All Lives Are Equal but Some Lives Are More Equal than Others

Staff Security and Civilian Protection in the Humanitarian Sector

in Journal of Humanitarian Affairs


In 2015, Action Contre la Faim launched a campaign calling on the UN to create a new post, that of a Special Rapporteur for the protection of humanitarian aid workers. Critics of the proposal claimed, inter alia, that creating such a post would imply that aid workers were a special category of civilians, worthy of protection over and above that accorded the wider population in the contexts in which they work.1 This raises an important issue which runs deeper than the campaign for a Special Rapporteur. The present article argues that, with or without such a post, the current situation is one in which humanitarian agencies treat aid workers as distinct and separate from the wider civilian population, and take significantly different measures for the safety of their staff from those they take for other civilians. For the most part, the distinction and associated differences are uncritically accepted, and this article sets out to challenge such acceptance by highlighting the nature of the differences, assessing possible explanations for the underlying distinction and considering its implications. Through this analysis, the article argues that this distinction not only reflects but also reinforces an unequal valuing of lives internationally.


In 2015, Action Contre la Faim launched a campaign calling on the UN to create a new post, that of a Special Rapporteur for the protection of humanitarian aid workers. Critics of the proposal claimed, inter alia, that creating such a post would imply that aid workers were a special category of civilians, worthy of protection over and above that accorded the wider population in the contexts in which they work.1 This raises an important issue which runs deeper than the campaign for a Special Rapporteur. The present article argues that, with or without such a post, the current situation is one in which humanitarian agencies treat aid workers as distinct and separate from the wider civilian population, and take significantly different measures for the safety of their staff from those they take for other civilians. For the most part, the distinction and associated differences are uncritically accepted, and this article sets out to challenge such acceptance by highlighting the nature of the differences, assessing possible explanations for the underlying distinction and considering its implications. Through this analysis, the article argues that this distinction not only reflects but also reinforces an unequal valuing of lives internationally.


In contemporary crises, a key aim of international humanitarian action is the protection of the civilian population. In the same contexts in which the protection needs of the local population are greatest, staff members of international humanitarian agencies may also come under threat themselves. Thus the organisations that seek to keep the local civilian population safe from physical violence are at the same time seeking to keep their own staff safe from physical violence. Despite the same broad objective, two distinct labels are used – ‘civilian protection’ and ‘staff security’ – and each designates a distinct set of policies and practices. Starting from the perspective that the reasons for such a distinction are not self-evident, the current article seeks to draw attention to the differences between staff-security and civilian-protection strategies, and to stimulate a conversation about the extent to which the differences are justified. The aim is not to argue for or against particular strategies for the safety of aid workers or the wider civilian population, or even to argue that the distinction between these two fields of practice should be removed, but rather to highlight and problematise this distinction, which is usually taken for granted.

Concerns about physical violence and safety are by no means new to international humanitarian agencies (Bugnion, 2003: 125–6; Taithe, 2016: 43–7). However, it is over the past thirty years that these concerns have been addressed by increasingly professionalised approaches (Gentile, 2011; Neuman, 2016a: 26-28; Stoddard et al., 2006: 21–35). The expansion and professionalisation of efforts to protect the local civilian population in contexts of armed conflict is evident in the range of policy statements, handbooks and guidelines (Global Protection Cluster Working Group, 2010; ICRC, 2008; InterAction, 2006; O’Callaghan and Pantuliano, 2007; Oxfam, 2005; Paul, 1999; Slim and Bonwick, 2005). Institutional staff-security policies also began to appear in the 1990s (Cutts and Dingle, 1995; ICRC, 1999). In 2000, the Humanitarian Practice Network published a Good Practice Review on Operational Security in Violent Environments (hereafter, GPR). Concerned not only with the safety of aid-agency staff but also with ensuring safe and continued programming and access, this became known as the ‘bible’ of humanitarian security, and an updated version was published in 2010 (Humanitarian Practice Network, 2010). Several more policy guidelines, manuals and documents focused on staff or staff and agency security have been published at the sector level over the past fifteen years (DG ECHO, 2006; Egeland et al., 2011; EISF, 2018; IASC, 2015; InterAction, 2015).

Efforts to keep humanitarian-agency staff and the wider civilian population safe in violent contexts thus evolved in parallel but as distinct fields of practice. All the policy documents cited above are concerned with one or the other, but not both, of staff security and civilian protection. This difference in language is so entrenched that even without specifying the referent object, ‘humanitarian security’ is used to designate efforts to keep aid agencies and their staff safe (Gassmann, 2005; Harmer, 2008; Humanitarian Practice Network and Security Management Initiative, 2010; Neuman and Weissman, 2016), while ‘humanitarian protection’ refers to efforts to keep the wider civilian population safe (Baines and Paddon, 2012; Dolan and Hovil, 2006; Pantuliano and Svoboda, 2016). The difference is not merely semantic. While staff members are a subset of the civilian population, and both staff security and civilian protection are about keeping people safe from violence, it is taken for granted that the strategies for keeping staff safe are different from those for the wider civilian population (see, for example, Humanitarian Practice Network, 2010: 2).

Starting from the perspective that this distinction sits uncomfortably with the equality inherent in the core humanitarian principle of humanity, and that the reasons for the distinction are not self-evident, I ask three questions. How does the policy and practice of ‘civilian protection’ differ from that of ‘staff security’? Why do they differ in this way? What are the consequences of this distinction? In addressing these questions, I draw on and contribute to a range of literature, not only on staff security and civilian protection but also on the nature and evolution of the humanitarian project more broadly.

The literature outlined above on best practices and institutional policies for staff security and civilian protection, together with work examining how these policies are implemented on the ground (see, for example, Beerli, 2018; Bradley, 2016; Hoffmann, 2017; Neuman, 2016b; Soussan, 2016; Sutton, 2018), provides much of the data for comparing ‘staff security’ and ‘civilian protection’ in this article. Such a comparison is novel, with existing comparisons focusing on the differential treatment of different categories of staff or of other civilians, but stopping short of comparing staff and other civilians. Larissa Fast, for example, laments the differential treatment accorded to refugees compared with the internally displaced, and to international staff compared with national staff, but says nothing of the differential treatment accorded to displaced persons on the one hand and staff on the other (Fast, 2015: 119, 127). The comparison in this article serves two purposes. First, comparing two phenomena helps us to better understand each phenomenon individually, and foregrounding the differences between them serves to illuminate the inclusions and exclusions in each field of practice. Second, in drawing attention to a distinction that is mostly uncritically accepted, the comparison aims to stimulate a conversation about its appropriateness.

A small body of literature asks why each of these fields of practice – staff security and civilian protection – is the way it is, with some work seeking to explain variation over time or across different agencies (see, for example, Bradley, 2016; Bradley, forthcoming; Neuman, 2016a; Schneiker, 2012; Taithe, 2016). I draw on this literature to identify likely explanations for the distinction between staff security and civilian protection, which I assess in order to argue that differential political constraints and opportunities are a key factor driving the differences between the two fields of practice. This article also builds on academic literature on the negative impacts of existing staff-security strategies on humanitarian action and on the limits of humanitarian approaches to the protection of civilians (see, for example, Bradley, 2016; Duffield, 2012; Fast, 2014; Ferris, 2011). However, even this more critical literature almost invariably focuses either on aid workers or on the wider civilian population, unquestioningly reproducing the hegemonic discourse by referring to ‘staff security’ and ‘civilian protection’. Rare exceptions consider the tensions between efforts to keep staff safe and the safety of the wider civilian population (Hoffmann, 2017; Sutton, 2018). The present article goes beyond identifying tensions between such efforts to contest their separation into two distinct fields of practice. In doing so, it contributes to critical humanitarian scholarship that highlights the tensions inherent within a humanitarian project that is centrally defined both by the idea of equality and by the idea of charity for distant others (Fassin, 2012; Harrell-Bond, 2002).

The article focuses on the work of the larger Western-based and civilian humanitarian actors – including UN agencies and international non-governmental organisations that are guided by the principles of humanity and impartiality. Such actors are not homogenous, and staff-security and civilian-protection practices vary across different agencies (Bradley, 2016; Schneiker, 2012). Nonetheless there are many commonalities, and this article focuses on general trends, albeit at the expense of some detail. ‘Staff security’ and ‘civilian protection’ comprise two clearly distinguished fields of practice within the broader field of international humanitarianism, with their own sets of handbooks, guidelines and best practices, and managed by their own professional staff. That said, security strategies are often different – and frequently less comprehensive – for national staff as compared with expatriate staff (Beerli, 2018; Stoddard et al., 2011). Even the categories of ‘national’ and ‘expatriate’ staff encompass a range of sub-categories across which there is variation in security strategies. Likewise, under the rubric of civilian protection, different measures may be taken for different segments of the civilian population, according to age, gender, legal status (e.g. refugees and internally displaced persons) and other characteristics (Carpenter, 2003; Dolan and Hovil, 2006). While the main line of comparison in this article is between safety measures taken for staff and those taken for the wider civilian population, I also draw attention to some of the differences within these two broad categories.

The rest of the article proceeds as follows. First, I compare the strategies and activities that constitute humanitarian-agency approaches to staff security and civilian protection to highlight the extent and nature of the differences between these two fields of practice. Second, I assess a range of plausible explanations for these differences and for the overall distinction. Third, I consider what the consequences of this distinction are both for aid workers and for the wider civilian population, for the relationship between humanitarian aid workers and the populations they seek to help, and for the implementation of humanitarian organisations’ social missions.

The Distinction between ‘Staff Security’ and ‘Civilian Protection’ in Policy and Practice

The grey literature on civilian protection identifies threat reduction and vulnerability reduction as protection objectives (IASC, 2016: 3; Slim and Bonwick, 2005: 52–3). Threat-reduction strategies aim to change the behaviour of those who would perpetrate violence, while vulnerability-reduction strategies implicitly accept a certain level of threat and aim to reduce people’s exposure to that threat. In addition, so-called remedial action to mitigate the consequences of violence after it has occurred is included as a component of civilian-protection strategies (Bradley, 2016: 119–26; IASC, 2016: 31). Policy guidance on staff security, and on operational security more broadly, commonly distinguishes three types of strategy: acceptance, deterrence and protection (Egeland et al., 2011; Humanitarian Practice Network, 2010). Acceptance aims to reduce the threat through soft measures, such as building relationships with local communities and stakeholders to obtain their consent for the agency’s presence and work (Humanitarian Practice Network, 2010: xv). Deterrence, by contrast, ‘attempts to deter a threat by posing a counter-threat, in its most extreme form through the use of armed protection’ (Humanitarian Practice Network, 2010: xvi). For staff, protection is classified a sub-category of security and is aimed at reducing vulnerability (Humanitarian Practice Network, 2010: xviii). Drawing on these typologies, I compare the policy and practice of ‘staff security’ with that of ‘civilian protection’ to assess how the balance between different kinds of objectives, and the measures taken in pursuit of those objectives, vary across these two fields of practice.

Threat Reduction

For most international humanitarian agencies, especially NGOs, acceptance is the preferred strategy for staff and organisational security, at least on paper. That said, it is not consistently well operationalised (Fast et al., 2013). Practices aimed at generating acceptance for humanitarian agencies and their staff include the use of emblems, logos, and agency T-shirts to distinguish themselves not only from military actors but also from other civilian actors, notably those that do not adhere to the principles of humanitarian action (Sutton, 2018). Acceptance strategies also include dialogue with various interlocutors. The International Committee of the Red Cross (ICRC) seeks to secure access and build relationships with a range of state and non-state interlocutors with the express purpose of engaging in a dialogue on civilian protection (Bradley, 2013; 2016: 162–8). By contrast, for many other agencies dialogue is used much more, if not exclusively, as part of an acceptance strategy to secure access for humanitarian staff and material aid (Bradley, 2016: 163; Collinson and Duffield, 2013: 18).

Many humanitarian agencies occasionally criticise authorities and other actors publicly. They may do so with respect to violent incidents or mistreatment of staff (a more adversarial, but still soft, acceptance strategy) or the wider civilian population, but in several cases attacks on staff have been publicly denounced while contemporaneous attacks on other civilians in the same country were not publicly commented upon (in Sri Lanka in 2009, for example; see Niland, 2014). When expatriate or national staff members are killed, the agency that employed them will invariably make a statement, but they do not always do so when other (non-fatal) staff-security incidents occur, and only rarely do they make public statements when other civilians are killed. Even in Médecins sans Frontières (MSF), the agency with témoignage at the heart of its mandate, the role of public statements has been contested among the senior management, and MSF mostly undertakes its medical work without making public statements about abuses in its zones of operations (Weissman, 2011). Both the ICRC and MSF have also run global campaigns concerned with attacks on medical missions and healthcare facilities.2 While framed primarily in terms of protecting people’s access to medical and healthcare services, the absence of campaigns of a comparable scale concerned with attacks on other civilians is notable.

While most agencies favour staff-security strategies centred on acceptance in theory, many – especially UN agencies – rely on harder strategies in practice (Collinson and Duffield, 2013: 12). For example, staff may travel under armed escort. Aid agencies are somewhat secretive about their use of armed escorts, making it difficult to assess how frequently they use them. However, research on private security firms, for example, suggests they are used more widely than generally recognised for purposes of staff security but are not contracted by either donor governments or operational humanitarian agencies to protect the wider civilian population (Singer, 2006). In some contexts, the government insists on humanitarian agencies travelling with military escorts as a condition of access to particular areas (Bradley, 2016: 148). Where UN peacekeeping forces are present, UN agencies (and sometimes international NGOs) may travel to unsecure areas under their escort, as has been the case with UNHCR in the Democratic Republic of Congo, for example (Bradley, 2016: 61). Notably, the main guidelines on the use of armed escorts by humanitarians specifically exclude consideration of the transportation of the wider civilian population (IASC, 2013: 2).

With the exception of the ICRC, most of what humanitarian agencies do under the rubric of civilian protection is not aimed at changing the behaviour of armed actors to reduce the threat they pose (Bradley, 2016: 101–28; DuBois, 2010). Despite the fact that aid agencies seek to reduce the threats faced by their staff, and that civilians themselves adopt threat-reduction strategies (Baines and Paddon, 2012; South et al., 2012), it is argued that there ‘is little humanitarian actors can do to eliminate the protection threats faced by those they seek to help’ (Jackson, 2014: 3). To the extent that agencies other than the ICRC seek to reduce the threat posed to the wider civilian population, this is often through ‘protection by presence’, based on the idea that the witnessing of abuses of the civilian population by staff of an international agency disincentivises such abuses. Such an approach is unlikely to be effective in contexts where staff themselves are targets of attack, and it is different from the strategies aimed at gaining acceptance for agency staff and operations in that it is passive rather than proactive.

Vulnerability Reduction

Both staff-security and civilian-protection strategies include measures aimed at reducing vulnerability. However, the forms of vulnerability reduction envisaged for staff are more proactive than those generally employed for the wider civilian population. Vulnerability reduction for staff is about ‘hardening the target’ (Egeland et al., 2011: 28), whereas reducing the vulnerability of civilians does not seek to ‘harden’ anything, and nor does the language of civilian protection refer to ‘targets’. In line with these differences, people with very different professional profiles are responsible for staff security as compared with civilian protection. An advert for a ‘Staff Safety and Resilience Manager’ for World Vision in South Sudan, for example, lists ‘police/law enforcement or military training’ among the requirements (World Vision, 2018). Military-consulting firms have been hired by humanitarian agencies to conduct threat assessments and to provide advice and training on staff security to international humanitarian agencies (Singer, 2006). By contrast, the skills required for protection managers and officers are much softer and do not include military expertise.

In many contexts, aid workers live and work in compounds fortified against outside attack (Collinson and Duffield, 2013). Perhaps the closest equivalent to fortification for the wider civilian population would be safety zones, which have been created relatively infrequently and with limited success – and sometimes deadly failure (Landgren, 1995). For the most part, locally hired staff do not live in agency compounds, so they only benefit from this kind of protection during working hours, and the wider civilian population does not benefit from it at all. The Protection of Civilians (PoC) sites in South Sudan are the exception that proves the rule, both because they are unique to that context and because they were not an initiative of international humanitarian agencies (Murphy, 2017).

Evacuation to a safer country is perhaps the ultimate in reducing someone’s vulnerability. The GPR includes a chapter on ‘evacuation, hibernation, remote management programming and return’, and evacuation is part of the standard toolkit for staff security, particularly expatriate staff security. Almost all major international humanitarian agencies have at some point used remote management, which involves the withdrawal of international (and sometimes national) staff to oversee activities from a different location, implying the transfer of responsibility – and often risk – to local staff or local partner organisations (Stoddard et al., 2010). International humanitarian agencies have also evacuated civilians relatively frequently in recent years, mainly from siege environments (Norwegian Refugee Council, 2016: 5). However, such evacuations are usually to another site in the same country, and the expectation is that local civilians should be evacuated in-country (Norwegian Refugee Council, 2016: 23; UNHCR, n.d.). On the other hand, civilians frequently ‘evacuate’ themselves, a phenomenon more commonly called displacement, and international agencies provide support (camps, shelter, material assistance) to displaced civilians, indirectly facilitating displacement. Facilitating the displacement of civilians internally or within the region can reduce their vulnerability, but they often continue to face significant insecurity during and after displacement (Bradley, 2017: 18–19). Moreover, such support is reactive – responding ex post to the movement of people – rather than proactive as in the case of evacuations.

Beyond facilitating displacement, much vulnerability reduction in civilian protection focuses on providing material assistance in such a way as to enable civilians to reduce their exposure to threats, as in the often cited example of providing fuel-efficient stoves to reduce women’s and girls’ exposure to threats of sexual violence while collecting firewood (Ferris, 2011: 108; O’Callaghan and Pantuliano, 2007: 35; Slim and Bonwick, 2005: 89). Projects focused on income-generating activities may reduce the need for dangerous travel or otherwise risky livelihoods strategies (Bradley, 2016: 182–3). Soft measures taken to reduce the vulnerability of humanitarian-agency staff include restrictions on what staff – especially but not exclusively international staff – can do in their free time, including curfews and no-go locations (Beerli, 2018). Interestingly, then, reducing vulnerability for the wider civilian population often involves increasing the options open to them to avoid threats, while vulnerability reduction for staff may involve restricting their options.

Remedial Action to Mitigate the Consequences of Violence after an Attack Has Occurred

Civilian-protection strategies often include a number of activities that are ‘reactive, rather than proactive, responsive rather than preventive’, aimed at mitigating the consequences of violent incidents after they have occurred (Bradley, 2016: 119). Reflecting this, protection analysis tends to be backward-looking, focused on what has happened, rather than on predicting what will happen (Bradley, 2016: 157). By contrast, the kind of analysis emphasised in staff security is forward-looking risk assessment (Humanitarian Practice Network, 2010: 29). Remedial action for the wider civilian population frequently includes the provision of psychosocial aftercare to the victims of violence, especially to victims of sexual violence, and international humanitarian agencies are criticised for not providing the same kind of support to staff members who have suffered violence or trauma in the course of their work (Hughes, 2015). This is not to suggest these agencies are providing more care to victims of violence within the wider population than they do to their own staff – indeed, such care is probably insufficient across the board – but rather to highlight the fact that this is often a central component in civilian-protection strategies, in the absence of the more proactive measures that are central to staff security. Moreover, for staff, such ‘aftercare’ is most often conceptualised as part of well-being rather than safety and comes under the responsibility of human resources rather than security management.

Table 1 summarises this analysis. Staff-security strategies seek to reduce the number of violent incidents directed at aid workers at the same time as they seek to reduce exposure to residual threats. By contrast, civilian protection strategies largely accept that violence will be directed against the wider civilian population and seek to reduce risk only through reducing the exposure of civilians to that violence. To the extent that both staff-security and civilian-protection strategies seek to reduce vulnerability or exposure to violence, the efforts of international humanitarian agencies to reduce the vulnerability of their staff often involve harder measures, while efforts to reduce the vulnerability of the wider civilian population are mostly limited to softer measures. Furthermore, much of what is done in the name of civilian protection is not about reducing either threat or vulnerability but about mitigating the consequences after civilians have suffered violence.

Table 1Analysis of staff-security and civilian-protection strategies
Staff-security strategiesCivilian-protection strategies
Threat reduction – changing the behaviour of armed actors to reduce the likelihood of violent incidentsPublic advocacy and private dialogue are central to acceptance strategies

Hard measures, e.g. armed escorts, are relatively common in some agencies
Limited public advocacy and private dialogue

Soft measures, e.g. passive protection through presence
Vulnerability reduction – reducing people’s exposure to threatsHard measures, e.g. fortified aid compoundsSoft measures, e.g. the provision of material assistance and legal status
Mitigating the consequences – providing support to the victims of violence after an incident has occurredLimited support after attacksSupport after attacks is a major component

Explaining the Distinction

In this section, I assess different plausible explanations for the distinction and differences described above. I argue that differences in threats and vulnerabilities are insufficient to explain the distinct approaches taken to staff security and civilian protection. Legal obligations and political opportunities can account for some of the differences but do not tell the whole story. Other differences are best explained as a consequence of differential tolerance of casualties.

Distinct Threats and Vulnerabilities

Different categories of civilians face different threats in different contexts and may be characterised by different vulnerabilities. For example, land activists have been particularly targeted in Colombia, and women are at greater risk of sexual violence than men in the Democratic Republic of Congo. Within the category of humanitarian staff, different individual profiles are exposed to different levels of risk according to, inter alia, age, ethnicity, gender, nationality and sexuality (EISF, 2018). In some cases, as in the bombings of ICRC and UN headquarters in Iraq in 2003, aid agencies and their staff are specifically targeted, and this could explain singling staff out from the rest of the civilian population on a case-by-case basis. However, it is not evident that the category of ‘staff’ is always and everywhere subject to a more distinct set of threats and vulnerabilities than other categories of civilian, such as could account for the wholesale distinction described in the previous section.

The extent to which humanitarian work is getting more dangerous, and aid workers increasingly targeted for attack, is contested (Fast, 2014; Stoddard et al., 2009; Weissman, 2016). There is a widespread perception that aid work is increasingly dangerous, but the differences between staff security and civilian protection do not seem to be driven by beliefs about differential levels of threat. Analysis of the data on attacks on aid workers mostly focuses on trends in the absolute numbers of attacks, comparing them across time and across country operating contexts.3 To the extent that comparisons are made based on estimates of the rates of attacks on aid workers, these comparisons are with on-the-job death rates in hazardous civilian professions in the United States (Stoddard et al., 2006: 4). Comparisons are not made with the risks faced by the wider civilian population. More generally, we know that civilian-protection activities are often targeted based on standardised categories of vulnerability rather than context-specific vulnerability analysis, and that national staff are frequently offered lesser protections than expatriate staff, even where they face greater risks (Carpenter, 2003; Dolan and Hovil, 2006; Stoddard et al., 2011).

Furthermore, a 2011 UN Office for the Coordination of Humanitarian Affairs (OCHA) report on ‘good practice for humanitarians in complex security environments’ acknowledges that local civilians ‘suffer most from conflict and violence’ and makes the case for aid-worker security measures not in terms of differential threat or vulnerability but in terms of the instrumental value of humanitarian staff: ‘where aid workers are attacked, the quality and quantity of aid is reduced, and beneficiaries suffer’ (Egeland et al., 2011: 5). Such a perspective is representative of the dominant narrative on aid-agency security, which focuses on mitigating risks to enable operations to continue or expand (Collinson and Duffield, 2013: 9). However, to the extent that staff-security strategies are better or more comprehensive than civilian-protection strategies, even this kind of formulation, which justifies additional measures for staff on the basis of the services they provide for others, sits uncomfortably with the principle of humanity, according to which ‘human value is based on life not utility’ (Slim, 2015: 56).

Legal Frameworks

Differences in legal status do not offer a convincing explanation of differences in security/protection strategy. International law does offer additional protections to specific categories of humanitarian staff under certain conditions, but for many agencies and their staff international law affords no protection beyond that afforded to the general civilian population (Mackintosh, 2007). Furthermore, there is no shortage of international law on the protection of civilians in which aid agencies can ground dialogue and advocacy, and yet many grave violations of international human rights and humanitarian law against the wider civilian population occur without most humanitarian agencies discussing these violations either in private dialogue with alleged perpetrators or in public communications. Moreover, as part of their acceptance strategies, international humanitarian agencies sometimes actively distinguish themselves from others with the same legal status (Sutton, 2018).

The legal obligations inherent in the notion of ‘duty of care’, and the risk of legal action against agencies that fail in their duty of care, have shaped staff-security strategies (Edwards and Neuman, 2016). They may explain why harder security measures, greater recourse to evacuation and a more forward-looking approach focused on assessing and managing risks characterise staff-security strategies as compared with civilian-protection strategies. They may also explain some of the differences in the measures taken for expatriate staff as compared with national staff. A duty of care requires employers to take all reasonable measures to ensure the safety of their staff and to avoid the risk of foreseeable injury. On at least one interpretation, an agency’s duty of care extends only to risks that arise from an individual’s employment with that agency, and this could exclude many of the risks that locally hired staff face (EISF, 2018: 14). There is no equivalent legal obligation to protect other civilians – even for those agencies with explicit protection mandates (e.g. ICRC and UNHCR). Indeed, many of the measures adopted with the intention of increasing accountability to affected populations – which in any case do not generally include legal mechanisms – pay scant attention to protection (Terry, 2000).

Political Constraints and Opportunities

In addition to differences in legal opportunities and obligations, there exist differences in political opportunities and constraints. There is more opportunity to evacuate expatriate staff internationally because they are fewer in number than the local civilian population and, crucially, they have the possibility of living and working elsewhere. Whereas expatriate staff by definition hold a passport of another country, evacuating large numbers of the local civilian population requires other states to accept them onto their territory, and such acceptance is rarely forthcoming. The fact that the evacuation of staff is generally understood to imply international evacuation (with the term ‘relocation’ more commonly used for in-country evacuation) while evacuations of civilians by humanitarian agencies tend to be within-country supports the idea that differential political constraints may be part of the explanation for this particular difference between staff security and civilian protection.

Differential political constraints and opportunities also go some way to explaining differential use of armed escorts. Sometimes the government of the affected state or, in the case of UN humanitarian agencies, the UN Department of Safety & Security insists on the use of armed escorts for travel to certain areas. Agencies may therefore accept escorts not because of security concerns but because they consider it a price worth paying for access to populations in need. That said, the ICRC has only rarely resorted to armed escorts, and in many contexts has better access than other agencies. Differential opportunities – or differential expectations about effectiveness – may also explain the differences observed in the use of advocacy and public criticism of authorities. Public criticism of those who target aid workers may be expected to attract more international attention and pressure on the perpetrators of violence than public criticism in response to attacks on other civilians.

To the extent that differences in political constraints and opportunities account for the differences between staff-security and civilian-protection strategies, the distinction between them can be understood as driven by a differential valuing of lives at the level of geopolitics. In other words, even if humanitarian agencies themselves value all lives equally, their actions are shaped by a world which does not.

Additional Explanations

Greater obligations and opportunities with respect to staff security as compared with the wider civilian population may explain the harder security measures and more frequent advocacy on behalf of aid workers, but analysis of policy guidance and discourse on evacuations and advocacy suggests that political constraints and opportunities are not the whole story. There seems to be an implicit understanding of violence against local civilians as normal and of violence against aid workers as exceptional.

Guidelines on evacuations of both staff and civilians advise consideration of the political impacts of evacuating people, but different issues are highlighted. Where evacuating staff leads to the suspension of programming, it is seen as sending a message about the severity of the situation to the authorities and the international community (Humanitarian Practice Network, 2010: 86). Relocating large numbers of civilians could also send a message about the severity of the situation, but when it comes to checklists of costs and benefits to be analysed as part of the decision-making process before evacuating the local civilian population, this point is not made. Instead, the risk of supporting the war aims of a party to conflict and, in the worst cases, contributing to ethnic cleansing is highlighted (Norwegian Refugee Council, 2016: 38–41). In many contexts, the evacuation of staff may also further war aims, but this point is absent from discussions of the pros and cons of staff evacuations. In sum, policy guidance on evacuating staff asks decision-makers to consider the potential positive political effects of evacuation, while policy guidance on evacuating civilians asks decision-makers to consider the potential negative political effects of evacuation; there is thus a bias in favour of the evacuation of staff and a bias against the evacuation of other civilians.

Analysis of the discourse around dialogue and advocacy decisions also suggests different thresholds of tolerance of violence. In terms of civilian protection, the arguments in favour of public criticism of actors who have abused human rights or violated international humanitarian law (IHL) are twofold. It is hoped that ‘naming and shaming’ will encourage perpetrators to improve their conduct towards civilians, and there is additionally a concern that remaining silent in the face of abuses implies some kind of complicity with those abuses. Against such public criticism are concerns about access and staff security. According to GPR, the ‘pursuit and preservation of acceptance may require that agencies stay silent about humanitarian or human rights abuses’ (Humanitarian Practice Network, 2010: 68). The same arguments can be made about staff, and yet we consistently see public statements made in response to the killing of aid workers and not in response to the killing of other civilians. Negotiations with armed actors tell a similar story: they are deemed essential for staff security and operational access but only ‘nice to have’ for civilian protection. In short, a certain level of abuse of the local civilian population can be tolerated, but the deaths of aid workers cannot.

Insights from the Comparison, and Consequences of the Distinction

In this final section, I consider what the consequences the distinction between staff security and civilian protection has for staff, other civilians and the humanitarian project more broadly. I argue that, just as many of the differences between staff security and civilian protection can be explained by the constraints imposed and opportunities offered by a world which values some lives more than others, these same differences serve to reinforce this unequal valuing of lives internationally. Furthermore, I argue that the differences in practice driven by an ‘us and them’, ‘strong and weak’, ‘saviour and victim’ mentality serve also to reinforce these dichotomies to the likely detriment of the safety and well-being of both humanitarian-agency staff and other civilians.

Scholarship on humanitarian-agency efforts to protect civilians does not usually advocate hard security measures for the local civilian population, and for very good reasons, but the lack of hard measures is explicitly or implicitly part of the explanation for the limits of such efforts (see, for example, Ferris, 2011). Critical literature on staff security points to the risk of fortification and the use of armed security guards contributing to a normalisation and, potentially, an escalation of violence (Collinson and Duffield, 2013: iv, 19–22). Furthermore, several authors argue that the increasing resort to hard security measures and fortified aid compounds has led to the ‘bunkerisation’ of aid and the paradox that aid agencies gain or maintain access in insecure environments at the same time as (especially expatriate) personnel are distanced from those they seek to assist (Collinson and Duffield, 2013; Duffield, 2012; Fast, 2014).

The inclusion of hard measures for staff at the same time as they are excluded for other civilians can be seen to have two additional consequences. First, providing armed protection to some people – staff – may not only reduce the risk faced by those people but may also serve to increase the risk by those who do not receive such protection – the wider civilian population (Singer, 2006: 117). In addition, where public resources, e.g. police, are used for the close protection of aid agencies and their staff, those resources may be diverted from the maintenance of general security, to the detriment of the civilian population. Second, such differential measures create not only physical but also psychological or emotional distance between staff and those they seek to assist. Without convincing evidence that aid agencies and their staff are subject to different and greater threats than the rest of the civilian population, employing hard security measures for their safety but not for anyone else sends a message that the lives of aid workers are more important or of greater value than everyone else’s. Such messages do not go unnoticed (see, for example, Schneiker, 2012: 255).

The differential approach to making public statements when aid workers are killed and only occasionally doing so to protest the killing of other civilians sends a similar message. One strand of the academic literature on civilian protection critiques the failure of humanitarian agencies to take a more political stance and/or to speak out against abuses of civilians (Hart and Lo Forte, 2013; Niland, 2014). The general argument is based on the assumption that public criticism would shame those abusing human rights or violating IHL to improve their conduct towards civilians. In fact, little is known about the conditions under which such advocacy is effective (Bradley, 2014). However, it is clear that the publicisation of atrocities against some people (staff) and not others (local civilians) sends problematic messages.

Speaking out about abuses of staff but not of the civilian population more broadly may be taken by the wider international community to imply that there is no mistreatment of the civilian population (or at least that it is less serious than it in fact is), making political efforts to address civilian targeting less likely. Calling out some killings and not others can also send a message to perpetrators of violence that it is more acceptable to target some people than others. When MSF emphasises, as part of the #notatarget campaign, that hospitals, health workers and patients must not be bombed, do they not also imply that bombing other civilian installations and persons is somehow more acceptable? More generally, differential advocacy strategies send the message that some lives are worth more than others or that some deaths are more surprising and newsworthy than others. Through strategies to distinguish themselves from other civilian actors in conflict contexts, humanitarian aid workers are arguably constructing a ‘civilian plus’ category for themselves, to the potential detriment of those who are ‘mere civilians’ (Sutton, 2018).

In adapting to the opportunities, obligations and constraints facing them, rather than challenging the ways in which geopolitics implicitly assigns more value to some lives than others, humanitarian agencies end up reinforcing the differential valuing of lives.


The humanitarian project is revolutionary in the sense that it seeks to inject an ethic of valuing all lives equally into the very contexts where some lives are considered by many to be expendable. Thus Hugo Slim writes of the ‘radical equality’ that characterises the principle of humanity that motivates humanitarian action (Slim, 2015: 56). Yet the analysis of staff security and civilian protection in this article shows that, in professionalising these two fields of practice, the humanitarian sector has come to reinforce the very inequality it seeks to challenge. A paradox lies at the heart of international humanitarianism, a project that is centrally defined by equality and universality at one and the same time as the idea of helping and saving distant others implies a clear distinction and inequality between ‘the life that is saved (that of the victims), and the life that is risked (that of the aid workers)’ (Fassin, 2012: 231).

To the extent that the practice of staff security is ‘better’, more comprehensive or more effective than the practice of civilian protection, for example, we can reasonably ask whether civilian protection could and should look more like staff security. To be sure, there are many sound arguments against adopting harder measures to protect the wider civilian population. What is not clear is why these arguments are any less sound with respect to aid workers. The principle of humanity does not imply always taking the same measures in pursuit of the safety of both staff and other civilians, but it surely requires differential measures to be justified in terms of differential threats or vulnerabilities. Where no such justification exists, and differential measures are adopted due to differential obligations and opportunities, those obligations and opportunities (or the lack of them with respect to some civilians) are worth highlighting and challenging.



See and






Baines, E. and Paddon, E. (2012), ‘“This Is How We Survived”: Civilian Agency and Humanitarian Protection’, Security Dialogue, 43:3, 23147. doi: 10.1177/0967010612444150.

Beerli, M. J. (2018), ‘Saving the Saviors: Security Practices and Professional Struggles in the Humanitarian Space’, International Political Sociology, 12:1, 7087. doi: 10.1093/ips/olx023.

Bradley, M. (2013), ‘International Humanitarian Law, Non-State Armed Groups and the International Committee of the Red Cross in Colombia’, Journal of International Humanitarian Legal Studies, 4:1, 10834.

Bradley, M. (2014), ‘Sri Lanka: Limited Humanitarian Action – Or a Lesson in the Limits of Humanitarian Action?’, International Development Policy, 6:1, 14954.

Bradley, M. (2016), Protecting Civilians in War: The ICRC, UNHCR, and Their Limitations in Internal Armed Conflicts. Oxford: Oxford University Press.

Bradley, M. (2017), ‘The Impact of Armed Conflict on Displacement’, report commissioned by Concern Worldwide. doi: 10.13140/RG.2.2.33905.94562/1.

Bradley, M. (forthcoming), ‘Unintended Consequences of Adjacency Claims: The Function and Dysfunction of Analogies between Refugee Protection and IDP Protection in the Work of UNHCR’, Global Governance, 25:3.

Bugnion, F. (2003), The International Committee of the Red Cross and the Protection of War Victims. Oxford: Macmillan.

Carpenter, R. C. (2003), ‘“Women and Children First”: Gender, Norms, and Humanitarian Evacuation in the Balkans 1991–95’, International Organization, 57:4, 66194. doi: 10.1017/S002081830357401X.

Collinson, S. and Duffield, M. (2013), Paradoxes of Presence: Risk Management and Aid Culture in Challenging Environments. London: Humanitarian Policy Group/Overseas Development Institute.

Cutts, M. and Dingle, A. (1995), Safety First: Protecting NGO Employees Who Work in Areas of Conflict. London: Save the Children.

DG ECHO. (2006), NGO Security Collaboration Guide. Brussels: Directorate-General for European Civil Protection and Humanitarian Aid Operations.

Dolan, C. and Hovil, L. (2006), Humanitarian Protection in Uganda: A Trojan Horse? London: Humanitarian Policy Group/Overseas Development Institute.

DuBois, M. (2010), ‘Protection: Fig-Leaves and Other Delusions’, Humanitarian Exchange, 46, 24.

Duffield, M. (2012), ‘Challenging Environments: Danger, Resilience and the Aid Industry’, Security Dialogue, 43:5, 47592, doi: 10.1177/0967010612457975.

Edwards, J. and Neuman, M. (2016), ‘Who Benefits from “Duty of Care”?’, in Neuman, M. and Weissman, F. (eds), Saving Lives and Staying Alive: Humanitarian Security in the Age of Risk Management (London: Hurst), pp. 826.

Egeland, J., Harmer, A. and Stoddard, A. (2011), To Stay and Deliver: Good Practice for Humanitarians in Complex Security Environments. New York: UN OCHA.

EISF. (2018), Managing the Security of Aid Workers with Diverse Profiles. London: European Interagency Security Forum.

Fassin, D. (2012), Humanitarian Reason: A Moral History of the Present. Berkeley, CA: University of California Press.

Fast, L. (2014), Aid in Danger: The Perils and Promise of Humanitarianism. Philadelphia, PA: University of Pennsylvania Press.

Fast, L. (2015), ‘Unpacking the Principle of Humanity: Tensions and Implications’, International Review of the Red Cross, 97:897–8, 11131. doi: 10.1017/S1816383115000545.

Fast, L., Freeman, C. F., O’Neill, M. and Rowley, E. (2013), ‘In Acceptance We Trust? Conceptualising Acceptance as a Viable Approach to NGO Security Management’, Disasters, 37:2, 22243.

Ferris, E. (2011), The Politics of Protection: The Limits of Humanitarian Action. Washington, DC: Brookings Institution Press.

Gassmann, P. (2005), ‘Rethinking Humanitarian Security’, Humanitarian Exchange, 30, 324.

Gentile, P. (2011), ‘Humanitarian Organizations Involved in Protection Activities: A Story of Soul-Searching and Professionalization’, International Review of the Red Cross, 93:884, 116591. doi: 10.1017/S1816383112000343.

Global Protection Cluster Working Group (2010), Handbook for the Protection of Internally Displaced Persons. Geneva: Global Protection Cluster Working Group.

Harmer, A. (2008), ‘Integrated Missions: A Threat to Humanitarian Security?’, International Peacekeeping, 15:4, 52839. doi: 10.1080/13533310802239824.

Harrell-Bond, B. (2002), ‘Can Humanitarian Work with Refugees Be Humane?’, Human Rights Quarterly, 24:1, 5185.

Hart, J. and Lo Forte, C. (2013), ‘Mandated to Fail? Humanitarian Agencies and the Protection Of Palestinian Children’, Disasters, 37:4, 62745.

Hoffmann, S. (2017), ‘Humanitarian Security in Jordan’s Azraq Camp’, Security Dialogue, 48:2, 97112. doi: 10.1177/0967010616683311.

Hughes, R. (2015), A Crisis of Anxiety among Aid Workers, New York Times, 8 March, (accessed 2 May 2018).

Humanitarian Practice Network. (2010), Operational Security Management in Violent Environments. London: Overseas Development Institute.

Humanitarian Practice Network and Security Management Initiative (eds) (2010), Humanitarian Exchange, 47, special issue.

IASC (2013), IASC Non-Binding Guidelines on the Use of Armed Escorts for Humanitarian Convoys. Geneva and New York: Inter-Agency Standing Committee.

IASC (2015), Saving Lives Together: A Framework for Improving Security Arrangements among International Non-Governmental Organisations/International Organisations and the United Nations. Geneva and New York: Inter-Agency Standing Committee.

IASC (2016), IASC Policy on Protection in Humanitarian Action. Geneva and New York: Inter-Agency Standing Committee.

ICRC (1999), Staying Alive: Safety and Security Guidelines for Humanitarian Volunteers in Conflict Areas. Geneva: International Committee of the Red Cross.

ICRC (2008), ‘ICRC Protection Policy: Institutional Policy’, International Review of the Red Cross, 90: 871, 75175.

InterAction (2006), Protection in Practice: A Guidebook for Incorporating Protection into Humanitarian Operations. Washington, DC: InterAction.

InterAction (2015), Minimum Operating Security Standards. Washington, DC: InterAction.

Jackson, A. (2014), ‘Protecting Civilians: The Gap between Norms and Practice’, in HPG Policy Brief 56. London: Humanitarian Policy Group/Overseas Development Institute.

Landgren, K. (1995), ‘Safety Zones and International Protection: A Dark Grey Area’, International Journal of Refugee Law, 7:3, 43658.

Mackintosh, K. (2007), ‘Beyond the Red Cross: The Protection of Independent Humanitarian Organizations and Their Staff in International Humanitarian Law’, International Review of the Red Cross, 89:865, 11330. doi: 10.1017/S1816383107000951.

Murphy, R. (2017), ‘The United Nations Mission in South Sudan and the Protection of Civilians’, Journal of Conflict and Security Law, 22:3, 367394. doi: 10.1093/jcsl/krx016.

Neuman, M. (2016a), ‘On Danger, Sacrifice and Professionalisation: MSF and the Security Debate’, in Neuman, M. and Weissman, F. (eds), Saving Lives and Staying Alive: Humanitarian Security in the Age of Risk Management (London: Hurst), pp. 2136.

Neuman, M. (2016b), The Duty of a Head of Mission: Interview with Delphine Chedorge, MSF Emergency Coordinator in Central African Republic, in Neuman, M. and Weissman, F. (eds), Saving Lives and Staying Alive: Humanitarian Security in the Age of Risk Management (London: Hurst), pp. 89102.

Neuman, M. and Weissman, F. (eds) (2016), Saving Lives and Staying Alive: Humanitarian Security in the Age of Risk Management. London: Hurst.

Niland, N. (2014), ‘Sri Lanka: Unrestricted Warfare and Limited Humanitarian Action’, International Development Policy. doi: 10.4000/poldev.1680.

Norwegian Refugee Council (2016), NRC Considerations for Planning Mass Evacuations of Civilians in Conflict Settings. Geneva: Norwegian Refugee Council.

O’Callaghan, S. and Pantuliano, S. (2007), Protective Action: Incorporating Civilian Protection into Humanitarian Response. London: Humanitarian Policy Group/Overseas Development Institute.

Oxfam (2005), Protection in Practice: An Introduction. Oxford: Oxfam.

Pantuliano, S. and Svoboda, E. (2016), ‘Humanitarian Protection – Moving beyond the Tried and Tested’, in Wilmot, H., Mamiya, R., Sheeran, S. and Weller, M. (eds), Protection of Civilians (Oxford: Oxford University Press), pp. 37290.

Paul, D. (1999), Protection in Practice: Field-Level Strategies for Protecting Civilians from Deliberate Harm. London: Relief and Rehabilitation Network, Overseas Development Institute.

Schneiker, A. (2012), ‘The Vulnerable Do‐Gooders: Security Strategies of German Aid Agencies’, Disasters, 37:2, 24466. doi: 10.1111/j.1467-7717.2012.01300.x.

Singer, P. W. (2006), ‘Humanitarian Principles, Private Military Agents: Implications of the Private Military Industry for the Humanitarian Community’, Brown Journal of World Affairs, 13:1, 10522.

Slim, H. (2015), Humanitarian Ethics: A Guide to the Morality of Aid in War and Disaster. London: Hurst.

Slim, H. and Bonwick, A. (2005), Protection: An ALNAP Guide for Humanitarian Agencies. London: Active Learning Network for Accountability/Overseas Development Initiative.

Soussan, J. (2016), ‘Qabassin, Syria: Security Issues and Practices in an MSF Mission in the Land of Jihad’, in Neuman, M. and Weissman, F. (eds), Saving Lives and Staying Alive: Humanitarian Security in the Age of Risk Management (London: Hurst), pp. 10926.

South, A., Harragin, S., Corbett, J., Horsey, R., Kempel, S., Fröjmark, H. and Carstensen, N. (2012), Local to Global Protection in Myanmar (Burma), Sudan, South Sudan and Zimbabwe, HPN Network Paper 72. London: Humanitarian Policy Group/Overseas Development Initiative.

Stoddard, A., Harmer, A. and DiDomenico, V. (2009), Providing Aid in Insecure Environments: 2009 update, in HPG Policy Brief 34. London: Overseas Development Institute.

Stoddard, A., Harmer, A. and Haver, K. (2006), Providing aid in insecure environments: trends in policy and operations. In HPG Report 23. London: Overseas Development Institute.

Stoddard, A., Harmer, A. and Haver, K. (2011), Aid Worker Security Report 2011. Spotlight on Security for National Aid Workers. London: Humanitarian Outcomes.

Stoddard, A., Harmer, A. and Renouf, J. S. (2010), Once Removed: Lessons and Challenges in Remote Management of Humanitarian Operations for Insecure Areas. London and New York: Humanitarian Outcomes.

Sutton, R. (2018), ‘The International Humanitarian Actor as “Civilian Plus”: The Circulation of the Idea of Distinction in International Law(PhD thesis), London School of Economics and Political Science (LSE), (accessed 3 February 2019).

Taithe, B. (2016), Danger, Risk, Security and Protection: Concepts at the Heart of the History of Humanitarian Aid’, in Neuman, M. and Weissman, F. (eds), Saving Lives and Staying Alive: Humanitarian Security in the Age of Risk Management (London: Hurst), pp. 3752.

Terry, F. (2000), The Limits and Risks of Regulation Mechanisms for Humanitarian Action. Paris: MSF CRASH.

UNHCR (n.d.), Humanitarian Evacuations in Violence and Armed Conflict. Geneva: United Nations High Commissioner for Refugees.

Weissman, F. (2011), ‘Silence heals…. From the Cold War to the War on Terror, MSF Speaks Out: A Brief History’, in Magone, C., Neuman, M. and Weissman, F. (eds), Humanitarian Negotiations Revealed: The MSF Experience (London: Hurst), pp. 17797.

Weissman, F. (2016), ‘Violence against Aid Workers: The Meaning of Measuring’, in Neuman, M. and Weissman, F. (eds), Saving Lives and Staying Alive: Humanitarian Security in the Age of Risk Management. London: Hurst.

World Vision (2018), Staff Safety and Resilience Manager, (accessed 10 May 2018).



All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 132 132 132
PDF Downloads 89 89 89

Related Content