The Biden administration is pressing ahead with the planned withdrawal of international forces from Afghanistan. Early gestures by the Taliban suggested it was willing to demobilize in exchange for a negotiated settlement. Late last year, the group even agreed to a set of rules for negotiating a comprehensive cease-fire.
But recent military operations have changed assessments of what is likely to follow the final withdrawal of U.S. troops on Sept. 11. In recent weeks, the Taliban has led an unexpected offensive, seizing dozens of districts and cities in Northern Afghanistan — along with Afghanistan’s main border crossing with Tajikistan. As the Taliban quickly moved across the region, local leaders coordinated peace agreements and Afghan troops exchanged U.S. vehicles and weapons for safe passage.
In response, the U.S. intelligence community has updated a previous, and more optimistic, assessment of the security transition, noting that the Afghan government might collapse within six months of withdrawal.
What explains these rapid shifts in perceptions of the Taliban’s strength and the Afghan government’s fragility? Our research on earlier U.S. withdrawals suggests the Taliban has been strategic in the use of violence, using restraint to influence military assessments of its capabilities to encourage more rapid withdrawals.
This strategy, however, suggests the worst violence is likely to emerge after September. These findings also suggest that even the more grim current assessments of the real balance of power between the Taliban and the Afghan government likely exaggerate the government’s military capabilities.
Here’s how we did our research
In a recent paper, we studied the two phases of the 2011-2014 withdrawal of international forces in Afghanistan. The first phase of withdrawal included transferring military authority to local forces. The second phase involved the physical departure of international troops, as the number of foreign forces decreased from 140,000 to 12,000, along with the closure, retrograde or transfer of nearly 800 military bases.
Our study used exceptionally granular geotagged and time-stamped data on different types of insurgent and security operations — we used data collected since the start of NATO activity in Afghanistan in 2001. This data is otherwise known as SIGACTS and has been used widely to study combat operations in Afghanistan and Iraq. We combined the data with survey records of 370,000 civilians gathered between 2008 and 2016 (called ANQAR), which detail perceptions of security transitions, perceptions of territorial control and the extent of local security provision. Our article provides further details on the methodology we used.
The transfer of military authority was split across five waves, giving an opportunity to compare trends in violence and public perceptions of security provision across districts that were assigned to different transfer dates.
The closure of bases was more haphazard, and also less well documented. To study this second phase, we leveraged a simple but important fact about the timing of physical exit: the geographic location of bases — their distance to military air hubs used to transport supplies out of Afghanistan — partly determined when bases were shut down.
Here’s what we found
Recall the initial security transition that began a decade ago had two phases. During the initial phase, we compared districts in which the local transition of authority had been implemented to other districts, before and after the transition. During this first phase, we noted both a decrease in actual violence outcomes and an increase in perception of security as reported by the SIGACTS and ANQAR survey data, respectively.
What about the second phase, as international troops departed in 2011 and later? We can gauge the effect of physical withdrawal of NATO troops on security conditions and civilian perceptions of military performance. Since the decision to close or hand over individual bases was highly discretionary and district specific, it is important to whether the sequencing of base closures across different regions was influenced by trends in violence. We take advantage of the geography of withdrawal to address this issue. Leveraging the fact that proximity to military air hubs prompted some bases to close earlier than others, we found that the physical withdrawal of foreign troops and base closure is associated with a drastic worsening of the conflict situation by 2016.
Our research suggests that violence decreased with the transfer of authority (the first phase) but increased with the physical exit of international troops (the second phase). What could explain this result? We conclude the most compelling account of these patterns of violence is that the Taliban was lying low, waiting for coalition forces to begin physically withdrawing before beginning to apply more military pressure on local forces. This accelerated withdrawal by giving a false impression of the capabilities of local groups to take on the security challenges of fighting alone.
The Afghanistan Papers, released by The Washington Post in 2019, give us an opportunity to explore these decisions through internal assessments. A Special Inspector General for Afghanistan Reconstruction report, reflecting on these exit interviews, came to the conclusion that various military benchmarks had failed to adequately measure the capabilities of local Afghan forces, masking their fundamental weaknesses. The report concluded these forces were “ill-prepared to deal with deteriorating security after the drawdown of U.S. combat forces.”
What happens now? The Taliban seems likely to take advantage of the power vacuum left behind with the final withdrawal. Taliban tactics over the past few months have revealed its interest in seizing border crossings as well as territory that was contested during its previous hold on government. Taking control of these areas might enable the Taliban to repress its political opponents while generating tens of millions in revenue before the end of the year. Without an unexpected surge by Afghan forces, the Taliban may be able to wrest power from the elected government of Afghanistan, and with it, 20 years of democratic reform.
Thiemo Fetzer (@fetzert) is professor of economics at the University of Warwick. Pedro Souza (@pedroclsouza) is senior lecturer of economics at Queen Mary University of London. Oliver Vanden Eynde (@olivervdeynde) is associate professor at the Paris School of Economics and Research Fellow at the French National Centre for Scientific Research (CNRS). His research centers on the political economy of conflict and public service provision, with a special focus on South Asia. Austin L. Wright (@austinlwright) is assistant professor at the University of Chicago’s Harris School of Public Policy. His research is primarily focused on the political economy of conflict and crime, with a special focus on Afghanistan.