Friday, July 11, 2008

Book review - Still Broken by A.J. Rossmiller

Still Broken: A Recruit’s Inside Account of Intelligence Failures, From Baghdad to the Pentagon

A.J. Rossmiller wanted to serve his country. Already interested in a career in intelligence and/or law enforcement, the events of September 11, 2001 eventually led Rossmiller to a job as an intelligence analyst at the Defense Intelligence Agency shortly after graduating college circa 2004. And not long after that, his conscience led him to volunteer for a minimum 6 month deployment to do DIA counterinsurgency intelligence analysis in Iraq. He had left hoping to make a contribution to the stabilization of Iraq, but after a dissapointing and frustrating experience there he returned to the States to continue doing intelligence analysis, only to have the remaining optimism he had left dashed against the reality of an intelligence system that is still broken from the top to the bottom.

If this sounds vaguely familiar, then you might be thinking of the review I wrote of Inside the Wire, which similarly offered an on the ground view of the way that our intelligence system is dysfunctional. I haven't searched around the net to see if anyone else has made this point, but these two books work really well as companion pieces, with both offering first hand accounts of how and why our approach to terrorism is counterproductive and ineffective. If you can find the time to read one, I recommend reading the other one as well.

In the Inside the Wire review I listed four main factors that the primary author noted for the problems in intelligence gathering at Guantanamo. For Still Broken I can narrow the factors down to two main categories: bureaucracy and ideologically driven/determined intelligence analysis.

At nearly every step of Rossmiller's journey he encounters some form of bureaucratic inefficiency. Teams were constantly in flux with assignments changing or up in the air. There was in-fighting between mid level management about who should be doing what. Reports wouldn't make their way to where they would be needed. Problems got passed around and ignored; rarely were they addressed and solved.

He watched his hard work -85 hours per 6 days - of critical and careful intelligence analysis turned in the field into Anyone who's picked up gets sent to prison. "Are you fucking kidding me?" was the response Rossmiller had when first informed of this procedure. Rossmiller expected that if those detained hadn't been setting IEDs off before they were sent to prison, they would certainly be more likely to do so after getting out.

The "screening" interrogations that Rossmiller watched amounted to inexperienced debriefers yelling and shouting repeatedly at detainees (some, or many, of which were randomly rounded up for being in the wrong place at the wrong time) that they were an insurgent and needed to confess what they were doing.

To give you an idea of this process, here's a passage from right after Rosmiller watched a detainee ask in futility what his crime was.

I turned away as the interrogator called for an escort to take the prisoner to the holding area for later transport to the prison. No evidence, no charge, no reason. What the hell were we doing? When he was gone, I asked the team if they couldn't have used the intelligence files to ask him about specifics.

"Oh, you mean this file?" The debriefer opened up the folder in front of him, exposing a single sheet of paper with the detainee's name and ID number on the top. "We don't have shit on these guys. They were in a spot with insurgent connections, so they're either insurgents or facilitating insurgents. Fuck it, just send 'em to Abu G and they'll sort it out there."
Can you already guess what the bureaucratic wrinkle is going to be? That's right! The people working at Abu Ghraib assumed that anyone sent to them was guilty. In Iraq (as it did at Gitmo), the U.S. has managed to make Kafka's vision of a bureaucratic hell in The Trial a reality.

Disregarding the injustice of such a policy, Rossmiller expresses how terribly stupid this is from a practical view

Considering that the single most vital element in fighting an insurgency is to eliminate popular support for the fighters, rounding up and incarcerating as many people as possible is exceptionally counterproductive. An occupying force can establish a secure environment in one of two ways: Either the citizens must be so fearful of the occupiers that they don't dare transgress, or the citizens have to believe the occupying force is better than the alternative.
Given that even were we to set aside the moral repercussions of attempting to make the Iraqi populace fear us to such an extent that they dare not engage in insurgency, it would still be practically impossible to achieve. So that leaves you with the second option, and incarcerating at random without any due process large segments of the population is not the way to achieve that goal.

Rossmiller also succinctly expresses the problem with those who lament that the press is not doing enough good news reporting about Iraq:

[S]uch criticism mistakes the micro for the macro. That is, building a school in the midst of a civil war is like planting a tree in the midst of a forest fire: a nice gesture, and intrinsically good, but not likely to provide much help overall.
At the end of his 6 month tour in Iraq, Rossmiller returned to the States and began work at the Pentagon in the Office of Iraq Analysis. Despite his having developed grave doubts about the way the occupation of Iraq was being conducted and retaining a frustration that the work he and his team did there was never properly utilized, he still maintained an optimism that he would be able to contribute to improving the situation. That would not last long.

There was still a problem of bureaucracy but Rossmiller managed to get the hang of the system and work around it to the best of his ability. However, something that really was holding him back was a lack of resources. Rossmiller was doing intelligence for the war in Iraq in the DIA’s main office but had no desk!

Of course, I still didn’t have a desk, and as the rest of my Baghdad companions began to filter in, space and resources got even tighter. Of the tons of problems with the office and with the analytical process in general, including fairly shocking examples of analytical modifications, terrible human resources, and atrocious morale, the thing that bothered many of us most frequently was the lack of material resources in the office. We simply could not believe that three years into a major war, the most significant military and foreign policy effort in decades, and with American soldiers dying on a daily basis, the main Department of Defense Iraq intelligence shop was in temporary office space without the resources we needed to succeed.

The office was budgeted for roughly a hundred people, and due to deployments, rotations, and a backlog of candidates for open slots, we actually had about eighty. For those eighty, we had approximately sixty computer stations. At any given time, there were people who simply could not do their jobs because we didn’t have enough computers, and the ones we did have were undersupplied.
The work Rossmiller did required 3 operating systems: Class A, Class B and Unclass (internet capability.) All the systems were necessary but only Class A was on all of the computers. Class B was available on 15 computers and Unclass on 12. They were in an office that didn’t even have internet on all the computers!

Yet the room adjacent to the OIA was empty and most of the computers there had all of the systems up. Rossmiller was given a bureaucratic reason for not using the space (i.e. belongs to another unit). Meanwhile, his team had to utilize shifts because of the lack of computers.

The adjacent space had been empty for months before I arrived; it remained so throughout the time I spent at OIA, and, according to friends still in the office, it’s unoccupied as of this writing. There still aren’t the computer resources we needed to do our job properly, and leaders are still telling personnel that it’s “in progress.” I suppose it’s not surprising that DoD and administration leadership who failed to provide proper armor for troops and transportation in Iraq would neglect the intelligence operation driving the war effort, but one despairs despite the predictability. I wonder how many operations we could have helped, how many better predictions we could have made, and, most important, how many lives might not have been lost if we just had the same number of computers and desks as people.
Rossmiller got some inkling of the DoD priorities by the fact that his office did have “about a dozen enormous flatscreen TVs blaring Fox News nonstop from the walls.”

Before the Dec 05 election the Bush administration was saying the election would lead to unity and stability. Rossmiller and colleagues disagreed. Their analysis indicated that the election would break down along ethnic and sectarian lines and that secular and centrist leaders would be shunned. Leadership in OIA reprimanded them for being “too pessimistic.” As the election approached, leadership continued to press for alterations in the analysis – for more positive assessments that better reflected the administration rhetoric. Rossmiller notes that office supervisors spoke directly to them rather than his supervisors, leaving him and other analysts with little to no options.

“This is too pessimistic. Again! Why are you guys always off message?”

When his bosses challenged their office bosses, they were ignored or attacked. When his bosses didn’t challenge them, they instead criticized Rossmiller and friends for not following the chain of authority (in other words, for not manipulating intelligence like they were told to.) Rossmiller explains what seperated this process from a normal process of debate

Two crucial elements differentiated the intelligence manipulation from healthy debate. First, while we analysts supported our assessments with reams of documentary evidence and citations of related previous accuracies, our leadership generally failed to engage us on the issues, instead questioning our analytical rigor and proffering the ubiquitous “too pessimistic” accusation and making appeals for “balance.”

Second, the changes were virtually always in the same direction: toward a more favorable evaluation of what Bush administration officials were hoping for.

One of the leaders in his office was willing to challenge intelligence manipulation. After receiving an e-mail criticizing his analysts for failing to come up with analysis that matched the stated goals of the Bush administration, Rossmiller's supervisor Major Nimick wrote back that their job was to assess facts, not to assess whether or not their findings are on message and noted that he was amazed that leadership would receive a briefing on Iraq and instead of figuring out how to deal with the truth they would attempt to change it.

Within months of sending this e-mail, he was stripped of his leadership position and put in an out-of-the-way administrative post. That demonstrated to the rest of us the danger of speaking truth to power.
Rossmiller and his team would get advance copies of speeches from administration officials (e.g. Bush, Cheney, Rumsfield, etc) and would be able to pick out the lies and false statements. But for this analysis to reach the top it would have to go through every person in the hierarchy without being vetoed.

Analysis could and would be overturned by superiors with zero analytical basis just to make it more optimistic. Example: Rossmiller had judged an Iraqi leader to be weak, and his coworkers and superiors agreed. He was ordered to change the assessment to say that the leader was strong, however, by a higher up J2 who had no knowledge of or experience with Iraqi issues. A few months later this “strong” leader was ousted in an “unpredicted" event.

Rossmiller sees two possibilities:

1.Despite constantly being wrong, administration leaders believed their analysis was too pessimistic.

2. The Bush administration has created an atmosphere of political loyalty to allies and ideology rather than truth where it is considered a stab in the back to report "bad” news.

Either way, the intelligence process is broken - it rewards failure and punishes success - with the second being the worse scenario. Rossmiller worries that these actions are becoming habits of bad intelligence analysis that will become institutionalized long after Bush is gone.

As a result of this manipulation, analysts who could think about finding work elsewhere started to drop out of the office. This problem also happened in the counterrorism unit, so the two most important offices to the mission in Iraq were the places that were the least desirable to work in.

Finally, after watching assessment after assessement get overturned for political reasons only to later turn out to have been right, Rossmiller reached a breaking point and quit his job. But he didn’t simply quit and leave, he attempted to make his leaving an effort to improve the situation by explaining precisely why he was leaving to his bosses so that the problem would at least be brought into the open.

Regardless of the impact that made on his office, this book is an important step towards galvanizing an effort to fix our still broken intelligence apparatus.

No comments: