5,000 Troops for 5 years: A no drama approach to Afghanistan for the next US president

What should America’s next president, Democrat or Republican, do about Afghanistan? With an electorate tired of “forever wars” and a new Pentagon strategy focusing on great-power rivalry with Russia and China, the temptation for presidential candidates to promise a complete U.S. departure is palpable.  

Instead, candidates should propose a path to downsize and then stabilize the U.S. military mission in Afghanistan so that it is roughly on the scale of the current U.S. deployment in Iraq. That would allow the United States a good chance of achieving its core counterterrorism goals in Afghanistan, as well as in Pakistan. The slogan “5,000 troops for 5 years” would be a reasonable distillation of the chief military elements of this approach. That time horizon would also give peace talks a realistic chance, unlike recent proposals that fancifully imagined a power-sharing accord between the Afghan government and the Taliban by the end of 2020. 

Introduction and main argument

By the time America’s next president is inaugurated on January 20, 2021, America’s role in the Afghanistan war will be approaching its twentieth anniversary. Afghans themselves will have been at war continuously at least twice as long, if one dates the beginning of the modern conflict to the Soviet invasion there in 1979.

Americans are understandably tired of this war. It has by any measure been a frustration, especially when measured against the more ambitious “nation-building” goals of the first Obama term. However, this fact will come as news to virtually no one—despite the December 9, 2019 Washington Post story that alleges a Vietnam-like coverup. U.S. officials have been consistently and publicly realistic about the difficulty of making progress in Afghanistan. Whether it was the Bush administration’s decision to maintain only a “light footprint” in Afghanistan because state-building there was seen as hard, or the McChrystal review of 2009 that identified Afghan government corruption as a severe threat to the mission on a par with the Taliban itself, or the subsequent Obama White House review that same fall that reluctantly and temporarily concluded with a decision to surge forces in Afghanistan (yet only for a short time, to keep pressure on the Afghans themselves to work harder), to the Obama and Trump reviews in more recent years that have consistently sought to find an exit strategy out of the country, the tenor of the American policy debate has consistently been sober.

But the mission has not been an abject failure. The Afghan government continues to hold all major and mid-sized cities as of this writing, and even more to the point for Americans, the United States has not again been attacked by a group that plotted or organized its aggression from within Afghan borders. The United States probably has the ability to do its part to sustain these modest, yet real, accomplishments at far lower cost in blood and treasure than before. The bad news is that there is likely no near-term exit strategy; this reality should be faced head-on. The good news is that, in strategic and military and budgetary terms, the cost of the mission is sustainable.

The United States needs a policy that recognizes Afghanistan for what it is—a significant, but not a top-tier, U.S. strategic interest—and builds a plan accordingly. That overall strategy should still seek peace, but its modest military element should be steady and stable, and not set to a calendar. Roughly 5,000 U.S. troops for at least five years could be the crude mantra.

A future force of 5,000 U.S. troops in Afghanistan, aided by 2,000 to 3,000 other NATO military personnel, would contrast with the late-2019 figure of 13,000 GIs there. It would be 95 percent less than the 100,000 U.S. troops commanded there by General David Petraeus and then General John Allen at the peak of the American presence in 2010-11. This lower level could probably be achieved by 2022, though the glidepath could be slowed if conditions required.

The advantages of this approach go well beyond the reduction in force numbers. By laying out a plan designed to last for several years, Washington would be avoiding the drama and the huge consumption of policy bandwidth associated with annual Afghanistan policy reviews that have typified the late Obama and early Trump years.

The number of U.S. forces could decline even further if a peace deal eventually were struck—in which case it might even wind up near zero eventually. The size of the deployment could also be reduced further if the Afghan military started to develop greater strength and battlefield momentum. But for planning purposes, “5,000 for 5” would be the core premise.

This idea for a smaller but enduring U.S. military presence in Afghanistan will go against much of the preferred sentiment of the Democratic base. But Democrats do not have the luxury of sweepingly opposing long-term military missions while castigating President Trump for ending them recklessly in the way he just nearly did, yet again, in Syria.

Of course, Trump administration policy could change dramatically before election day as well. Absent a comprehensive peace deal with the Taliban by next fall that also involves the Afghan government—an extreme longshot, given the lukewarm interest that the Taliban, as well as their sometimes sponsor Pakistan, likely has in the idea—President Trump will face a dilemma. He will have to choose between his original campaign promise to end forever wars and leave Afghanistan on the one hand, and his promise to defeat global terrorism and protect America on the other. Al-Qaida remains present in Afghanistan and just over the border in Pakistan, where drones flying from bases in Afghanistan can help keep a watch on things. The ISIS presence in Afghanistan has been growing, too. So far, neither has provided the capability or bastion that extremists would need to strike the United States. But their capacities could, and almost surely would, grow in the absence of America’s and NATO’s extensive capabilities in intelligence, airpower, unmanned systems, and special forces. Thus, Trump would be wisest to claim partial credit for a partial drawdown while maintaining enough capacity to check al-Qaida and ISIS.

With this approach, the United States would maintain several counterterrorism strongholds throughout the broader region where the jihadist scourge has been most serious. An enduring presence in Afghanistan would complement other mid-sized U.S. capabilities in places closer to the Middle East such as Iraq, Kuwait, Bahrain, Qatar, and Djibouti. In each of these places, as well as in Turkey, the United States stations roughly 1,500 to 10,000 uniformed personnel. Several thousand sailors are routinely in or near the Persian Gulf aboard ships as well. Together, all these forces represent 2 to 3 percent of the nation’s total active-duty strength. That is a significant but reasonable burden to bear in a region where America’s interests remain considerable, even if the primary focus of U.S. military planning has turned to scenarios involving Russia and China. Indeed, the United States competes with both China and Russia within the Middle East. So a strategy aimed at Beijing and Moscow presumably should not concede crucial territory to either country in that region.

Some will say that terrorism in and near Afghanistan can be checked even without an American military presence on the ground, even if our departure leads to all-out civil war and/or a victory by the Taliban. Perhaps any future al-Qaida or ISIS presence on Afghan soil could be handled with long-range strikes or occasional commando raids that emanate from ships in the Indian Ocean. Or perhaps we could be confident that such groups have no substantial future interest in basing themselves in Afghanistan.

But that latter argument ignores history, as well as the geographic suppleness of global extremist movements in general. Few saw the ISIS caliphate coming in Iraq and Syria before 2014 but then, all of a sudden, it was there. And the former argument shows a poor appreciation of how counterterrorism intelligence is developed—usually by cooperation with partners on the ground—as well as an unrealistic appreciation for the geographic remoteness and ruggedness of the Hindu Kush. Stand-off counterterrorism is generally an oxymoron.

It is true that the Afghanistan war is not being won. Nor is it lost, however. Although its influence has receded modestly in recent years, the government still controls the territory where more than 60 percent of the population lives, including all mid-sized and major cities. Taliban strongholds amount to 10 to 12 percent of the population, based on official U.S. government estimates from late 2018 (after which the United States started classifying such data, regrettably—however imperfect it may be, it gave some sense of trendlines). The remaining quarter or so of the country’s population lives in contested areas.¹ Moreover, no major attack on the United States has again come from Afghan soil since 9/11. Casualties to Afghans are way too high, but that is a strange reason to concede a war to the enemy who is responsible for most of those casualties.

With 5,000 American troops (and some additional civilians and contractors) in Afghanistan, the United States could maintain two or three major airfields and hubs of operations for intelligence, airpower, and special forces/commandos—at Bagram near Kabul in the nation’s center, near Kandahar in the south, and perhaps around either Khost or Jalalabad in the east. It also could maintain a modest military advisory and training presence in Kabul to help the Afghan army and police carry out the bulk of the fighting against extremists.²

The annual cost of this presence would be perhaps $7 billion to $8 billion—not trivial, but only 1 percent of the defense budget. It would require several billion dollars more in annual assistance to the Afghan government, to maintain its army and police and to pursue modest development goals, though other donors could be expected to provide most funds for the latter purposes.

Committing to such a presence for half a decade would also signal to Pakistan and the Taliban that the new president would not be expecting a “Hail Mary” peace deal as a viable near-term exit strategy. Perhaps such a promise would even improve the seriousness with which one or both might then engage in peace talks.

Of course, even if it often dominates Washington debates, the question of U.S. and therefore NATO troop levels in Afghanistan is not the only crucial policy question in regard to the future of that country. At least as important are two other general areas of effort: the attempts by the Afghan government to reduce corruption and improve governance for its citizens, thereby dampening support for the insurgency, and the on-again/off-again peace process with the Taliban. This policy brief does not offer comprehensive treatments of these matters, but provides several observations and suggestions on each.

Afghan government and security policy reforms

Arguably, in many ways the center of gravity of the Afghanistan mission is the sustainability of that country’s security forces in the face of a very long internal conflict, and very high loss rates. The losses are due largely to casualties on the battlefield—in addition to Afghan civilian fatalities of several thousand a year from the war, more than 5,000 Afghan security personnel have been killed annually due to enemy action for several years running. They also happen because many soldiers and police go AWOL or simply choose not to reenlist. As a result, Afghan forces are several tens of thousands of personnel short of authorized levels, and the quality of many fighters and commanders is not as good as a more-experienced and stalwart force might manifest.³

To avoid the possible collapse of the Afghan army and police, and thus catastrophic mission failure, we need to work with the Afghan government on ideas like these:

• Scale up the Afghan National Army Territorial Force (ANATF) concept. Since many Afghans prefer to defend their home territories rather than distant parts of the nation, this concept could help greatly with recruiting and retention. The force might reach the low tens of thousands of soldiers. (The Afghan Local Police concept is a similar program within the police, rather than the army, and has had mixed results to date—part of the rationale for placing the ANATF within the army.)
• Emulate the rotation and rest policies of the Afghan special forces within the regular army and police. Today, most soldiers and police rarely get leave time or down time. To make such changes possible, the Afghan government will have to give up protection of some remote regions of the country at least temporarily.
• Consolidate police checkpoints into fewer, better defended outposts so they are less vulnerable to being overrun by Taliban ambush. Taliban forces are often capable of marshaling many dozen or even a few hundred irregulars in one place for surprise attack, so even with the advantages of firepower and protection, any fixed location for the army and police should typically have at least dozens of personnel located within its perimeter (and quick response forces close by for reinforcement). The Afghan government could still aspire to reassert its control and influence of those areas where it closed checkpoints someday down the road. Afghan police forces have apparently closed about 150 of the more dangerous checkpoints around the country in recent months, but there is much more work to be done in this regard.
• Help the Afghan government acquire more battlefield medical evacuation capacity so that it can keep more of its wounded alive. Perhaps this could be achieved partly with private security contractors equipped with helicopters.
• Provide members of the Afghan parliament and other officials modest funds to hire small personal security details so they will make fewer demands on the regular police to protect them, thus allowing police to do their normal jobs more effectively.

Although some of these ideas are not entirely new, they should be given a new urgency as part of a redefinition of the center of gravity of the ongoing conflict in Afghanistan to focus on the well-being, capability, and most of all sustainability of the Afghan police and army. Precise trends in how much of the country the government does or does not hold matter less than questions about whether the army and police, which are presently adequate to secure major cities and most roads, can continue to do so.

The prospects for peace

Many are pinning their primary hopes for the Afghanistan mission on a negotiated settlement to the war. This seems sensible, after so many years of fighting, that has left every major party to the conflict bloodied and fatigued.

Yet there are reasons to be very wary of the peace process, as President Trump seems to have realized himself in the late summer of 2019 when he declared “dead” a process that seemed on the verge of achieving an initial agreement. In fairness to Trump, that initial deal wouldn’t have done much. Apparently, it would only have required U.S. troop cuts that were likely to happen anyway (down to about 8,600 American personnel) in exchange for a verbal commitment from the Taliban to break ties with al-Qaida, and perhaps to relent modestly in the pace of attacks in certain parts of the country. Even if it had been achieved, and complied with, the next stage in the peace process was sure to be far more difficult. It envisioned meaningful power-sharing between the Afghan government and the Taliban, even though at present the latter will not even sit down with or recognize the former. It also sought to achieve a complete U.S./NATO troop pullout, if so requested by the new coalition government, and to make all this happen in 2020!

We should not be surprised that the process has broken down. Both main Afghan parties to any deal—the government and the Taliban—expect to come out ahead in the overall distribution and balance of power in any agreement. Their aspirations may well be mutually incompatible. The Taliban, already of the view that it is winning on the battlefield, may simply stall for time, expecting that the United States (and thus its NATO and other foreign allies) will ultimately leave in frustration, as President Trump has telegraphed that he wishes to do. Yet at the same time, a recently reelected President Ghani (if that is indeed the ultimate verdict of the flawed but still meaningful September 28, 2019 Afghan elections) would himself likely feel that legitimacy and the balance of power favor him and his government.⁴

For its part, the Pakistani state, particularly its Inter-Services Intelligence agency (ISI), may continue to prefer an ultimate Taliban victory over any alternative outcome. Doubting America’s commitment to the region, and fearing India’s influence with a future Afghan government, it may continue to support, or at least condone and tolerate, Taliban operations planned from its own soil.⁵ This is an unwise and counterproductive policy for the ISI and the Pakistani state, but it is deeply entrenched. Prime Minister Imran Khan, though himself apparently supportive of the peace process (and an eventual, ensuing U.S./NATO military departure), may not succeed in changing this behavior of his own security and intelligence forces more than his predecessors.

Many peace processes do succeed in ending civil wars at some point, including in Mozambique, Cambodia, Colombia, El Salvador, and Nicaragua in recent decades. But often they take many years, and they may also require that one side to a conflict accept that it has been largely stymied or defeated on the battlefield. These conditions may not apply in Afghanistan.

To revive peace talks, Washington may need to inject big ideas into the process, rather than counting on Afghans to propose such ideas themselves. For example, initially, Taliban might be allowed to control and govern the rural areas they control—provided that basic human-rights standards are protected. They could even receive some limited foreign aid for schools, food, and medical care for citizens in those areas. Later, Afghanistan’s constitution could be modified to allow the direct election of provincial governors and city mayors by local citizens (today, all are appointed by the president, as a result of the constitution approved back in 2004 and written with a large degree of U.S. influence). Such a move toward decentralization would be consistent with much of Afghanistan’s history, in which the traditional regions around Kandahar, Herat, Mazaar-e-Sharif, and Jalalabad had a considerable degree of autonomy and clout.⁶ In such a situation, the Taliban might hope to win some elections in the country’s east and south, even if it is unlikely to win a national vote at the presidential or parliamentary level.⁷Indeed, the Taliban might even be allowed a hand in creating much of the local security presence in areas it controls politically (under at least loose central supervision).

Even with such disruptive ideas including constitutional reform on the table, peace will almost surely not come fast to Afghanistan. A negotiated settlement is probably the only way to end this war. But any expectation that it will happen fast or on terms that Washington and Kabul now prefer would be an unwise foundation for strategy.

Afghanistan and the aumf debate

Consideration of where we stand in the Afghanistan war should be one of the key factors informing any efforts by the 117th Congress and a newly-elected president to review and reconsider the 2001 Authorization for Use of Military Force (AUMF). That AUMF, now more than 18 years old, provides the main legal basis for operations against extremists from Afghanistan—the location from which the 9/11 attacks were first planned, and thus the logical first focus of that 2001 legislation—to Iraq, Syria, Yemen, Libya, Somalia, and even Niger. It has, in the eyes of many, been twisted and stretched multiple times to justify operations against entities that had nothing to do with the 9/11 attacks. Even for those of us who are more flexible and forgiving about how the legislation has generally been employed, it must be acknowledged that it is getting old and that it seems to provide a justification for “forever war” well beyond the locations of strongholds of al-Qaida and affiliates or allies. It was only these latter groups that the 2001 law aimed to target.

Advocates of a new AUMF tend to favor restrictions on the geographic scope, time duration, and targeting flexibility of any future U.S. military operations. Clearly, as the focal point for the initial planning of the 9/11 attacks and the sanctuary from which al-Qaida leadership then operated, Afghanistan would seem to qualify as within the reasonable geographic limits of any future law. But revised legislation needs to be drafted carefully. If focused on only a few specific groups, it would be vulnerable to the possibility that a wily future terrorist entity could simply change its name while maintaining much of the membership, operational networks, and underlying philosophy and goals of al-Qaida or ISIS. For example, even if “ISIS-Korasan” were included as within the purview of a new AUMF, that group might rename itself as something to the effect of “the South Asia Caliphate” under some future new leader. There would have to be a means for the intelligence community to evaluate whether any such new derivative or offshoot were itself threatening enough to the United States to be brought within the scope of a future AUMF; otherwise, the restrictions of a new law could be severely injurious to the security interests of the United States and its allies.⁸

And as for time duration, while it seems reasonable that any new AUMF should only cover a certain number of years, drafters of the legislation would have to be careful. Given Washington’s recent penchant for playing brinkmanship with everything from judgeships and political appointments to the annual federal budget process, one worries about a scenario in which an AUMF expires but Congress and the president fail to consider a successor bill in time to allow ongoing operations against very real threats. As such, any new AUMF should include a default mechanism allowing crucial military operations to continue if it expires without Congress having yet voted on a possible replacement.

Conclusion

The Afghanistan war remains frustrating and costly. There are still some 13,000 U.S. troops in country as of this writing in late 2019—not counting regional forces or temporary units. The cost to the United States of this ongoing operation is probably $20 billion a year, plus or minus several billion.⁹ American fatalities have been in the range of 10 to 20 per year since the major drawdown of forces was completed toward the end of President Obama’s term in office.¹⁰

On the other hand, the simple fact remains that the United States has not again been attacked by any group operating principally out of Afghanistan or environs since 9/11. A couple of Afghans or Afghan-Americans have attempted or carried out lone-wolf attacks since then, but on the core goal of protecting the homeland, the Afghanistan effort has not failed.

It may also be worth noting that Afghans helped America and its allies win the Cold War by defeating the Soviet Union on their territory in the 1980s, at huge human and national cost to the Afghan people. This is not a reason to reinforce a failing mission, or to overlook all the corruption and poor performance of many Afghans since 2001. But it is worth bearing in mind when Americans feel that they are being taken advantage of by allies around the world.

If the U.S. and NATO effort in Afghanistan is troubled, challenged, and frustrating, yet still succeeding in helping protect American and allied homelands, it should not be abandoned lightly. The prospects for a successful peace process require a measure of resoluteness as well. It defies reason to think that the Taliban will negotiate in good faith with the United States or the Afghan government if they believe, rightly or wrongly, that after outwaiting Washington, Afghanistan will then somehow magically fall into their laps after western forces leave.