No, 100,000 Palestinians haven’t died in Gaza
Kidney patients sit amid rubble at Shifa Hospital in Gaza City, early July 2025
Sergio DellaPergola writes in Haaretz on 6 July 2025:
Let’s assume a group of researchers sends a scientific article to a medical journal about a new wonder drug that will cure the human race of a dangerous disease. The article undergoes review by the journal and is rejected, but the researchers nevertheless publish it independently. Haaretz then devotes an article to the report, and some consumer buys the wonder drug and dies of their illness.
Ironically, that is similar to the article Nir Hasson published in Haaretz on June 26 about the number of people killed in the Gaza Strip. Hasson covered a report published by Prof. Michael Spagat – a researcher from the University of London who is an expert in estimating war deaths – and other researchers, including Palestinian pollster Dr. Khalil Shikaki, who is generally considered reliable.
According to Hasson’s article, the new research about the number of deaths in Gaza is now undergoing peer review; meanwhile, the researchers decided to publish it online. In reality, the article had already been rejected by the journal to which it was sent after the review process was completed.
The authors have the right to disseminate their research freely, but such a study has much less credibility not being peer-reviewed. It’s not inconceivable that another journal will publish the study. Today, the market for scientific publication is broad, generous and diverse, and almost any author can find some platform for what was rejected elsewhere. In practice, the boundary between articles that meet academic research criteria and those that reflect free narratives of some kind is increasingly being blurred.
Regarding the substance of the report about 100,000 dead in Gaza, it’s first necessary to acknowledge that the situation there is undoubtedly catastrophic and intolerable. We have witnessed an utter failure by the Israeli government, local actors in Gaza and international players to bring an end to the war that began with Hamas’ massacre in southern Israel on October 7, 2023.
The most terrible expressions of that failure are the fact that 50 Israeli hostages are still in Hamas’ hands and the enormous number of deaths in Gaza due to the army’s not proportional response to Hamas’ attack. Every day, the feeling of dread grows regarding the war’s effects on Gazans and Israelis alike. And in part for this reason, making as accurate an estimate as possible of the number of victims and the scope of the destruction is both a civic and a scientific duty.
The high number of deaths in Gaza stems from the army’s operations, but also from the worsening of conditions there: the exposure to risk, the lack of medical treatment for those who need it, the distress, the hunger, the epidemics and other causes. There are several ways to provide an accurate estimate of all of these factors, but in the current circumstances, the only source of data is Gaza’s Hamas-run Health Ministry, which cannot be considered credible because it is itself a political subject. Another possible source – the Palestinian Authority’s Central Bureau of Statistics, which is ostensibly a professional agency – is out of the picture entirely.
The first and most self-evident way to estimate the number of dead is by counting the bodies that arrive at morgues or are buried. Astoundingly, according to Gazan sources, this simple action was never performed.
The number of deaths according to Hamas’ Health Ministry is based only partly on counting bodies, and partly on the declarations of families seeking a death certificate. These certificates are essential to receive death benefits, and because of the enormous pressure, they are issued without sufficient investigation of individual cases.
Granted, some of the bodies are still buried under destroyed buildings. But even in the most horrific sanitary conditions, it’s inconceivable that there are still tens of thousands of buried bodies 21 months after the fighting began. And at least according to the IDF, residents were warned to evacuate their homes before heavy bombs were launched.
The second way to check the estimates provided by Hamas’ Health Ministry is by comparing lists of residents, including lists of refugees entitled to UN assistance. A comparison of this sort (see the article by Michel Guillot and colleagues that was published in the Lancet in January 2025) identified 34,344 people who were killed (21,974 men and 12,370 women) between October 7, 2023 and August 31, 2024, based on detailed checks by the Health Ministry. The ratio of men to women, with men being significantly overrepresented, reflects the fact that Hamas fighters are exposed to greater risks than the rest of the population. The age breakdown of the dead, according to this study, also fits that fact, as a significant percentage were aged 18 to 45.
In an ordinary year, the number of deaths in Gaza ranges from 6,000 to 7,000. In other words, the death rate is 3 to 3.5 per 1,000 residents. This is a fairly low rate, and it stems from the fact that Gaza’s population is young, with half of all residents below age 18. Children (which include some fighters) account for a significantly lower percentage of the dead proportionate to their share of the total population.
A second reason for the low death rate in ordinary years is that before the war, Gaza had dozens of hospitals. Even if their level wasn’t exceptionally high, the medical infrastructure was still incomparably better than it is in less developed countries. Life expectancy was thus similar to European countries like Poland.
According to the Gaza Health Ministry, the number of people killed during the war stood at between 40,000 and 50,000 before the fighting resumed in March following a cease-fire. As an example of the fact that, today, anything goes, an article published in the Lancet in July 2024 by Rasha Khatib and others claimed the number of dead in Gaza had already reached at least 186,000.
Spagat and his colleagues sought an alternative, independent method of estimating the number of victims. Their study surveyed 2,000 families together containing some 10,000 members from all five of Gaza’s districts, thereby ostensibly creating a representative sample of Gaza’s two million residents. The respondents (one person in each family) were asked about what has happened to their relatives since the war began. The answers address all cases of death, missing people, emigration or being taken prisoner by Israel known to family members.
Based on this method, Spagat and his colleagues estimated the number of deaths in Gaza at around 91,400. Of these, 75,200 were killed directly by wartime operations, while between 8,000 and 9,000 died from circumstances that aren’t directly connected to the army’s operations (and all this is on top of the 6,000 to 7,000 who die in an ordinary year).
The idea behind this research is a good one, and it’s original. But the question must be asked as to whether implementation met the standards of scientific research. And here, several critical questions arise. First, Gaza is in a state of chaos, destruction and forced evacuation of entire areas, with the evacuees living in emergency housing comprising tents and other temporary structures. This makes it very hard to define what a household is and what a family is.
Before the start of the war, households ranged from simple nuclear units to very complex and broad units. It can be assumed that each household occupied one residential unit, small or large. The mass movement to temporary residences has resulted in reconfigured housing units, such as a tent or a room.
Many of these reconstituted housing units are smaller than the original households. For example, a 12-person household – which previously lived in a two-floor building and suffered from the loss of an adult family member because the building was destroyed – may have moved to two tents, one with six people, and one with five; in other words, it divided into two housing units. The relative completing the survey could report the deceased adult because he or she will think of the original family composition. The fact that the average household size in the current survey is smaller than the size of the original household in population corresponds to this assumption. In other words, the division and redefinition of households as a result of the state of emergency in Gaza may cause the same death to be reported two or three times.
The survey included respondents who remained in their original locations and others who moved to elsewhere in Gaza, to temporary shelters, because their original residences were inaccessible. It may be assumed that the respondents in temporary shelters came from more severely damaged areas, where there were more deaths. But that is not the case in the survey’s data. For example, the Gaza district with the highest ratio of reported victims proportionate to the population (55 per 1,000) is Deir al-Balah, which was not evacuated at the time of the survey.
Belying expectations, the correlation of cases of violent and non-violent death by district is neither high nor clear, even though it might have been expected that greater destructions would directly or indirectly result in a larger number of evacuees and deaths.
The selection of respondents is another reason why the sample was not representative. The authors note that “Once a certain geographic counting cell reached ten households that agreed to participate, the count there was considered complete” and “In each household, the respondent was not randomly selected. Instead, interviews were conducted with one adult (aged 16 and up), irrespective of the respondent’s age or gender.” The objective, therefore, was to fill a quota in a particular area without maintaining sample randomness.
It should be emphasized that in statistical research if the survey does not keep the principle of randomness and mainly reflects willingness to participate, the results are invalid. The 10 households in each counting cell could be uniquely accessible for one reason or another, compared with other households in the same counting cell, and therefore bear unique stories.
Moreover, it is possible that the respondents were more willing to talk about their experiences. The pollster did not really control the randomness of the responses and let them continue until the quota was filled. At no point did they check how many people might report the same death. This procedure, known as multiplicity, is critical in surveys of this kind in order to prevent duplications, but it was not performed. In cases of surplus death – those not caused by direct injury – the report’s authors did not classify them by sex, age and cause of death.
Besides reporting deaths, the survey also collected data on prisoners, migrants to outside Gaza and missing persons in order to complete a picture of general suffering in Gaza. It may be assumed that the same statistical rules that apply to the survey’s main subject – deaths – also apply to the secondary subjects, such as prisoners and the missing. The writers estimate the number of prisoners at 9,000, with a 95 percent probability that it is between 6,000 and 12,000.
They also provide a contradictory figure from another source regarding the number of Gazan prisoners captured by Israel – 3,436. In fact, the number of 3,436 equals only 38.2 percent of the estimate of 9,000 obtained by the survey. Since the same ratio must also apply to the survey’s estimate of deaths, the latter number should fall from 91,400 to 34,380. This number is surprisingly close to another figure cited above following the only systematic check actually available from the Gaza Health Ministry. A simple test was sufficient to obtain a more reasonable result for the number of deaths in Gaza, but it was not done.
Regrettably, it is clear that the number of deaths in Gaza grows daily, and the results reported in the Spagat et. al survey, and in critiques of the survey, require revision. Nonetheless, it would have been better if, before publishing the figure of 100,000 deaths in Gaza, Haaretz had checked the facts, or at least added a question mark to the rounded figure.
This article is reproduced in its entirety