Natural Disasters: “Do Something!”

In this interview conducted by Claudine Vidal in 2012 and published in the book Agir à Tout Prix (Humanitarian Negotiations Revealed: the MSF experience), Rony Brauman speaks about emergency humanitarian aid set up following natural disasters. Time constraints, access to victims, cooperation with local institutions, misleading representations of the disasters' effects, controversial assessments of the number of casualties; various topics related to these interventions are discussed and illustrated with specific examples.

Has MSF always considered natural disasters part of its mission?

Alongside armed conflicts, natural disasters are the first category of intervention to be cited by the authors of MSF’s charter and by-laws. Moreover, among the events that led to the founding of MSF were the earthquake in Peru that killed 30,000 people in May 1970, and the Bhola cyclone that hit eastern Pakistan in November in the same year, leaving 250,000 to 500,000 people dead. Natural disasters have always taken centre stage for the organisation. If you remember, MSF was formed through the merger of two associations created in 1970: GIMCU (Groupe d’intervention médico-chirurgicale d’urgence— Group for Medical and Surgical Emergency Intervention), founded by former Red Cross volunteers in Biafra, and SMF (Secours médical français—French Medical Relief), set up by medical journal Tonus to respond to the disaster in eastern Pakistan. This was the time when emergency medicine was gaining momentum as a specific category of care and “collective accidents”, as they were curiously named in the charter, were the ideal field in which to practise it.

But GIMCU’s first experience in a disaster situation, Peru in 1970, was a failure: the French doctors arrived on the scene a week after the earthquake and, in spite of the scale of the disaster, didn’t encounter a single injured person. What they did find was that the countries in the region, including the United States, had already delivered emergency relief.

The lesson learnt from this first attempt at emergency intervention held sway for a long time and became a principle for all earthquakes: to implement a life-saving operation in such situations, medical assistance had to come on stream within the first forty-eight hours. Any later and the victims trapped under rubble, the injured suffering from multiple trauma—with or without crush syndrome—would have no chance of survival. So, MSF focused from the outset on reducing deployment time by ensuring emergency supplies (“kits”) were ready and waiting, and doing its utmost to get its teams out to the disaster area within twenty-four hours of the alert. But to no avail. It wasn’t until 2005 and the earthquake in Pakistan’s Kashmir that we actually operated on casualties for the first time—although we weren’t on site and operational immediately.

Earthquakes and other disasters have become more frequent in recent years. According to the CRED (Centre for Research on the Epidemiology of Disasters), the yearly average number of earthquakes causing more than ten fatalities increased from twenty-one between 1960 and 1990 to thirty at the beginning of the twenty-first century, with peaks recorded in 1990, 2003 and 2004. But only a few led to an international relief operation. In fact, we only respond to large earthquakes, when the initial estimate of fatalities is in the thousands and the national authorities call for international assistance. This is a useful reminder to us that, in spite of this type of disaster’s high rank in the hierarchy of emergency humanitarian assistance, MSF had had very little experience in the field until the beginning of the new millennium. Furthermore, as over 80% of earthquakes occur in the “Pacific Ring”, the distance from Europe makes the objective dictated by the precepts of emergency medical assistance of getting to the disaster site within forty-eight hours totally unrealistic. But distance and time to deploy do not explain everything, as we saw in 1990 when an earthquake leaving 37,000 dead hit Zandjan in Iran. MSF’s medico-surgical teams were on site twenty-six hours later, but as their sole medical activity was providing routine consultations and totally unrelated to the traumatology they were expecting, ten days later they packed their bags and left.

It took us some time to realise that earthquakes didn’t lead to a particularly high number of casualties, and that most of these received immediate treatment in local health facilities around the disaster area. Foreign medical teams, unless they were already on site, were in fact superfluous to requirement. Earthquakes were far from providing the situation par excellence that we had imagined for exercising emergency medicine, in spite of breathtaking figures evoking thousands, or even tens of thousands of casualties. However, given the symbolic importance of natural disasters in emergency assistance, it was almost inconceivable for an organisation claiming emergency response as its culture and expertise not to be part of the action. So, at the beginning of the 1990s, MSF changed direction and focused on its other skill, logistics, securing the supply of drinking water, for example, and when necessary setting up medical consultations in the places where the victims were assembled. The images of numerous surgical teams rushed off their feet and operating non-stop that we have witnessed since the earthquake in Port-au-Prince are so close to conventional representations of disaster medicine that we tend to forget that they are, in fact, relatively new, as they were seen for the first time in Kashmir in 2005.

What happened in Kashmir in October 2005? Did the relief operations launched in response to this disaster differ from previous experience?

When news of the earthquake reached us, MSF-France’s operations managers were initially extremely reluctant to intervene, for all the reasons I’ve just mentioned. But MSF-Belgium and MSF-Holland were in the country at the time and their teams were reporting back to us on the enormity of the disaster and particularly on the huge number of casualties. The province’s health facilities were all completely overwhelmed. According to official estimates, there were tens of thousands of critically and seriously injured people in need of orthopaedic and intensive medical care. However accurate these figures, and I’ll come back to this point later, it was clear that for the first time in an emergency situation the local facilities were submerged by the inflow of polytraumatised patients and unable to cope.

I think the explanation behind this sudden increase in the number of injured is the trend towards urbanisation, in other words, the densification of badly built dwellings in a high-risk seismic region. In Kashmir, people were no longer living in shantytowns, but in unsound houses made of poorly cemented breezeblocks and stones. While partial collapse of this type of construction results in crushed limbs, the victims are not buried under rubble, as they would be in buildings with several floors. But shantytowns at least have an advantage in that the wooden, plastic or sheet metal partitions used to build them cause little damage when they collapse. A reminder that all that is “natural” in a disaster is what causes it, i.e. the origin of the seismic or climatic event. Whereas the aftermath is the result of decisions made by people, such as the location or construction standards or insufficiently protected industrial installations in hazard zones. To return to the situation in Kashmir, urban densification was not simply the result of the rural exodus common to all countries, but was also part of a deliberate population distribution strategy linked to separatist intrigues and the ongoing territorial dispute with India since partition in 1947. Nonetheless, for the first time in the history of emergency responses to an earthquake, international medical and surgical teams had a real and major role.

In 2005, the press reported that access to the victims was often impossible. What practical solutions were found to overcome this?

Access to the region was indeed difficult at first because of its geography, but also because of its politics. Kashmir is a sloping plateau in the east and is easily accessible from India. But, on the Pakistani side, there is a barrier of escarpments which is difficult to cross. Landslides stopped us from using the roads and as it was early winter, poor weather conditions further complicated matters. We had to use helicopters, which are great for transporting personnel, but their low cargo capacity meant they were not suitable for a disaster of such magnitude. These practical difficulties were the main problems we encountered, if we don’t count the initial resistance of the Pakistani army, whose main concern at that point was to provide assistance to its own troops and maintain control over a province of strategic importance in its dispute with India.

The physical obstacles could have been overcome by bringing aid in via India, and the Indian government did indeed offer assistance to Pakistan. But such an offer was unacceptable to the Pakistani army, which refused it outright, although it did agree to a partial opening of the border. This doesn’t mean, however, that the army only concerned itself with its own personnel and territorial security, leaving the population without assistance. On the contrary, after a few days, it did more and more, bringing in aid supplies, treating and evacuating the injured by helicopter, and managing the coordination of the relief operations. Restrictions on movements were lifted and special permits were no longer required to travel around the tribal areas.

A multitude of local NGOs quickly got down to work, helping the victims get organised in collective centres and providing shelters. Some had highly competent personnel and were particularly well-equipped— in particular, the Al Rasheed Trust. An Islamic organisation ideologically close to the Pakistani Taliban, it set up a sixty-bed hospital for orthopaedic surgery and ran outreach and relief activities. Our collaboration with the army, the Health Ministry and Al Rasheed was excellent on the whole, much to the surprise of MSF’s management staff, who had expected things to be more complicated. The local Islamic organisations, which benefit from well-established social aid networks, took immediate action and supplied a considerable amount of aid.

Let’s not lose sight of the fact that most of the search for survivors and provision of food and shelter in the early stages of any disaster situation are always handled by local people and organisations. Contrary to conventional belief, it isn’t a state of shock that we witness but rather active solidarity, at least during the first few weeks. So, although there was nothing surprising about the extent of local mobilisation in Kashmir, it needs to be said that once they had seen that MSF wasn’t involved in any proselytising and that patients were being properly cared for, the Islamic NGOs were particularly cooperative. Islamic organisation members even praised the invaluable logistical assistance it had received from the American army.

The situation was one of close cooperation with the Ministry of Health, the army—whose helicopters we even used on occasion—and religious NGOs. Our constructive relationship with our natural partners, the health authorities and the Pakistani NGOs, raised no issues for MSF. However, despite the crucial role it played, as we have seen, the same could not be said of the army, viewed by MSF as a compromising partner. Some of the MSF operational leaders even suggested trimming down the teams in order to limit contact. This determination to reassert the distinction between military and humanitarian operators, motivated by concerns for the teams’ safety, ended up taking a back seat to imperatives for urgent action in a context marked by the ongoing emergency and the otherwise fruitful working relations with all the different actors, whatever conflicts may have opposed them in the past.

In situations of natural disaster, the national army is usually the best placed and the best equipped to respond and, apart from exceptional cases (such as in the zone controlled by the LTTE in Sri Lanka after the 2004 tsunami), is welcomed by the victims. So there is no reason to actively distance ourselves, as we rightly do in situations of conflict. This applies equally to medical relief and logistics provided by foreign armed forces.

Taking into account the material difficulties caused by the geography of the area, the deployment of the aid operation was dynamic and, within three weeks, had reached a level where the needs were being met. However, when it came to medical care MSF’s teams noted that, whereas the country’s response had been rapid and profuse, standards were not so satisfactory: amputations were numerous, probably overly so, and already conservative orthopaedic interventions—for saving injured limbs—were often below par. Let’s keep in mind that most of the surgery performed by MSF’s teams, who had not dealt with the initial influx of casualties, consisted in secondary surgery. However, I should point out that any reservations regarding the quality of the medical treatment stem from clinical impressions rather than from the findings of epidemiological studies, and that this was a context of damage-control surgery in the face of a very high number of wounded patients.

But overwhelmed medical facilities do not explain everything. In my opinion, we should also examine why war surgery techniques were used. Penetrating wounds caused by projectiles (bullets, shrapnel, etc.) can lead to complications, notably infections, which, in the uncertain environment of an armed conflict, may prompt the surgeon to perform more radical surgery. But wounds caused by crushing, the common lot of civilian surgery, permit the use of more conservative techniques. Yet, as we saw in Indonesia during the 2004 tsunami and again after the Haiti earthquake in 2010, the paradigm of war, or in this case a blitzkrieg, always seems to prevail. The medical teams are just as much influenced by this representation as the observers, as revealed by a remark made by a team of American relief workers: “Overworked surgeons […] amputated limbs and debrided infected tissue. […] For the next two days, we practiced continuous battlefield medicine”.Paul S. Auerbach et al., “Civil-military Collaboration in the Initial Medical Response to the Earthquake in Haiti”, The New England Journal of Medicine, Feb. 2010.We are justified in asking ourselves if this kind of representation has an impact on the techniques used, and studies are being conducted using medical data collected in Haiti, the only other natural disaster, along with the one in Kashmir, to have caused such massive numbers of casualties.Notably led by Prof Anthony Redmond and Dr Simon Mardel, HCRI, University of Manchester.

The very recent experience of mass surgery in such circumstances explains the current lack of systematised knowledge on the subject.

We also lack reliable quantitative data to draw up a comprehensive evaluation of the relief operation in Pakistan. The figures provided the day after the disaster—54,000 dead, 77,000 injured and hundreds of thousands made homeless—give an indication of the scale of the catastrophe, but should be viewed with caution, particularly from a medical standpoint. In the light of the absence of civilian registration and demographic data, the number of fatalities can only be a rough estimate.

The civil-military cooperation—read “military leadership” of the relief operations—was hailed as a success by the United Nations and the NGOs. The dividing up into sector-based groups of responsibility or “clusters” (logistics, health, sanitation, etc.), which the army had less trouble adapting to than the humanitarian operators (as was noted with some irony by the United Nations representative),Lieutenant General Ahmed Nadeem and Andrew Mc Leod, “Non-interfering coordination: the key to Pakistan’s successful relief effort”, Liaison Online 4, no. 1 (2008) was also a success.

However, de facto truces resulting from a natural disaster do not signal an end to hostilities, and we mustn’t lose sight of the political or even counter-insurrectional dimension of aid. The extremely sensitive deployment of US and NATO forces in response to the earthquake was explicitly dictated by such considerations. It encountered no visible opposition, as all the population was concerned with what was provided and not with who was providing it. As for the Islamic groups, they mostly kept silent, although some of them did express their approval publically. A study conducted by the US Institute for Peace concludes that the objective of “winning hearts and minds” remained theoretical, for the activist groups and for the United States and NATO, as momentary gratitude does not lead to political loyalty. But as this belief tends to hold sway, it results in more latitude for action, as nobody wants to be seen as the one depriving the people of valuable aid during a period of acute crisis.

What is your definition of a natural disaster?

A disaster disrupts the ordinary course of things. From the purely practical standpoint of an emergency medical organisation, this first means earthquakes and then severe climatic events—storms, cyclones and flooding—occurring in or close to densely populated areas. Earthquakes have been our sole topic of conversation so far, as they have recently become the main cause of emergency medical operations. But looking at things from a broader angle, and to use more commonly accepted definitions, a disaster can be defined as a sudden encounter between natural forces of harm and a people in harm’s way, where demands exceed the disaster-affected community’s capacity to cope or, in other words, it is “the product of the encounter between hazards and vulnerability”.Grégory Quénet, “Catastrophe naturelle”, in Yves Dupont (ed.), Dictionnaire des risques, Paris: Armand Colin, 2007.

The problem with these definitions resides in the definition of “natural”. The causal event may be natural, but the aftermath is closely linked to the way society is organised in the places where they occur. For example, you may remember that in Ethiopia (1985) and Niger (2005), the drought and the ensuing invasion of locusts were described by the authorities as a “natural disaster”, and the primary cause of a situation of acute malnutrition or famine. The stakes were high because attributing these consequences to this cause determined the response. MSF was expelled from both countries after a political controversy on these issues.To be more precise, MSF-France’s activities were suspended in Niger in 2008 on the orders of the head of state two years after the public controversies, but directly related to them.

 Remember the ironically evocative titles of the two books published by the organisation on the subject: Ethiopie. Du bon usage de la famine [Ethiopia: How to make best use of a famine], and A Not-So Natural Disaster, Niger 2005.Respectively, François Jean, (ed.), MSF, 1986; Xavier Crombé & Jean-Hervé Jézéquel (eds), London: Hurst and Co., 2009.

The cholera outbreak in Haiti during the winter of 2010 to 2011 was the source of an intense controversy of the same nature: the advocates of a “natural” hypothesis attributed its origin to plankton and opposed all those who claimed that the infestation was of human origin (caused by the emptying of a septic tank containing cholera germs into a river). Everyone agreed that the disease had only been able to result in so many fatalities (4,800 in total) because of the country’s deplorable hygiene conditions, but the circumstances that led to the outbreak were the subject of virulent dissension, even within MSF. The fact that the human origin was blamed on a contingent of United Nations peace-keepers, themselves embroiled in political clashes as a result of the election campaign underway at the time, only served to accentuate the political dimension of the epidemic. As it happens, an enquiry conducted by the United Nations later confirmed the second hypothesis.“Final Report of the Independent Panel of Experts on the Cholera Outbreak in Haiti”, May 2011.

Once again, it was not simply a matter of determining the origin of the epidemic; understanding its cause had practical consequences on how the immediate medico-sanitary response was organised.

The controversies seem to be as much due to the definition of natural disasters as to the evaluation of their consequences?

As we have just seen, the rebranding of a situation from major crisis to natural disaster can lead to controversy because of the political responsibilities that such a categorisation engages. But independent of any disagreement on this aspect, the consequences of a disaster can also be a source of controversy, particularly (but not exclusively), with regard to the epidemics they might cause, and hence the emergency resources that should be deployed. Because of the unprecedented media attention it attracted, the 2004 tsunami saw this question propelled into the public arena.

A few days after this exceptionally large-scale disaster, the WHO’s operations director announced: “We may see as many fatalities from disease as from the actual disaster itself”.“WHO warns up to five million people without access to basic health services”, 30 Dec. 2004.

 So the subject was raised of a possible second wave of mortality due to epidemics, which threatened to double the number of victims caused by the actual tsunami. It was brought up by the WHO at subsequent press conferences and passed on enthusiastically by the media, with the result that the relief effort focused on providing emergency assistance to save some 150,000 people supposedly in danger of imminent death. The success of such announcements, without scientific or empirical basis, stems from how well they fit in with the widespread belief that decomposing bodies are a source of infectious contamination. Yet as several research studies have shown, there have been no cases of a fatal epidemic in the wake of a disaster, whatever the scale.C. de Ville de Goyet, Stop Propagating Disaster Myths, The Lancet, 2000, 356: 762–4, Nathalie Floret et al., “Negligible Risk for Epidemics after Geophysical Disasters”, Emerging Infectious Diseases, www.cdc.gov/eid, Vol. 12, 4 Apr. 2006.

 Put quite simply, epidemics cause corpses, but corpses don’t cause epidemics. Some epidemic foci of digestive and respiratory infections may occur and require preventive and curative action, but their effects are nothing like the scaremongering announcements I just mentioned.

More generally, and for reasons similar to those I talked about earlier in relation to earthquakes, there was no life-and-death emergency after the tsunami. The horrendous ordeal suffered by a large number of survivors, some of whom lost everything, justified in itself the appeal for solidarity, and I’m certainly not disputing the need to respond to it. But the model adopted of “rescuing a population in peril” was totally inappropriate. At one point, we saw up to twelve surgeons gathered around just one casualty in Indonesia, right when we were talking in terms of hundreds of thousands of casualties! In practice, to be of real help to the victims of the disaster, the need was for financial and material resources to clear up and start rebuilding—quite different from launching an emergency medical operation. However, media pressure was such that it made it difficult for MSF to stay away. The field teams lost no time in raising the issue; some of the most experienced members had grasped what was happening within a few days. But withdrawing from the country would not have been understood in a situation so emotionally-charged, and the organisation’s leadership decided to switch the focus to non-medical aid.

Can the way a disaster is presented after the event make a difference then?

As we’ve seen, the scale of mobilisation shrank all the narrations, beliefs and prevailing representations of the event. Talk was of casualties, refugees, epidemics and, when UNICEF issued a statement, orphans too. We have already discussed casualties and epidemics, but the issue of refugees and orphans was much the same. I’ll say more about this in a moment, but first I want to emphasise that these four themes, recurrent during the first few weeks, formed a narration of the consequences usually observed in armed conflicts. In other words, with the benefit of a little hindsight, it becomes clear that we were unconsciously reacting to a natural disaster as if it were a war.

There were endless pictures and non-stop television images of the after-effects of the disaster, focusing on a few hundred people assembled in makeshift shelters, “showing” the existence of refugee camps, whereas, in reality, people were not gathering, but rather dispersing. Most of them wanted to stay as close to their homes as possible and were living with neighbours or family and moving back and forth between their former homes and their temporary accommodation. The same goes for the destruction caused by the tsunami. In Sri Lanka, for example (except in the hardest hit region in the north), it was concentrated along a narrow strip of land between 50–300 metres wide, depending on the lay of the land where the wave hit. So the survivors were in fact only a few minutes’ walk at most from the unaffected parts of the country, something we couldn’t tell from the pictures we were seeing. This kind of metonymical representation, of which aid workers are as guilty as journalists, is seriously misleading. I should also add that the thousands of Sri Lankan doctors and nurses, who arrived within hours to help their colleagues and fellow citizens, were no more visible as they were indistinguishable from the disaster victims. These misinterpretations were given such credence because they fit in so well with the preconceived notion mentioned earlier of disaster victims in a state of total shock, passively waiting for help to arrive.

As for the orphans described by the director of UNICEF as wandering the streets at the mercy of child prostitution gangs, this was a rumour spread all too hastily, but rapidly dissipated by other humanitarian organisations, and by UNICEF itself. Obviously there is no question that some children had lost their parents, but what I do contest is that they had been abandoned. I should perhaps explain at this point that the post-tsunami solidarity movement, often portrayed in the North as exemplary and cited as a reference, in fact left the concerned countries with memories of an agitated, arrogant and ineffectual mob. Despite its endeavours to distance itself from the prevailing discourse, MSF did not escape from harsh collective judgement.

But let me return for a moment to the schema of war superimposed on that of natural disaster. In spite of images that make them look very similar, they are in fact diametrically opposed. Disasters are concentrated into a very limited time period and a very restricted geographical area, whereas armed conflicts are spread over an extended time period and wide geographical area. Wars are drawn out affairs, erratic in their movements, killing and injuring in their path, causing the displacement and re-assembly of populations between one region and another, creating intense and relentless pressure, rampant and massive impoverishment and wide-spread destruction, including of health facilities. These vulnerability factors, producing all these effects and creating a high potential for epidemics, cannot be caused by a one-off event. A natural disaster, however horrendous, cannot engender the same consequences as a war.

Is there a clear association between the myths surrounding events after a disaster and political situations?

International emergency aid is loaded with a specific kind of symbolism that has nothing to do with its real usefulness, as we have just seen.

It is inevitably an intrinsic part of the pre-existing dynamics of international relations—and becomes an extension of them. For example, when Iran was hit by an earthquake in June 1990, the French government offered to send in specialised teams, even though the two countries had broken off diplomatic relations: the emergency aid brought to light the fact that Paris and Teheran had secretly resumed talks. The same can be said of China sending a plane full of aid supplies to Haiti after the earthquake in January 2010, in spite of the absence of diplomatic relations between the two countries owing to Haiti’s recognition of Taiwan. This was a first. Beijing had never before contributed towards disaster relief operations outside its regional sphere of influence in Asia. However, the fact that China now wants to assert its status as a global power meant taking part in the international relief effort. Just as the earthquake in Pakistan proved the existence of a “disaster policy”, there is also “disaster diplomacy”, whereby the special circumstances created by an emergency allow governments to dem-onstrate their strategic choices at little cost.

In this respect, the case of Cyclone Nargis, which hit Myanmar in 2008, merits attention. In May 2008, the Irrawaddy delta was swept by winds reaching 240 km per hour, followed by a wave four to six metres high, which surged up the river resulting in extensive loss of life and massive destruction in this densely populated and fertile region. The Myanmar junta, faithful to its obsession with maintaining order and as ever indifferent to the fate of its people, did not react, simply appealing to the United Nations for international aid and refusing any new foreign presence on its soil. However, right from the first few days, members of MSF and other NGOs already working in the country were able to travel to the area, assess the extent of the damage and launch the relief effort with the local resources at hand. At the same time, planes from neighbouring India, Thailand, Bangladesh and Malaysia, as well as from western counties acting on behalf of UN agencies, were landing in the capital city, Yangon. In the meantime, the press and western governments, apparently unable to see beyond the junta’s sovereigntist and isolationist rhetoric, were talking about restrictions and even a total blockade of outside aid. On 11 May, the NGO Oxfam issued a communiqué and the first few lines set the tone: “International agency Oxfam said today (11 May) that in the coming weeks and months the lives of up to 1.5 million people are in danger in the Myanmar cyclone zone because of the risk of disease and a public health catastrophe if clean water and sanitation are not urgently provided”.

Seen from the field, this scaremongering was far from justified. It was true that the army had been seen diverting aid for its own purposes or to make a profit out of distributing it but, as always, the population got itself organised on different levels. Local organisations and authorities, the Red Cross, Buddhist temples and wealthy businessmen all distributed water, food and equipment, and foreign aid began arriving via the NGOs. As for the injured and the threats of epidemic, I repeat what I said earlier about the tsunami; they were non-existent.

It was striking that most of the television coverage, whether videos made by local people or official television reports, all showed scenes of aid distribution almost everywhere. We saw endless short scenes of businessmen arriving with their lorries and handing out bottles of water, sacks of rice, etc. Elsewhere, Buddhist monks were similarly shown, as was the army, an NGO or the Myanmar Red Cross. Basically, we were seeing the usual images of food distribution and, here and there, one or two bodies. Watching the media coverage attentively, I realised that the commentaries accompanying the pictures were actually contradicting everything they were showing, insisting on the total absence of aid and the numbers of decomposing bodies, which were described as bacteriological time bombs on the brink of spreading their deadly emanations. When I asked some journalists during interviews on the subject what they thought about the dissonance between the pictures and the commentary, they said they hadn’t noticed it and were obviously suspicious of any challenge to the general alarmist view.

So it was in this context that threats of military intervention to impose aid by force first began to appear in the press. Gareth Evans, one of the authors of the UN’s “Responsibility to protect” concept, started the ball rolling on 12 May,Gareth Evans, “Facing up to our responsibilities, The Guardian, 12 May 2008followed two days later by Robert Kaplan, one of the most prominent neo-conservative strategists, who sketched the outline for armed intervention in an article entitled, “Aid at the Point of a Gun”.The New York Times, 14 May 2008.

 And on 19 May, French foreign minister Bernard Kouchner published an article reminding us that “the Secu-rity Council can at any time decide to intervene to force a passage for humanitarian aid, as has been done in the past”.Bernard Kouchner, “Birmanie: morale de l’extrême urgence”, Le Monde, 19 May 2008.

Three military vessels, British, French and American, were thus hastened to the Myanmar coast as a sign of their governments’ determination to prevent the supposed deaths of hundreds of thousands of innocent people.

It must be said that this time, unlike after the tsunami, the WHO posted on its website that corpses posed no risk and that survivors of the cyclone were in no danger of a deadly epidemic. But this was not enough to prevent the British Foreign Office from warning of the “peril”, or to dissuade the advocates of armed interventionism, governments and associations alike, from using it to encourage the Security Council to activate the “Responsibility to protect” mechanism.John D. Kraemer, Dhrubajyoti Bhattacharya, Lawrence O. Gostin, “Blocking humanitarian assistance: a crime against humanity?”, The Lancet, Vol. 372, 4 Oct. 2008.

Until the war in Libya in March 2011, instigated by the same governments (France, UK and the US), this was the only debate in which the Security Council had actually envisaged implementing this mechanism.

Did emergency relief organisations learn any new lessons from the earthquake in Haiti?

The January 2010 earthquake in Haiti was the second mass medico-surgical emergency after the one in Pakistan in 2005. MSF had been working in Haiti for several years when the disaster struck and so was in the right place to respond rapidly. Three surgical units were set up in a container and the first major operations were performed three days after the earthquake. During the first forty-eight hours, care had been provided in the streets. The inflatable hospital used in Pakistan was sent out, so we were operating in optimum conditions from day thirteen, which is the time it took to get this really imposing piece of equipment on site and up and running. By the way, the famous fortyeight-hour window beyond which casualties cannot survive can now be filed away under “conventional wisdom”, as the Pakistan precedent had already confirmed. MSF thus took up position alongside the multitude of local and international organisations, governmental and private, which had rushed to set up operations in Port-au-Prince and the surrounding region during the two weeks following the earthquake.

There was a lot of talk at the time about the chaos in which the “humanitarian expeditionary corps” was deployed. The lack of coordination and information on needs and the running of the relief operations were severely criticised in the press, but these criticisms don’t actually hold water. Firstly, because disorder is the hallmark of a disaster, all the more so when it hits a country’s capital and therefore its seat of power. Secondly, because the shortcomings of Haiti’s public institutions were already notorious and the country was without an army, which had been dissolved under US pressure during the “Restore Democracy” operation in 1995. Lastly (and most importantly), because the response to the urgent needs was focused on a limited area, it was carried out correctly, in spite of everything, with the notable exception of the shelters, which were both unsuitable and insufficient.

There are two medically-related issues that I would like to single out: the first, quite specific, concerns the use of techniques derived from war surgery, which tend to be more radical but can be inappropriate; the high number of military surgeons in such a setting, as well as the ever-present juxtapositions with the representation of war as mentioned earlier, give pause for thought. The other issue is more general and concerns the criteria adopted explicitly or otherwise by medical teams from different professional culturesFrédérique Leichter-Flack, “Sauver ou laisser mourir”, http://www.laviedesidees.fr. for deciding which cases, medical as well as surgical, should be given priority and which should not be treated. Do the exceptionally high workload and the logic of rationing induced by a disaster, which is where triage usually comes in, lead to laxity in procedures?http://www.theworld.org/2010/02/doctors-face-ethical-decisions-in-haiti/.

 We have only fragmented and flimsy data, so I won’t attempt to answer these questions. I just want to stress the need for a methodical reflection on them.

Why is estimating the number of victims in a disaster the subject of such frequent debate?

Estimating the number of victims is another major issue as the figure is a crucial emotional marker, the trigger that “allows us to feel the disaster”Sandrine Revet, “Anthropologie d’une catastrophe, Les coulées de boue au Venezuela”, Presses Sorbonne Nouvelle, 2007, p. 267. and determine where it features on the scale of gravity. Unlike what we see in many conflict situations, the disaster toll (usually an approximation) announced by the governmental authorities and the United Nations a few days after the event is accepted by the press and aid organisations as objective, in spite of its unreliability. Three days after the Haiti earthquake, the government announced that 50,000 bodies had been recovered. This figure was to increase day after day to reach 250,000, or even 300,000 a month later,“Death toll from the earthquake could reach 300,000, according to the president of Haiti”, Le Monde, 22 Feb. 2010. making the disaster one of the most serious ever.

These evaluations were based on an estimate of population density and the number of collapsed buildings in a given district, which left considerable room for uncertainty. Respect for the victims does not proscribe challenging figures drawn up in a chaotic environment and with no credible foundation. Heads of some of the UN agencies encountered six months after the earthquake privately agreed on a death toll of somewhere between 50,000 and 70,000, based mainly on the number of mass graves dug by Minustah, the only organisation charged with the task.Encounters by the author in Port-au-Prince in June 2010.

Similarly, after a survey of the different actors in their field, Handicap International Belgium’s head of mission estimated that the number of disabled people was closer to 1,000 than 5,000, the figure that had ended up becoming official after being bandied about in aid circles.

Reducing the estimated loss of human life is clearly a sensitive issue, as it ties into collective emotion. Bringing down the numbers exposes us to suspicions of hard-heartedness, or even hostility or shameful ulterior motives, whether in situations of natural disaster or, even more so, in other settings with a more direct political dimension, such as armed conflict, population displacement, or the quantification of atrocities.Peter Andreas, Kelly M. Green (eds), Sex, Drugs and Body Count, The Politics of Numbers in Global Crime and Conflict, New York: Cornell University Press, 2010.

 The death toll after the earthquake in Armenia in 1988, established at 100,000 deaths a few weeks after the disaster, was later reduced to 23,390 in figures published by the authorities. This reduction in the official death toll produced reactions of incomprehension, hostility even, as the original number had become a symbol of Armenian suffering and changing it was seen as a denial of this suffering. In practice, it is likely that such distortions and amplifications abound in many similar situations.

Estimating the number of victims—and the number of fatalities among them—is most definitely not a superfluous exercise, not only because this is the first question that everybody asks, but particularly because, however vague and fluctuating it may be while the aid is being set up, it allows a threshold effect to operate. It has been observed that we reason in terms of a major disaster justifying international-level deployment when the death toll reaches or exceeds ten thousand. The practical importance of such estimates from a relief agency’s point of view is, however, limited, but I raise this issue here to underline how the highly uncertain nature of the figures makes it extremely difficult to know which resources to activate, other than basing ourselves on the threshold mentioned earlier. In concrete terms, the specific information required to guide relief operations would be, on the one hand, the number and condition of the survivors in order to gauge the medical assistance requirements, as well as needs for other types of aid—shelter, food, water, telecommunications, damage clearance and transport—and on the other, information on what the other relief operators, local as well as international, are doing.

Although not victims of an executioner, disaster victims are caught up in high stakes, as the examples given above have shown. Funding, media coverage, rallying sympathy for traumatised people, all combine to produce an escalation that nobody plans, but which is fostered by the apparently indisputable nature of the cause defended—that of increasing emergency aid as high as it can go.

Translated from French by Mandy Duret