In an article in today’s New York Times, it is reported that the U.S. commander in Afghanistan, General Stanley McChrystal, is plannibg what is variously described as a “new strategy” for dealing with the Taliban or as simply a refinement of the current AfPak strategy based in counterinsurgency doctrine. Regardless of what it is called, it requires a calculation of forces that does not add up using the “old math” I learned in school: some “new math” is clearly needed.
The new emphasis of strategy in Afghanistan apparently is to place greater emphasis on securing and providing ongoing security for Afghan villages and villagers in areas of contention, which presumably means particularly in the rural parts of the country where the Pashtun are in the majority. This emphasis follows from the so-called COIN (conterinsurgency) doctrine found in the joint Army-Marine Counterinsurgency manual, Army FM 3-24 and Marine Corps Warfighting Publication 3-33.5. This shifts the center of activity away from hunting down and killing Taliban faster than they can be replaced (the way much of Vietnam was fought), because, in McChrystal’s own words, the supply of replacement fighters is “essentially endless.” This admission in itself is telling and sobering: if the recruitment pool of new Taliban is so great, how can we ever expect to prevail, since the endlessness of that reservoir suggests either that the Taliban is very popular or that our presence is greatly opposed (or a combination of the two).
Accepting this reformulation, however, turns one strictly into the teeth of the numbers. As noted earlier (at at the risk of being a “nag” on the subject), the Counterinsurgency Manual embraces the idea of pacification but also points out that it is very manpower intensive. To reiterate, an effective COIN force, in the manual’s own estimate, requires 20 counterinsurgents for every 1,000 in the population being protected. As previously noted, that literally means a force of 660,000 counterinsurgents, given the size of the Afghan population. Even fudging on the numbers, that adds up to over one half million forces confronting the Taliban, whose numbers of full-time fighters has been estimated at as little as 20,000, not including part-timers.
How do the numbers match up to fill this bill? Here we need some new math. Currently, there are 62,000 American forces in the country, scheduled to expand to about 68,000 by year’s end. With other NATO contributions, the number swells to about 100,000, although the NATO numbers are likely to shrivel. Current projections call for an Afghan National Army (ANA) force of around 134,000 by the end of 2011. Given the progress in recruiting and training those forces (and especially in making them ethnically representative enough for the Afghans themselves to think of them as “national”), we are talking about a total force of less than 250,000 by the end of next year. That does not even come close.
It is clear, as been suggested here, that McChrystal will return to Washington later this year hat-in-hand and doing his best William Westmoreland imitation to ask for more troops. Clearly, he cannot ask for the roughly 350,000 new forces necessary to meet the COIN doctrine’s requisites, so he will almost certainly ask for a more modest number of “trainers” to help expand the ANA. If my numbers are correct, however, it will require an Afghan force that is THREE TIMES the force already planned to come even close to a half million total counterinsurgents, which is on the low side (a total force of slightly less than a half million).
Where are all these extra forces going to come from? Thanks partly to Taliban harassment (as well as antipathy toward the government), it is proving difficult to meet current goals. How in the world can these be trebled? Moreover, who is going to pay for them if they can be found (we all know the answer to that one–us)? Further, all this will take time, and in the interim, it is difficult to imagine the Taliban will sit idly by and allow this expansion to occur: may they have won before this new force can be fielded?
Maybe I’m crazy, and maybe I am a victim of some old math that does not add up. At the same time, maybe those who think this is all going to work out well are privy to some new math that makes what seems impossible possible? If so, I’d like the tutorial.
One piece of math is not so hard to interpret. Recent polls show that over half the U.S. population opposes the war in Afghanistan, mostly because people believe the war is likely unwinnable. Does that mean there are a lot of Americans still held hostage to the old math? Or does it mean the majority understand arithmetic better than the planners in Washington?