Only the most trusted sources should be used for death tolls, and the most conservative estimates are often the most accurate.
Clionadh Raleigh is President and CEO of the Armed Conflict Location & Event Data Project.
When Hamas fighters attacked Israel on October 7, they didn’t just target Israeli army bases or civilian towns and homes. They also struck a weak spot in the popular perception of war: a global fixation on fatality figures.
Hamas’ apparent order of the day — to kill and capture as many Israelis as possible, whether military or civilian – was a terrorist act by any definition, and its brutality multiplied the shock. However, it would be a mistake to call this nihilism, as United States President Joe Biden did in an editorial. For as both Biden and Hamas know, in many cases, people only notice a conflict when struck by a death toll.
I say this because it’s my job to count. It was the astonishing unreliability of fatality statistics that years ago coaxed me to start collecting data on political violence in six African countries. The Armed Conflict Location & Event Data Project (ACLED), which grew out of that work, now tracks 243 countries and territories worldwide. And over the years, we’ve recorded over 1.5 million instances of political violence and protest events, making them publicly available for anyone to search and see — including fatality estimates.
If there’s one thing I’ve learned on my journey through the statistics of global conflict, it’s that fatality counts are usually the most inaccurate measures of violence. They’re biased, often taken out of context, and far more people are exposed to conflict than are killed.
Only the most trusted sources should be used for death tolls, and the most conservative estimates are often the most accurate. Thus, we must constantly update our criteria, sources and assumptions.
In my early interviews with local African officials, trying to get to the bottom of conflict reports, I often heard that the time, place and location of a reported event was correct. However, fatality reports varied greatly, and some individuals even admitted they had made them up for reasons like receiving money from their capitals, troops from headquarters or attention from faraway media.
And these local officials were right: The bigger the claimed death toll, the more attention they got. We struggle with aggregate death tolls resulting from authorities rounding up numbers to end in several zeros. And because they are “official,” and can’t be proved or disproved, these “zombie death tolls” quickly take on a life of their own.
For example, in his article on the Gaza conflict, Biden cited the Israeli civilian death toll at 1,200, but didn’t mention the quickly growing number of Palestinian deaths, which he later stated he had “no confidence” in. By then, the Hamas-aligned Palestinian Ministry of Health had counted 11,675 Palestinians killed in Gaza since October 7. ACLED’s provisional count for that time period is 8,754 for Gaza and the West Bank, and 1,189 for Israel — but our figures will rise as we’re able to link fatalities to specific events.
Biden’s choice of data underlines the point that as the political temperature climbs, fatality figures just become part of the conflict itself. And the weaponization of death tolls only exposes and deepens existing divisions rather than change minds or help resolve conflicts.
When planning its October 7 attacks, Hamas must have expected the furious Israeli bombing of Gaza and hoped that this would put the militant group itself, its goals and Palestine back on the global agenda. And if so, its gamble on the lives of its own civilians has paid off: By ACLED’s count, 90 percent of some 4,200 demonstrations around the world this last month were pro-Palestine.
Hamas may be losing the conflict, but it’s winning the argument.
I want our statistics to help build a more peaceful world and offer clarity by recognizing the differences between violent conflicts.
In our 2023 Conflict Index (updated prior to the Hamas attacks), Ukraine was the deadliest country in terms of accumulated fatalities, but our metrics also highlighted conflicts that many miss. For example, the number of active armed groups propelled Myanmar to the top of the overall list. Meanwhile, Mexico was the most dangerous country for civilians, measured by the number of direct attacks on unarmed groups and individuals.
Crucially, however, Palestine topped the list of places with the widest diffusion of conflict. High levels of violence — mostly confrontations between Palestinians and Israelis — had affected over 60 percent of Palestinian territory. Yet, at the same time, fatalities were relatively low. And tragically, while this was a clear warning sign in the data, it was one that most people ignored.