can hinder the development of real area, or subject-matter, expertise. There are, in addition, no incentives to discourage analysts from changing subjects frequently, preventing them from ever having enough time and study to become experts on any one topic. Richard Russell describes the almost palpable disdain held by managers at the CIA for PhDs, and the lack of encouragement to hire new analysts with doctorates. While there is certainly a need for generalists in the IC, the current career expectations discourage the formation of internal experts and force agencies to rely on external subject-matter experts. This is a serious issue for the new challenges that the U.S. faces in little-studied regions of the Muslim-majority world, since there are relatively few such subject-matter experts for the IC to rely upon and many of them, for ethical or political reasons, will not work with the IC.
An Open Source-Classified Divide. The IC naturally prizes classified intelligence and prioritizes it over other sorts of information. After all, if intelligence professionals simply read newspapers or other published sources to gather information, how would they differ from outside experts or reporters? But this natural tendency is unhelpful when it leads to neglect, or even downplaying, of the value of open-source materials. This is especially troubling when dealing with non-state actors like al-Qaeda and ISIS, which put out a great deal of public information about their plans, objectives, and even failings. The terrorists’ decisions to use Twitter, Telegram, and other social media, as well as to publish biographies of members, demonstrate that open source can also be a rich source of detailed information about jihadist groups, one that the IC, by discounting unclassified material, may be failing to fully exploit.
Strategic Thinking Lapses. The IC, driven by the demands of its customers as well as the community’s own hierarchy of importance, engages mostly in short-term, tactical analysis and thinking, rather than strategic and long-term consideration. Although it seems reasonable that current intelligence must be the focus of the majority of the IC’s analytical work, neglecting longer-term strategic analysis can create key vulnerabilities by de-emphasizing strategic warning, forecasting, and prediction. One exception to this general problem is Global Trends, produced by the National Intelligence Council every four years. However, this public document, which analyzes likely developments in national security issues over the succeeding two decades, is apparently a stand-alone product that might have little influence on analysis within the IC. No publicly released NIE, for instance, has ever cited Global Trends in its analysis or conclusions. In addition, a look at Global Trends documents since the publication’s inception in 1997 shows a clear focus on current challenges and a bias toward straight-line assumptions about future developments.
Cultural Challenges. Cultural problems can seem even more intractable than analytical challenges, since they are “baked into” the very essence and structure of institutions. In the IC, two cultural issues in particular have prevented the community from being an adaptable and learning organization.
Fear of Failure. With the caveat that there have indeed been actual intelligence failures, it is also true that the intelligence community is often the whipping boy for national security policy mistakes. It is far too easy for policymakers and political leadership to blame errors of judgment and policy on the IC rather than take responsibility for their own failings. The result is that the entire IC, and the CIA in particular, has developed a well-founded fear of failure and, consequently, of being blamed by the public and policymakers. This can lead to a number of serious problems, including risk aversion and an analytical tendency to overcompensate for previous errors by going too far in the other direction with subsequent assessments. In addition, the CIA’s response in 2013 and 2014 to the Senate detainee report shows that CIA leadership will go to great lengths to avoid being “stuck” with the blame for actions that are later called into question or outright discredited, a tendency perhaps sparked by the Church Committee process of the 1970s.
Problems with Learning. Mistakes should be opportunities for learning, but—despite some positive examples of learning and adaptation, best exemplified by George Tenet’s reform efforts during the 1990s and the lengthy process undertaken after 9/11—the IC as a whole has great difficulty looking at failures objectively and making adjustments to prevent mistakes in the future. It is also unclear whether there is an institutionalized learning process for the IC, one that would capture lessons learned from both successes and failures and pass them on to future generations of operators and analysts.
A Strategic Vision for the Intelligence Community
This set of challenges requires a series of deep reforms, not just one or two small tweaks to the community. Overall, the basic concept for reform is to help create a more flexible and adaptable institution that can evolve with the changing national security situation rather than requiring serious reforms every decade or so. This Backgrounder offers proposals to do this by creating stronger leadership for the IC while streamlining the bureaucracy, and by encouraging a good balance of cooperation, diversity of views, and learning by intelligence professionals throughout the entire community.
Meeting the Structural Challenges.
Overhauling the Bureaucracy. While the ODNI has improved cooperation and collaboration across the community, the DNI is relatively weak and must be given the power to actually control the IC, including the CIA. At the same time, the bureaucratic layers between analysts and their customers need to be reduced to allow more analytical freedom and diversity, and to prevent the massaging of information to fit a preferred narrative or achieve consensus by going to the lowest common denominator in making hard judgment calls.
- To empower the DNI, the equivalent of combatant commands that would be under the DNI’s direct control and supervision should be created. The basis for these “commands” would be the National Intelligence Council (NIC) and its members and National Intelligence Officers (NIO), both of which are already directly tied to the DNI. Under this proposal, NIOs would be “super-empowered” to oversee both collection and analysis on their issues across the IC. To ensure proper coordination and control of these matters, NIOs would meet regularly as the NIC, with the DNI as chair.
- Since there will be a new office in each agency tasked with strategic thinking and planning, the NIC will also need an office organically embedded within it that oversees foresight, strategic warning, and strategic thinking and planning throughout the IC. Given the NIC’s collection and analysis oversight function as well as its need for this new office, the council will require a sizeable staff associated with it, drawn from the very best analysts and collectors across the IC. The de facto mission managers of today who evaluate and advocate resource allocations to specific country or transnational portfolios would move under the NIC.
- The heads of the various agencies would then become more like the Joint Chiefs of Staff (JCS) than their current capacity, charged with manpower, building the futures of their agencies, research, and training and education for their workforces. They would also meet regularly with the DNI, who in this capacity would be acting somewhat like the Chairman of the JCS. Potential modifications to the Intelligence and Terrorism Prevention Act would need to be made to clarify the reporting relationship to the DNI.
- To ensure that the new NIOs would have good supervision of their regions and issues, there should be an IC-wide reorganization that structures the highest-level offices uniformly across the community. There would necessarily be some offices and bureaus that would exist in certain agencies and not in others, but the top-level organization of all the analytical pieces of the IC would match each other, as would the top-level organization of the collection parts.
- Finally, to ensure proper political oversight of this entire new structure, the NIC chairman would need to follow the same appointment process as other Senate-confirmed positions. In this case, the NIC chairman would be selected by the DNI, nominated by the President, and confirmed by the Senate.
Fixing Management Problems. The creation of this new structure for the IC makes flattening the current unwieldy bureaucracy all the more urgent. Yet there seems to be a serious dilemma for any attempt to rid the IC of its layers of management: No outside group of experts would be able to quickly grasp the complex system of administration of the IC and understand how best to reorganize the entire community for its many missions—while any group of insiders might be both too close to the problem for objectivity, and too timid in proposing changes.
- Bring in a world-class consulting firm and pay it to determine the best way to flatten the IC bureaucracy to ensure the most effective and efficient ratio of management to analysts and collectors. It would be like reinventing the wheel for any group of IC experts to attempt to take the place of an experienced firm that brings with it decades of experience in advising businesses how to reorganize themselves. A consulting firm would also be more objective and ruthless than any group drawn from within the IC. The firm’s proposals should be implemented in toto to streamline administration and management.
Unifying the Community. While reorganization of the IC leadership can ensure good cooperation and coordination at the very highest levels of the community, it will not break down the walls that exist among the all-source analysis agencies (such as the CIA, DIA, or the Bureau of Intelligence and Research). Exposure to the culture and viewpoints of other agencies will also help to foster a broader range of views and undermine groupthink around the IC.
- Cooperation with other all-source agencies should be encouraged and groupthink or cultural norms undermined by making certain that the very best analysts take part in Joint Duty Assignments. In addition, serving on the NIC staff at regular intervals, for at least six months at a time, should be a prerequisite for promotion to the SES level throughout the IC. By working side by side with peers from the various agencies, young analysts would come to understand the viewpoints of other institutions, see the biases that exist in their own agencies, and learn new ways of doing their profession.
Encouraging Diversity of Views. Given the dangers of groupthink suggested by the investigation at Central Command, as well as the failures to foresee the Arab Spring, the resurgence of al-Qaeda, and other national security challenges, much more needs to be done to encourage diversity of views and analytical conclusions. One traditional answer would be to establish red teams, and these should indeed be an integral part of the intelligence community’s analytical and product writing processes. Yet red teams alone will not provide well-sourced, reliable, and widely disseminated new ways of understanding ongoing challenges. Established views will always be seen as the norm, while the concepts proposed by red teams will be understood as exceptional, marginal, and therefore perhaps ignored in writing NIEs or other major products.
- In order to empower and disseminate alternative views more widely, red teams should generate counter-narratives and opposing analytical frameworks to feed into Team A and Team B exercises. This structure, used in 1976 to challenge received notions about Soviet doctrine and objectives, leveled the ground between a new analytical concept and the established views, and allowed a more objective testing of received wisdom. The participation of analysts from multiple agencies, as well as outside experts, would also allow fresh thinking on stubborn strategic challenges, as well as those that seem well-understood and manageable. These sorts of exercises should be uniformly applied when writing any major product, such as an NIE, but that they also should be encouraged as part of writing other, widely distributed products.
- These exercises will not, in themselves, guarantee that all sound and relevant viewpoints are captured and presented to the President and other customers. In addition to Team A/Team B exercises, the community should produce multiple major products, such as NIEs, when necessary. When there are serious disagreements among either relevant offices or even entire agencies, the dissenting entity, rather than being relegated to a footnote, will be encouraged to produce a competing product—that is, a dissenting NIE—for customers. The dissenting NIE would be distributed with the consensus NIE and to the same consumers. This will be true as well for the President’s Daily Briefing (PDB), which should include a text box expressing significant dissenting views when present.
- A culture of dissent should be encouraged even at the lowest analytical levels by rewarding helpful alternative views. Proposing a novel and sound concept, framework, or view that differs significantly from that of one’s peers, and that shows itself to be correct, should result in a step promotion, for instance. It might even be institutionalized across the IC by the DNI that one must have proposed a number of dissenting views—that proved correct—before one can be promoted. Managers should be trained to encourage young analysts to take chances and to reward disagreeing (respectfully and with sound analysis) with their peers, and the managers themselves should also be granted rewards and promotions when their charges propose sound alternative views.
- To further aid dissenting views, access to all relevant information or intelligence should be granted to every analyst working on a particular assessment. Today, the views expressed by all sides suffer from insufficient access to the most sensitive information that might in turn alter either the mainstream judgments or the dissent. (Note: The removal of several layers of bureaucracy, discussed above, should also allow a freer flow of information from the analysts and experts to the President and his principals. With fewer managers editing the analysis, the information to the customers should be less massaged, subject to groupthink, or politicized.)
- To help encourage real expertise, incentives should be established to encourage the hiring of PhDs and the granting of leaves of absence so that well-qualified junior analysts can work on doctorates and become experts on issues of concern. The threats the U.S. is facing will not disappear overnight. It is now that the U.S. must create the experts who will provide the necessary insight to deal with these issues over the long term. On a practical level, Joint Duty Assignments could be expanded to include these sorts of opportunities at academic institutions, as well as immersive experience at think tanks or high-tech firms in Silicon Valley.
- To maintain morale and to keep the majority of the workforce as generalists, no one should be prevented from moving to another subject area, but those who choose to stay in one subject area their entire careers should be rewarded with faster promotions (as their work indicates, of course). For promotion to the Senior Executive Service level, the DNI should consider making a PhD—in a relevant field—obligatory, or at least rewarded.
Ending the Open Source–Classified Divide. Classified information must remain the centerpiece for intelligence analysis, not only because it is the sole provenance of the IC, but also because it is often unequaled in value and, if correctly interpreted, can give a view of the enemy that is closer to reality than open-source information alone. At the same time, given the difficult problem set and the new complex security environment that the U.S. is facing, the IC can no longer neglect open-source materials as it has lately been wont to do. Instead, the IC needs to return to the earlier way when publicly available sources were exploited in the decades-long fight with the Soviet Union—as an additional resource that can be combined with classified information to present a more complete picture of these problem sets.
- The Open Source Center (OSC) should be significantly expanded to become an IC center of excellence, rather than its current incarnation as a largely CIA entity. The OSC needs a broadened mandate to provide open source tools and products. Combining open source materials with classified sources should become a routine part of writing every product throughout the IC. There should be continuous training of analysts to demonstrate the value of open sources and the OSC in particular and analysts should be asked to justify their neglect of open source materials if this occurs.
Instilling Strategic Thinking. Finally, there is the issue of too little strategic and long-term analysis by the IC. While it is true that this can be customer-driven, and that current intelligence must remain the primary focus of analysts, a neglect of longer-term thinking and writing will force the IC to remain reactive rather than proactive when dealing with the many challenges ahead. Because tactical analysis and strategic thinking are two completely different sorts of mental activities, which generally require two different sorts of analysts, it is necessary to create a separate structure within each agency—as well as one within the NIC—to adequately cover this gap within the IC.
- Create an office for long-range forecasting and strategic thinking and planning within each of the 16 agencies, perhaps called the Office of Strategic Analysis (OSA). Each OSA would be charged with considering non-current issues as broadly as possible to understand the entire complex strategic environment confronting the U.S. An OSA would also look at issues over the longer term, up to 10 years in the future, and consider broad trends affecting the national security sphere as a whole. To prevent these OSAs from being marginalized and excluded from giving input into the main documents produced by the IC, these analysts should be required to participate in the NIE, PDB, and other regular analytical product processes. They should also meet regularly with the equivalent office within the NIC to brainstorm, exchange views, and to work on the main analytical products written by each of the offices.
- The OSAs will be charged with producing a series of strategic documents, looking, for instance, one year, two years, and five years ahead, taking into consideration the national security landscape as a whole, not just narrow issues or current topics. These documents should build on each other and should be produced every year, with “lessons learned” taken from previous years’ products to improve and sharpen strategic thinking. The documents will then feed into an equivalent series of NIC documents that will attempt to take the soundest strategic analysis from the individual agencies and synthesize them into a high-level strategic analysis for the IC as a whole.
- While the current national security situation lasts, the OSAs should also be charged with producing, every six months, net assessments of the enemies the U.S. is facing and a comprehensive review of the current global position of the United States. These documents can then be used by operators, the Armed Forces, and others to feed into their strategic planning processes.
- Because there are equivalent offices within other departments around the interagency, the IC OSAs—either through the NIC OSA or through individual agency offices—should participate in National Security Council staff meetings on strategic planning and long-range forecasting with high-level members of these other entities, helping to create cross-fertilization of ideas, and perhaps even common documents, across the interagency.
Allowing Failure While Encouraging Learning. Despite the very best efforts to encourage diversity of opinions, to create the capacity for strategic analysis, and to allow analysts to speak as clearly as possible to policy leaders, there will always be intelligence failures and mistakes. The important point is to make learning from experience, whether good or bad, an integral part of the training process throughout the IC. While not guaranteed to fix poor analysis, it will help to prevent repeated misdiagnoses, wrong analytical frameworks, and poor predictions, and also to foster adaptability. In addition, much more remains to be done to encourage agencies to admit errors and fix them, while providing some cover when Congress or policymakers attempt to blame the IC for errors that are the result of mistaken policies.
- All managers, analysts, collectors, and operators should routinely write after-action reports (AAR). For operators and collectors, this should be done whether there is success or failure, while managers and analysts should only write them in cases of failure. The AARs on successes would be immediately placed in the new center described below, while those on failures would go through a lengthier process. First, these particular AARs would not be written jointly or as part of a group; everyone who participated in the failed project would write a separate, anonymous AAR that would not be shared with anyone else until the process discussed below. The individual AAR would describe precisely what occurred (the facts of the case), the analysis and analytical framework that were applied to the case, and what—in the opinion of the intelligence professional—caused the failure. These anonymous AARs would provide the basis for a special gathering of NIC staffers, outside experts, and internal managers and analysts to determine what went wrong, and the actions that need to be taken to prevent its reoccurrence. Their recommendations would be presented to the DNI as a memorandum for further action by the ODNI.
- To institutionalize this process for further learning and training, all AARs and memos, along with the actions taken to prevent reoccurrence, would be gathered in a Center for Lessons Learned housed at the NIC. Sessions at the center should be an integral part of the ongoing training process of analysts throughout their careers. In addition, analysts and managers who were participants in any failure would be asked to work with the DNI to implement the lessons learned in their case and to take further retraining, if the DNI believes it is warranted. Despite this lengthy learning process, the IC needs to encourage well-grounded risk-taking within both the analytical and operational sides of the IC. Unless malicious or negligent actions led to the error, no penalty should be assigned to the managers, analysts, or operators.
- To help protect the analytical branch in particular from political blame for errors, the IC as a whole should move to more specific percentage-based assessments rather than amorphous labels, such as “likely” or “unlikely.” A 50 percent assessment, for instance, would mean that both courses have exactly the same chance of occurring. Analysts will be encouraged to be honest and put the percentage as high as possible, with rewards for those who rate actions at 70 percent or above and are correct; but only certainties like “the sun rising in the East” would warrant a 90 percent rating. Such a rating approach would help to push the dilemmas of decision-making back on the policy leadership rather than intelligence analysts. Barring serious errors in judgment by the IC managers, analysts, or collectors themselves, the President and his principals should be the ones taking the risks and the ones who are blamed when their decisions do not work out as they had hoped.
A Final Word
The recommendations presented here could make the intelligence community into a very different enterprise—more unified and with a clearer chain of command, leaner bureaucracy, better analysis, and a concept for strategic analysis that preserves the necessity for current intelligence. But the most important change would be to help the new IC become a more adaptable, learning organization. These proposals are not the final reforms that are necessary for the intelligence community. History shows that there will be unforeseeable events and processes that will require flexibility and changes to structure, analysis, and the culture of the IC. If these proposed reforms accomplished only one thing—to make the IC an institution that welcomes change rather than rejects it for fear of punishment, and that can evolve with the complex national security environment—they would have succeeded.—Mary Habeck is a Visiting Scholar at the American Enterprise Institute. Charles D. Stimson is Manager of the National Security Law Program and Senior Legal Fellow in the Center for National Defense, of the Kathryn and Shelby Cullom Davis Institute for National Security and Foreign Policy, at The Heritage Foundation.