Abolish the ODNI

By John R. Bolton

Thursday, July 16, 2026

 

William Pulte’s appointment as acting director of national intelligence, a position he will hold while simultaneously remaining head of the Federal Housing Finance Agency, caused remarkably bipartisan shock in Washington. Donald Trump’s adviser Steve Bannon called the Pulte announcement “a f*** you to the Deep State,” which accurately reflects Trump’s contempt for the  intelligence community and his lack of interest in its work — except when it’s searching for evidence of foreign tampering in the 2020 presidential election.

 

Considerable ink has been spilled over Pulte’s lack of qualifications, and his new status is decidedly temporary. Nonetheless, Trump’s appalling selection presents an opportunity to review the fundamental question of whether the Office of the Director of National Intelligence (ODNI) should exist at all. The nomination of Pulte’s predecessor, Tulsi Gabbard, somehow failed to prompt such a review. Thank goodness for second chances, because the ODNI’s “one ring to rule them all” approach has not and will not solve the problems it was meant to address, and it has created new ones of its own.

 

Structuring the far-flung intelligence community (IC) and, even more important, defining and managing its relationship with policymakers, has frustrated presidents and their administrations since the “big bang” of American intelligence during World War II. Before then, intelligence was almost exclusively a function of the military, conducted by the War and Navy Departments and focused on the intentions and capabilities of potential adversaries. George Washington, as commander in chief of the Continental Army, was his own top spy-runner and a master of the game.

 

Before World War II, State Department ambassadors also gathered as much information as possible, openly and not so openly, as diplomats have done since time immemorial. They did not distinguish between what we today call “open-source” intelligence and what they gathered by cloak-and-dagger methods, and there was no separate intelligence service within State. There existed, however, a very strong disinclination to engage in “spying.” As Secretary of State Henry Stimson wrote in 1929, “Gentlemen do not read other gentlemen’s mail.” In reality, they did that and more. Diplomats were propagandists and clandestine operators. After all, as the British diplomat Sir Henry Wotton wrote in 1604, “An ambassador is an honest man sent to lie abroad for the good of his country.”

 

The contours of today’s struggle to structure the IC were set by the National Security Act of 1947, which among many other things established the Central Intelligence Agency. As the agency’s name suggests, its creators hoped to bring order to the proliferation of government intelligence units. These were particularly clustered in another of the statute’s major innovations, the Department of Defense, which was created by merging the War and Navy Departments and establishing a new Department of the Air Force. The overall direction of national security policy of course remained with the president, who would now be assisted by the act’s other major creation, the National Security Council.

 

President Harry Truman created the position of director of central intelligence (DCI) by executive order in 1946. The DCI also became the director of the CIA, which meant that one individual was charged with leading the work of all of the government’s intelligence components in Defense and in other cabinet departments. The DCI’s primary responsibility was to coordinate intelligence activities, avoiding unnecessary duplication and overlap and harmonizing the often conflicting analyses of the various intelligence operations.

 

This structure created two long-standing problems. First, the dual-hatted director was not automatically a cabinet-level official. All the bureaucratic components he was charged with leading were headed by cabinet secretaries who were often protective of their respective turfs, especially the heavyweight defense secretary, whose department housed the majority of intelligence activities. While the CIA grew quickly (“Bigger than State by ’48” was its battle cry), the DCI rarely won major turf fights. Second, other agencies, mostly within the Defense Department, contended that the dual-hatted role created an inherent conflict of interest, biasing the director in favor of the CIA’s institutional positions and against the rest of the IC.

 

These disputes percolated for decades until the 9/11 attacks, which provoked an important debate on why the IC failed to detect the impending terrorist plot. The government’s first response was to create the Terrorist Threat Integration Center, which was renamed the National Counterterrorism Center in 2004 and is now housed in the ODNI. Critics of the dual-hatted CIA director also seized the opportunity to elevate responsibility for coordinating the IC to a higher-ranking official. They argued that this was the only way to take on the mighty defense secretary.

 

The ODNI, created by statute in 2004 and launched the following year, was originally envisioned as a small office with IC-wide oversight. Its purpose was to enforce budget discipline and manage IC priorities, including intelligence-gathering and analysis and clandestine activities for the president. Over time, however, bureaucratic creep caused the ODNI’s responsibilities and staff to multiply. Claims of bureaucratic empire-building, duplication and overlap with other departments, and inevitable personality conflicts made things worse, especially in analytical fields.

 

Turf fights erupted, including entirely foreseeable clashes between the ODNI and every other agency. Even today, on Iran for example, the CIA and the ODNI are reportedly withholding information from each other. Moreover, the ODNI has itself begun competing for resources, undercutting the idea that it would be impartial in its decision-making among IC agencies. The NSC’s role has also decreased as IC components have fought among themselves for programs, staff, and money. Moreover, in most presidential administrations since 2005, both the DNI and the CIA director have attended NSC meetings and lower-level meetings of NSC members, sometimes agreeing and sometimes not.

 

The debate over the ODNI, however, largely missed the real issue: how best to serve the executive branch’s consumers of intelligence. After all, the IC is not a self-licking ice-cream cone: It exists not to satisfy its own curiosity but to acquire information that policymakers can put to practical use. Intelligence consumers — decision-makers in key national security departments and agencies — frequently believe that they are unable to influence the IC’s seemingly impenetrable budget and policy process. While decision-makers appreciate the enormous importance of the intelligence they do receive, they often feel excluded from the process that determines intelligence-collection priorities and resource allocation. In some respects, various intelligence agencies and bureaus within departments serve as proxies for their main customers in the scramble for staff and funding. But that is not the same as direct, frequent participation by intelligence consumers in decisions that allocate resources.

 

None of the foregoing is intended as a criticism of the personnel or leadership of the ODNI over the years. Nor can the long-term failure to find an optimal solution be primarily attributed to intelligence agencies. The real problem has been the absence of clear resolve from the president and the NSC to straighten things out. I confess that during my time as national security adviser, I struggled, often unsuccessfully, to keep the many policy trains on their tracks. The prospect of making long-term improvements anywhere, including in the IC, was just a tantalizing mirage.

 

With this abbreviated history in mind, how should we proceed? Given the importance of an effective IC to national security, there has been no lack of suggestions over the years. And there is certainly no perfect outcome. Nonetheless, the weight of decades of experience points to restoring the CIA director as the dual-hatted director of central intelligence, with one major modification. The National Security Council must be equipped to perform a larger oversight role — not only to increase IC coordination but to ensure that the community as a whole and each of its agencies is in fact providing policymakers with the intelligence they need to do their jobs effectively.

 

The principal misconception of the ODNI model is that coordination among intelligence agencies can only be accomplished by someone above the CIA director. There are 18 members of the IC. Remove the ODNI, and nine of the remaining 17 are components of the Defense Department. The other seven are part of the Justice Department (the FBI and the DEA); the Department of Homeland Security (an intelligence and analysis office and the Coast Guard); and the Departments of State, Treasury, and Energy (small analysis agencies).

 

Adding the ODNI to this mix simply compounded the problem and perhaps even further obscured it, notwithstanding more than two decades of good-faith efforts to make the arrangement work. This is not a question of bureaucratic size. The Trump administration proposed significant budget cuts and personnel reductions to the ODNI, as it did to other intelligence organizations. Budget and staff reductions, however, are not the real issue. Congress and the Trump administration have recognized that the United States requires more resources to address diverse and growing national security threats. Appropriately, defense budgets are rising, and intelligence agencies should not be shortchanged. Instead, the problem is one of sclerosis in the bureaucratic process. Adding another layer to a sclerotic system only increased the problem over time.

 

Another pro-ODNI argument, beyond questions of management and budgeting, is that the president benefits from having one principal intelligence adviser. But that role depends far more on the individual occupying it than on any organizational chart, and it can be performed equally well by a dual-hatted CIA director. Bill Pulte hardly compares with John Negroponte (the first DNI), or with Dan Coats (the DNI in Trump’s first term), or with prior CIA directors such as William Webster or Allen Dulles. This argument also assumes that the DNI, unlike Gabbard and Pulte, will actually do his job.

 

Moreover, while greater internal IC cohesiveness has its merits, policymakers facing hard choices often benefit from hearing different perspectives on intelligence. It is easy, and often correct, to say that duplication is wasteful. But in many cases, the unique goals and perspectives of individual agencies provide a more useful check on institutional biases than “coordination” and, worse, the homogenization of analytical products. This is one reason why, for example, military intelligence often differs from that of the CIA. It is not a question of one being invariably better than the other. Intelligence is not a crystal ball that reveals ineluctable pathways for national security policy. It is something that will always be disputed by policymakers with differing viewpoints.

 

There are legitimate concerns about whether the IC should devote more resources to collecting intelligence rather than analyzing it, a function performed by all IC components and particularly by the ODNI. The IC’s analytical responsibilities are unquestionably valuable. Agencies can answer questions ranging from the highly technical (What are those fuzzy bumps on a satellite picture?) to the most abstract (What is Vladimir’s Putin’s long-term objective in Ukraine?). But you can analyze a subject to death without reaching a conclusion, and policymakers often do not have that luxury. Analysis necessarily involves value judgments that also affect strategic and tactical calculations. It is not a sorcerer’s stone that determines policy but only one factor among many that policymakers must consider.

 

Before the ODNI was created, an “intelligence community staff” carried out the DCI’s government-wide functions. Much of that staff was simply transferred to the ODNI when it was launched and could easily be transferred back to serve under a restored DCI. The National Intelligence Council, for example, consists of experts on regional and functional issues such as arms control. It provides, as described on the ODNI website, “strategic, rigorous, and objective estimative all-source analysis.” But it can do this wherever it is housed bureaucratically. Similarly, the National Counterterrorism Center could be positioned administratively under the dual-hatted DCI without losing its analytical independence. The same is true for other coordinating functions of the ODNI such as counterintelligence, counterproliferation of weapons of mass destruction, and cybersecurity.

 

But beyond organizational charts and wiring diagrams is the imperative of satisfying intelligence consumers, particularly those far removed from critical decisions on the allocation of IC resources. How much should be devoted to collection and clandestine operations rather than analysis? How should resources be divided among countries or global regions? How much attention should be paid to nuclear, chemical, and biological weapons, or to conventional weapons intelligence, or to cyberspace, or to the multiplying forms of asymmetric warfare? As Congress envisioned in 1947, the NSC is the appropriate forum to resolve these issues, which are intimately linked to the policy decisions that the president and NSC members must make.

 

As Trump’s ODNI nominee Jay Clayton undergoes his confirmation hearings, senators should probe these issues and more. After two decades, we have an important opportunity to look back at the ODNI and make significant course corrections, thereby enabling a well-functioning  intelligence community to fulfill its critical national security responsibilities.