Former president of the NGO Médecins Sans Frontières and current director of studies at the MSF Foundation’s Centre de réflexion sur l’action et les savoirs humanitaires (CRASH), Dr Rony Brauman, 67, explains the difficulties involved in obtaining reliable figures on the human costs of natural disasters.
How do you count the human losses related to a natural disaster?
The methods vary widely and are only estimates, which are based on geographic and demographic criteria. We apply a hypothetical attrition rate (percentage of victims) to the population in the affected area.
This was the method used in Port-au-Prince that led to an estimate of around 300,000 deaths after the 2010 earthquake. This method, however, is imprecise in two ways: first, not all countries keep their demographic data up-to-date and second, the percentage of victims varies depending on the time of day the event occurs, the quality of construction, the housing density and the magnitude of the event.
As a result, the death toll systematically follows a steeply rising curve[KLT1] during major natural disasters ( tolls that exceed 1,000 deaths). They climb from an initial underestimate to an overestimate that can be very high. We generally see realistic figures three or four weeks later. This results from technical problems rather than from manipulating the numbers.
Do you know of any cases where death tolls were manipulated?
A government might minimise the figures to save face, or else exaggerate them to gain attention and get more financial and material aid. But unlike the death toll during armed conflicts, like those in Sudan and Darfur, the natural disaster toll is rarely controversial. That’s because they don’t play into a motive to worsen the other side’s barbarous image.
The toll from the 2010 earthquake in Haiti was probably closer to 60,000 than 300,000. Peacekeepers with the United Nations Stabilization Mission in Haiti (UNSTAMIH), who were responsible for digging mass graves since they had construction equipment, estimated that they actually buried 50,000 to 60,000 bodies and that the percentage of victims remaining under the rubble was extremely low.
The Haitian administration didn’t have the resources to be more accurate so we’ve kept this estimate by default because there was no ulterior motive.
In countries with stronger governments, the figures generally drop by a factor of three or four as soon as civil society groups or mortality surveys review the data. Two or three years after the 1988 earthquake in Armenia, the Armenian government reduced the initial figure of 100,000 deaths to 25,000, and there have been many cases where the figures were corrected at a later date.
In mid-December 2017, three months after Hurricane Irma struck Puerto Rico on 20 September 2017, the island’s governor announced that a survey would be conducted in response to claims by NGOs and victims’ families that the death toll of 64 was inaccurate. To my knowledge, this is the first time that disaster-related indirect mortality has sparked debate. This controversy surrounding the official figures is entwined in the complex relationship between Puerto Rico and the United States, which reflects the island’s defiant response to the contempt shown by the White House after Maria. Puerto Ricans, who are American citizens of a “free associated state”, were deeply hurt when President Donald Trump delayed coming to the island and called the death toll negligible compared to the more than 1,800 deaths caused by Hurricane Katrina in Louisiana in 2005.
Indirect mortality isn’t generally taken into account in natural disaster situations because it’s very low, and such disasters briefly strike limited areas, which gives victims access to medical care in unaffected areas nearby. However, we do justifiably count the excess mortality caused by armed conflicts, which have a lasting and profoundly harmful effect on much larger areas where, in addition to the direct casualties, people die from diseases and injuries that would have been treated during times of peace.