Avoiding a Kluster ****

National Intelligence Estimates (NIEs), are supposed to be a collaborative effort of every agency (in reality, every agency with a dog in that particular fight). Whether you value the findings in recently declassified NIEs or not, most of us have a pretty good idea of what any sort of product produced by committee is like. There is a saying about enjoying sausage and avoiding knowing how it is actually made that applies.


A key problem with NIEs – at least those that are not linked to charged political issues – is that you don’t always get all of the right people in the same room at the same time. Not everyone can make every meeting, sometimes key someone’s can’t make any meetings. The authority of a document that is used to advise Presidents tends to be diminished if you called it a “collaborative, consensus product from everyone who didn’t get caught in traffic.”
However, an NIE built via a mechanism likeKluster (insert joke about how the process is a cluster-**** already here) and ask yourself this question: why _not_ build intelligence assessments this way? Aside from the significant drop in impact to participants by working remotely, consider:

Each user’s successes, failures, reputation, areas of expertise, and overall history are considered. This encourages users to earn respect, to act positively, and most importantly, enables extremely educated decisions to be made using real world logic.

Can you think of a more practical way to minimization of politicization?

2 comments for “Avoiding a Kluster ****

  1. Michael Tanji
    February 28, 2008 at 3:29 PM

    I don’t claim that it is a common problem, but having lived through it I do know that it exists. Having half the right people weigh in in person and then the balance (or less) weigh in later isn’t as effective as all the right people all the time.

  2. Ralph H.
    February 28, 2008 at 1:51 PM

    I participated as an agency rep in four NIEs that I recall, and I was the principal drafter in a fifth. I don’t recall that getting the right people to the coordination sessions being a “key problem.” People missed meetings, but their input invariably showed up before or immediately after the meeting. At the end of the day I don’t think any viewpoints were left out by accident. Some were deliberately excised, of course.
    While detailed to the NIC I’m rather proud to have been one of the first, if not the first, drafter to plug in a laptop at the conference table and have the working text up on a screen so that the whole group could suggest and see immediate changes. A senior CIA analyst whose name I can’t recall went one better, using the “hidden text” feature of MS-Word to embed supporting intelligence report titles & document #s with every phrase, sentence, or paragraph that enunciated a conclusion, no matter how minor. This was in the late 1990s, and I can only hope that collaboration apps have made greater inroads in the last few years.

Comments are closed.