Tuesday, June 9, 2009

Why Do We Need Metrics?

The Center for A New American Security (CNAS) released a new document at the end of last week, and hopefully it will spark a discussion about measuring success or failure in Afghanistan. Even more hopefully, it will spark action following the discussion.

The discussion will hopefully refine the recommendations of some of the premier counterinsurgency theorists in the world today into actionable metrics for COIN, a subject that has been a sticking point in our execution of COIN, and potentially a shortcoming of FM 3-24 Counterinsurgency. Military officers, like business leaders, are rated on their success or failure in any environment based upon measurements. The question has been, "What can we measure that will tell us if we are doing the right things?"

In business, successful behaviors are relatively easy to measure. How much money did the business unit make? How much of that money was spent on making/delivering the product or service? How much money was left for profit afterwards? There are a lot of measurable factors that go into those main factors, but in the end there are lots of pertinent things to measure success or failure of any business unit, making managers easy to reward or disincent. In counterinsurgency the military officer is confronted by a seemingly nebulous environment, and he/she will often fall back on traditional military measurables, which have been demonstrated not to correlate to success in counterinsurgency. Worse, in their search for quantifiable meaning, officers will be forced to come up with equally or more meaningless measurables that can trick them into continuing unsuccessful behaviors.

The funny thing is that, while they are untrained doctrinally, the average Joe on the ground often sees the futility of the measurables that the officers above him are depending on. The guy on the ground may not understand what the hell "hearts and minds" is supposed to be all about, but he recognizes wasted action when he sees it. He may not have a better answer... sometimes he does... but he does know when his time is being wasted on unproductive behaviors or that potentially positive behaviors are being quashed in favor of an unproductive metric. All we have to do is look at the words of junior leaders who are in the suck or who have returned from it. These words help to diagnose our unproductive behaviors and our failure to train our junior leaders in the doctrine that they are expected to execute. They also diagnose our failure to choose objective measurables that mean anything.

Leaders, whether military or civilian, will strive to affect the measurable factors that they are measured against. Military officers begin their rating process by completing an "OER Support Form," or Officer Evaluation Report Support Form, in which they tell their boss what they are going to achieve during the rated period. The results of the OER affect their promotions... they mean money and career progression. Basically, they tell their boss what they will do and how they will measure their success or failure. They will choose metrics that are, first of all, measurable... usually easily measurable... and secondly, achievable. No one will set themselves up for failure. While we all agree that we are engaged in a fight against insurgents, we do not all agree on how to measure success or failure in such an environment.

We haven't all bought off on the appropriateness of the doctrine to actually fight against the insurgents, hence the COINdinista vs COINtra struggle.

This blog has pointed numerous times to the necessity to hold commanders accountable for their effects on success or failure during the time they spend in theater. In all previous conflicts, officers who were unable or unwilling to achieve the necessary results on the battlefield were relieved and replaced. Careers were stymied and ended. Until very recently, with the relief of GEN McKiernan, no such message was being sent in the current conflict. Very few officers have been relieved, and most maneuver unit commanders have declared excellent performance regardless of the security situation in their particular areas of responsibility. Since they set the measurables, which their higher commander have agreed to, they can point to these measurables and declare that they had positive effects on those metrics. Based on these, the declarations of success are warranted... but do they mean anything? Only one commander has been held accountable for the loss of "ground" in Afghanistan; and I posit that he was relieved in such a manner as much as a statement as for any lack of success that was worse than his predecessors... as well as to make room for a commander that GEN Petraeus believes will be more successful.

Lieutenant General Stanley McChrystal gave a hint during his confirmation hearings that the measurables are going to change. He said that we will not measure success by body counts, which we have been sliding towards, but by the percentage of the population shielded from violence. This is difficult to measure in its own right, and there are many things that will go into it. While, in a population-centric counterinsurgency, securing the population is job one, how do you measure something that hasn't happened? How do you measure the activities... and thereby incent their application... that contribute to providing security? Since COIN is really a political struggle where perceptions are important, how do you measure those?

This has been a key piece missing from the puzzle. It has resulted in rewarding failure. It has resulted in the continuation of failed behaviors such as staying rooted on what Tim Lynch has labeled "Big Box FOBs." This is a behavior which will never, as LTG McChrystal states is important, secure the population and prevent their being victimized and intimidated. You have to be there. Going home to the Big Box at the end of the day, only showing up to the village every once in awhile and demanding that they tell you where the Taliban are just doesn't prevent intimidation.

Commanders will do what they are incented to do. When they set the metrics, based on what they feel that they can do, and based on concepts that have nothing to do with preventing intimidation of the population, they will. They will choose metrics like force protection or tons of Humanitarian Assistance distributed, or missions run vs number of casualties or, as has been a trend lately, on enemy body counts. The population will still be subject to the predations of the Taliban or Taliban-like or affiliated groups; and we still lose ground. The point is that commanders need to be given metrics that work to measure positive counterinsurgent indicators. They will measure something. We need to ensure that the metrics that they use mean something to the counterinsurgency.

That's why metrics... good, meaningful metrics... are so important.


  1. Good questions Ole Blue.

    Metrics Matter. Unfortunately numbers can be manipulated. Political Correctness can be played in.

    It does matter how many bad guys are killed, but it also matters what kind of bad guys are killed, and if we can prove they were bad guys and if the the locals know they were bad guys.

    Col Julian gets this. He knows that the information war is as important as the ground war.

    I cannot support the firing of McKiernan because I've come to believe it was for the wrong reasons. I can support the appointment of McChrystal though I fear he is hamstrung from the start.

    As I say about my own site: It is not just about what we know, but what we can prove. It is not enough to know we killed the bad guys, we must be able to prove they were bad guys and not just civilians.

    Keep up the good work!

  2. The Thunder Run has linked to this post in the blog post From the Front: 06/10/2009 News and Personal dispatches from the front and the home front.

  3. Old blue, I made a comment on Tim Lynch's blog last night about this, but I wanted to add a note here as well.


    At the CNAS conference yesterday, GEN Petraeus stated that he felt as long as our FOBs/COBs overwatched rural villages in Afghanistan that would be enough to provide security.

    Having worked in Bermel with the Vampire crew last year and also having operated in many other remote locations on FOBs that were only a hundred yards from a village, I can say without a doubt that it is not enough just to have a FOB nearby, the occupants of that FOB have to be outside, in the village and interacting with the population all the time.

    Our FOBs separate us from the population, it is not easy for civilians to come to the FOB and have a talk with us, they would be viewed as collaborators and get beated/executed for it.

    The US ISAF forces in RC-East are stuck on their FOBs, it takes them hours or days to get approval to conduct a patrol or act on time sensitive intelligence. Half of the Platoon/Company Sized COBs on the border are nothing but static defense sites. They sit there, wait for a few 107MM rockets on timers to be launched at them, track it on the radar and do a counter fire mission on where the rockets came from even though the enemy is long gone. They just sit there and wait to get shot at!

    Instead they should use these COBs as rally points from which to launch dismounted patrols, 4th BCT 25th ID has started to do this, but 4th BCT 101st ABN basically accomplished nothing in the year they were in Afghanistan.

    One of GEN McChrystal's biggest challenges this year is to recognize our tactical failures in rural Afghanistan and to change the culture and risk adverse attitudes of our BCT/BN leadership. We need to get our Soldiers out into the villages and interacting with the population or at least make them more readily available to the locals. Being on a FOB within ear/eyesight is not enough, the walls aren't transparent enough, there is no trust with us hiding behind concertina and hescos.

    It sends a message to the population, "you're district/village isn't safe enough for me to be outside of my FOB interacting with you everyday." Whether or not it is actually safe or dangerous is immaterial, the perception alone makes the population feel unsafe because their local guardians (ANSF and ISAF) are hiding behind their body armor, vehicles, hescos and concertina wire.

    Junior leaders and transition team members recognize this, they understand it, but their chain of command does not. The policies, procedures and orders that the guy on the ground has to deal with from the upper echelons of command are distracting him from completing the mission and prosecuting a truly robust and synchronized COIN campaign that integrates information operations, lethal and non-lethal effects.

    GEN McChrystal needs to institute massive culture change amongst our field grade and General officers in Afghanistan, then we will start seeing some more widespread tactical success.

  4. Nice post! Thanks.

    In any big endeavor, we have a tendency to rush into the solution before we've really defined the problem. The hardest problems to define are those that seem so obvious at first. How many of us in Afghanistan really know what the problem is that we're here to solve? Does anyone?


All comments will be moderated due to spamming of old posts.