Thursday, July 15, 2004

The gang that couldn’t think straight

Now the Senate's report on the CIA's intelligence failures is out, the pundits are all commenting on it. I haven't read the report but there are several points everyone commenting seem to agree on. One of these is that we didn't have enough “good” HUMINT [human intelligence, which one commentator equates with spies which is incorrect]; that there was a “human intelligence failure” with respect to Saddam’s WMDs. For example, the Wall St Journal notes that:

“One very alarming explanation, says the report, is that the CIA had "no [human] sources collecting against weapons of mass destruction in Iraq after 1998." That's right. Not one source.” [ link: ]

The senate report goes on to say [as quoted by the WSJ] that when asked about this, a CIA officer explained that it was because such a source was “very hard to sustain" [to put it mildly]. Let’s set aside the CIA guy's comment for a moment, and reflect on the rest statement. It is the kind of thing politicians excel at; a technically correct statement that is — and this is the only semi-polite response possible — a load of crap. Here’s why:

First, the senate report states that a “global intelligence failure” occurred. Their conclusion would then seem to imply that the lack of a single CIA source somehow “explains” this “failure”. The report does not allege that no one had sources collecting against WMD in Iraq after 1998. What about these other intelligence agencies and their sources, and their conclusions that matched ours, and that ours we to some extent derived form? Apparently they don’t matter. If the CIA used British or Israeli or Jordanian sources after 1998, it’s still all their fault for not having their own. Or something… But at least the senate report’s statement remains factually true.

And what about Hans Blix? He certainly was in Iraq after 1998, he was certainly collecting against WMDs, he certainly found Saddam in “material breach” on the WMD issue, and he certainly did not find any evidence that the CIA’s [and everyone else’s] prior intelligence assessment was wrong. Blix didn’t work for the CIA so the statement above is still factually true, but so what? There certainly was HUMINT collected from Iraq on WMDs after 1998, probably from multiple sources, and it was significant. The conclusions in the senate report regarding HUMINT [as stated] reflect either criminal malfeasance, near-criminal stupidity or both. I suspect both.

Let me set one thing straight: HUMINT is not just "spies" -- it is ANY human derived intelligence. Interviews with defectors are HUMINT; reports by diplomatic personnel are HUMINT, the reports of the UN inspectors are HUMINT. But the senate report clearly means spies -- covert sources inside Iraq -- and that is clearly what the CIA guy was responding to. So what is the big deal about spies?

Let's play the spy game for a moment, shall we?

First, there's the rules -- or not. In US-USSR spy game there were rules, and sometimes they were followed. One of the rules concerned exchanging spies -- it sometimes happened. Sometimes the suspected spy was wired to a board and then fed alive, feet first, into the furnace of a crematorium. That reportedly was the fate of Lt. Col. Oleg Penkovsky [though officially he was said to have been shot], who spied for the CIA while inside Soviet military intelligence [GRU] and provided vital info to us during the Cuban Missile Crisis. But there were still rules, or the semblance of them, because the US and the USSR could exert leverage over each other; we each wanted things that made it sometimes in our interest not to hang [us] or incinerate [them] captured spies.

In Saddam's Iraq, there were no rules because there was no similar leverage and what happened to suspected spies made what happened to Col. Penkovsky seem downright cozy. Consider what Saddam did to poets and those suspected of speaking ill of him and you might get some idea, but I rather hope not. Consider the affect of this knowledge on anyone considered spying against him for us. Now let's get on with the game...

OK, so we want spies working inside in Iraq on WMD -- who do we want?

Well, weapons designers would be nice. They could tell us what they were building and how far along they were. But they wouldn't necessarily know where, or if, existing WMDs were deployed or when [or if] they were to be used. It would be better to have some Iraqi commanders for that, colonels and generals preferably. We also need to consider that our spies might be mistaken, confused, or they might be plants. we can't just tale their word of things; we should have corroborating evidence. To get this, access to suspected WMD facility would be good.

How many people could we expect to recruit for this? Given the risks, a small number: 5 or 10 well-placed sources to be generous. Let's be very generous and say 5 or 10 of each. Except when it comes to facility access, lets say 20 or 30 -- after all, this could be janitors, service and repair guys, the falafel man. Now let's think about how much info these spies could produce before they were caught; how many facilities they could search and how well...

But ... Uh-oh! I think we have a problem with our game! Didn't the UN inspectors interview dozens [not 5 or 10] of Saddam's weapon scientists? Didn't quite a number colonels and generals attest to the existence of deployed WMDs after the war? Didn't the UN search suspected Iraqi WMD facilities -- not 1 or 2, but dozens?-- for several years? Isn't it a fact that the pre-war WMD intelligence estimates were largely based on the inspector's findings, both before and after 1998?

Damn right they were! On HUMINT: military commanders, scientists, facilities searches. Exactly the kind of well-placed sources we sought for our fantasy spy game.

What else were they based on? Friendly foreign intelligence reporting -- this was a "global failure" remember? But what was that reporting largely based on? Oh yeah – HUMINT; their spies and well-placed sources.

So our prewar WMD assessment was about 90% based on what? HUMINT. And if it was all wrong, as so many believe, then all the highly placed spies in our fantasy game wouldn't have been any more right, because all the problems we associate with our real HUMINT sources also apply to our fantasy covert sources: they can be mistaken, they can be confused, they might be less than candid. So is there anything our fantasy spies might have been able to do that our current and historical HUMINT sources don’t do?

Well, suppose the pre-war WMD estimate is substantially correct and Saddam destroyed, sold off, concealed and or moved his stockpiles out of the country sometime before the war. After all — and this point seems to be routinely ignored — the CIA never said there would be WMDs in Iraq after the war; they said Saddam had them before the war, but lets put a time limit on that: say after 2001 [I suspect that’s about right though I haven’t seen the information cut-off date on the NIE published anywhere.] If we had some highly placed spies, maybe they could tell us what happened to those WMDs — where they were moved to, or concealed, or sold, or whatever.

But, wait a minute! We do have HUMINT sources claiming exactly that. We have claims that substantial WMD components, including weapons, were moved to Syria and concealed; we have people claiming to know exactly where. It’s just that no one seems to consider these claims credible. But what would make a covert source saying the same thing [or the opposite thing] more credible, just because he or she is covert? If the actual claimants are wrong or are trying to sell us a bill of goods for their own purposes, a covert source might be just the same.

The point is that HUMINT is HUMINT is HUMINT and it stands or falls on it’s merit, not because it is obtained covertly or not. We are swimming in HUMINT on Saddam’s WMDs and just because we achieve a covert penetration of Saddam’s government or military doesn’t by itself make the HUMINT obtained better — in fact, the preceding comparisons show the opposite: it is liable to be less complete, more sketchy, and a whole lot riskier to get than what we did obtain.

The “problem” with the HUMINT that was the basis for the CIA’s WMD assessment isn’t that it has been proven wrong; the problem is that many people don’t believe it. They don’t believe it because — and this is the real “problem”; the real “failure” — it fails to support the conclusions that they [in this case, the politicians] want to reach: that there were no WMDs; that there was an “intelligence failure” that they can exploit for political, ideological, and personal reasons.

So what are we to conclude? Mainly that the Senators don’t understand intelligence, and that they sure as hell don’t understand HUMINT. They do understand how to write polemics, how to get their names in the paper, how to dodge responsibility and how to deflect blame from themselves, and how to dishonestly exploit the intelligence community for venal purposes. They understand those things very well.

But the main thing they don’t understand is how their self-serving actions endanger Americans and innocent lives everywhere.

The Real Intelligence failure

What is the irony of the Senate’s report castigating the CIA for a “global intelligence failure”? It is that the conclusion that there was a “global intelligence failure” is based on a real, and singular intelligence failure: the work of Iraq Survey Group [ISG] run by David Kay.

The ISG investigation Dr. Kay ran is the “proof”, the sole pivot on which the “no WMDs in Iraq” argument rests. It was Dr. Kay’s interim report, distorted by the press, that created the “no WMDs” canard, fueled by Katy’s foolish statement that he believed that no WMD stockpiles [he didn’t say what a “stockpile” was] in fact existed. It now seems almost everyone, on whichever side of the war debate, believes it. But the startling fact — though it should not be — is that the failings the senators accuse the CIA of in reaching its pre-war WMD assessment were actually committed by the ISG during the botched post-war investigation.

According to an article by Frank J. Gaffney Jr., “Failures”, in The Washington Times []:
“The senators attributed these [the CIA’s pre-war] failings to "a combination of systemic weaknesses, primarily in analytic trade craft, compounded by a lack of information-sharing, poor management and inadequate intelligence collection.”
While I shall demonstrate in a future that to the extent these failings did [and do] apply to the CIA, they were imposed from outside: by the Clinton administration and by congress; they all apply to the ISG investigation to such a degree as to render its conclusion that no WMD stockpiles existed baseless and without merit. [To clarify: that is not to say that intelligence failures of the ISG prove they did exist, only that the ISG conclusions are useless in arguing that they didn’t.]

The ISG investigation was ill-conceived, poorly planned, and incompetently conducted. Other authors have described this in some detail [see here, here, and here]; I will merely touch on the salient points:

The ISG had no sound methodology established when they went in Iraq. They didn’t have the time or the resources for a systematic search of suspected or possible weapon sites; nor could they translate and analyze all the thousands of Iraqi documents they had access to. So they decided to interview Iraqi scientists in hopes of getting quick results. For whatever reason — naivete, ignorance, incompetence, arrogance; it hardly matters — the ISG, and Dr Kay, decided the Iraqis they were talking to were telling the truth, even when their statements did not jibe with what they’d told other investigators. The ISG should have been — must have been? — aware of this; it should have been a red flag. But to the extent they noticed these discrepancies, they appear to have assumed they were getting the truth and the other guys were not. This was a colossal blunder, especially as it has been documented that the ISG got less information in many from these scientists then UNSCOM did. ISG interviews were neither as long nor as detailed as UNSCOM interviews, and the interviewers were not as expert. If this is the way the ISG worked, I’m surprised if they don’t now own a lot of Florida swamp land.
The ISG did not have the proper skills for the task. The ISG didn’t correctly understand the scope of the task, and reportedly sent managers to Iraq instead of their analysts, who knew much more about Iraq's weapons. So there was, in the senate’s words, a basic failure in analytic trade craft.

There was mismanagement and a lack of effective cooperation in the ISG effort. The ISG refused to get help or include the real experts on WMDs in Iraq; the UNSCOM inspectors. The ISG spent far less time on the WMD issue than UNSCOM; its personnel were far less knowledgeable. It was reported that David Kay tried to include UNSCOM people in the ISG, but he wasn't that successful. This was a critical of failure of David Kay’s leadership that compromised his whole mission.

There has been a lack of effective review or oversight of the ISG findings. UNSCOM’s work was reviewed and vetted by a host of experts; so far ISG has not opened its findings to outside vetting. ISG has not appropriately shared data or explained why or how it’s conclusions can fly in the face of years of work by vastly better qualified inspectors, to say nothing of the considered assessment of the entire intelligence community.
Of course, the ISG was made up of elements of that same intelligence community. The CIA was involved with ISG, but not just the CIA certainly, and David Kay, who was from the CIA, must bear the brunt of the responsibility for the overall failure of the ISG. Certainly, the ISG did get some things right and did make positive contributions [mostly ignored or buried by the press], and the ISG did support important conclusions of the pre-war assessment. The ISG also supported the assessment that important WMD components, if not whole stock piles, were shipped to Syria.

But overall, the ISG investigation is an excellent example of the kind of intelligence failure that the CIA and the intelligence community as a whole has been and is being persecuted for. However, the ISG investigation was an isolated incident, involving a host of US, British, and Australian organizations that came together in a hasty and ultimately dysfunctional way. It was not typical of the pre-war intelligence situation, or of the CIA, nor of the US intelligence community as a whole.

So the central myth of our time — the global pre-war assessment of Saddam’s WMDs was based on a “global intelligence failure” — is in fact the product of a genuine multinational intelligence failure. The ISG’s intelligence failure is accepted only because it suits the political or ideological aims, or the malice, of the Bush administration’s critics. If allowed to grow unchecked, this myth could spawn a catastrophe that will dwarf 9/11 in its proportions.