Michaël Neuman and Fabrice WeissmanTranslated from French by Nina Friedman.

In 2013 the French section of Médecins Sans Frontières (MSF) created the position of “security focal point,” tasked with developing guidelines, procedures, a database and training courses for security management. In so doing, the section joined the rest of the MSF movement, which, like other large humanitarian organisations and corporations, had already set up safety and security units or departments dedicated to risk prevention and management.

Encouraged by Western donors, the field of security expertise for NGOs and UN agencies took off in the mid-1990s,See Claude Bruderlein and Pierre Gassmann, “Managing Security Risks in Hazardous Missions: The Challenges of Securing United Nations Access to Vulnerable Groups”, Harvard Human Rights Journal, vol. 19, 2006, pp. 63-93.
 resulting in the creation of the posts of security advisor and risk manager, both at headquarters and in the field. These were, to begin with, filled mainly by former military and police personnel.Ibid.
 These experts progressively set up regional coordination platformsSuch as the Afghanistan NGO Safety Office (ANSO), created in 2002; the NGO Coordination Committee for Iraq (NCCI), formed in 2003; the NGO Safety Project for Somalia, set up in 2004; and the Gaza NGO Safety Office (GANSO), created in 2007.
 and professional networksAmong them the International NGO Safety and Security Association (INSSA) in the United States (www.ingossa.org) and the European Interagency Security Forum (EISF) in Europe (www.eisf.eu).
 to develop standards, databases, manuals and training programmes with courses and self-instruction modules designed for all humanitarian workers.Such as InterAction’s Minimum Operating Security Standards, the Aid Worker Security Database, ODI’s Operational Security Management in Violent Environments best practice guidelines, and the Security Management Training Course developed by InterAction and RedR with USAID/OFDA support.
 Offering training, consultancy services and, in some cases, protection services, specialist companies and NGOs entered the booming humanitarian security market.See for example RedR, which specialises in training, and INSO (International NGO Safety Organisation), which is an NGO entirely devoted to security analysis and consultancy (www.ngosafety.org/about).

Private security companies employing veterans from the police, army and intelligence corps (such as the UK’s Control Risks Group and France’s Amarante) can now boast numerous humanitarian organisations among their clients—Médecins Sans Frontières amongst them.

In the space of twenty years, the dangers inherent in deploying relief operations in conflicts and natural disasters have come increasingly to be treated as risks that can be controlled using methods developed by security specialists. The root of fundamental transformations, this evolution has been considered inevitable, and even positive, by the vast majority of humanitarian organisations, who see it as proof of the growing professionalism of their sector. Backed up by quantitative studies and media news coverage, experts and aid agencies assert that relief workers are now exposed to dangers of unprecedented frequency and nature. Besides the risk of “collateral damage” which one necessarily faces when operating in war zones, they add the threat of being deliberately targeted by criminal or terrorist networks or repressive governments.“Over the past ten to fifteen years, the operational environment of NGOs has become increasingly dangerous. Serious incidents—killings, kidnappings and attacks that cause serious injuries—are on the rise as are politically-motivated attacks against humanitarian workers”, InterAction members warned, for example, in 2015 (www. interaction.org/work/security, last accessed 22 December 2015).
 Given this increased danger, humanitarian organisations appear to have no choice but to professionalise the management of their security by calling on the knowledge and practices developed by experts. They would thus be able to safeguard their operations while fulfilling their moral and legal obligations to their staff, who could potentially sue them for breach of the employer’s obligation to protect employees.

Faced with the violent deaths and kidnappings of several of its staff members in recent years in Syria, Central African Republic, the Democratic Republic of Congo and Somalia, some at MSF are also coming to believe that their work environment has become more dangerous. As a result, many association members, from the field to the Board of Directors, are advocating the development of security departments, procedures, training, tools and data collection—and bringing in the know-how of external experts.

The increasing influence of security specialists in humanitarian organisations has, however, elicited numerous questions and criticisms from practitioners and researchers alike. Most critics associate the expansion of security expertise with aid workers being walled off in fortified aid compounds,Mark Duffield, “Risk-Management and the Fortified Aid Compound: Everyday Life in Post-Interventionary Society”, Journal of Intervention and Statebuilding: Working in Challenging Environments, no. 4 (2010), pp. 453-474; Mark Duffield, Risk Management and Aid Culture in Sudan and Afghanistan Project, November 2011; Mark Duffield, “Challenging Environments: Danger, Resilience and the Aid Industry”, Security Dialogue, vol. 43, no. 5 (2012), pp. 475-492.
 far not only from danger but from the very people they are supposed to be assisting. Like anthropologist Mark Duffield, criminologist Arnaud Dandoy describes how in Haiti, for example, “in urban areas, as a direct result of the increased sway of the security discourse, humanitarian organisations have retreated behind the walls of fortified residences and offices” and instituted “no-go times” and “no-go zones” for their staff.Arnaud Dandoy, Insécurité et aide humanitaire en Haiti: l’impossible dialogue? Décrypter les enjeux des politiques sécuritaires des organisations humanitaires dans l’aire métropolitaine de Port-au-Prince, Port-au-Prince: Groupe URD, 2013 [authors’ translation].
 In Dandoy’s opinion, this social and spatial segregation of humanitarian workers reduces their ability to understand their environment and establish relationships of trust with the population and its representatives—which is the only way to create a secure environment conducive to action.The tendency to raise the height ofthe walls around humanitarian workers—rather than try to do without them—is also central to the critique by researcher Larissa Fast, who encourages aid actors to distance themselves from the dominant security norms in order to re-establish relationships of trust with the populations based on empathy and proximity (see Larissa Fast, Aid in Danger: The Perils and Promise of Humanitarianism, Philadelphia: University of Pennsylvania Press, 2014). According to sociologist Silke Roth, loss of motivation and disengagement await humanitarian workers subjected to a security regimen that distances them from the populations and deprives them of autonomy—in particular, the pleasure of taking what they believe to be calculated, justified risks (see “Aid Work as Edgework-Voluntary Risk-Taking and Security in Humanitarian Assistance, Development and Human Rights Work”, Journal of Risk Research, vol. 18, no. 2 (2015), pp. 139-55.)

While Médecins Sans Frontières spokespeople regularly condemn the “bunkerisation” of aid agencies and their “risk aversion,”“Where is everyone ? Responding to emergencies in the most difficult places”, MSF report, 2014.
 its members privately acknowledge that the association is not always successful in bucking the trend. Moreover, a number of operational managers dispute the reality of the increased danger advanced by the experts to justify the need for their services. Many complain of the mounting pressure they face to report on how they manage security and to apply the best practices recommended in manuals, such as organising and following up on training, creating and updating databases, drawing up crisis management procedures and guidelines, etc.See Report from MSF-France Head of Mission Week, May 2012; interviews con-ducted in 2013-2014 with ten programme managers and three operations directors from MSF-France, MSF-Belgium and MSF-Switzerland (Fabrice Weissman, “Sécurité et prise de risques en mission, Synthèse des premiers entretiens”, Paris: MSF-CRASH).
 Sometimes doubting the utility of such measures, they often feel obliged to implement them, if for no other reason than to calm the concerns and demands of their boards of directors, management and some field volunteers.

The doubts and controversies surrounding the imposition of security standards, guidelines, indicators and procedures on the workings of humanitarian organisations such as MSF are at the root of this book. How does professionalisation of the security sector help aid workers to cope with the dangers encountered in conflict situations and other crisis settings ? Is there an alternative to the dominant security culture ? These are the two questions that guide our reflection.

Methodology

This book is divided into three parts. In the first part, we attempt to understand how the debate on security and the role of experts has evolved, both in the humanitarian sector as a whole and within MSF. Bertrand Taithe explores how relief workers have apprehended the notions of risk, danger, security and protection since the nineteenth century while Michael Neuman relates how the security of teams working in the field has been problematised and debated by MSF-France board members and senior managers since the association’s inception.

In the second part, we examine the diagnosis and recommendations made by security risk management specialists. Fabrice Weissman offers an analysis of efforts to quantify violence committed against the aid sector, and, with Monique J. Beerli, provides a study of the security manuals for humanitarian workers published since the latter half of the 1990s.

In the third and final part, we provide an overview of contemporary MSF security practices, using Central African Republic, north Syria and an abduction in the Russian Caucasus as examples. In an interview with Michael Neuman, Delphine Chedorge describes the responsibilities of a head of mission in charge of security in Central African Republic in 2014. Judith Soussan recounts the security practices implemented by a field team deployed in the thick of the Syrian civil war in an area controlled by a succession of opposition groups (including the Islamic State of Iraq and al-Sham) between 2013 and 2014. In the final chapter, Duncan McLean tells of MSF’s controversial efforts to secure the release of a Dutch volunteer abducted in Dagestan in 2002.

The choice of these accounts merits some explanation. Our aim was not to describe the state of MSF’s security practices in an exhaustive or representative way, but to shed some light on their diversity and their possible discrepancies with the analytical and action frameworks proposed by security manuals or required by MSF management. We gave preference to what were deemed especially dangerous situations and those that are the focus of current debates on insecurity. The decision to examine a case of abduction was based on the serious impact of kidnappings and the lack of transparency surrounding them. The sometimes debatable legality of the practices used to free hostages, the reluctance of some victims to talk about the harsh conditions of their detention and the fear of making the job easier for future kidnappers by exposing how such cases were resolved combine to enforce a code of silence that is not conducive to analysis and deliberation. This is why we chose to discuss a case from some time ago: namely, the 2002 abduction of Arjan Erkel, the details of which have already been made public in several books and press articles, as well as the legal battle between MSF and the Dutch government.

Before going further, we should make it clear that, strictly speaking, this book does not address the causes of the insecurity affecting humanitarian organisations. While it concludes that it is impossible to establish whether there is a general increase or decline in insecurity, we readily agree that concerns about the perils facing humanitarian teams in many of the areas where they operate are indeed legitimate. Although we discuss on several occasions the type of dangers affecting MSF, we do not aim to provide an exhaustive list of the different situations in which the safety of aid workers is jeopardised. There is a plethora of literature supporting (or debating) the hypothesis that today’s humanitarian workers are deliberately targeted due to their lack of independence, impartiality or neutrality. Indeed, a critique of this theory is central to our earlier book, Humanitarian Negotiations Revealed: The MSF Experience.Claire Magone et al. (eds), Humanitarian Negotiations Revealed: The MSF Experience, London: Hurst & Co, 2011.

Our deliberations on humanitarian security practices have drawn us to numerous social science texts devoted to the history of risk management in Western societies, the sociology of management tools and the way in which other sectors (military, banks, development agencies, air traffic control organisations, etc.) handle their relationship to danger and uncertainty. In the pages that follow, we will rely on this extensive literature to present the different chapters of our book and propose some answers to the question posed at the beginning of this introduction: how does the dominant security culture help a humanitarian organisation such as MSF to cope with the dangers it encounters in conflict situations and other crisis settings ?

How Has the Culture of Security Risk Management Come to Gain Traction? The End of an Era

In his contribution, historian Bertrand Taithe points out that the notion of risk management was already quite prevalent among the humanitarian foundations and organisations that came into being during the nineteenth century. At that time, it applied to financial and institutional administration of charitable funds, frequently administered by bankers and businessmen anxious to demonstrate their responsible management. Individual exposure to danger in war zones was a completely different story—one of courage and bravery, often recounted in the form of heroic narratives. During the wars of the latter half of the nineteenth century, humanitarian security practices were based on negotiation, the mutual interests of the belligerents and the threat of public denunciation—some going so far as to publish the names of the officers in command of artillery batteries that bombarded hospitals. Although security was rarely what it should have been (far from it), humanitarian narratives often implied the opposite, speaking very little about the great difficulties experienced by those in the field, thereby helping to reinforce the notion that health facilities were neutral, inviolable sanctuaries.

According to Taithe, the paradoxical coexistence of risk management and the spirit of adventure, which persisted through most of the twentieth century, was made possible by the physical distance separating headquarters from the field, unsophisticated bureaucratic procedures and communication systems and the chivalrous spirit of early humanitarian workers. Relief organisation personnel were treated, and saw themselves, as associates of a noble adventure, rather than as employees who might demand of their employers the security guarantees to which fledgling labour law entitled them.

In Taithe’s view, the development of a security culture in the aid sector at the turn of the twenty-first century signalled an extension of the risk management approach hitherto limited to the financial administration of charitable institutions. It was, he claims, accompanied by a reconfiguration of the headquarters-field relationship. Thanks to improved communication systems, and in the name of employee protection, faraway decision-makers began exercising greater control over humanitarian workers, gradually eroding their autonomy and sense of responsibility. This phenomenon was facilitated by a view of the field afforded by remotely transmitted data that allowed headquarters to feel that they were in as good, or even better, a position than the teams to assess the situation and pilot operations. And so the security culture has done away with the fiction and narratives that fuelled the commitment of aid workers. The influence of risk management has led to disenchantment with humanitarian action, whose chivalrous spirit has been drowned in the icy waters of actuarial calculation and remote control.

MSF-France: Calling in the Specialists, Despite the Doubt

Michael Neuman’s study shows that the heroising spirit of chivalry lived on among MSF-France’s presidents and administrators until at least the late 1980s. During the association’s first twenty years, they considered exposure to danger an essential part of humanitarian engagement, that it conferred a certain nobility. Individual commitment is central to MSF’s first charter, whose final paragraph ends on this solemn note: “Anonymous and volunteers, [its members] seek no individual or collective satisfaction from their activities. They understand the risks and dangers of the missions they carry out and make no claim for themselves or their assigns for any form of compensation other than that which the association might be able to afford them.”MSF’s current charter, dating from 1992, maintains this paragraph in slightly modified form.
 Throughout the 1970s and 1980s, although numerous incidents were already occurring, headquarters had no real control over the day-to-day management of security in the field because of a lack of any direct means of communication.

As Neuman explains, the Board of Directors began viewing insecurity as a problem at the turn of the 1990s when the French section suffered its first violence-related deaths in a context marked by the expansion of relief operations in post-Cold War conflicts. The period coincided with the advent of portable suitcase satellite equipment that was soon replaced by mobile phones. Headquarters began communicating in real-time with field teams, even as bullets were raining down around them. It was then that the organisation issued its “golden rules”, framing how risks were to be taken on mission. With the reminder that one “could never count on humanitarian immunity” and that security depended, first and foremost, on understanding the context, positioning and contacts, these “golden rules” imposed three limits on volunteer engagement: the team must not be targeted; it should conduct curative and worthwhile activities; and headquarters took precedence over field in deciding to withdraw.

The interpretation and implementation of these rules were the focus of many debates during Board of Directors and Annual General Meetings throughout the 1990s and early 2000s. Directors would frequently go to the field to meet with the teams and make their own informed judgement, a tradition that was to wane during the 2000s. But there seemed to be agreement on some points: refusal to tolerate deaths and serious injuries occurring on a regular basis, rejection of blanket explanations for insecurity and distrust of the forms that the professionalisation of the security sector was taking.

For the managers of MSF-France, security had been primarily the business of logisticians and project coordinators (and their line managers). Logisticians were responsible for the material aspects of team protection: safe rooms and bomb shelters, transportation and communication equipment, protective equipment, hibernation stocks, evacuation plans, etc. Coordinators, “resourceful and diplomatic”, were responsible for the political aspects of security: context analyses, contacts and negotiations, possibly including suspension of activities, withdrawal and public denunciation—the latter having been used as a political tool since the association’s beginnings. In the 1990s, in an attempt to defend this logistical and political approach to security, MSF-France leaders refused to create specialist security expert positions or use training courses given by ex-military personnel. MSF did decide, however, to enhance the skills of operational managers in context analysis, in particular via training and the development of research into the relationship between humanitarian action and its political environment. During that same period, expertise in security logistics was formalised, as illustrated by the ever-increasing size of the chapters devoted to security in successive editions of MSF’s Aide a l'organisation d’une mission guidelines (“Guidelines on Setting Up a Mission”).

But risk management culture won in the end. In 2013, after several years of appropriating more and more from specialist training and best practice manuals, MSF-France created the position of “security focal point”. How did this change come about?

A More Dangerous World?

The remarkable powers of attraction of risk management are in large part due to a growing sense of insecurity within the aid world over the last twenty years. Yet, as Weissman explains in Chapter 4 of this book, it is impossible to deduce from quantitative studies on violence against aid workers whether the danger is increasing or lessening, or indeed whether aid workers are now targeted for political reasons relating to a perceived lack of independence or neutrality. Figures on humanitarian insecurity are not statistically significant and are used primarily for promotional purposes, to justify the existence and power of a new guild of security professionalsSee Monique J. Beerli, “Securitizing Professions ? A Sociology of Humanitarian-Security Professionals and their Practices of Protection”, IPS Doctoral Workshop, Ottawa, 25-27July 2014.
 and to construct a victim narrative around violence against aid workers, held up as the heroes and martyrs of contemporary wars.

However, according to datasets produced by the Universities of Oslo and Uppsala on armed conflicts, aid agencies have been operating in a world that is no more violent than it was at the end of the Cold War. Indeed, in the 2000s armed conflict killed on average five times fewer people annually than it did in the 1980s, and nine times fewer than in the 1950s.Human Security Report Project, Human Security Report 2009/2010: The Causes of Peace and the Shrinking Costs of War, New York: Oxford University Press, 2010, p. 121.
 Mass violence against civilians is apparently also on the decline (with the notable exception of 1993 to 1997, which was marked by an exceptional number of deaths in Rwanda and its neighbouring countries).

Although the overall death rate from conflicts is declining, there are obviously episodes of extreme violence with huge human casualties caused by mass killings, famine, and disease. Such has been the case, for example, in Central African Republic, South Sudan and Syria over the past three years, where humanitarian workers regularly face sometimes persistent periods of extreme insecurity.According to Uppsala University data, the war in Syria is primarily responsible for the four-fold increase (from about 22,600 to 101,000) in the number of violent deaths due to combat from 2011 to 2014. The year 2014 was the deadliest since the fall of the Berlin Wall, though the death rate was still only half that of an average Cold War year. Therése Pettersson and Peter Wallensteen, “Armed conflicts, 1946-2014”, Journal of Peace Research, vol. 52, no. 4 (2015), pp. 536-550.Furthermore, kidnap now poses a significant risk to aid workers, and foreigners in general, whatever their occupation. Indeed, the monetary and political value of Western nationals on the international hostage market has been radically inflated by conflicts between armed transnational Salafist groups and states. While the French Ministry of Foreign Affairs stopped announcing in 2009 the numbers of French nationals abducted (or released) in other countries, it did acknowledge that, between 2004 and 2008, the figure had increased from eleven to fifty-nine. During the same period, the number of countries where these abductions took place rose from five to fifteen.Christophe Cornevin, “Les rapts de Français explosent dans le monde”, Le Figaro, 25 January 2010.
 The threat now extends over much of Africa, the Middle East, the Caucasus and Central Asia and, since 2011, there has not been a single year that an MSF section has not seen one of its members disappear or be kidnapped.

In this sense, although attacks against humanitarian workers are no new phenomenon, humanitarian organisations’ current concern with security is far from unfounded.

Nevertheless, from the security perspective, the most dramatic change in the past twenty-five years has been the substantial increase in relief operations and the number of humanitarian personnel working in the midst of conflicts. For example, the World Food Programme’s staff increased tenfold (from approximately 1,500 to 11,400 permanent employees) between 1995 and 2014Source: WFP Executive Board, Annual Performance Reports, 1995 and 2014, http://executiveboard.wfp.org/board-documents, last accessed 22 December 2015.
 and MSF’s national and international staff grew from 12,000 in 1998 to 36,500 in 2014. Western governments engaged in containment and stabilisation policies in peripheral conflicts encouraged and financed this increase. As Mark Duffield and others have shown, since the end of the Cold War, Western-led interventions in war-torn or unstable countries have relied heavily on humanitarian organisations to contain crises and population displacements within their own borders and to support peacekeeping and state-building operations.See Mark Duffield, Global Governance and the New Wars: The Merging of Development and Security, London: Zed Books, 2001. For an MSF perspective, see Fabrice Weissman (ed.), In the Shadow of “Just Wars”: Violence, Politics, and Humanitarian Action, London: Hurst & Co., 2004; François Jean and MSF (ed.), Face aux crises..., Paris: Collection Pluriel, Hachette, 1993.

Normalising Increasing Exposure to Danger

As one might expect, the increase in the number of aid workers deployed to conflict zones and unstable areas has resulted in a greater number of deaths, injuries and kidnappings. The development of risk management in the aid sector appears to be an effort to curb this increase and, in so doing, render occurrences more acceptable according to a logic observed in the history of risk management in Western societies.22. See Soraya Boudia and Nathalie Jas, “Risk and ‘Risk Society’ in Historical Perspective”, History and Technology, vol. 23, no. 4 (2007), pp. 317-331.
 As historians Jean-Baptiste Fressoz and Dominique Pestre point out, the introduction of the concept of occupational risk in nineteenth-century labour law contributed to both the recognition and the normalisation of the new dangers workers were exposed to as a result of the Industrial Revolution. By treating these unprecedented perils as risks, the legislature made it possible to acknowledge their existence and to make them acceptable, thanks to a system of regulation (standards, indicators, procedures, etc.) and compensation (insurance).See Jean-Baptiste Fressoz and Dominique Pestre, “Risque et ‘société du risque’ depuis deux siècles”, in Dominique Bourg et al. (eds), Du risque à la menace. Penser la catastrophe, Paris: Presses Universitaires de France, 2013, pp. 19-56.

As in other areas,Historian Soraya Boudia explains that the extension of formalised risk management techniques to environmental issues in the United States in the early 1980s followed a similar logic : “to contain the physical and political spillovers caused by activities whose growth [the authorities] deemed inexorable despite their health and environmental costs”. The well-known report “Risk Assessment in the Federal Government: Managing the Process” (considered the “bible” on the subject) was based on research funded by the RAND Corporation and the Ford Foundation, in the wake of mounting environmental protests around oil spills, pesticides and nuclear power. They used studies from cognitive psychology on the perception of risk and how to render it acceptable to the public, and econometric studies basing risk management on the calculation of cost-benefit ratios. See Soraya Boudia, “La genèse d’un gouvernement par le risque”, in ibid., pp. 57-76.
 the growth of risk management in the aid sector both recognises and normalises the exposure to danger of large numbers of humanitarian workers now deployed in the heart of conflict zones. This drive for normalisation seems all the more necessary as aid organisations confront the obsolescence of the spirit of sacrifice associated with the ethos of the first generations of humanitarians,See on this subject Michel Tondellier, “L’action organisée face à la prise de risque: l’héroïsme au travail et son institutionnalisation”, Acteur, risque et prise de risque colloque, CNRS, Lille, 2004; Patrick Le Gal, “L’esprit de sacrifice dans l’armée professionnelle d’aujourd’hui”, in Christian Benoit et al. (eds), Le sacrifice du soldat. Corps martyrisé, corps mythifié, Paris: CNRS Éditions, 2009, pp. 56-59.
 a growing social demand for protectionAccording to sociologist Patrick Peretti-Watel, the contemporary mindset is marked by what sociologist Jean Kellerhals calls “providentialism”, i.e., the “primacy of the individual’s concern for protection against the social environment, and even against the consequences of his own decisions.” Demanding protection pp. [10-13] and looking for accountability when the former is lacking relates to what Giddens calls the “disembedding of social relations”—the fact that increasing interactions render us dependent on complex technical systems vouched for by professionals. In this regard, the pursuit of accountability reflects not just the new individualism, but also the public’s protest against the power of the experts and authorities upon whom it increasingly depends. See Patrick Peretti-Watel, La société du risque, Paris: La Découverte, 2010, pp. 47-50. and the judicialisation of social relations. In this regard, humanitarian organisations face the same pressures as banks and large corporations, which in the 1990s established formal and auditable risk management mechanisms to protect themselves from lawsuits and scandal in case of adverse events.See Michael Power, The Audit Society: Rituals of Verification, Oxford: Oxford University Press, 1997; Michael Power, Organized Uncertainty: Designing a World of Risk Management, Oxford: Oxford University Press, 2007.

Nevertheless, as Michael Neuman and Jonathan Edwards note in this publication, the risk of legal judgment against an employer for failing to protect an employee is still largely theoretical in the aid sector, though the recent ruling by a court in Oslo, which found the Norwegian Refugee Council guilty of gross negligence in its handling of the kidnapping of Steve Dennis and three other staff members in Dadaab, Kenya in 2012, might change the equation.Imogen Wall, “NRC kidnap ruling is ‘wake-up’ call for aid industry”, IRIN News, 25 November 2015, http://www.irinnews.org/report/102243/nrc--kidnap--ruling' is-wake-up-call-for-aid-industry

Yet MSF asks its volunteers and employees to sign contracts with ever more intricate clauses and appendices, to protect the organisation from liability in case of accident. It is also to protect against potential lawsuits that the Boards of Directors of MSF partner sectionsAs a reminder, MSF is an international movement with five operational sections and sixteen partner sections.
—legally liable in the event of a lawsuit by one of their section’s volunteers on mission—encourage MSF operational centres to adopt the formal risk management measures set out in humanitarian security manuals.

Security in Theory

The Management Approach to Security

First appearing in the 1990s, security manuals were introduced primarily as practical guidelines designed to alert field workers to the dangers they might encounter in war zones. A second generation of manuals, published in the 2000s, called for a managerial approach, with a stated objective of protecting humanitarian organisations from legal and reputational risk. This is the case of the highly influential ‘Good Practice Review Number Eight’ (GPR 8), published by the Overseas Development Institute (ODI), to which Monique J. Beerli and Fabrice Weissman devote a large portion of their chapter.

At first glance, these manuals look like compilations of recommendations and checklists formalising the know-how developed over time in the field (how to manage communication equipment, organise travel, secure sites, behave in the event of fighting nearby, etc.). Their innovation consists in promoting a “strategic and operational approach to security management” based on detailed calculation, planning and standardised procedures.

The latest manuals intend to replace subjective security assessments with scientific methods to eliminate the biases of human perception. To achieve this, they suggest apprehending risk in its mathematical form (risk = probability [threat, vulnerability] x impact)Harmer et al., Good Practice Review: Operational Security Management in Violent Environments, Humanitarian Practice Network, London: Overseas Development Institute, revised edition, 2010, xviii.
 and refining its calculation using databases. Reminiscent of the actuarial approach adopted by insurance companies,See Chapter 3, p. 37.
 this process tends to disregard the causes and meanings of particular events, in favour of a probabilistic approach relying on mathematical laws to detect risk factors—despite the fact that the events in question (for example, the murder or abduction of an aid worker, an attack on an ambulance or the bombing of a hospital) are far too rare and dissimilar to be modelled using statistical series.

The manuals then recommend defining a “security strategy”, preferably relying on a so-called “acceptance” approach. An acceptance strategy aims to cultivate the goodwill of a population and its representatives toward humanitarian workers through a defence of their image as “good people who do good work.” In this regard, the GPR 8 describes the press and journalists as a major risk factor: “A poorly worded, inaccurate or inflammatory statement can put staff in direct danger and may even result in expulsion from a country.”Harmer et al., op. cit., p. 159.
 To contain this risk, standard communication procedures should ensure that everyone in the organisation, from security guard to president, projects the same message—and that no unauthorised documents or statements leak out. Not just their words, but all other forms of humanitarian worker behaviour need to be standardised via codes of conduct and operational procedures to ensure that the intended strategy is correctly applied.

The Ideological Assumptions in Security Manuals

The particularity of second-generation security manuals is not so much that they recommend using rules, indicators and procedures—such regulation mechanisms predate the professionalisation of the security sector and the shift to a “managerial approach”. Their innovation is to substantially multiply these tools, and to convey in the guise of technical recommendations the specific ideological assumptions highlighted by Monique J. Beerli and Fabrice Weissman.

First, the GPR 8 and the guidelines modelled on it promote an apolitical view of security challenges. Considering security as a technical problem requiring technical solutions, they obscure the social and political conflicts as well as the power plays and interests that structure the arena where aid agencies negotiate their presence and protection. This apolitical approach is facilitated by the use of the notion of “risk” itself. As sociologist Patrick Peretti- Watel explains, risk is “danger that we consider random, without cause. It is danger for which it is less a matter of blaming culprits for past occurrences than of preventing future occurrences.”He continues: “The notion of risk then appears as a reducer of uncertainty, characteristic of the prospective activity of an individual who seeks to control his future or that of others, like an insurer or actuary. Risk is, in a word, a danger that proliferates, insofar as the notion leads to the multiplication, and thus dispersal, of causal links.” Peretti-Watel, op. cit., pp. 14-15 [authors’ translation].

The apolitical approach is manifest in its recommendations regarding the media—at best considered as merely a means to relay standard marketing campaigns projecting the image of virtuous, consensual humanitarian organisations; at worst, a threat to be neutralised. This distrust vis-à-vis the public sphere contrasts sharply with MSF- France’s practices of the 1970s to the 2000s, characterised by repeated public appeals aimed at reinforcing the association’s stances in its (often conflict-ridden) dealings with political and military powers capable of affecting its security.See Fabrice Weissman, “Silence Heals.”, in Claire Magone et al. (eds), op. cit.
 In this publication, the account of MSF-Switzerland’s efforts to obtain the release of its kidnapped volunteer in the Caucasus is a good example of this.

The manuals’ second assumption is the positivism that sees wars and crisis settings as the sum of risks that can be controlled by calculations and planning. Even according to management theories (on which the GPR 8 claims to base its approach), such confidence in the ability of reason to predict and control every possible phenomenon—provided the necessary time, means and expertise are allocated—appears obsolete. For the past fifteen years, authors such as Dominique Genelot have been recommending that businesses base their organisation and management on the notion of “complexity”,Dominique Genelot and Jean-Louis Le Moigne, Manager dans la complexité: Réflexions à l’usage des dirigeants, 4th edition, Paris: INSEP Consulting, 2011.
a term used to designate “anything that is completely or partially outside our understanding or control”. The manuals’ positivist approach is also at odds with the thinking of military theorists who, following Clausewitz, have considered uncertainty the chief characteristic of the battlefield. Faced with the “fog of war” and the unpredictable behaviour of the military machine, subject to the phenomenon of “friction”, many military theorists recommend—as does General Vincent Desportes—using tactical methods based on “trust in man and the flexibility of systems.”Vincent Desportes, Décider dans l’incertitude, 2nd ed. Paris: Economica, 2015. Carl von Clausewitz used the phrase “fog of war” to describe the unreliability of the information available to officers: “In war, much of the information is contradictory, even more is false, and the majority is uncertain; the facts are rarely fully known and their motivations even less so.” (cited in Desportes, p. 37). “Friction” refers to the “countless minor incidents—of the kind you cannot really foresee— [that] combine to reduce the general level of effectiveness in such a way that one never achieves the goal” (von Clausewitz, cited in Desportes, pp. 38-39) [authors’ translation].

Yet the distrust of man is the third assumption conveyed in the latest security manuals. Indeed, they manifest a mistrust that is threefold. Mistrust of populations that aid agencies are supposed to be helping, but whom manuals tell us to regard as potentially threatening, mistrust of the general public and opinion leaders, considered vectors of risk to reputation, and mistrust of the volunteers themselves. In this last regard, security manuals such as the GPR 8 disregard aid workers’ subjective judgement on security, preferring a matrix; they distrust their initiative, preferring standard operational procedures; and lastly, they doubt their loyalty, preferring waivers annexed to their contracts.

The adventurous ethos of the early humanitarians described by Bertrand Taithe—discernible in the notion of the “aristocracy of risk” at MSF—would thus seem to present a particular risk in the eyes of the experts. From the 2000s on, the security manuals contain numerous negative references to “dinosaurs” and “cowboys,” and to “adrenaline-addict (...) A-type personnalit[ies]”Koenraad Van Brabant, “Mainstreaming the Organisational Management of Safety and Security: A review of aid agency practices and a guide for management”, HPG Report 9, London: Overseas Development Institute, March 2001, p. 17.
 who are “overconfident that they can handle any security situation because they have been doing it for many years,”Harmer et al., op. cit., p. 123.
 when in fact they are throwbacks to a bygone era where “there were fewer threats, [and] greater respect for aid organisations.”Van Brabant, op. cit., p. 49.
 Their ideal volunteer does not measure the risks and perils of their mission and is not “engaged” with the action. They are docile and responsible; they trust the experts to analyse the risks objectively and to know how they should talk and behave (even as far as in their sex lives) in order to stay safe and protect their comrades and their organisation.

And yet there is a fourth assumption in the manuals: the legitimisation of a rational-legal ethics of sacrifice. According to the GPR 8, “good operational security management” cannot completely eliminate danger or losses. It should, however, ensure that “residual risk” is kept to a minimum via procedures aimed at reducing the probability and impact of incidents, and that such risk is “justified in light of the potential benefit of the project or programme”. Danger and sacrifice are acceptable, provided procedures are followed and the cost-benefit ratio is favourable.

The introduction of risk management in humanitarian action is in this way symptomatic of a broader phenomenon, which political scientist Béatrice Hibou terms “neoliberal bureaucratisation:”Béatrice Hibou, La bureaucratisation néolibérale, Paris: La Découverte, 2013.
 the invasion of social relations by forms of bureaucratic regulation issuing from the private sector and based on its own abstractions.Hibou gives the example of a nurse who spends a third ofher workday document¬ing it on standardised forms—a large percentage of which were created by non-medical consultants—and who, once home, must battle the various formalities necessary to manage her telephone subscription and other day-to-day bureaucratic obligations, the sense of which escape her. Hibou, op. cit., pp. 5-14.
 Hence in the aid world, the concept of actuarial risk is used to apprehend the dangers that humanitarian workers face in war zones and cost-benefit calculations to determine the acceptable level of exposure. The security manuals epitomise this strange mix of neoliberal ideology and technocratic planning fantasy peculiar to many tools of contemporary humanitarian action, such as the widely-used planning and management tool called the “logical framework.”See François Giovalucchi and Jean-Pierre Olivier de Sardan, “Planification, gestion et politique dans l’aide au développement: le cadre logique, outil et miroir des développeurs”, Revue Tiers Monde no. 198 (2009/2), pp. 383.

Security in Practice

An approach based on ethnographic observation is needed to accurately describe the practical uses of security tools, procedures, manuals and training and their actual impact on how MSF and similar organisations operate.Studying the logical framework’s impact on development practices, Giovalucchi and Olivier de Sardan underscore the contrast between its ideological hegemony and its limited use in practice. They explain: “the ideological meanings or cogni¬tive assumptions incorporated to varying degrees in the tool can be applied, ignored, transformed, bypassed or manipulated in actual practice. Thus, a rigorous analysis [of the tool] as an instrument of public action should both reveal the political ideology and cognitive model embedded in it and describe its practical uses.” Ibid. [authors’ translation].
 This is not the aim of the accounts included in this book, as their primary objective is to illustrate the discrepancies between the apolitical, positivist view of danger inherent in risk management and the experience of teams responsible for mission security or securing the release of a hostage. They also offer a glimpse into how the dilemmas created by risk-taking on mission (the subject of the “golden rules” set down in the 1990s) arose, and were resolved, in the situations in question.

The last three chapters of this book begin by describing the broad assortment of dangers that MSF teams face—dangers that cannot be solely attributed to a “lack of neutrality or independence,” to paraphrase the dominant interpretation provided in quantitative studies and security manuals. These dangers are associated with the terror strategies used by belligerents: the lynching of enemy wounded inside hospitals by militias in Central African Republic; a machine gun attack on a meeting of prominent citizens inside a hospital; the bombing of hospitals by the Syrian army; the kidnapping, execution and trafficking of hostages by Syrian armed groups; the assassination by these same groups of supposed traitors and apostates among patients and staff; and human trafficking in the Caucasus fuelled by an extremely brutal pacification campaign. But the dangers are also tied to the micro-histories of the missions and the individual behaviours of their members. It might be a social conflict that degenerates into death threats in Central African Republic; or a patient’s father who, feeling his son is not getting good care, points his weapon at a doctor in Yemen; or a head of mission who arouses the suspicions of the Russian secret service by acting as a guide for a delegation of American military personnel in the Caucasus.

These studies also show that the risk analyses performed by operational managers are very far from an objective process that neutralises the human factor in favour of mathematical rationality. Understanding of the context and the risks is influenced by the personal paths of heads of mission and project managers, their previous knowledge of the country, their interest in its history and its political actors, and their personal network of relationships. The diversity of personalities and circumstances is reflected in how they obtain information or set up a network of contacts. Some prefer to keep their distance from local society, for fear of being caught up in power struggles between clientelistic networks, while others prefer to create a network of friends who can help them understand the environment and, if need be, actually protect the mission by using their influence or passing on important and well-timed information.

Whatever their approach, operational managers have to handle extremely complex and volatile situations. The last three chapters illustrate the “fog” in which decision-makers must make their decisions, the “friction” associated with the functioning of the MSF machine, and the impossibility of relying on standard procedures when dealing with uncertainties. The limitations of guidelines and training are particularly obvious in the abduction discussed in the book, which underscores the degree to which uncertainty about the kidnappers’ identity and motives forced the negotiators to take perilous gambles—which meant disregarding the recommendations of private and government experts. In such uncertainty, profound differences can appear between successive field teams or between field, coordination and headquarters regarding the analysis of the context, the ensuing danger or the usefulness of the mission.

Security is the outcome of constant negotiation with political and military authorities on the choice of activities, the services rendered, the dividing-up of MSF income (salaries and rents), public pressure and the quality of the interpersonal relationships established by operational managers with those around them. We also see the pragmatism of teams that go so far as to delegate some degree of security management to a Salafist politico-military entrepreneur in Syria or to a priest in a Catholic mission in the CAR.

Furthermore, personal freedoms of field volunteers are limited by rules governing their movements and behaviour (dress, attitude, emotional life, etc.), rules created not only for security reasons but also to facilitate human resource management. The accounts show, indirectly, that fear, anxiety, guilt, elation and valuing courage and strength of character play an important role in how aid workers assess risk and the merit of their actions. But they also show that such emotions and merits usually remain unsaid. We see the taboo against certain types of violence that are considered dishonourable—torture and sexual assault, for example—and the lack of transparency between humanitarian organisations (including between MSF sections) about security incidents they experience. This lack of information—like the attitude of project coordinators who feel that they alone are responsible for security—sometimes makes it impossible for field staff to gauge the risks they face in going about their work.

The accounts herein also highlight how difficult it is to interpret the rules (such as the “golden rules”) regulating the scope of risk-taking in the field. When the Syrian government bombs hospitals in rebel areas, when jihadist groups declare that “foreign infidel NGOs are not welcome in Syria,” or when a Syrian MSF surgeon, known for his militant atheism, is abducted while on call and then killed, is the organisation being targeted ? Do MSF’s activities in such contexts justify the risk? Opinions differ between, and even within, sections. While activity statistics and efficiency are brought to bear in these debates, the determination of acceptable risk in Syria does not boil down to a cost-benefit calculation and complying with procedures, but includes, among others, the field teams’ feeling of “being where they should be.” And finally, these accounts underline the ambivalent role of headquarters. By its considerable involvement in security management, depending on the case, it contributes—from the teams’ perspective—to over- or underexposing them.

From Bunker to Humanitarian Martyr

The MSF teams mentioned in these studies were not exposed to the “bunkerisation” syndrome described by Arnaud Dandoy and Mark Duffield. We should point out that, unlike the United Nations and many large NGOs, the specialised units dedicated to security at MSF are still in the early stages of their development, and their prerogatives are no greater than those of the Operations Departments, which are still ultimately responsible for the risks that are taken. Yet one cannot help being concerned by the apparently unstoppable spread of the dominant risk management culture. Responding to the need to normalise humanitarian workers’ exposure to danger and protect their organisations against legal and reputational risk, it poses a threat to aid workers, holding out a promise of protection that it cannot fulfil while overshadowing the social and political dimensions of their security. And lastly, it drives their organisations towards authoritarianism.

“Bunkerisation” and operational paralysis are only one potential consequence of an ever-expanding risk management culture. At the other extreme, this culture can help render the growing number of dead, wounded and kidnapped acceptable by heroising humanitarian workers while bureaucratically normalising their exposure to danger. This heroisation takes the form of public campaigns aimed at denouncing violence against aid workers while simultaneously contributing to the symbolic representation of humanitarians as heroes and martyrs of contemporary wars.While the UN has encouraged honouring “those who have lost their lives in humanitarian service” on 19 August every year since 2008, the number of monuments dedicated to humanitarian workers who died on mission has been growing in recent years (especially in Great Britain, Canada and Australia). In addition, awareness campaigns condemning violence against aid workers (the ICRC’s “Health Care in Danger” campaign, or ACF’s “Protect Aid Workers” campaign) participate in the symbolic transformation of humanitarian workers into the heroes and martyrs of contemporary wars. Yet as anthropologist Jean-Pierre Albert writes, “the heroism is linked not to the outcome of the undertaking, but to the acceptance of risk and suffering, even death.” Thus, paradoxically, the spirit of sacrifice associated with the chivalrous ethos of the early humanitarians is co-opted to rationalise exposure to danger based on actuarial cost-benefit calculations. See Jean-Pierre Albert, “Du martyr à la star. Les métamorphoses des héros nationaux”, in Pierre Centlivres et al. (eds), La Fabrique des héros, Paris : Éditions de la Maison des Sciences de l’Homme, 1998.
 Whether overexposed or overprotected, humanitarian workers tend to be deprived of a sense of engagement in dangerous situations as their employers develop numerous procedures to protect themselves from legal and reputational risk in case of accident.

The alternative to this trend is not to reject, en bloc, security indicators and rules. It is to recognise that the dangerous situations in which humanitarian workers operate involve an unavoidable amount of uncertainty, making it necessary to take gambles.In this sense, humanitarian action belongs to the domain of prudential occupations described by Florent Champy, from whom we borrow the observations that follow. See Florent Champy, “Grand résumé de Nouvelle théorie sociologique des professions, Paris: Presses Universitaires de France, 2011”, SociologieS, 9 May 2012, http://sociologies.revues.org/3922, last accessed 22 December 2015.
 The experience of other professionals (such as doctors, firefighters or police officers)Ibid. faced with irreducible levels of uncertainty shows that the more these gambles are based on an empirical analysis of each particular situation (rather than on blanket explanations and general recommendations), the more they rely on practitioners’ experience and professional judgement (rather than on the automatic application of routines and formalised procedures), and the more they are subject to deliberation on the means and ends of the actions to be undertaken (rather than to authoritarian, sub rosa decisions), the less risky they are. The alternative to the dominant security culture means trusting the practical wisdom of humanitarian workers and helping it flourish by relating and analysing their experiences with danger.