An analysis by the Center for Immigration Studies of Bureau of Labor Statistics data for May 2021 shows that while the official unemployment rate (share actively looking for work) of both the native-born and immigrants has fallen significantly, it remains higher than before Covid. Moreover, the labor force participation rate — the share of working-age (16-64) people with a job or looking for one — remains lower than in May 2019. Both immigrants and natives without a college degree have fared much worse in terms of unemployment and labor force participation. Immigrants (legal and illegal together) are often referred to as the "foreign-born".
Among the findings:
- The unemployment rate for native-born Americans (ages 16-plus) was 5.5 percent in May, well above the 3.5 percent in May 2019 before Covid-19. Among immigrants (legal and illegal together), the rate was 5.6 percent, higher than the 2.8 percent in May 2019. (Table 1C)
- The number unemployed in May stood at 7.3 million for natives and 1.5 million for immigrants, both still much higher than in May 2019, though much lower than at the peak last spring. (Table 1B)
- In addition to the unemployed, 45.8 million working-age (16-64) native-born Americans and 9.4 million immigrants were out of the labor force in May — neither working nor looking for work. (Table 3B)
- Combined, 64 million natives and immigrants were not working in May 2021 — unemployed or out of the labor force — 4.3 million more than in May 2019.1 (Tables 1B and 3B)
- Excluding young people under age 25, there were still 45.1 million immigrants and natives not working in May. (Table 2B)2
- The huge number of working-age people not working is a clear indication that there is a large pool of potential labor for employers to draw upon to fill vacant positions.
- We estimate that the unemployment rate for illegal immigrants (sometimes referred to as "illegal aliens" or the "undocumented") was 4.8 percent, and that it was 5.9 percent for legal immigrants in May. However, our estimates by legal status are approximations only. (Table 5C)
Among the less-educated:
- The unemployment rate for native-born Americans (ages 25-plus) without a bachelor's degree was 6.2 percent in May, compared to 2.7 percent for those with at least a bachelor's degree. Among immigrants (ages 25-plus), 6.4 percent without a bachelor's degree were unemployed, compared to 4 percent with at least a bachelor's. (Table 1C)
- Those without a college degree comprise less than two-thirds of the population 25 to 64, but account for three-fourths of those not working — unemployed or not in the labor force. (Table 2A and 2B)
- The Covid-19 shutdown has exacerbated the long-term decline in the labor force participation rate (share working or looking for work) of the less-educated. In May 2021, only 67 percent of working-age (16-64) natives without a bachelor's degree were in the labor force, down from 68 percent in May 2019 (before Covid), 71 percent in May 2007, and 74 percent in May 2000. (Figure 2)
- Focusing on only men who are of prime working age (25 to 54) still shows a long-term decline in labor force participation. The share of native-born men ages 25 to 54 without a bachelor's degree in the labor force was 83 percent in May 2021, compared to 84 percent in May 2019, 88 percent in May 2007, and 89 percent in May 2000. (Figure 3)
- The unemployment rate (immigrant and native) in many jobs typically performed by the less-educated has improved significantly since last spring, but remains relatively high:3 (Table 4G)
- 9.1 percent for construction laborers;
- 9.7 percent for food preparers and servers;
- 9.9 percent for maids and housekeepers;
- 11.3 percent for freight, stock, and material movers;
- 10.7 percent for cashiers; and
- 6.3 percent for healthcare aides and nursing assistants.
Data and Methods
This report uses the public Current Population Survey (CPS) to examine the employment situation in the United States as of May 2021, with particular attention paid to differences between immigrants and the native-born. Based on a monthly survey of 60,000 households, the CPS is the nation's primary source for the unemployment rate and other labor force statistics. The sample is weighted to reflect the actual size and demographic makeup of the civilian non-institutionalized population. The raw data used in this analysis comes directly from the Census Bureau, which collects the data for the Bureau of Labor Statistics (BLS).4
Key Concepts and Terms. The labor force includes all workers, plus those who are not working but have actively looked for work in the four weeks prior to the survey. There are three measures of labor force attachment used in this analysis. First, we report statistics using the standard unemployment rate, referred to by the BLS as the U-3 rate. It is calculated by dividing the number of people actively looking for work by the number in the labor force (working or looking). Second, there is the labor force participation rate, which is the share of people in the labor force, typically confined to a particular age group, such as 16 to 64. It is calculated by taking the share working or looking for work and dividing by all persons in the age group of interest. Third, there is the share of the population that is employed, which is referred to as the employment rate. It reflects the share of the population that is working divided by all persons (working and not working) in the age group of interest.
Definition of Immigrant. The foreign-born in the CPS are all persons who were not U.S. citizens at birth. They include naturalized citizens, permanent residents (green card holders), temporary visitors, guestworkers, and illegal aliens. We use the term "immigrant" to encompass all of these foreign-born individuals. In Tables 1A, 1B, 1C, 2A, 2B, 2C, 3A, 3B, and 3C we report figures separately for non-citizens and naturalized citizens. Tables 5A, 5B, 5C, 6A, 6B, and 6C report figures for legal immigrants and illegal immigrants separately.
Identifying Illegal Immigrants. Prior research indicates that most illegal immigrants are included in Census data. To determine which respondents are most likely to be illegal aliens, CIS first excludes immigrant respondents who are almost certainly not illegal aliens — for example, spouses of native-born citizens; veterans; people who have government jobs; Cubans (because of special rules for that country); immigrants who arrived before 1980 (because the 1986 amnesty should have already covered them); people in certain occupations requiring licensing, screening, or a government background check (e.g., doctors, pharmacists, and law enforcement); and people likely to be on student visas.
The remaining candidates are weighted to replicate known characteristics of the illegal population (population size, age, gender, region or country of origin, state of residence, and length of residence in the United States). CIS has previously used the Department of Homeland Security (DHS) as the source of those known characteristics; however, DHS data were last published in 2015. For more recent data, we turn to 2018 estimates from the Center for Migration Studies (CMS), including its estimates of educational attainment.5 The resulting illegal population, which consists of a weighted set of CPS respondents, is designed to match CMS on the characteristics listed above. However, we do not adjust the number of illegal immigrants for undercount in the CPS.6 Estimates for legal immigrants are calculated simply by subtracting estimated counts of illegal immigrants from the total immigrant population.
All of our findings that separate immigrants by legal status should be considered rough approximations only. Because we do not have information about the illegal immigrant population in 2021, we are forced to take the known characteristics of illegal aliens from 2018 and apply them to the new labor market data in 2021. The primary weakness of this approach is that it assumes the illegal immigrant population still has the same demographic profile (population size, age, gender, etc.) as it did in 2018.
Matching Published BLS Tables. In this report, all statistics covering the entire population, such as the unemployment rate, match those published by the BLS. However, the counts and percentages for sub-populations will in some cases be slightly different. The reason is that the Census Bureau adds "perturbations" to the public-use microdata to protect respondent confidentiality. Fortunately, the Census Bureau states that any differences between published figures and those calculated from the public-use data will be so small that they "fall well within the sampling variability associated with CPS estimates".7 In other words, there should be no meaningful difference between statistics calculated from the pubic-use data and those published by the BLS.
Finally, the figures in this report are not seasonally adjusted. Unadjusted figures are conceptually simpler and easier for other researchers to replicate. Also, the limited number of statistics on the foreign-born published in the BLS monthly "Employment Situation" reports are, like those reported here, not seasonally adjusted.8
Potential Problems with the Data. The BLS reports a potential problem with the CPS since March 2020 because of the Covid-19 pandemic. First, interviewers who administered the survey miscoded some respondents as employed when they were on furlough, temporary layoff, or similar situations. They should have been considered temporarily laid-off and coded as unemployed. The number unemployed and the unemployment rate would be higher in March 2020 through May 2021 if these individuals were counted as unemployed. For May 2021, BLS states that if these misclassified workers were included in the unemployed population then the unemployment rate would have been 0.3 percentage points higher. But they also state that, "this represents the upper bound of our estimate of misclassification and probably overstates the size of the misclassification error."9 Second, the response rates for March 2020 through May 2021 were significantly lower than prior to Covid-19, though rates have improved since hitting a low in June 2020. These lower rates increase the sampling error of the survey. In May 2021, the response rate was 78 percent. The Bureau stated in May 2021 that, "While the rate was lower than the average of 83 percent for the 12 months ending in February 2020, it was considerably higher than the low of 65 percent in June 2020."10 Neither the Census Bureau, which collects the data, nor the BLS has altered the data in response to these issues. Our analysis takes the raw CPS data as provided; and our results match published non-seasonally adjusted figures, with the caveats about perturbations in public-use data discussed above.
1 Following the example of the Bureau of Labor Statistics, the unemployment figures in this analysis include all persons who are actively looking for work, including a modest number over age 64, which was 534,000 in May 2021. The figures for those outside of the labor force include only those 16 to 64. If we focus only on those 16 to 64 unemployed and out of the labor force the figure would be 63.4 million (Table 2B)
2 As indicated in end note 1, 534,000 people over age 64 in May 2021 were unemployed, and are included in the 45.1 million figure. Excluding the elderly actively looking for work and focusing only on the those 25 to 64 unemployed or out of the labor force would make the total 44.6 million (Table 2B).
5 "State-Level Unauthorized Population and Eligible-to-Naturalize Estimates", Center for Migration Studies, undated. CMS has published illegal immigrant estimates for 2019, but they have not yet released a detailed profile of the illegal population, so we continue to use their 2018 numbers.
6 In 2018, CMS estimated a total illegal immigrant population of 10.6 million, which includes an undercount adjustment for those missed in Census Bureau data. Our analysis of the CPS totals to 9.8 million illegal immigrants, reflecting a 7.5 percent undercount.
7 "Monthly Current Population Survey Public Use Microdata Files", U.S. Census Bureau, undated.
8 "Table A-7. Employment status of the civilian population by nativity and sex, not seasonally adjusted", U.S. Bureau of Labor Statistics, June 4, 2021.
9 See the text box in the June 4, 2021, press release titled, "Coronavirus (COVID-19) Impact on May 2021 Establishment and Household Survey Data" toward the bottom of the release.
10 See posting at the BLS website entitled, "Impact of the coronavirus (COVID-19) pandemic on The Employment Situation for May 2021.