Introduction

‘I remember [years ago] being in a refugee camp in Syria, and when there were demonstrations and people picked up sticks and were throwing stones, and we were like, ‘Alright, that’s it, we’re withdrawing until they settle down.’ We withdrew for two days until they came and apologised and then we went back in again. Sticks and stones are a piece of cake compared to what we face now.’¹

Relayed by a humanitarian worker interviewed for this article, the quote paints a vivid portrait of the way that many humanitarians view the shifting nature of their operational environments. Indeed, the salience of security management as a policy concern for humanitarian organisations has ballooned in recent decades. The aid industry has directed substantial efforts towards bureaucratising security management (Schneiker, 2015; Beerli, 2018), raising the public visibility of the security risks that humanitarians face in the field (see, for example, Action Against Hunger, n.d.), probing the causes of attacks against aid workers (Stoddard et al., 2012, 2017), and analysing and debating organisational approaches towards operating in insecure environments (Humanitarian Practice Network, 2010; Egeland et al., 2011; Healy and Tiller, 2014; Neuman, 2016; Jackson and Zyck, 2017). Whether humanitarian insecurity has actually increased in recent years is the subject of often heated debate among practitioners, analysts and policymakers (Dandoy, 2014; Weissman, 2016). After all, humanitarian actors have always faced security risks in their endeavours to deliver assistance to, and promote protection for, vulnerable people amid ongoing armed conflicts, natural disasters, or other crises (Dandoy and Pérouse de Montclos, 2013). It is clear that, regardless of whether humanitarian insecurity has actually statistically increased, security management as a policy issue will continue to play a significant role in the planning and implementation of humanitarian operations.

The particular aim of this article is to probe how the humanitarian sector is pushing towards new frontiers in security management and to examine the policy fault lines likely to shape humanitarian organisations’ responses moving forward. This analysis draws primarily from a set of extensive semi-structured interviews conducted with 118 individuals with a broad scope of professional experience in the humanitarian sector. The interviewee pool includes practitioners with substantial field experience working for United Nations (UN) agencies; international and national non-governmental organisations (NGOs); the Red Cross/Red Crescent Movement; and professional fora and associations in the humanitarian sector. Interviewees discussed humanitarian operations undertaken around the globe: in Africa, the Middle East, the Asia/Pacific region, Europe and the Americas/Caribbean.

Interviewees were primarily field-based humanitarian actors who have served in senior- or mid-level management or operational roles in humanitarian response. The gender breakdown of the interviewee pool is 78 male, 40 female. The interviewee pool consists largely of international staff – indeed, only eight interviewees were national staff – a limitation, given that national staff constitute the vast majority of the humanitarian workforce and since they are also disproportionately affected by security incidents. Nevertheless, the scope of the interviewee pool allows for an examination of the policy discourse at the global level, especially among those disproportionately represented in policymaking. The majority of the interviews, 104 in total, were conducted remotely via Skype or telephone. Additional interviews, totalling 14, were undertaken in person in Cambridge, Massachusetts (United States), Dubai (United Arab Emirates), and Beirut and Byblos (Lebanon).

This article is divided into five sections. The first three sections examine, in turn, legal accountability efforts – or the lack thereof – directed towards addressing humanitarian insecurity, public advocacy and confidential negotiation. The fourth section extrapolates lessons from these earlier sections relevant to the broader discourse on security management within the humanitarian sector. The final section offers concluding remarks.

Legal Accountability

On 3 October 2015, the Médecins Sans Frontières (MSF) trauma hospital in Kunduz, Afghanistan came under aerial assault. As a series of sustained airstrikes targeted the main building, first hitting the intensive care unit, MSF staff scrambled to treat the wounded while appealing to Afghan, US, and UN officials to stop the attack. By the time the firing stopped over an hour later, at least forty-two people had been killed – including MSF staff, hospital patients, and family members – over thirty were injured, and thirty-three were missing. MSF later alleged that the attack constituted a war crime and called for an independent and impartial investigation.

The bombing of MSF’s trauma centre in Kunduz, Afghanistan by a US Air Force AC-130U gunship quickly garnered headlines around the world, generating newfound attention to the targeting of humanitarian and medical actors in conflict zones, not only in Afghanistan, but also in Syria, Yemen and elsewhere, as well as renewed calls for legal accountability. As the incident highlighted, however, a notable gap exists between the lofty theoretical promise of international law – and its domestic corollaries – and the difficulties of achieving accountability in practice. On the one hand, the legal clarity of international law regarding the protection of humanitarian action in armed conflict from deliberate attack is inescapable, as enshrined, for example, in international humanitarian law (IHL), including the four Geneva Conventions of 1949 and their Additional Protocols of 1977 and 2005; the Rome Statute of the International Criminal Court (ICC); and customary international law (Brooks, 2015).

On the other hand, the interview findings of this study reveal a widespread sense of frustration among aid workers over the perceived inefficacy of the law in practice for the protection of humanitarian action. Even in light of the existence of not only international courts and tribunals but also informal accountability mechanisms such as UN reporting mechanisms and commissions of inquiry, there are concerns, as interviewees attested, about the viability and effectiveness of these mechanisms. Interviewees pointed to a lack of political will as a key reason for the paucity of adequate avenues for accountability for violence against aid workers. Moreover, the ICC lacks jurisdiction over many contexts where attacks against aid workers have occurred.² Interviewees also highlighted the long timeline typically required for legal cases, often spanning years or even decades. The pursuit of international justice, in the words of one interviewee, can be ‘an intergenerational battle’. Regarding justice at the domestic level, interviewees expressed the belief that, even if they pursued justice through domestic institutions in many contexts where they work, justice was actually infeasible. To illustrate with a particularly extreme example of the shortcomings of local legal institutions, a man charged with killing two MSF staff in Somalia, after being found guilty and sentenced to thirty years of prison time, was released after just three months (Karunakara, 2013).

Turning back to the Kunduz case, MSF engaged in sustained high-profile advocacy over the incident, garnering a great deal of public attention to the issue of violence against medical care, ultimately helping to foster the passage of UN Security Council Resolution 2286 (2016) on the protection of medical facilities. Yet, MSF struggled to find an appropriate venue to seek legal accountability for this specific incident. Action Against Hunger (ACF) has faced similar obstacles in its efforts to seek accountability for the execution of seventeen of its staff in their compound in Mutur, Sri Lanka in 2006. Just as MSF did in the wake of the Kunduz bombing, ACF alleged that the massacre constituted a war crime and called for an independent investigation (Action Against Hunger, 2006). While many suspected the involvement of state security forces in the massacre, the government of Sri Lanka attributed the killings to a rebel non-state armed group, and a commission of inquiry that the government established was later disbanded without issuing a public report (Amnesty International, 2009).

Despite the widespread lamentations about impunity for perpetrators of attacks against aid workers and calls from the humanitarian community for increased accountability, many practitioners caution against directly engaging with accountability mechanisms themselves. Humanitarian actors who struggle to gain and maintain acceptance from parties to conflict and access to populations in need are highly sensitive to the potential costs of further antagonising the perpetrators of attacks. This element is particularly salient in protracted conflict settings where the relationships between humanitarians and political or military actors who control, or might control, territory become long-term in nature. Indeed, some interviewees perceived that pushing for accountability could, at least in the short-term, increase risks for aid workers operating in the field.

Even in spite of the challenges of the practical application of the law, interviewees expressed an aversion to fully abandoning the pursuit of accountability, due to a belief that doing so would be tantamount to fully embracing impunity. Still, legal accountability is far from a panacea, but rather, one component of an array of tools to be employed in a holistic strategy to mould the operational environment into one that is more conducive to the implementation of humanitarian programmes. In the rest of this article, we examine two other important prongs – public advocacy and confidential negotiation – in such a multi-faceted approach.

Public Advocacy

On 11 July 2016, 80 to 100 soldiers forced their way into the Terrain compound housing international NGO staff in Juba, South Sudan. The soldiers fatally shot a local journalist and gang-raped several foreign women, singling out Americans and forcing the foreigners to watch. They also carried out mock executions, beat and robbed people and looted the compound. Throughout the attack, those at the compound appealed unsuccessfully for help to the UN peacekeeping force stationed less than a mile away, as well as to the US and other embassies. Eventually, South Sudanese security forces rescued all but three foreign women and around sixteen staff. Reports of the incident first became public only a month later, when several victims spoke with an Associated Press reporter, who then published the story. Subsequently, aid agencies and others forcefully criticised the UN for failure to act upon the calls for help from the foreigners at the compound, as well as for its public silence about the incident for nearly a month after it took place.

The Terrain compound attack accelerated broader discussions within the sector about not only accountability but also the extent to which humanitarian organisations should speak out about attacks against their staff, operations or facilities. At first blush, it may seem natural for humanitarian organisations to publicly condemn violence against aid workers. After all, if aid workers do not advocate for their own protection, who else will? There is indeed a concern that, if humanitarians fail to speak out in condemnation of attacks, this collective silence could constitute acquiescence to a ‘new normal’, as one interviewee put it, in which attacks against aid workers are prevalent and accepted as part of the cost of undertaking aid work.

Organisational silence can also place a burden on field workers who themselves feel compelled to speak out. After the Terrain compound attack, for example, some affected staff felt the need to go public after their organisations failed to do so. Yet, speaking out as an individual in such an instance can come at a great personal risk, not only of re-traumatisation but also to one’s career in the humanitarian field. This dynamic can be particularly heightened in cases involving sexual violence, especially due to the stigmatisation surrounding this form of violence and the predominance of men in security management roles (Mazurana and Donnelly, 2017).

While acknowledging the desirability of public advocacy at the organisational level, as well as the converse risks of remaining silent, humanitarian practitioners have also expressed various concerns about speaking out. One concern is about whether public advocacy is actually impactful, and what form of advocacy should humanitarians pursue. On the one hand, interviewees articulated scepticism about whether the plethora of data gathering and awareness raising initiatives in this area – such as the Aid Worker Security Database, the Aid in Danger project, the ICRC’s Health Care in Danger initiative, and MSF’s #NotATarget campaign – have actually played a role in shaping armed actors’ behaviour in relation to humanitarians. On the other hand, case study and comparative research paints a mixed and uncertain picture of the actual impact of more targeted public advocacy focused on particular combatants, contexts, and incidents. There are some contexts where advocacy has evidently influenced combatants’ behaviour (in particular, where parties to a conflict are concerned for their reputation and legitimacy in the eyes of the international community) and other instances when advocacy has appeared to be ineffective or even to escalate tensions (Leader, 2000; Atkinson and Leader, 2000; Bridges, 2010; Médecins Sans Frontières International Movement, 2013, 2014a, 2014b, 2014c, 2015a, 2015b; Bussmann and Schneider, 2016).

A second concern is that humanitarians’ public advocacy for their own protection could distract from the core mission of protecting civilians. As one interviewee commented, a tension arises because ‘aid workers don’t like to be seen as putting themselves above beneficiaries’. Nevertheless, many interviewees rejected the idea that advocacy for the protection of aid workers must come at a cost to the protection of civilians. Quite the contrary, these interviewees noted that the protection of humanitarian action and the protection of civilians more broadly are interlinked, and indeed, that the protection of humanitarian actors is an enabling factor for civilian protection. As one interviewee commented, ‘Our mission is to save lives and relieve suffering. And if we don’t have access to those populations, we’re not saving lives or relieving suffering, and therefore our mission is void.’

A third concern about speaking out on the issue of attacks against aid workers is the potential for blowback in the operational environment. This concern parallels the perceived risks of seeking legal accountability. Namely, many interviewees expressed the sense that speaking out about attacks on aid workers can come at a direct cost to humanitarian access or security by jeopardising their perceived neutrality, impartiality or independence. Furthermore, some practitioners expressed the fear that speaking out could lead to further attacks against humanitarians, especially in light of the aforementioned widespread lack of accountability for perpetrators.

A final difficulty relates to engaging in advocacy collectively as a country team, coalition or sector. Interviewees emphasised the desirability of inter-organisational efforts to speak with one voice about the unacceptability of violence against aid workers. But there is a broader-deeper trade-off at play. When more organisations are involved in drafting a statement in the wake of an attack, the statement can take a longer time to negotiate, tends to become less precise, and hence, could be less powerful. Conversely, involving fewer organisations could lead to a more forceful statement but sacrifices the power that could result from speaking with a more expansive, collective voice.

Donor dynamics further complicate these issues. For NGOs reliant on donor funds from the same states that might be targets of advocacy, it can be harder to take a forceful stance because these NGOs might perceive that they lack, as one interviewee commented, ‘the ability, or the comfort level, to really speak out’, except in ways that are ‘more measured’. The inter-organisational competition for donor support can also impede the public acknowledgement of security incidents, or even the sharing of information about incidents with donors, whether concerns about donor repercussions are well placed or not. As one interviewee attested, ‘If an organisation has to relocate for a week because of conflict and they come back to find their generators are gone, their vehicles are gone … they’re not necessarily keen on going out to their donors and explaining everything that’s happened, because they’re worried that the donors will hold them accountable.’

Thus, while many practitioners see the value in speaking out for the protection of civilians and, to an increasing extent, humanitarian actors, a wide range of internal and inter-organisational factors can stymie humanitarian organisations’ ability to use public advocacy as an effective means to alter the humanitarian security dynamics at play in the field environment. Given the limitations and frustrations surrounding both legal accountability and public advocacy, as this and the previous section have discussed, humanitarian actors often find themselves relying on other mechanisms for grappling with threats in the field. The next section examines a third mechanism – negotiation – that humanitarians employ sometimes in parallel with, sometimes in concert with, and sometimes in tension with public advocacy efforts.

Negotiation

In the midst of a civil war in a central African country, an armed criminal gang kidnapped several national staff working for an international NGO. The team in the field, scrambling to resolve the crisis, explored the possibility of paying the gang to secure their colleagues’ release. This option contravened the NGO’s explicit policies, but given the urgency and gravity of the situation, with their colleagues’ lives at stake, the aid workers in the field perceived that this option might be the only viable choice. Other aid agencies had experienced similar kidnapping crises in the same context. Rumours circulated about how much money organisations had paid, but publicly, agencies issued only blanket denials that money had been paid at all. Only through an informal, personal connection that one international staff member had with someone working for one of these other NGOs was the field team able to discern the ‘going rate’ for securing the hostages’ release.³

The vignette highlights the myriad difficulties that can arise in negotiations related to security incidents in the field. In this instance, the kidnappers’ monetary demands butted up against organisational policies that proscribed such payments. The sensitivity surrounding this issue also led to inter-organisational inhibitions around information sharing related to how different agencies operating in the same context had resolved similar incidents in the past with the same gang. Had other organisations paid? If so, how much? In this instance, the NGO did not ultimately need to pay the kidnappers to resolve the crisis, but the organisation seriously considered this option.

Such incidents are not necessarily uncommon. Rather, the anecdote is emblematic of the difficulties that many humanitarian organisations experience when negotiating not only hostage situations but also security guarantees from armed actors, the use of armed escorts in the field, and responses to attacks – including robberies, killings or other types of attacks – after they have occurred. Given that the lives of colleagues may be at risk, the stakes of such negotiations are high. But the tools at hand can seem limited due to the sometimes conflicting interests of aid workers and their interlocutors; the difficulties of internal deliberations, especially in situations involving ‘red lines’ set by organisational policies; and the challenging inter-organisational dynamics that can complicate cooperation and information sharing across organisational lines.

Not all negotiations bring forth the issues illustrated by the aforementioned vignette. Many negotiations that touch upon the protection of humanitarian action are largely relational in nature and might not require the humanitarian actor to consider what concessions to make or how far to compromise. Even in a hostage situation, when interlocutors appear to be driven by purely monetary interests, an ‘acceptance’-based approach rooted in widespread relationship-forging with key actors can imbue humanitarian actors with a valuable source of leverage. As one interviewee explained, ‘for the other side, there are always networks, social, and political realities that you can influence’.

Yet humanitarian actors cannot always expect negotiations to be purely relational. When interlocutors in the field push back in the discourse of negotiations, humanitarian actors sometimes need to make difficult choices about what they will and will not accept in terms of their operational parameters. The interviews revealed numerous issues on which humanitarian actors struck compromises when their ideal positions could not be met. For example, one interviewee discussed how the NGO for which he worked had a policy against the use of military escorts. Yet, in a context where, due to the insecure nature of the field environment, the use of military escorts was ‘the only way authorities would accept movement’, they reached a creative solution by which the NGO travelled behind the armed convoy, ten minutes later. Another interviewee described similar dynamics at play during a hostage situation. The NGO he worked for had a policy precluding payment to secure hostages’ release, but in practice during hostage negotiations, the organisation has several times found a third party – for example, a hostage’s family – to pay.

Determining when to bend, or even break, organisationally prescribed ‘red lines’ can lead to vexing internal deliberations. In the words of one interviewee, ‘The most difficult part of negotiating humanitarian access is not the players on the ground, it’s [negotiating with] the people you work for’. Inter-organisational dynamics can further fuel these difficulties. There is indeed a collective action problem that humanitarian organisations face. Joining together on collective ‘red lines’ and coordinating negotiation approaches across different organisations can empower humanitarian organisations to tilt an asymmetric power dynamic – by which humanitarian organisations need to reach an agreement with their interlocutors more than vice versa – in the humanitarians’ favour. There are, of course, numerous contexts where humanitarian organisations have adopted joint operating principles or protocols at the country level (Harmer et al., 2018). However, the humanitarian sector is fragmented, with disparate interpretations of where ‘red lines’ should be set, and organisations do not always adhere to collective red lines to which they have agreed. After a serious security incident prompts an organisation to temporarily suspend operations in order to receive assurances that such an incident will not occur again, in the words of one interviewee, ‘All it takes is for one organisation to say, “You know what, we’re just going to go back in tomorrow because our headquarters is requiring it, [or] our donor’s requiring it,”’ to render the suspension of operations irrelevant.

There are also other important aspects, as Clements (2018) has examined, of the asymmetric power dynamics at play when humanitarians negotiate. Hard pressure, in the form of physical threats, is more readily available to many humanitarian negotiators’ counterparts, especially armed actors. Humanitarian actors can, and in many cases do, use armed protection to level the otherwise asymmetric field. A reading of the history of humanitarianism reveals operational linkages between humanitarian actors and armed actors to have a long and rich historical pedigree (Pérouse de Montclos, 2014). Nevertheless, in the contemporary policy discourse, the sense is prevalent that armed protection and humanitarian action are, as one influential policy document notes, ‘uneasy bedfellows’, given the possibility that such a posture can be incompatible with humanitarian principles, harm local perspectives of humanitarian organisations, logistically constrain flexibility, escalate tensions and bring forth ethical dilemmas surrounding paying for private security (Humanitarian Practice Network, 2010: 74). As these concerns indicate, the issues raised are not merely about principles but are also operational in nature. If humanitarians do employ armed protection, the hard power dynamics might become more equal in nature, but only at a potentially great cost in terms of humanitarians’ ability to operate effectively.

Also relevant to humanitarian negotiations are softer forms of pressure, such as threats of public denunciation, scaling down or suspending operations, or withdrawing entirely from a country or location. In this sense, negotiation and public advocacy can function as two components of the same strategy, with public pressure used to push along an otherwise stagnant negotiation process. However, as this article earlier described, such modes of soft pressure can also harm relationships with key interlocutors, hindering an organisation’s ability to work in that context in the future. Interviewees discussed examples where their relationship-building efforts during negotiations were harmed by public advocacy from their organisation stemming from a higher organisational level. In this sense, negotiation and public advocacy may exist in concert or in tension with one another.

Furthermore, a threat to withdraw from a context can be an ethically vexing option to pursue, especially since it involves cutting off services to populations in need. Interviewees attested that pressing too hard, even in confidential dialogue, can have the opposite of the intended effect, inadvertently fuelling the security risks that humanitarian actors face. Making a viable threat to withdraw, and following through on such a threat, can thus be difficult for humanitarian organisations. The question then becomes: to what extent can humanitarian negotiators successfully wield other negotiation tools at their disposal – for example, relationship building and modes of seeking creative compromise – so that the outcome is not mere submission to the interlocutor’s wishes?

The pathway forward for addressing the challenges of negotiating the protection of humanitarian action has various dimensions. At the individual level, there is a need for humanitarian actors to commit themselves to a long-term process of cultivating their own negotiation capacity. The humanitarian sector has made some significant strides towards grappling with some of these challenges. Indeed, there has been a definitive embrace of the importance of negotiation in humanitarian operations (Grace, 2020). However, the interview findings suggest that much room for improvement remains. Many interviewees had never received any negotiation training or mentorship, and in relation to hostage negotiations particularly, felt that they and/or their organisations were ill equipped to the task. At the policy level, the compromises that many negotiations bring forth raise unsettled organisational and sector-wide questions about what humanitarians’ ‘red lines’ can and should be, and which trade-offs are worth making.

Moving beyond the Security Triangle

It is important to link the numerous issues that this article has addressed to the broader discourse on security management in the humanitarian sector. Specifically, this section examines legal accountability, public advocacy, and confidential negotiation through the lens of the security triangle, the popular security management framework articulated by Koenraad Van Brabant (2001). The security triangle, which has played an influential role in shaping discourse on security management in humanitarian operations, lays out three approaches (Figure 1). One approach is ‘acceptance’, which entails developing and maintaining positive relationships with communities and stakeholders, focusing on undertaking impactful, ‘principled humanitarian action’. A second approach is ‘deterrence’, or containing a threat with a counter-threat, for example, through armed guards, military escorts or diplomatic reprisals. A third approach is ‘protection’, or reducing vulnerabilities to any attacks that might occur, for example, through security training, body armour, fortified compounds or maintaining a low profile.

The security triangle

Source: Van Brabant (2001: 27).

Citation: Journal of Humanitarian Affairs 2, 1; 10.7227/JHA.030

View Full Size

The security triangle

Source: Van Brabant (2001: 27).

Citation: Journal of Humanitarian Affairs 2, 1; 10.7227/JHA.030

The security triangle

Source: Van Brabant (2001: 27).

Citation: Journal of Humanitarian Affairs 2, 1; 10.7227/JHA.030

The overarching question at stake is: how can and should humanitarian actors grapple with insecurity? Looking at this question in conceptual terms, there are two broad options. On the one hand, humanitarians can confront threats. Such measures aim to reduce operational risks by engaging with, and shaping the perceptions and incentives of, external actors present in the field environment. ‘Acceptance’ and ‘deterrence’ fall into this category. These approaches are oriented towards reducing the probability that actors in the field will feel inclined to target humanitarian workers for attack. Conceptually comparing ‘acceptance’ to ‘deterrence’, one can discern a crucial difference in terms of the depth of the transformation sought. ‘Acceptance’ aims for a ‘deep’ transformation, meaning that relevant actors in the field (for example, members of armed groups or gangs, militaries, or local political or business leaders) would shift from being inclined to attack – or being tolerant of attacks against – humanitarian actors, towards feeling no inclination to direct violence at humanitarians or even towards actively promoting the protection of humanitarian action. ‘Deterrence’ aims for a ‘shallower’ transformation in the sense that, if humanitarians successfully employ ‘deterrence’, certain local actors might still have some incentives to attack humanitarian actors but simply refrain from doing so out of concern for potential retaliatory consequences or the sense that an attack would simply be ineffective.

On the other hand, humanitarians can also mitigate vulnerabilities. Such measures – including physically fortifying humanitarians (often criticised as ‘bunkerisation’) or implementing bureaucratic risk management procedures to prevent humanitarian actors from operating in locations where security risks are too severe – aim to reduce humanitarians’ vulnerabilities but shy away from addressing the threats themselves. A ‘protection’ approach distinguishes this category in a sense that it aims – to quote the seminal report, To Stay and Deliver, published in 2011 by the UN Office for the Coordination of Humanitarian Affairs – ‘to reduce vulnerability to a threat, but does not affect the threat itself’ (Egeland et al., 2011: 28). To illustrate this conceptual distinction more concretely, a mitigation approach seeks to shield humanitarian actors from harm due to any armed attacks that may occur. A confrontational approach aims to dissuade the armed actor from launching an attack on humanitarians in the first place.

Of these three approaches, there is a widespread sense of the desirability of ‘acceptance’-based measures. Various interviewees spoke about employing ‘acceptance as our main security strategy’, ‘using the local community to be our protection’, and ‘rely[ing] so heavily on explaining who we are, and what we stand for, and how we work, and how decisions are made, and that’s the embodiment of humanitarian principles’.

Approaches rooted in ‘protection’ tend to be seen as a necessary evil of security management, to be employed in particularly volatile environments when other approaches appear infeasible or ineffective. Furthermore, creating physical barriers between humanitarian actors and their operational environment can even hinder ‘acceptance’-based modes of operation, making it difficult for humanitarians to undertake the types of engagements necessary to develop and sustain fruitful relationships with key local actors. It is perhaps due to the aforementioned conceptual distinctions – between not only ‘deep’ versus ‘shallow’ but also mitigating vulnerabilities versus confronting threats – that humanitarians have largely embraced ‘acceptance’ as the desired security management strategy. The most effective security management strategy, it seems, would be to create buy-in from local actors for humanitarian organisations’ aims and operations, or in other words, to grapple directly with the potential threats themselves.

The particular mechanisms examined in this article – legal accountability, public advocacy, and confidential negotiation – feed into these approaches in various ways. Legal accountability and public advocacy aim towards both ‘acceptance’ and ‘deterrence’. The vision inherent in erecting international laws that proscribe attacks against aid workers (and indeed all civilians), as well as engaging in vigorous public advocacy to condemn attacks when they do occur, is a normative environment where attacks against humanitarian actors will not be tolerated. The aim is to transform actors’ behaviour by crystallising the notion that violating such norms will yield legal or reputational consequences (i.e. ‘deterrence’) or even, additionally, that actors entirely internalise the norm of protecting humanitarian action (i.e. ‘acceptance’). Confidential negotiations can also feed into approaches rooted in ‘acceptance’ (through the relational dynamics at play during negotiation processes) or ‘deterrence’ (through threats of public denunciation or withdrawal, for example).

However, as the analysis of this article suggests, we are far from a world in which legal accountability, public advocacy and the relational and soft power aspects of negotiation entirely succeed in this manner. When ‘acceptance’-based or ‘deterrence’-based approaches fall short, humanitarian organisations have prioritised mitigating vulnerabilities. Indeed, modes of physical fortification and maintaining a low profile in line with a ‘protection’ approach remain prevalent in many insecure contexts. There are also more internally oriented dimensions of mitigating vulnerabilities that focus on ‘risk management’, for example, which seeks to manage, through developing bureaucratic procedures and context analysis capabilities, where humanitarian actors go and how they undertake their work in order to keep them safe. Other options in this vein are evacuation, scaling down, and ‘remote management’ (meaning withdrawing international staff and leaving implementation to national or local staff who still reside in the country), a controversial option that can unduly transfer risk from international to national staff while greatly reducing an organisation’s ability to monitor and control what exactly occurs in the field (Humanitarian Practice Network, 2010).

The analysis of this article points towards another important underexplored aspect of humanitarian organisations’ behaviour. Just as one can consider ‘risk management’ to be an internally oriented aspect of a ‘protection’ approach, ‘acceptance’ and ‘deterrence’ too have internally oriented counterparts that often remain overlooked in security management discourse. Whereas ‘acceptance’ and ‘deterrence’ both seek to transform the external environment, the concessionary dimension of negotiation processes, in particular, points towards important internal transformational aspects of grappling with humanitarian insecurity. Furthermore, just as one can categorise ‘acceptance’ and ‘deterrence’ as ‘deep’ or ‘shallow’, respectively, one can also envisage internally oriented approaches along this same spectrum.

An internally oriented – but ‘shallow’ – approach is ‘decoupling’, a phenomenon of organisational behaviour widely explored in sociology literature but yet to be applied to organisations operating in the humanitarian space. A ‘decoupling’ approach means that a fissure emerges between organisational policies and the actual practices of practitioners in the field. Interviewees discussed that, in many instances, there can be a ‘disparity’ or ‘big disconnect’ between, on the one hand, the policies that an organisation’s headquarters appears to believe are being implemented on the ground, or that the organisation publicly admits, and on the other hand, what actually occurs in the field (Boxenbaum and Jonsson, 2017). One can discern ‘decoupling’ at play, for example, when organisations publicly declare that they do not pay kidnappers who have taken staff hostage but either actively engage to find a third party to pay or even pay themselves without publicly admitting this fact. There is also the aforementioned example of the NGO with an organisational policy that proscribes the use of armed escorts, but when necessary to gain the authorities’ permission to operate, agreed to an armed escort that would travel a certain distance ahead of the NGO’s vehicles. In such examples, the publicly touted modus operandi contrasts with the reality of the compromises struck at the field level. ‘Decoupling’ is indeed ‘shallow’ in transformational terms in the sense that practices adapt only at the field level. The organisation, as a whole, does not morph in terms of what the organisation, overall, aspires to be. But a tension persists between, on the one hand, the image of itself that the organisation strives to project, and on the other hand, the pragmatic necessities of the field environment. Indeed, ‘decoupling’ can pervade almost every aspect of security management.

The contemporary discourse emphasises the notion that ‘acceptance’-based approaches and ‘principled humanitarian action’ are inherently linked and can facilitate access and security. But this thesis does not reflect the reality of contemporary security management. Instead, humanitarian negotiators are often left to define what is an ‘acceptable’ compromise on their own. Humanitarian organisations, at the policy level, tend to speak the language of principles, while at the ground level, aid actors engage daily in the actualities of interests and politics.

Humanitarian organisations might also consider an approach geared towards a ‘deep’ internal transformation. One could call such an approach ‘synthesis’, a concept drawn from the notion of dialectics commonly attributed to Georg Wilhelm Friedrich Hegel.⁴ In this conception of dialectics, an initial concept (‘thesis’) is negated by a second concept (‘antithesis’), and a third concept (‘synthesis’) unifies and reconciles the two. Relating this notion to the modern humanitarian sector, the initial thesis, prevalent in the contemporary policy discourse, is ‘acceptance’ as ‘principled humanitarian action’. The antithesis is the empirical reality of ‘decoupling’, meaning that, although humanitarian organisations publicly espouse this notion of ‘acceptance’, privately, at least at the field level, there is a widespread acknowledgement that compromise rather than principles is what actually facilitates access and security. ‘Synthesis’ as a security management approach would entail a public organisational acknowledgement of this reality, and indeed, internal and sector-wide debate about what does and does not constitute an acceptable compromise, and a perpetual reassessment of these issues across contexts and over time (Table 1).

In this sense, ‘synthesis’ has a double meaning in that it also refers to the polar opposite of ‘decoupling’. Whereas ‘decoupling’ insinuates a disconnect between policy and practice, ‘synthesis’ would entail embracing the reality to which ‘decoupling’ points. There is indeed an acknowledgement, as the interview findings indicate, that the humanitarian principles (especially impartiality, independence and neutrality), widely thought to facilitate access, can actually inhibit access and fuel insecurity. As the section of this article on negotiation explored, interlocutors in the field do not always wish for humanitarians to be principled. Rather, these interlocutors – for example, a criminal gang that kidnaps humanitarians in the hopes of making money – actually want humanitarians to serve these interlocutors’ own interests. This notion is in no way alien to policy discourse in the humanitarian sector. Indeed, various publications produced by the Centre de Réflexion sur l’Action et les Savoirs Humanitaires (MSF CRASH) have emphasised a different notion of ‘acceptance’ that places focus on finding acceptable compromises with armed actors, whose willingness to refrain from targeting aid actors is linked to their ability to instrumentalise humanitarians’ presence. This may occur, for example, where armed actors benefit from humanitarian services, the money that humanitarian organisations inject into local economies or an enhanced image and legitimacy garnered through facilitating humanitarian access (Magone et al., 2012; Neuman and Weissman, 2016). A sector-wide embrace of ‘synthesis’ would push further in this direction, leading to a new phase in the discourse on the interrelationship between principles, compromise and security management.

Conclusion

While violence against humanitarian action is certainly not new, practitioners and policymakers now devote substantial resources to grappling with the myriad challenges and dilemmas that humanitarian insecurity presents. As this article has examined, humanitarian organisations have increasingly embraced security management approaches that aim not only to mitigate the impact of risks in the field (i.e. ‘protection’ and ‘risk management’) but also to transform the operational environment into one more conducive to the implementation of humanitarian programming (i.e. ‘acceptance’ and ‘deterrence’). There is also the possibility, when external transformation is challenging or even infeasible, of looking inward. Indeed, internal transformations too play an important role in grappling with humanitarian insecurity (i.e. ‘decoupling’ and ‘synthesis’). As this article discussed, the pursuit of legal accountability, public advocacy and confidential negotiation are three important prongs of these approaches.

There remains, however, an overarching tension between the long-term vision and the short-term practicalities. Taking the long view, there is an enduring effort to shape the fabric of the normative environment in order to solidify the notion that violence against humanitarian actors will not be tolerated. As part of this drive, there are continued efforts to seek legal accountability (even with low expectations for tangible short-term results), engage in public advocacy (even despite uncertainty about the actual short-term impact) and maintain ‘red lines’ in negotiations (even in light of temptations to compromise). But when humanitarians do push for legal accountability, speak out publicly or push back too forcefully during negotiation processes, there is the risk of operational backlash. These efforts to resist what some practitioners see as the ‘new normal’ of humanitarian insecurity can actually increase security risks in certain operational contexts in the short-term.

The security management framework that this article presented – encompassing not just three approaches (as included in the security triangle) but rather six – can frame thinking about how the humanitarian sector continues to navigate this tension between long-term efforts and short-term imperatives. The discourse on security management has focused on the desirability of confronting threats through approaches – namely, ‘acceptance’ and ‘deterrence’ – geared towards transforming the external operational environment. But when these measures fall short, humanitarian organisations must consider the extent to which they themselves will change. The evident prevalence of ‘decoupling’ suggests an important disconnect between what humanitarian organisations aspire to be and what they actually are. The possibility of ‘synthesis’ points towards another avenue forward. This article has aimed to offer a more expansive security management framework to facilitate policy conversations about what is actually going on in the field, why these tensions are persisting and what alternatives to current approaches exist.

Acknowledgements

The interviews for this article were conducted under the auspices of the Harvard Humanitarian Initiative, with funding from the Swedish International Development Cooperation Agency, fund number: 210665. The authors wish to thank Anaïde Nahikian, who collaborated on designing the interview methodology and conducting the research interviews. Serena Hajjar, Meredith Blake, Edward Young and Jeremy Nelson provided useful support in transcribing interviews. The authors also thank Pauline Chetcuti and Meredith Blake for comments given on an earlier draft. The article also benefitted from feedback given during a presentation of an earlier iteration of this research at the ISSS-ISAC annual conference in 2017, convened in Washington, DC. Any errors are those of the authors.