The $28 Billion Question: How We Calculated The Cost Of Afghanistan Retrograde

The Pentagon says it doesn’t know what it cost to leave Afghanistan, so we did the math.

The $28 Billion Question: How We Calculated The Cost Of Afghanistan Retrograde

Six billion dollars. That’s what the Pentagon says it cost to get the U.S. military out of Afghanistan. It’s a hefty price tag. But it’s incomplete. In fact, the figure wildly understates what U.S. taxpayers actually spent. The Pentagon tallies up only those expenses incurred outside the operations of the military—so-called external costs, such as contracting with outside commercial carriers or paying Pakistani transit fees. What the Pentagon’s numbers don’t include is the work done by troops to pack up outposts and close down bases; to move equipment and orchestrate convoys; to identify and catalog which items to take home, which to leave behind, and which to dismantle; plus all the security elements required to protect these activities. In other words, the Pentagon’s figure excludes the core part of engineering and executing the retrograde.


Fast Company asked for those more specific numbers, but the Pentagon doesn’t perform the kind of cost accounting that the private sector would. The Pentagon says it only tracks high-level costs—such as how much fuel the entire Afghan operation used, for example, or the overall cost of equipment maintenance—but doesn’t break those costs down into the specific activities related to packing up, the way a private company would.

But there is a way to estimate the overall costs of the retrograde—and according to our calculations, the tally is at least four times as much as the Pentagon’s report, and perhaps as much as seven times the official figure. When we began to engage in the math, we wondered: Would the final tally be closer to $10 billion or $50 billion? Our most conservative estimate landed at $28 billion. Here’s how we got there.

We started with the total war budget for 2013 and 2014.

According to a report by the Congressional Research Service, the Afghanistan war budgets for 2013 and 2014 were $86 billion and $70.4 billion, respectively. (See table, right.) By definition, the cost of the retrograde is a subset of that overall spending. Our calculation process would involve peeling away the nonretrograde expenses, to reveal what was left.

We could have been more aggressive. We could have also included some piece of the war budget for 2012, since the retrograde started back then. But the efforts were comparatively small at that date. To be conservative, we left that year out entirely.

We also left out all stateside activities. Planners and analysts back in the U.S. were heavily involved in determining and coordinating logistics, helping to decide what was worth bringing back and what should be donated or destroyed in-country, managing transportation, developing tools and systems, and so on. But because we couldn’t access the cost of those activities—or any figures that gave us a sense of their scale—we decided to exclude them entirely. As with every other choice-point in this exercise, we chose the most conservative path.

Beginning total: $156.4 billion.


We then removed $30 billion in possible budget chicanery.

Some parts of a Pentagon budget may include ancillary expenses—costs for other activities that have been placed inside a larger allocation. According to Todd Harrison, a defense budget expert with the Center for Strategic and Budgetary Assessments, that’s what happened with the Afghanistan budget. When Congress slashed the Pentagon’s overall budget in 2014, the Afghanistan war effort, known as an “overseas contingency operation,” or OCO, was not counted against the budget caps. To avoid breaching the caps, the Department of Defense moved money from its base budget into the OCO budget. As a result of these accounting maneuvers, Harrison contends that Congress and the Pentagon reclassified $30 billion of non-OCO items into the OCO budget.

Not everyone agrees with Harrison, but since we’re looking to be conservative, we are attuned to any possible padding of costs. So we removed $30 billion from our initial starting cost.

New total: $126.4 billion.

Next, we removed $10.2 billion for items that were unrelated to retrograde.

A few major line items in the OCO budget seem clearly unrelated to retrograde: the Afghan Security Forces Fund and the Task Force for Business & Stability Operations. They came to $10.2 billion, which we subtracted from our total.

New total: $116.2 billion.

Next, we took out $6 billion in external retrograde costs.

The Pentagon’s $6 billion in external costs for the retrograde definitely needed to be part of our final figure. For now, we took that amount out and set it aside. What remained were shared costs like Force Protection, Military Intelligence, and Military Construction (see table), which would have supported retrograde as well as other war activities. After we determined what portion of the shared costs should be apportioned to retrograde, we would add the $6 billion back in.


New total: $110.2 billion.

We applied 20 percent of that to retrograde.

If there’s a leap in our estimate, it is in this step. But as leaps go, it’s a pretty small one. Our current running total, $110.2 billion, would have covered all of the operations under way in Afghanistan in 2013 and 2014. During those years, there were three major activities taking place: the training of the Afghan security forces, counterinsurgency and counterterrorism operations, and retrograde. What percentage of those overall costs should be assigned to retrograde? Once again, the Pentagon declined to respond to our requests for details.

An aggressive calculation would split those three areas equally, setting retrograde costs at one-third of the total, or $36.7 billion. We chose to maintain a conservative posture—not simply moving our estimate from one-third to one-quarter, but dropping all the way to one-fifth. That would mean retrograde used far fewer resources than offensive combat or training, just 20 percent of the total. Given the priority of the retrograde, particularly during the latter stages of withdrawal, this estimate may seem unduly cautious. Still, we’d rather err on the side of caution. One-fifth of $110.2 billion comes to a little over $22 billion.

New total: $22 billion.

Now we add back the $6 billion in external costs.

We then added our $22 billion for shared costs to the $6 billion in external costs provided by the Pentagon. This comes to a conservative estimate of $28 billion. (The more aggressive estimate: If you assume shared costs for retrograde were one-third of our war costs, the total would come to nearly $43 billion.)

Final total: $28 billion.


It’s unlikely that American taxpayers will ever know the exact cost of getting out of Afghanistan. But this estimate—which is conservative, and probably too low—is the closest anyone is likely to get, unless the Pentagon releases more data.

About the author

E.B. Boyd (@ebboyd) has holed up in conference rooms with pioneers in Silicon Valley and hunkered down in bunkers with soldiers in Afghanistan