Why Kunduz Fell
Daniel Fisher and Christopher Mercado
As recently as three years ago, Kunduz Province appeared to be a bastion of relative stability in Afghanistan. Clearing operations conducted by conventional infantry units, direct action SOF raids, and an apparently successful local militia program had wreaked havoc on the insurgency in Kunduz. The Taliban abruptly shattered that perception in the early morning hours of September 28, 2015 when they seized Kunduz City—the first provincial capital to fall since the U.S.-led invasion of 2001. Warnings signs reflecting latent instability in Kunduz were plentiful, however, and the fall of the provincial capital should not have come as a surprise. The reality, evidently unperceived by both U.S. and Afghan government officials, was that the Taliban had been patiently setting the conditions to capture Kunduz City as early as 2012. Though only a year has passed, the Taliban have yet again launched a coordinated attack on Kunduz and its provincial capital of Kunduz City.
Both of us served in Kunduz from 2011-12, at a time when it seemed as though the province would continue on a trajectory of increasing stability. Although the insurgency remained active during our time in the province, ISAF had created ample room to “hold” and “build.” District “shadow governors” were killed, and junior Taliban commanders were defecting and reconciling with the Kunduz government, which was steadily expanding its reach into the rural countryside. Our unit’s operations supported both of these phases, and we believe that they too contributed to short-term stability in the province. In light of the Taliban’s resurgence in Kunduz, however, it is clear that much of this apparent success was fleeting. This analysis attempts to explain why.
Figure 1: Map of Kunduz Province
Source: Google Maps
Hopes were high for Kunduz in mid-2010. The deployment of a U.S. infantry battalion (1-87 IN, of the 10th Mountain Division) helped Afghan forces to contest areas that were previously controlled by the insurgency. A relentless direct action SOF campaign simultaneously threw local Taliban leadership into disarray. Meanwhile, the cooptation of key militias previously fighting on the side of the Taliban provided the government with auxiliary forces that could challenge the insurgency in rural areas, where sympathy for the Taliban was strongest.
Lacking the ability to control substantial swaths of territory, the insurgency shifted underground, focusing its operations on a two-pronged campaign involving the assassination of provincial leaders and suicide attacks against the populace—the latter conducted primarily in Kunduz City. As our unit (2-18 IN, 170th IBCT) began to penetrate deep into districts sympathetic to the Taliban in 2011-12, the insurgency responded by increasing the rate of its IED attacks, purportedly aided by experts from Northwest Pakistan who provided technical explosives expertise. Although this reduced our capacity to fully capitalize on, and consolidate, security gains, the Taliban nonetheless remained largely marginalized in Kunduz, engaging in mostly opportunistic attacks against ISAF and ANSF in lieu of large-scale confrontations.
In February 2012, however, exogenous events intervened to push momentum in Kunduz back in favor of the Taliban. Angered by the recent Koran burnings at Bagram airbase, locals in Imam Sahib launched a protest outside the gates of Combat Outpost Whitehorse (previously COP Fortitude), the size and location of which had allowed ISAF to project security over the entire Northern half of the province. The demonstration, which began peacefully enough, rapidly turned violent when protesters threw at least one hand grenade into the COP, injuring seven U.S. service members. As a result of this incident, COP Whitehorse was closed, significantly reducing ISAF’s ability to control both Imam Sahib and Dashti Archi districts (located North and Northeast of Kunduz City, respectively).
With the surge complete and the ISAF drawdown well under way, the security situation in Kunduz began to deteriorate. In March, 2013, the Taliban assassinated Imam Sahib Police Chief Abdul Qayoum Ibrahimi. This was a substantial boon to the insurgency: the Uzbek Ibrahimi family had long exerted both de facto and official control over Imam Sahib. The assassination therefore had a disproportionately destabilizing effect on this important Northern district, home to key smuggling routes into Tajikistan.
In October 2013, German units transferred their Provincial Reconstruction Team (PRT) compound over to ANSF. Reflecting a pattern that now verges on the cliché, this final phase of the ISAF drawdown in Kunduz set the stage for the 2014 fighting season, during which the insurgency made substantial gains, particularly in the oft-contested districts of Chahar Dara and Dashti Archi. By September 2014, as the fighting season drew to a close, the Taliban were closing in on Kunduz City. Although the Afghan government painted a rosy picture following pre-presidential-election clearing operations conducted throughout the province, the real story was far bleaker. As one Afghan intelligence officer anonymously noted,
“The Taliban are five kilometers away from the provincial city centre and one kilometre from district centre. The operation in Chardara was just to try and save the city and the district centre from falling to the Taliban - to push them back and then try to open a base so that neither the district nor Kunduz city can be taken.”
Regaining control of Chahar Dara, the officer added, was an impossibility, noting, “We would need a lot more support.”
The limited efficacy of large-scale operations against insurgencies aside, one especially concerning aspect of the ANSF-led clearing efforts was the extent to which they relied on local militias to “hold” territory. The Afghan government was reportedly struggling to compensate Afghan Local Police (ALP) units, a circumstance that not only called the mid- and long-term loyalty of the militias to the Afghan government into question, but also led them to impose illegal taxes on local communities.
Meanwhile, Pakistani military operations in Waziristan in the early summer of 2014 significantly bolstered the Taliban’s strength in Kunduz. By this time, all the warning signs were in place: a year before the seizure of the provincial capital apparently caught U.S. and Afghan officials by surprise, many local residents were already expressing significant alarm. In the words of one businessman, “The Taliban could take the [Kunduz] city any time they want to. They just don’t want to bother with holding and managing it right now.” Settling in for the winter, the emboldened insurgency set its sights on the provincial capital.
In the spring and summer of 2015, the Taliban capitalized on the territorial gains it made during the previous fighting season, consolidating control over Chahar Dara and Dashti Archi, and extending their presence in Imam Sahib, Aliabad, and Khanabad districts—the latter of which had increasingly suffered from the overbearing, government-sponsored ALP. Having effectively surrounded Kunduz City over the course of the 2015 fighting season, the Taliban finally seized the provincial capital in September 2015.
Figure 2: Timeline - Fall of Kunduz
Why Kunduz Fell
Despite ISAF’s progress in the province before and during our time in Kunduz, several structural factors either remained unaddressed by Coalition Forces, or rapidly emerged in the months leading up to the seizure of Kunduz City, and thus contributed to current instability. Specifically, we attribute the fall of the provincial capital to five underlying causes. Below, we expand on each.
Kunduz was an economy of force operation within an economy of force operation. ISAF lacked the resources and commitment required for irreversible gains and both the Taliban and Afghans were watching.
As Afghanistan was an economy of force operation to the strategically more significant Iraq from 2003 until about 2009, so too was Kunduz an economy of force operation to the operationally more significant RC-East and RC-South. The distribution of ISAF forces in August 2010, at which time the Afghanistan surge was well under way, readily demonstrates this circumstance. At this critical juncture, the overall ISAF troop level in RC-North was 11,000. By contrast, total forces numbered 32,000 in RC-East, and 35,000 in RC-South.
Although the distribution of forces reflected the relative volatility of Afghanistan’s South and East, low troop strength in the North prevented ISAF from consolidating hard-won security gains. We experienced this phenomenon directly: Bravo Company, 2-18 IN was responsible for stabilizing an area with a population of more than 300,000 people with a company sized unit of approximately 160 Soldiers, only half of which could actively patrol out of our combat outpost (COP) in Imam Sahib at any one time. Although the combined number of ISAF and ANSF forces was much higher, the ANSF in our area of operations (as well as throughout the province) were largely relegated to static defensive positions, fixed site security, and checkpoints designed to hold terrain, rather than to project security across a broader area. One of the few exceptions was the ANP Provincial Quick Reaction Force. However, the QRF primarily conducted targeted raids rather than population-centric patrols.
Adopting an economy of force approach, particularly in a province as ethnically and tribally complex as Kunduz, also produced disproportionate influence among local militia groups, which began to fill gaps in security as early as 2009. By this time, under pressure from ISAF in Southern Afghanistan, the Taliban had migrated North, and now controlled significant swathes of Kunduz—including “most of Chahar Dara, smaller areas of Aliabad and Dashti Archi districts, and spots in the remaining districts,” according to the Afghanistan Analysts Network. The Taliban buildup was especially alarming to provincial officials due to the need to provide security at the ballots for the 2009 presidential election.
As a result, provincial leaders began forming militias—known as “Arbaki”—under a program sponsored by Afghanistan’s National Directorate of Security (NDS). In most, if not all cases, these militias served to consolidate local patronage networks—strengthening Tajik and Uzbek warlords connected to the provincial leadership, and redistributing responsibility for security away from ANSF. With their position as government-sponsored groups already established, the Arbaki would later form the basis of the U.S.-sponsored Kunduz ALP (Afghan Local Police) program, the composition of which would prove to have further adverse and unintended consequences for the distribution of security and legitimacy within the province.
Absent a credible deterrent, the Taliban regained significant footholds in the province, particularly in the districts of Chahar Dara and Dashti Archi, and patiently set conditions for a future strategic offensive to seize Kunduz.
As a result of the economy of force approach, the Taliban were able to retain significant footholds in Kunduz despite the presence of ISAF, most notably in the districts of Chahar Dara and Dashti Archi. In the former, German national caveats constrained the German Army’s activities primarily to multi-day patrol base operations without establishing a semi-permanent presence in areas of the district especially sympathetic to the Taliban. This cautious approach had the effect of temporarily flushing Taliban fighters out of Chahar Dara during major Bundeswehr operations. However, as neither Coalition Forces nor ANSF permanently held territory, the Taliban promptly and routinely returned.
Counter-IED measures employed to access Chahar Dara frequently took hours, allowing Taliban fighters ample time to exit the district along North-South running egress routes. Moreover, clearing operations created massive operational inefficiencies: Bundeswehr units allocated approximately half a day to enter and exit areas of the district that supported the Taliban, imposing an opportunity cost of time—a form of insurgent tax—on the Germans. Finally, in our experience, Bundeswehr units in Chahar Dara mostly conducted “presence patrols,” a label employed liberally and euphemistically to describe missions in which Soldiers do little more than merely walk around.
Although we were far more operationally aggressive in Dashti Archi, we were nonetheless limited by two factors that allowed the Taliban to regain a foothold in the district. First, although we were not compelled to clear our way into Archi after a series of initial operations conducted jointly with an attached engineer unit, the district’s ANP headquarters was a long distance from our combat outpost in Imam Sahib. Thus, like the Bundeswehr in Chahar Dara, we generally allocated a half day to ingress. Second, due to the district’s isolation, we were unable to patrol in Archi for more than a few days at a time before sustainment needs forced us to resupply. Consequently, we were restricted to similar “commute to work” operations during which we used the ANP district headquarters as a forward assembly area. Dashti Archi effectively became our company’s economy of force operation; the time and resources simply were not there to hold and build the terrain consistent with our doctrine.
Our operations therefore served to disrupt, but not fully reverse, the Taliban’s influence and control of the population in Archi. Personnel and resource shortfalls precluded the credible commitment necessary to deter the Taliban’s return, or induce their defeat. All of this was a direct result of being stretched thin, a circumstance exacerbated by the overall drawdown of U.S. forces in Afghanistan that occurred midway through our deployment.
Although the Afghan Local Police (ALP) program was intended to fill security gaps created by thinly-stretched ISAF and ANSF, the ALP were composed primarily of Tajiks and Uzbeks, a circumstance that served to marginalize the province’s Pashtun population.
The Afghan Local Police (ALP) program was intended to fill security gaps created by overstretched ISAF and ANSF units. However, although the militias established under the program curbed Taliban influence, they also aggravated ethnic tensions in a diverse province, thereby generating unforeseen security challenges. Indeed, the ethnic composition of the ALP ultimately appears to have been extremely problematic. As RAND analyst Jason Campbell notes, “Communities in predominantly Pashtun districts, where resentment of the government was high, largely refused to take part [in the ALP program]. To fill the gap, ALP units dominated by Tajiks and Uzbeks were employed.” This erected a barrier between ALP units and the populations they were charged to secure. Significantly, the province’s recent history had severely aggravated ethnic divisions. A report produced by Cooperation for Peace and Unity, an Afghan-led nonprofit organization, articulates the context:
“Initially following the fall of the Taliban there was a wave of ethnic persecution of the Northern Pashtun groups. The new regime left Pashtuns marginalised…. As a result of these tensions at least 20,000 Pashtuns were forced south from the north in 2002 and there was growing political affiliation along ethnic lines (AFP 2002).”
Thus, although ISAF, ANSF, and to some extent the ALP were able to temporarily secure Kunduz Province, neither ISAF nor the Government of Afghanistan was able to effectively win over to their side the province’s Pashtun population, whose support was vital to long-run government control. In other words, the appearance of stability in Kunduz masked the need for a political resolution among the province’s ethnic constituencies and created vulnerabilities which the Taliban exploited in the lead up to their 2015 offensive.
In addition to its ethnic composition, the ALP demonstrated other problems, allegedly committing human rights abuses in Kunduz. The Afghan National Police (ANP) were complicit in a number of the documented ALP offenses, further eroding support for the provincial government.
The ALP’s ethnic composition was far from its only problem, as some ALP units in Kunduz not only imposed illegal taxes on civilians but also committed human rights abuses. A United Nations report published in August 2015 quotes one especially jarring example provided by an Afghan woman in Kunduz:
“The ALP commander came to me and told me ‘You bitch! You must leave the house. Leave the area or I will kill your entire family.’ He beat me with the butt of his gun then with a shovel. He beat me on my waist, head, and shoulder. I was crying and screaming, but in vain. I passed out from the beating, and regained consciousness after two hours.”
Here, as elsewhere, it is admittedly difficult to attribute the offense specifically to the ALP, rather than to one of the many unsanctioned local militias (Arbaki) in the province not (or no longer) sponsored by the Afghan government. In many instances, however, such distinctions may have been altogether irrelevant to local civilians, who in either case believed by September 2015 that the provincial government had amply demonstrated its inability to control the militias, ALP or otherwise.
Indeed, the ANP in Kunduz were already complicit in ALP abuses. A year earlier, in September 2014, Kunduz Province Chief of Police Mohammad Khalil Andarabi led an operation in Dashti Archi, with the intent of decisively purging the district of Taliban fighters. Andarabi employed ALP units in the mission (in addition to several nominally illegal militias), which reportedly devolved into an extended series of abuses. The Afghanistan Analysts Network, an independent non-profit research organization, reported:
“The militias looted houses, stole motorbikes, jewelry and money from both men and women, as well as taking wheat and sheep. Some people fled their houses. Wazir Khan, a resident of another village raided by the armed groups, said that members of the ALP took motorbikes, four vehicles, cash and numerous mobile phones.”
Thus, by September 2015, the ALP, although perhaps the “least undesirable” alternative to the Taliban, had significantly undermined the credibility of the provincial government—a circumstance on which the insurgency was undoubtedly able to capitalize.
In the months leading up to the seizure of Kunduz City, tensions emerged among provincial government leaders. These tensions were caused in part by the rapidity with which Afghan President Ashraf Ghani moved at the beginning of his term to effect structural reforms within the Afghan government.
In attempting to address structural problems with the Kunduz Provincial Government, including the significant issue of ethnic tension exacerbated by the ALP, Afghan President Ashraf Ghani may have moved too quickly. Prior to the fall of Kunduz City, there appears to have been significant political turmoil among provincial leaders. The origin of the tension was evidently a video conference that Ghani held with the provincial leadership weeks before taking office. At the conclusion of the meeting, the President abruptly announced the appointment of a new Provincial Governor.
Although it is unlikely that the sudden appointment of Governor Mohammad Omar Safi, who took over from predecessor Ghulam Sakhi Baghlani in December 2014, won Ghani many friends among provincial leaders, Safi himself was a loyal supporter, having backed Ghani during the recent presidential elections. That Safi was a political ally of Ghani, however, generated unease at both the provincial and national levels. Moreover, Ghani reportedly did not consult Afghanistan Chief Executive Abdullah Abdullah, his political opponent during the election, in making the decision to appoint Safi—a slight, real or perceived, that reverberated throughout the provincial government. As the Afghanistan Analysts Network notes, “These factors and Kunduz’s well-known factionalism led to infighting from day one, with provincial officials trying to remove each other from their positions.”
Given these factors, and viewed retroactively, it may seem odd that the Taliban’s seizure of Kunduz City in 2015 came as a surprise. Indeed, it is evident that Kunduz remains an important intermediate objective in the Taliban’s strategy for reclaiming Afghanistan, as the recent, renewed offensive on Kunduz City demonstrates. That the successful assault did come as a shock suggests that the assumptions regarding the trajectory of stability in Kunduz Province—both those of U.S. policymakers and warfighters, and of the Afghan government—were severely flawed. This should serve as a significant warning to the United States. Is it possible that Kabul, too, could one day fall?
Given the recent uptick in insurgent attacks on the Afghan capital and with the number of districts controlled or contested by the Taliban on the rise, the answer is self-evident: absolutely. Indeed, the situation in Kunduz leading up to the fall of the provincial capital in many ways mirrors the present state of the country as a whole. As in Kunduz circa 2015, ANSF retain control over major urban centers and key infrastructure, but struggle to pacify rural areas where support for the Taliban remains strong. Local militias and police presently—and in many cases tenuously and corruptly—secure many of those areas, leaving them vulnerable to the insurgency. Meanwhile, political uncertainty and instability are pervasive, and the rise of the Islamic State has once again rendered Afghanistan a strategic backburner to Iraq (and, of course, to Syria). Many warning signs, in other words, are already in place.
"Afghan Withdrawal: Ministers Hand Over Germany's Kunduz Base," Spiegel, October 7, 2013.
"Afghanistan Midyear Report 2015 - Protection of Civilians in Armed Conflict," UNHCR/UNAMA, August 1, 2015.
Azam Ahmed, "Ashraf Ghani Races to Make His Mark on Presidency in Afghanistan," The New York Times, October 21, 2014.
Azam Ahmed. "Taliban Are Rising Again in Afghanistan’s North." The New York Times, October 22, 2014.
Obaid Ali, "The 2015 Insurgency in the North (3): The Fall and Recapture of Kunduz," Afghanistan Analysts Network, October 16, 2015.
"ANSF Drive Taliban Away from Kunduz Districts," TOLO News, August 24, 2014.
Jason Campbell, "The Fall of Kunduz and What It Means for the Future of Afghanistan," War on the Rocks, October 6, 2015.
Lola Cecchinel, "The End of a Police Chief: Factional Rivalries and Pre-election Power Struggles in Kunduz," Afghanistan Analysts Network, January 31, 2014.
Lola Cecchinel, "Taleban Closing in on the City: The next round of the Tug-of-war over Kunduz," Afghanistan Analysts Network, September 2, 2014.
Lawrence Devlin, Jacob Rinck, Christian Dennys, and Idrees Zaman, "Conflict Analysis: Kunduz City, Kunduz Province," Cooperation for Peace and Unity, March 1, 2009.
“From Arbaki to Local Police: Today’s Challenges and Tomorrow’s Concerns,” Afghanistan Independent Human Rights Commission (AIHRC), Spring 2012.
“The Future of the Afghan Local Police,” International Crisis Group, June 4, 2015.
Naveed Hussain, "Kunduz Debacle: Time for Soul-searching," The Express Tribune, October 6, 2015.
"International Security Assistance Force - ISAF." GlobalSecurity.org. August, 2010.
Sayed Jawad, "Afghan House Speaker’s Father and Brother Die in Kunduz Blast," Khaama Press, March 13, 2013.
Thomas Joscelyn, “Taliban forces enter Kunduz city center,” The Long War Journal, October 3, 2016.
“‘Just Don’t Call it a Militia’: Impunity, Militias, and the ‘Afghan Local Police,’ Human Rights Watch, September, 2011.
David Kilcullen, Blood Year (New York: Oxford University Press, 2016), 175.
Mujib Mashal and Najim Rahim, “Afghan Forces Push Taliban Out of Kunduz Center, Officials Say,” The New York Times, October 4, 2016.
Bethany Matta, "Afghan Forces Fail to Halt Taliban Resurgence," Al Jazeera, August 29, 2014.
Bethany Matta, "The Failed Pilot Test: Kunduz’ Local Governance Crisis," Afghanistan Analysts Network, June 5, 2015.
Philipp Münch, "Local Afghan Power Structures and the International Military Intervention," Afghanistan Analysts Network, November 1, 2013.
Eltaf Najafizada, "Taliban Seize Key Afghan District After Assaulting Forces," Bloomberg, June 21, 2015.
Sudarsan Raghavan, "A Year of Taliban Gains Shows That 'We Haven't Delivered,' Top Afghan Official Says." The Washington Post, December 27, 2015.
Sangar Rahimi and Alissa Rubin, "Koran Burning in NATO Error Incites Afghans," The New York Times, February 21, 2012.
Bill Roggio, “Afghan Ministry of Interior admits 9 districts under Taliban control,” The Long War Journal, June 29, 2016.
Sayed Salahuddin and Pamela Constable, “Kabul increasingly a target as Taliban launches shocking attacks,” The Washington Post, September 7, 2016.
Nick Walsh, and Masoud Popalzai, "Protester Killed, U.S. Service Members Hurt in Unrest over Burned Qurans," CNN, February 26, 2012.
 David Kilcullen, Blood Year (New York: Oxford University Press, 2016), 175.