Inequality in the Time of Quarantine: Why Cost Analysis of Covid-19 Quarantine Measures Should Incorporate Inequality

Maggie Baughman, guest contributor

Maggie Baughman

Princeton ‘21

The sudden outbreak of a global pandemic draws calculations of societal wellbeing into sharp focus, as governments across the world are pressed to weigh the costs and benefits of disease control measures under conditions of heightened risk and uncertainty. While economists have produced rough estimates of the cost of quarantine measures – stemming both from post-epidemic analyses of the SARS outbreak and preliminary estimates under the current COVID-19 crisis – they fail to account for the nuances of calculating social wellbeing in a highly stratified and socioeconomically unequal society. Combining literature on the unequal costs of epidemics for different socioeconomic groups with cost-benefit analyses of various quarantine measures, it is abundantly clear that the current models of the costs of quarantine fail to take into account the unequal costs placed on low-income and working-class individuals.

As the COVID-19 quarantine restrictions were first implemented in March of 2020, three economists, Martin Eichenbaum, Sergio Re elo, and Mathias Trabandt (ERT), took the step of translating quarantine cost functions into utility functions, integrating non-monetary aspects of utility into quarantine calculations. This model, published in a NBER working paper, was one of the first proposed during the Covid-19 crisis, combining traditional economic theory and established epidemiological modeling techniques. Using an SIRD model of disease transmission (where S is susceptible, I is infected, R is recovered, and D is deceased), they calculate the lifetime utility of an individual in each of these four classes based on their consumption, production, and likelihood to transition to a different state (for example, the likelihood of infection for a susceptible individual). To calculate the expected utility of different containment rates, they define social welfare, U0 , as a function of the weighted sum of these utilities: To maximize U0, the authors argue, containment rates must increase in parallel with disease transmission, minimizing the costs of the behavior of infected people as that number grows. This model was a step beyond the nationally popular predictions of fatalities or economic cost of recession, bringing together two fields to build a more nuanced, thoughtful model. 

The drawback of ERT’s  model, however, is that, in measuring an individual’s lifetime utility in terms of consumption and production, it inherently deprioritizes the lives of low-income individuals. The social welfare function defined above uses a utilitarian measure of social welfare (maximizing the total amount of welfare in society without considering distribution), and is therefore not inequality averse. Associating an individual’s utility solely with their consumption and production abilities prioritizes the lives of those who consume and produce more – higher income individuals. Similarly, the cost models developed above assume homogeneity in the population, and base their calculations of parameters on averages. They fail to account for the underlying factors of inequality that cause different members of the population to be more susceptible to disease, face greater medical costs, and earn less income. These gaps in existing models necessitate the development of a quarantine calculation that incorporates heterogeneity in the population, accounting specifically for socioeconomic inequality. 

INCORPORATING INEQUALITY INTO QUARANTINE COST MODELS:

In creating a fairer model for the economic impacts of quarantine measures, we should consider two factors. First, we should consider how the utility of quarantine changes at different socioeconomic levels, considering how model parameters are likely to differ. Second, in calculating social welfare under different quarantine measures, we should use an inequality averse social welfare function that accounts for the distribution of utility of quarantine across different income levels. 

Breaking down the value of quarantine into the individual cost of quarantine and the individual value of averted infection, we can consider how both factors are dependent on socioeconomic status. The individual cost of quarantine will not take into account the administrative costs but will consider lost wages. Here, we should consider conflicting wage possibilities for low-income workers: low-income individuals are more likely to be working “essential” jobs that are not covered by stay-at-home orders, making them less likely to face lost wages in that respect (though workers at essential jobs who are required to continue to work are more likely to be exposed to the virus, which would then force them to quarantine – demonstrating that working in an essential job is not a guarantee of continued wages), but are less likely to have paid sick leave or work from home alternatives. They are also more likely to be subject to the layoffs that have been hitting smaller businesses and sectors that are dominated by low income workers, as business owners lay off workers to remain afloat until government aid is released. Overall, we can estimate that low income workers are more likely to face lost wages due to universal quarantine than their wealthier counterparts, even though they are more likely to work jobs that are “essential.”

If we consider contact-based quarantine models (as used during the SARS epidemic), we see that low income workers are even more vulnerable due to their contact patterns. The 1918 infleunza epidemic demonstrated the increased vulnerability of low-income workers due to crowded living. This trend continues to this day, with income inversely related to crowding. Wage data indicates that the workers with the highest exposure risks are largely low-income – they make a median wage of less than 35,000 a year. Building from a University of Michigan model of lost wages as a consequence of contact rate, we can see that workers with higher contact rates are more likely to face isolation or quarantine as a result of interacting with infected individuals, and therefore suffer from lost wages. As a result, regardless of the type of quarantine implemented – contact or universal – low income workers face a higher likelihood of lost wages, and a higher cost of quarantine.

The value of averted infections is based on the total cost of SARS per person. This cost is broken down into the likelihood of hospitalization multiplied by the length of stay and the cost per day. As established above, low income individuals are more likely to come into contact with the virus. They are also more likely to have pre-existing health conditions, meaning that they have a higher probability of hospitalization should they contract the virus, and a higher likelihood of needing intensive care. Low income individuals are less likely to have healthcare coverage than their wealthier counterparts by a significant margin, making the per diem costs of healthcare for the individual higher. As a result of higher risk of complications, low income individuals also have a higher risk of death. This all suggests that the value of averting infection for lower income individuals is greater than for higher income individuals, since the costs of infection are higher. However, recalling the fact that lower-income workers are both more likely to work essential jobs that require continued work during the epidemic and higher risks of contracting the disease, and that they are more likely to live in crowded spaces with other individuals who are low-income (increasing the amount of contact both with co-residents and outside influences), the likelihood of averting the infection is smaller. While these factors have opposing influences on the value of averting infection, the clear takeaway is that low income individuals face different risks than higher income individuals, demonstrating the necessity of incorporating heterogeneous income levels into the model.

Calculating individual utility under quarantine using different parameters, rather than using averaged parameters across the population to generate a mean utility, allows for a more nuanced measure of social welfare. Rather than using Eichenbaum’s method of summing total utilities for each class (SIRD) of the population at different time increments, a utilitarian measure of social welfare that prioritizes total utility in the society, we have the necessary information to consider the distribution of welfare for different socioeconomic subsets of the population. A social welfare model of quarantine in a highly stratified society should inevitably incorporate “inequality aversion,” a parameter that penalizes systems for high degrees of welfare inequality. 

CONCLUSION:

Eichenbaum and his colleagues at the NBER made significant contributions to the conversation about the costs and benefits of quarantine, crossing disciplines to incorporate epidemiological modeling into economic predictions. However, the next step in these analyses is to incorporate social stratification, rather than treating the population as heterogeneous, facing the same risks and costs during the epidemic. 

Doctors, politicians, and reporters have made eloquent claims that the novel coronavirus does not discriminate – an epidemic cannot see social status. Indeed, the second wave of the 1918 influenza demonstrated just that – with high casualty rates of every social class. However, the costs of contracting a disease and the costs of disease control measures both vary significantly by income level. As public health officials make incredibly difficult decisions that affect millions of Americans, they turn to the reassurances of economic and epidemic calculations for guidance. The decision to incorporate inequality into these tools should not be a nuance to the academic conversation, but a vital piece of policy making practice. 

BIBLIOGRAPHY:

“158 Million Americans Told to Stay Home, but Trump Pledges to Keep It Short.” The New York Times, March 23, 2020, sec. World. https://www.nytimes.com/2020/03/23/world/coronavirus-updates-usa-world.html.

Breukers, Annemarie, Monique Mourits, Wopke van der Werf, and Alfons Oude Lansink. “Costs and Benefits of Controlling Quarantine Diseases: A Bio-Economic Modeling Approach.” Agricultural Economics 38, no. 2 (2008): 137–49. https://doi.org/10.1111/j.1574-0862.2008.00288.x.

“Coronavirus Could Cost the Global Economy $2.7 Trillion. Here’s How.” Bloomberg.Com. Accessed March 28, 2020. https://www.bloomberg.com/graphics/2020-coronavirus-pandemic-global-economic-risk/.

Day, Troy, Andrew Park, Neal Madras, Abba Gumel, and Jianhong Wu. “When Is Quarantine a Useful Control Strategy for Emerging Infectious Diseases?” American Journal of Epidemiology 163, no. 5 (March 1, 2006): 479–85. https://doi.org/10.1093/aje/kwj056.

Desilver, Drew. “As Coronavirus Spreads, Which U.S. Workers Have Paid Sick Leave – and Which Don’t?” Pew Research Center (blog), March 12, 2020. https://www.pewresearch.org/fact-tank/2020/03/12/as-coronavirus-spreads-which-u-s-workers-have-paid-sick-leave-and-which-dont/.

Eichenbaum, Martin S, Sergio Rebelo, and Mathias Trabandt. “The Macroeconomics of Epidemics.” Working Paper. National Bureau of Economic Research, March 2020. https://doi.org/10.3386/w26882.

Fisher, Max, and Emma Bubola. “As Coronavirus Deepens Inequality, Inequality Worsens Its Spread.” The New York Times, March 15, 2020, sec. World. https://www.nytimes.com/2020/03/15/world/europe/coronavirus-inequality.html.

FORSYTHE, KENNETH W. “Evaluating Economic Consequences of Livestock Diseases:: A US Perspective.” In The Economics of Quarantine and the SPS Agreement, edited by Kym Anderson, Cheryl McRae, and David Wilson, 265–84. University of Adelaide Press, 2001. https://www.jstor.org/stable/10.20851/j.ctt1t304rx.20.

“Is Quarantine Worth What It Costs?” The New York Times, March 17, 1854, sec. Archives. https://www.nytimes.com/1854/03/17/archives/is-quarantine-worth-what-it-costs.html.

“Is Quarantine Worth What It Costs?” Accessed March 28, 2020. http://timesmachine.nytimes.com/timesmachine/1854/03/17/75366060.html.

Jin, Beatrice, and Andrew McGill. “Who Is Most at Risk in the Coronavirus Crisis: 24 Million of the Lowest-Income Workers.” Politico, March 21, 2020. https://politico.com/interactives/2020/coronavirus-impact-on-low-income-jobs-by-occupation-chart/.

Jones, Sarah. “There’s No Such Thing As Unskilled Labor.” Intelligencer, March 21, 2020. https://nymag.com/intelligencer/2020/03/low-wage-work-is-essential-work.html.

Kaplow, Louis. “Concavity of Utility, Concavity of Welfare, and Redistribution of Income.” Working Paper. National Bureau of Economic Research, September 2003. https://doi.org/10.3386/w10005.

Levin Crimmel, Beth. “Health Insurance Coverage and Income Levels for the U.S. Noninstitutionalized Population under Age 65, 2001.” Medical Expenditure Panel Survey. Agency for Healthcare Research and Quality, May 2004. https://meps.ahrq.gov/data_files/publications/st40/stat40.pdf.

Lewis, Maureen. “The Economics of Epidemics.” Georgetown Journal of International Affairs 2, no. 2 (2001): 25–31.

Mubayi, Anuj, Christopher Kribs Zaleta, Maia Martcheva, and Carlos Castillo-Chávez. “A Cost-Based Comparison of Quarantine Strategies for New Emerging Diseases.” Mathematical Biosciences and Engineering: MBE 7, no. 3 (July 2010): 687–717. https://doi.org/10.3934/mbe.2010.7.687.

“Paid-Sick-Days-Good-for-Business-and-Workers.Pdf.” Accessed March 26, 2020. https://www.nationalpartnership.org/our-work/resources/economic-justice/paid-sick-days/paid-sick-days-good-for-business-and-workers.pdf.

Patrice, Gaines. “NAACP’s Virtual Town Hall Addresses Impact of Coronavirus on Communities of Color.” NBC News, March 17, 2020. https://www.nbcnews.com/news/nbcblk/naacp-s-virtual-town-hall-confronts-impact-coronavirus-people-color-n1160861.

Porter, Eduardo, and Jim Tankersley. “Shutdown Spotlights Economic Cost of Saving Lives.” The New York Times, March 24, 2020, sec. Business. https://www.nytimes.com/2020/03/24/business/economy/coronavirus-economy.html.

Solari, Claudia D., and Robert D. Mare. “Housing Crowding Effects on Children’s Wellbeing.” Social Science Research 41, no. 2 (March 2012): 464–76. https://doi.org/10.1016/j.ssresearch.2011.09.012.

Thompson, Derek. “The Coronavirus Will Be a Catastrophe for the Poor.” The Atlantic, March 20, 2020. https://www.theatlantic.com/ideas/archive/2020/03/coronavirus-will-supercharge-american-inequality/608419/.

Thomson-DeVeaux, Amelia. “Which Workers Are Most Vulnerable To The Economic Costs Of The Coronavirus?” FiveThirtyEight (blog), March 18, 2020. https://fivethirtyeight.com/features/which-workers-are-most-vulnerable-to-the-economic-costs-of-the-coronavirus/.

Torry, Harriet. “With Unmandated Sick Leave, U.S. Economy Is More Vulnerable to Pandemics.” Wall Street Journal, March 22, 2020, sec. Economy. https://www.wsj.com/articles/with-unmandated-sick-leave-u-s-economy-is-more-vulnerable-to-pandemics-11584885600.

Reason.com. “We Will Regret Not Taking the Economic Effects of Mass Quarantine More Seriously,” March 23, 2020. https://reason.com/2020/03/23/we-will-regret-not-taking-the-economic-effects-of-mass-quarantine-more-seriously/.