In 2020, governments globally implemented unprecedented policies to mitigate the COVID-19 pandemic, including lockdowns, travel restrictions, and business closures. These measures, while potentially life-saving, were associated with reduced economic activity as evidenced by decreased visits to workplaces and various establishments. While some argued that widespread economic disruption would have occurred regardless of restrictions, epidemiological studies suggest that these policies likely reduced mortality, although the causal link remains debated. The pandemic resulted in a significant global economic contraction, impacting global output and increasing poverty, disproportionately affecting low-income households and countries. This study aims to comprehensively analyze the prevalence of pandemic-related economic harm, its association with well-being, the relationship between harm and policy stringency, the variation in harm by socioeconomic status, and the comparative effects of different non-pharmaceutical interventions. The study leverages the Gallup World Poll, a globally representative dataset, which provides detailed demographic data, well-being measures, and information on pandemic-related economic impacts. The data's strength lies in directly measuring economic harm rather than relying on proxy measures like mobility. Policy intervention data are matched with individual responses, and a multilevel mixed model is used to analyze associations between individual and country-level factors and economic harm.

Existing literature documents the association between COVID-19 containment policies and reduced economic activity, using both global and country-specific analyses. Studies show the correlation between reduced mobility and economic activity with the implementation of these policies. However, some research suggests that significant economic disruptions would have still happened without the policies. While evidence suggests policies reduced mortality, the causal link remains inconclusive. Further, the pandemic's economic distress has been widely documented, with global output contracting significantly and global poverty increasing. Existing research consistently demonstrates the disproportionate impact of pandemic-related economic hardship on lower socioeconomic groups, a trend observed in previous pandemics. This paper expands on this existing body of knowledge by providing a more comprehensive global analysis.

The study uses data from the Gallup World Poll (July 9, 2020 – March 3, 2021) encompassing 321,386 individuals aged 15+ from 117 countries. The analysis focuses on five key survey items measuring economic harm: self-reported impact of COVID-19 on life, temporary job stoppage, permanent job loss, reduced work hours, and reduced income. Respondents were specifically asked to attribute these outcomes to the pandemic, mitigating omitted variable bias. Stringency of government restrictions is measured using the Oxford COVID-19 Government Response Tracker (OxCGRT), calculating cumulative stringency up to the survey date. The OxCGRT data evaluates policies on various dimensions, including business closures, school closures, stay-at-home orders, and other interventions. The study uses population-weighted averages for countries with sub-national data. The researchers validated their stringency measures by examining correlations with Google Community Mobility Reports (changes in visits to retail and restaurants), and changes in influenza transmission rates, verifying that stringent policies indeed reduced mobility and economic activity. Disease burden is measured using deaths per capita and model-based estimates of actual COVID-19 deaths from the IHME. The analysis employs multilevel mixed modeling to assess the associations between individual demographic characteristics, time-varying country-level variables, and measures of economic harm. The model accounts for intragroup correlation at the country and month-year levels. The individual-level variables include age, gender, immigration status, education, income, and urbanicity. Country-level variables are cumulative-to-date measures of COVID-19 restrictions, economic support policies, and COVID-19 deaths per capita. Interaction terms between stringency and income quintiles are used to examine heterogeneous effects. Additional analyses compare the effects of various non-pharmaceutical interventions on economic harm.

The study found that 42% of adults globally reported being significantly affected by the pandemic. Among those in the workforce before the pandemic, significant proportions experienced temporary job loss (51%), reduced hours (50%), income loss (49%), and job loss (27%). These outcomes varied significantly across countries and regions. A “harm index”, created by standardizing the five key harm variables, strongly predicted negative subjective well-being outcomes including changes in living standards, life evaluation, worry, and food insecurity. The primary analysis showed a robust positive association between policy stringency and economic harm. A one standard deviation increase in stringency predicted a 0.31 to 0.37 standard deviation increase in the harm index and a 14.2 to 15.9 percentage point increase in job loss. These relationships held even after controlling for individual demographics, disease burden, and economic support policies. The analysis also found that economic support policies were negatively associated with harm, but the impact of reported COVID-19 deaths was not significant. Additional analyses revealed that the impact of stringency on harm was significantly stronger for lower-income households, indicating a widening of economic inequality. Finally, comparing the effects of different policy interventions, the study showed that school closures, stay-at-home orders, and other economic restrictions were strongly associated with economic harm, but other public health interventions—such as contact tracing, mass testing, and protections for the elderly—were not.

The findings demonstrate a strong association between stringent COVID-19 suppression policies and increased economic harm, particularly for low-socioeconomic-status individuals. This supports the existing literature and provides more comprehensive evidence across various countries and contexts. The study highlights the importance of considering the potential trade-offs between public health goals and economic consequences when designing pandemic response strategies. The heterogeneous effects of stringency by income level underscore the need for policies that mitigate the disproportionate impact on vulnerable populations. The finding that certain non-pharmaceutical interventions, such as contact tracing, were not associated with economic harm suggests that alternative strategies might have been employed to minimize economic disruption while effectively controlling the virus. Future research should explore policy alternatives that minimize the economic fallout while maintaining public health goals.

This study provides compelling evidence that stringent COVID-19 suppression policies were associated with significant economic harm, disproportionately affecting low-income households. These findings underscore the importance of balancing public health goals with economic considerations during pandemic responses. Future research should focus on developing and evaluating alternative strategies that effectively control disease transmission while minimizing economic inequality and overall hardship. Further research is needed to explore the long-term economic and social impacts of the pandemic and the policies implemented to control it.

The study's findings are correlational and cannot establish causality. The non-random assignment of policies limits the ability to definitively conclude that stringency directly caused the observed economic harm. While the study controlled for various factors, omitted time-varying confounders at the country level could still influence the results. Data collection limitations, such as the shift from in-person to phone-based interviews during the pandemic, may have introduced some error. Further, the study does not provide a cost-benefit analysis of pandemic policies.