A mother who gave birth to her child at the obstetric and newborn care service in Dasht-e-Barchi before it was attacked in 2020.

Afghanistan : Should MSF accept the risk of targeted killings?

Fabrice Weissman

Graduated from the Institut d'Etudes Politiques de Paris, Fabrice Weissman joined MSF in 1995. He spent several years as logistician and head of mission in Sub-Saharian Africa (Sudan, Eritrea, Ethiopia, Liberia, Sierra Leone, Guinea, etc.), Kosovo, Sri Lanka and more recently Syria. He has published several articles and books on humanitarian action, including "In the Shadow of Just Wars. Violence, Politics and Humanitarian Action" (ed., London, Hurst & Co., 2004), "Humanitarian Negotiations Revealed. The MSF Experience" (ed., Oxford University Press, 2011) and "Saving Lives and Staying Alive. Humanitarian Security in the Age of Risk Management" (ed., London, Hurst & Co, 2016).

A shorter version of this piece first appeared in The New Humanitarian on November 26th 2020. 

The May 12th massacre at the MSF-supported maternity hospital in Dasht-e-Barchi (Afghanistan)  On May 12, 2020, the MSF maternity hospital in Kabul’s Dasht-e-Barchi district was attacked; the assailants murdered sixteen mothers in their beds, five of them about to give birth. Nine other people were killed, including a midwife employed by MSF and two children. https://www.lemonde.fr/idees/article/2020/06/18/a-kaboul-nous-refusons-qu-un-massacre-dans-une-maternite-soit-un-risque-ordinaire_6043330_3232.html raises, yet again, the question of our limits with regard to risk. What is an acceptable level of danger for humanitarian aid workers? How do we set limits? Why would MSF decide to leave Kabul but remain in Herat, for example, or leave Afghanistan but remain in Niger, Burkina Faso, Mali, or Somalia, where the teams also face extreme danger?

MSF-France has a specific policy to guide that reflection – namely the “Policy of Risk Taking and Security Management for MSF OCP staff”, endorsed in 2015 by its board and the partner sectionsMSF partner sections are MSF USA, MSF Japan, and MSF Australia.

Where did that policy come from? What does it say? What are the main issues and controversies regarding its implementation? Should it be changed in light of the recent killings? These are the questions I would like to address, briefly, building on the research done for the book “Saving Lives and Staying Alive. Humanitarian security in the age of risk management”, and more specifically on the chapter entitled “On Danger, Sacrifice and Professionalisation: MSF and the security debate”, written by Michaël Neuman, who has studied the evolution of security debates at the MSF France board from 1971 to 2015. 

The roots: we value exposure to danger, but we reject sacrifice and martyrdom (1970-80)

Since its inception in 1971, exposure to danger has been an integral part of MSF’s identity. What defines MSF, the doctors without borders, is its willingness to face danger in order to help people in distress. The inherently dangerous aspect of MSF’s mission is clearly acknowledged in its charter, which devotes one of its four paragraphs to recognizing that danger: 

As volunteers, members 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.

In the 1970s-80s, however, this glorification of danger and the chivalrous spirit was tempered by another principle: the rejection of sacrifice and martyrdom. MSFers are supposed to take risks, but come back alive. They are not on a suicide mission; they are not supposed to die for humanitarian ideals. There is no place for martyrdom at MSF. This was stated clearly by MSF presidents, including Bernard Kouchner, in the 1970s and ‘80s and again in the 1990s and 2000s.

But how does one reconcile the glorification of danger and the rejection of sacrifice? How does one draw the line between a mission that’s risky and one that’s “too risky”? 

Up to the early 1990s, it was primarily the field teams that drew the line. Headquarters was still very small and had no way to communicate quickly with the field (in Afghanistan, for example, letters traveled by donkey to Pakistan, where they were posted by air mail; they took at least three weeks to reach their destination). As a result, the decision on what was an acceptable level of risk usually fell to the field team, and more specifically the “coordinator” – a position created in the 1980s as part of MSF’s effort to become more “professional”.

In the 1980s, the most common form of security management was “embedment.” MSF teams in Afghanistan, Tigray, Eritrea, and Angola delegated their security to armed groups, who were responsible for their protection and logistical supplies. There were already numerous incidents: aerial bombings, kidnappings, and shootings (Tigray, Chad, Afghanistan, Uganda, Somalia, etc.)

Yet despite the field’s primary role, the Board of Directors felt that it had a collective responsibility to ensure that risky missions were not “too risky” as to be sacrificial. And on at least two occasions, the board decided to withdraw teams (in Uganda in 1981 and in Iran 1982), despite opposition from the field volunteers willing to stay. 

The turn of the 1990s

The 1990s were a turning point in terms of risk exposure, for a number of reasons. First and foremost, MSF experienced the first killings of its international volunteers: in Sudan, in December 1989, when a plane taking off from Aweil was shot down, killing the pilot (from Aviation Sans Frontières) and his three passengers (two MSF and one WFP staff). We still don’t know who fired the shot (the government or the armed opposition, the Sudanese People’s Liberation Army). Four months later in Afghanistan, insurgents killed a logistician in cold blood as part of a feud between rival commanders. In both instances, MSF halted its activities. 

These killings occurred in a context marked by the end of the Cold War, where humanitarian teams had greater opportunities for working on both sides of the front lines – something that required more autonomy in managing security than did clandestine “embedded” cross-border missions. This coincided with the advent of affordable portable satellite phones, which enabled HQ to be more involved in the day-to-day management of security. The first evacuation ordered by phone from Paris happened in 1992, after discussion with a team at Mogadishu airport in Somalia that, to the managers in Paris, seemed too panicked and disoriented to stay. Lastly, the 1990s were a time of rapid growth for MSF resources, with the number of expatriate and national staff deployed to war zones increasing two-fold in the latter half of the 1990s. 

It was under those conditions that MSF leaders felt the need to clarify and formalize a framework for security management in the field. A rough outline was proposed by Rony Brauman, then president of MSF France,  in his 1990 President’s Report, which for the first time had a section devoted to security incidents and management. The “Golden Rules” were formalized by Operations Director Brigitte Vasset, in the form of an aide-mémoire (the “Mémo”) that was included in the log-admin guidelines on how to set up and run a mission. Those “Golden Rules” were amended in 1992 and then resuscitated in 2015 by Benoît Leduc, the first MSF France Security Focal Point (SFP), who felt it important, as SFP, to be able to refer to an updated Board-endorsed security policy. 

The “Golden rules”

The 2015 Golden Rules are a revised, more detailed version of the original Mémo. What are its core principles? 

1.    There is no humanitarian immunity

The Mémo’s first statement says that “there is no humanitarian immunity”. The greatest danger is believing that humanitarian workers are protected by their principles, their logo, and the quality of their work. The killings in Afghanistan and Sudan at the turn of the 1990s demonstrated that providing valuable help to the population and being respected and appreciated by local civilian authorities and patients did not stop someone from shooting down an MSF plane, or assassinating a logistician. This rule contradicts the conventional discourse on acceptance, according to which doing good work and providing quality care in an impartial manner automatically ensures protection. 

2.    Our protection relies primarily on understanding the context and being able to build a network of contacts and relationships
We need to understand who threatens us and who protects us. Our security relies on our ability to find a protector (or protectors) with an interest in ensuring our safety, so as to benefit from our actions – the medical services and material assistance we provide (heath care for wounded combatants, social constituency, etc.), the money we spend (for salaries, rental contracts, local expenses, etc.), our access to mass media and our ability to portray the political authorities in a more or less positive light.  

This raises another concern: at what point do aid organizations become more useful to the military forces that protect/harass them than to the population? We have to be sure that primary beneficiaries of our interventions are actually the people we want to help, not the warring parties that protect or threaten us. We have to walk this fine line to avoid being victims or complicit in mechanisms that produce violence. 

3.    When insecurity is high, limit ourselves to curative activities carried out by smaller teams

The third rule states that the main way to reduce risk is to reduce the number of people exposed (hence the call for “smaller teams” in highly insecure contexts), and that these smaller teams need to be engaged in curative operations with obvious clinical outcomes (don’t take the risk for nothing…). Like the two previous ones, there is some opposition to this rule within the MSF movement. For example, consistent with the apolitical discourse around acceptance, our colleagues in Afghanistan believe that having very large projects is the surest way to improve our security; the greater the risk, the bigger we have to be, in order to be protected by the quality and the size of the services we offer. This illusion was cruelly dispelled by the massacres in KunduzIn 2015, an attack on the MSF hospital in Kunduz, Afghanistan killed 42 people (including 14 MSF staff) and injured dozens more. and Dasht-e-Barchi. 

4.    No heroes

The fourth principle reiterated that there is no place for martyrdom. In contrast to some UN agencies or the ICRC, which are mandated by States to work in conflict situations, NGOs like MSF have no mandate. They choose where to work and the level of risk they want to take. And as far as MSF is concerned, its members are not supposed to die for humanitarian ideas. Sacrifice is not part MSF’s mission statement. 

5.    If targeted, we leave

The fifth rule, established after the strafing of a clearly-marked MSF vehicle in Sri Lanka in 1991, asserted that “if we, MSF, are directly targeted, we leave”. In other words, MSFers are prepared to face the danger of being hit accidentally in war zones, but not of being deliberately targeted. 

How do we know if we are being targeted directly? The definition used thus far is pretty straightforward: if an armed group or government states and/or demonstrates its willingness to harm us and no one can protect us. The most obvious situation is when we have been deliberately hit, whatever the reason – for being a “bad doctor” deemed responsible for the death of wounded combatants, “spies” or “apostates” accused of supporting the “crusaders”, an “unfair employer” who dismissed a vindictive and powerful staff member, a “supporter of terrorist groups”, a “high value target” bringing media attention, “bounty” for kidnappers, or even “regrettable but necessary collateral damage.” Once a group or individual points a gun at us and pulls the trigger or otherwise demonstrates their intention to kill us – and we cannot be protected – we cross the red line into a “too-risky” mission. That line is defined as such by the current MSF France security policy: “MSF refuses to intervene in high security risk areas where there is an absence of authorities with which to negotiate our safety and/or when it is impossible to protect ourselves from groups demonstrating radical hostility against us.” 

Of course there are other types of “too-risky missions”: ones where the risk of being accidentally killed by random shelling, a stray bullet, friendly fire, etc. is so high that we decide to evacuate a mission as well. That was the case in northern Nigeria in January 2018, for instance, after the army bombed its own position in Rann, killing three MSF contractors. 

6.    When deciding whether to pull out, HQ can override the field’s decision. The top managers have the final word, and if necessary it goes up to the Board.

Rule #6 reaffirms MSF’s collective responsibility for managing security – more specifically, the coordination team’s authority over the field and the HQ’s authority over the coordination team when it comes to deciding to withdraw.  This is a bit tricky, since it implies that the top managers have the duty and power to call a halt, to stop teams that are in a “sacrificial mood,” but not to manage their exposure to danger on a day-to-day basis.  


These Golden Rules provide a framework for discussion, not an algorithm for decision-making. Their interpretation and implementation have always set off debate, starting with the red line that if we are targeted, we leave. 

In the mid-1990s, the situation in Burundi triggered heated discussion on the acceptable level of risk. In a context of mass violence against civilians, more than twenty foreign workers (from the ICRC and MdM, among others) were killed between 1995 and 1997. Given that radical groups had demonstrated their willingness and ability to kill international aid workers, the question was whether or not MSF should stay. 

Part of the Board of Directors and Operations Department was in favor of staying. They had three arguments: the fact that the killers had not yet targeted MSF, the magnitude of needs, and the field volunteers’ desire to stay, despite the risks. 

Another part of the MSF leadership disagreed. They argued that international aid workers doing the same work as we were doing in a nearby area had been targeted, and that there was good reason to believe that we might be next. They felt that the volunteers willing to stay were being driven by a “spirit of self-sacrifice” that was inconsistent with our rejection of martyrdom. 

The Board ultimately refereed the discussion and voted in favor of staying (11 for, 4 against), as long as there were volunteers willing to do so. During the debate, no one denied the fact that radical groups had demonstrated the will and ability to kill humanitarian workers – and that we were therefore targeted. The discussion was over agreeing, or not, to work under the threat of targeted killings. The decision to stay was de facto considered an exceptional override of the rule we had set for ourselves. As a matter of fact, the Human Resources department was refusing to brief volunteers leaving for Burundi (the Director of Operation had to do it). And the Board remained apprised of the situation, reexamining its decision on a regular basis – almost monthly – until a regime change in Burundi improved the security situation to some extent.  

The terms of those debates differed in some ways from how MSF’s presence in Afghanistan is currently being discussed. Proponents of a continuous presence in Afghanistan argue that – despite the deliberate killing of MSF patients and staff, and despite the Islamic state (IS)’s earlier suicide attack on the Save the Children compound in 2018 and their statement declaring aid workers a legitimate target – “we are not being targeted,” and no group has “demonstrated a radical hostility toward us.” Some top managers go so far as to propose that we “take the opportunity” offered by the massacre of women and children in our facility to try negotiate a kind of MoU with the IS that would exempt us from its policy of mass murder…

They also argue that even if targeted, we have to stay. That we have to consider the risk of targeted killings by radical groups the “new normal” we have to adapt to, which means revising our “Golden Rules” and abandoning the threat of targeted killing and lack of a reliable protector as a red line for excessive risk. 

Thus the debate has gradually shifted from “have we reached the red line” (are we being targeted?) to “should we cross the red line” (should we exceptionally stay despite the risk of targeted killings?) to “should we move the red line” (should we consider the threat of targeted killings a new normal and the loss of our colleagues the price we have to pay?).

This shift signals MSF’s drift toward an acceptance of sacrifice and martyrdom, consistent with what’s happening within the broader aid sector. As we have shown elsewhere, one of the primary consequences (functions?) of the portraying aid workers as heroes and of the adoption of a “risk management culture” is to make the growing number of dead, wounded, and kidnapped acceptable. 


To cite this content :
Fabrice Weissman, “Afghanistan : Should MSF accept the risk of targeted killings?”, 7 décembre 2020, URL : https://msf-crash.org/en/blog/humanitarian-actors-and-practices/afghanistan-should-msf-accept-risk-targeted-killings

If you want to criticize or develop this content, you can find us on twitter or directly on our site.


Add new comment

Cette question consiste à tester si vous êtes ou non un visiteur humain et à éviter les demandes automatisées de spam.