Analysis of data from 134 countries and territories shows wide differences between officially reported COVID-19 mortality rates and estimates based primarily on excess mortality.

Official COVID-19 death figures in many countries differed significantly from estimates based primarily on excess mortality. Ariel Karlinsky of Hebrew University and Professor Moses Shayo of Hebrew University and King’s College London analyzed data from 134 countries and territories. In 2020-2021, 5.08 million COVID-19 deaths were officially reported in the analyzed countries, while the researchers estimated the actual number of COVID-19-related deaths at 12.47 million. According to the authors, reporting errors occurred in approximately 45-55% of governments, with underreporting occurring much more frequently than overreporting. However, the results do not mean that every discrepancy was the result of deliberate manipulation.

In the article:

• how death data from 134 countries and territories were analyzed,
• how big was the difference between the officially reported and estimated number of deaths from COVID-19,
• why the state’s administrative capabilities do not explain all the discrepancies,
• what was the importance of institutional control mechanisms and the ability of citizens to verify information,
• why excess mortality can serve as an independent reference point for official statistics,
• what significance the results may have for the reliability of data used in public health.

COVID-19 as a global test of the credibility of official data

The COVID-19 pandemic presented governments with a similar challenge: the need to quickly measure the scale of the health crisis and communicate its severity to the public. One of the primary metrics remained the number of deaths attributed to COVID-19.

The problem, however, was that the official death toll depended not only on the actual mortality rate but also on the ability to diagnose infection, perform tests, determine the cause of death, the functioning of registration systems, and the quality of public statistics. Ariel Karlinsky and Moses Shayo therefore decided to examine the extent to which the official data corresponded to broader estimates based primarily on excess mortality.

The study, published in the Journal of Economic Growth , uses the COVID-19 pandemic as a unique global case study to compare the reporting of information by a large number of countries that faced a similar threat at a similar time.

Data from 134 countries and territories

Karlinsky and Shayo collected mortality data from 134 countries and territories. Information was sourced from national statistical offices, population registries, ministries of health, and other data sources.

Researchers compared officially reported COVID-19 deaths with estimates of actual mortality related to the pandemic. The primary independent benchmark was excess mortality, which is the number of deaths above the level that would be expected under normal circumstances.Information

Key data from the study

A total of 5.08 million COVID-19 deaths were officially reported in the 134 countries and territories analyzed between 2020 and 2021. The researchers estimated the actual number of COVID-19-related deaths at 12.47 million . According to the analysis, approximately 45-55% of governments misreported COVID-19 mortality , and underreporting was significantly more common than overreporting.

The discrepancy between official statistics and estimates based on excess mortality was therefore a significant element of the pandemic picture. However, the authors do not automatically consider every such difference to be evidence of deliberate data falsification.

The discrepancy does not necessarily mean deliberate manipulation

This is one of the most important caveats in interpreting the study. Individual countries vary greatly in their ability to diagnose diseases, register deaths, and maintain reliable and complete statistical systems. In countries with weaker administrative and diagnostic infrastructure, underestimation of deaths may result from technical, organizational, or reporting limitations.

The study authors therefore analysed whether differences in countries’ capacity to collect and record information could explain the observed discrepancies between the official and estimated death numbers.

According to the results presented, however, a country’s ability to conduct such measurements explained only a small portion of the observed variation. This means that differences in the quality of statistical and administrative systems alone were not sufficient to explain the scale of differences between individual countries.Information

An important interpretation caveat

The authors do not assume that every difference between the official and estimated death counts represents deliberate public misinformation. Some of the discrepancies may be due to diagnostic, registration, and statistical limitations. However, the study indicates that differences in administrative capacity between countries explained only a small portion of the observed variation.

The importance of control institutions and the possibility of data verification

Researchers observed a stronger relationship when analyzing the institutional and social constraints within which governments operate.

Countries with stronger institutional checks and balances showed significantly less underreporting of deaths. Lower levels of misreporting were also observed where citizens were better placed to evaluate and verify official information.

The study analyzed the public’s ability to evaluate information using factors including education level and internet access. The authors believe that not only the technical capacity of the administration to produce statistics may be important, but also the ability to independently evaluate and challenge them.

The authors point out that in many political systems, there may be incentives to present information in a way that benefits the government. However, from the perspective of the study’s findings, what is crucial is whether effective institutional and social mechanisms are in place to increase the likelihood of detecting and challenging inaccurate data.

Elections and historical political conditions

The analysis also found higher levels of reporting irregularities in countries where elections were held during the pandemic, as well as in countries with a communist past.

The authors emphasize, however, that these are descriptive findings. They do not allow for the conclusion that the conduct of elections or specific historical conditions caused the inaccurate reporting of mortality data.

Distinguishing correlation from causation is particularly important in this case. The results indicate the coexistence of certain characteristics between countries with greater data discrepancies, but they do not provide evidence of a causal mechanism.

Excess mortality as an independent benchmark

One of the key elements of the method used was the use of excess mortality as an independent reference point for official data.

The official number of COVID-19 deaths depended on the diagnosis of COVID-19 as the specific cause of death. This classification may have been influenced by testing and case reporting practices.

Excess mortality analysis, however, answers a different question: how many people died in a given period compared to the number of deaths that would be expected under normal circumstances?

This approach makes it possible to partially decouple the assessment of the impact of the crisis from the fact that each individual death was previously correctly identified and recorded as being related to COVID-19.

In the presented analysis, however, the estimate of COVID-19-related deaths was not a simple equation of all excess mortality with COVID-related deaths. As described in the study, the difference between the actual and expected number of deaths was taken into account, less excess deaths resulting from factors unrelated to COVID-19.

Why Independent Verification is Important

Comparing the official death count with an independent indicator allowed the authors to identify discrepancies that could not be easily explained solely by differences in countries’ ability to produce statistics.

The authors treat this mechanism as an example of a broader principle: the credibility of important official information can be greater when it can be verified using independent data.

Implications for epidemiology and public health

During the pandemic, case and death figures were not merely descriptive. They were used to inform decision-making at the national and international levels. According to the authors, data reported during COVID-19 influenced decisions regarding travel restrictions, among other things.

More broadly, governments regularly publish data on public health, the economy, unemployment, and other areas of public policy. The quality of such information can influence both domestic decisions and the actions of other countries and institutions.

The authors point out that incorrect official information can complicate international cooperation in areas such as foreign aid, climate action, and vaccination campaigns.

From an epidemiological perspective, the reliability of data used to assess the scale of the threat is particularly important. If mortality data are systematically underestimated or otherwise fail to reflect the true burden on the population, they can make it difficult to assess the course of the health crisis and compare its impact across countries.

The study is not an argument for rejecting public statistics.

The authors emphasize that the results should not be interpreted as an argument for automatic distrust of all official statistics.

Rather, the findings point to the importance of systems that allow for independent checking, questioning, and correction of information. Governments operating under stronger institutional and social constraints reported data with smaller discrepancies, while countries with weaker such mechanisms reported larger discrepancies.

In Karlinsky and Shayo’s analysis, the COVID-19 pandemic thus becomes not only the subject of a mortality study. It is also a natural, global case study that allows for an analysis of the conditions under which official information remains more or less reliable.

The importance of the results for future health crises

The study shows that the quality of information during a crisis does not depend solely on a state’s technical capabilities. Well-functioning registers, diagnostic systems, and statistical infrastructure remain important, but according to the results, they did not explain most of the differences observed between countries.

The conditions enabling an independent assessment of official data may also be important: institutional control mechanisms, public access to information and the ability of citizens to critically evaluate it.

The contemporary information environment further increases the importance of this problem. In the press release, the authors highlight the rapid spread of misleading information via digital platforms and AI-generated content. This makes mechanisms for correlating official statements with independent data even more important.

From the point of view of epidemiological surveillance, the COVID-19 pandemic highlights not only the problem of measuring the number of cases and deaths, but also the fundamental importance of being able to independently verify the data on which public health decisions are based.

Source

Source: Journal of Economic Growth , “Manipulation of information in times of crisis: Evidence from Covid excess mortality”, publication date: April 25, 2026.
DOI: https://doi.org/10.1007/s10887-026-09266-w

Source https://link.springer.com/article/10.1007/s10887-026-09266-w