Analyze this: inside the one spy agency that got pre-war intelligence on Iraq--and much else--right.

In August 1998, North Korea's military caught Western intelligence off-guard when it announced it had successfully tested a ballistic missile called the Taepo-Dong 1, considered a precursor to the kind of intercontinental weaponry that could threaten America's Pacific coast.

Conservatives in Congress condemned the CIA for being caught with its pants down. Brandishing a report by a committee headed by the once and future Secretary of Defense Donald Rumsfeld, which called rogue states with missiles the biggest threat to American security, these same conservatives pressed the Clinton administration to set aside money for a missile shield. In an effort to play catch-up, the CIA put a team of its best analysts on the case, sorting through satellite photos, tips from defectors, and electronic intercepts to determine how close North Korea was to possessing intercontinental missiles. Six months later, Director George Tenet delivered the CIA's conclusion in testimony before the Senate: Contrary to its own earlier analysis, the CIA now believed that North Korea would test an intercontinental missile in the "near future." In response to this new threat, the Clinton administration earmarked $6.6 billion over five years to develop a missile-defense system.

CIA analysts weren't the only ones poring over the data, however. Throughout the government, other intelligence agencies were looking at the same material, and one of these shops came to a markedly different conclusion. Over at the State Department's Bureau of Intelligence and Research (INR), analysts argued that the North Koreans were much farther off than the CIA believed. North Korea could potentially threaten the United States within a decade "only if it abandons its current moratorium on long-range missile flight testing," Tom Fingar, then-acting principal deputy, assistant secretary of INR, testified before Congress in February 2001. Although the White House and Congress accepted the CIA's analysis, INR ultimately proved to he correct. In the five years since Tenet's testimony, North Korea has yet to test an intercontinental ballistic missile.

This wasn't just a case of dumb luck for INR. Over the last decade, INR has frequently arrived at more prescient conclusions than the CIA and other intelligence agencies about the nature of threats to the United States. In 2001, when U.S. intelligence agents intercepted a shipment of aluminum robes hound for Iraq, CIA analysts concluded they were for uranium enrichment, proof that Saddam Hussein was building a secret nuclear-weapons program. The INR, working from the same body of intelligence, concluded (rightly, it turned out) that the tubes were more likely intended for conventional, not nuclear, weaponry.

Indeed, on the whole question of Iraq's nuclear capabilities, INR came consistently closer to the truth than did other agencies. The intelligence community's collective analysis on this issue was assembled by CIA Director Tenet in a 2002 National Intelligence Estimate, a document that contains the consensus of all the government's intelligence agencies: the CIA, FBI, NSA, and intelligence groups at State, Energy, and Defense, among others. The 2002 estimate included the crux of the case that the Bush administration would present to the American public and the world in arguing for war. And the document's conclusion--that Iraq was three to five years away from being capable of building nuclear weapons--convinced many Democrats and other skeptics that a war in Iraq might be justified. But when INR received a draft of the estimate, it balked, believing that other intelligence agencies had vastly overestimated the status and capability of Iraq's nuclear program. INR thought the whole report was flawed; rather than including minor objections to specific statements, it took its name off of the estimate and detailed its objections in one long endnote. Of course, when American soldiers and U.N. inspectors combed through Iraq in the wake of the U.S. military conquest, what they found--and didn't find--confirmed INR's suspicions. The rest of the intelligence community had gotten it wrong again.

INR hasn't always called it right. The agency joined the rest of the intelligence community in greatly overestimating Iraq's chemical and biological weapons capability. And there have been some matters which the CIA has gotten right and INR wrong. INR dissented when the CIA concluded that Pakistan was selling M-11 missiles to China in the 1990s, though it turned out that the sales actually were taking place. Still, INR's overall record has been impressive. Richard Clarke, a former counterterrorism advisor to three presidents and a former senior INR official, called INR "the intelligence-analysis organization with the best track record for accuracy."

In the wake of Congress' restructuring of the U.S. intelligence establishment, the new national intelligence czar faces a mountainous task: compelling the CINs cowed and moribund intelligence collectors to get out into the field, coordinating disparate and warring bureaucracies, and coping with an ever more shadowy and cunning terrorist threat. Perhaps the new czar's most critical job will be improving the performance of analysts at the nation's intelligence agencies--the people who look at the material, connect the dots; and decide what it means. Thankfully, INR provides a terrific model for how the analytic shops at different agencies should operate. The new czar need only learn its secrets: that its performance derives from its analysts' high quality and depth of experience, from their facility with foreign languages, from their human contacts within the regions they study, and, most importantly, from an institutional culture that does not just tolerate dissent--but actually encourages it.

No plans to grow

It may be one of the smallest and least-noticed intelligence agencies, but INR developed a track record early on in its history of getting the big questions right. The office was cleaved off from the World War II Office of Special Services (OSS): OSS' operations department became the CIA. But its analysts were moved into a new agency under the auspices of the State Department: INR. From the beginning, its mission was to provide the secretary of state with independent analysis of the intelligence gathered by Foreign Service officers and other agencies on the primary threats of the day. And though the CIA, NSA, and other agencies grew into mammoth structures during the Cold War, some adding analytic divisions to their operations forces while vastly expanding their collection capacities (spies, satellites), INR stuck to analysis. Interpreting data other people collect isn't terribly glamorous, but INR developed a reputation in spy circles for its accuracy and judgment.

In the 1950s, the bureau's assessments of Soviet military capability were more accurate than the often-inflated opinions coming from the CIA and the Pentagon, says John Prados, an intelligence expert and historian at the National Security Archives. "State Department intelligence," he writes, "helped to hold projections made by the military services within some sort of bounds." And during the 1960s, INR held fast to its belief--later proven true--that the Russians were not testing missiles with multiple, independently-targetable nuclear warheads, despite the Pentagon's insistence to the contrary, and the CIA's willingness to bow to that hawkish view. During the Vietnam War, INR analysts looked at the military's own data on the number of enemy prisoners captured, enemy attacks, and weapons confiscated, and concluded that the war was not going as well as defense officials were claiming. It was a heady moment for the civilian analysts at INR, taking on the military establishment on a military topic using the military's own data. The Pentagon was outraged, says Prados, but again INR turned out to be right.

Today, while the CIA operates out of a gleaming campus on the Potomac, and the NSA inhabits a similarly futuristic enclave in Fort Meade, Md., INR has a set of grad-student-like digs in Foggy Bottom. The nondescript offices in the State Department's headquarters sport repainted ceiling tiles that buckle at the seams, linoleum floors, and narrow hallways.

I went there recently to interview the agency's head, Tom Fingar, a tall man with hair more salt than pepper, round glasses, and a pensive demeanor. The office, spacious and wood-paneled as a dean's redoubt, matches the personality of its occupant, a former academic and China scholar. Today, INR is still small and obscure. At $50 million, its budget is "decimal dust," Fingar said, when compared to the estimated $40 billion spent by the rest of the intelligence community. Experts estimate the CIA has between 4,000 and 7,500 analysts, and are under presidential orders to increase that number by 50 percent; INR has approximately 160 analysts with no plans to grow.

Knowing the mad scramble for billions now underway as government agencies compete for funds that Congress can't seem to give away fast enough, I pressed Fingar about whether his agency could use more money. The director considered the question before answering. "Does it matter for our people to ride coach instead of first class? I don't think so. Would I like more? Yeah, a little more. I don't know what I'd do with more than a 10 percent increase," he concluded. "I could use more flexible money to let someone travel to get their language skills back. But we're in good shape."

"It's weird," former CIA officer Robert Baer told me a couple of days later, after praising INR's work. "But not having money in Washington seems to help."

My buddy Vladimir Putin

Funding might or might not have anything to do with INR's success. But when I asked dozens of intelligence operators worldwide why the office has been so successful, I got several consistent answers.

The first reason seems to be that INR analysts simply have more years on the job, focused on the same country or region, than do their counterparts elsewhere. The average tenure of an INR analyst is 15 years, whereas at the CIA, where analysts typically change beats as a way of advancing their careers, the average tenure is more like five years, according to sources. "The difference," former INR director Carl W. Ford, who previously spent years at the CIA, explained to me, "is having one experienced person looking at material versus 25 inexperienced people."

Second, INR gets a different kind of analyst. The CIA, under the gun to staff up mightily after its ranks were thinned by budget cuts in the 1970s and 1990s, tends to recruit kids right out of college and train them in their new "specialties." (All new CIA hires must be under 35 years of age, although that requirement is occasionally waived.) And while the CIA's young analysts occasionally travel to their countries of responsibility and bone up by reading at their desk, they have little first-hand experience of their regions. INR couldn't be more different. Among the civil servants who make up two-thirds of its staff are many scholars lured out of the academy who come with years of knowledge. Fingar is one of them: He spent a decade-and-a-half as a scholar at Stanford's U.S.-China relations program, speaks fluent Mandarin, and has traveled widely in China. The other third of INR'S staff are Foreign Service officers rotating through who usually have spent several diplomatic tours in the country or region they are focusing on at INR, and who thus have both a reservoir of knowledge about its personalities and history, and a deep well of personal contacts.

Typical is John Evans, who was a senior analyst at INR's Russia desk before taking his current post as ambassador to Armenia. During the early 1990s, Evans worked at the U.S. consulate in Leningrad, where he spoke on occasion with the city's mayor, Anatoly Sobchak. His usual contact, however, was Sobchak's deputy, Vladimir Putin. "Imagine the CIA," says Greg Thielmann, a former strategic director of INR, "with a whole stable of bright young college grads in their twenties looking at all the evidence about Putin--and, over at State, you've got a guy who's worked regularly with him."

The third reason INR tends to get it right is that because many of its analysts have been in the field as Foreign Service officers, they have a better sense of how intelligence is collected. Consequently, they know when it should be looked at critically in a way that CIA analysts--few of whom come over from CIA field jobs--don't. In most embassies, Foreign Service officers work down the hall from CIA case officers and in many cases do similar kinds of work: collecting information, sometimes from the same sources. "The difference between the State Department and the CIA is that the CIA pays for information, and State doesn't," explained former CIA officer Baer, and the reliability of paid-for information is often questionable. Moreover, to protect the identity of sources, CIA field reports are typically written with little explanation of where or from whom information came. INR analysts, who from experience have learned to be suspicious of some sources--for example, exiles--tend to give the CIA's more selectively sourced intelligence less credence. In general, Thielmann said, he was more skeptical of human intelligence that came from the CIA, and he thought that was true of other Foreign Service officers he worked with at INR. Although, Thielmann noted, everyone at INR was pretty skeptical. "I was surrounded by civil service analysts with the same skepticism, which came from watching bad reports out of Iraq and elsewhere. We took a jaundiced eye towards this stuff."

The fourth and most important reason for INR'S success is that the agency has a culture that tolerates dissent. The lasting criticism of the CIA that the 9/11 Commission produced was of the Agency's tendency to shoehorn evidence to fit the results that the higher-ups desired. The commission criticized the CIA for having a culture of compliance, not disagreement. If anything, INR displays the opposite impulse: One former INR official told me that the office is a "small band of curmudgeons."

There's no better example than the case of the aluminum tubes. When American intelligence operatives seized a shipment of these tubes headed for Iraq in late 2002, one CIA analyst with some professional experience of working with centrifuges concluded that the tubes were destined for use in uranium enriching facilities. Most of the rest of the intelligence community, working in an environment in which the president and his top appointees were demanding evidence of Iraq's continuing WMD programs, signed on with this view. But analysts at the Department of Energy's intelligence shop, whose expertise on these technical matters dwarfed that of the CIA, were skeptical, arguing that the tubes were far more likely intended for use in building conventional rockets since the Iraqis had used the same kinds of tubes to that purpose in the past. Though this wasn't the politically popular answer, INR agreed to DOE's analysis, the only other agency to dispute the CIA's line.

It is hard to overstate the importance of the aluminum tube issue. It was supposedly the smoking gun, the piece of the administration's case for war against Iraq that made all the other pieces fall together. "Remember," David Kay, the chief American arms inspector after the war, told The New York Times, "the tubes were the only piece of physical evidence about the Iraqi weapons programs that they had." In his speech to the United Nations making the case for war with Iraq, Secretary of State Powell split the difference, acknowledging that there was a debate over the tubes, but arguing strongly that they had been intended for uranium enrichment, a view his own analytic shop thought wrong. Other senior administration officials were even less circumspect.

A culture of dissent must be nurtured and protected if it is to thrive--and State has usually given INR the requisite political insulation. After Powell mentioned the CIA's conclusions about the aluminum tubes in his speech at the United Nations, retired INR analyst Greg Thielmann criticized Powell publicly, telling "60 Minutes II" that the speech was "probably one of the low points in [Powell's] long, distinguished service to the nation." But far from shunning Thielmann, the State Department kept him in its good graces: Last September, when Fingar followed Powell into State's grand, eighth-floor Benjamin Franklin Room to be sworn in as assistant secretary of State, one of the people invited to watch was Thielmann. For analysts, the symbolic embrace of Thielmann affirmed something they'd always been told INR believed: Productive dissent was encouraged. That's unique in intelligence circles. Every ex-Langley hand I interviewed agreed with former CIA operations officer Melissa Boyle Mayle, who said her agency would never be as forgiving towards one of its internal critics. "If you're out [at the CIA], you're out," she told me. "You're one of 'them.'"

Part of the reason that INR enjoys this level of protection is institutional. The State Department's staffers, by virtue of their responsibilities, simply must be open to the points of view of the countries with which they conduct daily diplomacy. There is an inherent tension within the State Department between cooperating with the White House and coordinating with the rest of the world. That tension creates a market within the department for objective analysis, a need for an honest broker. "This building understands we're useless if we're not objective," Fingar told me. "If you want an echo, close your door. Or sing in the shower."

Another part of the reason for INR's insulation has to do with the kind of people who have led it. INR analysts give former INR director Ford, for instance, tremendous credit for shielding the bureau from political pressure. "Carl earned the respect of people in the bureau for standing up for the work of the bureau," Thielmann told me. "He even stood up to the CIA."

Even secretaries of state have understood how valuable INR can be when it is protected. When the Bush administration asked Powell to deliver its case for war to the United Nations in January 2003, it turns out it already had prepared the speech for Powell to give. The CIA had been working on it since December 2002. Before he did, Powell secretly passed the text to INR analysts for a thorough going-over, cutting the more egregious claims in the speech. Keeping the transaction private protected INR.

In the company of wolves

Not everyone in Washington is a fan of INR. Many neoconservatives especially see the agency as a threat to the more vigorous military project they advocate. In The Washington Times in August 2003, former Reagan White House official Frank Gaffney Jr. lamented the purported bias of INR's career civil servant experts. "This bureau's intelligence products have tended to reflect the policy predilections of State's permanent bureaucracy, rather than the facts." But there's a simple bottom-line test for intelligence: Who called it right most often? And on the big questions, INR has consistently gotten right what other agencies have gotten wrong.

That's certainly the conclusion the 9/11 Commission came to. While the commission recommended that the CIA, DIA, and most other intelligence agencies be gathered together under the office of a new intelligence czar, it also urged that INR be kept away from the czar's jurisdiction because its independent analysis had worked so well. Congress agreed. When it passed legislation in December creating the Office of National Intelligence, INR was one of the few intelligence agencies left out of the reorganization.

Congress also concurred on another commission recommendation: that the new czar should try to foster in all intel agencies the same culture of dissent that thrives at INR. The new law specifies a checklist of procedures the czar must implement to encourage openness and criticism, from mandatory reviews to see if analytic judgements match the underlying intelligence data, to a government-wide ombudsman to handle any complaints by analysts of political pressure from higher-ups.

But in the end, the intelligence czar answers to the president, and will promote a culture of questioning only to the extent he is ordered to, and thus far, President Bush has not been known for his encouragement of dissenting views. Similarly, INR can continue to thrive as Washington's most prescient intel shop only if its independence is promoted and defended by the secretary of state. Already, rumors are rampant at Foggy Bottom that Condoleezza Rice, the president's choice to replace Powell, has orders to enforce more policy and message discipline at State, and plans to oust dissidents, including top officials at INR. On the eve of Bush's second term, then, the big question is this: Will the rest of the intelligence community end up looking like INR, or will INR end up looking like the rest of the intelligence community?

Justin Rood is a staff writer at Congressional Quarterly.