Official optimism, of course, still reigns. Marine Maj. Gen. Richard Mills, who until recently commanded the 1st Expeditionary Force in the Southwest--mainly Helmand and Nimruz provinces, considered partial success stories--says he believes that COIN can work even with limited troops and time. "Success can become a tidal wave, building momentum of its own," Mills told National Journal. "As people become more confident in their Afghan security forces, I think you're going to see a shift even in the troublesome areas.... It's beginning to show success."
As one piece of evidence of how the U.S. military "learns" (a key Nagl idea), Mills says he was not as interested in "body counts" as in other measures: the diminishing resources of the Taliban as they "search for old munitions and save expended cartridges"; the eagerness with which parents in his area sent their children to school; and the lower desertion rates among Afghan soldiers. At last count, the U.S. military said that the attrition rate had dropped from 70 to 30 percent. Mills says he saw, in just a year, dramatic improvement in "probably the most important metric: the separation between the insurgency and the population.... When you begin to see parents pushing their children to school, then that tells you that threat is diminishing."
Petraeus's successor, Marine Lt. Gen. John Allen, will follow a similar course. He perfectly fits the Nagl/Petraeus mold of the scholar-soldier. Adaptable and erudite (he has "one of the best libraries" in the military, according to retired Marine Col. Dan Kelly, an old friend), Allen also has considerable hands-on experience in counterinsurgency. He directed governance and stability operations in Iraq's Anbar province during the 2006-08 surge, helping to transform it from a hotbed of insurgency into a critical building block of stability.
But Allen may now have an impossible circle to square: He will preside over a drawdown at exactly the moment when COIN's architects say it has just begun to receive the minimum resources necessary. Under COIN theory, there should ideally be about 20 security officers for every 1,000 civilians, which would mean that NATO and the U.S. need 600,000 troops in Afghanistan rather than the 140,000 or so there now--a number set to decline next month.
Not surprisingly, according to military officials, right now COIN is not working to protect--and thereby win over--most of the Afghan population. Earlier this month, the United Nations announced that May was the deadliest month for civilians there since it began keeping count in 2007. The Taliban and other insurgents caused most of the deaths, but at least 12 percent were ascribed to mistakes by NATO troops and Afghan forces. Because Afghan civilians are dying at higher rates, Campbell says, Western forces are not seeing the population come over to their side in large numbers--something that's crucial for the strategy to succeed. The people, he concedes, believe that the coalition doesn't do enough to protect them, and neither the Afghan army nor some 20,000 police are yet up to that task. "We're not going to kill our way out," Campbell says.
Another problem for counterinsurgency is its mind-boggling cost--and its meager return on investment. Even with the United States spending $80 billion per year, Campbell says that "the full spectrum of COIN" is being applied only in a relatively small section of the country's south and southwest, centered on Kandahar, with a population of about 2.1 million. The 14 provinces of the battle-stricken east have a population nine times that size, constituting more than half of Afghanistan's total population of 30 million. Nagl may be right in saying that Regional Command East is "all we have to do," but it's bigger by far than anything the U.S. has yet tried. Campbell says it "will be the last place we leave." Meanwhile, it appears that Obama has decided not to press for a larger COIN approach in the east, which Petraeus wanted, so that he can withdraw from Afghanistan more quickly, according to a senior administration official. Petraeus "wanted to move some of the [south] folks into [the east] to repeat the counterinsurgency campaign there. The president doesn't think that's necessary," the official said.
Two further problems are the absence of a legitimate and trusted government that the U.S. and NATO can hand off to--this is absolutely critical and perhaps the most Vietnamesque aspect of Afghanistan--and the Taliban's safe haven across the border in Pakistan, which has possibly grown even safer as relations between the U.S. and Pakistan continue to deteriorate.
Before Obama settled on his current strategy and ordered the surge, Vice President Joe Biden argued vehemently in 2009 for a pared-back approach that discarded COIN, admitted that Afghan President Hamid Karzai wouldn't be much help, and focused mainly on attacking terrorists. Today, the problems of counterinsurgency are rapidly reducing American strategy to something that looks more and more like this. "Essentially what's happening now is the Biden plan on steroids," says retired Army Lt. Col. Douglas Ollivant, Campbell's former military aide and an ex-White House staffer focused on Afghanistan.
Ollivant calls this "a realist version of COIN": Practice counterinsurgency in places where it might work and leave the other places (for example, the unruly Pech Valley, from which Campbell simply withdrew) alone. If these regions become nests for Taliban, "then you do the Biden plan there" and launch missions, he says. "What Campbell tried to focus on was a more limited approach that tries to get the low-hanging fruit," Ollivant says. "It connects the cities that want to be part of the solution." COIN-istas like to compare their approach to the spread of ink spots on a paper towel; the secured (ink-spotted) area bleeds out to connect with other ink spots until important corridors of stability appear throughout the country. But the new skeptics, such as Ollivant, say that so many areas are irremediable that a few stray ink spots may never spread far enough--which is partly why a similar strategy failed in Vietnam.
Can such a partial solution, amounting to a kind of land-based island-hopping, ever work? Retired Marine Col. Thomas X. Hammes says it can't, and he is in a position to know. Hammes, now a fellow at the National Defense University's Center for Strategic Research, is a onetime COIN supporter who helped devise the current strategy. Outside pockets of success like Helmand province, "what Petraeus is doing is killing Taliban," Hammes says, and that's about it. Because of Karzai's failures and Pakistan's open door, a huge "strategic disconnect" exists between the COIN concept and the reality on the ground, he contends. "The counterinsurgency discussion led by John Nagl and his team" doesn't acknowledge that.
Afghanistan is far more primitive than Iraq, where most of the COIN proponents cut their teeth. "Social and economic conditions are not the same as in Iraq, or Malaya in 1950," Hammes says. Counterinsurgency requires at least "an 18th-century concept of representative government and a 19th-century concept of government providing services." In Afghanistan, he says, "we think we're going to leap from somewhere in the 7th-to-14th century into the 19th century--and do it in four years." And although insurgents crossed from Syria and Saudi Arabia into Iraq, there was nothing like Afghanistan's porous border and the sanctuary next door. Afghanistan "is a helluva lot larger," Hammes says. "Pakistan can't control the border as Syria or Saudi [Arabia] could." And it doesn't even want to. Campbell uses the word "biblical" to describe Afghanistan: "Iraq had water, infrastructure, oil, educated people. Afghanistan has none of that." Beyond that, Campbell says, there is almost no one who can replace Karzai, meaning there may be no credible government to hand things off to. "The bench is not deep at all," he says. "This is where it's completely different from Iraq."