International forces in Afghanistan are preparing to hand over responsibility for security to Afghan soldiers and police by the end of 2014. U.S. President Barack Obama has argued that battlefield successes since 2009 have enabled this transition and that with it, “this long war will come to a responsible end.” But the war will not end in 2014. The U.S. role may end, in whole or in part, but the war will continue—and its ultimate outcome is very much in doubt.

Should current trends continue, U.S. combat troops are likely to leave behind a grinding stalemate between the Afghan government and the Taliban. The Afghan National Security Forces can probably sustain this deadlock, but only as long as the U.S. Congress pays the multibillion-dollar annual bills needed to keep them fighting. The war will thus become a contest in stamina between Congress and the Taliban. Unless Congress proves more patient than the Taliban leader Mullah Omar, funding for the ANSF will eventually shrink until Afghan forces can no longer hold their ground, and at that point, the country could easily descend into chaos. If it does, the war will be lost and U.S. aims forfeited. A policy of simply handing off an ongoing war to an Afghan government that cannot afford the troops needed to win it is thus not a strategy for a “responsible end” to the conflict; it is closer to what the Nixon administration was willing to accept in the final stages of the Vietnam War, a “decent interval” between the United States’ withdrawal and the eventual defeat of its local ally.

There are only two real alternatives to this, neither of them pleasant. One is to get serious about negotiations with the Taliban. This is no panacea, but it is the only alternative to outright defeat. To its credit, the Obama administration has pursued such talks for over a year. What it has not done is spend the political capital needed for an actual deal. A settlement the United States could live with would require hard political engineering both in Kabul and on Capitol Hill, yet the administration has not followed through.

The other defensible approach is for the United States to cut its losses and get all the way out of Afghanistan now, leaving behind no advisory presence and reducing its aid substantially. Outright withdrawal might damage the United States’ prestige, but so would a slow-motion version of the same defeat—only at a greater cost in blood and treasure. And although a speedy U.S. withdrawal would cost many Afghans their lives and freedoms, fighting on simply to postpone such consequences temporarily would needlessly sacrifice more American lives in a lost cause.

The Obama administration has avoided both of these courses, choosing instead to muddle through without incurring the risk and political cost that a sustainable settlement would require. Time is running out, however, and the administration should pick its poison. Paying the price for a real settlement is a better approach than quick withdrawal, but both are better than halfhearted delay. For the United States, losing per se is not the worst-case scenario; losing expensively is. Yet that is exactly what a myopic focus on a short-term transition without the political work needed to settle the war will probably produce: failure on the installment plan.


The international coalition fighting in Afghanistan has long planned on handing over responsibility for security there to local Afghan forces. But the original idea was that before doing so, a troop surge would clear the Taliban from strategically critical terrain and weaken the insurgency so much that the war would be close to a finish by the time the Afghans took over. That never happened. The surge made important progress, but the tight deadlines for a U.S. withdrawal and the Taliban’s resilience have left insurgents in control of enough territory to remain militarily viable well after 2014. Afghan government forces will thus inherit a more demanding job than expected.

The forces supposed to carry out this job are a mixed lot. The ANSF’s best units should be capable of modest offensive actions to clear Taliban strongholds; other units’ corruption and ineptitude will leave them part of the problem rather than part of the solution for the foreseeable future. On balance, it is reasonable to expect that the ANSF will be able to hold most or all of the terrain the surge cleared but not expand the government’s control much beyond that. Although the Taliban will probably not march into Kabul after coalition combat troops leave, the war will likely be deadlocked, grinding onward as long as someone pays the bills to keep the ANSF operating.

Those bills will be substantial, and Congress will have to foot most of them. The coalition has always understood that an ANSF powerful enough to hold what the surge gained would be vastly more expensive than what the Afghan government could afford. In fiscal year 2013, the ANSF’s operating budget of $6.5 billion was more than twice as large as the Afghan government’s entire federal revenue. Most of the money to keep the ANSF fighting will thus have to come from abroad, and the lion’s share from the United States.

In principle, this funding should look like a bargain. According to most estimates, after the transition, the United States will contribute some $4–$6 billion annually to the ANSF—a pittance compared to the nearly $120 billion it spent in 2011 to wage the war with mostly American troops. The further one gets from 2011, however, the less salient that contrast becomes and the more other comparisons will come to mind. Annual U.S. military aid to Israel, for example, totaled $3.1 billion in fiscal year 2013; the amount required to support the ANSF will surely exceed this for a long time. And unlike Israel, which enjoys powerful political support in Washington, there is no natural constituency for Afghan military aid in American politics.

Afghan aid will get even harder to defend the next time an Afghan corruption scandal hits the newspapers, or Afghan protests erupt over an accidental Koran burning, or an American adviser is killed by an Afghan recipient of U.S. aid, or an Afghan president plays to local politics by insulting American sensibilities. Such periodic crises are all but inevitable, and each one will sap congressional support for aid to Afghanistan. I recently spoke to a gathering of almost 70 senior congressional staffers with an interest in Afghanistan and asked how many of them thought it was likely that the ANSF aid budget would be untouched after one of these crises. None did.

In the near term, Congress will probably pay the ANSF what the White House requests, but the more time goes on, the more likely it will be that these appropriations will be cut back. It will not take much reduction in funds before the ANSF contracts to a size that is smaller than what it needs to be to hold the line or before a shrinking pool of patronage money splits the institution along factional lines. Either result risks a return to the civil warfare of the 1990s, which would provide exactly the kind of militant safe haven that the United States has fought since 2001 to prevent.

Managing the congressional politics around sustaining Afghan forces after the transition was feasible back when Washington assumed that a troop surge before the transition would put the Taliban on a glide path to extinction. The United States would still have had to give billions of dollars a year to the ANSF, but the war would have ended relatively quickly. After that, it would have been possible to demobilize large parts of the ANSF and turn the remainder into a peacetime establishment; aid would then have shrunk to lower levels, making congressional funding a much easier sell. But that is not the scenario that will present itself in 2014. With an indefinite stalemate on the horizon instead, the politics of funding the ANSF will be much harder to handle—and without a settlement, that funding will outlast the Taliban’s will to fight only if one assumes heroic patience on the part of Congress.


Since outlasting the Taliban is unlikely, the only realistic alternative to eventual defeat is a negotiated settlement. The administration has pursued such a deal for well over a year, but so far the process has yielded little, and there is now widespread skepticism about the talks.

Many, for example, doubt the Taliban are serious about the negotiations. After all, in late 2011, they assassinated Burhanuddin Rabbani, the head of Afghan President Hamid Karzai’s High Peace Council and the Kabul official charged with moving the talks forward. Since the Taliban can wait out the United States and win outright, why should they make concessions? Others argue that the Taliban are interested in negotiations only insofar as they provide a source of legitimacy and a soapbox for political grandstanding. Still others worry that bringing together multiple Taliban factions, their Pakistani patrons, the Karzai administration, the governments of the United States and its allies, and intermediaries such as Qatar will simply prove too complex. Conservatives in the United States, meanwhile, doubt the Obama administration’s motives, worrying that negotiating with the enemy signals weakness and fearing that the White House will make unnecessary concessions simply to cover its rush to the exits. Liberals fear losing hard-won gains for Afghan women and minorities. And many Afghans, especially women’s groups and those who are not part of the country’s Pashtun majority, also worry about that outcome, and some have even threatened civil war to prevent it.

Yet despite these concerns, there is still a chance for a deal that offers more than just a fig leaf to conceal policy failure. The Taliban have, after all, publicly declared that they are willing to negotiate—a costly posture, since the Taliban are not a monolithic actor but an alliance of factions. When Mullah Omar’s representatives accept talks, other factions worry about deals being made behind their backs. Taliban field commanders wonder whether the battlefield prognosis is as favorable as their leaders claim (if victory is near, why negotiate?) and face the challenge of motivating fighters to risk their lives when shadowy negotiations might render such sacrifice unnecessary. The Taliban’s willingness to accept these costs thus implies some possible interest in a settlement.

There may be good reasons for the Taliban to explore a deal. Mullah Omar and his allies in the leadership have been living in exile in Pakistan for over a decade—their children are growing up as Pakistanis—and their movements are surely constrained by their Pakistani patrons. Afghans are famously nationalist, and the Afghan-Pakistani rivalry runs deep; exile across the border surely grates on the Afghan Taliban. Perhaps more important, they live under the constant threat of assassination by U.S. drones or commando raids: just ask Osama bin Laden or six of the last seven al Qaeda operations directors, all killed or captured in such attacks. And a stalemate wastes the lives and resources of the Taliban just as it does those of the Afghan forces and their allies. While the Taliban are probably able to pay this price indefinitely, and while they will surely not surrender just to stanch the bleeding, this does not mean they would prefer continued bloodletting to any possible settlement. The conflict is costly enough that the Taliban might consider an offer if it is not tantamount to capitulation.

What would such a deal comprise? In principle, a bargain could be reached that preserved all parties’ vital interests even if no one’s ideal aims were achieved. The Taliban would have to renounce violence, break with al Qaeda, disarm, and accept something along the lines of today’s Afghan constitution. In exchange, they would receive legal status as a political party, set-asides of offices or parliamentary seats, and the withdrawal of any remaining foreign forces from Afghanistan. The Afghan government, meanwhile, would have to accept a role for the Taliban in a coalition government and the springboard for Taliban political activism that this would provide. In exchange, the government would be allowed to preserve the basic blueprint of today’s state, and it would surely command the votes needed to lead a governing coalition, at least in the near term. Pakistan would have to give up its blue-sky ambitions for an Afghan puppet state under Taliban domination, but it would gain a stable border and enough influence via its Taliban proxies to prevent any Afghan-Indian axis that could threaten it. And the United States, for its part, would have to accept the Taliban as a legal political actor, with an extra-democratic guarantee of positions and influence, and the probable forfeiture of any significant base structure for conducting counterterrorist operations from Afghan soil.

From Washington’s perspective, this outcome would be far from ideal. It would sacrifice aims the United States has sought since 2001, putting at risk the hard-won rights of Afghan women and minorities by granting the Taliban a voice in Afghan politics and offering a share of power to an organization with the blood of thousands of Americans on its hands. Yet if properly negotiated, such a deal could at least preserve the two most vital U.S. national interests at stake in Afghanistan: that Afghanistan not become a base for militants to attack the West and that it not become a base for destabilizing the country’s neighbors.

As long as the Taliban are denied control of internal security ministries or district or provincial governments in critical border areas, the non-Taliban majority in a coalition government could ensure that Afghanistan not become a home to terrorist camps like those that existed before the war. Chaos without a meaningful central government, by contrast, would preclude nothing. And whatever fate Afghan women and minorities suffered under a stable coalition would be far less bad than what they would face under anarchy. A compromise with the Taliban would be a bitter pill to swallow, but at this point, it would sacrifice less than the alternatives.


Simply meeting with the Taliban is only the starting point of the negotiating process. To create a deal that can last, the U.S. government and its allies will need to go far beyond this, starting by laying the political groundwork in Afghanistan. Although negotiators will not have an easy time getting anti-Taliban northerners to accept concessions, the biggest hurdle is predatory misgovernance in Kabul. Any settlement will have to legalize the Taliban and grant them a political foothold. This foothold would not give them control of the government, but their legal status would allow them to compete electorally and expand their position later. Over the longer term, therefore, the containment of the Taliban’s influence will depend on political competition from a credible and attractive alternative—something the establishment in Kabul is not yet able or willing to provide.

The Taliban are not popular in Afghanistan; that is why they will accept a deal only if it guarantees them a certain level of representation in the government. But at least they are seen as incorruptible, whereas Karzai’s government is deeply corrupt, exclusionary, and getting worse. If Karzai’s successor continues this trend, he will hand the Taliban their best opportunity for real power. Should Kabul’s misgovernance persist and worsen, eventually even a brutal but honest opposition movement will make headway. And if a legalized Taliban were to eventually control critical border districts, enabling their militant Pakistani allies to cash in some wartime IOUs and establish base camps under the Taliban’s protection, the result could be nearly as dangerous to the West as the Afghan government’s military defeat. The only real insurance against that outcome is for Kabul to change its ways.

To date, however, the West has been unwilling to compel reform, preferring so-called capacity-building aid to coercive diplomacy. Such benign assistance might be enough if the problem were merely a lack of capacity. But Afghanistan is misgoverned because its power brokers profit from such malfeasance; they won’t change simply because the Americans ask them to, and unconditional capacity building just creates better-trained kleptocrats. Real improvement would require, among other things, that donors withhold their assistance if the Afghan government fails to implement reforms. But donors have shied away from true conditionality for fear that their bluff will be called, aid will have to be withheld, and the result will be a delay in the creation of a higher-capacity Afghan civil and military administration—the key to current plans for Western military withdrawal.

If the West cannot credibly threaten to withhold something Kabul values, then Afghan governance will never improve. It is late in the game to begin such an approach now; the West would have had more leverage back when its aid budgets were larger and military resources more plentiful. Still, credible conditionality could make even a smaller budget into a stronger tool for reform. Using conditionality properly, however, would mean accepting the possibility that the West might have to deliberately reduce the capacity of Afghan institutions if they refuse to reform—a task that is neither easy nor pleasant, but necessary if the West is going to be serious about a settlement.

The Obama administration will need to undertake serious political work in Washington as well as in Kabul. Any viable settlement will take years to negotiate and require the West to make real concessions, and such a process will offer ample opportunities for members of Congress to embrace demagoguery and act as spoilers. The Obama administration’s initial experience on this score is instructive: as an early confidence-building gesture, last year the administration offered to free five Taliban detainees at Guantánamo in exchange for the release of Sergeant Bowe Bergdahl, the Taliban’s only American prisoner. But U.S. lawmakers howled in outrage, the detainees were not released, the Taliban charged bad faith (both on the detainee issue and on the addition of new conditions from Karzai), and the negotiations collapsed. Serious negotiations toward a final peace settlement would provide countless opportunities for such congressional outrage, over much larger issues, and if legislators play such games—and if the administration lets itself be bullied—then a viable settlement will be impossible. Likewise, if Congress defunds the war too soon, unfinished negotiations will collapse as the Taliban seize victory on the battlefield with no need for concessions.

For talks to succeed, Congress will thus need to engage in two acts of selfless statesmanship: accepting concessions to the Taliban and prolonging unpopular aid to the Afghan military. The latter, in particular, would require bipartisan compromise, and achieving either or both goals may prove impossible. If they are going to happen, however, one prerequisite will be a sustained White House effort aimed at building the congressional support needed. The president will have to make a major investment in garnering political backing for a controversial Afghan policy, something he has not done so far.


As daunting as the obstacles to a negotiated settlement are, such a deal still represents the least bad option for the United States in Afghanistan. If the White House is unwilling to accept the costs that a serious settlement effort would entail, however, then it is time to cut American losses and get out of Afghanistan now.

Some might see the Obama administration’s current policy as a hedged version of such disengagement already. The U.S. military presence in Afghanistan will soon shrink to perhaps 8,000–12,000 advisers and trainers, and U.S. aid might decline to $4–$5 billion a year for the ANSF and $2–$3 billion in economic assistance, with the advisory presence costing perhaps another $8–$12 billion a year. This commitment is far smaller than the 100,000 U.S. troops and over $100 billion of 2011, and it offers some chance of muddling through to an acceptable outcome while discreetly concealing the United States’ probable eventual failure behind a veil of continuing modest effort.

Only in Washington, however, could $14–$20 billion a year be considered cheap. If this yielded a stable Afghanistan, it would indeed be a bargain, but if, as is likely without a settlement, it produces only a defeat drawn out over several years, it will mean needlessly wasting tens of billions of dollars. In a fiscal environment in which $8 billion a year for the Head Start preschool program or $36 billion a year for Pell Grant scholarships is controversial, it is hard to justify spending another $70–$100 billion in Afghanistan over, say, another half decade of stalemated warfare merely to disguise failure or defer its political consequences.

It is harder still to ask Americans to die for such a cause. Even an advisory mission involves risk, and right now, thousands of U.S. soldiers are continuing to patrol the country. If failure is coming, many Afghans will inevitably die, but a faster withdrawal could at least save some American lives that would be sacrificed along the slower route.

It would be preferable for the war to end a different way: through a negotiated compromise with the Taliban. Talks so complicated and fraught, of course, might fail even if the United States does everything possible to facilitate them. But without such efforts, the chances of success are minimal, and the result is likely to be just a slower, more expensive version of failure. Getting out now is a better policy than that.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now