Afghanistan 2001-2021: U.S. Policy Lessons Learned
Afghanistan 2001-2021: U.S. Policy Lessons Learned
U.S. Congressional Testimony / Asia 14 minutes

Afghanistan 2001-2021: U.S. Policy Lessons Learned

In a 17 November 2021 hearing before the U.S. Senate Committee on Foreign Relations, Crisis Group’s Asia Program Director Laurel Miller outlined five lessons learned from twenty years of U.S. involvement in Afghanistan and suggested some steps Washington can take in reformulating its Afghanistan policy.

Good morning, Chairman Menendez, Ranking Member Risch and distinguished members of the Committee. Thank you for the invitation to testify today on policy lessons that can be learned from American involvement in Afghanistan over the last two decades, and on recommendations for the immediate future of U.S. policy toward Afghanistan.

In the aftermath of the U.S. failure – after enormous sacrifice of lives and treasure – to defeat the Taliban, establish a self-sustaining Afghan democracy and economy, and ensure the durability of health, education and other social gains, it may be too easy to assume the United States will never do that again. But it is crucial not to brush past an examination of how and why this failure happened. The future is too uncertain to assume that the United States will never again encounter circumstances analogous to those that impelled its leaders to invest so heavily in Afghanistan for two decades. After the failed American war in Vietnam, counter-insurgency was anathema in national security policy; after nation-building in the Balkans, the George W. Bush administration initially derided the concept as an inappropriate use of U.S. resources, even though the policy had been relatively successful. Yet both became central to the American intervention in Afghanistan. “Never say never” may be the most basic lesson to learn from Afghanistan at this time. Therefore, a thorough accounting will be needed of U.S. policy decisions, the means chosen to implement them and their results.

Another basic lesson is that the United States could not have achieved its goals in Afghanistan solely through its own policies and actions, because both its partners and its adversaries had at least as much influence over the course of events. Likewise, the failure is not uniquely an American one. Nevertheless, there were strategic choices that were controlled by the United States and were especially consequential in leading to failure. I will focus my remarks on five lessons that can be learned from errors in these choices.

Five Key Lessons

1. Be very, very wary of regime change.

A narrative has taken hold that, because the motivation for the U.S. invasion of Afghanistan was counter-terrorism, the nation-building that ensued represented mission creep. In reality, the Bush administration’s 2001 decision not only to chase and punish the terrorist perpetrators of the 11 September attacks, but also to oust the Taliban regime that had harbored al-Qaeda leaders, necessarily required a nation-building mission. It would have been the height of irresponsibility to wipe away the existing regime in Afghanistan and make little effort to support the construction of a reasonably functional state in its wake. Indeed, in the early years, the United States was criticised for doing too little, not too much, to build up an Afghan state, including indigenous security forces.

After initially assuming away the complexities of regime change, the recognition quickly took hold that political disorder and the absence of state institutions could give rise to the persistence of conditions that led to Afghanistan being an exporter of security threats. But Washington was unprepared for the implications of regime change and struggled over how committed to be to managing them.

The United States started with a strategy based on the assumption that it could eliminate the Taliban simply by killing them – as if essentially all of them could be killed, with no replenishment of ranks – and that, after the Taliban’s elimination, other Afghan groups and factions would come together and decide their own political arrangements without much fuss. The central idea, in other words, was that the United States could invade, wipe the political slate clean, move on, and somehow the situation would sort itself out without considerable U.S. effort, which was to turn to military involvement elsewhere in the world. This idea was a theory with no empirical support. As the theory quickly proved itself false, the U.S. shift to nation-building began as early as April 2002.

By choosing to engage in regime change and install a new regime that would act in alignment with U.S. interests, the United States chose to engage in nation-building.

By choosing to engage in regime change and install a new regime that would act in alignment with U.S. interests, the United States chose to engage in nation-building. The specific aims and the resources devoted to the mission expanded over time as recognition of the challenges expanded and as paltry results seemed to demand greater effort, and the ways and means of implementation evolved. But, fundamentally, the policy die was cast at the very beginning. The United States had adopted a policy of constructing an Afghan state, of making a “partner” in Afghanistan, and of needing that partner to succeed at governing in order for U.S. policy to succeed at leaving Afghans, eventually, to sustain their own system and ensure their own security in accord with U.S. security interests. U.S. policy became dependent on Afghan government success.

2. If your strategy’s success depends on particular conditions, be sure you can create or control those conditions.

Several essential conditions for success of the U.S. military and civilian missions in Afghanistan were no mystery to U.S. policymakers, and yet the implausibility of creating those conditions was never adequately factored into shaping strategy. Instead of adjusting policy to reflect obstacles that were unlikely to be surmounted, the United States adopted a policy that required surmounting the obstacles, based on the belief, or hope, that willpower, military might and financial wherewithal would prevail.

Foremost among these obstacles was Pakistan’s policy. From the first days after 9/11, the U.S. relied on its presumed ability to get Pakistan to take steps to cooperate in eliminating the Taliban that Pakistan had made quite clear it did not see in its own interests to take. That clarity was evident in Pakistan’s supportive relationship with and material aid for the Taliban prior to 9/11, and also in what Pakistani officials said to U.S. officials afterward. There was no need to read between the lines, though there was a need to pay attention to how seriously Pakistanis meant what they said. And what they said consistently from the start and over the years since was that they wanted a government in Kabul that would be amenable to Pakistani interests, disagreed with a U.S. strategy of militarily eliminating the Taliban, and wanted to see the Taliban included in Afghanistan’s political dispensation. Just three days after 9/11, Pakistani President Pervez Musharraf agreed to cooperate with the United States in counter-terrorism, but said there would be implementation details to work out, including that “Islamabad, [Musharraf] said, wants a friendly government in Kabul”. There should have been no misunderstanding that “friendly” meant including the Taliban.

Pakistan’s consistent pursuit of its interests in Afghanistan as it perceived them – regardless of U.S. diplomatic remonstrances or financial enticements intended to convince Islamabad to change its national security calculations – contributed significantly to U.S. failure in Afghanistan. To be sure, Islamabad also facilitated the U.S. war, especially by providing air and land access to landlocked Afghanistan. But this duality of Pakistani policy only reflected that Islamabad had two distinct and irreconcilable policy goals: on one hand, maintain a constructive relationship with Washington beyond Afghanistan matters, and on the other, see the Taliban return to at least a substantial share of power in Kabul.

Although Pakistan enabled the United States to fight the Taliban, its officials regularly stated explicitly that Pakistan itself would not fight the Afghan war on Pakistani soil – meaning that Pakistan would not take steps to make the Taliban its own adversary. As a result, the Taliban enjoyed the crucial benefit for an insurgency of safe haven in a neighbouring state. Pakistani officials occasionally denied safe haven existed, but the denials were virtually irrelevant because the reality was known to the United States and Pakistan never suggested it would dismantle the safe havens that it generally declined to acknowledge.

Getting Pakistan to switch to a policy of opposing the Taliban proved unrealistic.

Sharing a long frontier with Afghanistan and being well-practiced in opaque means of providing support to the Taliban, helping the insurgency to survive and thrive was not difficult for Pakistan. Getting Pakistan to switch to a policy of opposing the Taliban proved unrealistic; there were naturally limits to how far the U.S. would go to pressure Islamabad, and the latter knew as much. Because strategic success in Afghanistan was not existentially important to U.S. national security, it would have been unwarranted and unrealistic for Washington to widen the war to include military action against Pakistan, a nuclear-armed nation of 225 million people some 8,000 miles away.

Being stymied in counter-insurgency as it was, the United States could instead have changed its policy much sooner than it did from one centered on war-fighting to one centered on diplomatic efforts to reach a political settlement among Afghans, bringing the Taliban into a share of power. That policy shift, from late 2018, was one that Pakistan, unsurprisingly, embraced and supported. Unfortunately, however, by that stage the Taliban had a clear upper-hand on the battlefield and U.S. leverage was greatly diminished by having made it evident that U.S. forces would be withdrawn sooner rather than later regardless of whether a political settlement was reached. The effort to motivate the Afghan contestants to negotiate the end of the war failed and, of course, the U.S. military withdrew nonetheless, in accordance with a bilateral deal signed between the U.S. and Taliban on 29 February 2020.

A second key condition resistant to U.S. efforts to change it was the weakness of the Afghan government Washington helped build and saw as its partner. The disunity and endemic corruption that plagued the Afghan government has been well-documented for many years. The strategic error was not in failing to recognise those problems existed but, rather, in expecting that they could be sufficiently ameliorated fast enough to deprive the insurgency of fuel and to align with any plausible duration of American political willingness to prop up the Kabul government. The political disunity reflected a competition for power driven by Afghan dynamics that the United States was unequipped to modify. And there is simply no historical precedent for an external actor to remake the patronage basis of a society through foreign policy and foreign aid measures.

A third condition outside the ambit of U.S. control was that Afghanistan was in 2001, and remains, one of the poorest and least institutionalised countries in the world, and one that is also landlocked and historically dependent on external resources. There was every reason to expect that the timescale would be generational for Afghanistan to develop a self-sustaining economy and a government able to fully, or nearly so, provide for its own security and public services. Decision-makers often assumed, however, that these developments could be sped up through funding and diplomatic pressure to fit U.S. policy urgency.

3. Recognise how much you do not know, but also embrace what you do know and change your policy accordingly.

Looking back at the earliest strategy decisions related to regime change cited above, policymakers appear woefully naïve about what the United States could achieve in Afghanistan. Perhaps U.S. lack of understanding of conditions within Afghanistan could be excused considering how little the United States had been engaged there during the preceding decade and the consequent paucity of U.S. government expertise regarding the country. What was more problematic – and where a lesson for future U.S. policy lies – was the failure to appreciate how little the conditions were understood and, therefore, the lack of a firm basis for confidence that the U.S. strategy made sense. The less you know, the greater the uncertainty about policy effects, and the greater the risk of unintended consequences.

As the U.S. intervention wore on, many essential facts emerged into view. Indeed, the seeds of failure were present for many years, and none of the factors that ultimately produced the collapse of the Kabul government and the disintegration of U.S. policy were unknown. Even if the specific timing of the Afghan government’s collapse could not be predicted, it was widely anticipated as at least a plausible scenario. The precarity of state institutions, the government’s extraordinary aid dependency (about 75 per cent of public spending was donor financed), the bubble effects of a wartime economy and crucial weaknesses within the Afghan security forces were all well known. Within the Afghan security forces, problems that would naturally affect morale and will to fight, such as irregularity of pay and inadequacy of equipment, living conditions and other forms of support, were regularly observed and publicly pointed to as critical weaknesses. These problems were not left unfixed because they were obscure; rather, they were very difficult to fix, through very slow processes at best.

The amalgamation of these and other problems led to routinely pessimistic publicly reported assessments by the U.S. intelligence community for the last dozen years, at least. Instead of shaping policy in accordance with these assessments, until the final push toward exit, decision-makers shaped policy in accordance with a hoped-for ability to prove the assessments wrong.

Although there were routine claims, including in testimony to the U.S. Congress, that progress was being achieved in improving the capabilities of the Afghan security forces and in setting a course toward winning the war, some called out the war as unwinnable as early as 2009. Public reporting of steady Taliban battlefield gains has been plentiful, especially since the major U.S. military drawdown of 2014. There was no shortage of public reporting, too, of the U.S. intelligence community’s negative assessments of the sustainability of Afghan government and security forces without a continued U.S. military presence and exceptional scale of financial support, as well as warnings more generally of the impending failure of U.S. policy.

Embracing rather than resisting the facts in plain sight should have led much earlier to a judgment that the war was not likely to be won and that the main effort should have been diplomacy, seeking a negotiated end of the conflict or at least of American involvement in it, years earlier than occurred.

4. Aid conditionality does not work if your strategy depends on the recipient’s success.

Because the United States was well aware that corruption in Afghanistan was fuelling support for the Taliban insurgency and political disunity was weakening the state, it tried repeatedly to address these and related problems by conditioning aid disbursements on improvements in these areas. The latest iteration of conditionality was the Afghanistan Partnership Framework agreed upon by the Kabul government and donors in November 2020.

The use of conditionality in Afghanistan suffered from a fatal flaw: because the United States had a policy that required the success of the Afghan government (as discussed earlier), it could not deprive the government of resources considered essential to ensure that success. Given the policy in place, cutting off vital aid would have been self-defeating. Afghan counterparts, of course, were well aware of this conundrum, and, understanding the limits of conditionality, were not highly motivated by it. The leverage, in other words, operated in both directions.

5. Recognise the limits of U.S. ability to impose its will where doing so is not existentially vital.

An overarching lesson to draw from Afghanistan, and one that will require rigorous examination to define thoroughly, is that the experience shows the limits of America’s ability to impose its will. Those limits can be seen in some of the more specific points highlighted above. But there are also broader questions to explore about the political judgments that were made to support the invasion and regime change in the first place, to sustain the military effort against the evidence of its poor results, and to end the intervention through the 2020 U.S.-Taliban agreement in a way that virtually assured the Taliban’s return to power. 

The Bush administration decided to invade Afghanistan not only to go after al-Qaeda but to punish the Taliban and make an example of them. It did not invade because Afghanistan itself was a place central to U.S. national security interests. For the next twenty years, the intervention cost nearly 2,500 American lives, tens of thousands of Afghan lives, enormous financial resources, and the time and energies of thousands of U.S. service members, diplomats, aid workers and others. And yet it remained – except for the threat of terrorism, to which the nation-building work was only tangentially related – peripheral to U.S. interests. That is not a circumstance conducive to success at so difficult a set of tasks. Afghanistan was, for most of twenty years, considered too important to fail but, ultimately, not important enough to stay forever, staving off the Taliban’s return.

What to Do Now?

In the wake of the U.S. military withdrawal, the Afghan government’s collapse, and the Taliban’s 15 August takeover, the United States must now entirely reformulate its policy. I will briefly suggest a few ideas related only to immediate steps.

Afghanistan is headed toward becoming the world’s greatest humanitarian crisis. Drought, increased displacement due to conflict, economic deterioration, the COVID-19 pandemic and other factors were worsening the humanitarian situation even prior to 15 August. Since then, the country has suffered an enormous economic shock. The suspension of U.S. and other foreign aid, freeze of state assets and effects of sanctions have produced widespread joblessness, hunger and a severe liquidity crisis.

The humanitarian and economic crisis already emerging in Afghanistan shows that it will not be possible to stand with the Afghan people ... while isolating the regime governing them.

A collision has occurred between two longstanding themes of U.S. policy. For years, U.S. officials told the Taliban that if they gained power through military means, rather than through a negotiated political settlement, they would rule only as a pariah regime, starved of resources. At the same time, the United States offered regular assurances that it would not abandon the Afghan people and that the lesson had been learned from the post-Soviet withdrawal period in the 1990s that washing its hands of Afghanistan could ultimately come to harm U.S. security interests. The humanitarian and economic crisis already emerging in Afghanistan shows that it will not be possible to stand with the Afghan people in any practical sense while isolating the regime governing them.

It will be important for the United States now to be clear-eyed about how best to advance its interests in Afghanistan, not allowing the pain and distastefulness of losing the war to stand in the way of an objective assessment of the importance of helping millions of Afghans. Greater impoverishment of Afghanistan under the Taliban is likely, but a glide path to a much lower level of international support rather than allowing the economy and public services to tip over a cliff would be more humane. That approach – which would entail some relaxation of sanctions and easing the complete cut-off of development aid – would also take account of U.S. participation in enabling over the last twenty years Afghanistan’s extreme aid dependency and, thus, the state’s precariousness.

An at least modestly more engaged approach – in terms of diplomacy and development – would also take account of the reality that isolation of the Taliban regime is not likely to produce results favourable to U.S. interests. Having proved resilient in the face of significant U.S. military pressure, the Taliban are highly unlikely to shape core policies or modify their ideology in response to financial pressure or the use of aid as leverage. They might, however, cooperate in limited areas, even potentially (if only secretly) on counter-terrorism – or at least such cooperation is a possibility to probe through engagement. Isolation, on the other hand, holds no chance of producing cooperation.

As the humanitarian and economic situation worsens in Afghanistan in the days and weeks ahead, politically difficult decisions will need to be made, and robust diplomacy will be needed to bring into alignment with U.S. policy, as much as possible, the policies of allies and of Afghanistan’s influential neighbours. The U.S. policy agility and pragmatism now needed in dealing with Afghanistan’s new rulers requires the support of the U.S. Congress.

Subscribe to Crisis Group’s Email Updates

Receive the best source of conflict analysis right in your inbox.