When the United States and its NATO allies launched military operations in Afghanistan in October 2001, the stated objectives were defined in unambiguous terms: dismantle the Al-Qaeda network, remove the Taliban regime that harboured it, and prevent the country from serving as a sanctuary for future attacks. Within weeks, those initial objectives appeared to have been achieved at remarkably low cost. The Taliban government collapsed, Al-Qaeda’s leadership dispersed, and a new political order was inaugurated at the Bonn Conference in December 2001. What followed, however, was not consolidation but an expanding and ultimately unresolvable entanglement – one that lasted two decades, consumed hundreds of billions of dollars, and ended in August 2021 with a collapse so swift it confounded even the most pessimistic assessments.
The trajectory of that entanglement can only be understood through the lens of its military dimension. The intervention began as a counterterrorism operation, evolved into a full-spectrum counterinsurgency campaign, and concluded as an increasingly desperate effort to build Afghan security forces capable of sustaining themselves after coalition withdrawal. At each transition – from counterterrorism (CT) to counterinsurgency (COIN), from COIN to Security Sector Assistance (SSA), and finally from SSA to drawdown – the military instrument was called upon to compensate for deficits that were structural rather than tactical: absent institutions, contested legitimacy, endemic corruption, and the chronic mismatch between the model of security forces being constructed and the social, economic, and administrative realities of the Afghan state.
NATO’s Comprehensive Approach (CA) envisioned precisely the kind of integration that might have addressed these structural conditions. In its canonical formulation, the CA posits that sustainable security cannot be achieved through military means alone – it requires the coordinated deployment of diplomatic, developmental, and governance instruments alongside military force, subordinated to political objectives and calibrated to the specific environment (NATO, 2024). The Bucharest Summit Declaration of 2008 committed the Alliance explicitly to this framework, acknowledging that “ the situation in Afghanistan will not be solved by military means alone” and that security, governance, and development efforts must be mutually reinforcing (NATO, 2008). The Lisbon Summit of 2010 reaffirmed this commitment as the Alliance prepared its transition strategy (NATO, 2010).
The gap between that doctrine and the operational reality in Afghanistan was not incidental. It was, as two decades of evaluations by the Special Inspector General for Afghanistan Reconstruction (SIGAR) have systematically documented, structural. Forces were built on Western models requiring levels of literacy, logistics capacity, and institutional continuity that Afghan society could not provide. COIN doctrine was applied without the governance foundations its own logic demanded. Civil instruments were consistently subordinated to military imperatives rather than integrated with them. And when political decisions – most consequentially the Doha Agreement of February 2020 – finally brought the military effort to its conclusion, the edifice constructed at such enormous cost collapsed in a matter of weeks.
This article examines the military dimension of the Comprehensive Approach in Afghanistan across three operational phases: the initial intervention and force construction period (2001-2005); the surge and SSA expansion (2006-2014); and the Resolute Support transition and ultimate collapse (2015-2021). It addresses a precise analytical question: to what extent were the principles of NATO’s Comprehensive Approach applied in the military domain of the Afghanistan intervention (2001-2021), as assessed by SIGAR evaluations? The central argument is that military failure in Afghanistan was not a product of tactical inadequacy but of structural misalignments – forces built on unsustainable models, a COIN doctrine applied without institutional foundations, and a civil-military relationship in which military imperatives consistently overwhelmed civilian instruments. These dysfunctions were not unforeseen; they were identified repeatedly and in detail by SIGAR across the duration of the intervention.
The article proceeds as follows. Section 2 establishes the theoretical framework regarding the Comprehensive Approach’s military dimensions. Section 3 outlines the qualitative methodology based on SIGAR data. Section 4 presents the chronological findings across the three operational phases, while Section 5 discusses the resulting cross-cutting structural tensions. Finally, Section 6 concludes with the implications for CA theory and practice in complex interventions.
The Comprehensive Approach (CA) acquired explicit doctrinal status with the 2008 Bucharest Summit Declaration, which committed the Alliance to deploying “a wide spectrum of civil and military instruments in a concerted effort” in addressing complex security challenges (NATO, 2008), a commitment consolidated at Lisbon in 2010 through the affirmation that sustainable security requires an approach that “effectively combines political, civilian and military crisis management instruments” (NATO, 2010). Within this architecture, the military instrument is conceived as one vector within an integrated effort – its function being to establish the security conditions under which governance and development can take place – rather than as the dominant element. CA assigns the military three interdependent functions: initial stabilization, Security Sector Assistance (SSA), and graduated disengagement as responsibility transfers. The literature identifies, however, an inherent tension: military capacity tends to fill the gaps left by fragile civilian institutions, thereby exceeding its mandate and substituting for them (NATO, 2024). The operationalisation of this tension is most visible in the SSA function, where the doctrine’s ambitions – building self-sustaining partner forces while preserving civilian primacy – are most directly tested against the realities of fragile-state intervention.
SSA operationalizes the capacity-building function of the Comprehensive Approach through the Train-Advise-Assist (TAA) triad: transmitting tactical skills, advising at the operational and institutional level, and providing direct mission support. A fundamental contradiction underlies the TAA model: systems built on sophisticated Western standards of equipment, logistics, and aviation generate structural dependency on external support that is incompatible with long-term autonomy – making sustainability a function of design choices made at the outset, not of subsequent effort. The effectiveness of SSA therefore depends on three structural conditions: a local partner with institutional legitimacy and political will, conditionality mechanisms linking assistance to governance standards, and a timeframe sufficient for genuine institutionalization rather than mere force expansion (SIGAR 17-62-LL, 2017). Where conditionality mechanisms are absent, the result is a culture of impunity in which local elites allocate resources on a patrimonial basis without accountability (SIGAR 16-58-LL, 2016a); readiness assessments then measure tangible inputs – personnel strength, equipment, training hours – while ignoring the determinative intangible factors of leadership, morale, and will to fight. The SSA model’s structural vulnerabilities do not, however, operate in isolation: they intersect with the doctrinal assumptions of the counterinsurgency framework within which capacity-building is typically embedded.
The counterinsurgency (COIN) doctrine centered on a “clear, hold, and build” policy. As Seth G. Jones explains, this approach aimed “to clear areas of Taliban presence, hold these areas securely, and build governance structures and infrastructure to encourage stability”, thereby pursuing the institutional reconstruction needed to eliminate the insurgency’s recruiting base (Jones, 2009). Its effectiveness rests on three assumptions that have proven structurally demanding in fragile-state contexts: a time horizon compatible with contributing states’ political cycles; a local partner government with genuine legitimacy and administrative capacity; and interagency coherence sufficient for tactical gains to translate into strategic stabilization. Emrah Ozdemir adds a further structural critique, noting that COIN “posits itself as a more inclusive… way of war, while it simultaneously constitutes the population as its battlespace” (Ozdemir, 2019) – a contradiction that inherently undermines mission legitimacy regardless of tactical execution. A fourth limiting assumption concerns the compatibility between the Western centralized state model and local political realities: where power is distributed among sub-state actors and legitimacy is locally rather than institutionally grounded, COIN’s governance-building logic rests on foundations that the intervention itself cannot supply. Each of these assumptions defines a condition that the Comprehensive Approach must satisfy through its civilian instruments – which places the question of civil-military coordination at the center of CA’s operational viability.
NATO doctrine prescribes horizontal coordination among diplomatic, development, and military instruments under a unified political framework, with civilian resources retaining autonomy from military logistics and military actors refraining from substituting for local state structures (NATO, 2024). The literature documents a recurrent inversion of this model in complex intervention contexts. Christopher Freeman argues that the Comprehensive Approach’s logic is vulnerable to systematic perversion, whereby “human security would be an outcome of military stabilization… rather than the actual aim” (Freeman, 2007) – reducing civilian objectives to by-products of military operations rather than independent strategic ends. This inversion has measurable structural consequences: Alessandro Monsutti documents that resource allocation patterns in post-conflict interventions consistently favour military channels, with reconstruction funding routed through defence departments at rates disproportionate to development agencies (Monsutti, 2012), subordinating the civilian agenda to military imperatives and replacing long-term development logic with short-term stabilization priorities. Coordination mechanisms conceived to bridge this gap – such as integrated civil-military teams or joint reconstruction structures – have in practice tended to reproduce military dominance rather than correct it, with civilian mandates encroached upon and European civilian efforts structurally constrained by resource asymmetry and national caveats (Hassan, 2023). The pattern identified in the literature thus points to a structural deficit internal to CA as an operational framework: the doctrine prescribes integration but contains no binding mechanisms to prevent military weight from overwhelming civilian instruments. Identifying the conditions under which this deficit becomes decisive requires a longitudinal evaluative source capable of tracking the interaction between military and civilian instruments across the full duration of an intervention.
The theoretical tensions identified across the preceding subsections – between CA’s prescriptive integration logic and the structural tendency toward military dominance, between SSA’s capacity-building ambitions and the sustainability constraints of the Western force-building model, and between COIN’s governance assumptions and the realities of fragile-state intervention – converge on a shared analytical gap: the absence of a systematic study examining the interaction among SSA performance, COIN application, and civil-military relations through the explicit lens of CA principles across the full duration of a single intervention. The academic literature on post-2001 Afghanistan, while extensive, relies heavily on anecdotal or journalistic data and lacks a longitudinal evaluative corpus capable of tracking structural patterns rather than isolated failures. The Special Inspector General for Afghanistan Reconstruction (SIGAR) addresses this gap directly: mandated for independent oversight of reconstruction expenditure, it generated over two decades an archive of quarterly reports, thematic assessments, and lessons learned publications covering all dimensions relevant here. SIGAR’s epistemic value lies in its access to classified information, its independence from the military chain of command, and the longitudinal consistency of its evaluations – enabling identification of the structural dysfunctions that discrete studies cannot capture. This justifies treating SIGAR reports as primary rather than secondary sources, and positions Afghanistan (2001-2021) as a case where the conditions for testing CA’s military dimension were present to an unusual degree – making the gap between declared doctrine and operational reality analytically consequential beyond the Afghan context alone.
This article employs a single-casequalitative research design grounded in systematic documentary analysis. Afghanistan (2001-2021) is treated as a critical case: the conditions for applying NATO’s Comprehensive Approach (CA) were present to an unusual degree – sustained coalition commitment, explicit doctrinal framing, two decades of institutional investment – and the gap between declared doctrine and operational reality was documented with exceptional longitudinal depth. The critical case logic holds that if CA’s military dimension failed here, despite the largest Security Sector Assistance (SSA) effort in post-Cold War history and explicit Alliance-level political commitment, the structural limitations identified carry analytical weight beyond the Afghan context.
The primary corpus consists of four major lessons learned and evaluation reports from the Special Inspector General for Afghanistan Reconstruction (SIGAR): Reconstructing the Afghan National Defense and Security Forces (SIGAR-17-62-LL), Corruption in Conflict (SIGAR 16-58-LL), What We Need to Learn (SIGAR 21-46-LL), and Why the Afghan Government Collapsed (SIGAR 23-05-IP). These are supplemented by two quarterly reports providing contemporaneous operational snapshots at analytically significant junctures: October 2016 (SIGAR Q3-2016, 2016b) and January 2020 (SIGAR Q4-2019, 2020). SIGAR reports are treated as primary sources on account of the institution’s statutory independence, access to classified operational data, and unmatched longitudinal scope. As SIGAR itself notes, while its oversight approach is generally technical – “we identify specific problems and offer specific solutions” – the cumulative body of work exposes a “staggering” list of systemic challenges that fundamentally compromised the reconstruction effort (SIGAR 21-46-LL). Secondary academic sources supply the theoretical frameworks structuring the analysis, providing interpretive perspectives on warlordism, civil-military imbalance, neopatrimonialism, and liberal peacebuilding that fall outside SIGAR’s institutional mandate.
The analysis applies thematic coding across five military dimensions: (1) Afghan National Defense and Security Forces (ANDSF) capacity and sustainability; (2) Counterinsurgency (COIN) doctrine versusoperational reality; (3) civil-military relationswithin the coalition effort; (4) operationalreadiness and logistical dependency; and (5) the impact of withdrawal sequencing onforce viability. These dimensions are appliedconsistently across three chronological phases– Phase I (2001-2005), Phase II (2006-2014), and Phase III (2015-2021) – enablingboth synchronic comparison and diachronictracking of how structural dysfunctionsevolved and compounded over time.
Four limitations bound the analysis. First, SIGAR’s accountability mandate is not identical to a strategic studies research lens, as it often evaluates progress through Western bureaucratic metrics. Second, reliance on public documents excludes classified data on intelligence coordination and special operations. Third, the single-case design does not support direct comparative generalization; rather, the findings are suggestive for the broader Comprehensive Approach literature. Finally, the three-phase periodization inevitably smooths over within-phase variation and uneven doctrinal change across different coalition members.
The American military intervention of October 2001 rapidly achieved its stated tactical objective – the dismantlement of the Taliban regime and the degradation of Al-Qaeda’s infrastructure – yet the operational choices that produced this swift victory generated structural dysfunctions whose effects proved irreversible across the full duration of the conflict. Driven by an aversion to nation-building, a strict focus on counter-terrorism, and a “light footprint approach” to the war (SIGAR 21-46-LL, 2021), the coalition transferred territorial control to warlord networks and local paramilitary forces whose predatory behavior fundamentally compromised the state-building project from the outset. Matthew P. Dearing documents how CIA operatives and U.S. special operations forces utilized these intermediaries as proxies. Instead of promoting stability, “ these clients failed to represent the liberal values foreign patrons endorsed, because the latter not only offered resources without conditions but also rewarded bad behavior” (Dearing, 2019). This dynamic institutionalized an ecosystem in which violence and resource capture were economically rational – a foundation completely incompatible with establishing a legitimate state monopoly on force.
The decision to deploy the Northern Alliance in the capture of Kabul carried political consequences that planners systematically disregarded. Edwards (2010) demonstrates that installing an ethnically unbalanced coalition, perceived by Pashtuns as an occupying power, produced a legitimacy fracture that directly fuelled the Taliban’s resurgence as a Pashtun resistance movement – converting immediate military success into long-term political defeat.
Construction of the Afghan National Army (ANA) began in 2002 on a Western-designed model that was maintenance-intensive and incompatible with Afghan societal realities. SIGAR retrospectively finds that the force was built to American standards – reliant on aviation assets, sophisticated logistics, and maintenance contractors – without assessing the Afghan state’s capacity to sustain these capabilities once external support was withdrawn (SIGAR 17-62-LL, 2017). This design decision constitutes the origin of the structural vulnerability that would determine the collapse of Afghan security forces in 2021. Military resources overwhelmed civilian instruments from the outset: Monsutti (2012) documents that approximately 60% of American reconstruction funding was channelled through the Department of Defense rather than development agencies, militarising the state-building process and undermining the coordinated approach the Comprehensive Approach (CA) prescribed.
The security vacuum produced by the limited military footprint outside Kabul left the provinces under warlord control, foreclosing durable state consolidation. As Hassan (2023) notes, the decision to restrict the initial International Security Assistance Force (ISAF) mandate to Kabul created a vacuum that neither subsequent Provincial Reconstruction Teams (PRTs) nor European missions were able to recover. The diversion of military and reconstruction resources toward Iraq from 2003 compounded this fragility, producing a strategic discontinuity that SIGAR identifies as a critical aggravating factor (SIGAR 21-46-LL, 2021).
The Taliban resurgence after 2005 and the Obama administration’s surge decision (2009-2010) marked a dramatic intensification of the military effort and Security Sector Assistance (SSA) programming. ISAF expanded, Afghan National Defense and Security Forces (ANDSF) budgets grew substantially, and structures such as The Combined Security Transition Command – Afghanistan (CSTC-A) and Train-Advise-Assist Commands (TAACs) proliferated. Capacity appeared to grow: personnel numbers increased, equipment accumulated, assessments reported progress. SIGAR systematically dismantles this illusion, demonstrating that measurement instruments captured tangible inputs – equipment, personnel strength, training hours – while ignoring decisive intangibles: leadership, will to fight, and institutional cohesion (SIGAR 17-62-LL, 2017).
Logistical corruption provides the most concrete illustration of the gap between appearances and operational reality. A 2013 SIGAR audit of the Afghan National Army’s petroleum, oil, and lubricants (POL) supply documented that $201 million in DOD fuel purchases remained “unaccounted for because records were shredded”, while the CSTC-A monitoring system operated on fundamentally incomplete data (SIGAR 13-4-AU, 2013). Funds intended to increase ANDSF operational independence were instead feeding corruption networks that siphoned fuel before it reached units, directly degrading combat readiness. As SIGAR documents, this was systemic rather than localized: contractors supplying American and NATO bases routinely paid protection fees to Taliban insurgents for safe passage. By tolerating an ecosystem where insurgents were paid to refrain from attacking convoys, the coalition effectively made the enemy “unofficial subcontractors to the U.S. government”, directly subsidizing the very adversary the mission was mandated to defeat (SIGAR 16-58-LL, 2016a).
The COIN doctrine applied during this phase encountered the structural limitations anticipated in the academic literature. The “clear, hold, build” principle required three simultaneous conditions Afghanistan could not provide: sufficient military presence to sustain territorial gains, a functioning civilian administrative apparatus to assume control of secured areas, and a rural population persuaded of central government legitimacy. As Seth G. Jones notes, while the strategy “saw some success, especially in urban centers”, it ultimately “clashed with the on-ground realities of limited institutional capacity” and competing power structures in rural areas where the Taliban enjoyed local support (Jones, 2009). Furthermore, the operational execution of COIN was systematically undermined by rapid, six-to-twelve-month personnel rotations. These short tours of duty fragmented relational continuity and local knowledge, as SIGAR evaluations highlight, the “constant turnover of U.S. and NATO trainers impaired the training mission’s institutional memory and hindered the relationship building required” for effective security sector assistance (SIGAR 17-62-LL, 2017). Ultimately, these personnel policies systematically undermined institutional knowledge transfer, subjecting the broader reconstruction effort to what SIGAR described as an “annual lobotomy” as departing staff took critical information with them (SIGAR 21-46-LL, 2021).
Civil-military relations remained structurally imbalanced despite the proliferation of coordination mechanisms. Provincial Reconstruction Teams (PRTs) functioned in practice under military dominance, characterized by a massive mismatch in resources; at the height of the surge, nearly 100,000 U.S. troops were deployed compared to just under 600 USAID personnel (SIGAR 21-46-LL, 2021). Consequently, civilian personnel were routinely subordinated to operational priorities, with USAID officials reporting they were often “bulldozed” by the military into implementing projects in insecure environments to serve tactical ends rather than long-term stability (SIGAR 21-46-LL, 2021). European civilian efforts shared a similar fate; the EU Police Mission (EUPOL) was structurally “dwarfed by USA’s military and reconstruction spending” and ultimately had little impact due to its limited capacity to operate in unsecured environments (Hassan, 2023).
Alessandro Monsutti confirms this dynamic, noting that the subordination of USAID to the Department of Defense perverted development assistance into a tactical instrument, allowing military operations to “increasingly impinge on the traditional space of bilateral development agencies” (Monsutti, 2012). Reflecting on this militarized approach, SIGAR retrospectively concluded that without genuine Afghan institutional reform and meaningful conditionality, the massive military escalation was fundamentally flawed, noting it would have been better “never to have surged at all” (SIGAR 23-05-IP, 2022). At the tactical level, the proliferation of Afghan Special Security Forces (ASSF) as a response to conventional unit weakness produced a further dysfunction. Because the regular Afghan National Army (ANA) lacked structural capacity, elite forces were subjected to severe misuse – such as manning static checkpoints – which overextended the ASSF and merely masked the ANA’s profound institutional incapacity (SIGAR Q3-2016, 2016b).
The transition to the Resolute Support Mission (RSM) in 2015 marked the formal reduction of the coalition’s combat presence and the transfer of security responsibility to Afghan forces – without any of the structural vulnerabilities documented in preceding phases having been resolved. The ANDSF remained dependent on external financing, on American contractors for aviation maintenance, and on Close Air Support (CAS) for major offensive operations. The coalition drawdown eroded not only tangible capabilities but critical intangibles: troop morale, popular perceptions of state legitimacy, and the will to fight among field commanders (SIGAR 23-05-IP, 2022).
The Doha Agreement of February 2020 constituted the strategic breaking point. Concluded exclusively between the United States and the Taliban without Afghan government participation, the agreement conveyed a devastating signal: the external patron was implicitly recognising the adversary as a legitimate actor and announcing a conditional withdrawal rather than a path to victory. The 2020 SIGAR quarterly report documents that military assistance had shifted to “point-of-need advising” – a fragmented approach structurally incapable of substituting for the institutional support framework being simultaneously withdrawn (SIGAR Q4-2019, 2020). The operational paradox of this period lies in the simultaneity of two contradictory processes: the planned doubling of Afghan Special Security Forces and the signing of an agreement that announced their imminent abandonment.
SIGAR 23-05 identifies the proximate cause of the August 2021 collapse as the inability of Afghan political leaders to internalise that the United States would actually leave – a failure that also reflects the fundamental breakdown of civil-military relations across the entire intervention. The Afghan government had been excluded from major strategic decisions and treated less as a sovereign partner than as an object of American foreign policy (SIGAR 23-05-IP, 2022). The hyper-centralised command structure, shaped by American preferences for singular interlocutors, left field commanders isolated and without autonomous resources at the moment of the final Taliban offensive.
The collapse of the Afghan Air Force (AAF) illustrates the structural mechanism most clearly. Built as a central pillar of ANDSF capability, Afghan military aviation depended entirely on American contractors for maintenance. Their post-Doha withdrawal rapidly degraded aircraft availability, eliminating the capability that had allowed ANDSF to compensate for numerical inferiority against Taliban advances (SIGAR 23-05-IP, 2022) – the predictable outcome of a force-building model that had systematically prioritised immediate performance over sustainability.
A cross-phase reading reveals a persistent pattern: at each doctrinal transition, the military instrument compensated for structural deficits that CA was designed to address through genuine inter-instrument coordination. Afghan institutional incapacity was treated as an operational variable to be managed through additional military resources rather than a political problem to be resolved. Corruption was tolerated as a transactional cost of partnership rather than confronted as a strategic threat. Civilian instruments – USAID, EUPOL, diplomatic assistance – operated systematically in the shadow of the military, underfunded and subordinated to operational timelines. These structural dysfunctions, documented longitudinally by SIGAR across two decades of evaluation, confirm that the failure resided not in tactics but in the architecture of the intervention as a whole (SIGAR 21-46-LL, 2021).
The analysis of the three interventionphases reveals a pattern that transcends discrete failures and constitutes a systemic structural dysfunction: at each critical juncture, the military instrument compensated for deficits that the Comprehensive Approach prescribed to be resolved through inter-agency coordination. This recurring substitution was not incidental – it reflected a fundamental asymmetry of capacity and mandate between coalition components that CA acknowledged theoretically but never corrected operationally.
The most consistent cross-phase pattern is that each military response to a structural vulnerability reproduced, at a greater scale, the very vulnerability it purported to address. In Phase I, the “light footprint” doctrine avoided the institutional commitment required for state-building by delegating local security to warlords. As Dearing observes, this approach created “paramilitaries [that] were influenced by material, social, and normative incentives that rewarded violent and predatory behavior”, thereby establishing predatory power structures that proved impossible to dismantle (Dearing, 2019). In Phase II, the surge injected massive resources to accelerate Afghan National Defense and Security Forces (ANDSF) capacity, but the absence of meaningful conditionality transformed those resources into fuel for corruption. Official audits confirm that this lack of oversight ended up “fueling the insurgency by financing insurgent groups and reinforcing grievances”, indirectly subsidizing the very adversary the coalition intended to defeat (SIGAR 16-58-LL, 2016a). In Phase III, American logistical and air support merely concealed the structural non-viability of Afghan forces. Consequently, the sudden withdrawal of this support instantly deactivated the capacity it had artificially sustained for a decade, leaving the Afghan state “fundamentally unprepared to manage the fight against the Taliban as the United States military and its contractors withdrew” (SIGAR 23-05-IP, 2023). Ultimately, each operational solution deepened the structural dependency it claimed to treat.
A second cross-sectional pattern concerns how corruption was managed throughout the intervention. Official evaluations indicate that decision-makers consistently perceived a trade-off between combating corruption and achieving short-term security objectives, routinely prioritizing the latter. As SIGAR explicitly concluded, “security and political goals consistently trumped strong anticorruption actions”, reflecting a pragmatic but flawed willingness to partner with malign actors to pursue counterterrorism imperatives (SIGAR 16-58-LL, 2016a). This framing treated corruption as an acceptable transactional cost rather than a lethal strategic threat. The consequence was that neopatrimonial networks consolidated at every phase: the warlords co-opted in Phase I adapted to become the political power brokers of Phase II. Malejacq describes this resilience as an ability to “shape-shift” and constantly reinvent themselves to capture international resources (Malejacq, 2016). Consequently, the rentier oligarchy they formed came to control both military logistics and civilian administration, effectively creating “an exclusive oligarchy devoted to its own enrichment and closely tied to the international coalition” (Rangelov & Theros, 2012). Viewed through this lens, the spectacular collapse of the ANDSF is not merely a failure of military capacity but a failure of political ecology. As Weeda Mehran observes, within Afghanistan’s neopatrimonial system, patronage networks “have been relatively effective in delivering services to those within the network” (Mehran, 2018). Therefore, Afghan security forces behaved rationally within the incentive structure of a predatory state, not irrationally against the Weberian standards the coalition claimed to be building toward.
The most analytically significant deficit concerns the internal structure of the Comprehensive Approach as an operational instrument. While the Bucharest (2008) and Lisbon (2010) summit declarations affirmed the principle of integrated civil-military coordination, analysis across all three phases shows that no binding implementation mechanism existed to prevent military dominance. Official evaluations explicitly confirm that the coalition “lacked a comprehensive approach… and a coordinating body to successfully implement the whole-of-government programs necessary” (SIGAR 17-62-LL, 2017). Consequently, Provincial Reconstruction Teams (PRTs) functioned in practice under military control, the EU Police Mission (EUPOL) operated with minimal resources and was structurally “dwarfed by USA ’s military and reconstruction spending” (Hassan, 2023), and USAID was routinely subordinated to the tactical requirements of the Department of Defense. As SIGAR noted, the massive resource mismatch meant that “civilian agencies simply could not compete with DOD’s resources and planning in the provinces”, resulting in the military making “critical decisions that should have been made by U.S. civilian officials with expertise in navigating complex political dynamics” (SIGAR 21-46-LL, 2021). The conclusion is not that CA failed due to discrete poor decisions, but that the doctrine lacked structural mechanisms for balancing military weight against civilian instruments – making failure predictable rather than contingent.
A fourth pattern concerns the inadequacy of the Western Security Sector Assistance (SSA) model as applied in the Afghan context. As official evaluations document, the Afghan National Defense and Security Forces were designed to Western standards of equipment, logistics, and air technology that the host nation could not independently sustain. SIGAR explicitly concluded that “providing advanced Western weapons and management systems to a largely illiterate and uneducated force without appropriate training and institutional infrastructure created long-term dependencies, required increased U.S. fiscal support, and extended sustainability timelines” (SIGAR-17-62-LL, 2017). This represented a fundamental design error rather than a mere implementation failure: the model optimized for short-term performance, relying heavily “on tangible outputs, such as staffing, equipping, and training levels”, which fatally “ masked intangible factors, such as corruption and will to fight” (SIGAR-17-62-LL, 2017). By prioritizing measurable tactical outputs over long-term sustainability conditions, the coalition built a force that could not function without its foreign patrons.
The paradox of simultaneously expanding the Afghan Special Security Forces while negotiating the Doha Agreement captures this contradiction in symbolic form. While the Department of Defense reported that ASSF elements were “on track to double in size by the end of 2020” to improve combat performance (SIGAR Q4-2019, 2020), the concurrent political decision to negotiate a withdrawal directly with the insurgency “ undermined the morale of the Afghan security forces” (SIGAR 23-05-IP, 2023). Investment in elite military capacity thus continued mechanically even after the political exit strategy had already rendered it strategically void.
The findings suggest that the Comprehensive Approach’s failure in Afghanistan was not the flawed implementation of a sound model, but an intrinsic limitation of the model under conditions of radical partner asymmetry. CA presupposes applicability conditions that its doctrine leaves unspecified: a host-state with minimally viable institutions, a coalition willing to subordinate tactical interests to integrated strategic objectives, and a sufficient time horizon for institutional reform to take hold. As official evaluations demonstrate, the latter condition was structurally violated, since “ the painstaking work of rebuilding institutions was simply never compatible with the urgency with which the U.S. government perpetually operated in Afghanistan” (SIGAR 21-46-LL, 2021). Ultimately, Afghanistan met none of these conditions at any phase.
This study’s empirical contribution to the CA literature – engaging debates on the militarization of civilian instruments raised by Christopher Freeman and the quantitative subordination of development aid to military budgets documented by Alessandro Monsutti – lies in demonstrating that the absence of these applicability conditions does not produce a mere implementation failure. Instead, it transforms CA into a rhetorical framework that legitimizes military dominance under the appearance of integrated coordination. Future research may test whether this pattern – the systematic substitution of coordination by military dominance – recurs across other dimensions of interventions in fragile states or in other NATO missions operating under a declared CA mandate, contributing to a broader reformulation of the doctrine’s conditions of applicability.
This article addressed a preciseanalytical question: to what extent were the principles of NATO’s Comprehensive Approach applied in the military domain of the Afghanistan intervention between 2001 and 2021, as assessed by SIGAR evaluations? The answer the analysis supports is unambiguous. CA remained, in the military domain, an unimplemented doctrinal aspiration. The gap between doctrine and operational reality was not an isolated failure of execution but a persistent structural dysfunction, documented across two decades of independent evaluations.
The article’s central argument – that military failure in Afghanistan was not a product of tactical inadequacy but of structural misalignments – is confirmed by the empirical record across all three phases. Phase I established flawed foundations: the light-footprint doctrine and the co-optation of warlords without conditionality produced a predatory client-state rather than a legitimate partner, invalidating the central premise of SSA from the outset. Phase II amplified the illusion of capability: the surge produced tactically measurable gains without reforming the intangible factors essential to sustainability, while logistical corruption indirectly financed the insurgency the coalition was fighting. Phase III demonstrated that the structural dependency built in the preceding phases was irreversible: the withdrawal of contractors disabled ANDSF, and the Doha Agreement delegitimised the Afghan government before its forces could operate independently.
The analysis refines the central argument in one important respect. The failure was not the aggregate of isolated poor decisions but the product of a systemic pattern in which each military solution adopted to address a structural vulnerability reproduced that vulnerability at greater scale. Corruption was not an accidentally tolerated dysfunction – it was the transactional cost of short-termtactical security, paid systematically and predictably. Civil-military imbalance did not deteriorate gradually through contingentfactors – it was structurally embedded from the mission’s inception, because CA doctrinecontained no mechanisms to counterbalancemilitary dominance over civilian instruments.
The article’s principal theoretical contribution lies in reframing the terms on which CA failure must be assessed. The conclusion is not that CA is a flawed doctrine, but that it presupposes conditions of applicability that the Afghanistan intervention never satisfied: a local partner with functional legitimacy, a genuinely operationalised interagency coordination architecture, and a sufficient time horizon for capability investments to reach sustainable autonomy. The absence of any one of these conditions transforms CA from a framework for genuine integration into an instrument for the discursive legitimation of military dominance. At the practical level, the analysis offers a longitudinal documentary framework – grounded in SIGAR reporting as an independent evaluative source – that can inform NATO doctrinal revisions and the design of future SSA missions.
The study’s limitations are methodological and documentary. Single-case design constrains the direct generalisability of findings to other intervention contexts. SIGAR’s mandate covers predominantly operational and financial dimensions of reconstruction, leaving coalition decision-making dynamics and Afghan perspectives on the processes under analysis insufficiently documented. Classified operational data relevant to assessing force readiness remain beyond the reach of academic research.
Several directions for future research follow from these findings. The pattern identified here – military solutions that structurally reproduce the vulnerability they address – warrants testing across other dimensions of the Afghanistan intervention and in comparable NATO-led missions, to determine whether the conditions of CA applicability identified in this study reflect context-specific constraints or a broader structural limitation of the doctrine. SIGAR’s longitudinal evaluative corpus stands in this respect as a rare methodological model: few major interventions have produced independent documentation of comparable systematicity and accessibility, making the Afghan case an enduring reference point in the CA literature.