In the initial weeks of the COVID-19 pandemic, employment in several industries was especially vulnerable because of shutdown policies imposed by states, as well as a drop in demand as people engaged in social distancing. As a general rule, these were industries that were not deemed essential and that provide goods and services requiring considerable interaction between workers and customers. A recent article by Matthew Dey and Mark A. Loewenstein, published in the April 2020 Monthly Labor Review,1 using a taxonomy developed by Joseph Vavra to identify vulnerable industries, provides estimates of the number of jobs and the wages paid in these vulnerable industries of the economy.2 A key finding of that article is that, in 2019, about 20 percent of all employees worked in these highly exposed industries. Furthermore, occupations with lower wages are more common in the highly exposed sector than elsewhere in the economy. (Throughout this article, we characterize the economy as being made up of two sectors, the highly exposed sector and the not highly exposed sector.)
The effects of the pandemic have now become widespread, and employment losses have not been confined to businesses in industries that were forced to shut down. Furthermore, some localities and states have now begun to lift stay-at-home orders and businesses in vulnerable industries have begun to reopen. However, a number of unknowns exist: Will customers return when industries that have been shut down are reopened? Will workers return? Will further breakouts occur that result in industries again being shut down? In the state of these uncertainties, industries initially identified as vulnerable may continue to face difficulties.
This article expands on the earlier analysis by looking at the demographic characteristics of workers in the highly exposed industries. We also look at family earnings and other family characteristics. In the final section of this article, we briefly examine Current Population Survey (CPS) April 2020 estimates to gauge how employment in the highly exposed sector and elsewhere was affected at the start of the pandemic. The estimates indicate that, between February and April, employment losses have been especially severe in the highly exposed industries.
Note: Analysis on more recent CPS estimates is available at https://www.bls.gov/ers/update-on-demographics-earnings-and-family-characteristics-of-workers-in-sectors-initially-affected-by-covid-19-shutdowns.htm.
The Dey and Loewenstein’s April 2020 article uses establishment data from the U.S. Bureau of Labor Statistics (BLS) Quarterly Census of Employment and Wages and Occupational Employment Statistics (OES) programs. In this current article, we use household data from the CPS. Conducted by the U.S. Census Bureau for the BLS, the CPS is a monthly survey of approximately 60,000 households. The CPS provides a comprehensive body of data on the labor force status of individuals (employed, unemployed, or not in the labor force), hours of work, and other demographic and labor force characteristics. In addition, for one-fourth of the monthly sample, information about the earnings of wage and salary workers is collected.3 The estimates in this article are obtained by averaging the data across months in 2019, with use of only a quarter of the sample for which earnings information was collected.
The current article uses the same industry classification scheme as that used in the initial article.4 (The list of census industries used for this analysis can be found in the appendix, table A-1.) Industries in the highly exposed sector include “Restaurants and Bars, Travel and Transportation, Entertainment (e.g., casinos and amusement parks), Personal Services (e.g., dentists, daycare providers, barbers), other sensitive Retail (e.g., department stores and car dealers), and sensitive Manufacturing (e.g., aircraft and car manufacturing).”5 Using CPS data, we find 27.5 million workers, or 19.4 percent of U.S. employment in 2019 (similar to the OES estimates), were in these highly exposed industries.
Employment estimates for various demographic groups are presented in table 1 and in tables A-2 and A-3 of the appendix. The figures and the ensuing discussion in this section are based on the estimates in these tables.
Figure 1 depicts the fraction of workers by race, gender, and Hispanic ethnicity in the highly exposed industries and in the rest of the economy. One sees that the racial composition of the two sectors is quite similar, with a slight overrepresentation of minorities in the highly exposed sector. One also sees that the gender composition between the two sectors is similar, with a slight overrepresentation of women. (However, tables A-2 and A-3 show a gender imbalance among younger workers. In the highly exposed sector, 40 percent of those ages 16 to 24 are women and only 35 percent are men.) Other demographic differences are more pronounced. As figure 1 shows, Hispanics are overrepresented in the highly exposed sector. Twenty-three percent of Hispanic workers are employed in the highly exposed sector. The corresponding estimate for non-Hispanics is 18.6 percent.
Age, marital status, and education attainment differences of workers between the two sectors are even starker. Figure 2 shows the age composition of workers in the highly exposed industries and the rest of the economy. Note that workers under the age of 25 make up 25.9 percent of employment in the highly exposed industries and only 10.3 percent of employment in the remaining industries. This, in turn, implies that 37.9 percent of workers under age 25 are in the highly exposed sector even though this sector accounts for a little less than 20 percent of overall employment.
Marital status estimates for workers age 25 and older appear in figure 3. A disproportionate share of workers in the highly exposed sector is never married. Workers age 25 and older make up 20.7 percent of never-married workers employed in the highly exposed sector, while the estimate for married workers is 14.6 percent.6
The highly exposed sector is also disproportionately composed of workers with lower educational attainment. As shown in figure 4, at lower levels of education, the share of workers age 25 and older in the highly exposed sector exceeds the share in the other sector, whereas the pattern is reversed at higher levels of education. Thus, as can be seen from table A-3, the higher the level of education, the lower the share of workers in the highly exposed sector. This share is 24.3 percent for workers age 25 and older without a high school degree and 22.4 percent for workers age 25 and older with just a high school degree. The share drops to 19.1 percent for workers with some college or an associate’s degree, 12.7 percent for workers with a bachelor’s degree, and 6.2 percent for workers with an advanced degree.
As illustrated in figures 5 and 6, the highly exposed industries have more hourly workers and part-time workers. While 71.7 percent of workers are paid hourly in the highly exposed industries, 54.9 percent of workers are paid hourly in the remaining industries. Approximately one-third of workers in the highly exposed sector usually worked part-time hours—less than 35 hours per week. In the rest of the economy, this estimate is 18.5 percent.
Wages are considerably lower in the highly exposed sector than elsewhere.7 As noted earlier, the highly exposed sector has more part-time workers than the other sector, and as can be seen in table 1, part-time workers earn less than full-time workers do.8 In addition, as shown in figure 7, the wages of part-time and full-time workers in the highly exposed sector are both lower than the wages of workers with similar working arrangements elsewhere. The median hourly wage of part-time workers in the highly exposed sector is $11.80, compared with $15.00 elsewhere. The median wage of full-time workers is $17.00 in the highly exposed sector, compared with $23.00 for the other sector.
The lower wages in the highly exposed sector translate into lower earnings. Median usual weekly earnings of part-time workers in the highly exposed sector are $260.00, compared with $386.00 elsewhere. Median usual weekly earnings of full-time workers are $700.00 in the highly exposed sector and $961.53 in the rest of the economy. Of course, averaged over all workers, earnings in the highly exposed sector are also lower because of the much higher proportion of part-time workers.
From table 1, one sees that approximately three-fourths of workers in the highly exposed sector live with other family members.9 (The proportion is the same for workers in the other sector.) Within these families, workers both contribute to overall family earnings and are able to receive support from other family members who also work. Looking at the composition of these families and the proportions of family earnings that come from the highly exposed sector yields insights into how vulnerable workers may be to possible shutdowns in response to the COVID-19 pandemic.
Workers in the highly exposed sector disproportionately come from single-parent families. From table 1, one sees that approximately 26.0 percent of workers from single-parent families are employed in the highly exposed sector. In contrast, about 18.0 percent of workers from married families with children, 19.0 percent of workers in families with no children, and 20.0 percent of workers living alone or with nonrelatives are employed in the highly exposed sector. Figure 8 shows the percentage of workers in the highly exposed and not highly exposed sectors who are living alone or with nonrelatives and the type of family they live in if they are in a family. Note that approximately 25.0 percent of workers in the highly exposed sector do not live in a family and another 11.4 percent live in a single-parent family. Taken together, these percentages suggest that if workers were unable to work in the highly exposed sector, as many as 36.0 percent of them would be unable to draw on earnings from other family members in the household.
For workers who do live in families, the highly exposed sector disproportionately employs workers from families with low earnings.10 Figure 9 shows in each family earnings quintile the proportion of workers who are employed in the highly exposed and not highly exposed sectors. From the numbers in table 1, one sees that 27.6 percent of workers whose family earnings (not adjusted for the size of a person’s family) are in the bottom quintile are employed in the highly exposed sector. For workers whose family earnings are in the second quintile, this estimate is 22.2 percent, and it is 19.9 percent for workers whose family income is in the third quintile. The percentage falls to 17.1 percent and 14.7 percent for workers whose family earnings are in the fourth and top quintiles. The finding that the highly exposed sector disproportionately employs workers from families with low earnings further illustrates that even workers in the highly exposed sector who live in families may only be able to obtain minimal financial support from other family members should they lose their jobs.
The analysis in the previous section has been conducted from the perspective of the worker and of how much support workers who lose their jobs may expect to receive from other family members in their household. A related question concerns the financial support that workers employed in the highly exposed sector provide to their families. For families with at least one employed family member, the estimates in table 2 show that a little more than 26 percent have at least one worker who is employed in the highly exposed sector and about half of these families have children. For one to assess the vulnerability of these families for each quintile of the family earnings distribution, table 3 shows the percentage of family earnings that stem from employment in the highly exposed sector.
Overall, the estimates in table 3 indicate that 10.9 percent of families receive 100.0 percent of their earnings from workers in the highly exposed sector. The estimates also suggest that families with the lowest earnings depend heavily on employment in the highly exposed sector. To illustrate, figure 10 shows the percentage of families in each earnings quintile that receive all of their earnings from the highly exposed sector. Almost 46.0 percent of families in the bottom quintile receive all of their earnings from the highly exposed sector. For families in the second quintile, this percentage is 24.2 percent. The percentage of families in the middle quintile is 14.7 percent. This percentage drops to 8.8 percent and 6.5 percent for families in the fourth and top quintiles, respectively.
Families with children are similarly vulnerable as families overall. Of families with children, 11.4 percent had 100.0 percent of their earnings coming from workers in the highly exposed sector.11 In many instances, children in families in which 100.0 percent of the earnings are from the highly exposed sector live in single-parent households. For each family type (married families with children, single-parent families with children, and families with no children), table 2 shows the percentage of family earnings that stem from employment in the highly exposed sector. The data in the table show that single-parent families are especially vulnerable to shutdowns in the highly exposed sector. Of these families, 19.0 percent obtain 100.0 percent of their family’s earnings from workers in the highly exposed sector. In contrast, 8.3 percent of married families with children and 10.5 percent of married families with no children receive all of their earnings from the highly exposed sector.12 Figure 11 shows the breakdown of full-exposure cases by family type—47.5 percent are families with no children, 25.3 percent are single-parent families, and 27.3 percent are married families with children.
The recently released CPS April 2020 estimates, which are summarized in table 4, confirm the vulnerability of workers in the highly exposed sector.13 In April, the economic shocks due to the pandemic were clearly not confined to the highly exposed sector but were felt throughout the entire economy. Employment disruptions were widespread throughout the entire labor market. Between February and April, overall employment (not seasonally adjusted) fell by 15.6 percent. However, the reduction in employment was especially severe in what we have identified as the initially highly exposed sector. The CPS estimates indicate that, between February and April, employment in the highly exposed sector fell by 38.2 percent, compared with 10.5 percent elsewhere. The same is true of the unemployment rate. Overall, the unemployment rate (not seasonally adjusted) increased from 3.8 percent to 14.4 percent between February and April. However, the unemployment rate in the highly exposed sector increased from 4.5 percent to 34.1 percent. Elsewhere, the unemployment rate increased from 3.6 percent to 10.3 percent.
The employment disruptions during the first few months of the pandemic have been spread unevenly among the demographic groups. A thorough analysis of how the pandemic affected all the various demographic groups is beyond the scope of this article. Here, we simply highlight how several groups have fared. To this point in time, women have been more affected than men have. Throughout the economy as a whole, female employment has fallen by 17.9 percent, whereas male employment has dropped by 13.5 percent (see table 4). Female employment in the highly exposed sector fell by a whopping 43.3 percent. Hispanics also suffered severe losses in employment. Hispanic employment fell by 20.1 percent overall and by 42.2 percent in the highly exposed sector. Young workers are another group that experienced a large fall in employment. Employment of workers ages 16–24 fell by 31.3 percent overall and by 48.1 percent in the more highly exposed sector. Less educated workers are another group suffering a larger-than-average fall in employment. Employment of individuals 25 and older with less than a high school diploma fell by 20.7 percent and those with a high school diploma, but no college, declined by 20.8 percent overall and by 35.1 percent and 40.4 percent, respectively, in the highly exposed sector.
Finally, looking at the experience of the various family types, one sees from table 4 that employment of workers in single-parent families fell by 24.3 percent in the economy as a whole. In the highly exposed sector, employment declined by 47.4 percent. Employment of individuals who are not living with a family member fell by 19.3 percent in the economy as a whole and by 42.8 percent in the highly exposed sector.
To combat the COVID-19 pandemic, industries that are most prone to being shut down are disproportionately composed of workers who are younger, are unmarried, and have less education. Workers in the highly exposed sector are more likely to be in part-time jobs and generally have lower wages and total earnings than do workers in other parts of the economy.
Workers in the highly exposed sector disproportionately belong to single-parent families or do not live in a family. When workers in the highly exposed sector do live with other family members, family earnings are often toward the bottom of the earnings distribution.
Families with workers in the highly exposed sector are particularly vulnerable to industry shutdowns. A substantial percentage of families receive all of their earnings from family members working in the highly exposed sector. This percentage is particularly high for families whose earnings are at the bottom end of the earnings distribution. Almost 46 percent of families in the bottom quintile of the earnings distribution receive all of their earnings from the highly exposed sector. For families in the second quintile, this number is 24.2 percent. Families with children are similarly vulnerable to shutdowns as are families overall, but single-parent families with children are especially vulnerable—about 19 percent of single-parent families obtain 100 percent of their family’s earnings from workers in the highly exposed sector.
The recently released CPS April 2020 estimates confirm the vulnerability of workers in the highly exposed sector. Employment losses were widespread but were especially severe in the highly exposed sector. Furthermore, the reductions in employment were spread unevenly among various demographic groups. In some demographic groups, employment decreased substantially overall and especially sharply in the more highly exposed sector. Particularly hard-hit groups include, Hispanics, younger workers, and workers with less education level. Workers who are not family members and workers in single-parent families also experienced a large fall in employment and an increase in unemployment.
Matthew Dey, Mark A. Loewenstein, David S. Piccone Jr, and Anne E. Polivka, "Demographics, earnings, and family characteristics of workers in sectors initially affected by COVID-19 shutdowns," Monthly Labor Review, U.S. Bureau of Labor Statistics, June 2020, https://doi.org/10.21916/mlr.2020.11
1 Matthew Dey and Mark A. Loewenstein, “How many workers are employed in sectors directly affected by COVID-19 shutdowns, where do they work, and how much do they earn?” Monthly Labor Review, April 2020, https://www.bls.gov/opub/mlr/2020/article/covid-19-shutdowns.htm.
2 Joseph S. Vavra, “Shutdown sectors represent large share of all U.S. employment” (Chicago, IL: Becker Friedman Institute for Economics at the University of Chicago, March 31, 2020), https://bfi.uchicago.edu/insight/blog/key-economic-facts-about-covid-19/.
3 Earnings information is not collected for the self-employed.
4 While the Quarterly Census of Employment and Wages and Occupational Employment Statistics (OES) surveys use the North American Industry Classification System (NAICS) industry classification, the Current Population Survey (CPS) uses census industry codes. A few situations exist in which differences between the NAICS and census industry definitions made it necessary to add or subtract industries identified as exposed in our initial article. As noted in the text, the list of Census industries used for this analysis can be found in the appendix, table A-1.
5 Vavra, “Shutdown sectors represent large share of all U.S. employment.” As noted in the initial article, one could quibble about certain industries, but we are convinced that reasonable modifications to the list are likely to have relatively minor effects on our overall findings.
6 These particular numbers do not appear in figure 3. Figure 3 depicts the demographic composition of the highly exposed and not highly exposed sectors. The numbers in the text refer to the sectoral composition of workers in a particular demographic category.
7 In our earlier article, we looked at occupational wages by using the OES and showed that lower paying occupations are more heavily represented in the exposed sector. In this article, using the CPS data, we look at the wages of individual workers by using the CPS data.
8 To obtain a more comprehensive picture, we have constructed an hourly wage for all workers (BLS only does this for hourly workers). We also have chosen to calculate usual median weekly earnings for part-time workers and full-time workers. BLS press releases report usual weekly earnings for only full-time workers.
9 For this analysis, families are defined as two or more individuals living together who are related by birth, marriage, or adoption. All related individuals in a household are considered one family, even if several generations of a family are living together (for example, a divorced mother living with her adult son and his infant daughter would be one family). The estimates for families with children refer to children under the age of 18. Individuals who are living alone or in a household with completely unrelated people (for example, a group of unrelated people sharing a house) are classified as not living with a family member. Unmarried partners and same-sex spouses also are not classified as living with a family member.
10 Earnings are what individuals receive from being employed. Although typically the largest component, earnings are just one component of family income. Other potential sources of family income include payments from government programs such as social security and public assistance programs, rental income, and dividend payments.
11 Children are those age 18 and under who are sons, daughters, stepchildren, or adopted children living in the household. Nieces, nephews, grandchildren, other related and unrelated children, and children not living in the household are not included as children.
12 However, the exposure of single-parent families does not vary much by the number of children in the household. For single-parent households with one child, 19.5 percent of families obtain all the family’s earnings from workers in the highly exposed sector, while 18.5 percent of single-parent families with two or more children do.
13 The definition of the exposed sector is admittedly subjective. With current CPS data, we can identify industries that are potentially misclassified. To this end, we deem an industry in the exposed sector as potentially misclassified if employment grew or decreased less than 15 percent from February to April. In addition, we deem an industry in the not highly exposed sector as potentially misclassified if employment decreased by 30 or more percent from February to April. Given these definitions, we find that only 6.2 percent of highly exposed sector employment and 5.5 percent of not highly exposed sector employment are potentially in misclassified industries.