A note about US troop deployment. Currently, there are 145,000 US troops deployed in Iraq. Troops levels in Afghanistan seem to be at around 40,000. That is 32,000 NATO troops, of which 12,000 are US, and 8,000 US troops under US command. Iraq and Afghanistan also happen to have similar population sizes at around 29 million.

At first glance, it seems that to achieve an efficient allocation of troops, the marginal utility of moving a troop between the two nations should be zero, that is, the marginal utilities of adding a troop to either one should be equal. This analysis would seem to imply a redeployment in Iraq, to a more defensive position, coupled with the transfer of, say, 20,000 troops from Iraq to Afghanistan.

However, one must account for risk-aversion in national priorities. If we try to divide our forces between two nations, there is a not small probability that we will fail in both. If we focus only on one, then the probability of double failure drops dramatically. I'd imagine many people would prefer one sure thing to a double failure, especially as the situation in Iraq looks increasingly hopeless, while the situation in Afghanistan, though slipping, is within our capability to repair. This logic could justify an even larger troop tranfer.

Barring a large increase NATO troops levels, this could be the last best choice.