Essays by Herb Meyer
Doing Intel Right
National Review Online — April 12, 2005
If size were a measure of insight, the WMD Commission's just-released 618-page report would be an even more valuable guide to understanding the cascading series of failures that have plagued our country's intelligence service than the 9/11 Commission's 567-page best-selling opus. But there is no correlation whatever between size and insight and — alas — the most striking feature of both reports is their focus on what is obvious and their silence about what lies beneath.
For example, the WMD Commission concludes that "We need an Intelligence Community that is truly integrated, far more imaginative and willing to run risks, open to a new generation of Americans, and receptive to new technologies." Well, yes — but this recommendation is equally valid for large corporations, universities, and the 2008 U.S. Olympic women's synchronized-swimming team.
The 9/11 Commission concluded — after nearly two years of contentious hearings and deep thinking — that the horrific attacks on New York and Washington weren't prevented by our country's $40 billion-a-year intelligence service because it had suffered a "systemic failure." No kidding — but most Americans had figured this out even before the south tower of the World Trade Center came crashing down.
While both reports provide a detailed narrative of what precisely went wrong, neither report explains why things went so wrong.
They Were Incompetent
The root cause of the intelligence community's failures during the last few years is that the people in charge were incompetent. Yes, it really is this simple.
And it's devastating, because it means that blame for our country's intelligence failures runs from the White House — whose occupants chose the leaders of our intelligence service — to the House and Senate Intelligence Committees, to the intelligence community's top-level officials themselves, who had more confidence than talent and who put all our lives at risk by taking jobs far beyond their abilities to perform. In other words, our entire political establishment screwed up. And that's the one thing members of blue-ribbon commissions — who invariably are themselves members-in-good-standing of the political establishment — can never bring themselves to say.
And that's too bad, because our lives depend on assuring that we have the best possible intelligence.
"Doing Intel" Right
Putting aside covert operations — a different subject for another day — an intelligence service does two things: It gathers information, and then it tries to figure out what that information means. In other words, it collects dots and then connects the dots into patterns that illuminate the future soon enough, and clearly enough, for the president to change the future before it happens. When we have an intelligence failure, it's because one of three things went wrong: The collectors didn't gather up the dots, the dots were collected but never reached the analysts, or the analysts were provided with the dots but then failed to see the pattern those dots formed. So every failure is either a failure of collection, a failure of organization, or a failure of analysis — or a combination thereof.
It's clear from both the WMD Commission and the 9/11 Commission reports that our intelligence service blew it across the board. But it's also very clear from both reports that at their core the two truly catastrophic failures — the CIA's inaccurate National Intelligence Estimate about Iraq's WMD program and the entire intelligence community's failure to grasp that al Qaeda had for years been planning to hijack airplanes and fly them into buildings — were analytic failures compounded by organizational problems that kept information the collectors had gathered from reaching the analysts.
More precisely, it's clear from both reports that in each of these failures three specific mistakes were made: Dissenting views weren't taken into account, "group-think" kept the leaders of our intelligence service from realizing that they were veering way off track, and "co-ordination" among the dozen or so agencies that comprise our intelligence service — between the collectors and the analysts — was awful.
Nobody gets it right all the time, of course, but these were not the sort of mistakes that were made by previous directors of central intelligence such as Allen Dulles, John McCone, William J. Casey, and James Woolsey. That's because these men knew how to do their jobs, and how to set things up to keep such mistakes from happening.
How Bill Casey Did It
During the Reagan administration, I had the privilege to work closely with Bill Casey. Because Casey had managed Reagan's 1980 campaign, the press often portrayed him as just another pol who's "reward" was the CIA. Well, not quite. During World War II Casey had headed secret operations for the Office of Strategic Services (OSS), which was far and away the most effective intelligence service in history. And he had kept up his interest in intelligence — and his contacts — through the post-war decades while making a fortune in law and venture capital and serving along the way in a range of senior U.S. government positions, including chairman of the Securities and Exchanges Commission. In short, Casey knew his stuff about both sides of the intelligence game — operations and analysis. We'll skip over the operations side and focus now on how he got the analysis pointing in the right direction.
Since the National Intelligence Estimates were the most important products of all — they are the judgments and projections that reflect the combined wisdom of all the dozen or so agencies that comprise the intelligence community — Casey poured a lot of energy into getting the estimates right.
During the Carter administration, the estimates were fairly bland because Carter's DCI, Stansfield Turner, wanted consensus among the intelligence agencies. Of course, you can always get consensus by writing judgments that are accurate but useless, such as "We believe the coming years in Latin America will be uncertain, and at times volatile." Casey went in the opposite direction: He wanted estimates that genuinely reflected the differing judgments within the intelligence community.
My job was to manage production of these estimates, and one day early on Casey issued these instructions: "Look, I want the president to see where we disagree and to understand what the argument is about. Sometimes these guys at the Defense Intelligence Agency have a very solid point but they don't say it very clearly. They're soldiers, not wordsmiths. When they come in with a dissent, you go meet with them. Understand what they're driving at, then write it yourself and clear it with them. Then let's put the dissents in italics in the estimates so that no one can miss them."
Casey was never bothered when an estimate read like a shouting match and revealed that the intelligence community was split wide open on an issue. And if the finished product didn't reflect his personal judgment, he could always send an "eyes-only" memo to the president and know it would be read.
In addition, we often bounced the drafts of our estimates off experts from outside the intelligence community. Over the decades Casey had accumulated an astonishing rolodex of contacts throughout the world in business, politics, government, and academia. I also knew a few people. When we had the draft of an estimate in hand, we would sort through our rolodexes and then arrange one-on-one meetings with whomever we thought was knowledgeable about this particular issue. Sometimes we would meet in Casey's office — and the analysts were always circling outside, looking worried that the expert would tear their work to shreds. More often — and the analysts never knew this — we would meet with an expert in the evening, at Casey's house or at mine. On a couple of occasions I went overseas to meet with an expert whose opinion we really wanted. We couldn't show the classified estimate draft to the expert, but we didn't need to. We just walked him or her through the draft. (And since these people weren't stupid, they understood precisely what we were doing.) We would come away from the conversation either confident that we were on solid ground or queasy enough to take the draft apart to see where we might be going wrong. Every so often, we wound up shredding the draft completely and starting all over again.
The "Fail-Safe" System
Even with all this, it's possible to go wrong because we humans have a tendency to hear what we want to hear and to filter out whatever doesn't suit us. So we had an internal "fail-safe" system to minimize the chances that we would be so much in love with our own judgments that no inconvenient fact could stop us. It was called the senior review panel (SRP) and it was composed of four savvy and immensely experienced individuals whom Casey had cajoled out of retirement to lend a hand. Among the four — membership in the SRP changed over time, of course — were one of the State Department's most revered career ambassadors, an Army general who had spent much of his active-duty service in military intelligence, and a former CIA officer who knew his way around the ultra-secret world of satellites and electronic intercepts.
Meetings with the SRP were among the most contentious, verbally violent meetings in which I've ever participated. The SRP members weren't building careers so they didn't care whom they offended, and they never let anyone forget that they were busting their chops solely to keep the intelligence community from screwing up. When they thought we had our heads up our rear ends, they said so. And they were usually right.
The SRP also weighed in on the estimate draft — saying, for instance, that our judgments weren't supported by the evidence, or that we were relying too much on one super-secret source whose credibility the SRP doubted.
Could we ignore the SRP's comments and refuse to make the changes they wanted? Yes, and sometimes we did. But there was a paper trail — their very highly-classified memos, to be precise — and we knew that if we screwed up and some blue-ribbon commission was established to investigate our failure, they would get those SRP memos and we would have a lot of explaining to do. In other words, Casey set up a system designed to make us all think twice about ignoring things we didn't want to hear.
No system is perfect, of course, and in the end it was the director of central intelligence who signed off on the finished product and bore responsibility for its contents. But Casey was a pro, and he knew how to build in a set of back-stop mechanisms to highlight dissent, check our judgments with outside experts, and keep us from succumbing to the "group-think" mentality that afflicts us all.
The other big failure cited by both the WMD and 9/11 Commissions involved the tricky business of "coordinating" among the various intelligence agencies — more precisely, between the collectors and the analysts.
If you've been a diligent reader of recent articles and commentary about the intelligence business, you probably have the impression that collectors collect what they can, then forward their reports, intercepts, and photos over to the analysts — who then read through all this stuff until their eyeballs bleed. This is a formula for disaster. If you don't have a clear grasp of what you're looking for, it's virtually impossible to discern a pattern from the overwhelming torrent of information that's washing over you. Just imagine standing in the middle of a Wal-Mart, realizing you've left your wife's shopping list on the kitchen table — and hoping that if you just walk up and down the aisles you'll somehow figure it out.
Here's how the intelligence business really works: You start with a hypothesis, which is a leap of imagination based on your expertise, your judgment, and your "gut-feel" about what you think is really going on. Then you figure out what you would expect to find if your hypothesis is correct, and you convey all this to the collectors so they can get to work.
The CIA's Biggest Failure
The greatest intelligence failure of the Cold War — which, oddly, no blue-ribbon commission has ever been created to investigate — was the CIA's failure to grasp that by the late 1970s the Soviet economy had stopped growing and had begun to implode. The reason this was the biggest failure — vastly more important than any miscalculation about Soviet missiles or nuclear subs — was that if the Soviet economy were in fact imploding, the Cold War itself was no longer stable because the Soviet Union wouldn't be able to sustain its decades-long, global struggle for supremacy. Either the Soviet Union would collapse quietly — rather like a business that goes bankrupt and shuts its doors — or it would become dangerously aggressive.
The CIA calculated that in the late 1970s and early 1980s the Soviet economy was growing at an annual rate of more than three percent. President Reagan didn't believe it, and neither did Casey. (In the early 1980s, our own economy was struggling to reach a 3 percent annual growth rate: If we couldn't do it, how could they?) But they needed to know for sure, so Casey made it a top CIA priority to take a hard look at the Soviet economy and to figure out what was really going on.
We started off by literally making a list of all the things we would expect to find if in fact the Soviet economy were in trouble: shortages of consumer goods, late payment of bills, slowdown of new non-military construction projects, and so forth.
The next step was to take the finished list we'd made to the collectors. Casey sent me to the CIA's operations directorate, the various branches of the Defense Intelligence Agency, the super-secret National Security Agency, and to a few other places we still don't talk about. At each stop I met with the collectors, explained our hypothesis, and gave them the list of indicators we'd compiled. These were more like seminars than meetings, because the more clearly collectors understand what you want, why you want it, and how you think, the better able they are to do their jobs. Collectors hate it when you ignore them and say, in effect, "Just go collect stuff and maybe what you collect is what I want to know about." You must bring them into your confidence and let them know where you are going — or hope to go.
Moreover, you must be sure the collectors know what to do with whatever information they may find. Each meeting ended with one last instruction from me: "If you find any of these indicators, don't just throw your report or your memo into the pile. Jump the system and send it directly to me — and if anybody squawks let me know and Mr. Casey will cover you. And if you find something that isn't on our list but you think may be relevant, send me that too or, if you prefer, just give me a call and I'll come back for a cup of coffee and we can go over it together." And when these rounds of meetings in Washington were completed, Casey sent me overseas to repeat this process with collectors at those among our allied intelligence services that kept a sharp eye on the Soviet Union.
"It Just Didn't Fit"
Literally within days, collectors were sending me astonishing reports. One of them told of a growing number of strikes — strikes! — at Soviet factories. I asked why this information hadn't surfaced before and the collector replied, "No one was interested. It just didn't fit." That report was on Casey's desk in five minutes, and on the president's desk later that same afternoon. Likewise with an equally astonishing report recounting an episode in which Soviet workers had stopped and surrounded a train that was carrying meat. Troops arrived and surrounded the workers, and the standoff had to be resolved by the Politburo itself — which decided to allow the workers to offload and take the meat rather than risk a shootout. Casey delivered this knockout report to the president in person.
Over time, we compiled enough information to be sure the Soviet economy was starting to implode. And that gave President Reagan the hard intelligence he needed to be confident that his strategy of pushing the Soviet Union to its breaking point would work. (The CIA bureaucracy dragged its feet all the way, and it took years to get the official estimates of Soviet economic growth revised downward. But the correct intelligence was far too important to remain hostage to a bureaucratic dogfight, so Casey jumped the system by allowing me to write a series of memos to him that outlined the Soviet economy's growing crisis, and he made sure those memos reached the president. By the time the CIA bureaucracy finally got it right, the Reagan policies were already in place.)
Had the CIA used this methodology in the years before 9/11, those attacks might well have been prevented. But the CIA never clearly formulated a hypothesis that al Qaeda was steadily gaining strength and sophistication, that attacks within the U.S. were increasingly likely, and that the terrorists liked to use bombs and airplanes. And it failed to develop a detailed list of indicators based on this hypothesis, to distribute the list to collectors throughout the various agencies, and to give them the name and telephone number of a specific individual to contact when they found something. So when an FBI agent in Minneapolis learned that a bunch of single men from the Mideast were paying cash for flying lessons and telling their instructors that they weren't interested in learning how to land, the agent had no specific official close enough to the top to send it to.
Of course, there are dangers in developing a hypothesis and telling the collectors what you want. They can try to please you by providing whatever bits and pieces they can find, even though these bits and pieces aren't an accurate reflection of the truth. Or, if the collectors cannot find the indicators you are seeking, you can cling to your beloved hypothesis and blame the collectors for their inability to find what you know is really there — rather than admit that your hypothesis is just plain wrong. It's always a judgment call, which is why it's so important that the individual in charge has not only brains, but intellectual integrity.
And sometimes, everything depends on a director who has all this, and also the courage to go with his instinct — that un-quantifiable gut feel for where the big payoff lies. One of Casey's great predecessors was John McCone, whom President Kennedy made DCI after the 1961 Bay of Pigs fiasco. McCone was an older version of Casey — savvy, street-smart, knowledgeable about the world and Washington, deeply religious, seriously rich. The two men liked each other a lot, and from time to time McCone — frail but razor-sharp — would stop by for a chat. One day after the two had finished their private conversation and Casey had left the building for a Cabinet meeting, McCone had some time to spare and he very kindly stopped to talk with me. Our conversation turned to the 1962 Cuban missile crisis, and McCone told me a story I'd never heard before.
Sending the U-2s
As the crisis built and as members of Congress were warning that Khrushchev was placing nuclear missiles in Cuba that could reach the U.S., the CIA bureaucracy insisted that there were, in fact, no Soviet missiles in Cuba. McCone ordered that a U-2 flight be sent over the island to find out, one way or the other. It would be a dangerous mission — obviously dangerous to the pilot, less obviously a potential causus belli if the spy plane got shot down and Congress demanded that the U.S. retaliate — and opposition to sending the U-2 was strong at the CIA and the Pentagon. McCone insisted, the U-2 flew, and when the film it took was developed and studied by the experts, there were no missiles. McCone ordered a second U-2 flight, and the opposition was even more ferocious. But he insisted, the mission was flown and — again, no missiles. I asked McCone what happened next. "Oh, my," he said. "I'm afraid I wasn't very popular in Washington that week, because I ordered a third flight." And when the film from that third flight showed no missiles, McCone ordered a fourth, and then a fifth. It was the sixth U-2 flight that found the missiles, and the rest is history.
McCone was grinning as he finished his story — he was enjoying himself enormously — so I asked the question he was so obviously waiting to be asked: "Sir, what made you keep sending U-2s?" There was a long pause, and it seemed that McCone was thinking back to something that had happened a long time ago and struggling to get it just right. When he finally spoke his voice was so low I had to lean forward in my chair to hear him. "You know, that's just the question President Kennedy asked me after it was all over. Mr. President, I told him, I kept sending U-2s because I knew — I just knew — those damn missiles were there."
Intelligence isn't org charts. It's people. Put the wrong people in charge and they will screw up no matter how perfectly our intelligence community is structured. Put the right people in charge and they will overcome whatever structural flaws there may be and get the job done. The future looks very promising. President Bush's choice for the new post of director of national intelligence is Ambassador John Negroponte, one of the most experienced and effective diplomats in American history. His deputy will be General Michael Hayden, who has been running the National Security Agency and who knows the collection business upside down, backwards and forwards.
The danger is that these two very talented officials won't have the time to actually "do Intel." Having been convinced by the two commissions that the recent intelligence failures were caused by structure rather than by people, official Washington is obsessed with restructuring the intelligence community. The 9/11 Commission recommended a sweeping reorganization, and most of these recommendations have already been accepted, but are only just now being implemented. The WMD Commission has made several dozen very sensible recommendations that carry the work of the 9/11 Commission forward by sharpening and refining the new organizational structure. After confirmation by the Senate, Ambassador Negroponte and General Hayden will be up to their elbows sorting it all out.
But if all their time and energy is drained away messing with budgets, fighting turf wars with the Pentagon and the now-downgraded CIA, testifying before the dozen or so congressional committees that have intelligence-oversight responsibilities, and coordinating the agencies that comprise the intelligence community, they won't have any time or energy left to do the one thing that will actually make a difference in the quality of the intelligence delivered to the president and his top advisers: to sit quietly in their offices, alone, with the draft of an estimate in one hand and a pencil in the other, weighing the estimate's evidence and conclusions, then matching what they are reading with their own judgment, experience, and instincts.
Their success will depend not on their administrative skills, but on their talent for "doing Intel" right. So will our lives.