Sunday, October 25, 2009

Troop Numbers Again

As I said, I'm ambivilant about the surge question. The bottom line is that I support the military's request for more troops notwithstanding my many worries. So the question is, can we win with those troops? Let's look at the data with my amateur number-crunching again.

Looking at the regions of Afghanistan, the capital district (3.5 million people) is fairly secure but critical so let's allocate the assumed minimum of 2% troop strength compared to the local population, for 70,000 security forces to secure the area.

Regional Command East (7-10 million) and Regional Command South (3.2 million) total 12 million people for this example. With extra troops needed to interdict the border, let's call it as needing 2.5%, or 300,000 troops at the high end. But if we are pulling back from the border, 2% might do, or 240,000 security forces.

Regional Command West (3 million people) and Regional Command North (7 million) are not peaceful, but let's assume a level that I assumed for the Shia south in Iraq and call it 1%, or 100,000 troops in these areas. Maybe we could get by with only 0.5%, or 50,000, since we don't face the equivalent of the Iran-backed Sadrists in those areas.

So we'd need 410,000 troops for a proper counter-insurgency campaign in the populated areas.

Maybe 360,000 would be enough, however. Especially if Pakistan controls their side of the border.

Or maybe we'd need 470,000 if we want to interdict the border seriously from the Afghanistan side.

So what do we have?

Right now we have 68,000 on the ground or to be there soon.

Add in about 32,000 other Western troops.

For the rest, let's say 85,000 Afghan army and 82,000 national police.

Interesting enough, we can throw in 71,000 contract security personnel, too. I'd never read anything until recently on this force.

I have no idea if there are many reliable local police. I assume from everything I've read that the answer is "no."

The total is about 340,000 security personnel.

Are you telling me that we can't get 20,000 more security personnel from local defense forces or regional defense forces for the minimal number? Or even 70,000 for the middle number if we really focused? I think we had Sons of Iraq in those numbers. And our program didn't count Kurdish Pesh Merga, so we should get up to 130,000 militia type local defense forces or regional defense forces to supplement national security forces, shouldn't we, if we needed the high number?

It is also quite possible that we don't need as many troops in the higher threat areas as the percentages I assign them. The Afghanistan insurgents and terrorists aren't as well armed or financed as those in Iraq were. We may be able to get by with lower ratios in some or all of the regions.

Remember, just as in Iraq, many of the forces can be glorified security guards. As long as better quality units are on call, which can call on even better quality troops if needed, plus our firepower and advisors, we can get the numbers to wage a proper counter-insurgency strategy.

Certainly we can't get all those indigenous forces at once, which is surely why our military wants 40,000 (or 60,000 or 80,000--or even just 30,000--if some reports of actual wishes are credible) more US troops, so we can buy time for training locals.

But even without a surge of forces, we aren't doomed to lose the war. We could start proper counter-insurgency in smaller areas that are critical to the enemy and expand from those areas once they are tamped down.

But given the speed our leaders are losing heart and the wholly predictable desertion of Democrats from the so-called "good war" now that they don't need it as a cudgel to lose the "bad" Iraq War (because we won there) I can fully understand why our military wouldn't want to count on the patience to muddle through in Afghanistan.

I always thought that with patience, we could have muddled through in Iraq without the American surge. But since patience might have required us to still be supporting the Iraqis in fighting the insurgents and terrorists rather than just mopping up a much-reduced enemy today, in retrospect that probably would not have worked given the 2006 and 2008 US election results. And at the time, given the situation, I backed that surge despite what I thought was a risk to losing domestic support here faster than we could achieve results. In the end, our surge did produce results faster than heightened American casualties eroded support back home.

And on the positive side, my back-of-the-envelope numbers for Afghanistan seem to indicate that a surge of just 40,000 should be enough to defeat our enemies in Afghanistan despite the conventional wisdom that we'd need 600,000 security forces to control Afghanistan. And Pakistan is fighting on their side of the border, so there is hope there.

So again, despite my concerns, I back the proposed surge and will hope my worries don't come to pass.

UPDATE: This story puts the Afghan security forces at 200,000, or about 30,000 more than I thought:

There are already more than 100,000 international troops in Afghanistan working with 200,000 Afghan security forces and police. It adds up to a 12-1 numerical advantage over Taliban rebels, but it hasn't led to anything close to victory.

The story puts the Taliban at 25,000. What I don't know is if this is the total or just the estimated full timers. As a rule, if memory serves me, you can probably count on 9 part-timers (who may observe, supply, shelter, and sometimes fight) for every full-timer. So are we talking a pool of 250,000 total Taliban or 2,500 full-time fighters in the field at any one time? I suspect the former.

This is simply not an insurmountable force to defeat, assuming the people generally support us and we have the patience to fight long enough to train locals to fight instead of our troops.