David Brooks alerts us to a new Foreign Affairs essay by Andrew Krepinevich, which argues that the United States should switch to a traditional counterinsurgency strategy in Iraq—a strategy which has historically "worked"—rather than the whack-a-mole approach the Army's taking in the Sunni provinces right now, a battle in which the insurgents seem to have fought the US military to a draw. Obviously this is too sensible ever to happen (how many times has the "hearts and minds" suggestion been hauled out, to no avail?), though it's interesting to discuss.

Krepinevich's suggestion: The US would concentrate instead on securing key enclaves—Baghdad, Kurdistan, southern Iraq—and withdraw completely from the Sunni countryside. No more fighting insurgents head-on. In the "secure" areas, the military would focus on aid and reconstruction, stamping out crime and violence (those first two are crucial), training the local police and military forces, and mingling with the locals, garnering intelligence. Hopefully this leads to fewer clumsy raids on houses, fewer innocent Iraqis detained, and newfound support from the local population. Hearts and minds. Hopefully, pacifying these areas would make it easier for the US to disarm the Shiite and Kurdish militias. (Although in Krepinevich's essay he wants to rely heavily on those militias—which seems like a pretty serious mistake, given stories like this.) Over time, as the Iraqi government's security forces come online in the "held" territories, the home team would slowly spread out to the countryside, like an ink stain or oil spot, and pacify the remaining areas. That's the theory anyway.

One weird part, though, is that in his essay Krepinevich thinks we can do this with 120,000 troops in Iraq, and significant reductions thereafter. How does he figure? The British had about 20 security personnel for every 1,000 persons in Malaysia, which he cites as a model of counterinsurgency. Baghdad alone has 6,000,000 people, so that would take 120,000 troops right there. Does the US really want to rely on the increasingly out-of-control militias in Iraq? Wouldn't that create as many problems as it solved? (Iraq would resemble El Salvador in the 1980s, which was not successful at all.) Also, presumably the US would need different types of soldiers: fewer heavy vehicles, more police types, more translators, a different type of force structure. Is that feasible?

More crucially, though, consider the political problems with this approach here at home. First, Bush would have to admit that the occupation is going poorly, to say the least, and that the military will now have to completely readjust its strategy and stay in Iraq for, oh, another decade. Second, a counterinsurgency strategy, presumably, works best when the contractors responsible for aid and reconstruction aren't looting and pillaging the country. Just saying... Third, in the short term, many more soldiers are likely to get killed if the US switches to a "traditional" counterinsurgency strategy. It's just the nature of the thing—the troops have to mingle with the locals, walk around in small units, integrate with trained Iraqi units, rely on translators of dubious loyalty. No more zooming by in armored Humvees. This may prove more effective in the long run, but in the short run, people are getting blown up. See how this strategy hold up after a few nights of that on the evening news.

The politics from Iraq's perspective are no less tricky, a point which Krepinevich is well aware of. How much more patience do the Iraqis really have for a long-term American presence? Granted, if the US secured Baghdad and the Shiite south, and finally got reconstruction and job-creation and crime-fighting right, the Shiites and other urban Iraqis might start to warm up to us. But I doubt it. Muqtada al-Sadr has gained a lot of popularity over the years by revving up his hit single: "U.S. out NOW!", and while much of his support comes from the fact that he can provide poor Iraqis with the necessities Americans can't (a model pioneered by Hezbollah and HAMAS), he also seems to have struck a nationalistic chord among many Shiites who don't have the patience for ten more years of occupation. Oh, and switching to counterinsurgency would mean jettisoning the political process now underway—so much for the constitution; how are they supposed to hold a referendum if the US withdraws from the Sunni regions?—which is a tricky proposition by any means.

Will Krepinevich's plan work? I don't know, maybe. He's obviously a smart guy and knows far more about this stuff than I can ever hope to learn. It would almost certainly be better than what the U.S. is doing now. But this is all sort of moot, really: Bush isn't going to change course, Rumsfeld isn't going to retool the Army to deal with counterinsurgency (not enough expensive weapons systems involved; also, see Jason Vest for why the Army hasn't adapted these techniques). At this point the only thing worth doing is to figure out the politics of this whole mess so that the people in Washington who green-lighted and carried out the trashing of Iraq never get near the levers of power again.