Lowest-Income Workers See Accelerated Earnings Growth During Pandemic
In many respects—from higher COVID-19 mortality rates to a scarcity of child care—the pandemic has disproportionately harmed low-income workers. Earnings growth, triggered by labor shortages and high turnover rates, could be a rare exception.
Throughout 2021, labor markets gradually tightened, with demands for higher pay accompanying the highest labor market turnover in two decades. According to the Bureau of Labor Statistics (BLS), wages increased in many industries as the labor market churned. We examine these gains and find that workers who earn the least saw the fastest rates of earnings growth during the pandemic, even after accounting for composition effects that reflect changes in who was working. This pattern of more rapid earnings growth also applies to Hispanic workers, who are more likely to hold lower-paying jobs, compared with non-Hispanic white workers.
Nominal Earnings Growth Favors Low-Income Workers
Chart 1 shows that, before the pandemic, growth in nominal earnings was comparable across all five earnings quintiles. But once the pandemic struck, the lowest quintile of earners saw the fastest rate of growth in earnings. Between the last quarter of 2019 and the last quarter of 2020, nominal earnings grew 7.2 percent for low-earning workers, compared with 5.6 percent for mid-earning workers and 3.4 percent for high-earning workers. This relatively rapid rise for low-income workers continued over the following year as their nominal earnings grew 7.7 percent, compared with 4.8 percent for mid-earning workers and 3.6 percent for high-earning workers.
Cumulatively in the two-year period, the gap in the total earnings gain between the top and bottom quintiles was more than 10 percentage points. Nevertheless, all five groups saw improvements in nominal earnings over that period.
The earnings gains in low-income groups concurs with the tightening of labor markets in 2021, now widely referred to as the Great Resignation. Workers left their jobs in large numbers, citing reasons such as low pay (63 percent), no opportunities for advancement (63 percent), feeling disrespected at work (57 percent) and child care issues (48 percent among those with a child younger than 18 in the household).
Did earnings for all quintiles of worker grow faster than inflation? Consumer price index from the BLS suggest overall inflation of 10 percent in the period of study—between fourth quarter 2019 and first quarter 2022. Using this measure, earnings growth for the four lower quintiles outpaced inflation, though only narrowly for the 40–60 and 60–80 percentile earners.
Due to the transient nature of pandemic inflation, it is useful to compare the earnings gain to core inflation trends. The Federal Reserve Bank of Dallas calculates the Trimmed Mean PCE inflation rate, a more stable measure of core inflation in the Consumer Price Index. Over the period of study, the Trimmed Mean inflation rate grew 5.8 percent. Using this measure, all income groups saw real earnings growth over this period.
Gains against inflation are also seen when we compare Hispanic workers, who are more likely to hold a job in one of the lower-income quintiles, with non-Hispanic white workers (Chart 2). Earnings growth for Hispanic workers is consistently higher than for white workers over the period, while gains for the latter group fail to beat inflation. The differences between Hispanic and non-Hispanic white workers remain statistically significant even after controlling for educational attainment.
Caveat: The Impact of Worker Entry and Exit
Earnings movement by quintile substantially reflects how people in those quintiles are faring over time. But it also partly reflects the effect of people leaving their jobs or coming into the workforce. This introduces what economists call a “composition” bias.
While this may sound esoteric, it could have very real implications for our analysis. The composition effect can bias our results upward or downward. As described in a recent Dallas Fed Economics article, Sean Howard, Robert Rich and Joseph Tracy found unusually large composition effects in the positive direction for second quarter 2020 and in the opposite direction for second quarter 2021. Their methodology is applied to this analysis in Chart 3.
An upward bias could occur, for example, when workers with earnings at the very bottom of the lowest-earning quintile suffered disproportionate job losses over the last two years, perhaps because they disproportionately contracted COVID-19 or couldn’t find child care. This would cause average earnings for the lowest quintile to rise, even if no individual’s earning actually went up. For our current analysis, we found a positive composition effect for the lowest-income group for most quarters during the pandemic, as seen in Chart 3.
Alternatively, suppose retirements occurred mainly among the highest-earning workers in the highest-earning quintile. If that happened, then the average earning in the quintile would fall, which would again provide the appearance that seasoned low-earning workers were gaining ground on their high-earning counterparts—without that actually being the case. Chart 3 shows negative composition effects for each of the five income groups in March 2022.
We find that, after removing the composition effect, both of our previous conclusions regarding gains for low-income and Hispanic workers (Charts 1 and 2) remain. Chart 3 distinguishes the composition effect (Chart 3A) from the aggregation effect (Chart 3B), the latter of which is free from the disruptions caused by the frequent labor turnovers during the pandemic. The conclusion drawn from the aggregation effect is consistent with conclusions in Chart 1.
Despite Earnings Growth, Large Gaps Remain for Low-Income Workers
Overall, workers in the lowest-income quintile saw the fastest rate of earnings growth during the pandemic. Earnings growth for the four lower quintiles outpaced inflation, though only narrowly for the 40–60 and 60–80 percentile earners. Even so, significant challenges lie ahead on the road to a more inclusive economy. Large gaps remain between low-income and high-income workers as well as between white workers and traditionally disadvantaged populations. And low-income families continue to experience job volatility, health risks and rising asset inequality.
- Statistical significance is calculated using a simple difference-in-difference model with ethnic group dummies, a post-COVID dummy and ethnicity-COVID interaction terms. Educational attainment variables include high school, college and graduate-level attainment.
The views expressed are those of the authors and should not be attributed to the Federal Reserve Bank of Dallas or the Federal Reserve System.