There’s a whole bunch of good stuff in the new CNAS report on Afghanistan and Pakistan, but also a few tangles of seriously confused thinking. The authors’ definition of “securing the population”, for example, is a mess. On the one hand, they define it — correctly — as “denying [the population's] passive or active support to the Taliban.” But then there’s this:
Afghan civilian casualties, whether at the hands of the coalition, the Taliban, or the Afghan government, will be the most telling measure of progress.
Okay, see if you can spot the difference:
In Iraq, killing non-combatants was part of the insurgents’ strategy. Hence, in 2006 and 2007, it wasn’t uncommon for 2,000 civilians to die every month. In Afghanistan, that was the number of civilian casualties for the whole of 2008 — with more than a third killed not by the Taleban but by foreign troops. It follows from this that:
- In Iraq, “securing the population” meant stopping it from getting killed by the insurgents.
- In Afghanistan, it means stopping it from supporting the insurgents.
It also follows that, contrary to what the paper claims, civilian casualties in Afghanistan are not “the key metric”. In fact, they’re no metric at all. The only thing they’re good for is measuring ISAF’s efforts to screw up less.
None of this is to say shielding the population from the Taleban shouldn’t be the top priority. You should do it, and you can do it; it’s just that there’s no meaningful way to measure your success when dealing with intangibles like “support” or “opposition” — unless you count opinion polls, but I’m guessing they’re not concrete enough for Exum et al.
Another thing I found curious is how CNAS would prioritise resources:
Focusing on protecting the population requires making difficult operational tradeoffs. There will never be enough coalition forces in Afghanistan to execute a perfect population-centric counterinsurgency, and it will take years to train and deploy Afghan forces in sufficient numbers and of sufficiently high quality to undertake this mission effectively. It is necessary, therefore, to focus available forces where the fewest number of government and coalition troops can protect the greatest number of Afghans. This will require the coalition to depart some areas it currently occupies. For example, U.S. forces might be compelled to withdraw from sparsely populated Taliban strongholds such as the Korengal Valley in order to better protect more of the population in Kandahar City or Lashkar Gah.
Maybe so — I haven’t been on the ground for almost three years, so who knows, maybe Afghanistan has suddenly become an urban society. Last I looked, though, it was nothing like Iraq, and I can’t help wondering whether here, too, memories of the Surge are confusing otherwise bright minds.
I agree with everything you say above….
Did you suggest any changes after you read the draft?
Well, the bit about redirecting efforts to Lashkar Gah and Kandahar are pretty much a Kilcullen idea (I took him to task about that a few months ago). I don’t know why they think most Afghans live in those areas, but that’s what they were working with.
That being said, I do agree that the Korengal really shouldn’t be a priority the way it is now. It is a transit corridor, nothing more, and the fight there ultimately has very little to do with the Taliban/Haqqani/HiG/LeT movements, and great deal more to do with other things.
And, like Christian, I agree with everything you say above.
I read a couple of early drafts. But no final drafts. I was basically a fact checker, all the “big ideas” belong to the authors, not to me. I can’t even remember what I said about the first couple of drafts.
Next we know, they’ll be suggesting strategic hamlets. I don’t recall that one being a resounding success…
Then again, maybe every paradigm shift in military thinking requires its own revisionist narrative.