Djibouti shipwreck toll rises to at least 139, with 84 still missing

The boat was travelling from Djibouti toward Yemen when it capsized near Guhere, also spelled Guehere, north of Obock. Survivors placed the sinking on the night of October 8–9; IOM’s follow-up gives the same overnight timing. A UN News account describes the early hours of Friday. October 11 is the date of the updated assessment, not the date the vessel went down. No precise capsizing time or wreck coordinates have been established.⁠1⁠2⁠11

IOM, the UN migration agency, explicitly attributes the rise in deaths to the recovery of 30 additional bodies. Its initial assessment reported at least 109 dead and more than 116 missing. The later figures—at least 139 dead and 84 missing—replace those earlier counts. They describe successive assessments of the same wreck, not separate groups of victims to add together.⁠1⁠2

The first count of the dead included 52 men, 42 women and 15 children, a breakdown also reported by the Associated Press. Those categories total 109 and belong to the initial assessment. Later reporting still identifies 15 children among the dead, but supplies no revised count of adult men and women or demographic breakdown of the additional recoveries.⁠4⁠7

Even the starting number for accounting for everyone aboard remains unsettled. IOM gives 305 passengers in both its original and follow-up assessments. AP also reported 305, as did AFP’s October 11 account carried by France 24. Xinhua’s October 11 report, however, gives 350, and the BBC also reports 350. The competing estimates have not been reconciled.⁠1⁠2⁠4⁠7⁠15⁠17

That disagreement matters independently of the changing toll. An earlier death count can be superseded by additional body recoveries, as IOM explains here. The passenger discrepancy has no equivalent explanation. Without a verified manifest or an account of how the estimates were produced, 350 cannot be treated as a confirmed total or silently substituted for the agency’s 305.

The survivor figures also require care. The initial IOM account reported more than 80 survivors, identified as Ethiopian. A later UN News account, republished by Scoop, says about 82 people survived. “About” is important: it is an estimate, not an exact final count. AFP describes the passengers as mostly Ethiopian, which does not establish that everyone aboard was Ethiopian or provide an updated nationality breakdown of survivors.⁠1⁠11⁠7

IOM initially said it was caring for 64 survivors in Obock. Those 64 were part of the survivor group, not another group to add to it. That was an earlier snapshot of assistance, not evidence that exactly 64 people remain in care now. No newer care count, complete account of survivors’ medical condition or explanation of where every survivor was taken has been established.⁠1

These distinctions are not bookkeeping for its own sake. The initial figures were lower bounds, while the later survivor figure is approximate. Forcing them into a precise passenger ledger would suggest a certainty the reporting does not support. A reliable accounting must establish who survived, who died and whose whereabouts remain unknown—not merely produce numbers that add up.

In IOM’s initial response, Alexandre Coissac acknowledged both bereaved families and those still waiting for news. His full institutional title is not established in the available attribution.⁠8

The cause of the capsize remains unknown. AFP describes the boat as overcrowded, but that description is not a confirmed causal finding. The available reporting does not establish the vessel’s capacity or a sequence of events explaining why it overturned.⁠7

IOM initially described the wreck as the deadliest ever recorded off Djibouti. That is a claim about the recorded history off this coast, not a worldwide record or a complete inventory of unrecorded deaths. Its later description calls the incident one of the deadliest recorded shipwrecks, without supplying a complete historical comparison. The supported toll is already catastrophic without extending the record claim beyond its evidence.⁠1⁠2