Blessing or Curse of Democracy?: Current Evidence from the Covid-19 Pandemic

Ryan P. Badman, Yunxin Wu, Keigo Inukai, Rei AkaishiarXiv 2021 · DOI · arXiv

Abstract

Background: A major question in Covid-19 research is whether democracies handled the Covid-19 pandemic crisis better or worse than authoritarian countries. However, it is important to consider the issues of democracy versus authoritarianism, and state fragility, when examining official Covid-19 death counts in research, because these factors can influence the accurate reporting of pandemic deaths by governments. In contrast, excess deaths are less prone to variability in differences in definitions of Covid-19 deaths and testing capacities across countries. Here we use excess pandemic deaths to explore potential relationships between political systems and public health outcomes. Methods: We address these issues by comparing the official government Covid-19 death counts in a well-established John Hopkins database to the generally more reliable excess mortality measure of Covid-19 deaths, taken from the recently released World Mortality Dataset. We put the comparison in the context of the political and fragile state dimensions. Findings: We find (1) significant potential underreporting of Covid-19 deaths by authoritarian governments and governments with high state fragility and (2) substantial geographic variation among countries and regions with regard to standard democracy indices. Additionally, we find that more authoritarian governments are (weakly) associated with more excess deaths during the pandemic than democratic governments. Interpretations: The inhibition and censorship of information flows, inherent to authoritarian states, likely results in major inaccuracies in pandemic statistics that confound global public health analyses. Thus, both excess pandemic deaths and official Covid-19 death counts should be examined in studies using death as an outcome variable.

Introduction

Early in the Covid-19 pandemic, a controversial narrative emerged in the social sciences: the "Curse of Democracy." Proponents of this view argued that authoritarian regimes, with their centralized power and ability to enforce draconian lockdowns, were better equipped to handle the crisis than individualistic, decentralized democracies.

However, research by Ryan P. Badman and colleagues suggests that this purported advantage may be a data artifact. By looking beyond official government reports and analyzing "excess deaths"—the difference between observed mortality and historical baselines—the authors demonstrate that authoritarian and fragile states were not necessarily more effective at saving lives; they were simply more likely to underreport fatalities.

Methods and Data Integrity

The core challenge in global pandemic research is data integrity. Official death counts, such as those provided by Johns Hopkins University, are susceptible to two major biases:

  1. Institutional Capacity: Fragile states often lack the medical and administrative infrastructure to accurately test for Covid-19 or record deaths.
  2. Information Control: Authoritarian regimes may intentionally manipulate data or inhibit information flow to preserve their international standing.

To quantify these discrepancies, the researchers utilized the Undercount Ratio: $$\text{Undercount Ratio} = \frac{\text{Excess Death Count}}{\text{Official Death Count}}$$

This ratio was compared against established political metrics, including the EIU Democracy Index, the Polity5 Index, and the State Fragility Index (SFI).

Results: The Reporting Bias

The study found a significant correlation between political systems and the accuracy of pandemic reporting. More authoritarian and more fragile states exhibited substantially higher undercount ratios.

Correlations between political indices (EIU, Polity5, SFI) and the log undercount ratio.

As illustrated in the figure above, there is a strong negative correlation ($r \approx -0.60$) between democracy indices and the undercount ratio. Conversely, the State Fragility Index shows a positive correlation ($r = 0.508$), indicating that less stable governments are prone to massive inaccuracies in reporting, whether due to intentional censorship or structural incapacity.

The "Blessing" of Democracy

When the analysis is corrected using excess mortality data, the "Curse of Democracy" vanishes. Using the EIU Democracy Index, the researchers found a weak negative correlation ($r = -0.223, p = 0.043$) with excess deaths per 100,000 people.

This suggests that democratic systems—with their transparent information flows, public accountability, and capacity for innovation—likely handled the crisis slightly better than authoritarian regimes. The perceived success of authoritarian states in official data is largely an illusion created by underreporting.

Conclusion

The findings underscore the "censorship confound" inherent in authoritarian governance. Inaccurate pandemic statistics not only mislead the public but also confound global public health analyses and policy decisions. The authors conclude with a critical recommendation for the research community: official death counts should not be used as the sole outcome variable in cross-national studies without accounting for the political and institutional biases that shape them.

Summary updated

Abstract

Background: A major question in Covid-19 research is whether democracies handled the Covid-19 pandemic crisis better or worse than authoritarian countries. However, it is important to consider the issues of democracy versus authoritarianism, and state fragility, when examining official Covid-19 death counts in research, because these factors can influence the accurate reporting of pandemic deaths by governments. In contrast, excess deaths are less prone to variability in differences in definitions of Covid-19 deaths and testing capacities across countries. Here we use excess pandemic deaths to explore potential relationships between political systems and public health outcomes. Methods: We address these issues by comparing the official government Covid-19 death counts in a well-established John Hopkins database to the generally more reliable excess mortality measure of Covid-19 deaths, taken from the recently released World Mortality Dataset. We put the comparison in the context of the political and fragile state dimensions. Findings: We find (1) significant potential underreporting of Covid-19 deaths by authoritarian governments and governments with high state fragility and (2) substantial geographic variation among countries and regions with regard to standard democracy indices. Additionally, we find that more authoritarian governments are (weakly) associated with more excess deaths during the pandemic than democratic governments. Interpretations: The inhibition and censorship of information flows, inherent to authoritarian states, likely results in major inaccuracies in pandemic statistics that confound global public health analyses. Thus, both excess pandemic deaths and official Covid-19 death counts should be examined in studies using death as an outcome variable.
BibTeX
@misc{badman2021blessing,
  author = {Ryan P. Badman and Yunxin Wu and Keigo Inukai and Rei Akaishi},
  title = {{Blessing or Curse of Democracy?: Current Evidence from the Covid-19 Pandemic}},
  year = {2021},
  eprint = {2105.10865},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2105.10865},
}
Figure