Late comments on “Triage” and the CNAS Conference

The Center for a New American Security held its annual conference last Thursday at the Willard Hotel. The event featured GEN Petraeus as the morning keynote speaker, followed by a series of panels related to the rollout of several new policy papers. The one I’m most interested in, pretty obviously, is Triage: The Next Twelve Months in Afghanistan and Pakistan, by Andrew Exum, Nate Fick, Ahmed Humayun, and Dave Kilcullen. Fick and Ex participated in a panel discussion — moderated by the former commander of U.S. forces in Afghanistan, LTG (ret) David Barno — alongside noted counterinsurgency skeptic Andrew Bacevich and COL Chris Cavoli, who commanded a battalion of the 10th Mountain in RC-East in 2006-07.

Before I get into my critique of the policy paper, I should go ahead and voice my biggest complaint about the conference: Bacevich, however eloquent and entertaining, was woefully unsuited to participate in the Afghanistan-Pakistan panel and only served to distract both the participants and the audience from a useful discussion of the operational recommendations contained in Triage. However much I may disagree with him, Bacevich is a thoughtful counterweight to the conventional wisdom about American grand strategy (to the extent that we can be said to have one) and foreign policy. What he is not, however, is an Afghanistan expert, a tactical or operational thinker (though of course he does have experience as an Army officer in the Vietnam era), or a constructive voice in the formulation of a winning approach in South Asia. As I’ve noted in the past, I think he was unfairly personal in his criticism of The Accidental Guerilla; in some sense I’m disappointed that Kilcullen’s absence means wewere deprived of a rhetorical confrontation between two of the most disparate and charismatic voices in the debate about how America should be engaged in the world.

Bacevich, as might be expected, used the time alloted for prepared remarks to rail against American involvement in Afghanistan (and really, anywhere). He did not address any of the specific recommendations of the CNAS paper, nor did he even feign the slightest interest in doing so. I don’t blame this on the professor, but rather on those who thought he might improve a debate in which he seemed almost certain to refuse participation. One could cynically suggest, as some friends have, that this is precisely what the organizers of the panel intended: knowing that the audience would be swept up in the COINdinista/COINtra drama, the shortcomings of the paper would be overlooked. Which isn’t to say that the CNAS guys came out unscathed: Ex was made to look positively uncomfortable, at one point responding to Bacevich’s challenge about the inadvisability and unaffordability of an extended COIN campaign in Afghanistan with a plaintive “what is the alternative?”

COL Cavoli, who I had the opportunity to meet and speak with briefly at the reception afterwards (and who implausibly recognized my name from the comment threads at Abu Muqawama), offered some thoughtful criticisms based on his experience downrange. It was a real disappointment that the Bacevich diversion meant that we couldn’t hear much from the guy who was both willing and able to talk about the future in Afghanistan.

As far as the report itself goes, suffice it to say that I was disappointed. Much of what ended up in the paper has been said or written in almost precisely the same form in Congressional testimony and other work by Kilcullen over the last several months, so I was disappointed that we didn’t see much fresh analysis. One recommendation where I basically agree with the authors is on the subject of drone strikes; Kilcullen and Exum published an op-ed in the Times a couple of weeks ago making the same argument, and I think there’s a lot to commend it. Having said that, COL Cavoli joined other uniformed critics in suggesting perhaps Hellfire attacks should not be ceded so easily, as they’re one of the few ways U.S. forces can maintain pressure on militants on the Pakistani side of the border. I think the equation has also changed somewhat in recent weeks as the Pakistani populace demonstrates an increasing frustration with the disruption caused by Taliban encroachment.

So I guess the question I’m asking right now (and I credit my friend Jason Fritz for this thought) is this: why, instead of abandoning the UAV campaign altogether in Pakistan, can’t we get better at this? Why can’t we apply a higher standard of evidence to suspected targets, or only engage when those targets are a greater distance from built-up areas? Maybe this has the same effect as giving up the strikes — I don’t know. But are we sure we want to trade away that chip — something that Pakistani militants are demanding the government press us to cease in exchange for halting their campaign against civilians — and get nothing for it?

On the other hand, I think Kilcullen would respond to COL Cavoli’s criticism by suggesting that the best way to put pressure on the enemy is to mobilize the civilian population against him, something that’s impossible when you’re blowing up their houses and weddings and whatnot. And obviously this is a long-term solution to what must seem at the battalion and brigade level like a relatively un-complex tactical problem.

In the interest of avoiding a several-thousand-word ramble, I’m just going to scattershoot some of my other criticisms.

“CIVILIAN SURGE”:

• “Use the ‘civilian surge’ to improve governance and decrease corruption in Afghanistan. Place civilian expertise and advisers in the Afghan ministries and — to a lesser degree — the provincial reconstruction teams, rather than in the embassies.”

I feel sort of mixed about this. For one thing, as Ex said during the discussion, we have to be realistic about the fact that we just don’t have that much excess civilian capacity, and there really isn’t going to be a “civilian surge.” And if there were, wouldn’t it make even more sense to push that advisory and mentorship capacity even further out, to the village and district level?

I’m conflicted on this because I think that while a bottom-up effort is more effective in the short-term and probably “better COIN,” we have to appreciate that all the grassroots progress in the world is basically meaningless if no solution is found to the endemic corruption in the higher levels of government.

STRENGTHENING POLICE…

• “Strengthen the Pakistani police, with an emphasis on areas — such as Punjab and Sindh — where the Taliban has not yet exerted control.”

If we take the first bit on its own, I wholeheartedly agree. Favoring the police over the military and intelligence services is consistent with both COIN best practices and simple logic; Kilcullen has elaborated these reasons in the past — closer ties to local populations, no distraction from external threats (India), etc. — so I won’t go into them here. And I have a lot of time for people who say “we’ve thrown $10B at the military over ther last eight years and where has it gotten us?”

…IN PUNJAB AND SINDH:

Where I disagree, however — and vehemently — is the sort of “secure the base” geographic focus on the provinces east of the Indus. I think this proposal is informed by a certain alarmism, as well as a failure to adequately comprehend the range of Pakistani popular opinion and feelings of relative (in)security. This is a failure I honestly can’t understand in light of the authors’ having consulted Nick Schmidle, who does very obviously get this stuff (way better than me, which isn’t saying much, or pretty much anyone else in this country). The authors elaborate further a few pages on:

• “While Pakistan has lost effective sovereignty in most of FATA and adjacent portions of NWFP, the militants are increasingly testing the government’s ability to enforce law and order in major urban population centers across the country. If Pakistan were on the verge of collapse, one indication of impending disaster would be the government’s inability to provide security in major urban areas where much of the nation’s middle class resides — Lahore and Rawalpindi in Punjab, and Karachi in Sindh… To be sure, short-term aid to the police forces is not a long term fix for Pakistan. In the coming year, however, the neglected Pakistani police forces must be bolstered so that they can credibly secure the populations of Punjab and Sindh from militant attacks.”

First things first: are the militants really “testing the government’s ability to enforce law and order in major urban population centers across the country”? Because that doesn’t seem clearly true to me. Are the populations of Punjab and Sindh “credibly secure… from militant attacks”? I’d suggest that they are, in the same sense that the occupants of Manhattan high-rises are credibly secure from terrorism. After all, how secure does one have to be to constitute credible security? Are we to believe that the NWFP and FATA can only be dealt with once the populations of the eastern provinces are entirely safe from terrorism? Because, lest we forget when reading recent alarmist warnings about (then) Taliban control of Buner — a mere 60 miles from Pakistan’s capital — Islamabad is a world away from what’s going on in Swat.

And so: is Pakistan “on the verge of collapse”? Is the government unable “to provide security in major urban population centers”? I don’t think we’re there yet, and I don’t think pouring money into the Pakistani police is going to protect against such a contingency. Militants will take control of Punjab and Sindh — if and when they do — by seizing territory, not through political concession or popular will. Spending limited resources on beat cops in Karachi seems extremely inadvisable when the fight in Afghanistan is being waged at least in part by elements in the considerably more dangerous and unstable parts of Pakistan. As COL Cavoli said, adding effort in one place means that effort, leverage, and pressure are diverted from another place.

Is helping to bolster good governance and internal security in Pakistan’s core an essential part of stabilizing that country over the long term? Yes, of course. But can we honestly base our third-party counterinsurgency efforts (or at least security force assistance, if you don’t want to call it COIN) in that country on an approach that essentially cedes disputed territory to the enemy while the best-trained and best-resourced security forces in Pakistan operate in what amounts to the strategic rear? It strains credulity to say that perfect security in Pakistan’s cities in essential to the continued existence of the state; ironically, this argument seems to me a microcosm of Bacevich’s: offensive operations are an unnecessary and ineffective distraction, we can consolidate gains and achieve security more cheaply by focusing our efforts on what amount to defensive measures.

HOST-NATION CAPACITY IN DEVELOPMENT:

One last note (and this one is sort of unconnected to the others, but I think it’s worth commenting on):

• “[I]nternational development should be less about building schools and other infrastructure than about the process by which international donors partner with local governments and institutions. Accordingly, international aid to Afghanistan should privilege those programs… that emphasize local actors and local solutions. International aid and development workers should be less concerned with running their own projects — the overhead costs of which often outweigh the projects themselves — than with strengthening Afghanistan’s weak bureaucracies to function by themselves.

Which is all well and good, and I think we can all agree that strengthening the Afghan government is essential to this whole process. Having said that, a focus on local programs will often mean that resources are wasted on ineffective implementation, not to mention the fact that aid agencies and NGOs are bound to have concerns about accountability.

Reading this suggestion, I couldn’t help but think of Secretary Rumsfeld and GEN Casey in the period immediately preceding the Surge in Iraq, consistently and insistently reminding everyone that the mission was to focus on making Iraqi security forces capable of providing for the country’s security. Which, again, was all well and good except that circumstances simply weren’t right for such a transformation to take place: the operating environment was not permissive, to say the least, and the competence of the ISF was eroded by sectarian strife and political turmoil.

All of which by way of saying that the development of host nation capacity is the absolute sine qua non of COIN, but you can’t just wake up one day and say “well, we’ll develop HN capacity by letting them do stuff for themselves.” Sure, good idea, but there’s got to be an implementable plan to make this happen, and I don’t see it here.

METRICS:

Lastly I suppose I should say a word about metrics. There’s been a lot of debate about this subject, and it was my fervent desire that Ex, DK, et al would come along and clarify things for us. Unfortunately that hasn’t happened. What we have instead is something that strikes me as a bit of a firehose approach: let’s write down every metric we can think of that might be useful without giving a coherent plan for measurement or evaluation. “Improved civilian control over the military would also indicate progress,” for example. “A clear improvement in this area, as evidence by clear subordination to civilian decisions, would be a key sign of progress. A related metric is the proportion of aid… that Pakistan actually spends on supporting the coalition, securing supply lines, and combating militants, rather than diverting it into the pockets of corrupt officials or spending it on assets more suited to fighting India than surpressing extremism.” Ok, but how exactly are we meant to document that improvement? Are we meant to believe that the civilian leadership in Pakistan will readily account for the percentage of funds delivered to the security services and then applied to ends that the civilians find inappropriate? That a government already nervous about its ability to effectively control the military would readily highlight its own weakness for its benefactors, thereby stanching the flow of aid dollars from international partners that deem Islamabad to be ingracious or incompetent?

This particular “metric” is representative of a broader section that I think is slightly ill-considered and overly vague. The CNAS authors note that President Obama promised clear indicators for success in his rollout of the new “Af-Pak” strategy, something that he has thus far failed to provide; there’s some irony then that CNAS promised the same thing from its own report, then disappointed. As LTG Barno said, fewer, clearer metrics are always better. Here we’re faced with many vague and often unmeasurable metrics — or at least ones that the authors haven’t compellingly argued can be measured in a meaningful way — and I’m not sure that we’re any closer to an answer in the end.

I guess you won’t be surprised to find that I was really disappointed by Triage. There were a number of smart people working on this paper, and they lined up even more smart guys to assist (Christian Bleuer, Josh Foust, and Schmidle). I just feel like this was a missed opportunity, that CNAS in its haste to contribute to the debate ended up putting out a product that fundamentally failed to offer workable, near-term solutions. “Ruthless prioritization” resulted in a wish list of operational adjustments that seems improbable, to say the least.

Still, the conference was a blast. I had a great time talking, eating, drinking, smoking and debating with friends who think seriously about these issues, and that’s always a treat. I got a chance to meet COL Cavoli and MAJ Jim Marckwardt (one of the co-authors of The Defense of Jisr Al-Dorea’a), got to speak with COL(P) Mike Garrett for a second time, and shared a room with Bacevich, Bob Kaplan, and MG Charles Dunlap without doing anything crazy or illegal. Good times!

---

Possibly related posts: (automatically generated)