View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

U.S. BUREAU OF LABOR STATISTICS
Bureau of Labor Statistics

Publications

Monthly Labor Review

HOME
Search MLR

ARCHIVES

FOR AUTHORS

ABOUT

SUBSCRIBE

GO

BOOK REVIEW

April 2 0 2 3

Fiscal policies as a response to recessions
Recession Ready: Fiscal Policies to Stabilize the American Economy. Edited by Heather Boushey, Ryan Nunn, and Jay Shambaugh. Washington, DC: The Brookings
Institution, Hamilton Project, 2019, 250 pp., download.
The National Bureau of Economic Research has documented 34 business cycles from 1854 to 2020 in the United States. The downturn phases of these business cycles are
usually characterized as recessions. During recessionary periods, declines in investment and employment are common. Recession Ready: Fiscal Policies to Stabilize the
American Economy is a recent addition to the literature on the effectiveness of fiscal policies in minimizing the impacts of economic downturns. Edited by Heather Boushey,
Ryan Nunn, and Jay Shambaugh, this book, a collection of papers penned by economists, focuses on two main topics. First, it documents the impacts of recessions and the
effectiveness of antirecessionary fiscal policies. Second, it presents several automatic stabilizers that could dampen the effects and length of recessions. Automatic stabilizers
are fiscal policies that automatically respond to macroeconomic fluctuations, such as declines in tax revenue during an economic downturn.
In the first chapter of the book, Boushey, Nunn, Jimmy O’Donnell, and Shambaugh review the economic impacts of recessions and the effectiveness of past fiscal responses.
The authors empirically show that recessions reduce gross domestic product (GDP), increase unemployment and underemployment, increase the probability of people leaving
the labor force, diminish the job prospects of recent college graduates, and reduce private and public investment. The authors estimate the effectiveness of past fiscal responses
to economic downturns, concluding that these responses have been slow and too short to counter the impacts of recessions.
In the second chapter, Louise Sheiner and Michael Ng find that federal spending is usually countercyclical, whereas state-level spending is usually procyclical (because many
states require their budgets to remain balanced). This finding leads the authors to favor federal-level fiscal spending over state-level spending. Automatic stabilizers are
considered an effective response to recessions because they quickly match the timing and duration of spending to recession indicators. Also, automatic stabilizers can target
those who are the most adversely affected by economic downturns, such as the unemployed and individuals with low incomes.
In the third chapter, Claudia Sahm proposes an automatic stabilizer involving lump-sum payments to all individuals, regardless of their income. These payments would be
triggered by a 0.5-percentage-point increase in the unemployment rate, which is an increase observed only around recessions. Sahm reviews the literature on the effectiveness
of lump-sum payments during past recessions, finding that such payments are spent more quickly than payroll deductions or multiple payments. The author also finds that the
amount of these payments should total 0.7 percent of GDP, which is approximately half the average spending decrease during past recessions. Such lump-sum payments
should be repeated each year in which the unemployment rate is at least 2.0 percentage points above the trigger level.
In the fourth chapter, Matthew Fiedler, Jason Furman, and Wilson Powell argue that federal spending on Medicaid and the Children’s Health Insurance Program (CHIP)
should automatically increase when a state’s unemployment rate reaches a predetermined level. They recommend this funding increase because of the negative social effects
of reduced state-level spending during recessions. For example, reduced education spending by states during recessions has hurt student achievement. Besides assessing the
direct effects of healthcare spending, the authors estimate that, over the long term, an automatic increase in spending on Medicaid and CHIP would increase GDP by 0.12
percent and reduce the unemployment rate by 0.1 percentage point.
In the fifth chapter, Andrew Haughwout proposes road-repair spending as a countercyclical stabilizer. The author documents this spending as a long-term investment that
provides service flows to individuals and businesses. Usually, spending on road repair is seen as a poor countercyclical measure because infrastructure projects involve
lengthy planning. According to Haughwout, to make infrastructure spending a timely response to recessions, states should consider developing a catalogue of planned roadrepair projects that would be federally financed. This planning would ensure quick project completion, making the spending more effective as a countercyclical policy.
In the sixth chapter, Gabriel Chodorow-Reich and John Coglianese propose an updated and expanded unemployment insurance (UI) program that uses automatic triggers
based on unemployment levels. The authors argue that such a program would better serve as a countercyclical antirecessionary policy. Because people who receive UI benefits
are usually in need of transfers, unemployed people are likely to spend these transfers relatively quickly, which would provide a countercyclical effect. Chodorow-Reich and
Coglianese’s recommendations include expanding benefits eligibility, encouraging the filing of more applications, extending benefits automatically during spells of high
unemployment, and increasing benefit amounts during recessions (which would increase the use of UI as a macroeconomic stabilizer).
In the seventh chapter, Indivar Dutta-Gupta proposes changes to the Temporary Assistance for Needy Families (TANF) program that would increase the program’s
effectiveness during recessions. The author argues for increases in both basic assistance and the number of subsidized jobs with supportive services. Under the latter
component of the proposal, unemployed individuals would be prepared for and placed in jobs in which their wages are partially or fully subsidized by federal spending. In
addition, these individuals would receive services, such as childcare and transportation assistance, that would increase their odds of becoming and staying employed. DuttaGupta finds that these changes to TANF would strengthen the program’s antipoverty and countercyclical effects.
In the eighth chapter, Hilary Hoynes and Diane Whitmore Schanzenbach argue that the Supplemental Nutrition Assistance Program (SNAP) can be altered to become a more
effective automatic stabilizer. The authors show that SNAP has reduced financial hardship and poverty and improved health outcomes, children’s educational attainment, and
children’s economic outcomes during adulthood. SNAP benefits have a fast fiscal effect, as 97 percent of benefits are spent within a month of being received. Hoynes and
Schanzenbach argue that increasing SNAP benefit levels and waiving SNAP work requirements during recessions will make the program a better automatic stabilizer.
Recession Ready is an accessible book focusing both on the past effects of recessions and on the policy responses to those effects. Readers new to these topics will find the
book to be a good starting place for their research. Professional macroeconomists will benefit from the book’s review of current research. Also, the book would be a useful

reading in courses on the business cycle or fiscal policy. Despite these strengths, Recession Ready does not discuss some policies that could also be used as automatic
stabilizers, such as work sharing proposals and public service employment.

ABOUT THE REVIEWER

Justin Holt
holt.justin@bls.gov
Justin Holt is an economist in the Office of Publications and Special Studies, U.S. Bureau of Labor Statistics.

U.S. BUREAU OF LABOR STATISTICS Division of Information and Marketing Services PSB Suite 2850 2 Massachusetts Avenue NE
Washington, DC 20212-0001
Telephone:1-202-691-5200 Telecommunications Relay Service:7-1-1 www.bls.gov/OPUB Contact Us

U.S. BUREAU OF LABOR STATISTICS
Bureau of Labor Statistics

Publications

Monthly Labor Review

HOME
Search MLR

ARCHIVES

FOR AUTHORS

ABOUT

SUBSCRIBE

GO

ARTICLE

April 2 0 2 3

Occupational projections overview, 2021–31
The Employment Projections program of the U.S. Bureau of Labor Statistics develops information about the labor market for the nation as a whole for 10 years in the future.
This article provides an overview of each occupational group, including projected employment change from 2021 to 2031, information about factors contributing to projected
employment change, information on median annual wage, and typical education or on-the-job-training requirements needed for occupational entry.
The Employment Projections (EP) program of the U.S. Bureau of Labor Statistics annually projects employment over a 10-year period for over 800 detailed occupations and
nearly 300 industries. Employment in the United States is projected to increase 5.3 percent during the 2021–31 decade, adding about 8.3 million new jobs.1 These projections
form the basis for data and outlook information in the Occupational Outlook Handbook (OOH).2 The EP data and the OOH are used by a wide variety of people, including
jobseekers, career counselors, education and training officials, and researchers.
This article is organized by 24 occupational groups,3 which highlight several of the detailed occupations that are projected to grow the fastest or projected to decline the
fastest. Additional information for these occupations may be found within the OOH. This article also illustrates common trends and factors within occupational groups and
across groups.
Table 1 displays the projected employment change from 2021 to 2031, sorted by percentage of total new jobs added. This table and all subsequent tables provide information
about employment change in two ways: numeric change and percent change. This is important to note because a fast rate of employment growth does not always translate into
many new jobs. For example, the math occupational group is projected to grow 28.7 percent from 2021 to 2031, the fastest of any occupational group. However, because of
this occupational group’s relatively small size, this percent growth accounts for only about 82,000 new jobs over the projections decade. In contrast, the healthcare
occupational group is projected to contribute the most new jobs of any group and projected to grow 12.6 percent. Because of its large size, the healthcare occupational group is
projected to add over 2 million new jobs over the decade. That is, the healthcare occupational group is projected to add more new jobs even though its growth rate is lower
than that of the math occupational group.

Table 1. Employment of OOH occupational groups, 2021 and projected 2031
Employment

Employment change (2021–31)

OOH occupational group

Percent of total new jobs projected to be added
2021

2031

Number

Percent

158,134.7

166,452.1

8,317.4

5.3

[1]

Healthcare

16,254.2

18,303.2

2,049.1

12.6

24.6

Food preparation and serving

11,761.8

13,081.6

1,319.9

11.2

15.9

Management

11,685.3

12,569.2

883.9

7.6

10.6

Transportation and material moving

13,350.7

14,212.6

861.8

6.5

10.4

Business and financial

9,987.4

10,702.5

715.1

7.2

8.6

Computer and information technology

4,665.2

5,348.0

682.8

14.6

8.2

Education, training, and library

9,151.2

9,809.3

658.2

7.2

7.9

Personal care and service

3,868.4

4,413.2

544.8

14.1

6.6

Installation, maintenance, and repair

6,038.7

6,342.6

304.0

5.0

3.7

Community and social service

2,843.2

3,137.8

294.6

10.4

3.5

Building and grounds cleaning

5,415.0

5,705.8

290.8

5.4

3.5

Construction and extraction

7,026.0

7,278.9

252.9

3.6

3.0

Legal

1,368.0

1,499.0

131.0

9.6

1.6

Life, physical, and social science

1,436.0

1,534.7

98.7

6.9

1.2

758.3

853.8

95.5

12.6

1.1

2,562.5

2,653.7

91.3

3.6

1.1

286.3

368.3

82.0

28.7

1.0

Protective service

3,482.2

3,554.8

72.6

2.1

0.9

Media and communication

1,111.9

1,180.5

68.6

6.2

0.8

918.8

939.4

20.5

2.2

0.2

Farming, fishing, and forestry

1,069.6

1,078.0

8.4

0.8

0.1

Production

8,787.1

8,623.5

-163.6

-1.9

-2.0

Sales

14,719.9

14,555.4

-164.5

-1.1

-2.0

Office and administrative support

19,587.0

18,706.2

-880.8

-4.5

-10.6

Total, all occupations

Entertainment and sports
Architecture and engineering
Math

Arts and design

[1] This entry is not applicable.
Note: Employment numbers are in thousands. Details may not sum to totals because of rounding.
OOH is the Occupational Outlook Handbook.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Overview
Of the 8.3 million new jobs projected to be added by 2031, almost one-quarter will be within the healthcare occupational group. The five occupational groups projected to add
the most new jobs from 2021–31 contribute about 70 percent of the 8.3 million new jobs. These five occupational groups include the following: healthcare occupations, food
preparation and serving occupations, management occupations, transportation and material moving occupations, and business and financial occupations. (See chart 1.)

Chart 1. Projected number of new jobs to be added from
2021–31, by OOH occupational group
Healthcare
Food preparation and serving
Management
Transportation and material moving
Business and financial
Computer and information technology
Education, training, and library
Personal care and service
All other occupation groups

Click legend items to change data display. Hover over chart to view data.
Note: OOH is the Occupational Outlook Handbook.
Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

View Chart Data

The concentrated nature of the employment growth lends itself to some key takeaways. Healthcare occupations are projected to add the most new jobs of any of the
occupational groups, contributing over 2.0 million new jobs to the total 8.3 million jobs during the projections period. This projected growth is mainly due to a growing
population, whose rising share of older people with chronic conditions is expected to lead to greater demand for healthcare services.
The growth in some industries synergizes with growth in other industries. Business growth and expansion (especially in healthcare, information technology (IT), and ecommerce) will drive demand for services provided by management occupations, transportation and material moving occupations, and business and financial service
occupations. These three occupational groups, combined, account for almost 30 percent of all new jobs expected to be added to 2031. The sectors that are projected to grow
faster than average, such as healthcare and IT, will consequently result in demand for managers in those areas. Continued growth of e-commerce should increase demand for
transportation and warehousing, supporting demand for package delivery services and material movers. The growth in digital marketing and e-commerce will support demand
for business and finance occupations that manage activities such as logistics, marketing research, and accounting.
Likewise, demand for the services provided by computer and information technology occupations and math occupations will stem from greater emphasis on collecting and
analyzing data, continuing growth in the digital economy, and an increasing need for information security. These two occupational groups, combined, account for a little over
9 percent of all new jobs to be added to 2031, and they are projected to be the two fastest growing groups. Overall, employment in math occupations is projected to grow 28.7
percent, the fastest of any group. Also, employment in computer and information technology occupations is projected to grow 14.6 percent from 2021 to 2031.
In contrast, significant employment gains for some occupational groups are primarily rebounds from the COVID-19 pandemic.4 Food preparation and serving occupations and
personal care and service occupations, combined, are projected to add almost 1.9 million new jobs, contributing about 22 percent of the total 8.3 million jobs projected over
the 2021–31 projections cycle. However, a large part of the projected growth for these two groups represents a recovery from the low employment level in 2021 because of the
lingering effects of the COVID-19 pandemic.
Not all occupational groups are expected to grow. Production occupations, sales occupations, and office and administrative support occupations are three occupational groups
expected to experience a decline in employment over the projections decade. Changes in technology, such as machines or software use that increases productivity or replaces
workers altogether, are expected to contribute to a decline in employment and suppress job openings.5 Nevertheless, the need to replace workers who change occupations or
leave the labor force is expected to create some job openings, even in occupations with projected employment declines.

Occupational groups analysis
The occupational groups in this article encompass every civilian job in the United States and can be broken down into individual occupations. Of these, more than 500 detailed
occupations in over 300 occupational profiles are covered in the OOH, accounting for about 4 out of 5 jobs in the economy. The OOH includes information on job outlook, job
descriptions, entry-level education, training information, and wage data.6
Occupations can be grouped by similar duties or purposes; for example, protective service occupations include police and sheriff’s patrol officers, security guards, and
correctional officers and jailers. Examining the growth of occupational groups reveals the key factors affecting employment over the projections period.
Projected employment information for each of the OOH occupational groups is outlined below, including information about factors that may be contributing to the projected
employment change at the group level. In this article, occupational groups are presented in order of the percentage of total new jobs projected to be added over the projections
period from the occupational groups projected to add the most new jobs to the occupational groups projected to lose jobs over the projections decade. Detailed occupations are
highlighted in each group section with the fastest growing occupations or the fastest declining occupations. The median annual wage and typical education needed for entry
into these detailed occupations are provided. Links to the OOH pages are provided for additional information on what they do, work environment, and a list of similar
occupations, among other information.

Healthcare occupational group

Overall employment in healthcare occupations is projected to grow 12.6 percent from 2021 to 2031, much faster than the average for all occupations; this increase is expected
to result in about 2.0 million new jobs over the decade, the most of any group.7 Healthcare workers will be needed to assist a growing number of older Americans stay healthy
and active and to provide services to those with chronic conditions, such as diabetes. Five of the top 30 fastest growing occupations are detailed occupations within the
healthcare occupational group: nurse practitioners, physician assistants, physical therapist assistants, home health and personal care aides, and occupational therapy assistants.
(See appendix A-1.)

Table 2. Top five fastest growing occupations within healthcare occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021
Total, all occupations
Healthcare

00-0000
29-0000 and
31-0000

2031

Number

Median annual wage, 2021 [1] Typical education needed for entry

Percent

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

16,254.2

18,303.2

2,049.1

12.6

47,070

[2]

Nurse practitioners

29-1171

246.7

359.4

112.7

45.7

120,680

Master's degree

Physician assistants

29-1071

139.1

177.5

38.4

27.6

121,530

Master's degree

Physical therapist assistants

31-2021

96.5

122.1

25.6

26.5

61,180

Associate's degree

Home health and personal care aides

31-1120

3,636.9

4,560.9

924.0

25.4

29,430

High school diploma or equivalent

Occupational therapy assistants

31-2011

43.4

54.5

11.0

25.4

61,730

Associate's degree

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.

[2] This entry is not applicable.
Note: Employment numbers are in thousands. Details may not sum to totals because of rounding.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Nurse practitioners are projected to experience the fastest employment growth of all occupations, with projected growth of 45.7 percent. Healthcare facilities are increasingly
using team-based healthcare models, which employ nurse practitioners, physician assistants, and other healthcare practitioners to provide patient care that would otherwise be
provided by a doctor. Many of the fastest growing occupations in healthcare work closely with patients to help them maintain or improve their quality of life. Physical
therapist assistants assist physical therapists, particularly in long-term care environments; physical therapist assistants have a projected employment growth of 26.5 percent
(much faster than the average for all occupations and the fastest growing occupation within the healthcare support occupational group). Occupational therapy assistants will be
needed to help occupational therapists in caring for patients with conditions and ailments, such as arthritis and strokes, that may affect their ability to do everyday activities.
Employment of home health and personal care aides is projected to grow by 25.4 percent and add about 924,000 jobs. Both the fast growth of the elderly population and their
desire to live in their own homes are expected to underpin demand for more in-home assistance. Demand for assistance and care for individuals in retirement communities,
assisted-living facilities, nursing homes, and other facilities is expected to contribute to the overall growth of this occupation.
Healthcare occupations had a median annual wage of $47,070 in May 2021, but wages vary widely; some healthcare occupations are among the highest paying, while others
have wages below the median annual wage. Wages are generally correlated with education as occupations with higher levels of typical entry-level education usually pay more.
Many of the occupations within this group require on-the-job training, internship, or residency experience.
Food preparation and serving occupational group

Overall employment in food preparation and serving occupations is projected to grow 11.2 percent from 2021 to 2031, much faster than the average for all occupations; this
increase is expected to result in about 1.3 million new jobs over the decade.8 Food preparation and serving occupations are projected to contribute 15.9 percent of the new jobs
added over the 2021–31 period. This mostly reflects job recovery from the COVID-19 recession of 2020 because pandemic restrictions had significant effects on employment
levels in restaurants. Only one of the top 30 fastest growing occupations projected from 2021 to 2031 is a detailed occupation within the food preparation and serving
occupational group, namely, cooks, restaurants. (See appendix A-1.)

Table 3. Top five fastest growing occupations within food preparation and serving occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Median annual wage, 2021

Typical education needed for

[1]

entry

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Food preparation and serving

35-0000

11,761.8

13,081.6

1,319.9

11.2

28,400

[2]

Cooks, restaurant

35-2014

1,255.6

1,715.6

459.9

36.6

30,010

No formal education credentials

Bartenders

35-3011

514.0

606.0

92.0

17.9

26,350

No formal education credentials

Dining room and cafeteria attendants and bartender
helpers

35-9011

355.2

415.1

59.9

16.9

27,170

No formal education credentials

Chefs and head cooks

35-1011

152.8

176.3

23.6

15.4

50,160

High school diploma or equivalent

35-9031

347.7

400.3

52.6

15.1

24,600

No formal education credentials

Hosts and hostesses, restaurant, lounge, and coffee
shop

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands. Details may not sum to totals because of rounding.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Some employment lost in the food preparation and serving occupations during the pandemic and projected to be recovered over the projections decade has already been
recuperated as employment grew rapidly throughout the first half of 2022.9 Many food service establishments, restaurants, school cafeterias, and food contractors for
businesses have already reopened and reemployed workers as consumer food spending patterns returned to their prepandemic trends.
In addition to the immediate recovery discussed above, new restaurant openings and expanded food deliveries will contribute to increases in demand for restaurant food,
supporting demand for the services provided by the food preparation and serving occupations. Cooks at restaurants are expected to add the most new jobs within this group as
consumers demand more high-quality food from restaurants, contributing about 459,900 new jobs over the projections decade.
The food preparation and serving occupations group is the lowest paid major group, with a median annual wage of $28,400 in May 2021. Most food preparation and serving
occupations require on-the-job training, typically lasting up to 30 days; however, typically no education credentials are needed for entry.
Management occupational group

Overall employment in management occupations is projected to grow 7.6 percent from 2021 to 2031, faster than the average for all occupations; this increase is expected to
result in about 883,900 new jobs over the decade.10 Only one of the top 30 fastest growing occupations projected from 2021 to 2031 is a detailed occupation within the
management occupational group, namely, medical and health services managers. (See appendix A-1.)

Table 4. Top five fastest growing occupations within management occupations, 2021 and projected 2031
Change

Employment
2021

Median annual wage,

(2021–31)

Occupation
2031

Number

Percent

2021 [1]

Typical education needed for entry

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Management

11-0000

11,685.3

12,569.2

883.9

7.6

102,450

[2]

Medical and health services managers

11-9111

480.7

616.9

136.2

28.3

101,340

Bachelor's degree

Lodging managers

11-9081

51.2

60.4

9.2

18.0

59,430

Financial managers

11-3031

730.8

854.0

123.1

16.8

131,710

Bachelor's degree

Entertainment and recreation managers, except
gambling

11-9072

21.6

25.2

3.6

16.5

62,000

Bachelor's degree

Computer and information systems managers

11-3021

509.1

591.5

82.4

16.2

159,010

Bachelor's degree

High school diploma
or equivalent

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

The projected employment change for management occupations varies depending on the demand for the services they provide and the need for the supervision of workers. For
example, as the demand for healthcare services increases, medical and health services managers will be needed to support this demand, driving employment growth of 28.3
percent from 2021 to 2031. Similarly, the employment of computer and information systems managers is projected to grow 16.2 percent over the projections period as the
need for IT services and enhanced security requirements continue to grow.
Both lodging managers and entertainment and recreation managers, except gambling, will also see fast employment growth. However, much of this projected growth will be
due to recovery from the COVID-19 recession. The return to prepandemic travel patterns will result in strong demand for lodging managers in hotels and other lodging
establishments at the beginning of the projections decade.

The median annual wage for this group was $102,450 in May 2021, which was the highest among the major occupational groups. A bachelor’s degree is the required level of
education for most jobs within this group. However, requirements vary from a high school diploma to a master’s degree across the management occupations. In addition to
postsecondary education, most occupations in this group require work experience in a related occupation. For example, financial managers require years of work experience in
a related occupation for entry.
Transportation and material moving occupational group

Overall employment in transportation and material moving occupations is projected to grow 6.5 percent from 2021 to 2031, about as fast as the average for all occupations;
this increase is expected to result in about 861,800 new jobs over the decade.11 The economy depends on transportation and material moving workers to transport freight and
passengers and keep supply chains moving. Expected growth in e-commerce will drive demand for transportation and deliveries. Taxi drivers is the only detailed occupation
from the transportation and material moving group among the top 30 fastest growing occupations projected to grow the fastest from 2021 to 2031. (See appendix A-1.)

Table 5. Top five fastest growing occupations within transportation and material moving occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Median annual wage, 2021 [1] Typical education needed for entry

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Transportation and material moving

53-0000

13,350.7

14,212.6

861.8

6.5

36,860

[2]

Taxi Drivers

53-3054

128.5

165.1

36.6

28.5

29,310

No formal education credentials

Flight attendants

53-2031

106.3

128.4

22.1

20.8

61,640

High school diploma or equivalent

Shuttle drivers and chauffeurs

53-3053

189.5

215.2

25.8

13.6

30,000

No formal education credentials

Driver/sales workers

53-3031

531.0

594.5

63.5

12.0

29,280

High school diploma or equivalent

Pump operators, except wellhead pumpers

53-7072

11.0

12.3

1.3

11.4

49,580

High school diploma or equivalent

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Some of the occupations within this group, such as taxi drivers, flight attendants, and shuttle drivers and chauffeurs, were affected by the COVID-19 pandemic; these
occupations are expected to experience much faster than average growth over the projections decade. They are expected to recover lost employment from the recession of
2020 as the general population returns to prepandemic travel patterns.
The general demand for delivery options is expected to increase, and the services provided by driver/sales workers are projected to grow as these workers may be needed to
deliver items such as food and medical supplies.
The median annual wage for this group was $36,860 in May 2021, which was lower than the median annual wage for all occupations of $45,760. Education requirements for
this group range from no education credential to a postsecondary nondegree award, but a high school diploma is generally the level of education needed for entry. Some form
of on-the-job training is needed to attain competency in most of the occupations within the transportation and material moving group.
Business and financial occupational group

Overall employment in business and financial occupations is projected to grow 7.2 percent from 2021 to 2031, about as fast as the average for all occupations; this increase is
expected to result in about 715,100 new jobs over the decade.12 Continued domestic and international business operations, along with a complex tax and regulatory
environment, are expected to create demand for a variety of business and financial services, including accounting, consulting, and investment advisory services. In addition,
increasing efforts to understand customers behavior and product demand and to evaluate marketing strategies will lead to growing demand for market research. Only one of
the top 30 fastest growing occupations projected from 2021 to 2031 is a detailed occupation within the business and financial occupational group, namely, logisticians. (See
appendix A-1.)

Table 6. Top five fastest growing occupations within business and financial occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Median annual wage, 2021

Typical education needed for

[1]

entry

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Business and financial

13-0000

9,987.4

10,702.5

715.1

7.2

76,570

[2]

Logisticians

13-1081

195.0

249.1

54.1

27.7

77,030

Bachelor’s degree

Farm labor contractors

13-1074

1.2

1.5

0.3

22.3

47,770

No formal education credentials

Financial examiners

13-2061

62.8

76.0

13.2

21.0

81,410

Bachelor’s degree

13-1161

792.5

942.8

150.3

19.0

63,920

Bachelor’s degree

13-1121

128.2

151.1

22.9

17.8

49,470

Bachelor’s degree

Market research analysts and marketing
specialists
Meeting, convention, and event planners

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Logisticians are expected to be in high demand; their employment is projected to grow 27.7 percent from 2021 to 2031, much faster than the average for all occupations. As
the growth of e-commerce makes logistics more dynamic and complex, logisticians will be needed to manage multiple supply chains and oversee purchasing, transportation,
inventory, and warehousing activities. The increasing use of data and market research across many industries will support the demand for market research analysts and
marketing specialists; employment in these occupations is projected to grow 19.0 percent over the projections decade.
Employment of farm labor contractors is projected to grow 22.3 percent from 2021 to 2031, much faster than the average for all occupations, as farms seek the assistance of
contractors in recruiting and hiring seasonal and temporary farmworkers. Employment of financial examiners is projected to grow 21.0 percent from 2021 to 2031 as the
services they provide are needed to help navigate the regulatory environment and reduce the cost of compliance.
The median annual wage for this group was $76,570 in May 2021, which was higher than the median annual wage for all occupations of $45,760. Most occupations in this
group require a bachelor’s degree and many require some form of on-the-job training. Farm labor contractors represent the only occupation in this group requiring no
education credentials for entry; however, farm labor contractors do require short-term on-the-job training.
Computer and information technology occupational group

Overall employment in computer and information technology occupations is projected to grow 14.6 percent from 2021 to 2031, much faster than the average for all
occupations; this increase is expected to result in about 682,800 new jobs over the decade.13 Before the onset of the COVID-19 pandemic in 2020, many computer and IT jobs
were already projected to be in high demand over the next decade, growing much faster than average. The pandemic only served to make IT workers even more important to
the future economy.14 Strong demand for IT security, software development, and new products and services associated with the Internet of Things (IoT) continue to drive
demand for the services provided by these computer and IT occupations. Three of the top 30 fastest growing occupations projected from 2021 to 2031 are detailed occupations
within the computer and information technology occupational group: information security analysts, web developers, and software developers. (See appendix A-1.)
Table 7. Top five fastest growing occupations within computer and information technology occupations, 2021 and projected 2031
Change

Employment
2021

Median annual wage, 2021

(2021–31)

Occupation
2031

Number

Typical education needed for entry

[1]

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Computer and information technology

15-1200

4,665.2

5,348.0

682.8

14.6

97,430

[2]

Information security analysts

15-1212

163.0

219.5

56.5

34.7

102,600

Bachelor’s degree

Web developers

15-1254

95.3

124.1

28.9

30.3

77,030

Bachelor’s degree

Software developers

15-1252

1,425.9

1,796.5

370.6

26.0

120,730

Bachelor’s degree

Computer and information research scientists

15-1221

33.5

40.6

7.1

21.3

131,490

Master’s degree

Software quality assurance analysts and
testers

15-1253

196.3

237.1

40.8

20.8

98,220

Bachelor’s degree

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Employment of information security analysts is projected to grow the fastest among the computer occupations, at 34.7 percent from 2021 to 2031, more than six times the rate
of growth that is projected for the total economy. As businesses continue to focus on enhancing cybersecurity, they will need information security analysts to secure new
technologies from outside threats or hacks, including IoT-connected devices. As e-commerce continues to expand, organizations will look to utilize the services provided by
web developers to create and maintain websites, which will result in projected growth of 30.3 percent from 2021 to 2031.

Software developers and software quality assurance analysts and testers are projected to be among the fastest growing computer occupations (26.0 percent and 20.8 percent,
respectively) as the services they provide will be needed to support the increasing number of products that use software. Employment of computer and information research
scientists is projected to grow 21.3 percent from 2021 to 2031 as the demand for new and better technology continues to grow.
The median annual wage for computer occupations was $97,430 in May 2021, higher than the median for all occupations in the economy. A bachelor’s degree or higher is
needed for entry-level positions in most occupations in this group, and some on-the-job training may be needed to attain competency in a few of the occupations.
Education, training, and library occupational group

Overall employment in education, training, and library occupations is projected to grow 7.2 percent from 2021 to 2031, about as fast as the average for all occupations; this
increase is expected to result in about 658,200 new jobs over the decade.15 Growth in education, training, and library occupations is influenced by school enrollments and
budgets. The number of people attending postsecondary institutions is expected to grow over the projections decade as students continue to seek higher education to gain the
knowledge and skills necessary to meet their career goals. Only one of the top 30 occupations projected to grow the fastest from 2021 to 2031 is a detailed occupation within
the education, training, and library occupational group: health specialties teachers, postsecondary. (See appendix A-1.)

Table 8. Top five fastest growing occupations within education, training, and library occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Median annual wage, 2021

Typical education needed for

[1]

entry

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Education, training, and library

25-0000

9,151.2

9,809.3

658.2

7.2

57,220

[2]

Health specialties teachers, postsecondary

25-1071

246.7

306.1

59.4

24.1

102,720

Doctoral or professional degree

Nursing instructors and teachers, postsecondary

25-1072

87.0

105.7

18.7

21.5

77,440

Doctoral or professional degree

Self-enrichment teachers

25-3021

347.1

408.3

61.3

17.6

43,580

High school diploma or equivalent

Preschool teachers, except special education

25-2011

483.1

556.0

72.9

15.1

30,210

Associate’s degree

Tutors

25-3041

203.4

232.9

29.5

14.5

36,470

Some college, no degree

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.

[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Health specialties teachers and nursing instructors and teachers are projected to be the two fastest growing occupations within this group; they are projected to grow 24.1 and
21.5 percent, respectively, over the projections decade as the increased demand for medical care will support demand for postsecondary teachers to educate workers.
Demand for preschool and child daycare services is expected to be robust as early childhood education is emphasized. As a result, employment of preschool teachers is
projected to grow 15.1 percent from 2021 to 2031, much faster than the average for all occupations.
The median annual wage for education, training, and library occupations was $57,220 in May 2021, more than the median for all occupations in the economy. College
coursework is required for most jobs within the field, although this varies with the level of instruction. There is no typical on-the-job training needed for the education,
training, and library occupations.
Personal care and service occupational group

Overall employment in personal care and service occupations is projected to grow 14.1 percent from 2021 to 2031, much faster than the average for all occupations; this
increase is expected to result in about 544,800 new jobs over the decade.16 Government-imposed closures for entertainment events in some parts of the country and some
consumer preferences to avoid these types of events resulted in reduced attendance levels or canceled events throughout 2021. Within the personal care and service
occupational group, six of the top 30 occupations projected to grow the fastest from 2021 to 2031 are detailed occupations. This includes the 5 occupations in table 9 and
entertainment attendants and related workers, all other. (See appendix A-1.)

Table 9. Top five fastest growing occupations within personal care and service occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Median annual wage, 2021 [1] Typical education needed for entry

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Personal care and service

39-0000

3,868.4

4,413.2

544.8

14.1

29,450

[2]

Ushers, lobby attendants, and ticket takers

39-3031

63.2

88.8

25.6

40.5

24,440

No formal education credentials

Motion picture projectionists

39-3021

2.0

2.8

0.8

40.3

29,350

No formal education credentials

Animal caretakers

39-2021

290.7

377.6

86.9

29.9

28,600

High school diploma or equivalent

Animal trainers

39-2011

52.9

67.2

14.3

27.1

31,280

High school diploma or equivalent

Personal care and service workers, all other

39-9099

104.4

130.4

26.0

24.9

29,610

High school diploma or equivalent

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

The much-faster-than-average growth for ushers, lobby attendants, and ticket takers and motion picture projectionists primarily represents the recovery of jobs from the effects
of the COVID-19 pandemic. These occupations are expected to recover early in the projections period as social gatherings and other activities resume.
Increasing pet ownership and spending on pets will continue to contribute to employment growth of animal caretakers and animal trainers.
The median annual wage for personal care and service occupations was $29,450 in May 2021, lower than the median for all occupations in the economy. Most occupations in
this group require a high school diploma or equivalent; however, on-the-job training is needed for many of the personal care and service occupations.
Installation, maintenance, and repair occupational group

Overall employment in installation, maintenance, and repair occupations is projected to grow 5.0 percent from 2021 to 2031, about as fast as the average for all occupations;
this increase is expected to result in about 304,000 new jobs over the decade.17 Demand for workers in these occupations will stem from the need to install, maintain, and
repair a wide variety of equipment, including cars, factory machinery, and equipment used in homes and hospitals. In addition, many buildings will need upkeep and renewal
as older homes and buildings typically require more maintenance or repair, especially for pipes, insulation, electrical systems, and air-conditioning and heating systems.
Specialized maintenance and repair of these aging systems alongside installation of new systems will support growth for many jobs in this group. One of the detailed
occupations within this occupational group, wind turbine service technicians, is among the top 30 occupations projected to grow the fastest from 2021 to 2031. (See appendix
A-1.)

Table 10. Top five fastest growing occupations within installation, maintenance, and repair occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Median annual wage, 2021 [1]

Typical education needed for entry

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Installation, maintenance, and repair

49-0000

6,038.7

6,342.6

304.0

5.0

47,940

[2]

Wind turbine service technicians

49-9081

11.1

16.1

4.9

44.3

56,260

Postsecondary nondegree award

Medical equipment repairers

49-9062

59.1

69.1

10.0

17.0

49,910

Associate’s degree

Industrial machinery mechanics

49-9041

384.8

447.9

63.1

16.4

59,840

High school diploma or equivalent

Commercial divers

49-9092

3.0

3.4

0.4

14.7

60,360

Postsecondary nondegree award

Recreational vehicle service technicians

49-3092

16.7

18.7

2.0

12.2

43,560

High school diploma or equivalent

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Wind turbine service technicians are projected to have the second-fastest employment growth of all occupations. However, there will be relatively few jobs added (about 4,900
over 2021–31) because of the small employment numbers of wind turbine service technicians in 2021. Wind power generation has grown over the past 10 years, and it will
require technicians to install, maintain, and repair wind turbines as this generating capacity ages.
The use of medical equipment for diagnosis and treatment will expand as the number of older adults and people with chronic diseases increase. Medical equipment repairers
will be needed to maintain and repair medical equipment. Expansion of automation in production activities will support demand for the services provided by industrial
machinery mechanics as they are needed to help keep machines in good working order.
The median annual wage for installation, maintenance, and repair occupations was $47,940 in May 2021, which was higher than the median for all occupations in the
economy. Most of the occupations within this group require a high school diploma or equivalent and on-the-job training.
Community and social service occupational group

Overall employment in community and social service occupations is projected to grow 10.4 percent from 2021 to 2031, faster than the average for all occupations; this
increase is expected to result in about 294,600 new jobs over the decade.18 As demand remains strong for mental health, addiction, and school and career-counseling services,
employment of community and social service occupations is projected to experience fast growth over the projections period.

Table 11. Top five fastest growing occupations within community and social service occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Percent

Median annual wage,

Typical education needed for

2021 [1]

entry

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Community and social service

21-0000

2,843.2

3,137.8

294.6

10.4

48,410

[2]

21-1018

351.0

428.5

77.5

22.1

48,520

Bachelor's degree

Community health workers

21-1094

67.0

77.7

10.6

15.9

46,590

High school diploma or equivalent

Marriage and family therapists

21-1013

65.3

74.3

9.1

13.9

49,880

Master's degree

Social and human service assistants

21-1093

420.6

472.9

52.4

12.5

37,610

High school diploma or equivalent

Healthcare social workers

21-1022

179.5

199.3

19.9

11.1

60,840

Master's degree

Substance abuse, behavioral disorder, and mental health
counselors

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Much-faster-than-average employment growth is expected for substance abuse, behavioral disorder, and mental health counselors as people continue to seek addiction and
mental health counseling services. An emphasis on promoting healthy behaviors, particularly those based on experiences from the COVID-19 pandemic, is expected to
increase demand for the services provided by community health workers over the projections decade.
Much-faster-than-average employment growth is expected for marriage and family therapists. Growth is expected because of the increasing use of integrated care, which is the
treatment of multiple problems at the same time by a group of specialists. In providing integrated care, marriage and family therapists are working with counselors, such as
substance abuse, behavioral disorder, or mental health counselors, to address patients' issues as a team.
The median annual wage for community and social service occupations was $48,410 in May 2021, which was higher than the median for all occupations in the economy.
Education requirements vary from high school diploma to a master’s degree, tending to be higher for more complex social needs. Several occupations within this group also
require on-the-job training, internship, or residency experience.
Building and grounds cleaning occupational group

Overall employment in building and grounds cleaning occupations is projected to grow 5.4 percent from 2021 to 2031, about as fast as the average for all occupations; this
increase is expected to result in about 290,800 new jobs over the decade.19 None of the detailed occupations within this occupational group are projected to be among the top
30 fastest growing occupations from 2021 to 2031. Building and grounds cleaning workers will be needed to keep up with continued demand for lawn care, landscaping, and
cleaning services from both commercial and residential spaces.
Table 12. Top five fastest growing occupations within building and grounds cleaning occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Median annual wage, 2021

Typical education needed for

[1]

entry

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Building and grounds cleaning

37-0000

5,415.0

5,705.8

290.8

5.4

30,240

[2]

Maids and housekeeping cleaners

37-2012

1,237.4

1,353.8

116.4

9.4

28,780

No formal education credentials

Pest control workers

37-2021

90.6

96.7

6.1

6.8

37,540

High school diploma or equivalent

Pesticide handlers, sprayers, and applicators, vegetation

37-3012

27.6

29.2

1.6

5.8

38,270

High school diploma or equivalent

37-1011

253.0

266.7

13.7

5.4

39,630

High school diploma or equivalent

37-3011

1,191.6

1,248.5

56.9

4.8

34,430

No formal education credentials

First-line supervisors of housekeeping and
janitorial workers
Landscaping and groundskeeping workers

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Maids and housekeeping cleaners are expected to see strong employment growth mostly because of recovery from the COVID-19 recession of 2020. The return to
prepandemic travel patterns will translate to strong demand for maids and housekeeping cleaners in hotels and other traveler accommodations at the start of the projections
decade. The job recovery also is expected in other establishments that were affected by the COVID-19 pandemic, including private households, hospitals, and nursing care
centers.

Landscaping and groundskeeping workers will see employment growth associated with increasing demand for lawn care and landscaping services from homeowners and from
large institutions, such as universities and corporate campuses.
The median annual wage for building and grounds cleaning was $30,240 in May 2021, which was lower than the median for all occupations in the economy. Many of the
occupations within this group do not require an education credential, but a high school diploma or equivalent is needed for entry into higher paying building and grounds
cleaning occupations. On-the-job training is needed for most of the building and grounds cleaning occupations.
Construction and extraction occupational group

Overall employment in construction and extraction occupations is projected to grow 3.6 percent from 2021 to 2031, about as fast as the average for all occupations; this
increase is expected to result in about 252,900 new jobs over the decade.20 Overall growth in the economy will increase demand for new buildings, roads, and other structures,
which will create jobs in construction and extraction occupations. Alternative-energy-related activities will contribute to the growth of construction occupations, including the
installation of electric vehicle (EV) charging stations, photovoltaic (PV) panels, and wind turbines.21 Two of the top 30 occupations projected to grow the fastest from 2021 to
2031 are detailed occupations within the construction and extraction occupational group: solar photovoltaic installers and roustabouts, oil and gas. (See appendix A-1.)

Table 13. Top five fastest growing occupations within construction and extraction occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Median annual wage, 2021 [1]

Typical education needed for entry

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Construction and extraction

47-0000

7,026.0

7,278.9

252.9

3.6

48,210

[2]

Solar photovoltaic installers

47-2231

17.1

21.7

4.6

27.2

47,670

High school diploma or equivalent

Roustabouts, oil and gas

47-5071

37.3

45.9

8.6

23.0

38,920

No formal education credentials

Rotary drill operators, oil and gas

47-5012

12.1

14.3

2.1

17.6

56,380

No formal education credentials

Service unit operators, oil and gas

47-5013

35.7

42.0

6.3

17.5

48,410

No formal education credentials

Derrick operators, oil and gas

47-5011

8.6

10.0

1.4

16.9

47,230

No formal education credentials

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

With the continued expansion and adoption of solar photovoltaic (PV) systems, the employment of solar PV installers is projected to grow 27.2 percent from 2021 to 2031.
This is much faster than the average for all occupations.
The exploration and extraction of oil and gas will support demand for the services provided by roustabouts, rotary drill operators, service unit operators, and derrick operators.
The median annual wage for this group was $48,210 in May 2021, which was higher than the median annual wage for all occupations. Many of these occupations typically do
not require education for entry, however most construction trades occupations do require a high school diploma or equivalent. Nearly all construction and extraction
occupations require on-the-job training, and many construction trades occupations require an apprenticeship.
Legal occupational group

Overall employment in legal occupations is projected to grow 9.6 percent from 2021 to 2031, faster than the average for all occupations; this increase is expected to result in
about 131,000 new jobs over the decade.22 Legal services are expected to be in demand and contribute to employment growth in this occupational group.
Table 14. Top four fastest growing occupations within legal occupations, 2021 and projected 2031
Change

Employment
2021

Median annual wage, 2021

(2021–31)

Occupation
2031

Number

Typical education needed for entry

[1]

Percent

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Legal

23-0000

1,368.0

1,499.0

131.0

9.6

82,430

[2]

Paralegals and legal assistants

23-2011

352.8

402.7

49.9

14.1

56,230

Associate’s degree

Lawyers

23-1011

833.1

913.3

80.2

9.6

127,990

Doctoral or professional degree

Arbitrators, mediators, and conciliators

23-1022

8.9

9.5

0.6

6.2

49,410

Bachelor’s degree

Title examiners, abstractors, and searchers

23-2093

61.2

62.4

1.2

1.9

47,310

High school diploma or equivalent

Total, all occupations

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Demand for specialized and expert legal services will contribute to the overall employment growth of lawyers and legal-related occupations.

Despite this need for legal services, continuing price competition over the projections decade may lead law firms to rethink project staffing to reduce costs to clients. For
example, paralegals and legal assistants are less costly than lawyers in performing a variety of tasks previously assigned to entry-level lawyers.
The median annual wage for this group was $82,430 in May 2021, which was higher than the median annual wage for all occupations of $45,760. Most occupations within
this group require at least a bachelor’s degree, and several of the occupations within this group may also require on-the-job training.
Life, physical, and social science occupational group

Overall employment in life, physical, and social science occupations is projected to grow 6.9 percent from 2021 to 2031, about as fast as the average for all occupations; this
increase is expected to result in about 98,700 new jobs over the decade.23 Increasing demand for expertise in the sciences, particularly in occupations involved in biomedical
research, psychology, and environmental protection, is projected to result in employment growth in this group. One of the detailed occupations within the life, physical, and
social science occupational group, epidemiologists, is among the top 30 occupations projected to grow the fastest from 2021 to 2031. (See appendix A-1.)

Table 15. Top five fastest growing occupations within life, physical, and social science occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Median annual wage, 2021

Typical education needed for

[1]

entry

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Life, physical, and social science

19-0000

1,436.0

1,534.7

98.7

6.9

72,740

[2]

Epidemiologists

19-1041

8.6

10.9

2.2

25.8

78,830

Master's degree

Medical scientists, except epidemiologists

19-1042

119.2

140.0

20.8

17.4

95,310

Doctoral or professional degree

Biochemists and biophysicists

19-1021

37.5

43.2

5.7

15.3

102,270

Doctoral or professional degree

Animal scientists

19-1011

3.7

4.2

0.4

11.8

65,090

Bachelor's degree

Forensic science technicians

19-4092

17.6

19.6

2.0

11.4

61,930

Bachelor's degree

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Employment of epidemiologists is projected to grow 25.8 percent from 2021 to 2031 because of the increased need to identify and mitigate the impact of diseases. However,
because the occupation is small, the fast growth of epidemiologists will result in only about 2,200 new jobs over the projections decade. An increase in the number of people
in older age groups will drive the demand to develop new medicines and treatments to prevent, cure, or manage disease. The increased demand, in turn, is expected to
contribute to the much-faster-than-average projected employment growth of medical scientists and biochemists and biophysicists.
Animal scientists are expected to be needed to research more sustainable farming methods, especially in livestock production. However, because the occupation is small, the
fast growth of animal scientists will result in only about 400 new jobs over the projections decade.
As scientific and technological advances are expected to increase the availability, reliability, and usefulness of objective forensic information used as evidence in trials, more
forensic science technicians will be needed. Because this is a small occupation, its fast growth is expected to result in only about 2,000 new jobs over the projections decade.
The median annual wage for life, physical, and social science occupations was $72,740 in May 2021, which was higher than the median annual wage for all occupations of
$45,760. Some form of postsecondary education is needed for entry-level positions in nearly all occupations in this group. Some occupations within this group require on-thejob training, internship, or residency experience.
Entertainment and sports occupational group

Overall employment in entertainment and sports occupations is projected to grow 12.6 percent from 2021 to 2031, much faster than the average for all occupations; this
increase is expected to result in about 95,500 new jobs over the decade.24 Strong demand from the public for entertainment options, including movies and television shows,
and the continued popularity of sports will contribute to job growth for the entertainment and sports occupations. However, some of the projected employment growth in these
occupations is due to recovery from the COVID-19 recession of 2020; this growth is likely to occur early in the projections decade as participation and attendance in
recreational activities, including organized sports and performances, resume. Four of the top 30 occupations projected to grow the fastest from 2021 to 2031 are detailed
occupations within the entertainment and sports occupational group: athletes and sports competitors; umpires, referees, and other sports officials; dancers; and choreographers.
(See appendix A-1.)

Table 16. Top five fastest growing occupations within entertainment and sports occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Median annual wage, 2021

Typical education needed for

[1]

entry

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Entertainment and sports

27-2000

758.3

853.8

95.5

12.6

49,470

[2]

Athletes and sports competitors

27-2021

15.8

21.5

5.7

35.7

77,300

No formal education credentials

Umpires, referees, and other sports officials

27-2023

13.2

17.4

4.2

31.7

35,860

High school diploma or equivalent

Choreographers

27-2032

6.3

8.1

1.9

29.7

42,700

High school diploma or equivalent

Dancers

27-2031

6.2

7.7

1.5

24.5

[3]

No formal education credentials

Coaches and scouts

27-2022

244.3

293.1

48.8

20.0

38,970

Bachelor's degree

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.

[3] Wages for some occupations that do not generally work year-round, full time, are reported either as hourly wages or annual salaries depending on how they are typically paid. The
median hourly wage for dancers was $18.78 in May 2021.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

In addition to the recovery from the COVID-19 recession of 2020, an increased public interest in professional sports will support demand for athletes and sports competitors;
umpires, referees, and other sports officials; and coaches and scouts.
Dancers and choreographers also were affected by the COVID-19 recession and are projected to experience strong growth as they recover early in the decade. However,
because these are small occupations, their fast growth is expected to result in only about 1,500 new jobs for dancers and 1,900 new jobs for choreographers.
The median annual wage for entertainment and sports occupations was $49,470 in May 2021, slightly higher than the median for all occupations in the economy. While
typical entry-level education requirements vary within this occupational group, most occupations in the entertainment and sports occupations group require on-the-job
training.
Architecture and engineering occupational group

Overall employment in architecture and engineering occupations is projected to grow 3.6 percent from 2021 to 2031, about as fast as the average for all occupations; this
increase is expected to result in about 91,300 new jobs over the decade.25 None of the detailed occupations within this occupational group are among the top 30 occupations
projected to grow the fastest from 2021 to 2031. Most of the projected job growth in this group is for engineers; their services will be in demand in areas such as
manufacturing, construction, and renewable energy.

Table 17. Top five fastest growing occupations within architecture and engineering occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Median annual wage, 2021 [1] Typical education needed for entry

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Architecture and engineering

17-0000

2,562.5

2,653.7

91.3

3.6

79,840

[2]

Chemical engineers

17-2041

26.9

30.7

3.7

13.9

105,550

Bachelor's degree

Industrial engineers

17-2112

301.0

331.6

30.6

10.2

95,300

Bachelor's degree

Bioengineers and biomedical engineers

17-2031

17.9

19.7

1.7

9.8

97,410

Bachelor's degree

Petroleum engineers

17-2171

22.8

24.6

1.9

8.3

130,850

Bachelor's degree

Civil engineers

17-2051

318.3

340.4

22.1

6.9

88,050

Bachelor's degree

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

The employment of chemical engineers is projected to grow much faster than average, at 13.9 percent from 2021 to 2031, as chemical engineering services are needed across
various manufacturing industries. However, because the occupation of chemical engineers is small, its fast growth will result in only about 3,700 new jobs over the projections
decade.
Increased adoption of industrial robotics and integration of automation will continue to create demand for industrial engineers to design efficient manufacturing processes,
resulting in projected employment growth of 10.2 percent over the projections decade. Employment of bioengineers and biomedical engineers is projected to grow 9.8 percent
from 2021 to 2031 as demand for biomedical devices and procedures, such as hip and knee replacements, continues to increase.
The median annual wage for architecture and engineering occupations was $79,840 in May 2021, which was higher than the median for all occupations in the economy. Some
form of postsecondary education is needed for entry-level positions in nearly all occupations in this group. A few occupations within this group require on-the-job training,
internship, or residency experience.

Math occupational group

Overall employment in math occupations is projected to grow 28.7 percent from 2021 to 2031, much faster than the average for all occupations; this increase is expected to
result in about 82,000 new jobs over the decade.26 Expected robust growth in data and the associated demand for data to be collected and analyzed are major factors behind
the strong projected employment growth for math occupations. Growth is anticipated as larger amounts of digital and electronic data are collected with the expanding digital
economy. Workers in math occupations will be needed to collect, organize, and analyze data to help optimize and improve business processes. Three of the top 30 occupations
projected to grow the fastest from 2021 to 2031 are detailed occupations within the math occupational group: data scientists, statisticians, and operations research analysts.
(See appendix A-1.)

Table 18. Top five fastest growing occupations within math occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Median annual wage, 2021 [1] Typical education needed for entry

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Math

15-2000

286.3

368.3

82.0

28.7

98,680

[2]

Data scientists

15-2051

113.3

153.9

40.5

35.8

100,910

Bachelor's degree

Statisticians

15-2041

34.2

45.3

11.2

32.7

95,570

Master's degree

Operations research analysts

15-2031

104.2

128.3

24.2

23.2

82,360

Bachelor's degree

Actuaries

15-2011

28.3

34.2

5.9

20.8

105,900

Bachelor's degree

Mathematical science occupations, all other

15-2099

4.4

4.6

0.3

6.6

62,460

Bachelor's degree

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Because organizations rely on data scientists and statisticians to mine and analyze the large amounts of information and data collected, employment in these occupations is
projected to growth by 35.8 percent and 32.7 percent, respectively, over the projections decade. Data scientists and statisticians are among the top 10 fastest growing
occupations overall. (See appendix A-1.) As technology advances and companies and government agencies seek efficiency and cost savings, demand for operations research
analysts should continue to grow.
Actuaries also are expected to experience much-faster-than-average employment growth as their services will be needed to develop, price, and evaluate a variety of insurance
products and calculate the costs of new risks. Employment of actuaries is projected to grow 20.8 percent from 2021 to 2031.
The median annual wage for math occupations was $98,680 in May 2021, higher than the median for all occupations in the economy. Postsecondary education is needed for
entry-level positions in math occupations. All occupations within this group do not require on-the-job training except for actuaries; they typically require long-term, usually
more than a year, on-the-job training.
Protective service occupational group

Overall employment in protective service occupations is projected to grow 2.1 percent from 2021 to 2031, slower than the average for all occupations; the increase is expected
to result in about 72,600 new jobs over the decade.27 Demand for various types of protective services is expected to persist over the projections decade and support
employment growth for many occupations in this group. These services include protection from fires, crimes, and injuries in sporting and recreational activities.

Table 19. Top five fastest growing occupations within protective service occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Percent

Median annual wage,

Typical education needed for

2021 [1]

entry

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Protective service

33-0000

3,482.2

3,554.8

72.6

2.1

46,590

[2]

Forest fire inspectors and prevention specialists

33-2022

2.9

3.4

0.6

19.4

42,600 High school diploma or equivalent

Lifeguards, ski patrol, and other recreational protective
service workers

33-9092

120.8

140.7

19.9

16.4

25,630

No formal education credentials

Crossing guards and flaggers

33-9091

85.1

92.9

7.8

9.2

31,450

No formal education credentials

Gambling surveillance officers and gambling investigators

33-9031

9.5

10.3

0.8

8.5

35,450 High school diploma or equivalent

School bus monitors

33-9094

53.5

57.1

3.6

6.7

29,100 High school diploma or equivalent

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Forest fire inspectors and prevention specialists will continue to be needed to help control fires and investigate the cause of fires. These inspectors’ and specialists’ services
will continue to be in demand as they will be needed to enforce outdoor fire regulations and other means of forest fire prevention. The severity of wildfires in several states

has increased in recent years, resulting in a greater need for these workers.28
Crossing guards and flaggers and school bus monitors will be needed to protect school children crossing the street, getting on and off the bus, and traveling by bus.
The median annual wage for protective service occupations was $46,590 in May 2021, slightly higher than the median for all occupations in the economy. Typical entry-level
education for most protective service occupations is a high school diploma or equivalent, and candidates typically receive on-the-job training.
Media and communication occupational group

Overall employment in media and communication occupations is projected to grow 6.2 percent from 2021 to 2031, about as fast as the average for all occupations; this
increase is expected to result in about 68,600 new jobs over the decade.29 Demand for media and communication occupations is expected because of the continued need to
create, edit, translate, and disseminate information through a variety of different platforms.
Table 20. Top five fastest growing occupations within media and communication occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation

Total, all occupations

00-0000
27-3000 and

Media and communication

27-4000

Number

Median annual wage, 2021 [1]

Typical education needed for entry

2021

2031

Percent

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

1,111.9

1,180.5

68.6

6.2

61,140

[2]

Interpreters and translators

27-3091

69.4

83.4

14.0

20.2

49,110

Bachelor's degree

Audio and video technicians

27-4011

68.6

79.3

10.7

15.7

48,820

Postsecondary nondegree award

Lighting technicians

27-4015

5.7

6.5

0.8

14.7

51,470

Postsecondary nondegree award

Film and video editors

27-4032

48.1

54.7

6.6

13.8

62,680

Bachelor's degree

Photographers

27-4021

125.6

136.8

11.2

8.9

38,950

High school diploma or equivalent

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Employment of interpreters and translators is projected to grow 20.2 percent from 2021 to 2031 as demand for translation services continues to grow with increasing
globalization and a more diverse U.S. population.
Media and communication equipment workers will be in demand to support the increased need for audio and visual support, including audio and video technicians, lighting
technicians, film and video editors, and photographers. This is in addition to an increase in demand for new content produced for streaming services.
The median annual wage for media and communication workers, including media and communication equipment workers, was $61,140 in May 2021, which was higher than
the median annual wage for all occupations of $45,760. Most occupations within this group require some form of postsecondary education and on-the-job training.
Arts and design occupational group

Overall employment in arts and design occupations is projected to grow 2.2 percent from 2021 to 2031, slower than the average for all occupations; the increase is expected to
result in about 20,500 new jobs over the decade.30 Some of the projected employment growth in arts and design occupations is due to recovery from the COVID-19 recession
of 2020 and is likely to occur early in the projections decade as recreational activities resume.

Table 21. Top five fastest growing occupations within arts and design occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Median annual wage, 2021

Typical education needed for

[1]

entry

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Arts and design

27-1000

918.8

939.4

20.5

2.2

48,220

[2]

Fine artists, including painters, sculptors, and
illustrators

27-1013

27.1

28.8

1.7

6.4

60,820

Bachelor's degree

Special effects artists and animators

27-1014

58.9

62.1

3.2

5.4

78,790

Bachelor's degree

Set and exhibit designers

27-1027

27.0

28.5

1.4

5.2

54,860

Bachelor's degree

Craft artists

27-1012

10.7

11.2

0.5

5.1

35,930

No formal education credentials

Merchandise displayers and window trimmers

27-1026

161.6

169.1

7.5

4.6

32,060

High school diploma or equivalent

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Employment of fine artists, including painters, sculptors, and illustrators is projected to grow 6.4 percent from 2021 to 2031, and employment of craft artists is projected to
grow 5.1 percent over the projections decade as recreational activities resume.
Special effects artists and animators will be needed to meet the demand for animation and visual effects in video games, movies, television, and smartphone applications,
seeing a projected employment growth of 5.4 percent from 2021 to 2031. The employment of set and exhibit designers and merchandise displayers and window trimmers is
projected to grow about as fast as the average for all occupations (5.2 percent and 4.6 percent, respectively) from 2021 to 2031.
The median annual wage for arts and design occupations was $48,220 in May 2021, slightly higher than the median for all occupations in the economy. Most occupations
within this group typically require a bachelor’s degree for entry, and some of these occupations require on-the-job training.
Farming, fishing, and forestry occupational group

Overall employment in farming, fishing, and forestry occupations is projected to show little or no change from 2021 to 2031; this limited growth is expected to result in about
8,400 new jobs over the decade.31 The need for domestic agricultural products should support demand for these workers to produce and supply food. Continued
mechanization of farming and forestry may limit employment in some occupations and benefit employment of other occupations in this group.

Table 22. Top five fastest growing occupations within farming, fishing, and forestry occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Percent

Median annual wage,

Typical education needed for

2021 [1]

entry

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Farming, fishing, and forestry

45-0000

1,069.6

1,078.0

8.4

0.8

29,860

[2]

45-2091

66.6

74.7

8.0

12.1

36,360

No formal education credentials

45-1011

53.3

56.7

3.4

6.4

48,640

High school diploma or equivalent

Agricultural workers, all other

45-2099

11.6

12.1

0.6

4.9

32,550

No formal education credentials

Animal breeders

45-2021

7.3

7.6

0.3

4.3

40,090

High school diploma or equivalent

Log graders and scalers

45-4023

4.5

4.6

0.2

3.5

37,820

High school diploma or equivalent

Agricultural equipment operators
First-line supervisors of farming, fishing, and
forestry workers

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

An expected increase in the use of agricultural equipment will require more agricultural equipment operators relative to farmworkers and laborers. Small farms that sell their
products directly to consumers through venues such as farmers’ markets might create opportunities for some agricultural workers, including animal breeders and other
agricultural workers.
The median annual wage for farming, fishing, and forestry occupations was $29,860 in May 2021, lower than the median for all occupations in the economy. Most of these
occupations do not require any formal education, but a high school diploma and on-the-job training are usually necessary in more specialized jobs in this occupational group.
Production occupational group

Overall employment in production occupations is projected to decline 1.9 percent from 2021 to 2031, a decrease of about 163,600 jobs over the decade.32 The increasing
automation of production processes is expected to continue to require fewer manufacturing jobs, which account for a large share of production occupations. Thirteen of the top
30 occupations projected to decline the fastest from 2021 to 2031 are detailed occupations within the production occupational group, including the 5 occupations shown in
table 23. (See appendix A-2.)

Table 23. Top five fastest declining occupations within production occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Percent

Median annual wage,

Typical education needed for

2021 [1]

entry

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Production

51-0000

8,787.1

8,623.5

-163.6

-1.9

37,710

[2]

Cutters and trimmers, hand

51-9031

8.2

5.9

-2.3

-28.4

30,230

No formal education credentials

Nuclear power reactor operators

51-8011

4.8

3.5

-1.3

-26.8

104,260

High school diploma or equivalent

Print binding and finishing workers

51-5113

42.2

31.8

-10.5

-24.8

36,590

High school diploma or equivalent

Prepress technicians and workers

51-5111

26.0

20.1

-5.9

-22.7

42,610

Postsecondary nondegree award

51-2011

34.3

27.7

-6.6

-19.4

49,480

High school diploma or equivalent

Aircraft structure, surfaces, rigging, and systems
assemblers

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Improved precision technologies enable machines to cut various materials, resulting in reduced demand for cutters and trimmers. Increased use of robotics will enable
assemblers and fabricators to work alongside robots (a development also known as collaborative robotics). These robots can perform more complex tasks such as drilling
holes, cutting materials, or painting equipment, which reduces demand for some assemblers and fabricators.
Employment declines in some production occupations also will result from shifts in product preferences. As nuclear energy power production faces steep competition from
renewable energy sources, decommissioning plans for several reactors are reducing demand for nuclear power reactor operators. Similarly, increased customer demand for
digital products compared with printed materials will reduce the demand for the services provided by print binding and finishing workers and prepress technicians.
The median annual wage for production occupations was $37,710 in May 2021, lower than the median for all occupations in the economy. Education requirements range from
no formal education to a postsecondary nondegree award, but a high school diploma and some on-the-job training are typically needed.
Sales occupational group

Overall employment in sales occupations is projected to show little or no change from 2021 to 2031, seeing a decrease of about 164,500 jobs over the decade.33 The
increasing use of digital marketing and advertising is contributing to the projected decline in employment of many sales occupations. Alternative methods of direct marketing
(including web advertisements, emails, and text messages) have emerged as substitutes for telemarketing and door-to-door sales marketing. Telemarketers are the only detailed
occupation from the sales occupational group among the top 30 occupations projected to decline the fastest from 2021 to 2031. (See appendix A-2.)

Table 24. Top five fastest declining occupations within sales occupations, 2021 and projected 2031
Change

Employment
2021

Median annual wage,

(2021–31)

Occupation
2031

Number

Percent

Typical education needed for

2021 [1]

entry

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Sales

41-0000

14,719.9

14,555.4

-164.5

-1.1

30,600

[2]

Telemarketers

41-9041

115.7

94.7

-21.0

-18.2

28,910

No formal education credentials

Door-to-door sales workers, news and street vendors,
and related workers

41-9091

54.7

49.0

-5.7

-10.4

29,390

No formal education credentials

Cashiers

41-2011

3,371.6

3,036.0

-335.7

-10.0

27,260

No formal education credentials

Advertising sales agents

41-3011

100.7

92.7

-8.0

-7.9

52,340

First-line supervisors of retail sales workers

41-1011

1,505.7

1,427.5

-78.2

-5.2

39,230

High school diploma or
equivalent
High school diploma or
equivalent

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Telemarketers; door-to-door sales workers, news and street vendors, and related workers; and advertising sales agents are all projected to experience a decline in employment
from 2021 to 2031. The growth of digital advertising also will reduce demand for paper advertisements and increase demand for virtual promotional products.
Cashiers are projected to have the largest decline of any occupation over the projections period. Employment of cashiers is expected to decline because of advances in
technology, such as the increased use of online sales, digital payment, and self-checkout systems.
The median annual wage for sales occupations was $30,600 in May 2021, lower than the median for all occupations in the economy. Most occupations in this group typically
require no education credentials, but some of the occupations may require a high school diploma. Also, some form of on-the-job training may be needed.
Office and administrative support occupational group

Overall employment in office and administrative support occupations is projected to decline 4.5 percent from 2021 to 2031, a decrease of about 880,800 jobs over the
decade.34 Office and administrative support occupations constitute the largest of all groups, composing about 12.4 percent of all jobs in 2021; however, this group is also
expected to lose the most jobs of any occupational group by 2031. Technological changes are expected to continue to negatively affect the future employment of office and
administrative support occupations. Computer and application software tools, digital data collection, and automated scheduling systems continue to be improved and used in
many office and administrative support tasks. Seven of the top 30 occupations projected to decline the fastest from 2021 to 2031 are detailed occupations within the office and
administrative support occupational group, including the 5 occupations in table 25, legal secretaries and administrative assistants, and order clerks. (See appendix A-2.)

Table 25. Top five fastest declining occupations within office and administrative support occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

Number

Median annual wage, 2021

Typical education needed for

[1]

entry

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

[2]

Office and administrative support

43-0000

19,587.0

18,706.2

-880.8

-4.5

38,050

[2]

Word processors and typists

43-9022

46.1

28.5

-17.6

-38.2

44,030

High school diploma or equivalent

Data entry keyers

43-9021

155.9

117.4

-38.5

-24.7

35,630

High school diploma or equivalent

Telephone operators

43-2021

4.0

3.0

-1.0

-24.5

37,630

High school diploma or equivalent

Switchboard operators, including answering service

43-2011

49.0

37.2

-11.8

-24.0

30,150

High school diploma or equivalent

Executive secretaries and executive administrative
assistants

43-6011

508.0

405.4

-102.6

-20.2

62,060

High school diploma or equivalent

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] This entry is not applicable.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.

Many office tasks (for example, data entry, telephone, and answering services) continue to be automated. As a result, employment in office and administrative support
occupations is projected to decline. For example, technological improvements will require fewer secretaries and administrative assistants as many secretarial tasks can now be
completed by other workers. Word processors and typists is the occupation with the fastest projected employment decline of all occupations as computer use continues to
enable many occupations to acquire typing skills. Employment in this occupation is projected to decline by 38.2 percent from 2021 to 2031.
Despite projected employment declines, openings in office and administrative support occupations are expected to result from the need to replace workers who transfer to
other occupations or exit the labor force, such as retirees.
The median annual wage for office and administrative support occupations was $38,050 in May 2021, lower than the median for all occupations in the economy. A high school
diploma or equivalent is the most common entry-level education requirement for occupations in this group, and some form of on-the-job training may be needed.

Discussion and analysis
While specific factors drive employment change for each detailed occupation, there are broader macroeconomic factors that can affect occupations within an occupational
group or even across occupational groups. About 8.3 million new jobs are projected to be added over the 2021–31 projections decade. Nearly one in four new jobs will be in
the healthcare occupational group. Computer and information technology occupations and math occupations are projected to experience much-faster-than-average
employment growth because of the strong demand for informational technology (IT) services and an expected robust growth in data analysis. The food preparation and
serving occupations and personal care and service occupations are projected to experience fast growth over the projections period. However, a part of this projected growth
represents recovery from a low employment level in 2021. Production occupations, sales occupations, and office and administrative occupations are the occupational groups
projected to decline over the projections decade because of changes in technology, business practices, and outsourcing activities.

Appendix A-1: Top 30 fastest growing occupations, 2021 and projected 2031
Employment

Change (2021–31)

Median annual wage, 2021 [1]

Occupation
2021

2031

Number

Percent

Total, all occupations

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

Nurse practitioners

29-1171

246.7

359.4

112.7

45.7

120,680

Wind turbine service technicians

49-9081

11.1

16.1

4.9

44.3

56,260

Ushers, lobby attendants, and ticket takers

39-3031

63.2

88.8

25.6

40.5

24,440

Motion picture projectionists

39-3021

2.0

2.8

0.8

40.3

29,350

Cooks, restaurant

35-2014

1,255.6

1,715.6

459.9

36.6

30,010

Data scientists

15-2051

113.3

153.9

40.5

35.8

100,910

Athletes and sports competitors

27-2021

15.8

21.5

5.7

35.7

77,300

Information security analysts

15-1212

163.0

219.5

56.5

34.7

102,600

Statisticians

15-2041

34.2

45.3

11.2

32.7

95,570

Umpires, referees, and other sports officials

27-2023

13.2

17.4

4.2

31.7

35,860

Web developers

15-1254

95.3

124.1

28.9

30.3

77,030

Animal caretakers

39-2021

290.7

377.6

86.9

29.9

28,600

Choreographers

27-2032

6.3

8.1

1.9

29.7

42,700

Taxi drivers

53-3054

128.5

165.1

36.6

28.5

29,310

Medical and health services managers

11-9111

480.7

616.9

136.2

28.3

101,340

Logisticians

13-1081

195.0

249.1

54.1

27.7

77,030

Physician assistants

29-1071

139.1

177.5

38.4

27.6

121,530

Solar photovoltaic installers

47-2231

17.1

21.7

4.6

27.2

47,670

Animal trainers

39-2011

52.9

67.2

14.3

27.1

31,280

Physical therapist assistants

31-2021

96.5

122.1

25.6

26.5

61,180

Software developers

15-1252

1,425.9

1,796.5

370.6

26.0

120,730

Epidemiologists

19-1041

8.6

10.9

2.2

25.8

78,830

Occupational therapy assistants

31-2011

43.4

54.5

11.0

25.4

61,730

Home health and personal care aides

31-1120

3,636.9

4,560.9

924.0

25.4

29,430

Personal care and service workers, all other

39-9099

104.4

130.4

26.0

24.9

29,610

Dancers

27-2031

6.2

7.7

1.5

24.5

[2]

Health specialties teachers, postsecondary

25-1071

246.7

306.1

59.4

24.1

102,720

Entertainment attendants and related workers, all other

39-3099

4.7

5.8

1.1

23.2

24,170

Operations research analysts

15-2031

104.2

128.3

24.2

23.2

82,360

Roustabouts, oil and gas

47-5071

37.3

45.9

8.6

23.0

38,920

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
[2] Wages for some occupations that do not generally work year-round, full time, are reported either as hourly wages or annual salaries depending on how they are typically paid. The
median hourly wage for dancers was $18.78 in May 2021.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Consumer Expenditure Surveys Public-Use Microdata.

Appendix A-2: Top 30 fastest declining occupations, 2021 and projected 2031
Change

Employment

(2021–31)

Occupation
2021

2031

00-0000

158,134.7

166,452.1

8,317.4

5.3

$45,760

Word processors and typists

43-9022

46.1

28.5

-17.6

-38.2

44,030

Parking enforcement workers

33-3041

8.6

5.4

-3.2

-37.1

46,590

Cutters and trimmers, hand

51-9031

8.2

5.9

-2.3

-28.4

30,230

Nuclear power reactor operators

51-8011

4.8

3.5

-1.3

-26.8

104,260

Print binding and finishing workers

51-5113

42.2

31.8

-10.5

-24.8

36,590

Watch and clock repairers

49-9064

2.2

1.7

-0.5

-24.7

44,250

Data entry keyers

43-9021

155.9

117.4

-38.5

-24.7

35,630

Telephone operators

43-2021

4.0

3.0

-1.0

-24.5

37,630

Switchboard operators, including answering service

43-2011

49.0

37.2

-11.8

-24.0

30,150

Electronic equipment installers and repairers, motor vehicles

49-2096

9.2

7.1

-2.2

-23.4

40,670

Prepress technicians and workers

51-5111

26.0

20.1

-5.9

-22.7

42,610

Roof bolters, mining

47-5043

1.9

1.5

-0.4

-21.5

59,770

Floral designers

27-1023

44.4

35.1

-9.3

-21.0

29,880

Manufactured building and mobile home installers

49-9095

3.9

3.1

-0.8

-20.3

36,360

Refractory materials repairers, except brickmasons

49-9045

0.7

0.5

-0.1

-20.2

54,250

Executive secretaries and executive administrative assistants

43-6011

508.0

405.4

-102.6

-20.2

62,060

Aircraft structure, surfaces, rigging, and systems assemblers

51-2011

34.3

27.7

-6.6

-19.4

49,480

Legal secretaries and administrative assistants

43-6012

157.8

127.5

-30.4

-19.2

47,710

Grinding and polishing workers, hand

51-9022

16.1

13.1

-3.0

-18.7

35,670

Drilling and boring machine tool setters, operators, and tenders, metal and plastic

51-4032

6.9

5.6

-1.3

-18.6

38,580

Forging machine setters, operators, and tenders, metal and plastic

51-4022

11.8

9.6

-2.2

-18.3

44,520

Timing device assemblers and adjusters

51-2061

0.6

0.5

-0.1

-18.3

37,780

Telemarketers

41-9041

115.7

94.7

-21.0

-18.2

28,910

Coil winders, tapers, and finishers

51-2021

11.4

9.3

-2.0

-17.9

38,360

Loading and moving machine operators, underground mining

47-5044

4.5

3.7

-0.8

-17.8

57,900

Milling and planing machine setters, operators, and tenders, metal and plastic

51-4035

15.2

12.5

-2.7

-17.7

46,850

Order clerks

43-4151

143.9

119.7

-24.2

-16.8

37,920

Nuclear technicians

19-4051

5.4

4.5

-0.9

-16.6

99,340

Structural metal fabricators and fitters

51-2041

63.6

53.6

-10.1

-15.8

45,480

Power plant operators

51-8013

29.2

24.7

-4.5

-15.5

80,850

Total, all occupations

Number

Median annual wage, 2021 [1]

Percent

[1] Data are from the Occupational Employment and Wage Statistics program, U.S. Bureau of Labor Statistics. Wage data cover nonfarm wage and salary workers and do not cover the
self-employed, owners and partners in unincorporated firms, or household workers.
Note: Employment numbers are in thousands.
Source: U.S. Bureau of Labor Statistics, Employment Projections program.
SUGGESTED CITATION:

Nicholas DeZarn, Stanislava Ilic-Godfrey, and Emily Krutsch, "Occupational projections overview, 2021–31," Monthly Labor Review, U.S. Bureau of Labor Statistics, April 2023,

https://doi.org/10.21916/mlr.2023.6

Notes

1 “Employment projections: 2021–2031 summary,” USDL-22-1805 (U.S. Bureau of Labor Statistics, September 8, 2022), https://www.bls.gov/news.release/ecopro.nr0.htm.
2 Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh.
3 The Occupational Outlook Handbook (OOH) groups occupations into 25 occupational groups, but BLS employment projections cover the civilian workforce only. Therefore, the military
occupational group was excluded from this article. The OOH uses the 2018 Standard Occupational Classification (SOC) system structure. However, the OOH groups occupations differently in
some scenarios to show similar groups together or to show a more detailed level of information. Healthcare practitioners and technical occupations (29-0000) and healthcare support
occupations (31-0000) are combined under healthcare occupations; computer and mathematical occupations (15-0000) are shown in the OOH at the three-digit SOC level for computer and
information technology occupations (15-1200) and mathematical science occupations (15-2000); and arts, design, entertainment, sports, and media occupations are shown at the three-digit
SOC level for arts and design workers (27-1000), entertainers and performers, sports and related workers (27-2000), and media and communication workers (27-3000 and 27-4000). All other
groups are a direct match between the data shown in the OOH and the two-digit SOC structure.

4 The COVID-19 pandemic affected occupational groups in different ways. Some employment lost during the pandemic and projected to be recovered over the projections decade has already
been recovered as employment grew rapidly throughout the first half of 2022. As a result, some occupational groups have fast projected growth that reflects short-term recovery rather than
long-term expected gains. The pandemic also has been a catalyst for some structural changes in demand for certain goods and services, which are expected to affect long-term demand for
employment in a select group of industries and occupations. For more information on the effects of the COVID-19 pandemic on the 2021–31 projections, see “Employment projections: 20212031 summary.”

5 Although declining employment dampens hiring as job separations outnumber openings, turnover generates more openings than occupation growth; even in declining occupations, demand
for replacement of outgoing workers continues to support available opportunities. For more information on occupational separations, see

https://www.bls.gov/emp/documentation/separations.htm; for replacements, see https://www.bls.gov/emp/documentation/replacements.htm.

6 More information about what is included in the OOH is available online at “Occupational information included in the OOH,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics,
last modified on September 8, 2022), https://www.bls.gov/ooh/about/occupational-information-included-in-the-ooh.htm.

7 “Healthcare occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/healthcare/home.htm.
8 “Food preparation and serving occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/food-

preparation-and-serving/home.htm.
9 “Table 1.1A. Employment by major occupational group, 2021, and projected 2031, including adjustments for realized gains (numbers in thousands),” Employment Projections (U.S. Bureau of
Labor Statistics, last modified on September 8, 2022), https://stats.bls.gov/emp/tables/emp-by-major-occupational-group-alt.htm.

10 “Management occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/management/home.htm.
11 “Transportation and material moving occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022),

https://www.bls.gov/ooh/transportation-and-material-moving/home.htm.
12 “Business and financial occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/business-and-

financial/home.htm.
13 “Computer and information technology occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022),

https://www.bls.gov/ooh/computer-and-information-technology/home.htm.
14 Sara Hylton, Lindsey Ice, and Emily Krutsch, “What the long-term impacts of the COVID-19 pandemic could mean for the future of IT jobs,” Beyond the Numbers: Employment &
Unemployment, vol. 11, no. 3 (U.S. Bureau of Labor Statistics, February 2022), https://www.bls.gov/opub/btn/volume-11/what-the-long-term-impacts-of-the-covid-19-pandemic-

could-mean-for-the-future-of-it-jobs.htm.
15 “Education, training, and library occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022),

https://www.bls.gov/ooh/education-training-and-library/home.htm.
16 “Personal care and service occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/personal-care-

and-service/home.htm.
17 “Installation, maintenance, and repair occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022),

https://www.bls.gov/ooh/installation-maintenance-and-repair/home.htm.
18 “Community and social service occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022),

https://www.bls.gov/ooh/community-and-social-service/home.htm.
19 “Building and grounds cleaning occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/building-

and-grounds-cleaning/home.htm.
20 “Construction and extraction occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022),

https://www.bls.gov/ooh/construction-and-extraction/home.htm.
21 Readers who are interested in the transition to electric vehicles can read the following: Javier Colato and Lindsey Ice, “Charging into the future: the transition to electric vehicles,” Beyond
the Numbers: Employment & Unemployment, vol. 12, no. 4 (U.S. Bureau of Labor Statistics, February 2023), https://www.bls.gov/opub/btn/volume-12/charging-into-the-future-the-

transition-to-electric-vehicles.htm.
22 “Legal occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/legal/home.htm.
23 “Life, physical, and social science occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 14, 2022), https://www.bls.gov/ooh/life-

physical-and-social-science/home.htm.
24 “Entertainment and sports occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/entertainment-

and-sports/home.htm.
25 “Architecture and engineering occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022),

https://www.bls.gov/ooh/architecture-and-engineering/home.htm.
26 “Math occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/math/home.htm.
27 “Protective service occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/protective-

service/home.htm.
28 “Climate change indicators: wildfires,” Climate Change Indicators (U.S. Environmental Protection Agency, last updated on March 21, 2023), https://www.epa.gov/climate-

indicators/climate-change-indicators-wildfires.
29 “Media and communication occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/media-and-

communication/home.htm.
30 “Arts and design occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/arts-and-

design/home.htm.
31 “Farming, fishing, and forestry occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/farming-

fishing-and-forestry/home.htm.
32 “Production occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/production/home.htm.
33 “Sales occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/sales/home.htm.
34 “Office and administrative support occupations,” Occupational Outlook Handbook (U.S. Bureau of Labor Statistics, last modified on September 8, 2022), https://www.bls.gov/ooh/office-

and-administrative-support/home.htm.

ABOUT THE AUTHOR

Nicholas DeZarn
dezarn.nicholas@bls.gov
Nicholas DeZarn is an economist in the Office of Occupational Statistics and Employment Projections, U.S. Bureau of Labor Statistics.
Stanislava Ilic-Godfrey
ilic-godfrey.stanislava@bls.gov
Stanislava Ilic-Godfrey is an economist in the Office of Occupational Statistics and Employment Projections, U.S. Bureau of Labor Statistics
Emily Krutsch
krutsch.emily@bls.gov
Emily Krutsch is an economist in the Office of Occupational Statistics and Employment Projections, U.S. Bureau of Labor Statistics.

RELATED CONTENT

Related Articles
Projections overview and highlights, 2021–31, Monthly Labor Review, November 2022.
Employment projections in a pandemic environment, Monthly Labor Review, February 2021.
Full employment: an assumption within BLS projections, Monthly Labor Review, November 2017.
Labor force projections to 2022: the labor force participation rate continues to fall, Monthly Labor Review, December 2013.
Related Subjects
Projections
Labor force
Earnings and wages
Compensation
Employment
Occupations
COVID-19

ARTICLE CITATIONS

Crossref

0

U.S. BUREAU OF LABOR STATISTICS Division of Information and Marketing Services PSB Suite 2850 2 Massachusetts Avenue NE
Washington, DC 20212-0001
Telephone:1-202-691-5200 Telecommunications Relay Service:7-1-1 www.bls.gov/OPUB Contact Us

U.S. BUREAU OF LABOR STATISTICS
Bureau of Labor Statistics

Publications

Monthly Labor Review

HOME
Search MLR

ARCHIVES

FOR AUTHORS

ABOUT

SUBSCRIBE

GO

ARTICLE

ERRATA

Automotive dealerships 2019–22: dealer markup increases drive new-vehicle consumer inflation

April 2 0 2 3

Automotive dealerships 2019–22: dealer markup increases drive new-vehicle consumer inflation
Using U.S. Bureau of Labor Statistics data and novel analytical methods, this article shows how automotive dealerships contributed to new-vehicle consumer inflation
through markup increases during the economic recovery from the COVID-19 pandemic. Dealerships have a major role in managing the inventory of unsold vehicles and
typically have a significant amount of unsold inventory rotating through their lots and garages. Being an inventory intermediary in the vehicle supply chain, and already
having subdued margins due to previous profit-margin compressions, dealerships were well positioned to expand profit margins from new-vehicle sales in the recent economic
expansion. These increases contributed substantially to new-vehicle consumer inflation over the last 3 years.
The automotive industry is an important part of the U.S. economy. During the recent economic expansion associated with the recovery from the COVID-19 pandemic, price
increases for new cars and trucks contributed moderately to overall consumer inflation. In the previous expansion, which followed the global financial crisis of 2008 and
lasted from 2009 to 2019, new-vehicle prices were largely subdued because of profit-margin compression at vehicle dealerships. However, the competitive dynamics and
trends that developed during this earlier expansion set the stage for dealerships to subsequently increase profits from the sale of new vehicles, contributing largely to newvehicle consumer inflation in the COVID-19 economic expansion.1 This latest development, analyzed in this article, reflects dealerships’ major role as an inventory
intermediary in the vehicle supply chain.2

Industry and theoretical background
The U.S. Bureau of Labor Statistics (BLS) publishes several price indexes that track price changes for different goods and services in the automotive supply chain. The
Consumer Price Index (CPI) for new cars and trucks (hereafter referred to as “CPI for new vehicles”) tracks prices paid by consumers for new vehicles. The Producer Price
Index (PPI) for motor vehicles (hereafter referred to as “PPI for new vehicles”) tracks prices paid by wholesalers, dealerships, intermediaries, and other businesses to
manufacturers of new motor vehicles. The Import Price Index (MPI) for automobile and light duty motor vehicle manufacturing (hereafter referred to as “MPI for new
vehicles”) tracks prices for imported cars and trucks. Finally, the PPI for vehicle sales (hereafter referred to as “PPI for dealership markups”) tracks margins or markups,3
which are the retail selling prices received by dealerships for cars and trucks (regardless of whether the vehicles were manufactured in the United States or imported) less their
acquisition prices.
A central theme of this analysis is that changes in retail markups drive the statistical and conceptual differences between changes in producer vehicle prices and changes in
consumer vehicle prices.4 Table 1 shows the trends in the CPI for new vehicles, the PPI for new vehicles, and the PPI for dealership markups for the last three business cycles:
December 2000 to September 2007 (peak to peak), September 2007 to December 2019 (peak to peak), and December 2019 to December 2022 (peak to December 2022).
When the CPI for new vehicles and the PPI for new vehicles move similarly in magnitude and direction, the PPI for dealership markups is relatively stable; when they diverge,
the PPI for dealership markups changes substantially.

Table 1. Trends in CPI for new vehicles, PPI for new vehicles, and PPI for dealership markups, December 2000–December 2022

Business cycle

Percent change

Percent change

Percent change for

Difference between percent

Absolute difference between

Absolue value of percent

for CPI for new

for PPI for new

PPI for dealership

changes for CPI and PPI for

percent changes for CPI and

change for PPI for

vehicles

vehicles

markups

new vehicles

PPI for new vehicles

dealership markups

December 2000 to
September 2007 (peak to
peak)
September 2007 to
December 2019 (peak to

-6.0

-8.8

2.6

2.8

2.8

2.6

8.4

22.4

-32.9

-14.0

14.0

32.9

20.7

7.9

144.7

12.8

12.8

144.7

peak)
December 2019 to
December 2022 (peak to
December 2022)

Note: CPI = Consumer Price Index; PPI = Producer Price Index.
Source: Author's calculations based on data from the U.S. Bureau of Labor Statistics.

Covering the period from 2007 through 2022, charts 1 and 2 display scatterplots showing the absolute value of the year-ending annual percent change for the PPI for
dealership markups (horizontal axis) and the absolute value of the difference between the year-ending annual percent changes for the CPI for new vehicles and the PPI for new
vehicles (vertical axis). (Chart 1 includes the pandemic years, whereas chart 2 excludes them.) The PPI for dealership markups changes the most when the CPI for new
vehicles and the PPI for new vehicles move in opposite directions, but it also fluctuates when the movements of those indexes differ substantially in magnitude. The
comparatively larger magnitude of change in the PPI for dealership markups is due to the fact that, by definition, a markup is a marginal differential output of inputs that exist
on much larger scales than the markup. These trends, which show how dealerships function as an intermediary materially affecting price transmission, are also corroborated by
financial reporting data from the U.S. Securities and Exchange Commission (SEC).5
Chart 1. Absolute year-ending annual percent change for PPI for dealership
markups and absolute difference between year-ending annual percent changes
for CPI and PPI for new vehicles, pandemic included, 2007–22
Diff. b/n CPI and PPI percent changes
10
9
8
7
6
5
4
3
2
1
0
0

20

40

60

80

100

120

140

Percent change for PPI for dealership markups
Hover over chart to view data.
Note: R-squared = 0.5998. CPI = Consumer Price Index; PPI = Producer Price Index.
Source: Author's calculations based on data from the U.S. Bureau of Labor Statistics.

View Chart Data

Chart 2. Absolute year-ending annual percent change for PPI for dealership
markups and absolute difference between year-ending annual percent
changes for CPI and PPI for new vehicles, pandemic excluded, 2007–19
Diff. b/n CPI and PPI percent changes
8
7
6
5
4
3
2
1
0
-1
0

2

4

6

8

10

12

14

16

Percent change for PPI for dealership markups
Hover over chart to view data.
Note: R-squared = 0.743. CPI = Consumer Price Index; PPI = Producer Price Index.
Source: Author's calculations based on data from the U.S. Bureau of Labor Statistics.

View Chart Data

The competitive dynamics that evolved in the decades preceding the COVID-19 pandemic uniquely positioned dealerships to expand profits from new-vehicle sales during the
recent economic expansion. During the 2007–19 business cycle, dealerships experienced substantial profit-margin compression on new-vehicle sales because they faced
higher prices from manufacturers and did not fully push those higher prices onto consumers. Instead, dealerships weathered the margins squeeze by expanding their value
proposition by selling more finance and insurance products that subsidized the sale of low-to-no-margin vehicles. Dealerships were caught in the margins squeeze for two
reasons. First, the interdependencies and market-power dynamics between dealerships and manufacturers resulted in manufacturers pushing large amounts of inventory onto
dealerships, with the latter being unable to negotiate on price. Second, after the Great Recession of 2007–09, consumers were very sensitive to high prices.6
The pandemic, along with the economic stimulus provided in response to it, created the perfect combination of exogenous shocks to invert the competitive dynamics of the
preceding expansion. First, manufacturers encountered significant supply shortages and supply-chain bottlenecks, both due to the pandemic-related shutdowns adopted across
the globe.7 Although this shock acutely affected manufacturers’ production, dealerships, which tend to have between 60 and 90 days of inventory on hand, initially had
inventory volumes that did not require the purchase of additional vehicles from manufacturers.8 Moreover, during the initial phases of the pandemic, economic actors across
the supply chain believed the recession would follow a typical path of sustained decreases in demand, so they cut output and scrambled to push inventories downstream to
dealerships.9 Second, U.S. fiscal and monetary policy provided unprecedented stimulus in response to the pandemic, and this stimulus resulted in a large increase in personal
savings.10 Third, the perceived economic cost of consuming services increased dramatically because of the consumer concern that services came with the increased cost of
potentially contracting COVID-19. Consequently, consumers’ total budgets, boosted by stimulus, pivoted substantially toward physical goods, with the net result being a sharp
increase in the consumption of those goods.11

The consumer spending changes and economic stimuluses associated with the pandemic merit special consideration because of their unprecedented scale and apparent impact
on new-vehicle demand. The confluence of the three aforementioned factors allowed dealerships to substantially increase markups during the 2020–22 economic expansion,
boosting their profits and sizably contributing to new-vehicle consumer inflation.

Data analysis
Graphical and statistical analyses of BLS price-index data for vehicles and dealership services indicate that, because of profit-margin increases at dealerships over the last 3
years, consumer prices for new vehicles outpaced manufacturer and import prices for new vehicles. Chart 3 shows that, from December 2019 to December 2022, the CPI for
new vehicles grew by 20.7 percent, the PPI for new vehicles grew by 7.9 percent, the MPI for new vehicles increased by 6.4 percent, and the PPI for dealership markups
increased by 144.7 percent. It should be noted that, during 2019, consumer, producer, and import prices for new vehicles remained largely unchanged, so the PPI for
dealership markups was also relatively stable over the year. During 2020, consumer prices only slightly outpaced producer and import prices, so the PPI for dealership
markups increased by only 31.9 percent in that year. However, the difference in trends for consumer, producer, and import prices accelerated sharply in 2021, resulting in the
margins index increasing by 119.0 percent in 2021 alone. That year saw the largest ever absolute difference between the year-ending annual percent changes for the CPI for
new vehicles and the PPI for new vehicles, as well as the largest ever year-ending annual increase (119.0 percent) for the PPI for dealership markups.
Chart 3. CPI, PPI, and MPI for new vehicles, PPI for dealership markups, and
estimated markup index, January 2019–December 2022
PPI for dealership markups
CPI for new vehicles
MPI for new vehicles
Estimated markup index

CPI, PPI, and MPI for new vehicles

PPI for new vehicles

PPI for markups and est. markup index

130

400

125

350

120

300

115

250

110

200

105

150

100

100

95

50

90
Jan 2019

0
Jul 2019

Jan 2020

Jul 2020

Jan 2021

Jul 2021

Jan 2022

Jul 2022

Click legend items to change data display. Hover over chart to view data.
Note: CPI = Consumer Price Index; PPI = Producer Price Index; MPI = Import Price Index.
Source: U.S. Bureau of Labor Statistics (BLS) and author's calculations based on BLS data.

View Chart Data

Besides showing the official PPI for dealership markups, chart 3 also presents an estimated markup index derived from a linear residual of the short-term price relatives
(STRs) of the PPI for new vehicles and the CPI for new vehicles (long-term price relatives (LTRs) are constructed from STRs defined in equation set 1 in the appendix).12
Similarly to the PPI for dealership markups, the estimated markup index grew sharply from December 2019 to December 2022, rising by 255.1 percent. And similarly to the
official price measures, the estimated markup index did not begin rapidly increasing until 2021. Under the assumption of an average markup of 5.0 percent in January 2019,
the PPI for dealership markups would suggest that the markup would have peaked at 14.7 percent in June 2022, and under the same assumption, the estimated markup index
would suggest that the markup would have peaked at 17.7 percent in September 2022. By December 2022, these estimated markups would have fallen to 11.9 and 16.6
percent, respectively. Both of these estimates are largely corroborated by SEC financial data, which show that average new-vehicle markups increased by 146.0 percent from
the first quarter of 2019 to the third quarter of 2022, reaching 13.1 percent before falling to 10.9 percent in the fourth quarter of 2022.13
Combining the methodologies presented in two other Monthly Labor Review articles,14 one can use the PPI for new vehicles, the PPI for dealership markups, and the MPI for
new vehicles to recreate the CPI for new vehicles. This can be accomplished through a weighted input price index (input price index including markups) that includes the
services provided at a dealership, with margin percentages serving as weights.15 Conceptually, the equation for calculating this index (see equation set 2 in appendix) treats the
PPI for new vehicles and the MPI for new vehicles as vehicle prices and the PPI for dealership markups as a markup, using actual (known) margin percentages as initial
weights and estimated margin percentages as weights beyond the initial period.16
Chart 4 and table 2 show the correlations between the input price index including markups and the official CPI for new vehicles. The statistical and graphical correlations
between the two series are strong, further demonstrating that profit-margin changes at dealerships explain the difference between the CPI for new vehicles and the PPI for new
vehicles. The test statistics and graphical results exceed thresholds established in BLS-domain-hosted literature characterizing methods to recreate official BLS price
measures.17 From December 2019 to December 2020, the CPI for new vehicles increased by 2.0 percent and the estimated input price index including markups increased by
2.5 percent, whereas the PPI for new vehicles increased by only 0.7 percent. Additionally, from December 2020 to December 2022, a period in which dealership markups
increased dramatically, the CPI for new vehicles increased by 18.4 percent and the estimated input price index including markups increased by 17.3 percent, whereas the PPI
for new vehicles increased by only 7.2 percent. The statistically univariate regression model between the STRs of the CPI for new vehicles and the STRs of the input price
index including markups is the only model showing a statistically significant correlation at the 1-percent level of significance (p-value of 0.00) and a meaningfully high
correlation coefficient of 0.57. All other models—those using only the PPI for new vehicles, only the MPI for new vehicles, only the PPI for dealership markups, or only the
input price index without markups—have p-values greater than 0.01 and low correlation coefficients.

Chart 4. CPI, PPI, and MPI for new vehicles and input price indexes with and
without markups, January 2019–December 2022
Index

CPI for new vehicles
PPI for new vehicles
MPI for new vehicles
Input price index without markups
Input price index including markups

130
125
120
115
110
105
100
95
Jan 2019

Jul 2019

Jan 2020

Jul 2020

Jan 2021

Jul 2021

Jan 2022

Jul 2022

Click legend items to change data display. Hover over chart to view data.
Note: CPI = Consumer Price Index; PPI = Producer Price Index; MPI = Import Price Index.
Source: U.S. Bureau of Labor Statistics (BLS) and author's calculations based on BLS data.

View Chart Data

Table 2. Correlations between CPI for new vehicles and PPI and MPI for new vehicles, PPI for dealership markups, and input price indexes with and without
markups, 1-month percent changes, January 2019–December 2022
Independent variable

Correlation coefficient

p-value

PPI for new vehicles (STR)

0.30

0.04

MPI for new vehicles (STR)

0.17

0.24

PPI for dealership markups (STR)

0.25

0.08

Input price index without markups (STR)

0.31

0.04

Input price index including markups (STR)

0.57

0.00

Note: CPI = Consumer Price Index; PPI = Producer Price Index; MPI = Import Price Index; STR = short-term price relative.
Source: Author's calculations based on data from the U.S. Bureau of Labor Statistics.

The strong graphical correlations between the official PPI for dealership markups and the estimated markup index, as well as between the CPI for new vehicles and the input
price index including markups, further demonstrate that markup increases at dealerships drove the gaps among the CPI for new vehicles, the PPI for new vehicles, and the
MPI for new vehicles. These gaps are crucial to understanding the drivers of new-vehicle consumer inflation during the COVID-19 pandemic, because they help isolate where
in the supply chain inflation occurred—in this case, at dealerships. To further corroborate the trends apparent in BLS data, chart 5 and table 3 illustrate profit-margin changes
at publicly traded dealerships from 2019 through 2022. Like BLS-derived estimates, SEC financial data for publicly traded dealerships show that markups grew from an
average of 5.0 percent in 2019 to an average of 13.1 percent in 2022. This observation further demonstrates the impact of profit-margin growth on consumer vehicle prices and
confirms the validity of the BLS-derived estimates presented earlier. Measuring profits as a percentage of the average vehicle price shows that the proportion of the consumer
price represented by dealership markups grew from 4.8 percent in 2019 to 11.5 percent in 2022.18
Chart 5. Cumulative percent change in markups for publicly traded dealerships,
first quarter 2019 to fourth quarter 2022
Percent

Industry total
AutoNation
Asbury Automotive Group
Group 1 Automotive
Lithia Motors
Sonic Automotive

300

200

100

0

-100
Q1 2019

Q3 2019

Q1 2020

Q3 2020

Q1 2021

Q3 2021

Q1 2022

Click legend items to change data display. Hover over chart to view data.
Source: Author's calculations based on U.S. Securities and Exchange Commission data.

View Chart Data

Q3 2022

Table 3. Percent change in markups for publicly traded dealerships, first quarter 2019 to fourth quarter 2022
Quarter

Industry total

AutoNation

Asbury Automotive Group

Group 1 Automotive

Lithia Motors

Sonic Automotive

Q1 2019

5.3

5.1

4.5

5.3

6.2

5.3

Q2 2019

4.9

4.7

4.1

4.7

5.9

4.9

Q3 2019

4.7

4.3

4.1

4.8

5.8

4.6

Q4 2019

5.2

4.8

4.5

5.3

6.1

5.2

Q1 2020

4.9

4.4

4.6

4.9

6.0

5.0

Q2 2020

6.0

5.6

5.3

6.4

7.2

5.3

Q3 2020

6.8

6.5

6.8

6.7

8.0

6.1

Q4 2020

7.0

6.9

7.3

7.0

7.5

6.5

Q1 2021

7.0

6.8

7.0

6.9

7.7

6.4

Q2 2021

10.2

10.3

10.0

9.8

11.1

8.8

Q3 2021

12.8

13.1

12.6

11.9

13.7

11.2

Q4 2021

14.5

14.6

14.7

13.2

15.6

13.2

Q1 2022

14.1

14.0

13.7

13.0

15.1

14.2

Q2 2022

13.8

13.7

13.4

12.8

14.4

14.3

Q3 2022

13.1

13.0

12.6

12.3

13.9

13.5

Q4 2022

11.9

12.0

12.0

11.1

12.5

11.7

Source: Author's calculations based on U.S. Securities and Exchange Commission data.

Although pandemic-related supply-chain bottlenecks and semiconductor shortages significantly affected the quantity of vehicles produced by manufacturers—and also had an
impact on producer prices for new vehicles—chart 3 shows that these disruptions had a stronger effect on consumer prices than on producer prices.19 From December 2019
through December of 2022, the compound annual rate of change for the PPI for new vehicles was 2.6 percent, which is only 0.9 percentage point higher than the average rate
of 1.7 percent during the 2007–19 business cycle. Over the same 2019–22 period, the compound annual rate of change for the MPI for new vehicles was 2.1 percent, which is
only 1.9 percentage points higher than the average rate of 0.2 percent during the 2007–19 business cycle. The increases in the PPI and MPI rates of change are overshadowed
by the change in trend for the CPI for new vehicles. From December 2019 through December 2022, the CPI for new vehicles grew at a compound annual rate of 6.0 percent,
which is 5.3 percentage points higher than the average rate of 0.7 percent during the 2007–19 business cycle. These stark differences further demonstrate that increases in
dealership profit margins were a stronger driver of consumer price changes than were manufacturer price increases.
Trends in vehicle production, total sales, and vehicle inventories help explain the subdued price transmission from manufacturers to consumers, and from consumers to
producers, in the automotive supply chain during the 2020–22 business cycle. Chart 6 shows that dealership inventories shrank substantially during the recent economic
expansion. Shortly after stimulus checks were issued in two rounds in late 2020 and early 2021 (the first stimulus round occurred in April 2020), dealership sales surged in
March and April of 2021, and monthly sales continued to remain above their pandemic lows during the remainder of 2021. However, chart 6 shows that dealerships did not
substantially increase their orders of new vehicles from manufacturers, which led to shrinking inventories and subdued production. The drop in inventories and the increase in
sales that occurred shortly after the second and third stimulus payments coincide with the rapid increase in consumer prices shown in charts 3 and 4.
Chart 6. Cumulative percent change in manufacturer new and unfilled orders for
motor vehicles and parts, total vehicle sales, and domestic auto production and
inventories, January 2020–December 2022
Percent
60

Manufacturer new orders for motor vehicles and parts
Manufacturer unfilled orders for motor vehicles and parts
Domestic auto production
Domestic auto inventories

Total vehicle sales

40
20
0
-20
-40
-60
-80
-100
-120
Jan 2020

Jul 2020

Jan 2021

Jul 2021

Jan 2022

Jul 2022

Click legend items to change data display. Hover over chart to view data.
Note: Manufacturer new and unfilled orders for motor vehicles and parts are in millions of dollars,
monthly, seasonally adjusted. Total vehicle sales are in millions of units, monthly, seasonally adjusted
(annual rate). Domestic auto production and inventories are in thousands of units, monthly, seasonally
adjusted.
Source: U S Census Bureau U S Bureau of Economic Analysis and author's calculations

View Chart Data

Normally, dealerships maintain large vehicle inventories that are costly to manage and that can force dealers to offer competitive price cuts to consumers. However, given
supply-chain disruptions and persistently high demand, dealership inventories shrank to all-time lows in 2021 and 2022, applying upward pressure on consumer prices.20 In
January 2020, average inventories stood at 526,000 units, but by February 2022, that number had dropped to a record low of 65,000 units, a decline of 87.6 percent. In 2021,
some dealerships averaged less than 2 weeks’ worth of inventory, down from a more typical 2 to 3 months of inventory before the pandemic. AutoNation, for example,
averaged only 9 days’ worth of inventory in 2021, compared with an average of 52 days in 2019 and 60 days in 2018.21 Shrinking retail inventories are crucial to
understanding the divergence between consumer and producer vehicle prices. Instead of relying on manufacturer supplies to meet consumer demand, dealerships drew down

their existing inventories. As a result, backward demand transmission from consumer-demand increases was insufficient to generate demand increases for manufactures, and
dealerships absorbed the existing demand through markup increases and inventory drawdowns.
During the pandemic downturn and the subsequent economic expansion, new-vehicle orders, unfilled orders, and factory output were highly volatile, contributing to the
inventory drought at dealerships. Chart 6 shows that, from February 2020 through May 2020, new-vehicle orders to manufacturers collapsed as economic actors across the
supply chain anticipated a large decrease in demand. Although new-vehicle orders quickly rose to prepandemic levels in June 2020, they failed to consistently remain above
those levels after some initial spikes in the summer of 2020, despite surges in consumer demand. Additionally, because new-vehicle orders are a flow that cumulates into a
stock of total orders, the 5 below-trend months from February to May 2020 had a cumulative impact on the stock of overall orders that the industry never overcame. Factories
also shut down entirely in March 2020, with output falling 99.8 percent in a single month. Finally, when new-vehicle orders recovered, manufacturers failed to fill them
because of global supply-chain disruptions. In 2020 and 2021, unfilled orders grew and manufacturing remained flat, causing further declines in dealer inventories.
Charts 7 and 8 show that automotive sales and automotive loans increased substantially during and immediately after the three rounds of stimulus payments in 2020 and 2021.
The 1-month percent change for automotive sales jumped to an all-time high of 38.2 percent in April 2020, and this jump coincided with the initial COVID-19 stimulus.
Additionally, the 1-month percent change for automotive loans reached an all-time high of 1.4 percent in April 2021, immediately after the stimulus of March 2021. The data
strongly suggest that many consumers used their stimulus payments to support the purchase of vehicles. Personal savings and government expenditures all increased
substantially during the months of the stimulus payments. These increases also coincided with increases in vehicle prices, decreases in vehicle inventories, and increases in
vehicle sales (see charts 3 and 6).
Chart 7. One-month percent change in automotive loans, all commercial banks,
billions of dollars, seasonally adjusted, January 2019–December 2022
Percent
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
Jan 2019

Jul 2019

Jan 2020

Jul 2020

Jan 2021

Jul 2021

Jan 2022

Jul 2022

Hover over chart to view data.
Note: The periods of stimulus payments are April 2020, December 2020–January 2021, and March 2021.
Source: Author's calculations based on data from the Board of Governors of the Federal Reserve System.

View Chart Data

Chart 8. One-month percent change in total vehicle sales, millions of units,
monthly, seasonally adjusted annual rate, January 2019–December 2022
Percent
50
40
30
20
10
0
-10
-20
-30
-40
Jan 2019

Jul 2019

Jan 2020

Jul 2020

Jan 2021

Jul 2021

Jan 2022

Jul 2022

Hover over chart to view data.
Note: The periods of stimulus payments are April 2020, December 2020–January 2021, and March 2021.
Source: Author's calculations based on data from the U.S. Bureau of Economic Analysis.

View Chart Data

Taken together, the trends in consumer prices, producer prices, dealership markups, automotive inventories, automotive production, and stimulus expenditures suggest that
dealership profit-margin increases in response to stimulus-driven demand contributed substantially to new-vehicle consumer inflation in 2020 and 2021. Given the
considerable evidence that profit-margin increases caused a divergence between the PPI for new vehicles and the CPI for new vehicles, the inflationary impact of those
increases can be estimated simply as the difference between the two indexes. If the CPI for new vehicles had moved in lockstep with the PPI for new vehicles from 2019 to
2022 (meaning dealer margins were stable), it would have increased by 7.9 percent, which is 12.8 percentage points lower than the actual 20.7-percent increase (or 38.3
percent of the total change). SEC financial data show that the share of markups in a vehicle’s retail price increased by 140.9 percent during the pandemic, rising from an
average of 4.9 percent in 2019 to an average of 11.5 percent in 2022. In total, profit-margin increases were responsible for 34.7 percent of dealerships’ total increase in
revenues from new-vehicle sales. Using the implicit counterfactuals from BLS and SEC data, one can estimate that dealership markups, working through price transmission,
contributed between 34.7 and 61.7 percent of total new-vehicle consumer inflation from 2019 through 2022. Given that new vehicles account for 4.3 percent of the overall

CPI, the transmission of dealership markups to consumer prices contributed roughly 0.3 to 0.5 percentage point to the overall 15.6-percent increase in the CPI from December
2019 through December 2022.
In addition to raising new-vehicle prices, dealerships also substantially increased prices for other services. Chart 9 shows that these price increases coincided with the stimulus
payments issued in late 2020 and early 2021. The dealership service index for “other receipts,” which mostly tracks the sale of financial, insurance, and extended warranty
products, began increasing rapidly in March and April of 2021. Because products captured by the “other receipts” index are not affected by the supply chain, their price
increases are entirely due to increased demand for dealership services. In February 2022, the “other receipts” index recorded a 12-month increase of 33.5 percent, an all-time
high. The fact that prices for other dealership services and dealer markups increased at the same time further demonstrates that dealerships experienced an influx of customers
in early 2021.
Chart 9. Producer Price Index for dealership services—other receipts, January
2019–December 2022
Index
170
160
150
140
130
120
110
100
90
Jan 2019

Jul 2019

Jan 2020

Jul 2020

Jan 2021

Jul 2021

Jan 2022

Jul 2022

Hover over chart to view data.
Source: U.S. Bureau of Labor Statistics.

View Chart Data

Conclusion
During the COVID-19 pandemic, dealership profit-margin increases drove new-vehicle consumer inflation and contributed modestly to overall consumer inflation. The
implicit relationships between BLS consumer and producer price data illustrate these inflationary dynamics. By relying on their existing inventories to supply consumers with
vehicles, dealerships shrank those inventories and gained more pricing power. The PPI for dealership markups is a moderator variable that bridges the gaps in the implicit
relationships among the CPI, PPI, and MPI for physical goods. These relationships may work in other industries and could offer a predictive path to estimating lagged
quarterly profits with the more timely BLS monthly price data.

Appendix: Equation sets
This appendix presents two equation sets for calculating the STR of the estimated markup index (equation set 1) and the input price index including markups (equation set 2).
Equation set 1

The STR of the estimated markup index is calculated as follows:

where

and

are the average vehicle transaction prices in, respectively, periods t – 1 and t – 2;

wholesale prices in, respectively, periods t – 1 and t – 2; and
The values of

and

and

are calculated as follows:22

The initial average vehicle wholesale price in the base period of January 2019 is calculated as

where

is 4.9 percent, and

Equation set 2

The input price index including markups is calculated as follows:

is 1.

and

are the average vehicle

are the PPI STRs for new vehicles in, respectively, periods t and t – 1.

where

and

are the PPIs for dealership markups in, respectively, periods T and T – t;

and

are the input price indexes without markups in, respectively, periods T and T – t; and

and

are the weights in, respectively,

periods T and T – t.
The input price index excluding markups is calculated as follows:

where

and

are the PPIs for new vehicles in, respectively, periods T and T – t;

respectively, periods T and T – t; and
The values of

,

In the last equation,

and
, and

and

are the MPIs for new vehicles in,

are the weights in, respectively, periods T and T – t.
are calculated as follows:

is the total dollar amount of domestically produced vehicles sold in the United States in period T, and

is the total dollar amount of

imported vehicles sold in the United States in period T.
SUGGESTED CITATION:

Michael Havlin, "Automotive dealerships 2019–22: dealer markup increases drive new-vehicle consumer inflation," Monthly Labor Review, U.S. Bureau of Labor Statistics, April 2023,

https://doi.org/10.21916/mlr.2023.7

Notes

1 Kevin M. Camp, Michael Havlin, and Sara Stanley, “Automotive dealerships 2007–19: profit-margin compression and product innovation,” Monthly Labor Review, October 2022,

https://doi.org/10.21916/mlr.2022.26.
2 This article focuses on new-vehicle prices because supply-chain disruptions most proximally affected the manufacturing of new vehicles, rather than immediately causing shortages in the
used-vehicle market. Used vehicles played an important role in the used-car shortage through their substitutability with new vehicles and merit attention in a separate analysis; however, the
dynamics of the used-car market are quite distinct from those of the new-car market, and the used-car shortage is subordinate to the initial supply-chain shocks.

3 The U.S. Bureau of Labor Statistics (BLS) Producer Price Index (PPI) program publishes several margins indexes for dealer services, including an index for total vehicle sales, new-vehicle
sales, and used-vehicle sales. This analysis uses the BLS index for total vehicle sales because corroborating evidence from corporate financial data (see chart 5) shows that this index was a
much better estimator of new-vehicle margins during the pandemic.

4 This relationship is demonstrated in Camp, Havlin, and Stanley, “Automotive dealerships 2007–19.”
5

See ibid. Company-specific information is from the 10-K forms filed with the U.S. Securities and Exchange Commission (SEC), which are stored in the SEC EDGAR database

(https://www.sec.gov/edgar/search/).

6 Camp, Havlin, and Stanley, “Automotive dealerships 2007–19.”
7 David Coffin, Dixie Downing, Jeff Horowitz, and Greg LaRocca, “The roadblocks of the COVID-19 pandemic in the U.S. automotive industry,” Working Paper ID-091 (U.S. International Trade
Commission, June 2022), https://www.usitc.gov/publications/332/working_papers/final_the_roadblocks_of_the_covid-19_pandemic_in_the_automotive_industry.pdf.

8 Christian Zimmermann, “Clocking the sales of cars and homes,” The FRED Blog (Federal Reserve Bank of St. Louis, July 23, 2018), https://fredblog.stlouisfed.org/2018/07/clocking-

the-sales-of-cars-and-homes/?utm_source=series_page&utm_medium=related_content&utm_term=related_resources&utm_campaign=fredblog.
9 Automotive manufacturers commonly seek to move excess inventories to dealerships because these inventories are costly to maintain.
10

“Personal saving” (FRED, Federal Reserve Bank of St. Louis, February 23, 2023), https://fred.stlouisfed.org/series/PSAVE.

11 Kristen Tauber and Willem Van Zandweghe, “Why has durable goods spending been so strong during the COVID-19 pandemic?” (Federal Reserve Bank of Cleveland, July 7, 2021),

https://www.clevelandfed.org/publications/economic-commentary/2021/ec-202116-durable-goods-spending-during-covid19-pandemic.
12 The estimated markup index was generated by inflating the average consumer and producer vehicle prices in 2019 by the CPI for new vehicles and the PPI for new vehicles, subtracting
those products from each other, and calculating a cumulative percent change from the derived margin. The dollar amounts in 2019 are algebraically and mathematically irrelevant to the results,
so they simply serve to make the linear combination more intuitive for the reader. The determinative assumption in this recreation is the base-period margin, which is assumed to be 5.0
percent, pursuant to SEC data. Although the calculation of a proper input index should also include the import index, the latter was excluded because (1) the import prices of vehicles trended
with producer prices, (2) vehicle imports had a small weight, and (3) the inclusion of the import index would have introduced complexity without changing the results.

13 These data are from the 10-K forms filed with the SEC, which are stored in the SEC EDGAR database (https://www.sec.gov/edgar/search/).
14 Michael Havlin, “From wholesalers to gas tanks: with gasoline, two plus two really does equal four,” Monthly Labor Review (forthcoming); and Jayson Pollock and Jonathan C. Weinhagen,
“A new BLS satellite series of net inputs to industry price indexes: methodology and uses,” Monthly Labor Review, September 2020, https://doi.org/10.21916/mlr.2020.22.

15 Pollock and Weinhagen, “A new BLS satellite series of net inputs to industry price indexes.”
16 Here, margin percentages, rather than markups, should be used as weights because the weights reflect how a markup relates to its proportion of the final price.
17 Don A. Fast and Susan E. Fleck, “Unit values for import and export price indexes: a proof of concept,” in Katharine G. Abraham, Ron S. Jarmin, Brian C. Moyer, and Matthew D. Shapiro,
eds., Big data for twenty-first-century economic statistics (Chicago, IL: University of Chicago Press, 2022), pp. 275–295, https://www.bls.gov/mxp/data/unit-values-import-export-price-

indexes.pdf; and Don Fast, Susan E. Fleck, and Dominic A. Smith, “Unit value indexes for exports—new developments using administrative trade data,” Journal of Official Statistics, Sciendo,
vol. 38, no. 1, March 2022, pp. 83–106, https://doi.org/10.2478/jos-2022-0005.
18 Company-specific information is from the 10-K forms filed with the SEC, which are stored in the SEC EDGAR database (https://www.sec.gov/edgar/search/).
19 Coffin, Downing, Horowitz, and LaRocca, “The roadblocks of the COVID-19 pandemic in the U.S. automotive industry.”
20

See Hannah Lutz, “Auto dealers find inventory loans more costly,” Automotive News, April 8, 2019, https://www.autonews.com/nada/auto-dealers-find-inventory-loans-more-costly;
and Ayelet Israeli, Fiona Scott-Morton, Jorge Silva-Risso, and Florian Zettelmeyer, “How market power affects dynamic pricing: evidence from inventory fluctuations at car dealerships,”
Management Science, vol. 68, no. 2, February 2022, pp. 895–916, https://www.hbs.edu/faculty/Pages/item.aspx?num=59497.

21 AutoNation 2019 10-K form (annual report) retrieved from SEC EDGAR database (https://www.sec.gov/ix?

doc=/Archives/edgar/data/0000350698/000035069820000042/an10k2019.htm).
22 The starting average vehicle transaction price and the starting average vehicle wholesale price are entirely irrelevant for the equation’s result. The determinative assumption is the initial
percent difference between the average vehicle transaction price and the average vehicle wholesale price, and this difference is assumed to be 4.9 percent in the initial period (January 2019).

ABOUT THE AUTHOR

Michael Havlin
mhavlin@fmc.gov
Michael Havlin is an economist at the U.S. Federal Maritime Commission.

RELATED CONTENT

Related Articles
Automotive dealerships 2007–19: profit-margin compression and product innovation, Monthly Labor Review, October 2022.
From the barrel to the pump: the impact of the COVID-19 pandemic on prices for petroleum products, Monthly Labor Review, October 2020.
A new BLS satellite series of net inputs to industry price indexes: methodology and uses, Monthly Labor Review, September 2020.
Related Subjects
Consumer price index
Auto industry
Producer price index
Consumer expenditures
Expansions
Prices
Inflation
COVID-19

ARTICLE CITATIONS

Crossref

0

U.S. BUREAU OF LABOR STATISTICS Division of Information and Marketing Services PSB Suite 2850 2 Massachusetts Avenue NE
Washington, DC 20212-0001
Telephone:1-202-691-5200 Telecommunications Relay Service:7-1-1 www.bls.gov/OPUB Contact Us

U.S. BUREAU OF LABOR STATISTICS
Bureau of Labor Statistics

About BLS

Errata

Errata
04/24/2023 - Automotive dealerships 2019–22: dealer markup increases drive new-vehicle consumer inflation
Several numbers have been corrected in the paragraph that follows chart 8. There were also small changes to numbers in chart 5 and table 3 for the fourth quarter of 2022,
which did not affect the overall trend of the data.
Corrections made on 05/23/2023.

* For errata data prior to 2020, please see our Archived Errata Table.

For all errata information, please see our Errata Home Page.

U.S. BUREAU OF LABOR STATISTICS Postal Square Building 2 Massachusetts Avenue NE Washington, DC 20212-0001
Telephone:1-202-691-5200 Telecommunications Relay Service:7-1-1 www.bls.gov Contact Us

U.S. BUREAU OF LABOR STATISTICS
Bureau of Labor Statistics

Publications

Monthly Labor Review

HOME
Search MLR

ARCHIVES

FOR AUTHORS

ABOUT

SUBSCRIBE

GO

BEYOND BLS

Beyond BLS briefly summarizes articles, reports, working papers, and other
works published outside BLS on broad topics of interest to MLR readers.
A p r il 2 0 2 3

Remote work to blame for rise in housing prices
Summary written by: Eleni X. Karageorge
Remote work is mainly responsible for soaring home prices and rentals, according to a recent study. In “Remote work and housing demand” (Economic Letter,
Federal Reserve Bank of San Francisco, September 26, 2022), authors Augustus Kmetz, John Mondragon, and Johannes Wieland show that housing prices rose 24
percent between November 2019 and November 2021, with remote work contributing to more than 60 percent of that increase. In addition, this surge in home
prices is similar for rent prices. As of August 2022, approximately 30 percent of work in the United States is still remote work. Between November 2019 and
November 2021, remote work increased to 16 percentage points.
The shift to remote work during the pandemic led workers to search for cheaper housing and more desirable amenities. Consequently, as workers left relatively
expensive areas looking for cheaper housing in less expensive cities, the overall price of homes increased. Workers’ desire for homes in warmer climates with
more space also affected advancing home prices.
For their analysis, Kmetz and coauthors researched the relationship between the share of remote jobs in 2020 compared with the share of prepandemic remote
work. They looked at core-based statistical areas (CBSAs)—geographic areas that consist of one or more counties associated with at least one urbanized area of at
least 10,000 people connected by commuting.
To show that they had an accurate measure of migration across CBSAs, the researchers isolated the effect of remote work on housing demand, separate from the
effect of prepandemic migration. Even after adjusting for this migration, the authors estimated that an additional percentage-point increase of remote work caused
a 1.5-percent rise in home prices. By tracking migration and its effect on housing demand, the researchers found that from November 2019 to November 2021, the
surge in remote work alone increased home prices by approximately 15 percent.
Their analysis also revealed that the types of jobs available in a city matter because many jobs are not conducive to remote work. A work-from-home environment
may increase demand for housing because jobs done previously in an office environment will likely use additional space and time at home. Cities considered more
desirable for remote work saw the biggest increase in home prices because the limited supply of homes could not keep up with the influx of demand. Areas with
higher shares of remote work experienced substantially higher housing prices than those areas with less remote work.
Kmetz and coauthors conclude that the fundamentals of housing demand have changed since the pandemic and that housing prices and inflation are likely to rise
in the future as the shift to remote work becomes permanent.

U.S. BUREAU OF LABOR STATISTICS Division of Information and Marketing Services PSB Suite 2850 2 Massachusetts Avenue NE
Washington, DC 20212-0001
Telephone:1-202-691-5200 Telecommunications Relay Service:7-1-1 www.bls.gov/OPUB Contact Us

U.S. BUREAU OF LABOR STATISTICS
Bureau of Labor Statistics

Publications

Monthly Labor Review

HOME
Search MLR

ARCHIVES

FOR AUTHORS

ABOUT

SUBSCRIBE

GO

ARTICLE

April 2 0 2 3

Developing a consumption measure, with examples of use for poverty and inequality analysis: a new
research product from BLS
This study provides an update on the work being done by the U.S. Bureau of Labor Statistics (BLS) to produce a research-based consumption measure and explores its
potential use in poverty and inequality analysis. For this study, and for most U.S.-based studies in the literature, the Consumer Expenditure Surveys serve as the base to
develop the measure. The consumption measure presented in this study accounts for the flow of services from homeownership and owned vehicles, in-kind transfers from
government and private entities, and for the full value of health insurance; very few researchers have accounted for all of these. The main contribution is to provide the
literature with an update on BLS activities, which include a plan to include home production in the consumption measure and information regarding an upcoming BLS
research series. Using the measures produced in this study, we find that consumption without health insurance is 16 percent lower, on average, than total expenditures. Using
a consumption measure, with or without health insurance, results in lower poverty rates than when using measures based on total expenditures or pretax income, and
consumption for all the measures is more equal than the distributions of total expenditures or income. Using an absolute poverty measure, we find a noticeable decrease in the
poverty rate from 2019 to 2021, when measured by consumption without health insurance.
Much of the economic well-being research focused on understanding the effects of behavior and policies on utility relies on the relationship between utility and household
consumption or utility and household income. That research primarily uses income data from surveys. In the United States and other countries with more advanced
economies, income data are more readily available from surveys and thus are more commonly used as a proxy for well-being than consumption.1 Additionally, researchers
might prefer income as a measure of well-being because it reflects access to resources that could be used for consumption, whereas well-being measured by consumption
could be artificially low because of preferences. However, income is more sensitive to short-run fluctuations, whereas consumption better reflects long-term resources and is
more likely to capture disparities that result from differences across households in access to credit or the accumulation of assets.2 There is also evidence that components of
consumption that are particularly important for poor people are well captured in household surveys, while many components of income important to poor people may not be
as well captured in such surveys.3 Furthermore, some researchers have suggested that consumption is more strongly correlated with other indicators of economic well-being
than is income.4 However, rather than promoting one measure over another, other research efforts support multidimensional measures of economic well-being, thereby noting
that unidimensional measures are not sufficient.5
Although consumption may be a better measure of economic well-being than income, determining the actual consumption level for an individual or a group of individuals—a
person or family that forms a consumer unit (CU), for example—is difficult because it depends on the individual’s or group’s particular circumstances, choices, use of
purchases, and time usage.6 In addition, data limitations make the problem of valuing consumption even more difficult. For example, combining expenditures data with timeuse data has been particularly challenging.7 As a result, many researchers have used expenditures as a proxy for consumption.8 The U.S. Bureau of Labor Statistics (BLS)
Consumer Expenditure Surveys (CE) have been the primary source of expenditures data at the CU level for researchers and government agencies since the late 19th century.
And, as such, the CE is the oldest BLS product that collects household or consumer expenditures as a measure of living conditions for the United States.9
BLS has long been interested in the creation of a measure of consumption with CE data as the base,10 as have other researchers.11 For example, Fisher, Johnson, and
Smeeding (2015)12 produced a measure of consumption to study inequality, and Meyer and Sullivan (2012, 2013)13 produced a consumption measure to study poverty and
inequality. However, recent experiences with the COVID-19 pandemic have highlighted the need for a more comprehensive consumption measure than has previously been
produced. Throughout the pandemic, family and household members have played an increasingly important role in the well-being of other members of their households
through the provision of services such as childcare, eldercare, more home-cooked meals, and education, all of which are nonmarket transactions and not captured with
expenditures alone. For these transactions, a comprehensive consumption measure needs to account for the time household members spend in home production for their own
consumption. These shortcomings of expenditures as an economic measure of well-being have motivated researchers at BLS to develop a more comprehensive consumption
measure that accounts for a broader set of in-kind benefits than accounted for in previous measures and for the value of home production for own consumption.
Our first attempt at BLS to produce a consumption measure was published in the May 2022 edition of AEA Papers and Proceedings as “Building a consumption poverty
measure: initial results following recommendations of a federal interagency working group.”14 In this article, we build upon our earlier work by expanding what is included in
consumption and refining our methods to produce the measure. The primary difference between the earlier consumption measure and the one presented in this article is that
we introduce a value of health insurance; this makes our most recent consumption measure more like the ones produced by Meyer and Sullivan (2012, 2013). This enables us
to produce three consumption measures: one that excludes health insurance, one that includes health insurance capped as a percentage of consumption, and one that includes
health insurance not capped. The measure based on capping health insurance expenditures for poverty measurement is based on a recommendation made by the Interagency
Technical Working Group on Evaluating Alternative Measures of Poverty (ITWG).15 As illustrations of how the measures can be used, we produce simple means for the
consumption measures and components and inequality statistics for 2019, 2020, and 2021. The consumption-based poverty and inequality statistics are compared with
statistics based on pretax income and CE-defined total expenditures.16

In our analysis, we find that the value of CU consumption without health insurance averaged about 16 percent less than CE total expenditures over the 3-year period, and that
the values of CU consumption with health insurance (capped and uncapped) was about 14 percent more than CE total expenditures in 2019 and 2020, and about 12 percent
more in 2021. For all consumption measures, compared with total expenditures and pretax income, poverty rates are lower when based on a relative concept. The absolute
concept that we used for this study sets all poverty rates the same in 2019; such an approach allows us to see how poverty rates changed over the period while holding the base
poverty rate the same. Our results show a noticeable decrease in the poverty rate from 2019 to 2021 when it is measured by consumption without health insurance. For all 3
years, consumption distributions are more equal than distributions of total expenditures and pretax income. Relative to 2019, we find that consumption poverty fell in 2020,
with the onset of the pandemic, and that distributions of consumption became more equal. Compared with 2020, for all the consumption measures, poverty rose in 2021 when
we used a relative measure, and inequality rose but not to the same levels as in 2019.
The consumption measures presented in this article represent work in progress. Next steps in the development of the BLS research consumption measure include the addition
of the value of time for home production. During the 2021–23 period, BLS sponsored research to impute values of select home-production activities using the BLS American
Time Use Survey and other data in combination with the CE. Once the value of home production is included, the goal of BLS to produce a more comprehensive measure of
consumption can be realized. In addition, BLS plans to publish a consumption-measure research series based on internal CE data and imputation; this series will be available
to the public through published tables of means. The series will not be an official BLS production series, but it can be referred to as an official research series. Depending upon
available resources, BLS plans to release an auxiliary public-use data file that includes research-based consumption values.

Background and related literature
In standard economic models, individual utility is a function of consumption. In the life-cycle model, individuals choose the level of consumption in each period to maximize
utility subject to lifetime income. The consumption level in a given period can be more or less than the income level in that period because individuals can use savings and
borrowing to smooth consumption over their life cycle. The implication of the life-cycle model is that income, consumption, and savings will follow a predictable pattern over
an individual’s life. Income is expected to exceed consumption during the prime working years and be less than consumption early in life and in retirement. Individuals can
also use savings and borrowing to smooth consumption in response to unexpected fluctuations in income. This view of consumption was developed by Modigliani and
Brumberg (1954) and extended by Friedman (1957).17 As noted by Jappelli and Pistaferri (2017), models of consumption developed in the 1970s through the 2000s further
account for intertemporal choice under uncertainty, as well as the tradeoff between leisure and home production.18
Given its direct relationship with utility, consumption is often considered a better unidimensional measure of well-being compared with income or expenditures. In a
framework focused on households, it is assumed that the members of the household collectively choose a level of consumption that maximizes utility for a given budget
constraint defined by available resources. Consumption can be viewed as an outcome variable reflecting what has been achieved, with income acting as one of many inputs.
This contrasts with economic well-being outcome measures that focus on what could be achieved, such as income, for example.
Many researchers have attempted to construct measures that reflect consumption. Early examples include Cutler and Katz (1991) and Slesnick (1991), who use a flow-ofservices approach to value durables consumption.19 Later studies that examined household consumption as distinct from expenditures include Johnson, Smeeding, and Torrey
(2005); Krueger and Perri (2006); Attanasio, Battistin, and Ichimura (2007); Heathcote, Perri, and Violante (2010); Aguiar and Bils (2011); Coibion, et al (2012), and
Attanasio, Hurst, and Pistaferri (2012).20 The studies from the economics literature most related to our attempt to construct a consumption measure at the household level are
those by Meyer and Sullivan (2013, 2023) and Fisher, Johnson, and Smeeding (2015).21 (The next section provides a more detailed discussion of the similarities and
differences between the Meyer-Sullivan and Fisher-Johnson-Smeeding measures and those presented in this article.)
There is a large body of literature produced by international and national organizations that provides guidance on the development of a consumption measure. At the
international level, guidance has been provided by the International Labour Organization (ILO), the Organisation for Economic Co-operation and Development (OECD), the
United Nations Economic Commission for Europe, and the World Bank.22 At the U.S. national level, the ITWG included in its 2021 report a comprehensive review of the
literature on the development of consumption measures and their use in poverty measurement.23 Then, in 2022, BLS published a report by Curtin et al. called “A conceptual
framework for the U.S. Consumer Expenditure Surveys,” which includes a measure of consumption.24

Data
In this section, we discuss the data and methods used to construct our consumption measures. We describe the methods used to produce the earlier consumption measure and
those reflecting the addition of health insurance and improvements to our methodology. We also provide a discussion of the scope of our measures and then compare our
measures with those produced by other researchers.
Consumer Expenditure Surveys and variables

The CE Interview Survey is the primary data source for the consumption measures presented in this article. The CE uses two survey instruments to collect expenditures at the
CU level: the Diary Survey and the Interview Survey. Respondents are selected to participate in only one of the two surveys. The Diary Survey focuses on collecting
expenditures for certain frequently purchased goods and services, such as food at home and food away from home, apparel, and other products and services purchased on a
regular basis. In contrast, the Interview Survey collects expenditures more comprehensively, capturing approximately 95 percent of total expenditures, which is why we use it
for our consumption measure.25 The Interview Survey is conducted throughout the year on a rolling basis and has a 3-month reference period. CUs complete the Interview
Survey up to four times at 3-month intervals.
For our analysis, we compare consumption with CE-defined total expenditures and a measure of income. To define income, we start with the CE definition of pretax income
available in the microdata files and subtract the value of Supplemental Nutrition Assistance Program (SNAP) benefits.26 This income measure matches the U.S. Census
Bureau measure of money income used to produce the official U.S. poverty statistics.27 Note that unemployment benefits are included in this definition, so the impact of
changes to benefit generosity are captured in the absolute magnitude of benefits collected in the CE from year to year. However, benefits related to the COVID-19 pandemic,
such as the Economic Impact Payments, the expanded Child Tax Credit, and other tax credits are not included in the income measure because they are income-tax
adjustments; however, the use of these credits are likely to be reflected in our consumption measures.
Consumption measure

In defining the scope of the consumption measure, we generally follow the ILO and OECD guidelines and the consensus recommendations of the ITWG.28 We deviate from
the ILO and OECD guidelines, which include education and health insurance in final consumption, and instead follow the ITWG recommendations by excluding education
but including values for health insurance. We also considered the research of others in the development of the consumption measures presented in this article.

In many countries, including the United States, household surveys are used to collect expenditure data that can be used, in part, to create a consumption measure. For most
categories, consumption will equal expenditures in a given period. However, not all components of consumption satisfy this equality. For durable goods, the value of
consumption is defined as a flow of services over the life of the product, rather than as the expenditures needed to acquire the good. International guidelines recommend that
the flow of services from housing and vehicles be represented by, for example, rental equivalence or user cost. For our measures—the earlier measure and those presented in
this article—we use rental equivalence for owner-occupied shelter as collected in the CE, and we impute elements of user costs for cars and trucks not collected in the CE (i.e.,
depreciation and the opportunity costs of capital).29
International guidelines and the ITWG also recommend that the value of in-kind benefits be counted in consumption; however, the value of these benefits is often not
collected in national household surveys. In the CE, the value of most in-kind benefits—those from the government (e.g., subsidized school meals and energy assistance) and
from employers (e.g., health insurance)—are not collected. One exception is that SNAP benefits are collected in the CE. Because SNAP benefits are administered in the form
of electronic benefit transfer (EBT), we assume that they are used like money income and are implicitly included in reported food expenditures. For other government and
employer in-kind benefits, imputed values are included in our measure of consumption. Government in-kind benefits for food, energy, and rental assistance were included in
the earlier BLS consumption measure; two of the consumption measures presented in this study include government and employer in-kind benefits in the form of health
insurance. Although we do not include home production in the current version of the consumption measure, we plan to include it in future versions.
Expenditures in categories such as education and health can be better thought of as investments rather than as providing for current consumption. Following the
recommendation of the ITWG,30 in this and in our earlier consumption measure, we exclude education expenditures. We also exclude out-of-pocket spending on medical
goods and services because the ITWG did not reach a consensus recommendation for the treatment of those expenditures. But health insurance provides current utility in the
form of risk protection, unlike expenditures for medical goods and services, which provide utility indirectly by their effect on health. The inclusion of health insurance in a
consumption measure for poverty analysis is controversial. Following the ITWG recommendation, and as presented in this article, we produce consumption measures with and
without health insurance. But one concern with including health insurance in the measure for poverty analysis is that the value of government-provided health insurance can
make up an outsized share of total consumption and almost singlehandedly push people out of poverty. To prevent overstating the effect of government-provided health
insurance, we produce a consumption measure that restricts health insurance to be no more than half of total consumption with health insurance (uncapped) and use this
measure in our poverty analysis. For consistency, we also use consumption with health insurance capped for our inequality analysis.
Several previous studies have created measures of consumption. Table 1 provides a comparison between the research consumption measures presented in this article and a
selection of other measures. The measures most closely aligned with our measure are those created by Meyer and Sullivan (2012, 2013) and Fisher, Johnson, and Smeeding
(2015).31 Although these measures are similar to ours, there are some important distinctions.32 With respect to the components of consumption, both Meyer and Sullivan and
Fisher, Johnson, and Smeeding used rental equivalence for owner shelter but not for vacation homes.33 We include rental equivalence for owner shelter and vacation homes in
our earlier consumption measure and in the current consumption measure. Our measures and those of Meyer and Sullivan replace reported rent with imputed market rents for
renters who report receiving government subsidies and for those who reside in public housing. Meyer and Sullivan assigned the value of rent as the maximum of the reported
rent versus the imputed rent.34 In contrast, we assign the imputed rent regardless of its size relative to the reported value. In addition, we impute rents for CUs living in rentcontrolled units, those living rent free, and those living in college dormitories. Fisher, Johnson, and Smeeding imputed government rental and public-housing subsidies and
added them to reported rents. For vehicles, Meyer and Sullivan estimated varying depreciation rates, but they did not account for opportunity cost.35 Fisher, Johnson, and
Smeeding accounted for opportunity costs but used a fixed rate (5 percent); in addition, they used a fixed rate of depreciation (10 percent).36 Finally, Meyer and Sullivan did
not include out-of-pocket health expenditures or education in their consumption measure. However, they included an imputed value of health insurance, which is capped at 30
percent of total consumption in their poverty analysis.37 We take a similar approach in this article, but we cap health insurance at 50 percent of consumption.38 In contrast,
Fisher, Johnson, and Smeeding included out-of-pocket health and education expenditures in their consumption measure.39

Table 1. Comparison of various consumption measures with those of the present study
Consumption or expenditures
Fisher,
Spending
category

Present
study[1]

Johnson,
and
Smeeding
(2015)

Nondurables

Meyer

Johnson,

and

Smeeding,

Slesnick

Sullivan

and Torrey

(2001)

(2013)

(2005)

Cutler
and Katz
(1991)

Unknown

Attanasio,

Coibion,

Aguiar

Heathcote,

Hurst, and

Gorodnichenko,

and

Perri, and

Pistaferri

Kueng, and Silvia

Bils

Violante

(2012)

(2012)

(2011)

(2010)

Attanasio,
Battistin,

Krueger

and

and Perri

Ichimura

(2006)

(2007)

Hassett
and
Mathur
(2012)

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Unknown

Yes

Yes

Yes

Yes

Yes

Yes

X

X

Yes

X

X

Yes

Unknown

X

X

X

X

X

X

X

X

X

X

X

X

Unknown

Yes

Yes

Yes

Yes

Yes

Yes

X

X

Yes

X

X

Yes

Unknown

Partial

Yes

Yes

Yes

Yes

Yes

Partial

X

X

X

Yes

Yes

Unknown

Partial

Yes

Yes

Yes

Yes

Yes

Yes

Yes

X

X

X

Yes

Unknown

Yes

X

X

X

X

X

X

X

X

X

X

X

Unknown

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

X

Yes

Yes

Unknown

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Unknown

Yes

Yes

Yes

Yes

Yes

Yes

Partial

Yes

Yes

X

X

Yes

Unknown

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Unknown

Yes

Yes

Yes

X

Yes

Yes

X

X

X

X

X

Yes

Unknown

X

X

X

Yes

X

X

X

X

Yes

X

X

X

Unknown

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Unknown

Yes

Yes

Yes

Yes

Yes

Yes

Unknown

X

Yes

X

Yes

Yes

Unknown

Insurance

Yes

Yes

Yes

Yes

Yes

Yes

Yes

X

Yes

X

X

Yes

Unknown

Vehicle rental

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

X

Yes

Yes

Unknown

Public
transportation

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Unknown

X

Yes

X

Yes

Yes

X

X

Yes

Yes

X

Yes

Unknown

Yes

X

Yes

X

X

X

X

X

X

X

X

X

Unknown

Fees and
admissions

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Unknown

Durable
equipment

Partial

Yes

Yes

Yes

Yes

Yes

X

X

Yes

Yes

X

Yes

Unknown

Food at home
Food away from
home
Alcoholic
beverages
Housing
Rental
equivalence for
owned home
Mortgage
interest and
principal
Rent for
renters
Maintenance,
repair, and
insurance
Other lodging
Rental
equivalence
for vacation
home
Utilities (e.g.,
electricity,
water, etc.)
Household
operations
(e.g., cleaning)
Home
furnishings
and equipment
Apparel
Transportation
Service flow
from owned
vehicles
Net outlays for
vehicles
Gasoline and
motor oil
Maintenance
and repair

Out-of-pocket
health

Yes

expenditures
Imputed value of
health insurance
Entertainment

[1] The measure presented in this study is similar to the one presented in Armstrong, et al., with the exceptions that the previous measure included expenditures for individuals outside the
consumer unit (e.g., gifts) and did not include a value of health insurance or a flow-of-services housing services for consumers living rent free or in college or university dormitories. See
Grayson Armstrong, Caleb Cho, Thesia I. Garner, Brett Matsumoto, Juan Munoz, and Jake Schild, “Building a consumption poverty measure: initial results following recommendations of a
federal interagency working group,” AEA Papers and Proceedings, vol. 112, May 2022, pp. 335–39, https://doi.org/10.1257/pandp.20221041.
Key: “Yes” means item is included, “X” means item is excluded, and “Partial” means that only part of the category is included in the measure. CE = Consumer Expenditure Surveys; CPS =
Current Population Survey; PSID = Panel Study of Income Dynamics.
Note: The information in this table, with the exception of the column for the present study, is taken from Jonathan Fisher, David S. Johnson, and Timothy M. Smeeding, “Inequality of
income and consumption in the U.S.: measuring the trends in inequality from 1984 to 2011 for the same individuals,” Review of Income and Wealth, vol. 61, no. 4, December 2015, pp.
630–50, https://doi.org/10.1111/roiw.12129; see the appendix (supporting information available only in the online version), especially table A3, “Comparison of consumption definitions
by terminology used to describe the measure”; see the references in Fisher, Johnson, and Smeeding for citation information for the various studies listed in this table.

Consumption or expenditures
Fisher,
Spending
category

Present
study[1]

Johnson,
and
Smeeding
(2015)

Personal care

Nondurables

Meyer

Johnson,

and

Smeeding,

Slesnick

Sullivan

and Torrey

(2001)

(2013)

(2005)

Cutler
and Katz
(1991)

Unknown

Attanasio,

Coibion,

Aguiar

Heathcote,

Hurst, and

Gorodnichenko,

and

Perri, and

Pistaferri

Kueng, and Silvia

Bils

Violante

(2012)

(2012)

(2011)

(2010)

Attanasio,
Battistin,

Krueger

and

and Perri

Ichimura

(2006)

(2007)

Hassett
and
Mathur
(2012)

Yes

Yes

Yes

Yes

Yes

Yes

Partial

Yes

Yes

Yes

Yes

Yes

Unknown

Yes

Yes

Yes

Yes

Yes

Yes

Partial

X

Yes

Yes

Partial

Yes

Unknown

X

Yes

X

Yes

Yes

Yes

X

X

Yes

Yes

X

Yes

Unknown

Tobacco

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Unknown

Miscellaneous

Yes

Yes

Yes

Yes

Yes

Yes

Yes

X

Yes

X

X

Yes

Unknown

Life insurance

X

X

X

X

X

X

Yes

X

X

X

X

X

Unknown

Both

Both

Both

Both

Both

Unknown

Urban

Unknown

Urban

Urban

Urban

Urban

Unknown

Yes

Yes

Yes

X

Unknown

X

X

Unknown

X

X

Yes

X

Unknown

4-quarter
consumer units

X

Yes

X

Yes

X

X

X

X

Yes

Yes

X

Yes

Unknown

Age restriction

X

X

X

X

Unknown

X

25-65

Unknown

25-64

25-60

25-60

Nonelderly

X

Single females

Yes

Yes

Yes

Yes

X

X

Yes

Yes

Yes

Yes

X

Yes

Yes

CE

CE

CPS

CE

X

CPS

PSID

CE

CE

CPS

X

CE

CPS

items
Reading
materials (e.g.,
books)
Education

Sample information
Urban or rural
Includes
incomplete
income reporters

Income data
source

[1] The measure presented in this study is similar to the one presented in Armstrong, et al., with the exceptions that the previous measure included expenditures for individuals outside the
consumer unit (e.g., gifts) and did not include a value of health insurance or a flow-of-services housing services for consumers living rent free or in college or university dormitories. See
Grayson Armstrong, Caleb Cho, Thesia I. Garner, Brett Matsumoto, Juan Munoz, and Jake Schild, “Building a consumption poverty measure: initial results following recommendations of a
federal interagency working group,” AEA Papers and Proceedings, vol. 112, May 2022, pp. 335–39, https://doi.org/10.1257/pandp.20221041.
Key: “Yes” means item is included, “X” means item is excluded, and “Partial” means that only part of the category is included in the measure. CE = Consumer Expenditure Surveys; CPS =
Current Population Survey; PSID = Panel Study of Income Dynamics.
Note: The information in this table, with the exception of the column for the present study, is taken from Jonathan Fisher, David S. Johnson, and Timothy M. Smeeding, “Inequality of
income and consumption in the U.S.: measuring the trends in inequality from 1984 to 2011 for the same individuals,” Review of Income and Wealth, vol. 61, no. 4, December 2015, pp.
630–50, https://doi.org/10.1111/roiw.12129; see the appendix (supporting information available only in the online version), especially table A3, “Comparison of consumption definitions
by terminology used to describe the measure”; see the references in Fisher, Johnson, and Smeeding for citation information for the various studies listed in this table.

Imputations for consumption

In cases in which in-kind benefits are not collected as part of the CE, a value of in-kind benefits must be imputed. For some types of in-kind benefits, such as government
rental assistance and health insurance (both employer provided and government), participation is captured in the CE; for these programs, we only need to assign a value to the
benefits. For other in-kind benefits, such as the National School Lunch Program (NSLP), the Women Infants and Children (WIC) program, and the Low Income Home Energy
Assistance Program (LIHEAP), the CE does not ask about participation. For these programs, participation and benefits must be imputed. Finally, in terms of the flow of
services from housing and durable goods, as noted previously, the CE asks about rental equivalence for owner-occupied housing; however, rental equivalence is not asked for
vehicles. Instead, we apply a user-cost approach to impute values of the service flows arising from vehicle (i.e., car and truck) ownership. For the consumption value of
consumer units who are not homeowners but are identified as renters, select imputations are also produced. In this section, we provide a summary of the imputation methods.
(See the appendix for a more detailed description of imputation methods.)
As previously noted, the CE does not collect program-participation or benefit values for NSLP, WIC, or LIHEAP. To impute benefit values to the CE, we start by modelling
participation for these programs and LIHEAP benefits as reported in the Current Population Survey Annual Social and Economic Supplement (CPS ASEC). Once
participation in these programs is imputed to CUs in the CE, a value of the in-kind benefits is assigned. This value comes from administrative sources for NSLP and WIC. For
LIHEAP, the reported benefit amounts from the CPS ASEC are used to impute a value of the benefits. Imputations for NSLP and LIHEAP for the measures presented in this
article are the same as those included in the earlier measure; however, improvements in methodology to produce WIC imputations are introduced.
In addition to owner-occupied housing, in certain instances it is expected that reported rents do not reflect the full consumption value of rented shelter. Regarding renters
living in units for which we consider their reported rents are less than market values, imputations for the value of shelter consumption are needed. We consider renters who are
paying additional rental-related expenses—for example, those for maintenance and repairs—as not reporting rents that represent their consumption. In addition, the CE
includes questions about whether the renter receives government assistance in paying rent, lives in public housing, lives in rent-controlled units, or lives rent free. However,
not asked is the value of the difference in what renters pay for shelter and the market value of similar units. We use regression models for those who pay full market value to
calculate the implied market value for those who pay less than the full value. The imputed market values represent the consumption value of rental shelter for these renters.
Improvements in the rent imputations are introduced with the current measure, and imputed rents for renters living rent-free were added.
The CE asks about different types of health insurance coverage and collects data on out-of-pocket insurance premiums associated with each type of insurance. The CE also
collects out-of-pocket expenditures on medical goods and services. As noted previously, for one of the measures presented in this article, the consumption value of health
insurance is included. To derive this value, we first impute the full value of health insurance on the basis of insurance type. For private health insurance, the full value is based
on the market price. This includes the out-of-pocket premiums that are captured by the CE and the employer contribution for employer-provided plans, or the value of any
subsidy received for individual plans. For public insurance programs, the full value is based on the cost to the government (including administrative costs).
Finally, for vehicles, we begin with the CE definition for all transportation expenditures, which includes the expenditures for the purchase of all new and used vehicles, vehicle
finance charges, gasoline and motor oil, maintenance and repairs, vehicle insurance, public transportation, and vehicle rental licenses and other charges. We remove
expenditures for the purchase of all vehicles and vehicle finance charges that were made during the 3-month reference period. These expenditures are replaced with a flow-ofservices value for cars and trucks that is imputed using a user-cost approach. The user cost is defined as the depreciation plus the opportunity cost of capital (current estimated

value of the car multiplied by an interest rate) plus maintenance and repair costs. To derive a flow of services from the ownership of vehicles, which we limit to cars and
trucks, we include imputed values of vehicle depreciation and opportunity costs of capital. Other components of user costs are already accounted for in the transportation
expenditures that we keep. We estimate vehicle depreciation using vehicle purchase information in the CE by comparing similar vehicles purchased at different ages. A
consumption flow-of-service value is imputed to CUs for the stock of cars and trucks owned for nonbusiness use. Although the flow of services from cars and trucks
was included in the earlier BLS consumption measure, improvements in the methodology have been introduced and are reflected in the current measure. Note that our user
costs will undervalue the flow of services from the stock of vehicles that are not cars or trucks; these other vehicles include, for example, planes, boats with motors, and
motorized campers.
Going from expenditures to consumption

To help readers understand the relationship between consumption and total expenditures as defined by BLS and presented in its published tables, we present a step-by-step
description. Because we are interested in the CU’s consumption level, and not the consumption among people who do not live in that CU, we first remove expenditures for
goods and services purchased to be given to someone outside the CU; these are identified as “gifts.” Next, we remove expenditures related to home ownership and the
purchase of vehicles because these values will be captured instead by the flow-of-services value for shelter and vehicles, respectively. And, as noted previously, we also
remove education and noninsurance medical expenditures because we view these as being more investment than consumption. For the measure of consumption with health
insurance (uncapped and capped), we add the imputed value of health insurance. Finally, we do not include certain expenditures that are included in CE total expenditures that
can be better thought of as financing future consumption (e.g., retirement contributions, life insurance purchases, etc.).

Methods for analysis
In this section, we present the methods used to examine the impact of moving to a consumption measure from CE total expenditures. Then, we describe the methods used to
examine the impact of using consumption as opposed to total expenditures or income for poverty and inequality analysis.
Means

We analyze five measures of expenditures and consumption: total expenditures as defined by the CE, total expenditures as defined by the CE excluding gift expenditures,
consumption without health insurance, consumption with the value of health insurance capped, and consumption with health insurance uncapped. To calculate the averages,
we pool four quarters of data starting with the second quarter of the current calendar year through the first quarter of the following year. Quarters refer to the calendar period
when the CE data were collected; for example, quarter one includes expenditures collected during interviews that took place from January to March; the reference period for
January interviews is October through December of the previous year, while March interviews reference expenditures made in December of the previous year through
February of the current year. Thus, when we refer to “quarterly” expenditures, these are actually 3-month values. For example, to calculate the quarterly averages for 2019, we
use data from the second quarter of 2019 through the first quarter of 2020. We present CU-level quarterly averages that are weighted using one-fourth of the CU weight
included in the CE data file; CE data are weighted quarterly and thus scaling by one-fourth accounts for the fact that four quarters of data are pooled together. Annual means
can be produced by taking the quarterly mean and multiplying by four, but these values will differ from the means reported in the published CE tables.40
Poverty and inequality analysis

We present results for three consumption measures: one that does not include health insurance, a second that includes health insurance capped, and a third that includes health
insurance uncapped. For comparison, we also present results for CE total expenditures and pretax income. The income measure that we use differs from the income measure
that appears in BLS published tables with CE expenditures data.41 Unlike the pretax income measure in the published CE tables, the one in this study does not include SNAP
benefits or food and rent as pay. For each measure, we create equivalized values using a three-parameter equivalence scale.42 Poverty and inequality statistics are based on
equivalized values.
For studying poverty, we produce two sets of poverty rates. The first set is based on purely relative thresholds. These thresholds are defined to be 60 percent of the median
equivalized value for each measure for each year. The second set of poverty rates is based on the implicit thresholds that result when all poverty rates are anchored to the same
rate for a single measure; this is a type of absolute measure. We anchored the poverty rates for all the measures in 2019 to be the same as the 2019 relative poverty rate for
consumption with health insurance capped. For 2020 and 2021, the implicit 2019 thresholds (based on equivalized-measure values) resulting from this anchoring are updated
to account for inflation. To approximate inflation, we create an annual index as the average of the monthly Chained Consumer Price Index for All Urban Consumers for all
items and all urban areas.43 When the respective measure for the CU is below the threshold, the CU is considered poor, and all members of that CU are considered poor. For
example, if the CU’s equivalized consumption without health insurance is below the poverty threshold for consumption without health insurance, then the CU and everyone in
it are considered poor. The poverty rates that we show refer to the percentage of people in the United States below the thresholds.
For studying inequality, we produce Gini indexes and Lorenz curves for each measure for each year. CUs are ranked on the basis of their equivalized value of each measure,
and population weights are used to produce the distributions. For each measure, the Lorenz curve is a plot of the cumulative share of the population plotted against cumulative
share of each measure. For example, 60 percent of the population accounts for 40 percent of overall consumption without health insurance.

Results
The results are calculated for 2019, 2020, and 2021 and are broken into three parts. The first section presents a discussion of the means. The second and third sections present
the results of the poverty and inequality analysis, respectively. Results for 2021 are labeled as “preliminary” because some of the underlying non-BLS data used to produce the
consumption measures are not finalized. Specifically, select 2021 health insurance values are subject to revision. In addition, WIC data are not finalized until up to a year after
initial release by the U.S. Department of Agriculture (USDA). When updates to these non-BLS data are released, we will produce revised estimates for 2021.
Means

Quarterly CU averages are presented in tables 2 and 3. Table 2 includes results for 2019, 2020, and 2021 for the three consumption measures and two expenditures measures.
Table 3 includes means for the consumption and expenditure subcomponents for 2020. (Detailed results for all 3 years are presented in table A-1 of the appendix for
comparison at the subcomponent level.)
As shown in table 2, the trend from 2019 to 2021 is similar across all five measures, although the levels are slightly different. Means are lowest for 2020 relative to 2019 and
2021; this pattern is expected because of changes in consumption and expenditure patterns during the first year of the COVID-19 pandemic.44 Quarterly means are lowest for
consumption without health insurance (from $12,158 to $13,562). The next highest means are for total expenditures that do not include those for gifts, followed by total
expenditures that include those for gifts. Quarterly means for both expenditure measures are about $2,000 higher than those for consumption without health insurance. The

quarterly means for consumption with health insurance capped and uncapped are more than $4,000 higher than the means for all years for consumption without health
insurance.

Table 2. Nominal quarterly means for measures of expenditures and consumption, 2019 to 2021
Measure

2019

2021[1]

2020

$14,717

$14,555

$16,196

CE-defined total expenditures (not including gifts)

14,509

14,386

15,955

Consumption without health insurance

12,395

12,158

13,562

Consumption with health insurance

16,792

16,767

18,062

Consumption with health insurance capped

16,716

16,675

18,004

CE-defined total expenditures

[1] Data for 2021 are preliminary.
Note: CE = Consumer Expenditure Surveys.
Source: U.S. Bureau of Labor Statistics.

Focusing on 2020, table 3 presents a complete breakdown, by subcomponent, of the quarterly means for total expenditures with and without gifts, consumption without health
insurance, and consumption with health insurance uncapped. The quarterly mean for total expenditures is $14,555. The quarterly mean drops, as expected, to $14,386, when
we remove expenditures for goods and services purchased to be given to people who live outside the CU (i.e., identified as gifts by BLS). When we move to consumption
without health insurance, the quarterly mean drops further, to $12,158. As revealed by the subcomponent means, the lower mean for total consumption without health
insurance relative to the means for both measures of total expenditures is largely due to the removal of health insurance and the deduction of out-of-pocket expenditures, as
well as switching from the net purchase price of vehicles to a flow of services. However, the decline was mitigated by increases in the consumption values of owned dwelling,
rented dwelling, and other lodging; these reflect the movement from expenditures to rental equivalence for owned housing and imputed rents for renter shelter. Including
health insurance increases the consumption measure to $16,767. Capping health insurance to 50-percent of consumption only slightly reduces the mean, to $16,675.

Table 3. Nominal quarterly means of expenditures and consumption, by subcomponent, 2020
Category

Total

Total expenditures,

Consumption without

Consumption with health

Consumption with health

expenditures

excluding gifts

health insurance

insurance uncapped

insurance capped

$14,555

$14,386

$12,158

$16,767

$16,675

[1]

[1]

11.31%

31.87%

29.86%

$2,139

$2,136

$2,142

$2,142

$2,142

112

112

112

112

112

5,037

4,995

[1]

[1]

[1]

3,133

3,114

[1]

[1]

[1]

Owned dwellings[3]

1,849

1,849

3,518

3,518

3,518

Rented dwellings[4]

1,100

1,089

1,188

1,188

1,188

184

176

308

308

308

1,049

1,044

1,047

1,047

1,047

362

357

[1]

[1]

[1]

48

48

[1]

[1]

[1]

313

309

309

309

309

493

480

[1]

[1]

[1]

72

72

[1]

[1]

[1]

422

408

408

408

408

251

237

237

237

237

2,430

2,404

[1]

[1]

[1]

1,208

1,188

[1]

[1]

[1]

[1]

[1]

829

829

829

Gasoline, other fuels, and motor oil

386

384

384

384

384

Other vehicle expenses

782

782

782

782

782

54

51

51

51

51

1,227

1,225

[1]

[1]

[1]

Health insurance[10]

918

917

[1]

4609

4517

Medical services

217

215

[1]

[1]

[1]

Prescription drugs

65

65

[1]

[1]

[1]

Medical supplies

29

28

[1]

[1]

[1]

626

615

[1]

[1]

[1]

55

55

[1]

[1]

[1]

571

560

560

560

560

Personal care products and services

68

67

67

67

67

Reading

16

15

15

15

15

284

223

[1]

[1]

[1]

76

76

76

76

76

135

124

124

124

124

Average quarterly expenditures or consumption
Percent of consumption that is imputed
Food[2]
Alcoholic beverages
Housing
Shelter

Other lodging[5]
Utilities, fuels, and public services[6]
Household operations
Child daycare expenses[7]
Out-of-pocket expenses, excluding child daycare
Household furnishings and equipment
Purchase of major kitchen appliances[8]
Out-of-pocket expenses, excluding household
furnishings and equipment
Apparel and services
Transportation
Vehicle purchases (net outlay)[9]
Depreciation and opportunity costs of owning
vehicles

Public and other transportation
Health

Entertainment
Motorized recreational vehicles (net outlay)
Out-of-pocket expenses, excluding motorized
recreational vehicles

Education[9]
Tobacco products and smoking supplies
Miscellaneous

[1] Not applicable.
[2] For consumption, includes National School Lunch Program and Women, Infants, and Children program. Also includes an adjustment for board for students who report living in a dorm.
[3] For consumption, includes rental equivalence for primary residence.
[4] For consumption, includes market value of rental units. Consumer units residing in a college dorm were assigned the national average value for dorms using data from the U.S.
Department of Education.
[5] For consumption, includes rental equivalence for vacation homes. Consumption also includes an adjustment for expenditures on dorms for students who report living in a dormitory.

[6] For consumption, includes energy assistance using Low-Income Home Energy Assistance Program..
[7] Not included in consumption because considered part of education.
[8] Not included in consumption because considered part of rental equivalence and rent.
[9] Item not included in consumption.
[10] For consumption, only the imputation for the full value of health insurance is included. For 2021, the value of health insurance is based on 2020 imputations adjusted for inflation.
[11] Definition excludes food and rent as pay. This definition differs from the definition of income used in the published Consumer Expenditure Surveys tables, which includes food and rent
as pay.
[12] Does not include the value of Supplemental Nutrition Assistance Program or food and rent as pay.
Note: CE = Consumer Expenditure Surveys.
Source: U.S. Bureau of Labor Statistics.

Category

Total

Total expenditures,

Consumption without

Consumption with health

Consumption with health

expenditures

excluding gifts

health insurance

insurance uncapped

insurance capped

1,592

1,592

[1]

[1]

[1]

121

121

[1]

[1]

[1]

1,471

1,471

[1]

[1]

[1]

564

564

[1]

[1]

[1]

CE-defined quarterly pretax income[11]

$21,156

$21,156

$21,156

$21,156

$21,156

Census Bureau-defined quarterly pretax income[12]

$21,094

$21,094

$21,094

$21,094

$21,094

131,542

131,542

131,542

131,542

131,542

20,158

20,158

20,158

20,158

20,158

52.14

52.14

52.14

52.14

52.14

People

2.47

2.47

2.47

2.47

2.47

Children under 18

0.58

0.58

0.58

0.58

0.58

Adults 65 and older

0.42

0.42

0.42

0.42

0.42

Earners

1.29

1.29

1.29

1.29

1.29

Vehicles (owned)

1.8

1.8

1.8

1.8

1.8

Vehicles (leased)

0.08

0.08

0.08

0.08

0.08

Men

47

47

47

47

47

Women

53

53

53

53

53

66

66

66

66

66

With mortgage

39

39

39

39

39

Without mortgage

27

27

27

27

27

33

33

33

33

33

Personal insurance and pensions[9]
Life and other personal insurance[9]
Pensions and Social Security[9]
Cash contributions[9]
Income:

Number of consumer units (in thousands)
Number of sample interviews
Consumer unit characteristics:
Age of reference person
Average number in consumer unit:

Vehicles:

Percent distribution:
Reference person:

Housing tenure:
Homeowner

Renter

[1] Not applicable.
[2] For consumption, includes National School Lunch Program and Women, Infants, and Children program. Also includes an adjustment for board for students who report living in a dorm.
[3] For consumption, includes rental equivalence for primary residence.
[4] For consumption, includes market value of rental units. Consumer units residing in a college dorm were assigned the national average value for dorms using data from the U.S.
Department of Education.
[5] For consumption, includes rental equivalence for vacation homes. Consumption also includes an adjustment for expenditures on dorms for students who report living in a dormitory.

[6] For consumption, includes energy assistance using Low-Income Home Energy Assistance Program..
[7] Not included in consumption because considered part of education.
[8] Not included in consumption because considered part of rental equivalence and rent.
[9] Item not included in consumption.
[10] For consumption, only the imputation for the full value of health insurance is included. For 2021, the value of health insurance is based on 2020 imputations adjusted for inflation.
[11] Definition excludes food and rent as pay. This definition differs from the definition of income used in the published Consumer Expenditure Surveys tables, which includes food and rent
as pay.

[12] Does not include the value of Supplemental Nutrition Assistance Program or food and rent as pay.
Note: CE = Consumer Expenditure Surveys.
Source: U.S. Bureau of Labor Statistics.

Poverty

Chart 1 presents the poverty rates defined using relative thresholds (set at 60 percent of each equivalized measure value). For exposition purposes, we focus on consumption
with health insurance capped, consumption without health insurance, total expenditures, and pretax income. Relative consumption poverty rates with or without health
insurance (about 11 and 14 percent, respectively) or health insurance capped (about 11 percent) are lower than relative poverty rates based on total expenditures (averaging 20
percent) and those based on pretax income (averaging 28 percent).45 Relative poverty rates for all measures fell in 2020 and increased slightly in 2021. Focusing on 2019
relative to 2021, we see that there was little to no change in relative consumption poverty.46 For consumption without health insurance, total expenditures (that include gifts)
and pretax income, there were declines in relative poverty.

Chart 1. Poverty rates for total population based on relative poverty
thresholds, 2019 to 2021
2019

2020

2021

Percent
35
30
25
20
15
10
5
0
Consumption with Consumption with
​health insurance ​health insurance
​capped
​uncapped

Consumption
​without health
​insurance

Total expenditures

Pretax income

Click legend items to change data display. Hover over chart to view data.
Note: Data for 2021 are preliminary.
Source: U.S. Bureau of Labor Statistics.

View Chart Data

Using the same relative thresholds as for the total population, chart 2 shows the poverty rates for the population who are less than 18 years of age. The rates for this younger
population relative to the rates for the total population are higher for all measures. Unlike for the total population, for which poverty increased from 2020 to 2021, for people
less than 18 years of age, there was a continued decline in consumption poverty and in total expenditures poverty in 2021.47 Income poverty increased in 2021, returning to its
2019 level. Note that the income measure we use does not include the expanded Child Tax Credit or the Economic Impact Payments issued during the pandemic. However,
receipt of these subsidies could be reflected in the consumption estimates.48
Chart 2. Poverty rates for people under age 18, using relative thresholds for
total population, 2019 to 2021
2019

2020

2021

Percent
35
30
25
20
15
10
5
0
Consumption with Consumption with
​health insurance ​health insurance
​capped
​uncapped

Consumption
​without health
​insurance

Total expenditures

Pretax income

Click legend items to change data display. Hover over chart to view data.
Note: Data for 2021 are preliminary.
Source: U.S. Bureau of Labor Statistics.

View Chart Data

Chart 3 presents the poverty rates based on an absolute concept of poverty, rather than a relative concept. Thresholds are absolute in that they are “fixed” in 2019 and updated
only for inflation. The thresholds for each measure are derived in such a way that the 2019 poverty rates for all the measures equal the 2019 relative poverty rate based on
consumption with health insurance capped. Thus, the starting poverty rates for all measures are set to equal 11.2 percent. Using this measure, we find that poverty fell from
2019 to 2021 for each of the consumption measures, with the largest drop occurring for consumption without health insurance. Poverty rates for total expenditures also
dropped, while pretax income poverty rose from 2020 to 2021.

Chart 3. Poverty rates for total population based on relative thresholds that
result in same total population 2019 consumption poverty rates, 2019 to 2021
2019

2020

2021

Percent
35
30
25
20
15
10
5
0
Consumption with Consumption with
​health insurance ​health insurance
​capped
​uncapped

Consumption
​without health
​insurance

Total expenditures

Pretax income

Click legend items to change data display. Hover over chart to view data.
Note: Data for 2021 are preliminary.
Source: U.S. Bureau of Labor Statistics.

View Chart Data

Chart 4 presents the poverty rates for children (under age 18) that are based on the same absolute thresholds as those used for the total population (with the poverty rate set at
11.2 percent). Child poverty rates are not anchored. As with the relative thresholds, poverty rates for children are higher than those for the total population. However, by 2021,
consumption-based child poverty rates had fallen more (from 2.4 to 4.0 percentage points) than did the poverty rates for the total population (from 1.0 to 2.3 percentage
points). In contrast, the pretax-income child poverty rate for 2020 to 2021 increased by more (2.0 percentage points) than did the total population income poverty rate (0.9
percentage point).
Chart 4. Poverty rates for people under age 18, using relative thresholds that
result in same total population 2019 consumption poverty rates, 2019 to 2021
2019

2020

2021

Percent
35
30
25
20
15
10
5
0
Consumption with Consumption with
​health insurance ​health insurance
​capped
​uncapped

Consumption
​without health
​insurance

Total expenditures

Pretax income

Click legend items to change data display. Hover over chart to view data.
Note: Data for 2021 are preliminary.
Source: U.S. Bureau of Labor Statistics.

View Chart Data

Inequality

Table 4 presents the Gini indexes for the different measures. Gini indexes are a summary measure of inequality. They range from 0 to 1, with higher values corresponding to
less equal distributions. Consumption is distributed more equally than total expenditures, which are distributed more equally than pretax income. Adding health insurance to
the consumption measure makes the distribution more equal. The greatest inequality across the 3 years occurred in 2019 for all measures except total expenditures. The
distributions of all the measures became more equal in 2020; but inequality increased in 2021.

Table 4. Gini indexes for consumption with health insurance (capped and uncapped), consumption without health insurance, total expenditures, and pretax
income, 2019 to 2021
Measure

2019

2021[1]

2020

Consumption with health insurance capped

0.250

0.241

0.247

Consumption with health insurance uncapped

0.247

0.239

0.245

Consumption without health insurance

0.295

0.282

0.289

Total expenditures

0.372

0.364

0.376

Pretax income

0.464

0.449

0.455

[1] Data for 2021 are preliminary.
Source: U.S. Bureau of Labor Statistics.

Chart 5 presents the Lorenz curves for 2020. Lorenz curves are a visual representation of inequality and show the cumulative share of the population (ranked by the value of
each measure from lowest to highest value) on the x-axis at or below the cumulative share of the measure as represented on the y-axis. Perfect equality is represented by the
45-degree line; the closer the Lorenz curve is to the line for perfect equality, the more equal the distribution. If a measure is equally distributed across the population, an equal
share of the population would account for an equal share of the measure—for example, 50 percent of the population would account for 50 percent of consumption. An
example of an unequal distribution is shown in chart 5 by the measure for consumption without health insurance: 50 percent of the population accounts for 30 percent of
consumption. As with the Gini index results, pretax income is the least equally distributed of the measures, and consumption with health insurance capped is the most equally
distributed. The Lorenz curves for consumption with health insurance and with health insurance capped are indistinguishable, and thus only one of the curves is presented.
Lorenz curves for 2019 and 2021 (based on preliminary data) exhibit similar patterns.
Chart 5. Lorenz curves for consumption with and without health insurance,
total expenditures, and pretax income, by own-rank deciles, 2020
Perfect equality
Consumption with health insurance capped
Consumption without health insurance
Total expenditures
Pretax income

Cumulative share
​of equivalized value
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Cumulative share of population ranked from lowest to highest value
Click legend items to change data display. Hover over chart to view data.
Source: U.S. Bureau of Labor Statistics.

View Chart Data

Conclusion
In this article, we construct an initial version of a comprehensive consumption measure for the United States and use it to study poverty and inequality from 2019 to 2021. Our
results show that the average value of consumption without including health insurance is lower than the average value of total expenditures. However, when we include health
insurance, capped or uncapped, average consumption exceeds average total expenditures. We also find that consumption poverty rates defined using relative thresholds are
lower than expenditure- or income-based poverty rates, regardless of whether health insurance is included. Consumption is also more equally distributed than expenditures or
income. Poverty rates across all our measures fell in 2020 relative to 2019, and the distributions of the measures became more equal. This general finding is likely an effect of
the COVID-19 pandemic, but verifying that claim is beyond the scope of this article.
BLS will continue to develop a comprehensive consumption measure for the United States with the CE providing the core data. The next main area of improvement is to
incorporate the value of home production. Because home production is not a market transaction and relies instead on household members spending time performing tasks that
add value to the household, we must impute time use to CUs participating in the CE. Home production includes the provision of childcare and eldercare. When purchased,
both services can represent heavy financial burdens for families, some of whom will be priced out of the market. In such cases, children and elders in these families often
receive care from other family members or friends through nonmarket transactions. We expect that the inclusion of home production in the CE will help us better understand
consumption dynamics during the COVID-19 pandemic, when many childcare centers were closed.

The consumption measures presented in this article represent research in progress. However, BLS plans to release to the public a research series of consumption values in
tables similar to the currently published data tables for CE total expenditures. These tables will include average consumption values. In addition, and dependent upon available
resources, BLS is considering making available an auxiliary public-use microlevel data file that researchers could use to reproduce these consumption measures or create new
ones of their own.49 Finally, BLS expects its ongoing development work to result in methodological improvements that will be reflected in the CE data and in the
consumption measure particularly.

Appendix: Imputation methods
In this appendix, we present the details of the imputation methods for NSLP, WIC, and LIHEAP benefits, market rents for which reported rents are expected to undervalue
consumption, health insurance, and vehicle flow of services that are used in the construction of our consumption measures.
NSLP, WIC, and LIHEAP

The basic approach for imputing these in-kind government benefits is based on that of Garner and Gudrais (2018) in their research on producing Supplemental Poverty
Measure (SPM) thresholds that account for in-kind benefits.50 For the BLS consumption measure, we draw upon more recent methods developed by BLS (described below)
and those developed by the Census Bureau to assign NSLP benefits to CPS-ASEC households for the SPM resource measure during the COVID-19 pandemic period.51
To produce the new consumption measure, the first step is to restrict, when necessary, the CPS-ASEC sample to households that are potentially eligible for in-kind programs.
The full sample is asked the LIHEAP questions, and thus there are no sample restrictions for the imputation of LIHEAP benefits. However, only subsets of CPS-ASEC
households are asked NSLP and WIC questions. The CPS ASEC only asks the NSLP-participation question of survey respondents in households with school-aged children
(ages 5 to 18) who usually ate a hot lunch at school. During CPS-ASEC data collection, if a Census Bureau field representative was asked what is meant by “usually,” the
response would be more than 50 percent of the time when the reference period was 2019; for the 2020 and 2021 reference periods, the response to what is meant by “usually”
would have been when students were in school prior to the pandemic or when schools remained open during the pandemic.52 The WIC question is only asked of households
with women ages 15 to 45 with no children (to include women who could potentially be pregnant) and women ages 15 and older with children under 5 years of age. These
same sample restrictions for the NSLP and WIC benefit program are applied to the CE, with one exception. In the CPS ASEC, the NSLP questions are only asked of
households with children who usually ate a hot meal at school; this information is not collected in the CE, which means we cannot restrict the CE sample along this
dimension. The CE and CPS-ASEC data are then pooled. In the pooled data, NSLP and WIC participation and LIHEAP benefits are not missing in the CPS ASEC, but they
are missing in the CE by design. A logistic regression model is used to impute participation when participation values are missing (in this case, the CE). For LIHEAP, a
logistic regression model is used to first impute participation and then for those with imputed receipt, an ordinary least squares regression model is used to impute LIHEAPbenefit values from the CPS ASEC to the CE. For the statistical models, the explanatory variables are demographic variables that are defined the same for the CPS ASEC and
the CE. For the imputed participation in these programs, the estimated model coefficients are used to generate participation probabilities for CE respondents, and then, using
these predicted probabilities, participation (yes or no) is assigned randomly.53
From the CPS-ASEC data, there are three possible outcomes for NSLP: (1) receive free or reduced school lunch, (2) pay full price for school meals, and (3) does not consume
school meals. All school lunches provided to children are subsidized. Thus, children in the second group are NSLP participants, but we refer to them as “paid” in that they paid
fully for their school lunches. For “free” and “reduced,” the CPS-ASEC question for NSLP participation does not distinguish between these two levels of meal support; thus,
we assign children to “free” and “reduced” by using a method similar to that used by the Census Bureau for assigning benefits to participating households when calculating
SPM resources. NSLP-program benefit values for the three categories—free, reduced, and paid—are based on data from the USDA.54
Two different methods were used to assign free and reduced benefits, one for 2019 and a different one for 2020 and 2021. The method used for 2019 is based on CU pretax
income (not including SNAP benefits) and a random assignment. If this pretax income measure is less than 150 percent of the official poverty threshold, children in the CU
are all assumed to receive free meals. CUs with an NSLP-participation assignment of free or reduced and income equal to or above this threshold are randomly assigned to
either being free or reduced NSLP participants.55 For 2019, the number of school days for which these benefits are assigned is based on the state average number of school
days in an academic year.56 In contrast, for SPM resources, the Census Bureau assigns NSLP benefits using a national average of 179 school days.
The method to assign NSLP benefits for 2020 and 2021 for the BLS consumption measure is an adaption of the method developed by Census Bureau staff when accounting
for NSLP benefits in SPM resources.57 The method accounts for both school closures and receipt of EBTs to assign school lunch values. It is based on a combination of school
operating status, SNAP receipt, and an additional question that was added to the CPS ASEC for 2020 and 2021: “Did your children continue receiving free/reduced price
meals through your school or school district if schools were closed during the pandemic?” For 2020 and 2021, free lunch NSLP benefits are assigned to consumers reporting
SNAP and with imputed NSLP participation. For 2020 only, reduced lunch NSLP benefits were assigned to CUs with imputed NSLP participation but no SNAP benefits.
There were no reduced benefits assigned for 2021. For 2020 and 2021, we only assign NSLP benefits to CUs with school-age children when they are expected to be in school,
in person. Because many schools were closed during the pandemic in 2020 and 2021, the USDA used EBTs to administer NSLP benefits. For our measure of consumption,
when EBTs were received, we assign a zero value to the NSLP benefit because we assume expenditures based on the use of the EBTs are already reflected in reported food
expenditures. This is the same assumption that underlies the treatment of SNAP in the CE because SNAP benefits are administered via EBTs and thus are considered “like
cash”; adding SNAP values to reported food expenditures would be double counting.58
For 2019 to 2021, if children in CUs were not assigned as receiving free or reduced lunch and were in school, in person, they were assumed to eat a hot lunch and to have
received the NSLP-benefit values for paid meals. For 2019, the new BLS consumption measure used the average number of school days in an academic year, by state, to
produce benefits for those who paid for their meals. For 2020 and 2021, we employed the method developed by the Census Bureau that accounted for school closings and
changes in the NSLP when imputing NSLP benefits for those who paid for their meals.
For WIC, once participation receipt is imputed to the CE, WIC benefits are assigned. Average quarterly WIC-benefit values imputed to the CE are based on the benefits,
infant-formula rebates, and infant participation. For those CE-respondent CUs who are imputed to have participants, the values for the in-kind benefits are assigned on the
basis of state level per-beneficiary averages for WIC food benefits and infant-formula rebates, as well as timing of the adoption of WIC being electronically administered via
EBT.59 For consumer units living in states that have fully implemented the distribution of WIC food benefits by EBT, only the infant-formula rebate is added to consumption.
To produce quarterly WIC benefits, average monthly values are produced and then multiplied by 3. For example, to compute the average monthly value for 2019 (represented
by the CE data for 2019 quarter two to 2020 quarter one), three-quarters of 2019 fiscal-year monthly averages are pooled with one quarter of the 2020 fiscal-year monthly
average. In contrast, for our earlier measure, fiscal year 2019 monthly averages were used. Average WIC food benefits are assigned to women and children. In addition,
average infant-formula rebates are added to WIC food benefits. The infant-formula rebate benefit is assigned to the fraction of all infants who are assumed to be only formula
fed.60 WIC benefits and participation rates are based on administrative data published by the USDA.61

Consumption of rental shelter

For the consumption measure, there are several cases for which we expect the CE-reported out-of-pocket shelter expenditures not to reflect the full consumption value of this
shelter. These include CUs receiving government rent subsidies, those living in public housing or those living in rent-controlled units, and those assumed to be paying less
than market rents for other reasons. The CE asks about participation in the government programs, but not about the value of the difference in what renters pay for shelter and
the market value of similar units, or about the market value of their units. Others who may pay less than the full consumption value of shelter include CUs consisting of people
who are living rent free and those who are paying less rent because they are providing services to the landlord in lieu of paying the full rent. In addition, it is unlikely that the
rents reported by college students living in dormitories reflect their consumption of shelter.
To impute rents for those receiving government rent subsidies, living in public housing, or living in rent-controlled units, we use the CE-participation variable for each
program. In addition, we impute rents for renters who we assume are not paying full market rents: specifically, those reporting expenditures for maintenance and repairs in
addition to reported rents. We also use the fact that some CUs consist of people living in dormitories to assign shelter consumption values.
We estimate a model of the full market rental value as a function of demographic, geographic, and housing-unit characteristics using a censored normal regression model,
where the observed rental payment is censored from above if the CU received rental assistance (i.e., received government rental assistance, lived in public housing, or lived in
a rent-controlled unit), or paid additional rent-related expenses. The estimation sample is restricted to traditional renters (i.e., respondents who report renting and do not report
participating in a government rental assistance program, and who do not report out-of-pocket expenditures beyond rent alone). The estimation sample also excludes CUs
consisting of people living in college or university dormitories. The rent used in the estimation model as the dependent variable is the logarithm of rent paid for the last month
in the reference period, rather than the quarterly reported rent. This last month’s rent is used so that the period more closely matches that of the period covered in owners’
reports of rental equivalence: how much owners think their homes would rent for currently (at the time of the interview) without furnishings and without utilities.62 The
imputed market rent multiplied by 3 is used in place of the CE-recorded rent for the last 3 months (the reference period) for all respondents who are identified as participating
in a government rental assistance program, occupying their residence without payment of rent, and those with out-of-pocket maintenance and repair expenditures in addition
to their reported rents. To derive shelter consumption for renters, out-of-pocket expenditures for tenant’s insurance are added to the rents reported by CUs assumed to be
paying full market rents, and to the imputed rents for all other renters.
The consumption shelter value for CUs that consist of people living in college or university dormitories is imputed differently. Because dormitories are distinctly different
from other rental units, we cannot apply our market-rent imputation model to these cases. Instead, we assign the national average dormitory cost, which we obtained from the
National Center for Education Statistics.63 We use this value to calculate an average monthly cost, which is then adjusted to reflect the 3-month cost for the reference period.
Because we do not expect dormitories to be rented year round, to get the 3-month reference-period cost, the monthly value is scaled by the number of months in the reference
period that overlap with the August–May school year. For example, a CU interviewed in July has a reference period of April, May, and June. This reference period has 2
months that overlap with the August–May school year, so the consumption value for dormitories during this reference period is twice the average monthly cost for a
dormitory. The same adjustment is made for the cost of board while living at college.
Health insurance

The CE Interview asks about different types of insurance coverage and measures the out-of-pocket insurance premium associated with each type of insurance. The CE also
measures out-of-pocket expenditures on medical goods and services. For people with insurance, out-of-pocket expenditures on medical goods and services will reflect
copayments and coinsurance. One limitation of the CE data on health insurance coverage is that for types of insurance other than Medicare, the only questions asked are
whether anyone in the CU is covered and the number of people who are covered. So, there is no way to identify the specific individuals covered.
For the consumption measure with health insurance, we first impute the full value of health insurance on the basis of insurance type. For private health insurance, the full
value is based on the market price. This includes the out-of-pocket premiums that are captured by the CE and the employer contribution for employer-provided plans, or the
value of any subsidy received for individual plans. For public insurance programs, the full value is based on the cost to the government (including administrative costs).
The methods and data sources used for the imputation of the full value of health insurance are based on the work of Garner, et al. (2022).64 For that project, imputed values
were added to CE health insurance expenditures to match the scope of the U.S. Bureau of Economic Analysis personal consumption expenditures (PCE) data, which, by
definition, include employer contributions to employer-provided health insurance and government insurance programs. The data sources used to impute the values for the CEPCE project are the same as those used to produce the consumption measures with health insurance (capped and uncapped) presented in this article. However, the values for
health insurance used in the production of these consumption measures and those used for the CE-PCE project are not directly comparable because the two measures differ in
scope.65
For employer-provided health insurance, we use data from the Medical Expenditure Panel Survey Insurance Component (MEPS-IC).66 MEPS-IC has data on the average
employer contribution and average employer share. For employer-provided coverage, we impute different amounts on the basis of the type of plan: single coverage, plus one,
and family. We assign individuals in the CE to different plan types on the basis of the number of people covered by the plan. So, individuals are assumed to have single
coverage if the plan covers one CU member, plus one if the plan covers two CU members, and family coverage if the plan covers more than two CU members.
People who purchase individual health insurance plans can receive subsidies in the form of a tax credit. The CE asks individuals who purchase individual plans whether they
receive a subsidy. We assign the average subsidy among those who receive a subsidy to CUs in the CE with individual plans who report receiving a subsidy. The data on the
average subsidy amounts come from the Centers for Medicare & Medicaid Services (CMS).67
For public programs, the imputed amount is the per-enrollee cost to the government, including administrative costs. For Medicare, the average cost (including administrative
costs) per beneficiary is calculated from the Trustee’s report for each of the Medicare programs: traditional Medicare (includes Parts A and B), Medicare Part C, and Medicare
Part D. These programs also have out-of-pocket premiums. There are two ways to account for out-of-pocket premiums. One way is to add the average cost less premiums to
the out-of-pocket premium amount reported in the CE. The other approach is to assign the average cost and ignore the out-of-pocket amounts. Which approach is preferable
depends on whether the variation in out-of-pocket premiums for these programs is due to variation in the amount of coverage purchased or to variation in the amount the
premium subsidized. If the coverage is the same, then adding the average cost less average premium amount will lead to a lower value for individuals who have received
subsidies for their premiums. This is the case for traditional Medicare, so we ignore the CE-reported premiums and impute a value on the basis of the average cost. We treat
Part D the same; although there is variation in Part D plans, the variation in out-of-pocket premiums due to subsidies is likely much larger. Part C can be thought of as a
bundle of traditional Medicare and supplemental coverage. For the traditional Medicare component, we impute an average cost and ignore the CE-reported Part B premiums.
Then, we add any additional premiums paid to reflect the supplemental coverage.
Other public programs are more straightforward because there is less variation in the types of coverage offered. So, we assign a single average-cost value to participants in
other public programs and ignore any out-of-pocket premiums. For Medicaid and the Children’s Health Insurance Program, the per-enrollee average cost comes from the CMS

National Health Expenditures tables.68 Per-enrollee costs for other public programs (e.g., TRICARE, Veterans Affairs benefits, and Indian Health Services) are calculated
from the budgets for each of the programs.69
Depreciation and opportunity costs of owning vehicles

For vehicles, we employ a user-cost approach to assign a consumption value that is based on the flow of services. The user cost is defined as the depreciation plus the
opportunity cost of capital (value of the car times the interest rate), plus maintenance and repair costs. We estimate the depreciation and opportunity costs; the other
components of user costs are already included in consumption as nonvehicle-purchase-related expenditures. Depreciation and opportunity costs are imputed to CUs on the
basis of vehicle ownership. For our research, we restrict vehicles to cars and trucks, with sport utility vehicles being categorized as trucks. For 2021, about 87 percent of
vehicles reported in the CE as being owned were classified as cars or trucks. The remaining 13 percent are other vehicles such as planes, boats with motors, and motorized
campers. Because the sample size for other vehicles is quite small and the CE does not collect data on characteristics for these vehicles, depreciation and opportunity costs for
“other vehicles” are not calculated and thus are not included in the consumption measure.
We estimate the vehicle depreciation rate using data on vehicles owned as reported in the CE. For all vehicles owned, the CE collects data on the vehicle characteristics,
including make, model, and year. The CE also asks when the vehicle was acquired and the purchase price of the vehicle. We estimate a depreciation rate by comparing the
purchase price for vehicles purchased at different ages while controlling for vehicle characteristics. We estimate an age-specific depreciation rate for vehicles 10 years old or
less. Vehicles over 10 years old are assumed to depreciate at a constant rate, as there are too few transactions involving older vehicles to estimate age-specific depreciation
rates. The imputed depreciation value is calculated as the current market value of the vehicle times the age-specific depreciation rate. The current market value is calculated as
the estimated new purchase price of the vehicle minus the estimated depreciation for prior years. The opportunity cost of capital is the product of an interest rate and the
estimated current market value of each car and truck owned by the CU. The interest rates are the year-specific Treasury Long-Term Average (Over 10 Years), InflationIndexed annual rates, not seasonally adjusted, published by the Federal Reserve Bank of St. Louis.70

Table A-1. Nominal quarterly means of expenditures and consumption, by subcomponent, 2019 to 2021
Total expenditures

Total expenditures,

Consumption without health

Consumption with health

excluding gifts

insurance

insurance uncapped

Category
2019
Average quarterly expenditures or consumption
Percent of consumption that is imputed
Food[3]

2020

2021[1]

$14,717 $14,555 $16,196
[2]

2019

2020

2021[1]

2019

2020

2021[1]

2019

2020

2021[1]

$14,509

$14,386

$15,955

$12,395

$12,158

$13,562

$16,792

$16,767

$18,062

[2]

[2]

[2]

[2]

[2]

12.16%

11.31%

10.95%

30.11%

31.87%

29.99%

$2,184 $2,139

$2,472

$2,175

$2,136

$2,462

$2,220

$2,142

$2,492

$2,220

$2,142

$2,492

134

112

142

134

112

142

134

112

142

134

112

142

4,862

5,037

5,367

4,815

4,995

5,312

[2]

[2]

[2]

[2]

[2]

[2]

3,010

3,133

3,300

2,987

3,114

3,723

[2]

[2]

[2]

[2]

[2]

[2]

Owned dwellings[4]

1,662

1,849

1,861

1,662

1,849

1,861

3,246

3,518

3,762

3,246

3,518

3,762

Rented dwellings[5]

1,110

1,100

1,177

1,199

1,089

1,168

1,214

1,188

1,290

1,214

1,188

1,290

238

184

263

224

176

244

332

308

362

332

308

362

1,012

1,049

1,058

1,008

1,044

1,052

1,012

1,047

1,056

1,012

1,047

1,056

390

362

404

385

357

399

[2]

[2]

[2]

[2]

[2]

[2]

75

48

60

75

48

60

[2]

[2]

[2]

[2]

[2]

[2]

315

313

344

310

309

340

310

309

340

310

309

340

451

493

604

435

480

588

[2]

[2]

[2]

[2]

[2]

[2]

65

72

97

65

72

97

[2]

[2]

[2]

[2]

[2]

[2]

386

422

508

371

408

491

371

408

491

371

408

491

310

251

363

296

237

311

296

237

311

296

237

311

2,649

2,430

2,776

2,616

2,404

2,751

[2]

[2]

[2]

[2]

[2]

[2]

1,135

1,208

1,281

1,116

1,188

1,268

[2]

[2]

[2]

[2]

[2]

[2]

[2]

[2]

[2]

[2]

[2]

[2]

854

829

794

854

829

794

Gasoline, other fuels, and motor oil

522

386

559

518

384

554

518

384

554

518

384

554

Other vehicle expenses

792

782

807

791

782

806

791

782

806

791

782

806

Public and other transportation

199

54

129

191

51

123

191

51

123

191

51

123

1,225

1,227

1,301

1,217

1,225

1,296

[2]

[2]

[2]

[2]

[2]

[2]

Health insurance[11]

886

918

927

886

917

927

[2]

[2]

[2]

4,397

4,609

4,500

Medical services

239

217

274

231

215

270

[2]

[2]

[2]

[2]

[2]

[2]

Prescription drugs

66

65

68

66

65

68

[2]

[2]

[2]

[2]

[2]

[2]

Medical supplies

34

29

32

34

28

32

[2]

[2]

[2]

[2]

[2]

[2]

643

626

801

628

615

780

[2]

[2]

[2]

[2]

[2]

[2]

26

55

83

26

55

83

[2]

[2]

[2]

[2]

[2]

[2]

617

571

718

601

560

697

601

560

697

601

560

697

Personal care products and services

96

68

102

96

67

102

96

67

102

96

67

102

Reading

14

16

19

14

15

18

14

15

18

14

15

18

316

284

269

245

223

209

[2]

[2]

[2]

[2]

[2]

[2]

77

76

84

77

76

84

77

76

84

77

76

84

126

135

149

118

124

137

118

124

137

118

124

137

1,572

1,592

1,741

1,572

1,592

1,741

[2]

[2]

[2]

[2]

[2]

[2]

129

121

119

129

121

119

[2]

[2]

[2]

[2]

[2]

[2]

1,443

1,471

1,623

1,443

1,471

1,623

[2]

[2]

[2]

[2]

[2]

[2]

506

564

608

506

564

608

[2]

[2]

[2]

[2]

[2]

[2]

CE-defined quarterly pretax income[12]

$20,726 $21,156 $21,890

$20,726

$21,156

$21,890

$20,726

$21,156

$21,890

$20,726

$21,156

$21,890

Census Bureau-defined quarterly pretax income[13]

$20,665 $21,094 $21,798

$20,665

$21,094

$21,798

$20,665

$21,094

$21,798

$20,665

$21,094

$21,798

132,068 131,542 133,653

132,068

131,542

133,653

132,068

131,542

133,653

132,068

131,542

133,653

20,406

21,280

20,158

20,406

21,280

20,158

20,406

21,280

20,158

20,406

Alcoholic beverages
Housing
Shelter

Other lodging[6]
Utilities, fuels, and public services[7]
Household operations
Child daycare expenses[8]
Out-of-pocket expenses, excluding child daycare
Household furnishings and equipment
Purchase of major kitchen appliances[9]
Out-of-pocket expenses, excluding household
furnishings and equipment
Apparel and services
Transportation
Vehicle purchases (net outlay)[10]
Depreciation and opportunity costs of owning
vehicles

Health

Entertainment
Motorized recreational vehicles (net outlay)
Out-of-pocket expenses, excluding motorized
recreational vehicles

Education[10]
Tobacco products and smoking supplies
Miscellaneous
Personal insurance and pensions[10]
Life and other personal insurance[10]
Pensions and Social Security[10]
Cash contributions[10]
Income:

Number of consumer units (in thousands)
Number of sample interviews

21,280 20,158

Consumer unit characteristics:
Age of reference person

51.62

52.14

51.87

51.62

52.14

51.87

51.62

52.14

51.87

51.62

52.14

51.87

2.46

2.47

2.44

2.46

2.47

2.44

2.46

2.47

2.44

2.46

2.47

2.44

Average number in consumer unit:
People

Total expenditures

Total expenditures,

Consumption without health

Consumption with health

excluding gifts

insurance

insurance uncapped

Category
2019
Children under 18

2020

2021[1]

2019

2020

2021[1]

2019

2021[1]

2020

2019

2021[1]

2020

0.57

0.58

0.56

0.57

0.58

0.56

0.57

0.58

0.56

0.57

0.58

0.56

Adults 65 and older

0.4

0.42

0.42

0.4

0.42

0.42

0.4

0.42

0.42

0.4

0.42

0.42

Earners

1.3

1.29

1.28

1.3

1.29

1.28

1.3

1.29

1.28

1.3

1.29

1.28

Vehicles (owned)

1.84

1.82

1.8

1.84

1.82

1.8

1.84

1.82

1.8

1.84

1.82

1.8

Vehicles (leased)

0.08

0.08

0.07

0.08

0.08

0.07

0.08

0.08

0.07

0.08

0.08

0.07

Men

48

47

47

48

47

47

48

47

47

48

47

47

Women

52

53

53

52

53

53

52

53

53

52

53

53

64

66

65

64

66

65

64

66

65

64

66

65

With mortgage

37

39

38

37

39

38

37

39

38

37

39

38

Without mortgage

27

27

27

27

27

27

27

27

27

27

27

27

34

33

33

34

33

33

34

33

33

34

33

33

Vehicles:

Percent distribution:
Reference person:

Housing tenure:
Homeowner

Renter

[1] Data for 2021 are preliminary.
[2] Not applicable.
[3] For consumption, includes National School Lunch Program and Women, Infants, and Children program. Also includes an adjustment for board for students who report living in a dorm.
[4] For consumption, includes rental equivalence for primary residence.
[5] For consumption, includes market value of rental units. Consumer units residing in a college dorm were assigned the national average value for dorms based on data from the U.S.
Department of Education.
[6]
For consumption, includes rental equivalence for vacation homes. Consumption also includes an adjustment for expenditures on dorms for students who report living in a dormitory.
[7] For consumption, includes energy assistance using Low-Income Home Energy Assistance Program.

[8] Not included in consumption because considered part of education.
[9] Not included in consumption because considered part of rental equivalence and rent.
[10] Item not included in consumption.
[11

For consumption, only the imputation for the full value of health insurance is included. For 2021, the value of health insurance is based on 2020 imputations adjusted for inflation.
[12] Definition excludes food and rent as pay. This definition differs from the definition of income used in the published Consumer Expenditure Surveys tables, which includes food and rent
as pay.

[13] Does not include the value of Supplemental Nutrition Assistance Program or food and rent as pay.
Source: U.S. Bureau of Labor Statistics.

ACKNOWLEDGMENT: We would like to thank former Commissioner of Labor Statistics William W. Beach, without whose support this research and the new
consumption measure we introduce would not have been possible. We would also like to thank Associate Commissioner Jeffrey Hill of the Office of Prices and Living
Conditions for his support throughout this project, as well as researchers Jonathan D. Fisher, David S. Johnson, Bruce D. Meyer, and James X. Sullivan, and our
colleagues at the U.S. Census Bureau: John Creamer, Liana Fox, and Emily A. Shrider. We thank all those who attended and presented at the BLS Consumption
Symposium in September 2021; the information shared and the many discussions we had greatly contributed to the development of the BLS consumption measure
presented in this article. Finally, we thank our research assistants for their excellent work in producing inputs for the consumption measure: Caleb Cho for vehicle
inputs and Juan Munoz for in-kind benefit and rent inputs.
The poverty and inequality statistics presented in this article are meant to provide only an example of how the consumption measure can be used; BLS is not producing
poverty or inequality statistics in an official capacity.

SUGGESTED CITATION:

Thesia I. Garner, Brett Matsumoto, Jake Schild, Scott Curtin, and Adam Safir, "Developing a consumption measure, with examples of use for poverty and inequality analysis: a new research product
from BLS," Monthly Labor Review, U.S. Bureau of Labor Statistics, April 2023, https://doi.org/10.21916/mlr.2023.8

Notes

1

The Luxembourg Income Study Database (LIS) and research series are excellent sources of information on income from household surveys; see the LIS Cross-National Data Center website

at https://www.lisdatacenter.org/. For countries with developing economies, consumption and consumption expenditures are more meaningful concepts for household survey respondents
than is income; thus, these measures are more often used for poverty and inequality analysis. See, for example, recent reports from the World Bank, including Piecing Together the Poverty
Puzzle, Poverty and Shared Prosperity series (Washington, DC: World Bank, 2018), https://www.worldbank.org/en/publication/poverty-and-shared-prosperity-2018.

2 See Bruce D. Meyer and James X. Sullivan, “Consumption and income inequality in the United States since the 1960s,” Journal of Political Economy, vol. 131, no. 2, February 2023,

https://doi.org/10.1086/721702; and Jonathan Fisher, David S. Johnson, and Timothy M. Smeeding, “Inequality of income and consumption in the U.S.: measuring the trends in inequality
from 1984 to 2011 for the same individuals,” Review of Income and Wealth, vol. 61, no. 4, December 2015, pp. 630–50, https://doi.org/10.1111/roiw.12129. In addition, researchers have
promoted the use of the joint distribution of income, consumption, and wealth as a better measure of well-being; see, for example, Jonathan D. Fisher, David S. Johnson, Timothy M.
Smeeding, and Jeffrey P. Thompson, “Inequality in 3-D: income, consumption, and wealth,” Review of Income and Wealth, vol. 68, no, 1, March 2022, pp. 16–42,

https://doi.org/10.1111/roiw.12509.

3 See Fernando Rios-Avila, “Quality of match for statistical matches using the American Time Use Survey 2013, the Survey of Consumer Finances 2013, and the Annual Social and Economic
Supplement 2014,” Working Paper 914 (Annandale-on-Hudson, NY: Levy Economics Institute of Bard College, September 2018), https://www.levyinstitute.org/pubs/wp_914.pdf; Laura
Wheaton, “Underreporting of means-tested transfer programs in the CPS and SIPP,” Research Report (Washington, DC: Urban Institute, February 2008),

https://www.urban.org/research/publication/underreporting-means-tested-transfer-programs-cps-and-sipp; Graton Gathright and Tyler Crabb, “Reporting of SSA program participation
in SIPP,” Working Paper (U.S. Census Bureau, 2014); Jonathan L. Rothbaum, “Comparing income aggregates: How do the CPS and ACS match the national income and product accounts,
2007–2012,” SEHSD Working Paper 2015-01 (U.S. Census Bureau, January 14, 2015); https://www.census.gov/library/working-papers/2015/demo/SEHSD-WP2015-01.html; Bruce
D. Meyer, Nikolas Mittag, and Robert M. Goerge, “Errors in survey reporting and imputation and their effects on estimates of Food Stamp Program participation,” Working Paper 25143
(National Bureau of Economic Research, October 2018), https://doi.org/10.3386/w25143; and Bruce D. Meyer and Nikolas Mittag, “Using linked survey and administrative data to better
measure income: implications for poverty, program effectiveness and holes in the safety net,” American Economic Journal: Applied Economics, vol. 11, no. 2, April 2019,

https://www.aeaweb.org/articles?id=10.1257/app.20170478.
4 See Bruce D. Meyer and James X. Sullivan, “Measuring the well-being of the poor using income and consumption,” Journal of Human Resources, vol. 38, Special Issue on Income Volatility
and Implications for Food Assistance Programs, 2003, pp. 1180-1220 https://www.jstor.org/stable/3558985; Bruce D. Meyer and James X. Sullivan, “Viewpoint: Further evidence on
measuring the well-being of the poor using income and consumption,” Canadian Journal of Economics, vol. 44, no. 1, February 2011, pp. 52–87, https://www.jstor.org/stable/41336351;
and Bruce D. Meyer and James X. Sullivan, “Identifying the disadvantaged: official poverty, consumption poverty, and the new Supplemental Poverty Measure,” Journal of Economic
Perspectives, vol. 26, no. 3, Summer 2012, pp. 111–36, https://doi.org/10.1257/jep.26.3.111.

5 See, for example, the World Bank’s Multidimensional Poverty Measure, which draws from other prominent poverty measures, particularly the Multidimensional Poverty Index (MPI) developed
by the United Nations Development Programme and Oxford University, https://www.worldbank.org/en/topic/poverty/brief/multidimensional-poverty-measure; and “Measuring wellbeing and progress: well-being research” (Organisation for Economic Co-operation and Development), https://www.oecd.org/wise/measuring-well-being-and-progress.htm.

6 In the U.S. Bureau of Labor Statistics (BLS) Consumer Expenditure Surveys (CE), a consumer unit is defined as follows: “A consumer unit comprises either: (1) all members of a particular
household who are related by blood, marriage, adoption, or other legal arrangements; (2) a person living alone or sharing a household with others or living as a roomer in a private home or
lodging house or in permanent living quarters in a hotel or motel, but who is financially independent; or (3) two or more persons living together who use their income to make joint expenditure
decisions. Financial independence is determined by the three major expense categories: Housing, food, and other living expenses. To be considered financially independent, at least two of the
three major expense categories have to be provided entirely, or in part, by the respondent.” See “Consumer Expenditure Surveys: Glossary,” entry for “consumer unit” (U.S. Bureau of Labor
Statistics, last modified February 13, 2015), https://www.bls.gov/cex/csxgloss.htm.

7 See Nancy Folbre, “Care data infrastructure: a U.S. case study,” Review of Income and Wealth, January 2023 (online version of record before inclusion in an issue),

https://doi.org/10.1111/roiw.12633.
8 Although we cannot trace the origin of this assumption, we have found examples of expenditures being used as a proxy for consumption as far back as Hall (1978), Sargent (1978), and Hall
and Mishkin (1982). These authors state that they use expenditures as a measure of consumption, but they do not provide any justification for doing so. The lack of justification suggests that
using expenditures as a measure of consumption was an accepted idea at the time. See Robert E. Hall, “Stochastic implications of the life cycle-permanent income hypothesis: theory and
evidence,” Journal of Political Economy, vol. 86, no. 6, December 1978, pp. 971–87, https://www.journals.uchicago.edu/doi/10.1086/260724; Thomas J. Sargent, “Estimation of dynamic
labor demand schedules under rational expectations,” Journal of Political Economy, vol. 86, no. 6, December 1978, pp. 1009–44, https://www.journals.uchicago.edu/doi/10.1086/260726;
and Robert E. Hall and Frederic S. Mishkin, “The sensitivity of consumption to transitory income: estimates from panel data on households,” Econometrica, vol. 50, no. 2, March 1982, pp. 461–
81, https://www.jstor.org/stable/1912638?seq=8.

9 For a review of these early efforts, see David S. Johnson, John M. Rogers, and Lucilla Tan, “A century of family budgets in the United States,” Monthly Labor Review, May 2001,

https://www.bls.gov/opub/mlr/2001/05/art3full.pdf.
10 For an updated version of the conceptual framework for the CE, see Scott Curtin, Adam Safir, Thesia I. Garner, Brett Matsumoto, and Jake Schild, “A conceptual framework for the U.S.
Consumer Expenditure Surveys” (U.S. Bureau of Labor Statistics, last modified October 7, 2022), https://www.bls.gov/cex/research_papers/garner-et-al-conceptual-framework-for-

CE.htm. An earlier version (September 2000) is available from the authors upon request.
11 For a select list of references, see those listed in “Consumer Expenditure Surveys: Consumption Research” (U.S. Bureau of Labor Statistics, last modified December 27, 2022),

https://www.bls.gov/cex/consumption-research.htm.
12 Fisher, Johnson, and Smeeding, “Inequality of income and consumption in the U.S.: measuring the trends in inequality from 1984 to 2011 for the same individuals.”
13

See Bruce D. Meyer and James X. Sullivan, “Winning the war: poverty from the Great Society to the Great Recession,” Brookings Papers on Economic Activity, Fall 2012, pp. 133–83,

https://www.brookings.edu/bpea-articles/winning-the-war-poverty-from-the-great-society-to-the-great-recession/; and Bruce D. Meyer and James X. Sullivan, “Consumption and
income inequality and the Great Recession,” American Economic Review, vol. 103, no. 3, May 2013, pp. 178–83, https://www.aeaweb.org/articles?id=10.1257/aer.103.3.178.
14 See Grayson Armstrong, Caleb Cho, Thesia I. Garner, Brett Matsumoto, Juan Munoz, and Jake Schild, “Building a consumption poverty measure: initial results following recommendations
of a federal interagency working group,” AEA Papers and Proceedings, vol. 112, May 2022, pp. 335–39, https://doi.org/10.1257/pandp.20221041.

15 See Final Report of the of the Interagency Technical Working Group on Evaluating Alternative Measures of Poverty, p. 29, https://www.bls.gov/evaluation/final-report-of-the-

interagency-technical-working-group-on-evaluating-alternative-measures-of-poverty.pdf. Although most of the work of the group was completed in 2020, the report was posted to the
BLS website in January 2021.

16 For the BLS definition of total expenditures in the CE, see “Consumer Expenditures and Income: Concepts,” Handbook of Methods (U.S. Bureau of Labor Statistics, last modified September
12, 2022), https://www.bls.gov/opub/hom/cex/.

17 See Franco Modigliani and Richard Brumberg, “Utility analysis and the consumption function: an interpretation of cross-section data,” in Kenneth K. Kurihara, ed., Post-Keynesian
Economics (New Brunswick, NJ: Rutgers University Press, 1954), pp. 388–436; and Milton Friedman, A Theory of the Consumption Function (Princeton, NJ: Princeton University Press, 1957).

18

Tullio Jappelli and Luigi Pistaferri, The Economics of Consumption: Theory and Evidence (New York, NY: Oxford University Press, 2017).

19 David M. Cutler and Lawrence F. Katz, “Macroeconomic performance and the disadvantaged,” Brookings Papers on Economic Activity, vol. 1991, no. 2, 1991, pp. 1–74,

https://doi.org/10.2307/2534589; and Daniel T. Slesnick, “The standard of living in the United States,” Review of Income and Wealth, vol. 37, no. 4, December 1991, pp. 363–86,
https://doi.org/10.1111/j.1475-4991.1991.tb00379.x.
20 David S. Johnson, Timothy M. Smeeding, and Barbara Boyle Torrey, “Economic inequality through the prisms of income and consumption,” Monthly Labor Review, April 2005,

https://www.bls.gov/opub/mlr/2005/04/art2full.pdf; Dirk Krueger and Fabrizio Perri, “Does income inequality lead to consumption inequality? Evidence and theory,” Review of Economic
Studies, vol 73, no. 1, January 2006, pp. 163–93, https://doi.org/10.1111/j.1467-937X.2006.00373.x; Orazio Attanasio, Erich Battistin, and Hidehiko Ichimura, “What really happened to
consumption inequality in the United States?,” in Ernst R. Berndt and Charles R. Hulten, eds., Hard-to-Measure Goods and Services: Essays in Honor of Zvi Griliches, National Bureau of
Economic Research Studies in Income and Wealth, vol. 67 (Chicago, IL: University of Chicago Press, 2007); Jonathan Heathcote, Fabrizio Perri, and Giovanni Violante, “Unequal we stand: an
empirical analysis of economic inequality in the United States, 1967–2006,” Review of Economic Dynamics, vol. 13, no. 1, January 2010, pp. 15–51,

https://doi.org/10.1016/j.red.2009.10.010; Mark A. Aguiar and Mark Bils, “Has consumption inequality mirrored income inequality,” Working Paper 16807 (Cambridge, MA: National Bureau
of Economic Research, February 2011), https://doi.org/10.3386/w16807; Olivier Coibion, Yuriy Gorodnichenko, Lorenz Kueng, and John Silvia, “Innocent bystanders? Monetary policy in the
U.S.,” Working Paper 18170 (Cambridge, MA: National Bureau of Economic Research, June 2012), https://doi.org/10.3386/w18170; and Orazio Attanasio, Erik Hurst, and Luigi Pistaferri,
“The evolution of income, consumption, and leisure inequality in the U.S., 1980–2010,” Working Paper 17982 (Cambridge, MA: National Bureau of Economic Research, April 2012),

https://doi.org/10.3386/w17982.

21 Meyer and Sullivan, “Consumption and income inequality and the Great Recession”; Meyer and Sullivan, “Consumption and income inequality in the U.S. since the 1960s”; and Fisher,
Johnson, and Smeeding, “Inequality of income and consumption in the U.S.”

22 See the following sources for the international publications: Report II: Household Income and Expenditure Statistics, Seventeenth International Conference of Labour Statisticians,
November 24–December 3, 2003 (International Labour Organization, 2003), https://www.ilo.org/public/libdoc/ilo/2003/103B09_182_engl.pdf; OECD Framework for Statistics on the
Distribution of Household Income, Consumption and Wealth (Paris: Organisation for Economic Co-operation and Development, 2013), https://dx.doi.org/10.1787/9789264194830-en;
United Nations Economic Commission for Europe, Guide on Poverty Measurement (New York: United Nations, 2017),

https://unece.org/DAM/stats/publications/2018/ECECESSTAT20174.pdf; and Giulia Mancini and Giovanni Vecchi, On the Construction of a Consumption Aggregate for Inequality and
Poverty Analysis (Washington, DC: The World Bank, March 2022),

https://documents1.worldbank.org/curated/en/099225003092220001/pdf/P1694340e80f9a00a09b20042de5a9cd47e.pdf. For further background information, see the following two
studies cited in the World Bank report: Angus Deaton and Salman Zaidi, “Guidelines for constructing consumption aggregates for welfare analysis,” Living Standards Measurement Study
Working Paper 135 (Washington, DC: The World Bank, May 2002), https://documents1.worldbank.org/curated/en/206561468781153320/pdf/Guidelines-for-constructing-

consumption-aggregates-for-welfare-analysis.pdf; and Jean Olson Lanjouw and Peter Lanjouw, “How to compare apples and oranges: poverty measurement based on different definitions
of consumption,” Review of Income and Wealth, vol. 47, no. 1, March 2001, pp. 25–42, https://doi.org/10.1111/1475-4991.00002.
23 See Final Report of the of the Interagency Technical Working Group on Evaluating Alternative Measures of Poverty. See also Bruce Meyer and David Johnson, “Poverty measurement for
the next generation: findings from the Interagency Technical Working Group on Evaluating Alternative Measures of Poverty,” W75-2021, webinar from Institute for Research on Poverty,
University of Wisconsin-Madison, April 21, 2021, https://www.irp.wisc.edu/resource/poverty-measurement-for-the-next-generation-findings-from-the-interagency-technical-

working-group-on-evaluating-alternative-measures-of-poverty/
24 See Curtin et al., “A conceptual framework for the U.S. Consumer Expenditure Surveys.”
25 Not collected in the Interview but in the Diary are items such as postage and prescription drugs.
26 The definition of pretax income used in the microdata differs from the definition of pretax income used in the published CE tables. Specifically, the definition used in the published CE tables
includes income from food and rent as pay, whereas the microdata definition of pretax income does not include these sources of income. Both definitions include Supplemental Nutrition
Assistance Program (SNAP) benefits.

27

See “Income and Poverty” (U.S. Census Bureau, last modified July 6, 2022), https://www.census.gov/topics/income-poverty.html.

28 See Report II: Household Income and Expenditure Statistics (International Labour Organization); OECD Framework for Statistics on the Distribution of Household Income, Consumption and
Wealth; and Final Report of the of the Interagency Technical Working Group on Evaluating Alternative Measures of Poverty.

29 For owned primary residences and owned vacation homes, we use the CE variable that is created for use in the production of the CPI based on quarterly rental equivalence values adjusted
for ownership. For owned timeshares, we use the same comparable variable but with the addition of an adjustment for duration of usage.

30 See Final Report of the of the Interagency Technical Working Group on Evaluating Alternative Measures of Poverty.
31 Meyer and Sullivan, “Winning the war: poverty from the Great Society to the Great Recession”; Meyer and Sullivan, “Consumption and income inequality and the Great Recession”; and
Johnson and Smeeding, “Inequality of income and consumption in the U.S.: measuring the trends in inequality from 1984 to 2011 for the same individuals.”

32

In addition to the differences noted above, there are also differences regarding the samples used to create the consumption measures. For example, Meyer and Sullivan included consumer
units that participated in the CE Interview during quarters that correspond to a calendar year (e.g., they use CE Interview data collected in the first calendar quarter of 2019 through the fourth
calendar quarter of 2019 for their 2019 consumption measure). Fisher, Johnson, and Smeeding restrict their sample to only those respondents with four consecutive interviews and create
annual consumption values by summing the quarterly values. Specifically, the estimation sample includes consumer units whose last interview was between April of the current year and March
of the following year, with the restriction that there were to be four interviews (e.g., 2009 annual estimates of consumption are estimated as the sum of quarterly values, with their last interview
as early as April 2009 or as late as March 2010). For the BLS measure, we base our sample on CE interviews conducted in the second quarter of the current calendar year through the first
quarter of the following calendar year to create quarterly measures of consumption for the current year (e.g., we use CE Interview data collected from second quarter 2019 through first quarter
2020 for our 2019 consumption measure). Data collected in each calendar quarter reference expenditures made during the previous 3 months; for example, data collected in the second
quarter of 2019 refer to expenditures made in the period from January through March 2019. Thus, our measure covers expenditures from January of the current year through February of the
following year (i.e., the 2019 consumption measure covers January 2019 through February of 2020).

33 Meyer and Sullivan, “Winning the war: poverty from the Great Society to the Great Recession”; Meyer and Sullivan, “Consumption and income inequality and the Great Recession”; and
Fisher, Johnson, and Smeeding, “Inequality of income and consumption in the U.S.”

34 Meyer and Sullivan, “Winning the war: poverty from the Great Society to the Great Recession”; Meyer and Sullivan, “Consumption and income inequality and the Great Recession.”
35 Ibid.
36 Fisher, Johnson, and Smeeding, “Inequality of income and consumption in the U.S.”
37

Meyer and Sullivan, “Winning the war: poverty from the Great Society to the Great Recession”; Meyer and Sullivan, “Consumption and income inequality and the Great Recession.”

38 The Interagency Technical Working Group on Evaluating Alternative Measures of Poverty (ITWG) recommends that health insurance be no more than half of total consumption. We chose to
use 50 percent, rather than 30 percent or some other value, because it is the least binding cap that is still consistent with the ITWG recommendations.

39 Fisher, Johnson, and Smeeding, “Inequality of income and consumption in the U.S.”
40 This discrepancy occurs because the published tables use integrated data from both the Interview and Diary Surveys, whereas the results presented in this article are strictly based on the
Interview Survey. Additionally, the published tables show calendar-year estimates—meaning, the expenditures used in the calculation are all within a specified calendar year. This study defines
the measure on the basis of a collection year, which we defined as the second quarter of a specified year through the first quarter of the following year; therefore, the quarterly average means
also include expenditures from outside a given calendar year. Finally, the adjustment to the quarterly weights, in order to produce the publication tables, is not exactly equivalent to the
adjustment made in this study.

41 For example, see table 1800, “Region of residence: Annual expenditure means, shares, standard errors, and coefficients of variation, Consumer Expenditure Surveys,
2021,” https://www.bls.gov/cex/tables/calendar-year/mean-item-share-average-standard-error/cu-region-1-year-average-2021.pdf. Food and rent as pay are included in the “other
income” category in the table.

42

This equivalence is the same one used by BLS to produce the Supplemental Poverty Measure (SPM) thresholds. See “Price and Index Number Research: 2021 Research Supplemental

Poverty Measure Thresholds” (U.S. Bureau of Labor Statistics, last modified June 23, 2022), https://www.bls.gov/pir/spm/spm_thresholds_2021.htm.

43 See “Consumer Price Index: Chained Consumer Price Index For All Urban Consumers (C-CPI-U)” (U.S. Bureau of Labor Statistics, last modified December 3, 2021),

https://www.bls.gov/cpi/additional-resources/chained-cpi.htm.
44 For further evidence of the impact of the pandemic, see “Changes to expenditures during the COVID-19 pandemic,” The Economics Daily, May 3, 2022,

https://www.bls.gov/opub/ted/2022/changes-to-consumer-expenditures-during-the-covid-19-pandemic.htm: “After the COVID-19 pandemic began, consumer spending in the second
quarter of 2020 was down 9.8 percent from the same period in 2019. One year later, in the second quarter of 2021, the pandemic was still affecting the economy, but businesses and
consumers had begun to adapt. That resulted in consumer expenditures that were 15.7 percent higher in the second quarter of 2021 than a year earlier.”

45 These contrast with estimates of poverty rates based on CE pretax income data and official poverty thresholds; rates are estimated to be 12.2 percent for 2019, 11.4 percent for 2020, and
11.7 percent for 2021. These rates are not much higher than the rates based on the relative thresholds and consumption with health insurance, both capped and uncapped. The CE-based
income poverty rates using official thresholds are similar to the official poverty rates published by the U.S. Census Bureau.

46 In this study, we do not examine whether the differences in poverty rates are statistically significant because no standard errors have been produced for these measures. BLS expects to
produce standard errors for them in the future.

47

This result is in line with the change in child poverty as measured by the SPM. See Kalee Burns and Liana E. Fox, “The impact of the 2021 expanded Child Tax Credit on child poverty,”

SEHSD Working Paper 2022-24 (U.S. Census Bureau, November 22, 2022), https://www.census.gov/content/dam/Census/library/working-papers/2022/demo/sehsd-wp2022-24.pdf.

48 Receipt of these payments resulted in increased expenditures, which will lead to an increase in consumption. See Jonathan Parker, Jake Schild, Laura Erhard, and David Johnson,
“Economic Impact Payments and household spending during the pandemic,” Brookings Papers on Economic Activity: BPEA Conference Drafts, September 8–9, 2022 (Washington, DC:
Brookings Institution, August 2022), https://www.brookings.edu/wp-content/uploads/2022/09/Parker-et-al-BPEA-Conference-Draft-BPEA-FA22.pdf; and Sophie M. Collyer, Thesia
Garner, Neeraj Kaushai, Jiwan Lee, Jake Schild, Jane Waldfogel, and Christopher T. Wimer, “Effects of the expanded Child Tax Credit on household expenditures: preliminary intent-to-treat
estimates from the Consumer Expenditure Survey,” BLS Working Paper 549 (U.S. Bureau of Labor Statistics, April 2022), https://www.bls.gov/osmr/research-

papers/2022/pdf/ec220040.pdf.
49 This release would be similar to the release of state weights for the CE public-use data file. See “CE research products: State Weight files” (U.S. Bureau of Labor Statistics, last modified
May 6, 2022), https://stats.bls.gov/cex/csxresearchtables.htm#stateweights.

50 The imputations of National School Lunch Program (NSLP) and Women Infants and Children (WIC) benefits that we use in this article are similar to those developed by Garner and Gudrais;
however, for the Low-Income Home Energy Assistance Program (LIHEAP), the methods differ. Garner and Gudrais did not use the reported LIHEAP values from the Current Population Survey
Annual Social and Economic Supplement (CPS ASEC); rather, they assigned benefits based on heating and cooling degree days by geography. See Thesia I. Garner and Marisa Gudrais,
“Alternative poverty measurement for the U.S.: Focus on Supplemental Poverty Thresholds.” Working Paper 510 (U.S. Bureau of Labor Statistics, September 25, 2018).

https://www.bls.gov/osmr/research-papers/2018/pdf/ec180100.pdf. References to earlier work that focused on estimating and including in-kind benefits in SPM thresholds follow: Thesia
I. Garner, Marisa Gudrais, and Kathleen S. Short, “Consistency in Supplemental Poverty Measurement: adding imputed in‐kind benefits to thresholds and impact on poverty rates for the United
States,” Working Paper (U.S. Bureau of Labor Statistics, October 2015), https://www.bls.gov/osmr/research-papers/2015/st150120.htm; Thesia I. Garner and Charles Hokayem,
“Supplemental Poverty Measure thresholds: imputing School Lunch and WIC benefits to the Consumer Expenditure Survey using the Current Population Survey,” Working Paper 457 (U.S.
Bureau of Labor Statistics, July 2012), https://www.bls.gov/osmr/research-papers/2012/ec120060.htm; and Thesia I. Garner and Charles Hokayem. Supplemental Poverty Measure
thresholds: imputing noncash benefits to the Consumer Expenditure Survey Using Current Population Survey—Parts I and II,” Working Paper, (U.S Bureau of Labor Statistics, 2011),

https://www.bls.gov/osmr/research-papers/2011/st110100.htm.
51

The U.S. Census Bureau research focuses on the assignment of NSLP benefits to CPS ASEC households to produce the SPM resource measure. See the following research papers for the
methods used to produce NSLP benefits during the COVID-19 pandemic: Em Shrider, “Alternative school lunch valuation in the CPS ASEC during COVID-19,” SEHSD Working Paper 2021-20
(U.S. Census Bureau, September 2021), https://www.census.gov/library/working-papers/2021/demo/SEHSD-WP2021-20.html; and Shrider, “School lunch and P-EBT valuation in the
2021 Supplemental Poverty Measure,” SEHSD Working Paper 2022-15 (U.S. Census Bureau, September 2022), https://www.census.gov/library/working-papers/2022/demo/SEHSD-

wp2022-15.html. To see the impact on poverty rates of including imputed in-kind benefits in resources, see John Creamer, Emily A. Shrider, Kalee Burns, and Frances Chen, Poverty in the
United States: 2021, Current Population Reports, P60-277 (U.S. Census Bureau, September 2022), https://www.census.gov/content/dam/Census/library/publications/2022/demo/p60277.pdf.
52 According to Em Shrider of the U.S. Census Bureau, when the March 2020 CPS ASEC was administered (with reference period 2019), if the respondent asked, “‘Usually’?” or “What do you
mean by ‘usually’?” the field representative (FR) would explain that it meant more than 50 percent of the time. For the CPS ASEC administered in March 2021 (reference period 2020) and
March 2022 (reference period 2021), the FR notes say that the word “usually” refers to days when school was held in person, such as during the prepandemic period, or in areas where
schools remained open during the pandemic. These directions are in the FR notes and are presented during training, but they are not made available to the public. Emily A. Shrider, email
communication with Thesia I. Garner, March 29, 2023.

53 A monotone regression method is used to impute NSLP, WIC, and LIHEAP participation and ordinary least squares regression is used to impute LIHEAP benefits to the CE. We use SAS
PROC MI for these imputations. For program participation, after the predicted probabilities are produced, a random number is then drawn for each respondent and imputation. If the random
number is less than the participation probability, the respondent is identified as participating in the respective program. For example, suppose the estimated model predicts that consumer unit j
has a 20-percent chance of participating in the NSLP. The SAS procedure will draw a random uniform [0,1] variable u. If u is less than 0.2, consumer unit j will be assigned a value of 1, and if u
is greater than 0.2, consumer unit j will be assigned a value of 0. SAS repeats this for every observation with missing values for NSLP. See “The MI procedure,” chap. 2 in SAS/STAT 14.1
User’s Guide (Cary, NC: SAS Institute, 2015), https://support.sas.com/documentation/onlinedoc/stat/141/mi.pdf.

54 Although NSLP benefits are based on data from the U.S. Department of Agriculture (USDA), the benefit values that we use are those calculated by the Census Bureau and combine the
NSLP values for meal reimbursement, bonus commodities, and entitlements. For USDA information about these programs, see “National School Lunch Programs: Rates of Reimbursement”
(U.S. Department of Agriculture, Food and Nutrition Service, last modified July 26, 2022), https://www.fns.usda.gov/cn/rates-reimbursement; and “USDA Foods in Schools: Value of
Donated Foods Notices” (U.S. Department of Agriculture, Food and Nutrition Service, last modified December 26, 2019), https://www.fns.usda.gov/usda-fis/value-donated-foods-notices.
The USDA values are presented by academic year; the Census Bureau uses data from 2 academic years to produce benefit values that align more closely with a calendar year.

55

Shrider, “Alternative school lunch valuation in the CPS ASEC during COVID-19”; see note in figure 1, p. 7.

56 For the estimation of 2019 NSLP benefits in the consumption measure, we use the average number of school days in the 2018–19 academic year, by state. For the most recent data
available relative to 2019, see table 234.20, “Minimum amount of instructional time per year and policies on textbooks, by state: selected years, 2000 through 2020,” in Digest of Education
Statistics: 2019 (National Center for Education Statistics, February 2021), https://nces.ed.gov/programs/digest/d19/tables/dt19_234.20.asp .

57 Shrider, “Alternative school lunch valuation in the CPS ASEC during COVID-19”; and Shrider, “School lunch and P-EBT valuation in the 2021 Supplemental Poverty Measure.”
58 Ibid. In contrast to the treatment of NSLP benefits that are administered by electronic benefit transfer (EBT) being set to zero, when producing the SPM resource measure, the Census
Bureau method assigns NSLP benefits that were administered by EBTs during months when schools were assumed to be closed.

59 See “WIC EBT Activities” (U.S. Department of Agriculture, Food and Nutrition Service, last updated December 2022), https://www.fns.usda.gov/wic/wic-ebt-activities.
60

This fraction is calculated as the number of WIC infants who are fully formula fed divided by the total number of infants participating in WIC. As evidence that most WIC infants are formula
fed and receive the fully formula-fed package and not the partially formula-fed package, in their 2018 food package report, authors Nicole Kline, Kevin Meyers Mathieu, Jeff Marr write, “Fully
formula-fed packages were prescribed for 30.9 to 66.0 percent of infants younger than 6 months old (800 ounces or more), and 54.5 percent of infants aged 6 months or older (at least 600
ounces). Quantities prescribed in the fully formula-fed [full nutrition benefit to maximum monthly allowance] ranges were most common across all infant age groups. Partially breastfed
packages were prescribed for 8.5 percent of 0- to 0.9-month-old infants (less than 200 ounces), 13.7 percent of 1- to 3.9-month-old infants and 10.9 percent of 4- to 5.9-month-old infants (at
least 200 but less than 600 ounces), and 4.8 percent of infants aged 6 months or older (at least 200 but less than 400 ounces).” See Kline, Mathieu, and Marr, WIC Participant and Program
Characteristics 2018 Food Packages and Costs Report (Alexandria, VA: U.S. Department of Agriculture, Food and Nutrition Service, 2020), p. 21, https://fns-

prod.azureedge.us/sites/default/files/resource-files/WICPC2018FoodPackage-1.pdf.
61 For 2019 through 2021, see “WIC Data Tables: Monthly Data—State Level Participation by Category and Program Costs” (U.S. Department of Agriculture, Food and Nutrition Service, last
updated February 10, 2023), https://www.fns.usda.gov/pd/wic-program.

62 In the CE, owners are asked the following question: “If someone were to rent your home today, how much do you think it would rent for monthly, unfurnished and without utilities?” To derive
the quarterly consumption of owner-occupied shelter using responses to this question, we multiply by 3.

63 The data we used are from table 330.10, “Average undergraduate tuition and fees and room and board rates charged for full-time students in degree-granting postsecondary institutions, by
level and control of institution: 1963–64 through 2015–16,” Digest of Education Statistics: 2016 (National Center for Education Statistics, February 2018),

https://nces.ed.gov/programs/digest/d16/tables/dt16_330.10.asp.
64 Thesia I. Garner, Robert S. Martin, Brett Matsumoto, and Scott Curtin, “Distribution of U.S. personal consumption expenditures for 2019: a prototype based on Consumer Expenditure
Survey data,” Working Paper 557 (U.S. Bureau of Labor Statistics, August 8, 2022), https://www.bls.gov/osmr/research-papers/2022/ec220120.htm.

65

For example, care provided in government-operated facilities is out of scope in the personal consumption expenditures data, so the imputed values for Department of Veterans Affairs (VA)
services, TRICARE (the health care program for uniformed service members, retirees, and their families), and Indian Health Services (IHS) only include the cost of care purchased from private
providers. By contrast, the cost of the care provided in government facilities is in scope for the BLS consumption measure presented in this article. For more information on the personal
consumption expenditures data, see “Consumer Spending” (U.S. Bureau of Economic Analysis, last updated February 24, 2023), https://www.bea.gov/data/consumer-spending/main.

66 See “Medical Expenditure Panel Survey Insurance Component” (Agency for Healthcare Research and Quality), https://datatools.ahrq.gov/meps-ic.
67 See U.S. Department of Health and Human Services, Centers for Medicare & Medicaid Services, https://www.cms.gov/.
68 Ibid.
69 For budget information on TRICARE, see “Under Secretary of Defense (Comptroller): Defense Budget Materials” (U.S. Department of Defense, last updated April 12, 2022),

https://comptroller.defense.gov/Budget-Materials/; for VA benefits, see Department of Veterans Affairs FY2022 Appropriations, Report 46964 (Congressional Research Service, last
updated June 28, 2022), https://crsreports.congress.gov/product/pdf/R/R46964; for IHS benefits, see U.S. Department of Health and Human Services, Indian Health Services,
“Congressional Justifications,” https://www.ihs.gov/budgetformulation/congressionaljustifications/.
70

See “Economic Research,” Federal Reserve Bank of St. Louis, https://fred.stlouisfed.org.

ABOUT THE AUTHOR

Thesia I. Garner
garner.thesia@bls.gov
Thesia I. Garner is the chief research economist in the Office of Prices and Living Conditions, U.S. Bureau of Labor Statistics.
Brett Matsumoto
matsumoto.brett@bls.gov
Brett Matsumoto is a research economist in the Office of Prices and Living Conditions, U.S. Bureau of Labor Statistics.
Jake Schild
schild.jake@bls.gov
Jake Schild is a research economist in the Office of Prices and Living Conditions, U.S. Bureau of Labor Statistics.
Scott Curtin
curtin.scott@bls.gov
Scott Curtin is a branch chief in the Office of Prices and Living Conditions, U.S. Bureau of Labor Statistics.
Adam Safir
safir.adam@bls.gov
Adam Safir is a division chief in the Office of Prices and Living Conditions, U.S. Bureau of Labor Statistics.

RELATED CONTENT

Related Articles
A conceptual framework for the U.S. Consumer Expenditure Surveys, Program Series Report, September 2022.
Building a consumption poverty measure: initial results following recommendations of a federal interagency working group, BLS Working Paper 548, March 2022.
Economic inequality through the prisms of income and consumption, Monthly Labor Review, April 2005.
A century of family budgets in the United States, Monthly Labor Review, May 2001.
Related Subjects
Income
Expenditures
Consumer expenditures
Economic and Social Statistics

ARTICLE CITATIONS

Crossref

0

U.S. BUREAU OF LABOR STATISTICS Division of Information and Marketing Services PSB Suite 2850 2 Massachusetts Avenue NE
Washington, DC 20212-0001
Telephone:1-202-691-5200 Telecommunications Relay Service:7-1-1 www.bls.gov/OPUB Contact Us

U.S. BUREAU OF LABOR STATISTICS
Bureau of Labor Statistics

Publications

Monthly Labor Review

HOME
Search MLR

ARCHIVES

FOR AUTHORS

ABOUT

SUBSCRIBE

GO

ANNOUNCEMENT

A p r il 2 0 2 3

Are you a college student with an article to publish?
Are you a college student with interesting research that you’d like to publish? If so, the U.S. Bureau of Labor Statistics is pleased to share a new opportunity
for students to publish their work in a well-known and respected journal, the Monthly Labor Review (MLR), in a new pilot project called the Student MLR.
What is the Student MLR? The Student MLR is pilot project dedicated to publishing social science research by undergraduate students. The Student MLR
provides an opportunity for students to refine their analytical abilities, receive comments from experienced professionals, develop their research conjectures for
graduate study, gain professional experience, and produce new knowledge. Subjects that the Student MLR publishes include, but are not exclusive to,
demographics, labor economics, prices, environment, community research, and social statistics.
Want more information? Please join us for an online information session on April 25, 2023, at 2:00 PM Eastern Time. Registration is required and available
online at https://www.eventbrite.com/e/bls-student-mlr-spring-information-session-tickets-590158970367.

U.S. BUREAU OF LABOR STATISTICS Division of Information and Marketing Services PSB Suite 2850 2 Massachusetts Avenue NE
Washington, DC 20212-0001
Telephone:1-202-691-5200 Telecommunications Relay Service:7-1-1 www.bls.gov/OPUB Contact Us

U.S. BUREAU OF LABOR STATISTICS
Bureau of Labor Statistics

Publications

Monthly Labor Review

HOME
Search MLR

ARCHIVES

FOR AUTHORS

ABOUT

SUBSCRIBE

GO

BEYOND BLS

Beyond BLS briefly summarizes articles, reports, working papers, and other
works published outside BLS on broad topics of interest to MLR readers.
A p r il 2023

COVID-19 and employment outcomes of people with disabilities
Summary written by: Yavor Ivanchev
During economic downturns, people with disabilities suffer more severe labor market impacts than their counterparts without disabilities. During the 2007−09
Great Recession, for example, job losses were more prevalent among people with disabilities, and the postrecession employment recovery for that group was
markedly slower. But did the same predicament reemerge during and after the 2020 recession caused by the COVID-19 pandemic?
This is the question addressed in Ari Ne’eman and Nicole Maestas’ recent article titled “How has COVID-19 impacted disability employment?” (National Bureau of
Economic Research, Working Paper 30640, November 2022). The authors note that, in theory, the effects of COVID-19 on the employment experiences of people
with disabilities are not clear cut. On the one hand, fear of viral exposure in face-to-face interactions, coupled with job discrimination, may have put workers in
this group at a disadvantage in the labor market. On the other hand, pandemic-induced increases in the use of telework, together with tight postrecession labor
market conditions, may have given those workers a much-needed employment boost.
Considering these possibilities, the authors use data from the Current Population Survey to examine employment trends for people with and without disabilities
before, during, and after the COVID-19 recession, focusing on the period from the first quarter of 2020 to the second quarter of 2022. The main dependent
variable in the analysis is the employment-to-population ratio (employment rate) for each group, with the numerator of that variable capturing only people who
are both employed and “at work” (rather than individuals who may be employed but furloughed). The authors’ models also include a series of demographic
control variables (age, gender, race, and education) and capture differences across occupational types (essential jobs, teleworkable jobs, etc.).
The empirical results presented in the article differ sharply from those reported for previous recessions, indicating that people with disabilities may have reaped
employment benefits from the transformational effects of the pandemic. Although people with and without disabilities both experienced significant declines in
their employment rates in the first half of 2020, the former’s rate saw a strong recovery in the subsequent 2 years. By the second quarter of 2022, that rate had
exceeded its prepandemic level by nearly 3.6 percentage points, whereas the rate for people without disabilities stood about 0.5 percentage point below the level
recorded before the pandemic. Over the recovery period, people with disabilities also experienced faster employment growth in percent terms and a stronger
rebound in their labor force participation rate.
Another important result reported in the article is that the relatively faster employment recovery for people with disabilities was driven mainly by those employed
in essential, nonfrontline occupations amenable to telework. According to the authors, this finding indicates that this group of workers benefited from the sharp
expansion of telework after the pandemic hit. However, the authors caution that it remains unclear how much of the group’s employment recovery could be
attributed to an increase in the number of people who became disabled because of the adverse health effects of the pandemic.

U.S. BUREAU OF LABOR STATISTICS Division of Information and Marketing Services PSB Suite 2850 2 Massachusetts Avenue NE
Washington, DC 20212-0001
Telephone:1-202-691-5200 Telecommunications Relay Service:7-1-1 www.bls.gov/OPUB Contact Us

U.S. BUREAU OF LABOR STATISTICS
Bureau of Labor Statistics

Publications

Monthly Labor Review

HOME
Search MLR

ARCHIVES

FOR AUTHORS

ABOUT

SUBSCRIBE

GO

ARTICLE

April 2 0 2 3

Federal government wage indexes
For nearly 50 years, the Employment Cost Index (ECI) has been providing the public with estimates of the change in employer labor costs. We explore the practicality of
constructing federal wage indexes, in the spirit of the ECI, using Office of Personnel Management (OPM) salary data. To accomplish this task, we aggregate OPM records
into occupation and industry groups. Although these salary data have a crosswalk for mapping OPM occupation codes into the Standard Occupational Classification system,
no corresponding crosswalk exists for industries. A key hurdle, therefore, involves creating a crosswalk that assigns industry codes to OPM establishments. We create this
crosswalk by developing an algorithm that uses Quarterly Census of Employment and Wages data and machine-learning tools to match agencies with a unique industry. With
this agency-North American Industry Classification System crosswalk, we calculate annual Laspeyres, Paasche, and Fisher wage indexes for several aggregations. The
resulting wage inflation rates are plausible and do not deviate substantially from the corresponding private industry and state and local wage inflation rates.
The Employer Cost Index (ECI) of the National Compensation Survey (NCS) has provided the public with estimates of changes in labor costs since December 1975. At the
ECI launch, only private industry estimates were published; however, in June 1981, ECI expanded to include state and local government workers. The federal government,
despite being the largest U.S. employer with over 3 million employees (see table 1), is presently out of scope for NCS data products. This article explores, as a proof of
concept, the practicality of constructing federal wage indexes using Office of Personnel Management (OPM) salary data. Since this analysis is purely exploratory, we do not
attempt to fully replicate ECI methodology, but instead use it as a guide.

Table 1. Number and percentage of civilian federal workers, by occupation and industry, second quarter of 2020, 2021, and 2022

Category

2020

2021

2022

second quarter

second quarter

second quarter

Number

Percent

Number

Percent

Number

Percent

Occupation

1,588,381

49.6

1,608,050

49.5

1,604,617

50.0

924,123

28.9

949,506

29.2

940,476

29.3

10,908

0.3

10,279

0.3

9,330

0.3

Office and administrative support

301,965

9.4

304,487

9.4

293,667

9.1

Service

251,655

7.9

255,296

7.9

244,730

7.6

Construction, extraction, farming, fishing and forestry

17,075

0.5

16,680

0.5

16,408

0.5

Installation, maintenance and repair

18,298

0.6

18,862

0.6

18,380

0.6

Production

12,549

0.4

12,496

0.4

12,115

0.4

Transportation and material moving

75,143

2.3

74,427

2.3

71,750

2.2

37,199

1.2

36,203

1.1

35,752

1.1

Transportation and warehousing

7,123

0.2

7,117

0.2

7,345

0.2

Elementary and secondary schools

9,012

0.3

9,176

0.3

9,299

0.3

Colleges, universities, and professional schools

3,273

0.1

3,330

0.1

2,774

0.1

39,908

1.2

40,945

1.3

49,224

1.5

875

0.0

878

0.0

852

0.0

5,975

0.2

6,174

0.2

6,126

0.2

80,381

2.5

82,526

2.5

83,502

2.6

3,005,275

93.9

3,052,558

93.9

3,005,394

93.6

11,076

0.3

11,176

0.3

11,205

0.3

Management, business, and financial
Professional and related
Sales and related

Industry
Wholesale and retail trade

Hospitals
Nursing and residential care facilities
Rest of health services
Rest of services
Public administration
Goods producing

Work schedule
Full time

3,097,080

96.8

3,147,790

96.9

3,115,654

97.0

Part time

103,017

3.2

102,293

3.1

95,819

3.0

3,200,097

100.0

3,250,083

100.0

3,211,473

100.0

Total

Source: Authors’ calculations using data from the Office of Personnel Management.

To construct federal wage indexes, we must overcome one major hurdle: records from the OPM data must be categorized into industry (see appendix table A-1) and
occupation groups (see appendix table A-2) that are consistent with NCS aggregations used for the ECI.1 The latter is straightforward because the U.S. Bureau of Labor
Statistics (BLS) uses a crosswalk classification system to map OPM occupations into the Standard Occupational Classification (SOC) system. The former, in contrast, is more
difficult because the OPM data do not contain industry codes. To address this problem, we use the department and agency information in the OPM data and machine-learning
tools to match OPM and Quarterly Census of Employment and Wages (QCEW) establishments.2 An algorithm is developed to select a unique North American Industry
Classification System (NAICS) code for each agency observed in the OPM data. This final mapping yields a desired agency-to-NAICS crosswalk that we use to calculate
Laspeyres, Paasche, and Fisher wage indexes for a variety of aggregations.3

Wage index number formulas
We have many index number formulas to choose from, including the commonly used Laspeyres and Paasche indexes and the less commonly used Dutot or Jevons indexes.4
For exploratory purposes and brevity, we focus on the Laspeyres, Paasche, and Fisher indexes.
Given wages and employment for periods 0 (base period) and 1 (comparison period), the Laspeyres and Paasche wage index number formulas use a fixed “basket” of jobs
(employment) to compute the ratio of total wage costs for period 1 to total wage costs for period 0. The Laspeyres index uses the fixed basket to be period-0 employment,
whereas the Paasche index uses the fixed basket to be period-1 employment. These formulas are given by

and

where

and

are the Laspeyres and Paasche indexes,

is hourly wage,

is the expenditure share, and i is job 1, 2, ..., n. The expenditure share is given by

where

is employment, i and j are jobs

, and t is period 0,1.5 In theory, employers can be expected to substitute away from more expensive workers. Since the

Laspeyres index uses a period-0 fixed-employment basket, the Laspeyres index theoretically overstates wage inflation. Conversely, since the Paasche index uses a period-1
fixed-employment basket, the Paasche index theoretically understates wage inflation.
The Fisher wage index is given by the geometric mean of the Laspeyres and Paasche indexes as

Along with the Törnqvist index, the Fisher index is considered to be “superlative,” with a base and comparison period treated symmetrically to better capture labor
substitution effects.6

Data
BLS has four quarters of OPM data: first quarter of 2019 and second quarter of 2020, 2021, and 2022. For this analysis, we omit the data from the first quarter of 2019 for two
reasons. First, 2019 (first quarter) to 2020 (second quarter) straddled the start of the COVID-19 pandemic, which saw large and uncharacteristic changes in the labor market.
Second, 2019 (first quarter) to 2020 (second quarter) was a five-quarter period that included two federal salary increases. The data cover workers employed at the end of each
quarter. Note that the data are reported to OPM by human resource offices across the federal government and may be subject to some error. If the federal workforce were
incorporated into the ECI, data would need to be collected quarterly from OPM.
OPM data include individual federal employees, annual salary, OPM occupation, full-time or part-time status,7 grade, agency, city, and state. BLS’s OPM data include workers
on military bases (which we exclude) but not postal service employees.8 These data do not include any benefit-cost data (e.g., health insurance, retirement, nonproduction
bonuses). All salaries are given as annual full-time salaries, so hourly wages are computed by dividing salary by 2,087.9 Missing from OPM data are industry data (NAICS
codes), so we use QCEW data and some machine-learning tools to construct an agency-to-NAICS concordance.
Also missing from the OPM data are establishment identifiers. So, we identify them by what we observe: agency, city, and state data, which can be used as imperfect proxies
for an establishment. When an agency has just a single establishment within a city, city and state work as a perfect proxy. But if an agency has multiple establishments within
a city, city and state are imperfect because multiple establishments are identified as a single establishment.
With an agency-to-NAICS crosswalk and a method for identifying establishments, we then map SOC and NAICS codes into occupation and industry groups (sometimes
referred to as pseudo-SOC [PSOC] and pseudo-NAICS [PNAICS]). (See appendix tables A-1 and A-2.) Mean wages and total employment are computed for each basic ECI
cell (a grouping by PSOC, PNAICS, and job) or subcell (a grouping by PSOC, PNAICS, subcell category, and job). Summary statistics, including employment counts and
percentages of total employment from the OPM, are presented in table 1.
Since this analysis is purely exploratory, we do not attempt to reproduce the method for computing the ECI but instead use its basic conceptual framework for computing
wage cost indexes for common index number formula.10 For the ECI, the unit of observation is a quote (such as an establishment, occupation, work status, or grade). These
quotes are aggregated into cells consisting of an ownership sector, industry group (PNAICS), and occupation group (PSOC). Cells can be further divided into subcells that
may include full- or part-time status, region, division, union status, and so forth.

NAICS codes
Missing from the OPM data are NAICS codes. We construct an agency-to-NAICS crosswalk using QCEW-reported NAICS codes for federal government establishments. The
OPM data have standardized, descriptive text for each department and agency. In the QCEW, the department, agency, and NAICS codes are reported individually by each
establishment. These reports are subject to variations in establishment practice and can include spelling errors and varying abbreviations. For these reasons, matching the OPM
establishments with QCEW establishments is not straightforward.
To construct an agency-to-NAICS crosswalk, we begin by aggregating individual employee data in the OPM data to agency by location. We then match each OPM agency and
location with each QCEW establishment by year or quarter, state, and county. For each of these matches, cosine similarities are then calculated for term frequency–inverse
document frequency (or TF–IDF) vectorized department descriptions and agency descriptions. This approach essentially amounts to the construction of a cardinal measure of
similarity between two vectors. A number of options exist for constructing these vectors for a given match’s descriptions. We have explored bag-of-words unigrams (an
unordered list of the individual words from the descriptions) and character n-grams (a contiguous sequence of n characters from a piece of text). We ultimately chose character
n-grams because they account for the issue of spelling errors or variations. A key problem with selecting a vectorization strategy is the lack of an objective standard. That is, in
the absence of an objective standard, any choice between vectorization strategies possesses some level of arbitrariness.
For a given vectorizer, we use the mean of the cosine similarities for department and agency, weighting by QCEW-reported mean employment and upweighting and
downweighting by the relative deviation between employee counts in the OPM establishment-level data and QCEW-reported mean employment. We assume here that larger
establishments are more reliable but may also be “punished” for large differences in the reporting of a variable that should be similar. The QCEW department or agency with
the best weighted cosine similarity is chosen as the match.
Finally, since each department or agency should uniquely match a NAICS code, we compare the weighted cosine similarity among all establishments for a department or
agency and select the NAICS code for the establishment with the best matching weighted cosine similarity. As constructed, the crosswalk is not without flaws, with a mean
agency-size weighted score of 0.76 (standard deviation 0.161) and ranging from nearly the worst (0.002) to the best (1.000). The cumulative distribution of cosine similarity
scores, weighted by agency size (see chart 1), shows that the bulk of matches are fairly reliable (>0.8), with very few that are clearly unreliable (<0.4). Moreover, the federal
government distribution of PNAICS in the OPM dataset roughly matches that for the QCEW data (see chart 2).

Chart 1. Distribution of cosine similarity scores for selected Quarterly
Census of Employment and Wages matches, weighted by agency size
Cumulative distribution
1.0

0.8

0.6

0.4

0.2

0.0
0.0

0.2

0.4

0.6

0.8

1 – cosine similarity
Hover over chart to view data.
Source: Author’s calculations.

View Chart Data

Chart 2. Comparison of the distribution of PNAICS codes with OPM and
QCEW data, second quarter of 2020
OPM

Proportion

QCEW

1.0
0.8
0.6
0.4
0.2
0.0
9200

81R0

4300

420A

6220

62R0

6110

G000

6113

6230

PNAICS
Click legend items to change data display. Hover over chart to view data.
Note: OPM = Office of Personnel Management, PNAICS = pseudo-North American Industry
Classification System, and QCEW = Quarterly Census of Employment and Wages.
Source: Authors’ calculations.

View Chart Data

For computing exploratory wage indexes, this imperfect crosswalk is sufficient. But to publish indexes using OPM data will require dedicated analyst labor to create a more
accurate crosswalk.

Wage index calculations
To compute wage indexes, we first partition the OPM microdata into establishments (department, agency, and city and state) and jobs (occupation, full- or part-time status, and
grade).11 Next, we compute average hourly rates and number of employees for each job within an establishment. The establishment-job data are then matched between the
second quarter of 2020 and the second quarter of 2021 and between the second quarter of 2021 and the second quarter of 2022. The resulting matched data are partitioned by
cell (PNAICS and PSOC) and period. We then calculate weighted average wages and total employment. Finally, we aggregate these data into wage indexes with the use of the
Laspeyres, Paasche, and Fisher formulas. To compute subcell wage indexes, we partition the matched establishment-job data by subcell (PNAICS, PSOC, subcell category).
Then, we calculate weighted average wages and total employment and aggregate them into subcell wage indexes. Note that for the published ECI, the base period is fixed and
all comparisons are relative to the current base quarter (currently the fourth quarter of 2005). In contrast, for each matched pair of OPM datasets (e.g., the first quarter of 2020
to the second quarter of 2021), the base period is the earlier time (e.g., the first quarter of 2020) so that the time series of indexes for each cell and subcell is what is termed
“chained.”
Laspeyres, Paasche, and Fisher wage index calculations are shown in tables 1 through 6 for the basic cell aggregation and for a variety of subcell aggregations. We find that
our computed rates of inflation are reasonable. Note that the calculations of the Laspeyres, Paasche, and Fisher wage indexes are quite close and, in some instances, equal up
to the fourth decimal. This result is similar to other research results.12 This present research also showed that the expected pattern in which the Laspeyres index exceeds the
Paasche index is frequently reversed.13 Finally, a comparison of the federal Laspeyres index with the official ECI is given in table 7. Perhaps unsurprisingly, the exploratory
federal ECI is more closely aligned with the state and local ECI.

Table 2. Wage index calculations of basic cell, 2020 second quarter to 2022 second quarter
Period

Laspeyres

Paasche

Fisher

2020 Q2 to 2021 Q2

1.0131

1.0131

1.0131

2021 Q2 to 2022 Q2

1.0342

1.0341

1.0341

Note: Q2 = second quarter. Wage index data are aggregated into basic cells consisting of ownership sector, industry group, and occupation group.
Source: Authors’ calculations using data from the Office of Personnel Management.

Table 3. Wage index calculations of full-time and part-time work schedules, 2020 second quarter to 2022 second quarter
Work schedule

Period

Full time

Part time

Laspeyres

Paasche

Fisher

2020 Q2 to 2021 Q2

1.0130

1.0129

1.0129

2021 Q2 to 2022 Q2

1.0337

1.0337

1.0337

2020 Q2 to 2021 Q2

1.0366

1.0361

1.0363

2021 Q2 to 2022 Q2

1.0427

1.0425

1.0426

Note: Q2 = second quarter.
Source: Authors’ calculations using data from the Office of Personnel Management.

Table 4. Wage index calculations, by Census divisions, 2020 second quarter to 2022 second quarter
Census division
New England

Middle Atlantic

East South Central

South Atlantic

East North Central

West North Central

West South Central

Mountain

Pacific

Period

Laspeyres

Paasche

Fisher

2020 Q2 to 2021 Q2

1.0123

1.0122

1.0122

2021 Q2 to 2022 Q2

1.0491

1.0490

1.0490

2020 Q2 to 2021 Q2

1.0144

1.0144

1.0144

2021 Q2 to 2022 Q2

1.0355

1.0355

1.0355

2020 Q2 to 2021 Q2

1.0183

1.0184

1.0184

2021 Q2 to 2022 Q2

1.0370

1.0368

1.0369

2020 Q2 to 2021 Q2

1.0097

1.0097

1.0097

2021 Q2 to 2022 Q2

1.0320

1.0320

1.0320

2020 Q2 to 2021 Q2

1.0159

1.0158

1.0159

2021 Q2 to 2022 Q2

1.0341

1.0341

1.0341

2020 Q2 to 2021 Q2

1.0152

1.0152

1.0152

2021 Q2 to 2022 Q2

1.0378

1.0375

1.0377

2020 Q2 to 2021 Q2

1.0145

1.0145

1.0145

2021 Q2 to 2022 Q2

1.0345

1.0346

1.0345

2020 Q2 to 2021 Q2

1.0150

1.0150

1.0150

2021 Q2 to 2022 Q2

1.0398

1.0395

1.0397

2020 Q2 to 2021 Q2

1.0207

1.0206

1.0206

2021 Q2 to 2022 Q2

1.0437

1.0436

1.0437

Note: Q2 = second quarter.
Source: Authors’ calculations using data from the Office of Personnel Management.

Table 5. Wage index calculations, by Census region, 2020 second quarter to 2022 second quarter
Census region
Northeast

South

Midwest

West

Period

Laspeyres

Paasche

Fisher

2020 Q2 to 2021 Q2

1.0139

1.0139

1.0139

2021 Q2 to 2022 Q2

1.0388

1.0388

1.0388

2020 Q2 to 2021 Q2

1.0155

1.0155

1.0155

2021 Q2 to 2022 Q2

1.0361

1.0360

1.0360

2020 Q2 to 2021 Q2

1.0111

1.0111

1.0111

2021 Q2 to 2022 Q2

1.0322

1.0322

1.0322

2020 Q2 to 2021 Q2

1.0180

1.0179

1.0180

2021 Q2 to 2022 Q2

1.0412

1.0411

1.0412

Note: Q2 = second quarter.
Source: Authors’ calculations using data from the Office of Personnel Management.

Table 6. Wage index calculations by size class, 2020 second quarter to 2022 second quarter
Size class by number of employees

Period

1 (<50)

2 (51 to 100)

3 (101 to 500)

4 (>500)

Laspeyres

Paasche

Fisher

2020 Q2 to 2021 Q2

1.0189

1.0189

1.0189

2021 Q2 to 2022 Q2

1.0402

1.0402

1.0402

2020 Q2 to 2021 Q2

1.0117

1.0116

1.0116

2021 Q2 to 2022 Q2

1.0407

1.0408

1.0408

2020 Q2 to 2021 Q2

1.0104

1.0104

1.0104

2021 Q2 to 2022 Q2

1.0349

1.0349

1.0349

2020 Q2 to 2021 Q2

1.0133

1.0132

1.0132

2021 Q2 to 2022 Q2

1.0342

1.0341

1.0342

Note: Q2 = second quarter.
Source: Authors’ calculations using data from the Office of Personnel Management.

Table 7. Comparison of federal Laspeyres index with official Employer Cost Index, 2020 second quarter to 2022 second quarter
Period

Private industry

State and local

Exploratory federal

2020 Q2 to 2021 Q2

1.0315

1.0202

1.0131

2021 Q2 to 2022 Q2

1.0554

1.0341

1.0342

Note: Q2 = second quarter.
Source: Authors’ calculations using data from the Office of Personnel Management.

Conclusion
This analysis demonstrates the practicality of using OPM data to compute a federal government wage component of the ECI. Other elements of the ECI may also be feasible if
benefit-cost and hours data can be acquired. Given the magnitude of the U.S. federal workforce, its inclusion would expand NCS coverage as well as filling a void in
information about federal workers. Although the annually announced federal pay increase provides some information about federal employment cost growth, it is an imprecise
indicator—actual cost growth depends on the flow of employees into and out of federal service and the mix of employee tenures. The calculation of a wage or employment
cost index would provide BLS data users useful measures of the growth of federal employment costs.
Further exploration of OPM data for use with the NCS will be enhanced by access to benefit-cost data. Even though acquiring benefit-cost data might be infeasible, we believe
that the construction of federal wage indexes would prove a valuable addition to the NCS. The addition of the federal workforce to the NCS will require an analyst-validated
NAICS crosswalk, which we view to be an attainable goal considering the findings presented in this article.

Appendix: North American Industry Classification System codes by industry and Standard Occupational Classification codes by occupation
Table A-1. Government industry group definitions, including codes
PNAICS

NAICS

Industry

G000

21, 23, 31 to 33

Goods producing

4400

221

Utilities

420A

42 to 45

Wholesale and retail trade

4300

48, 49

Transportation and warehousing

6110

6111

Elementary and secondary schools

6112

6112

Junior colleges

6113

6113

Colleges, universities, and professional schools

61R0

61, excluding 6111 to 6113

Rest of educational services

6220

622

Hospitals

6230

623

Nursing and residential care facilities

62R0

621, 624

Rest of health services

9200

92, excluding 928

Public administration

81R0

51 to 56, 71 to 81, excluding 814

Rest of services

Note: PNAICS = pseudo-North American Industry Classification System, and NAICS = North American Industry Classification System.
Source: U.S. Bureau of Labor Statistics.

Table A-2. Occupation group definitions, including codes
PSOC

SOC

Occupation

110

11, 13

Management, business, and financial

120

15, 17, 19, 21, 23, 25, 27, 29

Professional and related

210

41

Sales and related

220

43

Office and administrative support

300

31 to 39

Service

405

45, 47

Farm, fishing, forestry, construction, and extraction

430

49

Installation, maintenance, and repair

510

51

Production

520

53

Transportation and material moving

Note: PSOC = pseudo-Standard Occupational Classification, and SOC = Standard Occupational Classification.
Source: U.S. Bureau of Labor Statistics.
SUGGESTED CITATION:

Travis A. Cyronek and Theodore To, "Federal government wage indexes," Monthly Labor Review, U.S. Bureau of Labor Statistics, April 2023, https://doi.org/10.21916/mlr.2023.9

Notes

1 Each basic Employer Cost Index (ECI) “cell” is categorized into industry and occupation groups. ECI cells are further separated into subcategories or “subcells.” These subcategories include
full- or part-time work, Census division or region, establishment size, metropolitan or nonmetropolitan, New York–Chicago–Los Angeles area, union status, and time and incentive status. Our
analysis includes only subcells for full- or part-time work, Census division, region, and establishment size.

2 An establishment is defined as an economic unit that produces goods or services, usually at a single physical location, and that is engaged in one or predominantly one type of economic
activity. For more information, see U.S. Bureau of Labor Statistics glossary https://www.bls.gov/bls/glossary.htm#E.

3 U.S. Census, “General information about price indexes” (U.S. Census Bureau, n.d.), https:/www.census.gov/construction/cpi/pdf/generalinformationaboutpriceindexes.pdf.
4 For a list of index formulas, see Wikipedia: The Free Encyclopedia, “List of price index formulas,” https://en.wikipedia.org/wiki/List_of_price_index_formulas; and U.S. Census,
“General information about price indexes.”

5

Typically, Laspeyres and Paasche index number formulas are expressed as a ratio of total wage costs, given period-0 and period-1 fixed employment baskets,

and

After some manipulation of these formulas, the Laspeyres and Paasche indexes can also be expressed as the function of wage relatives and expenditure shares, as given in the main text.

6 Since the Törnqvist and Fisher indexes are close approximations of one another (formulas produce numbers that are close to one another), we do not use the slightly more complicated
Törnqvist index number formula.

7 OPM defines part-time work as between 16 and 32 hours a week and full-time work as more than 32 hours a week. In addition to full-time and part-time work, a number of other work
schedules include full-time seasonal, part-time seasonal, intermittent, and intermittent seasonal. Our analysis only includes full-time and part-time workers.

8 We excluded military bases because they can have establishments such as schools, hospitals, entertainment venues, and so forth. Although nurses and teachers might be straightforward to
classify into hospitals and schools, occupations such as janitors and secretaries would be challenging. U.S. Postal Service employee data are separately available from OPM and potentially
could be included in the future.

9 “Fact sheet: computing hourly rates of pay using the 2,087-hour divisor” (U.S. Office of Personnel Management, n.d.), https://www.opm.gov/policy-data-oversight/pay-leave/pay-

administration/fact-sheets/computing-hourly-rates-of-pay-using-the-2087-hour-divisor/.
10

Underlying the ECI is the Laspeyres index number formula.

11 Technically, ECI jobs are also differentiated by union status and time or incentive status. Union status is unavailable in our data, and to our knowledge, incentive pay is not widely used in the
federal government.

12 Michael K. Lettau, Mark A. Loewenstein, and Aaron Cushner, “Is the ECI sensitive to the method of aggregation?” Monthly Labor Review, June 1997,

https://www.bls.gov/opub/mlr/1997/06/art1full.pdf; and Michael K. Lettau, Mark A. Loewenstein, and Steve P. Paben, “Is the ECI sensitive to the method of aggregation? an update,”
Monthly Labor Review, December 2002, https://www.bls.gov/opub/mlr/2002/12/art3full.pdf.
13 Ibid. For an explanation of this pattern reversal, see specifically Lettau et al. “Is the ECI sensitive to the method of aggregation?”

ABOUT THE AUTHOR

Travis A. Cyronek
cyronek.travis@bls.gov
Travis A. Cyronek is a research economist in the Office of Compensation and Working Conditions, U.S. Bureau of Labor Statistics.
Theodore To
to.theodore@bls.gov
Theodore To is a research economist in the Office of Compensation and Working Conditions, U.S. Bureau of Labor Statistics.

RELATED CONTENT

Related Articles
The Linked Employment Cost Index: a first look and estimation methodology, Monthly Labor Review, July 2020.
Wage and job-skill distributions in the National Compensation Survey, Monthly Labor Review, February 2017.
Introducing 2012 fixed employment weights for the Employment Cost Index, Monthly Labor Review, August 2016.
Changes in the publication of seasonally adjusted Employment Cost Index series, Monthly Labor Review, March 2013.
Revising the Standard Occupational Classification system for 2010, Monthly Labor Review, August 2010.
Related Subjects
Government
Labor force
Jobs
Employment cost index
industry
Compensation
Compensation costs
Occupations
Industry and Occupational studies

Error processing SSI file
Error processing SSI file