More than six years since the terrorist attacks on 9-11 the intelligence community continues to employ a substandard analytical practice that virtually guarantees shoddy and inaccurate analysis. What am I talking about? An analyst within the CIA (or DIA or INR) who writes an article for the Presidential Daily Brief or other community wide daily intelligence brief is not currently required to coordinate with analysts outside of their organization. What’s so bad about that? The failure to coordinate and obtain the clearance of other analysts prevents policymakers from getting the best analysis and information available. Perhaps this helps explain the mess we encountered with the October 2002 National Intelligence Estimate on Iraq.
Sorry to sound like an old guy, but I need to explain what I mean in talking about “coordination”.
When I was an analyst I was required to coordinate any article I wrote for the National Intelligence Daily and the Presidential Daily Brief with my counterparts at the State Department’s Bureau of Intelligence and Research (INR) and the Defense Intelligence Agency (DIA). For example, I was the Honduran analyst.
If I wrote about the threat of Cuban backed terrorism from Nicaragua, I first had to share what I wrote with the analysts at CIA who worked on Nicaragua, Cuba, or terrorism. That meant I took my draft to three different offices (remember, this was before email). Why? My bosses wanted to make sure that the CIA spoke with one voice. They did not want Larry Johnson’s personal views being shared with the President. My supervisors demanded that the information in my intelligence articles was accurate and reflected everything we knew about the current state of intelligence. This part of the coordination process covered only inside the CIA.
Once we had an agreed upon CIA version, I was then required to send the draft to the Honduran analyst at INR and the analyst at DIA who covered Honduras. (Both women by the way.) S ometimes they drove me nuts. They did not agree with how I worded a paragraph or with a particular conclusion. I had a choice. Either I accepted their changes or we escalated the dispute to a branch chief. If the INR or DIA analyst was not satisfied with our proposed fixes they were allowed to write a “dissent”.
A dissent is shorthand for a different point of view. For example, I could say “Iraq is trying to buy uranium yellowcake from Niger”. While INR would write, “No, Iraq is not trying to buy uranium yellowcake from Niger and cannot process the yellowcake currently in its possession”. This ensures the policymaker will understand there may be a dispute about particular matters. If there is not dispute then they have a reasonable expectation that they are reading a consensus view of the intelligence community.
More often than not I accepted the changes proposed by my counterparts. Sometimes I did not like it. We got into screaming matches. Sometimes these gals thought of something I had not considered and helped dramatically improve the quality of the article and the analysis. Despite the rough and tumble and frustration inherent in this process, the end result was a piece of good analysis that reflected the collective judgment of the analysts who were the substantive experts on the topic at hand.
That is not the case today and has not been the case for at least 8 years. I still have not been able to determine who instituted this change–was it Woolsey, Deustch, or Tenet? Don’t know, but it was a damn stupid change. Analysts at CIA, DIA, and INR are now free to write articles that are disseminated throughout the intelligence community without having to coordinate with each other and get clearance on their pieces. And you wonder why we have intelligence failures? This is a contributing factor.
So what does it take to fix this? Not much. We do not need more bureaucracy. We do not have to spend more money. We do not need to hire more analysts. The Director of National Intelligence simply needs to tell the intelligence community to get off of its lazy ass and ensure that every article that is circulated outside of an intelligence agency–especially the PDB–should be fully coordinated and cleared by the relevant analysts of CIA, DIA, INR, and FBI. This one is simple. What is amazing is that it is not being done.
It seems you are pushing for a process that mandates cohesion and may tend to deter cherry picking since dissents are included in a final article / report making it to the president.
On the other hand, the widespread distribution seems to increase cross-awareness among the many sections, increasing the likelihood that important facts will percolate to the top. Is it manageably read by all those who get the copy?
The process you described with dissents also resembles an appeals court decision process.
Do you see any advantages to analysts' articles being "disseminated throughout the intelligence community"? Their analyses are not tantamount to raw, collected intelligence facts, true?
The example of the Yellow Cake uranium rests on which facts make it to analysts, not just which analysts' views make it into the final intelligence product. Does the distinction matter in your view?
October 31, 2007 12:40 AM | Reply | Permalink
There's no simple answer to this. I'm still struggling with writing how intelligence dissemination (including net assessment and strategic gaming works). Among other things I do have posted, intelligence analysis does address several factors relevant here:
Potentially, yes to the first. I don't know if the report was a raw spot report or a coordinated document. See below. My impression, not knowing the exact process, is that a reality check did not come from an analyst, probably at DoE intelligence, that aluminum would not be the preferred material. The question should have been asked: "Could Saddam have gotten maraging steel tubes"? As it was, someone did apparently suggest the real application was in light rocket casings.
So, some questions:
Was the original note about the aluminum tubes disseminated as a spot report to the policy level? If so, interagency coordination is usually not done; the other analysts receive the note if it was sent out?
If it was coordinated, it is worth having a cleared Congressional staffer look at the actual report and where it was coordinated. Apparently, there was a dissent (Army National Ground Intelligence Center, IIRC) that they fitted a Soviet-style light rocket.
Were the tubes and the yellowcake mentioned in the same report? Did anyone see the two together, add 1 + 1, and get 3? 3 is the answer if no one checked for the other indicators of separation.
More significantly, anyone (Department of Energy intelligence?) familiar with centrifuges would know maraging steel is the preferred materials for tubes. There are other indications of centrifuge cascade construction, such as a large power generating facility nearby. I haven't heard anything discussed openly about these and other relevant indicators.
It is a very real point, Mike, that many analyst reports are not really read unless the analyst does someting (e.g., a phone call that you really need to read this, or the analyst has a working relationship with at least a staffer for the policymaker that needs the information. I'm still working on the article including dissemination and estimates.
OTOH, I knew, without looking it up, that aluminum wasn't the preferred material, and there has to be a large power source near anything that does centrifuge cascade separation, and I'm really not a nuclear specialist. Was, perhaps, a trade analyst with no technical background looking at the report?
As a stray comment, the Militarily Critical Technologies List (MCTL), which probably had a relevant note in several sections, is now classified. I have an earlier, unclassified version somewhere in storage, but maybe I can find a copy online.
--
Howard
*equal opportunity offense to both extremes*
"Those who cannot remember the past are condemned to repeat it" [George Santayana]
October 31, 2007 8:13 AM | Reply | Permalink
When I saw the word "dissent" , "Minority Report" came to mind.
dc
October 31, 2007 10:00 AM | Reply | Permalink
"Perhaps this helps explain the mess we encountered with the October 2002 National Intelligence Estimate on Iraq."
The history of the run up to invading Iraq shows that Bush wasn't interested in anyones "intelligence" but his own.
And we see where that's got us today.
You don't have to be a blind conservative not to see it, just an ignorant one to deny it.
October 31, 2007 10:30 AM | Reply | Permalink
Why is more thorough analysis not being done? Because Cheney doesn't give a rat's ass about it. To him, "intelligence" is something to "massage" so that it backs up what he already wants to do. He's not at all interested in rigorous analysis of intelligence and of course all Shrub wants to do is ride his bike and talk with his mouth full. These insane, lethal clowns aren't a government -- they're the Marx Brothers on acid.
October 31, 2007 10:45 AM | Reply | Permalink
Maybe I'm an old guy too Larry but I have to tell you that the process you described sound very reasonable to me and leads me to ask - do they no longer do this? And if not, why? This really seems like one of those "common sense" situations. Then again that's a quality that's proven to be remarkably lacking in anyone associated with this administration and the agencies working with it on it's watch.
I know we've had some atrocious intelligence failures within the last decade (in particular under this administration's watch) and I can only assume that this is in part a result of an overall politicizing of the intelligence community. I do understand that there are "politics" in everything. What I'm referring to is more along the lines of what this administration appears to be doing in the courts and the Attorney General's Office - stacking them with cronies and yes-men. Larry do you think that if there has been a serious breakdown in the sort of intelligence vetting you describe here that this might be one of the causes?
I've also felt for several years now that there is a growing trend towards incompetence and unaccountability all over this country, both in the political and the private sectors. I know it might be a dangerous generalization but I feel like we are seeing more and more examples of it every day. We see people that seem completely incapable appointed to important positions. And when they inevitably fail there's a hallow apology at a staged press conference and the guilty party is quietly replaced by yet another incompetent individual. And then, after six months, the replaced individual is tapped on the shoulder and asked to fill yet another position which they are incapable of performing and the circus rolls on. It all seems maddeningly irresponsible and maddeningly obvious.
October 31, 2007 11:24 AM | Reply | Permalink
Great points, Larry. It would be worthwhile to find out who was responsible for the change, and what reasons were given for it.
"If the only tool you have is a hammer, you tend to see every problem as a nail." ~~ Abraham Maslow
October 31, 2007 11:24 AM | Reply | Permalink
My guess is this was probably changed at Bush's or Cheney's request. Bush is the personality type who doesn't handle conflicting opinions well. He doesn't like opinions. If someone tells him something, he expects it to be the final word on that subject, not an opinion based on available intelligence.
If you give him a paper on Honduras that has an analysis and a dissenting opinion, he's going to hand it back and say "Who is right?" It doesn't help that if you give him a paper on Honduras that conflicts with what he knows to be true about Honduras (based on what someone tole him before, which he accepted as true), he is going to hand it back and say "This is wrong."
So what Bush probably gets now is the single analysis that best represents what he already knows to be true. That's the only analysis he respects. He believes that he has an infallible ability to detect the truth. Analysis can't conflict with what he knows to be true, because once he accepts something as true, it cannot ever become false without implicating his ability to discern truth. And without that ability, he is an empty vessel adrift on an ocean of opinions. He is lost and alone.
October 31, 2007 1:12 PM | Reply | Permalink
My guess is this was probably changed at Bush's or Cheney's request.
===========================================
Nah
Per LJ:
That is not the case today and has not been the case for at least 8 years. I still have not been able to determine who instituted this change–was it Woolsey, Deustch, or Tenet?
October 31, 2007 1:40 PM | Reply | Permalink
Larry, you say analysts are "not required" to coordinate but are they prohibited from coordinating?
I'm surprised at least some analysts don't realize the importance of getting other viewpoints on their work and don't seek out their counterparts for those viewpoints,
October 31, 2007 7:28 PM | Reply | Permalink
Whether or not coordination is required depends, in part, on the time criticality of the information. All analysts are not CIA; think of the military as well. There are Flash (and higher) priority reports that need to get to the highest authority fast, even if they turn out to be wrong.
PINNACLE (OPREP-3P)Here are some of the categories for US and Canadian OPREP-3 reports. "OPREP-3 reports use command and control channels to notify commanders immediately of any event or incident which may attract international, national, US Air Force, or significant news media interest. The report is established to provide time-sensitive information on which to base an appropriate response to any significant event or incident that has occurred or is in progress.
OPREP-3 reports categorize events or incidents according to their nature. Flagwords associated with each category aid prompt transmission, processing, and distribution of the reports by alerting people to their importance.
This report provides information to the National Command Authorities, NMCC, and appropriate commanders on any incident or event of national level interest which does not require notification using other OPREP-3 reports.
PINNACLE NUCFLASH(OPREP-3PNF).
This reports the actual or possible detonation of a US nuclear weapon which may cause the outbreak of a nuclear war. This report has the highest precedence in the OPREP-3 reporting structure.
PINNACLE FRONT BURNER(OPREP-3PFB).
This reports any attack on North America or harassment of NORAD forces.
PINNACLE BROKEN ARROW (OPREP-3PBA).
This reports an accidental event involving nuclear weapons or nuclear components but does not create a risk of nuclear war.
--
Howard
*equal opportunity offense to both extremes*
"Those who cannot remember the past are condemned to repeat it" [George Santayana]
October 31, 2007 8:47 PM | Reply | Permalink
Um, did you answer my question? Seriously, I don't see how any of what you said gets to the question of whether CIA analysts are prohibited from coordinating with other analysts.
October 31, 2007 9:01 PM | Reply | Permalink
There isn't always time to coordinate. The various 24/7 Watch Centers have indications checklists of things that have to be reported right now.
Are you assuming all intelligence reports are analytical studies? They aren't. NRO and NSA, for example, are basically collection rather than analytic agencies. Their watch centers still have things that are to be reported immediately.
The level of coordination on something very hot is usually to convene a conference call among the watch centers, and then call in specialists as necessary.
With perhaps some restrictions for exceptionally compartmented material, there's no prohibition against coordination. Sometimes, there isn't time to do it.
If someone bans it just for the hell of it, they are going against just about everything in analytic tradecraft. Seriously, do you want some references? You might go to the CIA online library and look through the Sherman Kent Occasional Papers, or some of the Center for the Study of Intelligence. Here's something I've written on intelligence analysis; I'm still writing on dissemination.
--
Howard
*equal opportunity offense to both extremes*
"Those who cannot remember the past are condemned to repeat it" [George Santayana]
October 31, 2007 9:15 PM | Reply | Permalink
OK, I'm going to try this one more time: I'm not saying there is always time to coordinate; I'm not saying all intelligence reports are analytical studies. All I'm asking is whether in cases where it would be possible for such coordination, analysts have been forbidden from doing it. Larry said they were not required to coordinate. I merely asked if they had been forbidden to coordinate. It's a simple question.
November 1, 2007 1:10 PM | Reply | Permalink
To the best of my knowledge, except for things like the Office of Special Plans that were outside the intelligence community anyway, no one has been forbidden to coordinate.
That doesn't mean that for specifc studies, there have not been case-by-case restrictions. These could be for reasons of extreme security, which have a history of biting back in an uncomfortable place. There could also be pure personalities involved, either of the analysts or the managers of the analysts.
There is a good deal of current literature from the intelligence community, however, that indicates they need both better tools and doctrines to improve coordination. NSA, for example, has been the most aggressive agency in using collaborative Wikis, but NSA is not primarily an analytic agency. They certainly do participate in analysis related to their areas of expertise. DIA probably works with them fairly well.
CIA apparently has been the most reluctant to use collaborative tools, due to a culture of controlling information. Whether this changes with the new role of the CIA under the DNI is yet to be seen.
INR has a tradition of coordination, but their diplomatic roots have something to do with that. Interestingly, they are, in one respect, the antithesis of NSA: they are a pure analytic agency with no collection ability other than diplomatic reporting.
--
Howard
*equal opportunity offense to both extremes*
"Those who cannot remember the past are condemned to repeat it" [George Santayana]
November 1, 2007 1:28 PM | Reply | Permalink
Thank you. That is good to know.
November 1, 2007 7:57 PM | Reply | Permalink