SFA = (COIN + SFA) ??

Since Abu Muqawama's software seems to hate my comments these days, I'm just going to post a response over here.

Exum has a brief post this morning about Richard Haass' op-ed today on U.S. options in Afghanistan. Here's his comment about one of Haass' proposed courses of action:

[Haass:] "One would reduce our troops’ ground-combat operations and emphasize drone attacks on terrorists, the training of Afghan police officers and soldiers, development aid and diplomacy to fracture the Taliban."

[Exum:] I don't know how much Haass knows about training security forces, but his first "alternative" would require an investment in Afghanistan as massive as the one we're already making. So I think it's more an operational alternative than a strategic one.

I guess I know a fair bit about training security forces, at least from a policy perspective if not a practical one. But more importantly, I know a fair bit about math. So I guess what I don't understand here is how a security force assistance/foreign internal defense mission (in which a complement of U.S. forces would focus on training ANSF, without engaging in combat operations except as an embedded component of the units they're training and advising) plus development assistance could possibly cost as much as SFA/FID plus development assistance plus a counterinsurgency campaign.

SFA/FID = $A
COIN = $B
$A + $B = $A ???

We're already training ANSF and waging a COIN campaign (including combat operations, obviously). Exum doesn't think that halting the COIN campaign would save money? The only way the math works on this is if you believe that we're not really doing that much training right now, or devoting that many resources to the mission.

4/82 might disagree. (More on this later.)