The idea of extending the alliances may have come, in part, from Stephen Biddle,
We have to start using the military not as a device to secure everything uniformly but as a device for creating incentives and disincentives—sticks and carrots—to push along the process of local cease-fires with particular factions." For instance, he said, we would have to tell each faction: "We will defend you if you cooperate; if you don't cooperate, we will attack you."....
I'm puzzled by his idea here. The deals with the Sunni sheiks are explicitly opportunistic. Assuming that the alliances of convenience whip the jihadists, there is nothing preventing the Sunnis and Shiites (and Americans) from going back to killing one another.
It is worth noting that Biddle himself has serious doubts about the whole notion. In his interview with Gwertzman, he said the odds that the surge and the new strategy might work—that is, that they might produce "something like stability and security in Iraq"—are "maybe one in 10."
Whether those odds are worth gambling on, he said, depends on whether you're averse or prone to risk. Biddle described himself as risk-averse. Therefore, if the decision was up to him, he'd pull the troops out. President Bush, he said, "is clearly very tolerant of risk." And so he's pouring more in.
I think this framework lays out a good deal of the decision trees and risk assessments people are making on the question of whether or not to support a significant or complete withdrawal of US combat forces from Iraq in a militarily responsible timeframe. James' fear is that there will be great carnage in Iraq without the presence of fifteen to twenty brigades and that this carnage will be much higher than the ongoing level of violence and chaos that is now the baseline trend. Therefore a low probability chance of success has a much lower downside than withdrawal in his mind.
This is a reasonable and informed judgment. I think taking 10:1 odds is seldom a good choice in anything but for the horses during a crowded field race. I also think that the difference in carnage with fifteen to twenty US brigades in Iraq and very few to no US brigades in Iraq will mainly be a function of time instead of magnitude. In this interpretation, we're screwed no matter what, so might as well minimize total costs and vulnerability to threats by withdrawing.
The second part of Kaplan's article illustrates why any strategy going forward is a very long shot strategy and thus why I believe it is time to get out.
If the U.S. military had, say, 100,000 more troops to send and another 10 years to keep them there; if the Iraqi security forces (especially the Iraqi police) were as skilled and, more important, as loyal to the Iraqi nation (as opposed to their ethnic sects) as many had hoped they would be by now; if the Iraqi government were a governing entity, as opposed to a ramshackle assemblage that can barely form a quorum—then maybe, maybe, this plan might have a chance.
Each of these IF statements need to be postively met, else the entire argument is false and impratical. The five IF statements (US troops =~250,000 ish, 10+ years of US deployments @ 250,000 ish, Loyal Iraqi Army, Loyal Iraqi Police, Effective Iraqi Government) are all being violated right now. The US troop levels and decade long committment are mostly indepedent of the last three IF statements although there is some feedback present; however the problems of the Iraqi Army and Police feed into an ineffective government, and the positive feedback loop recycles itself and dumps governmental instability and incapacity into the security forces.
Breaking one of these logjams would be nice and a great accomplishment for the team that leads that effort. However it will be ineffective in the strategic view unless the other four problems are also resolved. And what is the probability of that? In my mind that probabililty is very, very, very low, so any potential benefits of success have to be massively discounted by this reality.