Unemployment Rates During the COVID-19
June 15, 2021
Pandemic
Gene Falk, Coordinator
The Coronavirus Disease 2019 (COVID-19) pandemic has had a significant effect on
Specialist in Social Policy
labor market metrics for every state, economic sector, and major demographic group in

the United States. This report provides information on unemployment rates, labor force
Paul D. Romero
participation rates, and nonfarm payrolls in the United States during the ongoing
Research Assistant
pandemic. It presents CRS analysis of overal unemployment rate trends during the

pandemic. The report first examines these trends national y, and at the state and
Jameson A. Carter
industrial levels. Next, it examines how unemployment rates varied across demographic
Research Assistant
groups. The report then repeats this analysis, where appropriate, for the labor force

participation rate, which sheds light on the size of the workforce wil ing and available
Isaac A. Nicchitta
for work. The final portion of the report analyzes the impact the pandemic has had on
Research Assistant
overal employment and by sector.

Emma C. Nyhof
Among other findings, this report shows the following:
Research Assistant

 In April 2020, the unemployment rate reached 14.8%—the highest rate observed
since data collection began in 1948. In May 2021, unemployment remained

higher (5.8%) than it had been in February 2020 (3.5%).
 The labor force participation rate declined to 60.2% in April 2020—a level not seen since the
early 1970s—then began a partial recovery in May 2020. The labor force participation rate was
61.6% in May 2021, 1.8 percentage points below the level in January 2020, before the pandemic
and the economic recession.
 Nonfarm payrolls shed 22.1 mil ion jobs between January 2020 and April 2020, with employment
declining to 86% of its pre-recession level. In May 2021, aggregate employment remained 7.3
mil ion jobs below its pre-recession level.
 The COVID-19 pandemic has impacted economic sectors disparately. The leisure and hospitality
sector lost the largest number of jobs since January 2020, and persons last employed in this sector
have consistently exhibited some of the highest unemployment rates throughout the pandemic.
Additional y, the education and services sector and the government sector have exhibited the
second and third-largest losses in jobs since January 2020, despite relatively low unemployment
rates among persons last employed in these sectors.
 The COVID-19 pandemic has impacted demographic groups disparately. Although al
demographic groups were affected, persons identifying as Black or Hispanic and younger
workers generally experienced relatively high peaks in unemployment and relatively steep
declines in labor force participation over the course of the pandemic. Additional y, persons with
lower educational attainment have general y experienced relatively higher unemployment rates
and lower labor force participation throughout the pandemic.

Congressional Research Service

link to page 25 link to page 27 Unemployment Rates During the COVID-19 Pandemic

Figure 18. Job Loss During the COVID-19 Pandemic in the Public Sector ............................. 21

Contacts
Author Information ....................................................................................................... 23

Congressional Research Service

Unemployment Rates During the COVID-19 Pandemic

Introduction
The National Bureau of Economic Research has declared February 2020 as the start of the current
economic downturn, marking the end of the longest period of expansion in U.S. history.1 This
expansion followed the Great Recession (December 2007 to June 2009), a downturn widely
considered to be the worst since the Great Depression (August 1929 to March 1933).2 The
unemployment rate rose quickly in March 2020, and by April 2020 it had greatly surpassed its
previous peaks observed during and just after the Great Recession. This rise in unemployment
was caused by an unprecedented loss of 22.1 mil ion jobs between January 2020 and April 2020.
Many individuals left the labor force over this period, and by April 2020 the labor force
participation rate3 declined to a level not seen since the early 1970s.
This deterioration in the U.S. labor market corresponded with various advisory or mandated stay-
at-home orders implemented in response to the Coronavirus Disease 2019 (COVID-19) pandemic
and other pandemic-related factors affecting U.S. demand.4 States and localities implemented
these orders5 to mitigate the risks of COVID-19 after it was declared a pandemic disease by the
World Health Organization on March 11, 2020.6
This report discusses the state of the U.S. labor market using data from the Bureau of Labor
Statistics (BLS). The three primary sources are the Current Population Survey (CPS), the Local
Area Unemployment Statistics (LAUS) program, and the Current Employment Statistics (CES)
program. In addition to the usual caveats about estimates (see the “General Data Caveats”
section), there were additional data chal enges caused by the COVID-19 pandemic (see the
“COVID-19 Pandemic-Related Data Issues” section). The pandemic led to lower survey response
rates by businesses and households, and BLS detected an error in their categorization procedures
that likely underestimated unemployment early in the recession.7 Labor force participation rates
were not affected by this categorization error and met BLS standards of accuracy, despite
depressed response rates.8 BLS also identified an error in the nonfarm enrollment data processing

1 T he National Bureau of Economic Research; see https://www.nber.org/cycles.html for their historical series of
expansions and contractions. For more on their process for determining expansions and contractions, see
https://www.nber.org/cycles/recessionsfaq.html#:~:text=
What%20is%20an%20expansion%3F,more%20than%20a%20few%20months.& text=
Expansion%20is%20the%20normal%20state,economy%3B%20most%20recessions%20are%20brief .
2 T he Great Recession was a particularly long recession, characterized by a steady and large increase in unemployme nt
and unprecedented decreases in labor force participation. T he unemployment rates observed during the Great
Recession, however, never surpassed those of the early 1980s. For more on labor force metrics during the Great
Recession see CRS Report R45330, Labor Market Patterns Since 2007, by Sarah A. Donovan and Marc Labonte.
3 Defined as the percentage of persons in the overall adult population who either have a job or are looking for a job.
4 See CRS Insight IN11388, COVID-19: U.S. Economic Effects, by Rena S. Miller and Marc Labonte.
5 For a list of state-level stay-at-home orders and estimates of the impact of these orders on risk mitigation, see Amanda
Moreland, Christine Herlihy, and Michael A. T ynan et al., Tim ing of State and Territorial COVID-19 Stay-at-Hom e
Orders and Changes in Population Movem ent, Centers for Disease Control, Morbidity and Mortality Weekly Report
Vol. 69 No. 35, Washington, DC, September 4, 2020, pp. 1198-1203, https://www.cdc.gov/mmwr/volumes/69/wr/
mm6935a2.htm?s_cid=mm6935a2_w.
6 World Health Organization, Coronavirus Disease 2019 (COVID-19), Situation Report 51, March 11, 2020, p. 1,
https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200311-sitrep-51-covid-19.pdf.
7 See CRS Insight IN11456, COVID-19: Measuring Unemployment, by Lida R. Weinstock.
8 For BLS impact summaries of COVID-19 on these measures, see https://www.bls.gov/covid19/effects-of-covid-19-
pandemic-and-response-on-the-employment-situation-news-release.htm.
Congressional Research Service

1

link to page 7 Unemployment Rates During the COVID-19 Pandemic

system wherein some businesses were improperly included in estimates, although BLS has since
determined the impacts of this error were insignificant.9
This report general y finds the following:
 The unemployment rate peaked10 in April 2020, at a level not seen since data
collection started in 1948, before declining to a level in May 2021 that stil
remained 2.3 percentage points above the rate observed in February 2020.
 In April 2020, the labor force participation rate declined to levels not seen since
the early 1970s. Labor force participation has improved since then to 61.6%,
which remains 1.8 percentage points below its pre-recession level.
 Nonfarm payrolls shed 22.1 mil ion jobs between January 2020 and April 2020.
In May 2021, aggregate employment remained 7.3 mil ion jobs below its pre-
recession level.
U.S. Unemployment Rate: Historical Trends
Prior recessions typical y developed with gradual y increasing economic distress. The current
recession was caused by the COVID-19 pandemic, which was an abrupt and exogenous shock to
the economy. The pandemic resulted in rapidly implemented efforts to limit contact among
individuals and many shutdown orders. Therefore, the trends in the unemployment rate in the
current recession differ from those in prior recessions (see Figure 1). Rates observed during prior
recessions rose relatively gradual y over the course of an economic downturn and then peaked.
The current recession exhibited an unprecedented sharp increase in the unemployment rate (10.3
percentage points) from February to April 2020.11 Following April, the rate declined rapidly (6.4
percentage points from April 2020 to August 2020) as temporarily furloughed workers returned to
work. Despite these rapid declines, the unemployment rate remains at an elevated level (5.8%)
compared to February 2020. The share of workers on furlough has declined since peaking in April
2020, while the share of permanently laid off workers has steadily increased.12 Although
economic projections have general y improved since early in the recession, the Congressional
Budget Office (CBO) has projected that unemployment rates over 5.0% wil persist over the next
two years.13

9 For a description of this error, see https://www.bls.gov/ces/notices/2021/ces-sample-rotation-issue-caused-by-
pandemic-related-challenges-to-enrollment.htm.
10 T hroughout this report, peak refers to the highest level of unemployment between January 2020 and May 2021. It
does not account for months outside this range.
11 For information on the differences between the congressional response to the current recession compared to the
congressional response during the Great Recession in the Unemployment Insurance system, see CRS Report R46472,
Com paring the Congressional Response to the Great Recession and the COVID -19-Related Recession: Unem ploym ent
Insurance (UI) Provisions, by Katelin P. Isaacs and Julie M. Whittaker.
12 CRS analysis of BLS data, which can be found at https://www.bls.gov/webapps/legacy/cpsatab11.htm. Workers on
temporary layoff declined from 18.0 million in April 2020 to 1.8 million in May 2021 as the number of permanent job
losers increased from 2.6 million in April 2020 to 4.0 million in May 2021.
13 See https://www.cbo.gov/about/products/budget-economic-data#4 for CBO’s 10-year economic projections of
unemployment rates, as of February 2021.
Congressional Research Service

2

Unemployment Rates During the COVID-19 Pandemic

Figure 1. Historical Unemployment Rate
Seasonal y adjusted monthly data, January 1948 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: Shaded regions indicate recessionary periods as identified by the National Bureau of Economic Research.
The unemployment rate’s relatively rapid decline since April 2020 may have been aided by laws
passed in response to both the recession and the pandemic. Congress has passed three rounds of
stimulus checks for families,14 expanded nutrition assistance programs,15 and enacted increases in
refundable tax credits (which are not scheduled to be disbursed until July 2021).16 These
provisions increase families’ disposable income and those that have taken effect may have
increased consumer spending, enabling businesses to better endure the recession; a 2020 study
from NBER found that higher replacement rates of lost wages led to higher consumer spending.17
Congress also enacted the Paycheck Protection Program, which provides loans that can be fully
forgiven if the majority of funds borrowed are used to maintain payrolls.18
Additional y, Congress expanded Unemployment Insurance (UI) program benefits and extended
length of coverage.19 Some say this policy could directly lead to the unemployment rate
remaining above what it would be otherwise because past research has shown UI extensions can

14 For more information, see CRS Insight IN11605, COVID-19 and Direct Payments: Comparison of First and Second
Round of “Stimulus Checks” to the Third Round in the American Rescue Plan Act of 2021 (ARPA; P.L. 117 -2), by
Margot L. Crandall-Hollick.
15 For more information, see CRS Report R46681, USDA Nutrition Assistance Programs: Response to the COVID-19
Pandem ic, by Randy Alison Aussenberg and Kara Clifford Billings.
16 For more information, see CRS Report R46680, The American Rescue Plan Act of 2021 (ARPA; P.L. 117 -2): Title
IX, Subtitle G—Tax Provisions Related to Prom oting Econom ic Security, by Molly F. Sherlock, Margot L. Crandall-
Hollick, and Jane G. Gravelle.
17 Miguel G. Casado, Britta Glennon, and Julia Lane, et al., The Effect of Fiscal Stimulus: Evidence from COVID-19,
NBER, Working Paper 27576, Cambridge, MA, August 2020, https://www.nber.org/papers/w27576.
For more on this topic and theories of recessionary policy, also see CRS Report R46460, Fiscal Policy and Recovery
from the COVID-19 Recession, by Jane G. Gravelle and Donald J. Marples.
18 For more information, see CRS Report R46397, SBA Paycheck Protection Program (PPP) Loan Forgiveness: In
Brief, by Robert Jay Dilger and Sean Lowry.
19 For more information, see CRS Report R46687, Current Status of Unemployment Insurance (UI) Benefits:
Perm anent-Law Program s and COVID-19 Pandem ic Response, by Julie M. Whittaker and Katelin P. Isaacs.
Congressional Research Service

3

link to page 9 Unemployment Rates During the COVID-19 Pandemic

extend the duration of unemployment for a relatively smal segment of unemployed workers.20 As
of the cover date of this report, 26 states have announced plans to opt out of expanded UI before
its expiration in September 2021. Many of these states have cited the potential work disincentives
of expanded UI as one reason for ending the program.21 A recent study examining data from April
to July 2020 found that while UI benefit increases reduced employment by 0.2% to 0.4%, overal
spending by recipients increased by 2.0% to 2.6%.22 The authors note that while these expansions
directly decreased employment among recipients, that increased spending among recipients may
have insulated the labor market from further deterioration.
These (and other) policies may have affected unemployment rate trends in several ways; however,
the causal impact of policy choices on the unemployment rate is beyond the scope of this report.
Comparing the Great Recession and the COVID-19 Recession
During the Great Recession, the unemployment rate increased from 5.0% in December 2007 (the
start of the recession) to 9.5% in June 2009 (the end of the recession) (see Figure 2). The
unemployment rate peaked at 10.0% in October 2009, four months after the recession official y
concluded. In the current recession, the unemployment rate increased from 3.5% in February
2020, to 4.4% in March 2020, and then peaked at a high of 14.8% in April 2020. Since then, the
unemployment rate fel to 5.8% in May 2021. This increase represents the quickest month-over-
month increase in the unemployment rate and the peak represents the highest overal
unemployment rate since the CPS data started being collected in 1948.23 The decline in the
unemployment rate of 6.4 percentage points between April 2020 and August 2020 represented the
largest decline in the unemployment rate over a four-month period since the data collection
began.

20 T he study referenced here showed that UI extensions during Great Recession increased the unemployment rate of 0.4
percentage points. Henry S. Farber and Robert G. Valletta, Do Extended Unem ploym ent Benefits Lengthen
Unem ploym ent Spells? Evidence from Recent Cycles in the U.S. Labor Market, NBER, Working Paper 19048,
Cambridge, MA, May 2013, https://www.nber.org/papers/w19048.
21 For more information, see CRS Insight IN11679, States Opting Out of COVID-19 Unemployment Insurance (UI)
Agreem ents, by Julie M. Whittaker and Katelin P. Isaacs.
22 Peter Ganong et al., Spending and Job Search Impacts of Expanded Unemployment Benefits: Evidence from
Adm inistrative Micro Data, Becker Friedman Institute, Working Paper No. 2021 -19, Chicago, IL, February 2021,
https://bfi.uchicago.edu/wp-content/uploads/2021/02/BFI_WP_2021-19.pdf.
23 T here are many differences in labor force statistics observed during the Great Recession, its aftermath, and the
COVID-19 recession. For more on this and for information on labor market patterns since 2007, see CRS Report
R45330, Labor Market Patterns Since 2007, by Sarah A. Donovan and Marc Labonte.
Congressional Research Service

4

link to page 10

Unemployment Rates During the COVID-19 Pandemic

Figure 2. U.S. Unemployment Rate
Seasonal y adjusted monthly data, November 2004 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
COVID-19 Recession: Unemployment Trends
The COVID-19 pandemic has affected the unemployment rates for every state, economic sector,
and major demographic group. In the early stages of the current recession, unemployment rates
disproportionately increased among economic sectors delivering in-person services. Some
demographic groups are overrepresented in such sectors, contributing to higher unemployment
rates for those workers.24
Unemployment Rates by State
Figure 3 displays state-level monthly unemployment rates from January 2020 to April 2021 (the
state-level data for May 2021 have not been released as of the cover date of this report). The
figure shows that no state was immune from economic damage early in the pandemic.25 At the
onset of the current recession, the unemployment rate for every state and the District of Columbia
surpassed levels seen during the Great Recession. The variation in economic damage was due to a
number of factors, including the proportion of jobs in sectors that provide nonessential services to

24 Guido Matias Cortes and Eliza Forsythe, “ T he Heterogeneous Labor Market Impacts of the Covid-19 Pandemic,”
Upjohn Institute, May 2020; and Robert Fairlie, “ T he Impact of Covid-19 on Small Business Owners: Evidence of
Early-Stage Losses from the April 2020 Current Population Survey,” NBER Working Paper No. 27309, June 2020.
25 Felipe Lozano-Rojas et al., “Is the Cure Worse than the Problem Itself? Immediate Labor Market Effects of COVID-
19 Case Rates and School Closures in the U.S.,” NBER Working Paper No. 27127, May 2020; Eliza Forsythe et al.,
“Labor Demand in the T ime of COVID-19: Evidence from Vacancy Postings and UI Claims,” NBER Working Paper
No. 27061, April 2020.
Congressional Research Service

5

link to page 12

Unemployment Rates During the COVID-19 Pandemic

in-person customers,26 individual concerns of contracting COVID-19 causing declines in personal
consumption,27 and the implementation of stay-at-home orders and business closure policies.28
Figure 3. State Unemployment Rate
Seasonal y adjusted monthly data, January 2020 to April 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: The National Bureau of Economic Research identified February 2020 as the first month of the current
recession. The month-over-month changes are point estimates and have not been tested for significance. The
state-level data for May 2021 have not been released as of the cover date of this report.
The unemployment rate in most states peaked in April 2020 and since declined. In April 2021, the
states with the highest unemployment rates were Hawai (8.5%), California (8.3%), New Mexico
(8.2%), New York (8.2%), and Connecticut (8.1%). The states with the lowest unemployment
rates in April 2021 were Nebraska (2.8%), New Hampshire (2.8%), South Dakota (2.8%), Utah
(2.8%), and Vermont (2.9%).
Unemployment Rates by Sector
Figure 4 displays the change in sector unemployment rates from January 2020, before the start of
the recession, to May 2021. Sector unemployment rates define the unemployment rate among

26 Matthew Dey and Mark 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.
27 Austan Goolsbee and Chad Syverson, “Fear, lockdown, and diversion: comparing drivers of pandemic economic
decline 2020,” NBER Working Paper No. 27432, June 2020.
28 Sumedha Gupta et al., “Effects of Social Distancing Policy on Labor Market Outcomes,” NBER Working Paper No.
2780, May 2020.
Congressional Research Service

6

link to page 12 link to page 12 Unemployment Rates During the COVID-19 Pandemic

individuals whose last job was in a particular sector. The figure shows that some sectors were
disparately impacted by the recession, although the data are not seasonal y adjusted. Without
seasonal adjustments, it is difficult to determine the extent to which unemployment trends are
related to the recession or to seasonal trends. Readers should interpret trends shown in Figure
4with some caution as this report does not test for statistical significance of these differences.
Congressional Research Service

7

Unemployment Rates During the COVID-19 Pandemic

Figure 4. Unemployment Rates by Sector
Non-seasonal y adjusted monthly data, January 2020 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: Statistical significance is not calculated for these trends and it is unclear how dependent these trends are
on regular seasonal variation. Sectors are defined by the North American Industry Classification System (NAICS)
and can be found at https://www.bls.gov/iag/tgs/iag_index_naics.htm. The figure shows unemployment rates for
wage and salary workers. Two sectors do not have displayed peak unemployment rates. The mining sector
experienced two peaks of 19.3% in November 2020 and February 2021, while the agriculture sector experienced
its peak before the recession and pandemic began (12.5% in January 2020).
Congressional Research Service

8

link to page 14 Unemployment Rates During the COVID-19 Pandemic

Workers whose last job was in the leisure and hospitality sector experienced a higher peak in
unemployment (39.3% in April 2020) than did workers who were previously employed in any
other sector; they also had the highest unemployment rate in May 2021 (10.1%). However,
elevated unemployment rates are not constrained to sectors providing in-person services. Workers
whose last job was in the mining or extraction sector have experienced steadily increasing
unemployment since the onset of the recession; in May 2021 they exhibited the second highest
rate (9.6%) among al workers across sectors. The lowest May 2021 rates were among workers
whose last job was in the government (2.2%), financial activities (3.0%), or education and health
services (3.4%) sectors. These sectors have exhibited relatively low unemployment rates
compared to most other sectors from February 2020 through May 2021.29 Within sectors, certain
types of workers were more likely to lose their jobs than others early in the recession. For
example, some studies from early in the pandemic suggest that low-wage workers in the leisure
and hospitality sector and other services sectors experienced disproportionately large employment
losses.30
Unemployment Rates for Full- and Part-Time Workers
As shown in Figure 5, part-time workers experienced a higher peak unemployment rate (24.5%
in April 2020) than full-time workers (12.8% in April 2020). This disparity modestly reversed as
the recession progressed, as the unemployment rate for part-time workers in May 2021 (5.1%)
was less than the unemployment rate for full-time workers (5.8%).
There are a few considerations that may provide additional context. First, some workers who last
worked part-time jobs may have left the labor force, and hence are not counted in the official
unemployment statistics used in this report. It is unclear whether that is the case. Additional y,
there was a considerable increase in the number of part-time workers who reported that they
would have preferred to work full-time but work part-time because their hours were reduced or
they could only find part-time jobs.31 This could be reflected as a reduced unemployment rate
among part-time workers. Further, BLS has observed that labor underutilization has remained
elevated for workers, including those who have been working part-time for economic reasons.32

29 T hese data are not seasonally adjusted and do not account for the likely seasonal variation in employment within the
education and health services sector.
30 Alexander Bartik et al., “ Measuring the labor market at the onset of the COVID-19 crisis,” NBER Working Paper
No. 27613, July 2020; and Guido Matias Cortes and Eliza Forsythe, “ T he Heterogeneous Labor Market Impacts of the
Covid-19 Pandemic,” Upjohn Institute Working Paper, May 2020.
31 T he number of workers working part-time for economic reasons increased from 4.4 million in February 2020 to 5.3
million in May 2021 on a seasonally adjusted basis. See https://www.bls.gov/web/empsit/cpseea07.htm.
32 See https://www.bls.gov/news.release/empsit.t15.htm for U-6 unemployment rates. U-6 is a measure of the total
unemployed, plus all persons marginally attached to the labor force, plus total employed part time for e conomic
reasons, as a percentage of the civilian labor force plus all persons marginally attached to the labor force. For more on
this measure, see CRS In Focus IF10443, Introduction to U.S. Econom y: Unem p loyment, by Lida R. Weinstock.
Congressional Research Service

9

link to page 15

Unemployment Rates During the COVID-19 Pandemic

Figure 5. Unemployment Rates for Part- and Full-Time Workers
Seasonal y adjusted monthly data, January 2020 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: Both groups experienced their peak unemployment rate in April 2020.
Unemployment Rates by Sex and Age
As seen in Figure 6, unemployment rates tended to increase more for younger workers and were
higher for women early in the recession. Between February and April 2020, the rate for women
ages 16-19 increased by 25.3 percentage points to 36.3%; in contrast, the rates for men of the
same age increased by 16.2 percentage points to 28.2%. Since then, the gap between younger
men and women has reversed.
The unemployment rate for teenaged men (10.1%) was higher than the rate for teenaged women
(9.0%) in May 2021, although both of these rates roughly equal pre-recession levels. While
unemployment rates for younger workers remain relatively high compared to older workers, the
May 2021 rates for men and women across the remaining age groups have declined to levels
similar to each other. The rate for men ages 20-24 (10.9%) was slightly higher than the rate for
women of the same age (9.2%). The large disparities observed in April 2020 between younger
men and women were not observed in older age groups. The unemployment rate in May 2021 for
women ages 25 to 54 (4.9%) was lower than the rate for men (5.5%) in that age band, while the
rate for women ages 55 and over (5.0%) was slightly higher than that of men ages 55 and over
(4.6%).
Congressional Research Service

10

link to page 16

Unemployment Rates During the COVID-19 Pandemic

Figure 6. Unemployment Rates by Sex and Age
Seasonal y adjusted monthly data, January 2020 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: Every group experienced their peak unemployment rate in April 2020.
Unemployment Rates by Racial Group and Hispanic Ethnicity
As seen in Figure 7, the unemployment rates for Black, Asian, and White33 workers increased
sharply in early 2020. But whereas the unemployment rate for White workers peaked in April
2020, the rate for Black and Asian workers continued to rise through May 2020. The May 2021
rates for Black (9.1%), Asian (5.5%), and White (5.1%) workers were al higher than their
respective rates in January 2020. The rate for Black workers has declined 7.6 percentage points
since peaking in May 2020, compared to a decline of 9.4 percentage points for Asian workers and
9.0 percentage points for White workers across the same period.

33 Black, Asian, and White are the three racial categories used in BLS, T able A2: Employment status of the civilian
population by race, sex, and age. See https://www.bls.gov/news.release/empsit.t02.htm.
Congressional Research Service

11

link to page 16

Unemployment Rates During the COVID-19 Pandemic

Figure 7. Unemployment Rates by Racial Group
Seasonal y adjusted monthly data, January 2020 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: Black and Asian workers experienced their peak unemployment rate in May 2020. White workers peak
rate occurred in April 2020.
People of any race can identify as being either Hispanic or non-Hispanic in the CPS. The
unemployment rate for Hispanic workers rapidly increased by 13.7 percentage points to 18.5%
from February to April 2020. For non-Hispanic workers the unemployment rate increased by 10
points to 13.6%. As seen in Figure 8, Hispanic workers continue to experience elevated
unemployment rates. In May 2021, nonseasonal y adjusted unemployment rates experienced by
Hispanic (6.8%) and non-Hispanic (5.2%) workers were higher than those experienced prior to
the recession. While unemployment remains elevated compared to January 2020, these rates are
much lower than the peak exhibited in April 2020.
Figure 8. Unemployment Rates by Hispanic Origin
Non-seasonal y adjusted monthly data, January 2020 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Congressional Research Service

12

link to page 17

Unemployment Rates During the COVID-19 Pandemic

Notes: Statistical significance is not calculated for these trends, and it is unclear how dependent these trends are
on regular seasonal variation.
Unemployment Rates by Education
In general, workers with lower levels of educational attainment have higher rates of
unemployment. This pattern has been amplified during the current recession, as seen in Figure 9.
The unemployment rate for workers with less than a high school diploma peaked in April 2020
(21.0%), which was higher than the peak for those at al other education levels. The May 2021
rate for workers with less than a high school diploma (9.1%) was also higher than the rate for al
other education levels. Workers with a Bachelor’s degree or higher, the highest educational level
classified here, had the lowest peak unemployment rate (8.4% in April 2020) and the lowest May
2021 rate (3.2%) among al education levels.
Figure 9. Unemployment Rates by Education
Seasonal y adjusted monthly data, January 2020 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: Al groups experienced their peak unemployment rate in April 2020.
U.S. Labor Force Participation Rate:
Historical Trends
While the unemployment rate measures the prevalence of unemployment in the labor force, it
does not consider the state of the labor force itself. When persons stop looking for work, they exit
the labor force, decreasing the number of persons who are either working or actively looking for
Congressional Research Service

13

Unemployment Rates During the COVID-19 Pandemic

work. Such persons are not counted in the unemployment rate, by definition, but are an important
group to examine when evaluating unemployment. CRS therefore uses the labor force
participation rate to further contextualize unemployment rates observed during the COVID-19
pandemic. The labor force participation rate measures the percentage of noninstitutionalized
people ages 16 and older who are either looking for work, or working.34
Over much of the past two decades, the labor force participation rate has general y declined.
Following several years of modest growth through the early and mid-90s, labor force
participation rates started to plateau, hit a historical peak in April 2000 (67.3%), and then
declined. Labor force participation declined further following the Great Recession, before
stabilizing and steadily increasing starting in October 2015. This decline can be attributed to
several factors, although one prominent reason is the ongoing retirement of the baby boomer
generation.35 Despite the increase exhibited following the Great Recession, the labor force
participation rate in January 2020 (63.4%) prior to both the COVID-19 pandemic and economic
recession remained below its historical peak.
Between February 2020 and April 2020, the labor force participation rate exhibited an
unprecedented decline of 3.1 percentage points as 8.3 mil ion people left the labor force. The
participation rate partial y recovered between May 2020 and August of 2020 before stagnating. It
remains below the pre-recession rate (63.4%) in May 2021 (61.6%).
Figure 10. Historical Labor Force Participation Rate
Seasonal y adjusted monthly data, January 2020 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: Shaded regions indicate recessionary periods as identified by the National Bureau of Economic Research.

34 For definitions of the labor force, labor force participation rate, unemployment rate, and other relevant terms, see
https://www.bls.gov/cps/definitions.htm#lfpr.
35 See Michael Dotsey, Shigeru Fujita, and Leena Rudanko, Where is Everybody? The Shrinking Labor Force
Participation Rate, Federal Reserve Bank of Philadelphia, Economic Insights Vol 2. Issue 4, Philadelphia, PA, 2017,
pp. 17-24, https://www.philadelphiafed.org/-/media/frbp/assets/economy/articles/economic-insights/2017/q4/
eiq417.pdf.
Congressional Research Service

14

link to page 19 link to page 20

Unemployment Rates During the COVID-19 Pandemic

COVID-19 Recession: Trends in Labor Force
Participation
The COVID-19 pandemic affected the labor force participation rates in every major demographic
group. The analysis in this section compares the pre-recession (January 2020) labor force
participation rate to the current labor force participation rate, calculating the difference between
the two for each month between January 2020 and May 2021. Figure 11 shows the sharp decline
in the labor force participation rate for individuals ages 16 years and older between February
2020 and April 2020. During this period, 8.3 mil ion individuals left the labor force. The overal
rate recovered between May 2020 and August 2020 before stagnating. The labor force
participation rate in May 2021 remains 1.8 percentage points below its pre-recession level.
Figure 11. Labor Force Participation Rate During COVID-19 Pandemic
Seasonal y adjusted monthly data, January 2020 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: LFPR = Labor Force Participation Rate; ppt = percentage points. Changes in LFPR since January 2020
have not been tested for statistical significance.
Labor Force Participation Rate by Age and Sex
Figure 12 displays the change in the labor force participation rate since January 2020 by age and
sex. Between January 2020 and April 2020, every group experienced a decline in their labor force
participation rate. Women aged 16 to 19 (-7.0 percentage points) and men aged 20 to 24 (-9.2
percentage points) experienced the largest declines in labor force participation between January
2020 and April 2020. Men and women in the 25-54 and 55-and-older age groups experienced
smal er declines in labor force participation but have seen their recovery stagnate.
Congressional Research Service

15

link to page 21

Unemployment Rates During the COVID-19 Pandemic

Figure 12. Labor Force Participation Rates by Age and Sex
Seasonal y adjusted monthly data, January 2020 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: LFPR = Labor Force Participation Rate; ppt = percentage points. Changes in LFPR since January 2020
have not been tested for statistical significance.
Labor Force Participation Rate by Race and Ethnicity
Figure 13 displays the change in the labor force participation rate for the Black, Asian, and White
racial groups since January 2020. Between January 2020 and April 2020, each group experienced
a sharp decline in their labor force participation rate. In May 2021, the participation rate for each
group remained below its January 2020 value. In particular, participation rates for Black
individuals (-1.8 percentage points) and White individuals (-1.9 percentage points) remained wel
below their January 2020 values in May 2021.
Congressional Research Service

16

Unemployment Rates During the COVID-19 Pandemic

Figure 13. Labor Force Participation Rate by Race
Seasonal y adjusted monthly data, January 2020 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: LFPR = Labor Force Participation Rate; ppt = percentage points. Changes in LFPR since January 2020
have not been tested for statistical significance.
Individuals of any race can identify as Hispanic or non-Hispanic. The data on labor force
participation rate for these two ethnic groups are not seasonal y adjusted. Therefore, this report is
constrained to a comparison across values for January 2020 and the most recent month for which
the data are available. This comparison shows that Hispanic individuals had higher participation
rates in both January 2020 and May 2021 than non-Hispanic individuals, but the decline in labor
force participation over this period was greater for Hispanic persons. The labor force participation
rate for Non-Hispanic individuals in May 2021 was 1.3 percentage points below its value for
January 2020. For Hispanic individuals, the participation rate in May 2021 was 2.7 percentage
points below its January 2020 value.
Figure 14. Labor Force Participation Rate by Hispanic Origin
Non-seasonal y adjusted monthly data, January 2020 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Congressional Research Service

17

link to page 22

Unemployment Rates During the COVID-19 Pandemic

Notes: LFPR = Labor Force Participation Rate; ppt = percentage points. Statistical significance is not calculated
for these trends and it is unclear how dependent these trends are on regular seasonal variation. Changes in LFPR
since January 2020 have not been tested for statistical significance.
Labor Force Participation Rate by Educational Attainment
Figure 15 displays the difference in the labor force participation rate from its January 2020 level
for groups with different levels of educational attainment. Labor force participation fel for al
groups between January 2020 and April 2020. The largest decline was experienced by those with
a high school diploma but no college (-4.3 percentage points), which was followed by those with
less than a high school diploma (-3.3 percentage points). In May 2021, every group remained
below their labor force participation rate in January 2020. Those with less than a high school
diploma (-2.8 percentage points) and those with a high school diploma (-3.1 percentage points)
remained wel below their labor force participation rates in January 2020.
Figure 15. Labor Force Participation Rate by Educational Attainment
Seasonal y adjusted monthly data, January 2020 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: LFPR = Labor Force Participation Rate; ppt = percentage points. These data reflect the civilian
population that is 25 years and older. Changes in LFPR since January 2020 have not been tested for statistical
significance.
COVID-19 Recession: Nonfarm Payrolls
The number of nonfarm workers on payroll further contextualizes the relatively high
unemployment rates observed during the COVID-19 pandemic. Unemployment rates and labor
force participation rates do not measure the extent of job disruption occurring during the
pandemic. By comparing the number of jobs before the recession to recent data, CRS can
evaluate the extent of job loss and put unemployment rates into a broader economic context.
Congressional Research Service

18

link to page 23 link to page 24

Unemployment Rates During the COVID-19 Pandemic

Figure 16 displays the gap between the January 2020 and May 2021 employment levels for al
supersector industries in the North American Industry Classification System (NAICS).36 In May
2021, there were 7.3 mil ion fewer jobs than there were in January 2020. The largest portions of
this gap were made up by the leisure and hospitality (2.4 mil ion), education and health services
(1.0 mil ion), and government (1.0 mil ion) sectors. While losses were concentrated in a handful
of major sectors, employment gaps existed for al sectors in May 2021.
Figure 16. Change in Employment by Sector
Seasonal y adjusted data for May 2021; relative to January 2020

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: A single purple box reflects roughly 10,000 jobs. Data for May 2021 are preliminary. “Other services” is
a BLS aggregation of three subsectors: repair and maintenance, personal and laundry services, and membership
associations and organizations.
Trends in Employment in the Private Sector
Figure 17 displays the monthly change in seasonal y adjusted nonfarm jobs from January 2020 to
May 2021 for six major sectors of the private sector.37 The trough in job loss occurred in April
2020 for each sector in the figure. The leisure and hospitality sector experienced greater job
losses (-8.3 mil ion) than any other sector. Additional y, the leisure and hospitality sector had the
largest employment gap in May 2021 (-2.4 mil ion) compared to any other sector in that month.

36 Sectors are defined by the NAICS and can be found at https://www.bls.gov/iag/tgs/iag_index_naics.htm.
37 For the purpose of this visualization, CRS selected the six private sectors with the largest (i.e., most negative)
average losses in seasonally adjusted nonfarm jobs from January 2020 to May 2021.
Congressional Research Service

19

link to page 25

Unemployment Rates During the COVID-19 Pandemic

The trade, transportation, and utilities sector had the second most jobs lost (-3.4 mil ion),
followed by education and health services (-2.7 mil ion), professional and business services (-2.3
mil ion), other services38 (-1.4 mil ion), and manufacturing (-1.4 mil ion). In May 2021, al six of
these sectors had fewer jobs compared to their respective employment levels in January 2020.
Figure 17. Job Loss During the COVID-19 Recession in the Private Sector
Seasonal y adjusted monthly data, January 2020 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: M = mil ion. Data for April and May 2021 are preliminary. Changes in employment since January 2020
have not been tested for statistical significance.
Trends in Employment by Government Sector
Figure 18 displays the monthly change in seasonal y adjusted nonfarm jobs from January 2020 to
May 2021 for three levels of government employment: federal, state, and local. While the number
of federal government jobs increased during 2020 (peaking at +309,000 jobs in August),39 state
and local governments both experienced significant job losses. Local governments experienced
the largest job losses of any government level, peaking at 1.2 mil ion jobs lost in May. There were
911,000 fewer local government jobs in May 2021 than in January 2020. State government jobs

38 “Other services” is a BLS aggregation of three subsectors: repair and maintenance, personal and laundry services,
and membership associations and organizations. Sectors are defined by the North American Industry Classification
System (NAICS) and can be found at https://www.bls.gov/iag/tgs/iag_index_naics.htm.
39 T his peak could be due to temporary employment for those conducting activities related to the 2020 Decennial
Census.
Congressional Research Service

20

Unemployment Rates During the COVID-19 Pandemic

fel throughout 2020 and peaked in October (-340,000 jobs). In May 2021, state government
employment was 167,000 jobs below its January 2020 employment level.
Figure 18. Job Loss During the COVID-19 Pandemic in the Public Sector
Seasonal y adjusted monthly data, January 2020 to May 2021

Source: Created by CRS using data from the Bureau of Labor Statistics (BLS). Extracted using the Labor Force
Statistics data series at https://www.bls.gov/data/.
Notes: M = mil ion. Data for April and May 2021 are preliminary. Changes in employment since January 2020
have not been tested for statistical significance.
Data Limitations and Caveats
National level data presented in this report are from the Current Population Survey (CPS) or
Current Employment Statistics (CES) survey and state level data are from the Local Area
Unemployment Statistics (LAUS) program. The CPS is a sample survey of about 60,000
households conducted by the Census Bureau for BLS. The CES is a sample survey of about
144,000 business and government agencies conducted by BLS. LAUS is a BLS program that
calculates state-level unemployment rates using multiple data sources, including the CPS and
CES.40
Estimates from al three sources are subject to sampling and nonsampling error.41 Sampling error
occurs when the survey sample is not representative of the underlying population, while
nonsampling error describes errors often associated with data collection.42 Sampling error is a
result of statistical theory that underlies any estimate generated through surveys. While the CPS
and CES samples are selected to be representative of the nation, the possibility remains that it
does not accurately estimate certain nationwide statistics.43 Nonsampling error refers to al

40 In addition to the CPS, LAUS uses the Current Employment Statistics survey, state Unemployment Insurance claims
counts, the Quarterly Census of Employment and Wages program, and data from the Census Bureau’s American
Community Survey and Population Estimates Program; https://www.bls.gov/lau/laumthd.htm.
41 For further discussion of error, see the “Reliability of the Estimates” section of the Employment Situation report’s
T echnical Note at https://www.bls.gov/news.release/empsit.tn.htm. For a description of LAUS estimation procedures,
see https://www.bls.gov/lau/laumthd.htm.
42 For more information, see https://www.bls.gov/opub/hom/topic/error-measurements.htm.
43 For more information, see https://www.bls.gov/opub/hom/topic/sampling.htm.
Congressional Research Service

21

Unemployment Rates During the COVID-19 Pandemic

sources of error that are not due to sampling. They can result from incorrect or biased collection
and processing of the data. For example, nonsampling error can occur if a surveyor incorrectly
records responses or a respondent incorrectly responds to a question.
COVID 19 Pandemic-Related Data Issues
The COVID-19 pandemic increased nonsampling error in the CPS and CES due to a number of
factors. For example, BLS reported that both surveys experienced lower response rates.44 BLS
additional y noted that business closures initial y interfered with the ability for businesses to
respond to CES inquiries. (The bureau has since made statements affirming the robustness of its
CPS and CES estimates despite these lower response rates.45) Furthermore, BLS detected an error
in its CPS categorization procedures that likely underestimated unemployment early in the
recession.46 Specifical y, large numbers of workers were classified as employed but not at work
when they should have been recorded as unemployed on temporary layoff.
Per agency policy, BLS did not adjust CPS records, but it did provide adjusted estimates of the
unemployment rate. This report does not use these adjusted estimates as they are not official BLS
estimates. BLS estimated that its categorization error underestimated seasonal y adjusted
unemployment by roughly 0.9 percentage points in March 2020, 4.8 points in April, 3.1 in May,
1.2 in June, 0.9 in July, 0.7 in August, 0.4 in September, 0.3 in October, 0.4 in November, 0.6 in
December, 0.6 in January 2021, 0.5 in February 2021, 0.4 in March 2021, and 0.3 in April 2021.47
In May 2021, the error underestimated seasonal y adjusted unemployment by an estimated 0.3
percentage points.48 These estimates evaluate what the impact would be in the worst-case
scenario, as the true impact is uncertain. BLS released a statement regarding the underestimate,
noting that, “these assumptions probably overstate the size of the misclassification error.” In later
months, BLS made efforts to correct this classification error during data collection and
processing.49
Additional y, BLS recently identified a data processing error in the CES, which began in July
2020 but remained undetected. The error unintentional y caused some businesses to be
inappropriately included in the sample and used for estimates. BLS has since determined the
impacts of this error were insignificant.50
LAUS was impacted by both the low response rate and the categorization error due to its
connection with the CPS and CES. Considering that LAUS is dependent on a number of other
data sources that were impacted by COVID-19 in their own right, the net effect of the pandemic
on LAUS estimates is unknown.51

44 See the FAQ BLS produced on this topic for more on the impact of COVID-19 on data collection by month at
https://www.bls.gov/covid19/home.htm.
45 See https://www.bls.gov/covid19/employment-situation-covid19-faq-april-2020.htm.
46 See CRS Insight IN11456, COVID-19: Measuring Unemployment, by Lida R. Weinstock.
47 See https://www.bls.gov/covid19/effects-of-covid-19-pandemic-and-response-on-the-employment-situation-news-
release.htm.
48 See https://www.bls.gov/covid19/employment-situation-covid19-faq-may-2021.htm.
49 Among other protocols, the Census Bureau monitored survey responses in August and marked those they felt could
be misclassified. T hese responses were then re-evaluated. For more on BLS and Census efforts to reduce the
misclassification, see https://www.bls.gov/covid19/employment -situation-covid19-faq-august-2020.htm#ques9.
50 For the BLS notice on this error, see https://www.bls.gov/ces/notices/2021/ces-sample-rotation-issue-caused-by-
pandemic-related-challenges-to-enrollment.htm.
51 For more on the impacts of COVID-19 on LAUS and its inputs, see https://www.bls.gov/covid19/effects-of-covid-
Congressional Research Service

22

Unemployment Rates During the COVID-19 Pandemic

General Data Caveats
Other data considerations include the following:
 Lack of seasonally adjusted data: Seasonal y adjusted data are published by BLS
for selected labor force indicators to better account for seasonality in the trends.52
Without seasonal adjustments, it is difficult to distinguish between trends related
to the recession and seasonal trends. Where adjusted data are not available, this
report presents unadjusted data that do not account for underlying seasonal
variation.
 Reference week: In general, CPS data are collected for the calendar week
containing the 12th of the month. CES data are collected for the pay period
including the 12th of the month. This could lead to incongruity between actual
labor force conditions over the course of a month and the conditions observed, as
wel as inconsistencies between CES-reported data and CPS-reported data.
 CPS and LAUS unemployment rate comparability: While the LAUS program
uses the same unemployment concepts as the CPS and uses the CPS as an input,
LAUS estimates are based on multiple sources (including administrative data).
Consequently, CPS and LAUS estimates are not directly comparable.

Author Information

Gene Falk, Coordinator
Isaac A. Nicchitta
Specialist in Social Policy
Research Assistant

Paul D. Romero
Emma C. Nyhof
Research Assistant
Research Assistant

Jameson A. Carter

Research Assistant

Acknowledgments
The four Research Assistants in CRS’s Domestic Social Policy Division were responsible for the analysis
and writing of this report, under the guidance of Gene Falk, Specialist in Social Policy. Questions from
congressional staff should be directed to Mr. Falk.

19-pandemic-on-employment-and-unemployment-statistics.htm.
52 See CPS and LAUS documentation for more on seasonal adjustment at https://www.bls.gov/cps/seasonal-
adjustment -methodology.htm and https://www.bls.gov/lau/lauseas.htm.
Congressional Research Service

23

Unemployment Rates During the COVID-19 Pandemic

Disclaimer
This document was prepared by the Congressional Research Service (CRS). CRS serves as nonpartisan
shared staff to congressional committees and Members of Congress. It operates solely at the behest of and
under the direction of Congress. Information in a CRS Report should n ot be relied upon for purposes other
than public understanding of information that has been provided by CRS to Members of Congress in
connection with CRS’s institutional role. CRS Reports, as a work of the United States Government, are not
subject to copyright protection in the United States. Any CRS Report may be reproduced and distributed in
its entirety without permission from CRS. However, as a CRS Report may include copyrighted images or
material from a third party, you may need to obtain the permission of the copyright holder if you wish to
copy or otherwise use copyrighted material.

Congressional Research Service
R46554 · VERSION 15 · UPDATED
24