Carnegie Mellon University

US troops at Kabul airport

October 11, 2021

Were Twenty Years in Afghanistan “Costless”?

By Daniel Silverman and Kyle Larson

Bill Brink
  • Communications Director

There has been a significant outcry that the decision to withdraw US forces from Afghanistan was a mistake, especially given the chaos that unfolded there during the spring and especially over the summer of this year. A key pillar of this argument is that our presence in Afghanistan was virtually “costless” in terms of American lives lost. We argue here that this idea of a costless “forever war” is deeply misleading.

Proponents of continued American intervention in Afghanistan point out that there had been very few US military casualties for the last several years of the mission. In the words of former UK Secretary of State for International Development Rory Stewart, staying in Afghanistan was “the easiest thing to continue to do for the Afghan people” as there had been “zero casualties, zero risk over the last several years.” In a similar vein, New York Times columnist David Brooks backed continued intervention on the grounds that, “[o]ver the past few years, a small force of American troops has helped prevent some of the worst people on earth from taking over a nation of more than 38 million – with relatively few American casualties.”

Putting aside the financial and reputational costs of the intervention, it is true that US military deaths had been relatively low since 2015, when the US substantially scaled back the mission in Afghanistan. As shown in Figure 1, the number of American soldiers killed per year in Afghanistan peaked at 710 in 2010, declining steadily thereafter and averaging just sixteen per year from 2015-20. If indeed there were a meaningful benefit to continued intervention – and critics of the withdrawal generally point toward a humanitarian one – and it was low cost to do so, then why would we leave?


A superficial look at US casualties over time in Afghanistan is misleading

Despite the apparent logic of this line of thinking, it is crucial to ask why our casualties dropped so low and what this shift represented in strategic terms.

US casualties swelled in the early-to-mid Obama years because that was when we conducted the “surge” in Afghanistan. The number of US troops in Afghanistan grew from about 30,000 in 2008 to roughly 110,000 in 2011, and American military leaders were upbeat about our prospects. We sought to totally vanquish the Taliban and extend state control across the country. We tried to win.

Thinking we had prevailed, or perhaps realizing we could not, President Obama began a troop drawdown in 2011, with the intent of handing over security responsibilities to the Afghans in 2014. We began letting Afghan forces take the lead on the ground while conducting an increasingly aggressive bombing campaign to keep them afloat. At this point, we ceased trying to defeat the Taliban or control all of Afghanistan. While our leaders still projected confidence and would never admit it publicly, our strategy shifted to trying not to lose.

Understanding this difference is critical to thinking about what we were doing in Afghanistan, the costs we incurred while doing it, and the wisdom of continuing.

The US was fighting not to lose in Afghanistan. That wasn’t sustainable forever.

Two points are worth stressing about this post-2014 strategy. First, US intervention was only low-cost given that we were trying to avoid defeat. In the best-case scenario, supporters of continued American intervention were arguing for keeping a corrupt, weak, and illegitimate client state from total collapse. Moreover, this would have only occurred as long as we stayed in the country, as recent events have made clear that the Afghan army had no capacity to face the Taliban alone. This argument essentially boiled down to keeping our troops in Afghanistan and bombing the Taliban indefinitely in order to keep a flawed and failing proxy state on life support.

Second, it is likely that even this limited objective was not sustainable forever. Indeed, the Taliban was already gaining ground. Figure 2 shows that the Afghan government’s control over territory in Afghanistan declined steadily from 2015 until 2021, with just 18 percent of Afghan districts not under government control in 2015 and 67 percent not under government control prior to the US withdrawal, despite the increasingly intense US bombing. The US strategy of a minimal ground footprint and extensive air support did succeed in reducing American casualties, but it wasn’t succeeding at preventing the Taliban advance.


Moreover, this was despite the Taliban strategically holding back. A detailed quantitative analysis  found that even as early as 2014, the Taliban began trying to seize control of districts only once the Americans had fully abandoned them. Recent events confirmed this discovery, with the Taliban’s swift and decisive conquest immediately after US withdrawal. It would have been far harder to cling to the country through heavy bombing alone if we had no plans to withdraw, and the Taliban were fighting their hardest to retake it.

There is no such thing as free lunch (or costless war). Afghanistan is no different.

Where does this leave us? Much can be criticized about the specific manner of US withdrawal. But in the big picture, this should leave us very skeptical about superficial claims that our continued intervention was costless, or even relatively low-cost. These claims ignore the fact that the existing intervention was low-cost in American lives precisely because it had sharply limited aims and was only achieving them temporarily. In other words, there was no long-term, low-cost option in Afghanistan. Don’t be fooled by flawed or disingenuous claims that we could eat our cake and have it too. 

Daniel Silverman is an Assistant Professor of Political Science in the Institute for Politics and Strategy at Carnegie Mellon University.

Kyle Larson is a Senior Research Associate at the Chicago Project on Security and Threats (CPOST) at the University of Chicago.