The belief that development and reconstruction activities are central to stability and security is by no means novel. The need for ‘integrated’ approaches or ‘coherence’ in post-conflict environments had been largely acknowledged by humanitarian and military actors alike, particularly in the aftermath of Rwanda and other humanitarian crises of the 1990s (Collinson and Elhawary 2012). However, ‘stabilisation’ in foreign policy, military strategy, and development aid assumed significantly greater prominence after the events of 9/11. Such approaches were highly contentious, perhaps nowhere more so than in Afghanistan where troop contributing nations (TCNs) to the International Security Assistance Force (ISAF) sought to utilise development and reconstruction activities to undermine the Taliban and enhance the legitimacy of the Afghan government. The stabilisation approaches employed by international forces and their governments in Afghanistan have become a model for many Western countries and for NATO, yet aid agencies’ experiences in dialogue with these forces and in the context of stabilisation have often been deeply negative, affecting the way they operate and interact with military forces globally.

This article summarises research, conducted in Kabul and elsewhere between October 2012 and January 2013, on dialogue between aid agencies and military forces in Afghanistan from 2002 through 2012. Through the Afghanistan case, this study seeks to better understand the challenges of civil-military dialogue – dialogue between military forces and independent humanitarian actors – in the context of combined international and national military forces pursuing stabilisation.

Methodology and Terminology

This analysis is based on an extensive desk review of literature on Afghanistan, with a focus on the military strategy and humanitarian operations conducted in the country between 2002 and 2012. Initial consultations were also conducted with experts currently or formerly engaged in civil-military coordination structures to refine the focus of the study and identify individuals for qualitative interviews. This was supplemented by research conducted in Afghanistan as well as phone interviews with a wide range of current and former military and civilian officials from a diverse set of ISAF troop-contributing countries, donors, diplomats, UN officials, aid workers, Afghan government officials, and analysts. Approximately 24 aid workers (predominantly internationals), 14 military actors, 10 donor representatives, and six independent experts with direct knowledge or experience of civil-military interaction in Afghanistan were interviewed between October 2012 and February 2013. It should however be noted that to some extent these distinctions are arbitrary; several individuals had experience in multiple categories (for example working with aid agencies or donors as well as on Provincial Reconstruction Teams or with military forces). In total, 54 individuals were interviewed. Although these conversations were guided by a semi-structured questionnaire, often the content of discussions was adapted to match the key areas of expertise and personal experiences of the interviewee.

While several studies have focused on specific provinces, Provincial Reconstruction Teams (PRTs), or specific aspects of civil-military relationships such as governance or protection, this article aims to provide an overview of civil-military relations during the course of a decade. This consequently limited the level of detail that could be covered, which in turn was complicated by the fragmented nature of civil-military cooperation efforts, the variety of stabilisation approaches pursued at national/provincial/district level, and their complex evolution over time. In order to gain greater understanding of the evolution of civil-military dialogue over time, the narrative of this article follows a chronological approach.

For the purposes of this article, the expression ‘aid agencies’ refers to both humanitarian and multi-mandate (humanitarian and development) not-for-profit organisations. These agencies, including the UN, Red Cross/Red Crescent, and international and national NGOs, follow recognised humanitarian principles. This includes the principles of humanity (saving human lives and alleviating suffering wherever it is found), impartiality (taking action solely on the basis of need, without discrimination between or among affected populations), and independence (autonomy from the political, economic, military, or other objectives that any actor or party to a conflict may harbour with regard to the areas where humanitarian actors are working). Some humanitarian and multi-mandate organisations, though not all, will be guided by neutrality (not favouring any side in a conflict or other dispute). The authors defined ‘civil-military coordination’, in line with the UN Office for the Coordination of Humanitarian Affairs (OCHA)/Interagency Standing Committee (IASC) Guidance (2008), as the essential dialogue and interaction between civilian and military actors in humanitarian emergencies that is necessary to protect and promote humanitarian principles, avoid competition, minimise inconsistency, and, when appropriate, pursue common goals.

The Origins of Stabilisation in Afghanistan

The concept of stabilisation is neither well defined nor consistently understood; it has assumed a number of forms and names over time (pacification, stabilisation, peace-support operations, or reconstruction) (Mac Ginty 2012). Nonetheless, its broad objectives and approaches have remained largely the same (Barakat et al 2010). At the core of stabilisation theory is the assumption that conflict and weak governance pose a threat to international peace and stability. A related assumption is that such conflicts are fuelled by underlying grievances towards the state, driven by state neglect and poverty – the corollary being that development projects, particularly those that improve service delivery and offer economic opportunities and improved governance, can ‘stabilise’ conflict situations. These components are both short-term – with the immediacy required to prevent a return to violence – and long-term – to address the root causes of a conflict through an overall improvement of the quality of governance. Development assistance is meant to enhance public approval of the government, thus ‘buying time that serves to reduce the chances of the state slipping back into violence’, while longer-term goals comprise ‘a potentially transformative, comprehensive, and long-term project, possibly entailing substantial social, political, and economic change’ (Gordon 2010: S369; Collinson et al 2010: S277).

After 9/11, stabilisation assumed greater prominence – and, at times, near-fanatical support – among and across Western governments, transforming the approach and structure of both military and civilian agencies. Stabilisation discourses influenced military doctrine and foreign aid and fundamentally changed the relationship between the military and civilian components of these governments. Many governments adopted integrated civilian–military or ‘whole of government’ approaches to dealing with so-called ‘fragile states’, supported by stabilisation units within the military or civilian foreign assistance departments. The result in conflict-affected states, not least Afghanistan, was that civilian assistance became inextricably linked with – and often guided by – political and military objectives.

The two largest troop-contributing countries to ISAF, the US and the UK, both embraced the concept, albeit in different ways. Stabilisation ideas both fed into the early designs of the mission in Afghanistan and developed in response to those countries’ experiences there. For the US, stabilisation has always remained within the remit of the Department of Defense (DoD) and has not been integrated in the same structural way with the US Agency for International Development (USAID) or the US Department of State. In terms of doctrinal development, the US Army and Marine Corps’ Counterinsurgency Field Manual 3–24 identified stabilisation operations as one of the three components of a COIN strategy (offence, defence, stabilisation). The Army’s Stability Operations Field Manual 3–07 developed a definition of stabilisation as ‘the process by which underlying tensions that might lead to resurgence in violence and a breakdown in law and order are managed and reduced, while efforts are made to support preconditions for successful long-term development’ (Department of the Army 2008: 1–12). In the UK the Post-Conflict Reconstruction Unit was developed in 2004 (later renamed the Stabilisation Unit in 2007). The Stabilisation Unit reports to the Ministry of Defence, Foreign and Commonwealth Office, and Department for International Development, and is staffed by all three branches of government. There have also been changes in DoD doctrine, such as the development of stabilisation ideas in military counterinsurgency strategy through military Joint Doctrine Publications (JDPs) 3–40, 3–50, and 3–52 (Gordon 2010: 370). Many other troop contributing nations (TCNs) have also set up specialised ‘stabilisation units’; for example, Canada established the Stabilisation and Reconstruction Taskforce (START) within the Department of Foreign Affairs and International Trade (DFAIT) and Australia created the Crisis Prevention, Stabilisation and Recovery Group within the Australian Agency for International Development (AusAID).

Stabilisation in the Early Years of the International Intervention

On 7 October 2001, coalition troops were deployed to Afghanistan under the US-led Operation Enduring Freedom (OEF).1 The following December, following the rapid collapse of the Taliban government, a number of prominent Afghans2 met under UN auspices in Bonn to form an interim government, the Afghan Interim Authority (AIA), to be followed after its six-month mandate expired by a two-year Transitional Authority (TA). The Bonn Agreement recommended the deployment of a UN-mandated international force to maintain security. On 22 December 2001, UN Security Council Resolution 1386 authorised the creation and deployment of ISAF under Chapter VII of the UN Charter to ‘assist the Afghan Interim Authority in the maintenance of security in Kabul and its surrounding areas’. The first ISAF troops were deployed in June 2002 and, while initially operating under the rotating command of coalition member states, ISAF was placed under the control of NATO in August 2003.3

After the fall of the Taliban regime, diplomatic and aid agency presence, which was limited under the Taliban government, dramatically increased. On 28 March 2002, UN Security Council Resolution 1401 established the UN Assistance Mission in Afghanistan (UNAMA) to support ‘focused recovery and reconstruction’. UNAMA pursued a ‘light footprint’ approach with an initially limited presence outside of Kabul while Afghan and international aid agencies expanded operations throughout the country. By 2006, there were over 800 aid agencies operating in Afghanistan (Olson 2006). Even so, aid expenditure was comparatively low – less than 10% of what it had been in post-war Bosnia and less than a quarter of what was devoted to post-conflict East Timor (Waldman 2008). The military presence too was limited, with 9,700 troops deployed at the end of 2002 in a country with an estimated population of 28 million.

In the early years of the international intervention in Afghanistan (2002–2008), the primary instruments of stabilisation were Provincial Reconstruction Teams (PRTs). Though predominantly military in composition at the outset, PRTs were defined by ISAF as ‘a civil–military institution’ (ISAF 2007: 5). PRTs were intended to fill an interim role until the nascent Afghan government could provide services and security itself. The origins of the PRT model can be traced back to US-led Coalition Humanitarian Liaison Cells (CHLCs) (Borders 2004). These entities, comprising 10–12 troops, were tasked with something between intelligence gathering and the implementation of quick impact projects (QIPs). Although they were conceived as being institutions with significant civilian leadership/composition, PRTs generally comprised 50–100 predominantly military personnel with many PRTs being solely military in composition at the outset.

An ‘ink-spot strategy’ was envisioned whereby PRTs would support the extension of security and government authority by establishing small zones of stability that would then spread, eventually joining up with other stabilised areas until the entire region was secured. In addition to providing security, PRTs were initially tasked with coordinating reconstruction, including conducting assessments, identifying potential projects, and coordinating NGOs, UN actors, the Afghan Transitional Authority, and others (Stapleton 2003).4 PRTs were intended as a cost-effective alternative to fully-fledged nation building, which would have required greater military presence and financial resources.

After a generic concept was formalised, PRTs were established by the US and TCNs at a rapid pace. All led by the US, the first PRTs were set up in Paktia, Kunduz, Bamiyan, and Balkh by 2003, and in Parwan, Herat, Nangarhar, and Kandahar by early 2004. The Kunduz PRT was later transferred to Germany and a German-led PRT was subsequently established in Badakshan. By the end of 2004, there were a total of 19 PRTs in Afghanistan and by the end of 2008 there were 26 PRTs led by 13 TCNs.

While some PRTs were initially appreciated by many Afghans, a large share of interventions, including QIPs and other activities focused on ‘winning hearts and minds’, lacked the requisite planning and involvement from communities and Afghan institutions to ensure appropriateness and sustainability. Part of the problem was that many PRT military staff lacked the skills for ‘development’: for most there was no PRT-specific pre-deployment training, nor was there any form of systematic handover or debriefing for PRT commanders, with the problem being exacerbated by frequent rotations resulting in little ability to develop any institutional memory. PRT structure and activities also varied widely across lead nations, with little effective coordination among themselves or with aid agencies, reflecting the variations in local security conditions and in the amount of financial resources available to each PRT. The result was what a US government report described as ‘a wide variety of entities with the same name’ and ‘no clear definition of the PRT mission, no concept of operations or doctrine, no standard operating procedures’ (US House of Representatives 2008: 18).

The structure and activities of individual PRTs were shaped most directly by the local operating environment. In the north, German-controlled PRTs had little insecurity to contend with during this period. The UK by contrast encountered significant challenges when it moved from the north and assumed command of the PRT in Helmand province, forcing the British to adopt a dramatically different approach to deal with much greater levels of insecurity (Gordon 2010). Lead nations were also inhibited by their own bureaucratic restrictions and legal constraints, including ‘national caveats’ that restricted particular forces from specific security-related functions without explicit approval from their governments. For example, German troops were prohibited from staying outside of their camps overnight and so could not carry out long-range patrols. All patrols had to be accompanied by an armoured ambulance and German aircrafts had to return to their bases before nightfall (Merz 2007). Despite pressure from NATO, the German government refused to lift these restrictions.

The size and structure of early PRTs varied widely, although most had a relatively small civilian component – estimated to be on average around 5–10% of the total staff of a typical PRT in 2004 (Save the Children 2004). German PRTs ranged up to 300 troops with a limited number of civilians, while early US PRTs averaged 100 military staff and around five civilians. In theory, all US PRTs were supposed to include State Department and USAID representatives, yet in many PRTs these posts were not filled. There were only around 40 USAID employees in Afghanistan in 2003, most of them located at the embassy in Kabul (Gall 2003).

To some extent, these staffing structures reflected the models of stabilisation each lead nation employed. Overwhelmingly military PRTs, like those of the US, were highly militarised and short-termist in their approach. US PRTs initially focused on QIPs designed to win hearts and minds. Civil Affairs teams hired private contractors to execute the work, which included the construction of schools, clinics, wells, and other small projects intended to establish good relations with Afghans and collect intelligence. Support to governance translated into promoting the authority of those local power holders who were perceived to be favouring the government, frequently former warlords or militia commanders of questionable loyalty.

Where PRTs were multinational – as in the case of the Uruzgan PRT led by the Netherlands between 2006 and 2011 with significant Australian support – differences in ideology and approach became acutely problematic. The Dutch approach, seen by many of those interviewed for this study as one of the most effective PRT models5, focused on addressing local grievances and conflicts. Researchers and cultural advisors were employed to map the origins and dynamics of local conflicts and help develop stabilisation approaches, with limited use of force and only on the basis of concrete intelligence. In many instances aid was discreetly given to individuals or initiatives seen as critical to security, including tribal elders disgruntled with the government, with little oversight and minimal visibility. The thinking was that such aid was useful in providing security, but only if it could not be traced back to the Dutch for fear that this would undermine the legitimacy of the local actors whose power they were seeking to bolster.

By contrast, the Australians implemented much more visible stabilisation projects, often in the wake of or in the same location as Australian combat action. Australian stabilisation projects during this period, implemented jointly by the military and AusAID, were low-budget (around US$10,000), short-term QIPs aimed at securing support for the action of Australian forces. Many Australian military staff reportedly felt that the Dutch approach was too ‘soft’ and ‘politically naïve’, and believed that they were only able to pursue such an approach thanks to the combat operations conducted by US and Australian Special Forces (Fishstein 2012: 8).

The lack of adequate training and support, combined with a high level of staff turnover, made dialogue between PRTs and aid agencies difficult. Effective coordination was rare and often strained, with the role and objectives of PRTs being unclear to many aid workers. At a meeting in Herat in 2004, the PRT commander, addressing a meeting with NGOs stated, ‘You don’t need to love us, you just need to work with us.’ One frustrated aid worker responded, ‘You only have another eight months here and yet you want to tackle long-term development issues such as unemployment.... You will only work where other NGOs cannot and yet you are working where we are all working. What do you actually see as your added advantage here – in relation to what the NGO community is already providing?’ (Save the Children 2004: 24–25).

While Afghans initially appreciated some PRTs, they were ultimately an inadequate solution to the insecurity that spread throughout the country after 2001. Between 2002 and 2006, insurgent attacks increased by 400% and casualties by 800% (Jones 2008). Security deteriorated significantly from 2006 onwards; in that year alone, bomb attacks nearly doubled on the previous year, suicide attacks increased six-fold, and over 1,000 civilians were killed or injured (Human Rights Watch 2008). The dual role of PRTs became increasingly schizophrenic, as did their attempts to win hearts and minds. In one incident in Ghazni province in 2004, PRT officials attempted to offer condolences to villagers and offered to dig a well weeks after they had fired rockets into the village, killing nine children. Villagers reportedly responded with anger and confusion, later telling researchers ‘we want them to leave – we don’t want their help…let them keep their well’ (Save the Children 2004: 24).

Aid Agencies: Objections and Strategies for Coordination

Despite pressure to support their activities, PRTs from the very beginning were criticised both by the humanitarian and development community in Afghanistan, and by the Afghan government. A primary objection was that PRTs, and the broader stabilisation approaches of which they were a part, aimed to militarise and politicise assistance by aligning aid with stabilisation objectives rather than addressing the needs of affected people (Stapleton 2003). As a result, aid was grossly skewed towards insecure provinces or provinces where troops were present. In 2006, the Lithuanian PRT in Ghor province spent approximately US$462,000 on development projects, while the Italian PRT in neighbouring Herat spent US$4.5m (Abbaszadeh et al 2008).

Another concern was that tasking the military with delivering aid would create confusion among insurgents and civilians, blurring the lines between aid actors and the military. These arguments centred on the principle of distinction: many were concerned that an inability to differentiate between military and civilian aid actors would have dangerous consequences for the safety of aid workers on the ground.6 The Geneva Conventions stipulate the responsibilities of occupying powers (a category that would have arguably applied to international forces in Afghanistan, at least in the early stages of the conflict)7 and their obligations towards civilian populations, including ensuring access to food and medical supplies as well as sufficient health and hygiene conditions. Nonetheless, many worried that the delivery of aid with the goal of gathering intelligence or fostering loyalty to pro-government forces would force civilians to make an impossible choice between badly needed assistance and their own safety.

Aid actors’ efforts to coordinate locally with PRTs showed mixed results, and their attempts to limit the role of PRTs in hearts and minds activities were largely unsuccessful. While most objected to PRTs doing ‘development’, there was no common opinion on precisely what PRTs should do. Some felt it would be impractical to demand that PRTs abandon reconstruction work altogether, while others felt any military involvement in reconstruction or development was unacceptable. Additionally, many aid agencies directly or indirectly supported stabilisation and some international and local agencies accepted funding directly from PRTs to implement projects. More commonly, agencies accepted funding from the donor agencies of PRT lead nations to work in provinces where PRTs were present and in sectors that were identified as being integral to consolidating military gains, including in ‘target’ districts identified on the basis of security and military concerns. Each agency appears to have dealt with the perceived lack of independence and impartiality deriving from cooperation with PRTs in its own way. Some felt compromises were acceptable in ‘peaceful’ provinces but not where international forces were heavily engaged in combat. Others accepted funding from donor governments involved in the conflict but refused to utilise it in provinces where their troops were present.

These problems were exacerbated by insufficient capacity, leadership, and coordination on civil-military matters among aid agencies. UNOCHA (UN Office for the Coordination of Humanitarian Affairs), present in Afghanistan since 1991, closed its office in 2003 and humanitarian affairs were subsumed under an integrated mission, UNAMA. Even within the Humanitarian Affairs Unit of UNAMA there reportedly was only one international staff member responsible for civil-military affairs for the majority of the unit’s existence. Nonetheless, UNAMA tried to resolve coordination issues with PRTs and establish greater clarity on their role.

Despite the challenges, there were productive engagements with the military during this period. These included the joint civil-military PRT Executive Steering Committee (ESC), which provided policy guidance on PRT operations, and the Civil-Military Working Group (CMWG). The PRT ESC, established in 2004, met only a handful of times before ceasing to exist altogether around 2007, but it did issue several policy notes to guide PRT policy and interaction with external actors with regards to disarmament, development activities, and coordination with humanitarian actors. PRT Policy Note 3 of 2009, for example, states that the use of ‘foreign military and civil defence assets in disaster relief activities must be in extremis circumstances only, to be utilised as a last resort requested by either the Government of Afghanistan or the United Nations Humanitarian Coordinator (HC), and in accordance with the Afghan specific Civil–Military Guidelines’. However, the almost non-existent coordination of PRTs by ISAF made uniform policy change a formidable challenge, and it is unclear how or to what extent ESC policy notes reached PRT commanders. Without effective dissemination to commanders on the ground, these notes would have been largely irrelevant.

The CMWG sought to build on the work of the PRT ESC by establishing Afghanistan-specific Civil Military (Civil Military Working Group 2008), endorsed in 2008 by the UN Humanitarian Coordinator, the UN Humanitarian Country Team, the Agency Coordinating Body for Afghan Relief (ACBAR), and ISAF. The Guidelines sought to adapt internationally recognised principles to the unique challenges aid agencies faced in an operating environment dominated by concerns over PRT activity and growing insecurity. They also sought to curb military practices that could lead to confusion between civilian and military actors. According to many interviewees who were involved in the process, the adaptation of the Guidelines to the context of Afghanistan with the agreement of ISAF were key outcomes in and of themselves. Yet the Guidelines faced opposition, largely from within the aid community, and were not sufficiently disseminated to military actors.

The CMWG subsequently declined, being disbanded altogether by 2011. Following the approval of the guidelines, the CMWG appeared to turn its focus to issues of minimal consequence to the affected populations. One example was the use of unmarked white ‘civilian’ vehicles by military personnel, which the CMWG campaigned against. Aid actors contended that white vehicles in particular were associated with NGOs and the UN, and that their use by the military had led to a number of cases where civilian convoys had been mistakenly attacked by insurgents (Cornish and Glad 2008). Although rarely invoked in arguments or public statements on the issue, IHL states that combatants must distinguish their vehicles as military, not civilian. Interviews with military actors confirmed that white vehicles were essentially used as force protection, allowing troops to travel in lower-profile vehicles that were believed to be less at risk of attack. Aid agencies succeeded in obtaining a concession from ISAF in 2009 dictating that TCNs distinguish their vehicles. However, the order did not specify precisely how this should be done or whether they should be distinguished as specifically military vehicles at all. Shortly after the directive was issued, one interviewee reported seeing Afghan support personnel applying a single brown stripe to the sides of white vehicles at an ISAF base in Kabul.

Diverging – and at times diametrically opposed – objectives were pursued by aid actors and the military through the CMWG. The military often saw the CMWG as a means of gaining information and cooperation from aid agencies or as an opportunity to present their narrative of the conflict. But the more the military pursued its goals, the more aid agencies pulled away. Despite the fact that agencies did not have a unified stance on this matter, outspoken advocacy by several agencies portrayed this as a cynical attempt to co-opt aid agencies into a military strategy and as an assault on humanitarian principles.

Aid agencies sought to use the CMWG to argue against the core tenets of the military strategy – entailing, for instance, the engagement of military actors in development-like activities – that they were unlikely to be able to change in any fundamental way, rather than as a means for discussing and communicating issues around violations of IHL or calling upon principles of civil–military coordination. An additional challenge was the lack of understanding of such principles among aid agency staff. Likewise, ISAF often sent personnel to the CMWG who had little influence or decision-making power; attendance was reportedly limited to civil–military coordination (CIMIC) staff (CJ9) rather than staff from ISAF’s strategy and planning unit (CJ5) (BAAG/ENNA 2008). Over time, the CMWG deteriorated to such an extent that both aid agencies and military officials stopped attending the meetings.

As dialogue was breaking down, concerns around the protection of civilians grew. Security sharply deteriorated in 2006 and the number of civilian casualties increased; Human Rights Watch (2008) reported that 929 civilians died as a result of the conflict that year, a quarter of the deaths attributable to international forces. In 2007, the UNAMA Human Rights Unit, operating under the auspices of the Office of the High Commissioner for Human Rights (OHCHR), in cooperation with the Afghan Independent Human Rights Commission (AIHRC), began to comprehensively investigate and record incidents where civilians were harmed. As the Taliban extended their control throughout the south and east, and subsequently into some western, northern, and central provinces, attacks on aid agencies increased. By the end of 2007 the UN considered nearly half of the country’s districts too dangerous for UN personnel to access directly (Meo 2007).

Civil-Military Dialogue and the ‘Surge’

To address growing insecurity, the US authorised a troop ‘surge’ in 2009, nearly doubling its force presence in Afghanistan. In addition to major ‘clearing operations’, the new military strategy focused on counterinsurgency (COIN), supplemented with greater funding and numbers of civilians deployed from TCNs to support these efforts. More than ever before, the military strategy focused on winning hearts and minds, solidifying the very approach aid actors objected to.

In August 2009, ISAF Commander General Stanley McChrystal issued new COIN guidance. The core assumption of this approach was that military victory could only be achieved ‘by persuading the population, not by destroying the enemy’ (ISAF 2009: 1); in other words, Bolstering effective governance and services, thus improving civilian support for the government, was more critical than military engagement with the insurgency. The new COIN guidance dictated that troops should ‘embrace the people’ and ‘leverage economic incentives and routine jirgas with community leaders to employ young men and develop peaceful means to resolve outstanding issues’ (ISAF 2009: 4).

Other manuals and directives further elaborated this strategy. The US Army’s ‘Commanders’ Guide to Money as a Weapons System’ defined aid as ‘a nonlethal weapon’ to be utilised to ‘win the hearts and minds of the indigenous population to facilitate defeating the insurgents’ (US Army Combined Arms Center 2009). The amount of aid devoted to these objectives rapidly increased: annual CERP funding rose from US$200m in 2007 to US$1bn in 2010 (SIGAR 2012a). So did the structures comprising US PRTs, including District Support Teams (DSTs), civilian-led joint civil–military teams comprising State Department, US Department of Agriculture, and USAID officials, and US National Guard Agri-Business Development Teams (ADTs)8 (USAID 2010).

The number of civilian officials deployed to support military-led governance and development efforts also increased. Following an interagency review, the US unveiled a new integrated military and civilian approach in March 2009 and a civilian ‘surge’ or ‘uplift’ that nearly tripled the presence of US civilian officials. Deployed across eight US government agencies in Afghanistan, the number of US civilian officials increased from 261 in January 2009 to 989 by February 2011, reaching 1,040 by June 2011 – at a cost of US$2bn between 2009 and 2011 (SIGAR 2011; 2012b). Other countries, to lesser degrees, followed suit. The UK, for example, increased development aid to Afghanistan by 40% in 2010 and the number of Department for International Development (DFID) staff in the country increased from 41 in 2007 to 75 in 2012 (International Development Committee 2012).

As part of a ‘clear-hold-build’ approach, derived in part from British strategy against communist insurgents in Malaya in the 1950s and adapted by the US in Iraq, civilian officials were meant to assist in the ‘build’ phase by supporting governance and service provision once areas had been ‘cleared’ of insurgents by the military. Other TCNs similarly revised their strategies. In 2009, the UK shifted its strategy to focus on a governance-led approach, with the intention of ‘stabilising’ Helmand ‘through containment of the military threat posed by the Taliban while convincing the Helmand population that there would be an enduring Afghan government presence that was increasingly responsive to its needs and concerns’ (Gordon 2010: S375). Other countries also announced more comprehensive and integrated military, diplomatic, and assistance strategies, and many allocated greater proportions of their aid budgets to areas where their troops were stationed.

As troop presence expanded, insecurity intensified and spread through previously stable provinces. Agency operating space eroded: access to large parts of the south and east, and portions of the west, was nearly impossible for many international agencies. There was increased pressure on aid agencies to support development and governance aspects of the military strategy, which aid agencies often saw as little more than ‘battlefield clean up’, and thus generally refusing to take part in it. The appetite for dialogue rapidly diminished as many agencies avoided interaction with the military.

Aid agencies increasingly sought distance from UNAMA, once critical in coordinating civil military dialogue, feeling that its close association with ISAF and the Afghan government undermined perceptions of their neutrality and independence. Upon request by Afghan and international NGOs who felt humanitarian coordination needed to be distinct from UNAMA, OCHA was re-established in Afghanistan in 2009.9 Initially, however, OCHA suffered from many of the same challenges of limited staffing and capacity that UNAMA did and struggled to perform its civil-military functions.

The new military approach also posed significant risks for aid workers and those they aimed to help, drawing civilians further into the conflict. As ISAF attempted to implement COIN, focused on service delivery and engaging local populations, attacks on anyone suspected of supporting these efforts (i.e. for-profit contractors, Afghans working for the government, or ISAF) increased. There is also strong evidence that insurgents increasingly came to see aid agencies as being associated with the military effort (Jackson and Giustozzi 2008; Glad 2009).

While there is no inherent contradiction between the COIN doctrine espoused by ISAF and IHL and other principles underpinning civil–military coordination, the ways in which the military strategy was implemented – and the consequences for civilians – were heavily criticised by aid agencies. There are also strong indications that the military strategy further undermined in practice whatever respect was left for IHL or civil–military guidelines. According to one military official serving in Kandahar at the time, ‘humanitarian and civmil guidance weren’t of use, it was overridden by COIN’. The appetite for dialogue rapidly diminished as many aid agencies sought to avoid the military, either to limit the perception of association or simply because many felt that any discussion would ultimately be futile. In 2010, the Afghanistan NGO Safety Office (ANSO) advised NGOs against engaging in civil–military coordination, warning them that they had ‘nothing to gain and much to lose from interacting with IMF [International Military Forces] who are only interested in leveraging advantage from your activities’ (ANSO 2010).

Outside Kabul, the increased troop presence made dialogue between the military and aid actors more complicated and less effective. PRTs were now part of an increasingly complex array of official military and civilian organisations contracted to carry out development activities. Poor coordination among ISAF TCNs and the varying approaches and philosophies pursued by PRT lead nations became acutely problematic. For civilians or aid workers attempting to establish dialogue or resolve problems (for example, trying to ascertain the status of staff members detained by military forces), identifying the appropriate interlocutor was increasingly difficult. There were some efforts to ensure a basic level of co-existence between military and civilian actors. ACBAR organised a small NGO–ISAF contact group with the primary goal of keeping lines of communication open between aid agency directors and senior ISAF commanders, as well as providing a forum to resolve instances of violations of civil–military guidelines. The group, which comprised a handful of directors from international (primarily US-based) aid agencies, met on a monthly basis. However, it was disbanded after the directorship of ACBAR changed hands and the new director felt that such close relations with ISAF were no longer desirable.

With increasing civilian casualties, the most successful civil-military engagement focused on civilian protection, and more specifically on the reduction of civilian casualties attributed to ISAF. While there had been significant dialogue on these issues in response to growing numbers of civilian casualties and increasing anger among Afghans regarding the issue, the adoption of COIN and its rhetoric of ‘protecting the population’ allowed for the creation of a new opportunity for aid actors and human rights advocates to engage on these issues, while the link between ISAF and OEF under a single chain of command, COM-ISAF,10 helped to streamline dialogue. In contrast to other civil–military dialogue efforts during this period, UNAMA Human Rights/OHCHR showed significant leadership and was particularly active on these issues. Many of the actors involved, including UNAMA Human Rights/OHCHR, but also human rights NGOs such as Human Rights Watch and the Center for Civilians in Conflict, adopted an approach that can be best described as ‘strategic argumentation’ – appealing to key tenets of COIN and shared concerns over civilian harm, alongside international law. Evidence and data was critical in persuading military officials to adopt tighter controls on the use of force, as was cultivating relationships with key military officials at various levels. Investigations of civilian harm routinely conducted by UNAMA Human Rights/OHCHR and AIHRC, as well as publicly available bi-annual UNAMA Human Rights/OHCHR reports from 2008 onwards, helped to exert pressure on military forces and increased accountability and transparency.

From 2008 onwards, and particularly after 2009, ISAF tightened its rules of engagement, introduced new Tactical Directives, and reinforced COIN guidance restricting the use of force and underscoring the importance of avoiding civilian harm (often referred to as ‘courageous restraint’). How much of this was due to advocacy or dialogue and how much may have occurred naturally as a consequence of the adoption of COIN is unclear. Nonetheless, ISAF was responsible for 316 civilians deaths in 2012 – down from 828 in 2008.11 Airstrikes accounted for nearly two-thirds of all civilian deaths in Afghanistan in 2008. However, following a Tactical Directive issued by the ISAF Commander in 2009, in large part the result of sustained lobbying and advocacy, airstrikes dropped dramatically; by 2012 they accounted for 4% of all civilian deaths. ISAF also introduced systems to improve accountability and oversight, including an internal civilian casualty-tracking cell.

The Legacy of Civil-Military Relations

Experiences in Afghanistan show a conflict – if not on a theoretical level at least on a practical one – between stabilisation and internationally recognised guidelines and principles of civil-military interaction that aim to safeguard IHL and humanitarian space. In situations where the military aggressively seeks to co-opt civilians, lack of adherence to these principles is likely to be even more extreme.

Aid agencies also have an obligation to adhere to their own principles if they would like to see them respected and to ensure that their actions do not actively undermine them. In Afghanistan, some aid agencies prioritised presence and funding over principles, or appeared to assume that Afghanistan’s post-Taliban recovery from conflict would be relatively straightforward. According to Soren Jessen-Petersen (2011: 4), former Assistant Commissioner of UNHCR, ‘some humanitarian organisations worry more about being present and visible in a major operation than the reasons why they should be there – in other words, to provide impartial and independent protection and assistance to the victims of conflict’.

Many rationalised their choices or sought to mitigate damage by, for example, limiting their direct contact with the military. It is unclear if this was sufficient. Working in geographic areas determined by TCN political/military interests led to an association of aid agencies with one side of the conflict. Even where agencies insist that such programmes were based on need, and while many genuinely benefitted Afghans, aid agencies knowingly furthered the political and military objectives of pro-government forces. This undermined their ability to advocate for truly neutral and impartial assistance from donors and for adherence to the guiding principles of civil-military interaction with the military.

Important lessons can also be drawn about the role of UN actors. UNAMA Human Rights/OHCHR cultivated relationships with key stakeholders, received significant amounts of high-level UN support, and developed a neutral position focused on the impact of the conflict on civilians. Substantial evidence was used to bring about policy change. Others were markedly less successful in influencing change, at least during the surge period. In the case of OCHA (and its predecessor, the UNAMA Humanitarian Coordination Unit), lack of support from senior UN officials, lack of capacity in terms of staffing and systems, and an unclear role posed formidable challenges to effective change. The effect of the integration between OCHA and UNAMA is debatable. Regardless of the structure of the mission, strong in-country humanitarian leadership was arguably not possible without consistent principled leadership from above and a genuine, complementary prioritisation of humanitarian concerns within UNAMA.

It appears that advocating against PRTs, stabilisation, or COIN was largely ineffective. Didactic arguments based on the perceived rights and special status of aid agencies were also largely ineffective, and often resulted in military actors becoming frustrated. By contrast, where dialogue was rooted in IHL and strategic argumentation – as was the case with advocacy efforts focused on civilian harm which appealed to a shared interest between civilian and military actors to reduce that harm – it was markedly more persuasive. However, such engagement is complex and time-consuming, requiring a significant level of capacity that many aid agency staff simply did not have.

The lack of unity among aid agencies and the lack of a clear unified humanitarian voice further undermined efforts at effective dialogue. Part of this, predictably, arose from competition for resources and competing agendas and from the diverse mandates and objectives of aid actors. While difficult to achieve, a unified and sustained aid agency position would have undoubtedly been more effective in engaging the military compared to the ad hoc and contradictory initiatives that often prevailed in Afghanistan. There were also significant tensions or differences in approach between some international actors and Afghan aid agencies.

NATO and TCN governments also have much to learn from experiences in Afghanistan. Implementing development interventions in areas of conflict in a partial manner, with the goal of furthering the chances of one side’s military victory and with the involvement of armed forces, is not only dangerous for every actor involved but often proves to be self-defeating. There is little evidence that NATO, TCNs, or their donor agencies have critically examined the danger posed by these strategies or learned any lessons from the largely negative experience of stabilisation in Afghanistan.

While it would be tempting to recommend that TCN governments and donor agencies conduct assessment exercises with regard to stabilisation and PRT experiences, such activities would be unlikely to have much impact. Seeking to acquire greater evidence on effectiveness and risks would only be useful insofar as such policy decisions are based on objective evidence. In Afghanistan and other stabilisation contexts, the role of evidence in policymaking and programme design appears minimal. Nonetheless, more objective evidence on the impact of stabilisation is required to gather a greater understanding of the risks and limitations involved – even if such evidence is unlikely to be generated by donor governments themselves.

Questions remain about what will happen to PRT assets and military-led interventions after the drawdown of international combat forces. The long-running problems of insufficient technical capacity, Afghan government involvement, and oversight associated with military-led assistance complicate any ‘handover’ to Afghan institutions. Given the poor quality or short-term nature of many of these projects, it is unclear what will be handed over at all. Regarding future dialogue with Afghan forces, it is important to note that in contrast to ISAF, arguably one of the strongest and most sophisticated fighting forces in the world, the Afghan security forces are nascent. Although they have made progress in recent years, they continue to struggle with basic issues around command and control, resources, and effectiveness. The Afghan government and the country’s military forces have been largely absent from dialogue on civil-military issues until recently, which will inevitably render any attempt at establishing effective working relations with key individuals and institutions in the country rather challenging. While it is relatively clear that Afghan forces are unlikely to pursue the same kind of militarised aid activities that are currently performed by international actors in the country, their capacity and willingness to engage in dialogue remains unclear.