HOME / war stories: Military analysis.

We Finally Have a Strategy for AfghanistanUnfortunately, that may not be enough.

By Fred KaplanPosted Tuesday, Dec. 23, 2008, at 6:53 PM ET

Afghan President Hamid Karzai It's time to start getting nervous about Afghanistan.

In recent days, Adm. Mike Mullen, chairman of the Joint Chiefs of Staff, has talked about doubling the number of U.S. troops in the country from 30,000 to 60,000—way more than the three brigades (roughly 12,000 extra troops) that Barack Obama endorsed on the campaign trail.

A case could be made that reinforcements are needed. But it's not clear—to anyone, including many officers—whether this will mark the pivotal boost or the start of a quagmire.

RSS

The good news is that, seven years after the Sept. 11 attacks and nearly three years after the resumption of full-scale war with the Taliban, we are finally beginning to formulate a strategy—and we have officers in place who think strategically.

As history shows, however, smart generals and shrewd strategists don't necessarily yield victory—especially in Afghanistan.

The biggest problem is that the country's fate ultimately lies outside its borders. As long as Pakistan's northwest territories remain a lawless free-for-all, with Taliban and al-Qaida fighters crossing the border at will, Afghanistan will never be stable. And as long as Pakistan faces a threat from India to the east, its leaders will never deploy enough troops to quash the insurgents in the northwest territories.

In short, we could do everything perfectly in Afghanistan, but it wouldn't matter unless the region-wide conflicts could be brought under some control.

Again, the good news is that all the relevant players—President-elect Obama, Adm. Mullen, Secretary of Defense Robert Gates, Gen. David Petraeus, and their top aides—understand this. But knowing the dimensions of a problem is only the first step to solving it. And each one of this problem's aspects—countering the insurgency in Afghanistan, stabilizing Pakistan, and calming tensions between Pakistan and India—is very difficult.

The problem of Afghanistan is the easiest—or at least the easiest to calculate—in the sense that it's to some extent susceptible to military power. But, as Gates and Petraeus have said several times, it's not entirely a military problem; there can be no "victory" in the standard meaning of the word. A good ending, if there is one, will involve a negotiated settlement in which "reconcilable" Taliban—those who joined the insurgency for nonideological reasons—are lured over to the Afghan government's side.

This is why, according to today's Wall Street Journal, top U.S. officers are starting to aid local militias in the fight against the Taliban. This approach has the merit of realism: Afghanistan is a tribal society; power is focused on the militias; securing the population, at this point, can be done only through them.

However, we're not going to win over any chieftains unless we can demonstrate that we might win. This is the main reason for a boost—nobody's calling it a "surge"—in U.S. troop levels. We need some quick tactical victories against the Taliban. Air power can't do it: Bombing kills too many civilians; and that turns the Afghan people against us, against the Afghan government, and toward the insurgents. So more ground forces are the only way.

But three caveats are worth noting. First, as Dexter Filkins reports in his excellent book The Forever War, Afghan militias are notorious opportunists; they switch sides at the slightest shift—in who's winning or who's paying more—or sometimes just at whim. They might be won over, but maybe not for long.

Second, even 60,000 U.S. troops aren't enough to win over all the tribes and militias, much less to stabilize the whole country. It's only enough, at best, to secure a few more towns, to close off a few more border crossings, and to clobber a few more Taliban units. The hope is that some high-profile successes will set in motion a cascade of further successes, brought on by alliances with the Afghans themselves.