Abstract
North Atlantic Treaty Organization’s (NATO’s) active cooperation with non-governmental organisations (NGOs) on the protection of civilians (PoC) challenges the academic assumption that NATO’s decision-making structures are generally closed to civil society participation. What explains the decision of NATO officials to work with NGOs when developing a Policy for the PoC, and how did this cooperation take place? Combining insights from resource dependency and organisational learning theories, we explore whether and how concerns about public legitimacy and operational effectiveness can explain that NATO engaged with NGOs during its learning process on PoC. Methodologically, we combine data from 26 interviews with NATO officials and NGO representatives with a document analysis of 648 NATO publications. Our analysis shows that NATO’s collaboration with NGOs in the adoption of a PoC policy was very much a bottom-up process. Military commanders began engaging with NGOs on the ground during NATO’s International Security Assistance Force operation in Afghanistan (2001–2014). NGOs then gradually gained access to the formulation and institutionalisation of the PoC policy. While legitimacy concerns triggered NATO’s initial cooperation with NGOs, effectiveness concerns related to policy implementation facilitated stronger cooperation once the process was underway.
Keywords
Introduction
In recent years, North Atlantic Treaty Organization (NATO) has closely cooperated with non-governmental organisations (NGOs) on the protection of civilians (PoC). During NATO’s International Security Assistance Force (ISAF) in Afghanistan (2001–2014), NGOs’ extensive regional and linguistic expertise proved critical to ISAF commanders. As early as 2008, monthly briefings were held between ISAF military staff, the United Nations Assistance Mission in Afghanistan and NGOs – such as Human Rights Watch and the Afghanistan Independent Human Rights Commission – to discuss new tactical directives and methods for civilian casualty tracking. Interestingly, in addition to this on-the-ground cooperation, NATO (2020a) also increasingly engages with NGOs at the political level in the development of its PoC standards, including through ad hoc workshops and trainings, as well as participation in military exercises. For instance, since 2018, a biannual NATO-UN PoC training is organised in collaboration with the Finnish Defence Forces International Centre, the Dutch NGO PAX for Peace, and the International Committee of the Red Cross (ICRC; National Defence University, 2021).
While there is a robust literature on cooperation between intergovernmental organisations (IGOs) and NGOs, NATO’s cooperation with NGOs on PoC is somewhat surprising because it extends beyond the more commonly studied realm of on-the-ground cooperation. NATO has actively engaged with (mainly humanitarian) NGOs in operations in Afghanistan, Bosnia-Herzegovina and Kosovo (Gheciu, 2011; Halterman and Irvine, 2014; Lischer, 2007; Williams, 2011, 2014). Yet, at the political level, prominent studies have found that NATO’s strong intergovernmental decision-making procedures generally remain more closed to input from NGOs (Mayer, 2014; Steffek, 2013). A recent study by Wright (2023) confirms the idea that NATO was not the most welcoming IGO for civil society and NGOs on its Women, Peace and Security Agenda. This suggests that we need more comprehensive research on why and at which levels (operational, strategic, political) NATO engages with NGOs.
In explaining why IGOs open up to NGOs, the literature often refers to resource dependency theory, according to which IGOs have resource needs that can only be fulfilled by external actors. Yet, this does not explain the process by which IGOs become aware of their resource insufficiencies and come to integrate NGOs into their policymaking. To examine this process, we combine insights from resource dependency and organisational learning theories and ask: What explains the decision of NATO officials to work with NGOs when developing a PoC policy, and how did this cooperation take place?
As a result of the cooperation between NATO and NGOs, as of 2016 – with the adoption of its Policy for the PoC – NATO (2016b) defines the PoC as: [. . .] all efforts taken to avoid, minimize and mitigate the negative effects that might arise from NATO and NATO-led military operations on the civilian population and, when applicable, to protect civilians from conflict-related physical violence or threats of physical violence by other actors, including through the establishment of a safe and secure environment.
1
We expect that NATO officials’ decision to work with NGOs for a PoC policy was motivated by both ‘push’ and ‘pull’ factors: not only do we expect NGOs to push for access to the process, but we also expect that NATO’s concerns about public legitimacy and effectiveness acted as ‘pull’ factors to collaborate with NGOs. To examine NATO’s motivations for including NGOs in the process, we follow NATO’s (2022) own classification of its lessons learned processes, examining stages of analysis, implementation and institutionalisation.
Empirically, we build on data from 26 interviews with mid- and senior-level NATO officials and NGO representatives, and an analysis of NATO policy documents and secondary sources. Our analysis shows that NATO’s collaboration with NGOs in the adoption of a PoC policy was very much a bottom-up process. NATO military officials first started to cooperate with NGOs on PoC issues during the ISAF operation in Afghanistan. This cooperation was driven by operational effectiveness and public legitimacy concerns (‘analysis’ stage). Gradually, NATO civilian officers also started to engage in discussions with NGOs to effectively formulate NATO’s PoC policy (‘implementation’ stage). As of 2017, this cooperation focused on institutionalisation of the PoC policy within NATO through collaboration with NGOs in NATO trainings and exercises that include elements of PoC (‘institutionalisation’ stage).
Our findings contribute to broader discussions about IGO–NGO cooperation, as well as policies for the PoC in conflict. They demonstrate that NATO, as a politico-military defence alliance, is not so different from other IGOs in terms of how it cooperates with NGOs. Our findings also echo previous work (Tallberg et al., 2018) that showed how the added value for IGOs of involving NGOs in their policymaking goes beyond increasing perceptions of public legitimacy; NGOs can also be included to increase the effective implementation of specific policies, as shown by the PoC policy implementation process.
IGO–NGO cooperation in global governance studies
Understanding cooperation (and competition) between state, non-state and transnational actors has become a central research objective for scholars in the field of global governance studies (Avant et al., 2021 [2010]; Eilstrup-Sangiovanni and Westerwinter, 2022; Tallberg et al., 2013; Westerwinter et al., 2021). NGOs are increasingly central actors in this global governance complexity (Lipschutz, 1992; Martens, 2005; Reimann, 2006; Steffek, 2013). Yet, the population of NGOs is diverse and mandates vary widely, from service provision to advocacy and lobbying (Steffek, 2013; Williams, 2014). This article focuses on the relations between IGOs and those NGOs that seek to influence their policymaking. NGO involvement in IGO policymaking is generally seen as a positive development, as it increases the democratic legitimacy of IGOs’ decisions and actions (Sithole, 2016; Steffek, 2013; Tallberg, 2010).
Particularly the United Nations has attracted scholarly attention because of its institutionalised cooperation with NGOs on human rights, women’s rights, environmental, humanitarian, peace and development issues (Donini, 1995; Joachim, 2003; Kimber, 2023; Martens, 2005, 2006; Narain, 2012). Other IGOs, such as the Organisation for the Security and Co-operation in Europe (OSCE) and the Council of Europe, have also opened up to input from NGOs (Agné et al., 2015; Mayer, 2008). NATO has received much less scholarly attention when it comes to its relationship with NGOs, largely because of its long-term reluctance to show openness to NGO involvement (Mayer, 2008).
In the IGO–NGO literature, IGO member states are often expected to be reluctant to involve NGOs in intergovernmental decision-making processes for fear of losing power in the negotiations (Steffek, 2013). NGOs’ own actions are often expected to be more important in determining their access to the IGO (Steffek, 2013). Researchers have argued that NATO does not provide NGOs with systematic access to the political headquarters and does not have an NGO accreditation procedure (Mayer, 2008; Williams, 2014). At the same time, however, NATO has engaged with NGOs and the ICRC at the operational level (Gheciu, 2011; Halterman and Irvine, 2014; Lischer, 2007; McQueen, 2005; Williams, 2011, 2014). For instance, NGOs and the ICRC played a major role in NATO’s creation of ‘safe areas’ in Bosnia in the 1990s (McQueen, 2005). The creation of Provincial Reconstruction Teams as part of NATO’s comprehensive approach in Afghanistan is another example of how on-the-ground interaction with NGOs took place (Williams, 2011). Moreover, receptiveness to cooperation on the ground works both ways: in Afghanistan, NATO’s commitment to non-military activities created friction with humanitarian NGOs who did not want to be perceived by the Afghan population as having close ties to the military (Williams, 2011).
In contrast to NATO’s cooperation with NGOs at the operational level, cooperation at the political level appears to be a more recent phenomenon and has only recently received some scholarly attention. Particularly notable is the work of Wright (2023), who showed the transformative role of NGOs in NATO’s agenda for Women Peace and Security (WPS). This suggests that NATO is increasingly opening to civil society in policymaking. Yet, Wright focused predominantly on how civil society actors perceived and experienced this relationship. We do not know when or why NATO itself becomes receptive to input from civil society in its policymaking. In particular, the receptiveness of NATO military staff is worth exploring, as the civil–military relations literature suggests that interaction between civil society and military institutions can be challenging because of their different organisation and objectives (Allen et al., 2021; Gheciu, 2011; Ruffa and Vennesson, 2014). For NATO specifically, it has been argued that its ‘complex organisational structure, with civil-military cooperation at its core’, can both benefit and hamper policy change (Von Hlatky, 2018: 365).
In sum, we know that NATO, as a politico-military alliance, has a strong intergovernmental decision-making process and a strong military component, which can hinder engagement with civil society. Yet, we also see that in its development of a PoC policy, NATO has gradually opened up for input from NGOs. Why and how this cooperation emerged remains unexplained.
Common explanations for IGO–NGO cooperation
Several explanations have been put forward to explain why and when IGOs and NGOs collaborate. The orchestration literature offers an IGO-centric explanation for the phenomenon of IGO–NGO cooperation: IGOs enlist and use intermediary actors in pursuit of the IGO’s goals, for example, by using NGOs as enforcers of international law rather than trying to govern actors themselves (Abbott et al., 2015). IGOs may engage in orchestration to make up for a lack of resources, whether ‘regulatory competence, operational capacity or legitimating authority’ (Abbott et al., 2021: 152). As such, they recruit NGOs to act as allies and carry out governance work on behalf of IGOs in various fields (Tallberg, 2015).
Other scholars focus on the strategic behaviour of NGOs, examining when and how NGOs shape political processes and policies within IGOs. For instance, some NGOs decide to join forces with other civil society organisations to participate in specific policy areas, as has been the case with the promotion of women’s rights in the United Nations and NATO (Shepherd, 2015; Wright, 2023). Competition may also lead NGOs to strategically differentiate their actions to ensure their survival. This can lead to a divide between (larger, older) NGOs that tend towards more conventional strategies, and smaller NGOs engaging in less costly, more confrontational strategies (Eilstrup-Sangiovanni and Sharman, 2022).
Finally, resource dependency approaches are widely used to explain why IGOs cooperate with other international actors, combining IGO- and NGO-centric factors. The assumption is that all organisations depend on external resources for their functioning and survival (Pfeffer and Salancik, 1978). When focusing on IGO–NGO relations, resource dependency highlights both ‘push’ factors from NGOs, and ‘pull’ factors from IGOs at different stages of the policy cycle. On the ‘push’ side, NGOs want access to IGO policymaking because it allows them to influence policy issues or, in certain cases, to receive funding that is essential to their survival. On the ‘pull’ side, NGOs can provide scarce resources that the IGO needs to produce or implement its policies (Mayer, 2008; Steffek, 2013).
Two main types of ‘scarce’ resources have been identified in literature. First, IGOs are more likely to open up to third parties when they lack the resources to effectively achieve their goals (Harsch, 2015). NGOs can provide knowledge or expertise through ‘information politics’, relying on their international networks and country-specific knowledge (Hill et al., 2013; Keck and Sikkink, 1998; Tallberg et al., 2018). Such functionalist, goal-based arguments have found support in studies of why IGOs such as the World Bank open up to NGOs (Kelly, 2011). Second, IGOs are also more likely to open up to external input when they are concerned about their organisational reputation or public legitimacy. Engaging with NGOs may bring the work of IGOs closer to the citizens, provide these IGOs with legitimacy and create an image of accountable and transparent governance (Dellmuth and Tallberg, 2015; Grigorescu, 2007; Scholte, 2004; Steffek, 2013; Steffek et al., 2008; Tallberg, 2010).
However, this knowledge about the resources NGOs can bring does not tell us much about the process by which IGOs ‘discover’ these resource gaps, and how they then open up to these external actors. To grasp this process, we apply insights from the organisational learning literature. Differentiating between various learning stages allows us to not only show why, but also how NATO opened up to cooperating with NGOs on its PoC policy.
Organisational learning in NATO
Organisational learning studies draw heavily on the seminal work of Levitt and March (1988), who have shown that organisations apply routines that are based on and adapted to their past experiences to achieve their goals. The capacity to accumulate, record, store and interpret knowledge is assumed to be essential for organisations to perform and to learn. However, organisational learning and change, and therefore the decision to open up to stakeholders, take time and are often incremental processes. This is especially true for IGOs with strong member state oversight and many veto points (Dijkstra et al., 2019). The success of a learning process often depends on the political commitment and resources allocated to the organisation’s learning process, and on signals from the external environment (Bossong, 2013).
Organisational learning insights have also been applied to understand learning in IGOs with military mandates, including NATO and the European Union (EU). For instance, Michael E. Smith (2017, 2018) adopted insights from organisational learning to explain the evolution of the EU’s Common Security and Defence Policy (CSDP) as well as its capacity and the obstacles to learn from CSDP missions and operations. Also, Heidi Hardt (2017) showed the informal dynamics of learning within NATO, and particularly how the International Staff creates and maintains institutional memory.
Most of these studies share a process understanding of learning. First, organisational learning starts with knowledge acquisition about performance, from which lessons are identified that can be used to correct errors or improve future actions. Second, this knowledge is disseminated within the organisation and used for policymaking through formal processes, such as feedback and training. Third, lessons are learned when knowledge is institutionalised across the different structures of the organisation and recorded in official documents, concepts, and doctrines (Dyson, 2019; Smith, 2018). We intend to explain why and how NATO involved NGOs in the creation of a PoC policy by empirically unpacking these three learning stages.
Importantly, given NATO’s double hat as a military and political (civilian) organisation, learning takes place at multiple levels of the organisation. For example, NATO’s (2022: 12) Joint Analysis and Lessons Learned Centre conducts analyses of military operations, after which the North Atlantic Council (NAC; the civilian structure) decides whether these analyses will be used to create new policies (Hardt, 2017). We, therefore, follow the terminology of the NATO (2023a) organigram and focus on learning as it takes place within and between the different levels of NATO. 2
Moreover, the literature suggests that learning in organisations with civilian and military structures, such as NATO, is affected by bureaucratic politics, divergent organisational cultures and approaches to policy assessments (Bossong, 2013; Dyson, 2019; Dyson and Pashchuk, 2022; Hardt, 2017; Smith, 2018). First, the different priorities and goals of the bureaucracies of the IGO structures can reduce the effectiveness of the learning process (Bossong, 2013; Dyson and Pashchuk, 2022; Smith, 2018). For example, effective learning can stumble over budgetary disagreements or internal power struggles. Second, different cultures between civilian and military structures (Bossong, 2013; Hlatky, 2018; Smith, 2018), as well as within militaries (Dyson and Pashchuk, 2022), can affect the types of lessons identified and the extent to which they are learned by the IGO. Third, misinformation about which units are responsible for the learning process fails to support the learning process (Hardt, 2017). Fourth, personnel turnover can make it difficult to retain lessons learned in organisational memory (Bossong, 2013; Dock et al., 2021; Hardt, 2017; Smith, 2018). Finally, effective learning requires coordination between civilian and military structures, as well as a strong civilian leadership capable of overseeing the learning process (Bossong, 2013; Dyson, 2019; Dyson and Pashchuk, 2022; Smith, 2018).
Combining these insights from organisational learning with what we know from resource dependency, we argue that NATO staff opened up to NGOs because organisational learning processes revealed resource gaps, which were filled by including NGOs in the process. Specifically, we expect that resource needs to differ between NATO’s civilian and military structures, that is, the two levels at which NATO operates for policymaking. On one hand, IGO policy staff is generally expected to care more about the public image and public legitimacy of the organisation. They are sensitive to public support for their actions and the organisation as a whole (Eilstrup-Sangiovanni and Westerwinter, 2022; Mayer, 2008; Steffek, 2013; Tallberg et al., 2013). We, therefore, expect that officials in the NATO civilian structure open up to NGOs primarily for the public legitimacy that they can provide. We understand public (or social) legitimacy as ‘the acceptance of IOs’ right to rule by states and societies’ (Dellmuth and Tallberg, 2015: 459).
On the other hand, previous research that has highlighted the strong military ‘lessons-learned’ culture with a focus on ensuring operational effectiveness (Dyson, 2019; Hardt, 2017; Recchia, 2020). We also expect that officials in the NATO military structure open up to NGOs primarily because their expertise can improve the effectiveness of NATO operations. Public legitimacy and operational effectiveness expectations are summarised in Table 1.
Summary of expected resource concerns in NATO.
Data and methods
We followed a two-step empirical approach. First, we used evidence from public documentation: NATO publications (retrieved before July 2023), NGO reports and peer-reviewed articles on PoC. This allowed us to establish a first timeline of the development of NATO’s Policy for the PoC, explore NATO’s motivations for engaging with NGOs and identify organisations who were involved in NATO’s PoC work. Second, we complemented this data with insights from 26 interviews with mid- and senior-level NATO civilian and military staff officers, NGO representatives and other NATO stakeholders with expertise on PoC, conducted between February and November 2023 (see Appendix 1). We used these interviews to gain a deeper understanding of NATO officials’ motivations to engage with NGOs in developing a NATO PoC policy.
Public documentation
We conducted a mapping of a set of 32 research reports published by the Center for Civilians in Conflict (CIVIC), PAX, the ICRC and the Stimson Center (a US think tank) on the content and the operationalisation of NATO’s PoC policy. While NGO reports provide valuable insights, they are typically not peer-reviewed and may not hold the same level of scientific validity as academic research. Therefore, we also examined key peer-reviewed publications on NATO’s Policy for the PoC to corroborate the evidence present in the above-mentioned research reports (Hill and Manea, 2018; Keenan and Beadle, 2015; Rynning, 2017).
NATO’s publications represented the largest share of our documents and were analysed in detail using specific codes. A first group of documents was selected by searching for mentions of ‘non-governmental organisation’ and ‘NGO’, resulting in 396 documents. A second group of documents includes mentions of ‘protection of civilians’ and ‘civilian casualties’, resulting in 252 publications. We then analysed the content of these NATO documents. For the group of documents on NGOs, we considered the political context and the type of relationship (e.g. discussion, cooperation), the mentioned purpose of this relationship, as well as evidence of concerns about public legitimacy or operational effectiveness. In our search for evidence of ‘legitimacy concerns’, we looked for mentions of support to civilian populations and exchange of views with different stakeholders. For evidence of effectiveness concerns, we looked for mentions of ‘effectiveness’, ‘success’ and goal achievement language. We also checked for the level at which NGOs are involved (civilian, political or operational) and mention of other international actors. Our coding strategy effectively maps the key tenets of NATO’s engagement with third actors, as summarised in Table 2.
Coding of NATO’s publications with mentions of NGOs.
Following the terminology of NATO’s (2022) ‘Lessons Learned Handbook’, we then analysed NATO’s learning process by coding documents on civilian protection and civilian casualties. NATO’s first learning stage consists of an ‘analysis’; that is, NATO staff officers make an observation, for example, about a potential problem the organisation is facing, and they analyse this observation. This leads to the identification of a lesson and initiates the second learning stage, ‘implementation’. During this stage, the NATO chain of command tasks a person or organisation (‘Tasking Authority’) to produce and implement an action plan, report on that plan, before ‘remedial actions’ can be validated. A remedial action is defined as ‘an activity or set of activities that corrects an issue identified for improvement or facilitates the implementation of a best practice’ (NATO, 2022: 19). It ensures that the identified lesson becomes learned. Adopting a policy is a typical case of remedial action, as it helps to ‘change existing ways of doing things based on the learning’ (NATO, 2022: 10). In a final stage, the lesson learned is institutionalised within NATO and will be used by exercise planners, operations planners and training (NATO, 2022). Institutionalisation consists of ‘for example, incorporating the change into doctrine and procedures’ (NATO, 2022: 10).
We linked mentions of civilian casualty incidents and civilian protection practices prior to NATO’s (2016b) PoC policy to the analysis stage. Mentions of NATO’s PoC policy indicate an implementation stage, as such a policy regroups ‘remedial actions’ ensuring that future NATO operations follow better civilian protection practices. Finally, we connected the mention of exercises, education, and training that include objectives of civilian protection to an institutionalisation stage. Table 3 summarises our interpretation of NATO’s learning stages on civilian protection.
Interpretation of NATO’s learning stages on the protection of civilians.
Interviews
In the second stage of data collection, we used snowball sampling to identify interviewees who had been involved in NATO’s learning process on PoC, and could explain why and how NATO cooperated with NGOs throughout this learning process. This enabled us to map which NATO structures and NGOs were involved in NATO’s learning process on PoC. We conducted 26 interviews with NATO International Staff and International Military Staff, NGO representatives, ICRC representatives, NATO subject matter experts and other experts. Given the duration of the learning process and staff turnover, not all interviewees were present during the entire process from analysis to institutionalisation.
We first asked interviewees questions about when they first heard about NATO’s Policy for the PoC. We then delved deeper into assessing our expectations of legitimacy and effectiveness by asking interviewees why NATO engaged with NGOs on PoC issues, how this cooperation took place in practice and what they considered to be costs and benefits of engaging with NGOs. Our NATO interviewees were for most part still working for NATO and preferred to speak under the condition of anonymisation.
Two limitations are worth addressing. First, using a snowball sampling method meant that we mainly interviewed people with good knowledge of NATO’s Policy for the PoC. NATO interviewees were mainly active in the Human Security division or as civil–military cooperation (J9) officers, and three of our interviewees were former members of NATO’s high-level leadership. Yet, capturing member states’ perspectives is, therefore, beyond the scope of this study. Second, we did not have access to classified documents that could reveal high-level deliberations on NATO’s Policy for the PoC.
NATO’s learning process with NGOs on civilian protection
Based on the document analysis and interviews, we identify three key stages of NATO–NGO cooperation on PoC. We label these stages using NATO’s (2022) ‘Lessons Learned Handbook’ terminology (see Table 3). Until 2014, NATO–NGO cooperation was mostly operational (analysis stage). Between 2015 and 2016, ad hoc cooperation between NATO’s civilian structure and NGOs enabled NATO to formulate and adopt a civilian protection policy (implementation stage). Since 2017, NATO’s military structure has opened up to NGOs and other stakeholders for the operationalisation and institutionalisation of the policy (institutionalisation stage). We first present the overall trends from the document analysis, before turning to why and how NATO cooperated with NGOs during each individual learning stage.
Figure 1 displays an increasing trend in the number of NATO documents mentioning NGOs over time. More than 65% of the documents mentioning NGOs were published between 2002 and 2011, suggesting that this is the period when NATO gave more public visibility to the work of NGOs. In comparison, between 2012 and 2023, the number of publications where NGOs are talked about is much lower, with a maximum of nine publications per year. It is difficult to distinguish NATO’s learning stages on PoC in this group of documents, as these documents cover a total of 45 different topics, including a variety of policy issues, bilateral country relations, and NATO operations (see Table 5, Appendix 2). Documents that mainly address issues of operational effectiveness (N = 98) tend to largely refer to ‘cooperation’ with NGOs (N = 67). Other documents where public legitimacy seems to be a central issue (N = 45) tend to refer more to ‘discussions’ with NGOs (N = 17). Table 7 summarises these results (see Appendix 2).
Distribution of NATO documents mentioning NGOs over time.
Figure 2 shows the distribution of NATO documents that include the terms ‘civilian protection’ (N = 163) and/or ‘civilian casualties’ (N = 89). There is a clear peak in the number of these publications in 2011 (N = 79). This corresponds with NATO’s Operation Enduring Freedom in Libya, which was mandated to protect Libyan citizens. In this group of 252 documents, 32 can be associated with an analysis stage, 11 with an implementation stage and 11 with an institutionalisation stage. In addition, documents indicate that NATO engages with many different stakeholders on the PoC, such as the United Nations, the EU, and the ICRC (see Table 6, Appendix 2).
Distribution of NATO documents mentioning PoC and civilian casualties over time.
Overall, these initial findings from our document analysis suggest that NATO’s public recognition of the work of NGOs, as well as of concerns about civilian casualties and civilian protection, was a gradual process. These findings also already hint at a combination of both operational effectiveness and public legitimacy as NATO’s considerations for ‘discussing’ or ‘cooperating’ with NGOs. In the following section, we unpack this process to explain the emergence of this recognition and to show in more detail how NATO started to collaborate with NGOs at different levels.
Stage 1: analysis (pre-2014). The issue of civilian casualties in ISAF
NATO’s receptiveness to NGOs regarding civilian casualties originated from the military’s experience on the ground during ISAF, and progressively moved to the civilian level where PoC was placed on NATO’s political agenda. Our document analysis reveals that 21 of the 32 documents published during the analysis stage on PoC can be linked to ISAF (65%). Moreover, 97 out of 396 documents with mentions of NGOs also relate to the ISAF operation or operational activities in Afghanistan (24%), making it the operation with the highest frequency of mentions in both groups. There is a peak in the publication of documents mentioning NGOs in the context of ISAF in 2006 (N = 18) and 2007 (N = 13). This suggests that ISAF played a crucial role in NATO’s PoC learning process, specifically during the analysis stage.
Our interviews confirm this observation and show how on-the-ground cooperation with NGOs during ISAF was motivated by both legitimacy and effectiveness concerns. NATO became interested in the topic of civilian protection following NGO and media reports of large-scale civilian harm incidents caused by ISAF in 2006–2008 (interview 9). These incidents had the potential to undermine the public legitimacy of the NATO operation, particularly in NATO member states. As a former military advisor to the ISAF commander noted: ‘civilian casualties was obviously an area of great concern that affected public support for the government and of course, public support for the national security forces’ (interview 18). The support of the Afghan population for ISAF, which was ‘at the heart of the mission’, was also a concern (interview 4). At the time, CIVIC and the ICRC, as well as the UN Assistance Mission in Afghanistan, were engaging with NATO at the combat and command level of ISAF (interview 11). The ICRC, which had worked with successive ISAF commanders, eventually signed a Memorandum of Understanding with NATO’s (2016a) strategic commands (interview 14). In addition to on-the-ground engagement, CIVIC staff were conducting research about civilian harm incidents and advocating for better civilian protection practices. In their advocacy, they gained access to ‘senior commanders at ISAF’ (interview 11).
This all arguably led to more awareness about the issue of civilian casualties in ISAF, as shown by the first tactical directives issued by ISAF Commander General McNeill in June 2007 on the importance of exercising discretion and proportionality to avoid causing civilian causalities (interview 18), and the creation of a Civilian Casualty Tracking Cell in 2008. 3 These operational measures aimed to reduce the number of civilian casualties caused by ISAF (interviews 18, 24). 4 Moreover, monthly briefings were organised between ISAF military staff, the United Nations Assistance Mission in Afghanistan and NGOs such as Human Rights Watch and the Afghanistan Independent Human Rights Commission (interview 18). These meetings allowed the military to receive feedback and check the accuracy of the information gathered by the Civilian Casualty Tracking Cell. They also served to update partner organisations about new tactical directives issued by ISAF commanders (interview 18). We see here that both legitimacy and effectiveness concerns were at play. Engaging with NGOs on the ground could help NATO to reduce civilian casualties and provide greater legitimacy to ISAF in the eyes of the local population, which was key for the success of the operation. Nevertheless, NATO (2007, 2008, 2009) documents published at the time address the fact that some NGOs present on the ground did not engage with the military because they wanted to remain seen as neutral actors.
Although NATO’s cooperation with NGOs mainly took place at the operational level before 2014, there was already some, but less frequent, collaboration between the civilian structure and NGOs on civilian protection issues. First, from 2007 onwards, the Operations Division started organising biannual roundtables with NGOs in Brussels and New York to discuss about civilian harm issues. Although many NGOs were critical of NATO, some officers from the Operations Division were interested in hearing those critical views (interview 12). They were aware that engaging with NGOs was crucial as ‘the easiest way to erode [public] support [for ISAF] was the civilian casualty incidents’ (interview 12). For instance, many NGOs were strongly opposed to the idea of NATO establishing a Post Operations Emergency Relief Fund inspired by the fund made available to US commanders in Afghanistan for solatia payments.
Second, the Operations Division’s good relations with the NGO CIVIC and with the ICRC led to the establishment of non-binding guidelines on civilian protection in 2010. As some Allies became more vocal about their willingness to improve NATO’s PoC practices, Operations Division staff began to reach out to the NGO community. Our NATO interviewees who were involved in that process at that time referred particularly to engagements with CIVIC 5 (interviews 2, 9, 12). Yet, Human Rights Watch, Amnesty International, Save the Children and the Norwegian Refugee Council also took part in the discussions (interview 12). Moreover, Operations Division staff working on human rights issues broadly speaking already had a ‘very good relationship’ and a ‘very open dialogue’ with the ICRC, with whom they met twice a year in Brussels and Geneva (interview 12). In contrast, it took CIVIC ‘a long time and a lot of meetings and [. . .] trips to Afghanistan’ to be recognised as a ‘trusted advisor’ by high-level NATO officials (interview 9). Such recognition allowed CIVIC to influence the policy developments that followed.
Between 2008 and 2010, the Operations Division staff invited CIVIC representatives for brainstorm sessions, which led to the idea of developing non-binding civilian protection guidelines – an alternative to a policy that would have had little chance of being accepted by all Allies at the time (interview 12). CIVIC and the ICRC contributed to writing these guidelines and a representative of CIVIC even briefed the Operations Policy Committee on their importance. Eventually, the ‘Non-Binding Guidelines for Payments in Combat-Related Cases of Civilian Casualties or Damage to Civilian Property’ were approved by the NAC in June 2010 (interview 12). Such success is reflected in the fact that a year later, CIVIC staff were invited to contribute insights into civilian concerns to NATO’s Libya operations, allowing the NGO to show its practical ‘added value’ (interview 11). In the document analysis, only one document mentioning NGOs can be linked to NATO’s 2011 operation in Libya. In contrast, the NATO operation in Libya is by far the most covered one in the group of documents on PoC and civilian casualties, with 79 documents. It is remarkable that NGOs were so little rarely mentioned in this context, given that the Operation Division regularly engaged with NGO staff at that time (interviews 9, 11, 12). Overall, we see that NATO’s receptiveness to NGOs began on-the-ground as part of an analysis of operational experiences, before moving to the civilian level and feeding even more into the learning process on civilian protection.
Stage 2: implementation (2015–2016). NATO adopts a policy for the PoC
Our document analysis suggests a shift from analysing PoC issues at the operational level to implementing a policy in the civilian structure after the end of ISAF. For example, from the 16 documents published in 2016 mentioning PoC, six are official statements or texts on NATO’s PoC policy and four refer to NATO’s cooperation with the UN in peacekeeping operations. Four of the documents mentioning NGOs in 2015–2016 do so in the context of military exercises. Moreover, the official text of the Policy for the PoC does not specifically mention NGOs but emphasises the need for ‘engagement’ and ‘communication’ with ‘civil society’ four times (NATO, 2016b).
The interest of NATO civilian staff in formulating a PoC policy can be traced back to 2015. That year, the Operations Policy Committee was briefed by CIVIC representatives on the transfer of ISAF civilian protection best practices to the Resolute Support Mission, and the Joint Analysis and Lessons Learned Centre published an internal report entitled ‘Protection of Civilians, how ISAF reduced civilian casualties’ (interview 15). Such engagement was arguably driven by public legitimacy and operational effectiveness concerns that arose during ISAF. While writing the report, NATO analysts were in contact with representatives of the ICRC and the UN Assistance Mission to draw on their knowledge as organisations that work on a daily basis on PoC issues (interview 15). According to one NATO official, the report highlighted the ‘strategic costs’ of having (initially) failed to protect civilians in Afghanistan and the effectiveness of civilian harm mitigation training and exercises later during ISAF (interview 2). The PoC policy initiative also received the support of a group of 15 nations, labelled the ‘Tiger Team’ (interview 14). They invited the Operations Division to propose a PoC policy in October 2015 and sent the NATO Secretary General a letter highlighting the importance of developing a PoC policy (Dock et al., 2021). At the same time, the military exercise ‘Trident Juncture 2015’, which took place in October–November 2015, and which had some civilian protection elements, included the participation of NGOs (interview 21). According to the section head of the Operations Division PoC at the time, ‘there have been some Lessons Identified when working with NGOs during the exercise, and ‘there [was] a need to exchange feedback [with NGOs and IOs] . . . to ensure the overall safety and protection of civilian populations’ (NATO Joint Analysis and Lessons Learned Center, 2015), which suggests a concern for operational effectiveness. Eventually, the NAC tasked the Operations Policy Committee to formulate a civilian protection policy in December 2015.
To draft the policy, the Operations Policy Committee built on the network of organisations with whom it had been cooperating on children and armed conflict issues since 2012. This included the ICRC, CIVIC, the Stimson Center and the United Nations. At the same time, a mapping of NATO’s existing range of activities in terms of civilian protection in the civil space was conducted at ACT and involved contractors working as subject matter experts (interviews 1, 6, 7). For writing the actual policy, the Operations Policy Committee organised four briefing sessions with several stakeholders, including the United Nations, the ICRC, CIVIC, the Peacekeeping and Stability Operations Institute from the US Army War College, and the Norwegian Defence Research Establishment. NATO policy officers engaged with these stakeholders because of their expertise on PoC, and many had on-the-ground experience that was also valuable for NATO officers to learn from (interviews 1, 2, 7, 14, 20, 21, 22, 24). These organisations shared their views and suggestions on the content and language of the policy (interviews 1, 2, 7, 14, 20), which arguably brought more credibility to the draft policy (interviews 2, 8, 22). A first policy draft was sent to the International Military Staff for input and negotiations, before being sent to the NAC in April 2016. After a total of seven meetings with the stakeholders mentioned above, the NAC approved the NATO (2016b) Policy for the PoC in July 2016.
Although interviewees highlighted the perceived added value of including NGOs in the process, the extent to which NGOs’ input was integrated into the final policy remains unclear. Interviews suggest that the PoC policy had to be kept broad enough to be accepted by 30 NATO member states (interviews 2, 22). According to the same report of the Stimson Center, the final policy was less ambitious and used less precise language than the original version drafted by ACT and ACO (Dock et al., 2021). For some NGOs, this led to frustration that too few of their recommendations were included in the final policy (interviews 7, 8).
In sum, 2015–2016 saw the implementation of lessons learned from ISAF, with the adoption of a PoC policy for which NATO civilian officers consulted with NGOs and other stakeholders. This policy stemmed from the discussions about the public legitimacy of NATO operations in Afghanistan, where public reports pointed to high numbers of civilian casualties. The primarily ad hoc consultations aimed to develop a policy that would implement effective measures to enhance civilian protection in future NATO operations. More effective PoC policy was in turn linked to the legitimacy of those future operations. This cooperation revealed the value of NGOs’ views and recommendations on what should be included in the policy.
Stage 3: institutionalisation (since 2017). NATO’s military structures operationalise the PoC policy
The institutionalisation of the PoC policy by NATO’s military structure started in 2017. Interestingly, our document analysis shows NATO’s appreciation of the ICRC’s work. Several documents refer to NATO’s (2017, 2018a) cooperation with the ICRC to develop a PoC policy and to its participation in NATO exercises and training related to PoC. In addition, interviews show that NGOs such as CIVIC and PAX were also actively involved in the operationalisation of the policy. This cooperation appeared to be more pragmatic and driven by effectiveness concerns.
First, NATO’s strategic commands worked with NGOs throughout the process of translating the PoC policy into military terms, namely for writing a military concept, handbook and bi-strategic directive. Cooperation with NGOs then focused on the application of the guidelines of these documents in exercises and training. The process was coordinated by a team of four to five people from the Civil–Military Co-operation (CIMIC) branch in the office situated at the Supreme Allied Headquarters Europe. CIMIC acts as a bridge between external civilian organisations and ACO (interviews 2, 6, 7, 13, 20, 21, 22). Interviewees also agreed that NGO participation, especially in military training and exercises, is common at NATO (2018b) as it is part of the organisation’s military doctrine for civil–military cooperation. This section further zooms in on how NATO officers engaged with NGOs for implementing the PoC policy between 2017 and 2022.
NATO’s (2023b) strategic commands adopted an Action Plan early 2017 to define how to integrate PoC into ‘training, doctrine, education and exercises and into the planning and conduct of operations’. At first, CIMIC organised a few workshops to get the input from 15 organisations, including NGOs, on a Military Concept for the PoC (interviews 7, 9). A tabletop exercise was even organised with NGOs to see how the policy would work in practice (interview 7). Once the Military Concept for the PoC by the Military Committee was adopted in June 2018, CIMIC organised additional consultation rounds with NGOs and other stakeholders to obtain their views on the content of a military handbook (interviews 7, 10, 13). Among them were representatives of CIVIC, PAX, the ICRC and other organisations with specific expertise in the other topics covered in the handbook, such as Save the Children for Children in Armed Conflict issues and Transparency International for anti-corruption issues (interviews 7, 13). The expertise and knowledge of these organisations in their respective fields were valued by NATO officers (interviews 1, 6, 7, 9, 13, 14, 21). Meanwhile, because of its long-standing relationship with NATO, the ICRC also could access the work done on civilian protection to add in the Allied Joint Publications, which set the doctrine that the Joint Force Commands apply in operations. Eventually, ACO published a Handbook for the PoC in March 2021, bridging policy and military concept with the aim of creating a ‘PoC mindset’ (NATO, 2021). Furthermore, the ICRC was consulted by NATO in the writing of a bi-strategic directive on PoC, which was signed by the ACO and ACT in 2022 (interviews 14, 20, 21). This document sets out a division of tasks in PoC policy (interview 14).
NATO–NGO cooperation on the concept and handbook was not all positive. Lacking access to classified information, some organisations influenced member states to push agendas that were not in line with NATO’s official agenda (interview 7). Moreover, given that it drew on pre-existing relationships, some interviewees argued that the CIMIC office acted as a ‘gatekeeper’ to NGOs’ demands on PoC, as it filtered out which viewpoints and information to include. This ultimately affected how informed NATO’s operational commands were about NGO concerns and suggestions on the issue (interviews 6, 7, 13). For example, NGOs started to push for PoC to be incorporated in the work done by intelligence, planning and operations officers. However, several interviewees explained that PoC is still confined to CIMIC knowledge, which makes its operationalisation less effective (interviews 2, 6, 9, 26). Despite this, this stage saw more regular and specific cooperation with NGOs on the PoC policy.
From 2019 onwards, NATO’s strategic commands kept working with NGOs in exercises and training to operationalise the new PoC guidelines set out in the concept and handbook. This work mostly took place at the operational command level, with fewer interactions with the CIMIC office (interviews 7, 13, 21, 23). Depending on their expertise, stakeholders supported the development of training, while other organisations’ input was more focused on implementing the policy in the operations or developing additional standard operating procedures (interview 7). In April 2019, the Finnish Defence Forces International Centre (an academic institution that is part of the National Defence University in Helsinki) hosted and conducted a first ‘NATO-UN PoC course’, which was developed together with the Dutch NGO PAX (PAX for Peace, 2019). Some of the organisations that participated in this training were Save the Children, the Stimson Center, the United Nations Office for Coordination of Humanitarian Affairs and the ICRC. The participants in this training were officers assigned to NATO positions, military officers assigned by their countries and civilian staff officers (interviews 1, 6). Since then, the 10-day course has been conducted twice a year, including in an online format during the COVID-19 pandemic (interviews 1, 21).
In addition to the trainings, NATO also started including elements of civilian protection in the objectives of its military exercises, to which some NGOs were systematically invited to join. ‘Steadfast Jupiter Jackal’, the first exercise including civilian protection objectives in a scenario of high-intensity conflict, was organised in December 2020 (NATO, 2020b). According to some interviewees, it was sometimes difficult to gain access to discussions, planning, exercises and training organised by the military side (interviews 9, 13), with the civilian structure seen as a more appropriate partner for NGOs on PoC. The lack of strategy for the involvement of NGOs on the military side – despite it being a key tenet of the PoC Action Plan – has left some with a certain sense that the involvement of NGOs has been a box-checking exercise (interview 9). In 2021, during another military exercise, NGO staff did identify gaps in the military translation of the PoC policy for a scenario of security force assistance (interview 13). Another example is when PAX and the Stimson Center organised a 4-day tabletop exercise on civilian protection in October 2021. One interviewee explained that the exercise was tested at ‘one of the core level commands’ to see whether the planners could deal with tough civilian protection issues and showed that they could not (interview 9). By including NGOs, NATO ensures that its military staff understand what actions would be appropriate if NATO were one day to operate in a specific environment, familiar to NGOs because of their on-the-ground experience (interviews 16, 22, 25, 26). More specifically, interviews suggest that NATO has been interested in NGOs’ ‘external advice’ (interview 7) ‘outside views’ and receiving their ‘frank feedback’ (interview 2) on training and exercises. To this day, NATO continues to regularly invite a selected number of organisations, including NGOs, United Nations agencies and the ICRC, to contribute to the PoC elements of its Article 5 and non-Article 5 exercises (interviews 7, 20, 21, 23, 24, 25, 26). Such engagement suggests that NATO is willing to collaborate with NGOs because their resources can ensure that the PoC policy is more effectively operationalised in training and exercises.
For the period 2017–2022, our findings align with our expectation that NATO officers in the military structure opened up to NGOs because their knowledge, expertise, and feedback could contribute to the effectiveness of NATO’s PoC. The benefits of this cooperation are ‘immense’ for NATO, according to a NATO civilian staff officer (interview 10), and it is a ‘win-win situation’ according to an NGO representative (interview 13). Since 2022, the Policy for the PoC is part of NATO’s broader concept of ‘Human Security’. Interviewees agreed that NATO’s challenge is now to further define what the military contribution to this concept is. Interviews show that NATO’s work on Human Security has led to renewed consultations with those NGOs that focus on PoC, such as CIVIC and PAX (interviews 1, 2, 6, 9, 13, 14, 22, 25). NATO civilian officers appear to be interested in the ‘technical credibility’ of these organisations, for instance about the latest situation regarding civilians that are affected by Russia’s war in Ukraine (interview 22). It remains to be seen whether and how NATO will keep engaging with NGOs on the cross-cutting topics covered by Human Security.
Discussion
Our findings have implications for both organisational learning and resource dependency scholarships. On learning processes, the analysis first and foremost shows that NATO’s engagement with NGOs on civilian protection issues followed a bottom-up approach, starting at the operational level during ISAF in Afghanistan. Successive ISAF commanders began to address the issue of civilian casualties, and officers in the civilian structure became interested in bringing the solutions applied on-the-ground to NATO’s policy agenda. As highlighted in the literature (Bossong, 2013; Dyson, 2019; Dyson and Pashchuk, 2022; Smith, 2018), leadership contributed to NATO’s learning process. Here, it was the military rather than the civilian leadership that initiated NATO’s learning process on the PoC. Moreover, we find that there has historically been limited coordination between NATO’s civilian and military structures regarding their engagement with NGOs on PoC, with the institutionalisation of the 2016 PoC Policy as an exception. We find only minimal evidence of a transfer of lessons learned between different levels or between the civilian and military structures within NATO. Arguably, this can hamper effective learning on PoC within the organisation (Bossong, 2013; Dyson, 2019; Dyson and Pashchuk, 2022; Smith, 2018).
Looking at the motivations for NATO–NGO collaboration on the development of a PoC policy, the analysis shows that resource dependencies mattered. NGOs working specifically on PoC ‘pushed’ for access, but ‘pull’ factors and resource concerns ultimately led to more structured and frequent work with NGOs. However, we did not find a clear division of legitimacy and effectiveness concerns between NATO’s civilian and military structures. Rather, concerns about effectiveness and legitimacy are closely interlinked. During ISAF, NGOs ‘pushed’ for better civilian practices in operations, and ISAF commanders ‘pulled’ them onboard to ensure operational effectiveness and public support. Next, these NGOs were progressively granted access to NATO’s civilian structure for discussions. By 2015, NATO welcomed NGOs’ recommendations, which built on their on-the-ground expertise, to formulate more effective PoC policies. Since 2017, NGOs are regularly invited to contribute to military exercises and PoC trainings to ensure the effective operationalisation and institutionalisation of NATO’s Policy for the PoC.
Surprisingly, despite the importance of public legitimacy concerns during the initial analysis stage and concerns around civilian casualties during ISAF, few publicly available documents mention NGOs as being part of NATO’s PoC learning process (18 out of 252 documents). This suggests that collaboration with NGOs was not used to publicly legitimise NATO’s work on civilian protection. In other words, while legitimacy concerns triggered initial cooperation, effectiveness concerns facilitated stronger cooperation once the process was underway. Our findings are summarised in Table 4.
Summary of the findings.
Conclusion
The PoC is a matter of concern for the entire society and is, therefore, best addressed through collaboration between state and non-state actors, as well as civilian and military actors. While NATO is sometimes portrayed as an organisation that is relatively closed to civil society participation (Mayer, 2008; Wright, 2023), this article provides nuance to this assumption.
Relying on resource dependency and organisational learning theories, we asked: What explains the decision of NATO officials to work with NGOs when developing a PoC policy, and how did this cooperation take place? Our study demonstrates that NATO officials decided to work with NGOs on the PoC because of both legitimacy and effectiveness concerns. NATO commanders began to learn about the importance of better protecting civilians during ISAF, where incidents of civilian casualties risked eroding the public legitimacy of the operation and ultimately its success. Cooperating at the operational level with NGOs that had the knowledge and personnel that could help ISAF forces to reduce civilian casualties, for example, through the adoption of new tactical directives, addressed these concerns. As soon as the added value of this operational cooperation with NGOs became clear to the civilian structure, NATO (2016b) officers started to engage with NGOs to effectively implement and institutionalise the Policy for the PoC. NGOs such as CIVIC and PAX were involved in the process, as well as other organisations such as the ICRC, United Nations agencies and the EU. In practice, this engagement took place at both the civilian and military structure levels. At the civilian level, ad hoc discussions were organised where NGOs could share their views and on-the-ground expertise that helped NATO officials to formulate a PoC policy. At the military structure level, NGOs were brought in by NATO to contribute to the effective operationalisation of the policy: NGO representatives highlighted gaps in the operationalisation of the policy and helped to improve it. In other words, our findings show that, although NATO is often portrayed as unwelcoming to NGOs, in practice it has worked with these organisations over the past two decades. This cooperation appears to have had a rather positive effect in raising awareness of the need to better protect civilians in conflict and has influenced policymaking.
NATO’s renewed focus on collective defence could be a game-changer for its PoC policy, which is largely based on operational experiences in Afghanistan. As Russia’s war against Ukraine has shown, civilians are being deliberately targeted in modern conflicts and respect for principles of International Humanitarian Law is alarmingly eroding. The massacres committed by the Russian forces in the Ukrainian town of Bucha in March 2022 are one stark reminder that, in any future conflict on NATO territory, it is imperative that NATO countries have a clear and uniform strategy for protecting civilians. Embedding PoC guidelines into NATO’s standard operating procedures and ensuring NATO’s readiness to apply them in different possible military scenarios is a process that takes time. For now, the effectiveness of NATO’s PoC training remains to be seen, as does NATO’s work towards a ‘PoC mindset’ in the Alliance. As NATO membership has expanded to 32 Allies, it is time to examine how the Alliance’s PoC practices are being operationalised at the national level.
Footnotes
Appendix 1
Appendix 2
Codes co-occurring with legitimacy and operational effectiveness concerns in document group on ‘NGOs’ (N = 396).
| Legitimacy (N = 45) | Operational effectiveness (N = 98) | |
|---|---|---|
| Level of cooperation | ||
| Civilian structure (N = 81) | 13 | 5 |
| Military structure (N = 11) | 1 | 4 |
| Operational level (N = 54) | 0 | 29 |
| Type of relationship with NGOs | ||
| Discussion (N = 109) | 17 | 5 |
| Cooperation (N = 145) | 8 | 67 |
| NATO stakeholder | ||
| Academia (N = 26) | 3 | 1 |
| Diplomats, politicians (N = 10) | 1 | 0 |
| European Union (N = 67) | 7 | 30 |
| International organisations (World Bank, OSCE, OECD, others) (N = 68) | 13 | 13 |
| Journalists, experts (N = 32) | 0 | 6 |
| Think tanks (N = 20) | 5 | 0 |
| Mission or issue | ||
| Afghanistan, ISAF, Resolute Support Mission (N = 97) | 7 | 41 |
| Kosovo, KFOR, Yugoslavia (N = 29) | 0 | 8 |
| Comprehensive Approach (N = 19) | 0 | 6 |
Acknowledgements
The helpful feedback of participants of the panel ‘Civil-Military Relations in Challenging Times’ at the 2023 European Initiative for Security Studies, and of the workshop on ‘Cooperation and Legitimacy in International Relations’ at the 2023 Annual Political Science Workshops of the Low Countries, were highly appreciated.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by the Gerda Henkel Foundation [Grant for project AZ 12/KF/22].
