Death tolls from Israel and Gaza are important. Can you trust them?

Real-time fatality counts from conflict zones like Gaza are almost never right. Pay attention to them anyway.

Bodies in white bags lie on the ground while a crowd of men looks on.

Dead bodies at the morgue of Nasser Hospital in Gaza on October 25, 2023. Mustafa Hassona/Anadolu via Getty Images Keren Landman is a senior reporter covering public health, emerging infectious diseases, the health workforce, and health justice at Vox. Keren is trained as a physician, researcher, and epidemiologist and has served as a disease detective at the US Centers for Disease Control and Prevention.

At a press conference on October 25, PBS Newshour reporter Laura Barrón-López asked US President Joe Biden a stark question. More than 6,000 Palestinian deaths had been reported in Gaza since October 7, she said. Did this suggest Israeli Prime Minister Benjamin Netanyahu was ignoring Biden’s message to avoid civilian deaths?

In his response, Biden questioned whether the fatality numbers, which came from the Hamas-controlled Gaza Ministry of Health, accurately captured the reality on the ground. “I have no notion that the Palestinians are telling the truth about how many people are killed,” he said.

Biden’s remarks were met with intense anger by some commentators who found them overly dismissive of death and suffering; others noted Biden’s own administration has been relying on those figures internally throughout and before the conflict.

Two days later, in an unusual move this early in a conflict and seemingly in response to Biden’s remarks, Gaza’s Ministry of Health released a list containing the names and identity numbers of the nearly 7,000 people it says have died in the conflict so far.

Historically, the Gaza Health Ministry’s figures have been found largely accurate. News organizations, human rights groups, and international governments and bodies (including the United Nations) cite them in the moment; and human rights groups that have worked to verify the ministry’s data in previous conflicts have found it generally reliable. Vox reports these figures, as it reports the Israeli government’s stated death tolls.

For those occupying a grim corner at the intersection of political science and epidemiology, lists like these are just the beginning. “When we’re in the midst of something, it’s really, really hard to know” exactly how many have been killed, said Therese Pettersson, a senior analyst and research coordinator at the Uppsala Conflict Data Program (UCDP), a Swedish organization that has been gathering and publishing verified data on conflict-related fatalities for 40 years and is seen as one of the world’s most reliable sources on these types of figures. “As time passes, information will become better.”

She says the reality is that in the early days of a conflict, fatality numbers are incredibly important, incredibly politically powerful — and, unfortunately, incredibly hard to get right. Gazan health officials, for example, have cautioned that death tolls will likely grow, given the number of people trapped under rubble.

Pettersson and other experts in this space urge people to try to balance a few truths when it comes to fatality figures reported during conflicts: Early figures are often inaccurate, and can be exaggerated for political reasons. At the same time, they give us a crucial sense of the devastating scale of loss. In previous conflicts, for instance, the UN has found Gazan health officials’ toll accurate within 4 percentage points. And while a more precise understanding of a violent conflict’s true death toll will emerge in time, one thing is already clear: There is widespread death and suffering in Gaza as a result of the bombardment and fighting.

Although the Gaza Health Ministry’s numbers may historically be reliable, experts still view them as preliminary

Biden didn’t explain why he questioned the Gaza Health Ministry’s estimate of the conflict’s death toll, although it’s possible his remarks were related to what happened after an explosion at the al-Ahli Hospital on October 17. Even if it was not directly related to Biden’s comments, the incident at least shows how easy — and consequential — it is to make erroneous estimates and attributions around deaths related to individual incidents in the midst of a war.

In the hours following that event, news outlets worldwide reported that Israel was responsible for the blast and that it had killed more than 500 people, attributing the information to Gaza’s health ministry. But in a matter of hours, that became hotly disputed. Israel released new evidence alleging that an errant rocket from Hamas-aligned terrorist organization Palestine Islamic Jihad had caused the disaster. A rare US statement on intelligence-gathering sided with Israel; in the days since, news organizations have cast doubt on at least some of the evidence and continued to scrutinize the cause of the explosion.

Meanwhile, US estimates — although low-confidence — suggested the death toll from the hospital explosion was between 100 and 300. The health ministry’s revised final death toll was 471. The episode has been cited as a potential outlier in the health ministry’s general reliability.

But as investigative journalist David Zweig reported in a recent edition of his newsletter, the “500 deaths” figure is actually a misquotation of the health ministry and the likely result of some mistranslated Arabic and a game of journalistic telephone.

This is all to say: There were a lot of sources of uncertainty during this incident.

Omar Shakir of Human Rights Watch, which has been monitoring human rights abuses in Gaza for three decades, told the Guardian the group has “generally found the data that comes out of the ministry of health to be reliable.”

As one of the parties involved in the conflict, Hamas would arguably be incentivized to claim a large number of civilian casualties due to Israeli strikes (more on that below). However, the group has less control over Gaza’s Ministry of Health than it does over political and security agencies in Gaza, according to an Associated Press report. Health ministry employees come from a mix of factions, including Hamas but also the secular nationalist Fatah party, and some are independent. Hamas does not pay their salaries, nor, they say, does it influence the casualty figures they report.

Pettersson said that, historically, the UCDP has trusted Gazan authorities — “but we have also been able to verify their reports with, for example, reports from [the human rights information organization] B’tselem or other types of news reports.” But at the moment, there’s scant news media coverage happening within Gaza due to low electricity supplies and communications services, as well as the danger of working in the area. While there is limited cross-referencing from independent media, what does exist confirms widespread suffering. AP reporters, for instance, have “viewed large numbers of bodies at the sites of airstrikes, morgues and funerals.”

Notably, Israeli fatalities due to this conflict have been covered to an extraordinarily fine degree of detail by many different media outlets, making its casualty numbers much easier to corroborate, said Pettersson. The imbalance of information may be due in part to the imbalance in functioning communication infrastructure — Israel’s is still working, while Gaza’s has been fragile or, at times, completely out (a 34-hour communications blackout last weekend was blamed on a shutdown of phone and communication by Israel). It may also be related to the fact that while active violence leading to death has for the most part stopped in Israel, it remains ongoing in Gaza. An additional factor: Working as a journalist in Gaza is both currently and historically more dangerous than it is in Israel, due to frequent air attacks and Hamas’s history of harassing and using violence against journalists who attempt to report on its activities. At least 30 journalists have been killed in Gaza since this most recent conflict began, many in Israeli airstrikes.

Whatever the reasons, the result is that Israel’s casualties have been easier to verify down to the individual than Gaza’s.

There are good reasons to be skeptical of fatality numbers that emerge during conflict

Fatality numbers released early in the course of violent conflict are often inaccurate, said a number of experts who spoke with Vox. That’s partly because they’re hard to get. Violent conflict often destroys much of the infrastructure that would normally make it possible to reliably count deaths, said Paul Spiegel, a physician and director of the Center for Humanitarian Health at Johns Hopkins Bloomberg School of Public Health. Morgues and health care facilities, generally important sources of reliable casualty information, may be understaffed or too physically damaged to keep good records (although AP reporters spoke with hospital administrators in Gaza who said they record identifying information on every wounded person occupying a bed and every dead body arriving at a morgue, which feeds into a central database).

Additionally, active or impending conflict often prevents human rights organizations’ field staff — the people who would typically verify the numbers and identities of the dead in a disaster setting — from being safely able to do their work.

There’s another important reason that early figures are worth double-checking: They’re liable to be exaggerated, either upward or downward, by parties whose political aims may be aided by death counts that skew one way or another.

“Warring parties themselves have some interest in portraying the conflict in a certain way,” said Pettersson. In most conflicts, parties benefit from minimizing publicly reported deaths of their own fighters, while maximizing publicly reported deaths of their civilians. “I’m not saying that [Hamas] is exaggerating — we don’t know that really, it’s hard to know anything. But there is an interest to do that, to make it sort of fit into the narrative of Israel being the aggressors and Hamas and Gaza’s civilians being the victims,” she said.

The Israeli side would have this incentive too. “Each side will have reasons, usually political in nature, to either minimize or overemphasize,” Spiegel said.

Historically — in conflicts in 2008, 2014, and 2021 — the health ministry’s fatality numbers closely matched death tolls resulting from independent research by United Nations humanitarian agencies. The current conflict is far more complex than those prior conflicts were, and far fewer nongovernmental agencies are currently able to do that independent verification work in Gaza. However, it is reasonable to expect that when organizations like B’tselem verify deaths in the future, they will find numbers similar to what the ministry is now releasing — if not higher, given how many people remain unaccounted for.

Meanwhile, combatant fatalities, if publicized instantaneously, provide information that can be used by an opposing side to determine whether they are correctly targeting battle stations, said Pettersson.

The list published by the Gaza Ministry of Health did not distinguish between combatants and civilians, though it has previously stated that nearly two-thirds of those killed are women and children.

The interest warring parties have in manipulating real-time fatality data is not unique to this conflict. In the ongoing war between Ukraine and Russia, something very similar has often played out, said Pettersson. But the dynamic capitalizes on the way we try to understand these kinds of complex events. “That’s also how our brains work,” she said. “Who’s the bad guy and who’s the good one?”

Early fatality numbers can have important political consequences

Public opinion that gets mobilized by early conflict-related casualty data can have real and significant impact on how the conflict itself plays out, said Lawrence Gostin, who directs the O’Neill Institute for National and Global Health Law at Georgetown Law.

“Public opinion in Berlin, London, Paris, and Washington matters a lot in terms of what political leaders will do,” whether that’s sending aid or other assistance to the region, or voting for certain actions that affect the conflict in the United Nations Security Council.

The series of events that followed the al-Ahli Hospital explosion may be one of the best recent examples of how public opinion about fatalities in a conflict can change the course of that conflict. The explosion took place hours before President Biden was set to arrive in the region to meet with Israeli and Arab leaders. But public protests against Israel erupted worldwide; citing anger at Israel’s supposed role in the blast, Arab leaders canceled their planned summit. That meant delaying important conversations about the logistics of delivering humanitarian aid to Gaza and, potentially, about paths toward peace in the region.

To capitalize on the dynamic at work in episodes like this one, people with strong allegiances to warring parties may share data about their casualties before it has been confirmed. That’s why it’s so important, when news breaks of a fresh wave of violence in a larger conflict, to be aware that early numbers may be colored by bias — especially when they’re not corroborated by other, independent media sources.

Even imperfect fatality figures can be helpful

Over time, Pettersson says, the exact details of a conflict’s lethality come to light. But when can the general public feel confident that has happened?

In the early days of a conflict, the UCDP begins gathering fatality data from open-source materials, including news media, nongovernmental agencies, Telegram, and whatever Twitter is going by. It publishes these on the 20th of each month as “candidate event datasets.”

The UCDP isn’t usually able to verify that data until much later — generally a month or more after events have taken place, said Pettersson. Typically, they’ll work to verify the deaths by going back to the primary source that reported each death, whether that’s a journalist, a warring party, or a witness.

Often, this verification takes place in partnership with organizations that are verifying deaths with primary sources on the ground in the conflict area. In Gaza, the UCDP often works with B’tselem — which maintains a database of conflict-related deaths — to complement and triangulate data. But it uses data from other sources, too, including Reuters, Human Rights Watch, Amnesty International, Crisis Group, and a range of smaller organizations.

The end result is a list of verified organized violence-related fatalities differentiated by combatants and civilians. Because its definitions are rather strict, the list inevitably underestimates deaths attributable to any given conflict. The UCDP publishes this data annually.

(Importantly, UCDP doesn’t count deaths that result indirectly from the conflict, like deaths due to conflict-related famine or sanitation problems. Not because they’re not relevant, said Petterson — it’s just not what her organization does.)

There’s some debate over whether feverish media coverage of early fatality estimates is ultimately helpful to people suffering from violent conflict.

On one hand, these numbers help us get a sense of the scale of the tragedy unfolding in a war-torn region. We don’t need precise figures to know that when many people are dying, many more are suffering. Early casualty numbers are “at the top of the pyramid,” said Gostin: When a conflict kills a large number of people early on, that signifies a much larger number of people who are extraordinarily vulnerable and need immediate humanitarian aid. Right now, over 1.4 million people are displaced in Gaza, according to the UN.

Indeed, these figures may be important for determining how much help an area needs. Some guidelines for providing humanitarian aid use estimates of early mortality in certain age groups to determine the urgency of the response needed, said Spiegel.

Still, it’s not always clear that ceaseless, real-time media coverage of evolving conflicts best serves individuals in war zones. “It’s not always good to have this instant sort of information flow,” especially if that information isn’t verified, said Pettersson. “We don’t know how true it is — and then we react on it.”


No votes yet.
Please wait...

Leave a Reply

Your email address will not be published. Required fields are marked *