Researchers from the Hebrew University of Jerusalem compared official reports with excess mortality data in 134 countries and territories. The countries examined reported 5.08 million deaths from COVID-19 in 2020–2021, while the researchers estimate that the actual number of deaths reached 12.47 million. Israel is among the countries where no significant gap was found between the official report and the estimate. ([Springer Nature][1])
During the coronavirus pandemic, the number of sick and dead people became a key indicator of the severity of the crisis and the success of governments in dealing with it. However, new research by Ariel Karlinsky and Prof. Moshe Shaiyu of the Hebrew University suggests that in many countries around the world, the official numbers presented a very partial picture.
The researchers collected mortality data from 134 countries and territories and compared the number of deaths officially attributed to COVID-19 in 2020–2021 with an estimate of the actual number of deaths, based mainly on Excess mortality – The number of people who died exceeded the number that would have been expected based on mortality patterns in the years preceding the pandemic. The study was published in the journal Journal of Economic Growth and peer-reviewed. ([Springer Nature][1])
The result is striking: the countries included in the sample together reported 5.08 million deaths from Corona, while the researchers estimate that12.47 million – almost 2.5 times. The researchers estimate that between 45% and 55% of the governments examined reported data that did not match the actual death toll estimate. In the vast majority of cases where a significant discrepancy was found, the trend was towards underreporting rather than inflating the number of deaths.
Not every gap means a government is lying.
The researchers emphasize an important point: it is impossible to conclude that every difference between the official death toll and the mortality estimate necessarily indicates deliberate falsification.
The ability to identify COVID-19 as a cause of death, conduct testing, record deaths, and maintain reliable statistical systems varies greatly from country to country. So Karlinsky and Shayo examined whether a country's ability to collect and record data could explain the disparities.
The conclusion was that it explains Only a small part Even after the researchers took into account countries’ ability to diagnose, count and report deaths, widespread underreporting remained. They also tested whether countries simply differed in the criteria by which they attributed deaths to a particular disease, and found that such differences did not explain the results.
Why use excess mortality?
The official death toll from COVID-19 depends on determining that the cause of death was COVID-19. This requires a diagnosis, and sometimes appropriate medical testing and certification.
The total number of deaths, however, is much harder to hide or reclassify. The researchers used historical mortality data to calculate how many people would have died in each country if the pandemic had not occurred, and compared that to the actual number of deaths.
From the excess mortality, they subtracted, as much as possible, deaths attributable to other factors – including wars, natural disasters and other indirect effects – to get a better estimate of the number of direct deaths from coronavirus. ([Springer Nature][1])
The method is not without its uncertainties. The authors note that some countries lack complete data on other causes of death, and data on deaths in wars and natural disasters are also not necessarily perfect. However, they performed various sensitivity tests and found that the main conclusions hold. ([Springer Nature][1])
62 countries reported less than the estimate – and only ten more
Of the 131 countries for which a statistical confidence interval could also be calculated, in 59 no significant difference was found between the official number and the estimate.62 countries The official death toll was significantly lower than the number of deaths estimated by researchers, and only in ten cases was a report higher than the estimate.
Among countries with underreporting, the median gap was equal to 12% of the total number of deaths expected in the country during this period, with the average reaching 14.5%. In some of the most extreme countries, the gap reached more than a third of the total expected deaths.
Israel in the group of countries that reported accurately
From Israel's perspective, the result is actually positive. Israel appears in the group The 59 countries that the study classifies as reporting accurately, meaning that the official death toll from Corona is within the confidence interval of the researchers' estimate.
This group also includes, among others, Canada, Germany, Japan, the United Kingdom, Australia, South Korea, Spain, Leopard, and Norway. The Palestinian Authority also appears in the group in which the gap is not significant.
The figure also connects to findings published in Israel during the pandemic: an analysis by the Central Bureau of Statistics found in 2020 that deaths attributed to COVID-19 explained almost all of the excess mortality measured during that period.
Strong institutions make it difficult to hide information
One of the key questions in the study was what distinguishes countries that reported relatively accurately from those in which large gaps were discovered.
The most prominent connection was not to the level of GDP or dependence on tourism, but toThe institutional and social constraints imposed on the governmentThe stronger the system of checks and balances, the lower the extent of underreporting.
The researchers also found that reporting gaps were smaller in countries where the public was more able to verify official information for themselves – an indicator they assessed, among other things, through levels of education and internet access.
According to the researchers, there is no proof of a causal link here: it is not possible to conclude from the data alone that institutions are weak. Caused The government may be hiding deaths. However, the correlation remains strong even after taking into account the technical capacity of countries to register deaths.
Elections and Communist Legacy
The study also found higher underreporting rates in countries where elections were held during the pandemic years, as well as in countries with a communist past.
The authors are careful to frame these findings as descriptive, rather than as evidence that the election or political history caused the manipulation. However, the link to communist legacy remains even when the researchers take into account contemporary measures of institutions and constraints on executive power. ([Springer Nature][1])
Not just a story about the Corona virus
For Karlinsky and Shayo, the broader significance of the study does not end with the pandemic having already passed its emergency phase.
Governments publish data on economic growth, unemployment, health, pollution, greenhouse gas emissions, poverty, and many other areas. In most cases, the public, the media, and even foreign governments are forced to rely at least in part on the information published by the authorities themselves.
The coronavirus pandemic provided a kind of "natural experiment": almost all governments in the world were forced to deal with a similar crisis at the same time, and it was possible to compare the data they published to a relatively independent index - total mortality.
The conclusion the researchers offer is not that official data should not be trusted, but that trust alone is not enough. Official information becomes more reliable when there are institutions, researchers, journalists, and the public who are able to independently verify it, challenge it, and demand correction when the data does not match reality.
Questions and Answers
How many people died from corona according to the study?
In the 134 countries and territories examined, 5.08 million deaths were officially reported in 2020–2021, but the researchers estimate that the actual number of deaths was about 12.47 million. ([Springer Nature][1])
Does the study prove that governments intentionally falsified the data?
Not in every case. The researchers also looked at differences in the ability to register and diagnose deaths. Limited capacity explains some of the gap, but not most. In many countries, a gap remains that the researchers interpret as possible evidence of data manipulation.
Did Israel also report fewer deaths than the actual number?
Not significantly. Israel was included in the group of countries whose official reporting was statistically consistent with the researchers' mortality estimate.
What about more accurate reporting?
The strongest link found was to institutions that limit government power and the public's ability to check information – including through education and internet access. The study suggests correlation, not causation. ([Springer Nature][1])
Scientific source: Karlinsky, A. & Shayo, M., Manipulation of information in times of crisis: evidence from Covid excess mortality, Journal of Economic Growth 31, 433–462 (2026). The full scientific article at Springer
More on the subject on the science website
- Ministry of Health: Excess mortality due to Corona is starting to stand out in the data
- A thorough study of 1.2 million Clalit insured people: The coronavirus vaccine is effective
- No country in the world is prepared to deal with the corona virus, and most of them are unable to receive data in real time
- How the coronavirus is changing science
- A fast-moving pandemic: Researchers simulated an entire pandemic in the lab
For the scientific article: Opening the scientific article
