Tuesday, February 16, 2010

Protecting civilians in extremis

This Washington Post report gives a much more in depth look at the rules of engagement established by GEN McChrystal to stem mounting civilian casualties. While I completely agree that some constraints had to be put into place, this struck me as somewhat drastic:

Not seeing any civilians on a video feed from a drone or through one's rifle scope is no longer enough. Under a tactical directive McChrystal issued last summer, troops must verify that there are no civilians inside a house by watching it for at least 72 hours to establish a "pattern of life" before an airstrike will be authorized.

Emphasis mine. Even if armed insurgents were observed going into the house or going into a different house. Or stay or move in the vicinity of houses. Given some of the scenarios Marines are facing now in this new offensive (as reported in this article), this seems somewhat extreme. 72 hours of watching a house?? Operations in Iraq and Afghanistan have been economy of force missions - trying to do more with less. Even with the assistance of UAVs to keep eyes on larger tracts of land not controlled by forces on the ground, 72 hours is a long time to observe anything and devotes significant resources that could be used elsewhere. I don't see how this helps prevent civilian casualties when less amount of time would suffice for the command to justify the use of force (oh, I don't know, maybe four to six hours?). I'm all about observing the situation and minimizing civilian casualties, but 72 hours??

I also highly doubt these limitations are guiding drone strikes across the border in Pakistan - given that those operations do not fall under the purview of ISAF (to my knowledge at least) and that they are usually conducted by the CIA (I should note that I don't know if they have these rules or not, but given the reports of these strikes lately it doesn't seem to be the case). Essentially, it suggests that killing "high payoff targets" of both Pakistani Taliban and al Qaeda requires less rigor than destroying imminent threats to our soldiers and Marines on the ground. Understanding the strategic effects of killing high-ranking enemy, there are pretty significant effects to wiping out a platoon of insurgents armed, grouped, and about to attack your own forces. I have to say that would make me pretty angry if I were on the ground.

This also hamstrings commanders and removes the onus of decision making from him. That is what commanders get paid to do - make decisions based on the information he has in the interest of mission accomplishment. There is plenty of room for error of course, but this order essentially removes the ability of commanders to neutralize threats in any built up area. What if the Taliban guys just move from house to house every 48 hours? That will tie up large amounts of resources, either UAVs or putting boots on the ground to clear the area.

I'm not very keen on this order, and it's not even my ass hanging in the wind because of it. This strikes me as extreme and only used to appease the Afghan government while debilitating U.S. forces' ability to destroy enemy forces. This is no way to win a war.


  1. The ROE do seem kind of insane in how stringent they are. I wonder what the airstrike and artillery ROE are for the various U.S. SOF commands, which do not fall under ISAF.

  2. I had a similar reaction when I read the 72 hour bit this morning. I wish it weren't the case, but it fits with the recent evolution in my thinking about COIN provisos as strategic communications more than guidance to enhance operational effectiveness.

    Do McChrystal, Nicholson, et al really believe everything they're saying?

    Shouldn't we have more empirical data on whether a 72-hour observation period does anything to decrease the likelihood of civilian casualties, or whether this alleged decrease has any meaningful second- and third-order effects for accomplishment of the mission? Are we just taking these things as matters of faith?

  3. Q: If a unit is taking fire a built-up area and is pinned down and the patrol leader determines that he need to call in an airstrike on the buildings, will that request be denied on account of the 72-hour rule?

    I suspect the answer is no. In that case, this rule strikes me as no big deal. If the answer is yes, then we've lost our minds.

  4. Tintin- I thought part of the big deal about bringing in McChrystal is that SOF now falls under ISAF. I heard this because now it is much easier to get ISR assets because SOF competes with all other units, not preempts them like before.

    As to the ROE issue, I support the restrictive ROE. In most cases, and this is what ROE is for, US units are not desperately pinned down fighting for their lives, but are responding to wildly inaccurate popshots. In this case, a strict ROE is a great thing.

    Also, alot of the talk on ROE just seems really narrow minded. Americans forget that killing children, even on accident, sets a terrible precedence. We have to limit civilian casualties or we will never get out of Afghanistan.

  5. There a number of very good questions here that I, obviously, do not have the answers to. Hopefully more will come to light over time.

    @Michael C - I am all about a more restrictive ROE as well. The number of civilians killed by ISAF air strikes was terrible. However. This seems absurdly restrictive and probably my biggest issue is what I highlighted about taking decisions out of the commander's hands with a blanket policy like this. He should (operative word) know best on how to weigh the consequences of his actions.

    This ROE seems to be trying to avoid all civilian casualties - which is impossible. Or at least make it look like we're trying to avoid all civilian casualties. We can't do that in any way, but I'm more inclined to let commanders make the decisions they get paid to make, not let IO policy in Kabul (in my mind) unnecessarily risk soldiers' lives and tie commanders' hands.