What is the True Death Toll in Gaza?
Image by Mohammed Ibrahim.
According to the latest figures released by the Gaza Ministry of Health (GMoH) on December 23rd 2025, Israel had killed at least 70,937 Palestinians and wounded 171,192; of those identified fatalities, 53% were women, children or elderly.
Despite the decimation of Gaza’s healthcare system, the GMoH has demonstrated remarkable resilience in gathering, verifying, and publishing data on a rolling basis, including comprehensive updates on the death toll in Gaza. In an interview with Drop Site News, Michael Spagat, a professor of economics at the University of London with a specialization in the quantitative analysis of armed conflict, said that it was “very significant that [the GMoH] is making their data available and it’s highly unusual. As a consequence, we understand the death toll in the Gaza war better than we do in almost all wars of the past few decades”. This has not stopped biased actors from casting doubt on the official casualty numbers, as Joe Biden infamously exclaimed early on in the genocide when he said that he had “no confidence in the number that the Palestinians are using”. Israeli officials have also consistently dismissed these totals as “Hamas propaganda”, although it is noteworthy that Herzi Halevi, Israel’s former Chief of Staff, recently stated that “more than 200,000 Palestinians have been killed or injured in the war in Gaza”, a number that closely aligns with GMoH estimates.
Still, legitimate questions can be raised about the GMoH numbers, and it is important to have as accurate an account of the overall death toll, not only to reflect the atrocious reality on the ground and to hold Israel accountable, but to also ensure that every death is properly accounted for and documented in a dignified manner. In light of that, this article addresses the following question: what is the true death toll in Gaza? In order to address this question, we first need an understanding of the GMoH’s methodology in counting deaths, which I describe below. I then argue that the GMoH’s official death toll very likely represents a significant undercount of the true death toll based on our best available studies, while critiquing a recent analysis that estimates that the true death toll is an order of magnitude higher than the official numbers. I then provide some evidence for estimating the percentage of civilian deaths, before concluding with some observations about the true death toll in Gaza.
The GMoH’s methodology in counting deaths
The GMoH relies primarily on a central collection system to count deaths, which aggregates reports from public hospitals and morgues as well as from private and foreign-run facilities into an electronic database. Under normal circumstances, system access to the civil registry is available, allowing the GMoH to release detailed lists of all fatalities, which include names, birth dates and ID numbers. Importantly, only deaths from traumatic injury are reported by the GMoH (so-called “direct deaths”); “indirect deaths” due to war (e.g. starvation, disease) are not typically captured.
In previous Israeli wars on Gaza, this system worked remarkably well, and the GMoH’s figures were deemed to be “reliable, according to the UN and human-rights organizations, which have scrutinized them, and come up with their own estimates”; indeed, a discrepancy of only a few percent separated the official GMoH figures from the UN’s own independent estimates for the wars in 2008, 2014 and 2021.
However, as a result of Israel’s systematic assault on Gaza’s hospitals during this genocide, especially after the Al-Shifa hospital was occupied by Israeli forces on November 10th 2023, the casualty recording system collapsed. According to Zaher Al Wahidi, director of the GMoH’s Health Information Center, hospitals in the South of Gaza lost access to the civil registry, and hospitals in the North saw their computer systems completely destroyed by Israeli forces. With no access to the civil registry, staff in Southern hospitals had to manually enter details of deaths. In the North, with no computer systems at all, hospital staff relied on paper records when they could, or more often, when the hospital was overwhelmed with casualties and medical staff displaced or dead, non-medical staff helped count the dead as they passed through the hospital without being able to identify them, at best tagging and photographing bodies for later identification. These headcount numbers representing unidentified fatalities were added to the deaths appearing on the detailed lists, resulting in the official reported death toll.
To help address these issues, the GMoH introduced an online portal launched in January 2024 for families to report deaths. This online form allows family members to provide details of individual deaths of loved ones, including whether the body passed through a hospital, in which case one unidentified death would be subtracted from the total death toll. If instead, the form indicated that the body was buried without hospital involvement or was missing (because, for example, it was still under the rubble), additional deaths would be added to the total from the hospitals. A judicial process was established in October 2024 to ensure that any deaths claimed via the online form were verifiable in order for families to obtain formal death certificates; until then, the GMoH added all deaths submitted via the form to their totals without judicial review, leading to subsequent removals of previously recorded deaths if formal verification proved unachievable. Finally, as the central collection system was gradually restored by mid-2024, errors like missing IDs or typos were fixed, thus improving the overall quality of their releases.
The GMoH death toll: an undercount of the true death toll
Despite the GMoH’s best efforts at recording deaths, it is very likely the case that the official death toll is a significant undercount of the true total. This is so because it does not capture indirect deaths, and is furthermore incomplete as a record of all direct deaths, not only because of Israel’s relentless assault on Gaza’s health care system and the inability to recover inaccessible bodies, but also because of the GMoH’s insistence on maintaining high standards in their recording of official deaths, despite operating under the most difficult of circumstances. This conclusion is based on our best available studies, which are discussed in this section.
The Jamaluddine et al study
The peer-reviewed research article “Traumatic injury mortality in the Gaza Strip from Oct 7, 2023, to June 30, 2024: a capture–recapture analysis”, published in The Lancet by Jamaluddine et al, focuses on estimating direct deaths in Gaza between October 7th 2023 and June 30th 2024.
The official death toll reported by the GMoH over that period was 37,877 deaths. To address the possibility of some direct deaths not being officially captured, the authors employed a capture–recapture analysis, a statistical technique originally developed in ecology to estimate the size of animal populations, but adapted for conflict mortality estimation and used in places like Kosovo, Colombia, and Sudan. This method uses overlapping data from multiple independent lists to infer total deaths, including those not captured in any list. The authors relied on three separate lists – the GMoH hospital records of identified fatalities, the GMoH online form collecting community-reported deaths, and obituaries from social media platforms – and used statistical models to determine how the lists overlapped and how many deaths might be missing from all lists.
The results reveal an estimated 64,260 deaths, indicating a 41% undercount by the GMoH (= (64,260 – 37,877) / 64,260), with women, children, and elderly accounting for 59% of total deaths. The authors discuss several limitations with the study, including standardization challenges, the exclusion of missing people from the analysis, missing information in records, the degree of independence between lists, and the fact that indirect deaths are not being captured. Still, despite these limitations, this study makes substantive contributions to generating independent estimates of the true death toll, and clearly suggests that the GMoH total is a significant undercount.
The Spagat et al study
The paper “Violent and Nonviolent Death Tolls for the Gaza War: New Primary Evidence,” authored by Spagat et al, presents independent estimates of war-related mortality in Gaza for both direct and indirect deaths from October 7th 2023 to January 5th 2025 based on the Gaza Mortality Survey (GMS), a large-scale household survey conducted in partnership with the Palestinian Center for Policy and Survey Research.
Given the obvious challenges of data collection in the midst of a war-zone and an ongoing genocide, the authors had to carefully design the survey to take into account mass displacement of the majority of people in Gaza, as well as lack of direct access to restricted areas, in order to ensure that the survey sample remained representative of Gaza’s full population. Face-to-face interviews were conducted with 2,000 households using questionnaires, focused on the respondent’s household as it was on October 6th 2023 and the fate of household members after October 7th 2023 – were they alive, dead, or missing? For deaths, the questionnaire asked whether the cause was due to violence or non-violence, the latter category enabling the authors to estimate indirect deaths due to the Gaza genocide after subtracting projected baseline deaths based on pre-genocide trends.
The authors estimate 75,200 direct deaths, compared to the GMoH’s figure of 49,090 over that same period, indicating a 35% undercount (= (75,200 – 49,090) / 75,200), with women, children, and elderly accounting for 56% of total deaths, consistent with GMoH percentages. In addition, indirect deaths from the genocide totaled 8,540, resulting in an indirect to direct death ratio of 0.11 to 1. Although surprisingly low given the level of destruction of Gaza’s healthcare system, Spagat, in response to a question from a commentator on X, provides some justification for their low number, stating that “at the beginning, the population [was] rather healthy; e.g., high vaccination rates and the public health system was strong. A lot of aid flowed in, albeit with interruptions, until the big blockade after our field work was finished. Luckily no major infectious disease outbreaks”. Spagat does however caution that the situation has likely gotten much worse since their study was completed.
Potential upward biases in the study include oversampling of relatively high-mortality households, which was addressed through survey protocols and sensitivity analyses. A source of downward bias is the impossibility of sampling households with no remaining live members, a not-insignificant phenomenon in Gaza (by October 7th 2024, at least 1,238 Palestinian families were completely wiped out), as well as missing persons not factoring into the authors’ estimate of the overall death toll (they were classified into a separate category).
Still, despite these limitations, this study provides the first survey-based estimate of the overall death toll in Gaza, showing similar levels of undercounting by the GMoH as that reported by the Jamaluddine study using completely different methods, reinforcing the belief that a significant number of deaths are not being officially captured.
The Cockerill et al analysis
Matthew Ghobrial Cockerill, a doctoral candidate in History at the London School of Economics, recently published an analysis for Action on Armed Violence (AoAV) that audited certain fatalities originally included in the overall death toll by the GMoH, before the latter removed them in subsequent releases. Between August 2024 and March 2025, around 3000 fatalities were removed from the GMoH’s official death toll, with 1,079 of them being children. While many unscrupulous pro-Israeli commentators made unsubstantiated claims of data fabrication by the GMoH, Cockerill’s analysis, made possible by the tireless work of researchers in Gaza, not only provided a detailed rebuttal to such allegations, but also provided strong empirical evidence that the official GMoH death toll is an underestimate of the true total.
The investigation, using “open-source intelligence, social media documentation, and corroborated news reporting”, confirmed that of the 1,079 children removed from the official GMoH death toll, 36% of them were killed by violent means, with another 24% of them having died because of unknown causes (but almost certainly killed); deaths were confirmed “only when a connected (usually familial) source corroborated it, or when a secondary source described the fatality with details, such as the date of death, that were not contained in the MoH list”. It is worth noting that only 3% of the cases examined resulted in erroneous or premature claims.
Why would the GMoH remove so many verifiable child deaths from their official lists? Cockerill convincingly argues that the answer lies in “an unusually high evidentiary bar […] that requires that a name be included only if a government official has personally seen the body or if a judicial ruling has confirmed the death […] Many of the now-deleted child fatalities were initially submitted via an online form by surviving relatives, a common method in crisis zones, but were later removed when formal verification proved unachievable”. One possible reason for the latter involves the onerous bureaucracy associated with the formal judicial process to get the GMoH to officially recognize a death that did not pass through the central collection system, which likely discouraged some families already struggling under genocide from completing the formal process, leading to the removal of otherwise verifiable deaths.
A cautionary tale
Although there is strong evidence that the official GMoH death toll is a significant undercount of the true total, there is little evidence to support a recent analysis conducted by Hil and Polya, who estimated the true death toll from the Gaza genocide over the period between October 7th 2023 and April 25th 2025 to be 680,000, including 380,000 deaths of children under five. Those headline numbers went viral on social media, accumulating hundreds of thousands of likes and shares, and were even mentioned by Francesca Albanese, the UN Special Rapporteur on the Occupied Palestinian Territories. However, Hil and Polya’s analysis can legitimately be critiqued, as it relies on questionable assumptions that result in estimates that are unrealistically high (in fact, these unrealistic analyses should be critiqued by partisans of the Palestinian cause, as they otherwise hurt credibility).
In order to come up with the number of total direct deaths, the authors start with the death toll calculated by the Jamaluddine study, which estimated direct deaths between October 7th 2023 and June 30th 2024 (a period of 267 days) to be 64,260, and then linearly extrapolate that number to cover the period between October 7th 2023 and April 25th 2025 (a period of 566 days) to obtain ~136,000 direct deaths (= 64,260 * (566 / 267)). Their methodology implies an identical number of direct deaths per day in the 299 days between July 1st 2024 and April 25th 2025, as that in the 267 days between Oct 7th 2023 and June 30th 2024, an assumption for which there is no good evidence, and for which there is good counterevidence, in the form of the official GMoH death rate. Indeed, the death rate per the GMoH between October 7th 2023 and June 30th 2024 is 142 deaths per day (= 37,877 / 267), while the death rate between July 1st 2024 and April 25th 2025 is 45 deaths per day (= (51,439 – 37,877) / 299) – a factor of 3 difference – which clearly suggests that one cannot linearly extrapolate in this way (the death rate change over time tends to be correlated with the “intensity” of the different phases of the genocide, with the highest death rate occurring in the first three weeks given the unprecedented Israeli aerial bombardment of Gaza).
In order to come up with their estimate of total indirect deaths, the authors take their estimate of direct deaths and apply a 4 to 1 indirect to direct ratio to get a total of 544,000 indirect deaths (= 136,000 * 4), resulting in a total of 680,000 total (direct and indirect) deaths by April 25th 2025. This ratio comes from a non-peer-reviewed correspondence in The Lancet published by Khatib et al in July 2024, where they estimate the total number of indirect deaths in Gaza by “applying a conservative estimate of four indirect deaths per one direct death”.
First, it is important to mention that Khatib et al are silent on the time period over which these indirect deaths are to materialize (a weakness in their article), but their number is meant to be a projection into the (distant) future, as they state that “even if the conflict ends immediately, there will continue to be many indirect deaths in the coming months and years from causes such as reproductive, communicable and non-communicable diseases” [my emphasis]. Hil and Polya, on the other hand, simply assume that this ratio applies within their specified time period, without any stated justification – a highly questionable assumption given the much smaller ratio from Spagat’s survey over roughly the same time period.
Second, Khatib et al’s 4 to 1 ratio can itself be legitimately questioned. In order to justify their ratio, they refer to a 2008 report on the Global Burden of Armed Violence (GBoAV), which in turn summarizes estimates of the indirect to direct ratio in 13 previous (non-Gaza) conflicts around the world (see Table 2.3 on p.40, where the ratio on the lower end is 0 to 1 for Kosovo, while the ratio on the higher end is 15.7 to 1 for Sierra Leone). Khatib el al provide no justification for why this small subset of previous conflicts are good proxies for Gaza, or why a 4 to 1 ratio is appropriate in the case of Gaza; they simply decree that it’s a conservative estimate. Interestingly, both Spagat and Jamaluddine criticize the 4 to 1 ratio, and provide counter-arguments in their respective papers. Spagat highlights examples with much lower ratios, such as the aforementioned Kosovo example as well as Yemen, while Jamaluddine makes the observation that the ratios closest to 4 to 1 in Table 2.3 are the conflicts in Burundi (1993–2003) and Timor-Leste (1974–99), arguing that using such a high ratio for Gaza “might be inappropriate due to obvious differences in the pre-war burden of disease”. This is not to say that such a ratio is impossible; simply that the authors fail to make a convincing argument as to how such a ratio obtains.
Finally, in order to come with their estimate of 380,000 indirect deaths of children under five, Hil and Polya multiply the 544,000 number by 70%, based on “exhaustive analysis of avoidable deaths from deprivation in all countries from 1950 onwards [that] reveals that under-five-year-old infant deaths make up about 70 per cent of avoidable deaths in impoverished countries”. Given that Gaza’s pre-genocide population was around 2.23 million and children under-five made up 15% of the total, the total number of children under-five equates to 334,000 (= 2,230,000 * 0.15); clearly then, Hil and Polya’s estimate cannot be correct, implying that their underlying assumptions are faulty.
How many civilians have been killed?
The GMoH (and all previously discussed studies) do not differentiate between combatants and civilians when discussing the overall death toll – understandably so, as they are not in a position to make that assessment. However, it is a legitimate distinction, and the number of civilians can be approximated using different methods.
A recent investigation by the Guardian revealed figures from a classified Israeli military intelligence database with the names of 8,900 dead (and “probably dead”) fighters from Palestinian Islamic Jihad (PIJ) and Hamas (as of May 2025), suggesting that 83% of fatalities in the Gaza genocide were civilians. Although this database is inaccessible to the public, and little additional information was provided about completeness and methodology of the overall compiled fighter list, alternative methods discussed below that approximate the percentage of civilians killed suggest that the 83% number is not unrealistic.
Matthew Ghobrial Cockerill makes the case that at least 74% of Gazan fatalities identified by the GMoH are civilians in an analysis dated October 28th 2024. To calculate this number, he starts by making the (certain) assumption that all young boys (aged 0 – 14), elderly men (65+), and women and girls of all ages are civilians, three categories that make up 53% of total fatalities identified by the GMoH. He then assumes that of the remaining male fatalities (aged 15 – 64), at least the same amount of male civilians in that age range have been killed as females in that same age range, resulting in another 21% of total fatalities being civilians. He convincingly justifies that (conservative) assumption by making the correct observation that in conflicts the world over, civilian men killed outnumber civilian women, including in previous Israeli wars on Gaza according to the database maintained by B’Tselem, where “about 150% more male civilians aged 15 – 64 were killed than female civilians aged 15 – 64 across [Cast Lead and Protective Edge]”.
Another conservative approach to estimate the split between civilians and combatants among male fatalities aged 15 – 64 consists of applying previous splits that obtained in past Israeli wars on Gaza, an approach that results in very similar civilian percentages as Cockerill’s analysis. Based on the B’Tselem database, about 45% of male fatalities aged 15 – 64 in Israeli wars on Gaza between 2008 and 2022 were civilians. Based on the most recent detailed fatalities list released by the GMoH (on July 31st 2025), a total of 60,199 individuals were killed, with 29,152 being young boys, elderly men, and women and girls of all ages, while 31,047 were males aged 15 – 64, implying a total civilian toll of about 72% (= (29,152 + 0.45 * 31,047) / 60,199).
One final point is worth making. One of the most convincing pieces of data supporting the claim that Israel’s assault on Gaza constitutes genocide is the sex and age-related distributions of those killed by Israel (reflecting a significant “excess” of civilian deaths, especially among children). Unlike previous wars on Gaza, these distributions clearly fit the pattern typical of other genocides, as discussed in the peer-reviewed paper “Accounting for Uncertainty in Conflict Mortality Estimation: an Application to the Gaza War in 2023-2024” published in Population Health Metrics by Gomez-Ugarte et al (see Figure 2). Indeed, whereas the overwhelming majority of conflict-related deaths in typical wars occur among males (with a majority of those deaths occurring among “fighting-age” males), the pattern in Gaza is characterized by a significant number of deaths among children (~30% of all deaths), women and elderly, a pattern much more typical of genocide.
Conclusion
Our best studies, conducted nine and fifteen months into the genocide, suggest that the official death toll published by the GMoH is a 35% to 40% undercount of the true direct death toll. Assuming, not unreasonably, that the same level of under-reporting continued until today given that there’s no reason to believe that the GMoH had significantly improved their ability to capture all deaths before that, the true direct death toll is likely to be well over 100,000 (= 70,937 / (1 – 0.35); note that a similar approach was used to estimate the true direct death toll by researchers at the Max Planck Institute for Demographic Research, yielding similar results). It is important to highlight that at least 75% of these deaths are civilians, although the true percentage is likely higher (given the implied number of dead combatants).
In addition to direct deaths, there is a significant number of indirect deaths that must be added. Although the exact number is difficult to approximate, we can say with reasonably high confidence that the indirect to direct ratio will be higher than Spagat’s estimate of 0.11 to 1, given the significant deterioration in living conditions in Gaza since his household survey was conducted.
As merely one example, the UN-backed IPC, based on a report from the Famine Review Committee (FRC), confirmed in August that famine (with reasonable evidence) was occurring in the Gaza governorate, where almost 1 million people lived (~15% of which were children under-five) before the Israeli assault on Gaza City in September. Given that FRC experts inferred “from the available data that mortality thresholds for Famine have already been exceeded in [the] Gaza Governorate” based on IPC Guidance on Famine Classification (i.e. ≥ 2 deaths per 10,000 people or ≥ 4 deaths per 10,000 children under-five), this implies that their experts believed that at least ~200 people (= 937,604 * 2 / 10,000) or ~50 children under-five were dying of starvation or from the interaction of malnutrition and disease every day in the Gaza governorate over their period of analysis. Although these inferred numbers can be debated given the much smaller number of officially reported starvation deaths, surely a significant level of under-counting of indirect deaths has occurred, and will likely continue to occur even after a true cessation of hostilities, as additional people will succumb to malnutrition, disease and weather given the overwhelming destruction of Gaza’s life-sustaining infrastructure. The magnitude of these future indirect deaths will be dependent on how quickly adequate amounts of humanitarian aid can be brought into Gaza (which Israel is currently purposely limiting, despite the terms of the “ceasefire”), and how long it will take to rebuild Gaza’s critical infrastructure.
In closing, it is important to highlight that every death matters, and that every victim of this genocide deserves to be remembered. The process of identifying and naming every individual victim will likely take the GMoH years to complete, but doing so is essential, not only because it endows every victim with a modicum of dignity, but also because it ensures that Israel can be held fully accountable for the true scale of this calamity.
Disclaimer: The opinions and views expressed by the author are solely his and do not reflect in any way the positions of any organization the author is affiliated with.
The post What is the True Death Toll in Gaza? appeared first on CounterPunch.org.