Yet as NATO leaders prepare to gather in Chicago, where the war in Afghanistan will be Topic A, there is a growing presumption that, shortly after NATO finishes building up the Afghan security forces, we should start dismantling them. Allegedly because of binding budgetary constraints, a force that will soon reach its combined goal of 352,000 uniformed personnel (not counting intelligence services or community-watch organizations known as Afghan Local Police) may be quickly cut back to the suspiciously precise number of 228,500, starting around 2015.
Yes, this is a potential problem. And to say I lack confidence that our president wants to win more than he just wants to make sure there is a decent interval before losing would be an understatement.
Much depends on whether in this last year of American operations we can pound down the Taliban in the east enough to make the enemy weak enough for Afghanistan to handle. Much depends on whether the Taliban have sanctuaries in Pakistan to regenerate their capabilities.
In a country of 30 million, O'Hanlon says this is not enough, even bolstered by a small NATO contingent of up to 20,000 troops:
Counterinsurgency doctrine, based on experience from many previous cases, suggests that in a country of 30 million, like Afghanistan, as many as 600,000 soldiers and police officers could be required.
The 2% figure is in doubt based on past experience where counter-insurgents won with less than 2%, but it is a handy rule of thumb. But I prefer to address the numbers issue by threat regions. In Iraq, that led me time and time again to conclude we had enough troops to win that war. I have tried to do the same for Afghanistan, as I did here most recently.
Let me take that admittedly crude analysis and update it, assuming a population of 30 million and adjusting my population figures up 10% or so.
Looking at the regions of Afghanistan, the capital district (4 million people) is fairly secure but critical to hold, so let's allocate 1.5% troop strength compared to the local population, for 60,000 security forces to secure the area.
Regional Command East (11 million) where our main effort will end this year--with extra troops needed to interdict the border--let's call it as needing 2%, or 220,000.
Regional Command South and Southwest (3.5 million) is into the hold phase, but since it is the home ground of the Taliban, let's assume 1.5% are needed, for 52,000.
Regional Command West (4 million people) and Regional Command North (7.5 million) are not peaceful, but we could assume a level that I assumed for the Shia south in Iraq back during the intense fighting and call it 1%, or 115,000 troops in these areas. Maybe we could get by with only 0.5%, or 58,000, since we don't face (despite Iranian meddling) the equivalent of the Iran-backed Sadrists in those areas and because past claims that the Taliban would flow north and west after being defeated in the south, haven't panned out.
That totals 390,000 required to secure Afghanistan. If you assume Afghanistan will have 352,000 regular forces (armed forces and police), add in 20,000 NATO troops, and--for kicks--10,000 armed contract security forces--we get darned close at 382,000. I would be easy to make up the difference with 8,000 Afghan Local Police.
But we won't have 352,000 regular Afghan forces. We'll have 228,500. The problem is, paying for that is beyond Afghanistan's capabilities and NATO isn't eager to pay the full bill. This is a real problem. Whatever security forces Afghan can afford have to be enough to hold their ground.
Clearly, the assumptions have to be that the amount of troops needed per region will decline from today's levels and even from my post-surge assumptions above. This could very well be true.
But one major problem is that I don't see Pakistan--especially in the east--not being a sanctuary for the Taliban. Even if I can assume that most regions of Afghanistan can be secured with only 0.5% troop presence compared to the local population, the area of Regional Command East across from those Pakistan sanctuaries can't get down to that level. Even if we pound down the Taliban in Regional Command East, Afghanistan would need more forces there. Let's say 1% there in this scenario. If all the non-Regional Command East areas need only a 0.5% troop presence, that means 95,000 troops. One percent in the east means 110,000. That's 205,000 total. In that optimistic post-battlefield victory scenario, the NATO forces wouldn't even need to engage and could be a fire brigade for emergencies in addition to training and adviser missions.
Is that realistic to assume? I don't know. The big problem is that Regional Command East has so many people that the security level of that region is absolutely critical in figuring out what we need. And the east is not yet in the hold phase and the east is most vulnerable to enemy forces in Pakistan sanctuaries. This is a big probablem and a huge variable. If you assume we'll need 390,000 nationwide rather than the optimistic 205,000 in the post-battlefield victory scenario, 228,500 Afghan national forces plus 20,000 NATO forces and 10,000 contract security personnel give us only 258,500. That's a shortfall of 132,000 security personnel.
Now, during the Iraq War, I constantly reminded people that we didn't need all security forces to be of the caliber of the 170,000 peak force of American troops or their equivalents (essentially the British, in any number). Many who stand guard duty in static locations where 99% of the time there is no major enemy threat simply don't need to be special forces. But somebody has to be there even if all they can do is call for help before firing a few shots and running from a force too large to handle.
Can we scrounge up 132,000 Afghan Local Police or tribal militias that can be trusted sufficiently to count as part of the government's tally?
We can make assumptions about what is needed. Clearly we have done so. Maybe we are correct. But if we are correct, are we correct that the numbers planned are enough to win or enough to lose far enough in the future to prevent President Obama from being blamed for the loss?
