Exum, Broadwell, and COIN Metrics

In a recent post, Andrew Exum noted that though strategic challenges were quite serious in Afghanistan that “Counterinsurgency, as practiced at the tactical level, is the best I have ever seen it practiced.” I have to admit, I have no idea what this means.

Exum then had an exchange with Foust on this issue, which, to be blunt was not really illuminating.

What I presume it means, in practice, is that American forces were conducting operations and engaging the population in ways that match Exum’s subjective understanding of what constitutes good counter-insurgency. Or maybe it means COIN practices in line with doctrine? Frankly, the issue remains unclear to me.

Regardless, in order to assess utility of this argument, we need to explore three lines of reasoning.

First, is there any particular reason to believe that Exum has uniquely powerful insights about the conduct of counter-insurgency. He calls himself, “a specialist in small wars and insurgencies.” Is he?

I respect his service to our country. But in terms of his service at least, I find it hard to give him any particular credit.  He was a junior officer, serving most of his time as a lieutenant. I don’t think Exum ever even had a company command.  Did he?  I don’t want to dismiss his expertise on the ground. Two tours as a platoon commander is two tours more than I have, I admit. But it is also much less grounding than many, many, many people.  Is every retired 0-3 a specialist in small wars and insurgencies by dint of service in Iraq and Afghanistan? Col. Harry Tunnell has more experience than that — and also two rounds of PME under his belt — is he also an expert in small wars and insurgencies?

Again, I genuinely respect Exum’s service. But it is just an insufficient grounding to simply give him the benefit of the doubt. Beyond that, Exum has written a memoir, a blog, and a number of think tank papers which range from awful to good.  He’s read the basic literature, as far as I can tell. But his understanding of the academic literature on conflict is quite limited. He’s a reasonably sharp person.  He has outstanding access. But there is nothing in his career to suggest a unique insight into the nature of insurgency or the conduct of counter-insurgency.  In short, he’s not an “authority” on whom we can confidently rely. Or if he is, then what do we do about the literally thousands of other authorities out there?

Since, he is not an “authority,” the question then becomes whether he actually has any useful or systematic data to bring to the table.  The answer here is, sorta.  Exum frequently cites data, but he does so in the manner of someone who is repeating talking points, not making a systematic argument.  He never specifies his propositions in a clear matter, nor does he explain why the data he is citing is relevant.  This is part of a bigger discussion of metrics in counter-insurgency.  The short version is that there is a real metrics problem because many of the key propositions remain contested, others tautological.  Again, at the risk of oversimplifying, it remains unclear what data is relevant.  A few examples suffice to highlight this point:

–Does an increase in the level of violence in a given area reflect a strengthening of an insurgency or does it reflect taking the fight to the enemy?  Or does it depend?  And if it depends, what does it depend on?

–We think we know that increased collaboration between COIN forces and the local population is an important element of success.  But what if collaboration increases and insurgent activity also increases? As a practical matter, can you just count the number of turn-ins and tips? Or should collaboration only be considered positive it is seems to reduce areas of insurgent control or operations? If collaboration grows slower than insurgent activity, how is that a net plus?  In theory, the more tips you get, the more it demonstrates support from the local population. But in a complex conflict environments – as characterize most civil conflicts – how does one distinguish between tips that reflect a commitment to central government control and those that reflect political jockeying among various local factions? The latter sort of tip, of course, can have a destabilizing rather than stabilizing effect since it effectively drags COIN forces into local feuds.

–How much of our impression of better intelligence reflects, actual, you know, better intelligence? Exum is impressed:

“Our intelligence at the tactical level is greatly improved. Eighteen months ago, as I traveled around Afghanistan for the former commander here, intelligence officers were outstanding in terms of providing information on the enemy: size, disposition, composition, most likely course of action, etc. When it came to providing political intelligence on “white actors” or explaining local tribal dynamics, though, most intelligence officers did not have much to offer. What a difference 18 months makes. This time around, when an intelligence officer began a briefing, he or she usually began by explaining the human geography of their area of operations and only later focused on the insurgency as a part of that human geography. I am so impressed with how sophisticated the analysis provided by intelligence officers today is when compared with not too long ago.”

But how does anyone know whether that answer is actually correct?  It is almost certainly true that we are now gathering on a wider range of what we believe to be relevant data. But I just have no sense of whether this data is accurate, or for that matter even relevant.  At the end of the day, doesn’t all this “better COIN at the local level” need to translate into something concrete?  And if so what?

So, Exum is neither an authority on his own, nor does he demonstrate a particularly sophisticated understanding of the challenges of empirically validating his claims.  The issue is not that he right or wrong. I really have no idea. The problem is that he isn’t convincing. Other visitors to Afghanistan have come back with more skeptical assessments.  Still others with still more positive ones.  At the risk of making light of the issue, it does seem to me that the divide between the two camps is at least in part a function of access with ISAF. The more access the more positive you are, on the whole.  Why is that?

There are multiple competing interpretations, which I am sure are obvious to one and all. ISAF has better data. Or they might just have better talking points.  Analysts might be shading their views to maintain access, or even doing so subconsciously.  Or they might be resentfully expressing skepticism because of being denied access.  All I know is that when O’Hanlon, Exum, Biddle, the Kagans, etc. they come back talking of challenges and progress.  And when Michael Cohen or Josh Mull go they come back with their skepticism reinforced.  As an outside observer on this dynamic, I can only note that none of this fills me with great confidence in the analytical judgments of any of the participants in the debate.

I wish I knew how to resolve any of these complicated data issues in order to make better judgment.  But I just don’t.  The reality is that all of our metrics and interpretations of them are based on poorly defended theories of what matters.  Which is fine.  That is the nature of the beast.  All we can do is make judgments.  But the problem is that there are a lot of lives and money on the line in making the right calls, and we’re just fooling ourselves if we think we actually understand any of this in a systematic manner.

Last point… Exum’s arguments are particularly pernicious in a way.  By trying to suggest that local, operational successes can be  assessed independently of strategic outcomes he is reinforcing the main problem in our discussions of COIN.  Look, good COIN is COIN that defeats the insurgency.  Any use of force has to be measured, ultimately, by the political outcomes it achieves.  This is particularly true since the nature of these conflicts are intimately political.  You can judge whether, say, an amphibious operation was successful based on whether you establish a bridgehead, by the number of losses, etc.  But what in the world does it mean to do “good COIN” if insurgent activity or control expands in the process.  You just can’t assess it independent of the outcomes.

This is a particularly problematic issue because, I would argue, much of what passes for COIN theory reflects little more than the reification of perceived operational needs.  If I send troops into a region to fight an insurgency, the only way they can plan operations is with some local intelligence and cooperation.  So, in order to allow tactical units to operate, then need to build bridges to the local community, and the best way to do that is by — on one hand — offering goods and services and — on the other — making collaboration somewhat safer.  I get all of that.  But what remains unclear to me is how this turns into strategic success, and here I think the theory becomes much, much fuzzier.  There is just this assumption that collaboration is somehow linked or caused by or causes something called “legitimacy,” which I am not sure anyone can define with great precision.  And then there is another assumption that this “legitimacy” will manifest itself in diminished prospects for the insurgency.

Which all gets me back to Exum, who still seems to see the world from the vantage point of his platoon. Where the goal is kicking in the right door without getting the men under his command shot up and without causing collateral damage that will provoke additional attacks on his men. Got it. Admirable. Good to have some people in a position of influence who do have that vantage point.  But it is crucially important to be able to differentiate between tactical/operational success and strategic outcomes. I don’t think Exum – or actually any of the other COIN proponents do that very well.

Now, about Broadwell, who I suggested was guilty of a “banality of evil” in her reporting about the destruction of an entire Afghan village. In the comments to that post, Carl Prine took me to task:

There probably isn’t anyone more concerned than I am about the blurring of roles that we find with Broadwell and Ricks (are they journalists? Pundits? Flacks? Salespeople for a certain CNAS brand of warmaking?).

But I’ve come to know Paula and I’ll something in her defense: She’s honest. Regardless of what you think of her policy prescriptions and analyses, they’re sincerely rendered. She’s also not a professional reporter, so her initial attempts to convey meaning from her jumbled efforts shouldn’t be judged too harshly.

I can recall being a cub reporter after leaving the USMC and fumbling in front of many thousands of people. It’s not easy. She is trying.

Part of that process likely will include showing more empathy for all the affected peoples in these reports AND a bit less trust in official sources. But you must give her a chance and not write off what really is someone trying hard to understand a very complex milieu as Arendt’s freighted “banality of evil.”

Paula isn’t in a vile conspiracy with Eichmann, Bernard.

You’re a teacher. She’s a PhD candidate. Rather than harshly criticizing her and ridiculing her, why not reach out to her and use your many years of experience as an instructor to help her do it better?

I can assure you that she’s worth that effort and that she’s not evil. Even if I profoundly disagree with some of her methods and conclusions, she’s a decent woman.

Prine pushes several of my buttons here.

First, I resent the notion that I am somehow “kicking down” in my comments about Broadwell.  If she were indeed just a young, struggling grad student, I would certainly have shown more discretion.  Similarly, if I believed that Andrew Exum were just a young, struggling grad student, I would also be more generous in my assessments.  But that is simply not an accurate description of either Broadwell or Exum’s standing in the field  and role in the debate.  Is there anyone who believes that I somehow have a greater standing or reputation than they do?  Indeed, I think I flatter myself when I claim to be their equals.  I am a cranky professor type, whose blog gets 200 hits a day on a good day. I don’t advise ISAF. General Petraeus doesn’t know me from a hole in the wall.  The notion that because of my elevated standing, I need to pull my punches in critiquing Broadwell or Exum is laughable.

Second, I have no idea if Broadwell is a decent, honest person. Maybe she is, maybe she isn’t. To be blunt and at the risk of straying into inside baseball kind of discussions, her reputation on that score is mixed.  But whatever.  Look, the issue with the “banality of evil” is precisely the way in which ordinary, decent people can becomes so invested in the goals of their state that they cease to cast a critical eye on acts that would otherwise trigger moral alarm bells. Seeing an entire village destroyed and then characterizing the anger of one of the villagers as “theatrics” falls into that category. Sorry, but it was an appalling dispatch from the field.  Morally bankrupt, not to mention operationally suspect.

7 comments to Exum, Broadwell, and COIN Metrics

  • TJM

    If someone renders an opinion unsupported by footnotes, data, or expertise, then the opinion is likely for purposes of influencing public opinion, rather than helping to refine policy or strategy. So while I kind of cringe at the personal nature of your post, I can’t help but agree with you and I also think that your response is appropriate. You have not argued “Exum is wrong and here is why.” Instead, you’ve effectively put forth an argument that would reduce the likelihood that his argument would influence the target audience by simply rebutting, “Wait, who is this guy? What makes him an expert? And where is his evidence?” (Though your message will likely have a smaller audience and thus a small impact).

    You know things aren’t going well when the debate focuses on people rather than ideas. But that’s where we’re at. What we do in Afghanistan will not be guided by what makes the most sense for our national interests or for Afghanistan’s. We will do whatever we perceive is the path of least resistance through a swamp of public opinion, and constraints & limitations on our military and diplomatic power. And that will be more significantly impacted by influencing opinion than by improving our understanding and refining our policies.

    That’s my non-expert opinion.

  • Carl Prine

    Bernard, I suspect that I understand the “field” of COIN pretty well.

    Steve Metz, Dave Betz and anyone else who rhymes with “etts” probably would be members of the council of COIN cardinals, even if they’re cynics. Paula isn’t nearly in their company.

    That is — unfortunately — now painfully evident.

    She’s a PhD candidate, someone with commercial interests in pursuing an apparently authorized hagiography of Petraeus, a reserve major in the US Army and a stay-at-home mom. She apparently now wishes to be something of a journalist.

    Well, she ain’t one of those yet.

    She is not in the same position as Exum (he is paid by a think tank to provide often callow advice, typically on behalf of messianic military figures). Nor are her credentials or arguments even as impressive as her bete noire, Foust.

    I publicly have been very critical of Paula’s Afghanistan pieces. I have spoken as someone who takes war reporting very seriously, and I think she did a very poor job at the most rudimentary tasks of conveying information to a general readership.

    But I also believe that I should salt that concern with a generous bushel of help. If she’s trying to become something of a reporter, she needs an apprenticeship in the ethics and craftsmanship of the trade (alas, Tom Ricks wouldn’t be the first of my choices to instruct in either of those salient competencies; to put it mildly).

    I reached out to her on that and she ended our friendship. So when I advise you to attempt it, it’s not a naive perspective but rather one informed by a song of experience. It likely will be rejected, too. She’s doubling down on wrong and clumsy, but she’s an adult and we must watch it play out before an increasingly larger audience.

    If I erred in my suggestion, it was merely because I see you as a far more learned and important figure in this debate than she (or Ricks) is. If this is my sin, I hope that you’re flattered by the trespass.

    As for “banality of evil,” I’m more than familiar with the term and I believe you pushed it too far with her for rhetorical purposes. It carries connotations that, in this instance, I think are unfair.

    I stand by that.

  • Carl Prine

    And I’ll say something about Exum, too.

    While I’ve had profound disagreements with him, I still respect him. Not for his wartime service, although it was honorable, but because I believe (perhaps wrongly) that he’s committed to discovering the truth — admitting along the way that he is sometimes wrong — and that he strongly feels that what he’s doing will save lives, not only of Afghan citizens but also of US troops.

    Perhaps I’m getting soft, but I have a bit of pity for those who breathe their ideas through the lungs of the blogosphere. Not every thought put to bytes is a good one, but it’s there for everyone to ridicule. I’m probably wrong about Exum, but I would rather have him swinging at the plate everyday than standing in the box watching the pitches go past him.

    He’s also done something that I’ve long advocated for the COINdinistas, perhaps because I did it myself: Go live with guerrillas.

    The hardest but most important task to teach a commander or analyst is the craft of empathy. Broadwell obviously has yet to learn it. But Exum’s work on HezbAllah actually is quite promising. Had the US invaded southern Lebanon, he probably would sound more interesting when he wrote out his thoughts.

    I can’t put Broadwell into that category. Not yet. But she’s new at this.

  • Carl:

    I think those are fair points. And it is certainly not my intention or desire to silence Exum. I agree that he has a lot to contribute — though, I wish he would be as honest and candid in public as he is in private. I genuinely loathe hypocrisy. But more importantly, I think he’s a guy who needs to be pushed. Hard. He’s risen so far, so fast that he is no longer conscious of his own limitations. Much of it has gone to his head, and it is perhaps the biggest threat to his development as a thoughtful analyst. I agree that his work on Lebanon seems quite promising.

    As for Broadwell, I doubt she cares what I think. I’ll stand by my characterization of her as a significant figure in the COIN world. Google demonstrates that Broadwell has more visibility that Betz and Metz combined. That is what being a Petraeus crony will do for you.

  • Carl Prine

    I have a rule when quoting experts, Bernard. When I know more than the person I’m interviewing, then he or she ain’t an expert.

    I don’t know as much as Betz or Metz about COIN. My blind cat Cosmo, however, knows more about the subject than Broadwell.

    Ricks and Broadwell and Linda Robinson and others might be the best known by a popular audience, but they obviously aren’t experts on the subject. One readily understands that when one is forced to read them (which I think is considered an alternative form of penal servitude in Utah).

    Perhaps you shouldn’t attempt to conflate their reputations by engaging them as equals or betters? I don’t know the answer to that, but it sure seems to me that you’re punching down. It might help if the community of COIN scholars would police their own, but professors and commissioned officers can be so clemently collegial.

  • [...] dittoing Bernard Finel's argument on the COIN tactical to strategic disconnect: By trying to suggest that local, operational [...]

  • mclaren

    Really excellent article. Your discussion of the slippery meaning of these metrics brings Vietnam to mind. There seems to be a Western Euro-American tendency to try to reduce war to numbers, and, if the right numbers get hit, then the military conflict gets labeled as a “victory.”

    This actually seems to work reasonably well for Euro-American historical conflicts, oddly enough. Lanchester’s Linear Law seems to accurately describe many historical Western conflicts (unaimed fire; but also applies to random artillery barrages), while Lanchester’s Square Law seems to apply well to modern warfare (aimed fire).

    But these mathematical laws don’t seem to apply that well to non-Western conflicts, or to certain Western historical wars. Napoleon’s campaigns in Russia and Spain, for example, provide instances where Napoleon won every engagement but ultimately lost the conflict. Examples of modern 4GW like Mao’s Long March, in which Mao’s forces lost some 90% of their men but nonetheless eventually prevailed, suggest that these kinds of mathematical measures of modern warfare can be very slippery.

    Martin van Creveld has suggested (“The Transformation of War”) that the advent of nuclear weapons represents a disonctinuity in the history of warfare. After the H-bomb, conventional wars can no longer be waged in the same way or use the same assumptions as pre-1945.

    This leads to some real questions about the meaning of these kinds of COIN metrics. If violence goes down, does it mean COIN was effective…or does it merely mean that the U.S. army successfully bribed the insurgents to lay low until the Americans leave? Or does less violence perhaps mean that both sides in a civil war have ethnically cleansed one another into separate enclaves? What does victory mean in COIN? If U.S. forces declare victory over the insurgents but a virulently anti-Western fundamentalist Islamic regime takes power six months after the Americans leave, is that really a victory for COIN?

    These kinds of COIN metrics seem designed to create a series of numbers that field commanders can “hit.” Once the right numbers get hit, victory is declared — conveniently without having to worry about the end result on the ground in Afghanistan. Maybe metrics have become popular because they allow us to declare victory by the numbers in the midst of what is an inevitable strategic American defeat?

    Maybe everyone from Petraeus on down knows full well that there’s zero chance that Afghanistan is going to develop a functioning Western-style democracy or any kind of recognizable civil society with property rights, etc., so they lay out a set of metrics everyone can point to as a substitute for victory. Maybe Douglas MacArthur was wrong: maybe the substitute for victory in COIN is metrics.

Leave a Reply