THE POST of director of national intelligence (DNI) has had an unhappy five-year history. Until now it has been easier to blame the successive occupants of the position than to acknowledge the fundamental flaws of the office. Much commentary about the recent ouster of Dennis Blair, for example, has focused on his lack of chemistry with the president, his riling of the Israel lobby by attempting to appoint Chas Freeman to head the National Intelligence Council and a deficiency in political street-fighting skills compared to those of CIA Director Leon Panetta. All those no doubt contributed to Blair’s troubles. But one indication that the principal problems are those of the office rather than the occupant is that the job, in such a short time, has now chewed up three able public servants, each of whom excelled in their principal professions (the diplomatic service in the case of the first DNI, John Negroponte, and the military in the cases of Blair and his predecessor, Mike McConnell). More telling still is the difficulty in persuading other able people to take the job. Reportedly, the first to refuse was the future secretary of defense, Robert Gates, one of the most adept officials in Washington at protecting his own reputation; he can certainly recognize a losing hand when he sees one.

None of this should be surprising, considering the position’s provenance. For the office owes its existence less to any need for intelligence reform than to the anguish that followed the 9/11 terrorist attacks. With this, Americans yearned for something—anything—that could reassure them that an underlying problem had been identified and fixed and that no comparable disaster would ever happen again. They had no patience for details of how existing governmental machinery did or did not work, or how a proposal for changing it would or would not make things better. The mood of the moment left room for very few facts let alone any analysis. What mattered instead were raw emotion, the people’s overpowering appetite for whatever could be sold as a fix that would keep them safe and the need for national catharsis.

The 9/11 Commission became the main instrument for that release, and it performed the role masterfully. Rather than resisting the onslaught of emotion, it harnessed the people’s ire by beginning its series of televised hearings with the appearances of 9/11 family members. The commission’s leadership described its main priorities not as the accuracy of its report or the long-term effectiveness of its recommendations but instead as maintaining good relations with the victims’ relatives and squelching any partisan commission splits. Anger over 9/11 gave the commission prominence few other ad hoc inquiries have ever achieved, and it was that same anger that propelled its ill-conceived intelligence-reorganization proposal into law.

The commission’s rollout strategy, assisted by a public-relations firm, was clever indeed. It unveiled its proposal in the midst of the 2004 presidential-election campaign, in the expectation that the political parties would try to outbid one another in supporting the recommendations. What politician would not want to be identified with the idea of making Americans safe from future terrorist attacks? The plan worked brilliantly. Democratic candidate John Kerry, after barely enough time to have turned the front cover of the commission’s report, called for all of its recommendations to be adopted. His Republican opponent, then-President George W. Bush, had no choice but to follow suit.

In short, the 9/11 Commission presented its findings amid the worst possible environment for carefully examining them. It was an environment of passion and politics, not careful judgment. The almost total absence, with rare exceptions (notably the trenchant analysis of Judge Richard Posner), of any questioning of the commission’s output is, with the perspective of several years, stunning. The public and the press, not just presidential candidates, swallowed that output whole, with nary a hint of skepticism. The writer John Updike reflected the national response to the 9/11 Commission Report when he declared the document and the King James Bible to be the only masterpieces ever written by committee. This unthinking, uncritical acceptance meant that the report’s major faults went unnoticed. These included large holes and inaccuracies in the commission’s narrative, a disconnect between the narrative and the recommendations, and numerous questions about how the proposals would work in practice.

The idea of creating a director of national intelligence was the centerpiece of the commission’s faulty reorganization scheme. It held numerous attractions. The idea was, first of all, available; it had been floating around for years. Arguments for and against it were well rehearsed—and in the unquestioning post-9/11 environment, no one would inconveniently point out that since the proposal predated the attack, the idea obviously was not a response to what the investigation uncovered about the terrorist plot. Instead, establishment of a new office would be a highly visible change to which people could hitch their hopes for future safety. And it was appealing for some of the same general reasons that reorganizing bureaucracy has been a perennial Washington favorite: it was a concrete legislative measure that both the commission and Congress could point to as their response to failure. Establishment of a new office separate from existing agencies such as the CIA and the FBI also was consistent with the dominant sentiment of blaming those agencies for 9/11.

The result is that this job, which the 9/11 Commission sold as a fix to assuage national anguish about terrorism, is, to put it succinctly, a lemon.

ABOVE ALL, the position of DNI embodies the theme that greater centralization of the intelligence community is a good thing. Yet such a conclusion is in itself questionable. For all that is said about how different agencies need to communicate more and work together better, just as much is said—often by the same people—about the need to avoid groupthink and to encourage alternative analytical viewpoints. The 9/11 Commission’s plan did not reflect any careful analysis of this general question. There was an appetite for visible organizational change of some sort; the existing intelligence community was perceived as disjointed; the available idea for reform (the DNI proposal) was seen as a streamlining mechanism; and so centralization it was.

In a less emotional and more deliberative environment, the notion that rearranging lines and boxes on the intelligence community’s organization chart would help to crack the inherently tough nut of uncovering the next big terrorist plot would have been unconvincing, and appropriately so. Of course interagency cooperation is important. But counterterrorist officers do not, as their first act upon arriving for work in the morning, examine the organization chart on their cubicle walls to decide who will enjoy their assistance and who will not. Plenty of interagency cooperation takes place every day, but the organizational pathways have almost nothing to do with it. It is far more a matter of informal communication among people who share the fundamental mission of curbing terrorism and saving lives.

The best known instances of insufficient interagency communication in the 9/11 case—involving tardiness in placing names on watch lists—did not point to any need for additional procedures or machinery, let alone any that a DNI would create or enforce. Instead, they resulted from inadequately diligent use of procedures already in place, or from officers so swamped by the inevitably large amount of noise in the system that even the slightest inattention was enough to miss a signal whose significance would become clear only in hindsight.

The point was underscored by the attempted bombing of a Detroit-bound airliner in December 2009. The incident generated a reprise of all the post-9/11 recriminations about dots not being connected and different strands of information not being woven together. Except that this time, the principal targets of the recriminations were the DNI and the 9/11 Commission’s other major creation, the National Counterterrorism Center. This near miss, which became one of the most discussed terrorist incidents in the United States since 9/11, finally led to some serious questions about whether the commission’s reorganization scheme had been an improvement after all. But even then, few drew the more general lesson: that no amount of reform or rearrangement of wiring diagrams can crack the tough nuts, eliminate the signal-to-noise problems, and avoid oversights and errors that are within the normal range of human fallibility.

IN THE end, creating the DNI position has only made matters worse, as it has simply added one more level of bureaucracy: the Office of the Director of National Intelligence (ODNI). Physically, the ODNI resides comfortably in its own headquarters in suburban Virginia. Bureaucratically, it sits uneasily atop all of the other, previously existing components of the community.

First off, let us not miss the fact that prior to these reforms there already was an intelligence head: the now-divided position of Director of Central Intelligence (DCI). For anyone making a thoughtful case for consolidating power in the hands of a single overseer, the cleanest, most obvious, least disruptive way of empowering that person would have been to grant additional authority to the DCI. Of course this could never happen because it would not have satisfied political demands in the wake of 9/11. Because the DCI also had direct control of the CIA and was known generally as the “CIA director,” any such additional authority would have been seen as a reward for the CIA, when what was being demanded was to punish that agency. So a host of arguments were used to justify bifurcation of the DCI’s job.

A common view was that for the same official to lead both the CIA and the intelligence community was somehow anomalous, unbalanced or unfair. This view overlooked why the Central Intelligence Agency exists. The explanation is in the word “central.” The CIA was—until the establishment of the ODNI—the only intelligence organization not part of some other department, beholden only to national, not departmental, requirements and interests. If there was any one place where everything having to do with intelligence was supposed to come together, the CIA was it. The community-leadership functions of the DCI reflected this. Dissolving that position represents a deconsolidation, not consolidation, of the intelligence community’s work.

Proponents, too, incorrectly applied a span-of-control argument: that leading both the CIA and the intelligence community was too big a job for one person to handle. But the DCI, like any other senior leader in either the government or the private sector, did not personally do everything; he delegated. To help run the CIA he had an executive director of the agency, as well as a deputy director of central intelligence who focused much of his time on agency business. A second deputy director of central intelligence was solely concerned with community-wide matters, assisted by an intelligence-community staff. If such delegation does not negate the “too big” argument, then what does it imply about the DNI? If he really is in charge of the whole intelligence community, then his job is every bit as big as the DCI’s ever was. If he is not in charge of it, then what did the reorganization accomplish?

THE REORGANIZATION reflected a bigger misunderstanding still: why an intelligence community exists at all. Much pro-centralization sentiment seems to view the community, like Italy before the Risorgimento or Germany before Bismarck, as an entity that an unfortunate historical legacy has left in pieces but would be better and stronger if cemented together. Many of the community’s components exist, however, because their primary mission is to serve the intelligence needs of particular government departments, including functioning as the intelligence staff for a given secretary, responding to requests from individual bureaus and highlighting intelligence of particular relevance to a department’s responsibilities. These components include the Defense Intelligence Agency (DIA) in the Department of Defense, the Bureau of Intelligence and Research in the Department of State, and intelligence components in the Departments of Energy, Treasury and Homeland Security. Certainly, it makes sense to tap their expertise and to treat them as a community for some purposes (such as participating in the preparation and review of certain assessments and estimates), but their raison d’être and thus chief responsibility always will be to serve their own departments. No matter how ruthless a reform the rest of the intelligence community may undergo, the heads of those departments will understandably insist on retaining primary control over their intelligence arms.

Other intelligence agencies are properly under the control of a cabinet department for different, but still valid, reasons. Some intelligence-community components within the Department of Defense have broader missions, yet it is not a historical accident that they reside within that department because, for example, the imagery-processing capabilities of the National Geospatial-Intelligence Agency (NGA) do substantial double duty in fulfilling the military’s mapping requirements. In the same vein, the director of the National Security Agency (NSA) wears (since May 2010) a second hat as head of the military’s new Cyber Command. Given the extensive technological overlap between cyberwarfare and the NSA’s signals-intercept business, the dual-hatting makes sense.

The nature of the intelligence community and the missions of its components mean that an executive head who genuinely directs the whole shebang will always be a chimera. If we expect the DNI to play such a role, he inevitably will fail. The previous arrangement, in which the DCI provided leadership to the community, may have represented about as much direction as this inherently motley collection of components can bear. The DCI, and those elsewhere in the government who did business with him, had little doubt about his role. The DNI—lacking any institutional base other than the extra layer of bureaucracy known as the ODNI—will always have trouble figuring out what his own role is, or ought to be.

A LACK of clarity in the DNI’s legal authorities only adds to his difficulty in trying to scratch out a meaningful mission for himself. The DCI’s authorities accumulated through more than fifty years of legislation and executive orders. The drafters of those many documents, not anticipating what would happen in 2004 (when Congress created the position of DNI), did not use language that specified which of those powers belonged with the DCI because he ran the CIA and which belonged with him because he led the intelligence community. The hastily prepared legislation that split the DCI’s job into two did not come close to articulating which of the job’s functions devolved to the DNI and which to the current position of director of the CIA. Government lawyers have been negotiating over this ever since.

In trying to assert themselves amid this confusion, the DNI and his staff have naturally exploited the only solid line on the organization chart that extends downward from the ODNI to a preexisting agency: the one to the CIA. The lines to the other components of the intelligence community are only dotted, meaning some sort of leadership is supposed to be taking place but direct control resides elsewhere. (The DNI does not even fully control the National Counterterrorism Center. In one of the 9/11 Commission’s more convoluted bits of bureaucratic engineering, the head of the center has two bosses, reporting to the DNI on intelligence matters but directly to the White House on strategic planning for counterterrorism.) The inevitable tussles between the ODNI and the CIA have not only generated heat and distrust but also led to some results that cannot possibly improve governmental effectiveness. This has included, for example, the ODNI kicking CIA representatives off some interagency committees in order to claim the intelligence seat, even though the CIA was still directly and heavily involved in the subject at hand.

A few of the tussles have become public. The most notable one—which went to the White House for arbitration—concerned who should appoint overseas intelligence representatives. Some have pointed to this as an example of the kind of petty squabbling among intelligence bureaucrats over turf that makes vigorous reform all the more necessary. Yet the proponents of change fail to point out that the turf battle resulted directly from confusion created by the 9/11 Commission’s proposals. This was a good example of how intelligence reform, or supposed reform, feeds on itself. One round of it generates new problems that in turn feed a cottage industry of ideas for the next round. The whole question of the DNI’s overall future and role is a larger example of the same phenomenon.

CONSISTENT WITH the usual pattern of endlessly self-sustaining intelligence “reform,” the dominant response to the glaring problems with the DNI is to seek fixes to the fix. There has been no shortage of suggestions. Some probably would help, but all would have offsetting disadvantages.

Most criticism of the DNI has ignored the unachievable aspects of the expectations placed on him and the inherently contradictory aspects of the situation in which he has been placed. One line of criticism, for example—voiced by, among others, members of the 9/11 Commission—is that the ODNI has become too big, bloated and bureaucratic. Something leaner and meaner was intended, say the critics. But scrapping any staff functions of the ODNI means either that those functions do not get performed or that to carry them out the DNI must rely on the resources and goodwill of other agencies in the intelligence community. It is appealing to think that the DNI could draw on any capabilities throughout the community that he wants to, much as a military commander can draw on those of any units subordinate to his command. But because of the nature of the intelligence community, the DNI always will be more of a supplicant and dependent than a commander. The ODNI staff is the only hired help that he can always count on.

Another suggestion has been for the DNI to shed some of his roles. One area concerns the third major responsibility of the old DCI job, besides directing the CIA and leading the intelligence community: being the principal intelligence adviser to the president. This aspect of the DCI’s duties devolved to the DNI. Generally seen as part of this responsibility—and the part that in practice has consumed by far the largest portion of the DNI’s time and attention—is direct participation in the daily morning intelligence briefing of the president. Getting rid of this function, with all of the preparation time that goes with it, would free the DNI to devote his energy to leadership of the community.

But this change is exceedingly unlikely. During much of the history of the morning brief, subordinates handled the briefing chores, with directors (i.e., DCIs) seldom participating. It is only in more recent years that personal involvement in this morning ritual has come to be seen as a directorial function. But now that it is seen that way—and given that face time with the president is a highly prized commodity symbolizing access and influence—giving up the function probably would reduce the status of the DNI. Whatever he gained in freed-up time to lead the intelligence community he would lose with entrance of the perception that he no longer had the ear of the president. Residents of Washington probably would say he had been “Zorned”—after the name of a recent head coach of the Redskins football team who was widely viewed as having been emasculated when team management relieved him of his play-calling duties.

The President’s Intelligence Advisory Board recently weighed in with a different idea having a similar objective, which was for the ODNI to turn certain functions over to other agencies in the community—such as giving the CIA the job of evaluating security at overseas installations and having the DIA run the National Intelligence University—in the interest of enabling the DNI “to focus on strategic missions.” This would help to achieve the objective of leanness, if not meanness, in the ODNI. The functions are so secondary, however, that it is hard to believe that they, unlike the morning-briefing duty, constitute much of a diversion of the DNI’s time and attention. Moreover, they would hive off work in some of the very areas, such as common security and training standards, where the ODNI probably has done some good.

Most talk about making the job of the DNI more doable has centered on increasing the position’s statutory authority, including over personnel and budgets. A congressional fix along these lines might undo the one imprint that Congress placed on the 9/11 Commission’s plan, which it otherwise largely rubber-stamped: to ensure that the legislation of 2004 did not significantly curtail any control by the Department of Defense over its intelligence components. Every extra bit of legal authority would help any DNI, of course. But not much more authority could be granted without bumping up against legitimate issues of departmental cohesion and the jurisdiction of cabinet secretaries. This is not just a matter of congressional resistance and political realities. It is a matter of retaining a semblance of order and rationality in the larger organization chart of the executive branch. To try to turn the inherently messy, crosscutting intelligence community into something more like a unified entity under a single commander would do violence to that goal.

HERE’S THE best idea: do nothing—for now. Although this advice may seem contrary to the judgment that creating the DNI was a mistake, it reflects three other important truths.

One is that whether an initiative was a good idea in the first place is different from how to handle a problem that is in our laps now. In that respect the question of what to do about the DNI is similar to the issue of what to do about ill-considered wars. Acknowledgement that launching a war was a mistake is not the same as an argument that we ought to withdraw from it immediately. Likewise, recognition that a previous reorganization was not well thought-out does not constitute a case for rescinding it now. We cannot turn back the clock and undo past mistakes; we can only limit additional damage.

A second truth is that no amount of fixing can overcome the inherent challenges that underlie intelligence failures. No reorganization will eliminate the tenacity with which determined adversaries safeguard secrets or the impossibility of reliably forecasting foreign decisions yet to be made. No amount of bureaucratic engineering will enable intelligence services to achieve the omniscience that Americans so often seem to expect of them. It is natural, but unwarranted, to think that the next fix we make will be better than all the solutions that came before.

The third truth applies to any thought of another reorganization to correct the flaws of the last one. Shuffling authority always entails substantial costs in the form of disruption. These include everything from officers having to find out new telephone numbers of contacts in other components to superiors and subordinates who are thrown together for the first time having to develop new working relationships. For any reorganization to be warranted it has to offer substantial gross improvement, so that when the costs of disruption are factored in, it is still a net gain. That standard is difficult to meet, and much “reform” does not meet it.

Officers in the intelligence community are more adaptable than the common image of hidebound bureaucrats suggests. Many of them have had to adapt to previous reorganizations, and over the past five years they have been adjusting to the reforms that created the DNI. They have been implementing the kinds of practical, incremental adjustments that are better at making a flawed apparatus work than would any new scheme that is conceived a priori and imposed on them from the outside. Let them get on with their work.

THE POST of DNI will continue to be a mostly frustrating and thankless job. But the current framework, despite its major faults, still can be used to add value, as the ODNI already has in a few areas, mostly involving technical and supporting functions. Promoting interoperability of information technology is perhaps the best example.

There inevitably will be—regardless of what kind of bureaucratic apparatus is in place—additional intelligence failures, including ones deemed major enough to cause uproar. It also is inescapable that part of that uproar will be politically irresistible demands to overhaul U.S. intelligence. In other words, there will be more iterations of the endless cycle of recrimination and reform that has been going on for decades. The next phase—when a new reorganization, despite the costs of disruption, probably will be unavoidable—will be the occasion for trying to repair some of the damage of the last escapade. At that time reformers can consider whether to attempt a thorough consolidation of intelligence functions into a single national-intelligence service. Senator Pat Roberts (R-KS) proposed such a consolidation when he was chairman of the Senate Intelligence Committee, but his proposal never received the attention it deserved because the 9/11 Commission’s plan came to monopolize the reformist playing field. Other options would be to return to a DCI system, or possibly something else. The most important thing will be to subject whatever ideas gain traction to the kind of careful scrutiny that never happened with those of the 9/11 Commission.

In the meantime, the existing intelligence-community apparatus can serve as a monument to the costs of catharsis. Let its difficulties remind us of what can happen when, as in the aftermath of 9/11, we want to be reassured more than we want to be right.

Paul R. Pillar is director of graduate studies at Georgetown University’s Security Studies Program and a former national intelligence officer for the Near East and South Asia.