Global Evidence on the Economic Effects of Disease Suppression During COVID-19

doi:10.21203/rs.3.rs-2592296/v1

Download PDF

Article

Global Evidence on the Economic Effects of Disease Suppression During COVID-19

https://doi.org/10.21203/rs.3.rs-2592296/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 08 Jan, 2024

Read the published version in Humanities & Social Sciences Communications →

Version 1

posted

You are reading this latest preprint version

Governments around the world attempted to suppress the spread of COVID-19 using restrictions on social and economic activity. This study presents the first global analysis of the welfare impacts of those policies, using Gallup World Poll data from 321,000 randomly selected adults in 117 countries. Using several measures of economic harm, impacted individuals experience a loss in subjective-well-being and are more likely to have low socio-economic status. A one-standard deviation increase in policy stringency predicts a 0.28 standard deviation increase in economic harm, corresponding to a three-percentage point increase in the share of workers experiencing job loss. These results are supported by robustness checks and validation exercises. A decomposition shows that stay-at-home orders and other economic restrictions were strongly associated with economic harm, but other non-pharmaceutical interventions were not. Furthermore, we show that adults with lower socio-economic status were disproportionately harmed in countries with more stringent policies.

Social science/Economics

Business and commerce/Economics

Social science/Development studies

Scientific community and society/Developing world

In 2020, governments around the world started putting in place extraordinary policies to mitigate the spread of the coronavirus, including stay-at-home orders, restrictions on travel and gatherings, and closures of schools and workplaces. These polices have been found to be associated with reduced economic activity, in the form of visits to workplaces, parks, restaurants, and non-grocery retail establishments, both in global analyses (Hale et al, 2021), and based on detailed evidence from single countries or a subset of countries (Deb et al 2020; Boone and Ladreit 2021; Lozano Rojas et al 2020; Aminjonov, Bargain, and Bernard 2021; Carvalho et al 2020; Coibion, Gorodnichenko and Weber 2020; Gathergood et al 2021, including illegal economic activity (Nivette et al 2021)). Using U.S. data from early in the pandemic, other scholars have argued that widespread economic disruptions would have happened in the absence of restrictionsy (Goolsbee and Syverson 2021; Forsythe et al 2020; Gupta, Simon, and Wing 2020).

Epidemiological theory and empirical evidence suggest that these policies likely reduced the number of deaths (Liu et al 2021; Chernozhukov, Kasahara, and Schrimpf 2021; Violato et al 2021; Qi, Zhang, and Zhang 2022), although given the methodological challenges, the evidence on the causal links between mobility restrictions and COVID-19 mortality is not entirely consistent (Herby, Jonung, and Hanke 2022; Spiegel and Tookes 2022). Similarly, public health policies that do not restrict economic activity—such as contact tracing (Fetzer and Graeber 2021) and surgical mask use (Abaluck et al 2022)—have been found to reduce infections.

Regardless of the causes, pandemic-era economic distress has been widely documented. Global output contracted by 3.4 percent in 2020, with output contractions observed in 95 percent of countries, a scale that rivals that of the Great Depression (World Bank, 2022), and global poverty increased (Mahler et al., 2021; Kim et al 2021). Furthermore, low-income countries faced widescale income losses (Egger et al 2021; Josephon, Kilic, and Michler 2021), Within countries the economic effects of the pandemic have been worse for households with relatively low socio-economic status, as measured by income rank or educational attainment (Rothwell and Smith 2022; Narayan et al. 2022; World Bank 2022; Bundervoet et al. 2021; Kugler et al. 2021), which is consistent with past pandemics (Furceri et al 2022).

This paper contributes to the literature by providing the most comprehensive analysis to-date on several related aspects of the pandemic: the prevalence of economic harm caused by the pandemic and how it relates to subjective wellbeing and financial security; the uneven distribution of that harm by socio-economic status within countries; and the differential effects of different types of COVID-19 restrictions on job loss, income loss, and similar outcomes. The analysis relies heavily on the Gallup World Poll, which used randomized sampling procedures in 117 countries representing nearly three-quarters of global population. From July 2020 to March 2021, the survey collected detailed demographic data on income and education, subjective wellbeing measures, and information on several economic outcomes in which respondents were explicitly asked if they were caused by the pandemic. The resulting database provides the only harmonized quasi-global database available to study individual and country level employment and income outcomes.

Validating a novel measure of economic harm

Before describing the primary results, we establish grounds for accepting the validity of our key measures. Further information is provided in the Methods section and Supplementary Text. First, we create a “harm index” as the standardized individual-level mean of responses to five survey items about how respondents’ lives have been affected by the coronavirus situation. They are as follows: whether their lives have been affected, whether they lost their job or business temporarily or permanently (two distinct items), whether they worked fewer hours, or whether they received less money. Using the global sample, each item is standardized to have a mean of zero and a standard deviation of one.

In the Supplementary Text, we show that the economic harm index—and its component parts—strongly predict four measures of subjective well-being, covering 1) changes in subjective living standards, 2) current life evaluation, 3) experiences of worry, and 4) lack of money for food (wording is provided in Supplementary Table 3). Essentially, we regress these outcomes on the harm index, controlling for respondent demographics and county fixed effects and find large and significant conditional correlations, indicating that our measure of economic harm has consequences for subjective well-being and financial distress (Supplementary Fig. 1).

Next, we check the reliability and validity of World Poll data on employment losses against alternative sources. World Poll data on the job loss rate are broadly aligned with administrative data on changes in the official unemployment rate (correlation is 0.52 in 52 countries). Yet, in addition to broader coverage, the World Poll measures are superior in two respects: harmonization in measurement and a causal link with COVID-19. With rare exceptions in normal times, “unemployment” requires that adults are out of work but seeking and able to work. If that latter two conditions are unmet, the person is considered out of the labor force, but not unemployed. COVID-19, however, resulted in many people losing their job but temporarily halting efforts to find a new one—for various reasons. Statistical offices around the world took different non-harmonized approaches to classifying such persons, resulting in biased unemployment rate levels and changes. Moreover, COVID-19 was not the only causal factor affecting social and economic conditions around the world, so the World Poll data also improve conceptual validity by asking respondents to attribute their economic harm to the pandemic and allowing them to express it along several dimensions (see Supplementary Text for further discussion of these issues).

Finally, we show that our primary policy measure also meets basic validity criteria, as discussed in Hale et al (2021). Stringency is weakly and positively related to COVID death rates but more closely related to measures of social-distancing, particularly those involving declines in visits to restaurants and small businesses (Supplementary Text and Supplementary Fig. 2). We examine two additional and related outcomes. In more stringent countries, self-reported social contact (available from a non-representative alternative survey covering a smaller number of countries) tends to be lower and reported cases of seasonal flu fell further from baseline season-adjusted trends—for the subset of countries with high quality flu data. This provides further evidence that the behaviors associated with respiratory disease transmission (e.g. social-contact) fell further where disease-suppression policies were strongest, but flu case data are likely more informative than COVID-19 case count data, since flu surveillance systems were well-established before 2020, and COVID surveillance relied on novel tests that were neither available uniformly globally nor across regions within countries. Taken together, this evidence suggests a plausible link between stringency and economic outcomes.

Stringency measures and economic harm: main results

We now proceed with the main research question – whether more stringent restrictions are associated with a greater degree of economic harm. The stringency of mitigation policies is measured by the COVID-19 Government Response Tracker (Hale et al 2021). Harm is aggregated from the World Poll microdata, using sample weights to ensure representative coverage. The analysis regresses harm on stringency (see Methods).

Column (1) of Table 1 reports a bivariate correlation between policy stringency and economic harm. The estimate is equivalent to an 11-percentage point increase in the mean of the economic harm index (or 0.42 standard deviations) for every 1 standard deviation increase in the policy stringency index.

The unconditional correlation may be confounded by other factors. In particular, one can think heuristically of the degree of economic harm from restrictions to economic activity as a function of 3 main factors (i) the stringency of lockdown policies; (ii) the ability of workers to continue working remotely; and (iii) the extent of mitigation measures deployed by the government to alleviate the welfare impacts of the economic crisis. In columns (2)–(4) we control, one by one, for the variables that serve as proxies for these main transmission channels.

To capture the ability of workers to work remotely, we employ the measure of broadband connections per capita, which we predict will minimize harm from stringency by fostering remote work (Neidhofer et al., 2021; Narayan et al, 2022; World Bank, 2022). In line with our heuristic priors, we find higher levels of broadband connectivity to be strongly and negatively correlated with the degree of economic harm – a 1 standard deviation increase in broadband connections per capita is associated with an 18 percentage point decrease in the harm index. The correlation between the stringency index and economic harm become somewhat smaller once we control for broadband connectivity, falling from 11 to 8 percentage points, but it remains highly significant.

Next, we control for the degree of government’s economic support, as measured by the Oxford economic support index. The index captures the record of government providing direct cash payments to people who lose their jobs or cannot work, including payments to firms that are linked to payroll/salaries, as well as the record of the government freezing financial obligations for households (e.g. stopping loan repayments, preventing services like water from stopping or banning evictions). The estimates suggest that the degree of economic support is not correlated with the extent of economic harm, which is related to the fact that we are conditioning on broadband connectivity, and both the extent of mitigation measures and broadband connectivity are strongly correlated with GDP per capita. The bivariate correlation between harm and economic support is negative, as we would expect, at -0.31.

In column (4), we further condition on the measure of COVID-19 deaths per 100k population to check whether the degree of economic harm is a function of the health impact of the COVID-19 pandemic. We find that conditional on the policy stringency index, the death toll of the COVID-19 pandemic is not correlated with our measure of economic harm. We return to this point in the robustness analysis.

Finally, we consider that large pre-pandemic social-economic status gaps (measured by income inequality) may have affected both the policies and the outcomes. Therefore, we control, in column (5), for the pre-pandemic level of inequality as measured by the Gini index. Doing so, reduces the coefficient on stringency very modestly (from 0.08 to 0.07). The model shows that higher levels of pre-pandemic inequality are associated with greater economic harm – a 1 standard deviation increase in the Gini index is associated with an 8-percentage point increase in economic harm (0.28 standard deviations). Even with this included in the mode, a one-standard deviation increase in stringency predicts a 0.27 standard deviation increase in harm. This is our preferred estimate of the relationship between stringency and harm.

Table 1

Restrictions on economic activity and the harm index: cross-country evidence
	(1)	(2)	(3)	(4)	(5)
Standardized Stringency index	0.11***	0.08***	0.08***	0.08***	0.07***
	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)
Broadband connections per capita		-1.24***	-1.23***	-1.19***	-0.87***
		(0.11)	(0.13)	(0.16)	(0.17)
Economic support index			-0.00	-0.00	-0.00
			(0.02)	(0.02)	(0.02)
Log of COVID deaths per 100k				-0.01	-0.02
				(0.01)	(0.01)
Income inequality (Gini index)					0.92***
					(0.22)
Constant	-0.04	0.19***	0.18***	0.19***	-0.18*
	(0.02)	(0.02)	(0.03)	(0.04)	(0.09)
Observations	113	111	111	110	107
Adjusted R-squared	0.16	0.61	0.61	0.60	0.66
Notes: Dependent variable – harm index. Standard errors in parentheses. * p < 0.01, p < 0.05, * p < 0.1

Sensitivity analysis of main results

To test the robustness of our primary findings, we conduct four types of robustness checks, each of which is described in greater detail in the Supplementary Materials: 1) we decompose the harm index into its components and otherwise replicate the main analysis; 2) we vary the measurement of key constructs and the underlying theoretical heuristic; 3) we run a modified analysis using a panel of countries and country-fixed effects; 4) we replicate the analysis using out-of-sample data and external sources within the United States, taking advantage of detailed state-level data; additionally, we use pre-COVID U.S. political-party strength at the state level to identify exogenous variation in state-level COVID stringency.

In the first check, we re-run model 5 from Table 1 (our preferred specification) but change the outcome from the harm index to one of the five underlying constructs (i.e. lost work hours, temporarily laid off, lost income, affected a lot by COVID, or lost job/business). Stringency is positively and significantly related to each outcome at 95% confidence intervals (Supplementary Table 4). In standard deviation terms, the smallest effect of stringency is on lost income (0.18 st. dev.) and the largest is affected a lot (0.35 st. dev.). The harm index effect falls in between at 0.28.

Next, we explore other threats to validity by varying the sample and measurement. We run ten alternative models, using our preferred specification as the baseline. Results are available in Supplementary Table 5. Model 1 includes six binary variables for U.N. regions, under the assumption that disease may spread more readily within regions than across them. Model 2 uses more refined sub-regional effects (14 units). Model 3 replaces the official death count with a model-based estimate from (IHME 2021). The fourth uses seroprevalence estimates of infection rates instead of deaths per capita (which should be more accurate than case-count infection rates, given uneven testing access and motivation). The fifth replaces the economic support index with GDP per capital (PPP-adjusted). The sixth, includes controls for population density and median age, since the former predicts true infection-risk, all else equal, and the latter predicts actual health-risk. The seventh adds age, density, and controls for compliance with social-distancing (using Google Mobility visits to restaurants and similar places); the eight replaces the Google mobility measure with self-reported social distancing from a Facebook-based survey (Fan et al 2020); the ninth restricts the sample to low-income and lower middle-income countries (41 countries), and the tenth restricts to high and upper-middle-income countries (65 countries). Stringency remains significant at 95% confidence levels in each model and the coefficient is largely consistent.

Third, we use an out-of-sample replication of our model within the United States, taking advantage of state-level variation in policy stringency (also from Hale et al) and a U.S. Census Bureau data collection effort that asked adult respondents if anyone in their household had experienced a loss of work-related income in the last 4 weeks. Without using World Poll data, we are able to measure the same constructs used in Table 1 and conduct the analysis across 50 states and the District of Columbia, which has its own quasi-state level control over public health policy. Similar to our preferred model, we include controls for the teleworking rate, median household income, deaths per capita, and household income inequality (Gini coefficient).

The U.S. analysis is remarkably consistent with our global analysis. Regressing income loss rates on stringency, the coefficient on stringency is the same for both the U.S. sample and the global sample (0.03, see Supplementary Table 6), though the variance is much smaller within the United States. In our global sample, a one-standard deviation in stringency predicts 0.18 standard deviation increase in the percent of workers reporting income-loss. In the U.S. sample, a one-standard deviation increase in stringency predicts a 0.49 standard deviation increase in income loss.

The U.S. analysis allows one other advantage in causal identification. As has been widely noted (see Canes-Wrone, Rothwell, and Makridis 2022 for one example), partisan political power is highly correlated with COVID-19 related policies and voter preferences, even conditional on the measured disease burden—which is unrelated to stringency practices or preferences. Thus, state partisanship provides an exogenous source of variation in policy that is independent of the disease burden. We construct an instrumental variable for partisan power that is increasing in Republican Party influence over state institutions. This variable predicts much lower stringency. Our two-stage least squares estimator for the effect of stringency on income loss is highly significant (p < .01), with a coefficient of 0.06. This estimate can be interpreted as a causal effect in so far as factors affecting state-level economic performance during the pandemic did not vary by party control of the state, except in how they restricted social and economic activity (Supplementary Table 6).

Fourth, we run models that allow stringency and harm to vary within countries so we can control for country-factors that do not vary over the pandemic. This is possible because the World Poll used multiple waves of data collection. An important difference in this analysis is that we use the current unemployment rate and current deaths per capita instead of our cumulative harm and disease burden measures. We do this because cumulative measures include lagged country-level effects; thus, their inclusion would reintroduce the bias we are aiming to mitigate. The analysis finds that workers were more likely to be unemployed during the periods in which their country adopted tighter restrictions. A one standard deviation increase in the stringency index predicts a 0.9 percentage point increase in the unemployment rate, which has a standard deviation of 5 percentage points (Supplementary Table 7). Using this same set-up, we show that stringency also reduced visits to restaurants and reduced flu-transmission relative to historic expectations. This is consistent with stringent policies affecting actual behavior (e.g. social contact) relevant to the economy. Details and limitations of this analysis are discussed in the Supplementary Text.

Alternative policy responses and economic harm

We consider that restricting social and economic activity was not the only tool available to public health officials. Widespread testing, meticulous contact tracing, social-distancing focused on the elderly, travel restrictions, and use of facial coverings outside of one’s home are alternatives to universal social-distancing and some do not necessarily limit economic behavior. The Oxford database tracks these and several other policies, and we tested these in our preferred model that controls for COVID-19 deaths, broadband access, government capacity, and household income inequality (column 5 of Table 1). We also run a parsimonious model with only GDP per capita and broadband access.

In predicting the harm index, we find that some, but not all policy measures are positively associated with economic harm and many have no significant relationship. Using our preferred specification, the six measures for which we find positive and significant correlations with economic harm are the stringency index, stay-at-home orders, internal travel restrictions, the closure of public transportation, school closures, and the cancelation of public events. At the same time, protection of the elderly, vaccination policy, contract tracing, public information campaigns, and mask orders are not associated with greater economic harm (Fig. 1, left bottom panel). These results are largely consistent with a parsimonious model, except testing policies and workplace closings become significant and positively associated with harm.

We run the same analysis for the job loss rate (Fig. 1, right panels). In addition to the stringency index, school closures, the closing of public transportation, restrictions on gatherings, and stay-at-home orders are significantly associated with a higher job loss rate. On the other hand, we do not find a statistically significant positive correlation between job losses and measures such as testing policy, contact tracing, and masks orders or protecting the elderly. Workplace closings are significantly associated with harm in the parsimonious model, but in the preferred model, the effect becomes insignificant.

Taken together, these findings suggest that disease mitigation effort does not necessarily predict economic harm; while no policy significantly predicts less harm, those that predict greater harm all directly restrict economic and social activity. Policies targeting health-interventions rather than economic activities, on the other hand, seem relatively harmless to the macroeconomy. These include contract tracing, testing, and mask orders.

Socio-economic Status And Heterogenous Effects Of Stringency: Individual-level Analysis

The previous discussion considered the average country-level effects of policy stringency on economic harm. Here, we consider that public health policies, even when they are implemented uniformly at the national level, may not affect all households and individuals the same way. To do so, we return to the full individual-level microdata from the World Poll and examine whether national restrictions played out differently across population groups.

This analysis aims to extend this literature on the distributional impacts of the COVID-19 pandemic in several important ways. First, it offers the most comprehensive global coverage, representing nearly three-quarters of the global population (5.7 billion adults) in a large randomly-selected, representative sample. Second, we are able to capture distributional impacts using more holistic measures of economic harm, whereas most of the earlier studies, especially in developing countries, focused primarily on job losses. Third, we are able to estimate heterogeneous impacts across income quantiles within countries, whereas most of the earlier studies in developing country contexts had to rely on proxies such as the education level of respondents or their place of residence to make inference on the socio-economic gradient in the COVID-19 impacts.

In order to examine the socio-economic gradient associated with the policy stringency measures, we start by examining how the degree of economic harm varies across different population groups. We estimate multivariate regression with several alternative dependent variables, including (i) whether the household has been affected a lot by COVID-19; (ii) whether one experienced job loss on account of COVID-19 (temporarily or permanently), and (iii) our combined harm index. Our model regresses harm on the demographic indicators listed in Fig. 2, while additionally controlling for religious preference, country-effects, and year-month effects.

The results reveal considerable heterogeneity in the impact of economic restrictions on job losses and welfare within countries, with the most striking result being the higher economic harm among those at the bottom of the within-country income distribution. The degree of economic harm is highest in the bottom income quintile, monotonically decreasing for higher income levels.

These differences are large and economically significant. For instance, the probability of reporting a lost job due to COVID-19 is 17 percentage points higher in the bottom quintile than in the top quintile, on average. In similar vein, those with secondary education were 5 percentage points more likely to lose their jobs compared to those with tertiary education, and those with primary education were 9 percentage points more likely to lose jobs than those with secondary education. The degree of economic harm is higher for those with lower levels of education. These results are consistent with the findings of several earlier studies that similarly find greater job losses among the more vulnerable population groups (Bundervoet et al., 2021; Kugler et al., 2021; Narayan et al., 2022; World Bank, 2022).

When we aggregate the various dimensions of the welfare impacts of COVID-19 into our harm index, the estimates reveal that the impact of stringency measures on economic harm is 20 percentage points higher in the bottom income quintile compared to the top income quintile (0.27 standard deviations). Likewise, for those with only primary education the predicted value of the economic harm index is 9 percentage points higher relative to those with tertiary education (0.12 standard deviations).

Women were more likely than men to report being “affected a lot” by COVID-19, but no more likely to score higher on the harm index, nor to lose their job permanently, and slightly less likely to experience a temporary layoff. Middle-aged adults and residents of cities were more likely than adults aged 60–64 (the reference group) and young adults to experience harm. Likewise, urban dwellers were more likely to experience harm than those in rural areas. To summarize these results, the pandemic has resulted in disproportionate harm to households with lower socio-economic status, especially if middle-aged and living in cities.

Given the link between harm and stringency established above, we would also like to test whether the socio-economic status relationship with harm widens or narrows as country utilize more stringent COVID restrictions. To investigate this question, we construct quartiles of the Oxford policy stringency index, based on the overall distribution of the index in our sample, and re-estimate our models separately for each quartile of the policy stringency index distribution. For illustrative purposes, Fig. 3 compares the estimates across the 1st and 4th quartiles of the stringency index, with the first quartile corresponding to the lowest levels of stringency and the 4th quartile to the highest levels of stringency.

The estimates suggest that socio-economic status gaps become even larger as the degree of policy stringency becomes higher. For instance, households in the bottom income quintile are 13 percentage points more likely to lose a job relative to households in the top income quintile when the levels of economic stringency measures are relatively low (1st quartile), but as restrictions to economic activity become severe (4th quartile), job loss becomes 24 percentage points more likely in the bottom income quintile compared to the top quintile. The relative risk of job loss for those with a primary education compared to those with a tertiary education is roughly 5.2 percentage points higher in countries with severe restrictions compared to those with more relaxed restrictions (13.7 vs 8.5). We find similar results for the harm index and whether the respondent was affected a lot.

The heterogeneity in the degree of economic harm across different age groups also becomes more pronounced as stringency increases, though the error margins overlap. Overall, those in the middle of the age distribution report higher rates of harm.

Analytically, this approach is equivalent to regressing harm on an interaction term that multiples an indicator of country-level stringency with every demographic measure and observing large effects between stringency and low-income status. An alternative approach would be to interact the stringency index with only household income (as measured in within-country quintiles), omitting other controls that may be correlated with socio-economic status (see Supplementary Text and Supplementary Fig. 3). That analysis confirms that the socio-economic status gap widens as stringency increases, and in some cases, the results suggest that there is no socio-economic status gap for the subset of countries with very low stringency scores (like Sweden). Moreover, higher stringency does not predict higher levels of harm for households near the top of the income distribution, only those in the middle and lower rungs, which may shed light on the political demand for restrictions.

The restrictions to mobility and economic activity deployed by governments to contain the spread of coronavirus presumably saved many lives, by delaying transmission until effective treatments and vaccines became available. At the same time, the COVID-19 pandemic has also resulted in significant economic hardship for many households around the world, and restrictions may have exacerbated the scale and severity of hardship. Our research uses the best available data to describe these outcomes, how they varied within and across countries, and to what extent they were linked to pandemic restrictions and other public health policies.

Setting aside policy issues, we document that the economic effects of COVID-19 were broadly harmful to subjective wellbeing and financial security, and that this hardship was greater—around the world—for people with lower socio-economic status, as measured by educational attainment or within-country income rank.

Turning to the results of our policy analysis, we find remarkably robust evidence that countries adopting more stringent disease suppression policies experience higher rate of economic harm. This is true across a broad range of models, using a reasonable set of controls. These controls included policies to counter-balance shutdowns with government support, the actual disease burden, pre-COVID income inequality, and the capacity of households to work remotely. The results hold while adjusting for the observed and self-reported social-contact behavior of the population, including changes in flu-cases and visits to restaurants. Within the United States, we replicate these findings and find further evidence that stringency causes wider income losses by using pre-COVID partisan political power to identify policy stringency. We also find that—controlling for country fixed-effects—countries saw economic damage increase during the periods of most intense stringency. Finally, the socio-economic status gaps widen as stringency increases, suggesting that more stringent policies are most harmful to household with low levels of income or education. These findings partially confirm the work of previous scholars and add richer, more comprehensive evidence to support that work.

Additionally, we find that economic harm was reduced as the capacity to work remotely increased and economic harm—and stringency—were both positively correlated with pre-pandemic measures of household income inequality. The disease burden, surprisingly, was not related to economic harm, even using alternative model-based measures. GDP per capita predicted less economic harm; economic support policies, as measured by the Oxford team, did not. Importantly, stringent economic and social restrictions predict more severe harm globally, but other public health efforts—such as mask mandates, contract tracing, widespread vaccination, and special protections of the elderly did not predict economic harm.

A limitation of this analysis is that disease-suppression policies—like all government laws—are not randomly assigned, even though many of the policies studied were novel in their application as of 2020. Our extensive robustness checks, use of alternative models, and replication of the results across U.S. states should mitigate against omitted variables bias, but we cannot completely rule it out. It is also difficult to conceive of how any analysis could. Few variables successfully predict the disease burden of COVID, making instrumentation difficult. We exploited one potentially exogenous source of sub-national (state level) policy variation within the context of a large country—the United States—with two dominant parties that arrived at opposing positions on COVID-19 restrictions.

A key question for future research is whether alternative policies could have been enacted—or could be enacted during future pandemics—that are able to save lives and mitigate disease without generating widespread economic harm, nor disproportionate harm to the poorest households.

The hypotheses tested in this paper require data on three main components of our empirical models: (i) measures of economic harm or welfare impact; (ii) measures of the stringency of economic restrictions imposed by governments in the wake of the COVID-19 pandemic, and (iii) measures of the disease burden of COVID-19. These are described in turn in this section. Summary statistics from the county level and individual data are available in Supplementary Table 1. The analysis was conducted using Stata 17.0.

Economic harm measures

The main source of data supporting the analysis in this paper comes from surveys fielded by the Gallup World Poll between July 9, 2020 and March 3, 2021, with 321,386 observations of people aged 15 and older in 117 countries/territories.² The survey included demographic information as well as items related to health and wellbeing that were designed to be nationally representative for each country in the sample.³ The relevant ethics statement is provided in the Supplemental Materials.

The main focus of this analysis is on five survey items that broadly measure social or economic harm from COVID-19. The first item, fielded to all survey respondents solicits an answer to the question “In general, to what extent has your own life been affected by the coronavirus situation?” We recode responses as a binary variable which takes a value of one if the response is “a lot,” and zero if respondents reply “some” or “not at all.”

The second item, applicable only to people working at the start of the pandemic, solicits whether respondents have experienced each of the following as a result of the coronavirus situation?

Temporarily stopped working at your job or business
Lost your job or business
Worked less hours at your job or business
Received LESS money than usual from your employer or business

Respondents are instructed to answer “Yes” or “No”, or “Does not apply” if they did not have a job leading up to the pandemic. The World Poll includes a large number of other respondent-level variables, which we only briefly describe below in context. On average, 42 percent of adults responded that they were affected a lot; 24 percent reported permanently losing their job; 48% reported a temporary job loss, and 47% reported lost income (Supplementary Table 1).

The advantage of the above survey items for the purpose of assessing the welfare impacts of the COVID-19 pandemic, vis-à-vis more traditional measures of economic changes, such as employment status or income, stems from the fact that the respondent is asked to attribute the severity of the overall impact, or the different aspects of job and income losses to the pandemic in a causal sense. This is important because many factors other than COVID-19 could cause people to lose their job, leave the labor force or experience emotional stress. Thus, while the analysis in this paper relies on cross-sectional variation, the framing of the key questions related to impacts help at least partially guard against common omitted variable bias concerns in such settings.

A second advantage of these items is that respondents are well-placed to know if an economic change in their life was caused by the pandemic. The event itself was highly salient and often highly disruptive to daily life, with clear time boundaries, tied to events like international and national emergency declarations and stay-at-home orders. In many cases, employers may have specifically told employees that the cause of their layoff was the pandemic, but even in the absence of that messaging, respondents would be well-aware of the timing of their job loss and the circumstances leading up to it, which may include the business being shut down or customers cancelling contracts or no longer showing up.

Stringency of economic restrictions measures

The analysis relates the above measures of economic harm to the stringency of restrictions on economic activity imposed by governments. We measure stringency of lockdowns by using data from the Oxford COVID-19 Government Response Tracker (Hale et al 2021). The database evaluates national and sub-national government policies along various dimensions.

Data are coded on an ordinal scale, such that values are increasing with stringency. For business closures, 0 means no measures, 1 means recommended closing or recommended work from home, 2 requires the closing of some sectors, and 3 requires the closing of all but essential workplaces, such as healthcare offices and grocery stores. The data are collected for every day since the start of 2020, which facilitates our analysis. Since we are studying the cumulative economic effects, we want a measure of the cumulative lockdown, which we measure as the average stringency over the entirety of 2020. Hale et al (2021) constructed a “stringency index” that is the mean value of the stringency of the 8 containment policies and one health-related policy. The latter measures the degree to which public officials urged caution through coordinated efforts to promote social distancing and related behavioral changes across traditional and social media (Supplementary Table 2).

In our decomposition analysis, we focus on these containment measures, as well as “health systems policies” which include regulations of facial coverings, contact tracing, testing, emergency investment in healthcare, investment in vaccines, vaccination policy, and protections of the elderly.

In most cases, each country has a single daily measure for each indicator. Sub-national data is available for only Brazil, Canada, the United Kingdom, and the United States. For these countries, we obtain population data from national statistical offices for the sub-regions and use these as weights, so that the national value is a population-weighted average of the subregional policies. ⁴

It should be noted that stringency measures, such as those related to social distancing and restrictions to mobility, need not result in economic harm if the degree of compliance is low, either because people are unable or do not want to comply with the measures and the authorities lack the capacity to enforce them. Thus, it is important to ascertain that stringency measures are actually binding and do, in fact, lead to reduced mobility and economic activity. To verify this, we rely on several sources of data that speak to mobility and social distancing dynamics.

One direct way of looking at restrictions to mobility is with the aid of data from Google Community Mobility Reports. Using mobile phone location software, these data show the percentage change in visits to various places from a pre-COVID baseline (January 3 to February 6, 2020). We focus only on visits to retail and restaurants, described by Google LLC (2021) as: “places like restaurants, cafes, shopping centers, theme parks, museums, libraries, and movie theaters.” Roughly half of the adult population in the sample reported direct contact with non-household members, and visits to restaurants and similar places were down 25 percent on the average day through the end of 2020.

One additional proxy measure for the degree of social distancing that is directly relevant to COVID-19 transmission is related to changes in the transmission of a parallel respiratory virus that was common before the pandemic. A large drop in the transmission of this parallel virus would suggest major behavioral changes relevant to disease transmission, whereas the absence of change in its transmission would suggest limited behavioral changes. Influenza is close to ideal in providing this analytic opportunity. The most significant problem is that flu cases are measured only based on testing, and flu testing conditional on symptoms—like COVID testing—is likely to vary by country. At the same time, by comparing pre-COVID flu rates to COVID-era flu rates we control for unchanging country-level testing infrastructure and practices. We, therefore, believe these data provide a valid measure of changes in social distancing that include both policy-induced and non-policy induced behaviors. These data are from the World Health Organization’s FluNet and include total positive influenza cases per week by country for each year from 2016 to the 45th week of 2021. The use of weekly data allows us to adjust for seasonal effects, which vary by hemisphere and countries within hemispheres. We are interested in the percentage change in weekly cases during flu season before and after the pandemic, ending the analysis on the 12th week of 2021 to coincide with the World Poll data collection period. To identify the flu season for each country, we calculate the weekly share of cases from 2016 to 2019 and classify any week with at least 1% of annual cases as being part of flu season. For the United States, this would include weeks 1 to 16 and weeks 47 through 52 of every year. For Australia, in the Southern Hemisphere, it would include weeks 20 to 41. Unfortunately, FluNet data is far from being comprehensive and many countries are either entirely missing or only report a few weeks out of the year. We require 90% reporting coverage during flu-season weeks before and after the pandemic. We find that flu cases in the 2020–2021 flu season were just 18 percent of the mean number of flu cases measured from 2016 to 2019 flu season in 74 countries.

As part of our robustness checks, we analyze the relationship between stringency and changes in flu rates at the country-month level. For this analysis, we calculate positive flu cases in the current year (2020) relative to the previous year (2019) and the year before that (2018) and take the average of these two rates before aggregating this average to months. This gives us a measure of flu rates relative to previous years that varies by month and country.

We include an additional measure of subjective social distancing. In partnership with Facebook, the University of Maryland fielded a large-scale global daily survey of Facebook users (the COVID-19 World Symptom Survey Data), reweighted to be representative of the national population (Barkay et al 2020). Using the University of Maryland API (Fan et al 2020), we were able to get weighted data for the percentage of respondents who have reported having had direct contact (longer than 1 minute) with people not staying with them within the past 24 hours. Using data aggregated across 103 countries in 2020, 46 percent of respondents report direct contact with a standard deviation of 10 percent.

Disease burden measures

The disease burden is measured in terms of deaths per capita, which is preferable to the number of COVID-19 cases. We do not claim that every national health system is equally likely to capture and correctly identify every COVID-19 death, but nearly every country has formal systems to record the causes of death. By contrast, the probability of seeking testing conditional on the experience of symptoms is highly contingent on factors that vary widely by country, such as cost, guidelines on testing priorities, and the availability of tests. Asymptomatic testing, moreover, also varies widely by country. In short, data on COVID-19 cases per capita are very noisy measures of the disease burden relative to deaths per capita. Our data on both come from the European Centre for Disease Prevention and Control. We use cumulative deaths per capita through 2020, because we are interested in cumulative economic effects throughout the second half of 2020 and into early 2021.

To further guard against measurement error—and potential bias stemming from lack of reporting—we include model-based estimates of deaths from COVID-19 in our analysis from the University of Washington’s Institute for Health Metrics and Evaluation (IHME 2021). That analysis uses actual all-cause mortality data for 56 countries, subtracts out known increases in deaths, and determines estimates of actual COVID deaths. The research team then models the ratio between reported deaths and actual deaths for every country to arrive at a measure of total COVID deaths. We regard these as credible alternative measures of disease burden, as argued in (Wang et al 2022).

Other variables

To measure income inequality, we use data from All the Ginis (Milanovic 2019). Since the timing of coverage varies widely by country, we calculate mean household income Gini coefficients by country from 2000 to 2017, using the year as an analytic weight. This approach maximizes country coverage while prioritizing measurements from the most recent years. The other measures referred to in our modelling work draw upon conventional and well documented sources from the World Bank (population density) and United Nations (median population age).

² While China was part of the World Poll, it is not included in the analysis, since these items were not fielded on its survey.

³ A weighting variable was calculated by Gallup statisticians in multiple stages for each respondent to allow researchers to generate nationally representative estimates for each country. Base weights were calculated based on the probability of selection and post-stratification weights were calculated to adjust for non-response bias by gender, age, and, in some countries, education or socio-economic status.

⁴ Stringency measures are missing for several countries in our sample, including Guinea, North Macedonia, Montenegro, and Kosovo. We found no relationship between our harm index and whether stringency, deaths per capita, inequality, or flu growth variables were missing. The correlation between the harm index and whether these variables are missing is approximately zero.

Abaluck, J., Kwong, L. H., Styczynski, A., Haque, A., Kabir, M. A., Bates-Jefferys, E., ... & Mobarak, A. M. (2022). Impact of community masking on COVID-19: a cluster-randomized trial in Bangladesh. Science, 375(6577), eabi9069.
Aminjonov, Ulugbek, Olivier Bargain, and Tanguy Bernard. 2021. “Gimme Shelter. Social Distancing and Income Support in times of Pandemic.” IZA DP No. 14967.
Barkay, Neta, Curtiss Cobb, Roee Eilat, Tal Galili, Daniel Haimovich, Sarah LaRocca, Katherine Morris, Tal Sarig. 2020. "Weights and Methodology Brief for the COVID-19 Symptom Survey by University of Maryland and Carnegie Mellon University, in Partnership with Facebook Version 3-19, available https://covidmap.umd.edu/document/css_methods_brief.pdf
Boone, Laurence and Colombe Ladreit. 2021. "Fear of COVID and non-pharmaceutical interventions: an analysis of their economic impact among 29 advanced OECD countries" COVID Economics (73) March.
Bundervoet, T., Dávalos, M. E., & Garcia, N. 2021. The Short-Term Impacts of COVID-19 on Households in Developing Countries: AN overview based on a harmonized data set of high-frequency surveys. Policy Research Working Paper No. 9582, Washington DC: The World Bank.
Canes-Wrone, Brandice, Rothwell, Jonathan T., & Makridis, Christos. 2022. “Partisanship and Policy on an Emerging Issue: Mass and Elite Responses to COVID-19 as the Pandemic Evolved.” Working Paper, available at http://dx.doi.org/10.2139/ssrn.3638373
Carvalho, Vasco M., Juan R. Garcia, Stephen Hansen, Álvaro Ortiz, Tomasa Rodrigo, José V. Rodríguez Mora, and Pep Ruiz. "Tracking the COVID-19 crisis with high-resolution transaction data." Royal Society Open Science 8, no. 8 (2021): 210218.
Chernozhukov, Victor, Hiroyuki Kasahara, and Paul Schrimpf. 2021. “Causal Impact of Masks, Policies, Behavior on Early Covid-19 Pandemic in the U.S.” Journal of Econometrics, Pandemic Econometrics, 220 (1): 23–62. https://doi.org/10.1016/j.jeconom.2020.09.003.
Coibion, Olivier, Yuriy Gorodnichenko & Michael Weber. 2020. “The Cost of the Covid-19 Crisis: Lockdowns, Macroeconomic Expectations, and Consumer Spending.” NBER Working Paper 27141.
Deb, Pragyan, Davide Furceri, Jonathan D. Ostry, Nour Tawk. 2020. “The Economic Effects of COVID-19 Containment Measures.” IMF Working Paper WP/20/158.
Egger, Dennis, Edward Miguel, Shana S. Warren, Ashish Shenoy, Elliott Collins, Dean Karlan, Doug Parkerson, et al. 2021. “Falling Living Standards during the COVID-19 Crisis: Quantitative Evidence from Nine Developing Countries.” Science Advances 7 (6): eabe0997.
Fan, Junchuan, Yao Li, Kathleen Stewart, Anil R. Kommareddy, Andres Garcia, Jinyi Ma, Zheng Liu, Joe O’Brien, Adrianne Bradford, Xiaoyi Deng, Samantha Chiu, Frauke Kreuter, Neta Barkay, Alyssa Bilinski, Brian Kim, Tal Galili, Daniel Haimovich, Sarah LaRocca, Stanley Presser, Katherine Morris, Joshua A Salomon, Elizabeth A. Stuart, Ryan Tibshirani, Tali Alterman Barash, Curtiss Cobb, Andi Gros, Ahmed Isa, Alex Kaess, Faisal Karim, Roee Eliat, Ofir Eretz Kedosha, Shelly Matskel, Roee Melamed, Amey Patankar, Irit Rutenberg, Tal Salmona, David Vannette. 2020. COVID-19 World Symptom Survey Data API. https://covidmap.umd.edu/api.html
Forsythe, Eliza, Lisa B. Kahn, Fabian Lange, and David Wiczer. "Labor demand in the time of COVID-19: Evidence from vacancy postings and UI claims." Journal of public economics 189 (2020): 104238.
Fetzer, T., & Graeber, T. (2021). Measuring the scientific effectiveness of contact tracing: Evidence from a natural experiment. Proceedings of the National Academy of Sciences, 118(33), e2100814118.
Furceri, Davide, Prakash Loungani, Jonathan D. Ostry, and Pietro Pizzuto. "Will COVID-19 have long-lasting effects on inequality? Evidence from past pandemics." The Journal of Economic Inequality (2022): 1-29.
Google LLC. 2021. "Google COVID-19 Community Mobility Reports". https://www.google.com/covid19/mobility/ Accessed: March 31, 2021.
Goolsbee, A., & Syverson, C. (2020). Fear, lockdown, and diversion: Comparing drivers of pandemic economic decline 2020. Journal of public economics, 193, 104311.
Gupta, Sumedha, Kosali Simon, and Coady Wing. "Mandated and Voluntary Social Distancing during the COVID-19 Epidemic." Brookings Papers on Economic Activity (2020): 269-316.
Hale, Thomas, Noam Angrist , Rafael Goldszmidt, Beatriz Kira , Anna Petherick , Toby Phillips, Samuel Webster, Emily Cameron-Blake , Laura Hallas, Saptarshi Majumdar, and Helen Tatlow. “A global panel database of pandemic policies (Oxford COVID-19 Government Response Tracker).” Nat Hum Behav 5, 529-538 (2021).
Herby, J., Jonung, L. and Hanke, S., 2022. A literature review and meta-analysis of the effects of lockdowns on COVID-19 mortality. Studies in Applied Economics, (200).
Institute for Health Metrics and Evaluation. 2021. “Estimation of total mortality due to COVID-19”, http://www.healthdata.org/special-analysis/estimation-excess-mortality-due-covid-19-and-scalars-reported-covid-19-deaths
Josephson, A., Kilic, T. & Michler, J.D. Socioeconomic impacts of COVID-19 in low-income countries. Nat Hum Behav 5, 557–565 (2021).
Kim, Lydia Y.; Lugo, Maria Ana; Mason, Andrew D.; Uochi, Ikuko. 2021. Inequality under COVID-19 : Taking Stock of High-Frequency Data for East Asia and the Pacific. Policy Research Working Paper; No. 9859. World Bank, Washington, DC.
Kugler, Maurice, Mariana Viollaz, Daniel Duque, Isis Gaddis, David Locke Newhouse, Amparo Palacios-Lopez, and Michael Weber. 2021. “How Did the COVID-19 Crisis Affect Different Types of Workers in the Developing World?” IZA Discussion Paper No. 14519
Liu, Yang, Christian Morgenstern, James Kelly, Rachel Lowe, CMMID COVID-19 Working Group and Mark Jit. 2021. “The impact of non-pharmaceutical interventions on SARS-CoV-2 transmission across 130 countries and territories.” BMC Medicine 19.
Lozano Rojas, F., X. Jiang, L. Montenovo, K. I. Simon, B. Weinberg, and C. Wing (2020). Is the cure worse than the disease? Immediate labor market effects of covid-19 case rates and school closures in the US. Technical report, National Bureau of Economic Research.
Narayan, Ambar; Cojocaru, Alexandru; Agrawal, Sarthak; Bundervoet, Tom; Davalos, Maria; Garcia, Natalia; Lakner, Christoph; Mahler, Daniel Gerszon; Montalva Talledo, Veronica; Ten, Andrey; Yonzan, Nishant. 2022. COVID-19 and Economic Inequality: Short-Term Impacts with Long-Term Consequences. Policy Research Working Paper;No. 9902. World Bank, Washington, DC. © World Bank.
Neidhöfer, G., N. Lustig, and M. Tommasi. 2021. “Intergenerational Transmission of Lockdown Consequences: Prognosis of the Longer-Run Persistence of COVID-19 in Latin America.” Journal of Economic Inequality 19 (3): 571-98.
Nivette, A.E., Zahnow, R., Aguilar, R. et al. (2021). A global analysis of the impact of COVID-19 stay-at-home restrictions on crime. Nat Hum Behav 5, 868–877.
Ma, Lin; Shapira, Gil; de Walque, Damien; Do, Quy-Toan; Friedman, Jed; Levchenko, Andrei A.. 2021. The Intergenerational Mortality Tradeoff of COVID-19 Lockdown Policies. Policy Research Working Paper;No. 9677. World Bank, Washington, DC.
Mahler, D. G., N. Yonzan, C. Lakner, R. A. Castaneda Aguilar, and H. Wu. 2021. “Updated Estimates of the Impact of COVID-19 on Global Poverty: Turning the Corner on the Pandemic in 2021?” Data Blog (blog). June 24, 2021. https://blogs.worldbank.org/opendata/updated-estimates-impact-covid-19-global-poverty-turning-corner-pandemic-2021.
Milanovic, Branko. 2019. "Description of All The Ginis Dataset," https://stonecenter.gc.cuny.edu/files/2019/02/Milanovic-all-the-ginis-dataset-description.pdf
Qi, J., Zhang, D., Zhang, X. et al. Short- and medium-term impacts of strict anti-contagion policies on non-COVID-19 mortality in China. Nat Hum Behav 6, 55–63 (2022).
Rothwell, J., & Smith, E. (2021). Socioeconomic Status as a Risk Factor in Economic and Physical Harm from COVID-19: Evidence from the United States. The ANNALS of the American Academy of Political and Social Science, 698(1), 12-38.
Spiegel, Matthew, and Heather Tookes. 2022. "All or nothing? Partial business shutdowns and COVID-19 fatality growth." PloS one 17, no. 2.
Violato, C., Violato, E.M. and Violato, E.M., 2021. Impact of the stringency of lockdown measures on covid-19: A theoretical model of a pandemic. PloS one, 16(10), p.e0258205.
Wang, Haidong, Katherine R. Paulson, Spencer A. Pease, Stefanie Watson, Haley Comfort, Peng Zheng, Aleksandr Y. Aravkin et al. 2022. "Estimating excess mortality due to the COVID-19 pandemic: a systematic analysis of COVID-19-related mortality, 2020–21." The Lancet 399 10334: 1513-1536.
World Bank. 2022. Global Economic Prospects, January 2022. Washington, DC: World Bank. doi: 10.1596/978-1-4648-1758-8.

There is NO Competing Interest.

DBPRSupplement.docx
Supplementary Information - updated 23/2
SupplementalMaterialsNHB.docx
Global Evidence on the Economic Effects of Disease Suppression During COVID-19

Download PDF

Journal Publication

published 08 Jan, 2024

Read the published version in Humanities & Social Sciences Communications →

Version 1

posted

You are reading this latest preprint version

Global Evidence on the Economic Effects of Disease Suppression During COVID-19

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Results

Validating a novel measure of economic harm

Stringency measures and economic harm: main results

Sensitivity analysis of main results

Alternative policy responses and economic harm

Socio-economic Status And Heterogenous Effects Of Stringency: Individual-level Analysis

Discussion

Methods

Economic harm measures

Stringency of economic restrictions measures

Disease burden measures

Other variables

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1