HQ Staff Work - Metrics, or, Damned Lies and Statistics

MIL INT particularly likes a report published over at SWJ by Maj Grant Martin (Army SOF). The paper is about the use of "Operational Design" a planning methodology currently popular in some circles. But the report's true value it seems to me is an interesting insight into being on a HQ staff and what a group of smart folks in a difficult situation see as important or not.

In one (edited) section Maj Martin talks about the futility of metrics if the wrong thing is being measured.

Our group took all of these issues in during a briefing from the assessment folks. We discovered quite quickly that our metrics were measuring short-term progress and not a lot of long-term progress that would enable “transition”. In fact, we could not find many metrics at all that gave us a feel for how the Afghans were coming along in terms of taking on more and more responsibility.

Lastly, we recommended that we team up with the operational HQs and develop metrics and an assessment plan that would test the Afghan's propensity to run things themselves as well as inform us if we were heading down the wrong path- basically telling us if our assumptions were wrong.

Our conclusion on this effort was that our recommendations did not make it to many audiences within the command, but some of our group assumed that our conclusions were not what the command even wanted to hear. First, quantitative assessments were easier to understand for outside audiences. Second, the current metrics had a track record and both the U.S. Congress was used to them and the Operational Research folks were tied to them. The thought process was that any new metrics would take a long time to tell us anything. What they advocated was using the old metrics- which they said had only been around a few years anyway, in order to establish a baseline and start making some conclusions based on the trends that we were only now beginning to be able to spot. Thirdly, there was no political appetite for setbacks- and prioritizing Afghan progress in terms of them taking over operations entailed risk of short-term negative metrics. Lastly, in terms of teaming with the operational HQs, it was thought that they had a very different view than we did on what the priorities needed to be in Afghanistan. Because of those opinions, many in our group thought that our ideas had not met a friendly reception, if they had gotten one at all.

The Design Group at this time went into a frustrating and cynical period of time. We were frustrated that we were unable to understand the Coalition's actions vice our words in Afghanistan and that we were unable to get any traction at the higher levels on re-looking our underlying assumptions.

This is just one snippet of a very interesting paper - MIL INT encourages readers to take it on in full. The argument above goes to MIL INT's arguments regarding US and NATO efforts that forget TE Lawrence and ignore friendly locals while we 'do the job' for them.

Maj Martin's group goes on toe correctly assess the strategic picture and makes a series of recommendations culminating in this last (of many excellent) recommendations and the HQ response:

We decided to recommend that our command empower the lower levels out in the regions and task them with managing and transitioning the Afghan forces themselves, instead of trying to run everything from Kabul. Our concept put all of the training centers and schools under the regional subordinates, put a lot of our general officer slots out there instead of in Kabul, and tasked those still in Kabul mainly with just advising in the ministries. The real work would be transferred out to the regional teams. Interestingly enough, the same course of action, albeit with a few minor differences, was the conclusion of the planning team led by the planning section. And, interestingly enough, the recommendations were rejected outright.

This is the operational equivalent to Killcullen's recent interview at the strategic level where difficult realities are wished away and arguments are made for more of the same - just with added resources.