Special Inspector General for Afghanistan Reconstruction

Lessons Learned Program

August 2021

What We Need to Learn:Lessons from Twenty Years of Afghanistan ReconstructionInteractive Summary

Read the Full Report


Afghan army commandos stand on a sand bank in Marjah’s Balakino Bazar neighborhood, on February 24, 2010. (AFP photo by Patrick Baz)

The U.S. government has now spent 20 years and $145 billion trying to rebuild Afghanistan, its security forces, civilian government institutions, economy, and civil society. The Department of Defense (DOD) has also spent $837 billion on warfighting, during which 2,443 American troops and 1,144 allied troops have been killed and 20,666 U.S. troops injured. Afghans, meanwhile, have faced an even greater toll. At least 66,000 Afghan troops have been killed. More than 48,000 Afghan civilians have been killed, and at least 75,000 have been injured since 2001—both likely significant underestimations.

The extraordinary costs were meant to serve a purpose—though the definition of that purpose evolved over time. At various points, the U.S. government hoped to eliminate al-Qaeda, decimate the Taliban movement that hosted it, deny all terrorist groups a safe haven in Afghanistan, build Afghan security forces so they could deny terrorists a safe haven in the future, and help the civilian government become legitimate and capable enough to win the trust of Afghans. Each goal, once accomplished, was thought to move the U.S. government one step closer to being able to depart.

While there have been several areas of improvement—most notably in the areas of health care, maternal health, and education—progress has been elusive and the prospects for sustaining the progress that was made are dubious. The U.S. government has been often overwhelmed by the magnitude of rebuilding a country that, at the time of the U.S. invasion, had already seen two decades of Soviet occupation, civil war, and Taliban brutality.

Since its founding in 2008, SIGAR has tried to make the U.S. government’s reconstruction of Afghanistan more likely to succeed. Our investigations held criminals accountable for defrauding the U.S. government; our audits and special projects reports identified weaknesses in programs before it was too late to improve them; our quarterly reports provided near real-time analysis of reconstruction problems as they unfolded; and our lessons learned reports identified challenges that threaten the viability of the entire American enterprise of rebuilding Afghanistan, and any similar efforts that may come after it. SIGAR has issued 427 audits, 191 special project reports, 52 quarterly reports, and 10 comprehensive lessons learned reports. Meanwhile, SIGAR’s criminal investigations have resulted in 160 convictions. This oversight work has cumulatively resulted in $3.84 billion in savings for the U.S. taxpayer.

After conducting more than 760 interviews and reviewing thousands of government documents, our lessons learned analysis has revealed a troubled reconstruction effort that has yielded some success but has also been marked by too many failures. Using this body of work, as well as the work of other oversight organizations, SIGAR has identified seven key lessons that span the entire 20-year campaign and can be used in other conflict zones around the globe.

Read More (PDF)


Afghans sit on food donations in the back of a pick-up before distributing to needy families in Nawa District, Helmand Province, in September 2010. (Marine Corps photo by Sgt. Mark Fayloga)

Lesson 1: Strategy

President George W. Bush meets with the National Security Council in the Situation Room of the White House on September 20, 2001. (White House photo)
President George W. Bush meets with the National Security Council in the Situation Room of the White House on September 20, 2001. (White House photo)

The U.S. government continuously struggled to develop and implement a coherent strategy for what it hoped to achieve.

The challenges U.S. officials faced in creating long-term, sustainable improvements raise questions about the ability of U.S. government agencies to devise, implement, and evaluate reconstruction strategies. The division of responsibilities among agencies did not always take into account each agency’s strengths and weaknesses. For example, the Department of State is supposed to lead reconstruction efforts, but it lacked the expertise and resources to take the lead and own the strategy in Afghanistan. In contrast, DOD has the necessary resources and expertise to manage strategies, but not for large-scale reconstruction missions with significant economic and governance components. This meant no single agency had the necessary mindset, expertise, and resources to develop and manage the strategy to rebuild Afghanistan. For the U.S. government to successfully rebuild a country, especially one still experiencing violent conflict, civilian agencies will need the necessary resources and flexibility to lead in practice, not just on paper.

This poor division of labor resulted in weak strategy. While initially tied to the destruction of al-Qaeda, the strategy grew considerably to include the defeat of the Taliban, an insurgent group deeply entrenched in Afghan communities, then expanded again to include corrupt Afghan officials who undermined U.S. efforts at every turn. Meanwhile, deteriorating security compelled the mission to grow even further in scope. U.S. officials believed the solution to insecurity was pouring ever more resources into Afghan institutions—but the absence of progress after the surge of civilian and military assistance between 2009 and 2011 made it clear that the fundamental problems were unlikely to be addressed by changing resource levels. The U.S. government was simply not equipped to undertake something this ambitious in such an uncompromising environment, no matter the budget. After a decade of escalation, the United States began a gradual, decade-long drawdown that steadily revealed how dependent and vulnerable the Afghan government remains.

Read More (PDF)

We were devoid of a fundamental understanding of Afghanistan. We didn’t know what we were doing.

National Security Council “War Czar,” Lieutenant General Douglas Lute

Lesson 2: Timelines

President Obama meets with his national security team on Afghanistan and Pakistan in the Situation Room of the White House, on October 20, 2010. (White House Photo by Pete Souza)
President Obama meets with his national security team on Afghanistan and Pakistan in the Situation Room of the White House, on October 20, 2010. (White House Photo by Pete Souza)

The U.S. government consistently underestimated the amount of time required to rebuild Afghanistan, and created unrealistic timelines and expectations that prioritized spending quickly. These choices increased corruption and reduced the effectiveness of programs.

The U.S. reconstruction effort in Afghanistan could be described as 20 one-year reconstruction efforts, rather than one 20-year effort. U.S. officials often underestimated the time and resources needed to rebuild Afghanistan, leading to short-term solutions like the surge of troops, money, and resources from 2009–2011. U.S. officials also prioritized their own political preferences for what they wanted reconstruction to look like, rather than what they could realistically achieve, given the constraints and conditions on the ground. Early in the war, U.S. officials denied the mission resources necessary to have an impact, and implicit deadlines made the task even harder. As security deteriorated and demands on donors increased, so did pressure to demonstrate progress. U.S. officials created explicit timelines in the mistaken belief that a decision in Washington could transform the calculus of complex Afghan institutions, powerbrokers, and communities contested by the Taliban.

By design, these timelines often ignored conditions on the ground and forced reckless compromises in U.S. programs, creating perverse incentives to spend quickly and focus on short-term, unsustainable goals that could not create the conditions to allow a victorious U.S. withdrawal. Rather than reform and improve, Afghan institutions and powerbrokers found ways to co-opt the funds for their own purposes, which only worsened the problems these programs were meant to address. When U.S. officials eventually recognized this dynamic, they simply found new ways to ignore conditions on the ground. Troops and resources continued to draw down in full view of the Afghan government’s inability to address instability or prevent it from worsening.

Read More (PDF)

It’s like pouring a lot of water into a funnel; if you pour it too fast, the water overflows that funnel onto the ground. We were flooding the ground.

David Marsden, former USAID official

Lesson 3: Sustainability

Members of PRT Kunar prepare to examine a bridge under construction in Marawara District, Kunar Province, in 2005. (Photo by Harold Ingram)
Members of PRT Kunar prepare to examine a bridge under construction in Marawara District, Kunar Province, in 2005. (Photo by Harold Ingram)

Many of the institutions and infrastructure projects the United States built were not sustainable.

Reconstruction programs are not like humanitarian aid; they are not meant to provide temporary relief. Instead, they serve as a foundation for building the necessary institutions of government, civil society, and commerce to sustain the country indefinitely. Every mile of road the United States built and every government employee it trained was thought to serve as a springboard for even more improvements and to enable the reconstruction effort to eventually end. However, the U.S. government often failed to ensure its projects were sustainable over the long term. Billions of reconstruction dollars were wasted as projects went unused or fell into disrepair. Demands to make fast progress incentivized U.S. officials to identify and implement short-term projects with little consideration for host government capacity and long-term sustainability. U.S. agencies were seldom judged by their projects’ continued utility, but by the number of projects completed and dollars spent.

Over time, U.S. policies emphasized that all U.S. reconstruction projects must be sustainable, but Afghans often lacked the capacity to take responsibility for projects. In response, the U.S. government tried to help Afghan institutions build their capacity, but those institutions often could not keep up with U.S. demands for fast progress. Moreover, pervasive corruption put U.S. funds sent through the Afghan government at risk of waste, fraud, and abuse. These dynamics motivated U.S. officials to provide most assistance outside Afghan government channels. While expedient, the approach meant that Afghan officials were not getting experience in managing and sustaining U.S. reconstruction projects over the long term. As a result, even when programs were able to achieve short-term success, they often could not last because the Afghans who would eventually take responsibility for them were poorly equipped, trained, or motivated to do so.

Read More (PDF)

The Afghan government had no motive to create their own budget process, because the [donor] spigot was turned on full force.

USAID official

Lesson 4: Personnel

U.S. Army officer and Afghan National Police practice tactical movement as part of the Focused District Development training program at the Kandahar Regional Training Center. (NTM-A photo by David Votroubek)
U.S. Army officer and Afghan National Police practice tactical movement as part of the Focused District Development training program at the Kandahar Regional Training Center. (NTM-A photo by David Votroubek)

Counterproductive civilian and military personnel policies and practices thwarted the effort.

The U.S. government’s inability to get the right people into the right jobs at the right times was one of the most significant failures of the mission. It is also one of the hardest to repair. U.S. personnel in Afghanistan were often unqualified and poorly trained, and those who were qualified were difficult to retain. DOD police advisors watched American TV shows to learn about policing, civil affairs teams were mass-produced via PowerPoint presentations, and every agency experienced annual lobotomies as staff constantly rotated out, leaving successors to start from scratch and make similar mistakes all over again. These dynamics had direct effects on the quality of reconstruction. There were often not enough staff to oversee the spending, and certainly not enough who were qualified to do so. This was particularly true for civilian agencies, such as State or the U.S. Agency for International Development (USAID), which should have been leading the effort but were unable to meaningfully perform that role. This compelled the better-resourced DOD to fill the void, creating tensions with civilian agencies that often had different ideas but fewer staff to offer.

Read More (PDF)

I got this job because I had a pulse and a master’s degree.

USAID official

Lesson 5: Insecurity

Private security contractors protect the construction of the Khost-Gardez road, on March 30, 2010. (USAID photo)
Private security contractors protect the construction of the Khost-Gardez road, on March 30, 2010. (USAID photo)

Persistent insecurity severely undermined reconstruction efforts.

The absence of violence was a critical precondition for everything U.S. officials tried to do in Afghanistan—yet the U.S. effort to rebuild the country took place while it was being torn apart. For example, helping Afghans develop a credible electoral process became ever more difficult as insecurity across the country steadily worsened—intimidating voters, preventing voter registration, and closing polling stations on election day. In remote areas where the Taliban contested control, U.S. officials were unable to make sufficient gains to convince frightened rural Afghans of the benefits of supporting their government. Insecurity and the uncertainty that it spawns have also made Afghanistan one of the worst environments in the world to run a business. The long-term development of Afghanistan’s security forces likewise saw a number of harmful compromises, driven by the immediate need to address rising insecurity. The danger meant that even programs to reintegrate former fighters faltered, as ex-combatants could not be protected from retaliation if they rejoined their communities.

Read More (PDF)

[Rural Afghans] didn’t feel like their government could protect them, and we couldn’t change that no matter how much we built, how many people we employed, or how much they liked us for it.

USAID official

Lesson 6: Context

U.S. soldiers with the 1st Battalion, 5th Cavalry Regiment man an observation post overlooking Highway 1 in Wardak province, on Aug. 12, 2013. (U.S. Army photo by 1st Lt. Yau-liong Tsai)
U.S. soldiers with the 1st Battalion, 5th Cavalry Regiment man an observation post overlooking Highway 1 in Wardak province, on Aug. 12, 2013. (U.S. Army photo by 1st Lt. Yau-liong Tsai)

The U.S. government did not understand the Afghan context and therefore failed to tailor its efforts accordingly.

Effectively rebuilding Afghanistan required a detailed understanding of the country’s social, economic, and political dynamics. However, U.S. officials were consistently operating in the dark, often because of the difficulty of collecting the necessary information. The U.S. government also clumsily forced Western technocratic models onto Afghan economic institutions; trained security forces in advanced weapon systems they could not understand, much less maintain; imposed formal rule of law on a country that addressed 80 to 90 percent of its disputes through informal means; and often struggled to understand or mitigate the cultural and social barriers to supporting women and girls. Without this background knowledge, U.S. officials often empowered powerbrokers who preyed on the population or diverted U.S. assistance away from its intended recipients to enrich and empower themselves and their allies. Lack of knowledge at the local level meant projects intended to mitigate conflict often exacerbated it, and even inadvertently funded insurgents.

Read More (PDF)

In the majority of districts, we never even heard the real problems of the people. We made assumptions, conducted military operations, brought in government staff, and assumed it would lead to security and stability.

Jabar Naimee, former governor of Kunduz, Wardak, Khost, and Laghman provinces

Lesson 7: Monitoring and Evaluation

A USAID-funded program evaluator interviews an elder in Aybak District, Samangan Province, in May 2014. (USAID photo)
A USAID-funded program evaluator interviews an elder in Aybak District, Samangan Province, in May 2014. (USAID photo)

U.S. government agencies rarely conducted sufficient monitoring and evaluation to understand the impact of their efforts.

Monitoring and evaluation (M&E) is the process of determining what works, what does not, and what needs to change as a result. Conceptually, M&E is relatively straightforward, but in practice, it is extremely challenging. This is especially true in complex and unpredictable environments like Afghanistan, where staff turnover is rapid, multiple agencies must coordinate programs simultaneously, security and access restrictions make it hard to understand a program’s challenges and impact, and a myriad of variables compete to influence outcomes. The absence of periodic reality checks created the risk of doing the wrong thing perfectly: A project that completed required tasks would be considered “successful,” whether or not it had achieved or contributed to broader, more important goals.

SIGAR’s extensive audit work on sectors spanning health, education, rule of law, women’s rights, infrastructure, security assistance, and others collectively paints a picture of U.S. agencies struggling to effectively measure results while sometimes relying on shaky data to make claims of success. The U.S. government’s M&E efforts in Afghanistan have been underemphasized and understaffed because the overall campaign focused on doing as much as possible as quickly as possible, rather than ensuring programs were designed well to begin with and could adapt as needed. As a result, the U.S. government missed many opportunities to identify critical flaws in its interventions or to act on those that were identified. These shortcomings endangered the lives of U.S., Afghan, and coalition government personnel and civilians, and undermined progress toward strategic goals.

Read More (PDF)

It was impossible to create good metrics. We tried using troop numbers trained, violence levels, and control of territory, and none of it painted an accurate picture. At the end of the day, there was nothing for us to latch on to except for number of attacks against civilians, ANDSF, and [coalition partners].

Senior U.S. official


As part of Operation Medusa, soldiers from 4th Platoon, Bravo Company, Royal Canadian Regiment, burst into a suspected Taliban stronghold in Pashmul, a small village north of Panjwai District, on September 11, 2006. (Photo by Graeme Smith)

In conclusion, this report raises critical questions about the U.S. government’s ability to carry out reconstruction efforts on the scale seen in Afghanistan. As an inspector general’s office charged with overseeing reconstruction spending in Afghanistan, SIGAR’s approach has generally been technical; we identify specific problems and offer specific solutions. However, after 13 years of oversight, the cumulative list of systemic challenges SIGAR and other oversight bodies have identified is staggering. As former National Security Advisor Stephen Hadley told SIGAR, “We just don’t have a post-conflict stabilization model that works. Every time we have one of these things, it is a pick-up game. I don’t have confidence that if we did it again, we would do any better.”

This was equally apparent after the Vietnam War, when a war-weary and divided country had little appetite to engage in another similar conflict. After Vietnam, for example, the U.S. Army disbanded most active duty civil affairs units and reduced the number of foreign area officers, the Army’s “regionally focused experts in political-military operations.” Special Forces moved away from counterinsurgency and instead focused on conducting small-scale operations in support of conventional forces. And USAID’s global staff was gradually cut by 83 percent.

In other words, according to former Vice Chief of Staff of the Army General Jack Keane, “After the Vietnam War, we purged ourselves of everything that had to do with irregular warfare or insurgency, because it had to do with how we lost that war. In hindsight, that was a bad decision.” After all, declining to prepare after Vietnam did not prevent the wars in Iraq and Afghanistan; instead, it ensured they would become quagmires.

Rather than motivating the U.S. government to improve, the difficulty of these missions may instead encourage U.S. officials to move on and prepare for something new. According to Robert Gates, former secretary of defense from 2006–2011:

I have noticed too much of a tendency towards what might be called ‘Next-War-itis,’ the propensity of much of the defense establishment to be in favor of what might be needed in a future conflict. . . . Overall, the kinds of capabilities we will most likely need in the years ahead will often resemble the kinds of capabilities we need today.

The post-Afghanistan experience may be no different. As this report shows, there are multiple reasons to develop these capabilities and prepare for reconstruction missions in conflict-affected countries:

  1. They are very expensive. For example, all war-related costs for U.S. efforts in Afghanistan, Iraq, and Pakistan over the last two decades are estimated to be $6.4 trillion.
  2. They usually go poorly.
  3. Widespread recognition that they go poorly has not prevented U.S. officials from pursuing them.
  4. Rebuilding countries mired in conflict is actually a continuous U.S. government endeavor, reflected by efforts in the Balkans and Haiti and smaller efforts currently underway in Mali, Burkina Faso, Somalia, Yemen, Ukraine, and elsewhere.
  5. Large reconstruction campaigns usually start small, so it would not be hard for the U.S. government to slip down this slope again somewhere else and for the outcome to be similar to that of Afghanistan.

Nevertheless, after the last two decades in Afghanistan and Iraq, State, USAID, and DOD have all signaled they do not see large-scale missions as likely in the future. The Stabilization Assistance Review approved by all three agencies in 2018 noted, “There is no appetite to repeat large-scale reconstruction efforts, and therefore our engagements must be more measured in scope and adaptable in execution.” Just as after Vietnam, today U.S. policymakers and the public they serve may have sound reasons for avoiding another prolonged conflict and reconstruction mission. However, that does not mean such an endeavor is avoidable in the future.

As SIGAR’s Stabilization report notes, “there will likely be times in the future when insurgent control or influence over a particular area or population is deemed an imminent threat to U.S. interests.” If the U.S. government does not prepare for that likelihood, it may once again try to build the necessary knowledge and capacity on the fly. As seen in Afghanistan and Iraq, doing so has proven difficult, costly, and prone to avoidable mistakes.

As former U.S. Ambassador to Afghanistan Ryan Crocker observed, “You have to start working on it before you need it.” One former senior DOD official likewise noted that rebuilding another country requires advanced skills that must be cultivated ahead of time. “You wouldn’t invent how to do infantry operations [or] artillery at the start of a war. You need [to already have] the science behind [reconstruction] and people who think about it 24/7.”

Building on SIGAR’s body of work, as well as the work of inspector general offices across the government, this report points to conceptual, administrative, and logistical work that should be done between large-scale reconstruction efforts to increase the U.S. government’s chances of success in future campaigns.

The nature and range of the investment necessary to properly prepare for these campaigns is an open question. In previous lessons learned reports, SIGAR has made recommendations for existing U.S. government offices to create a database of qualified personnel to call up when necessary, build interagency doctrine for security sector assistance, and establish anti-corruption offices within key agencies. As former U.S. envoy to Afghanistan James Dobbins observed, properly preparing “doesn’t mean that you have to have a standing capability to immediately train [an entire army], but you need to have the know-how and an ability to surge those kinds of resources.” Others have argued that such an ability requires a permanent office with the authority and funding to prepare for, plan, execute, and evaluate all reconstruction missions.

U.S. agencies should continue to explore how they can ensure they have the strategic planning capabilities, reconstruction doctrine, policies, best practices, standard operating procedures, institutional knowledge, and personnel structures necessary for both large and small reconstruction missions.