Abstract
This study uses 2018 national survey data to characterize the proportion of essential workers in the US overall and living in economically vulnerable households, defined as income <$40 000 or at least 1 member uninsured or older than age 65.
The label of “essential worker” reflects society’s needs but does not mean that society has compensated those workers for additional risks incurred on the job during the current pandemic. When an essential worker contracts severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), they pose a risk to the other members of their household. These members may be elderly or lack health insurance, and the household may have limited resources to care for a sick family member.1,2 We assessed the proportion of essential workers in the US population and described the economic vulnerability of their households.
Methods
We analyzed data from the Public Use Microdata Sample of the 2018 American Community Survey (ACS). The US Census Bureau chooses a random sample of US addresses each month, contacts selected households by mail, and follows up by phone and personal visit to address nonresponse.3 The Department of Homeland Security’s Cybersecurity and Infrastructure Security Agency (CISA) released an “Essential Critical Infrastructure Workforce” advisory list of occupations necessary to “continuity of functions critical to public health and safety” in March 2020 and updated that list in April 2020.4 We matched industry and subindustry codes in the ACS to the 6-digit Standard Occupation Codes (indexed by the US Bureau of Labor Statistics) in the CISA advisory list to identify essential workers.5 We assessed the proportion and demographic characteristics (age, sex, and race as given in response to a multiple-choice question) of essential workers by industry.
We defined 3 risk categories to estimate the economic vulnerability of households: (1) low household income (below $40 000); (2) uninsurance (at least 1 person in the household is uninsured); and (3) household presence of 1 or more persons aged 65 years or older. We categorized households facing at least 2 of these 3 risks to be high risk and examined the prevalence of risk in households with essential workers by industry. The data were weighted to be nationally representative. We used Stata version 16.0 (StataCorp) for all analyses. The Harvard University institutional review board does not require formal review for analysis of public use data.
Results
A total of 3 214 539 individuals responded to the ACS in 2018 (response rate, 92.0%)6; we estimated that 101 188 748 of them were essential workers. As shown in Table 1, essential workers were estimated to comprise 40% of the US adult population. Of essential workers, 46% were female, 14% were black, 17% were Hispanic, 11% were uninsured, and 8% were aged 65 years or older. Eight of the 21 industry categories from the ACS accounted for 73% of essential workers. Health care accounted for a larger proportion (15%) of essential workers than any other industry: 65% of health care workers held essential jobs. Black individuals were overrepresented in several essential industries, notably transportation (23%), public administration (18%), and health care (18%).
Table 1. Essential Worker Characteristics (n = 101 188 748)a.
| Essential workers by industry | Unweighted No. (N = 2 562 591) | Weighted % | |||||||
|---|---|---|---|---|---|---|---|---|---|
| All adults who are essential workers in industry | Essential employment in industry | Industry employment that is essential | Female | Black | Hispanic | Uninsured | Aged ≥65 y | ||
| Health care | 144 578 | 5.9 | 14.6 | 64.7 | 82.6 | 17.8 | 12.7 | 6.5 | 8.4 |
| General hospital | 66 666 | 2.7 | 6.6 | 69.6 | 80.4 | 15.9 | 10.7 | 3.5 | 8.1 |
| Nursing homes | 17 850 | 0.7 | 1.8 | 77.5 | 86.3 | 27.8 | 9.8 | 10 | 9.1 |
| Physician offices | 14 555 | 0.6 | 1.4 | 56 | 89 | 10 | 14.4 | 6.2 | 8.5 |
| Outpatient centers | 14 550 | 0.6 | 1.5 | 61.2 | 80.7 | 14 | 14.9 | 5.8 | 7.9 |
| Home health | 11 520 | 0.5 | 1.2 | 66.2 | 88.8 | 29.7 | 17.5 | 14.8 | 11.3 |
| Other | 19 437 | 0.8 | 2.0 | 51.4 | 79.1 | 14.3 | 16.4 | 8.3 | 6.9 |
| Retail | 119 321 | 5.0 | 12.5 | 48.3 | 46.5 | 13.2 | 16.6 | 11.9 | 7.6 |
| Grocery | 26 768 | 1.1 | 2.8 | 56.6 | 43.6 | 12.8 | 17.3 | 13 | 7 |
| Pharmacy | 9068 | 0.4 | 0.9 | 74.9 | 67.8 | 12.1 | 12.5 | 6.5 | 7.1 |
| Other | 83 485 | 3.5 | 8.8 | 44.7 | 45 | 13.4 | 16.9 | 12.2 | 7.8 |
| Manufacturing | 127 275 | 5.0 | 12.4 | 49.9 | 30.3 | 11.1 | 16.4 | 9 | 8.2 |
| Construction | 95 253 | 4.0 | 10.0 | 67.8 | 7.7 | 6.9 | 26.7 | 24 | 6.8 |
| Accommodation and food | 66 266 | 3.0 | 7.5 | 41.7 | 53.3 | 15.5 | 23.9 | 20 | 4.1 |
| Restaurant | 56 034 | 2.5 | 6.3 | 41.4 | 50.5 | 14.7 | 22.4 | 20 | 3.5 |
| Other | 10 232 | 0.5 | 1.2 | 43 | 67.1 | 19.5 | 31.2 | 19.8 | 6.8 |
| Public administration | 58 397 | 2.2 | 5.5 | 53 | 44.2 | 18.3 | 11.7 | 3.1 | 10.1 |
| Scientific and technical | 59 067 | 2.3 | 5.8 | 39.7 | 39.6 | 7.4 | 9.1 | 5.1 | 8 |
| Transportation | 51 694 | 2.2 | 5.4 | 46.4 | 32.1 | 22.8 | 18.9 | 11.2 | 7.9 |
| Other | 265 854 | 10.5 | 26.2 | 16.5 | 49.9 | 13.4 | 17.8 | 10.6 | 9.2 |
| All essential adults | 979 918 | 39.9 | 100.0 | 67.1 | 46.4 | 13.6 | 17.2 | 11.3 | 8.1 |
Abbreviation: ACS, American Community Survey.
Weighted estimate of number of essential adults = 101 188 748. Statistics drawn from the 2018 ACS (unique respondents = 3 214 539). Each column contains the relevant percentage for each subgroup. For all analyses, data are reweighted to be representative of the US. Essential workers are defined using 6-digit occupation codes as compiled by the Labor Market Information Institute. For 20% of adults in the ACS, 5-digit, instead of 6-digit, occupation code was reported. For the first 2 columns of the table, we assume that within an industry, the share of essential workers among 5-digit reporters is the same as the share among 6-digit reporters within an industry. We ran 3 separate regressions of essential worker status (A), and nonessential worker status (B) and 5-digit reporting status (C) on a constant. We then use the Stata packages RSUEST and NLCOM to obtain point estimates and standard errors of the share of essential workers within an industry: Number of Essential Workers = A + C*[A/(A+B)]. For the last 4 columns that focus on characteristics within industries, individuals who report 5-digit occupation codes are ignored. Because of the large size of the sample, none of the 95% CIs were more than 2 percentage points on either side of the point estimate.
A total of 1 410 976 households were represented in the 2018 ACS with an estimated 51% of those households including an essential worker. Table 2 summarizes household characteristics of essential workers: 25% of essential workers were estimated to have low household income, 18% to live in a household with at least 1 uninsured person, and 18% to live with someone aged 65 years or older. We estimated that 48% of essential workers lived in a household with at least 1 risk and 13% of essential workers lived in high-risk households.
Table 2. Household Characteristics (n = 1 410 976)a.
| Essential workers by industry | Weighted % | ||||
|---|---|---|---|---|---|
| Income <$40 000 | Uninsured in household | Aged ≥65 y in household | Risk | ||
| High | Any | ||||
| Health care | 22.3 | 13.4 | 19.1 | 9.9 | 44.3 |
| General hospital | 16 | 9.1 | 18.4 | 6.5 | 36.8 |
| Nursing homes | 34 | 19.1 | 20.4 | 15.6 | 57.1 |
| Physician offices | 19.9 | 13.1 | 18.7 | 8.4 | 43 |
| Outpatient centers | 19 | 12.7 | 17.3 | 7.7 | 41 |
| Home health | 41.2 | 23.2 | 24.2 | 21.4 | 65.7 |
| Other | 22.3 | 15.9 | 17.7 | 10 | 45.5 |
| Retail | 30.1 | 20.2 | 19.2 | 14.6 | 54 |
| Grocery | 32.3 | 21.6 | 20.3 | 15.8 | 57.2 |
| Pharmacy | 25.6 | 13.2 | 17.9 | 9.8 | 46.4 |
| Other | 29.9 | 20.6 | 19.1 | 14.7 | 53.9 |
| Manufacturing | 22.1 | 15.9 | 16.9 | 10.9 | 43.4 |
| Construction | 28 | 31.4 | 15.6 | 18.5 | 55.3 |
| Accommodation and food | 38.6 | 29.8 | 14.7 | 20.1 | 61.7 |
| Restaurant | 37.8 | 29.5 | 13.8 | 19.4 | 60.5 |
| Other | 42.7 | 31.2 | 18.8 | 23.5 | 67.5 |
| Public administration | 14 | 8.3 | 19 | 6 | 35 |
| Scientific and technical | 13.1 | 9.3 | 15.6 | 5.9 | 31.8 |
| Transportation | 25.8 | 19.2 | 19.4 | 13 | 50.6 |
| Other | 25 | 17 | 18.6 | 12.4 | 47.4 |
| All essential adults | 24.8 | 18.2 | 17.8 | 12.6 | 47.6 |
Abbreviation: ACS, American Community Survey.
Statistics drawn from the 2018 ACS (unique respondents = 3 214 539; unique households = 1 410 976). Households were flagged as containing an essential worker if it housed at least 1 individual who was essential according to their 6-digit classification and the Labor Market Information Institute data. All estimates have been weighted to be representative of the US. Each column contains the mean for each subgroup. Because of the large size of the sample, none of the 95% CIs were more than 2 percentage points on either side of the point estimate.
Discussion
Using 2018 data, an estimated 40% of US adults would be categorized as essential workers during the current pandemic, with 13% living in high-risk households. One limitation of the study is that the definition of high risk was developed for this study and may not match other studies or comprehensively capture true risk. In addition, some jobs expose essential workers to greater risk than others. Policies to support workers who are putting themselves and their families at risk to benefit others should be considered, and there is urgent need for economic analyses of the costs and benefits of approaches such as stimulus payments, short-term health care coverage, and prioritized testing.
Section Editor: Jody W. Zylke, MD, Deputy Editor.
References
- 1.Blau F, Koebe J, Meyerhofer P Essential and frontline workers in the COVID-19 crisis. Econofact Report. April 30, 2020. Accessed May 16, 2020. https://econofact.org/essential-and-frontline-workers-in-the-covid-19-crisis
- 2.COVID-19 and workers at risk: examining the long-term care workforce. Kaiser Family Foundation Issue Brief. April 23, 2020.
- 3.American Community Survey Design and Methodology Chapter 7: data collection and capture for housing units. January 2014. Accessed May 16, 2020. https://www2.census.gov/programs-surveys/acs/methodology/design_and_methodology/acs_design_methodology_ch07_2014.pdf
- 4.US Department of Homeland Security, Cybersecurity & Infrastructure Security Agency Advisory memorandum on identification of essential critical infrastructure workers during COVID-19 response. April 17, 2020. https://www.cisa.gov/sites/default/files/publications/Version_3.0_CISA_Guidance_on_Essential_Critical_Infrastructure_Workers_1.pdf
- 5.LMI Institute SOC codes for CISA critical infrastructure workers. Rev 1. Accessed April 28, 2020. https://www.lmiontheweb.org/more-than-half-of-u-s-workers-in-critical-occupations-in-the-fight-against-covid-19
- 6.American Community Survey response rates. Accessed May 16, 2020. https://www.census.gov/acs/www/methodology/sample-size-and-data-quality/response-rates/index.php