Ten years after 9/11, the least reformed part of America’s intelligence system is not the CIA or the FBI, but Congress. The September 11 terrorist attacks sparked major efforts to transform executive branch intelligence agencies such as the CIA and the FBI. These include the creation of the Office of the Director of National Intelligence (ODNI), the most sweeping intelligence restructuring since the establishment of the CIA in 1947; the formation of the Department of Homeland Security (DHS), which combined twenty-two agencies and two hundred thousand employees to provide “one face at the border”; dramatic initiatives to transform the FBI from a law enforcement to a domestic intelligence agency; and the proliferation of more than seventy regional, state, and local terrorist fusion centers to integrate terrorist-threat reporting across the country.

Although reforms have generated some major successes—including the killing of Osama bin Laden this year—not all intelligence improvement efforts have actually produced improvements. Some reforms have failed. Many have not gone far enough or fast enough. Others have proven counterproductive, creating more red tape and fatigue than results. Recent terrorist plots, including the 2009 Fort Hood shootings, the 2009 Christmas Day underwear bomber, and the May 2010 Times Square car bomb plot, remind us all too well that serious weaknesses remain in the American intelligence system.

wanted posters
U.S. soldiers carry wanted posters for Osama bin Laden near Mazar-e Sharif, Afghanistan, in 2001. Ultimately, the successful hunt for the terrorist leader required no tanks, aerial bombardment, or invasion. It came down to a small, elite commando team and years’ worth of painstakingly collected and analyzed intelligence.

Despite this record, it is clear that the seventeen agencies that comprise the United States intelligence community are expending considerable energy attempting to adapt to ever-changing terrorist threats. As one senior FBI official put it, “This is all I do, OK? Twenty-four/seven, 365 days a year. I don’t have a wife. I don’t have kids. It’s all I think about.” The same is true in the field. “The burnout rate in my Al-Qaeda squad is terrible,” noted one FBI agent in May 2010. “And these are agents who have done other CT [counterterrorism] work, where the pace is already tough. They’re just getting crushed by the load.” Just above the doorway that leads to the CIA’s Counterterrorism Center hangs a sign that reads, “Today’s date is September 12, 2001.” Spend any time there, or in a military unit in Afghanistan, the New York Police Department, the Office of the Director of National Intelligence, or any of the other federal intelligence agencies charged with collecting and analyzing intelligence, and you will quickly realize just how many people are working feverishly to adapt to what they call, simply, “the mission.”

“This is all I do, OK?” one senior FBI official said of the antiterror effort. “Twenty-four/seven, 365 days a year . . . it’s all I think about.”

Congress is another story. While Congress has been instrumental in many post-9/11 executive branch reforms, Congress has been largely unable to reform itself. In 2004, the 9/11 Commission called congressional oversight “dysfunctional,” and warned that fixing oversight weaknesses would be both essential to American national security and exceedingly difficult to achieve. One year later, the commission’s report card gave efforts to improve intelligence oversight a D. That same year, a second blue-ribbon commission (chaired by Judge Laurence Silberman and former senator Chuck Robb), which was tasked with examining what went wrong with estimates of Iraqi weapons of mass destruction, joined the call for oversight reform. That commission’s final report concluded that “many sound past proposals for intelligence reform have withered on the vine. Either the intelligence community is inherently resistant to outside recommendations or it lacks the institutional capacity to implement them. In either case, sustained external oversight is necessary” (emphasis mine). Although many of the Silberman-Robb Commission’s executive branch reforms were adopted, its congressional reforms were not embraced.

In the twenty-first-century threat environment, intelligence eclipses military firepower as the most important line of defense.

By fall 2007, the Senate Select Committee on Intelligence had grown so deeply concerned that it held a hearing on itself. I testified at that hearing, along with Lee Hamilton, who served as the 9/11 Commission’s vice chairman and earlier as chairman of the House Permanent Select Committee on Intelligence. Pointing at all the senators around the room, Hamilton delivered an ominous warning:

To me, the strong point simply is that the Senate of the United States and the House of the United States is [sic] not doing its job. And because you’re not doing the job, the country is not as safe as it ought to be. . . . You’re dealing here with the national security of the United States, and the Senate and the House ought to have the deep down feeling that we’ve got to get this thing right.

Hamilton’s words prompted vigorous nods of agreement across the aisle, but never made headlines or produced major changes. Instead, the committee’s own oversight-reform centerpiece—consolidating appropriations and authorization powers—quickly and quietly died. In May 2009, then–House Speaker Nancy Pelosi declared in a press conference that oversight had become so feckless that the only way to change Bush administration intelligence policies was to oust Republicans at the polls. And in January 2010, eight years after 9/11, a third blue-ribbon commission, the Commission on the Prevention of Weapons of Mass Destruction Proliferation and Terrorism, concluded that congressional intelligence and homeland security reform efforts were still failing. Notably, nearly all of the proposed oversight changes required simply modifying internal congressional rules and committee jurisdictions, not passing new laws.

By fall 2007, the Senate Select Committee on Intelligence had grown so deeply concerned that it held a hearing on itself.

It’s well worth examining the roots of weak intelligence oversight to discern why deficiencies have persisted for so long, despite the clarion call for change after 9/11 and the unprecedented importance of intelligence in today’s threat environment. As many have noted, during the Cold War the Soviet adversary was easy to find but hard to kill, making military firepower the key to success. Today, the situation is reversed; our terrorist enemies are hard to find but easy to kill. It is now the weak who threaten the strong. Driven by fanaticism, hidden from view, and armed with Internet connections and deadly weapons that can fit into a suitcase or vial, small bands of transnational terrorists can wreak catastrophic damage as never before. In the twenty-first-century threat environment, intelligence has eclipsed military firepower as the nation’s most important line of defense.

Nowhere was the importance of intelligence more on display than in the May 1 operation to kill Osama bin Laden at his Pakistani hideout. Success required no tanks, aerial bombardment, or invasion. The “boots on the ground” did not number in the hundreds of thousands, or even the hundreds. Instead, the operation to nab the world’s most dangerous terrorist came down to just two critical elements: a small, elite Navy SEAL team and actionable intelligence that had been painstakingly collected and analyzed over a period of years.


Intelligence agencies cannot go it alone. Legislative oversight, done well, ensures that the intelligence community gets the resources it needs and deploys those resources to maximum effect. Good oversight sets agencies’ strategic priorities and pushes them to improve by asking tough questions and demanding better answers. Good oversight also maintains accountability by ensuring compliance with the law and generating public trust and support for agencies that must, by necessity, hide much of what they do.

Oversight done poorly, however, can hurt our national intelligence effort. Congressional micromanagement of intelligence activities serves no one’s interests well, distracting agencies from their missions and miring legislators in the day-to-day rather than focusing attention on the bigger and more important questions of what American intelligence agencies should be doing, how well they are doing it, and where they should be heading in the future. Accountability efforts can and often do go awry, saddling intelligence officials with onerous reporting requirements that sap time and attention from their day jobs to answer queries, write letters, and produce reports no one reads. In 2009, for example, the Department of Homeland Security spent 66 work-years responding to congressional questions, giving 2,058 briefings and sending 232 witnesses to 166 hearings. “It’s disgraceful,” said Representative Peter King, who chairs the House Homeland Security Committee. “There’s no good reason.”

Pennsylvania Ave
A surveillance camera keeps watch in Washington. Many observers acknowledge that the intelligence community often does not fully, proactively, or clearly disclose what it knows or does.

In addition, too often oversight consists of blaming agencies when things turn ugly and the cameras are rolling rather than partnering with them to prevent mishaps in the first place or working constructively to address weaknesses when no one is watching. Responsiveness can also be a double-edged sword, either keeping unelected intelligence officials from running amok or making them run in too many directions at once. Intelligence oversight, in short, is a critical component of national security in the twenty-first century. Getting it right is hard, and getting it wrong is dangerous.

Since 9/11, most explanations of Congress’s weak intelligence oversight have blamed the executive branch. Critics have focused particularly on the Bush administration and the extent to which officials withheld information from Congress about secret and controversial programs such as the National Security Agency’s warrantless wiretapping, the CIA’s use of harsh interrogation methods, and the establishment of CIA black sites to detain suspected terrorists abroad.

Although some episodes have a partisan and political cast, it is important to note that many people, including Republican staffers and career intelligence officials, have acknowledged that the intelligence community often does not fully, proactively, or clearly disclose what it knows or does. Executive branch secrecy is an important and complicated part of the oversight story, but it is by no means the only part. Congress also has struggled to bolster its own intelligence oversight capabilities for years, with limited success. Observers have been quick to examine how the executive branch has asserted broad powers and guarded information, but slow to understand why Congress has not strengthened its own oversight tools. What former Senate Intelligence Committee chairman John D. Rockefeller IV has called the “long and sordid history” of congressional oversight weaknesses is not widely known, even though it began before 9/11 and continues today.

In 2009, Homeland Security spent 66 work-years responding to congressional questions, giving 2,058 briefings and sending 232 witnesses to 166 hearings.

Many of Congress’s biggest oversight problems lie with Congress. Let me be clear: by Congress I mean the institution, not the political wrangling between parties or personalities that usually makes headlines. Much has been written about party polarization and the decline of the bipartisan foreign-policy consensus since the Cold War. While intelligence policy making has undoubtedly become more partisan and rancorous in recent years, and while individual personalities matter, I find that the root causes of intelligence-oversight dysfunction cross party lines, presidential administrations, individual congressional leaders, and eras.

Simply put, Congress has never expended as much effort overseeing intelligence as other policy areas. Whether Democrats or Republicans controlled the House and Senate; whether government was unified (with a single party controlling the presidency and both houses of Congress) or divided between parties; whether a committee was run by a charismatic chair or a weak one; or whether parties were more or less polarized from each other has never mattered as much as we think. Indeed, party ideology scores show quite clearly that congressional intelligence-committee members have not been drawn from the extreme wings of the Democratic and Republican parties for the past twenty years. Instead, the intelligence committees have been populated by legislators who are typically more moderate than their party colleagues in the House and Senate.

The real story is not about personalities, political animus, or ideological conflict. It’s about how Congress has collectively and persistently tied its own hands.

The sources of intelligence oversight deficiencies run deep. They rest in the electoral incentives that drive all legislators—from liberal Democrats to conservative Republicans—to behave in the ways they do, and in committee turf battles that almost always protect the status quo. The real story is not about personalities, political animus, or ideological conflict. It is about how Congress has collectively and persistently tied its own hands in intelligence oversight for a very long time.

Two institutional weaknesses are paramount: rules, procedures, and practices that have hindered the development of legislative expertise in intelligence; and committee jurisdictions and policies that have fragmented Congress’s budgetary power over executive branch intelligence agencies. These two weaknesses did not arise by accident. They were self-inflicted. In both areas, electoral incentives and internal congressional turf battles have led Congress to limit its own oversight capabilities even when the problems are well known and the national security stakes are high.

Ten years after 9/11, the United States has an intelligence oversight system that is well designed to serve the re-election interests of individual legislators and protect congressional committee prerogatives, but poorly designed to serve the national interest.


Here are the themes I explore at length in my new book, Eyes on Spies: Congress and the United States Intelligence Community:

  • What good oversight looks like. This question is more important and knottier than it seems. Political scientists and policy makers have never had a consensus working definition of “oversight” or what constitutes “effectiveness.” But a review of the history of intelligence oversight; interviews with intelligence officials, legislators, and congressional staff; and quantitative oversight metrics all suggest that something is amiss. However one defines good oversight, Congress has not been doing it in intelligence for a very long time.
  • The state of the research. An examination of research in both political science and intelligence studies finds that both literatures have insights and limitations in their understanding of weaknesses in intelligence oversight. Political scientists seek generalities that obscure key realities of intelligence. Scholars in intelligence studies, on the other hand, emphasize the distinctive features of espionage and overlook the roots of oversight dysfunction. The result is that one literature is too broad, the other too narrow. Taken together, however, they provide essential elements for understanding why intelligence oversight has remained so problematic for so long.
  • Why intelligence policy is so hard to make. Turning from logic to data leads to a comparison of intelligence oversight to other policy areas. Surprisingly, this kind of comparative analysis has never been done before. By applying the two most widely known oversight models (police patrol and fire alarm) to a range of policy issues and developing several original oversight datasets, Julie Quinn and I find that Congress is not overseeing nearly as much in intelligence as in other policy domains. Electoral incentives are the explanation. The desire to win re-election naturally leads legislators to consider voter preferences, heed interest-group demands, and cater to district industries and constituencies. But the average re-election-minded representative is motivated to spend as little time as possible on intelligence. Why? Because it is the worst of all worlds: a complicated policy area that requires large upfront investments of time to master but promises low political payoffs and a non-trivial degree of political risk. In short, intelligence oversight is always an uphill battle because the issue is always a political loser.
  • What’s wrong with the current model. I have examined more closely the flaws of intelligence oversight and found two crucial institutional deficiencies: limited expertise and weak, fragmented budgetary authority. These are serious. In politics, knowledge and money are two of the most potent weapons. The intelligence committees lack both. Here, too, rational self-interest explains why.
  • The future of intelligence oversight. To anyone considering the policy implications, the picture is not encouraging. If my analysis is correct, the sources of oversight dysfunction lie with electoral incentives and institutional prerogatives. These will not disappear any time soon. As David Mayhew argued thirty years ago, congressional rules and structures are designed in ways that maximize the re-election interests of individual members. Intelligence after 9/11 is no exception.
overlay image