The State of the Craft: Is Intelligence Reform Working?

W hen the Senate confirmed James Clapper Jr. as the nation’s new director of national intelligence in early August, he became the beneficiary of landmark legislation passed in late 2004 and designed to strengthen control and direction of the sprawling American intelligence community (IC). But even if he is successful in his work, Clapper will almost certainly be able to exercise less control over the IC than George Tenet did before he stepped down as director of central intelligence while the legislation was being drafted. The reasons for this lie in a complicated story, one whose outcome will impact the future role American intelligence plays in our complex and dangerous world.

In the summer of 2004, the American intelligence community was waging a relentless and successful global war against al-Qaeda. Its analysis that post-invasion Iraq would be inherently unstable was proving spot-on. One of its officers, Stephen Kappes, had just run back-channel communications that convinced the Libyans to abandon their nuclear and chemical weapons programs. Little matter. The 9/11 Commission was about to publish its final report urging the most fundamental realignment of the intelligence community since it was created by the National Security Act of 1947, and, as the nation was embroiled in a heated presidential election season, there was little doubt that something was going to happen. Senator John Kerry endorsed the findings of the commission within hours of their release. President Bush followed a few days later.

Most in the intelligence community at the time (including me) thought this was a bad idea. Our operational tempo was extremely high and we all knew that any major restructuring would be a drain on time and energy. But we also knew that we had not prevented the horrific attacks of 9/11 and that we’d had a clean swing and a miss in 2002 with the national intelligence estimate on Iraq’s weapons of mass destruction.

Defensive comments about correctable errors in analytic tradecraft or arguments based on the inherent difficulty of intelligence work (“all other major intelligence agencies also believed that Iraq had a WMD program”) were not going to carry much weight. The American people had forgiven us for getting some things wrong, but they wanted to see some visible changes.

It’s hard for Congress to legislate better analysis or more aggressive intelligence collection or perfect covert operations. Its choices are more limited. It can move money (it had already given the IC a lot); it can move people (and recruiting was at record rates); or it can restructure organizational charts and strengthen authorities.

In the Intelligence Reform and Terrorism Prevention Act of 2004, Congress chose the third option. Once you cut through the empty and emotionally charged criticisms of “Cold War mentalities,” “stovepipes,” and “bureaucratic turf,” it was pretty clear that the Hill was attempting to recalibrate for the intelligence community the critical balance that any complex organization needs—the balance between freedom of action for the parts and unity of effort for the whole. Too little autonomy for the parts leads to inaction, inflexibility, hesitation, and lost opportunities. Too little unity of effort means that individual excellence is not synchronized, harmonized, exploited, or leveraged.

In short, we were going to strengthen the center of the community. We were also going to relocate and rename the center. The director of central intelligence (DCI) would become the director of national intelligence (DNI), emphatically not the head of the CIA. He would not even be allowed to have his offices at Langley, as the law would forbid the DNI to be housed at any existing intelligence agency.


I n 2004, George Tenet was actually a powerful figure in the U.S. government and its intelligence community. As head of the National Security Agency, I was directed to act by him more times than by any combination of people in the Department of Defense, or anywhere else in government for that matter. Tenet had the ear of the president, with whom he met six days a week. He had a booming personality and a work ethic to match. He also headed up the CIA, and even in 2004 that C still stood for Central .

When Tenet called me, he would usually begin the conversation with “Mike, my guys were just in here...,” and usually end it with “and here’s what I want you to do.” In most of these conversations, Tenet’s “guys” were not personnel from his relatively small Community Management Staff, but his operational, analytical, and technical folks from CIA proper. In other words, in terms of creating unity of effort and operational cohesion in the intelligence community, the strongest “glue” that we had was the fact that the head of the community, the DCI, also headed up the most operationally relevant agency within it, the CIA.

So the diagnosis that the DCI was not especially strong was wide of the mark. But the corollary conclusion—that the source of his strength (that he was the head of the CIA) brought with it inherent limitations—was closer to the target. When I was the director of the CIA (the first occupant of that suite who was not also DCI), I would tell the CIA workforce that one of the advantages of the new DNI structure was that I could concentrate on being director. Indeed, hardly a day passed in that job when I did not wonder aloud how any of my predecessors could ever have given both tasks their due.

There is also an argument that any director of the CIA must view the world through a CIA lens. Can the head of our human intelligence (HUMINT) service be counted on to make wise resource trade-offs between HUMINT, which he directly controls, and other areas like signals intelligence (SIGINT), which he does not? My experience says that he can, but the question is neither unfair nor unexpected.

Many of us felt that if we were going to take direct control of the CIA away from the new head of the community, we really had to make sure that the legislation dealt the new office a very powerful hand and that it did it formally and specifically. That’s what I told Senators Susan Collins and Joseph Lieberman when I met with them in the summer of 2004. That’s why Jim Clapper and I warned the House Permanent Select Committee on Intelligence (HPSCI) in late summer of 2004 that a “feckless” DNI would actually make matters worse.

Remarks like that last one got us invited to a lunch with Secretary of Defense Donald Rumsfeld that included his deputy, Paul Wolfowitz, and his under secretary for intelligence, Stephen Cambone. Clapper and I were lined up across the table from the Pentagon leadership and were invited to make our case. In essence we argued that there was a real danger of Congress creating a leader of the IC who truly had less power than DCIs had historically been able to wield. We both agreed that this could be disastrous and argued for legislative language that would codify a robust role for the DNI—even over those big national collection agencies inside the Defense Department, namely, the National Security Agency (NSA), the National Geospatial-Intelligence Agency (NGA), and the National Reconnaissance Office (NRO). We even allowed ourselves to imagine a future where those agencies could be outside the Pentagon and directly under the DNI.

I’m sure that Rumsfeld didn’t think these were idle or theoretical comments as he was talking to the sitting heads of two of these agencies (I was head of the NSA, Clapper was director of the NGA). As lunch ended and the secretary and his team were leaving, I plaintively commented that we could be headed for a disaster unless the Pentagon could find it in its heart to be “generous.”

T he Intelligence Reform and Terrorism Prevention Act has an odd lineage. It was developed in the Senate by two members (Collins and Lieberman) who were not members of the intelligence committee. The legislation moved fairly quickly and was not the product of a long series of hearings and deliberations. And the intelligence committees of the House and Senate did not have a central role in the legislation.

One committee that did play an active role was the House Armed Services Committee. At the eleventh hour in the legislative process, Chairman Duncan Hunter inserted language into the bill stipulating that—notwithstanding any robust authorities that Collins and Lieberman might be able to insert into the legislation elsewhere (such as budget control)—the DNI could not abrogate the authorities of the cabinet officers of departments in which elements of the intelligence community were located. Section 1018 of the IRTPA, as it was called, was a determined push to protect the secretary of defense’s prerogatives when it came to his critical combat support agencies: the NSA, NGA, and NRO.

Lieberman and Collins, and prominent House member Representative Jane Harman, pushed back, and in classic Washington fashion the section was modified to leave it to the president to provide guidance on what the DNI could and could not do with respect to cabinet officers. Section 1018 created an opening to at least limit the provision’s impact since it specifically called on the president to issue implementing guidance to resolve potential ambiguities.

That had never been done, but in the summer of 2007 the National Security Council (NSC)—at the strong urging of Mike McConnell, then director of national intelligence—decided to rewrite Executive Order 12333, the basic document through which the president organizes and controls the intelligence community. The order had been written in 1981 under President Reagan, had rarely been modified since, and was badly in need of an update. This seemed an ideal opportunity to remedy what many of us believed was a weakness in the 2004 law.

The first draft of the new executive order (EO) put on the table at the NSC in the spring of 2008 boldly proclaimed that the DNI, in carrying out his responsibilities under the EO or under the law, should be presumed not to be abrogating the authorities of the various department heads. That would have tilted tough issues in the direction of the DNI and given him more room to act boldly (and without consensus).

As CIA director, I argued strongly for the provision. I wasn’t totally altruistic (I was the only agency head not “protected” by a cabinet official’s prerogatives), but I did point out that, absent that kind of presumption, the DNI and especially his staff would focus more and more on the CIA for the worst of all reasons—because they could.

I argued strongly and (except for Mike McConnell) mostly alone. After all, the room was largely filled with current and former cabinet officials. The final version of the new EO 12333 pointedly repeated that, in carrying out his responsibilities under the EO or under the law, the DNI shall not abrogate the authorities of the various department heads.

The Armed Services Committees had also acted boldly in 2003 to strengthen Defense Department equities when they created the position of under secretary of defense for intelligence, (USDI). This new post was effectively a senior Pentagon official between the nation’s intelligence chief (then the DCI, now the DNI) and several of his biggest collection agencies. When Rumsfeld later delegated his “authority, direction and control” of the major Defense Department intelligence agencies to the under secretary for intelligence, he effected a major reorganization and power shift in the IC. (In one admittedly extreme example, a memo emanating from the USDI’s office in 2010 seemed to suggest that the DNI could not even catalog the language proficiency or needs of intelligence community elements in the Department of Defense.)

It’s clear why Rumsfeld wanted a strong under secretary for intelligence. He knew the importance of intelligence. He was dissatisfied with what he was getting (although that may have been related more to the difficulty of the task than to any organizational defect). He wanted more personal influence over these key organizations. Indeed, I had been at the NSA job more than two years before I had spoken with any secretary of defense—not an ideal state of affairs. Besides, our country was at war and it should come as no surprise that the Pentagon character of the NSA, NGA, and NRO became more pronounced as the years at war rolled on, even though a significant portion of their mission remained national and their first initial remained N .

Moreover, the NSA—the workhorse of American intelligence collection, which I frequently claim is responsible for at least half of all American intelligence—had more and more harnessed its considerable power to the needs of the battlefield. I did this as director of the NSA and my successor, General Keith B. Alexander, has clearly followed suit, emphasizing communications externals (e.g., the behavior of a phone: is it always calling “dirty” numbers?) and geolocation (where is the phone right now? ). Such an approach is especially useful when you want to kill the communicant, less so when you want to extract intelligence from the content of a particular communication. And now, with the creation of Cyber Command and a dual role for the NSA director as its chief, even more of the energy at Fort Meade will be directed to the Title 10 war-fighting mission.

Much of this has great legitimacy, but all of this cuts across the grain of legislation designed to strengthen the “center,” to give the leader of the national intelligence community more strength, and to give the DNI more say over the current operations and future direction of the entire intelligence community. Beyond soft spots in the legislation and the crosscurrents of Washington, there were circumstances of history, timing, and performance that made this harder than it should have been.


T he new law gives the director of national intelligence two tasks: acting as senior intelligence adviser to the president and enhancing the intelligence community’s coordination and integration. Each task is monumental in its own right; together they are more than any one director could manage effectively. Hence the law gives the DNI a Senate-confirmed principal deputy. I was the first principal deputy supporting John Negroponte and was in the post for about a year. The position was vacant for well more than a year after I left and, indeed, has been vacant for nearly half the life of the DNI. This isn’t a trivial oversight. The DNI can easily be consumed by the morning briefing and other policy meetings downtown, which essentially made the office of the DNI a new post without a second-in-command during critical periods of its formation.

The law also recognizes that the most important relationship within the intelligence community is the one between the director of national intelligence and the director of the CIA. It’s so important that the law says that the former shall recommend to the president the candidate for the latter. But in the five-year history of the DNI, a CIA director has been picked by a sitting DNI only once (when Negroponte recommended me) and the two of us overlapped in these positions for only about six months. The pattern continued when DNI-designate Dennis C. Blair, by all indications, played no role in selecting Leon Panetta as the new administration’s head of the CIA.

This was always going to be hard—and we clearly made it harder than it should have been—but it was not impossible. At a colloquium held on the fifth anniversary of the DNI, I opined that this structure was as good as most and better than some and that, in any event, it was not so flawed that we needed yet another restructuring. Three remarkably talented men—Negroponte, McConnell, and Blair—obviously thought the task feasible and worth doing.

Indeed, shortly after Blair had left the DNI position, Negroponte and former Director of Central Intelligence James Woolsey said as much during an NPR interview. Reacting to Woolsey’s positive commentary on both the DCI and DNI models, Negroponte added, “I agree with Jim that either system could be made to work, and I think that we can’t turn the clock back now.” Negroponte could have recalled for his audience some real achievements of the new model in increased information sharing, coordination, setting collection priorities, and folding the FBI more fully into the intelligence effort. This latter task was especially important since it would have been far more difficult to achieve in our political culture with a DCI who was also the head of the nation’s foreign espionage service.

But with weaknesses in the law, competing concepts of what it meant, and the sheer magnitude of the American intelligence task (ours are the only services that are truly global or need to be), success depended on a series of intangibles. Foremost among these are the DNI’s political deftness, his closeness to the president, and his relationship with the director of the CIA.


B lair came to the job with energy, commitment, and some clear ideas of what he wanted to do. He picked a deputy with deep policy experience to cover many of the high-profile policy meetings downtown in order to carve out more of his own time for the tedious, lower-profile, and often thankless task of running the intelligence community. But even his best-motivated steps often left him digging out of political holes. His selection of Charles W. Freeman Jr. to be the head of the National Intelligence Council, designed to stir up the analytic community, succeeded only in stirring up the Israeli lobby, and ultimately the White House. Blair created further trouble for himself when he authored a quite accurate memo to the entire intelligence community stating that, whatever his personal discomfort with enhanced interrogation techniques might be, those techniques had led to the collection of valuable intelligence; and again when he testified on the Hill that the Detroit underwear bomber should have been questioned by the “High Value Detainee Interrogation Group.” The latter upset Congress and forced the president’s press team to admit that the administration had not actually gotten around to forming the group yet.

Blair’s lack of closeness to the president and his White House team seems to have been caused by more than just his being periodically off message. Without his office being notified, the DNI was quietly moved down the protocol list early in the administration. The DNI team has also frequently complained about a lack of overall guidance from the White House. One senior staffer described the administration’s management technique as not unlike one of those electronic pet fences: you don’t know you’ve gone beyond your limits until you feel the electric shock.

Obama and Blair did not know one another before assuming their respective offices and did not meet frequently during the transition. Beyond this, it was clear to most in the intelligence community that, as much as this or any administration values good intelligence, this team was not all that anxious to look “under the hood.” The general attitude seemed to be one of treating intelligence like a public utility. Most of us, when we enter a room and throw on a light switch, expect illumination—not a grand debate over the virtues of 110 versus 220, the physics of power generation, or even the relative merits of building codes. Just light, please. With a variety of crises imposing themselves on the administration, the inner workings of intelligence—whenever they became an issue—were clearly viewed as a distraction from important work.

In the news more than he wanted to be, and not having a close relationship with the president or other seniors in the West Wing, Blair now had to establish a working relationship with his most important agency head: the director of the CIA (DCIA). Even in the best of times, the DNI-DCIA relationship is a challenging one. The law puts the former at the center of the American intelligence community. History and tradition and even many current operations suggest that the DCIA, however, has pride of place, and the agency’s collective culture is very reluctant to admit otherwise. While director of the agency, I would frequently say that since the DCIA and the DNI were on or near the same grid reference, the only way to prevent fratricide was to separate us at different altitudes. The DNI should work at the higher altitudes—set policy, give overall direction, manage conflict—whereas the DCIA should work the lower—coordinate, conduct, operate. The DCIA must guarantee the DNI transparency; the DNI has to give the DCIA space to operate. Sounds simple. But Mike McConnell and I—who had known each other for years, were friends, and between us had nearly three quarters of a century of intelligence experience—still found that this relationship took a lot of effort.

In his confirmation hearing before the Senate Intelligence Committee, Panetta was asked how he viewed his relationship with the DNI. Tough question, actually, since the Intelligence Reform Act says that the DCIA “reports” to the DNI but carefully avoids mentioning “authority,” “direction,” and “control.” Concluding his responses to a series of probes, Panetta summarized, much to the satisfaction of the committee, “The DNI is my boss.”

It must not have seemed that way to Blair only months later, however, when Panetta, then head of the agency, told his station chiefs to ignore a directive they had received days earlier from the DNI on the obscure subject of DNI representatives in foreign capitals. CIA folks had been quick to call foul since Blair pulled the trigger on his memo without a promised final consultation with Panetta (who was away traveling).

The issue had been lingering for years. When we formed the DNI in the spring of 2005, we felt the need to designate his representatives overseas and easily defaulted to the senior CIA officer in each country. Other agency heads wanted at least the possibility for their seniors to serve in this role (which was a leadership and coordinating function, but far from a directive one) and lobbied for a formal memo to give the DNI that option. This is the kind of thing that never needs to be written down. Who could question that the DNI could name his own representatives? But the instances where it would be a good idea to name someone other than the station chief were so few (if they existed at all), they could be handled as one-offs—with quiet phone calls to affected ambassadors and agency heads. Trying to formally codify it would just rouse a lot of ghosts.

It became a manhood issue all around. As director of the agency, I fought it as energetically as my successor apparently did. I argued that the law actually gave the CIA the responsibility to coordinate foreign intelligence relationships and naming someone other than the station chief would unnecessarily confuse our partners. I remember that we used the term “nuclear red line” to underscore the depth of our feelings on the question. But when my staff would urge me to take this issue out of the intelligence family to someone in the West Wing, I was quick to point out that if the issue got to that level, we would lose. There was no way, I would point out, that the White House could not back the DNI on such a question. Whatever the merits of the underlying arguments, they would pale in comparison to the overarching issue of DNI authority. When the issue did get to the White House, however, I was proven wrong: the Obama administration sided with the CIA director. The DNI lost, and lost publicly.

If White House backing for Blair seemed weak on this issue, it all but disappeared in the aftermath of the attempted bombing of a Northwest Airlines flight on Christmas Day last year. There was the usual finger-pointing, and some genuine shortcomings were uncovered, but the striking issue was not the fault-finding, but rather who was doing it. It was not being done by the putative head of the American intelligence community, but by John O. Brennan, a retired CIA officer, now the president’s homeland security adviser. The DNI and affected agency heads were informed of Brennan’s findings only shortly before the president went public with them and, according to press accounts, the announcement was delayed as they furiously (in all senses of that word, apparently) pushed back against some of its conclusions. Brennan became the intelligence face of the administration in the days and weeks that followed, and even penned an op-ed for USA Today that aggressively responded to those who would criticize the administration’s actions.

Although he had commissioned his own, more thorough review of the intelligence leading up to the Detroit incident, until he was summoned to appear before congressional committees, the DNI was nowhere to be seen. It seemed only a matter of time before Blair would be asked to step down. It was an open secret that the President’s Intelligence Advisory Board had been tasked to look at the overall DNI situation. When I was interviewed by them I recommended that they have the White House photographer take a picture and widely circulate it: Blair and the president, the only two visible in the frame, huddled closely, poring over a map or a document. The president and his intelligence adviser sharing secrets and sharing views. It was only that image—backed, of course, by a corresponding reality—that could give Blair what he needed to truly function as DNI.

It was not to be.


J im Clapper is a wonderful officer. He is my friend and has been my mentor and my partner. Arguments against his confirmation, based on his long years of military service, are specious and without merit. But even as the administration changes DNIs, it also appears to be changing the definition of the job. In his remarks nominating Clapper for the position, Obama noted that Clapper, as USDI, had “successfully overseen the military and civilian intelligence personnel and budgets that make up the bulk of our sixteen-agency intelligence community.” The apparent (if implied) conclusion was that the president would not be expecting Clapper to be doing those things in his next job.

The next day’s Washington Post featured a story clearly sourced to White House officials describing the president’s “invaluable go-to person” on many intelligence questions. Except they weren’t writing about the new DNI nominee; they were writing about John Brennan. Although officials denied that Brennan was a de facto DNI, there is no denying that Brennan has used his knowledge of the intelligence community to direct or question specific offices without always informing the involved agency head, let alone the DNI.

Finally, on the same Sunday as the Post article, Secretary of Defense Robert Gates, while traveling in the Caucasus, gave reporters a thoughtful and seemingly well-prepared analysis of what the DNI job really was. He punctuated his remarks with multiple references to “temperament,” “positive chemistry,” “get[ting] along,” and bringing people along by “accommodating their interests” and getting them to “voluntarily work together.” He suggested that the position of DNI is closer to that of “a powerful congressional committee chair than it is to a CEO.”

As a career CIA officer and former DCI, Gates knows his subject. He had been offered the DNI job by Bush but turned it down because he feared the position lacked adequate power. He had written an op-ed during the debate over the 2004 law expressing support for an empowered intelligence chief. His call now for Clapper to lead what amounts to a coalition of the willing may be the best that can be hoped for under the current circumstances.

This is a far cry, however, from what most expected the new law to create—a strengthened center—although it does resemble what Clapper and I feared would happen: a leader of the intelligence community, cut loose from the resources of the CIA, in many ways actually weaker than the DCI. If Clapper cannot make it work, there are no obvious remedies in the current structure. Panetta has neither the staff, the authority, nor the interest to lead the community. And it would be inappropriate for Brennan to even try to do so as he is serving in an unconfirmed position, beyond congressional oversight, and comingling in the West Wing with colleagues responsible for policy and political decisions.

The nation is asking a lot of Jim Clapper, probably more than it has a right to ask. He will do well. And good people often overcome weak structures. But consistently relying on extraordinary heroism for routine success is hardly wise policy. And it is especially unwise in an area as critical as intelligence. The DNI and the people he will lead deserve better. We all need to keep that in mind as we rush to dissect the community after the next perceived “failure.”

Michael V. Hayden was director of the CIA from 2006 to 2009.

OG Image: