The Incoherence of COIN Advocates: Stephen Biddle Edition

Stephen Biddle is the single best defense analyst working today. His arguments are usually carefully considered and well supported empirically. For a generation of younger defense intellectuals, he is very much the gold standard, the model to emulate.

His recent essay in the American Interest (Is It Worth It? The Difficult Case for War in Afghanistan) has been widely cited as the best defense for expanding the American commitment there.  The problem is that while Biddle claims that the decision is a close call, it is only close by virtue of what can only be described as sloppy reasoning.

There are three key problems with Biddle’s essay.  First, his definition of American interests in Afghanistan is incoherent.  Second, he bolsters his case by arguing against a strawman.  Third, he makes the bizarre assumption that being better at counter-insurgency (COIN) is the same as being good enough at it to win.  I will deal with all three in turn.

The Stakes

First, Biddle argues that our interests in Afghanistan are:

The Stakes: The United States has two primary national interests in this conflict: that Afghanistan never again become a haven for terrorism against the United States, and that chaos in Afghanistan not destabilize its neighbors, especially Pakistan.

Biddle’s rhetorical trick is frankly a little seedy here.  It turns out, he does not actually believe the first interest is actually all that relevant.  He writes:

The first interest is the most discussed—and the weakest argument for waging the kind of war we are now waging


Thus it is still important to deny al-Qaeda sanctuary on the Afghan side of the Durand Line. But the intrinsic importance of doing so is no greater than that of denying sanctuary in many other potential havens—and probably smaller than many. We clearly cannot afford to wage protracted warfare with multiple brigades of American ground forces simply to deny al-Qaeda access to every possible safe haven. We would run out of brigades long before bin Laden ran out of prospective sanctuaries.

Biddle acknowledges, in short, that the counter-terrorism mission could justify interventions all over the place, and he also acknowledges that we lack the resources to achieve this mission everywhere.  So why mention it at all?  I can’t get inside Biddle’s head.  At best, listing this as the first of our “stakes” in the country is a bit of pandering to conventional wisdom.  The 9/11 connection still drives a reflexive commitment to Afghanistan.  At worst, this is an effort at waving the “bloody shirt” and generating an emotional appeal to support a weak argument.

A major challenge is that while Biddle’s arguments sounds so balanced, and he seems so willing to consider the pros and cons, it is easy to miss the core flaw with the counter-terrorism rationale, namely that it just makes no sense.  Yes, the idea for 9/11 seems to have been conceived in Afghanistan.  But the plot was executed through follow-up planning sessions in Karachi and Kuala Lumpur.  The key implementers of the plot (Mohamed Atta, Ramzi Binalshibh, Marwan al Shehhi, and Ziad Jarrah) were based in Hamburg.   Nawaf al Hazmi and Khalid al Mihdhar spent time in San Diego and took flying lessons there.  Al Qaeda was based in Afghanistan, but the plot was fully transnational.  There was no attack launched from Afghanistan except in the vaguest sense of having been discussed there and approved.  Indeed, it is amazing how little of the 9/11 plot actually required a base in Afghanistan at all.

Worse, even if we were to assume that the attack did indeed “originate” in a meaningful way in Afghanistan, the strategic argument for expanding our commitment to Afghanistan is fuzzy.  First of all, there is very little evidence that military occupation prevents terrorist plots and networks.  Indeed, there is a great deal of evidence suggesting quite the contrary.  Our experience in Iraq and Afghanistan — where insurgents have been able to build and deploy over 80,000 IEDs while under U.S. occupation — calls into question the ability of occupying forces to root out terror networks.  There is actually compelling evidence that some terrorist networks, notably groups that carry out suicide attacks, are the result of military occupations rather than contained by such deployments.

Even worse, Biddle “stake” is meaningless: “…that Afghanistan never again become a haven for terrorism against the United States…”  Never?  Not in 10 years?  How about a 100?  Or a 1,000?  Has Biddle invented a magical “control the future” button?  It would have been great if he’d been around in 1919 to establish a settlement where Germany would never again threaten world peace.  And what does “haven for terrorism” mean?  Does that mean that no plot is ever discussed there?  That no training camps exist?  That no one who has ever lived in Afghanistan be involved?  I can’t even begin to fathom how to operationalize that goal.  It is the kind of phrase one expects from a candidate on the stump, not from the foremost defense analyst of our generation.  Indeed, as an argument, it is perilously close to gibberish.

Perhaps it is unfair to go on and on about this first stake, since Biddle himself acknowledges it is the weakest argument.  So let us turn to the second, “that chaos in Afghanistan not destabilize its neighbors, especially Pakistan.” First let me cite an area of agreement with Biddle.  I also believe that “state collapse” in Pakistan would be a disaster of the first order.  And furthermore, I agree with him that, “Pakistani state collapse, moreover, is a danger over which the United States has only limited influence.”  But Biddle’s connection of the risk in Pakistan to our involvement in Afghanistan is shockingly tenuous.  He writes:

If we cannot reliably influence Pakistan for the better, we should at least heed the Hippocratic Oath: Do no harm. With so little actual leverage, we cannot afford to make the problem any worse than it already is. And failure in Afghanistan would make the problem in Pakistan much harder.
If the Taliban regained control of the Afghan state, their ability to use the state’s resources to destabilize the secular government in Pakistan would increase the risk of state collapse there.

Again, little of this argument makes sense.  Biddle assumes that the Taliban — if it controlled Afghanistan — would be either willing or able to devote significant resources to destabilizing Pakistan.  I see no reason to accept this assumption.  The Taliban would face the same problems the Karzai government faces in holding Afghanistan together, but with a fraction of the resources.  When Biddle says “state resources,” you’d think there was an army or an airforce or even a potent intelligence service that could be deployed in an attack on Pakistan.  In reality, the Taliban — if it did seize power, which is no sure thing — would likely find itself starved for resources to maintain itself in power.  Indeed, it is probably as likely that efforts to retain control would drain resources currently devoted to the campaign in Pakistan.  This may be one reason, by the way, why Pakistan was under less threat from Islamist radicals from 1996 to 2001 when the Taliban controlled Afghanistan then in the years since.

The one argument that resonates slightly is the notion that giving radicals a sanctuary in Afghanistan would be problematic.  And it would, if they didn’t already have a sanctuary.  The cross-border Pushtun movement that we loosely characterize as the “Taliban” already has a sanctuary in western Pakistan.  It has access to millions of dollars in drug revenue, secure bases essentially protected from American attack (other than by air) by Pakistani nationalism, and the support of some significant chunk of the population.  In short, the threat to Pakistan is already about as bad as it can get — which is either good news or bad depending on one’s assessment of the position of the Pakistani regime.

The Strawman

A second fundamental problem with Biddle’s essay is his assumption that either we can escalate or we can withdraw.  Like those TV pundits who are always screaming that we have to “go big or go home,” Biddle limits the choices to two extremes.  It is fair for him to choose “go big” as one option.  It is his preference, after all.  But it is absolutely not fair to use “go home” as the default alternative.

The choice is not just between escalation and withdrawal — though withdrawal should be on the table.  The debate should include various limited involvement strategies — some including U.S. forces and some not.

There is an interesting point of comparison.  The Soviet Union withdrew from Afghanistan in early 1989.  The puppet regime of Mohammad Najibullah remained in power until March 1992.  Najibullah’s regime was less legitimate than the current regime of Hamid Karzai.  Indeed, the Najibullah regime only collapsed after Soviet assistance to Afghanistan ended in September 1991 and especially when the Soviets (then Russians) stopped supply subsidized oil to Najibullah’s regime.

Is there any reason to assume that the Karzai regime is more fragile than the Najibullah regime was?  Actually, the evidence suggests the reverse.  The Najibullah regime maintained itself in power despite the continuation of significant American assistance to anti-Najibullah forces (under the concept of “positive symmetry”).  The Najibullah regime had never been legitimized by real elections.  He was a puppet for a brutal occupation that caused as many as one million civilian deaths.

If the United States continued to provide significant aid to Karzai or a successor government, there is every reason to believe that the current government of Afghanistan could improve on the performance of the Najibullah government.  And if it can’t, doesn’t that raise questions about whether it makes sense to fight, die, and spend in support of such a weak regime?

In addition to aid, the United States could also provide training, intelligence support, close air support, and a multitude of other forms of assistance to bolster the position of the Afghan government.

So why does Biddle limit the debate to “reinforcement or withdrawal”?  Again, I can’t get inside Biddle’s head.  At best, it is a massive failure of imagination.  At worst, an attempt to discredit critics by presenting a caricature of what the policy debate is or ought to be.

COIN Past and Present

Another distressing part of Biddle’s argument is his assessment of U.S. capability to wage a successful counter-insurgency campaign.  To summarize his arguments:

In general, the historical rate of great power success in COIN is not encouraging—around 25 percent.
Great powers’ poor track record in COIN is due partly to the inherent difficulty of the undertaking but also to poor strategic choices. We can learn from experience, and we can change strategies and methods. Indeed, the U.S. military has learned a great deal about COIN in recent years. The new Army-Marine counterinsurgency doctrine is the product of a nearly unprecedented degree of internal debate, external vetting, historical analysis and assessment of recent experience.
The forces implementing COIN doctrine are also much improved over their Vietnam-era predecessors—and even over their immediate predecessors in Iraq in 2003–04. The U.S. military of 2009 has become uncommonly proficient at counterinsurgency, combining stronger doctrine with extensive COIN combat experience, systematic training and resources that dwarf most historical antecedents. More should be done to improve U.S. COIN capability, but we are now vastly better at this than, for example, the Soviets were in the 1980s, and much more proficient than most historical great power counterinsurgents have been.
Perhaps most important, we are blessed in Afghanistan with deeply flawed enemies.

The problems here are legion.  First, the fact that most counter-insurgencies fail illuminates a key methodological challenge for the field.  The reality is that many of the “lessons” we’ve learned from history are negative lessons.  The assumption is that if you refrain from doing what seems to have caused others to lose, you have a better chance to win.  Learning from mistakes is undeniable important, but it is rarely a positive guide to action.  Unfortunately, much of what we think we’ve “learned” is an application of the principle that good things will occur if you do the opposite of what previously led to failure.

The problem is that the reason that counter-insurgencies rarely win — particularly third party interventions — is structural, not strategic.  It has little to do with what the counter-insurgents did, but rather who they were.  No one wants to see foreign soldiers on their soil.  Even in the best of circumstances, cultural differences cause tensions.  Third parties get dragged into internecine disputes unwittingly, alienating potential allies.  More generally counter-insurgencies rarely win because it is easier to blow up a school than to build one, and because while government forces are expected to provide a range of services, insurgents get credit for any services they provide.  Insurgencies, once established, run downhill.  Counter-insurgencies, always, run uphill.  Hard does not mean impossible, but Biddle’s assumption that the problem is primarily about strategic choice rather than structural condition is simply wrong.

Second, Biddle dramatically overstates the quality of the “internal debate, external vetting, [and] historical analysis and assessment of recent experiences.”  The fact is, the debate has been remarkably closed. Virtually everyone supporting the COIN consensus orbits in some way around General David Petraeus.  He is a brilliant and charismatic man, and he has very effectively marketed his ideas, but throughout the process critics and doubters have been systematically marginalized and purged.  The empirical case for the current orthodoxy — often summarized as population-centric counter-insurgency — is so thin as to be startling.  With the possible exception of Iraq, there is literally no historical case where the current orthodoxy has been applied and been successful.  The reality is that the current doctrine is plausible, but is historically untested and vulnerable to many conceptual critiques.  In other words, it is almost pure theory.

Third, saying that we are now better at COIN and our enemies are worse insurgents than the Vietcong is interesting, but irrelevant.  The only relevant test case is the one case where the COIN orthodoxy has been tested, and that is Iraq.  While it may be true that we’re better than we were in the 1960s, and it is possible that the VC were better than the Taliban, the fact is that the Taliban are unmistakeably better than the hodgepodge of Sunni extremists, foreign fighters, and Ba’athist holdouts that made up the core of the insurgency in Iraq.  Furthermore, our ability to do COIN in Afghanistan is lower than it was in Iraq.  We have fewer men.  The terrain in rougher.  There are fewer urban centers we can effectively defend.  Afghanistan is poorer.  We’ve been at war longer, and our will is probably lower.  Though it isn’t part of the orthodoxy, we don’t have the support a majority ethnic group willing to fill in the gaps of our COIN doctrine with the use of brutal force — when the history of the Iraq insurgency is written, it will feature significantly the role of Shi’a death squads and ethnic cleansing in breaking the back of the Sunni insurgents.  Indeed, in Afghanistan,we find ourselves at odds with the Pushtuns, who form the single largest ethnic block in Afghanistan.  In short, a fair and relevant comparison with past cases suggests that our task will be harder than we expect and the chances of success lower.


In short, there is not a single section of Biddle’s essay that hold up to serious scrutiny.  And Biddle’s argument is unquestionably the most cogent of attempts to defend escalation in Afghanistan on strategic grounds.

2 comments to The Incoherence of COIN Advocates: Stephen Biddle Edition

Leave a Reply