9 June 2015

Determining U.S. Commitments in Afghanistan

by Stephen Watts and Sean Mann
May 01, 2015

As the Obama administration's tenure winds down and the United States withdraws nearly all of its troops from Afghanistan, debates about the nature and scale of future U.S. involvement in Afghanistan continue.1 President Obama has committed to withdrawing all but a minor residual force by the end of 2016. On the other hand, Afghan President Ashraf Ghani has called for a sizeable U.S. military presence long after President Obama leaves office, and recent setbacks in Iraq have strengthened the hand of Congressional leaders,

U.S. military officers, and others who call for an enduring commitment.2

Skepticism in the United States about an ongoing U.S. military role in Afghanistan abounds among both the public and many foreign policy experts; indeed, such skepticism was recently labeled the “conventional wisdom.”3 In 2014, for the first time since the United States first committed troops to Afghanistan, a plurality of Americans (49 percent) thought it was a mistake to have ever sent forces to that country.4 As of July 2014, only approximately one-quarter of Americans thought President Obama was withdrawing U.S. forces from Afghanistan too quickly; nearly three-quarters thought the current timeline of withdrawal by the end of 2016 was either appropriate or too slow.5

This pessimism is rooted in three arguments. It derives first, and most viscerally, from disappointment in the results of the past thirteen years of international intervention. Although Afghanistan has made advances (to an extent not generally perceived in public discourse in the United States),6 these gains are overshadowed by continued violence and a sense that all such improvements are likely to dissipate as soon as the international community withdraws.

Second, skeptics of continued U.S. commitment point to the diffusion of al-Qaeda and other radical transnational Islamic terrorist threats. Recent events in Iraq, Syria, Yemen, Libya, Mali, Nigeria, and elsewhere have highlighted the scope of the dangers these groups pose. With “core al-Qaeda” so heavily degraded in Afghanistan and Pakistan, why should the United States continue to focus there when other regions arguably pose greater threats to U.S. interests?

Finally, even if a continued substantial U.S. commitment to Afghanistan is justified on strategic grounds (i.e., one beyond the vestigial force of up to 1000 troops in a predominantly advisory role that is currently contemplated in U.S. plans), some observers question whether U.S. domestic politics can maintain financial and military support over the course of at least another decade. Particularly when Afghanistan suffers likely setbacks—future corruption scandals, disputed parliamentary elections, battlefield losses to the Taliban— would Congress continue to write checks to support the Kabul government?7

These concerns are justifiable. Afghanistan does not deserve a blank check, nor can U.S. politics sustain one. But Afghanistan does deserve continued U.S. support, and a sustainable commitment requires rethinking how the United States conducts military interventions to support fragile states—a transformation in approach that will not come easily to a foreign policy and defense community used to more decisive outcomes.

This article first sets out the reasons why Afghanistan is—conditionally— worth continued U.S. commitment well beyond the end of the Obama administration. It then examines the challenges that the United States will face in realizing its goals through a discussion of both Afghanistan and similar environments. Although not insuperable, these challenges are substantial. U.S. decision makers will have to enter into a long-term commitment to Afghanistan with the understanding that the stakes in the country make a relatively low-cost gamble on its future advisable, but the odds of failure are nonetheless sobering. Finally, it lays out the basic elements of a long-term strategy for Afghanistan, including ways to keep costs within the bounds justified by expected gains and a discussion of the “red lines” that should trigger U.S. withdrawal if breached.

Worth Continued—Although Modest—Commitment

Afghanistan had already been devastated by over two decades of civil war when U.S. forces arrived in October 2001, less than a month after al-Qaeda's attacks. At least one in six Afghans had fled their homes for Iran or Pakistan, creating the world's largest refugee crisis. Hundreds of thousands of Afghans had been killed; millions had been wounded. An estimated 13 percent of children died before the age of five, and few children between the ages of five and seventeen attended school—just 29 percent of boys and 0 percent of girls in 2001. Annual per capita income averaged $115, less than half that of Nepal, the second- poorest country in Asia.8

The relative security that the international presence brought, combined with growing state capacity and massive foreign aid, has contributed to the return of more than five million refugees and allowed for some degree of recovery. By 2012, child mortality rates had fallen to 9.8 percent, more than half of all children (40 percent of them girls) were in school, and annual per capita income averaged $687. A 2009 survey asked Afghans to identify the “most harmful” period of recent Afghan history: 38 percent named communist rule from 1979–1992; 22 percent the subsequent civil war from 1992–1996; and 33 percent Taliban rule from 1996–2001. Just 3 percent identified the post-2001 conflict as the most harmful.9

Of course, the United States did not go into Afghanistan to develop its economy or end its civil war. It did so to defeat al-Qaeda and the Taliban regime that gave the group refuge. Thirteen years in, the results are frustratingly incomplete. Al-Qaeda's leaders escaped to Pakistan early on, though many were eventually captured or killed. Some core members reportedly remain in a few remote parts of eastern Afghanistan and Pakistan, cautious survivors of a drone campaign that has left them temporarily incapable of organizing attacks, but looking ahead to when U.S. pressure ends. The Taliban, on the other hand, were decisively defeated by late 2001, only to regroup and launch an insurgency in the years thereafter. Insurgent violence grew steadily until 2010, and has since plateaued. From 2010 to 2013, annual Afghan civilian deaths due to conflict hovered between 2,750 and 3,150, although in 2014 they jumped to 3,699, largely due to increased ground engagements between insurgents and Afghan forces. Insurgents were responsible for 72 percent of civilian fatalities in 2014, up from 56 percent in 2010. U.S. military fatalities have decreased nearly 90 percent from 2010 to 2014, reflecting their ongoing withdrawal from the battlefield.10

At the same time, the much expanded and less well-equipped Afghan National Security Forces (ANSF) now bear the brunt of insurgent violence. Despite high casualty rates—especially among the police—and other significant challenges, the ANSF has defied expectations to become a passably capable, resilient, and nationally representative force.11 Most recently, the ANSF provided better security with far less assistance from international troops during the 2014 elections compared to the 2009 and 2010 elections.12

Unless the United States withdraws its support or the new Afghan government fractures due to ethnic tensions, the current stalemate between the government and anti-government forces (above all the Taliban, but also a variety of other groups such as the Haqqani network) will likely continue for at least several years. Such a stalemate falls far short of what many had hoped for in the early days of the U.S. intervention.

It does, however, accomplish what the United States needs in the region—at least until a more durable settlement is possible. First, unless the

U.S. relationship with the Afghan government disintegrates, it provides critical bases for U.S. counterterrorism operations. While the United States could continue these operations from sites outside of Afghanistan, these alternative sites come with substantial drawbacks that would severely constrain U.S. intelligence and military operations.13

Second, a relatively moderate, pro-U.S. government in Kabul, combined with continued active U.S. military engagement, reduces the extent to which Afghanistan can be used as a sanctuary for groups attempting to destabilize Pakistan. Pakistani officials fear that Afghanistan will become a sanctuary for the Tehrik-i-Taliban Pakistan and other groups fighting the Pakistani state as the last of the U.S. soldiers depart. A weakened state and intensified conflict in Afghanistan could result in large numbers of refugees fleeing to Pakistan and provide an opening for militants heading in the other direction. While most militants fighting the government of Pakistan are based in Pakistan itself, several groups have found safe haven on the Afghan side of the border since at least 2010.

There have also been disturbing indications that the Afghan intelligence service has sought to recruit Pakistani militants to offset Pakistan's support for militants fighting the Afghan state.14 This practice could become more prevalent if U.S. support for and influence over the Kabul government were to decline precipitously. The evolution of the flow of fighters between Pakistan and Afghanistan will depend greatly on the policies of the government of Pakistan. Recent signs suggest that Islamabad has adopted a more constructive approach to the conflicts in its frontier regions, which if continued would greatly improve the odds of an acceptable outcome on both sides of the border. Given the instability of the government of Pakistan, however, the United States cannot count on these recent improvements to continue.15

Coping with State Fragility

That said, an extended period of military stalemate between the government and anti-government forces is hardly ideal either for U.S. decision makers or for Afghans. But judging by recent events in Afghanistan and by the history of similarly weak states, the best the United States may be able to do is accept the reality of the current stalemate and seek to nudge it toward increasing levels of stability and political inclusiveness over an extended period of time—likely at least the next five to ten years—with the ultimate goal of a negotiated settlement.

Unfortunately, in the short to medium term, there is little prospect for a settlement with anti-government forces, even if the United States committed wholeheartedly to achieving one.16 Peace deals are generally understood to have two primary prerequisites. First, they require mutual agreement on the relative balance of power between the warring parties (usually in practice obtained the hard way—through a long and painful military stalemate that makes clear to all participants the need for compromise). Second, they require mechanisms to ensure that the parties live up to their commitments in the peace deal, without which any of the warring parties' promises lack credibility.17 In Afghanistan, neither of these conditions is present.

Neither the Taliban nor anyone else is certain how government forces will fare as the Western withdrawal proceeds. Insurgents will want to test the resilience of the Afghan government as international support dwindles, hoping to secure vastly more returns through continued fighting than they could hope to secure through negotiations at this time. These groups have been at war with the United States for over a decade; they and their predecessors have been fighting more or less continuously since the late 1970s. Waiting a handful more years to see how fortunes change after the foreign military drawdown is not a major sacrifice for these actors.

Moreover, the government in Kabul faces a major credibility gap. In the years following the 2001 U.S. intervention, a number of actors in the Taliban—or factions affiliated with them—sought to reach an accommodation with the government. Many of these overtures were rebuffed or exploited for gain by government officials, often resulting in the persecution of those who had sought peace.18 Abuses by the Karzai administration and its allies in recent years have done nothing to improve the estimation of the Kabul government in the eyes of anti-government forces. Until the new administration led by President Ghani in conjunction with Abdullah Abdullah can demonstrate an ability to credibly commit to political compromises with dissident groups, an overarching peace deal is highly unlikely.

It is for these reasons that many observers of Afghanistan are highly pessimistic about the chances for a peace deal in the short term. The Afghan High Peace Council Chair, the outgoing U.S. commander in Afghanistan, and others involved in negotiations did not expect the Taliban to discuss peace seriously anytime soon, given insurgents' hope that victory will follow the removal of international troops.19 Even those who viewed instances of Taliban restraint during the recent elections as a positive sign for peace talks caution that an agreement will not arrive shortly.20 Certainly, the Taliban have not altered their rhetoric or actions following Karzai's departure, rejecting the new government as an unacceptable `sham' and carrying out several deadly attacks on inauguration day.21 More recently, while the new Afghan government has raised hopes for negotiations, the Taliban continues to publicly oppose talks.22

Afghanistan is not unique in the challenges it faces in securing a durable peace deal; in fact, its situation is quite common. Decades ago, civil conflicts typically ended in victory for either government or rebel forces. Today, stalemate is the single most common military outcome for countries that have weak state capacity and fail to politically incorporate significant subpopulations, as Afghanistan has—an outright government victory occurs in only one case in five.23 Moreover, such countries are at high risk of descending back into war, even if they are able to secure a peace in the short term.24

Improved governance typically provides the surest path out of such perpetual conflict. It does so by offering goods and services that give everyone a stake in peace and by providing mechanisms to assure contending parties that they will not be excluded from these benefits (i.e., through credible power- sharing arrangements). Political scientists are divided, however, about what aspect of good governance best secures peace—whether it is the capacity of the state to provide goods and services, or the political inclusion and power-sharing that help to ensure such goods and services are distributed equitably.25 The obvious answer is that both matter, although this does not help decision makers set priorities.

Unfortunately, prescriptions focused on good governance offer little reason for hope in the short term for Afghanistan. Substantially improving state capacity typically takes decades.26 The past decade of massive resource infusions from the international community has led to large strides for Afghanistan, but the government remains beset by a host of capacity shortfalls, and corruption remains rampant. Although such corruption, currently at stratospheric levels,27 may decline as the amount of foreign assistance falls off in coming years, it is hard to see how other aspects of state capacity can improve in the coming ten to twenty years with much-reduced levels of international support.

Improvements in the quality of political inclusion, on the other hand, are possible and would help the government credibly commit to a power-sharing arrangement that includes elements of the Taliban and potentially other groups (such as the Haqqani network and Hezb-i-Islami Gulbuddin) in an eventual peace deal. Politics in Afghanistan have absorbed many once-violent struggles for power, and politicians have repeatedly sought pragmatic consensus over strife. Though informal powerbrokers make crucial decisions behind the scenes, they still largely adhere to the rules set in the constitution on issues such as legislative and executive powers, term limits, and political appointments.28

Despite the nearly exclusively Pashtun nature of the Taliban movement, the current conflict has so far been remarkably free of explicitly ethnic and sectarian violence. The government has proven fairly successful at recruiting Pashtuns— at least from eastern Afghanistan—into the security forces, and has only rarely engaged in indiscriminate violence against communities suspected of supporting insurgency. For its part, while the Taliban has largely failed at recruiting non- Pashtuns, it continues to posture as a nationalist rather than an ethnic insurgent movement, and has generally refrained from explicitly ethnic and sectarian attacks. This contrasts sharply with rising sectarian violence in the rest of the region, as well as Afghanistan's experience in the 1990s, which included ethnic massacres carried out by the Taliban. While minority groups have a justifiably intense mistrust of the Taliban, the relative absence of ethnic violence could ease acceptance of a negotiated end to the conflict.29

Political inclusion at the national level, however, remains fragile and at least somewhat dependent on international support. Furthermore, societal divisions at the local level continue to run deep, as seen in patterns of ethnic and tribal bloc voting, inter-village grievances, and unresolved resource disputes.30 Populations excluded from access to government patronage or disproportionately targeted by pro-government forces have often turned to the insurgency, particularly in the areas hardest hit by the conflict.

Without a markedly more capable Afghan government, wider and more effective forms of political inclusion likely will prove necessary to bring the insurgency to a close through a peace deal—or at least contain it within a limited number of peripheral regions. Unfortunately, most studies of externally- induced democratization, liberalization, power-sharing, and other forms of political reform suggest serious limitations on what the United States can do to press for greater inclusion in Afghanistan. Historically, outside actors enjoy the greatest success in pressing political reforms on other states when the linkages (historical ties, trade relations, and so on) between the countries are extensive and where the partner nation is highly vulnerable to external leverage. The United States has few such links to Afghanistan, and U.S. leverage has historically been quite limited when the partner nation is of strategic value.31

This review of the broader record of insurgency and conflict termination suggests that three critical facts are likely to characterize Afghanistan's future. First, the country is probably in for an extended period of continued conflict amid military stalemate; neither victory nor a durable peace deal looks obtainable in the short term. Second, short-term pessimism about the prospects of a peace deal does not imply long-term pessimism about a negotiated solution. Over time, two things will change. If the Afghan government is able to sustain a hurting stalemate well after the international drawdown has occurred, then the current uncertainty about relative capabilities will gradually disappear; it will become apparent to the Taliban and other anti- government actors that they cannot obtain their goals through continued fighting, making a negotiated settlement more likely. Moreover, if the Afghan government can slowly improve its record on political inclusion, then it will be able to more credibly commit to an eventual peace deal or deals with insurgent groups. Third, although the country's best path out of its “conflict trap” is through credible commitments to more inclusive governance, there is relatively little the United States can do from the outside to move that process along quickly. What, then, does a successful U.S. strategy for Afghanistan look like?

A Long-Haul Policy in Afghanistan

If a durable peace deal is highly improbable in the short term and immediate withdrawal is highly unpalatable, the United States must instead find a way to sustain a modest level of commitment to Afghanistan over a lengthy period of time. Such a U.S. policy for the long haul should accomplish three things: it should strive for conflict termination in the medium and long term; prevent the collapse of the Afghan government or the re-consolidation of transnational Islamic terrorist networks in the short term; and maintain U.S. domestic political support for such a policy—likely for at least the next decade, although at resource levels much lower than currently required.

In the short term, the two biggest threats to the Afghan government are fiscal collapse or the fractionalization of Afghan security forces (and if the first occurs, the second will almost certainly follow).32 It is important to note that the Soviet-backed Afghan government only collapsed after Moscow withdrew its financial support, not after it withdrew its ground forces.33 Although continued financial support from the United States and its allies is necessary, efforts to limit corruption to less-than-catastrophic levels and to reinforce cohesion within the ANSF will likely be equally important. Although both better and worse trajectories can easily be imagined, the most plausible future scenario involves an Afghan government that is weakened but able to endure more or less indefinitely with international support. In this scenario, the government loses control of many peripheral regions in Afghanistan, but maintains sufficient coherence to govern “core” regions in conjunction with regional powerbrokers, as well as to protect the infrastructure (e.g., roads, electrical generation and supply) necessary to support those regions.

Such an outcome would require U.S. military and civilian support to accomplish three goals. First, support must preserve the cohesion of the ANSF (recent events in Iraq are instructive in this regard)34 and prevent the loss of key cities or lines of communication to anti-government forces. Second, it would be necessary to ensure that transnational terrorist networks do not again strengthen their global reach in portions of Afghanistan and Pakistan outside of these governments' effective control. And third, U.S. diplomats and aid workers should work with the government to prevent catastrophic corruption and nudge decision makers toward greater levels of political inclusion over time, eventually including peace deals with those anti-government forces willing to compromise.

U.S. military support to accomplish such goals would generally take the form of a long-term, sizable advisory presence, coupled with periodic air strikes as necessary, both to prevent enemy forces from massing and to disrupt and degrade terrorist groups. The United States has already drawn down its military forces to 9,800, with a further planned reduction to around 5,000 at some point in 2016. After 2016, President Obama has stated that “our military will draw down to a normal embassy presence in Kabul, with a security assistance component, just as we've done in Iraq.”35 What exactly a “normal embassy presence” may be has been the subject of much speculation. A recent news report suggested a figure of up to 1,000 personnel, although at this point specific numbers appear to be at best well-informed guesses.36

Given recent events in Iraq, however, a “normal embassy presence” seems unwise in Afghanistan. Instead, the United States should maintain a small military presence consisting of advisors, intelligence and targeting capabilities, drones for occasional strikes of high-value targets, and force protection as well as logistical units sufficient to support U.S. personnel. Although the exact requirements should be left to military planners, a very similar mission in Iraq— an environment at least as challenging as Afghanistan—has numbered approximately 3,000 personnel.37 The U.S.-led Combined Joint Task Force–Horn of Africa, stationed in Djibouti since 2002, has provided an advisory presence throughout the nations of the Horn and acted as a hub for special operations; its overall personnel numbers have generally fluctuated between 1,000 and 1,600, augmented by numerous contractors. The United States has also maintained similar advisory missions numbering in the hundreds of uniformed personnel and contractors in the less challenging environments of Colombia and the Philippines for over a decade. Continued troop contributions from allies are likely to further augment the U.S. presence. As with current operations in Iraq, the United States could bring in greater air power if necessary to prevent insurgents from massing to threaten critical population centers or lines of communication—but such an expansion of operations should be undertaken only to protect truly critical objectives.38

Just as U.S. military personnel will continue to draw down in the coming years, U.S. military and civilian aid to Afghanistan will and should continue to decline to more sustainable levels. The United States recently announced that it would seek funding for the ANSF to remain at its current “surge strength” of 352,000 through 2017, after which it might draw down over a couple of years to a less costly “enduring force” of 228,000—less than two-thirds of its current strength—as originally planned.39 The 228,000 level envisioned for the ANSF is estimated to cost $4.1 billion annually. The United States pledged to initially cover $2.3 billion of the cost of this smaller force, with other international donors pledging $1.3 billion.40 Likewise, U.S. non-security assistance has already dropped from its 2010 peak of $7.1 billion to an estimated $1.6 billion in 2014.41 Further declines are almost certain in subsequent years.

What would such reductions mean on the ground in Afghanistan? The United States and other international actors must be careful not to withdraw aid too abruptly for fear of precipitating either a fiscal crisis or a crisis of confidence in the Afghan government's ability to endure. Reductions in finances to the ANSF must be managed in a way that will neither create massive sustainment challenges (e.g., failure to deliver fuel, ammunition, and other critical supplies on a large scale to front-line forces) or reduce ANSF personnel numbers much beyond the rate at which natural attrition would. Reductions in civilian aid must be phased in slowly and ideally should be shifted toward more on-budget assistance and local contracting to minimize the negative economic effects on Afghanistan.42 Appropriately phased in, however, these reductions need not be catastrophic.

At 228,000 forces, the ANSF would still remain the 27th-largest security force in the world.43 Today, the Afghan government remains in control of the major cities and highways as well as provincial capitals. Most remarkably, the government also remains in control—at least nominally and on most days—of every single one of the over four hundred district centers in the country, many of which are in remote, insecure areas far from Afghan army bases. In the past year, insurgents have attacked several of these district centers en masse, sometimes even overrunning one temporarily before being turned back by the ANSF.

In the coming year or two, with reduced assistance from international troops, the ANSF will no longer be able to repel all such attacks. As districts in a few isolated areas fall, the Afghan government will have to concede the fiction that they can maintain uniform control over the entire country, and will likely retrench to a more sustainable security posture which leaves portions of the periphery under insurgent control. As long as international support and government morale are not overly shaken by the sight of insurgent gains, however, the ANSF will still be able to hold the major cities, highways, provincial capitals, and at least the great majority of district centers. Stalemate is likely to proceed from there.

Similarly, although civilian assistance would fall considerably, it would remain an important portion of the Afghan economy. U.S. assistance has long stood at over a third of Afghan GDP and covers an even larger proportion of the Afghan government budget,44 with the United States projected to cover three- quarters of the Afghan security budget in 2015.45 Such assistance provides considerable incentive for mid-level Afghan political leaders to continue to support the Afghan government (albeit while cutting deals with anti- government actors and maneuvering to secure maximum advantage from both relationships).46

With limited alternative revenue sources available to the Afghan government, this assistance will also provide the United States and other international actors considerable leverage. According to World Bank figures, Afghanistan is one of the most aid-dependent countries in the world—a fact that not even substantial declines in levels of foreign assistance would fundamentally change.47 Moreover, with nothing but a small, residual U.S. military force on the ground, U.S. threats to reduce international assistance are likely to be much more credible than they were at the height of the U.S. commitment, potentially giving the United States much more leverage than it has had in earlier years.48

The costs of such a U.S. commitment would not be negligible, but they would be manageable. Taking into account Afghan government revenue, other donor contributions, and planned ANSF reductions, annual U.S. assistance is expected to fall from $6.6 billion in 2014 to under $4 billion by 2016, with further reductions almost certain in subsequent years.49 These anticipated levels of assistance are in the range of what the United States provides to Israel every year, and only a small fraction of the annual cost of the Afghan war during the surge. Within five years, these amounts should move lower still as Afghanistan's government gradually transitions from the maximalist objectives of the U.S. surge from 2009–2012 to a more sustainable model. Costs should hover closer to the $700 million per year that the United States provided to Colombia on average from 2001–2010 than to the more than $3 billion the United States provides Israel.50

The United States has also proven its ability to maintain troops numbering in the hundreds or low thousands deployed for a decade or more in support of foreign partners including in Colombia, Djibouti, Kosovo, the Philippines, and the Sinai Peninsula. Nor are recent events in Iraq, where complete U.S. withdrawal was followed by renewed violence and instability, likely to be lost on members of Congress and others who might otherwise press for a full termination of the U.S. military mission.

However, an open-ended commitment to financial support of a fragile government combined with perpetual low-intensity U.S. military operations is not sustainable—either domestically in the United States or with Afghan partners. Some path to conflict termination must be visible, even if it is a longer-term prospect.51 Safeguards or “red lines” on abuses by the Afghan government—whether catastrophic levels of corruption or large-scale abuses by Afghan security forces—must be in place. These imperatives suggest that the United States must condition its aid on respect for those red lines and continue to work for greater political inclusion over an extended period of time—working with reformers where possible and imposing real costs in terms of withdrawn aid where Afghan leaders threaten to undo existing power-sharing arrangements. Working with the government of Pakistan is also essential. Although there have been some encouraging signs from Islamabad, the prospects for a fundamental change in how Pakistan deals with Afghanistan are still uncertain.


Absent significant changes in Pakistani government policy and behavior, which are beyond the purview of this article, the odds of an outright military victory in Afghanistan are poor. It is not that insurgencies can never be defeated militarily; it is that the Afghan government is too weak, and the opportunities available to the Taliban and other insurgent groups too great, to make it practical in this case. Nor are the odds of a negotiated peace deal good in the short term. The two prerequisites for a peace deal—a painful military stalemate and a government that is capable of credibly committing to real concessions to acceptable portions of the armed opposition—are a long-term prospect. Any continued U.S. commitment, therefore, will also have to be oriented to the long term. That prospect implies three critical features of continued U.S. engagement.

First, U.S. engagement must be politically sustainable over the long haul in the United States. As a rule, such a commitment will require keeping both financial costs and troop commitments low—in many ways similar to what the United States has done in Colombia (2000–2008 in its most intensive period) and the Philippines (2001–2015), although admittedly on a larger scale and with lower chances of success than in these two relatively well-functioning states.

Second, the United States will have to balance its desire to “disrupt and degrade” transnational terrorist threats, as well as insurgent groups that threaten Pakistan's stability, with the recruiting advantages that continued U.S. military operations afford to these groups. So long as the Afghan government can function tolerably well, refrain from widespread abuses by its security forces, and be slowly nudged in the direction of greater political inclusivity, the strategic advantages of U.S. efforts to disrupt and degrade militants in the country likely outweigh the risks posed by offering them a theme to motivate their recruiting efforts.52 By the same logic, if future governments in Afghanistan prove utterly dysfunctional or abusive, the United States must be willing to terminate its commitment. Otherwise, the United States truly is making an open-ended commitment to war without a visible pathway to an acceptable end-state. Such a commitment benefits neither the United States nor the Afghan people.

But this reality implies a cautionary note about continued U.S. support to Afghanistan: it is a probabilistic enterprise. When the United States committed military support to the governments of Colombia and the Philippines, the odds were low that these governments would collapse entirely or that they would use their security forces for truly large-scale abuses (the smaller-scale abuses by some security forces of both governments notwithstanding). In Afghanistan, these outcomes are very real risks. Can U.S. decision makers sustain a limited, but not inexpensive, gamble through considerable tumult over the next decade or more without giving in to the twin temptations of early withdrawal or doubling down and escalating when the odds of success cannot justify such costs? The answer to this question is critical, because it is the question posed not only by Afghanistan, but potentially by other contingencies in the future.

For a relatively low price, the United States can likely prevent the government of Afghanistan from being defeated by insurgent forces, but will be unlikely to secure an outright victory. Instead, the United States will face an extended period of military stalemate and state fragility. During this period, the United States can degrade militant groups who threaten U.S. interests. In doing so, however, the United States risks an open-ended security commitment which may do as much to undermine U.S. interests as accomplish them, unless the Afghan government can be nudged toward more inclusive politics.

In Afghanistan, both the odds of success and the stakes appear to justify such a gamble. In addition to fighting terrorism and indirectly bolstering Pakistan's stability, the United States has an interest in securing a positive outcome so as to strengthen its and NATO's reputations for resolve. And the evolution of Afghan security forces over the past decade, combined with Afghan patterns of pragmatic accommodation across ethnic and patronage-network divides, suggest reasonable chances of success. But U.S. decision makers should be reluctant to make such bets very often.


1. We are indebted to Stephen Biddle, Ambassador Jim Dobbins, Seth Jones, Alexander Lennon, and Niloufer Siddiqui for their insights and suggestions on this piece.

2. Michael D. Shear, “Ghani Wants U.S. Troops to Stay in Afghanistan Longer,” The New York Times, March 24, 2015,http://www.nytimes.com/2015/03/25/world/asia/ashraf-ghani- of-afghanistan-wants-us-troops-to-stay-longer.html?hp&action=click&pgtype=Homepage

Mark Landler, “U.S. Troops to Leave Afghanistan by End of 2016,” The New York Times, May 27, 2014,http://www.nytimes.com/2014/05/28/world/asia/us-to-complete-afghan- pullout-by-end-of-2016-obama-to-say.html.

3. Prakhar Sharma, “How Long Should Afghanistan Matter?” Foreign Policy, June 2, 2014,http://southasia.foreignpolicy.com/posts/2014/06/02/how_long_will_afghanistan_matter. See also “Time to Pack Up,” New York Times, October 13, 2012, http://www.nytimes.com/ 2012/10/14/opinion/sunday/time-to-pack-up.html?_r=0; and Jeff Merkley, Rand Paul, and Tom Udall, “Let's Not Linger in Afghanistan,” New York Times, July 4, 2011, http://www. nytimes.com/2011/07/05/opinion/05merkley3.html?_r=0.

4. Frank Newport, “More Americans Now View Afghanistan War as a Mistake,” Gallup, February 19, 2014,http://www.gallup.com/poll/167471/americans-view-afghanistan- war-mistake.aspx?utm_source=position8&utm_medium=related&utm_campaign=tiles.

5. Quinnipiac University Poll, “Iraq—Getting in Was Wrong; Getting Out Was Right, U. S. Voters Tell Quinnipiac University Poll,” July 3, 2014, http://www.quinnipiac.edu/ images/polling/us/us07032014_ulps31.pdf.

6. Peter Bergen, “What Went Right,” Foreign Policy, March 4, 2013, http://foreignpolicy. com/2013/03/04/what-went-right/. We expand on this point at greater length below.

7. Stephen Biddle, “Ending the War in Afghanistan: How to Avoid Failure on the Installment Plan,” Foreign Affairs 92, no. 5 (September/October 2013), 49–58.

8. The Cost of War: Afghan Experiences of Conflict, 1978–2009 (Oxfam International, November 2009),https://www.oxfam.org/sites/www.oxfam.org/files/afghanistan-the- cost-of-war.pdf. Statistics for mortality under the age of five, school enrollment, and per capita income are for Afghanistan in 2001. World Bank, “World Development Indicators,”http://data.worldbank.org/.

9. Ibid.

10. United Nations, “Afghanistan, Annual Report 2013: Protection of Civilians in Armed Conflict,” United Nations Assistance Mission in Afghanistan and UN Office of the High Commissioner for Human Rights, February 8, 2014,http://unama.unmissions.org/Portals/ UNAMA/humanpercent20rights/Feb_8_2014_PoC-report_2013-Full-report-ENG.pdf; “Afghanistan Civilian Casualties 2009–2012,” The Guardian, http://www.theguardian. com/news/datablog/2010/aug/10/afghanistan-civilian-casualties-statistics; Ian Livingston and Michael O'Hanlon, “Afghanistan Index,” Brookings Institution, May 14, 2014, http:// www.brookings.edu/about/programs/foreign-policy/afghanistan-index.

11. The ANSF largely meets its ethnic targets, including for Pashtuns, which make up 39 percent of soldiers, 38 percent of NCOs, and 41 percent of army officers. Despite substantial improvement in recent years, some problems remain—including under- representation of Afghans from the south and west of the country, as well as a lack of ethnic representativeness in the higher police ranks. U.S. Department of Defense, “Progress Toward Security and Stability in Afghanistan,” April 2014, pp. 38 and 50, http://www.defense.gov/pubs/Oct2014_Report_Final.pdf.

12. Joshua Partlow, “Violence Data Show Spike in Afghan Presidential Election,” The Washington Post, April 14, 2014,http://www.washingtonpost.com/world/asia_pacific/ violence-data-show-spike-during-afghan-presidential-election/2014/04/14/b16169e2- c322-41da-9293-202de09cd513_story.html.

13. Ken Dilanian and David S. Cloud, “U.S. Seeks New Bases for Drones Targeting Al Qaeda in Pakistan,” Los Angeles Times, February 16, 2014, http://www.latimes.com/ world/asia/la-fg-drone-bases-20140216-story.html.

14. Matthew Rosenberg, “U.S. Disrupts Afghan Tack on Militants,” New York Times, October 28, 2013,http://www.nytimes.com/2013/10/29/world/asia/us-disrupts-afghans- tack-on-militants.html?pagewanted=all&_r=0.

15. We are indebted to Niloufer Siddiqui for her insights on these points.

16. Vanda Felbab-Brown, “The Political Games in the Taliban Negotiations,” Brookings Institution, June 19, 2013,http://www.brookings.edu/research/articles/2013/06/19- taliban-negotiations-after-karzai-suspends-peace-talks-with-us-felbabbrown.

17. On the importance of common information about the balance of power and the credibility of commitments, see James D. Fearon, “Rationalist Explanations for War,” International Organization 49, no. 3 (Summer, 1995), pp. 379–414.

18. Michael Semple, Reconciliation in Afghanistan, (Washington, DC: United States Institute of Peace Press, 2009).

19. For remarks by High Peace Council Chair Salahuddin Rabbani, General Dunford, and others, see Yaroslav Trofimov, “Hard-Line Taliban Leaders Exploit U.S.-Afghan Rift,” Wall Street Journal, January 30, 2014,http://www.wsj.com/articles/SB1000142405270 2304428004579350592183403828. For similar remarks by Michael Semple, see “Behind the Peace Talks with Taliban: Who's Looking For What,” NPR, February 6, 2014,http://hereandnow.wbur.org/2014/02/06/peace-talks-taliban.

20. International Crisis Group, “Afghanistan's Political Transition,” Asia Report no. 260, October 16, 2014, p. 29,http://www.crisisgroup.org/en/regions/asia/south-asia/afghani stan/260-afghanistan-s-political-transition.aspx.

21. “Ashraf Ghani sworn in as new Afghan president,” BBC News, September 29, 2014, http://www.bbc.com/news/world-asia-29375438.

22. Mehreen Zahra-Malik and Hamid Shalizi, “Pakistan and Afghan officials say Afghan Taliban signal readiness for talks,”Reuters, February 19, 2015, http://www.reuters.com/ article/2015/02/19/us-afghanistan-pakistan-taliban-idUSKBN0LN0A520150219.

23. Stephen Watts, Jason Campbell, Patrick Johnston, Sameer Lalwani, and Sarah Bana, Countering Others' Insurgencies: Understanding U.S. Small-Footprint Interventions in Local Context, (Santa Monica, Calif.: RAND Corporation, RR-513-SRF, 2014), http://www. rand.org/pubs/research_reports/RR513.html; for historical trends on victory and negotiated settlements, see Monica Duffy Toft, Securing the Peace: The Durable Settlement of Civil Wars, (Princeton: Princeton University Press, 2010).

24. Barbara F. Walter, “Why Bad Governance Leads to Repeat Civil War,” Journal of Conflict Resolution, (March 31, 2014),http://jcr.sagepub.com/content/early/2014/03/30/ 0022002714528006.full.pdf+html.

25. For an overview of this literature, see Human Development Report 2011: Conflict, Security, and Development, (Washington, D.C.: World Bank, 2011). For contending perspectives, see Walter, , “Why Bad Governance Leads to Repeat Civil War”; and Håvard Hegre and Håvard Mokleiv Nygård, “Governance and Conflict Relapse,” Journal of Conflict Resolution, forthcoming.

26. See, for instance, Lant Pritchett, Michael Woolcock, and Matt Andrews, “Capability Traps? The Mechanisms of Persistent Implementation Failure,” Center for Global Development, Working Paper no. 234, December 2010, http://www.cgdev.org/ publication/capability-traps-mechanisms-persistent-implementation-failure-working- paper-234.

27. Transparency International ranks Afghanistan the 175th-worst country (out of the 177 it surveyed) for corruption. See Transparency International, “Corruption by Country/ Territory,” http://www.transparency.org/country#AFG.

28. Scott Seward Smith, “The Future of Afghan Democracy,” Stability: The International Journal of Security and Development3, no. 1 (2014), http://www.stabilityjournal.org/ article/view/sta.dk.

29. Thomas Ruttig, “Negotiations with the Taliban: History and Prospects for the Future,” National Security Studies Program Policy Paper, New America Foundation, May 2011,http://www.newamerica.net/sites/newamerica.net/files/policydocs/Ruttig_Nego

30. Fred Bezhan, “Controversial ID Cards Expose Ethnic Divisions in Afghanistan,” Radio Free Europe/Radio Liberty, December 18, 2013; Anna Badkhen, “Eternal Enemies, One Mile Apart,” Foreign Policy, May 2, 2010,http://foreignpolicy.com/2010/05/02/eternal- enemies-one-mile-apart/.

31. Steven Levitsky and Lucan A. Way, “Linkage Versus Leverage: Rethinking the International Dimension of Regime Change,” Comparative Politics 38, no. 4 (July, 2006), pp. 379–400; Thad Dunning, “Conditioning the Effects of Aid: Cold War Politics, Donor Credibility, and Democracy in Africa,” International Organization 58, (Spring 2004), 409–423.

32. These two factors are central to canonical analyses of revolution; see especially Theda Skocpol, States and Social Revolutions, (New York: Cambridge University Press, 1979). On the risk of fiscal crisis in Afghanistan, see Richard Hogg, Claudia Nassif, Camilo Gomez Osorio, William Byrd, and Andrew Beath, Afghanistan in Transition: Looking Beyond 2014, (Washington, D.C.: World Bank, 2013); William A. Byrd, “Afghanistan's Looming Fiscal Crisis: What Can Be Done?” United States Institute of Peace, Peace Brief no. 177, August 27, 2014. http://www.usip.org/publications/afghanistan-s- looming-fiscal-crisis-what-can-be-done.

33. Olga Oliker, Building Afghanistan's Security Forces in Wartime: The Soviet Experience, (Santa Monica, Calif.: RAND Corporation, MG-1078-A, 2011), http://www.rand.org/ pubs/monographs/MG1078.html.

34. Yasir Abbas and Dan Trombly, “Inside the Collapse of the Iraqi Army's 2nd Division,” War on the Rocks, July 1, 2014,http://warontherocks.com/2014/07/inside-the-collapse- of-the-iraqi-armys-2nd-division/.

35. The White House, “Statement by the President on Afghanistan,” video and transcript, May 27, 2014,http://www.whitehouse.gov/photos-and-video/video/2014/05/27/president- obama-makes-statement-afghanistan#transcript.

36. Shear, “Ghani Wants U.S. Troops to Stay in Afghanistan Longer.”

37. Hope Hodge Seck, “Senator to CMC: Marines might have to go back to Iraq,” Marine Corps Times, March 10, 2015,http://www.marinecorpstimes.com/story/military/capitol- hill/2015/03/10/graham-to-commandant-you-may-have-to-go-back/24704499/.

38. U.S. policy currently anticipates limiting air strikes to protection of U.S. personnel. Obviously this policy could be changed if the government of Afghanistan was threatened with collapse, much as has occurred recently in Iraq. It is also unclear precisely what the current U.S. policy means in practice. Would the presence of U.S. Special Operations Forces embedded with ANSF be sufficient reason to call in air strikes? If so, it is unclear just how limiting the current policy actually is.

39. Office of the Press Secretary, The White House, “Fact Sheet: Chicago Summit – Sufficient and Sustainable ANSF,” May 21, 2012, https://www.whitehouse.gov/the- press-office/2012/05/21/fact-sheet-chicago-summit-sufficient-and-sustainable-ansf.

40. Ibid. The Afghan government contribution was originally planned to start at $500 million in 2015, though they are unlikely to reach this target in the short term, suggesting that costs for the United States may be slightly larger than originally estimated. Jonathan Schroden et al, Independent Assessment of the Afghan National Security Forces, (Alexandria, Va.: CNA, January 2014), pp 159–160.

41. Special Inspector General for Afghanistan Reconstruction, “Quarterly Report to the United States Congress,” July 30, 2014, pp. 212–3, http://www.sigar.mil/quarterlyreports/.

42. The World Bank, Afghanistan in Transition: Looking Beyond 2014, Volume 1: Overview, (Washington, DC: The World Bank, May 2012,) p. 6.

43. The Military Balance 2014, (International Institute for Strategic Studies, February 2014).

44. Combined annual economic and security aid figures stood at over 33 percent of Afghan GDP each year from 2004 to 2013, the last year for which GDP data is available. Special Inspector General for Afghanistan Reconstruction, “Quarterly Report to the United States Congress.”; GDP from “World Development Indicators,” World Bank,http://data.worldbank.org/indicator/NY.GDP.MKTP.CD.

45. Security assistance and total ANSF budget are drawn from Office of the Secretary of Defense, “Justification for FY 2015 Overseas Contingency Operations Afghanistan Security Forces Fund,” Department of Defense, June 2014, p 3.

46. Thom Shanker, “NATO Reduces Scope of Its Afghanistan Plans,” New York Times, October 27, 2013,http://www.nytimes.com/2013/10/28/world/asia/less-extensive-mission- planned-in-afghanistan.html?ref=international-home&_r=0.

47. In 2012, official development assistance from all sources represented 32 percent of Afghanistan's total economy, the most for any country save Liberia and four small island states (Tuvalu, Marshall Islands, Micronesia, Solomon Islands)—and that figure does not even include military subsidies. “World Development Indicators,” World Bank.

48. One of the earliest explorations of the seemingly paradoxical relationship between increased commitment and decreased leverage was Douglas Blaufarb's classic analysis of Vietnam; see Douglas S. Blaufarb, The Counterinsurgency Era: U.S. Doctrine and Performance, (New York: The Free Press, 1977).

49. Special Inspector General for Afghanistan Reconstruction, “Quarterly Report to the United States Congress,” January 30, 2015, p 199. http://www.sigar.mil/pdf/quarterlyreports/ 2015-01-30qr-appendices.pdf.

50. Michael Shifter, “A Decade of Plan Colombia: Time for a New Approach,” Politica Exterior, June 21, 2010, English available at Inter-American Dialogue, http://www. thedialogue.org/page.cfm?pageID=32&pubID=2407.

51. On the importance of being able to communicate a credible path to success as a precondition for sustaining public support for military operations, see Christopher Gelpi, Peter D. Feaver, and Jason Reifler, “Success Matters: Casualty Sensitivity and the War in Iraq,” International Security 30, no. 3 (Winter, 2005/2006), 7–46.

52. The extent of the strategic effects of efforts to “disrupt and degrade” terrorism networks is highly contested, as is the extent of the recruitment advantages provided to terrorism groups by such counterterrorism strikes. For recent sophisticated treatments of the effects of military operations on war termination and public support for rival parties, see Michael Tiernay, “Killing Kony: Leadership Change and Civil War Termination,” Journal of Conflict Resolution, forthcoming; and Jason Lyall, Graeme Blair, and Kosuke Imai, “Explaining Support for Combatants during Wartime: A Survey Experiment in Afghanistan,” American Political Science Review 107, no. 4 (November 2013), pp. 1–27.

No comments: