The American Interest
Policy, Politics & Culture
object(WP_Session)#92 (5) { ["session_id:protected"]=> string(32) "9f3c8d10d2fdb35f469d546ec23244f0" ["expires:protected"]=> int(1414135931) ["exp_variant:protected"]=> int(1414135571) ["container:protected"]=> array(1) { ["ai_visit_counter"]=> int(0) } ["dirty:protected"]=> bool(true) }
Unfinished Business

Ten years after 9/11, intelligence reform is still a work in progress.

Published on February 2, 2012

In the wake of September 11, 2001, the U.S. government, largely driven by the report of the 9/11 Commission, undertook a major reform of its intelligence community. It was not as though an awareness of the system’s deficiencies was absent before the 9/11 attacks; as the records of several high-powered Federal commissions attest, there had been several efforts to achieve reform. 

The basic structural problem stemmed from the piecemeal proliferation of 16 intelligence agencies during the Cold War, and the evolution of their missions and technical capabilities. Over time, what is called the “Intelligence Community” had become a fragmented and inefficient organizational arrangement in which capabilities and responsibilities and budget authorities were profoundly misaligned. There was often wasteful competition and duplication in which intelligence was often not shared among different agencies, and integrated operations were the exception, not the rule. Yet despite a reasonable consensus on the nature of the problem, reform efforts before 9/11 produced only modest improvements before a shock to the system laid bare its shortcomings. Given the turnover of administrations and of the membership of congressional select intelligence committees, and how the inherent secrecy of the business precluded public debate and understanding of the structural nature of the shortcomings, the critical mass for change didn’t come until that disaster.

The Intelligence Reform and Terrorist Prevention Act (IRTPA) of 2004 sought to reshape the ungainly product of the incremental process that began in 1947 into an effective, integrated system appropriate to the diverse threats and technological environment of the 21st century. IRTPA, however, was written, debated and passed under heavy time pressures. It was in consequence laden with compromises that bequeathed to the first and subsequence Directors of National Intelligence (DNI) many ambiguities. Ironically, I myself testified during its formation against its passage as it was written, arguing that it left too many unresolved issues and did not establish the authority of the DNI firmly enough. There is nothing surprising about this. Though few remember it today, the seminal National Security Act of 1947 was amended not once but four times in the years immediately following its passage. That’s how our system works. The July 2008 revision of Executive Order 12333, which initially implemented IRTPA, filled in some of the gaps in the original legislation. In the more than three years since that time, however, little further progress has been made, and in fact, under the current administration, with a few exceptions, the overall momentum toward integration has slowed. 

The fundamental concept of putting an official like the Director of National Intelligence in charge of all aspects of the nation’s intelligence, the concept at the heart of the 2004 Intelligence Reform and Terrorist Protection Act, is sound. I saw firsthand on many issues how, with White House and Congressional support, the DNI can advance improvements in the intelligence community (IC) that enable it to keep pace with the dynamic threats America faces. We are not yet where we need to be, however. Intelligence reform must continue with further legislative and executive order provisions, and with strong support for the office of the DNI.

There are two components of any successful reform that attempts to combine previously independent agencies under a new higher level official. One is the structural work of legislation, executive orders, and establishing revised authorities and responsibilities. The other is the establishment of the official and personal authority of the new higher level official in practice. Whatever is written on paper, unless Congress and the White House support that official’s authority, and unless the agencies in the new structure give up their previous direct relationships and follow his or her direction, the reforms will be ineffective. Both of these components of intelligence reform require further work before the country will achieve the overall intelligence capability it needs. 

One good example that illustrates what has been accomplished by IRTPA and what needs to be done is the concept of mission management. A mission manager’s job is to drive integration and improvement of all aspects of the performance of the IC on issues of particularly high priority to national security. The IRTPA explicitly established mission managers both for regional intelligence challenges—Iran and North Korea—and for functional challenges: counterproliferation and counterterrorism. When I became DNI in early 2009, I found that several of our mission management teams were working well, but that many others were struggling and some did not even exist, even for high priority intelligence challenges. We raised the visibility and priority of these efforts, and my successor, Director James Clapper, has greatly expanded the concept of mission management. There are now some 16 National Mission Managers. 

Are these efforts working? I found that being effective as a mission manager required high-level and detailed support from the DNI himself in order to overcome the many remaining obstacles to integrated intelligence, such as agency restrictions on sharing of intelligence, agency control of personnel assignments to missions and agency control of budget execution. It took high-level attention and intervention to align agency priorities with overall national intelligence priorities. Perhaps over time the agencies will naturally support the mission managers, but this is not yet the case. It will be a challenge for the DNI and his staff to ensure that the 16 national mission managers receive the high-level support they need to be effective. This will not happen without a strong and aggressive DNI, despite the promising but still gradual change in the culture of IC personnel in favor of deeper habits of cross-agency cooperation. 

Mission management is just one of several areas in which, not legislation, but the actual working experience of the still-new role of the DNI is critical and in which progress is being made. During my time as DNI, I established four strategic goals in this regard: enabling wise national security policies; supporting effective national security action; delivering balanced and improved technical and budgeting capabilities; and, above all, pushing the system to operate as a single integrated team. These goals resulted in new and improved approaches to delivering the President’s Daily Briefing, better support for the mission in Afghanistan and, more broadly, the mission against al-Qaeda, a rationalized science and technology acquisition process (especially in the critical area of cyber security) and a significantly revamped budget preparation processes for the intelligence community. These latter two processes will become vastly more important in an era of flat or declining budgets for national intelligence. We now have a firm basis to build on in both of these areas.

I was unable to make the progress I hoped for in certain other non-legislative areas. The DNI is not yet properly placed in the decision-making processes regarding national security. Some believe that the DNI can operate behind the scenes of policy deliberations at the highest level without personally participating. A regional or functional senior intelligence analyst, according to this line of argument, should attend the high-level policy meetings. Still others have the odd notion that the President’s intelligence officer should give an intelligence assessment at the beginning of a policy discussion and then leave the room. Thanks to this notion, I was excluded from the final White House meetings that decided this Administration’s Af/Pak policy in the autumn of 2009. 

This way of thinking about the role of the DNI does the President and the country a disservice. The DNI is responsible for intelligence judgments on major issues. He or she knows the strengths and weaknesses of the organizations and individuals who formed those judgments, and is responsible for allocating overall intelligence resources. He or she also has cabinet-level understanding of the country’s and administration’s needs. Therefore it is the DNI who must advise his fellow NSC members and the President on the big issues. He or she must be included in all the policy deliberations so that they do not stray from a realistic understanding of both adversaries and friends. 

Structural Issues


hat said, there is no way to permanently establish the specific manner in which future DNIs will actually operate within a given administration. As with the National Security Council system over the years, such considerations will always depend on the personality and choices of the President and his key cabinet associates. Nevertheless, there are several structural issues that can be stabilized for the future. Some of these reforms must involve legislation. Others can make do with Executive Orders.

Two main fault lines run through the structure created by the IRTPA: the relationship of the DNI to the intelligence agencies in the Department of Defense and the relationship of the DNI to the CIA. There are also occasional cracks between the responsibilities of the Department of Homeland Security and the IC, as well as between the law enforcement responsibilities of the FBI and the intelligence responsibilities of the community as a whole.

Let’s look at the first fault line, between the DNI and the Defense Department. Although located within the Department of Defense, which calls them “combat support agencies”, the National Security Agency, the National Geospatial-Intelligence Agency, the Defense Intelligence Agency and the National Reconnaissance Office. By volume, they provide the bulk of their tactical-level intelligence to the Department of Defense, which has the greatest demand for it in planning and conducting operations in the field. However at the same time they are able to provide top-notch support throughout the government for both policymaking and effective action. There are at least two major reasons for this. First, the traditional divide between military intelligence and national intelligence no longer exists. All major national security challenges facing the United States mix military objectives with diplomatic and economic objectives. The major intelligence agencies in the Defense Department realize that this cannot be otherwise, that military success depends on understanding the political, cultural and economic aspects of foreign countries as well as their order of battle. These Defense intelligence agencies therefore support both military commanders and units and civilian officials with functional intelligence in their areas of specialty on all aspects of countries and groups of interest to the United States. Second, Defense intelligence agency leaders have broad backgrounds and outlooks. They are generally senior military intelligence officers (the current director of the National Geospatial Intelligence Agency is a civilian defense official) with multiple joint operational assignments and first-hand experience in supporting different types of intelligence “customers.” They know how to direct their agencies to support a full range of officials, whether in Washington or in the field.

One area, however, where the relationship between the intelligence community and the Department of Defense needs to be improved is in operations against terrorist groups. Currently, the Department of Defense conducts operations against al-Qaeda and other terrorist groups as “traditional military activities” under its Title 10 authorities. Direct action, led by the CIA, is conducted as covert action under Title 50. Over the past decade, the military officers and intelligence officials leading these operations have learned to cooperate well. However, the currently divided authorities take time and inordinate legal and staff work to work out chains of command. These one-off arrangements have on occasion caused delays in execution that have resulted in missed opportunities. These operations would be more effective if they were combined into integrated joint interagency task forces (JIATFs) whose commanders would have both sets of authorities and whose staffs would have skill sets in both areas. For time-sensitive operations against agile enemies, unity of command and integration of capabilities are always more effective than stovepiped cooperation, no matter how well rehearsed. 

There would need to be a new legislative basis for such a joint task force: a “Title 60.” Under Title 60 operations against a terrorist group, either a senior CIA officer or a senior military officer would command the task force with a deputy from the other organization. The staff would include both military and Intelligence Community officers, and the task force would have the authority to use all the capabilities of both the Department of Defense and the CIA and other intelligence agencies. It would gather intelligence both through the intelligence agencies’ collection capabilities and military reconnaissance systems. It would command aircraft and drones—whether Defense- or CIA-owned—for both surveillance and attack missions. It would command special operations strike teams and conventional military task forces, work directly with foreign military organizations and intelligence organizations, and recruit and train paramilitary forces under intelligence authorities. It would draw on the Defense Department strengths in planning, resourcing and conducted long campaigns, but have the budget flexibility and contracting capability of the CIA to come up with quick and imaginative responses to immediate problems. 

Title 60 legislation would have to deal with some difficult issues that have kept Title 10 authorities separate from Title 50 authorities in the past. Most of these issues are rooted in Cold War history, and several have become completely obsolete. For example, these operations would be secret, but would they also be covert—that is, would they be conducted under the concept of plausible deniability? That concept, codified in Title 50, made sense when the U.S. government wanted to take lethal action against the Soviet Union or its proxies without the risk of escalation should the action be officially acknowledged. This concern is irrelevant to contemporary counterterrorist operations, which are often conducted in areas where weak states cannot enforce their own sovereignty. The May 2011 raid that killed Osama bin Laden was technically conducted as a covert action under Title 50, but I can think of no operation in recent memory that was less intended to be plausibly denied. 

Other issues are similarly manageable. Would military forces involved in these operations be subject to the Geneva Convention and enjoy Status of Forces protections? Yes, they should, but in reality the Geneva Convention is of little practical protection for the safety of service members in dealing with groups like al-Qaeda, al-Shabaab in Somalia and the Haqqani network along the Afghanistan border.

To whom would these Title 60 task forces report? Like all joint commanders in the Department of Defense, they would submit plans outlining their objectives, their strategy, the resources they would require and the authorities they would exercise. The plans would be fully staffed through the Defense Department and the IC, and would be approved by the President after final staffing by the National Security Council. Then, in execution, the task forces would report to the President through the National Security Council staff. 

Finally, how would Congress be notified? Both House and Senate defense and intelligence oversight committees should receive briefings on Title 60 operations. For one operation during my time as DNI, we formed a team of Defense Department and IC members to explain an operation to the leadership of all four committees. That procedure worked well in that case, and could be used routinely. 


he second fault line, the division of authority between the DNI and the CIA, was difficult during my tenure as DNI. Before IRTPA established the Director of National Intelligence, the Director of the CIA had the additional job of Director of Central Intelligence, responsible for coordinating the activities of all the different intelligence agencies. In practice, most Directors of the CIA found their time dominated by the responsibilities of running the CIA, and in interagency issues, the CIA most often had its way. 

Many individuals in the CIA understand the importance of intelligence integration. They work constructively with other intelligence agencies and the new DNI organization as a team, understanding that the CIA is important but should not necessarily be dominant. However, the CIA has often used the leverage and influence left over from its days of dominance to act independently, undercutting the authority of the DNI and attempting either to gain leadership of community intelligence activities or to act independently of them. The CIA often used its direct relationships with the White House, Congress and the intelligence services of other countries for this purpose. 

Most of the clashes I had with the CIA came when I challenged its separate means of influence and leverage. When, after careful and prolonged study and consultation, I signed a directive that specified that CIA station chiefs would in the great majority of cases (but not all) continue to be representatives of the DNI in integrating the interests of the entire IC within a given country, Director Leon Panetta appealed my decision to the White House. When I attempted to direct a more disciplined approach to the formulation and supervision of covert action activities, the CIA continued to work directly with the National Security Council staff, arguing that any attempt to impose principles, standards and procedures on covert action would impair its effectiveness. In both cases, the White House worked out compromise solutions that left the CIA with a great deal of autonomy and weakened the DNI’s authority.

The CIA’s basic organization also acts against the broader national interest in integrated intelligence. It is one of the peculiar legacies of history that the national clandestine service, responsible for both recruiting spies and covert action, is in the same agency as the national analytical organization. In most other countries they are separate. There are real disadvantages to this arrangement. The action-oriented, can-do culture of the national clandestine service dominates and sometimes intimidates the reflective, critical analytical culture of the directorate of analysis. To have the primary analytical organization in the IC paired with the organization that gathers human intelligence also makes less sense than it did in the past. While human intelligence will play an indispensable role in the future, a greater proportion of intelligence reports will be signals intelligence gathered by the National Security Agency. It is therefore time to divide the CIA into two separate agencies: a human intelligence and covert action service and a central all-source analytical agency, both reporting to the DNI. 

For the National Clandestine Service, human intelligence gathering should be its primary focus, with covert action a secondary mission. Right now, the CIA is conducting a major worldwide campaign against al-Qaeda, but this campaign will not continue forever. In time, covert operations will decline to an historically more normal level. In addition, the sophisticated paramilitary capability within the Defense Department, developed since 9/11 by Special Operations Command, is better equipped than the CIA to handle prolonged paramilitary activities, which by their nature do not remain covert for long. 

The Defense Intelligence Agency also gathers human intelligence and its activities are closely coordinated with the CIA. Although military human intelligence officers should be organized, trained and equipped by their parent military services, it makes sense to assign them for operational purposes into this new, separate human intelligence agency, whose deputy should be a military intelligence officer. Human intelligence provides absolutely vital context and often-priceless detail for assisting policymakers in Washington and operators in the field. We need a separate agency for this mission.

 For its secondary mission of covert action, a combined CIA and military national clandestine service should report to the DNI rather than directly to the National Security Council staff. Covert action programs under the supervision of inexperienced National Security Council staffers have been the cause of major setbacks to American national security interests when they have gone astray. The President and his staff should approve them and be kept informed of their progress, but the responsibility for supervising such operations, should belong to his confirmed principal intelligence officer, the DNI. Those who argue against this arrangement sometimes cite the importance of the secrecy, flexibility, speed and responsiveness of the direct CIA-White House link, but these are exactly the qualities that have caused disaster in the past, from the Bay of Pigs to Iran-Contra to so-called black sites. Another layer in the chain of command need not add delay. The Department of Defense conducts equally quick, flexible and sensitive special operations under the Secretary of Defense and his staff without the Special Operations Command or the military services reporting directly to the White House. 

The Directorate of Analysis in the CIA is the most capable all-source analytical agency in the intelligence community. Despite the well-known mistakes in analysis the CIA has made, not least the faulty National Intelligence Estimate on Iraqi Weapons of Mass Destruction, policymakers over the years have justifiably valued its analytical skills, databases and deep expertise. Creation of a separate, all-source analytical agency reporting directly to the DNI would be a very positive step. 

One element of this reorganization would be incorporating part of the all-source analytical capability of the Defense Intelligence Agency into the new agency. Most DIA analysts would still support military commanders and units, but all-source analysts with military specialties should be combined with the all-source analysts of the CIA. As already noted, the end of the Cold War rendered obsolete the historical distinctions between national and military intelligence. Combining military and civil analysts into a single, all-source analytical organization to support civilian and military policymakers and operations in the field would be a major step forward. 

The major civil/military fault line I encountered is between the Department of Homeland Security, an organization nearly as new as the Directorate of National Intelligence itself, and the National Security Agency—specifically in the critical area of cyber security. I believe a specific organizational change would make a big difference in closing up this fault line. 

Tasked with responsibility for protecting Defense Department networks, the National Security Agency has unmatched size and depth of expertise in cyber security. The Department of Homeland Security lacks this technical depth, but it has responsibility for the security of the information networks of most of the Federal government and also for the Federal role in securing the information networks of the country’s critical infrastructure. In my view, DHS needs to draw on NSA’s technical skills to protect government and critical infrastructure information systems. 

There is strong opposition in the country at large and in the Congress to giving NSA any role in protecting these civilian information systems. Some worry that by assisting government agencies outside the Defense Department and private companies in protecting their networks, NSA will gain access to government and private systems to gather information on American citizens illegally. These fears, though entirely groundless both technically and procedurally, could be allayed by designating an NSA deputy to report in a dual-hatted relationship to the Secretary of Homeland Security for the purpose of cyber security for government and critical infrastructure networks. New procedural and organizational safeguards could also ensure that NSA cannot gather intelligence through its activities supporting DHS. 


here is also an array of difficult and sensitive issues involving overlapping responsibilities between the Department of Justice (specifically, the operations of the Federal Bureau of Investigation) and those of the IC. To illustrate, let’s begin with a discrete issue and then broaden the discussion to encompass more general considerations.

Our cyber security agenda is a full one. One aspect of this agenda involves computer network attack—one of the more difficult and frustrating subjects of my time as DNI. The specific details of the computer attack activities and capabilities of the IC are classified, but it is both possible and important to understand the principles involved at an unclassified level. The United States needs to decide when and how to employ these means to defend itself, especially in light of the fact that adversaries are already using the internet against us. Jihadist websites spew anti-American venom and incite deadly violence against Americans. Al-Qaeda operatives use the internet to surveil American targets, plan attacks and even conduct them. If the IC can stop or slow this activity, it should do so. 

Nevertheless, some in the government prefer to monitor such activity rather than to stop it, in order to obtain evidence for use in a court of law. Beyond this basic difference of perspective, determining the legally authorized basis, the responsibilities within government, and the process for planning and conducting attacks on hostile computer/internet networks is far from settled. Because the global communications system is so new, so interconnected among foreign and American persons and companies, and so much still in development, we have been unable to settle upon a practical and durable legal basis for computer network attacks. Such operations do not fit into the established categories of either traditional military activities or covert action. Computer network attacks can cause collateral damage, from slowing or denying internet service to permanently damaging computer control systems. The internet is so complex that it is difficult to predict collateral damage with precision. It is not surprising that the U.S. government has moved slowly.

The problem is that the technology and our adversaries are not similarly inhibited, so we risk falling behind both the threats and the opportunities. The Executive Branch and Congress must develop working approaches to computer network attacks, learning as they go. The nation simply cannot wait for complex legal opinions that, in any event, cannot translate into real operational capability. Nor can we decide on roles and missions on the basis of bureaucratic rivalries. We need to experiment now with low-risk but useful operations to build capabilities in this area. The Director of National Intelligence, as well as the Secretary of Defense and the Attorney General, must drive this issue, which ultimately will require White House decisions and Congressional authorizations. 

Beyond the specific issue of computer network security, our border defines the responsibilities of America’s foreign-oriented intelligence agencies, on the one hand, and its domestic law enforcement and security agencies, on the other. Intelligence agencies, with rare and carefully authorized and supervised exceptions, do not operate within this country. They do not spy on Americans. Domestic government agencies operating in the United States (these include some DHS authorities as well as the FBI and other Justice Department agencies) are authorized to investigate American citizens, provided certain well-established procedures for safeguarding their privacy and civil liberties are followed. 

The functional “wall” between our foreign and domestic intelligence agencies is a product of our philosophy of government and our Federal system. It is why the United States has never had the equivalent of Britain’s MI5, which blurs the distinction between domestic and foreign counterterror operations. Historically, the burdens of this arrangement have been light; they are no longer. Al-Qaeda and its affiliates, to take only one prominent example, have members, “wannabes” and sympathizers, including some American citizens, who can travel with relative ease into and out of this country to conduct surveillance, plan attacks and carry them out. The government-wide challenge that 9/11 so tragically illustrated is to detect and track threats to the United States by means of the integrated, geographically defined capabilities and activities of all intelligence and law enforcement organizations. 

A great deal of progress has been made since 9/11 in sharing information. One of the reasons for this progress is that IRTPA gives the DNI the responsibility of defining what constitutes “national intelligence” under the guidance of the Attorney General. Under this authority, the DNI can designate information collected anywhere within the Federal government as national intelligence, which means that it can be shared with all other intelligence agencies and used in analyzing threats to the country. 

Clearly, the most valuable national intelligence is the huge collection of databases of routinely collected government information searchable by computer algorithm. For example, an IC analyst may know that a potential terrorist attacker with the first name of “George” is between 23 and 28 years old, has lived in Atlanta and has recently traveled to Yemen. That analyst would like to be able to very rapidly query the travel records of the Customs and Border Protection Service, the investigative records of the Atlanta Police Department, the visa records of the State Department, and dozens of other government databases to learn George’s full name, current location, and whether he has been involved in any other suspicious activity. Right now our analysts can conduct only some of these searches. Only some, too, are one-click operations; many are still very complicated. The algorithms available to help in that search are still inadequate to the volume of data that must be screened. We share a lot more than we used to, but we are still far from the point at which an intelligence analyst can quickly and easily find all relevant information with a single query. 

Obviously, expanding the scope of national intelligence raises important privacy and civil liberty concerns. Government officials at all levels collect enormous quantities of information about Americans and visitors to this country in the course of doing their jobs. Should this information be readily available to intelligence analysts so that they can ferret out the few Americans who threaten the rest? The answer to that difficult question defines what the law means by “national intelligence.” I did not have enough time as DNI to widen that definition to increase the timely flow of information among foreign intelligence and domestic agencies, or to improve the technical means to do so. We made important progress after the December 2009 attempted bombing of an aircraft over Detroit, but there is much more to be done. As with so many additional steps required in intelligence reform, a strong and aggressive DNI, with support from the White House and Congress, is a prerequisite for progress. 

Other Unfinished Business: DNI Authorities 


side from the structural issues that concern the DNI’s relationship to the Department of Defense, the Department of Homeland Security and the Department of Justice, a few other aspects of unfinished reform also require attention. Most of the authorities to continue integration already exist either in the IRTPA or in the implementing Executive Order 12333, but several additional authorities for the DNI in the areas of personnel and budget and the DNI’s staff would speed integration. These matter are critical if we are to continue to build a more integrated but still federated system led by a strong Director of National Intelligence who is supported by an experienced and innovative staff.

As DNI, I could have further accelerated this process had I enjoyed the support of the President and his staff. However, their past experience, priorities and the White House-centric style of this Administration’s national security governance never offered me the opportunity. With a supportive White House and Congress, however, there are real possibilities for progress. Personnel issues and budget authority are two cases in point.

Senior personnel appointments in the intelligence agencies are one of the keys to successful integration. The joint duty provision of the IRTPA mandates that an intelligence officer complete an assignment outside his or her parent agency in order to be promoted to a senior rank. This provision will raise the overall level of joint commitment by senior leadership, but there needs to be individual consideration of the very top leaders. 

At present, the IRTPA requires the DNI’s concurrence for appointing top leaders of all 16 intelligence agencies or entities. I recommend that DNI concurrence also be required for the second- and third-level positions. These officials are key to integrated intelligence once the top leadership has set the basic direction. The DNI needs to ensure that community-minded officers lead the individual intelligence agencies. This provision could be established either by legislation or by executive order. 

The Director of the CIA is a special case in this regard. The IRTPA grants the DNI the authority to recommend the CIA Director to the President, but does not insist that the DNI concur in the selection. In practice this has happened only once, when DNI John Negroponte recommended General Michael Hayden to be CIA Director. In 2008, the President-elect consulted me in the process of naming a CIA Director, but the input of his campaign staff turned out to be more important than mine. 

In my view, the CIA Director (or the Directors of the two organizations that I recommend to take the place of the CIA as presently constituted) should be a career officer and serve for a fixed term. The CIA is a complicated, sensitive organization. An officer who knows it from years of service would have credibility and an instant following in the organization, yet also have ideas to improve it and the ability to work with the rest of the IC. Historically, experienced CIA officers have been among the best directors: Robert Gates, William Colby, Richard Helms. A professional CIA Director concentrating on running and improving the Agency would more likely work more smoothly with a DNI who brought outside skills and perspectives. There are several senior CIA officers in the Agency right now with the potential to be outstanding directors. 

I recommend legislation mandating that the CIA Director be a career CIA officer recommended by the DNI and appointed by the President for a fixed term. The same legislation could stipulate that the CIA be split into an all-source analytic agency and a clandestine and covert service, both headed by career officials appointed for a fixed term and under the DNI’s direction, authority and control. 

Additionally, as things stand, a single sentence in the IRTPA states that the Director of the CIA “shall report to the Director of National Intelligence regarding the activities of the Central Intelligence Agency.” This sentence needs to be strengthened. It should specify that the Director of the CIA must submit plans and reports on all CIA activities to the DNI for approval before going to the White House or to Congress, and that the Director obtain the approval of the DNI for significant communications with the intelligence services of other countries. In my experience, these were the areas in which the CIA exercised a degree of autonomy that both undercut DNI authority and made policy without the formal consideration of the DNI or the National Security Council. 

The authors of the IRTPA thought they were providing the DNI with one of his most powerful tools in the form of budget authority. As with much of IRTPA, this tool has not been used to its fullest and is in any case limited. Unfortunately, that most dysfunctional old problem, in which the DCI was responsible for the entire IC but lacked the budget authority to go along with it, has been replicated to some extent in the new arrangement.

One reason the DNI’s budget authority has not been exercised fully has been that the growth of overall intelligence funding during the past 15 years has made financial discipline unnecessary. But as budgets become more constrained, tradeoffs will become essential. Another reason has been the lack of a strong analytical capability to support budget decisions, but that is no longer a significant impediment. 

As things stand now, the DNI and his Chief Financial Officer have less authority than the Defense Department’s Secretary and Comptroller. If we are to have an integrated IC, we must provide the DNI with a comparable level of authority. I therefore recommend giving the DNI full comptrollership authority. This would mean that agency heads could not bypass the DNI’s office in dealing with either the Office of Management and Budget (OMB) or with congressional authorization and appropriations committees. Of course, information would flow between and among agencies, OMB and congressional staff, but this change would require the OMB and the committee staffs always to deal through the DNI. 

This reason this is so important is that without full comptrollership authorities, the Office of the DNI cannot hope to use budget resources effectively across the IC. There are many cases, for example, in which money should be moved from one agency to another in order to solve problems. The DNI has rarely been able to accomplish this function, especially in the year of execution, and then to monitor and enforce a successful solution of the problem. Instead, agencies have gotten into the habit of requesting additional budget allocations when receiving new assignments. This habit won’t avail in more austere circumstances, so the IC comptroller must have the authority to move money based on the best analysis available and DNI decisions. Congress must grant, and OMB directives must confirm, the DNI’s authority to transfer National Intelligence Program funding across agencies throughout the budget cycle. 

Finally in this regard, I recommend transferring the funding under the military intelligence program (MIP) in the Department of Defense budget to the national intelligence program (NIP) in the DNI budget. Before I left, James Clapper, then the Undersecretary of Defense for Intelligence (USDI), and I made great strides in coordinating MIP/NIP actions. For example, we split-funded several important programs, issued a single planning guidance document, and included both Defense and the IC in major studies. However, the MIP is subject to Defense Department decisions that have nothing to do with intelligence; the OSD Comptroller, for example, can subject it to overall Defense Department taxes, or he can overrule agreements that have been worked out between the DNI and the USDI. Defense Department-DNI cooperation on many classified programs must continue, of course, but the funding for intelligence should all come through the DNI and the national intelligence program. 


 am often asked if the DNI is a failed concept, if it is just too difficult a job. My answer is a resounding “no.” President Lincoln once remarked that it’s not a good idea to change horses while trying to ford a stream. He said that in the context of his (victorious) 1864 re-election campaign, but the same counsel applies to intelligence community reform. We are now on the right horse, and on the right course. We do not need, as some have recommended, a cabinet-level Department of Intelligence. We certainly must not go back to the fiefdoms that constituted the IC before 9/11. We need to finish what we have begun, and doing so is well within reach. If we fail at this, however, we will increase the potential for nasty surprises like 9/11 in our future. We must not fail. 

Dennis C. Blair has joined several non-profit groups in the areas of the promotion of democracy, American energy security and electoral reform since resigning as DNI in 2010. This essay is adapted from the author’s prepared statement before the U.S. Senate Committee on Homeland Security and Governmental Affairs on May 19, 2011.