Wednesday, June 30, 2010

The foundational flaw of our Afghanistan strategy, universally ignored by its proponents

Supporters of the current approach in Afghanistan seem to have settled on one primary rationale for the strategic necessity of this war: if we leave Afghanistan, it will again become a haven for terrorists to find sanctuary, train, and prepare to launch attacks against Americans. Leaving aside for a moment the reality that such a sanctuary already exists on the Pakistani side of the Durand Line, let's consider the implications of this rationale.

If we start from the assumption that Afghanistan's territory must be controlled by someone who will deny sanctuary to terrorists (as George Packer, Ross Douthat, and John Nagl have all recently insisted it must), we're basically left with two options: control Afghanistan with U.S. forces, or build Afghan capacity to the level required for the host government and ANSF to execute this task with minimal support. Our purported current approach is to do the former in the near term, while simultaneously accomplishing the second for a transition of responsibility over the medium to long term.

Douthat rather disingenuously suggests that the administration "hasn't been choosing between remaining in Afghanistan and withdrawing from the fight. It's been choosing between two ways of staying": COIN and Biden's "CT-plus." This is a curious way of framing the options, and one wonders if Douthat believes that the limited U.S. presence in Iraq after this year similarly constitutes just another "way of staying" rather than a pretty significant departure. But I digress. Here's why Douthat thinks we're sticking around:
First, the memory of 9/11, which ensures that any American president will be loath to preside over the Taliban’s return to power in Kabul. Second, the continued presence of Al Qaeda’s leadership in Pakistan’s northwest frontier, which makes it difficult for any American president to contemplate giving up the base for counterterrorism operations that Afghanistan affords. Third, the larger region’s volatility: it’s the part of the world where the nightmare of nuclear-armed terrorists is most likely to become a reality, so no American president can afford to upset the balance of power by pulling out and leaving a security vacuum behind.
And so how do we avoid leaving that vacuum? Through the COIN approach, Douthat suggests, "which seeks to lay the foundations for an Afghan state that's stable enough to survive without our support."

But it doesn't just have to survive, does it? It needs to be capable of denying its territory to terrorists. As we should all know by now, governments that are friendly to terror organizations aren't usually the problem -- it's governments that are too weak to do anything about those organizations that are the real trouble.

So if these are our worries -- the memory of 9/11, a base for CT operations, and "the nightmare of nuclear-armed terrorists" -- then any U.S. departure is going to be contingent on creating a government and security force that can secure our interests in these areas to the same degree that U.S. presence can, right?

Here's the thing: is there anyone who believes that the Afghan government will be capable of functioning in even the limited capacity required to deny sanctuary to those committed and confirmed terrorists who already exist within, say, the next decade?

If so, why?

As you'll no doubt have heard by now, the Special Inspector General for Afghanistan Reconstruction (SIGAR) released a report yesterday on the status of international efforts to train ANSF. The report was not good: it concluded that the capability rating system being used by ISAF to evaluate ANSF "basically has not been a dependable system," seeing as "even top-rated Afghan units could not operate independently and that the ratings of many security forces overstated their actual capabilities." NATO Training Mission-Afghanistan (NTM-A), the ISAF organization charged with the ANSF training mission, has since changed its metrics and overall rating system to try to get more useful data.

In his Senate confirmation hearing yesterday, GEN Petraeus addressed the report and tried to "clarify" in a manner that downplayed the significance of SIGAR's conclusions. The old Capability Milestone (CM)-based system conveyed information about the levels of manning, equipping, and training of Afghan units; that is, the system that was being used for the first NINE YEARS OF THIS WAR was based on how well the Table of Organization and Equipment was filled out, or how many boxes you could check on a training plan. The CM system didn't tell you anything about the actual FIGHTING ABILITY of the unit being rated. (For metrics geeks, you could say this was a Measures of Performance-oriented rating system, rather than Measures of Effectiveness.) And apparently it took a SIGAR report for us to figure out that this wasn't a terribly effective mechanism.

But that's a bit of a digression, isn't it? Our rating system isn't exactly the point here -- the capability of the ANSF is what we should care about. And size, but in this context, size is really only significant as a factor in determining capability to accomplish a certain mission set or secure a defined end-state. So this becomes a sort of a math problem: can 400,000 ANSF (ANA + ANP) do the business, assuming we get to the target size within the expected time period?

Can 400,000 ANSF deny sanctuary to al-Qaeda? Can 400,000 ANSF ensure that America won't be attacked by individuals who train or plan their attacks from Afghan territory? (We already know they can't do anything about the AQ folks on the other side of the Hindu Kush, but again, we're going to leave that aside for now.) Can 400,000 ANSF do the job that 130,000 ISAF troops plus ~245,000 ANSF are currently doing? So, basically, what's the time frame on which we can expect ANSF to effect a one-for-one, straight-up capability match with the ISAF troops who will be departing the country?

Seriously: is there anyone who thinks this can be done in a decade?

And if not, then why is this our strategic concept? Why aren't we working on some other plan to mitigate the nearly certain shortfalls that will exist when U.S. troops pull chocks and head home?
Or perhaps the better question: why do we insist on repeating the tired and utterly unsubstantiated line that Afghanistan will be vital to U.S. national interests until its territory can be wholly secured against the presence of terrorists and militants?

33 comments:

  1. This was one of the main arguments in favor of the CT approach from last year (which, to me, gets the strategic picture spot-on).

    First of all, is there any evidence that AQ will return to Afghanistan if they already have a safe haven in Pakistan?

    Secondly, even if the Taliban did take over in Afghanistan, would they still be friendly towards AQ? The relationship between the TB and AQ has been shaky at best...

    ReplyDelete
  2. Starbuck -- Agree with your two points, of course, though I left them aside in the original post so as just to talk about capability.

    But if you ask George Packer, then there is NO DOUBT AT ALL that U.S. departure in large numbers would result in THE WORST OF ALL POSSIBLE WORLDS. After all, David Rohde's Taliban captors said some al-Qaeda-ish things once!

    ReplyDelete
  3. Again, why is denying some territory to al Qaeda (however defined) on the left side of the Durand line instead of to the right is of vital importance to the US?

    Some might quibble over this is an important goal. Others might concede it has some relevance, but only for certain costs. Others might contend that ISI's need for strategic depth will lead Pakistan to do that which it is in Pakistan's interests to do, and al Qaeda will simply exploit the depth that is secured in the ungoverned stretches of NW Pakistan or eastern Afghanistan.

    But we shouldn't simply assume that this is a vital need, even if others loudly proclaim that it is.

    ReplyDelete
  4. Anon @ 1142: But we shouldn't simply assume that this is a vital need, even if others loudly proclaim that it is.

    Completely agree. My point was to isolate the capability angle and demonstrate just exactly how bankrupt the entire justification is, not simply because it's a poor rationale, but because even if you grant its necessity, it's basically impossible.

    And strategy is about the art of the possible.

    ReplyDelete
  5. Gulliver, I'm not sure that's even a measure of performance, it may only be a measure of effort (yes, metric geek, guilty, and it's only getting worse...).

    ReplyDelete
  6. For what it's worth, the fall 2009 DOD report that Petraeus quotes ("manning, equipping, and training" only) on the CM system had a very tenuous relationship with the reality of the CM system up to the spring of '09 (up to the end of the McKiernan era, when I left theatre, anyway).

    Below are the (unclas) metrics I was supposed to evaluate ANA S2 performance on in my reports at the time. Evaluators wanted to know, could they:

    *Develop humint sources and fuse humint into operations
    *Provide situational understanding of the enemy to higher, lower, own HQs
    *Provide reasonably accurate understanding of enemy capabilities
    *Track historical and background information on threats within their sector
    *Determine threat patterns and trends
    *Employ and manage the brigade recce company effectively
    *Provide intelligence to enable ops through IPB
    *Develop enemy COAs as part of planning
    *Conduct intelligence gathering from detainees

    These are obviously not just manning/equipment/training questions: we weren't just counting heads, or jeeps, or course reports to answer the questions above. So you have to conclude that up to a year ago either the intent of the system was being wildly misread by the actual people charged with applying it, myself included, or alternatively that these are post facto rationalizations for why the system used to evaluate Afghan units up until this year was in general terms an inaccurate predictor of combat performance and generally tended to overstate true ANSF combat competence.

    Put another way, the current explanation "Capability Milestone" system was not actually meant to evaluate "Capability" was not fully hoisted in by those actually involved in using it, pre-McChrystal.

    ReplyDelete
  7. Should have previewed. Last sentence should have read:

    "Put another way, the current explanation *that the* 'Capability Milestone' system was not actually meant to evaluate 'Capability' was not *getting* fully hoisted in by those actually involved in using it, pre-McChrystal."

    ReplyDelete
  8. Abu M's short Tweet (in response) beats this long blog opus, I'm afraid. The flaw involves our bipartisan policies TO THAT PLACE WHICH SHALL NOT BE NAMED, ever, by any Beltway Sahibs.

    Basically, I'm with zenpundit and Pundita on this. Yak all you want about COIN vs. CT: that ain't the problem.

    Interesting points, BruceR!

    ReplyDelete
  9. COIN is the way out. We implement the strategy, draw down and the leave at the appropriate moment. It is only then that we find out whether it has worked.

    The only other options are immediate withdrawal abandoning are obligations to the Afghan people a guaranteeing at least as big a mess as there was before or switching to a more overtly kinetic strategy, 'fighting the enemy over there rather than over here' with the added implication of indefinite occupation.
    The ISI is, of course, the Gorilla in the room, perhaps we would do well to invest more in a lasting peace between India and Pakistan.

    To me, it seems that the big failure in training the ANA and ANP is that we've been trying to create competent security forces, from scratch, out of an undereducated and undernourished population and on the cheap. I do not see how we can expect individuals and units to perform as well as ISAF troops unless they have the same training and ethos instilled in them.

    ReplyDelete
  10. "The ISI is, of course, the Gorilla in the room, perhaps we would do well to invest more in a lasting peace between India and Pakistan."

    Oh Gawd. I'm sure the Beltway Sahibs will handle such a brief with distinction, delicacy, and care.

    Sharm-el-Sheikh anyone?

    ReplyDelete
  11. Ack - to be fair Matthew Doye, I agree with your larger point.

    ReplyDelete
  12. One last run at CM:

    Note also the Nov/09 DOD report Petraeus cites actually says something like "the CM rating only measures manning and equipping and training and thus units that are rated low could still actually be much MORE competent than the stated definitions" (CM1=able to operate independent of ISAF, save enablers, etc.) In other words, nine months ago, pre-Marja, the official DOD line was "really, the Afghans could be much better than our evaluators are giving them credit for."

    So now, after it's clear that might be questionable, after multiple rotations have reported no difference between an ANA CM1 unit and a CM3 unit in its effect on the fight, that exact same quote, that previous DOD justification for why ratings for many units were still so POOR after several years of work in army building, has now been turned around to say that we were never really measuring ANSF combat effectiveness all along, anyway.

    In sum, DOD's official line in response to SIGAR is to effectively concede that all the ANSF metrics that had been issued prior to this point, and relied on heavily by Cordesman, O'Hanlon, the Canadian government, previous briefings to Congress, etc., in their war-progress reports were essentially misconstrued, if not totally worthless.

    Oh, and the brand new evaluation system (rolled out very shortly after COMISAF got its first briefing on the coming conclusions of the SIGAR report) ... is classified. Check.

    This really was great work by SIGAR, I have to say. This is what inspector general's offices are supposed to be for.

    ReplyDelete
  13. Bruce -- Great input, thanks for chiming in.

    Madhu -- Abu M's short Tweet (in response) beats this long blog opus, I'm afraid. The flaw involves our bipartisan policies TO THAT PLACE WHICH SHALL NOT BE NAMED, ever, by any Beltway Sahibs.

    For the benefit of everyone else: Ex's response was "I think the flaw starts with P and ends with Stan." I don't think, however, that he was making the same point Madhu is.

    So here's my question for you, Madhu: what do you recommend we do? Do you think the solution is to take military action against Pakistan?

    Matthew Doye -- COIN is the way out. We implement the strategy, draw down and the leave at the appropriate moment. It is only then that we find out whether it has worked.

    No offense, but this is absolutely preposterous. Military planning isn't (and can't be) based on the idea that we come up with our best guess, cross our fingers, and execute.

    The only other options are immediate withdrawal abandoning are obligations to the Afghan people a guaranteeing at least as big a mess as there was before or switching to a more overtly kinetic strategy, 'fighting the enemy over there rather than over here' with the added implication of indefinite occupation.

    Those, quite obviously, are not the only other options. That aside, nothing you've written even comes close to addressing the question I've posed: why are we doing this if not even the most ardent proponents can say with a straight face that it's going to work?

    ReplyDelete
  14. "So here's my question for you, Madhu: what do you recommend we do? Do you think the solution is to take military action against Pakistan?"

    Hilarious Gulliver. What constitutes military action?

    No, no, no: Quit dumping aid into that joint: military or otherwise. I've been consistent on that point. I may be wrong, but I'm consistent.

    Also, have it in the back of your mind that maybe, just maybe, we are not playing on the same page, no matter how nice General Kayani is to General Petraeus. We have different national goals. So, recognize it, plan accordingly.

    Sucks about the logistics, though.

    ReplyDelete
  15. Yes, of course, we have different goals and interests. So what happens when we "quit dumping aid into that joint"? How does that move us toward our desired endstate in south Asia?

    ReplyDelete
  16. I give up.

    No, wait.

    I never give up.


    You tell me: How is aid moving us toward a desired endstate in South Asia, and what is that desired endstate? Stability in Pakistan and trade with the region? A bulked up India that is a counterweight to China? No more hosting nasties and sponsoring proxies?

    What is the proof, data, metrics that our aid is doing what we say it's doing? What's the latest social science, poly sci mumbo jumbo on that? Every year, another article about how the money went where we didn't want it to go. So, show me the data that aid is working.

    Do we even know where the money is going? Is Pakistan more stable because of all the years of Western aid? Why is cutting off aid to a dysfunctional government so scary to you solutions industry people? I know I'm being mean, but I'm pissed!

    (Not at you, of course. Just generally upset. Maybe I should take a break from the COIN-y blogs. Maybe this stuff ain't good for me. Kudos to you that do it for a living. I'd slit my wrists.)

    ReplyDelete
  17. Fair questions all, if the point of this post had been to talk about aid effectiveness or Pakistan's role in U.S. counterterrorism policy. But alas...

    I asserted that the fatal flaw of our Afghan policy is that it can't possibly work, even in an optimistic telling of the future.

    You suggested that the real problem, in fact, is Pakistan.

    So here's what I want to know: is the denying-terrain-to-terrorists mission not that important (my view), or is there a way to do it that "solves the Pakistan problem"?

    I think that if you're in the business of trying to deny terrain to terrorists, it's going to take Pakistani cooperation (or at least complicity). Maybe you disagree, or maybe you think that the terrain-denial mission isn't that important. I'm just curious about how you dramatically change our policy towards Pakistan and still execute that theoretically vital mission.

    ReplyDelete
  18. Okay, last one for the day (hey, it's a blog comment section - it's the appropriate place to play around with ideas):

    http://zenpundit.com/?p=3444

    and

    http://ramanstrategicanalysis.blogspot.com/2010/06/enhanced-chinese-interest-in-pakistan.html

    (So, let's end on where we agree: if we "dialed down" things in Afghanistan, as zen says, then we wouldn't need to be so engaged in Pakistan. Not abandoning the areas as previously, just more low key. How's that for a point of discussion? Rebuttals always welcome.... :) )

    ReplyDelete
  19. Wait, I reread your last comment? What? I don't get it?

    Sincere question.

    - Madhu

    ReplyDelete
  20. Gulliver,

    As I read through your post, nodding my head in agreement at almost every turn, I was surprised you missed the obvious example - which is Pakistan. The very best outcome, from a governance standpoint, is an Afghan government that approaches Pakistan in terms of stability and capability. There is nothing to suggest such an Afghan government would be any better at policing its territory than Pakistan is able to do currently, and that's assuming a best-case scenario. And so the reason that sanctuary denial via governance is hopeless is because Pakistan, for all it's capability, isn't able to prevent a sanctuary within its own borders. What makes anyone believe that an Afghan government, given ten or even 50 years, could do better?

    ReplyDelete
  21. Yeah, yeah, yeah. I always break my "last comment" promise several times over.

    So, if I'm wrong about the aid thing (hey, I probably am, I've gotten so many things wrong in the past that I know for a fact the public policy person lifestyle is SO not for me:)


    "10.The US finds itself in a thankless situation. The more the aid it gives to Pakistan to deal with the terrorists, the more the incentive for Pakistan to keep the terrorists alive and active to keep alive the fears of the US. If it reduces its aid to Pakistan, there is a danger of Pakistan not doing even what it is doing now to deal with the terrorists.


    11.The only way the US can get out of this vicious circle is by taking in its own hands the responsibility for destroying the terrorist infrastructure in Pakistani territory instead of depending on Pakistan for this."

    (B. Raman again.)

    Yikes. That sounds awful. Hmm, maybe the COINistas are right! There are no good options and COIN is the least bad one!

    ReplyDelete
  22. Andy -- Sure, Pakistan is a fair comparison, but not a watertight one: there are questions about Islamabad's interest in waging a COIN campaign, never mind capability. But generally, yes, it's fair to question why Afghanistan should be any different.

    Madhu -- Maybe we've talked this one into the ground, but I don't get what you don't get. My question for you is this: if you think it's essential for us to be in Afghanistan to deny possible sanctuary to terrorists (and I don't know if you DO think we should remain in Afghanistan in significant numbers, or if this is your rationale for why we should), then how exactly is it that you expect what we're doing there to accomplish this end, Pakistan or no Pakistan? Is your point to say that we can't do it without Pakistani cooperation or complicity, which we're not currently getting, even with significant aid and a friendly foreign policy? If so, then we agree. (That said, I think our "friendly foreign policy" towards PK could be less schizophrenic, more focused on Pakistan's legitimate interests, and less plainly pro-non-Muslim -- at least in Congress -- than it currently is. But I digress again.)

    ReplyDelete
  23. 11.The only way the US can get out of this vicious circle is by taking in its own hands the responsibility for destroying the terrorist infrastructure in Pakistani territory instead of depending on Pakistan for this.

    Here's the abridged list of worst U.S. security policy ideas in the world:

    1. Take military action (beyond what is currently taking place, and even then only with continued tacit approval/complicity from the Pakistani government) in Pakistan

    2. Anything Elliot Abrams says

    3. Invade Iran

    ReplyDelete
  24. Gulliver - we have talked this into the ground. I'm saying the hypothetical Pakistan or no Pakistan makes no sense. If Afghanistan were in a different neighborhood, we'd have a completely different scenario. So I don't understand how you can divorce Afghanistan from her physical neighborhood. That's what confused me. There is no "no Pakistan," in this scenario. So why bring it up?

    Finally, I have no clue. Good luck General Petraeus!

    ReplyDelete
  25. Sure, Pakistan is a fair comparison, but not a watertight one: there are questions about Islamabad's interest in waging a COIN campaign, never mind capability.

    Pakistan has never "controlled" it's border areas. They are still administered under the same code use by the British before them which relies on patronage to maintain order. That system is in the process of breaking. There is no Pakistan rule of law in these areas. These areas are de-facto colonies even though maps show them as integral to the Pakistani state. For the Pakistanis to directly control these restive colonies would require conquering them. It's not simply a question of Pakistani will - though that is obviously questionable.

    ReplyDelete
  26. Andy -- Yeah man, I'm familiar. Maybe "will" wasn't exactly the right word, but I'm using it to encapsulate both intent and desire. What you've written about the FATA is precisely why Pakistan isn't a perfect comparison: the Pakistanis have proved both unwilling to and incapable of ending the terror threat through territorial control.

    ReplyDelete
  27. Heh, turns out I didn't even use "will" in the first place, I wrote "interest." Anyway, I think you know what I mean.

    ReplyDelete
  28. Yeah, like I said, I agree with you. My only point is that I think you could have highlighted the comparison in your argument. While the comparison isn't perfect, obviously, factors that inhibit territorial "control" are operative in both countries. There is very little to suggest an Afghan government would be superior to Pakistan in this regard.

    ReplyDelete
  29. "... the system that was being used for the first NINE YEARS OF THIS WAR was based on how well the Table of Organization and Equipment was filled out, or how many boxes you could check on a training plan. The CM system didn't tell you anything about the actual FIGHTING ABILITY of the unit being rated."

    Well, to be fair to ISAF, that is pretty much the half-baked system that our own forces use. That is the root of my disdain for our task-condition-standard approach.

    ReplyDelete
  30. Well, to be fair to ISAF, that is pretty much the half-baked system that our own forces use. That is the root of my disdain for our task-condition-standard approach.

    True, and I feel you on that one.

    ReplyDelete
  31. If the Taliban regain control of Afghanistan, but are not friendly towards al-Qaeda, would this be an acceptable outcome? Strategically? Politically?

    ReplyDelete
  32. Late addition to the discussion, wich has been interesting. The points about the evaluation parameters for the ANA was an eyeopener.

    One idea Ive been spouting for a long time is that there should be a UN force with real capability recruited from muslim countries. Or a Foreign Legion. I always wondered why there were no Indonesian security companies involved instead of Blackwater as well..

    ReplyDelete