“Those who have not been seen”, Starvation in the Northwestern Zone of Tigray

By Duke Burbridge

After five months of suspended food aid, starvation is everywhere in Tigray, but conditions are particularly bad in the Northwestern zone. The World Food Program’s (WFP) estimate from March 15, 2023 was that 1.08 million people in the Northwestern zone were in urgent need of food assistance. The WFP uses a population estimate of 1.14 million, so 95% are either hungry or starving to death. In a recent interview, Biruh Zeferu, Western Tigray’s Head of Public Mobilization, told Tigrai TV, that he was receiving about three reports of starvation-related deaths every day just from the IDP population of the Northwestern zone. The number of families starving to death in their homes are unknown, particularly in the areas still under military occupation by the Eritrean military or Amhara regional forces. 

The factors and conditions leading to starvation-related deaths in Tigray are complex, but rarely mysterious. During the occupation of Tigray, pro-government military and paramilitary groups committed horrible atrocities against the civilian population, including androcide and widespread sexual– and gender-based violence, as well as the systematic destruction of the food, agricultural, and health systems, widespread looting, and a humanitarian blockade. All of these factors contributed to the displacement of millions and food scarcity and starvation in Tigray for about two years. 

The Cessation of Hostilities Agreement (CoHA) signed last November in Pretoria was supposed to end the outside occupation of Tigray, the violence and looting, and the blockade. This would have, in turn, ended the displacement of more than a million civilians and allowed international aid agencies to step in and deliver life-saving goods and services to families in Tigray who currently lack the means of sustaining themselves after two and a half years of siege warfare. 

Unfortunately for Tigrayan civilians, there has only been meager progress in the past year towards ending the military occupation of Tigray by the Eritrean military in the north and Amhara forces in the south and west, where they continue to commit ethnic cleansing and block aid. In the unoccupied areas of Tigray where humanitarian access was eventually restored last year, food aid was only brought to scale in March. Then, at the end of March, food aid was halted, but this time by the WFP and USAID. Due to the ongoing occupation and aid suspension, most of the currently displaced population cannot return to their homes and farms. 

Food and nutrition aid in the Northwestern Zone

The week prior to the suspension of food aid was the first time since the war began that the WFP met their six-week food distribution objective for the Northwestern zone. While this was a remarkable achievement, many people in need did not receive food. It is worth noting that there was also a major theft from a warehouse in Sheraro, but that should not have affected the Food Security Cluster’s weekly distribution numbers in the chart below because the food was never distributed. By the first week in May this achievement was almost completely erased and by June no one in the Northwestern zone had received any rations in the past three months. 

While WFP and USAID both claimed that nutritional services for children at the final stages of starvation would continue during the food suspension, they only actually continued for a month. The chart below shows the decline of access to life-saving services for the children of Tigray, the Northwest zone, and select districts and towns that remain accessible (Shire, Sheraro, Endabaguna, and Zana¹These locations were chosen not only because they were accessible, but also due to the fact that the WFP and IOM use different sets of Woredas in the Northwestern zone with slightly different boundaries. It is my understanding that WFP’s is correct. The number of directly comparable woredas and towns are therefore limited. ) As shown below, the drop has ranged from 86% to 93% based on the latest available data from the Nutrition Cluster. In the displaced population alone, not counting host communities or returnees, more than 39,000 of the  48,000 children under five years old live in these towns and woredas. In August, only 229 children were admitted for SAM of which 79 were currently displaced, which suggests that only 0.2% of the 39,000 babies and toddlers who are displaced in the Northwestern Zone, have access to life-saving treatment if they slip into the last stages of starvation, God forbid. 

The number of children admitted for SAM in March was not enough to cover the need, despite the fact that WFP had met its food distribution objective for the first time since the beginning of the war. By July, the need would have been significantly greater and the failure to deliver nutritional support would have been significantly more lethal for the children of Northwestern Tigray. 

The obvious cause of the sharp decline in SAM admissions is a lack of supplies. With no formal acknowledgement from WFP, USAID, or UNICEF that treatment for child starvation has been effectively cut in Sheraro and Endabaguna and significantly reduced elsewhere, the cause of the disruption remains unknown.

Finally, the provision of supplemental food for treatment of moderate malnutrition in children and acute malnutrition in pregnant and lactating women (PLW) may have been cut completely, but there are also so many data irregularities that it is impossible to tell. According to the Nutrition Cluster in March more than 200,000 children and PLW were reached with treatment for acute malnutrition in Tigray. This number fell by nearly half in May to 100,000, then to 47,000 in June, then in July no services were reported in Tigray. This may be an error or failure to report, but it is a serious cause for alarm five months into a food aid suspension.

Food aid targeting

The food aid response in Tigray is split between the WFP and the JEOP. The WFP covers the entire Northwestern zone, Mekelle, and the lower half of the Southern zone, while the JEOP, led by Catholic Relief Services, serves everywhere else, except the Western zone, which the federal government of Ethiopia claims to serve, but they do not generally track or report food distribution.

When reporter Abel Tsgabu (@AbelTsgab) from Tigrai TV went to the Northwestern zone recently, displaced families in Shire and Sheraro told him that it had been eight months since the last time they received any aid. According to a multi-sectoral assessment performed in Endabaguna Town in April, only 5,453 IDPs had received food aid in the last distribution round. This is only 78% of the IDPs who had been living in Endabaguna for the past two years and less than 20% of the town’s current IDP population.

These reports are very alarming, because, in March alone, the WFP reported distribution of 1.08 million food rations in the northwestern zone, according to the Food Security Cluster. This should have covered their entire caseload. There should be no large, displaced populations in accessible areas of the Northwestern zone that did not receive food assistance this year. 

So why didn’t they get aid? 

It must be acknowledged that people in desperate situations where their families are starving sometimes lie. This seems unlikely. At least in the case of Endabaguna, the aid distribution data were provided by aid providers, rather than beneficiaries. Additionally, reports of people not receiving aid since December are widespread, even beyond the Northwestern zone, which suggests that people are telling the truth about not receiving aid. This timing is surprising, because WFP Executive Director Cindy McCain said that the majority of diversion happened in December and January. However, it appears that many of the most vulnerable people, at ground zero of the diversion crisis, are reporting that they received aid at this time of elevated diversion, but not after.

Nevertheless, the aforementioned reports of theft and diversion strongly suggest that WFP is having problems along the supply chain. However, this is also a problematic explanation in this particular case. Sheraro was the site of a major theft of food aid from a warehouse this year, but the Food Security Cluster’s distribution data would not have been affected because the food in a warehouse in Sheraro should still have been considered ‘dispatched’, not ‘distributed.’ If food aid is considered by WFP to be distributed when it arrives in a warehouse in Sheraro, then the entire food distribution monitoring system is flawed. Generally speaking, the Food Security Cluster uses more sound methods. If food was distributed to beneficiaries who then sold some of it for other household needs like medicine or in exchange for milling services, then people would have still, theoretically, would have reported receiving aid. 

A more likely explanation becomes apparent through comparison of (a) the WFP caseload for food assistance in the Northwestern zone against (b) the actual need. The WFP caseload for the Northwestern Zone can be found in the Nutrition Cluster’s June 2023 Advocacy paper. This appears to be the only public document to provide a district level breakdown of the WFP caseloads for food and nutrition aid in Tigray. This data is all estimated and not constrained by access challenges. The chart in the annex of this advocacy document shows how the EFSA is used to determine the total population in need. The data from the WFP woredas in the Northwestern Zone includes their estimates of the displaced population of each district. 

The actual need for food and nutrition services is more difficult to determine with any accuracy. However, a recent study by the IOM shines a new light on displacement in Tigray, making it possible to verify the assumptions made in the advocacy paper about the IDP population. Round 33 of the IOM Displacement Tracking Matrix (IOM-DTM) IDP Site Assessment was fielded between April and June 2023. This was the first round to include Tigray in two years. The assessment was limited in geographic scope due to the military occupation of large areas of Tigray by the Eritrean military and Amhara regional government. However, it is the most recent, geographically broadest, and most comprehensive study of displacement in the accessible districts of Tigray since the resumption of violence last August. 

The first thing that stands out from a comparison of the data in these three documents is that the WFP is using an estimate for the IDP population in the Northwestern zone that is significantly lower than what the IOM found in June. According to the IOM assessment of IDP sites in Tigray, there were around 316,000 IDPs in 140 hosting sites in the Northwestern zone. The IOM assessment was comprehensive, but not exhaustive. The IOM team was not excluding areas that are under control of the Eritrean military in the north or the Amhara regional forces and Fano militia in the south. The WFP estimate of the IDP population of the Northwestern Zone is much lower, less than 237,000, which includes areas along the border with Amhara and Eritrea that were not accessible to the IOM assessment team. If the areas excluded from the IOM assessment were also excluded from the WFP assessment, the WFP estimate falls to around 180,000. This would mean that, according to the IOM’s figures, and estimated 135,000 IDPs in the Northwestern Zone are effectively cut off from food or nutrition assistance in unoccupied areas, regardless of the suspension.

In the four areas from the previous chart, the estimate appears to miss more than 80% of the IDP population. This strongly suggests that the amount of food being provided prior to the suspension fed a fraction of the actual number of displaced people who urgently needed food. 

The IOM estimate appears to be sound. For example, the IOM assessment found that there were 27,766 IDPs being hosted in Endabaguna, of whom around 8,300 were children (0-14 years old). This total was verified by a Tigrayan source who visited family this month in Endabaguna, who reported that the current number of IDPs was around 28,000, with newly displaced families arriving daily (on September 19, for example, 10 new IDPs arrived.) 

WFP’s failure to integrate IDP data could have grave consequences in Endabaguna judging from their reported condition during an assessment conducted in April by a ‘multi-sectoral’ team that included 31 staff members from the major thematic aid clusters (food, camp management, health, etc.) and found that “zero food was distributed and IDPs and partners claim that there is mass starvation.” More than a month later, Mihet Kristos at the Addis Standard interviewed the Interim Administrator of the Northwestern zone who reported that the IDPs in Endabaguna had still not received food. My source spoke with aid workers who reported that WFP officers come occasionally for observation, but still, no food has been distributed. For this population in urgent need of support, my source was told that healthcare is available on a very limited basis and the UNHCR has provided IDPs with 1,000 Ethiopian Birr for rent for IDPs, about $18. That is the extent of the humanitarian response. The source also reports that 18 civilians had recently died of starvation in Endabaguna, including young children.

Not counting IDPs 

This is unfortunately not the first instance of WFP missing the IDP population in the Northwestern Zone. The two previous food security assessments of Tigray from last year did not reach enough IDPs to draw any conclusions about the most vulnerable populations. Both times, there were UN datasets and alternative methods readily available to close this gap. It remains open because it is not being taken as seriously as it should be given the grave implications. 

Only 4% of households sampled in WFP’s January 2022 Emergency Food Security Assessment (EFSA) were currently displaced. This important limitation was acknowledged in the analysis, which reasoned that the “lower than expected (sic)” representation of displaced households was because respondents “strongly identified as ‘part of the local community’ if even one household member remained in the community during the past year.” (p.12) In WFP’s June 2022 EFSA (published August 2022), representation of displaced households fell to just 2% of the sample. The assessment team again described the “lower-than-expected” representation of displaced households and again theorized that many respondents were displaced but would not admit it because they “strongly identified as ‘part of the local community’”. (p.11) 

The most recent EFSA was conducted in February 2023 and included a dedicated IDP cluster. This was the largest food assessment conducted in Tigray by the WFP since the war began, reaching 4,360 households in 48 woredas across Tigray including IDPs from five urban areas. This was also the first EFSA that WFP declined to publish. The IDP data is highly problematic because the IDP cluster is too small. In the Northwestern zone, IOM documented 140 IDP hosting sites in 12 districts; the February EFSA only included five IDP camps in one town, Shire. Conditions in Shire are bad, but it is also fully accessible and serves as a hub of humanitarian operations in the Northwestern zone. The IDP experience in Shire is not generalizable. Conditions 60 miles west in Sheraro are similar in some ways but different in others, same with Endabaguna to the south of Shire and the district of Zana in the east. 

Only three IDP sites in Shire reported food distribution on site in the past three months, which is an alarmingly small number. However, they were the only sites in the cluster to have reported any distribution in the past three months. The table below is from the IOM-DTM dataset. It shows that only one out of every three children under five who are displaced in Shire are registered to hosting sites where more than 25% of households received food during the last distribution. However, the data also shows that Shire is host to the only IDP sites in the cluster where food was distributed to more than 25% of the population. Almost all children in the cluster who are displaced outside of Shire are registered to hosting sites where the last food distribution either never happened or was too long ago to recall. 

As the operational lead for the food response in the Northwestern Zone, WFP cannot perform their critical role in the prevention of starvation deaths with: (a) an estimate of the IDP population that is significantly lower than the actual population and (b) a food security assessment that overlooks the IDP population. These two issues are not related to diversion of aid and they appear to be having a severe impact on conditions of life for the people who are trying to survive in the Northwestern zone. 

The Human Toll

According to the recently released findings of the first medically-verified study of starvation-suspected deaths in Tigray from November 2022 to July 2023, there were more than 1,300 starvation deaths in just nine woredas and the IDPs populations of five towns. There is also cause to believe that this figure is rapidly increasing. As this chart from the study shows, the monthly death toll was lowest in March before the suspension of food aid. Deaths increased slightly in April and May, and then sharply in June and July. 

In a Telegram exchange with another Tigrai TV journalist, Mustofa Tuha, offered this insightful reflection after a recent visit to Shire: “I mean those 600 people died in Shire alone. In this way, there are many known and unregistered people who have died on every street, those who have not been seen, and who have not been registered, but those who die of hunger are many. And they say displaced people are scattered all over Tigray outside Shire.” 

Sadly, Tuha’s fears are valid. According to the recent-starvation suspected mortality study 90% of deaths since the Pretoria Agreement (November 2022) occurred outside of medical care. The real number of deaths will never be known, as Tuha correctly notes, because most of the people who are displaced in the Northwestern zone are either unregistered or if they are registered, they lack access to food assistance. And he is right, Tigrayans are not just displaced to one or two population centers. To say that displaced families are “scattered all over Tigray” is an accurate bottom-line of what IOM found in the 643 sites that they assessed. Tigrayans who are displaced are also largely unseen, in my opinion, because the WFP is not looking hard enough for them. Tuha went on to say that now between 3 and 5 IDPs were dying in Shire every day. 

Zana was only of the nine woredas included in the study of starvation related deaths in Tigray. While the preliminary findings presentation only breaks down total deaths by woreda (as opposed to starvation-related deaths), Zana was found to have had a higher mortality rate per 100,000 than Samre and Seharti, where starvation and child mortality has been extensively documented already.  

WFP and USAID took drastic action to cut food aid to Tigray to minimize the risk of losing any more supplies. When it became evident that the access to life-saving treatment for severe malnutrition would no longer be available to Tigrayan children at the last stage of starvation, WFP and USAID should have taken similarly drastic actions to minimize the risk of losing any more human lives.