The research team assembled mortality data from 134 countries and territories, drawing on national statistical offices, population registries, health ministries and a wide network of other primary sources. Rather than relying on the cause-of-death classifications that each government submitted, the study anchored its analysis in excess mortality—the number of deaths recorded above what would have been expected in a normal year, based on pre-pandemic trends from 2015 onward. This methodological choice is central to the study’s power. Official Covid death counts depended on identifying Covid as the specific cause of death, a process vulnerable to gaps in testing availability, diagnostic practices, and reporting conventions that varied enormously across health systems. Excess mortality, by contrast, asks a simpler and harder-to-falsify question: how many people died compared with how many would normally have been expected? By carefully subtracting excess deaths attributable to non-Covid factors—such as reduced traffic fatalities during lockdowns or increased deaths from disrupted healthcare—the researchers constructed an independent estimate of true Covid mortality for each country.

The scale of the gap they uncovered is striking. Across the countries studied, governments collectively reported 5.08 million Covid deaths during 2020 and 2021. The researchers’ excess-mortality-based estimate puts the true figure at approximately 12.47 million—roughly two and a half times the officially acknowledged toll. Using a carefully constructed statistical framework, the authors estimate that between 45 and 55 percent of governments misreported Covid mortality in some form, with underreporting vastly more common than overreporting. The study then went a critical step further. The researchers deliberately avoided the assumption that every reporting gap reflects deliberate manipulation, recognizing that countries differ profoundly in their ability to diagnose illness, register deaths and maintain functioning vital statistics systems. A poor country with fragmented civil registration might simply miss many deaths without any political motive. So the team tested whether standard measures of state capacity could account for the observed discrepancies—and found that capacity explains only a small fraction of the variation. Something else was shaping what governments chose, or managed, to report.

That something, the study argues, is the structure of political incentives and the strength of the checks imposed on governments. When the researchers examined institutional constraints, a far stronger pattern emerged. Countries with robust systems of checks and balances—independent judiciaries, free presses, functioning opposition parties and accountable bureaucracies—showed dramatically less underreporting. Misreporting was also lower in countries whose citizens were better positioned to evaluate official information, measured through indicators including educational attainment and internet access. In societies where large numbers of people could observe burial activity, compare local hospital conditions with national claims, and cross-check official statements against independent sources, governments appear to have faced a real cost to publishing misleading figures. The mechanism the authors describe is not necessarily one of dramatic falsification; more often, underreporting likely took the form of passive failure to investigate, delayed or incomplete death certification, or narrowed definitions of what counted as a Covid death. But the aggregate effect was the same: where accountability was weak, the official record diverged sharply from demographic reality.

The study also uncovered two provocative descriptive associations. Countries that held national elections during the pandemic tended to show higher levels of misreporting, a pattern consistent with the incentives incumbent governments face when voters are heading to the polls and bad news carries electoral consequences. Countries with a communist legacy likewise exhibited elevated misreporting, potentially reflecting enduring institutional traditions in which official statistics serve political ends as much as informational ones. The authors are careful to emphasize that these relationships are correlational and descriptive rather than causal—they identify patterns, not proof of mechanism. Even so, the findings fit within a broader body of political economy research showing that governments routinely face temptations to present economic and social statistics favorably, and that what distinguishes accurate reporters from inaccurate ones is not intent alone but the presence or absence of institutions and audiences capable of detecting and punishing distortion.

The methodological contribution of the work deserves particular attention, because it extends well beyond the pandemic. Excess mortality functions as what social scientists call an independent benchmark—a measure generated by demographics and death registration rather than by the political process under scrutiny. When an official claim can be tested against a benchmark produced through a separate channel, the room for undetected manipulation shrinks considerably. This principle has applications far beyond mortality statistics. Claims about economic growth, unemployment, inflation, agricultural yields and environmental conditions all become more verifiable when independent evidence—satellite imagery, night-time lights, trade partner data, household surveys—can be brought to bear. The Covid episode demonstrated this dynamic at global scale under emergency conditions: because demographers and statisticians could reconstruct death totals from civil registries and historical baselines, the gap between claims and reality became measurable, and eventually undeniable.

The consequences of misreported mortality were not confined within national borders. During the pandemic, reported figures shaped international decisions about travel restrictions, border closures, vaccine allocation and the assessment of emerging variants. A country reporting low death counts might be treated as a safe travel partner or a low-risk source of new variants when the underlying reality was far grimmer. The study argues that inaccurate official information ripples outward, complicating international efforts in domains such as foreign aid targeting, climate action and global vaccination campaigns. Policymakers allocating scarce resources rely on the statistics governments provide; when those statistics are systematically distorted, resources flow to the wrong places and collective problems become harder to solve. In an era of fast-moving digital information flows and AI-generated content, the researchers note, the challenge of distinguishing reliable official data from favorable spin has become only more acute.

Importantly, the authors resist the conclusion that official statistics should now be treated with blanket suspicion. The study’s message is more nuanced and, in a sense, more constructive: trust in official information should be earned and maintained through verification, not granted by default or withdrawn wholesale. Governments facing strong institutional and social constraints generally reported accurately, and many countries—even some with limited resources—produced mortality data that closely tracked demographic reality. The lesson is not that governments lie, but that the reliability of what governments say depends on the surrounding ecosystem of accountability. Free media, professional statistical agencies insulated from political interference, transparent methodologies, engaged and informed publics, and independent researchers with access to raw data together form the infrastructure of truthful official communication. Where that infrastructure exists, even a crisis as chaotic as a global pandemic produced largely honest numbers. Where it was absent, the numbers bent toward political convenience.

Covid offered researchers an unusually clean natural experiment: governments across virtually every political system on Earth confronted a broadly comparable shock at roughly the same moment, making cross-national comparison meaningful in a way few crises allow. What the comparison reveals is a window into a much older and larger question—what makes official information trustworthy at all? Karlinsky and Shayo’s answer is that trustworthiness resides less in the technical capacity to count than in the political and social architecture that makes distortion costly. An informed public that can detect falsehood, and institutions that can challenge it, turned out to be better predictors of accurate death reporting than wealth or administrative sophistication. As the world confronts future pandemics, climate-related disasters and economic shocks that will again demand honest, timely data from every government simultaneously, the study’s central finding stands as both a warning and a guide: the numbers will only be as good as the accountability that surrounds them, and investing in strong institutions and informed citizens is, in a very literal sense, an investment in knowing the truth.

News Publication Date: 9-Sep-2026

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

References: Karlinsky, A., & Shayo, M. (2026). Manipulation of information in times of crisis: Evidence from Covid excess mortality. Journal of Economic Growth. https://doi.org/10.1007/s10887-026-09266-w