There are still lots of places in Kandahar Province with “unpaid” “ANP.” Someone gave the local village/tribal leader and his friends some guns. They are accountable to no one and at best a nullity when it comes to improving security.

Now if your issue is there’s absolutely no functioning district governance, and no actual trained Afghan police or soldiers there or likely to arrive any time soon, and there’s Western soldiers that are combat-ineffective because they’re standing around looking for someone to work with, you could make the argument that this sort of thing could work. The remaining big issue then would be guiding reintegration of the force back into the “legit” ANP down the road. It remains, however, a fairly rare case, at least in South Afghanistan up until now.

You make an interesting suggestion, but one of the problems with using tribe all the time (and re-enforcing it with incentives etc) is that a) tribes are pretty localized–there are a whole bunch of Barakzais, but the Barakzais in Kandahar are more or less seperate from the ones in Helmand, and so on and so forth. Especially if you are creating a program like this for a national force (though I suppose you could do it a micro-level as well) you would end up having to focus on small localities (X tribe in Y district, apart from X tribe in Z district) and treating them as wholly local groups based on geography. Or, on the other hand, you would continue treating e.g. all the Barakzais from all over Afghanistan as a group that acts collectively, and you’re back to the original problem.

The other issue that I see is one you actually kind of bring up–ethnicity. Talking about tribes at all only puts Pashtuns in the mix. This fact seems to get lost in many of these discussions (or conveniently glossed by calling the Hazaras, Tajiks, Uzbeks et al “tribes” which makes about as much sense as calling the French a tribe).

While it’s true that many of the most troubles areas in Afghanistan right now are Pashtun-heavy, as Dorronsoro warns in his recent CEIP piece the insurgency is expanding to non-Pashtun areas at an alarming rate, not to mention that familiar areas like Ghazni are only about 50% Pashtun. Solutions for areas like this that focus entirely on tribes either don’t map onto other groups or will fail to include them, creating new problems for each one you manage to solve.

So here’s my really crazy way to bridge the gap that someone needs to point out the holes in for the sake of my sanity. Is there a way to build the ANA and ANP to draw strength from a tribal identity in and of itself, rather than trying to do so via a particular leader? There’s plenty of lit. in the ethnic conflict field (particularly D. Posner’s work on Africa) that suggests that ethnic divisions become politically salient only if there are institutions that incentivises one particular identity over another. So make it that when you join one of the national security services, you can identify as a particular identity. Your presence in the security force, and the success of the missions you undertake is then combined in some type of “tribal team score” that is publicly available, and translates into some type of monetary perk (he with the highest per capita recruitment gets roads first kind of deal).

Do i think this is going to make for a very good free and open society? absolutely not. It entrenches tribalism within the security forces, creates incentives that force people into the security forces against their will, is vulnerable to central government corruption to name only a few, but it seems better than the current system…

Without an existing functioning political structure to work with, “reconstruction” is a meaningless feel good phrase. The residents of Afghanistan have to put their country together on their own terms. Lecturing by State Department and USAID functionaries isn’t going to mean much to the locals.

So that’s the brilliant counter argument. Get ready.

(It does go up to 11 and this is an actual model of my skeleton. You just can’t argue).