Is Hamid Karzai worth the fight in Afghanistan? We'd better learn the answer soon -- or give up the counterinsurgency game.

It's a bad moment for counterinsurgency strategy and its adherents. The surge that U.S. President Barack Obama ordered in Afghanistan appears to have sunk into a quagmire, one that critics of the policy foresaw at the time. Indeed, the polemic underneath Rolling Stone's profile of Gen. Stanley McChrystal was that the "runaway general" had hornswoggled a gullible president into signing off on a hopeless mission. The author, freelance journalist Michael Hastings, described the adversary as "Afghan kids who pose no threat to the U.S. homeland" and compared the nation-building effort there to "trying to stop the drug war in Mexico by occupying Arkansas and building Baptist churches in Little Rock." Of course, if that's so, then McChrystal probably should have been court-martialed rather than forcibly retired.

After reading the article, I compiled a document I called "COIN toss," listing the arguments for and against continuing the counterinsurgency effort. As someone who holds out some faint hope for the administration's strategy, I was dismayed to see that while I came up with 10 reasons to abandon COIN (counterinsurgency), most based on observable failures of the strategy, I could think of only five reasons to keep it, most based on hope and scant signs of progress. The list of cons included: "Karzai is too corrupt," "Karzai doesn't believe in it," "the Taliban is too strong," "Afghans hate the American presence," and "American troops won't do it" (the one argument Hastings powerfully vindicated). The pros included "social and economic indicators are rising," "Afghans hate the Taliban," and "it's too early." I don't think we can call that a tie.

Still, at the bottom of each list I had written the great equalizer: On one side, "We can afford to lose," and on the other, "We can't afford to lose." If U.S. and NATO troops really are facing kids who can't see beyond their neighborhood, or even fundamentalists who will be satisfied by stripping away all vestiges of modernity from Afghanistan, then the war is simply unnecessary for Americans. Americans lived with Taliban control of Afghanistan in 1997, and ashamed though they might feel for having raised the hopes of the Afghan people only to abandon them, Americans would probably live with it again. Perhaps they would pay no graver cost in leaving Afghanistan than they did in pulling out of Vietnam in 1975.

But I doubt it. While communism was rapidly discrediting itself as a fighting faith in the 1970s, jihadism is a vibrant cause that would experience profound validation from a forced U.S. withdrawal from Afghanistan. As Anthony Cordesman of the Center for Strategic and International Studies recently wrote, the Afghan Taliban is "far better linked to Al Qa'ida and other international extremist groups" than it was only a few years ago; should they gain real power, "they are likely to become such a sanctuary and a symbol of victory that will empower similar extremists all over the world." This is not to say that a monolithic jihadism will spread outward from Afghanistan in a contemporary version of the domino theory, but rather that a Taliban victory there is likely to attract and inspire Islamist radicals everywhere.

That doesn't mean the United States "can't afford to lose," but rather that the costs of accepting failure could be very high. So it is imperative to ask whether the obstacles to success, however defined, can be overcome. Right now, as my note-to-self indicated, those obstacles seem overwhelming: Even those who argue for some version of "stay the course," like the New Yorker's George Packer, view the available alternatives as even worse than the apparently doomed counterinsurgency effort. It's rapidly becoming intellectually embarrassing to profess any faith in the effort at all.

But let me try. The Rolling Stone article made a strong, if inadvertent, case that for all his many virtues, McChrystal was never the right man to carry out his own strategy. COIN doctrine requires a radical deference by military commanders to civilian goals, and to civilian leaders: You cannot, as advocates endlessly repeat, kill your way to victory. McChrystal's now-notorious contempt for the silver-tongued, glad-handing, endlessly ambivalent senior White House officials he had to deal with was of a piece with his nonchalance toward the inevitably messy, time-consuming, and compromised political objectives of the war. This is the man who spoke of setting up "government in a box" in each new liberated district. That's a blinkered view of governance, especially in a place like Afghanistan. McChrystal may be an enlightened soldier, but he's still a soldier.

During the time I spent in Afghanistan this April, I watched the painstaking and often just painful effort of giving birth to local government, and to a social contract between citizen and state. It felt more like the flowering of a seed -- at best -- than the unpacking of a box. You could argue in fact, as a general rejoinder to COIN strategy, that the organic time scale of such a process is just too gradual to match any military timetable Americans will accept. The American and Afghan officials I spent time with didn't think so. They thought that they could actually make a meaningful difference before mid-2011, when Obama's drawdown of troops is scheduled to begin; but they did fear that their effort would be wasted unless the Afghan state committed itself to making local government work nationwide, by sending resources and by delegating authority to provincial and district officials.

But that's going to require pushing Karzai to take governance seriously, or at least get out of the way so that local power brokers can do so. The civilian leaders to whom I talked understood that; I don't know whether McChrystal's team did. McChrystal earned Karzai's regard by treating him with great deference, and he used that capital to induce the Afghan president to sign off on, or at least not directly oppose, NATO's military operations. That was an important transaction, but it was a soldier's transaction. There was no one to push, seduce, or bribe Karzai to, for example, replace some of his most corrupt and brutal allies in Kabul or the provinces, in part because Karzai learned to play off the military against civilian authorities.

In cashiering McChrystal, Obama said that Gen. David Petraeus will pursue the same policy as his predecessor. I hope that's not true; I hope that when Petraeus undertakes his own review he'll conclude that the military tail is wagging the civilian dog. U.S. and NATO troops continue to rely for security and logistical support on some of the most brutal and venal figures in Afghanistan, thus securing short-term advantage at the cost of deepening the alienation of the Afghan people. How can Karzai be pressed to move against corruption if U.S. forces are themselves reinforcing it? And Karzai must be pressed to release his death grip on political power, allowing parliament and the courts to exercise authority and accepting that power must be decentralized in a country with a long tradition of local autonomy. If ordinary Afghans are to take real risks to defend the state from the Taliban, then the government they directly experience has to be empowered -- which is another reason why Karzai must first replace some of the worst actors at the local level.

Can Petraeus, whose dickering with members of Iraq's Sunni Awakening movement shows a flair for negotiation, push Karzai to make concessions he apparently doesn't believe in? That could be a Sisyphean effort. The only way to persuade Karzai to buy into NATO's war is probably to agree to buy in to his, which is to say the effort to persuade Taliban commanders to put down their arms, join the government, and thus preserve Karzai's own position. The carrot might have to be paired with a stick: If Karzai remains intransigent over the next six months or so, the United States will have to accept that the counterinsurgency effort cannot succeed and begin an earlier withdrawal of troops. Petraeus is probably the wrong man for such a messy, fluid, fragile deal. It's the kind of bargain those mealy-mouthed politicians McChrystal and his team despised are so good at striking.

John Moore/Getty Images

Terms of Engagement

It's Not Too Late to Save Kyrgyzstan

Russia and the United States weren't able to stop the recent outbreak of violence and ethnic cleansing in Osh. But there's still time to prevent the worst.

In 2005, the world's heads of states, gathered at the United Nations headquarters in New York, agreed that they had a responsibility to protect their own peoples from mass atrocities -- and that the responsibility would fall to the larger community when a state proved unable or unwilling to prevent such crimes. Since that time, violence reaching the legal threshold of crimes against humanity (the other specified constituent crimes are genocide, war crimes, and ethnic cleansing) has been perpetrated in Sudan, Sri Lanka, and the Democratic Republic of the Congo, and arguably in Kenya, Burma, and Zimbabwe, among other places. In almost every case, the world has failed to muster an even remotely effective response. So far, it looks like we can add Kyrgyzstan to the list; but it's not too late to get things right.

In the explosion of ethnic violence that rocked the southern city of Osh starting June 10, as many as half the country's 800,000 Uzbeks were forced to flee their homes before marauding mobs of ethnic Kyrgyz, apparently abetted by government troops. An untold number, which the New York Times now puts at "thousands," have been killed. Arson, rape, and other atrocities have been widespread. Late last week, the very fragile government in Bishkek finally seemed to gain control over the army; or perhaps, with the victim population largely terrorized and dispersed, the violence simply burned itself out. But this might be the lull before another storm: Uzbeks may seek revenge, in turn provoking new attacks from Kyrgyz or from the Kyrgyz-dominated security forces. Naomi Kikoler of the Global Center for the Responsibility to Protect calls the violence "a textbook case of R2P," as the norm has come to be abbreviated. Her organization as well as Human Rights Watch and the International Crisis Group have called for urgent international action.

In the first moments of the crisis, Kyrgyzstan's interim president, Roza Otunbayeva, issued a desperate call to Moscow to provide troops. Instead, Moscow referred the matter to its own regional body, the Collective Security Treaty Organization (CSTO), which then declined to authorize action. Russia had cited R2P -- with transparent cynicism -- to justify its 2008 invasion of South Ossetia and Georgia. Why not now, with a willing government and a genuine crisis? It is morally satisfying to say, as human rights advocates often do, that nobody is acting because nobody cares. And that's often true: The death of several hundred thousand Darfuris weighed far less with most states than assuring "African solutions to African problems" or preserving commercial ties with an oil-rich regime. But Kyrgyzstan, the world's only country with both a U.S. and a Russian military base, is scarcely a geopolitical orphan; and because the government, in this case, is not perpetrating the atrocities, potential actors do not have to defend a regime at the cost of neglecting citizens.

Whatever its strategic or moral preferences, Russia faced an extremely daunting calculus, as Peter Zeihan, an analyst for Stratfor, which provides "strategic intelligence," pointed out in a recent article. Kyrgyzstan is 1,800 miles from the Russian heartland and has a rugged geography and little economic value. There is also the huge complicating factor of Uzbekistan, which borders the inflamed area, aspires to be the regional hegemon, and would almost certainly have objected to the Kyrgyz request. Indeed, any dispatch of Russian troops to the periphery would have instantly reminded both Russians and their neighbors of very painful experiences in Afghanistan and Chechnya. The CSTO is widely viewed in the West as Russia's Potemkin NATO, but Moscow would have been loath to act unilaterally in the face of opposition from other members (one of which is Uzbekistan). Finally, ethnic Uzbeks were the victims -- and "Russians and Uzbeks don't like each other," as Stephen Sestanovich, a Russia expert at the Council on Foreign Relations, puts it bluntly.

And why didn't the United States act? There have been reports that Otunbayeva also asked Washington for help. A senior administration official says that the rumors are untrue, though an unofficial appeal might have been made to U.S. diplomats. Such a request, he insists, "would have been taken very seriously," but it's hard to believe that President Barack Obama would have ordered troops transiting the NATO base at Manas for Afghanistan to form a convoy to Osh. Indeed, asking neighbors to send troops to suppress atrocities within days of their outbreak, especially if they might have to take on a local army, is a test the international community is very likely to fail.

But military force is only one, and not necessarily the most effective, response to "R2P situations." In 2008, swift diplomacy prevented post-electoral violence in Kenya from turning into mass slaughter; tough sanctions at the very outset might have halted the killings in Darfur. The current lull in violence in southern Kyrgyzstan, and the eagerness for help displayed by a very weak but democratic regime, gives international actors a second chance to get things right, and without an urgent military intervention.

But will they? The White House, along with other donors and the United Nations, is now focused on helping to organize humanitarian assistance. The deeply insecure environment around Osh has raised the question of whether troops might be needed to secure a "humanitarian corridor," but because this was precisely the limited mandate of the U.N. peacekeepers who were sucked into a vortex in Bosnia and Somalia, among other places, officials are exploring this approach with extreme caution. A senior U.N. official says that shops in Osh have begun to reopen, raising the hope that supplies can be distributed without military protection.

The threat of renewed violence won't subside just because refugees are being fed and housed; returning them to their homes might, in fact, exacerbate the situation. The U.N. official to whom I spoke was surprisingly sanguine about reconciliation, hoping that a June 27 referendum that would give permanent status to the interim government would strengthen the regime's capacity to act. The Security Council discussed Kyrgyzstan once last week, and neither Secretary General Ban Ki-moon nor council members have shown much urgency on the subject. But the White House has been engaged in virtually round-the-clock consultations on mechanisms to prevent another outbreak of violence, along with Russia, various allies, and the Organization for Security and Cooperation in Europe, whose current head is Kazakhstan. Here, too, it might be necessary to ask the Security Council to authorize some kind of "stabilization mission," perhaps involving police or constabulary forces rather than peacekeepers. The administration official said policymakers are considering "a dozen alternatives" and haven't yet come up with the right one, much less marshaled whatever forces would be needed to implement it.

This official also said that the one positive thing to come out of an otherwise tragic situation is the close cooperation between Russian and U.S. diplomats and political leaders, and Kyrgyzstan will certainly be on the agenda when Obama and President Dmitri Medvedev meet June 24. That certainly is a bonus, and perhaps even a validation of the "reset" with Russia. But if the two former adversaries work together to make a really, really sincere effort that fails to stop a new round of violence, that will be to no one's credit.

Russia may be cynical on R2P, but this U.S. administration is not. In his National Security Strategy released last month, Obama vowed to remain "proactively engaged in a strategic effort to prevent mass atrocities and genocide." Last summer, when the Sri Lankan government killed thousands of civilians in its campaign to wipe out a vicious terrorist group, the administration worked quietly, and as it turned out ineffectively, to stanch the violence. Kyrgyzstan presents an easier test -- about as easy as R2P situations are likely to get. This time, no points for trying hard.

Postscript: Last week I reported that Haji Abdul Jabar, the governor of Arghandab, a district just north of Kandahar, had been assassinated by the Taliban. I questioned whether anyone would be brave and foolish enough to take his place. Kevin Melton, the U.S. Agency for International Development official working with the local government, informs me that Haji Muhammad, the "shura leader," immediately agreed to do just that. Those who would urge the Obama administration to abandon its counterinsurgency effort must be prepared to tell the likes of Haji Muhammad that the United States doesn't think the game is worth the candle.