Social Security: Economic Growth and the Funding Shortfall

August 6, 2020 R46479

Social Security: Economic Growth and the
August 6, 2020
Funding Shortfall
Barry F. Huston
The Old-Age, Survivors, and Disability Insurance (OASDI) program, commonly known Analyst in Social Policy
as Social Security, is facing a projected financial shortfal . The program provides

monthly benefits to retired or disabled workers and their eligible family members and to
the eligible family members of deceased insured workers. Approximately 99% of
For a copy of the ful report,
please cal 7-5700 or visit
program revenues are used to pay these monthly benefits, which constitute a substantial
www.crs.gov.
portion of income for a large segment of recipients.
The OASDI program is financed primarily by payroll taxes on covered earnings up to an annual limit, as wel as
federal taxes paid by some beneficiaries on a portion of their OASDI benefits. OASDI program revenues that are
not needed to pay current benefits are invested in government securities held by the Federal Old-Age Survivors
Insurance (OASI) and Federal Disability Insurance (DI) Trust Funds (together, the trust funds), which are
managed by a Board of Trustees (the trustees). The interest earned on these assets provides a third stream of
revenue to the OASDI program.
Under the current law and benefit structure, the program is estimated to be able to pay full scheduled benefits on a
timely basis until 2035; the ability to pay full benefits would require the trust funds to redeem their asset reserves.
When the trust funds’ asset reserves are depleted, the program’s benefit payments wil depend entirely on
continuing tax revenues (i.e., the depletion of asset reserves does not mean Social Security wil be completely
broke and unable to pay benefits). Policy options that would help to avoid such a shortfal general y fal into two
categories: (1) those that increase revenues (e.g., payroll tax increase) and (2) those that decrease benefits (e.g.,
increase in normal retirement age). In addition, some argue that economic growth could prevent the projected
financial shortfal . They reason that increasing economic growth would translate into increases in employment
and real wages, which, through the payroll tax, would beneficially affect the program’s financial status, possibly
avoiding any projected financial shortfal .
General y speaking, increases in employment (i.e., the number of workers subject to the payroll tax) and increases
in real wages (i.e., after-inflation earnings subject to the payroll tax) result in higher OASDI program revenues.
Some economic factors—such as the labor force participation rate, employment, real wage growth, and
productivity—have been shown to significantly impact the financial status of Social Security. In their 2020 annual
report to Congress, the trustees estimated how their intermediate assumptions for economic growth would affect
the program’s projected financial status. The trustees showed that if the underlying factors of economic growth
were to increase, the program’s financial status would improve. However, they also showed that economic growth
alone would not resolve the projected financial shortfal . The 2020 annual report reflects the trustees’
understanding of the Social Security program at the beginning of 2020. Thus, is does not account for any potential
effects of the Coronavirus Disease 2019 (COVID-19).
The trustees demonstrated that the projected funding shortfal is due in part to economic and demographic factors.
For instance, the changing age distribution in the United States indicates that those in covered employment are
supporting an increasing number of people collecting benefits. This demographic aging, resulting in an older age
distribution, is also affecting economic growth through its effects on labor force participation and employment.
Under the trustees’ intermediate assumptions, they estimate that growth in hours worked and productivity are
unlikely to result in an economic growth rate that, alone, would avoid the projected financial shortfal and restore
projected solvency. Thus, although increases in real economic growth—growth in employment, hours worked,
and productivity—would benefit both workers and the program as a whole, these alone are not projected to result
in a permanent solution to the OASDI program’s projected financial shortfal .
Congressional Research Service

Social Security: Economic Growth and the Funding Shortfall

Introduction
Social Security & Coronavirus Disease 2019 (COVID-19)
The Board of Trustees of the Social Security Trust Funds are required by law to submit an annual report to
Congress. Among other things, the report general y covers the projected financial status of the trust funds for a
75-year projection period. The projections require the trustees to make economic, demographic, and program-
specific assumptions about future conditions. The 2020 annual report reflects the trustees’ understanding of the
program at the beginning of 2020. Coronavirus Disease 2019 (COVID-19) wil likely have an effect on the Social
Security program. However, given the then early stage of the pandemic and uncertainty about its potential
magnitude and its likely effects on Social Security at the time of the annual report’s release, possible impacts from
COVID-19 were not incorporated into the 2020 annual report.
The Old-Age, Survivors, and Disability Insurance (OASDI) program, commonly known as Social
Security, provides monthly benefits to retired or disabled workers and their eligible family
members and to the eligible family members of insured deceased workers. These monthly
benefits constitute a substantial portion of income for a large segment of recipients. As of January
2020 there were over 64 mil ion recipients, of which approximately 70.5% are retired workers.1
The OASDI program operates primarily as a pay-as-you-go program in which revenues are paid
out as monthly benefits. The largest source of financing (89.0% of program revenues in 2019) is
the Social Security payroll tax paid, in equal parts by employers and employees, on covered
earnings up to an annual limit.2 The other sources of financing are income taxes (3.4% of
program revenues in 2019) paid on a portion of Social Security benefits by some beneficiaries
with income above a certain threshold and interest earned on reserves held in the Social Security
Trust Funds, or the “trust funds” (7.6% of the program’s income in 2019).3 In 2019, the OASDI
program’s largest cost was monthly benefits, which accounted for 98.9% of program costs.4
The OASDI program’s ability to pay full scheduled benefits is dependent on a favorable
relationship between revenues and cost, commonly known as a positive balance.5 Depending on

1 Social Security Administration (SSA), Office of the Chief Actuary, “ Fact Sheet on the Old-Age, Survivors, and
Disability Insurance Program,” January 2020, at https://www.ssa.gov/oact/FACT S/index.html.
2 SSA, Office of the Chief Actuary, “T he 2020 OASDI T rustees Report,” The 2020 Annual Report of the Board of
Trustees of the Federal Old-Age and Survivors Insurance and Federal Disability Insurance Trust Funds, April 22,
2020, p. 7, at https://www.ssa.gov/OACT /T R/2020/tr2020.pdf (hereinafter cited as Board of T rustees, 2020 Annual
Report). Under current law, the Social Security payroll tax is applied to covered earnings up to an annual limit, or
taxable maximum. T he taxable maximum is indexed to national average wage growth for years in which a cost -of-
living adjustment is payable; the taxable maximum for 2020 is $137,700. T he payroll tax rate for employees in covered
employment is 6.20% and matched by employers, resulting in a combined payroll tax rate of 12.4% for covered
earnings up to the taxable maximum.
3 Under current law, there are two separate trust funds for the Social Security program: (1) the Old-Age and Survivors
Insurance (OASI) T rust Fund and (2) the Disability Insurance (DI) T rust Fund. T hey are discussed here on a combined
basis as the Social Security T rust Funds, or the trust fun ds. Monies credited to each trust fund cannot be lent or
transferred to the other trust fund without authorization from lawmakers. OASI benefits can be paid only from the
OASI T rust Fund, and DI benefits can be paid only from the DI T rust Fund. For more in formation on the taxation of
Social Security benefits, see CRS Report RL32552, Social Security: Taxation of Benefits.
4 Board of T rustees, 2020 Annual Report, p. 7. T he remaining 1.1% of program costs are accounted for by
administrative costs and the Railroad Retirement Board (RRB) financial interchange. For more information on the
RRB, see CRS Report RS22350, Railroad Retirem ent Board: Retirem ent, Survivor, Disability, Unem ploym ent, and
Sickness Benefits.
5 In 2019, the Social Security program’s total revenues were larger than the program’s total cost. However, the
program’s total cost exceeded its noninterest revenues (i.e., income from the payroll tax and taxation of benefits). T his
is commonly referred to as a cash flow deficit. Periods of cash flow deficit can be offset by interest income or the
Congressional Research Service
1

Social Security: Economic Growth and the Funding Shortfall

trust fund asset reserves, however, the program may continue to pay full scheduled benefits
during times when costs exceed revenue, or periods of cash flow deficit.6 Given the OASDI
program’s dependence on a favorable revenue and cost relationship, the program’s ability to meet
its scheduled payments is necessarily connected to economic factors in the U.S. economy, such as
growth in employment, average hours worked, and productivity.7
The Social Security program is facing a projected financial shortfal . That is, the program is
projected to have sufficient funds needed to pay scheduled benefits until 2035, after which the
level of benefits payable is dependent on continuing tax revenues. Under current law, the OASDI
Trust Funds’ Board of Trustees, or the trustees, manage the trust funds and submit annual reports
to Congress regarding the program’s expected status.8 In recognition of the projected financial
shortfal , the trustees estimate that a hypothetical payroll tax increase of 3.14 percentage points, a
hypothetical benefit reduction of 19%, or some combination of both would need to immediately
be enacted to avoid such a shortfal .9 Some lawmakers argue that the best method to avoid the
projected shortfal is through maintaining economic growth in the United States. That is, if real
economic growth were to be maintained or increased, then the associated increases in
employment and wages could reduce a funding shortfal .10
This report presents the current status of the trust funds, projections for the factors in real
economic growth, and how projections for these growth factors are estimated to affect the
program’s financial status.
Current Status of the Trust Funds
The long-range financial status of the OASDI program is measured by the actuarial balance,
defined as the difference between the summarized cost rate and the summarized income rate over
a 75-year projection horizon.11 The summarized cost rate and the summarized income rate are
expressed as a percentage of taxable payroll.12 At the end of 2019, the Board of Trustees

redemption of trust fund assets. For more details, see CRS Report RL33028, Social Security: The Trust Funds.
6 For more information on the ability of the program to pay benefits during periods of cash flow deficit, see CRS
Report RL33514, Social Security: What Would Happen If the Trust Funds Ran Out? .
7 Stephen C. Goss, “T he Future Financial Status of the Social Security Program,” Social Security Bulletin, vol. 70, no. 3
(August 2010), p. 124.
8 42 U.S.C. §4001(c).
9 Board of T rustees, 2020 Annual Report, p. 5.
10 For instance, in a 2018 statement, Secretary of the T reasury Steven Mnuchin stated that “The Administration’s
economic agenda—tax cuts, regulatory reform, and improved trade agreements—will generate the long-term growth
needed to help secure these [Social Security and Medicare] and lead them to a more stable path.” Department of the
T reasury, “Statement by Secretary Mnuchin on the Spring 2018 Meeting of the Social Security and Medicare Board of
T rustees,” press release, June 5, 2018, at https://home.treasury.gov/news/press-releases/sm0404. Similarly, in a 2018
Hearing of the Subcommittee on Social Security, several lawmakers expressed views that pro -growth policies such as
the T ax Cuts and Jobs Act (P.L. 115-97) would result in higher economic growth and improve the Social Security
program’s financial status. See U.S. Congress, House Committee on W ays and Means, Subcommittee on Social
Security, Hearing on Exam ining Social Security’s Solvency Challenge: The Status of Social Security’s Trust Funds ,
115th Cong., 2nd sess., June 7, 2018 (Washington, DC: GPO, 2018).
11 T he summarized cost rate is the ratio of the present value of cost to the present value of the taxable payroll for the
projection period. T he sum marized incom e rate is the ratio of the present value of scheduled noninterest income to the
present value of taxable payroll for the projection period. See Board of T rustees, 2020 Annual Report, p. 248.
12 Taxable payroll is the weighted sum of taxable wages and taxable self-employment income. When this sum is
multiplied by the Old-Age, Survivors, and Disability Insurance (OASDI) program payroll tax rate, it results in the total
amount of payroll taxes. See Board of T rustees, 2020 Annual Report, p. 249.
Congressional Research Service
2

link to page 8 link to page 8 link to page 8 link to page 8 Social Security: Economic Growth and the Funding Shortfall

estimated the long-range actuarial balance for the OASDI program to be -3.21% of taxable
payroll.13 This negative balance indicates an actuarial deficit (i.e., a funding shortfal ). Over the
next 75 years, the Board of Trustees projects OASDI program costs to exceed program revenues
by 3.21% of taxable payroll. To maintain program solvency throughout this period, available
policy options would be to reduce costs, increase revenues, or some combination thereof to the
magnitude of 3.21% of taxable payroll.14 (This report focuses on the trustees’ intermediate set of
assumptions, which reflect their best estimates of future demographic and economic trends.)
Another measure of the OASDI program’s long-range financial status projects how long the
program could continue to pay full scheduled benefits, or the projected date for trust fund reserve
depletion. Under their 2020 intermediate assumptions, the trustees projected that the trust fund
reserves wil be depleted in 2035.15 This indicates that current and future beneficiaries would
receive the full amount of their scheduled benefits until sometime in 2035. After that point,
continuing tax revenues would determine benefit amounts.16 Continuing tax revenues are
projected to cover about 79% of scheduled benefits.17
Components of Real Economic Growth
Real economic output, defined as the amount of goods and services produced in the economy, is a
function of economic factors. Specifical y, real economic output (i.e., real gross domestic product
[GDP]) is a product of employment, hours worked, and productivity. It follows that the growth in
real economic output is approximately composed of the growth in its factors—employment, hours
worked, and productivity. 18
Total Employment
The growth in total employment is a major factor in the U.S. economy’s real economic growth.
Historical and projected total employment and growth in total employment (i.e., annual
percentage change) are shown in Figure 1 and Figure 2, respectively.
Figure 1 shows historical total employment in the United States, reflecting the number of
individuals in the U.S. economy who are in paid employment. Figure 2 shows the annual
percentage change—or growth—in total employment. Both figures show that total employment in

13 T he actuarial deficit of 3.21% of taxable payroll differs from the required 3.14 percentage point increase in the
payroll tax increase, as it reflects behavioral response changes to payroll tax rate changes. See Board of T rustees, 2020
Annual Report, p. 5.
14 In June 2019, the Congressional Budget Office (CBO) projected the actuarial balance to be -4.6% of taxable payroll
(see CBO, The 2019 Long-Term Budget Outlook, at https://www.cbo.gov/system/files/2019-06/55331-LTBO-2.pdf).
CBO estimates that its assumptions for population (longer life expectancies and higher disability rates), earnings
inequality (lower amount of aggregate covered earnings subject to the payroll tax), lower long-term real interest rates,
slower growth in nominal economic output, and differences in analytical approaches account for the differen ce in the
projected actuarial balance as compared with the trustees. For more information, see CBO, “Social Security,” CBO’s
Long-Term Social Security Projections: Changes Since 2018 and Com parisons with the Social Security Trustees’
Projections, December 19, 2019, at https://www.cbo.gov/system/files/2019-12/55914-CBO-Social-Security-
Comparison.pdf.
15 Board of T rustees, 2020 Annual Report, p. 3. CBO projects a combined OASDI trust fund depletion date of 2032.
See CBO, “Budget,” The 2019 Long-Term Budget Outlook, June 25, 2019, at https://www.cbo.gov/system/files/2019-
06/55331-LTBO-2.pdf.
16 For more information on trust fund reserve depletion and partial benefits, see CRS Report RL33514, Social Security:
What Would Happen If the Trust Funds Ran Out? .
17 Board of T rustees, 2020 Annual Report, p. 5.
18 Board of T rustees, 2020 Annual Report, p. 111.
Congressional Research Service
3

link to page 8 link to page 8

Social Security: Economic Growth and the Funding Shortfall

the United States general y increased throughout the historical 1960-2019 period. However, as
marked by the shaded bars, total employment and employment growth general y decreased during
years in which the economy experienced a recession.
As seen in Figure 1, total employment in the United States is expected to continue to increase
throughout the 75-year projection period at a slower pace than it has historical y. This is reflected
in Figure 2 by a projected growth rate in total employment of less than 1%. That is, the growth
rate in total employment is expected to be relatively stable, growing at less than 1% per year.
Assumptions for the projection period reflect the trustees’ understanding of the OASDI program
at the start of 2020; thus, it does not include potential effects of Coronavirus Disease 2019
(COVID-19).
Figure 1. Historical and Projected U.S.
Figure 2. Historical and Projected
Total Employment
Growth in U.S. Total Employment
(in thousands)
(annual percentage change)

Source: Congressional Research Service (CRS), from data provided by the Bureau of Labor Statistics (BLS),
“Current Employment Statistics – CES (National),” at https://www.bls.gov/ces/; Social Security Administration, The
2020 Annual Report Of The Board Of Trustees Of The Federal Old-Age And Survivors Insurance And Federal Disability
Insurance Trust Funds, April 22, 2020, at https://www.ssa.gov/OACT/TR/2020/tr2020.pdf (2020 Annual Report
[AR]).
Notes: Projections are based on the Board of Trustees’ intermediate assumptions in the 2020 AR. Assumptions
for the projection period reflect the trustees’ understanding of the OASDI program at the start of 2020; thus, it
does not include potential effects of Coronavirus Disease 2019 (COVID-19). Shaded areas represent a year in
which a recession occurred, as designated by the National Bureau of Economic Research.
The size of the population and the labor force participation rate (LFPR), or the share of the
population participating in the workforce, determine total employment. Thus, two important
factors that contribute to changes in employment are fluctuations in (1) the size of the total
population, or cohort, of persons of working age and (2) that population’s participation in the
labor market, as measured by the LFPR.
Population Change
Total employment is closely linked to a country’s population in the long run. As the population of
the United States has grown over time, so has its total employment. Total employment in the U.S.
labor force measures the subset of the population that is working or available to work.
General y, during periods of high employment, a larger fraction of a population is comprised of
persons in their prime working ages. In the U.S. context, “baby boomers” (general y defined as
those born between 1946 and 1964) grew into a large pool of working-age adults. As time
elapsed, the large cohort continued to age, moving through their employment years and into their
retirement ages. They were replaced by smal er cohorts of working-aged adults due to
Congressional Research Service
4

link to page 26 link to page 26 link to page 26 link to page 26 link to page 26 link to page 26 Social Security: Economic Growth and the Funding Shortfall

demographic factors. Thus, as the baby boomers make up a large portion of the population, their
aging and exit from the paid labor force wil negatively impact the LFPR.19 Projections from the
Bureau of Labor Statistics suggest that the decreasing trend in the LFPR wil continue. The LFPR
is typical y highest for the 24 to 54 age group (commonly known as the prime working age), and
is general y over 80%.20 This cohort was dominated by baby boomers for many years. However,
the last baby boomers—those born in 1964—exited the prime working age group in 2018. The
size of the cohort that moved into the 24 to 54 age group after the baby boomers was smal er, and
the number of workers who newly participated in the labor force were not able to compensate for
the number of baby boomers exiting the prime working age group. Although the overal
population continues to grow, meaning some post-baby boom cohorts may be larger in number,
the eventual exit of the baby-boom generation from the workforce entirely wil continue to put
downward pressure on the LFPR.21
Over the last several decades, the total U.S. population has been growing slowly at a decreasing
rate. The Appendix shows this trend and two factors that are largely attributed with slowing
population growth: (1) decreasing fertility and (2) decreasing net immigration. A population
expands with increases in fertility (i.e., new births) and net immigration (i.e., new residents).22 As
the fertility rate (Figure A-2) is projected to remain low and net immigration (Figure A-3) is also
projected to remain low, it is likely that the total population (Figure A-1) wil demonstrate slow
growth as wel . The trustees’ projection of a decreasing growth rate in total population is mirrored
in U.S. Census Bureau projections, which show a similar slowing trend projected over the next
several decades.23
Changes in Labor Force Participation Rates
The percentage of the civilian noninstitutional population 16 years and older that is working or
actively looking for work is commonly measured using the LFPR: the percentage of the civilian
noninstitutional population 16 years and older that is working or actively looking for work.24
Figure A-4 shows the LFPR increased from 1960 to 2000, indicating that over this period, the
share of the total population in the labor force was increasing. Figure A-4 also shows that these
increases in the LFPR were, among other things, the result of the increasing share of women
joining the labor force, which more than offset the fal ing share of men participating in the labor
force.

19 Mitra T oossi, BLS, “Labor force projections to 2024: the labor force is growing, but slowly,” Monthly Labor Review,
December 2015 (hereinafter cited as T oossi, “Labor Force Projections to 2024”).
20 T oossi, “Labor Force Projections to 2024.”
21 T oossi, “Labor Force Projections to 2024.” In 2018, approximately 10,200 baby boomers attained age 65 per day, an
age commonly associated with retirement. See U.S. Government Accountability Office, The Nation’s Retirem ent
System : A Com prehensive Re-Evaluation Needed to Better Prom ote Future Retirem ent Security, GAO-19-342T ,
February 6, 2019, p. 6, at https://www.gao.gov/assets/700/696766.pdf.
22 Increased life expectancy also plays a role in the age distribution. See CRS Report R45990, Social Security:
Dem ographic Trends and the Funding Shortfall.
23 Sandra L. Colby and Jennifer M. Ortman, Projections of the Size and Composition of the U.S. Population: 2014 to
2060, U.S. Census Bureau, P25-1143, March 2015, p. 1, at https://www.census.gov/content/dam/Census/library/
publications/2015/demo/p25-1143.pdf. Further annual data for Census Bureau projections can be found at
https://www.census.gov/data/tables/2017/demo/popproj/2017-summary-tables.html.
24 Steven F. Hipple, U.S. Bureau of Labor Statistics (BLS), “Labor force participation: what has happened since the
peak?,” Monthly Labor Review, September 2016, at https://www.bls.gov/opub/mlr/2016/article/labor-force-
participation-what-has-happened-since-the-peak.htm (hereinafter cited as Hipple, “ Labor Force Participation”).
Congressional Research Service
5

link to page 26 link to page 12 link to page 12 link to page 12 link to page 12 link to page 12 Social Security: Economic Growth and the Funding Shortfall

A wide spectrum of reasons may explain why a sizeable portion of the population cannot or
chooses not to seek work. Reasons range from individual-level characteristics, such as age,
disability, or personal preference, to structural mismatches in labor demand, skil sets, and
geography. As seen in Figure A-4, the LFPR peaked in 2000 and has general y declined since. In
the years since the peak, the LFPR has decreased for most demographic groups commonly
associated with working ages. From 2000 to 2015, the LFPR declined for teenagers, young adults
aged 20-24, and women and men aged 25 to 54; the LFPR rose for men and women ages 55 and
older.25 Although this decrease is partly attributable to cyclical effects (e.g., economic recessions),
research suggests that demographic changes (e.g., aging of the population) is also a major cause
for the decrease in the LFPR.26
Average Hours Worked
The growth in average hours worked is another major contributor in the U.S. economy’s real
economic growth. Historical and projected average weekly hours worked and growth in average
weekly hours worked are shown in Figure 3 and Figure 4, respectively.
Figure 3 shows the historical average weekly hours worked per week in the United States, which
reflects the amount of time for which participants in the paid labor force are compensated. Figure
4 shows the annual percentage change in the average weekly hours worked per week. As seen in
the figures, the average number of weekly hours worked per U.S. labor participant has general y
decreased throughout the historical 1960-2017 period. For instance, the average U.S. worker was
paid for about 34 hours per week in 2017, the most recent year for which data are available,
whereas the average U.S. worker was paid for about 37 hours per week in 1960.
Changes in Average Weekly Hours
The decline in the average weekly hours worked is not a new or recent phenomenon. Early
research pointed to changes in paid time off as the main cause for the reduction in average weekly
hours worked. Specifical y, researchers found that from 1947 to 1978 the amount of time
al ocated to workers for paid leave (e.g., vacation or sick leave) and paid on-premise time (e.g.,
breaks or travel time) increased. The amount of paid time off during this period almost doubled,
from 5.4% of paid time hours in 1947 to 10.6% of paid time off hours in 1978.27 In the
succeeding years, Congress approved measures that could increase the amount of paid time off
for workers. As examples, Congress enacted legislation in 1983 commemorating Martin Luther

25 Hipple, “Labor Force Participation.”
26 Estimates suggest that somewhere between one-half to two-thirds of the decrease in the labor force participation rate
(LFPR) since 2007 can be attributed to an aging population. A paper by the Council of Economic Advisers found that
approximately one-half of the decrease in the LFPR was due to an aging population (see Council of Economic
Advisers, The Labor Force Participation Rate Since 2007: Causes and Policy Im plications, July 2014, p. 3, at
https://obamawhitehouse.archives.gov/sites/default/files/docs/labor_force_participation_report.pdf), whereas other
research puts the estimate as high as two-thirds (see William R. Cline and Jared Nolan, Dem ographic versus Cyclical
Influences on US Labor Force Participation , Peterson Institute for International Economics, Working Paper no. 14 -4,
July 2014, p. 5, at https://www.piie.com/publications/wp/wp14-4.pdf). Research published by the Federal Reserve
Bank of Chicago also found that about half of the decline in LFPR could be explained by demographic factors.
Additionally, this research showed that the effects on LFPR differ when analyzed by demographic group, such as
gender and education (see Daniel Aaronson and Jonathan Davis, “Explaining the decline in U.S. labor force
participation rate,” March 2012, at https://www.chicagofed.org/publications/chicago-fed-letter/2012/march-296).
27 Herbert R. Northrup and T heresa Diss Greis, “T he Decline in Average Annual Hours Worked in the United States,
1947-1979,” Journal of Labor Research, vol. IV, no. 2 (Spring 1983), p. 103.
Congressional Research Service
6

Social Security: Economic Growth and the Funding Shortfall

King Jr.’s birthday to create a federal holiday,28 and the Family and Medical Leave Act of 1993
(as amended, P.L. 103-3) guaranteed unpaid leave for eligible employees for certain family and
medical needs.29 Additional y, several states also have enacted paid family leave plans. Though
the plans are relatively new, the general result is the option for eligible employees to access
partial y or fully compensated leave for caregiving needs such as childcare.30
Changes in the age composition of the workforce may also explain a portion of the decrease in
U.S. laborers’ average weekly hours worked. A study looking at U.S. Census data from 1950
through 2000 showed that average weekly hours worked per person was highest for those aged 25
to 54.31 Therefore, an increase in the number of workers at the prime working age relative to other
ages is likely to increase the average weekly hours worked by the overal workforce. This may
help to explain why average weekly hours worked increased in the 1990s, when an influx of baby
boomers fully entered the workforce, but declined 5% from 1950 to 2000. The large number of
baby boomers would have heavily influenced average hours, which is essential y an average of
hours worked weighted by the different age groups of the paid labor force. From 1900 to 2005 the
average hours worked by those in prime working ages (25 to 54) remained relatively stable but
declined for other age groups.32 Time spent not working was replaced with an increase in school
hours for young workers (aged 14 to 24) and an increase in leisure hours for older workers (aged
55 and older).33
A change in the composition of labor demand may also explain the decreasing trend in average
hours worked per week. Since 1950, the growth in service-providing jobs has outpaced the
growth in goods-producing (i.e., manufacturing) jobs. In 1950, for example, service-producing
jobs accounted for 63% of total nonfarm employment, whereas service-producing jobs accounted
for 86% of total nonfarm employment in 2019. 34 Research showed that from 1972 to 1999, the
service-providing sector divisions that have experienced the largest increase in employment also
showed the lowest average in weekly hours worked.35 Employment figures from 2008 through
2019 confirmed that employment growth in the service-providing sector—the largest sector in
U.S. employment—was stil outpacing that of the goods-producing sector.36 Over this period, the

28 P.L. 98-144.
29 For more information on the Family and Medical Leave Act, see CRS In Focus IF10329, The Family and Medical
Leave Act (FMLA).
30 Paid family leave plans vary by state; for more information on paid family leave plans, see CRS Report R44835,
Paid Fam ily and Medical Leave in the United States.
31 Ellen R. McGrattan and Richard Rogerson, “Changes in Hours Worked, 1950 -2000,” Federal Reserve Bank of
Minneapolis Quarterly Review, vol. 28, no. 1 (July 2004), pp. 23 -27. This result corresponds to what many refer to as
the prim e working age, ages 25-54. Individuals of this age range are likely to have finished formal education and be not
yet approaching retirement. See Rob Valletta and Nathaniel Barlow, The Prim e-Age Workforce and Labor Market
Polarization, Federal Reserve Bank of San Francisco, September 10, 2018, at https://www.frbsf.org/economic-
research/files/el2018-21.pdf.
32 Valerie A. Ramey and Neville Francis, “A Century of Work and Leisure,” American Economic Journal:
Macroeconom ics, vol. 1, no. 2 (July 2009), pp. 190-191 (hereinafter cited as Ramey and Francis, “ Century of Work and
Leisure”).
33 Ramey and Francis, “Century of Work and Leisure.”
34 BLS, “Current Employment Statistics – CES (National),” at https://www.bls.gov/ces/ (hereinafter cited as BLS,
“CES National”).
35 Katie Kirkland, BLS, “On the decline in average weekly hours worked,” Monthly Labor Review, July 2000, p. 27.
36 BLS, “Employment by major industry sector,” at https://www.bls.gov/emp/tables/employment-by-major-industry-
sector.htm.
Congressional Research Service
7

link to page 12 link to page 12 link to page 12 link to page 8 link to page 12

Social Security: Economic Growth and the Funding Shortfall

average hours worked per worker in the service-providing sector remained unchanged at an
average 33.3 hours per week.37
Figure 3 and Figure 4 show how the Board of Trustees assumes average weekly hours worked
wil change over their 75-year projection period. As evidenced by most observations appearing
below the 0.0% line in Figure 4, the trustees expect average weekly hours worked to continue to
decrease throughout the projection period, albeit at a relatively slow rate.
Figure 3. Historical and Projected
Figure 4. Historical and Projected
Average Weekly Hours
Growth in Average Weekly Hours
(in hours)
(annual percentage change)

Sources: CRS, from data provided in the 2020 Annual Report (AR), Table V.B1, at https://www.ssa.gov/OACT/
TR/2020/tr2020.pdf; Robert C. Feenstra, Robert Inklaar, and Marcel P. Timmer, “The Next Generation of the
Penn World Table,” American Economic Review, vol. 105, no. 10 (October 2015), at https://www.aeaweb.org/
articles?id=10.1257/aer.20130954.
Notes: The Penn World Table (PWT) is used for historical data, and projections are based on the Board of
Trustees’ intermediate assumptions in the 2020 AR. Assumptions for the projection period reflect the trustees’
understanding of the OASDI program at the start of 2020; thus, it does not include potential effects of
Coronavirus Disease 2019 (COVID-19). Shaded areas represent a year in which a recession occurred, as
designated by the National Bureau of Economic Research.
Aggregate Hours Worked
The total employment in the U.S. economy represents the amount of labor employed in the
production of goods and services. The average hours worked per week in the U.S. economy
represents the effective usage rate for the labor. When the total employment is multiplied by the
average weekly hours worked, the product is the aggregate hours worked in the entire economy
per week. As shown in Figure 2 and Figure 4, the annual percentage change, or growth rate, for
total employment and average weekly hours worked is projected to be relatively stable
throughout the 75-year projection period. Consequently, the number of aggregate weekly hours
worked in the economy is projected to stable as wel .
Projections for the future of the labor force necessarily incorporate projec tions for the role of
technological advancement and automation. There is considerable debate as to the effects of
automation on the supply or demand for paid labor. On one hand, some argue that automation wil
eventual y displace labor. On the other hand, some argue that automation displaces some labor
while it creates the need for labor elsewhere, or reinstates paid labor. That is, automation serves
as both a substitute for labor and a complement to labor. The former results in a decrease in the

37 BLS, “CES National.”
Congressional Research Service
8

link to page 14 link to page 14 Social Security: Economic Growth and the Funding Shortfall

demand for labor, whereas the latter can result in an increase in the demand for labor.38 This
possible increase helps to explain why past advancements in automation (e.g., the industrial
revolution) did not result in sustained unemployment.39
General y, past advances in automation have helped to improve the efficiency of labor, not
replace labor.40 Advances in automation technologies, however, are specifical y designed to
replace labor, not to augment it (e.g., multipurpose machines such as industrial robots that are
automatical y controlled and reprogrammable).41 Whereas past improvements in automation
served to complement human labor and increase productivity, al owing the combination of
humans and automation to produce more per hour of work, new automation (e.g., industrial
robots) serves as a substitute for human labor, al owing for equal or greater productivity with a
decrease in human labor. One study has shown that the introduction of industrial robots into the
U.S. labor market has led to a reduction in the overal employment-to-population ratio and a
reduction in overal wages.42
Workers in goods-producing sectors experience comparatively larger average drops in aggregate
weekly hours worked during recessions than those in the service-providing sector, but the decline
in aggregate weekly hours worked for service-providing workers has sharpened over the last
several recessions. For instance, the average weekly hours in the service-providing sector
decreased by 1.2% in the 2007-2009 recession versus a decrease of 0.3% in the 1990-1991
recession.43 Although more research is needed to determine if this is a trend, future economic
downturns could nevertheless exacerbate this decline considering the service-providing sector has
seen the majority of U.S. job growth.44
Productivity
The growth in total factor productivity is another major determinant in the U.S. economy’s real
economic growth. Total factor productivity, or multifactor productivity, is the efficiency at which
measured inputs are utilized in producing output of goods and services, measured as output per
unit of combined input.45 The historical and projected total factor productivity index and its
growth rate can be seen in Figure 5 and Figure 6, respectively.

38 David H. Autor, “Why Are T here Still So Many Jobs? T he History and Future of Workplace Automation,” Journal
of Econom ic Perspectives, vol. 29, no. 3 (Summer 2015), p. 5.
39 Ben Vermeulen, Jan Kesselhut, and Andreas Pyka et al., “T he Impact of Automation on Employment: Just the Usual
Structural Change?,” Sustainability, vol. 10, no. 1661 (May 2018), p. 2.
40 Daron Acemoglu and Pascual Restrepo, The Wrong Kind of AI? Artificial Intelligence and the Future of Labor
Dem and, National Bureau of Economic Research (NBER), Working Paper no. 25682, March 2019, p. 3, at
https://www.nber.org/papers/w25682.pdf.
41 Daron Acemoglu and Pascual Restrepo, Robots and Jobs: Evidence From US Labor Markets, NBER, Working Paper
no. 23285, March 2017, p. 2, at https://www.nber.org/papers/w23285.pdf (hereinafter cited as Acemoglu and Restrepo,
Robots and Jobs).
42 Acemoglu and Restrepo, Robots and Jobs, pp. 4-5. T he employment-to-population ratio, an alternative measure of
labor market strength, is the ratio of total civilian employment to the civilian noninstitutional population. For more
information on this ratio, see CRS Report R44055, An Overview of the Em ploym ent-Population Ratio.
43 Steven Kroll, “T he decline in work hours during the 2007 -2009 recession,” Monthly Labor Review, April 2011, pp.
56-57.
44 For more information on the growth of the service-providing sector, see CRS Report R45330, Labor Market Patterns
Since 2007.
45 Combined inputs, such as labor and capital, are used directly to produce output. BLS, Office of Productivity &
T echnology, “ Productivity Glossary,” at https://www.bls.gov/mfp/optglossary.htm.
Congressional Research Service
9

link to page 14 link to page 14 link to page 14

Social Security: Economic Growth and the Funding Shortfall

Productivity is general y cyclical, exhibiting higher growth during periods of increasing economic
output (e.g., expansions) and exhibiting negative growth during periods of decreasing economic
output (e.g., recessions).46 On average, productivity grew in the 1960 to 2019 period, averaging
an annual growth of 1.76%. From 1960 through 1976, the growth rate in total factor productivity
averaged 2.44% annual y. Military-related research and technological advancements during
World War II are largely cited as causes for the long-term average growth rate during this
period.47 Productivity demonstrated a marked slowdown from 1977 through 1995 (average annual
growth of 1.28%). Productivity then increased for the 1996 through 2005 period (average annual
growth of 2.50%), attributable to many high-tech innovations from the 1970s and 1980s (e.g.
fiber optics, personal computers, internet).48
As Figure 6 shows, total factor productivity growth has averaged 1.06% since 2005. This below-
trend productivity growth, or missing productivity, is commonly referred to as the productivity
puzzle. Figure 5 and Figure 6 also show projections for future productivity growth. These
projections assume that growth in the long-run total factor productivity wil average 1.63%,
which is closer to the historical long-run average of 1.76% and above the post-2005 average of
1.06%.
Figure 5. Historical and Projected Total
Figure 6. Historical and Projected
Factor Productivity Index
Growth in Total Factor Productivity
(index, 2012=100)
(annual percentage change)

Sources: CRS, from data provided by Federal Reserve Bank (FRB) of St. Louis at https://fred.stlouisfed.org/; 2020
Annual Report (AR), Table V.B1, at https://www.ssa.gov/OACT/TR/2020/tr2020.pdf.
Notes: Projections based on the Board of Trustees’ intermediate assumptions in the 2020 AR. Assumptions for
the projection period reflect the trustees’ understanding of the OASDI program at the start of 2020; thus, it does
not include potential effects of Coronavirus Disease 2019 (COVID-19). Shaded areas represent a year in which a
recession occurred, as designated by the National Bureau of Economic Research.
The Productivity Puzzle
Research has suggested many reasons for why productivity growth slows from its long-run
average, including arguments that productivity slowdowns

46 Michael Chernousov, Susan E. Fleck, and John Glaser, “Productivity trends in business cycles: a visual essay,”
Monthly Labor Review, June 2009, p. 50.
47 Roger W. Ferguson Jr. and William L. Wascher, “Distinguished Lecture on Economics in Government: Lessons
from Past Productivity Booms,” The Journal of Economic Perspectives, vol. 18, no. 2 (Spring 2004), p. 9 (hereinafter
cited as Ferguson Jr. and Wascher, “ Distinguished Lecture on Economics in Government”).
48 Ferguson Jr. and Wascher, “Distinguished Lecture on Economics in Government,” p. 11.
Congressional Research Service
10

Social Security: Economic Growth and the Funding Shortfall

 result from financial imbalances (e.g., overvalued equity prices and resulting
corrections) and overinvestment into a specific sector (e.g., telecommunications);
 are associated with exogenous disruptions (e.g., oil shocks, banking crises);49 or
 are a natural result of productivity growth level ing off after periods of high rates
of innovation and technological change.50
This last argument contends that some innovations are more important for economic growth than
others. For instance, innovations in electricity, the internal combustion engine, and transportation
were widely applicable across industries and resulted in large productivity gains throughout the
1970s. These innovations were general purpose, whereas many innovations after the 1970s were
narrower (e.g., information technology) in their application and resulted in smal er productivity
gains.51
Productivity growth remains below its long-run average even with the adoption of numerous
sophisticated technologies.52 Decreases in new general purpose technologies lead many
economists to conclude that productivity growth wil be slower than its long-run average even
though levels of technology remain high.53
Labor Productivity and Real
Figure 7. Labor Productivity and Real Earnings
Earnings
Productivity is also measured by labor
productivity—the ratio of total
economic output, adjusted for price
changes, to the total number of hours
worked.54 Increases in labor
productivity provide employers the
means to raise wages. That is,
productivity and wages are closely
linked.

Since the 1970s, the trends in labor
Source: CRS, from data provided by the Bureau of Labor
productivity and real (after-inflation)
Statistics.
Notes: Shaded areas represent a year in which a recession
occurred, as designated by the National Bureau of Economic
Research.

49 Ferguson Jr. and Wascher, “Distinguished Lecture on Economics in Government,” pp. 19 -21.
50 Ferguson Jr. and Wascher, “Distinguished Lecture on Economics in Government,” pp. 19 -21.
51 Robert J. Gordon, “T he Ascent and Descent of Growth,” in The Rise and Fall of American Growth: The U.S.
Standard of Living Since the Civil War (Princeton, NJ: Princeton University Press, 2016), pp. 2-7.
52 Daron Acemoglu and Pascual Restrepo, Artificial Intelligence, Automation, and Work, NBER, Working Paper no.
24196, January 2018, p. 13, at https://www.nber.org/papers/w24196.pdf.
53 Dale W. Jorgenson, Mun S. Ho, and Jon D. Samuels, Education, Participation, and the Revival of U.S. Economic
Growth, NBER, Working Paper no. 22453, July 2016, p. 3, at https://www.nber.org/papers/w22453.pdf. T he economist
Robert Gordon posited additional reasons why U.S. future productivity —and overall economic growth—will remain
slow, including changing demographics, rising inequality, globalization, global warming, and rising household and
government deficits. See Robert J. Gordon, Is U.S. Growth Over? Faltering Innovation Confronts the Six Headwinds,
NBER, Working Paper no. 18315, August 2012, pp. 18 -19, at https://www.nber.org/papers/w18315.pdf.
54 Susan Fleck, John Glaser, and Shawn Sprague, “T he Compensation-Productivity Gap: A Visual Essay,” Monthly
Labor Review, January 2011, p. 59 (hereinafter cited as Fleck, Glaser, and Sprague, “ Compensation -Productivity
Congressional Research Service
11

link to page 15 link to page 17 link to page 8 link to page 12 link to page 14 link to page 17 Social Security: Economic Growth and the Funding Shortfall

earnings have diverged (Figure 7).55 Economists general y cite two factors for the increasing gap
between productivity and real earnings. First, productivity is indexed using the implicit price
deflator, whereas earnings are indexed using the Consumer Price Index (CPI).56 In recent decades,
the CPI has grown faster than the implicit price deflator.57 Indexing general y al ows the
comparison of factors across time by accounting for changes in price (i.e., inflation). The fact that
labor productivity and real hourly compensation are indexed using different measures accounts
for some divergence between the two. Second, the labor share of income—the portion of
economic output that employers spend on workers’ earnings—has general y declined since
1980.58 This result has affected income for the Social Security program. Had workers real
earnings kept pace with gains in labor productivity, the resulting increase in real hourly
compensation would have increased the program’s revenue and increased future benefits.
Real Economic Output
The total value of an economy’s output is measured by GDP, defined as the total value of al
goods and services produced by a nation’s economy.59 Figure 8 shows real economic output for
the United States. As seen, the U.S. economy is general y expanding but periods of recessions
may result in periods of decreased real output, or contraction. For any given period, real GDP is
approximately equal to the product of total employment, average hours worked, and total factor
productivity in that period. Thus, the growth rate in real GDP is approximately equal to the sum
of the growth rates for total employment, average hours worked, and productivity.60
That is, if the annual percentage changes—growth rates—shown in Figure 2, Figure 4, and
Figure 6 were summed, the sum would equal the real economic growth rates shown in Figure 9.
The figures below also show the projections for real economic output and real economic growth.

Gap”).
55 Fleck, Glaser, and Sprague, “Compensation-Productivity Gap,” p. 57.
56 T he implicit price deflator, or gross domestic product (GDP) price deflator, is a measure of inflation in the prices of
goods and services produced in the United Sates, including exports. See Bureau of Economic Analysis, “ GDP Price
Deflator,” at https://www.bea.gov/data/prices-inflation/gdp-price-deflator.
57 Fleck, Glaser, and Sprague, “Compensation-Productivity Gap,” p. 60.
58 Fleck, Glaser, and Sprague, “Compensation-Productivity Gap,” p. 63.
59 For more information on GDP, see CRS In Focus IF10408, Introduction to U.S. Economy: GDP and Economic
Growth.
60 Board of T rustees, 2020 Annual Report, p. 111.
Congressional Research Service
12

link to page 18 link to page 18 link to page 18

Social Security: Economic Growth and the Funding Shortfall

Figure 8. Historical and Projected Real
Figure 9. Historical and Projected
Gross Domestic Product (GDP)
Growth in Real GDP
(in tril ions, chained 2012 dol ars)
(annual percentage change)

Source: CRS, from data provided in the 2020 Annual Report (AR), Table V.B2, at https://www.ssa.gov/OACT/TR/
2020/tr2020.pdf; Bureau of Economic Analysis (BEA), Real Gross Domestic Product; Federal Reserve Economic Data,
at https://fred.stlouisfed.org.
Notes: The BEA is used for historical data; projections are based on the Board of Trustees’ intermediate
assumptions in the 2020 AR. Assumptions for the projection period reflect the trustees’ understanding of the
OASDI program at the start of 2020; thus, it does not include potential effects of Coronavirus Disease 2019
(COVID-19). Shaded areas represent a year in which a recession occurred, as designated by the National Bureau
of Economic Research.
Table 1 shows the average annual growth rates for real GDP grouped by economic or business
cycle.61 For instance, the U.S. economy experienced positive real GDP growth from 1973 to
1979. The economy experienced a recession, or contraction, in 1979 but recovered and
experienced another period of sustained growth until 1990. Table 1 also shows the average
annual growth rates for the factors of real GDP grouped by the economic cycles. The U.S.
economy demonstrated its highest average annual growth rate during its 1969 to 1973 expansion,
averaging an annual growth in real GDP of 3.6%. Since then, the average annual growth rates in
GDP have general y decreased. As discussed, growth rates in total employment and average hours
worked have contributed to this slowdown in economic growth. To the extent that these
contributing factors show slow growth, and total factor productivity remains below its long-term
trend, overal real economic growth wil demonstrate slow growth rates as wel .
Table 1 also shows the trustees’ projections for future economic growth. The trustees project
annual average growth in real economic output to be 2.0% throughout the projection period. This
forecast matched the Congressional Budget Office’s (CBO’s) projection for growth for 2020 that
was published in January 2020. At the time, CBO assumed persistent low growth in total factor
productivity would lead to a further slowdown in overal growth and projected average annual
growth in GDP to be 1.7% for 2021 through 2030.62

61 For more information on the economic or business cycle, see CRS In Focus IF10411, Introduction to U.S. Economy:
The Business Cycle and Growth.
62 See CBO, “10-Year Economic Projections,” Budget and Economic Data (January 2020), at https://www.cbo.gov/
about/products/budget-economic-data. In May 2020, CBO released new economic projections that accounts for
changes in economic activity due to COVID-19.
Congressional Research Service
13

link to page 18 link to page 18 link to page 18 Social Security: Economic Growth and the Funding Shortfall

Table 1. Historical and Projected Real Growth and Determinants of Growth Rates
(average annual percentage change)
Total
Average Hours
Year(s)a
Employment
Worked
Productivity
Real GDP
Historical Economic Cycles
1969-1973
1.8
-0.87
2.65
3.6
1973-1979
2.4
-0.54
1.07
3.0
1979-1990
1.7
-0.10
1.41
3.0
1990-2001
1.2
0.10
1.85
3.2
2001-2007
1.1
-0.50
2.19
2.8
2007-2019
(ongoing)b
0.6
-0.01
1.07
1,7
Projectionsc
2020-2093
0.4
-0.05
1.63
2.0
(intermediate)
2020-2093
(low-cost)
0.6
0.05
1.94
2.7
2020-2093
0.2
-0.15
1.33
1.3
(high-cost)
Source: Congressional Research Service (CRS), from data provided in the 2020 Annual Report, Tables V.B1 and
V.B2, at https://www.ssa.gov/OACT/TR/2020/tr2020.pdf.
Notes: Figures may not sum to total due to rounding. The annual percentage change in real gross domestic
product (GDP), or real GDP growth, is the sum of the annual percentage changes in total employment, average
hours worked, and total factor productivity.
a. The economic cycles are measured from peak to peak. For instance, the 1969-1973 economic cycle is
measured from its beginning peak in 1969 to its ending peak in 1973; the subsequent cycle began in 1973
and ended with the next peak in 1979.
b. At the end of 2019, the economic expansion that began in 2007 (as measured from the peak of the most
previous complete economic cycle) was ongoing.
c. The Board of Trustees presents three alternative (intermediate, low-cost, and high-cost) sets of
assumptions for demographic, economic, and program-specific factors. The intermediate set of assumptions
represent the trustees’ best estimates of likely future conditions. Assumptions for the projection period
reflect the trustees’ understanding of the OASDI program at the start of 2020; thus, it does not include
potential effects of Coronavirus Disease 2019 (COVID-19).
Productivity, Demographics, and Growth
The U.S. population has been experiencing a shift in age structure toward older ages and an
increase in the median age, termed demographic aging.63 Decreasing fertility and increasing life
expectancy are the main factors for this demographic trend. On one hand, increased life
expectancy would indicate a longer period in retirement, thus a lower LFPR. On the other hand,
increasing life expectancy would indicate an increasing supply of labor as people live longer and
need to save more for retirement. Most research has supported the former interpretation—
demographic aging leads to an overal downward trend in labor force participation.64 As

63 CRS Report R45990, Social Security: Demographic Trends and the Funding Shortfall.
64 Loretta J. Mester, “Demographics and T heir Implications for the Economy and Policy,” speech given at the Cato
Congressional Research Service
14

link to page 14 link to page 14 Social Security: Economic Growth and the Funding Shortfall

discussed, a decrease in the supply of labor, or total employment, would have negative
implications for economic growth.
Demographic aging may also affect productivity. Although some slowdown in productivity (see
Figure 5 and Figure 6) may be cyclical, research suggests demographic aging affects this
slowdown. Data indicate that productivity increases when a worker enters the workforce, then it
stabilizes, and then it declines as a worker nears retirement.65 It is possible then that individual
productivity has not changed, but the changes in the workforce’s age structure have changed
overall productivity.
Potential Economic Output
Potential Impact of Coronavirus Disease 2019 (COVID-19)
The 2020 Annual Report reflects the trustees’ understanding of the program’s financial status at the start of 2020.
That is, “the report does not reflect the potential effects of the COVID-19 pandemic on the Social Security or
Medicare programs.”66
A variety of factors stemming from COVID-19 may have economic repercussions that affect the program’s
financial status. For instance, a decrease in payrol tax revenues resulting from lower-than-expected employment
or an increase in costs resulting from higher-than-expected beneficiaries would result in depletion of trust fund
asset reserves sooner than as projected under the trustees’ assumptions reflected in the 2020 report.
Alternatively, higher-than-expected revenues or lower-than-expected costs could delay the projected date of the
trust fund depletion.
Neither the short-range CBO projection nor the long-range trustees’ projection for real economic
growth include a projected economic contraction.67 Both projections assume growth rates below
the long-term average, but the growth rates are stil positive. Regardless of how likely an
economic contraction might be, much of the research since the 2007-2009 financial crisis
suggests that deep contractions may exhibit long-lasting and negative impacts on an economy’s
potential output.
Potential output, or potential GDP, is an economy’s normal level of production given its resources
and technology.68 Once a recession ends, the economy is expected to return to pre-contraction
levels. Deep recessions, such as the 2007-2009 financial crisis, can lead to a persistent decrease in
potential output: “Potential output fal s because a recession reduces capital accumulation, leaves
scars on workers who lose their jobs, and disrupts the economic activities that produce
technological progress.”69 Said differently, a level of potential GDP would no longer be possible
due to the potential continued impact of a deep recession. Thus, were real economic growth to

Institute’s 35th Annual Monetary Conference: T he Future of Monetary Policy, Washington, DC, November 16, 2017, p.
4, at https://www.clevelandfed.org/en/newsroom-and-events/speeches/sp-20171116-demographics-and-their-
implications-for-the-economy-and-policy.aspx (hereinafter cited as Mester, “ Demographics and T heir Implications for
Economy and Policy”).
65 Mester, “Demographics and T heir Implications for Economy and Policy,” p. 6.
66 Board of T rustees, 2020 Annual Report, p. 1.
67 In other words, the projections assume future GDP is always at potential. The trustees’ high-cost assumption includes
an assumed recession. Under the high-cost set of assumptions, all economic, demographic, and program -specific
factors are assumed to be high-cost. T hat is, the effects of slower economic growth is not isolated from other high-cost
assumptions. Board of T rustees, 2020 Annual Report, p. 18.
68 Laurence M. Ball, Long-Term Damage From the Great Recession in OECD Countries, NBER, Working Paper no.
20185, May 2014, p. 1, at https://www.nber.org/papers/w20185.pdf (hereinafter cited as Ball, “ Long-T erm Damage
from Great Recession”).
69 Ball, “Long-T erm Damage from Great Recession,” p. 1.
Congressional Research Service
15

link to page 6 link to page 26 link to page 7 link to page 7 Social Security: Economic Growth and the Funding Shortfall

unfold in a manner similar to CBO’s or the trustees’ projections, any decrease in actual growth
could shift the future path of output to a lower level.
The Trust Funds and Economic Growth
The “Current Status of the Trust Funds” section of this report discusses how the Social Security
program is projected to have funds necessary to pay full scheduled benefits until sometime in
2035. Once the trust funds assets are depleted, the programs’ payments wil depend on ongoing
tax revenues (i.e., the depletion of asset reserves does not mean Social Security wil be
completely broke and unable to pay benefits).70
As part of their annual report, the trustees publish sensitivity analyses for several of their
assumptions. These analyses al ow the effects of a change in one assumption to be captured and
its effects on the trust funds to be ascertained,71 al owing policymakers to understand that not al
economic growth factors would affect the program’s finances to the same extent. For example, an
increase in real economic growth resulting from an increase in the labor force does not
necessarily have the same impact on the program’s finances as a similar-sized increase in real
earnings. In analyzing the effects of economic growth on the program’s finances, it is helpful to
determine the source of the growth that generated the change (e.g., change in total employment
versus changes in productivity or real earnings).
Changes in the Size of the Labor Force
As discussed earlier, total employment is a factor in economic growth. It follows then that an
increase in total employment would have a positive effect on economic growth. As discussed,
population size is the most significant factor in determining the long-term size of the labor force.
An increase in the labor force would have a positive effect on economic growth and the trust
funds. In their 2020 Annual Report, the trustees provide sensitivity analyses for two demographic
factors—total fertility rate (TFR) and total net immigration (TNI)—that affect the size of the
labor force and how changes in those demographic factors would affect the financial status of the
trust funds.
Total Fertility Rate
The TFR is the average number of children that would be born to a woman throughout her
lifetime if she were to experience, at each age of her life, the birth rate observed in that year.72
The size of the labor force could be increased, with about a two decade lag, were the TFR in the
United States to increase, al else equal. That is, an increase in the TFR would result in more
people, such that were the LFPR (see Figure A-4) to remain relatively stable, more people would

70 For more information on the level of scheduled benefits payable should the trust fund assets become depleted, see
CRS Report RL33514, Social Security: What Would Happen If the Trust Funds Ran Out? .
71 T he discussed examples—changes in labor force (i.e., fertility and immigration) and real wage growth (i.e.,
productivity)—demonstrate the changes in the financial status of the Social Security program given the singular effects
of one factor of growth. T hey do not, however, demonstrate the possible combined effects if more than one factor of
growth were to change at the same time. T he definition of real economic growth (see “ Components of Real Economic
Growth”), indicates that changes in one factor of growth would have a multiplicative effect on overall growth and an
additive effect on growth rates.
72 Samuel H. Preston, Patrick Heuveline, and Michel Guillot, Demography: Measuring and Modeling Population
Processes (Malden, MA: Blackwell Publishers, 2001), p. 95 .
Congressional Research Service
16

link to page 21 link to page 21 link to page 21 link to page 26 link to page 21 link to page 21 link to page 21 Social Security: Economic Growth and the Funding Shortfall

be available for work. As discussed, an increase in the number of workers, and a subsequent
increase in employment, would result in an increase in real economic output.
The trustees’ intermediate assumption for the long-range TFR is 1.95, 15% higher than the
estimated TFR observed in 2019.73 With this assumption, and their other intermediate
assumptions, the trustees project a long-range actuarial balance of -3.21% of taxable payroll
(Table 2) and a projected trust fund reserve depletion date of 2035. Table 2 also gives the
projected actuarial balance and trust fund reserve depletion date under different assumptions for a
future TFR. Under an assumption for a TFR of 2.15—approximately 28% higher than the
estimated TFR observed in 2019—the trustees project an actuarial balance of -2.79% of taxable
payroll and a projected trust fund reserve depletion date of 2035. Table 2 shows that although an
increase in the TFR would result in a stronger financial position (i.e., smal er actuarial deficit),
the trust funds’ reserves are stil projected to deplete in 2035. The trustees estimate that each
increase of 0.1 in the TFR would increase the long-range actuarial balance by roughly 0.22% of
taxable payroll.74 Said differently, each increase of 0.1 in the TFR would decrease the actuarial
deficit by 0.22% of taxable payroll. Thus, it follows that the long-range assumption for the TFR
would need to increase by about 1.46, from 1.95 to 3.41, to eliminate the projected financial
shortfal . The United States last experienced a TFR of 3.4 or higher in 1962 (see Figure A-2).
Table 2. Total Fertility Rate and Projected Financial Measures
(% of taxable payrol )

Total Fertility Rate (TFR)

1.75a
1.95b
2.15c
Summarized Income Rate (2020-2094)
13.89
13.85
13.80
Summarized Cost Rate (2020-2094)
17.55
17.06
16.59
Actuarial Balance (Income less Cost; 2020-2094)
-3.66
-3.21
-2.79
Year of OASDI Trust Fund Reserve Depletion
2035
2035
2035
Source: CRS, from data provided in the 2020 Annual Report, Tables VI.D1, at https://www.ssa.gov/OACT/TR/
2020/tr2020.pdf.
Notes: The TFR for a year is defined as the average number of children that would be born to a woman in her
lifetime, at each age of her life, if she were to experience the birth rate observed for a specified year. OASDI =
Old-Age, Survivors, and Disability Insurance.
a. A TFR of 1.75 is the trustees’ high-cost assumption for the 75-year projection period. From the standpoint
of the program’s finances, this assumption would be the most unfavorable
b. A TFR of 1.95 is the trustees’ intermediate assumption for the 75-year projection period. From the
standpoint of the program’s finances, this assumption represents the trustees’ best guess of future
conditions.
c. A TFR of 2.2 is the trustees’ low-cost assumption for the 75-year projection period. From the standpoint of
the program’s finances, this assumption would be the most favorable.
Changes in the size of employment due to changes in the fertility of American women are
unlikely to substantial y change year-to-year. Were the TFR to differ greatly from the trustees’

73 Board of T rustees, 2020 Annual Report, pp. 87-88. In each of the trustees’ annual reports, the total fertility rate
(T FR) for the most recent calendar year is estimated (e.g., 1.68 in 2019). T he U.S. T FR has demonstrated a wide range.
T FR can be affected by many factors, including social attitudes, demographic composition, economic conditions, and
birth control practices. A T FR assumption of about 1 .95 is close to the long-run average since 2020 end of the period
resulting in the baby boom (i.e., the U.S. T FR has average 2.0 since 1964).
74 Board of T rustees, 2020 Annual Report, p. 181.
Congressional Research Service
17

link to page 23 link to page 23 Social Security: Economic Growth and the Funding Shortfall

assumptions, it would stil take time for the effects to change the total employment size and
influence economic growth.
Total Net Immigration
Immigration could have a more immediate impact, as the size of the labor force could be
increased with an increase in net immigration, especial y if the immigrants were of working age.
The total net immigration (TNI) is the sum of immigration from both lawful permanent residency
and other-than-lawful permanent residency.75 An increase in TNI would result in more people,
such that were the LFPR to remain relatively stable, more people would be available for work.
Similar to the TFR, an increase in workers, and a subsequent increase in employment, would
result in an increase in real economic output. One difference between an increase in TNI and in
the TFR is that an increase in TNI would have a more immediate effect on the number of persons
available for work.
The trustees’ intermediate assumption of the long-range TNI is 1,261,000 persons per year,
approximately 14% lower than the estimated TNI observed in 2019.76 The trustees’ intermediate
assumptions project a long-range actuarial balance of -3.21% of taxable payroll (Table 3) and a
projected trust fund reserve depletion date of 2035. Table 3 shows that under a higher average
annual TNI—1,598,000 immigrants per year—the projected actuarial balance would be -2.96% of
taxable payroll, and the projected date for trust fund reserve depletion would be 2035. That is, the
program would be in a stronger financial position (i.e., smal er actuarial deficit) if an increase in
immigration were to result in an increase in the size of the labor force, but the projected date for
trust fund reserve depletion would not change. The trustees estimate that each increase of 100,000
in average annual TNI would improve the long-range actuarial balance by 0.08% of taxable
payroll.77 Under this assumption, it would require an increase in the average annual TNI
assumption of 4.0 mil ion immigrants, from 1.26 mil ion to 5.27 mil ion, to eliminate the
projected financial shortfal .

75 Lawful permanent residency results from persons who are granted lawful permanent resident (LPR) status or those
who change their status to LPR (e.g., those with legal visas but without LPR status). Other -than-lawful permanent
residency results from persons not granted LPR status, such as undocumented immigrants or st udents with temporary
visas. Board of T rustees, 2020 Annual Report, pp. 88-89.
76 In the 2019 Annual Report, the trustees cite two main reasons for this assumption. First, the T rump Administration
has decreased the ceiling for annual refugees. Second, previous reports assumed that some people deferred immigration
(i.e., other-than-lawful permanent residency immigration) due to the 2008 -2009 recession. Recent immigration
estimates have not supported this assumption, so the 2019 Annual Report has removed the a ssumption of a temporary
surge in immigration. The 2019 Annual Report of the Board of Trustees of the Federal Old -Age and Survivors
Insurance and Federal Disability Insurance Trust Funds, April 22, 2019, p. 74, at https://www.ssa.gov/OACT /T R/
2019/tr2019.pdf.
77 Board of T rustees, 2020 Annual Report, p. 184.
Congressional Research Service
18

link to page 23 link to page 23 link to page 23 link to page 25 Social Security: Economic Growth and the Funding Shortfall

Table 3. Total Net Immigration and Projected Financial Measures
(% of taxable payrol )

Average Annual Total Net Immigration (TNI)

946,000a
1,261,000b
1,598,000c
Summarized Income Rate (2020-2094)
13.88
13.85
13.91
Summarized Cost Rate(2020-2094)
17.35
17.06
16.78
Actuarial Balance (income less cost; 2020-2094)
-3.47
-3.21
-2.96
Year of OASDI Trust Fund Reserve Depletion
2034
2035
2035
Source: CRS, from data provided in the 2020 Annual Report, Tables VI.D3, at https://www.ssa.gov/OACT/TR/
2020/tr2020.pdf.
Notes: Average annual TNI is the annual total net immigration from lawful permanent residents and other-than-
lawful permanent residents averaged over the 75-year projection period.
a. A TNI of 946,000 per year is the trustees’ high-cost assumption for the 75-year projection period. Of the
three assumptions shown in the table, this assumption would be the most unfavorable from the standpoint
of the program’s finances.
b. A TNI of 1,261,000 per year is the trustees’ intermediate assumption for the 75-year projection period. Of
the three assumptions shown in the table, this assumption represents the trustees’ best guess of future
conditions.
c. A TNI of 1,598,000 per year is the trustees’ low-cost assumption for the 75-year projection period. Of the
three assumptions shown in the table, this assumption would be the most favorable from the standpoint of
the program’s finances.
Changes in Real Wages
Although the 2020 Annual Report does not include a sensitivity analysis for productivity, it does
include a sensitivity analysis for real earnings growth, shown as the real-wage differential in
Table 4. In the long run, real wage growth is largely determined by productivity growth.78 So,
assuming the usual relationship between productivity and real wages were to continue, an
increase in real wages would imply an increase in productivity and increase real economic
growth.
Thus, if wages increase in line with productivity gains, the effects of productivity assumptions
may be gleaned from examining real (after-inflation) earnings growth.
Productivity, a key factor in real economic growth, has large implications for program finances
through its effects on real earnings growth. On the program level, earnings determine the amount
of payroll taxes; on the worker level, earnings determine initial benefits.79 An increase in
productivity growth, or real earnings growth, would have a positive impact on economic growth
and the Social Security program’s finances. Specifical y, an increase in real earnings growth
would lead to an increase in benefits for current and future workers but have no effect for current
beneficiaries.80 That is, an increase in real earnings growth could have an immediate positive

78 CBO T estimony of Douglas Holtz-Eakin, House Committee on Ways and Means, Subcommittee on Social Security,
The Role of the Econom y in the Outlook for Social Security, hearing, 109th Cong., 1st sess., June 21, 2005, p. 5, at
http://www.cbo.gov/sites/default/files/cbofiles/ftpdocs/64xx/doc6492/06-21-prefunding.pdf.
79 Payroll taxes are the largest source of program income. For more information, see CRS Report RL33028, Social
Security: The Trust Funds.
80 While initial benefits are determined by an individual’s wages and indexed to average wage growth in the economy,
subsequent benefit increases are indexed to inflation. For more information, see CRS Report R43542, How Social
Congressional Research Service
19

link to page 25 link to page 25 link to page 18 link to page 18 Social Security: Economic Growth and the Funding Shortfall

effect on the program’s financial status through increased income but not result in immediate
higher costs.81 An increase in real earnings growth would thus increase the long-range actuarial
balance (i.e., smal er actuarial deficit).82
The trustees’ intermediate assumption for real earnings growth is shown in Table 4 as the real-
wage differential. The real-wage differential is the difference between the annual percentage
increase in average wages and the annual percentage increase in the CPI. For instance, the
intermediate assumption for the long-range real-wage differential is 1.2%, indicating the trustees
expect after-inflation wage growth to be 1.14%. Under these assumptions the trustees project a
long-range actuarial balance of -3.21% of taxable payroll (Table 4) and a projected trust fund
reserve depletion date of 2035. The table shows that under a higher real-wage differential (i.e., if
wage growth outpaced inflation by1.76 percentage points), the projected actuarial balance would
be -2.09% of taxable payroll, and the projected date for trust fund reserve depletion would be
2037. Thus, as the real-wage differential, or real earnings, increases, it results in a stronger
financial position (i.e., smal er actuarial deficit) and a later projected date for trust fund reserve
depletion. The trustees estimate that each 0.1 percentage-point increase in the differential would
increase the long-range actuarial balance by about 0.18% of taxable payroll.83 Under this
assumption, it would require an increase in the real wage differential of roughly 1.78 percentage
points, from 1.14% to 2.92%, to eliminate the projected financial shortfal .84 The United States
has experienced real-wage differentials of this magnitude before (e.g., in 1998-1999) but not on a
sustained basis. For instance, over the last five-year period (i.e., 2015 through 2019) the real-
wage differential averaged 1.55%; since 1960, no five-year period has experienced a real-wage
differential of at least 2.92% (the five-year period of 1996 through 2000 had a real-wage
differential of 2.88%).85 Furthermore, Table 1 shows that of the past U.S. economic cycles, the
1969 through 1973 period resulted in the highest average annual growth in real GDP (i.e., 3.6%).
The real-wage differential during this period was 1.08%.86

Security Benefits Are Com puted: In Brief.
81 T his result assumes that increases in real earnings would occur at levels subject to the Social Security payroll tax.
Since 1983, the aggregate amount of earning subject to the payroll tax h as declined (see Ratio of OASDI Taxable
Payroll to Covered Earnings, at https://www.ssa.gov/OACT /T R/2020/2020_Long-
Range_Economic_Assumptions.pdf). For more information on the taxable earnings base, see CRS Report RL32896,
Social Security: Raising or Elim inating the Taxable Earnings Base.
82 Social Security Advisory Board, 2019 Technical Panel on Assumptions and Methods, Report, September 2019, p. 63,
at https://ssab.gov/Portals/0/TPAM2019/TPAM%202019%20FINAL%20REPORT_508.pdf?ver=2019-09-24-092004-
910.
83 Board of T rustees, 2020 Annual Report, p. 185.
84 In a 2016 analysis, the Urban Institute published the effects on the program’s finances resulting from an increase in
productivity that would lead to indefinite 3.4% average real economic growth (as shown in Table 1, the trustees project
average real economic growth to be 2.1%). T his would represent a real economic growth rate that has not been
experienced in the U.S. economy since the 1969 -1973 economic expansion. T he analysis showed that such an increase
in real economic growth would delay the trust fund asset reserve depletion date by about 30 years. However, according
to the analysis, even an average real growth rate of 3.4% would not result in avoidance of a financial shortfall. Richard
W. Johnson and Karen E. Smith, “ Can economic growth really fix Social Security?,” Urban Institute, February 5, 2016,
at https://www.urban.org/2016-analysis/can-economic-growth-really-fix-social-security.
85 Board of T rustees, 2020 Annual Report, pp. 106-108.
86 Board of T rustees, 2020 Annual Report, p. 105.
Congressional Research Service
20

link to page 25 link to page 25 link to page 25 Social Security: Economic Growth and the Funding Shortfall

Table 4. Real-Wage Differential and Projected Financial Measures
(% of taxable payrol )

Real-Wage Differential

0.52%a
1.14%b
1.76%c
Summarized Income Rate (2020-2094)
14.03
13.95
13.68
Summarized Cost Rate (2020-2094)
18.39
17.06
15.77
Actuarial Balance (Income less Cost; 2020-2094)
-4.36
-3.21
-2.09
Year of OASDI Trust Fund Reserve Depletion
2033
2035
2037
Source: CRS, from data provided in the 2020 Annual Report, Tables VI.D4, at https://www.ssa.gov/OACT/TR/
2020/tr2020.pdf.
Notes: The real-wage differential is the difference in the annual percentage increase in average wages in covered
employment and the annual percentage increase in the Consumer Price Index (CPI).
a. A real-wage differential of 0.52% is the trustees’ high-cost assumption for the 75-year projection period. Of
the three assumptions shown in the table, this assumption would be the most unfavorable from the
standpoint of the program’s finances. Assuming a 2.40% increase in the CPI, the Board of Trustees’
intermediate assumption for price growth, results in an average annual wage increase of 2.92% in covered
employment.
b. A real-wage differential of 1.14% is the trustees’ intermediate assumption for the 75-year projection period.
Of the three assumptions shown in the table, this assumption represents the trustees’ best guess of future
conditions. Assuming a 2.40% increase in the CPI results in an average annual wage increase of 3.54% in
covered employment.
c. A real-wage differential of 1.76% is the trustees’ high-cost assumption for the 75-year projection period. Of
the three assumptions shown in the table, this assumption would be the most favorable from the standpoint
of the program’s finances. Assuming a 2.40% increase in the CPI results in an average annual wage increase
of 4.16% in covered employment.
Conclusion
Economic growth has a significant impact on the Social Security program’s financial status and
the trust funds. The financial status of the Social Security program, as measured by the
combination of the actuarial deficit and the trust fund balance, indicate how long the program is
able to pay full scheduled benefits. In the absence of trust fund asset reserves, the level of payable
benefits is that which is able to be supported by ongoing tax revenues. The financial health of the
Social Security program would respond favorably to increases in the contributors to economic
growth—total employment (i.e., a combination of population growth and increase in LFPR),
average hours worked, and productivity. However, given the current demographic (i.e.,
population aging) and economic forces (i.e., low productivity and low real wage growth) at work
in the United States, it is unlikely that these factors could combine in such a manner that would
result in real economic growth capable of avoiding the projected shortfal estimated under current
conditions. The trustees’ intermediate assumptions, their best estimates as to the program’s future
experience, suggest that low growth is contributing to the projected financial shortfal (the
trustees’ intermediate assumptions reflect their understanding of the program at the beginning of
2020 and do not reflect any potential effects of COVID-19). Although increases in real economic
growth would benefit both workers and the program as a whole, these alone would not likely
result in a permanent solution to the projected financial shortfal .
Congressional Research Service
21

Social Security: Economic Growth and the Funding Shortfall

Appendix. Total Employment
Figure A-1. Historical and Projected
Figure A-2. Historical and Projected Total
U.S. Total Population
Fertility Rate (TFR)
(annual percentage change)
(average number of children born)

Source: Congressional Research Service (CRS),
Source: CRS, from data provided in the 2020
from data accompanying the 2020 Annual Report
Annual Report (AR), Table V.A1, at
(AR), at https://www.ssa.gov/OACT/HistEst/
https://www.ssa.gov/OACT/TR/2020/tr2020.pdf.
Population/2020/Population2020.html;
Notes: Projections based on the Board of Trustees’
2017 National Population Projections Tables, US
intermediate assumptions in the 2020 AR. The TFR is
Census Bureau, at https://www.census.gov/
the average number of children born to a woman if
programs-surveys/popproj.html.
she were to experience the birth rates observed
Notes: Projections based on the Board of
from ages 14 to 49.
Trustees’ intermediate assumptions in the 2020
AR.
Figure A-3. Historical and Projected
Figure A-4. Historical Labor Force
Total Net Immigration (TNI)
Participate Rate (LFPR)
(in thousands)
(% of 16 and over civilian population)

Source: CRS, from data provided in the 2020

Annual Report (AR), Table V.A2, at
Source: CRS, from data provided by the Bureau of
https://www.ssa.gov/OACT/TR/2020/tr2020.pdf.
Labor Statistics.
Notes: Projections based on the Board of
Notes: The LFPR is the labor force as a percent of
Trustees’ intermediate assumptions in the 2020
the civilian noninstitutional population.
AR. TNI includes lawful permanent residents and
other-than-lawful permanent residents (e.g.,
undocumented immigrants).

Congressional Research Service
22

Social Security: Economic Growth and the Funding Shortfall

Author Information

Barry F. Huston

Analyst in Social Policy

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

Congressional Research Service
R46479 · VERSION 1 · NEW
23

Download PDF

Download EPUB

Revision History

Aug. 6, 2020

HTML · PDF

Metadata

Topic areas

Economic Policy

Domestic Social Policy

Report Type: CRS Report

Source: CRSReports.Congress.gov

Raw Metadata: JSON