Source: Emnet Negash

Take-home message: As our estimate of the civilian deaths in the Tigray war is regularly mentioned in the media, it seems important to share our evolving understanding and updated (lower) number of civilian deaths as a result of the Tigray war and blockade. We concluded that the IPC/FEWS categorization, on which our Tigray statistics are mainly based, overestimates hunger mortality. Along with developing information on the ground, this would point to a total number of civilian deaths ranging from 162,000 to 378,000. This is still a staggering figure, and the revised estimate does not acquit the Ethiopian government and its allies of the premeditated war crime of besieging and starving the population.

Several international media articles have given attention to our estimate of the civilian deaths in the Tigray war. We have always given a range of civilian deaths, roughly between 300k and 600k civilian victims, indeed in Tigray alone. Progressive insights, particularly related to the number of famine deaths, may indicate an overestimation of famine deaths in the IPC/FEWS classification. Jointly with emerging evidence on the ground, this would hint at a total number of civilian deaths that could rather be on the lower side of our earlier estimated range.

The IPC issued a famine warning in May 2021. They have detailed and structured data around how many people are affected by hunger in Tigray. Although not perfect, this gave us a starting point to calculate how many people might have died of starvation. IPC stopped providing data afterwards (due to inaccessibility to the region), but we had FEWS NET and WFP reports to work with as well, although they are less detailed.

Now, with better access to Tigray, there are some sample areas where civilian casualties were tallied. Christian Putsch, for the German newspaper Die Welt, visited the tabiya (municipality) Tashi in the Saharti woreda (district), where the tabiya leader had tallied all civilian victims of the war: 350 dead, of which 50 in a massacre and the others by starvation and lack of medication. That is one out of 30 inhabitants.

In the Dogu’a Tembien district, we had our own qualitative investigations on the impact of the war on the social coherence and the natural environment in ten, generally small villages. The following civilian casualties were mentioned:

  • Atsa: “There was no health service during the war. Many people have died due to lack of medicine and medical treatment.”
  • Addilal: “Six civilians were killed by soldiers and an airstrike;” “Several mothers died during their delivery. Very recently, three mothers died. In the last two years, about 20 community members passed away [due to war conditions].”
  • Addi Qoylo: “Several civilians who escaped first and returned later to inspect their homes were killed by soldiers;” “Though most of the village community was displaced, the few people who stayed at home, were killed by the soldiers.”
  • Miheni: “Many civilians were killed by soldiers;” “The soldiers killed four people (from this village only) who were on their way to escape to the caves to save their lives;” “We were denied burying them. We buried the bodies after 7 days. One of the bodies was eaten by hyenas.”
  • Kolal: “An airstrike killed two people of the village, a man and a girl who were in the market in Togogwa and wounded many.”
  • Awulo: “The Eritrean soldiers killed 14 residents in the village;” “My son fell down from the cliff and died while escaping from the soldiers.”
  • Zeyi: “There was critical shortage of medicine in the last two years. Villagers died because there was no access to health facilities.” 
  • Gumuara: “One day, Eritrean soldiers passed through this village. On their way they encountered a priest who was going to the church and they killed him.”
  • Aregen: “Some people have been killed;” “Fourteen farmers have died due to hunger;” “About eight mothers died while delivering due to shortage of medication.”  

Except for Aregen and Addilal, these are all small villages with 100-200 households. The stories are harrowing, everybody is bony skinny, but fortunately nothing in the interviews would point to a decimation of the population, i.e. 10% of the inhabitants losing their life.

From these reported numbers, the sample sites tend to indicate between 2% and 5% of civilian deaths per location. Of course, this is only a very small sample, and there are several caveats.

Tallying of death toll by local administrators or inhabitants may miss deaths due to hunger or inaccessibility to healthcare, because the cause may go unnoticed. The IPC approach takes such cases into account, in a statistical way.

Further, in several villages, the situation was and is still very bad and at the worst of the blockade, people have moved to towns or even to the Amhara region and Eritrea, hoping that they could better survive there. Deaths along the road or at the place of destination would not have been accounted for. For instance, in the Aregen tabiya in the Dogu’a Tembien district, the inhabitants told us that 600 people (some 10% of the population) had left the tabiya in search for survival.

These reports, from Saharti and Dogu’a Tembien, concern better accessible areas; the situation is thought to be worse in the low-lying drier woredas, which often happened to be the most severe battlefields, such as Kola Tembien, Weri’i or Zana; and places such as Irob where no humanitarian access is allowed to date. We also have no information at all from the areas that are still under Amhara or Eritrean occupation.

Reversely, our own research in Dogu’a Tembien brought us to Dabba Selama, one of the better endowed villages. There, the earlier food reserves of the farmers were not discovered and destroyed, and the local solidarity and age-old survival skills made that no starvation deaths were reported.

Overall, the samples show an estimate of 2% to 5% civilian deaths per location, which is well below our earlier estimate of 5% to 10% that heavily drew on death tolls associated with the different phases of the IPC classification.

An indicator that possibly too high death tolls are associated with the IPC classes is the recent study by the London School of Hygiene & Tropical Medicine in collaboration with WHO and UNICEF on excess deaths due to starvation in Somalia. In this study, they tried to estimate the overall excess deaths in 2022 compared to the death rate in 2016, before the drought. They had access to hundreds of mortality reports from hospitals and health centres, and they could also reference their results with historical famine data. The researchers acknowledge that the data is not 100% accurate, but it can be used as a ‘what is known’ scenario. Here, the death rates associated with the 2022 IPC data on Somalia would project at least 115,000 excess deaths (own calculations). The report mentions 43,000 excess deaths, only 38% of the projected IPC data.

Applying a similar correction on our earlier calculations for Tigray (starvation adjusted to 38% of the calculations) results in an estimate of 96k to 218k famine-related deaths by December 2022. Adding 30k-100k deaths due to lack of healthcare (hence, halving our earlier estimate) and 36k-60k deaths due to direct killings of civilians (massacres, bombardments, …), ends up with a total estimated civilian death rate in Tigray of 162k to 378k people.

The available data and most of the narratives coming out of Tigray, make us believe that the situation is not totally as bad as what the IPC/WFP data (and hence our earlier calculations) suggested, luckily. Yet, 162,000 to 378,000 civilian victims in Tigray is a huge number and this new estimate does not at all absolve the Ethiopian government and its allies from the deliberate war crime of starving the population of an entire region through a near-medieval siege.

Further reading: