Skip to main navigation Skip to search Skip to main content

Manipulation of information in times of crisis: evidence from Covid excess mortality

  • Ariel Karlinsky*
  • , Moses Shayo
  • *Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Governmental information manipulation has been hard to measure and study systematically. This paper provides comparable country-level estimates of the extent of misinformation in official data, in a large and diverse set of countries. We hand-collect data from a variety of sources in 134 countries to estimate misreporting of Covid mortality in 2020-2021. We find that information manipulation by governments during this period was remarkably widespread: between 45%–55% of governments misreport the number of deaths. Misreporting is large in magnitude: together, the countries we study reported 5.08 million Covid deaths during 2020-2021, whereas actual Covid mortality was more than double, at 12.47 million. The lion’s share of misreporting cannot be attributed to a country’s capacity to accurately diagnose and report deaths. Contrary to some theoretical expectations, information manipulation takes the form of downplaying rather than exaggerating the severity of the crisis. What characterizes governments that manipulate information? We examine the role of macroeconomic incentives, culture, audience sophistication, and institutions. The evidence strongly suggests that misreporting is higher where governments face fewer social and institutional constraints; in countries holding elections during this period; and in countries with a communist legacy.

Original languageEnglish
JournalJournal of Economic Growth
DOIs
StateAccepted/In press - 2026

Bibliographical note

Publisher Copyright:
© The Author(s) 2026.

Keywords

  • Covid
  • Information manipulation
  • Institutions
  • Mortality

Fingerprint

Dive into the research topics of 'Manipulation of information in times of crisis: evidence from Covid excess mortality'. Together they form a unique fingerprint.

Cite this