Intelligence Reform: Progress, Remaining Deficiencies, and Next Steps by Jordan Tama

Paper commissioned by the Princeton Project on National Security

October 2005

No aspect of the U.S. foreign policy infrastructure has received more attention since 9/11 than intelligence — and rightfully so.1 As former 9/11 Commission Vice Chairman Lee Hamilton has stated: “The single most important tool that we have in preventing terrorist attacks is intelligence.”2 In addition to its central role in combating terrorism, good intelligence is essential to defeat the insurgency in Iraq, stem the proliferation of weapons of mass destruction (WMD), shape smart policies toward the Muslim world, plan for the rise of emerging states such as China and India, and understand the security and economic implications of global dangers such as climate change and highly infectious diseases. Yet the track record of the U.S. intelligence community (IC) includes both notable successes and serious failures. The IC excelled at monitoring the Soviet Union’s military capabilities; has frequently provided critical information to American military commanders during wartime; and has helped to capture or kill hundreds of jihadist and insurgent leaders in Afghanistan, Iraq, and elsewhere. But the IC did not forecast India’s 1998 nuclear tests or the 9/11 terrorist attacks, inaccurately assessed Iraq’s WMD capabilities before the war, and today knows disturbing little about Iran’s or North Korea’s nuclear programs. It may not be fair, with the benefit of hindsight, to expect the IC to have gotten all of these issues right. But for roughly $44 billion a year, Americans naturally expect more.3 The 9/11 and Iraqi WMD failures have led to hundreds of proposals for intelligence reform from blue-ribbon commissions, congressional inquiries, and individual experts. They have also led to the adoption of the most far-reaching IC reforms since the passage of the National Security Act in 1947. These adopted reforms include the establishment of the position of Director of National Intelligence (DNI), the creation of the National Counter Terrorism Center (NCTC) and National Counter Proliferation Center (NCPC), the formation of an intelligence-oriented National Security Service in the Federal Bureau of Investigation (FBI), and the breaking down of many walls preventing cooperation and information sharing among the IC’s 15 federal agencies.4

On the whole, these reforms have improved America’s intelligence capabilities and begun the long process of establishing unity of effort across the sprawling intelligence

1

enterprise. But their long-term impact remains to be seen, and many of the most glaring intelligence problems remain. These problems include deficiencies in integration and coordination, collection, analysis, information sharing, the relationship between the IC and policymakers, congressional oversight, personnel policies, and innovation. Addressing these shortcomings adequately requires moving beyond structural reform to pursue policies that change the cultures and routines of agencies and individuals.

The next steps in intelligence reform must also include deeper changes in the IC’s relationship to people and institutions outside government and in its conception of the model intelligence officer. To maximize its expertise and influence, the IC should evolve into a multilayered network that reaches beyond intelligence agencies to include other federal, state, and local agencies, as well as selected nongovernmental organizations (NGOs) and experts.5 In the future, some intelligence officers should still be career civil servants, but many should be drawn, for shorter stints of service, from other professional backgrounds, including NGOs, the private sector, and academia. This paper describes major recent proposals for intelligence reform, reviews changes in the IC since 9/11, assesses continuing intelligence deficiencies, and offers recommendations to address remaining gaps and problems.6 The paper’s analysis and proposals are based on 16 interviews with intelligence officials and experts (listed at the end of the paper), as well as primary and secondary source research.

1. Major Proposals for Intelligence Reform since 9/11

The impulse for intelligence reform was generated by 9/11, which was widely seen as an intelligence failure, and was accelerated by the Iraq war, now viewed as having been based on poor or manipulated intelligence. This part of the paper summarizes the main findings and recommendations of five major bodies — four American and one British — that have issued reports on intelligence over the past four years. While each of these bodies was established as a result of the deficiencies exposed by 9/11 or the Iraqi WMD case, they all produced reports with conclusions and proposals of broader significance for intelligence. These recent reports drew on the work of other panels that examined intelligence during the decade preceding 9/11. According to intelligence reform scholar Amy Zegart, no fewer than twelve major commissions, government bodies, or think tank task forces studied intelligence between 1991 and 2001, and nearly all of them issued recommendations emphasizing the need for major change in the IC.7 But at a time when intelligence was not a top public or political concern, few of their recommendations were implemented.8 The bodies that issued reports after 9/11 repeated, often in modified form, some of the proposals of their predecessors, but a different political climate gave their proposals greater traction. Many of these proposals now have been, or are being, adopted. Readers familiar with these reports may wish to skip to Part 2 of the paper.

2

A. The Joint Inquiry report The first major investigation of intelligence shortcomings after 9/11 was conducted by a Joint Inquiry of the House and Senate intelligence committees. The Joint Inquiry’s report, released on December 10, 2002, included a narrative account of the 9/11 attacks and the following major findings and recommendations:9 Findings

• The U.S. counterterrorism effort before 9/11 suffered from a lack of an effective domestic intelligence capability and from divisions between domestic and foreign intelligence and between intelligence and law enforcement.10

• Failures among agencies to share information were a result of turf battles and differences in agency missions, legal authorities, and cultures.11

• The White House did not adequately prioritize intelligence goals and the Director

of Central Intelligence (DCI) did not have the ability to enforce priorities throughout the IC.12

• The IC did not conduct enough high-quality strategic analysis about the threat

posed by Al Qaeda.13

• The IC operated without a comprehensive strategy for combating the threat posed by Al Qaeda.14

• The IC did not develop or use human sources effectively to penetrate Al Qaeda,

relying too much on foreign liaison services for human intelligence (humint) on the group.15

Recommendations

• The government should establish a DNI with “the full range of management, budgetary and personal responsibilities needed to make the entire U.S. intelligence community operate as a coherent whole.”16

• The President should ensure that clear, consistent, and current priorities are

established and enforced throughout the IC.17

• The government should establish capabilities for the timely sharing of intelligence within the IC and with appropriate other federal, state, and local authorities.18

• The FBI should improve its strategic analytical capabilities.19

• There should be more extensive use of “joint tours” — rotations by intelligence

officials among agencies to broaden their perspectives.20

3

• The IC should recruit a more ethnically and culturally diverse workforce to

improve its ability to penetrate terrorist groups.21 B. The 9/11 Commission report The 9/11 Commission built on the work of the Joint Inquiry. Its report, released on July 22, 2004, provided a more detailed narrative of the 9/11 attacks and recommendations covering not only intelligence but also homeland security and foreign policy.22 Many of its intelligence findings and recommendations echoed those of the Inquiry, but the Commission’s report also included important new conclusions and proposals: Findings

• The 9/11 attacks “revealed four kinds of failures: in imagination, policy, capabilities and management.”23

• Prior to 9/11 there was not a coordinated effort across government to manage

terrorism cases and issues effectively.24

• Before 9/11 the IC did not conduct enough strategic analysis of Al Qaeda.25

• The DCI does not have enough capacity to set priorities, move resources, or establish standards throughout the IC.26

• Information is not adequately shared across intelligence agencies or across the

foreign-domestic divide.27

• Much of the IC has excessive security rules that lead to overclassification and excessive compartmentalization of information.28

• The congressional intelligence committees do not have the budgetary power,

political influence, or sustained capability they need to conduct effective oversight.29

Recommendations

• An NCTC should be created to unify strategic intelligence and operational planning related to Islamic terrorists.30

• A DNI should be established to manage the IC, with control over national

intelligence program budgets, personnel, and information-sharing standards.31

• National intelligence centers should be established to conduct joint collection and analysis on select high-priority issues.32

4

• The Central Intelligence Agency (CIA) should strengthen its analytic and humint capabilities, improve language training, and increase the diversity of its operations officers.33

• The IC should advance information sharing through the creation of decentralized

networks that can be accessed by people in all intelligence agencies.34

• The exercise of imagination should be routinized — for instance, through “red- teaming” (devil’s advocate thinking or analysis from the enemy’s perspective).35

• Congress should establish either a joint congressional committee for intelligence

or House and Senate intelligence committees that combine authorization and appropriations powers.36

• The FBI should develop an integrated national security workforce with an

institutional culture grounded in intelligence and national security expertise.37 C. The Senate Intelligence Committee report As the 9/11 Commission report was released, public discussion of intelligence shortcomings began shifting from a focus on 9/11 to an examination of apparently inaccurate prewar government statements and assessments about Iraq’s WMD capability. The Senate Intelligence Committee issued the first major report analyzing the IC’s prewar assessments on July 7, 2004.38 Its main findings and recommendations included the following: Findings

• Most of the major judgments of the October 2002 National Intelligence Estimate (NIE) on Iraq “either overstated, or were not supported by, the underlying intelligence reporting.”39

• The IC suffered from a “groupthink” dynamic that led collectors, analysts, and

managers to selectively interpret information about Iraq’s capabilities and to fail to question their underlying assumptions.40

• IC officials did not “accurately or adequately explain to policymakers the

uncertainties” behind the NIE judgments.41

• The IC had no human sources in Iraq after 1998 and depended too heavily on UN inspectors, defectors, and foreign government services to obtain humint on Iraq.42

Recommendations

• Intelligence analysts should clearly convey to policymakers the difference between what they know, what they don’t know, and what they think.43

5

• IC managers must encourage analysts to challenge their assumptions and fully

consider alternative arguments.44

• The IC “must develop and recruit unilateral sources with direct access to terrorist groups to confirm, complement or confront foreign government service reporting on these critical targets.”45

D. The Butler Committee report Seven days after the release of the Senate Intelligence Committee report, a British committee chaired by Lord Butler issued a report on British WMD intelligence.46 While many of its findings and recommendations were specific to Britain, some of them were relevant to the United States. These conclusions and proposals of broader relevance included the following: Findings

• Between 1991 and 1998 the International Atomic Energy Agency (IAEA) and UNSCOM (UN Special Commission) learned more about Iraq’s WMD programs than did national intelligence agencies.47

• Acquiring intelligence on secret WMD programs is particularly difficult because

they are usually subject to strong state control, very few people have complete knowledge of them, and many WMD materials and activities have dual uses.48

Recommendations

• Western countries should recognize and build on the value of intelligence generated by international organizations, such as the IAEA.49

• To prevent the development of groupthink, intelligence agencies should routinely

employ red-teaming and provide chances for the expression of dissent.50

• Policymakers and intelligence officials must recognize the limits and incompleteness of intelligence, rather than artificially constructing intellectually satisfying accounts.51

• Intelligence assessments should not necessarily place greater weight on secret

sources of information than on information obtained through nonsecret means.52 E. The WMD Commission report On March 31, 2005, the presidentially-appointed WMD Commission issued its report on U.S. intelligence.53 Many of its findings and recommendations echoed those of the

6

earlier investigations into 9/11 and Iraq, but it added some new conclusions and proposals: Findings

• The Iraq WMD failure stemmed from major deficiencies in collection, analysis, and communication with policymakers about the nature of the evidence.54

• Political pressure did not cause analysts “to skew or alter any of their analytical

judgments,” but an environment existed that did not “encourage skepticism about the conventional wisdom.”55

• The IC still knows “disturbingly little about the weapons programs and even less

about the intentions of many of our most dangerous adversaries,” including North Korea and Iran.56

• The IC has achieved important successes since 9/11, such as correctly assessing

the state of Libya’s WMD programs and using innovative techniques to discover and penetrate the A.Q. Khan proliferation network.57

• The IC as a whole remains generally unmanaged, with agencies operating largely

autonomously and collaborating only episodically.58

• Analytical tradecraft in the IC is subpar, the IC lacks enough analysts with expertise in Islamic extremism, and analysts rely too much on secret information.59

• Information sharing has “improved substantially” since 9/11 in counterterrorism,

but similar progress has not occurred in other key areas, including counterproliferation.60

• The IC adapts and innovates too slowly in technology, collection methods, and

operations.61 Recommendations

• The President should support the DNI in taking his powers over IC budgets, personnel, programs, and priorities “to the limit.”62

• Instead of establishing additional national intelligence centers (as the 9/11

Commission proposed), the DNI should designate “mission managers” who are responsible for developing strategies for all aspects of intelligence related to priority intelligence targets.63

• The President should establish an NCPC to oversee collection and analysis on

WMD across the IC.64

7

• The DNI should establish a new human resources authority to develop IC-wide

personnel policies aimed at encouraging joint assignments; improving job training; enhancing personnel incentives; and recruiting more people with needed technical, scientific, and linguistic expertise.65

• Congress should create intelligence appropriations subcommittees in both houses

and reduce the IC’s reliance upon supplemental budgets.66

• The DNI should set up a management system to harmonize the acquisition of technical collection systems with the prioritization of intelligence targets.67

• The FBI and Defense Department (DOD) humint operations should be made

subordinate to a new CIA Human Intelligence Directorate in order to improve humint coordination.68

• An Open Source Directorate should be established in the CIA that would make

open source information available to the entire IC.69

• A unit should be set up under the National Intelligence Council (NIC) to perform only long-term and strategic analysis.70

• A government-sponsored nonprofit research institute should be created outside

the IC to encourage diverse and independent intelligence analysis.71

• A National Intelligence University should be established to train IC analysts in standardized tradecraft.72

• A National Security Service subject to DNI authority should be established in the

FBI to manage and direct all FBI resources engaged in intelligence, counterterrorism, and counterintelligence.73

2. Recent Changes, Remaining Deficiencies, and Recommendations

Many changes have been made to the intelligence community since 9/11, partly in response to reform proposals and partly through the independent initiative of the executive and Congress. The landmark reform accomplishment was the enactment of the Intelligence Reform and Terrorism Prevention Act (hereinafter the Intelligence Reform Act) on December 17, 2004, but there have also been many other changes. This part of the paper summarizes the Intelligence Reform Act and assesses progress, describes continuing problems, and makes proposals in various areas of IC operations.

8

A. The Intelligence Reform Act The Intelligence Reform Act contained many provisions drawn or modified from the Joint Inquiry and 9/11 Commission reports, as well as a number of additional measures. Its important provisions included the following:

• Establishing a Senate-confirmed DNI with the authority to “develop and determine” the National Intelligence Program74 budget; “ensure the effective execution” of the budget; establish objectives and priorities for the IC; direct the tasking of collection, analysis, production, and dissemination of national intelligence; transfer or reprogram funds from one agency to another; and develop personnel policies to enhance the capacity for joint operation.75

• Establishing the NCTC to integrate foreign and domestic counterterrorism

intelligence collection and analysis, and to conduct “strategic operational planning” for counterterrorism operations.76

• Authorizing the President to establish an NCPC and national intelligence centers

on other priority issues.77

• Establishing a Joint Intelligence Community Council (JICC), chaired by the DNI and including the Secretaries of State, Treasury, Defense, Energy, and Homeland Security, as well as the Attorney General, to advise the DNI on budgetary and other matters and ensure the timely execution of DNI policies and directives.78

• Requiring the President to establish an “Information Sharing Environment” to

facilitate the sharing of terrorism information among federal, state, local, and private sector entities.79

• Creating a Privacy and Civil Liberties Oversight Board within the President’s

executive office to ensure that privacy and civil liberties concerns are appropriately considered in the implementation of counterterrorism policies.80

• Mandating that the DNI establish processes for ensuring that elements of the IC

conduct alternative analysis.81

• Requiring that the DNI appoint an individual or entity to ensure that finished intelligence products are objective, independent of political considerations, and based on proper analytic tradecraft standards.82

• Mandating that the DNI develop a comprehensive education, recruitment, and

training plan to meet the IC’s linguistic needs.83

• Mandating service in more than one element of the IC as a condition for promotion to certain positions.84

9

• Requiring the President to designate a single entity to oversee the security clearance process throughout the IC and develop uniform standards for access to classified information.85

B. Integration and coordination Many of the Intelligence Reform Act’s provisions, including the establishment of the DNI and NCTC, were designed primarily to improve coordination and encourage a greater unity of effort across the IC. The Act itself, as well as the 9/11 Commission report, referred explicitly to the goal of replicating in the IC the advances in jointness created by the Goldwater-Nichols Act of 1986, which reorganized the Defense Department.86 It remains to be seen to what extent the new reforms will spark greater integration and collaboration. The DNI’s authority and influence: The scope of integration and coordination will depend in part on how much power the DNI is able to wield over the various agencies. Former 9/11 Commission members and staff argue that the Intelligence Reform Act gives the DNI enough legal authority to exert influence over the agencies, but they would have preferred the Act to define his authority in clearer and stronger terms. They emphasize that strong presidential backing of the DNI will be essential to enable him to exercise his authority effectively.87 Skeptics of the law assert that the law does not give the DNI enough authority to direct agencies such as the National Security Agency (NSA), National Geospatial-Intelligence Agency (NGA), and National Reconnaissance Office (NRO), which also report to the Secretary of Defense. Former Deputy Director of Central Intelligence John McLaughlin described the law’s provision that its implementation must “respect and not abrogate the statutory authorities of the heads of departments” as “a loophole you can drive a tractor through.”88 It is too early to tell how powerful the DNI will ultimately be, but the law provides him with more legal authority than his predecessors in the position of DCI possessed and encouraging signs suggest that he will be able to exercise substantial influence over the IC. President Bush has given DNI John Negroponte the responsibility of preparing the President’s Daily Brief (PDB) and allowed him to take the seat formerly held by the DCI at National Security Council (NSC) meetings. Bush has also supported Negroponte in a couple of early turf battles with the Pentagon and FBI. In June, Chairman of the House Armed Services Committee Duncan Hunter (R-CA) sought, with encouragement from the Pentagon, to have the House pass a measure that would give the Armed Services Committee veto power over a decision by the DNI to transfer personnel from intelligence agencies to his office or to any new national intelligence center. But the White House supported Negroponte by intervening successfully to dissuade Hunter from advancing the measure.89 President Bush also backed Negroponte by adopting a policy that gives the DNI the partial power of appointment of the head of the FBI’s new National Security Service, which Bush established in June and which will include the FBI’s intelligence, counterterrorism, and counterintelligence divisions. The National Security Service will

10

be funded through the National Intelligence Program and its head will be hired by the FBI Director and Attorney General only with the concurrence of the DNI.90 This is the same arrangement that the Intelligence Reform Act established to govern appointments of heads of other intelligence agencies within Cabinet departments: the department head must gain the DNI’s concurrence in order to appoint the agency director.91 The arrangement strengthens the DNI’s authority by giving him powers that DCIs did not possess. Negroponte has also expressed an intention to assert his authority strongly, stating that he will deal with the DOD intelligence agencies directly, rather than through the Secretary of Defense, and telling all chiefs of station overseas that they work for him.92 Still, as the WMD Commission pointed out, the DNI’s power will be sharply limited by his lack of direct line authority over all of the IC agencies that remain part of separate departments.93 House Intelligence Committee Chairman Peter Hoekstra (R-MI) has recently indicated that the Pentagon is already carrying out new intelligence activities through programs intended to escape the DNI’s oversight.94 Moreover, legal authority is one thing, and political influence is another. As one former CIA official reportedly said: “The budget authority, when all is said and done, is going to come down to a fight between Negroponte and Rumsfeld and the OMB and the President.”95 The real balance of power among Negroponte, Rumsfeld, and other Cabinet officials should become more evident when administration negotiations over IC budgets for FY 2007 take place. Some critics of the reform law worry further that the DNI will not have a large enough staff to manage the IC effectively.96 Indeed, the DNI’s staff is expected eventually to total 500 to 700 people, a tiny fraction of the size of the large intelligence agencies, whose staffs number in the tens of thousands. But if the DNI successfully asserts his control over the CIA, NCTC, and NCPC — entities that, unlike other parts of the IC, do not report to another department head — he will have tremendous capabilities at his disposal. The extent of the DNI’s power, however, is not an accurate measure of how integrated the IC will become. To create more unity of effort, the DNI must use his power effectively to promote strategic management, common personnel and information sharing standards, and a culture of collaboration throughout the IC. Managing strategically: First, the DNI must, with clear guidance from the NSC, establish priorities for the IC and develop a strategic management plan for ensuring that all of the agencies share and advance them. Under the Bush administration, the NSC establishes IC priorities, based on a recommendation from the CIA.97 That process will presumably now be changed to have the DNI make the recommendation to the NSC, but the NSC’s role in establishing the priorities is likely to continue. In order for this process to work effectively, the NSC must devote adequate attention to it. In the past, the NSC has not made the provision of strategic guidance to the IC a priority, in part because of its

11

general preoccupation with more immediate concerns.98 The President and National Security Advisor must ensure that the NSC takes this responsibility more seriously. In carrying out this responsibility, the NSC must carefully balance the need to avoid establishing too many priorities and the need to ensure that important issues receive attention. According to the Joint Inquiry report, during the late 1990s the Clinton administration gave so many directives to the IC that there was inadequate distinction among priorities of greater or lesser importance and, as a result, “everything became a priority.”99 To avoid this pitfall, the NSC, through the DNI, must clearly rank the strategic issues on which agencies should focus. At the same time, the NSC and DNI must ensure that top priorities do not squeeze out other important intelligence needs. Since 9/11, the IC has been so focused on counterterrorism, Afghanistan, and Iraq that it is giving inadequate attention to problems related to China, Russia, Africa, or Latin America.100 There is also a danger that current intelligence priorities distract the IC from analyzing the potential security implications of global dangers, such as climate change, infectious diseases, or energy shortages. Striking the right balance between current top priorities and other needs is therefore a major challenge. Deputy DNI Michael Hayden is reportedly now examining how the IC should set priorities, a good sign that the DNI recognizes its importance.101 With priorities in hand, the DNI must work to develop collection and analytical strategies and to allocate and move resources appropriately. The WMD Commission sensibly recommended that the DNI establish a management structure in which the IC’s collection capabilities “are harmonized with intelligence priorities and deployed in a coordinated way.”102 It is especially important for the DNI to ensure that the development and acquisition of collection technologies fits a strategic plan. Too often, as former U.S. weapons inspector David Kay has pointed out, the government has allowed the national technical agencies — the NRO, NSA, and NGA — to demand resources for objectives that do not advance a common IC strategic agenda.103 Former NSA Director William Odom argues that this problem is particularly acute in the NRO, which, he charges, frequently develops projects to advance the economic interests of the aerospace industry rather than to achieve a national intelligence goal.104 The DNI must sharply reduce this type of waste and ensure that funds are only allocated to those projects advancing national priorities. He must also cut waste by reducing the large amount of duplication of activities among IC agencies, except when such redundancy is used to deploy multiple collection or analytical approaches against a given target or topic. The reduction of waste through better management should help pay for some of the costly reforms proposed in other sections of this paper. Congress must assist the DNI in these tasks by fulfilling its own responsibility for sound and sustained oversight of the IC. As discussed in the section on Congress below, congressional oversight today suffers greatly from a fragmented committee structure in which the intelligence committees, despite their superior expertise, have little influence over the intelligence budget. Agencies such as NRO can therefore get funding for programs approved by the armed services and appropriations committees, even if their

12

funding request is opposed by the intelligence committees and does not match national intelligence priorities. Giving the intelligence committees more authority would strengthen the capacity of Congress to help the DNI establish priorities, match resources to those priorities, and reduce waste.105 The NCTC, NCPC, and any other mission managers established by the government must also play a central role in strategic management. This requires them to establish priorities in their issue area; take stock of all IC activities related to that area; and develop management plans for integrating efforts, filling gaps, and ensuring that the allocation of resources matches their priorities. Developing IC-wide standards: Second, the DNI must develop and enforce IC-wide standards on issues including information sharing, training of personnel, rotations, classification of information, and security clearances. The Bush administration has taken appropriate first steps toward these objectives by establishing the new positions of Chief Information Officer and Chief Human Capital Officer under the DNI. The White House also announced in June 2005 that the DNI will create a personnel rotation plan and an IC- wide National Intelligence University system.106 Facilitating and requiring more rotations among agencies is important to educate intelligence officials about the workings of other agencies and give officials a broader perspective. Today, in much of the IC, the pursuit of a temporary assignment in another agency is still viewed as a career-limiting move.107 The DNI must overcome this agency resistance to rotations by laying out clear rules for implementing the Intelligence Reform Act’s provision mandating service in more than one element of the IC as a condition for promotion to certain positions. Specifically, the DNI should mandate that service in more than one agency be necessary for promotion to any position in the IC’s senior ranks. Such rules would help develop intelligence community jointness analogous to what the Goldwater-Nichols reforms accomplished for the uniformed services. Reforming IC security clearance and classification procedures must also be a priority for the DNI. Currently, agencies have a wide range of such procedures, leading to the undesirable result that agencies classify the same information at different levels.108 Uniform rules to govern clearances and classification would enable the IC to share information more readily. Changing cultures: Third, the DNI must attempt to channel the many cultures of intelligence agencies into a shared culture of collaboration. Nearly all intelligence experts agree that the various agencies carry out their work in fundamentally different ways and view each other with distrust or scorn.109 Senate Homeland Security and Government Affairs Committee Minority staffer David Barton explained: “Each agency still has blinders on and possesses its own distinct, insular culture.”110 These cultural divisions weaken the U.S. intelligence enterprise by reducing cooperation and fueling turf battles. While there has been substantial progress in improving cooperation among agencies since 9/11, turf battles continue to be a significant problem.

13

A congressional staffer described one current conflict: “The FBI is in a desperate battle to outdo the DHS [Department of Homeland Security]. It will send out alerts about threats to the U.S. — and that’s DHS’ job. Agencies are trying to carve out territory.”111 Such battles exist throughout the IC and could become fierce among the DNI, NCTC, NCPC, and all of the separate agencies. Deputy DNI Michael Hayden acknowledged in April that the creation of a collaborative culture is the most serious challenge facing the DNI.112 Reducing turf battles requires, above all, changing the mindsets of intelligence officials. This is hard to do through structural change, but common standards for information sharing, training, and rotations can lead to cultural change over time. Additionally, the turnover of large numbers of employees itself should facilitate significant cultural change, as people with fresh perspectives replace personnel that remain wedded to traditional ways of doing business. Senate Homeland Security and Government Affairs Committee Counsel Rajesh De noted: “The top-level officials seem to understand the need to share more, but many career officials further down remain turf conscious. It takes a new generation of analysts and officers to promote a new culture of sharing and breaking down of barriers.”113 The movement among agencies of personnel who have worked at the NCTC, in the DNI’s office, or at other new joint intelligence bodies should also promote a more collaborative culture. There are as well some encouraging precedents of cultural change. As 9/11 Public Discourse Project President Chris Kojm remarked: “Everyone despairs of reform at the FBI, but I don’t. You look at the FBI under [former Director J. Edgar] Hoover and look at it under [current Director Robert] Mueller, and they’re very different organizations.”114 Indeed, today the FBI is far more dedicated to the rule of law and attentive to civil liberties than it was 40 years ago. More recently, the NSA and NGA have developed a culture of collaboration since 9/11. The agencies have worked systematically to try to promote greater integration by rewiring computer networks, retraining staff, reorganizing operations, and redesigning security procedures. One of the results of these efforts is that the agencies now coproduce joint intelligence reports and serve together in military commands. Joan Dempsey, executive director of the President’s Foreign Intelligence Advisory Board, recognized these advances in a recent speech, stating: “The NGA and NSA are acting closer together now than any intelligence organizations in history.”115 Their collaboration provides a valuable model for the rest of the IC. Balancing integration and decentralization: While developing a more integrated and better coordinated IC, the government must recognize the value of decentralization. The need to coordinate IC activities should not drive the United States to create a fully centralized system in which the entire IC operates in the same way. Some amount of decentralization allows for a greater diversity of collection and analytic approaches and perspectives, which is critical to promote innovation and prevent groupthink from taking root. In making this point, a Markle Foundation Task Force noted that the government is more likely to find signs of potential terrorist activity if multiple analysts are examining information from different points of view.116 The goal, then, must be to advance unity of

14

effort and a culture of collaboration, while giving agencies room to pursue national intelligence objectives in a variety of ways. C. Collection Management and coordination, of course, are only useful if they lead to effective intelligence collection, analysis, and distribution. This section discusses gaps and challenges in the three primary collection disciplines of humint, signals intelligence (sigint), and imagery intelligence (imint), as well as the role of open sources and the need to fuse collection disciplines. Human collection: Since 9/11, Congress and the executive have attempted to rebuild American human intelligence capabilities, authorizing substantial increases in resources for CIA, FBI, and DOD humint operations. The CIA graduated the largest-ever class of new case officers this year, and both the Pentagon and FBI have ramped up their humint collection.117 In November 2004, President Bush attempted to accelerate this expansion by directing the CIA to increase its cadre of operations officers by 50% over the next five years (according to a recent report the CIA has 1200 case officers stationed overseas).118 More recently, in June 2005, the House voted to shift other resources in an intelligence reauthorization bill from the development of new reconnaissance satellites to investment in humint capabilities and analysis.119 Also in June, the White House announced that a senior CIA official would become the national human intelligence manager, in charge of coordinating CIA, DOD, and FBI operations.120 John Negroponte plans to hire a deputy in his own office to oversee that coordination.121 In describing these changes, senior intelligence officials stated that the new Deputy DNI would provide broad, strategic direction of humint and that the new CIA manager would set standards designed to minimize conflicts among agencies, rather than assert direct operational control.122 In a sign of growing coordination, the CIA, FBI, and Pentagon reportedly reached agreements with each other delineating their respective areas of humint authority and responsibility.123 After years of neglect, however, U.S. humint remains seriously deficient. Increases in numbers and improvements in coordination will not resolve its underlying problems. One such problem is that the IC is too attached to quantitative, rather than qualitative, measurements of success, typically rewarding officers for how many agents they recruit instead of the quality or accuracy of the information those agents provide.124 This is a major shortcoming because the collection of inaccurate information can do serious harm, as it did in the Iraqi WMD case. The United States also still possesses very little humint on North Korea and Iran — even though those countries have been high-priority intelligence targets for years. Additionally, the IC remains woefully short of officers — in the CIA, DOD, and FBI — who speak Middle Eastern and South Asian languages. Even among new hires of officers, only a small number are fluent in Arabic or other languages sorely in demand.125 Moreover, language ability alone is not enough to enable officers to blend into Muslim

15

societies. To penetrate dangerous Islamist groups, officers need deep cultural knowledge and experience.126 Lee Hamilton argued that the government still has not adapted to the challenge of conducting humint on hard targets: “We refuse to recognize how hard it is to penetrate nonstate actors and how ill-prepared or unwilling we are to do it. The problem is that everyone in the CIA is white.”127 The failure to hire more Muslim-Americans or others with Middle Eastern or South Asian roots — even since 9/11 — reflects in large part the difficulty they face in getting security clearances.128 Former CIA case officer Robert Baer has colorfully described this problem:

If you are, say, an American born in Islamabad who happens to have a second cousin working in the Pakistani intelligence service, the chances of getting security clearance to join the agency [CIA] are close to nil. Third- generation Americans with no known foreign relatives but who have spent much of their lives overseas have a better chance, but the odds are still slim, especially if those overseas years were spent studying in a place like Cairo… To reject such people solely because they aren’t provincial is yet another way the agency cuts off its nose to spite its face.129

To improve its humint capabilities, the United States should modify its security clearance procedures to make it easier for the IC to hire qualified candidates from Middle Eastern and South Asian backgrounds. Baer suggests establishing a two-level clearance system for CIA officers. One level would consist of a top-secret clearance comparable to that which officers receive today, while the other level would be for people who spend most of their lives outside the United States, including people who marry foreign spouses. The officers in this latter category would only get a secret clearance — meaning they would not have access to sigint, imint, or nuclear intelligence.130 Such a multilevel clearance system would give the United States greater capacity to penetrate hard targets. Hiring officers who live mostly overseas would also benefit America because such officers are likely to develop more in-depth political and cultural expertise on the countries in which they live.131 Baer’s proposal has been resisted by some CIA officials, who do not want to weaken security standards, but CIA Director Porter Goss recently stated that the agency has made changes to its hiring process to enable it to hire more immigrants and people who have engaged in extensive foreign travel.132 The NSA demonstrates that a multilevel security clearance system can work; its security clearance rules even allow non-American citizens to be hired in limited cases.133 In addition to recruiting more officers with experience living abroad, the IC should sharply increase its use of non-official cover officers (NOCs), who work cover jobs in the private sector or an NGO rather than as a U.S. diplomat. A NOC is much more likely than an officer with a diplomatic cover to be able to gain access to nonstate organizations such as terrorist groups. Training effective NOCs is particularly difficult, however.134 Former CIA Inspector General Frederick Hitz noted: “A good non-official cover officer is worth his weight in gold, but it’s a very tough job.”135 It also costs the government more to support NOCs because they require special arrangements for pay and health care.136 But that cost is worth paying if it will produce valuable intelligence.

16

The IC should also reach out more to Americans who might not be interested in working in intelligence but have connections to other countries and might have useful information and insights about them. Former NIC Chairman Robert Hutchings commented: “A country as multicultural as ours should have huge advantages in intelligence, but we don’t exploit them in government.”137 Hutchings noted that the IC could learn much about Iran and North Korea from Iranian- and Korean-Americans, and that intelligence officials need to be more open to contact with such ordinary citizens outside government. The IC should also tap the collection capacity of immigrant or second-generation Americans through state and local law enforcement agencies, particularly police departments. The New York Police Department (NYPD) already has advanced counterterrorism and intelligence capabilities, including officers stationed overseas and direct communications with security services and police departments in many countries. Some NYPD officers also collect intelligence by reading foreign media, visiting mosques, and meeting jihadists in online chat rooms.138 William Finnegan described the NYPD’s impressive intelligence capabilities in a recent New Yorker article: “In some key areas, such as languages that are critical to counterterror work, the NYPD, drawing on a city of immigrants, has deeper resources than the federal agencies traditionally responsible for fighting international terrorism.”139 Indeed, the Pentagon recently borrowed 17 computer literate Arabic speakers from the NYPD to assist its own intelligence work.140 Finnegan notes as well that NYPD officers are typically more adept at blending into immigrant communities to collect intelligence than FBI officers, who have less local knowledge and experience.141 The FBI, DHS, and other federal agencies should recognize the potential intelligence capacity of police departments across the country, and should work to integrate federal, state, and local intelligence collection efforts. Even with an enhanced humint capability, the IC will still need to work closely with foreign intelligence services to gather information. Michael Hurley, Senior Director for Policy at the 9/11 Public Discourse Project, emphasized that these liaison relationships are essential in many places: “Some allied countries have large intelligence and security services that are focused on collecting intelligence on the Islamist extremists in their territory. We need to strike a balance between unilateral human intelligence operations and close cooperation with those intelligence organizations that, by means of their personnel, resources, control of the terrain, and knowledge base, can make a strong contribution to the common effort.”142 To leverage the intelligence networks and capabilities of friends and allies, the United States should engage in both bilateral and multilateral cooperation. Bilateral intelligence cooperation has long been the IC’s preferred means of cooperation, but multilateral networks can be established with a select group of countries trusted by the United States. Already such multilateral cooperation is occurring: the United States, France, Britain, Germany, Canada, and Australia have established a multinational center in Paris called Alliance Base, which facilitates the sharing of intelligence and joint planning of covert operations.143 Intelligence officials report, moreover, that the vast majority of successful

17

identification, tracking, and capturing or killing of suspected terrorists outside Iraq and Afghanistan has occurred through joint intelligence work.144 The United States must be very careful, however, when working with foreign services — including those of allies — to ensure that they are providing reliable information and are not leaking information provided by America to others. The United States also needs good unilateral intelligence sources to complement bilateral and multilateral relationships so that it does not depend completely on the cooperation of foreign intelligence services whose loyalties might be suspect or divided. With respect to Pakistan, for instance, the United States cannot be confident it is receiving accurate information from the Pakistani intelligence services in the absence of its own intelligence operations in the country. Porter Goss recently stated that the CIA is reducing its reliance on foreign intelligence services by developing more unilateral intelligence operations.145 The United States should recognize as well that it can gain valuable intelligence from international inspection regimes. As the WMD Commission noted, the United States learned more about Iraq’s WMD programs from the IAEA and UNSCOM during the 1990s than through national intelligence means. Similarly, the United States has learned more about Iran’s nuclear program in recent years from the IAEA than from the IC. Rather than criticizing the IAEA for not taking a tougher stand on Iran, America should view the IAEA as a valuable collection tool and work to strengthen its inspection capacities so that it can learn more about secret nuclear programs. Similarly, the United States should support the establishment and enforcement of strict inspection regimes for biological and chemical weapons programs to increase the likelihood that hidden programs will be uncovered. Technical collection: Technical intelligence collection, including both signals and imagery intelligence, has not received as much attention in the intelligence reform debate since 9/11 — in part because it is less well understood by outsiders and remains shrouded in great secrecy. It is clear, however, that technical collection currently has major shortcomings. With respect to sigint, the NSA is struggling to keep up with advances in encryption technology and other telecommunications deception techniques.146 Former NIC Vice Chairman Gregory Treverton argues that this problem is so severe that by 2015 strong encryption could mean the effective end of sigint as it is currently known.147 That judgment is probably too harsh, as investments in new technology are occurring that should revitalize the sigint enterprise. But the NSA is years behind schedule and hundreds of millions of dollars over budget on a major modernization program, called Trailblazer, intended to improve its capacity for intercepting and sorting phone, e-mail, and other communications.148 Imagery intelligence confronts serious problems as well. Imint was very valuable during the Cold War, when the IC was focused on monitoring Soviet arms capabilities that were often visible from the sky. As the WMD Commission noted, however, imint is often ineffective against today’s enemies, who have learned to conceal WMD activities by

18

conducting them underground, or, in the case of some biological and chemical programs, hiding them in buildings with no suspicious signatures.149 Imint also faces the same filtering challenge that confronts the sigint enterprise: processing the vast quantities of information gathered by satellites and other systems to retrieve only the data that analysts need.150 Additionally, the development of very sophisticated commercial satellites raises questions about whether the government should be investing in developing its own costly satellites when they might have only marginal advantages over commercial ones.151 The biggest government program for a new generation of reconnaissance satellites, Featuring Imagery Architecture, has tripled in price to about $25 billion while producing less than promised.152 Congressional concern about the cost and utility of this program led the House to divert FY 2006 funds from it to humint and other intelligence activities.153 Both sigint and imint efforts also contain serious inefficiencies. The inefficiencies stem in part from the fact that the development and acquisition of collection systems is conducted by the NRO, while the processing and analysis of imint and sigint is done by the NGA and NSA, respectively. According to critics such as William Odom and some House congressional officials, the NRO’s investments often do not match the priorities of the NGA, NSA, or IC as a whole, but instead are driven by the commercial interests of contractors who develop collection systems.154 To address this problem, Odom has proposed abolishing the NRO and transferring its activities to the NSA and NGA.155 While that particular proposal might not be politically feasible, it is clear that the IC must overhaul the technical collection enterprise so that resources are better matched with priorities and investments only go to systems that the country needs. Open-source collection: The IC must recognize that it is not necessary to turn to secret means to assess many intelligence issues. Gregory Treverton frames this issue by distinguishing between puzzles — which have answers — and mysteries — which do not.156 For instance, the following are puzzles: Where is Osama bin Laden? How many nuclear weapons does North Korea possess? By contrast, the following are mysteries: Will Al Qaeda become stronger or weaker? Will North Korea’s government collapse? Secret intelligence can be extremely helpful for solving puzzles. It is of little use, however, for investigating mysteries; instead, openly available information, expertise, and judgment are likely to lead to the best assessments. Given this distinction, Treverton proposes that humint efforts focus only on closed potential foes, such as North Korea; closed and dangerous programs in open societies, such as India’s nuclear program; and terrorists or other nonstate enemies.157 The same criteria could apply to sigint and imint collection. For nearly all other strategic issues, as discussed below in the section on analysis, the IC should rely primarily on open sources. The collection of information from open sources presents distinct challenges because the vast amount of information available in the public domain must be filtered into a usable form. Some collection from open sources, of course, should consist simply of reading newspapers, web sites, and other media. To process more vast quantities of information, the IC should invest heavily in data mining technologies.158 One such technology could be software that searches all published media in the world for information of relevance to analysts. The RAND Corporation sponsored a study that proposed a more sophisticated

19

system that would gather information related to intelligence targets from secret and open sources, continually filter the findings to identify unusual behavior or information, and pass unresolved matters to human analysts.159 Such a system might not be technically feasible yet, but the IC should invest in developing one when technology permits. Fusing collection disciplines: The IC needs to integrate the operation of the different collection disciplines so that humint, sigint, and imint become what former CIA official John Gannon has called “multi-int.”160 Currently, as the WMD Commission asserted, the IC doesn’t have “the long-term, coordinated collection strategies that are necessary to penetrate today’s intelligence targets.”161 The United States must not just fuse data after it has been collected, but also collect data by employing collection disciplines in combination. As discussed above, the NSA and NGA have gone a long way towards integrating their operations, but their activities remain largely separate from human and open source collection. All of these disciplines need to be used in sync throughout the process of gathering and analyzing intelligence. Collecting on military and nonmilitary targets: The IC must ensure that the military’s collection demands do not squeeze out other national collection needs. Since the 1990s, the trend in the IC has been toward increased DOD control of collection priorities, not just in the technical collection agencies but also in the CIA. This has led to a dominant focus on tactical intelligence for military commanders and a neglect of collection on strategic issues.162 The DNI must ensure that the IC collects on long-term strategic issues as well as to satisfy tactical requirements. Protecting privacy and civil liberties: Finally, the collection enterprise must have clear guidelines and safeguards for the protection of privacy and civil liberties. The Patriot Act gave new powers to the government for gathering intelligence on people inside the United States through wiretaps, surveillance, the acquisition of medical and library records, and other means. The House and Senate have passed different versions of Patriot Act reauthorization bills this year, which are being reconciled in conference committee as of this writing. Both bills renew nearly all of the Act’s provisions, but the Senate version places new limits on some powers granted to the executive branch by the Act. For instance, the Senate bill requires that the FBI Director or Deputy Director approve efforts to obtain medical or library records, and that the government demonstrate to a judge that a request to monitor any phone or computer used by a suspect is relevant to a national security investigation.163 These types of requirements are valuable because they balance carefully the need to give the government adequate tools to prevent terrorism and the need to ensure that those tools do not unnecessarily restrict Americans’ privacy or liberties. The government must also respect the human dignity and rights of non-Americans who are detained and interrogated. Interrogation is not usually discussed as part of the intelligence reform debate, but it should be, particularly since it has become an important part of intelligence gathering in the war on terror. In Iraq, Afghanistan, and elsewhere, military and CIA interrogations have led to severe abuses of detainees, including some that have resulted in death and serious injury.164 The CIA has also reportedly set up

20

secret detention facilities to hold and interrogate about 30 better-known terrorism suspects, whose conditions of confinement are not known to the public.165 In October 2005, the Senate voted 90-9 to regulate the detention, interrogation, and treatment of prisoners held by the military in an amendment to the military appropriations bill for FY 2006. This amendment would ban the use of “cruel, inhuman or degrading treatment or punishment” against anyone in U.S. government custody — reaffirming the existing American and international prohibition of torture — and would require all U.S. troops to use only interrogation techniques authorized in a new Army field manual.166 The measure is an appropriate step toward providing clearer guidelines for detentions and interrogations, but it faces stiff opposition from Republican leaders in the House and a veto threat from the White House. To provide more oversight related to individual rights, the Intelligence Reform Act adopted a recommendation of the 9/11 Commission to establish a Privacy and Civil Liberties Oversight Board in the President’s executive office. The White House has acted slowly in setting up the Board, however, and only named its members in June 2005. This delay suggests that the administration does not strongly support the Board’s role. Concerned that privacy and civil liberties are not being adequately protected, Representatives Carolyn Maloney (D-NY), Christopher Shays (R-CT), and Tom Udall (D-NM) have cosponsored a bill that would turn the board into an independent agency with bipartisan membership.167 Such a change would help ensure that the protection of individual rights is monitored by officials whose incentives differ from those of the executive branch. D. Analysis To be of use for policymakers, collected information needs to be analyzed. Yet the IC has long neglected analysis, particularly strategic analysis, while funneling the vast majority of its resources to collection. Inaccurate IC assessments of Iraq’s WMD programs represent only the most glaring example of a number of deficiencies in analytical tradecraft and capability. The IC has begun addressing many of these deficiencies, but much more work needs to be done. Preventing groupthink: In examining the Iraq WMD assessments, the Senate Intelligence Committee, Butler Committee, and WMD Commission all pointed to the pervasive problem of groupthink. There are several things the IC should do to prevent groupthink from taking root and to encourage diverse and insightful analysis. First, the IC should have multiple agencies or individuals analyze a given issue to allow a range of perspectives to be brought to bear on it. Some amount of decentralization and redundancy in analysis is therefore valuable. Currently, the CIA, FBI, Defense Intelligence Agency (DIA), State Department Bureau of Intelligence and Research (INR), and DNI all have analytic capability, which can allow for a healthy competition of views. In the Iraq WMD case, however, the views of dissenters, such as INR and the Department of Energy, were often not given an adequate airing.

21

Second, a mechanism must be in place to bring the different analyses together into a single IC assessment, while preserving recognition of uncertainties and disagreements. This has long been the role of the NIC, which oversees the production of National Intelligence Estimates (NIEs) and is now chaired by the Deputy DNI for Analysis Thomas Fingar. The recognition of uncertainties is particularly important because intelligence judgments on hard targets are usually based on inconclusive information.168 In the case of the October 2002 Iraq NIE, caveats and disagreements were relegated to footnotes, and as a result the estimate gave the impression that its key judgments were more certain than they were. To prevent this problem, former NIC Chairman Joseph Nye proposed that NIEs and other assessments should clearly state what is known, unknown, and uncertain, and should express degrees of confidence in their judgments.169 Since 2003, the IC has attempted to adopt this approach — estimates now include degrees of confidence and acknowledge ambiguities and disagreements more clearly than before.170 One example of this improvement was a recent NIE on Iran’s nuclear program, which included alternative theories that could explain some of the suspicious activities discovered in Iran over the past three years.171 The DNI has also changed the process for preparing the President’s Daily Brief (PDB) so that it includes more diversity of analysis and recognition of dissents.172 Third, the IC should encourage analysts to challenge prevailing views. This can be done through red-teaming, in which analysts make the case for alternative positions or interpretations. President Eisenhower used this approach effectively in an exercise known as the “Project Solarium,” in which his advisers were divided into three groups that debated the merits of alternative grand strategies toward the Soviet Union.173 The IC has begun doing more red-teaming in recent years, and the Intelligence Reform Act called for establishing an individual or entity in the DNI’s office to ensure that the IC conducts alternative analysis when appropriate.174 In the end, however, individual analysts must have primary responsibility for questioning dominant views. As Joseph Nye said commented: “After putting together an assessment, analysts must ask: What would it take to make this assessment wrong?”175 Answering that question requires the exercise of imagination called for by the 9/11 Commission. Alexis Albion, 9/11 Public Discourse Project Director of Policy, noted that the leaders of intelligence agencies must allow their analysts to do this type of imaginative work and should provide them with appropriate training for it.176 Finally, the IC should seek additional perspectives from outside experts. Joseph Nye began this effort as NIC Chairman by brainstorming assessments with outsiders in unclassified meetings and commissioning think tanks to write estimates parallel to those done within the IC.177 Robert Hutchings further expanded the NIC’s contact with people outside government during his chairmanship from 2003-05. For instance, under his leadership the NIC held public meetings with private sector, academic, and NGO experts in six different countries in preparing the report Mapping the Global Future, which

22

attempts to gauge the direction of international trends and imagines what world politics might be like in 2020.178 Strengthening strategic and long-term analysis: Mapping the Global Future is remarkable because it is one of a small number of IC products that feature long-term or strategic thinking. This is a shortcoming at least as serious as the problem of groupthink. Throughout the IC, analysts are constantly asked to answer tactical questions, such as: Where is terrorist leader X likely to be? Will the insurgency step up its attacks before the upcoming election? But analysts are rarely asked to answer strategic questions, such as: Is U.S. counterterrorism policy reducing or increasing the threat of terrorism? What are the costs and benefits of inducing insurgent leaders to join the political process? The IC’s focus on immediate concerns drives out consideration of the broader context of policy and of longer-term political and social trends. The IC also does not devote enough attention to assessing the likely intentions of adversaries, often focusing only on their capabilities. This tendency has grown worse over the past decade, as tactical intelligence, particularly support to military commanders, increasingly dominates both collection and analysis.179 The IC remains particularly weak in domestic strategic analysis. Since 9/11, the FBI has sought to improve its analytic capability by establishing an intelligence directorate and hiring roughly 700 additional analysts.180 But a recent report of the Justice Department’s Inspector General found that the FBI still faces problems hiring, training, and retaining analysts, and that it is not using its analysts effectively. Analysts reported that they are not respected by special agents and are sometimes given menial tasks such as escorting visitors or collecting trash that needs to be incinerated.181 Another recent report, by the Congressional Research Service, found that some FBI analysts had yet to receive basic training in analytic skills and that some analysts in FBI field offices do not have Internet access on their desktops.182 Resulting job dissatisfaction has fueled a substantial attrition rate: approximately 10% of FBI analysts have left their jobs in each of the past three years.183 These deficiencies have probably contributed to a glaring shortcoming: the FBI has not yet produced a comprehensive assessment of the terrorist threat in the United States.184 To address these problems, the FBI must give analysts better training, ensure that they are given work that suits their qualifications, and educate special agents about analysts’ role and importance. Strategic analysis also needs to be given more attention in other parts of the IC. The WMD Commission smartly recommended the establishment of an organization under the NIC that would perform only long-term and strategic analysis to shield these tasks from the constant pressure of day-to-day demands.185 Some strategic analysis — for instance, assessing China’s likely political evolution or the impact of counterterrorism policies on Jihadist groups — requires personnel with deep expertise in particular issues, countries, cultures, or regions.186 As a former DIA counterterrorism analyst testified to the Joint Inquiry: “In the case of an analyst responsible for tracking a Middle Eastern terrorist group, this person will need to have an expertise or at least a good working knowledge of terrorism itself, the group that they have for an account, regional and country issues

23

present in the group’s operating area, which can be quite extensive, and Islamic history, culture and the sects thereof.”187 Thomas Fingar reportedly recognizes the need for such expertise and is examining ways for the IC to hire and develop more analysts who are real experts.188 Former White House counterterrorism coordinator Richard Clarke argues that the IC should recruit fewer young college grads as new analysts and instead sign up more senior experts for short-term, renewable contracts.189 Such short-term contracts would enable the IC to hire people who do not wish to have a long government career. At the same time, the IC should continue to hire well-qualified young analysts, but it should allow them to master an issue, country, culture, or region by keeping them on the same account for an extended period of time.190 To create a further incentive to conduct strategic analysis, at least one PDB per week should include an article containing strategic or long-term thinking. Additionally, the IC might benefit from developing a Federally Funded Research and Development Center (FFRDC) with a mandate to conduct strategic analysis. The staff of such an organization would possess security clearances, but they would not be government employees and would have closer connections to the academic and nonprofit worlds.191 The government should also invest in developing academic and linguistic expertise on priority issues by funding university programs related to them — for instance, in Islamic studies, Arabic, and Chinese — as it did with Soviet Studies programs during the Cold War.192 Since 9/11 the number of Americans studying Middle Eastern and South Asian languages has increased substantially, but it remains far too low to supply government with an adequate stream of fluent speakers.193 Increasing the study of these languages in schools is critical because it takes an analyst 33 months on average to become fluent in a non-Roman alphabet language when learning on the job.194 More broadly, intelligence officials and analysts must recognize that publicly available information is usually more valuable than secret information for answering strategic questions. Today, too many officials and analysts continue to overemphasize secret information to the exclusion of valuable open source information.195 The Bush administration has attempted to address this problem by establishing a position of Assistant Deputy DNI for Open Sources to oversee open source collection and analysis.196 The WMD Commission further recommended that an Open Source Directorate be established in the CIA. But former CIA Inspector General Frederick Hitz and John Gannon have pointed out that such a body could have the undesirable effect of separating secret and open intelligence.197 Instead of establishing new institutions for open source analysis, the IC should train individual analysts to make full use of both open and secret sources when making assessments. E. Information sharing Good analysis and useful information become more valuable when they are shared with other analysts as well as the officials and policymakers who must act on them. As Lee Hamilton has stated: “The key to intelligence is getting the right piece of information to

24

the right person at the right time.”198 Since 9/11, information sharing has improved substantially, but many deficiencies remain. Progress since 9/11: Progress on this front has included increased information sharing among national intelligence agencies; federal, state, and local agencies; and intelligence and law enforcement authorities.199 The latter change was facilitated by the Patriot Act, which broke down barriers preventing information sharing between prosecutors and intelligence officials. FBI Director Robert Mueller has also issued a new policy guideline: “Share by rule, withhold by exception.”200 There has been progress as well in developing an integrated government watch list of suspected terrorists. The new Terrorist Screening Center has consolidated information from many separate agency watch lists into a single database, though it is not yet free of problems.201 The Bush administration’s recent decision to appoint a Chief Information Officer in the DNI’s office should further advance information sharing. Continuing deficiencies: Sharing is still not adequately institutionalized, however. The WMD Commission report noted that agencies act like they own the information they collect and that sharing “depends too much on physical co-location and personal relationships as opposed to integrated, community-wide information networks.”202 Former Attorney General Richard Thornburgh, who chairs a National Academy for Public Administration panel that closely follows the FBI, also recently observed that information sharing remains inadequately grounded in laws and institutions.203 State and local officials perceive the FBI and DHS have a long way to go in their willingness to share information.204 The DHS is developing a network for sharing classified information with federal, state, and local intelligence and law enforcement agencies, but the network still has major flaws and it will be years before all relevant authorities are connected to it.205 Some agencies have serious problems just sharing information within their own offices. Over the past few years, the FBI poured more than $100 million into a system known as the Virtual Case File, which was intended to integrate the FBI’s files electronically, but the system ran into major technical problems and was abandoned earlier this year. Today, the FBI is still without an integrated database to manage its cases, as its chief information technology officer, Zalmai Azmi, recently acknowledged: “What we don’t have is an efficient way of working with our information. What I mean by that is that we’re still paper-based.”206 Developing a networked community: To promote better information sharing, the IC should develop a decentralized network that includes the systems of each agency involved in intelligence and is accessible to relevant state and local authorities. The 9/11 Commission proposed this structure, in which agencies would keep their own networks, but those networks would become searchable across agency lines, with secrets protected through the network’s design.207 That recommendation echoed the model of a highly respected Markle Foundation Task Force.208 John McLaughlin expressed the magnitude and urgency of this task: “The real challenge today is searching and fusing data from many sources. We need a Manhattan Project for a new information architecture.”209

25

Many intelligence officials worry that such a network would compromise classified information, but it could be designed so that only officials with appropriate clearances would automatically have access to secret information. Officials with insufficient clearance could only access classified information with approval, on a case-by-case basis, from managers charged with controlling access.210 The DNI must make the development of this type of network a top priority because individual agencies will not move to develop it on their own. As Alexis Albion noted: “Part of the information sharing problem is political will and culture — bureaucracies still don’t want to give up control of information.”211 Agencies also lack an adequate incentive to invest their own resources in the development of an IC-wide network; they would rather free ride off of the efforts of other agencies or the DNI. The network should be designed so that the IC can allow unclassified parts of it to be accessible to people outside government. In this way, the IC could move toward establishing a global intelligence community. Robert Hutchings has advanced this idea, arguing that the IC should use unclassified channels to promote communication between government officials and experts outside government.212 As discussed in the section above on analysis, such outside expertise is particularly useful in strategic or long-term analysis. Individuals outside government could be given access to parts of the network that include unclassified information and analysis on strategic or long-term questions, without receiving access to parts of the network containing classified information. They would then be able to participate in an ongoing dialogue with other experts and government analysts, allowing a wider range of perspectives and expertise to be applied to key security and intelligence challenges. The protection of privacy and civil liberties must be a fundamental principle of the development of an information sharing network. The Markle Foundation Task Force noted that “any government undertaking to build such an information network would not be sustainable if the government did not build public trust by including protection of well-established civil liberties throughout that system.”213 This principle is particularly important because the network should include relevant information from corporations and other nongovernmental actors. Clear policies need to be in place to govern the acquisition, use, and retention of privately collected data.214 The Privacy and Civil Liberties Oversight Board should monitor these aspects of information sharing closely. F. The role of policymakers Even if the intelligence community collects, analyzes, and shares information superbly, its work will be futile unless that information is used by policymakers appropriately and effectively. In the Iraqi WMD case, policymakers placed improper pressure on the IC and distorted intelligence assessments. Although the WMD Commission found that political pressure did not cause analysts to alter their judgments, high-level Bush administration officials repeatedly asked the IC to produce evidence of Iraqi WMD programs and connections to Al Qaeda, and that constant drumbeat made it more difficult for analysts to do their work objectively. Chris Kojm described how this politicization

26

operates subtly upon analysts: “People in the IC know where the train is going and know what type of analysis will be looked upon with favor.”215 To prevent politicization and maximize the benefits of intelligence, policymakers must do several things. First, they must give the IC the freedom to make assessments independent of political pressure. There is a risk that the DNI could become the arm of the President. The President and DNI must ensure that he remains an independent actor, rather than becoming an advocate for administration policy. The DNI must also create a climate for independent thinking by subordinate officials and analysts by giving them autonomy to reach their own conclusions. Second, policymakers have to use intelligence capabilities wisely. This means asking good questions of the IC, bringing the IC into the policymaking process, and allowing intelligence judgments to influence the shaping of policy. Robert Hutchings criticized the tendency for one-way communication: “Policymakers are much too passive. They just react to what the intelligence agencies give them. They need to ask about the likely effect of policies being considered and how we could measure whether policies are succeeding.”216 To integrate intelligence more into the policymaking process, the government should establish mechanisms to give intelligence officials a greater role in policy discussions at the NSC, State Department, Pentagon, and other agencies. Greater exchange between policymakers and analysts could also be promoted by assigning some CIA analysts to Cabinet departments so that they would interact with policymakers on a regular basis.217 That would allow analysts to understand better what type of information policymakers need to make decisions and would give policymakers a better sense of what type of questions they should ask of the IC. Third, policymakers must act on intelligence responsibly. In the Iraqi WMD case, some high-level officials did not relay intelligence reporting to the public accurately. If policymakers distort intelligence information or assessments, members of Congress with access to the same intelligence should publicly correct them.218 Using intelligence responsibly also means reading important intelligence assessments and responding with appropriate action. Fewer than half a dozen Senators even read any of the 2002 NIE on Iraqi WMD activities, and Bush administration officials failed to adjust their postwar planning following two January 2003 NIC assessments projecting substantial resistance after a U.S. invasion by remnants of the old Baathist regime, disaffected Sunnis, and foreign fighters.219 Indeed, a July 2004 review by former intelligence officials concluded that the Bush administration “apparently paid little or no attention” to the prewar assessments.220 Inattentive or inflexible policymakers can squander even the most outstanding intelligence reporting and analysis. G. The role of Congress Since 9/11, Congress has advanced intelligence reform by holding hearings, conducting inquiries, and passing the Intelligence Reform Act. But Congress has made only minor changes to its own intelligence practices, even though they are widely viewed as dysfunctional — even by members of Congress and their staff.221

27

The primary problem in Congress is that the House and Senate intelligence committees, which contain the greatest congressional expertise on intelligence, have minimal influence over intelligence authorization or appropriations. This is the case in part because the intelligence committees share oversight responsibility with other committees that have jurisdiction over Cabinet departments containing intelligence agencies. The armed services committees, for instance, oversee the more than 80% of intelligence spending that is located in DOD.222 Yet those other committees do not make intelligence oversight a priority. One Senate Armed Services Committee member reportedly commented that the committee spends only five minutes on the intelligence budget each year.223 As a result, congressional oversight is woefully inadequate.224 This deficiency is compounded by a second problem: nearly all intelligence spending, including funding for the CIA and all intelligence agencies under DOD, is appropriated through national defense appropriation acts. These laws are written by the defense appropriations subcommittees of the House and Senate appropriations committees. (Funds for the intelligence units of the FBI, DHS, and State Department are appropriated through other acts.) Defense appropriations subcommittees give little attention to the intelligence budget while focusing on the rest of the defense budget, which is approximately ten times as large. Furthermore, they frequently override authorized priorities and appropriate much of the IC’s budget in supplemental spending bills, which makes it difficult for the IC to plan ahead.225 In these ways, Congress fails to use its intelligence expertise effectively in both the authorization and appropriations processes. As a result, the IC contains tremendous waste and intelligence shortcomings often do not get adequate attention. The 9/11 Commission report argued that strengthening congressional oversight is critical: “The other reforms we have suggested — for an NCTC and a DNI — will not work if congressional oversight does not change too.”226 The report recommended establishing either a joint congressional committee for intelligence or separate House and Senate intelligence committees that combine authorization and appropriations powers. The Commission also called for appropriating the IC’s funding to the DNI, not the Secretary of Defense; establishing intelligence subcommittees devoted to oversight; and abolishing term limits for intelligence committee members.227 The WMD Commission offered more modest recommendations to address the same problems, proposing that both houses create intelligence appropriations subcommittees and that Congress reduce the IC’s reliance on supplemental funding by developing annual budgets that include intelligence needs for the entire year.228 Congress has adopted few of these reform proposals, however, and it has taken only small steps to improve its intelligence oversight. These steps include the following: the Senate gave its intelligence committee responsibility for confirming all civilian nominees for intelligence posts; the House Intelligence Committee established an oversight subcommittee; and the Senate Intelligence Committee created an oversight subcommittee, abolished term limits for its members, and expanded its staff.229 The Senate also voted to establish an intelligence appropriations subcommittee, but that

28

change has not been implemented. Congress has taken no action on the more critical recommendation to establish intelligence appropriations subcommittees because members of the defense appropriation subcommittees and armed services committees have used their influence to block reforms that might decrease their power.230 Congress’ intelligence oversight responsibility is tremendously important because Congress is the only body capable of monitoring intelligence systematically. Neither the media nor the public can do the job because they do not have access to classified information — except when it is leaked.231 Congress must, in particular, conduct routine oversight of all aspects of the intelligence enterprise, rather than only focusing on oversight following an intelligence crisis or failure.232 This day-to-day oversight work does not grab headlines, but it is critical to U.S. national security. To conduct routine oversight effectively, Congress must streamline its committee structure and give greater authority to the intelligence committees. It should also establish intelligence appropriations subcommittees, some of whose members should serve on both the appropriations subcommittee and the intelligence committee. Oversight could be further improved through greater integration of the staff of national security and intelligence committees.233 By establishing some positions for staff to be shared between the intelligence and armed services committees, Congress could ensure that intelligence concerns are given greater attention in the defense appropriations process. H. Personnel The aphorism that any organization is only as good as its people applies fully to the IC. Frederick Hitz commented: “It’s a Washington tactic to try to solve problems by making structural change and throwing money at them, but that doesn’t get to the heart of the matter, which is people.”234 Today, the IC faces significant problems attracting and retaining highly qualified people because of dissatisfaction with pay, long work hours, and general perceptions that government does not properly reward a job well done. Many CIA operations officers earn just $60-70,000 a year for difficult and sometimes dangerous work, and the high rate of staff turnover at the FBI is partly a function of burnout.235 The arduous and lengthy security clearance process further discourages many job candidates. As noted in the section on collection above, clearance procedures make it very difficult for the IC to hire native speakers of languages used in the Middle East and South Asia. The IC is also failing to recruit enough specialists with certain types of technical and scientific expertise. The WMD Commission report stated that technical expertise in weapons systems “has fallen sharply in the past 10 years,” and that in biotechnology expertise the IC is “well behind the private sector.”236 The DNI must make the development of better IC-wide personnel policies one of his top priorities. The Bush administration has already announced that the DNI’s Chief Human Capital Officer will pursue the establishment of a joint personnel rotation system and more creative performance incentives. The new National Intelligence University (NIU)

29

system, whose structure and programming are still being developed, has the potential to address another personnel need by providing better training of analysts and officers. The NIU system is unlikely to provide the type of high-quality academic instruction offered by top public policy schools, but it could serve a useful role if its curriculum focuses on analytical and espionage tradecraft, interagency collaboration, and other issues specific to the IC.237 To supplement these steps, the DNI’s office should lead an effort to review and reform security clearance procedures to make them more flexible and fast-moving. The NSA offers a valuable model with its multi-level security clearance system and rapid pace of processing clearances — the agency completes them on average in just 79 days.238 Accelerating the clearance process should allow IC agencies to expand hiring of people for short stints of service or for repeated tours of duty at different points in their career. To date, the model intelligence officer or analyst has been a lifer — someone who spends 20 years or more in the IC. This model makes it very difficult for the IC to hire talented people who would like to serve for only a few years. As the career path of highly skilled people becomes increasingly mobile, the IC must change its vision of the model intelligence analyst or officer to include not just career civil servants but also people who come to the IC with valuable experience in other organizations, including corporations, NGOs, and academia, and who cycle between the public sector and other jobs throughout their careers. Adopting this new vision would help the IC achieve the important goal of hiring more officers or analysts with deep experience and expertise. At the same time, the IC must offer challenging and attractive career paths to those analysts and officers who are interested in devoting much of their career to government service. Today many people who spend their entire career in the same agency experience stagnation and dissatisfaction. In addition to establishing and requiring more rotations, the DNI should make it easier for officers and analysts to follow a career trajectory that includes long-term jobs in multiple agencies and opportunities for significant merit-based promotions and raises. Intelligence agencies should be able to describe this career path to job candidates to enhance their recruitment appeal.239 I. Adaptation and innovation Finally, the IC must develop a culture that fosters regular adaptation and innovation.240 Despite the IC’s significant reforms in the wake of 9/11, many agencies remain wedded to parochial perspectives and outdated practices, and the IC maintains elements of a risk- averse culture. While some changes have been made to encourage greater risk-taking — for instance, in 2002 the CIA rescinded a 1995 directive that placed restrictions on the recruitment of informants with criminal backgrounds — intelligence officers and analysts generally do not get rewarded for taking risks.241 The DNI should establish a task force to consider ways to routinize adaptation and innovation.242 One mechanism could be more frequent after-action reviews, in which the IC examines its performance in a certain area to identify weaknesses that need to be addressed.243 Currently, the IC does not conduct such “lessons learned” exercises nearly enough.244 The IC should also create incentives to encourage more risk-taking, so long

30

as that risk-taking is subject to adequate oversight and conforms to U.S. laws and regulations. Amy Zegart remarked: “Sometimes people should be rewarded for taking a risk even if they fail.”245 More systematically, the IC should develop feedback loops to drive institutional learning. For instance, analysis that reveals a need for more intelligence on a certain topic should lead the IC to modify its collection practices to gather that information. Today, analysts are not well enough integrated into the process for establishing collection priorities to enable that kind of feedback to take place.246 The DNI should create effective feedback loops as part of his more general effort to integrate the IC and manage it strategically.

Conclusion

This paper has focused on continuing deficiencies in intelligence and offered some recommendations for addressing them. That focus on shortcomings should not obscure the real achievements of the intelligence community. Since 9/11, the IC has played a central role in important national security successes, including the identification and targeting of Al Qaeda leaders, the discovery and penetration of the A. Q. Khan nuclear proliferation network, the conduct of countless combat and policing operations in Afghanistan and Iraq, and the prevention of additional terrorist attacks in the United States. The IC has also adopted many significant reforms that facilitate better integration, coordination, and information sharing. Moreover, the IC faces staggering demands. Policymakers are asking it to provide intelligence on very hard targets, such as Islamist terrorist groups, North Korea, and Iran. When faced with such targets, we must recognize the limits of intelligence. Even if the IC does almost everything right, it still might not be able to learn how many nuclear weapons North Korea possesses because North Korea’s political system and society are extremely closed. It is also unrealistic to expect the IC to detect every impending terrorist attack. Intelligence can never accomplish all that policymakers ask it to do. As a result, there are limits to what intelligence reform can achieve. Recognizing the IC’s successes and limitations, however, must not prevent government from making the intelligence enterprise work as well as it possibly can. Every improvement in the accuracy, insight, and effectiveness of the IC’s work can strengthen U.S. national security and save American lives. Congress and the executive must ensure that intelligence reform is not a brief episode in the wake of scandal or media attention, but rather an ongoing process guided by continuing and responsible leadership at the highest levels of government.

31

Interviews Conducted for this Paper Alexis Albion, Director of Policy, 9/11 Public Discourse Project David Barton, Minority Professional Staff, Senate Homeland Security and Government Affairs Committee Rajesh De, Counsel, Special Bipartisan Staff, Senate Homeland Security and Government Affairs Committee Hagar Hajjar, Legislative Fellow, Office of U.S. Rep. Christopher Shays Lee Hamilton, President, Woodrow Wilson International Center for Scholars Frederick Hitz, Diplomat in Residence, Princeton University Michael Hurley, Senior Director for Policy, 9/11 Public Discourse Project Robert Hutchings, Diplomat in Residence, Princeton University Christopher Kojm, President, 9/11 Public Discourse Project John McLaughlin, Senior Fellow, Johns Hopkins School of Advanced International Studies Edward Mills, Legislative Assistant, Office of U.S. Rep. Carolyn Maloney Joseph Nye, Sultan of Oman Professor of International Relations, Harvard University Benjamin Rhodes, Special Assistant to the President, Woodrow Wilson International Center for Scholars Amy Zegart, Associate Professor of Public Policy, University of California at Los Angeles Congressional official involved in intelligence issues Former intelligence official

32

Notes

1 I am grateful to Elizabeth Colagiuri, Mickey Edwards, Dawn Hewett, Robert Hutchings, Lorelei Kelly, Julia Kernochan, Joseph Nye, Benjamin Rhodes, Anne-Marie Slaughter, and Allison Stanger for helpful comments on earlier drafts of this paper. 2 Lee H. Hamilton, “Reflections upon the Findings and Recommendations of the 9/11 Commission and Intelligence Reform,” Remarks to the George Washington University Elliott School Advisory Board, October 21, 2004. 3 Walter Pincus recently reported that the intelligence authorization bill passed this year by the Senate Intelligence Committee allocates about $44 billion for the IC, though the budget figure remains classified. Walter Pincus, “GOP Senators Look to Shift Spy Management from CIA,” Washington Post, October 1, 2005. 4 The 15 agencies are the Central Intelligence Agency, Defense Intelligence Agency, National Security Agency, National Geospatial-Intelligence Agency, National Reconnaissance Office, and intelligence units of the Federal Bureau of Investigation, Department of Homeland Security, State Department, Energy Department, Treasury Department, Army, Navy, Air Force, Marine Corps, and Coast Guard. 5 This conception of a multilayered network was suggested to me by Robert Hutchings in a talk and interview. Robert Hutchings, “The Morning After: The Politics of Intelligence Reform,” International Relations faculty workshop, Princeton University, May 16, 2005; Interview of Robert Hutchings, May 17, 2005. 6 I do not discuss covert action or CIA paramilitary operations in the paper because they are not strictly intelligence activities. 7 Amy Zegart, “September 11 and the Adaptation Failure of U.S. Intelligence Agencies,” International Security, Vol. 29, No. 4 (Spring 2005), 78-111. 8 Zegart calculates that only 35 on the 340 recommendations made by the 12 reports were implemented by government. Zegart, “September 11 and the Adaptation Failure of U.S. Intelligence Agencies,” 88. 9 U.S. Senate Select Committee on Intelligence and U.S. House Permanent Select Committee on Intelligence, Joint Inquiry into Intelligence Community Activities before and after the Terrorist Attacks of September 11, 2001, Unclassified version of report, December 2002 (hereinafter the Joint Inquiry report). 10 The Joint Inquiry report, 34-39, 78. 11 Ibid., 77-78. 12 Ibid., 39-40, 49. 13 Ibid., 59-60. 14 Ibid., 39-40. 15 Ibid., 90-91, 110-112. 16 The Joint Inquiry report, Unclassified errata print of recommendations, December 2002, 2. 17 Ibid., 3. 18 Ibid., 4. 19 Ibid., 7. 20 Ibid., 11-12. 21 Ibid., 12. 22 National Commission on Terrorist Attacks upon the United States, The 9/11 Commission Report (New York: W.W. Norton & Co.), 2004 (hereinafter the 9/11 Commission report). 23 Ibid., 339. 24 Ibid., 358, 400. 25 Ibid., 342-343. 26 Ibid., 410. 27 Ibid., 353. 28 Ibid., 417. 29 Ibid., 419. 30 Ibid., 400-407. 31 Ibid., 407-416. 32 Ibid., 411. 33 Ibid., 415.

33

34 Ibid., 416-419. 35 Ibid., 346. 36 Ibid., 419-423. 37 Ibid., 425-426. 38 Senate Select Committee on Intelligence, Report on the U.S. Intelligence Community’s Prewar Intelligence Assessments on Iraq, July 7, 2004 (hereinafter the Senate Intelligence Committee report). 39 Ibid., 14. 40 Ibid., 18-24. 41 Ibid., 16. 42 Ibid., 24, 34. 43 Ibid., 16. 44 Ibid., 24. 45 Ibid., 34. 46 Report of a Committee of Privy Counselors, Chairman, Lord Butler of Brockwell, Review of Intelligence on Weapons of Mass Destruction, July 14, 2004 (hereinafter the Butler Committee report). 47 Ibid., 8. 48 Ibid., 7, 12. 49 Ibid., 141. 50 Ibid., 16, 143. 51 Ibid., 14. 52 Ibid., 146. 53 Commission on the Intelligence Capabilities of the United States regarding Weapons of Mass Destruction, Report to the President of the United States, March 31, 2005 (hereinafter the WMD Commission report). 54 Ibid., 9-10. 55 Ibid., 11. 56 Letter from the WMD Commission to the President accompanying the report, March 31, 2005. 57 The WMD Commission report, 11. 58 Ibid., 314, 317. 59 Ibid., 12-13. 60 Ibid., 14. 61 Ibid., 5, 17. 62 Letter from the WMD Commission to the President accompanying the report. 63 The WMD Commission report, 317. 64 Ibid., 19. 65 Ibid., 19-20. 66 Ibid., 20, 338. 67 Ibid., 22. 68 Ibid. 69 Ibid., 23. 70 Ibid., 25. 71 Ibid., 26. 72 Ibid., 26-27. 73 Ibid., 30-31. 74 The National Intelligence Program (NIP) includes all intelligence programs other than those exclusively supporting the needs of the military. Intelligence programs only supporting the military include Tactical Intelligence and Related Activities (TIARA), which covers programs supporting a particular armed service, and the Joint Military Intelligence Program (JMIP), which comprises programs that support multiple- service or theater-level defense needs. See William Odom, Fixing Intelligence: For a More Secure America, 2nd edition (New Haven: Yale University Press, 2004), 205-206. 75 United States Senate Committee on Governmental Affairs, “Summary of Intelligence Reform and Terrorism Prevention Act of 2004,” December 6, 2004, 1-4, <http://hsgac.senate.gov/_files/ConferenceReportSummary.doc>. 76 Ibid,, 8. 77 Ibid., 8-9.

34

78 Ibid., 9. 79 Ibid., 5-6. 80 Ibid., 6-7. 81 Ibid., 7. 82 Ibid., 7-8. 83 Ibid., 9. 84 Ibid., 3. 85 Ibid., 10. 86 Ibid., 3; The 9/11 Commission report, 403. 87 Interview of Lee Hamilton, June 15, 2005; Interview of David Barton, June 13, 2005; Comments by John Lehman, 9/11 Public Discourse Project panel, “Challenges Facing the DNI,” Woodrow Wilson International Center for Scholars, June 13, 2005. 88 Interview of John McLaughlin, June 26, 2005. 89 Scott Shane, “Intelligence Chief to Retain Personal Powers,” New York Times, June 10, 2005; Interview of Lee Hamilton. 90 Alfred Cumming and Todd Masse, “Intelligence Reform Implementation at the Federal Bureau of Investigation: Issues and Options for Congress,” Congressional Research Service, August 16, 2005, 7; Walter Pincus and Dan Eggen, “Spy Chief Gets More Authority over FBI,” Washington Post, June 30, 2005; Siobhan Gorman, “Bush Boosts Power of Intelligence Chief, Forms FBI Spy Office,” Baltimore Sun, June 30, 2005. 91 For a more complete discussion of the DNI’s statutory authorities, see Richard A. Best Jr., Alfred Cumming, and Todd Masse, “Director of National Intelligence: Statutory Authorities,” Congressional Research Service, April 11, 2005. 92 Douglas Jehl, “Nominee Promises Tighter Control over U.S. Intelligence Agencies,” New York Times, April 13, 2005; Comments by Siobhan Gorman, 9/11 Public Discourse Project panel, “Challenges Facing the DNI.”| 93 The WMD Commission report, 314. 94 Douglas Jehl, “Republicans See Signs That Pentagon Is Evading Oversight,” New York Times, September 29, 2005. 95 Quoted in Gorman, “Bush Boosts Power of Intelligence Chief, Forms FBI Spy Office.” 96 Comments by Robert Hutchings, “The Morning After: The Politics of Intelligence Reform;” Interview of Robert Hutchings. 97 Interview of John McLaughlin. 98 Robert Hutchings made this point in e-mail correspondence, August 15, 2005. 99 The Joint Inquiry report, 49. 100 Former CIA official John Gannon has argued that the IC is not paying enough attention to China. Robert Hutchings noted that the IC is thin on expertise related to Latin America and Africa. Comments by John Gannon, 9/11 Public Discourse Project panel, “CIA and FBI Reform,” Woodrow Wilson International Center for Scholars, June 6, 2005; Interview of Robert Hutchings. 101 Comments by Siobhan Gorman, 9/11 Public Discourse Project panel, “Challenges Facing the DNI.” 102 The WMD Commission report, 22. 103 Comments by David Kay, Hearing of the Senate Select Committee on Intelligence, August 18, 2004. 104 Odom, Fixing Intelligence, 120-129. 105 Benjamin Rhodes made this point in e-mail correspondence, August 31, 2005. 106 White House document, “Administration’s Response to Recommendations of the Commission on the Intelligence Capabilities of the United States regarding Weapons of Mass Destruction,” June 29, 2005, accessible at < http://www.whitehouse.gov/news/releases/2005/06/20050629-5.html>. 107 Interview of Amy Zegart, June 22, 2005. 108 Ibid. 109 Gregory Treverton asserts that the cultural differences among — and with — agencies are so great that the idea of an intelligence “community” is “somewhere between a fiction and an aspiration.” Gregory F. Treverton, Reshaping National Intelligence in an Age of Information (Cambridge: Cambridge University Press, 2003), xiii. 110 Interview of David Barton. 111 Interview of congressional staffer involved in intelligence issues, July 1, 2005.

35

112 Scott Shane, “On First Day on Job, Intelligence Director Sizes up His Challenges,” New York Times, April 23, 2005. 113 Interview of Rajesh De, June 23, 2005. 114 Interview of Chris Kojm, June 14, 2005. 115 Bob Drogin, “Two Agencies Melding Minds on Intelligence,” Los Angeles Times, December 31, 2004. 116 Markle Foundation Task Force on National Security in the Information Age, Creating a Trusted Information Network for Homeland Security, December 2003, accessible at <www.markletaskforce.org>. 117 The Economist, “Special Report: Intelligence Reform,” March 19, 2005, 31. 118 Ibid. 119 Douglas Jehl, “Spy Satellite Programs Are in Upheaval,” New York Times, September 30, 2005. 120 White House document, “Administration’s Response to Recommendations of the Commission on the Intelligence Capabilities of the United States regarding Weapons of Mass Destruction.” 121 Mark Hosenball, “Spies: Who’s in Charge?” Newsweek, October 10, 2005; Pincus, “GOP Senators Look to Shift Spy Management from CIA.” 122 Douglas Jehl, “Little Authority for New Intelligence Post,” New York Times, October 14, 2005. 123 Pincus, “GOP Senators Look to Shift Spy Management from CIA.” 124 Siobhan Gorman, “Intelligence Gap,” National Journal, March 19, 2005, 827-831; Lindsey Moran, “More Spies, Worse Intelligence?” New York Times, April 12, 2005. 125 The Economist, “Special Report: Intelligence Reform,” 31. 126 Sydney G. Freeburg Jr., “Men and Machines,” National Journal, March 19, 2005, 835. 127 Interview of Lee Hamilton. 128 Comments by John Gannon and Chitra Ragavan, 9/11 Public Discourse Project panel, “CIA and FBI Reform.” 129 Robert Baer, “Wanted: Spies like Us,” Foreign Policy, March/April 2005, 69. 130 Ibid., 70. 131 Jack Devine, “America Badly Needs More and Better Spies,” Financial Times, April 14, 2005. 132 Greg Miller, “Goss Cites ‘Real Progress’ in His Year As CIA Chief,” Los Angeles Times, September 23, 2005. 133 Douglas Jehl, “CIA Reviews Security Policy for Translators,” New York Times, June 8, 2005. 134 Gorman, “Intelligence Gap,” 831. 135 Interview of Frederick Hitz, June 16, 2005. 136 Richard A. Best Jr., “Intelligence Issues for Congress,” Congressional Research Service, June 3, 2005, 5. 137 Interview of Robert Hutchings. 138 William Finnegan, “The Terrorism Beat: How Is the NYPD Defending the City?” New Yorker, July 25, 2005. 139 Ibid, 60. 140 Ibid, 64. 141 Ibid, 68. 142 Interview of Michael Hurley and Alexis Albion, June 13, 2005. 143 Dana Priest, “Help from France Key in Covert Operations,” Washington Post, July 3, 2005. 144 Ibid. 145 Miller, “Goss Cites ‘Real Progress’ in His Year As CIA Chief.” 146 The Joint Inquiry report, 55; The WMD Commission report, 15-16. 147 Treverton, Reshaping National Intelligence in an Age of Information, 220. 148 Scott Shane, “Eavesdropping Isn’t Easy, The Master at It Says,” New York Times, August 17, 2005. 149 The WMD Commission report, 16. 150 Rich Tuttle, “NGA Project Would Ease Burden of Intelligence Analysts,” Aerospace Daily and Defense Report, June 28, 2005. 151 Best, “Intelligence Issues for Congress,” 4. 152 Tim Wiener, “Air Force Seeks Bush’s Approval for Space Arms,” New York Times, May 18, 2005. 153 Jehl, “Spy Satellite Programs Are in Upheaval.” 154 Odom, Fixing Intelligence, 120-129; Scott Shane, “Panel Seeks to Shift Money from Satellites to Agents,” New York Times, June 15, 2005. 155 Odom, Fixing Intelligence, 120-129.

36

156 Treverton, Reshaping National Intelligence in an Age of Information, 11-18. 157 Ibid, 2. 158 Gregory F. Treverton and Peter A. Wilson, “True Intelligence Reform Is Cultural, Not Just Organizational Chart Shift,” Christian Science Monitor, January 13, 2005. 159 John Hollywood et al., Out Of the Ordinary: Finding Hidden Threats by Analyzing Unusual Behavior (Santa Monica, CA: RAND Corporation), 2004. 160 Gorman, “Intelligence Gap,” 832. 161 The WMD Commission report, 12. 162 Treverton, Reshaping National Intelligence in an Age of Information, 64-65; Interview of Amy Zegart. 163 Nicole Gaouette, “Extension of the Patriot Act OKd,” Los Angeles Times, July 30, 2005; Baltimore Sun, “Patriot Act Renewal,” October 5, 2005. 164 See, for instance, Jordan Tama, “Is Torture Justified in the Fight against Terrorism?” International Affairs Review, Vol. 13, No. 2 (Fall 2004), 80-85. 165 Dana Priest, “Senate Urged to Probe CIA Practices,” Washington Post, April 23, 2005. 166 Eric Schmitt, “Senate Moves to Protect Military Prisoners despite Veto Threat,” New York Times, October 6, 2005. 167 Interview of Edward Mills and Hagar Hajjar, June 13, 2005. 168 Robert Hutchings, “Strategic Choices, Intelligence Challenges,” Speech at the Woodrow Wilson School of International and Public Affairs, December 1, 2003, in Calvin M. Logue et al., editors, Representative American Speeches, 2003-2004 (New York: H. W. Wilson Co., 2004), 3-12. 169 Interview of Joseph Nye, June 30, 2005; Joseph Nye, “How to Give Our Leaders the Facts They Need,” Financial Times, September 24, 2004. 170 Interview of former intelligence official, June 20, 2005; Douglas Jehl and Eric Schmitt, “Data Is Lacking on Iran’s Arms, US Panel Says,” New York Times, March 9, 2005. 171 Dafna Linzer, “Iran Is Judged 10 Years from Nuclear Bomb,” Washington Post, August 2, 2005. 172 Douglas Jehl, “Intelligence Briefing for Bush Is Overhauled,” New York Times, July 20, 2005. 173 See Michèle A. Flournoy and Shawn W. Brimley, “Strategic Planning for U.S. National Security: A Project Solarium for the 21st Century,” Paper Commissioned by the Princeton Project on National Security, accessible at <www.wws.princeton.edu/ppns>. 174 Interview of John McLaughlin; Richard A. Best, Jr., “The Director of National Intelligence and Intelligence Analysis,” Congressional Research Service, February 11, 2005, 5; Porter J. Goss, Testimony before the Senate Select Committee on Intelligence, “Global Intelligence Challenges 2005: Meeting Long- Term Challenges with a Long-Term Strategy,” February 16, 2005. 175 Interview of Joseph Nye. 176 Interview of Michael Hurley and Alexis Albion. 177 Treverton, Reshaping National Intelligence in an Age of Information, 133. 178 National Intelligence Council, Mapping the Global Future, December 2004. 179 Interview of Amy Zegart; Comments by David Kay, Hearing of the Senate Select Committee on Intelligence, August 18, 2004. 180 Cumming and Masse, “Intelligence Reform Implementation at the Federal Bureau of Investigation;” David Johnston, “Antiterror Head Will Help Choose an FBI Official,” New York Times, June 12, 2005. 181 U.S. Department of Justice Office of the Inspector General, The Federal Bureau of Investigation’s Efforts to Hire, Train, and Retain Intelligence Analysts, Audit Report 05-20, May 2005, x. 182 Cumming and Masse, “Intelligence Reform Implementation at the Federal Bureau of Investigation,” 30- 31. 183 U.S. Department of Justice Office of the Inspector General, The Federal Bureau of Investigation’s Efforts to Hire, Train, and Retain Intelligence Analysts, xii. 184 Comments by Chitra Ragavan, 9/11 Public Discourse Project panel, “CIA and FBI Reform.” 185 The WMD Commission report, 25. 186 Stephen Grey, “Follow the Mullahs,” Atlantic Monthly, November 2004, 44-45. 187 The Joint Inquiry report, 69. 188 Interview of Chris Kojm. 189 Richard Clarke, “Building a Better Spy,” New York Times Magazine, May 22, 2005, 20. 190 Interview of Frederick Hitz. 191 Treverton, Reshaping National Intelligence in an Age of Information, 247.

37

192 Interview of Amy Zegart. 193 The Economist, “Know Thine Enemy,” May 7, 2005. 194 Mark Lowenthal, “The Role of Intelligence in the United States Today,” Remarks at a Potomac Institute Seminar on the Revolution in Intelligence Affairs, New Mexico State University, September 15, 2003. 195 The WMD Commission report, 13. 196 White House document, “Administration’s Response to Recommendations of the Commission on the Intelligence Capabilities of the United States regarding Weapons of Mass Destruction.” 197 Interview of Frederick Hitz; Comments by John Gannon, 9/11 Public Discourse Project panel, “CIA and FBI Reform.” 198 Remarks by Lee H. Hamilton, “Reflections upon the Findings and Recommendations of the 9/11 Commission and Intelligence Reform.” 199 Interview of former intelligence official, June 20, 2005; Comments by Richard Thornburgh, 9/11 Public Discourse Project panel, “CIA and FBI Reform.” 200 Comments by Richard Thornburgh, 9/11 Public Discourse Project panel, “CIA and FBI Reform.” 201 US Department of Justice Office of the Inspector General, Review of the Terrorist Screening Center, Audit Report 05-27, June 2005. 202 The WMD Commission report, 14. 203 Comments by Richard Thornburgh, 9/11 Public Discourse Project panel, “CIA and FBI Reform.” 204 Chitra Ragavan and Carol Susan Hook, “Fixing the FBI,” U.S. News & World Report, March 28, 2005. 205 Lara Jakes Jordan, “Homeland Security Information Network Criticized,” Washington Post, May 11, 2005. 206 Chitra Ragavan, “Case Management,” U.S. News & World Report, June 20, 2005. 207 The 9/11 Commission report, 416-419. 208 Markle Foundation Task Force on National Security in the Information Age, Creating a Trusted Information Network for Homeland Security. 209 Interview of John McLaughlin. 210 The 9/11 Commission report, 418; The WMD Commission report, 572. 211 Interview of Alexis Albion and Michael Hurley. 212 Talk by Robert Hutchings, “The Morning After: The Politics of Intelligence Reform.” 213 Markle Foundation Task Force on National Security in the Information Age, Creating a Trusted Information Network for Homeland Security, 6. 214 Ibid., 9-10. 215 Interview of Chris Kojm. 216 Interview of Robert Hutchings. 217 Stan A. Taylor and David Goldman recommend this in “Intelligence Reform: Will More Agencies, Money and Personnel Help?” Intelligence and National Security, Vol. 19, No. 3 (Autumn 2004), 428. 218 This idea was suggested to me by a former intelligence official in an interview. 219 Robert Hutchings cited these examples in e-mail correspondence, August 15 and 16, 2005. 220 Douglas Jehl, “Report Says White House Ignored CIA on Iraq Chaos,” New York Times, October 13, 2005. 221 Lee Hamilton noted in an interview: “I don’t know anyone on the intelligence committees who thinks they’re doing a good job.” 222 This discussion of congressional oversight draws on Best, “Intelligence Issues for Congress,” 7. 223 Remarks by Lee Hamilton, “Reflections upon the Findings and Recommendations of the 9/11 Commission and Intelligence Reform.” 224 Current and former members of Congress discussed congressional shortcomings in intelligence at a 9/11 Public Discourse Project panel, “Congressional Reform,” Woodrow Wilson International Center for Scholars, July 11, 2005. 225 Letter from Tom Kean and Lee Hamilton to U.S. Reps. Christopher Shays and Carolyn Maloney, March 14, 2005. 226 The 9/11 Commission report, 420. 227 Ibid., 420-421. 228 The WMD Commission report, 20, 338.

38

229 Comments by Jane Harman, 9/11 Public Discourse Project panel, “Challenges Facing the DNI;” Interview of Rajesh De; Walter Pincus, “Senate Realigns Intelligence Procedures,” Washington Post, December 23, 2004. 230 Interview of Edward Mills and Hagar Hajjar. 231 Walter J. Oleszek makes this point in “A Perspective on Congress’s Oversight Function,” Congressional Research Service, December 23, 2004, 14. 232 Robert Hutchings made this point in his talk, “The Morning After: The Politics of Intelligence Reform.” 233 Mickey Edwards made this suggestion in comments at a meeting of the Princeton Project on National Security Working Group on Foreign Policy Infrastructure and Global Institutions, August 12, 2005. 234 Interview of Frederick Hitz. 235 Interview of Lee Hamilton; Gorman, “Intelligence Gap,” 832; Comments by Richard Thornburgh, 9/11 Public Discourse Project panel, “CIA and FBI Reform.” 236 The WMD Commission report, 13. 237 Robert Hutchings argued in e-mail correspondence that it is unlikely that the NIU would teach anything of academic merit, August 16, 2005. 238 Drogin, “Two Agencies Melding Minds on Intelligence;” Jehl, “CIA Reviews Security Policy for Translators.” 239 For more detailed recommendations on intelligence personnel, see Center for American Progress, Better Spies, Better Intelligence: A Progressive Strategy for Creating a Professional Intelligence Corps, April 2005. 240 Amy Zegart describes how the IC did not adapt effectively during the 1990s in Zegart, “September 11 and the Adaptation Failure of US Intelligence Agencies.” 241 Best, “Intelligence Issues for Congress,” 5; Interview of Amy Zegart. 242 IC veteran Deborah Barger has taken this argument further, calling for a Revolution in Intelligence Affairs that would fundamentally transform the IC. Deborah G. Barger, Toward a Revolution in Intelligence Affairs (Santa Monica, CA: RAND Corporation), 2005. 243 The WMD Commission report recommended conducting these types of “lessons learned” exercises, 20. 244 Interview of Amy Zegart. 245 Ibid. 246 Comments by Siobhan Gorman, 9/11 Public Discourse Project panel, “Challenges Facing the DNI.”

39