If the case of Edward Snowden -- the former employee at Booz Allen Hamilton (a defense contractor for the National Security Agency) who smuggled classified information out of his workplace and provided it to news organizations -- has revealed anything, it is that the U.S. intelligence services made some mistakes as they reformed after 9/11 and the Iraq war. It is true that many of the changes they made have been for the better. But some, such as increased collaboration and information sharing, have now backfired. Snowden’s case, and possibly that of army private Bradley Manning, who is on trial for sharing classified documents with WikiLeaks, highlight the problems caused by involving more people, more agencies, more contractors, and more classification in intelligence gathering and analysis -- all trends that have accelerated in the past decade. In light of this, and of the first decline in intelligence spending since 2001, it is time to rethink how the U.S. intelligence community functions yet again.

First, the very structure of the intelligence community should be updated. The most important post-9/11 reform -- indeed, the most important reform since the creation of the Central Intelligence Agency in 1947 -- was the establishment in 2004 of an Office of the Director of National Intelligence. The ODNI is supposed to coordinate the United States’ 16 separate intelligence agencies to prevent important intelligence, i.e., Osama bin Laden’s plans to attack, from slipping through the cracks. However, its creation turned out to be a half measure. Instead of bringing together the spooks at the CIA, the analysts at the Department of Energy Office of Intelligence, and the agents at the Drug Enforcement Administration, for example, it only added a superfluous layer of bureaucracy (some 1,600 staffers in all) to an already clunky intelligence community.

The reason: ODNI was given little actual control over the operations and budgets of the sprawling intelligence infrastructure, which is still dominated by much more powerful agencies such as the CIA, the National Security Agency, and the Defense Intelligence Agency. This became clear in 2009 when Dennis Blair, who was then the head of ODNI, tried to assert his authority to pick the senior U.S. intelligence representative in each overseas embassy, instead of relying on the CIA station chief to fill that role. In most cases, the CIA would still have retained the chief-of-station position, but in some countries where the CIA didn’t have someone to take the seat, it might have gone to the NSA or even the DEA. Leon Panetta, who was director of the CIA at the time, balked. A bruising turf battle ensued and Panetta emerged victorious, which only underscored the ODNI’s lack of clout.

A more powerful ODNI would not have prevented the Snowden or Manning leaks. But giving the office more clout would be a first step in rationalizing the way U.S. intelligence operates overall. Congress needs to either grant the body real control over the intelligence budget and enact Blair’s proposed reform, or else transform it into a smaller, more focused organization built around the existing National Intelligence Council, which is arguably the only well-functioning unit in the ODNI. The National Intelligence Council currently operates as an in-house think tank that focuses on long-term strategic analysis and is staffed by top-flight analysts from all the different intelligence agencies as well as academia and the private sector.

The Snowden case also revealed the need to revamp security classification guidelines: fewer documents should be classified, and the number of people who have access to real secrets should be limited. It is absurd that the most basic and routine information continues to be classified. For example, the existence of a CIA training facility, Camp Peary (a.k.a. the farm), in Williamsburg, Virginia, is supposedly secret even though Camp Peary has a Wikipedia page that lists its exact coordinates. A good place to start would be to declassify many “confidential” and “sensitive” documents -- the lowest levels of classification. Serious consideration should be also be given to declassifying “secret” documents, whose contents also exist in public literature and don’t contain sources that need to be protected.

Today’s rampant over-classification, with trillions of pages of digital files locked away every year, has ripple effects. More classified data means that more people need security clearances. Today, according to the Associated Press, 4.9 million government workers and contractors have security clearances. Statistically speaking, there are bound to be some bad apples, Manning and Snowden among them. Setting a higher bar for what needs to be secret would allow the government to give access to fewer people.

The most obvious place to start is restricting access for contractors such as Snowden. After 9/11, the intelligence agencies rapidly expanded their workforces and had to contract out some roles to fill gaps, including in language and technical skills such as building and operating surveillance drones. According to The Washington Post, contractors now constitute roughly 30 percent of the intelligence community’s workforce and some 70 percent of its budget. That is too much. Intelligence collection and analysis are inherently sensitive tasks that should be reserved primarily for federal employees who have been more extensively vetted and trained than is typical of contractors.

Yet even the intelligence community’s own vetting and hiring processes are due for a change. Personnel practices continue to be designed, above all else, to weed out potential security risks -- not necessarily to recruit the most qualified analysts or case officers. These practices have not prevented damaging leaks of information; they have, however, made it exceedingly hard to hire “hyphenated” Americans -- the very citizens who might be helpful in intelligence operations against China or Iran. And, as the Snowden and Manning cases have shown (along with those of a whole slew of other homegrown traitors ranging from Robert Hanssen, the FBI agent who spied for the Soviet, then Russian, intelligence services for 22 years, and Aldrich Ames, the CIA officer whose nine years of spying for the Soviet Union and Russia was outmatched only by Hanssen’s), the assumption that naturalized citizens are somehow more likely to be compromised than native-born hires is deeply flawed. Foreign-born personnel do work and excel throughout the community, but it is more difficult for them than it should be.

Not only should the intelligence community hire more foreign-born Americans with language and cultural knowledge, it should also hire more mid-career professionals, who have had extensive and direct experience in a target country, whether through academia, business, charity, journalism, or in some other capacity. That runs counter to the community’s long-standing preference for hiring students right out of college or graduate school, believing that mid-career hires are more likely to have been compromised at some point by a foreign intelligence service. Still, older professionals bring greater knowledge and maturity to their tasks and require less on-the-job training. They also have a long track record that the agencies can review, unlike a young unknown out of college.

Finally, the intelligence community needs to revamp its analytical products, such as its National Intelligence Estimates, and the way it collects human intelligence (HUMINT) -- namely, traditional spying with case officers recruiting agents. The “failures of imagination” noted by the 9/11 Commission remain pervasive in the intelligence community. Rather than appreciating bold judgments, analysts swung toward extraordinary caution. Their work often focuses on amalgamating all potentially relevant data and judgments, and is so heavily caveated it is of little use to anyone.

Quality is further diminished by efforts to present a unified position across the entire intelligence community -- an effort coordinated by ODNI. The result is lowest-common-denominator work. Even National Intelligence Estimates, the crown jewel in the community’s analytical efforts, are often no more profound than what the major think tanks produce -- and usually have little impact on policymakers. The agencies should instead prize fewer but higher-quality reports.

And that can be accompanied by a renewed emphasis on HUMINT. The entire intelligence community has become too reliant on technological sources of intelligence, such as the NSA databases that Snowden tapped. Surprising as it might be, there are still many places that technology simply cannot reach. Consider the limitations of using technology to delve inside the mind of terrorists to figure out where they may strike next. If the terrorists are at all savvy in their security practices, they will avoid leaving an obvious electronic trail. Lack of good HUMINT, in fact, explains the biggest intelligence failures of recent times -- the failure to prevent 9/11 or adequately assessing Saddam Hussein’s WMD.

To fill in for technology’s blind spots, the U.S. intelligence community needs higher-quality human collectors, including greater numbers of case officers who operate under more inventive “nonofficial cover.” That is, posing as business executives or journalists, rather than operating under diplomatic protection at U.S. embassies abroad. It is true that it is inherently more dangerous to deploy these types of case officers. It is also expensive to build a credible nonofficial cover. But it is precisely case officers working under nonofficial cover who are more likely to penetrate terrorist cells. Here, hiring mid-career professionals would be especially important: newly recruited clandestine service officers can use their former professions to establish convincing cover for their activities. Yet the CIA currently imposes a maximum age limit of 35 for recruits to its clandestine service, although exceptions are made in some cases.

All this is not to say that the U.S. intelligence community is broken. In fact, whereas other intelligence services excel at certain niches and in certain regions, no other intelligence service in the world has such a broad range of global capabilities as the United States. A malcontent such as Snowden is an important, but very rare anomaly: the intelligence workforce is overwhelmingly bright, dedicated, hard-working, and patriotic. But that does not mean there is no room for reform. It would take only a few tweaks at the margins to vastly enhance the effectiveness of U.S. intelligence agencies -- and prevent major leaks in the future.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • SUE MI TERRY is a former CIA analyst who has worked at the National Security Council and the National Intelligence Council. She is now a senior research scholar at the Weatherhead East Asian Institute at Columbia University.
  • More By Sue Mi Terry