The company they keep.

A few years ago, several senior officials of the Central Intelligence Agency gathered in a conference room at Langley to ponder the worst situation an intelligence service can confront. One by one, the Company's agents in the Soviet bloc--spies code named Tickle, Blizzard, Gentile, and Pyrrhic--had been uncovered and apprehended by the Communists. Those picked up included the London station chief of the K.G.B., a Red Army general, and most of the C.I.A.'s top Soviet agents of the 1980s. HUMINT (spy talk for human intelligence) operations in Russia, the primary target of the C.I.A., were in ruins. And no one knew why. As the top intellicrats of the U.S. government scratched their heads, one jokingly remarked. "Well, someone in this room must be a mole." Everyone laughed at the preposterous notion--including a C.I.A. veteran named Aldrich Hazen Ames.

The arrest of Rick Ames, chief of the Soviet counterintelligence branch, provoked outrage from national security hawks and derision from C.I.A. critics. Both groups had the same question: How the hell could the Agency have not caught on to Ames' espionage when he was driving a Jaguar, buying a fancy house with $540,000 in cash, hacking into C.I.A. computers, lying to his superiors about his overseas travels, and having trouble passing lie detector tests? Part of the answer--one that nobody has paid attention to--is that the C.I.A. is far too much of a private club, one in which its members take care of each other and pledge allegiance to their own community. This clubbiness protected Ames who, as the son of a C.I.A. officer, was a legacy. In addition to posing a security problem, the clique mentality that protected Ames for so long also prevents the Agency and its outsider overseers from dealing with the shortcomings, large and small, of the C.I.A.

All government bureaucracies perpetuate a certain exclusivity. But shrouded in secrecy, the C.I.A.--like other spy services--is culturally more insular than most agencies. Such secrecy is bound to have an effect: It draws members of the club closer together and further distances them from the civilian, non-secret world. (In the Ames case, the outside world included the spy-hunters of the F.B.I., from whom the C.I.A. withheld information regarding Ames' worrisome encounters with a lie detector.) Two directors of Central Intelligence have noted (in their memoirs) the deleterious effect of such clandestine bonding. William Colby, director from 1973 to 1976, observed that many C.I.A. people dropped out of non-Agency society and immersed themselves exclusively in the cloak-and-dagger life." They formed "a real fraternity.... They increasingly separated themselves from the ordinary world and beveloped a rather skewed view of that world.... And out of that grew...an inbred, distorted, elitist view of intelligence that held it to be above the normal processes of society, with its own rationale and justification, beyond the restraints of the Constitution, which applied to everything and everyone else."

Stansfield Turner, President Carter's much maligned C.I.A. director, mused on the impact of a covert life: "Hiding your accomplishments, leading a double life, regularly facing moral issues...can all take their toll. In many ways, a clandestine career can be said to deform the person involved." And the institution itself.

In the course of writing a book on one longtime, highly decorated, and highly controversial C.I.A. officer, Theodore Shackley, I have interviewed more than 100 former Agency employees. I found many to be intelligent and thoughtful, as well as candid about the failings of the Agency. But what is striking is the number of stories I heard in which one or more Agency employees realized that something was wrong with Agency operations but did nothing about it. It comes as no surprise that a bureaucracy--and as a bureaucracy the C.I.A. probably has more in common with the U.S. Department of Agriculture than it does not--is populated with people who adhere to a get-along philosophy. Yet in the C.I.A. the natural bureaucratic impulse to protect the institution is compounded by the bond of secrecy. And a culture is spawned that shields the Agency from F.B.I. investigators, congressional busybodies (who are supposed to watch over the intelligence community), and citizens who seek assurances that behind the veil nothing too untoward is being done in their name and with their tax dollars.

The more closed a community, the more difficult it is for its members to pursue allegations of wrongdoing and to speak out. One good example of this principle is a minor episode that occurred early in Shackley's career. In the mid-1950s, Shackley, who as a young officer had impressed his C.I.A. superiors, was posted to Berlin, the most prestigious overseas assignment available at the time. William Harvey, a legendary officer, ran the base where hundreds of Agency employees mounted operations to recruit spies behind the Iron Curtain. Shackley was in charge of a group of case officers who targeted Poland and Czechoslovakia. Their successes were few. Most of the C.I.A.'s Soviet bloc espionage work in the 1950s amounted to little more than dubious, doubled, or dead agents.

Shackley, though, managed to find a Polish source who provided a steady stream of information--nothing grand, yet useful nonetheless. But two fellow C.I.A. officers who had been transferred from Berlin complained to a base chief in another German city that Shackley was overplaying the operation, that the agent had not always said what Shackley reported. "I stewed about this for a long time," the base chief recalled in an interview with me. "A snowstorm of reports came out of this agent, and these guys said Shackley was making them up.... I can't think of any higher crime for an intelligence officer. If you can't have integrity, you may as well not have an intelligence agency."

But what to do? The base chief realized that if he raised a flap he would be in a losing game. Shackley would deny he had rigged the intelligence to make himself look good. Harvey could be expected to scream and remind all that he had booted the two accusers from the Berlin post. The base chief, who did not speak Polish, was in no position to conduct an independent evaluation. He considered bumping the whole problem to his superiors in Washington. But, he figured, they would only face the same no-win situation. The base chief let it go. He knew that an intelligence service attracts ambitious people, trains them to be duplicitous, assigns them tasks that are not amenable to intrusive oversight, and so must rely on their personal integrity. To do anything to question the integrity of an officer was to question the most crucial aspect of the system, and he was not willing to do that. There the matter ended. Had the charges been true--and they may have been--Shackley's career should have ended. Instead, he went on to positions of great influence where he was responsible for intelligence in some of the most significant hotspots of the Cold War--Miami, Laos, and Vietnam.

Decades later, Turner discovered that there was still much internal reluctance to coming down too hard on a fellow member of the club. Five weeks after he took over the C.I.A. in 1977, he was flabbergasted to read in The Washington Post that Edwin Wilson, a C.I.A. official who had become a rogue arms dealer with ties to Libyan strongman Muammar Qaddafi, still had active connections with Agency officials. Two mid-level officers had aided Wilson's business deals, and a pair of senior officers (including Shackley, then the associate deputy director for operations) had regularly socialized with him. Turner demanded that the inspector general's office, the Agency's internal watchdog unit, investigate.

Turner was more shocked to learn that months before, during the tenure of C.I.A. chief George Bush, the I.G. had examined this and learned of Wilson's relationship with the C.I.A. men. No one previously had alerted Turner to this; no one had done anything; no one even seemed concerned. Turner whistled into his office the top officials of the Agency: deputy director E. Henry Knoche, deputy director for operations William Wells, I.G. John Waller, Shackley, and others. Gentlemen, Turner said, the question is what to do about the two C.I.A. employees who had worked with Wilson. The choices were exoneration, punishment, or dismissal. All the career Agency officials suggested a modest punishment, arguing that dismissal would demoralize the Agency. Only Robert "Rusty" Williams, a C.I.A. outsider brought in by Turner to be his executive assistant, favored canning the pair. Turner stared at his senior aides--he could not believe their attitude--and said sarcastically, "Majority wins. They're fired." (Turner did not officially punish Shackley for his contacts with Wilson, but he eventually transferred him to a less influential post.) It was an eye-opening moment for Turner. These fellows, he thought, are too damned protective of each other.

Spy Anxiety

In researching a small but significant slice of C.I.A. history, I came across several instances of officers covering up for other members of the fraternity. When a base chief in Vietnam was caught fabricating agents and padding his expenses, he was allowed to pay back the funds but was not turned over to the Justice Department. In 1973, Shackley, then chief of the Western Hemisphere Division, had to deal with a Senate subcommittee investigating a 1970 C.I.A. and I.T.T. plot to undermine Salvador Allende, the democratically elected Marxist president of Chile. When Shackley ordered a subordinate to arrange for an I.T.T. official to testify falsely--this I.T.T. official later pleaded guilty to lying to Congress--Shackley's deputy chief opposed the move but did nothing to stop it. In Laos in the mid-1960s, Shackley, then chief of station there, engineered a military engagement that was a complete disaster: Over 2,000 Laotian soldiers were lost in a battle against the North Vietnamese Army and the Communist Pathet Lao forces. Afterward, he ported to Washington that the fault lay with the Laotian commanders--not the plan he had supported. But one of the C.I.A.'s best Laotian hands wrote his own report noting that Shackley had pushed on the Laotian military a plan that knowledgeable people in Laos had predicted would fail. Instead of sending his report to Langley, this officer put it in his safe and then later burned it. He felt there was no point in snitching on Shackley, who was rising fast through the ranks. No one in Washington, he believed, would bother to listen.

The covert world's clannishness allows its members not only a little slack. It also keeps one of the club's biggest secrets under wraps. Ames, though, shared this secret when he pleaded guilty to espionage and tax evasion in a federal district court outside Washington in April. After expressing regret for his treachery, he cited two factors that led him to betray his country. First, he had "come to dissent" from U. S. national security policy and the decades-long shift to the right in American politics. Second--and here comes the secret--Ames had concluded that "the espionage business, as carried out by the C.I.A. and a few other American agencies, was and is a self-serving sham, carried out by careerist bureaucrats who have managed to deceive several generations of American policy makers and the public about both the necessity and value of their work." Thousands of case officers and tens of thousands of agents around the world, Ames maintained, have been spinning their wheels. "The information our vast espionage network acquires at considerable human and ethical costs," he charged, "is generally insignificant or irrelevant to our policy makers' needs. Our espionage establishment differs hardly at all from many other federal bureaucracies, having transformed itself into a self-serving interest group immeasurably aided by secrecy."

It is too bad that it took a traitor to reveal the most important secret held by the secret-keepers. Ames certainly has no love lost for the Agency, but his words have the ring of truth. After Ames' statement, Representative Dan Glickman, chairman of the House Intelligence Committee, noted that the mole's comments should not be ignored. Glickman and Senator Dennis DeConcini, chairman of the Senate Intelligence Committee, both contacted Ames' lawyer and asked if Ames would testify before their committees. (Those appearances have been put off until the F.B.I. and the C.I.A. finish debriefing Ames.) The lawmakers' request was greeted by hoots from the spy world. Richard Helms, a former C.I.A. director, denounced the prospect of "having a traitor on the Hill to vent his spleen.... I deplore it." Incumbent Director of Central Intelligence R. James Woolsey decried "the fact that some people would take Ames seriously as an authority on the C.I.A. and what its value is."

On the subject of the C.I.A.'s effectiveness and worth, many Agency veterans I interviewed said more or less the same thing as Ames did. Few put it in such harsh terms. But they acknowledged they had spent decades in the espionage netherworld and never accomplished much--and had not seen colleagues do any better. But you always had to look busy. So if you're a C.I.A. case officer based in Colombia and can't bag a Soviet official as a spy, then you have to come up with something else: perhaps penetrate the local communist party, bribe a journalist to reveal his or her sources, or recruit the capital's police chief. Most of this mattered little, but it gave the impression the C.I.A. was busy prosecuting the Cold War.

In key theaters, the C.I.A. compiled a poor performance record on espionage. In the 1950s, it flopped miserably in its efforts to recruit truly significant Soviet bloc spies. In the 1960s, it failed to infiltrate Fidel Castro's ruling class. (Some C.I.A. veterans claim that the espionage program run by its Miami station, which was headed by the seemingly ubiquitous Shackley, discovered evidence that the Soviets were installing missiles in Cuba in 1962, but the Agency's own historical records indicate that the telling information came from a routine interview with a Cuban refugee conducted at an Army processing center in Florida.)

Intelligence in Vietnam was a bust. The C.I.A. never penetrated the higher reaches of the enemy, nor did the Agency fully convey the weaknesses of the Saigon regime to Washington. (Its analysts in Langley did consistently and correctly predict that Lyndon Johnson's bombing campaign would fail to win the war--unless the bombing was greatly intensified.) In a 1991 interview with an oral historian, Richard Helms, discussing the C.I.A. in Vietnam, offered a stunning indictment: "We were dealing with a complicated cultural and ethnic problem which we never came to understand. In other words, it was our ignorance or innocence, if you will, which led us to misassess, not comprehend, and make a lot of wrong decisions." How reassuring: The C.I.A.--like the rest of the national security bureaucracy--bungled cluelessly during the most important and bloody conflict of the Cold War. And on the all-important targets of the Soviet Union and China, the Agency gathered few agents who made a great difference. Its espionage failed to uncover the true secret of the mighty Soviet empire: It was hollow.

It is easy to bash the Agency. Its advocates always offer up the familiar chestnut: Everyone knows our failures, no one knows our successes. But any evidence of significant success is quite hard to come by. And it's not difficult to find Agency veterans who offer no such evidence.

Standing before a judge and wearing prison garb, the fallen Ames observed, "Now that the Cold War is over and the Communist tyrannies are largely done for, our country still awaits a real national debate on the means and ends--and costs--of our national security policies.... To the extent that public discussions of my case can move from government-inspired hypocrisy and hysteria to help even indirectly to fuel such a debate, I welcome and support it."

He is right: A public discussion of how and why the United States spies is desperately needed. Now if the defenders of the clandestine status quo were smart, they actually would be happy to see Ames testify before Congress on the perils of the clubby covert culture. Members of the spy set might even want to encourage that behind the scenes. For if they could link the call for evaluation and change to a despised turncoat whose acts of betrayal caused deaths, perhaps they could forestall a debate that, if conducted honestly and openly, the die-hard protectors of the C.I.A. and its budget would find tough to win.