Social Security: The Effects of Wage and Price
June 16, 2021
Indexing on Benefits
Barry F. Huston,
Social Security is a social insurance program that protects insured workers and their family
Coordinator
members against a loss of income due to old age, disability , or death. The program is comprised
Analyst in Social Policy
of two components: Old-Age and Survivors Insurance (OASI) and Disability Insurance (DI).

They are commonly referred to on a combined basis as OASDI.
Paul S. Davies
Specialist in Income
In April 2021, there were approximately 65.0 million Social Security beneficiaries collecting an
Security
average monthly benefit of $1,429. Monthly Social Security benefits are determined by federal

law. The Social Security benefit formula is first applied to calculate a worker’s primary insurance
amount (PIA), which is the basic monthly benefit amount payable at the Social Security full
Tamar B. Breslauer
retirement age before any applicable adjustments based on early or delayed retirement or other
Senior Research Librarian
factors. The computation process involves three main steps:

1. First, a summarized measure of lifetime Social Security–covered earnings, called the
average indexed monthly earnings (AIME), is computed. Each year of a worker’s taxable
earnings after 1950 are increased by the growth in average earnings in the economy from the year of work until two
years before eligibility for benefits, which for retired workers is age 60. The highest 35 years of these earnings are then
averaged to provide the AIME.
2. Second, a progressive benefit formula is applied to the AIME to compute the primary insurance amount (PIA). This
step requires a worker’s AIME to be sectioned into three brackets (or segments) of earnings using dollar amounts known
as bend points. Each bracket of earnings is replaced by a fixed percentage. The dollar bend points that create these
brackets, however, are adjusted annually to account for growth in overall economy-wide earnings.
3. Third, the PIA is adjusted based on the age at which a beneficiary chooses to begin receiving benefits. For retired
workers who claim benefits at the full retirement age (FRA) and for disabled workers, the monthly benefit equals the
PIA. Retired workers who claim earlier than the FRA receive monthly benefits lower than the PIA (i.e., an actuarial
reduction), and those who claim later than the FRA —up to age 70—receive benefits higher than the PIA (i.e., a delayed
retirement credit).
After a beneficiary’s first year of eligibility (i.e., age 62), monthly benefit amounts are adjusted for inflation. Thus, the actual
benefit amount depends on worker-specific factors such as past earnings and the age at which benefits are claimed , as well as
cohort-specific factors such as wage growth and price growth in the overall economy. The cohort-specific factors affect
workers of a certain birth year differently than workers of other birth years. Given this, a change in wage or price growth may
affect a birth cohort’s portion of earnings that are replaced by monthly benefit amo unts (i.e., replacement rate). The current-
law benefit formula generally results in stable replacement rates across cohorts. From year to year, the average benefits that
new beneficiaries receive increase at approximately the same rate as average earnings in the economy.
Recent news articles have suggested that deviations from long-term trends in wage and price growth experienced in 2020
could affect Social Security benefit amounts for the 1960 birth cohort—those turning 60 in 2020 (i.e., the year to which past
earnings are wage indexed). This report examines the ways in which this birth cohort’s monthly benefit amounts would be
affected by wage and price indexing.

Congressional Research Service

Social Security: The Effects of Wage and Price Indexing on Benefits

Introduction
Recent news articles have raised concerns that a decrease in the national average wage in 2020
wil lead to a corresponding decrease in the Social Security benefits for those reaching age 60 in
2020.1 While Social Security benefits are based on a worker’s earnings record in covered
employment, the calculations are also affected by changes in national wages and prices.
Average monthly Social Security benefits general y increase because of wage-indexing and price-
indexing. The Social Security benefit formula indexes a worker’s earnings to overal economy-
wide earnings. As a result, replacement rates—the portion of earnings that initial benefits
replace—increase at approximately the same rate as average earnings in the economy. Benefit
amounts collected after the earliest eligibility age (EEA) general y increase because of price-
indexing. Benefits after EEA are indexed to the Consumer Price Index for Urban Wage Earners
and Clerical Workers (CPI-W) through an annual cost-of-living-adjustment (COLA).
Although wages and prices general y increase over time, instances may arise where wages, prices,
or both decrease. Under such conditions, as may result from the recent recession caused by the
COVID-19 pandemic, benefit amounts would be affected. Additional y, benefit amounts for
family members of workers would also be affected.
As of April 2021, there were approximately 65.0 mil ion Social Security beneficiaries collecting
an average monthly benefit of $1,429. Retired-worker and disabled-worker beneficiaries
accounted for 84.1% of the beneficiary population. The largest single category of beneficiaries
was retired workers (71.6%), with an average monthly benefit of $1,552. The second-largest
category was disabled workers (12.5%), with an average monthly benefit of $1,280. Monthly
Social Security benefits are determined by federal law. These monthly benefits constitute a
substantial portion of income for a large segment of recipients. As such, changes in the long-term
trends in wage and price growth could have a substantial impact on retirement income security
through their effects on monthly benefit amounts. This possibility has led to increased
congressional interest and has prompted hearings by the House Ways and Means Subcommittee
on Social Security.2
This report outlines how the Social Security benefit formula was developed and how the formula
is affected by changes in wage and price growth in the economy. The report provides examples on
how wage growth would affect Social Security benefits for workers with different earnings
patterns and how price growth would affect benefit amounts after eligibility. The report also
outlines legislative options that are available to policymakers.
Brief History of Social Security Benefit Calculation
Prior to 1972, the average level of monthly Social Security benefits required congressional action
to increase. That is, the average level of benefits—the general level of pre-retirement earnings

1 Andrew G. Biggs, “If You Were Born in 1960, You Pay Extra for Covid-19,” Wall Street Journal, May 11, 2020,
https://www.wsj.com/articles/if-you-were-born-in-1960-you-pay-extra-for-covid-19-11589235056; and John
Waggoner, “ Born in 1960? Your Social Security Retirement or Disability Benefits Could T ake a Hit ,” AARP,
November 18, 2020, https://www.aarp.org/retirement/social-security/info-2020/pandemic-impacts-1960-birth-year-
benefits.html.
2 On July 17, 2020, the Subcommittee on Social Security held a hearing titled The Impact of COVID-19 on Social
Security and Its Beneficiaries. See https://waysandmeans.house.gov/legislation/hearings/impact -covid-19-social-
security-and-its-beneficiaries.
Congressional Research Service

1

Social Security: The Effects of Wage and Price Indexing on Benefits

replaced at the full retirement age by the Social Security benefit—was the result of ad hoc
legislation.3
A worker’s benefit amount was calculated by first determining his or her average monthly wage
(AMW) as the average of nominal earnings from al years of employment after 1950.4 Next, the
primary insurance amount (PIA) was calculated by separating the AMW into several brackets and
applying a replacement factor to each bracket. For example, the Social Security Amendments of
1958 (P.L. 85-840), effective for benefits payable in January 1959, established two brackets
($0.00-$110.00 and $110.01-$400.00) and two replacement factors (58.85% and 21.40%) applied
respectively. When Congress decided to increase benefits, it could add additional brackets and/or
increase replacement factors. For instance, the Social Security Amendments of 1964 (P.L. 87-64),
effective for benefits paid in January 1965, resulted in an average benefit increase of 7% by
establishing three brackets ($0.00-$110.00, $110.01-$400.00, and $400.01-$550.00) and three
replacement factors (62.97%, 22.90%, and 21.40%) applied respectively.5 Thus, the 1964
amendments increased benefits by creating a new bracket (i.e., earnings between $400.01 and
$550.00) and increasing the replacement factors (e.g., the replacement factor for the first $110.00
of earnings increased from 58.85% to 62.97%). Under this method, no parameters used to
calculate benefits were indexed.
This method of benefit calculation had a major drawback. During periods of high inflation (i.e.,
larger than average increases in the price index), the purchasing power of beneficiaries could be
substantial y degraded.6 To prevent the purchasing power of beneficiaries from declining,
Congress would need to continue to pass legislation. As can be seen in the above examples,
increasing the number of brackets and their respective replacement rates could result in a benefit
increase for both current and future beneficiaries. Mathematical y, the current and future benefits
were increased by the exact same percentage. This method of benefit calculation coupled the
effects of average wage growth and average price growth. That is, the benefit increases passed by
Congress included an adjustment for price growth and an adjustment for wage growth. So long as
wage and price growth remained relatively stable, this method (i.e, ad hoc legislation to maintain
purchasing power) kept initial benefits aligned with wage growth and current benefits aligned
with price growth.7 However, the economic conditions experienced in the 1970s (inflation) led to
higher-than-expected benefits for future beneficiaries, thereby creating higher-than-expected
program costs. Congress made several changes in subsequent years to address issues stemming
from this coupling.

3 For more information on Social Security amendments that resulted in benefit increases, see CRS Report RL30920,
Social Security: Major Decisions in the House and Senate Since 1935 , by T amar B. Breslauer and William R. Morton.
4 T he definition of average monthly wage was expanded many times to incorporate the expanding nature of Social
Security itself. For instance, in 1954 the definition of AMW was updated to exclude periods of disability. For more
information, see T able 2.A10 in Social Security Administration (SSA), Annual Statistical Supplem ent, November 2019,
https://www.ssa.gov/policy/docs/statcomps/supplement/2019/2a8-2a19.html.
5 For more information on historical changes to AMW brackets and replacement factors, see T able 2.A16 in SSA,
Annual Statistical Supplem ent, November 2019.
6 However, without allowing future wage or price increases to affect benefit levels—that is, unless Congress acted—the
long-range costs of the Social Security program (i.e., aggregate benefits paid) were well known.
7 Larry DeWitt, Daniel Beland, and Edward Berkowitz, Social Security: A Documentary History (T housand Oaks, CA:
SAGE Publications, 2008), p. 286.
Congressional Research Service

2

Social Security: The Effects of Wage and Price Indexing on Benefits

The Social Security Amendments of 1972
Among other provisions, the Social Security Amendments of 1972 (P.L. 92-336) increased Social
Security benefits by an average of 20% and provided for future automatic increases in benefits
when the Consumer Price Index (CPI) rose by more than 3.0%.8 In 1972, inflation was at
relatively low rates and expected to decline, but by 1973, inflation had surpassed expectations.9
The automatic increases that were part of the 1972 amendments made the benefit calculations
more sensitive to changes in the relationship between wages and prices. This produced higher
benefits for future beneficiaries and also increased program costs. As Larry DeWitt, Daniel
Beland, and Edward Berkowitz note in Social Security: A Documentary History:
As the economics of the 1970s produced disruptions in the historic relationship between
prices and wages, this had an adverse effect on Social Security benefits. In a period of
stagflation, overall program costs soared and initial benefit levels for future beneficiaries
rose much higher than planned.10
The Social Security Amendments of 1977
The Social Security Amendments of 1977 (P.L. 95-216) decoupled the effect of wage growth and
price growth on Social Security benefits that were shown to have introduced instability in
replacement levels. Decoupling the benefit calculation was also expected to relieve some
financial pressure on the system. It was projected that, without decoupling legislation, a worker’s
future benefits could exceed his or her pre-retirement earnings.11 The 1977 amendments
essential y designed the current-law benefit formula by establishing a benefit formula that indexes
workers’ initial benefit levels to wage growth and then indexes their future benefits to price
growth.
Omnibus Budget Reconciliation Act of 1986 (OBRA 86)
Automatic COLAs became effective in 1975 after being established in 1972 as part of the Social
Security Amendments of 1972 (P.L. 92-336). Initial y, the COLA formula required inflation to be
at least 3% during the specified base period before a COLA could be triggered. As part of OBRA
86 (P.L. 99-509), lawmakers eliminated the 3% trigger, requiring instead that inflation (or wage
growth in certain cases) be greater than 0% during the specified base period for a COLA to be
payable. This requirement effectively al owed for a COLA smal er than 3% to be paid while
continuing to protect benefits from being decreased during periods of declining prices (i.e., the
Social Security COLA cannot be negative and thus cannot reduce benefit levels).
The following section explains the current-law benefit calculation in more detail.

8 At the time, there was only one CPI measure. T o finance future benefit increases resulting from price growth, the
amendments automatically increased the taxable wage base. For more information, see CRS Report RL32896, Social
Security: Raising or Elim inating the Taxable Earnings Base, by Zhe Li. P.L. 92-603 is also referred to as the Social
Security Amendments of 1972.
9 James Kelley and Joseph Humphreys, Final Report on the Social Security ‘Notch’ Issue, Commission on the Social
Security “Notch” Issue, Appendix, 1994, https://www.ssa.gov/history/notchbase.html.
10 DeWitt, Beland, and Berkowitz, Social Security: A Documentary History, p. 20. Stagflation exists when an economy
experiences both high inflation and high unemployment. T he U.S. economy experienced such conditions during the
1970s. For more information on stagflation, see CRS Report R41656, Changing the Federal Reserve’s Mandate: An
Econom ic Analysis, by Marc Labonte.
11 DeWitt, Beland, and Berkowitz, Social Security: A Documentary History, pp. 298-321.
Congressional Research Service

3

link to page 24 Social Security: The Effects of Wage and Price Indexing on Benefits

Social Security Benefit Calculation Under Current
Law
The Social Security benefit formula is used to compute benefit amounts paid to both insured
workers and their eligible dependents and survivors. Benefit amounts are based on a worker’s
past Social Security–taxable (i.e., covered) earnings and can be affected by program factors that
are indexed to growth in national wages.12 For example, lifetime earnings in covered employment
are wage-indexed to account for growth in economy-wide earnings over a worker’s career.
The benefit formula is applied in three main steps: (1) computing the average indexed monthly
earnings, (2) converting these earnings into the primary insurance amount, and (3) applying
adjustments based on the Social Security benefit claiming age.13
To account for inflation, an annual COLA is applied to benefits collected after the first year of
eligibility based on changes in the CPI-W.14
Average Indexed Monthly Earnings
In the first step of the benefit computation, a summarized measure of lifetime Social Security-
covered earnings is computed, cal ed the average indexed monthly earnings (AIME). In this step,
the Social Security Administration (SSA) first indexes a worker’s lifetime covered earnings to
reflect changes in national wage levels, as measured by SSA’s Average Wage Index (AWI).15 The
AWI includes al wages that are subject to federal income tax. Thus, the AWI includes wages
earned from covered and non-covered employment, as wel as earnings in excess of the taxable
maximum.16 This indexing is done by increasing each year of a worker’s taxable earnings after
1950 by the growth in average earnings in the economy, as measured by the AWI, from the year
of work until two years before eligibility for benefits, which for retired workers is at age 60.
(Workers are first eligible for benefits at age 62.17) For example, the SSA’s AWI grew from
$32,155 in 2000 to $41,674 in 2010. So if a worker earned $20,000 in 2000 and turned 60 in

12 Some program factors may also be affected by COLAs, as some factors change only when a COLA is payable.
13 Retired-worker benefits can be affected by other adjustments. For example, the windfall elimination provision can
reduce benefits for individuals who receive pensions based on employment not covered by Social Security, and benefits
can be temporarily withheld under the retirement earnings test if a beneficiary under the full retirement age continues to
work and earns above a certain amount. Social Security benefits can also be subject to income tax, thereby affecting the
beneficiary’s net income. For more information on possible adjustments, see CRS In Focus IF10203, Social Security:
The Windfall Elim ination Provision (WEP) and the Governm ent Pension Offset (GPO) , by Zhe Li. For more
information on how Social Security benefits may be affected by income tax, see CRS Report RL32552, Social
Security: Taxation of Benefits, by Paul S. Davies. For more information on the retirement earnings test, see CRS Report
R41242, Social Security Retirem ent Earnings Test: How Earnings Affect Benefits, by Zhe Li.
14 Table A-4 lists selected items used in the Social Security benefit formula and their relationship with CPI -W and
AWI.
15 SSA, Office of the Chief Actuary (OCACT ), “ National Average Wage Index,” https://www.ssa.gov/oact/cola/
AWI.html.
16 20 C.F.R. §404.211(c). Under current law, the Social Security payroll tax is applied to covered earnings up to an
annual limit, or taxable maximum ($142,800 in 2021). T his level of earnings is both the contribution base (i.e., the
amount of covered earnings subject to the Social Security payroll tax) and the benefit base (i.e., amount o f earnings
used to determine benefits). As such, it is commonly referred to as the contribution and benefit base.
17 SSA uses the national average wage indexing series to ensure that future benefits reflect the general rise in the
standard of living over the course of a worker’s earning history. For details, see OCACT , “ Index Earnings Used to
Compute Initial Benefits,” in “National Average Wage Index.”
Congressional Research Service

4

link to page 9 link to page 9 Social Security: The Effects of Wage and Price Indexing on Benefits

2010, the 2010 indexed wage for 2000 would be $20,000 × ($41,674/$32,155), or $25,921.
Earnings from later years—for retired workers, at ages 60 and above—are not indexed.18 After
indexing, the highest 35 years of earnings are summed, and the total is divided by 420 (the
number of months in 35 years) to determine a worker’s AIME.19 The indexing process ensures
that a worker’s or family member’s benefit wil reflect increases in the average wage growth
observed over the worker’s earnings history.
Primary Insurance Amount
The second step requires applying a segmented replacement formula to the AIME to compute the
primary insurance amount (PIA). The PIA is essential y a worker’s basic monthly benefit amount
payable at the Social Security full retirement age. It does not reflect any applicable adjustments
based on early or delayed retirement or other factors.20 The PIA is also used to determine the
basic monthly benefit amount for a worker’s eligible dependents or survivors, subject to certain
adjustment factors. The benefit formula is progressive. As a result, workers with higher AIMEs
receive higher benefits, but the benefits received by people with lower earnings replace a larger
share of their career-average earnings.
In this step, a worker’s AIME is divided into segments—or brackets—by “bend points.” Brackets
of a worker’s AIME are replaced at different rates, or replacement factors, the sum of which is the
PIA (Table 1). The replacement factors—90%, 32%, and 15%—are fixed in law, but the bend
point dollar amounts, or brackets, to which they apply are indexed annual y to the AWI. Under
current law, the benefit formula used to determine the PIA for a worker who attains age 62,
becomes disabled, or dies in 2021 is shown in Table 1.
Table 1. Computation of a Worker’s Primary Insurance Amount (PIA) in 2021
For a Hypothetical Worker with an Average Indexed Monthly Earnings (AIME) of $6,500
PIA for a Worker with an
Factors
Three Brackets of AIME in 2021
Illustrative AIME of $6,500
90%
first $996 of AIME, plus
$896.40
32%
AIME over $996 and through $6,002, plus
$1,601.92
15%
AIME over $6,002
$74.70
Total: Worker’s PIA (by law, rounded down to
$2,573.00
nearest 10 cents)
Source: CRS.
Notes: The bend points shown in the table apply to workers who first become eligible in 2021. Under current
law, the PIA is rounded down to the nearest dime (42 U.S.C. §415(a)(1)(A)).

18 Earnings through age 60 are indexed to average wage growth. Earnings from age 61 and later are counted at nominal
value.
19 If the retired worker has fewer than 35 years of covered earnings, years of zero earnings are entered in the
calculation. T he number of benefit computation years for disabled or deceased workers may be fewer than 35 years.
20 42 U.S.C. §415(a).
Congressional Research Service

5

link to page 10 Social Security: The Effects of Wage and Price Indexing on Benefits

Adjustments for Claiming Age
Third, an adjustment may be made based on the age at which a beneficiary chooses to begin
receiving payments.21 Retired workers may claim benefits when they turn 62 years old, which is
the EEA. Workers who claim benefits before reaching the full retirement age (FRA) would be
subject to a permanent reduction in monthly benefits, commonly referred to as an actuarial
reduction. It equals five-ninths of 1% for each month (6⅔% per year) of the first three years of
early claiming and five-twelfths of 1% for each month (5% per year) beyond 36 months. The
FRA for workers born in 1960 or later is 67. Thus, a worker in the 1960 birth cohort who claimed
benefits at age 62—the earliest eligibility age and five years before his or her FRA—would
receive 70% of his or her PIA.
For retired workers who claim benefits at the FRA and for disabled workers, the monthly benefit
equals the PIA. However, if workers delay claiming, they receive a permanently higher monthly
benefit. In other words, workers who claim benefits after reaching FRA receive delayed
retirement credits. For people born in 1943 and later, that credit is 8% for each year of delayed
claiming after the FRA, up to age 70.22 Thus, a worker in the 1960 birth cohort with an FRA of 67
who chose to claim benefits at age 70 would receive 124% of his or her PIA (3 years × 8% per
year).
Cost of Living Adjustments (COLAs)
After a beneficiary's first year of eligibility, subsequent benefits are adjusted for price growth.
The annual COLA is based on changes in the CPI-W, updated monthly by the Bureau of Labor
Statistics. For Social Security, the COLA equals the change in the CPI-W from the third quarter
of the prior year to the third quarter of the current year. The COLA announced in October
becomes effective in December of the current year and is payable in January of the following
year. Social Security payments always reflect the benefits due for the preceding month. Table 2
il ustrates the COLA computation for 2020 applied to benefits paid in January 2021.
Table 2. Computation of the Social Security Cost-of-Living Adjustment (COLA),
January 2021
Consumer Price Index for Urban Wage Earners

and Clerical Workers (CPI-W) Index Values
July 2019
250.236
August 2019
250.112
September 2019
250.251
Average for third quarter of 2019 (rounded to the nearest
one-thousandth of a point):
250.200
July 2020
252.636
August 2020
253.597

21 For more information on claiming age, see CRS Report R44670, The Social Security Retirement Age, by Zhe Li.
Also, after a beneficiary’s first year of eligibility, subsequent benefits are adjusted for price growth. For more
information, see CRS Report 94-803, Social Security: Cost-of-Living Adjustm ents, by Paul S. Davies and T amar B.
Breslauer.
22 For people born before 1943, the delayed retirement credit varies from 3.0% to 7.5%, depending on the year of birth.
See “Delayed Retirement Credit” in OCACT , “Early or Late Retirement?,” http://www.ssa.gov/OACT /quickcalc/
early_late.html#late.
Congressional Research Service

6

Social Security: The Effects of Wage and Price Indexing on Benefits

Consumer Price Index for Urban Wage Earners

and Clerical Workers (CPI-W) Index Values
September 2020
254.004
Average for third quarter of 2020 (rounded to the nearest
one-thousandth of a point):
253.412
Percentage increase or decrease from the third-quarter
average for 2019 to the third-quarter average for 2020
((253.412-250.200)/250.200)*100% =
(rounded to the nearest one-tenth of 1% for the final
1.3%
application, when positive, as required by law):
Social Security COLA (zero if the percentage change is
1.3%
negative):
Source: Department of Labor, Bureau of Labor Statistics data series for the CPI-W for 2019 and 2020,
https://www.bls.gov/cpi/data.htm.
Notes: The reference base period for the CPI-W is 1982-84 (i.e., the period when the index equaled 100). The
COLA determination fol ows procedures set forth in Section 215(i) of the Social Security Act (42 U.S.C. §415(i)).
In most years, prices increase, resulting in a positive COLA being applied to the next year’s
benefits. However, if there is no percentage increase in the CPI-W for the specific period
described above, no COLA is payable, and Social Security benefits are not adjusted. No COLA
was payable in January 2010, January 2011, or January 2016. Section 215(i) of the Social
Security Act protects Social Security benefits from being decreased during periods of negative
price growth.23
Social Security Benefits and the Average Wage Index
The current-law benefit formula is affected by both positive and negative wage growth. While in
most years, wage growth is positive, the benefit formula is affected by negative wage growth in
the economy—that is, a decreasing AWI—in two ways. First, given the lag time in computing the
AWI, a worker’s earnings are wage-indexed up to age 60, whereas earnings from later years—at
ages 60 and above—are not.24 This means that negative wage growth in the economy in the year
in which a worker turns 60 would result in a lower AIME than if wage growth increased or
remained level. For instance, using the example from an earlier section, the SSA’s AWI grew
from $32,155 in 2000 to $41,674 in 2010. Assuming, for il ustrative purposes, that the AWI
demonstrated negative 5% growth from the previous year instead, the AWI in 2010 would have
been $38,675 (i.e., a 5% decrease from the 2009 AWI of $40,711). So, under this assumption, if a
worker earned $20,000 in 2000 and turned 60 in 2010, the indexed wage for 2000 would be
$20,000 × ($38,675/$32,155), or $24,055. The lower AWI of $38,675 would be used to index
every year of the worker’s earnings. Thus, for each year of the worker’s earnings, his or her
wage-indexed earnings would be less than if AWI had experience typical positive growth.
A second way in which benefit amounts may be affected by negative wage growth is in the
calculation of the PIA. In calculating the PIA, two dollar-based bend points are used to section
the AIME into three brackets. The two bend points are adjusted annual y for average wage growth
in the economy using the AWI. The bend points are based on the year in which an individual

23 42 U.S.C. §415(i). T he average CPI-W for the third quarter of 2010 increased relative to the third quarter for 2009.
However, the average amount was still less than in the third quarter of 2008. As such, 2008 remained the base year for
the calculation, and thus there was no COLA determined in December 2010. See CPI -W values at https://www.ssa.gov/
oact/ST AT S/cpiw.html.
24 Wages earned at ages 60 and 61 are entered into the formula at their nominal value (i.e., not indexed to the AWI).
Congressional Research Service

7

link to page 20 link to page 20 link to page 20 Social Security: The Effects of Wage and Price Indexing on Benefits

reaches the earliest eligibility age, which for retired workers is age 62. However, given the lag
time in computing AWI, the value from two years prior (i.e., when the worker was 60) is used to
determine the bend points. Since the replacement factors are fixed, lower PIA bend points
necessarily result in lower PIAs.
PIAs for Hypothetical Earners Born in 1960
This section provides several il ustrative examples of the PIAs of hypothetical earners to
demonstrate how PIAs are calculated under current law and how different levels of AWI for 2020
could affect the PIAs. Following methods developed by the SSA’s Office of the Chief Actuary,
this report uses the career-averaged earnings of hypothetical earners to il ustrate PIAs. Wages for
hypothetical earners are expressed at each age as a percent of SSA’s AWI and are shown in
Figure A-1.25 In this section, examples of benefit calculations are shown for very low, low,
medium, and high lifetime hypothetical earners as wel as maximum earners.26 To demonstrate
the effect of the possible low AWI for 2020, the hypothetical workers are assumed to have been
born in 1960 and began working at age 21 in 1981. Additional y, the hypothetical workers are
assumed to have worked continual y through age 60, the age at which past earnings are wage-
indexed, or 2020. Using the scaled factors in Figure A-1, the earnings for hypothetical workers
are calculated (Table A-1). This can be accomplished for al years of the 1960 birth cohort except
for 2020 and 2021, the years in which the workers are 60 and 61, respectively. As the 2020 value
for AWI is not scheduled to be publicly known until November 1, 2021, the following il ustrative
examples assume a range of possible outcomes to demonstrate how national average wage growth
can affect benefits.27 Thus, in the following examples, the AWI for 2020 (i.e., the year in which
the 1960 birth cohort turns 60) was assumed to grow or decline by 3.5%, 0.0%, -1.0%, -2.0%, -
3.0%, -4.0%, -5.0%, or -6.0%.28 For al calculations, the AWI for 2021 was projected to grow by

25 OCACT , Scaled Factors for Hypothetical Earnings Examples Under the 2020 Trustees Report Assumptions, April
2020, https://www.ssa.gov/OACT /NOT ES/ran3/an2020-3.pdf.
26 A maximum earner is a worker who has earnings at or above the contribution and benefit base (i.e., the maximum
level of earnings subject to the Social Security payroll tax in a given year) for each ye ar starting at age 22 through the
year prior to retirement. OCACT , 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. 152, https://www.ssa.gov/
OACT /T R/2020/tr2020.pdf. T he contribution and benefit base for 2021 is $142,800 (see SSA, “ 2021 Social Security
Changes,” https://www.ssa.gov/news/press/factsheets/colafacts2021.pdf).
27 As stated in Actuarial Note 133, “T he Social Security Act requires that the average wage be promulgated in the
Federal Register by November 1 of the year following the year in which the wages were earned.” For more
information, see Michael D. Clingman and Jeffrey L. Kunkel, “ Average Wages for 1985-90 for Indexing Under the
Social Security Act,” OCACT , September 1992, https://www.ssa.gov/oact/NOT ES/note133.html.
28 T his selection of AWI growth rates encompasses a wide range. For the 1960 birth cohort, the AWI has had an
average growth of 3.84% per year. Additionally, the Social Security Board of T rustees projected the AWI to grow at
3.5% in 2020. T his projected value was under the trustees’ intermediate assumption —reflecting the trustees’ best
estimate of future experience—but did not reflect any potential effects of COVID-19 (OCACT , 2020 Annual Report, p.
118). On November 24, 2020, OCACT updated the baseline for the actuarial status of the trust funds and projected the
growth for AWI in 2020 to be -4.01% (see Stephen C. Goss, Chief Actuary, SSA, and Karen P. Glenn, Deputy Chief
Actuary, SSA, memo to Andrew Saul, Commissioner, SSA, and David Black, Deputy Commissioner, SSA, November
24, 2020, T able 2, https://www.ssa.gov/OACT /solvency/UpdatedBaseline_20201124.pdf). In a January 27, 2021, letter
to Senator Grassley, the Congressional Budget Office (CBO) projected that AWI would decline by 0.5% from 2019 to
2020. In September 2020, CBO had initially estimated a decline of 3.8% (see https://www.cbo.gov/system/files/2021-
01/56973-AWI.pdf). See also CBO, Baseline Projections: Social Security Old-Age and Survivors Insurance, February
2021, p. 3, https://www.cbo.gov/system/files/2021-02/51308-2021-02-socialsecurity.pdf.
Congressional Research Service

8

link to page 13 link to page 14 link to page 12 Social Security: The Effects of Wage and Price Indexing on Benefits

8.11%.29 Table 3 demonstrates how different growth rates in the AWI for 2020 would affect the
AIMEs of hypothetical workers born in 1960.
Table 3. Projected Average Indexed Monthly Earnings (AIMEs) for Hypothetical
Earners in the 1960 Birth Cohort by Earnings Levels and Changes in the 2020
Average Wage Index (AWI)
Percent
Very Low
Low Earner
Medium
High Earner
Maximum
Change in
Earner AIME
AIME
Earner AIME
AIME
Earner AIME
2020 AWI
3.5%
$1,164.00
$2,096.00
$4,660.00
$7,456.00
$11,501.00
0.0%
1,125.00
2,025.00
4,502.00
7,204.00
11,135.00
-1.0%
1,114.00
2,005.00
4,457.00
7,132.00
11,040.00
-2.0%
1,102.00
1,985.00
4,412.00
7,060.00
10,925.00
-3.0%
1,091.00
1,965.00
4,367.00
6,988.00
10,821.00
-4.0%
1,080.00
1,944.00
4,322.00
6,916.00
10,716.00
-5.0%
1,069.00
1,924.00
4,277.00
6,844.00
10,611.00
-6.0%
1,057.00
1,904.00
4,232.00
6,771.00
10,507.00
Source: CRS.
Notes: Wage-indexed earnings are rounded to the nearest cent, and AIMEs are rounded down to the nearest
dol ar (see 20 C.F.R. §404.211).
Under the different assumed values for AWI growth in 2020, Table 4 shows how the PIA—and
the bend points used to calculate PIA—would change for al workers and how the PIA would
change for each hypothetical earnings level. The PIA general y represents the benefit amount a
worker would collect at FRA without any additional adjustments.
In the first scenario, the AWI increases by 3.5%. For each hypothetical earnings level, the
beneficiary would receive more than the amount the same hypothetical earner born a year earlier
would have received due to a higher AIME as wel as higher bend points in the PIA. In the second
scenario, the AWI is assumed to have zero growth—that is, it would remain the same as 2019.
Under this assumption, benefit levels for the hypothetical earners are roughly equivalent to those
for the same hypothetical earners born a year earlier. In the remaining scenarios, the AWI is
assumed to decrease at rates between -1.0% and -6.0%. In each of these cases, the hypothetical
earners would experience lower AIMEs than if wage growth were to increase or remain level.
Additional y, the bend points used to calculate PIA are shown to decrease as wel , resulting in
lower PIAs. These examples demonstrate the two ways in which negative wage growth in the
economy affects benefit amounts.

29 Under the OCACT updated baseline, the updated projected growth rate for AWI in 2021 is 8.11% (see footnote 27).
T he growth rate assumed under the trustees’ intermediate assumptions in the 2020 Annual Report is 4.4%. As with
other assumptions in the report, this did not reflect any potential effects of the COVID-19 (OCACT , 2020 Annual
Report, p. 118).
Congressional Research Service

9

link to page 15 link to page 5 link to page 22 Social Security: The Effects of Wage and Price Indexing on Benefits

Table 4. Projected Primary Insurance Amounts (PIAs) for Hypothetical Earners in
the 1960 Birth Cohort by Earnings Levels and Changes in the 2020 Average Wage
Index (AWI)
Percent
First PIA
Second
Very Low
Low
Medium
High
Maximum
Change in
2020
Bend
PIA Bend
Earner
Earner
Earner
Earner
Earner
AWI
Point
Point
PIA
PIA
PIA
PIA
PIA
3.5%
$1,031
$6,212
$970.40
$1,268.70
$2,089.10
$2,772.40
$3,379.10
0.0%
996
6,002
937.60
1,225.60
2,018.30
2,678.60
3,268.20
-1.0%
986
5,942
928.30
1,213.40
1,998.10
2,651.80
3,236.50
-2.0%
976
5,882
918.70
1,201.20
1,977.90
2,625.00
3,204.70
-3.0%
966
5,822
909.40
1,189.00
1,957.70
2,598.20
3,173.10
-4.0%
956
5,762
900.00
1,176.50
1,937.50
2,571.40
3,141.40
-5.0%
946
5,702
890.70
1,164.30
1,917.30
2,544.60
3,109.60
-6.0%
936
5,642
881.10
1,152.10
1,897.10
2,517.60
3,079.00
Source: CRS.
Notes: Under each hypothetical change in the 2020 AWI, the bend points shown in the table would apply to
workers who first become eligible (i.e., turn 62) in 2022. Under current law, the PIA is rounded down to the
nearest dime (42 U.S.C. §415(a)(1)(A)), and bend points are rounded to the nearest dol ar (42 U.S.C.
§415(a)(1)(B)(i i)).
Effects of wage indexing can result in what is commonly referred to as a notch effect. This occurs
when one cohort of beneficiaries receives a different level of benefits compared to an age-
adjacent cohort.30 Many news articles have argued that relatively higher levels of unemployment
in 2020 wil lead to a decrease in AWI, causing a worker born in 1960 to receive a lower
replacement rate than a similar worker who is one year older or younger.31 The potential size of a
notch effect between the 1959 birth cohort and the 1960 birth cohort is shown in Table 5 for each
AWI assumption. For each hypothetical earnings level, if AWI were to decrease in 2020, the
hypothetical earner from the 1960 birth cohort would receive a lower PIA than a comparable
hypothetical earner from the 1959 birth cohort.32

30 One example of a notch effect was experienced by Social Security beneficiaries born between 1917 and 1921. As a
result of changes to the benefit formula in the 1972 Social Security Amendments ( P.L. 92-336), many of the
beneficiaries born in these years believe they are not receiving fair benefits as compared to beneficiaries born in earlier
years. For more information on t he notch effect, see CRS Report RS22678, Social Security: The Notch Issue, by Dawn
Nuschler.
31 A lower initial benefit would be compounded as it would result in a nominally lower COLA dollar increase eve ry
year. For more information, see CRS In Focus IF11599, Social Security Benefits and the Effect of Declines in Average
Wages and Prices, by Barry F. Huston and Paul S. Davies. See footnote 1 for examples of news articles.
32 Although an increase in the AWI could also be thought of as creating a notch effect in a narrow, technical sense, it
would generally not be considered a notch because the AWI (and thus AIMEs and PIAs) is generally expected to
increase from year to year. Over its history, the AWI has increased in all but one year (2009) at an average rate of 4.5%
(see Table A-2).
Congressional Research Service

10

link to page 15 link to page 23 Social Security: The Effects of Wage and Price Indexing on Benefits

Table 5. Comparison of Projected Primary Insurance Amounts (PIAs) for
Hypothetical Earners in the 1959 and 1960 Birth Cohorts by Earnings Levels and
Changes in the 2020 Average Wage Index (AWI)

Very Low Earner Low Earner Medium Earner High Earner Max Earner

PIAs for 1959 Birth Cohort

$936.70
$1,223.70
$2,014.10
$2,675.40
$3,262.70
Percentage
Change in 2020
Dollar Change in PIAs for 1960s Birth Cohort (2020 PIAs Less 2019 PIAs)
AWI
3.5%
$33.70
$45.00
$75.00
$97.00
$116.40
0.0%
0.90
1.90
4.20
3.20
5.50
-1.0%
-8.40
-10.30
-16.00
-23.60
-26.20
-2.0%
-18.00
-22.50
-36.20
-50.40
-58.00
-3.0%
-27.30
-34.70
-56.40
-77.20
-89.60
-4.0%
-36.70
-47.20
-76.60
-104.00
-121.30
-5.0%
-46.00
-59.40
-96.80
-130.80
-153.10
-6.0%
-55.60
-71.60
-117.00
-157.80
-184.70
Source: CRS.
Note: Under the scenario where there is no change in 2020 AWI (i.e., 0.0% growth in AWI from 2019 to 2020),
there is stil a change in PIAs between the 1959 and 1960 birth cohorts. This difference is attributable to the
different earnings periods that are indexed. The 1959 birth cohort began work in 1980, while the 1960 birth
cohort began work in 1981.
In addition, a potential decrease in AWI in 2020 would also affect dependents and survivors who
claim benefits based on the earning records of workers who were born in 1960. Benefits for
family members that are based on the benefits of a retired worker that are indexed during a year
of declining wage growth would also be affected, as those benefits are calculated as a percentage
of the worker’s PIA.33
Social Security Benefits and the COLA
Recent values of the Social Security COLA, il ustrating the years in which no COLA was
payable, are shown in Table 6. Historical values of the CPI-W and the Social Security COLA are
provided in Table A-3.
Table 6. Average Third-Quarter Consumer Price Index for Urban Wage Earners and
Clerical Workers (CPI-W) and Resulting Social Security Cost-of-Living Adjustment
(COLA), Selected Years
Average CPI-W for the Third
Year
Quarter
Resulting COLA
2007
203.596
2.3%
2008
215.495
5.8

33 For more information on benefits for workers’ family members, see CRS Report R42035, Social Security Primer, by
Barry F. Huston.
Congressional Research Service

11

Social Security: The Effects of Wage and Price Indexing on Benefits

Average CPI-W for the Third
Year
Quarter
Resulting COLA
2009
211.001
no COLA
2010
214.136
no COLA
2011
223.233
3.6
2012
226.936
1.7
2013
230.327
1.5
2014
234.242
1.7
2015
233.278
no COLA
2016
235.057
0.3
2017
239.668
2.0
2018
246.352
2.8
2019
250.200
1.6
2020
253.412
1.3
Source: CRS calculations using data from Department of Labor, Bureau of Labor Statistics.
Notes: The Social Security COLA is payable for benefits due in January of the fol owing year.
It is commonly understood that the COLA affects current beneficiaries because it is used to adjust
benefits already in payment. However, the COLA also affects individuals who are eligible for
retirement benefits but are not yet receiving them. The lack of a COLA in a given year can lead to
a birth cohort of beneficiaries having markedly different benefit levels compared with age-
adjacent birth cohorts. For example, benefits for individuals born in 1947 (i.e., those who became
eligible for retirement benefits at age 62 in 2009) were lower relative to earlier birth cohorts given
the combined effects of a higher than usual 5.8% COLA payable in January 2009 (effective
December 2008) and no COLA payable in January 2010 (effective December 2009). Recal that a
decline in the CPI-W (or negative price growth) does not result in a negative COLA, whereas a
decline in the AWI would result in lower AIMEs and PIAs (as discussed earlier).
Several Social Security program elements are indexed to wages rather than prices but may
increase only when a COLA is payable. These program elements include the contribution and
benefit base (CBB, or the maximum level of earnings subject to the Social Security payroll tax in
a given year); the retirement earnings test exempt amounts; and the substantial gainful activity
threshold for blind Social Security beneficiaries.34 In years in which no COLA is payable, the
value of these program elements is unchanged. For example, had a COLA been payable for Social
Security benefits due in January 2016, the CBB would have increased from $118,500 in 2015 to
$122,700 in 2016.35 Because no COLA was payable in January 2016, the CBB remained
unchanged. Then, when a COLA was payable for Social Security benefits due in January 2017,
the CBB for 2017 was increased based on its 2015 value to reflect two years of wage growth.

34 For more details, see CRS Report R46658, Social Security: Benefit Calculation, by Barry F. Huston; CRS Report
R41242, Social Security Retirem ent Earnings Test: How Earnings Affect Benefits, by Zhe Li; and CRS Report R44948,
Social Security Disability Insurance (SSDI) and Supplemental Security Incom e (SSI): Eligibility, Benefits, and
Financing, by William R. Morton.
35 Such a hypothetical increase in the CBB is dependent on positive wage growth. It is possible for a COLA to be
payable during a period in which the AWI is decreasing.
Congressional Research Service

12

link to page 22 Social Security: The Effects of Wage and Price Indexing on Benefits

Projections and Legislative Options
Projections for AWI and COLA
In their 2020 Annual Report, the Social Security Board of Trustees’ intermediate assumptions
projected the AWI to grow at 3.5% in 2020 and projected COLA values of 2.3% for 2020 and
2.5% for 2021.36 The trustees’ intermediate set of assumptions general y reflect their best
estimates of future experience. However, the 2020 Annual Report does not include any potential
effects resulting from COVID-19. (The assumptions reflect the trustees’ understanding at the start
of 2020.) The actual 2020 COLA (for benefits payable in 2021) was later revealed to be 1.3%.37
Since the publication of the 2020 Annual Report, projections for AWI have changed. Updated
COLA projections have not been released.
In September 2020, the Congressional Budget Office (CBO) projected that AWI would decrease
by 3.8% from 2019 to 2020.38 In a January 27, 2021, letter to Senator Grassley, CBO revised its
estimate to project that the AWI would decrease by 0.5% from 2019 to 2020.39 On November 24,
2020, SSA’s Office of the Chief Actuary updated the baseline for the actuarial status of the trust
funds and projected that the AWI would decline by 4.01% from 2019 to 2020.40
Options to Address a Decline in AWI
Policymakers have several legislative options to address the negative effect that a decline in the
AWI would have on Social Security benefit amounts. One option is to do nothing, as was done in
2009 (see Table A-2). In 2009, the AWI decreased by -1.51%, affecting the benefit amounts for
those turning 60 in that year (i.e., the 1949 birth cohort). Alternatively, Congress could administer
ad hoc benefit increases to birth cohorts that are adversely affected by changes in the AWI,
effectively “resetting” a cohort’s lifetime benefit levels. Stil another solution would be to use the
same growth rate from 2009 in years where the AWI declines. This method would effectively
make the negative effects of a decline in AWI no worse than the last cohort to experience negative
wage growth in the year they turned 60. Yet another alternative could be to use wage index data
only from the first quarter of 2020 (i.e., pre-COVID-19). These measures are essential y one-time

36 OCACT , 2020 Annual Report, p. 118. In the annual reports, the trustees present three alternative (intermediate, low-
cost, and high-cost) sets of assumptions for demographic, economic, and program -specific factors. T he intermediate set
of assumptions represents the trustees’ best estimate of likely future conditions. Relative to the set of intermediate
assumptions, the low-cost set of assumptions is significantly more optimistic, whereas the high-cost set of assumptions
is significantly more pessimistic.
37 SSA, “Social Security Announces 1.3 Percent Benefit Increase for 2021,” October 13, 2020, https://www.ssa.gov/
news/press/factsheets/colafacts2021.pdf,
38 In answers to questions for the record, CBO Director Swagel noted, “As a result, CBO projected in its September
report on the budget outlook that the AWI will be about 7 percent lower in 2020 than it estimated at the beginning of
this year. T he labor market recovery has proceeded faster than anticipated when that projection was made, suggesting
that the decline in the AWI will probably be considerably less than the 7 percent projected in September.” For more
information, see CBO, “Answers to Questions for the Record Following a Hearing Conducted by the Senate Committee
on the Budget on CBO’s Budget Projections,” December 18, 2020, https://www.cbo.gov/system/files/2020-12/56908-
senate-updated-budgetoutlook.pdf.
39 Letter from Phillip L. Swagel, Director of CBO, to Senator Grassley, January 27, 2021, https://www.cbo.gov/system/
files/2021-01/56973-AWI.pdf.
40 See Goss and Glenn, T able 2.
Congressional Research Service

13

Social Security: The Effects of Wage and Price Indexing on Benefits

solutions that, if adopted, would not prevent the issue from recurring (i.e., post-1960 birth
cohorts).
Policymakers could also opt for a permanent solution. For example, one approach would be to
use a multiyear average of wages. However, a multiyear average could be subject to consecutive
years of negative wage growth.
In the 116th Congress, policymakers introduced proposals that followed different approaches.
Senators Kaine and Cassidy proposed S. 4180, Protecting Benefits for Retirees Act, which would
have amended Section 209(k) of the Social Security Act.41 Essential y, this bil would have made
the AWI used in calculations in one year at least as high as the AWI used in calculations for the
previous year.42 For example, if the AWI for 2020 decreases, it would have been determined to be
equal to the 2019 AWI for purposes of computing Social Security benefits. The Social Security
benefits for the 1960 birth cohort would not be negatively impacted. Instead, the AWI applied in
the previous year would be applied to these workers’ Social Security benefits.
A similar approach was provided by Representative Larson in H.R. 7499 and by Senator Hirono
in S. 4986, the Social Security COVID Correction and Equity Act. Among other things, this bil
would have required the use of any prior year’s AWI if it exceeds the current year’s AWI in the
calculation of a worker’s AIME and for indexing the bend points used to calculate the PIA.
Effectively, this bil would have required the largest AWI among al previous years to be used in
benefit calculations for al years after 2019.43
As of the publication date of this report, no proposals were identified to address the AWI issue in
the 117th Congress.
Options to Address the COLA
Policymakers also have several legislative options to address the COLA. Searching Congress.gov
for legislative text that references the cost-of-living section of the Social Security Act (Section
215(i)) or the paral el section in the U.S. Code (42 U.S.C. §415(i)) retrieved over 30 bil s from the
116th Congress.44 However, some of these bil s do not include provisions to change the COLA.45
The following summary of bil s reflect options proposed to amend the COLA in the 116th
Congress.46
One option involves mandating a specific COLA for a particular year. For example,
Representative DeFazio in H.R. 8598 would have established that the COLA for 2020 be 3%.

41 42 U.S.C. §409(k).
42 T his change to the formula would be financed from the Social Security trust funds, as the bill does not authorize
transfers from the General Fund of the U.S. T reasury.
43 T his provision, and other provisions in these bills that would increase the cost of t he Social Security program, would
be financed by transfers from the general fund.
44 T he complete search results can be viewed at https://go.usa.gov/x64tr.
45 For example, P.L. 116-178 directed the Department of Veterans Affairs to increase, as of December 1, 2020, the
rates of veterans’ disability compensation, additional compensation for dependents, the clothing allowance for certain
disabled veterans, and dependency and indemnity compensation for surviving spouses and children by the same
percentage increase as the cost -of-living increase for Social Security recipients. Other bills, such as S. 580 and H.R.
1496, address the allowances for former Presidents.
46 Some bills may not have been retrieved from this search, as not all bills may have had the bill text available when the
search was conducted.
Congressional Research Service

14

Social Security: The Effects of Wage and Price Indexing on Benefits

This proposal also has a hold-harmless clause so that the Social Security trust funds would be
replenished with general funds as if this provision had not taken effect.
Another option involves issuing a compensating payment in years without a COLA.
Representative Lee’s Social Security Safety Dividend Act of 2019 (H.R. 46) would have directed
the Department of the Treasury to disburse to recipients of Social Security benefits or certain
other federal benefits a $250 payment in years where no COLA is payable.
Representative Smith in H.R. 8600 would have also addressed the possibility of no COLA by
directing that a minimum COLA of 3% be granted for 2020. Representative Smith’s bil also
recommended a permanent change in the calculation of the COLA to use the Consumer Price
Index for the Elderly (CPI-E) instead of the CPI-W as used under current law.47 Representative
Engel in H.R. 3389 (and Senator Whitehouse in S. 1923) would have similarly revised the
method used to calculate COLAs for certain Social Security benefits by ordering the Bureau of
Labor Statistics to develop a price index that tracks the spending patterns of older consumers for
purposes of calculating the COLA (i.e., the CPI-E), while also requiring that individuals over 62
must receive, at minimum, a 3% annual cost-of-living increase in any year.48
Other options, il ustrated by Representative Walberg in H.R. 7765, proposed that either the CPI-
W be applied as under current law or, if the amount would be higher, the COLA would be based
on a 10-year average of previous COLAs.
Similar to the AWI, no proposals were identified to address the COLA issue in the 117th Congress.

47 Some proposals, such as those identified here from the 116th Congress, would base the COLA on the CPI-E, which is
generally projected to increase Social Security COLAs relative to current law. Other proposals would base the COLA
on the Chained CPI for All Urban Consumers, which is generally projected to reduce Social Security COLAs relative
to current law. No bills were introduced in the 116 th Congress that would have adopted the chained CPI. More
discussion of alternat ive methods for calculating the COLA can be found in CRS Report R43363, Alternative Inflation
Measures for the Social Security Cost-of-Living Adjustm ent (COLA), by Julie M. Whittaker.
48 T he option to calculate the COLA using the CPI-E instead of the CPI-W appears in several legislative proposals from
the 116th Congress, including H.R. 4121, H.R. 2654, H.R. 2302, H.R. 1170, H.R. 860, S. 478, S. 269, and S. 112.
Congressional Research Service

15

link to page 22

Social Security: The Effects of Wage and Price Indexing on Benefits

Appendix. Supporting Information
Figure A-1. Scaled Factors by Hypothetical Earnings Level and Age
Percent of Average Wage Index (AWI)

Source: Social Security Administration, Office of the Chief Actuary, Scaled Factors for Hypothetical Earnings
Examples Under the 2020 Trustees Report Assumptions, Table 6, April 2020, https://www.ssa.gov/OACT/NOTES/
ran3/an2020-3.pdf.
Notes: There is no scaled factor for a maximum earner. Maximum earners are assumed to have earned at or
above the contribution base in each respective year (see Table A-2).
Table A-1. Hypothetical Wages for 1960 Birth Cohort by Earnings Level
Very Low
Low
Medium
High
Maximum
Year
Age
Earner
Earner
Earner
Earner
Earner
1981
21
$1,019.21
$1,831.82
$4,063.06
$6,514.68
$29,700.00
1982
22
1,307.82
2,354.08
5,216.75
8,355.52
32,400.00
1983
23
1,722.03
3,093.57
6,888.14
11,017.97
35,700.00
1984
24
2,162.10
3,888.55
8,648.40
13,843.89
37,800.00
1985
25
2,540.20
4,575.72
10,143.97
16,250.54
39,600.00
1986
26
2,875.42
5,179.22
11,484.37
18,378.45
42,000.00
1987
27
3,316.77
5,951.76
13,248.66
21,190.49
43,800.00
1988
28
3,712.14
6,689.58
14,848.54
23,761.54
45,000.00
1989
29
4,080.21
7,336.34
16,320.83
26,089.22
48,000.00
1990
30
4,457.93
8,032.69
17,852.76
28,577.02
51,300.00
1991
31
4,820.36
8,659.21
19,237.83
30,797.98
53,400.00
1992
32
5,229.28
9,403.52
20,894.17
33,439.84
55,500.00
Congressional Research Service

16

link to page 21 link to page 21 link to page 22 link to page 12 link to page 14 Social Security: The Effects of Wage and Price Indexing on Benefits

Very Low
Low
Medium
High
Maximum
Year
Age
Earner
Earner
Earner
Earner
Earner
1993
33
5,413.04
9,738.85
21,652.18
34,652.74
57,600.00
1994
34
5,700.85
10,237.77
22,755.88
36,414.16
60,600.00
1995
35
6,052.89
10,870.49
24,186.84
38,689.06
61,200.00
1996
36
6,452.56
11,609.43
25,810.24
41,306.76
62,700.00
1997
37
6,938.78
12,478.83
27,755.11
44,402.69
65,400.00
1998
38
7,388.53
13,333.99
29,611.84
47,361.62
68,400.00
1999
39
7,891.69
14,229.42
31,627.69
50,610.40
72,600.00
2000
40
8,424.56
15,177.08
33,698.25
53,923.63
76,200.00
2001
41
8,724.31
15,670.83
34,864.31
55,769.73
80,400.00
2002
42
8,878.31
15,994.26
35,513.23
56,827.82
84,900.00
2003
43
9,163.47
16,487.44
36,653.89
58,659.84
87,000.00
2004
44
9,660.76
17,396.49
38,643.03
61,814.59
87,900.00
2005
45
10,051.20
18,106.94
40,278.70
64,408.97
90,000.00
2006
46
10,590.49
19,016.49
42,284.64
67,678.62
94,200.00
2007
47
11,071.10
19,919.90
44,284.41
70,830.81
97,500.00
2008
48
11,325.78
20,419.48
45,344.46
72,542.87
102,000.00
2009
49
11,154.98
20,111.54
44,660.64
71,489.59
106,800.00
2010
50
11,418.63
20,586.87
45,757.87
73,179.25
106,800.00
2011
51
11,776.41
21,188.95
47,105.65
75,343.26
106,800.00
2012
52
12,055.49
21,717.62
48,310.62
77,252.67
110,100.00
2013
53
12,119.80
21,860.53
48,568.99
77,701.40
113,700.00
2014
54
12,457.05
22,404.09
49,828.19
79,715.81
117,000.00
2015
55
12,698.04
22,894.95
50,888.35
81,430.98
118,500.00
2016
56
12,549.67
22,618.60
50,247.34
80,405.47
118,500.00
2017
57
12,630.79
22,795.82
50,623.82
80,967.92
127,200.00
2018
58
12,723.58
22,892.01
50,842.16
81,399.59
128,400.00
2019
59
12,713.50
22,938.40
50,962.19
81,528.68
132,900.00
2020a
60
-
-
-
-
-
2021a
61
-
-
-
-
-
Source: CRS.
Notes: Very low, low, medium, and high earners are assumed to work at specified ages with earnings equivalent
to the respective scaled earners as shown in Social Security Administration, Office of the Chief Actuary, Scaled
Factors for Hypothetical Earnings Examples Under the 2020 Trustees Report Assumptions, Table 6, April 2020,
https://www.ssa.gov/OACT/NOTES/ran3/an2020-3.pdf. Al dol ar values are shown in nominal terms (i.e., not
indexed). Maximum earners are assumed to have earned at or above the contribution base in each respective
year (see Table A-2).
a. The average wage index (AWI) of 2020 and al subsequent years is unknown until it is calculated by SSA on
November 1, 2021 (see footnote 27). In Table 4, a range of different AWI levels were assumed to calculate
each hypothetical earner’s earnings and wage-indexed earnings.
Congressional Research Service

17

Social Security: The Effects of Wage and Price Indexing on Benefits

Table A-2. Average Wage Index (AWI), Contribution and Benefit Base, and Primary
Insurance Amount (PIA) Bend Points for Select Years
Annual
Contribution and
Benefit Base
Second PIA Bend
Year
AWI
Change
(Taxable
First PIA Bend Point
Point
(AWI)
Maximum)
1981
$13,773.1
10.07%
$29,700
$211
$1,274
1982
14,531.34
5.51%
32,400
230
1,388
1983
15,239.24
4.87%
35,700
254
1,528
1984
16,135.07
5.88%
37,800
267
1,612
1985
16,822.51
4.26%
39,600
280
1,691
1986
17,321.82
2.97%
42,000
297
1,790
1987
18,426.51
6.38%
43,800
310
1,866
1988
19,334.04
4.93%
45,000
319
1,922
1989
20,099.55
3.96%
48,000
339
2,044
1990
21,027.98
4.62%
51,300
356
2,145
1991
21,811.60
3.73%
53,400
370
2,230
1992
22,935.42
5.15%
55,500
387
2,333
1993
23,132.67
0.86%
57,600
401
2,420
1994
23,753.53
2.68%
60,600
422
2,545
1995
24,705.66
4.01%
61,200
426
2,567
1996
25,913.90
4.89%
62,700
437
2,635
1997
27,426.00
5.84%
65,400
455
2,741
1998
28,861.44
5.23%
68,400
477
2,875
1999
30,469.84
5.57%
72,600
505
3,043
2000
32,154.82
5.53%
76,200
531
3,202
2001
32,921.92
2.39%
80,400
561
3,381
2002
33,252.09
1.00%
84,900
592
3,567
2003
34,064.95
2.44%
87,000
606
3,653
2004
35,648.55
4.65%
87,900
612
3,689
2005
36,952.94
3.66%
90,000
627
3,779
2006
38,651.41
4.60%
94,200
656
3,955
2007
40,405.48
4.54%
97,500
680
4,100
2008
41,334.97
2.30%
102,000
711
4,288
2009
40,711.61
-1.51%
106,800
744
4,483
2010
41,673.83
2.36%
106,800
761
4,586
2011
42,979.61
3.13%
106,800
749
4,517
2012
44,321.67
3.12%
110,100
767
4,624
2013
44,888.16
1.28%
113,700
791
4,768
Congressional Research Service

18

link to page 11 link to page 11 Social Security: The Effects of Wage and Price Indexing on Benefits

Contribution and
Annual
Benefit Base
Second PIA Bend
Year
AWI
Change
First PIA Bend Point
(AWI)
(Taxable
Point
Maximum)
2014
46,481.52
3.55%
117,000
816
4,917
2015
48,098.63
3.48%
118,500
826
4,980
2016
48,642.15
1.13%
118,500
856
5,157
2017
50,321.89
3.45%
127,200
885
5,336
2018
52,145.80
3.62%
128,400
895
5,397
2019
54,099.99
3.75%
132,900
926
5,583
2020
-
-
137,700
960
5,785
2021
-
-
142,800
996
6,002
Source: CRS.
Notes: Dashes indicate data not available. As discussed in “Social Security Benefits and the Average Wage
Index,” given the lag time in computing AWI, bend points are computed using the AWI value from the prior two
years.
Table A-3. Consumer Price Index for Urban Wage Earners and Clerical Workers
(CPI-W) and Social Security Cost-of-Living Adjustment (COLA) for Selected Years
CPI-W, Annual
Annual Percentage
Year
Average Value
Change in CPI-W
Social Security COLA
1985
106.9
3.5%
3.1%
1986
108.6
1.6
1.3
1987
112.5
3.6
4.2
1988
117.0
4.0
4.0
1989
122.6
4.8
4.7
1990
129.0
5.3
5.4
1991
134.3
4.0
3.7
1992
138.2
2.9
3.0
1993
142.1
2.8
2.6
1994
145.6
2.5
2.8
1995
149.8
2.8
2.6
1996
154.1
2.9
2.9
1997
157.6
2.2
2.1
1998
159.7
1.3
1.3
1999
163.2
2.2
2.5
2000
168.9
3.5
3.5
2001
173.5
2.7
2.6
2002
175.9
1.4
1.4
2003
179.8
2.2
2.1
Congressional Research Service

19

Social Security: The Effects of Wage and Price Indexing on Benefits

CPI-W, Annual
Annual Percentage
Year
Average Value
Change in CPI-W
Social Security COLA
2004
184.5
2.6
2.7
2005
191.0
3.5
4.1
2006
197.1
3.2
3.3
2007
202.8
2.9
2.3
2008
211.1
4.1
5.8
2009
209.6
-0.7
0.0
2010
214.0
2.1
0.0
2011
221.6
3.6
3.6
2012
226.2
2.1
1.7
2013
229.3
1.4
1.5
2014
232.8
1.5
1.7
2015
231.8
-0.4
0.0
2016
234.1
1.0
0.3
2017
239.1
2.1
2.8
2018
245.1
2.5
2.8
2019
249.2
1.7
1.6
2020
-
-
1.3
Source: Department of Labor, Bureau of Labor Statistics (BLS), Consumer Price Index (CPI) Databases, Urban
Wage Earners and Clerical Workers (Current Series), https://data.bls.gov/PDQWeb/cw; and Social Security
Administration, Office of the Chief Actuary, “Cost-of-Living Adjustments,” https://www.ssa.gov/OACT/COLA/
colaseries.html.
Notes: The average annual CPI-W value is the arithmetical average of that year’s monthly CPI-W values, U.S.
city average, al items, not seasonal y adjusted, as reported by BLS (1982-1984 = 100). The Social Security COLA
is calculated by measuring the percentage increase in the CPI-W between the third quarter of the current
calendar year and the CPI-W of the third quarter of the previous year. The percentage change is rounded to the
nearest one-tenth of 1% (0.1%) and cannot be less than zero. Dashes indicate that data are not available.
Table A-4. Selected Items Used in the Social Security Program and Their
Relationship with the Consumer Price Index for Urban Wage Earners and Clerical
Workers (CPI-W) and/or Average Wage Index (AWI)
Possible
Possible
Change
Change
Citation for
Indexing
When CPI-
When AWI
Items
Indexation
Measure
W Declines
Declines
Social Security Cost-of-Living
42 U.S.C. §415(i)
CPI-W
No
—
Adjustment
Contribution and Benefit Base
42 U.S.C.
AWI
No
No
§430(b)
Quarter of Coverage
42 U.S.C.
AWI
No
No
§413(d)
Average Indexed Monthly Earnings
42 U.S.C.
AWI
—
Yes
§415(b)(3)(A)
Congressional Research Service

20

Social Security: The Effects of Wage and Price Indexing on Benefits

Bend Points for Primary Insurance
42 U.S.C.
AWI
—
Yes
Amount
§415(a)(1)(B)
Bend Points for Family Maximum
42 U.S.C.
AWI
—
Yes
Benefits
§403(a)(2)
Source: CRS review of statutory and regulatory provisions, as wel as sub-regulatory guidance, pertaining to
each item.
Notes: Dashes indicate not relevant.

Author Information

Barry F. Huston, Coordinator
Tamar B. Breslauer
Analyst in Social Policy
Senior Research Librarian

Paul S. Davies

Specialist in Income Security

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
R46819 · VERSION 1 · NEW
21