Earlier this year, the share of the adult Black population with a job—or the Black employment-to-population ratio (EPOP)—surpassed the share of the adult white population with a job for the first time in modern history. While the Black EPOP has since declined somewhat, it’s worth examining the reasons behind the sharp increase in employment over the pandemic recovery and the narrowing of the gap between Black and white workers in 2023. In my analysis, I conclude the following:
- The Black EPOP appears close to the white EPOP in recent years in part because the Black population is significantly younger (i.e., more likely to work) than the white population. After accounting for their different age distributions, Black workers have significantly lower EPOPs than white workers.
- However, the tremendous bounceback in the labor market was a significant boon for Black workers. While Black workers suffered greater losses in the pandemic recession, the employment gap narrowed because Black workers experienced a stronger jobs recovery than white workers.
Figure A shows the employment-to-population ratio (EPOP) for Black and white workers between January 2020 and June 2023. These data come from the Current Population Survey (CPS), which is a survey of households that is subject to some volatility because of its relatively small sample size. That means more weight should be put on longer-term trends, rather than month-to-month values. Here, the trends are clear. We can see that the white EPOP was about 2.0 percentage points higher than the Black EPOP in February 2020 before the pandemic hit. In the depth of the pandemic recession, that differential grew to 3.0 percentage points. While the white EPOP bounced back a bit faster in the summer of 2020—widening the gap further—Black employment then started growing faster than white employment to such a degree that the Black EPOP surpassed the white EPOP earlier this year.
Black employment-to-population ratio sharply rises, hitting white employment-to-population ratio in 2023: Employment-to-population ratio for population age 16 and older, by race, January 2020–June 2023
| Black | White | |
|---|---|---|
| Jan-2020 | 58.9% | 61.3% |
| Feb-2020 | 59.3 | 61.3 |
| Mar-2020 | 57.6 | 60.1 |
| Apr-2020 | 48.7 | 51.7 |
| May-2020 | 49.6 | 53.5 |
| Jun-2020 | 50.8 | 55.5 |
| Jul-2020 | 51.5 | 55.8 |
| Aug-2020 | 52.5 | 57.2 |
| Sep-2020 | 52.6 | 57.3 |
| Oct-2020 | 53.7 | 58.1 |
| Nov-2020 | 54.1 | 57.9 |
| Dec-2020 | 54.0 | 57.9 |
| Jan-2021 | 54.7 | 57.8 |
| Feb-2021 | 54.2 | 57.9 |
| Mar-2021 | 54.9 | 58.1 |
| Apr-2021 | 55.2 | 58.2 |
| May-2021 | 55.4 | 58.3 |
| Jun-2021 | 55.9 | 58.2 |
| Jul-2021 | 55.8 | 58.6 |
| Aug-2021 | 56.2 | 58.7 |
| Sep-2021 | 56.5 | 58.9 |
| Oct-2021 | 56.3 | 59.1 |
| Nov-2021 | 56.8 | 59.5 |
| Dec-2021 | 56.6 | 59.7 |
| Jan-2022 | 57.7 | 59.9 |
| Feb-2022 | 58.1 | 60.1 |
| Mar-2022 | 58.2 | 60.2 |
| Apr-2022 | 58.6 | 59.9 |
| May-2022 | 59.1 | 59.9 |
| Jun-2022 | 58.5 | 59.9 |
| Jul-2022 | 58.3 | 60.0 |
| Aug-2022 | 57.9 | 60.1 |
| Sep-2022 | 58.5 | 60.1 |
| Oct-2022 | 58.4 | 60.0 |
| Nov-2022 | 58.8 | 59.8 |
| Dec-2022 | 58.9 | 60.2 |
| Jan-2023 | 59.5 | 60.2 |
| Feb-2023 | 59.8 | 60.1 |
| Mar-2023 | 60.9 | 60.2 |
| Apr-2023 | 60.0 | 60.3 |
| May-2023 | 59.6 | 60.2 |
| Jun-2023 | 58.9 | 60.4 |
Note: Using latest data from the Bureau of Labor Statistics that do not create mutually exclusive categories of race and ethnicity: Black and white data do not exclude Hispanic adults.
Source: EPI analysis of Bureau of Labor Statistics’ Current Population Survey public data.
Let’s take a step back and disentangle some demographic issues that may be related to the trends and levels of employment that we see in Figure A. Figure B displays the age distribution of both the Black and white adult populations, 16 years and older. What should be immediately apparent is that the white age distribution is significantly older than the Black age distribution. In all the age groups below 50, the Black distribution is relatively larger than the white. In fact, about 61% of the Black population is below 50 while only about 49% of the white population is below 50. This is significant because—starting at 50 years old—the older the population is, the less likely they are to work.
I should be clear I’ve changed the race categorizations slightly from what the Bureau of Labor Statistics (BLS) reports and what I presented in Figure A. In the remainder of the figures in this blog post, I’m using the EPI microdata extracts from the CPS and utilize EPI’s typical mutually exclusive race/ethnic categories used throughout much of our research and data library; therefore, in this post, Black refers to Black non-Hispanic and white refers to white non-Hispanic. I’m also combining the first three months of the year, January to March 2023.
Black population distribution younger than white: Share of population in each age group, by race, 2023
| Black | White | |
|---|---|---|
| 16-19 | 7.8% | 5.3% |
| 20-24 | 9.8% | 7.1% |
| 25-29 | 9.5% | 7.0% |
| 30-34 | 10.0% | 7.6% |
| 35-39 | 8.4% | 7.6% |
| 40-44 | 8.3% | 7.3% |
| 45-49 | 7.3% | 6.8% |
| 50-54 | 7.5% | 7.7% |
| 55-59 | 7.3% | 8.1% |
| 60-64 | 7.5% | 8.9% |
| 65-69 | 5.7% | 8.3% |
| 70-74 | 4.3% | 7.0% |
| 75+ | 6.5% | 11.1% |
Notes: Data for January to March 2023. Black and white race categories are also non-Hispanic.
Source: Authors’ analysis of Current Population Survey basic monthly microdata, EPI Current Population Survey Extracts, Version 1.0.40 (2023), https://microdata.epi.org.
Figure C displays the employment-to-population ratio by race for the same set of age groups. Regardless of race, it should be apparent that EPOPs rise through the teen years and early 20s, are highest between ages 25 and 49, and then start declining at age 50 before falling precipitously after age 59. But when we compare by race, Black EPOPs are lower than white EPOPs at nearly every age level. This is because of a long history of occupational segregation, persistent discrimination, and unequal bargaining power. Given that Black workers are younger on average and Black EPOPs are lower than white EPOPs across the age distribution, the fact that the Black EPOP appears close to the white EPOP over the last several years can in part be attributed to their younger age distribution.
At nearly every age level, Black employment-to-population ratio is lower than white: Employment-to-population ratio by age group and race, 2023
| Black | White | |
|---|---|---|
| 16-19 | 25.7% | 35.4% |
| 20-24 | 61.8% | 68.5% |
| 25-29 | 75.8% | 81.5% |
| 30-34 | 80.1% | 83.1% |
| 35-39 | 78.8% | 83.0% |
| 40-44 | 78.3% | 84.5% |
| 45-49 | 79.5% | 82.4% |
| 50-54 | 73.7% | 79.6% |
| 55-59 | 67.8% | 73.1% |
| 60-64 | 50.0% | 56.5% |
| 65-69 | 29.2% | 33.2% |
| 70-74 | 18.6% | 18.6% |
| 75 + | 8.8% | 8.2% |
Notes: Data for January to March 2023. Black and white race categories are also non-Hispanic.
Source: Authors’ analysis of Current Population Survey basic monthly microdata, EPI Current Population Survey Extracts, Version 1.0.40 (2023), https://microdata.epi.org.
Figure D puts the different parts of that story together, using the age distribution and EPOPs for different age groups to construct the actual EPOPs in both 2019 and 2023 (using January to March 2023 data) as well as three simulated EPOPs for 2023. The first two sets of bars on the left show actual Black and white EPOPs in 2019 and 2023. The Black EPOP rose between 2019 and 2023 while the white EPOP ticked down.
Now, let’s take a step back and look at the simulations. The first simulation holds constant the 2019 EPOP levels for each age group, but incorporates changes to the age distribution for Black and white adults in 2023. We see here that both the Black and white EPOPs have declined by about one percentage point since 2019. The white population is aging slightly faster than the Black population, but not significantly so. This means that the change in EPOPs between 2019 and 2023 is primarily not driven by changes in the relative age distribution between Black and white adults.
Now, let’s turn to the fourth set of bars. Here, I’ve held the 2019 age distribution constant for both groups, but I have allowed the EPOPs to change as they did between 2019 and 2023. These bars look very similar to the relationship between actual 2023 EPOPs by race. The Black EPOP has now surpassed the white EPOP when the aging population isn’t taken into account. That means that Black workers experienced a relatively stronger recovery than white workers.
Black employment-to-population ratio rose in the recovery, but remains lower than white when adjusting for age distribution: Employment-to-population ratio (EPOP) by age group and race, under alternative scenarios
| Black | White | |
|---|---|---|
| 2019 | 58.1% | 59.9% |
| 2023 (actual) | 59.5% | 59.0% |
| 2023 (age adjusted only) | 57.2% | 58.8% |
| 2023 (EPOP adjusted only) | 60.4% | 60.1% |
| 2023 (apply white age distribution) | 54.7% | 59.0% |
Notes: Data for January to March 2023. Black and white race categories are also non-Hispanic. The first two sets of bars represent the actual EPOPs in 2019 and 2023, respectively. The third set of bars shows what the 2023 EPOPs would be if the population aged to 2023, but the EPOPs didn’t change. The fourth set of bars shows what the 2023 EPOPs would be if the population didn’t age between 2019 and 2023. The fifth set shows the Black and white EPOPs if they both used the white age distribution.
Source: Authors’ analysis of Current Population Survey basic monthly microdata, EPI Current Population Survey Extracts, Version 1.0.40 (2023), https://microdata.epi.org.
One thing I’d note is that the strength of the recovery for Black workers cannot be attributed to a stronger recovery, in general, for younger workers. In fact, the Black EPOP grew for workers younger and older than age 50 between 2019 and 2023, so the age distribution itself is not an explanation for the faster employment growth. It may come as no surprise that the rate of growth could be faster simply because Black workers suffered greater losses in the pandemic recession, but Figure D compares the peak before the pandemic recession to today, accounting for the deep recession and recovery. In this peak-to-peak analysis, Black workers are now unexpectedly in a stronger position now than they were before the pandemic hit.
These gains are likely the result of a tight labor market spurred by policy investments during the pandemic recession. Because the historical 2:1 Black-white ratio in the unemployment rate remains fairly fixed throughout the business cycle, losses and gains are sharper for Black workers. However, tight labor markets can erode race-based gaps in employment-to-population ratios. Policy investments (e.g., unemployment insurance expansions, economic impact payments, and aid to state and local governments) created an immediate and strong bounceback from the pandemic recession, compared with the pursuit of austerity and prolonged recovery from the Great Recession.
Finally, the last set of bars tells a story about the Black-white EPOP differential, not about the change since 2019. As discussed in Figures B and C, the white age distribution is older than the Black age distribution, and EPOPs are higher for white adults than Black adults. Therefore, it’s not surprising that when the Black 2023 EPOP is reweighted to the older white age distribution, the simulated Black EPOP is much lower.
In conclusion, after accounting for their different age distributions, Black workers have significantly lower EPOPs than white workers. However, the tremendous bounceback in the labor market was a significant boon for Black workers. If that strength continues, the labor market may finally deliver sustained and equitable gains and further narrow the disparities that exist today.
Notes
As of this writing, EPI’s extracts are only available through March 2023 because of some changes to the BLS methodology that we are still sorting out for April and May. It is very unlikely that including those months will change the conclusions reached here, particularly since the Black EPOP has softened somewhat since then. If anything, the gaps are larger than what I’m showing here.
For consistency with the data available for 2023, I’m also restricting the 2019 sample for the first three months of the year so there are no concerns about seasonality.
Sign up for EPI’s newsletter so you never miss our research and insights on ways to make the economy work better for everyone.
