These figures do not contradict each other. They measure different populations, over different areas, across different periods, with different methods. Set side by side, what they show is that no one has ever been mandated to count Wollega as a whole.
| Source and method | What it counts | Figure | Status |
|---|---|---|---|
| ACLEDMedia-coded events. Export of 3 September 2026. | All conflict deaths, all parties, four Wollega zones, 2018 to August 2025 | 4,646 | Recorded floor |
| ACLEDSame export, one event category. | Events coded as violence against civilians, outside combat, same area and period | 1,997 | Recorded floor |
| AAGE registerNominal identification, entry by entry. | Individuals named, across the documented massacre sites | 1,426 | Method note pending |
| Amhara Association of AmericaAdvocacy documentation. | Amhara civilians, the whole of Oromia, year 2022 only | 1,566+ | Corroborated in part |
| OHCHRUnited Nations reporting. | Killings in Amhara and Oromia regions together, year 2023 only | 1,106 | Confirmed |
The instrument matters more than the total. For the war in northern Ethiopia, ACLED recorded 9,861 deaths, while a peer-reviewed study led from the London School of Hygiene & Tropical Medicine calculated more than 102,000 excess deaths in Tigray alone, of which 72 per cent from violence — a divergence of an order of magnitude between two instruments that do not measure the same thing. Event coding counts deaths tied to reported incidents; a population-based study estimates excess mortality, including deaths no single incident can carry. The two are not a ratio. What the comparison establishes is a measurement problem, on the one Ethiopian conflict where both instruments were applied. No such study has ever been conducted for Wollega. Every figure above is therefore a floor, and the distance between them is not a dispute about facts. It is the shape of an investigation that has not taken place.