America’s war in Afghanistan will soon enter its 17th year. Most Americans tired of the war long ago, but its odd invisibility in American domestic politics has allowed it to carry on, mostly out of sight, with little real political pressure to change a stalemated status quo.
This condition of stasis may not last much longer, however. The war’s 17th year promises both opportunity and peril—and the former is due at least partly to the latter. Afghanistan’s growing political dysfunction and the opportunism of its neighbors create a risk that it could go the way of Syria, or of Afghanistan itself in the 1990s: a breakup of the state into warring factions with rivalrous neighbors sucked in to support conflicting proxies. Anyone can see what this has done to Syria since 2011, and South Asians old enough to be in leadership positions remember what Afghanistan was like after Najibullah fell in 1992. This specter presents a credible threat of equal opportunity disaster: the simultaneous realization of every actor’s worst-case scenario. Few would choose this condition, but it does focus the mind; the threat of a Syrian abyss may be the goad that enables long-stalled negotiations to end the fighting on terms the United States can accept.
Indeed, the prospects for an acceptable peace in Afghanistan are now better than they have been in a very long time—perhaps since the Bonn Conference of 2001. But there are still many ways the negotiations could collapse and turn the worst-case scenario from goad to reality. Only some of the risk factors are under U.S. control, and certainly Americans lack the leverage once offered by 100,000 troops and $100 billion of annual expenditure. But U.S. choices do make a difference, and poor ones could slam shut a door that might otherwise be opening, dooming Afghanistan to a replay of the 1990s.
Three critical contributions are now needed from the U.S. government: patience, flexibility, and sustained politico-economic engagement. None have been hallmarks of the U.S. experience in Afghanistan since 2001, and skeptics could be forgiven for wondering how well these policy traits fit the current moment in American domestic politics. But none are impossible, and an acceptable end to a long, horrible war now rides on them.
Why Care?
The most basic question, as usual, is perhaps the most important, and it is the one at the root of much of the war’s frustrating current condition. Does Afghanistan matter enough to the United States to warrant continued engagement at any significant cost?
The problem here is that Afghanistan poses stakes for Americans that are real, but limited. First of all, and most obvious, if the Kabul government falls, al-Qaeda, the Islamic State, or other terrorist groups could re-establish base camps in Afghanistan from which they could target the United States or other Western countries, as al-Qaeda did in 2001. This is a real threat. But there is nothing unique about Afghanistan as a base for terrorists to attack the West; dozens of ill-governed spaces around the world could serve this purpose.
The more important U.S. stake in Afghanistan is the stability of its nuclear-armed region, and especially Pakistan. Nuclear-armed Pakistan, with a nuclear-armed Indian rival on its eastern border, has been enmeshed in a counterinsurgency war of its own since 2004. Pakistan and Afghanistan are separated by a notoriously porous border, home to a large ethnic Pashtun population that is closely associated with both the Afghan and Pakistani insurgencies. Americans have long worried that Afghan Taliban base camps among Pakistani Pashtuns make the Afghan insurgency hard to defeat, but a much greater danger is if Afghan base camps make the Pakistani Taliban harder to defeat. A virulent insurgency in any nuclear-armed state is a serious danger to international stability; a virulent Pakistani insurgency with established base camps across a porous Afghan border amid the detritus of a failed Afghan state is a much bigger danger. The risk that collapse in Kabul could destabilize an already unstable, nuclear-armed region is a real threat to American interests.
That said, this peril, too, has its limits. Many things would have to break badly for an Afghan collapse to lead to nuclear violence. Kabul would have to fall; Islamabad would have to fall; the Pakistani security forces would have to split; nuclear weapons would have to fall into terrorist hands; and these terrorists would have to find a way to deliver one to a major city. The compound probability here is small, but it is not zero. And the resultant catastrophe if the breaks go badly for the United States would be of historic magnitude. This does not make Afghanistan a problem worth sending 500,000 American soldiers to solve—but neither does it make Afghanistan irrelevant or unimportant. Afghanistan lies somewhere in the middle on the scale of security interest magnitude: a real, but limited, threat to U.S. interests.
Expensive Means to Limited Ends
Real but limited U.S. interests in Afghanistan have created a long and frustrating debate over how to respond. Because U.S. interests are real, neither the Obama nor the Trump Administration have been willing to just pull out. If they did, and things went badly, a catastrophe would be their fault. Neither Administration has been willing to take the risk of going down in history as the people responsible for a nuclear detonation in Times Square. But decisive victory in the war would be very expensive in lives, dollars, and time; this cost has long seemed to exceed what the stake was worth. The result has been a long, floundering search for some real-but-limited means to secure real-but-limited stakes wherein the cost would be commensurate with the benefit.
Mostly, this search has led to assorted variations on “light footprint” strategies where the United States provides training, advising, equipment, commando raids, and air power, while the Afghan government provides the ground forces and suffers most of the casualties. The Obama Administration experimented with a heavy footprint alternative, ordering a surge to an ultimate strength of 100,000 U.S. troops by 2011. But the surge announcement was accompanied by a promise of reversal and drawdown to begin after less than two years, and for most of the Obama presidency the U.S. military was in the process of lightening, not thickening, the U.S. presence.
The logic of this reinforce-then-drawdown approach was that the surge would hit the Taliban hard enough to reverse their momentum and put the war on a glide slope toward ultimate Taliban defeat. In the meantime, U.S. aid would build a greatly expanded Afghan National Security Force (ANSF) to whom the war could then be handed off as U.S. troops gradually withdrew. A sustained, open-ended 100,000-soldier presence seemed more expensive than the war was worth, but perhaps a brief phase of excessive effort followed by much cheaper indigenous war-making could secure limited interests at a price worth paying.
The first part actually worked: The U.S. surge did reverse Taliban momentum, and U.S. troops were able to stabilize threatened Afghan districts when deployed in sufficient strength and kept in place to defend what had been cleared. The problem was mostly the second part of the plan. 100,000 soldiers seems like a lot, but it was enough to clear only about half of Afghanistan’s threatened southern and eastern districts at a time. To clear all of threatened Afghanistan in the allotted time would have taken at least twice that number of troops; alternatively, the same troops could have cleared twice the ground by swinging from the south to the east once the south was cleared, relying on the ANSF to hold the cleared south afterwards, but that would have taken at least double the time. With only limited perceived stakes on the table, the Obama Administration was unwilling to double either the troops or the time. And the result was that the surge ended with the job left unfinished. The ANSF was handed not a war on a glide slope to Taliban defeat, but a war against a Taliban enemy that still held large stretches of strategically important Afghan real estate.
The job of defeating an enemy this strong was—and remains—beyond any plausible ANSF capability. This should not be surprising. In weakly institutionalized political systems like Afghanistan’s, state militaries are often riven with corruption that saps combat motivation, undermines equipment availability, and interferes with the development of technical military expertise. Without a judicial system or legislature that can adjudicate conflict among armed elites, regimes in places like Afghanistan commonly rely on an internal balance of power to create political order and prevent factionalism from spilling over into armed violence. In such settings, the primary purpose of the army and police is not to defend the borders or defeat an insurgency—it is to maintain this internal balance within the political elite, and to do this typically requires a mixture of cronyism and corruption in the armed forces.
Corruption buys the loyalty of the largest armed militia in the country: the army. Cronyism reinforces the armed forces’ loyalty by installing as senior commanders not trained technocrats but relatives, co-ethnics, political supporters, or representatives of allied political factions. Together such techniques bind the armed forces to the civilian leadership; just as important, they limit the threat that a powerful, technically proficient, politically disinterested army would pose to every other armed body in the country—most of which are in the hands of warlords and other elites. If the Afghan army really did professionalize itself by replacing cronies with technocrats, rooting out corruption, and promoting based on merit rather than political alignment, it would pose an existential threat to dozens of warlord militias that now extract resources for the benefit of their followers, and see the right to do so as their due given the strength of their armed following. Even worse, the first wave of new military technocrats would threaten their own corrupt, cronyist superiors. In a political system where order is the product of the internal balance of armed power and not judicial or legislative institutions, the very process of reform is dangerously destabilizing. The result is a military that is actually very good at its primary purpose—maintaining internal political stability among armed elites—but is very poor at what Americans mistakenly suppose is its purpose: defeating an insurgency.
The unsurprising result of this is that light-footprint U.S. training and assistance has had deeply disappointing effects. U.S. equipment and logistical support is commonly redirected into the black market for the financial benefit of officers; training is used as a form of largesse to reward loyalists; U.S. financial aid underwrites ghost soldiers who exist on the payroll but not in the field. There are exceptions: Elite ANSF commando units too small to be a threat to the internal balance of power, for example, can be allowed to professionalize and often perform well in combat. But the ANSF as a whole cannot defeat the Taliban insurgency—and more U.S. advisers or a bit more U.S. aid cannot change that. Thus, reliance on a much-cheaper indigenous ANSF cannot secure even limited aims inexpensively.
Stalemate and its Discontents
What reliance on the ANSF can do is to preserve a rough stalemate—or at least, preserve it as long as outsiders pay most of the bills. The ANSF will never be the U.S. Marine Corps, but neither are the Taliban, and the ANSF now fields over 300,000 soldiers and police. The ANSF also benefits from increasingly intense U.S. air support, which dropped more than 2,500 bombs in support of ANSF operations in the first six months of 2018 alone, a pace that would more than double the war’s previous annual peak of 2,170 in 2011. An ANSF that vastly outnumbers the Taliban insurgency and is backstopped by this kind of U.S. firepower has proven generally (if barely) able to hold most of the ground it took over from the Americans after 2011.
At the same time, the Taliban have expanded their control in traditional strongholds such as the central Helmand River Valley, and have mounted short sallies into urban areas such as Konduz or Ghazni before being ejected by counterattacks (mostly by Afghan commandoes and U.S. special forces). But they have yet shown no ability to take and hold any of the major urban areas that are the core of government power in the country. Seven years after the U.S. drawdown began, the Taliban still control less than 15 percent of the country’s 407 districts (they contest another 30 percent). The trends have been turning gradually in the Taliban’s favor, but at this rate it would take decades for them to complete the job were military factors alone to be the deciding ones. If the ANSF stays in the field and operating, the military prognosis is thus a continued grinding stalemate for many years to come.
The problem is that the ANSF probably can’t stay in the field forever. Their morale is increasingly a concern: ANSF loss rates are high and growing, and the Afghan government is finding it ever harder to fill the ranks.
Their funding is another major concern. A 300,000-person security force is vastly larger than what the Afghan government can afford on its own. The ANSF’s FY 2013 operating budget of $6.5 billion was more than twice the Afghan government’s entire federal revenue. The lion’s share of the money needed to keep the ANSF in the field must thus come from abroad, especially from the U.S. Treasury, which now pays almost $5 billion a year for Afghan soldiers and police. In a sense, this is a bargain: It’s vastly cheaper than the $100 billion a year the U.S. government spent when there were 100,000 American troops in Afghanistan. But the political constituency in the U.S. Congress for Afghan soldiers is minimal, and the current Commander-in-Chief is clearly frustrated with any cost at all. It took a major sales job from the Defense Department earlier this year to persuade Donald Trump not to pull all U.S. troops and funding from the war, and a mercurial and impulsive President could easily change his mind. If the plan for securing U.S. interests is to fund the ANSF forever to keep the war on indefinite life support, it’s a plan that could be ripped to shreds by a single White House tweetstorm.
Were the ANSF to be defunded or break under the strain of stalemated war-making, Afghanistan would quickly collapse into chaos. The government in Kabul would fall, but an immediate Taliban restoration would be less likely than an atomized, multi-sided civil war in which factions of the old ANSF fought over the rump with warlord militias, the Islamic State, and the Taliban’s respective factions—much as happened in Afghanistan in 1992, the last time central government authority collapsed.
In the midst of such chaos, deals would very likely be made between local Afghan factions and Pakistani insurgents to call in old IOUs or broker new alliances that would enable access to strategic base areas on the Afghan side of the border for militants seeking to topple the regime in Islamabad. Such arrangements have already appeared on a small scale in border areas where Afghan government control is weakest. This would define defeat: The central war aims for which the United States has fought since 2001 would be forfeit, and the sacrifices and costs of the past rendered vain.
If the war simply drags on, sooner or later this will be the result. The only plausible alternative to eventual outright defeat at this point is not military victory, which is impossible at expenditure levels that Americans will ever accept, but a compromise political settlement with the Taliban.
To date, however, the negotiation prospects have looked bleak. A handful of initial contacts showed some promise, but a combination of confusion on the allied side (as Afghan President Hamid Karzai and U.S. officials failed to communicate successfully on the arrangements for talks) and a leadership succession crisis on the Taliban side (when the Taliban spiritual and political leader Mullah Omar’s death was announced) froze the process in 2015. Talks remained stalled for more than two years afterwards.
Change on the Horizon?
This politico-military deadlock might now be breaking, however—for better or worse. In November 2017, unofficial Track-Two meetings between Taliban representatives and former U.S. officials Christopher Kolenda and Robin Raphel signaled significant softening in the Taliban position. It dropped its prior insistence on total U.S. withdrawal as a precondition for talks, and it promised to oppose terrorism in a postwar Afghanistan. The U.S. government then dropped its former insistence that any talks be Afghan-led, and agreed to the Taliban’s demands for direct, bilateral meetings between U.S. government and Taliban representatives in Doha (formal “negotiations” must still be Afghan-led, but the U.S. government is now willing to conduct direct bilateral “talks”). The Taliban reciprocated a temporary ceasefire announced by Afghan President Ashraf Ghani in June, and the resulting truce led to an outpouring of Afghan support for an end to the fighting. Widespread demonstrations in favor of peace led to scenes of Taliban fighters and government policemen celebrating in the streets over ice cream. These are baby steps, at best representing the beginning of what would be a long peace process if one formed at all. But against the backdrop of years of stalemated warfare with little or no progress toward settlement, the pace of change in the last few months has been breathtaking.
Why now? And does this apparent thaw have a future?
Part of the explanation for this recent movement may lie in the resolution of the Taliban succession crisis following the announcement of Mullah Omar’s death in 2015. Omar had actually died two years earlier, but the Taliban kept his demise secret to avert precisely the internal political crisis that broke out when the Afghan government announced his death to the world on July 29, 2015. His immediate successor was then killed in an American air strike in November 2015, yielding further turmoil. Eventually, former cleric Hibatullah Akhundzada emerged as the new leader, but with potential rivals in the wings he dared not risk appearing weak by promoting settlement negotiations; instead, he escalated the war. Over time, however, the new leadership has consolidated power, making successful negotiations plausible.
Another part of the explanation may lie in the end of the U.S. government’s previous insistence on shooting itself in the foot with repeated public deadlines for troop drawdowns. For the six years between President Obama’s West Point speech of December 2009 and his Administration’s announcement in July 2016 that further withdrawals would be halted, U.S. troop strength in Afghanistan was always subject to a timetable stipulating dates for incremental withdrawals. Each of these drawdown dates created the possibility that a weak ANSF might collapse with the next reduction in U.S. support, giving the Taliban an incentive to withhold concessions and wait to see whether they could get what they wanted for free. When President Obama finally ended this policy in 2016, the Taliban faced the prospect of a potentially permanent U.S. presence, which the Trump Administration eventually reinforced with its own declaration this year that there would be no announced date for withdrawal. One could debate this promise’s credibility, but at least there are now no looming deadlines to encourage the Taliban to wait.
Perhaps the most important explanation, however, may be the looming peril of state collapse in Afghanistan if the war continues much longer. Internal politics in weakly institutionalized states are often fragile, and Afghanistan’s are increasingly so. The country’s last presidential election, in 2014, was deeply flawed. Many Afghans believe the winner, Ashraf Ghani, stole the election through large-scale fraud; to avert potential violence in the streets, U.S. officials brokered a compromise in which the second-place finisher, Abdullah Abdullah, was granted an ambiguous official status as the country’s “chief executive.” The rival camps were encouraged to divide subordinate offices between them, and both agreed to hold a constitutional convention within two years to formalize Abdullah’s new office.
The result has been four years of increasing acrimony as the two camps have bickered over the spoils of office and effectively paralyzed the government. In the process, Abdullah has been increasingly marginalized, and the constitutional convention has been indefinitely postponed. Meanwhile Ghani has pursued an anti-corruption campaign that Abdullah’s supporters see as a means of targeting them. Last year a coalition of warlords, including one of Ghani’s own vice presidents, Abdulrashid Dostum, formed an alliance to oppose him, escalating the dispute along lines reinforced by the warlord alliance’s access to armed followers and deep connections to elements in the ANSF.
Presidential elections are now scheduled for early 2019; if held, they are liable to be even more divisive than the last round. The competing camps are now deeply embittered even as the U.S. government’s ability to play honest broker has atrophied. The Taliban’s increased influence in the Pashtun south and east, too, is an even greater threat to electoral legitimacy. In a country where Ghani’s Pashtun political base is located disproportionately in the increasingly Taliban-threatened south and east, where voting is dangerous, while Abdullah’s Tajik and Uzbek base lives disproportionately in the safer north, where voting is easier, any result can be seen as illegitimate by the loser. After seeing themselves as railroaded following a stolen election in 2014, locked out of real power, and persecuted by the illegitimate winner, Abdullah’s northern base is in no mood for a repeat—yet Ghani’s Pashtun base would see an Abdullah victory as stolen via the inability of their voters to reach safe polling places. In this increasingly combustible mix, no electoral outcome will be stable; it is all too easy to imagine violence breaking out along factional lines in ways that outsiders could not tamp down this time around.
If a disputed election in an increasingly polarized and distrustful political system results in factional violence, the ANSF itself would come under severe pressure. The armed forces have long harbored competing northern Uzbek/Tajik and eastern Pashtun factions as one of the more consequential of its various internal fault lines. If pulled in different directions by orders to suppress a defeated party’s violent post-election demonstrations, the pressure could finally push an already-strained security force beyond its breaking point.
At the same time, Afghanistan’s neighbors are playing increasingly meddlesome roles on behalf of their own preferred proxies. Pakistan, of course, has long been deeply involved in supporting its Taliban allies in Afghanistan. But Russia, too, is playing an increasingly direct role in support of both the Taliban and several of Russia’s preferred non-Taliban warlords. Moscow is doing this partly as a means of hedging Russian bets against possible Afghan state failure, and partly to impose costs on the United States amid a deteriorating U.S.-Russian relationship. Iran, with similar incentives, is playing a similar role with its own internal allies. India has longstanding ties to a variety of northern Afghan factions dating back to the civil war of the 1990s and beyond; the Trump Administration has dangled the threat of supporting a larger Indian role in Afghanistan as a cudgel to inspire Pakistani cooperation against the Taliban, but unless deftly handled this threat could just pour gasoline on glowing coals. Pakistan’s chief interest in Afghanistan has long been to avert Indian encirclement via an Afghan-Indian alignment; a more prominent Indian role with northern proxies could easily inspire a less cooperative Pakistani response.
Together, these dynamics pose a serious risk that the coming election could be the straw that breaks the camel’s back, splitting the government and security forces, pulling meddlesome neighbors even deeper into proxy warfare, and putting Afghanistan onto a short road to Syria-scale chaos. If so, the worst-case scenario for all the major players would simultaneously come into play. It would obviously be bad for U.S. interests. But it would be little better for Russia, Iran, or Pakistan—each of whom would now confront a failed state with ungoverned spaces offering safe haven to their own enemies either directly on their border (for Pakistan and Iran) or in close proximity (for Russia). Faced with this possibility, moreover, each must protect itself with aid to sympathetic proxies—but simultaneous aid to opposed Afghan proxies would just deepen the crisis at higher levels of armament.
For Pakistan in particular, this scenario would pose grave dangers, creating both hostile insurgent camps in Afghan sanctuary plus Indian assistance to anti-Pakistan militants to Pakistan’s strategic rear. Pakistan has long sought to thwart the American project of creating a strong Afghan state that might align with India and threaten Pakistan with strategic encirclement, but a failed state on its border that drew India into the maelstrom could be even worse.
Even the Taliban apparently recoils at this prospect. In their Track-Two conversations with Kolenda and Raphel, Taliban representatives cited the rising danger of a Syria scenario as a major concern and as a reason to negotiate. Patriotism aside, a rational Taliban strategist would have good reasons to take this view. Internationalized proxy wars like Syria’s are notoriously hard to terminate once they metastasize. The Syrian war has raged for seven years now, and historically such conflicts have typically lasted seven to ten years, often laying waste to whatever was there before and leaving little behind but bitterness and ruin for whoever inherits the remains. The Taliban came into existence in the first place to end just such a conflagration in the 1990s, but despite their relative success then, they never fully pacified the country: Their Northern Alliance enemies still held about 15 percent of the country when the United States intervened in 2001 and toppled the Taliban regime. A repeat performance now could be even harder for them, with deeper outside involvement brewing that could involve the United States, too, this time around as a backer of its own preferred warlord proxies.
This time, too, the Taliban would have to contend with the Islamic State in Afghanistan—an enemy of both the government and the Taliban, a bitter rival to the Taliban for the status of defender of the faith, and a growing threat that has fought multiple pitched battles against the Taliban even while waging war against the government and the Americans. The Taliban might once have thought they could defeat the ANSF and take control of a coherent state; that now seems unlikely. Hence they have good reason to explore negotiated alternatives to another decade of what would become a multi-sided, internationalized civil war. The Taliban have already been at war with the United States for 17 years, losing untold thousands of their followers, forcing them to live as exiles in Pakistan and raise their children as Pakistanis for almost a generation, and running an annual risk of personal violent death in American air strikes (just ask Akhundzada’s predecessor Mullah Mansoor). The Taliban would not surrender just to save their country from a Syria scenario, but a compromise settlement that could avert all this without looking like a surrender instrument ought to be increasingly worthy of serious Taliban attention. It makes sense for them to explore a deal while there is still a government to deal with.
The Way Forward
What, then, should the United States, and others, do to exploit this opportunity and avert a mutual worst-case scenario? Several initiatives would help.
First, the United States must be prepared to accept serious compromises, and to encourage its Afghan government ally to do the same. The Taliban may be willing to parley, but they will not sign a surrender instrument. Compromise with an actor like the Taliban is not an appealing prospect, but military victory is a chimera and stalemate cannot be maintained forever. That leaves compromise or outright defeat as the only realistic prospects.
What does that mean in practice? Compromise will require legalizing the Taliban as a legitimate actor in Afghan politics, releasing Taliban prisoners from Western prisons, withdrawing all foreign troops and closing all U.S. bases if the Taliban continue to insist on this, and eventually offering them some extra-democratic set-aside of guaranteed offices, ministries, and/or parliamentary seats.
With today’s hyper-partisan U.S. domestic politics, compromise on this scale could be hard to sell; in 2012, a much-smaller proffer to exchange some Afghan prisoners for the Taliban-held American Sergeant Bowe Bergdahl collapsed under Republican outrage at a Democratic President’s willingness to make such a deal (Bergdahl was finally released two years later). But if a U.S. President can manage the domestic politics, an international agreement might be within reach: Afghan President Ghani has already offered to legalize the Taliban, exchange prisoners, review the Afghan constitution, and provide passports and safe passage for Taliban negotiators. The actual difference in the two sides’ bargaining positions has now grown surprisingly small: The Taliban have already signaled willingness to forswear terrorism, break with al-Qaeda, and to accept an Afghan state governed by something similar to today’s Afghan constitution, all critical elements of longstanding U.S. and Afghan government positions. And the June ceasefire demonstrated that the Taliban are sufficiently unified to make a deal stick if it can be reached. Still, the remaining differences are hardly trivial (including the nature of any eventual power sharing deal and the specifics of what a sufficiently Islamic state would comprise), and an actual deal will require concessions from allied governments as well as the Taliban.
Second, the United States will have to remain engaged both politically and economically in any postwar Afghanistan to stabilize the terms of a negotiated settlement. A legalized Taliban with a role in Afghan governance would not violate core U.S. interests, but an outright Taliban takeover would. A meaningful U.S. postwar assistance budget conditioned on Afghan government behavior is an important hedge against that possibility. A post-settlement Taliban’s best chance for complete power would be if the current government’s kleptocratic scale of corrupt extraction from the economy eventually alienates enough voters for a legalized Taliban to win national elections. Thoroughgoing reform is beyond American means today, but conditional aid on a plausible scale can at least discourage the worst kinds of government land grabs and other grand-mal corruption, and thus help keep a lid on a legalized Taliban’s electoral prospects.
Ongoing conditional aid can also provide a disincentive for rearmament or other Taliban violation of settlement terms: If a deal is generally better for the Taliban than war, and if Taliban participation in a coalition government affords them a share of international largesse, then a threat to withdraw aid can offer the incentive at the margin needed to discourage violations. Of course, this does not mean the U.S. government must spend $100 billion a year to keep Afghanistan stable. But Afghanistan was a major recipient of international aid for much of the 20th century, and a stable postwar Afghanistan will probably have to be, too—which means the U.S. government will need to continue to provide aid, and to remain sufficiently engaged politically to know whether and when to threaten aid withdrawal if its preconditions go unmet.
Finally, an acceptable end to the war will require strategic patience from American statesmen. This will not be a quick or easy negotiation. Progress has been surprisingly rapid recently, but to bring this to fruition will probably take years, not months. A gradual process of confidence-building measures is required, to be followed only later by concrete talks over difficult issues such as power sharing, constitutional change, disarmament, demobilization, and reintegration. Before any deal can be signed with the Taliban, too, domestic support within Afghanistan must be built for prospective settlement terms, a process that could be long and fraught in itself. If Americans tire of a long process that moves slowly and fails to yield a quick, showy deal, then no deal will be had.
To provide this kind of time will require a near-term agreement to postpone Afghanistan’s upcoming election and to substitute an interim caretaker government while negotiations with the Taliban unfold—perhaps with Ghani frozen in place but with a promise that he would not run for re-election afterwards, or perhaps with some other formula. The now-scheduled elections could easily serve as the match that sets Afghanistan’s political tinder ablaze; if so, the state could collapse long before negotiations could conclude. The risk of such a conflagration is an important incentive for all parties to negotiate seriously, but the timelines of political risk and negotiated solution are now badly misaligned. An election in early 2019 would undermine peace prospects no matter how it comes out. If the election yields chaos it will derail a long negotiation; if the election yields a stable transition it will remove an important goad for talks. Any caretaker arrangement will itself have to be negotiated, and U.S. leverage to strong-arm recalcitrant power brokers will be needed for any such setup. Without strong, sustained U.S. involvement it is hard to see how Afghanistan’s many factions could be brought in line behind any given choice for a caretaker, even a temporary one.
Moreover, this internal negotiation over the terms of any caretaker government may require an early ceasefire with the Taliban to enable it to unfold: ANSF effectiveness in the midst of prolonged uncertainty over who is in charge could well be even lower than today, risking serious losses if exposed to Taliban offensives before a new caretaker is appointed and lines of command clarified. A truce of this kind would be far from a final settlement to the war, and so could begin quickly, as the June ceasefire did. But it is important that it begin. If the Taliban decide not to facilitate a settlement with an early truce and instead try to exploit Afghan political instability for short-term battlefield gains, then settlement could become impossible—and if so, then the Taliban, like everyone else, will witness its worst-case scenario of chaotic state collapse unfold. To the extent that the Taliban’s recent interest in negotiation is sincere, this interest will have to include a willingness to accept a truce while the Afghan government assembles a negotiating partner for them. None of this is impossible, but neither is it easy, and the elections create a demanding timetable. To get a ceasefire by early 2019 and a delay in Afghanistan’s elections, the first steps must be taken soon.
Successful talks could also require international mediation, perhaps via the United Nations, and an explicit role for Pakistan among other Afghan neighbors in a parallel regional negotiating track. This, too, could be tough to swallow for a U.S. Administration that has been no friend of the United Nations and which has positioned itself to get tough on a duplicitous Pakistani ally rather than to welcome them into multilateral negotiations.
Given all this, skepticism about the prognosis for a negotiated end to the war is still a defensible position: Any talks will be long, hard, and perhaps ultimately unsuccessful; in the meantime, the United States will have to continue military assistance and economic aid on something like today’s scale. Some may judge the cost of playing out this string to be higher than the real-but-limited stakes involved for the United States. The cost-benefit calculus for war in Afghanistan has always been a close call. But the costs are now a great deal lower than they were in 2009-11, and the prospects for a tolerable settlement are now better than they have been in years, perhaps since 2001. An Afghan Syria is now a distinct possibility. But so is a settlement that could terminate the war, limit the danger of Afghan instability spilling over into the region, and end the suffering of an Afghan people who have suffered much over generations of almost continuous warfare. These are important benefits, to Afghans, to other South Asians—and to Americans.