When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

Agency decision

Ask Donna

What actually matters in this document.

Text

Peter J. Brady

Steven Bass

Investment Company Institute*

1401 H Street N.W.

Washington, DC 20005

pbrady@ici.org

Draft: May 10, 2023

Abstract

This study uses administrative tax data to build a unique panel dataset that follows the 1945

birth-year cohort from 2000 (when aged 55) through 2017 (when aged 72). These data allow us

to observe changes in the amount and composition of individuals’ income from before they are

eligible to claim Social Security retirement benefits until after they are eligible for maximum

Social Security benefits and generally must begin taking distributions from their IRAs and DC

plans. We examine changes with age in the inflation-adjusted amount of total and spendable

income, as well as changes in the composition of income as labor income declines and Social

Security benefits and retirement income (IRA distributions and income from pensions and

annuities) increase. We find that the typical individual maintained more than 90 percent of their

age 55-59 spendable income—that is, the income available to spend after paying taxes and

saving for retirement—through age 72. Those with lower age 55-59 income typically had higher

spendable-income replacement rates. We find much higher incidence of retirement income than

is typically reported in household survey data. By age 72, 75 percent received retirement income

directly or through a spouse, with incidence higher than 80 percent for the top 60 percent of the

age 55-59 income distribution. The median share of income from Social Security was 47 percent

at age 72. Reflecting the design of the US Social Security system, those with lower age 55-59

income tended to rely more on Social Security benefits in retirement while those with higher

age 55-59 income tended to rely more on retirement plan distributions.

* This research was conducted as part of the Statistics of Income Joint Research Program. Views presented

are those of the authors and do not necessarily represent the views of the Internal Revenue Service or the

views of the Investment Company Institute or its members. We thank Kevin Pierce for his assistance with

this project.

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

1. Introduction

This study uses administrative tax data to answer two questions about retirement in

America. First, how does spendable income change as workers transition into retirement?

Second, how does the composition of income change during this transition as workers move

from relying primarily on earnings from labor to relying primarily on Social Security benefits,

retirement plan distributions, and other income? In addition, we investigate how the answers to

these questions differ based on income.

The motivation for trying to answer these questions is that proposals to overhaul the US

Social Security system and the taxation of retirement plans are ubiquitous. We think such

proposals to change the current system should be based on an accurate assessment of how the

current system is performing.

Many proposals are motivated by a belief that the current system provides inadequate

resources to retirees and that the voluntary component of our retirement system—employersponsored defined benefit (DB) and defined contribution (DC) plans and individual retirement

arrangements (IRAs)—is primarily responsible for that failing.1 The belief that the US retirement

system produces inadequate retirement resources is based largely on research projecting that

most Americans will not accumulate enough resources for retirement (see, for example,

Munnell, Chen, and Siliciano 2021 and VanDerhei 2019). The belief that employer plans and

IRAs are not doing enough to help workers prepare for retirement is based largely on analyses

of household survey data that claim too few workers have access to retirement plans (see, for

example, Munnell and Bleckman 2014) and too few retirees receive income from employer

plans (see, for example, Social Security Administration 2016).

There is other research, however, that challenges these beliefs. Both Scholz, Sheshadri, and

Khitatrakun (2006) and Hurd and Rohwedder (2015) conclude that most—though not all—

American households appear to be adequately preparing for retirement. Further, studies which

See, for example, a series of editorials in Bloomberg News in early 2022 (Bloomberg News. America’s Retirement Crisis

Is a Financial Crisis Too, March 28, 2022; Saving for Retirement Is Harder Than It Needs to Be, April 4, 2022;

Piecemeal Reform Won’t Solve the U.S. Retirement Crisis, April 11, 2022; and How to Fix the Broken U.S. RetirementSavings System, April 18, 2022).

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

compare tax data to household survey data (Brady and Pierce 2012; Bee and Mitchell 2017; and

Brady and Bass 2021) show that household surveys undercount (non-Social Security) retirement

income—that is, distributions from employer-sponsored DB and DC retirement plans, IRAs,

and annuities. These studies indicate that many more workers accumulate resources from

employer plans and IRAs during their working career and many more retirees receive income

from retirement plans than is indicated in household surveys.

We take a novel approach to assess the US retirement system using unique panel data that

we developed from administrative tax data. Rather than analyzing cross-sectional data on

individuals approaching retirement and assessing the adequacy of the resources they have

accumulated to date, we follow a cohort of individuals through the transition into retirement

and analyze changes in the amount and composition of their income. Specifically, we begin

with a representative sample of the US population aged 55 at year-end 2000 and follow all who

survived through year-end 2017 (when they would have been aged 72).

The administrative tax data allow us to track the same individuals for 18 years and analyze

changes in income on an annual basis for a large representative sample of the US population—

something not possible with existing household survey data. Despite focusing solely on

individuals aged 55 at year-end 2000, our panel sample includes well over 100,000 individuals.

In addition, unlike household survey data, attrition from the sample is not an issue. Individuals

remain in our panel regardless of whether they file a tax return, change filing status, or report

any income to the IRS.

Administrative tax data also measures income more accurately than household surveys.

There is an extensive literature documenting the underreporting of income in household

surveys.2 Although this study is not focused on directly comparing the two data sources,

consistent with previous studies that are, we find considerably more income from IRAs, DB and

DC retirement plans, and annuities than is typically reported in household survey data.3

See, for example, Rector, Johnson, and Youssef (1999) and Meyer, Mok, and Sullivan (2009). For a wider discussion

of the literature comparing household surveys to other data sources, see Brady and Bass (2021).

For studies directly comparing tax and household survey data on retiree income, see Schieber (1995), Brady and

Pierce (2012); Bee and Mitchell (2017), and Brady and Bass (2021).

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

Our study builds on the method developed by Brady et al. (2017). That study followed

individuals aged 55 to 61 at year-end 1999 through 2010 and examined changes in the amount

and composition of income from one year prior to claiming Social Security through three years

after claiming Social Security. Whereas that study included only working individuals who filed

a tax return in 1999 and who were not yet receiving Social Security benefits, this study includes

all individuals identified in the tax data, regardless of whether they worked or filed a return in

2000. And whereas that study focused on those who claimed Social Security benefits by 2007

and only examined changes in income over a five-year period, we examine income changes for

the full population over an 18 year period. This allows us to follow individuals from age 55,

before they are eligible to claim Social Security retirement benefits and when retirement plan

distributions are generally subject to an early withdrawal penalty, until age 72, after the ages at

which delaying Social Security claiming no longer increases monthly benefits and required

minimum distributions (RMDs) from IRAs and employer plans begin. We know of no other

studies that have taken this approach to the topic.

We find that the typical individual maintained more than 90 percent of their age 55-59

spendable income through age 72, with 25 percent having age 72 spendable income replacement

rates of 125 percent or more and 25 percent having replacement rates of 67 percent or less.

Spendable income is the income available to spend after paying taxes and saving for retirement.

Non-Social-Security retirement income—that is, distributions from IRAs, employersponsored DB and DC plans, and annuities—was much more common than reported in

household surveys. At age 72, 75 percent of the panel received retirement income either directly

or through a spouse, with a median amount of $15,300 per individual.

Most retirees relied on both Social Security and retirement plan distributions. Considering

both income received directly or through a spouse, 97 percent of the panel received Social

Security income and 75 percent received retirement income at age 72, with 74 percent receiving

both. The median share of income from Social Security was 47 percent at age 72, with 25 percent

of the panel getting 75 percent or more of their income from Social Security, and 25 percent

getting 29 percent or less.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

When individuals are ranked by their age 55 through age 59 income, we observe

substantial differences by income. Higher-income individuals were more likely to delay

claiming Social Security and were more likely to continue working at older ages. Social Security

benefits represented a higher share of total income at age 72 for those with lower age 55-59

income while retirement plan distributions represented a higher share of total income at age 72

for those with moderate to moderately high age 55-59 income.

Despite the fact that those with lower age 55-59 income typically claimed Social Security

earlier, were less likely to work at age 72, and got a higher share of their income from Social

Security at age 72, they typically replaced a higher share of their age 55-59 spendable income. At

age 72, the median spendable income replacement rate was greater than 100 percent for the

bottom 25 percent, between 90 and 95 percent near the middle, and less than 80 percent for only

the top 10 percent of the age 55-59 income distribution.

The paper is organized as follows. Section 2 describes the data we use in our analysis.

Section 3 examines changes in income by age and spendable income replacement rates. Section

4 analyzes changes in the composition of income with age. Section 5 concludes the analysis.

2. Description of Data

This study uses US Internal Revenue Service (IRS) administrative tax data from tax years

2000 through 2017. These data include information from both federal individual income tax

returns filed by taxpayers and information returns issued by third parties and sent to both

taxpayers and the IRS. Information returns are used to report income (such as Form W-2, which

reports wages), expenses (such as Form 1098, which reports mortgage interest expense), and

other tax-relevant information (such as Form 1099-Q, which reports distributions from qualified

education savings plans).4 We also incorporate Social Security Administration (SSA) data on

gender, date of birth, and date of death (if applicable).

For a full listing of information returns used in this study and their description, see Appendix Table A.1.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

Our overall population of interest is US citizens and resident aliens5 who—provided their

gross income exceeded the filing thresholds—would have been required to file a 2000 Form

1040 (inclusive of Form 1040, Form 1040A, and Form 1040EZ), excluding residents of US

territories.6 This includes US citizens and resident aliens living in a state (inclusive of the 50

states and the District of Columbia), living outside the US, or living overseas as a member of the

US armed forces. We exclude residents of US territories because bona fide residents of US

territories generally do not file a Form 1040 with the IRS.7

To examine changes in annual income as individuals transition into retirement, we create a

unique panel dataset that follows individuals from an age when most income is from labor (age

55) to an age when Social Security benefits and retirement plan distributions are the primary

sources of income (age 72). The data allow us to observe individuals for seven years before they

turn age 62 (when they are first eligible to claim Social Security retirement benefits). The data

also allow us to observe the same individuals after they turn age 70 (when further delays in

claiming Social Security benefits no longer increase monthly benefits), and after they turn

age 70-½ (when they are generally required to begin taking minimum distributions from their

employer plans and IRAs).

2.1 Construction of the Balanced Panel

We derive our representative sample by sampling individuals rather than tax returns or

households. We use the individual as our unit of analysis because the focus of our research is

measuring income changes during the transition into retirement, and the composition of an

individual’s tax return or household may change from year to year. We include multiple

Resident aliens include individuals with a green card or who had a “substantial presence” in the US—inclusive of

the 50 US states and the District of Columbia. For more information on US income tax treatment of both resident and

nonresident aliens, see Internal Revenue Service (2018a).

For more information on filing requirements, see Internal Revenue Service (2017a) and Internal Revenue Service

(2017b).

US citizens and resident aliens who are bona fide residents of Guam, the US Virgin Islands, and the Northern

Mariana Islands are not required to file a Form 1040 with the IRS. US citizens and resident aliens who are bona fide

residents of American Samoa and Puerto Rico are only required to file a Form 1040 if they received income from a

source outside of the territory. For the definition of a bona fide resident and more information on filing requirements

for individuals with income from US possessions, see Internal Revenue Service (2018c).

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

individuals from a single tax return in the sample only if they independently meet our sampling

criterion.

To build our panel dataset, we first create a representative sample of the US population in

2000 by combining three separate subsamples: one for filers (primary or secondary taxpayers

listed on a return), inclusive of both non-dependent and dependent filers; one for dependent

nonfilers; and one for non-dependent nonfilers.8 Tax returns allow us to identify filers and the

dependents they claim. Information returns allow us to identify non-dependent nonfilers.9

We then select from the tax-year 2000 sample all individuals who were born in 1945 (who

would have been aged 55 by the end of the year). This group consists of roughly 136,000

observations representing 2.7 million individuals. We follow these individuals through the

earlier of 2017 or the year in which they die.10

Finally, we include in the balanced panel used for our analysis only those alive at the end of

2017, when they would be 72 years old. Of the 2.7 million individuals born in 1945 and alive at

the beginning of 2000, 2.2 million (or 81 percent) were alive at the end of 2017 (Figure 1).

As illustrated by their share of our balanced panel, the inclusion of nonfilers in the sample is

critical for measuring the income of the elderly (Figure 2). The nonfiler share of the 1945 birthyear cohort—inclusive of both dependent nonfilers and non-dependent nonfilers—generally

increases with age, from 6 percent of individuals aged 55 in 2000 to 24 percent of individuals

aged 72 in 2017. The lone exception is 2007 when, because individuals had to file a 2007 tax

return to receive a 2008 stimulus payment, only 5 percent of the sample did not file a return.

2.2 Assigning Marital Status to Nonfilers

In every year, all members of the panel are categorized as either joint or non-joint. For

individuals who file a return in a given year, married individuals filing a joint return are

Brady and Bass (2023) describes the method used to build a representative sample of the population in detail.

For a list of the information returns used to identify nonfilers, see Table A.1 in the appendix. One group of

individuals that may not be captured by the tax data is individuals solely dependent on public assistance. This is

because benefit payments from such programs as Temporary Assistance to Needy Families (TANF), Supplemental

Security Income (SSI), and Veteran Affairs (VA) are not reported to the IRS—neither on tax returns nor on

information returns.

Until the year of their death, individuals remain in the data even in years where no tax information is available for

them. In these years, they would be counted as an observation but would have no income.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

categorized as joint and all other filers (single, head of household, qualified widow[er], and

married filing separately) are categorized as non-joint. For individuals who do not file in a

given year, we assign marital/filing status based on returns filed in other years. We first look to

the most recent prior year a return was filed. If the individual has not filed a return in any past

year, we look to future years. If a nonfiler’s most recent prior or closest future return was a joint

return and the individual’s spouse is alive in the nonfiling year, we classify them as joint. We

classify nonfilers as non-joint if: (1) their most recent prior return was a joint return and the

individual’s spouse is dead in the nonfiling year; (2) the most recent prior or closest future

return was a non-joint return; or (3) they have never filed a return.

Over two-thirds (69 percent) of individuals are joint at the start of the panel but the share

declines with age, to 63 percent at age 72 (Figure 3).

2.3 Income and Tax Measures

We derive our total income measure from tax data, but it differs from the tax code’s

definition of income because we are primarily focused on measuring income available to spend

after paying taxes and saving for retirement. As such, we include some types of income

excluded from taxable income—such as tax-exempt interest and the nontaxable portion of Social

Security benefits. To the extent possible, we exclude from income all retirement plan

contributions and include in income all retirement plan distributions—regardless of whether

contributions were from an employer or an employee, and regardless of their tax treatment.

This means we exclude from income not only tax-deferred employee contributions to employer

plans and IRAs, but also Roth contributions and non-Roth after-tax contributions. It also means

we include in income not only taxable non-Roth distributions, but also Roth distributions and

the portion of non-Roth distributions that represents basis. Finally, because our spendable

income measure does not account for state income taxes, we exclude from income taxable state

income tax refunds.

The income and taxes of filers are primarily derived from tax returns while the income and

taxes of nonfilers are derived solely from information returns. For a nonfiler with non-joint

marital status, we use the information returns of the individual. For a nonfiler with joint marital

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

status, we use the information returns of both the individual and the spouse. Appendix Table

A.2 describes in detail how we calculate our income and tax measures for both filers and

nonfilers.

Our measure of total income is the sum of six types of income: labor (wage and salary, selfemployment earnings, unemployment compensation), Social Security (disability benefits and

retirement benefits), retirement (IRA distributions and income from pensions and annuities),

investment (taxable interest, tax-exempt interest, dividends, gains/losses),

business/farm/rents/royalties (business and farm income in excess of self-employment

earnings; income from rents, royalties, partnerships, S-corps, and trusts), and other (net

alimony [alimony received less alimony paid] and other income).11

Our measure of spendable income is total income less total federal taxes. Total federal

taxes are comprised of federal income taxes and the employee share of payroll taxes. For filers,

federal income taxes are taken from Form 1040 and payroll taxes are based on a combination of

amounts withheld on information returns and Form 1040. For nonfilers, both federal income

and payroll taxes are the amount withheld on information returns. We do not attempt to

estimate state and local taxes. We also do not impute federal excise or corporate income tax

burdens.

The primary measure of income we use to analyze the incidence and amount of income, in

total and by type, is per capita income, which allocates the joint income of married couples

equally to each spouse. For a primary or secondary taxpayer on a joint return, per capita income

is the income derived from the tax return divided by two. For a nonfiler with a joint marital

status, per capita income is the sum of the individual’s income and the spouse’s income derived

from information returns divided by two. For a primary taxpayer on a non-joint tax return, per

capita income is simply the income derived from the tax return. Similarly, for nonfilers with a

non-joint marital status, per capita income is simply the income derived from the individual’s

information returns.

Our income measure does not include public assistance income because benefit payments from such programs as

Temporary Assistance to Needy Families (TANF), Supplemental Security Income (SSI), and Veteran Affairs (VA) are

not reported to the IRS—neither on tax returns nor on information returns.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

In addition to per capita income, we also report own income for labor, Social Security, and

retirement income. For individuals with joint marital status, we use information returns to

allocate income to the spouse who received the income. For individuals with a non-joint filing

status, there is no difference between own income and per capita income.

Importantly, neither per capita income nor own income adjusts for family or household

size. All income reported on a tax return is allocated to filers—the primary taxpayer in the case

of non-joint returns and the primary and secondary taxpayers in the case of joint returns. The

number of dependents claimed on a tax return has no impact on either measure, as no filer

income is allocated to dependents. Dependents have income only if they file their own return or

have income reported on information returns.

2.4 Tax Rates

For individuals who have income in a given year, we calculate effective tax rates. We

categorize individuals as having income if they have nonzero per capita total income in any of

our six broad income categories (labor, Social Security, retirement, investment,

business/farm/rents/royalties, and other) or in any of the components of investment income

(taxable interest, tax-exempt interest, dividends, and gains/losses). An individual’s average

effective tax rate is calculated as taxes paid divided by total income, with both taxes and income

measured on a per capita basis.12 Effective tax rates are reported for federal income taxes,

payroll taxes, and total federal taxes.

2.5 Medians and Percentile Measures

The medians presented in this study are approximate, as true medians could represent

disclosure of an individual’s tax data. To calculate approximate medians, we average the 48th,

49th, 50th, 51st, and 52nd percentile values and then round that average (to the nearest dollar

for amounts less than $100, the nearest $10 for amounts from $100 to less than $10,000, the

nearest $100 for amounts of $10,000 or more, and two decimal places for percentages). We then

For individuals with positive taxes and negative total income, we set the effective tax rate to 100 percent. For

individuals with negative total income and either zero or negative taxes, we set the effective tax rate to zero.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

only report these approximate medians for groups with 100 or more observations. We use the

same method to report other percentile measures.13

3. Changes in Total and Spendable Income with Age

In this section, we examine changes in income from the 2000 tax year, when individuals in

our data were aged 55 at year end, through the 2017 tax year, when individuals in our data

were aged 72 at year end. In addition to individuals getting older, their income may also be

affected by the business cycle. Over this period, recessions occurred from March through

November of 2001 (the 2001 recession) and from December 2007 through June 2009 (the great

recession), as indicated by the gray shading in the figures (see, for example, Figure 4).14

Although we report changes in total income, we are primarily focused on changes in

spendable income—that is, income available to spend after paying taxes and saving for

retirement. Economic theory predicts that—rather than maintaining income in retirement—

individuals wish to maintain consumption in retirement.15 We cannot measure consumption or

spending with the tax data, but we can measure income available to spend.

Because we focus on spendable income, our replacement rate measure differs from

traditional replacement rate measures. Traditional replacement rates measure total income—

before accounting for taxes or savings—in the first year of retirement relative to total income in

the year, or years, immediately before they retired.16 The replacement rate measures used in this

Others who wish to use this measure for their own research may cite this article for authority or simply refer to the

measure as the Brady-Bass Adjusted Median (BBAM).

The National Bureau of Economic Research determines the official dating for US business cycle expansions and

contractions. See https://www.nber.org/research/data/us-business-cycle-expansions-and-contractions.

Optimization over the life cycle generally requires that the marginal utility of consumption be equal in each time

period. If certain other conditions are met, this would also imply that an individual would prefer to smooth

consumption over time. See Engen, Gale, and Uccello (2005) and Scholz, Seshadri, and Khitatrakun (2006) for a

more formal description of life-cycle models and for a discussion of retirement savings adequacy. To the extent that

some spending—such as a portion of spending on clothing and travel—are properly characterized as a cost of

working rather than consumption, and to the extent retirees substitute home production for market production —for

example, preparing lunch at home versus purchasing lunch at a fast food establishment or cafeteria, consumption

can be maintained even if spending declines. See Hurst (2008) for a discussion of consumption changes in

retirement. In addition, retired people may maintain their marginal utility by diminishing their consumption and

increasing their time devoted to leisure. So, a constant marginal utility across the life cycle can be maintained even

if consumption declines.

See Brady (2010) for a discussion of the origins of the traditional replacement rate measure and its shortcomings.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

study are based on those developed in Brady (2010) and measure replacement at older ages of

the inflation-adjusted spendable income—after accounting for federal taxes and retirement plan

contributions—that individuals had in their late 50s.

There are two reasons why it is important to focus on changes in spendable income rather

than changes in total income. First, spendable income generally falls by less than total income in

retirement because taxes fall more than proportionately with income.17 Second, how much taxes

fall in retirement will vary depending on an individual’s pre-retirement income, so rule-ofthumb replacement rates of total income will not be appropriate across the income distribution.

Specifically, we find that those in the middle of the age 55-59 income distribution typically

experienced the largest declines in average effective tax rates in retirement.

Through age 72, we find that the typical individual maintained more than 90 percent of

their age 55-59 spendable income, with those with lower age 55-59 income typically having

higher replacement rates. At age 72, the median replacement rate was greater than 100 percent

for the bottom 25 percent of the age 55-59 income distribution, between 90 and 95 percent near

the middle, and less than 80 percent for the top 10 percent.

All dollar amounts in the paper are adjusted for inflation and reported in constant 2017

dollars. For ease of exposition, dollar amounts reported in the text are generally rounded to two

significant digits. The values plotted in the figures are available in an accompanying Excel

spreadsheet.

3.1 Incidence and Amounts of Income

In every year of the panel, most individuals had income (Figure 4).18 There was a slight

decline in the share of the population with income from 99.1 percent at age 55 (2000) to

97.7 percent at age 61 (2006), with most of the decline occurring by age 57 (during and

Our measure of total income already excludes (pre-tax and after-tax) retirement plan contributions. For studies that

measure income prior to both taxes and savings, another reason less income is needed after retirement is that

individuals no longer need to save for retirement.

Individuals are categorized as having income if they have nonzero per capita income in any of our six broad

income categories (labor, Social Security, retirement, investment, business/farm/rents/royalties, and other) or in any of the

components of investment income (taxable interest, tax-exempt interest, dividends, and gains/losses). For the

derivation of these income measures, see Table A.2 in the appendix.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

immediately after the 2001 recession). The share then increased and averaged 98.5 percent from

age 62 (2007) through age 72 (2017), with the great recession appearing to have little impact on

the share of the population with income.

Among those with income, median inflation-adjusted total income fell by over 20 percent

from age 55 (2000) to age 72 (2017), with most of the decline occurring around the two

recessions (Figure 5). Over the entire period, median per capita total income fell $9,300, from

just over $44,000 at age 55 to just under $35,000 at age 72. Median income fell $4,100 between

age 55 (2000) and age 59 (2004) and then remained fairly stable through age 62 (2007) before

falling an additional $3,300 between age 62 (2007) and age 65 (2010).

In contrast, median spendable income fell by less than 10 percent from age 55 to age 72, as

taxes fell more than proportionately with income (Figure 5). Median per capita total federal

taxes fell by $6,100, from a bit more than $7,100 at age 55 to a bit less than $1,100 at age 72. As a

result, median per capita spendable income fell by only $3,400, from just under $37,000 at age 55

to just over $33,000 at age 72.

The median total federal tax rate fell substantially between age 55 and age 72, with the

typical individual paying much lower income taxes and paying no payroll taxes—either directly

or through a spouse (Figure 6).19 As a percentage of total income, median federal income tax

rates fell from 10.5 percent at age 55 to 2.1 percent at age 72. Median payroll tax rates declined

from 6.6 percent of total income at age 55 to zero by age 68.

Only a relatively small portion of the reduction in taxes we observed was caused by tax law

changes. Legislation cut taxes over this period, especially around the time of the two recessions,

but some of these changes were temporary and had either expired or been partially offset by

other change by 2017. We estimate that, all else equal, tax law changes would have reduced

2017 effective federal income tax rates by less than 3.0 percentage points relative to 2000 and

would have had no impact on 2017 payroll tax rates.20

For an explanation of the tax rate measures, see Section 2.4.

See the appendix for an explanation of our estimates and a review of major tax law changes from 2000 through

2017.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

Taxes fell more than proportionately with income primarily because of changes in the

amount and composition of income with age.21 Income taxes typically decline relative to income

over these ages because (1) total income generally declines with age, and (2) only a portion of

Social Security benefits are included in taxable income.22 In combination, these two changes

reduce the share of total income subject to income tax and reduce the share of taxable income

subject to higher marginal tax rates. Payroll taxes typically decline relative to income over these

ages because labor income falls as a share of total income.

3.1.1 Ranking Individuals by Age 55 to 59 Income

To examine how changes in spendable income differ by the amount of income individuals

had when we first observe them, we rank all individuals alive at the end of 2004 (when they

would have been age 59) by their average age 55 to age 59 per capita total income. We then split

those with positive age 55-59 total income into ventiles (20 equally sized groups), with ventile 1

having the lowest income and ventile 20 having the highest income. Those without positive

age 55-59 total income, representing 1.1 percent of the population aged 59 at year-end 2004, are

included in our tabulations for the entire sample but are not reported separately in the ventile

tabulations.

Note that not all who are alive at age 59 and who are ranked by income survive to age 72,

so the ventiles do not represent an equal share of the balanced panel (Figure 7). Overall, 84

percent of those alive at age 59 survived to age 72. Survival rates varied by age 55-59 income,

For an illustration of how both payroll and income taxes change over the lifecycle with constant tax laws, see Brady

(2016).

The percentage of Social Security benefit payments included in gross income is based on a taxpayer’s modified

adjusted gross income (MAGI), which includes half of Social Security benefit payments plus other income included

in gross income. For single, head of household, and qualifying widow(er) returns: if MAGI is $25,000 or less, no

Social Security benefit payments are included in gross income; if MAGI is between $25,000 and $34,000, the lesser of

50 percent of Social Security benefit payments or 50 percent of MAGI in excess of $25,000 is included in gross income;

if MAGI is in excess of $34,000, the lesser of 85 percent of Social Security benefit payments or 85 percent of MAGI in

excess of $34,000 plus $4,500 [=50%*($34,000-$25,000)] is included in gross income. For joint returns, the MAGI

thresholds are $32,000 and $44,000, respectively. These thresholds are not indexed for inflation. For more information

on the taxation of Social Security benefits, see Internal Revenue Service (2018b).

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

however, ranging from 72 percent for ventile 2 to 93 percent for ventile 20. As a result, the share

of the balanced panel varies from 4.3 percent for ventile 2 to 5.5 percent for ventile 20.23

Median total income is plotted for each ventile in each year in Figure 8. By construction,

average age 55-59 total income increases with ventile, from less than $4,000 for ventile 1 to more

than $200,000 for ventile 20. Income after age 59 has no impact on an individual’s income

ventile, but the ventiles maintained their median total income rank through age 72.

3.1.2 Changes in Income and Taxes by Income Ventile

Higher income ventiles experienced the largest declines in median total income between

age 55 and age 72 (Figure 8). Median total income fell 42 percent between age 55 and age 72 for

ventile 20 and fell 30 percent for ventile 19. Declines were smaller over these ages for lower

income ventiles, with median total income falling 20 percent for ventiles 10 and 11 and falling

9 percent for ventile 5. Median total income actually increased for the lowest two ventiles, with

median per capita total income of ventile 1 more than doubling from $4,400 at age 55 to $9,500

at age 72.

As a percentage of total income, total federal taxes declined the most for the middle income

ventiles (Figure 9a). Median total federal tax rates fell by more than 14 percentage points from

age 55 to age 72 for ventiles 7 through 11. Ventile 9 experienced the largest decline, falling 15.1

percentage points, from 16.5 percent at age 55 to 1.4 percent at age 72. Ventiles 1 through 6

experienced smaller declines in tax rates despite having zero median total federal taxes when

aged 72 because their median tax rates were lower at age 55. Higher income ventiles had higher

tax rates both at age 55 and at age 72 and experienced less of a decline in tax rates than the

middle income ventiles.

Both income and payroll tax rates fell the most for the middle income ventiles. Median

income tax rates fell more for the middle income ventiles than for the lower income ventiles

because they had higher income tax rates at age 55 and, thus, the rates had farther to fall

(Figure 9b). The income tax rates of the middle income ventiles fell more than those of the

Those without positive age 55-59 total income have a survival rate of 82 percent and represent 1.0 percent of the

population aged 72 at year-end 2017.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

higher income ventiles for two reasons. First, they experienced larger declines in taxable income

as a share of total income because Social Security benefits were a higher share of their total

income and—because the portion of Social Security benefits included in taxable income

increases with income—they included a smaller portion of those benefits in taxable income.

Second, declines in taxable income typically translate into larger tax changes for those with

lower income because tax rate brackets are narrower and have larger discrete rate jumps for

those with low and moderate income.24

Although median payroll tax rates were zero for all ventiles at age 72, the middle income

ventiles experienced the largest rate reductions because their payroll tax rates were the highest

at age 55 (Figure 9c). At 7.4 percent of total income, median payroll tax rates at age 55 were

highest for ventiles 6 through 12. Payroll tax rates at 55 were lower for lower income ventiles

because non-labor income (Social Security, retirement, and other income) typically makes up a

larger share of their income. Payroll tax rates at 55 were lower for higher income ventiles both

because non-labor income also makes up a larger share of their income and because only labor

earnings up to the Social Security maximum taxable earnings ($127,200 for an individual in

2017) are subject to the top payroll tax rate.25

Taxes had little impact on spendable income for the lowest income ventiles but falling taxes

moderated the decline in spendable income for the other income ventiles (Figure 10). The

largest moderating effects were for ventiles 6 through 14, with the decline in median spendable

income from age 55 to age 72 less than half the decline in median total income. For example,

inflation-adjusted median spendable income of ventiles 10 and 11 fell by 8 percent from age 55

to age 72, compared with 20 percent declines in median total income.

In 2017, for example, the marginal tax rates and taxable income ranges for the lowest five income tax brackets for

single individuals were: 10 percent (less than $9,325); 15 percent ($9,325 to $37,950); 25 percent ($37,950 to $91,900); 28

percent ($91,900 to $191,650); and 33 percent ($191,650 to $416,700).

All labor earnings (wage and salary earnings and self-employment earnings) are subject to the 1.45 percent

Medicare or Hospital Insurance (HI) tax, but only labor earnings up to the maximum taxable amount are subject to

the 6.2 percent Social Security or Old-Age, Survivors, and Disability Insurance (OASDI) tax. Note that we count the

additional Medicare tax and the net investment income tax, which are used to finance Medicare, as income taxes

rather than payroll taxes. See Table A.2 in the appendix for an explanation of our payroll tax calculation.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

Reflecting changes in both total income and taxes with age, the largest declines in

spendable income were experienced by the highest income ventiles (Figure 10). Between age 55

and age 72, median spendable income fell 33 percent for the highest income ventile and fell 21

percent for ventile 19. Lower income ventiles experienced smaller declines with age, with

median spendable income at age 72 the same or higher than it was at age 55 for ventiles 1

through 5, down less than 10 percent for ventiles 6 through 12, and down by less than

15 percent for ventiles 13 through 18.

3.2 Spendable Income Replacement Rates

The replacement rate measures used in this study are intended to measure replacement of

inflation-adjusted spendable income. Our replacement rates differ from traditional replacement

rate measures in two ways. First, we measure replacement of spendable income rather than

replacement of total income. Second, we compare age 55-59 income to (inflation-adjusted)

income through age 72 rather than attempting to identify a single point in time when retirement

occurs.

For each individual with positive age 55-59 spendable income, spendable income

replacement rates are calculated as the ratio of inflation-indexed spendable income at a given

age to average inflation-indexed spendable income from age 55 (in 2000) through age 59 (in

2004).

More formally, spendable income replacement rates are calculated as:

𝑌𝛼𝑖

𝑖

𝑅𝛼 = 1 59 𝑖

∑

𝑌

5 𝛼=55 𝛼

Where R = spendable income replacement rate

Y = inflation-adjusted per capita spendable income

i indexes individuals

α indexes year-end age, ranging from age 55 (in 2000) to age 72 (in 2017)

At every age through age 72, the typical individual maintained more than 90 percent of the

average spendable income they had in their late 50s. Among all individuals with positive

age 55-59 spendable income, the median spendable income replacement rate was close to 100

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

percent through age 62 and then fell slowly—to 97 percent at age 67, 94 percent at age 70, and

92 percent at age 72 (Figure 11, orange line). Replacement rates varied across individuals, with

25 percent having age 72 replacement rates of 125 percent or more and 25 percent having age 72

replacement rates of 67 percent or less (Figure 11, blue lines).

Those with lower age 55-59 total income tended to maintain a higher percentage of their

spendable income at older ages (Figure 12).26 Spendable income typically increased after age 61

for the lowest income ventiles, with median replacement rates (orange lines) for ventiles 1

through 5 still greater than 100 percent at age 72.

At age 72, median replacement rates fell with age 55-59 income, remaining above 90

percent through ventile 12 and above 85 percent through ventile 17 (Figure 13). The sharpest

declines in spendable income were experienced by those with the highest age 55-59 income,

with the median replacement rate 78 percent for ventile 19 and 62 percent for ventile 20 at

age 72.

A similar pattern by ventile is seen throughout the replacement rate distribution. At the 75th

percentile (top blue line), replacement rates at age 72 were well above 100 percent for the lowerincome ventiles and only dropped below 100 percent for the highest-income ventile (Figure 13).

At the 25th percentile (bottom blue line), the relationship between replacement rates and

age 55-59 income was less pronounced, with ventiles 4 through 16 all around 70 percent at

age 72.

3.3 Discussion of Replacement Rate Results

Although the replacement rate measures used in this study are based on those developed

in Brady (2010), the measures used in this study likely understate the true replacement of

spendable income in retirement because the data do not allow us to measure all taxes and all

savings. Our spendable income measure accounts for federal income and payroll taxes, but state

income taxes are also likely to decline as a share of total income in retirement.27 Our total and

In Figure 12, note that the x-axis ranges in the charts for ventiles 1, 2, 3, and 20 differ from that of the other income

ventiles.

See discussion and simulation results in Brady (2010) and Brady (2016).

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

spendable income measures account for savings accomplished through IRAs and employersponsored DB and DC plans (by excluding employer and employee contributions from income

and by including distributions in income) but they do not account for other savings. Accounting

for other savings would reduce pre-retirement spendable income and—to the extent that nonretirement-plan assets were drawn down in retirement—boost spendable income in

retirement.28

Even if the true spendable income replacement rate could be measured, it would still leave

open the question of what would represent retirement resource adequacy. The assumption that

retirees would want to replace 100 percent of their pre-retirement spendable income is a

reasonable starting point, but there are reasons the optimal replacement rate would be lower

than 100 percent.29 For example, there are certain work-related expenses – such as commuting

expenses, purchases of lunch and coffee, and buying work clothes – that need not be incurred

when an individual no longer works. Perhaps most importantly, if workers raised children

prior to retirement, necessary expenses presumably would be lower if these individuals or

couples were no longer supporting minor children in retirement.30

3.4 Summary

Despite median total income declining 21 percent from age 55 to age 72, median spendable

income fell by only 9 percent. The difference was attributable to a drop in taxes in retirement,

with the typical individual paying much lower income taxes and paying no payroll taxes.

Effective tax rates typically fell the most for individuals in the middle of the age 55-59

income distribution. Median effective total federal tax rates were generally quite low at age 72—

zero for the lowest 30 percent and below 5.0 percent for the lowest 60 percent of the age 55-59

income distribution. Those in the middle of the income distribution typically experienced the

For many, their most important asset is their home. For a discussion of how accounting for owner-occupied

housing would affect replacement rate measures, see Brady (2010).

See discussion in note 15.

For example, Scholz and Seshadri (2007) find that, controlling for lifetime earnings, children are an important

determinant of household wealth.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

largest drop in tax rates, however, because taxes were already low at age 55 for the lowest

income ventiles.

Those with lower age 55-59 income typically maintained more of their spendable income at

older ages. At age 72, the median replacement rate was greater than 90 percent for the lowest 12

income ventiles (that is, the bottom 60 percent of the age 55-59 income distribution). Substantial

drops in total and spendable income were typical only for those with the highest age 55-59

income, with the median spendable income replacement rate 62 percent for the highest income

ventile at age 72.

4. Composition of Income

This section of the paper investigates how the composition of income changes as the panel

ages and how the composition of income varies with age 55-59 income. The primary focus of

this section will be on the three sources of income on which most individuals rely for nearly all

of their income over the period of the study: labor (wage and salary, self-employment earnings,

and unemployment compensation), Social Security (disability benefits and retirement benefits),

and retirement (IRA distributions and income from pensions and annuities) income.

For labor, Social Security, and retirement income, we examine both own and per capita

income. Measures of own income allow us to analyze individual decisions about working,

claiming Social Security benefits, and taking distributions from retirement plans. Measures of

per capita income provide a fuller picture of the resources of married individuals.

Our analysis suggests that retirement is better thought of as a period of transition rather

than a single point in time. Most individuals did not move directly from having only labor

income to having only Social Security and retirement income, and married couples were even

less likely to do so. The share of the population that stopped working between age 61 and age

67 was about one-third of the share of the population who claimed Social Security benefits from

age 62 through age 66.

The most common ages at which we first observe individuals receiving their own Social

Security income were age 62 and age 66. Age 62 is the earliest age at which individuals can

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

claim their own retirement benefits. Age 66 is the full benefit retirement age for the 1945 birthyear cohort. The age of first own Social Security income tends to increase with age 55-59 income.

By age 72, most individuals received both Social Security income and retirement income. At

age 72, 67 percent directly received distributions from IRAs, pensions, and annuities and

75 percent received them directly or through a spouse. Social Security benefits were received by

nearly all individuals and represented about half of per capita total income for the typical

72 year-old.

Reflecting the design of the US Social Security system, those with lower age 55-59 income

tended to rely more on Social Security benefits in retirement while those with higher age 55-59

income tended to rely more on retirement plan distributions.

4.1 Incidence and Amounts

For analysis, we separate income into two broad categories. The first category includes the

three sources of income that most of the population rely on throughout their lives: labor, Social

Security, and retirement income. The second category includes all other sources of income:

investment (taxable interest, tax-exempt interest, dividends, gains/losses),

business/farm/rents/royalties (business and farm income in excess of self-employment earnings;

income from rents, royalties, partnerships, S-corps, and trusts), and other (net alimony [alimony

received less alimony paid] and other income). For ease of exposition, we will refer to the first

category as Labor+SS+Retire income and we will refer to the second category as

non-Labor+SS+Retire income.

Most individuals had both types of income throughout the period studied, but between age

55 (in 2000) and age 72 (in 2017) the incidence of Labor+SS+Retire income increased by

3 percentage points while the incidence of non-Labor+SS+Retire income decreased by 14

percentage points (Figure 14). At age 55, 77 percent of the population had both types of income,

18 percent had only Labor+SS+Retire income, and 5 percent had only non-Labor+SS+Retire

income. At age 72, reflecting the combination of a higher incidence of Labor+SS+Retire income

and a lower incidence of other income, the share of the population with both types of income

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

fell to 66 percent, the share with only Labor+SS+Retire income rose to 32 percent, and the share

with only non-Labor+SS+Retire income was close to zero.

Although most had both types of income, Labor+SS+Retire income made up the bulk of the

typical individual’s total income throughout the period studied. Among those with

non-Labor+SS+Retire income, the median amount peaked at just over $1,000 in 2007 (at age 62).

4.1.1 Labor, Social Security, and Retirement Income

The share of the population who derive a portion of their income from labor—either

directly or through a spouse—declined markedly from 90 percent at age 55 in 2000 to 34 percent

at age 72 in 2017 (Figure 15, top left panel). At age 55, 78 percent of individuals received labor

income directly and another 12 percent of individuals had no labor income but had a spouse

who did. The share of the population with their own labor income fell by about 2.0 percentage

points a year between age 55 and age 60. Declines then accelerated, peaking above 5.0

percentage points per year between age 65 and age 67. By age 72, only 23 percent of the

population had their own labor income. The share of the population with labor income either

directly or through a spouse fell by a similar amount, although it declined more slowly at

younger ages. The share of the population who had no labor income but who had a spouse who

did increased after age 55, peaked at around 15 percent between age 63 and 67, and then

decreased to 11 percent by age 72.

In addition to fewer individuals having labor income, the median income of those who

continue to work declined at older ages (Figure 15, top right panel). Adjusted for inflation,

median own labor income fell by about two-thirds, from $44,000 at age 55 to $14,000 at age 72,

with more than half of the decline occurring between age 61 and age 67. Among those with

labor income either directly or through a spouse, there was a similar decline in median per

capita labor income, from $40,000 at age 55 to $11,000 at age 72.

As the share with labor income declined, the share with Social Security income and

retirement income increased.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

By age 72, 96 percent of the population received their own Social Security income

(Figure 15, middle left panel).31 At age 55, 4.6 percent of the population already received their

own Social Security income.32 This share increased with age, with the increase accelerating

when survivor benefits became available at age 60 and again at age 62—the early claiming age

for own or spousal retirement benefits. The two largest increases in the share of the population

receiving their own Social Security income were at the early claiming age of 62 (when it

increased from 11 percent to 39 percent) and at the full benefit retirement age of 66 (when it

increased from 67 percent to 87 percent).

After age 70, only 1.2 percent of the population did not receive their own Social Security

income but had a spouse who did. At age 55, 5.7 percent of the population had only spousal

Social Security income—either because the spouse received disability benefits or because the

spouse was older and had claimed retirement benefits. This share increased to 18 percent of the

population by age 61 but then declined sharply.

By age 72 (in 2017), median own Social Security income was $17,200 and median per capita

income was $16,400 (Figure 15, middle right panel). Median inflation-indexed Social Security

benefits generally increased with age through age 70 and then remained about unchanged from

age 70 through age 72. Median own income was reduced in years when there was a large

amount of claiming—most notable from age 62 through age 66—as many individuals received

benefits for less than a full year. This effect was less notable for median per capita income

because the share of married couples with both spouses receiving Social Security income

increased with age—which, in turn, increased median per capita income relative to median own

income.

In this paper, own Social Security income refers to any benefits sent directly to an individual and which would be

reported as being sent to the individual on Form SSA-1099. These would include an individual’s own benefit—that is,

disability and retirement benefits to which the individual is entitled based on their own work history. It would also

include spousal or survivor benefits to which an individual is entitled based on the work history of a spouse.

Social Security income includes both disability benefits and retirement benefits. Before age 60, when individuals

may become eligible for survivor benefits (which are considered retirement benefits), nearly all own Social Security

income is in the form of disability benefits. Beginning at age 66—the full benefit retirement age for individuals born

in 1945—nearly all own Social Security income is in the form of retirement benefits (inclusive of own, spousal, and

survivor benefits).

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

Although less prevalent than Social Security income at age 72, retirement income was more

prevalent at age 55. By age 72, 75 percent of the population received retirement income (IRA

distributions plus pension and annuity income) either directly or through a spouse (Figure 15,

lower left panel). The share of the population who received their own retirement income

increased from 17 percent at age 55 to 67 percent at age 72. At age 55, another 9.5 percent of the

population did not have their own retirement income but had a spouse who did. The share with

only spousal retirement income increased to 14 percent of the population by age 69 but then

declined to 8.2 percent at age 72.

The share of the population with their own retirement income increased throughout the

period, but it increased more quickly at certain ages related to pension rules. The share

increased by 6.0 percentage points at age 60, presumably related to retirement distributions no

longer being subject to a 10-percent early withdrawal penalty after age 59½. There was also a

5.4 percentage point increase at age 65, the normal retirement age for many DB pension plans.

The most rapid increase, however, was the 13 percentage point increase from age 69 to age 71,

presumably related to distributions generally being required from IRAs and DC retirement

plans after age 70½.

By age 72 (in 2017), median own retirement income was $16,200 and median per capita

income was $15,300 (Figure 15, bottom right panel). Although median own retirement income

peaked at age 61, median per capita retirement income peaked at age 72.

The peak in median own retirement income at age 61 could be related to certain

beneficiaries of government and private-sector DB pension plans. Some with retirement income

at younger ages may have received benefits from a DB pension plan with an early retirement

supplement that ends when Social Security benefits become available at age 62—such as former

federal government workers covered by the Federal Employee Retirement System (FERS)33 or

former private-sector workers covered by certain DB pension plans. It may also have been

related to changes in the composition of those receiving retirement income with age. For

example, it may be the case that those with retirement income at younger ages were more likely

See Office of Personnel Management (2011).

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

to have been government workers not covered by Social Security—such as former federal

government workers covered by the Civil Service Retirement System (CSRS) and certain former

state and local government workers—who typically have more generous pension benefits as a

result.

4.1.2 Labor+SS+Retire Income

When the incidence and amounts of labor, Social Security, and retirement income are

examined separately, it may be difficult to make sense of the results. When we examine what

individuals get from the combination of these three income sources, a more coherent picture

emerges about the transition individuals make from working to retirement.

From age 55 to age 72, we observe a decline in the incidence of labor income and an

increase in the incidence of both Social Security and retirement income, but annual changes

were not highly correlated (Figure 16, top left panel). For example, the incidence of own Social

Security income increased 40 percentage points between age 61 and 63, from 11 percent to 51

percent of the population. Over these same ages, the incidence of own labor income declined by

only 8 percentage points, from 65 percent to 57 percent of the population.

One reason the annual incidence changes were not highly correlated is that some

individuals may not have had their own labor income in the year (or years) before they first

received their own Social Security income. Indeed, the increased share of the population who

received their own Labor+SS+Retire income after age 61 indicates that some individuals had

neither their own labor income nor their own retirement income prior to claiming Social

Security benefits.

Another reason the annual incidence changes were not highly correlated is that some

individuals may have continued to work after claiming Social Security or after beginning to

receive retirement income (Figure 16, bottom left panel). For example, of those aged 66 who

received Social Security, 42 percent also received their own labor income. That share remained

above 30 percent until age 70. Of those aged 55 with their own retirement income, 79 percent

had their own labor income. That share remained above 50 percent until age 65 and above 30

percent until age 69.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

By age 72—and focusing on income received directly or through a spouse—few relied on a

single source of Labor+SS+Retire income (Figure 16, bottom right panel). The most common

combination, at 48 percent of the population, was having both Social Security benefits and

income from pensions, annuities, and IRAs. Another 26 percent of the population had these two

sources of income plus labor income. Only 16 percent of the population had Social Security

income alone.

These patterns suggest that retirement is better thought of as a transitional period rather

than a single point in time. Focusing on how any single source of income changes with age may

provide a misleading picture, as many do not move instantaneously from being fully dependent

on labor income to being fully dependent on Social Security benefits and retirement plan

distributions. Some stop working well before claiming Social Security and some continue to

work several years after claiming. Some begin to receive retirement income well before claiming

Social Security and while they continue to work, while others delay drawing down their

retirement assets until sometime after they claim Social Security or stop working.

A more consistent, and perhaps more understandable, pattern emerges by looking at the

share of the population with income from at least one of these three sources (Figure 17, left

panel). Looking at those with their own Labor+SS+Retire income, incidence was relatively flat

from age 55 through age 61. Associated with claiming of Social Security benefits, this share

increased 14 percentage points between age 61 and age 66, from 83 percent to 97 percent. The

share of the population who received Labor+SS+Retire income either directly or through a

spouse was more stable over the entire period, as it increased only 4 percentage points between

age 61 and age 66, from 94 percent to 98 percent.

The differences between own incidence and own or spouse incidence indicates that most of

the increase in own Labor+SS+Retire income incidence over these ages came from married

individuals who, prior to age 62, did not have their own income but had a spouse who did. On

average from age 55 through age 61, 83 percent of the population received their own

Labor+SS+Retire income and another 11 percent did not but had a spouse who did. On average

from age 70 through age 72, the share of the population with their own Labor+SS+Retire income

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

increased to 98 percent while the share with only spousal Labor+SS+Retire income declined to

0.3 percent.

Conditional on having the income, median Labor+SS+Retire income declined by less

measured on a per capita basis then on an own basis (Figure 17, right panel). Median own

Labor+SS+Retire income fell 33 percent, from $44,000 at age 55 to $30,000 at age 72. Median per

capita Labor+SS+Retire income decreased by only 24 percent over these same ages, from $43,000

to $32,000. Per capita income fell by less because fewer individuals had only spousal

Labor+SS+Retire income at age 72. All else equal, an individual who changes from having no

own income to having a small amount will reduce the conditional own median. That same

individual would increase the conditional per capita median, however, if they have a spouse

with sufficiently high income.34

In addition to declining by less, the per capita conditional median was higher than the own

conditional median of Labor+SS+Retire income after age 61 (Figure 17, right panel). When

looking at the components of Labor+SS+Retire income individually, median own labor and

retirement income were higher than their per capita equivalents at all ages, and median own

Social Security income was higher than its per capita measure at all ages other than age 62.

When looking at the combination of the three types of income, we also see that median own

Labor+SS+Retire income was higher than its per capita measure from age 55 through age 61.

Beginning at age 62, however, median per capita Labor+SS+Retire income was higher.

The reason that median per capita was higher than median own Labor+SS+Retire income

after age 61 was that married individuals with low own income often had a spouse with high

own income. For the components of Labor+SS+Retire income, the fact that the same amount of

income was spread across more individuals lowered the per capita median relative to the own

median and nearly always dominated other factors that might offset this effect. This was also

true for the combination of the three types of income from age 55 through age 61. As the share

with only spousal Labor+SS+Retire income declined after age 61, however, the effect of spousal

All else equal, a married individual who went from having no own income to having own income equal to ε would

𝜀

increase conditional median per capita income provided their spouse had own income greater than 𝑦̅ – , where 𝑦̅

equals the conditional median per capita income when the married individual had no own income.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

sorting began to dominate the relationship between own and per capita income. That is, for

those aged 62 or older, median per capita income exceeded median own income because a

sufficient number of married individuals with lower than median own income had a spouse

with higher than median own income. And the effect of this sorting was enough to outweigh

the impact on the per capita median of spreading the same amount of income over more

individuals.

4.1.3 Labor, Social Security, and Retirement Income by Income Ventile

Patterns of work, Social Security claiming, and retirement plan distributions differ based on

income. Those with higher age 55-59 income generally were more likely to work at older ages

and were more likely to delay claiming Social Security benefits. For those who continued to

work, however, median labor income typically fell more with age for higher income ventiles. By

age 72, nearly all received Social Security benefits across the income distribution, whereas the

share who received retirement income generally increased with age 55-59 income. Conditional

on having the income, median Social Security income varies less across income ventiles than

either median labor income or median retirement income.

Incidence

At age 55, the probability of having own labor income generally increased with income,

peaking for the fourth highest ventile (Figure 18, top left panel). Incidence increased from 44

percent for ventile 1 to 89 percent for ventile 17 before declining to 78 percent for ventile 20.

The share with own labor income fell with age across ventiles but tended to fall more

slowly for those with the highest age 55-59 income and, as a result, the top three income ventiles

were the most likely to have their own labor income by age 67. Of the total decline in own labor

income incidence between age 55 and 72, nearly 60 percent of the decline had occurred by age

61 for the lowest income ventile. In contrast, only about 20 to 25 percent of the decline had

occurred by age 61 for the top 16 income ventiles (that is, the top 80 percent of the age 55-59

income distribution). After age 61, the decrease in incidence was about evenly split between

ages 61 to 66 and ages 66 to 72, with, generally, a bit more of the decline occurring at younger

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

ages for the lower income ventiles and a bit more of the decline occurring at older ages for the

higher income ventiles.

By age 72, those with the highest age 55-59 income were much more likely to have their

own labor income. Incidence of own labor income at age 72 was below 20 percent for the bottom

five ventiles, between 20 and 25 percent for ventiles 6 through 17, and increased to 37 percent

for the highest income ventile.

The share with labor income either directly or through a spouse was higher at every age

than the share who receive it directly, as some without labor income had a spouse who worked

(Figure 18, top panels). At age 55, the incidence of own or spouse labor income was higher than

90 percent for the top 15 ventiles (that is, the top 75 percent of the age 55-59 income

distribution). At age 72, the incidence of own or spouse labor income was 30 percent or less for

the bottom five ventiles, between 30 and 38 percent for ventiles 6 through 17, and increased to

52 percent for the highest income ventile.

The age at which we observed individuals receiving their own Social Security income

tended to increase with age 55-59 income (Figure 18, middle left panel). Across all income

ventiles, most (74 percent to 87 percent) of the Social Security claiming we observed (that is,

claiming after age 55) occurred from age 62 through age 66. Within that age range, however,

lower income ventiles were more likely than higher income ventiles to claim at age 62, and

higher income ventiles were more likely than lower income ventiles to claim when older,

particularly at the full benefit age of 66. Similarly, outside of that age range, lower income

ventiles were more likely to have their own Social Security income prior to age 62, and higher

income ventiles were more likely to begin receiving their own Social Security benefits after

age 66.

Until the full benefit retirement age, the lowest income ventiles had the highest own Social

Security incidence. From age 55 to 65, ventile 2 had the highest incidence of own Social Security

income. From the full benefit age of 66 to age 72, ventiles 5 to 9 had the highest incidence.

At age 72, ventiles 15 through 19 had lower incidence of own Social Security income than

all but the two lowest income ventiles. Most of those with lower age 55-59 income without their

own Social Security presumably did not work long enough to qualify for benefits, nor did they

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

have a spouse who did.35 In contrast, those with higher age 55-59 income without benefits were

most likely former government workers who, because they were covered by an alternative

pension system, were not required to participate in Social Security.

The share with Social Security income either directly or through a spouse was higher than

the share who received it directly, especially at younger ages (Figure 18, middle panels). Across

all ventiles, the share who did not receive their own Social Security benefits but had a spouse

who did peaked at age 61. After age 61, this share declined more rapidly for lower income

ventiles, with the highest income ventiles not experiencing rapid declines in spousal only Social

Security incidence until the full benefit retirement age of 66.

By age 70, nearly all had Social Security income. At age 72, only 79 percent in the lowest

income ventile received Social Security benefits directly or through a spouse. For the other

19 ventiles (that is, the top 95 percent of the age 55-59 income distribution), this share averaged

98 percent and ranged from 95 percent for ventile 2 to 99 percent for ventile 9.

The incidence of retirement income generally increased with age 55-59 income but,

depending on age, peaked in ventiles 15 through 18 and was noticeably lower in the highest

income ventile until age 71 (Figure 18, bottom panels). At age 55, more than 20 percent of

ventiles 11 through 18 already received their own retirement income and more than 30 percent

received retirement income directly or through a spouse. Incidence increased each year across

the income distribution but increased more slowly for the highest income until after age 69. For

example, own retirement income incidence at age 69 was 20 percentage points higher for ventile

15 than it was for ventile 20 (70 percent compared with 50 percent) but was about the same by

age 72 (84 percent for both).

By age 72, retirement income was common across the income distribution. The share of the

population who received retirement income either directly or through a spouse was greater

than 60 percent for the top 16 income ventiles (that is, the top 80 percent of the age 55–59

Individuals are required to have the equivalent of 10 years of covered employment to qualify for Social Security

benefits. Individuals who do not qualify based on their own work history, however, would be eligible to receive

spousal or survivor benefits if they were married, for 10 years or longer, to a worker who qualified for Social Security

benefits. See note 31 for a further explanation of how we measure receipt of own Social Security income.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

income distribution), above 80 percent for the top 12 ventiles, and 90 percent or more for the top

8 ventiles.

Median Amounts

In addition to fewer individuals working at older ages, the earnings of those who

continued to work typically declined with age for all but those with the lowest age 55-59 income

(Figure 19). At age 55 and expressed in constant 2017 dollars, median own labor income ranged

from $6,200 for ventile 1 (bottom left panel) to more than $170,000 for ventile 20 (top left panel).

At age 72 (in 2017), median own labor income ranged from $8,900 for ventile 2 to $53,000 for

ventile 20.

Conditional on working, median inflation-indexed own labor income fell the most for the

highest income ventiles between age 55 and age 72 (Figure 19, left panels). Median own labor

income fell by 69 percent or more for the top 10 ventiles (Figure 19, top left panel). In contrast,

median labor income declined by less than 50 percent for ventiles 2 through 5 and actually

increased for ventile 1 (Figure 19, bottom left panel).

Conditional on either the individual or their spouse working, we see changes by ventile in

median per capita labor income between age 55 and age 72 (Figure 19, right panels) that were

similar to the changes in median own labor income.

Social Security benefits increased with age 55-59 income but, reflecting the progressive

benefit formula, varied less by income ventile than either labor or retirement income. Among

those who received benefits directly, median own Social Security income at age 72 ranged from

$9,300 for the lowest income ventile to $26,000 for the highest income ventile (Figure 20, top left

panel). Conditional on either the individual or their spouse receiving benefits, median per

capita Social Security income at age 72 ranged from $8,700 for ventile 1 to just over $22,000 for

ventile 20 (Figure 20, bottom left panel). By contrast, among those with retirement income when

age 72, both own and per capita median retirement income ranged from under $4,000 for the

three lowest income ventiles to just over $39,000 for the highest income ventile (Figure 20, right

panels).

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

4.1.4 Labor+SS+Retire Income by Income Ventile

The incidence of own Labor+SS+Retire income increased substantially between age 55 and

age 72 across the income distribution (Figure 21, top panel). At age 55, incidence of own

Labor+SS+Retire income ranged from 55 percent for ventile 1 to more than 90 percent for

ventiles 12 through 18. Own incidence by ventile remained about the same through age 61, with

only the lowest and highest income ventiles experiencing noticeable declines. Beginning at

age 62, the incidence of own Labor+SS+Retire income increased noticeably, particularly among

lower income ventiles. By age 72, incidence of own Labor+SS+Retire income was more than 95

percent for all but the lowest income ventile and more than 99 percent for the top 15 ventiles

(that is, the top 75 percent of the age 55–59 income distribution).

For all but the lowest income ventiles, the vast majority of the increase in own

Labor+SS+Retire income incidence over these ages was attributable to individuals who did not

have the income at younger ages but had a spouse who did. The share of the population with

only spousal Labor+SS+Retire income declined markedly after age 61 (Figure 19, middle panel).

From age 55 through age 61, the share of ventiles with only spousal Labor+SS+Retire income

ranged from 7 percent to 17 percent, with the lowest and highest ventiles nearer to the top of

that range. Beginning at age 62, the share with only spousal income from these sources declined

substantially and is close to zero by age 72 across all ventiles.

As a result, the increase in the share of the population who received Labor+SS+Retire

income either directly or through a spouse was much less dramatic over these ages. Incidence of

own or spouse Labor+SS+Retire income increased by more than 4.0 percentage points for only

the lowest three income ventiles, with ventiles 9 through 19 experiencing an increase of less

than 2.0 percentage points.

Changes in the amount of Labor+SS+Retire income were very similar to the changes we

observed in total income, with the largest declines experienced by those with the highest

age 55–59 income (Figure 22). Median per capita Labor+SS+Retire income fell by nearly half

between age 55 and age 72 for ventile 20 and fell by more than one-third for ventile 19. Over

these same ages, median per capita Labor+SS+Retire income fell by 22 percent for ventiles 10

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

and 11, fell by 11 percent for ventile 5, and actually increased for the lowest two income

ventiles.

Because the highest income ventiles experienced the largest declines in median

Labor+SS+Retire income, differences across ventiles fell with age. At age 55 and expressed in

constant 2017 dollars, median per capita Labor+SS+Retire income ranged from $5,800 for

ventile 1 to more than $160,000 for ventile 20. At age 72, median per capita Labor+SS+Retire

income ranged from $9,700 for ventile 1 to $84,000 for ventile 20.

4.1.5 Non-Labor+SS+Retire Income

A large share of the population received non-Labor+SS+Retire income, but the amounts

were typically modest. Over the period studied, incidence of non-Labor+SS+Retire income fell

14 percentage points, from 82 percent at age 55 in 2000 to 68 percent at age 72 in 2017 (Figure 23,

top panel, blue line). Among those with the income, median non-Labor+SS+Retire income was

about $500 per year, on average, but ranged from roughly $300 to $1,000 depending on the year

(Figure 23, middle panel, blue line).

Although the incidence of non-Labor+SS+Retire income fell considerably, we believe a good

portion of that decline was related to market returns—particularly the fall in interest rates and

the reduction in interest paid on checking and savings accounts—rather than age. Similarly, we

believe changes in median amounts were more related to market returns than age.

Among the components of non-Labor+SS+Retire income, taxable interest had the largest

decline in both incidence and amounts. The share of the population with taxable interest income

fell by 20 percentage points between age 55 and age 72, from 73 percent to 53 percent (Figure 23,

top panel, orange line). From a peak of nearly $400 in 2007 (at age 62), median taxable interest

fell to less than $100 after 2012 (Figure 23, middle panel, orange line).

The pattern of taxable interest incidence and amounts by year appear related to market

interest rates—especially the unprecedented extended period of low interest rates after 2008.

For example, the federal funds rate—which anchors interest rates paid on bank deposits and

short-term debt securities—averaged over 6.0 percent in 2000, fell in response to the 2001

recession, and then increased, averaging 5.0 percent in 2006 and 2007, before falling sharply

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

during the great recession (Figure 23, bottom panel, orange line).36 From 2009 to 2015, the

federal funds rate averaged less than 0.2 percent. In addition to lowering the target federal

funds rate, the Federal Reserve System also engaged in so-called “quantitative easing” which

lowered longer-term interest rates, with the third round of quantitative easing starting in

September 2012.

Another indication that the declining incidence of taxable interest was related to market

interest rates—rather than, say, solely due to the drawdown of assets earning taxable interest

with age—is that incidence declined across the income distribution (Figure 24, middle panel).

There was a similar percentage point decline in incidence across all but the highest income

ventiles, with the share of the decline occurring after 2007 (age 62) increasing with income.

The incidence of other non-Labor+SS+Retire income also fell as the cohort aged, falling

9.0 percentage points between age 55 and age 72, from 63 percent to 54 percent (Figure 23, top

panel, green line). Of the components of other non-Labor+SS+Retire income, the incidence of

business income—that is, the business income of the self-employed—fell the most, from 23

percent of the panel at age 55 in 2000 to 13 percent at age 72 in 2017 (not shown).37

Among those with the income, median other non-Labor+SS+Retire income varied over the

business cycle (Figure 23, middle panel, green line), driven by variations in gains/losses and, to

a lesser extent, dividends. Median gains/losses ranged from gains of $1,000 in 2000 to losses of

$1,700 in 2009 (not shown). Median dividends were also cyclical but peaked in 2017 and had a

narrower range, from $280 in 2002 to $1,000 in 2017 (not shown). Perhaps not surprisingly given

the importance of gains/losses and dividends, the amount of other non-Labor+SS+Retire income

appears to have been related to equity market returns (Figure 23, bottom panel, green line).

The federal funds rate is the rate charged on overnight interbank loans of excess reserves. As part of its monetary

policy, the Federal Open Market Committee (FOMC) of the Federal Reserve System sets a target federal funds rate

and engages in market transactions to keep the federal funds rate in the target range.

Although most business and farm income is included in our measure of self-employment earnings, and thus

included in labor income, individuals with this income also will receive at least some business and farm income.

Individuals with business or farm losses have the income because self-employment earnings cannot be negative, thus

the entire amount of the any loss would be counted as business or farm income. Individuals with positive business or

farm income would have a small residual amount of business or farm income due to the way self-employment

earnings are calculated by the tax code. See appendix for an explanation of that calculation.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

Those with lower age 55-59 income were less likely to have non-Labor+SS+Retire income in

any given year and experienced the biggest declines in incidence between age 55 and age 72

(Figure 24, top panel). At age 55, incidence ranged from 57 percent of individuals in ventile 2 to

99 percent of individuals in ventile 20. Between age 55 and age 72, incidence fell by more than

20 percentage points for the lowest income ventiles, by around 15 percentage points for the

middle income ventiles, and by less than 10 percentage points for the highest income ventiles.

As a result, incidence at age 72 ranged from 28 percent for ventile 1 to 97 percent for ventile 20.

Incidence of non-Labor+SS+Retire income declined more for lower income ventiles for two

reasons. First, lower income ventiles experienced a larger percentage point decline in the

incidence of business income from self-employment (a component of other nonLabor+SS+Retire income). Second, conditional on having non-Labor+SS+Retire income, lower

income ventiles were less likely to have both taxable interest income and other nonLabor+SS+Retire income. Thus, reduced incidence of either type of income is more likely to

translate into reduced overall incidence for the lower income ventiles.

The incidence of taxable interest income generally increases with age 55-59 income but, as

already noted, all but the highest income ventiles experienced similar percentage point declines

in incidence between 2000 and 2017 (Figure 24, middle panel). For example, incidence fell from

57 percent in 2000 (age 55) to 36 percent in 2017 (age 72) for ventile 5 and fell from 86 percent to

65 percent for ventile 15.

Those with higher age 55-59 income were more likely to both have other nonLabor+SS+Retire income at age 55 and to continue to have it throughout the period analyzed

(Figure 24, middle panel). Lower income ventiles experienced larger declines in incidence

primarily because of larger declines in the incidence of business income from self-employment

(not shown).38 The share of the population with business income was similar across ventiles at

age 55, but incidence falls more with age for the lower income ventiles.

Although non-Labor+SS+Retire income was common across the income distribution,

substantial amounts were typical only among those with the highest age 55-59 income (Figure

See note 37.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

25, top panel). Although it varied over time, median non-Labor+SS+Retire income averaged less

than $200 for the lowest 7 ventiles, less than $500 for the lowest 14 ventiles, and less than $1,000

for the lowest 17 ventiles. For ventile 20, in contrast, median annual non-Labor+SS+Retire

averaged just under $25,000.

After peaking in 2007 (age 62), median taxable interest declined substantially for all income

groups (Figure 25, middle panel). By 2017 (age 72), median taxable interest was $100 or less for

the lowest 17 ventiles (that is, the bottom 85 percent of the age 55–59 income distribution).

Median other non-Labor+SS+Retire income was more cyclical, but when the peak occurred

differs with income. For 16 of the lowest 17 income ventiles, median amounts peaked in 2017 at

age 72. For the top three income ventiles (that is, the top 15 percent of the age 55–59 income

distribution), the peak was in 2007 at age 62. For example, median inflation-adjusted other nonLabor+SS+Retire income of the top income ventile hit $27,500 in 2000, $37,300 in 2007, and was

about $23,000 in 2014 and 2017.

4.2 Income Shares

To better quantify the importance of different types of income, this section analyzes the

share of total income from different sources. These measures capture both the likelihood that

individuals have a particular source of income—either directly or through a spouse—and the

amount received by those who have the income.

For each individual we calculate income shares as:

𝑆𝛼𝑖 =

𝑋𝛼𝑖

𝑍𝛼𝑖

Where S = share of per capita total income from a given source39

X= per capita income from a given source

Z = per capita total income

i indexes individuals

α indexes year-end age, ranging from age 55 (in 2000) to age 72 (in 2017)

For individuals with positive source income and negative total income, the income share from that source is set to

100 percent. For individuals with negative source income—which can only occur in the case of non-Labor+SS+Retire

income and some of its components—the income share from that source is set to zero. Because both negative total

income and negative source income are uncommon, individuals to whom these income shares are assigned are

unlikely to represent the 25th percentile, median, or 75th percentile.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

4.2.1 Income Shares by Age

From age 55 (in 2000) through age 72 (in 2017), most individuals received most of their

income from some combination of three sources: labor earnings, Social Security benefits, and

retirement income (that is, distributions from IRAs, pensions, and annuities). In all years, the

typical individual got more than 99 percent of their total income from these sources, with the

median share from Labor+SS+Retire income averaging 99.7 percent from age 55 in 2000 to age

72 in 2017 (Figure 26, orange line). Over the same period, the 25th percentile of the

Labor+SS+Retire income share of total income averaged 93.7 percent, falling below 90 percent

only at ages 61 and 62 in 2006 and 2007 (Figure 26, lower blue line).

The median share of total income from the combination of own and spousal labor income

was 95 percent at age 55 and 56 but then begins to decline and was zero by age 69 (Figure 27,

top panel). At least 25 percent got nearly all their income from labor through age 61, and 25

percent still got 11 percent or more of their income from labor at age 72. Another 25 percent got

less than two-thirds of their income from labor at age 55, and at least 25 percent had no labor

income by age 64.

As labor income declined, Social Security and retirement income increased in importance

(Figure 27, middle and bottom panel). The typical individual received neither Social Security

nor retirement income—either directly or through a spouse—before age 62. The median share of

income from Social Security hit 47 percent by age 72, with 25 percent of the panel getting 75

percent or more of their income from Social Security, and 25 percent getting 29 percent or less.

The median share of income from retirement hit 27 percent by age 72, with 25 percent of the

panel getting 52 percent or more of their income from retirement plans, and 25 percent getting

1.3 percent or less.

4.2.2 Income Shares by Age and Income Ventile

It was only in the highest age 55-59 income ventiles where non-Labor+SS+Retire income

provided more than a de minimus share of total income for the typical individual at any point

during the study period (Figure 28). By income, the median Labor+SS+Retire income share

averaged less than 99 percent for only the top three income ventiles (that is, the top 15 percent

of the age 55-59 income distribution). Similarly, other than the lowest income ventile where

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

some have no Labor+SS+Retire income at younger ages, the 25th percentile of the

Labor+SS+Retire income share averaged less than 90 percent for only the top three income

ventiles. For the highest income ventile at age 72, the median share was 81 percent, with 25

percent getting more than 98 percent and 25 percent getting less than 49 percent of their total

income from Labor+SS+Retire income.

All but the lowest and highest income typically got nearly all their income from labor at

age 55 but labor income was generally more important at older ages for those with higher

age 55-59 income (Figure 29). At age 55, the median share of an individual’s per capita total

income from labor was 95 percent or more for ventiles 3 through 17. The median share declined

with age, falling most quickly for those with lower age 55-59 income but eventually hitting zero

for all but the highest income ventile. From age 64 through age 72, the highest income ventile

had the highest median share of income from labor.

Reflecting the progressive benefit formula, Social Security income was more important for

those with lower age 55-59 income (Figure 30). The median Social Security income share

increased earlier and more sharply for those in the lowest income ventiles. At age 72, the

median share of per capita total income from Social Security was 100 percent for the two lowest

income ventiles, fell below 50 percent by ventile 11, and was below 33 percent for the highest

four income ventiles (that is, the top 20 percent of the age 55-59 income distribution).

Retirement plan distributions were most important for those with moderate to moderately

high age 55-59 income (Figure 31). The median share of per capita total income from retirement

was positive at age 60 for ventiles 10 through 19 and the median share remained highest for

these ventiles through age 72. At age 72, typical individuals in the four lowest income ventiles

(that is, the bottom 20 percent of the age 55-59 income distribution) got little to no retirement

income. The median retirement income share generally increased with income at age 72,

peaking at 50 percent for ventile 18.

The median retirement income share generally increased with age but there was a notable

jump up for most ventiles between ages 69 and 71, likely related to the start of RMDs from IRAs

and DC pensions. This was particularly true for the highest income ventile, with the median

retirement income share increasing from 11 percent at age 69 to 27 percent at age 71.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

4.3 Summary

Over the period studied, most individuals received nearly all their income from three

sources: labor earnings, Social Security benefits, and retirement plan distributions. Across all

years, only those with the highest age 55-59 income typically received substantially amounts of

non-Labor+SS+Retire income (which includes sources such as interest, dividends, gains/losses,

and income from partnerships and S-corps).

We observe a decline in the incidence of labor income and an increase in the incidence of

both Social Security and retirement income from age 55 to age 72, but annual changes in

incidence were not highly correlated. For example, although 76 percent of the panel claimed

Social Security benefits from ages 62 through 66, incidence of own labor income declined by

only 26 percentage points between age 61 and age 67.

Changes in income composition with age suggest that, for many individuals, retirement is a

transitional process rather than a single point in time. Some individuals stop working well

before claiming Social Security and some continue to work well after claiming. Some begin

receiving retirement income while they continue to work while others delay drawing down

retirement assets until they are required to do so. Married individuals, who represent about

two-thirds of the panel, may have a spouse who receives Social Security or retirement income

before they do, or have a spouse who continues to work after they no longer do.

By age 72, most individuals in the panel received both Social Security and retirement

income. Considering both income received directly or through a spouse, 97 percent of the panel

received Social Security income and 75 percent received retirement income, with 74 percent

receiving both. Among those with the income at age 72, median per capita Social Security

income was $16,400 and median per capita retirement income was $15,300.

The progressive formula for Social Security benefits results in those with lower age 55-59

income relying more on Social Security income at age 72 and contributes to Labor+SS+Retire

income being more evenly distributed at age 72 than it is at age 55. The variation of median

retirement income across income ventiles at age 72 was similar to the variation of median labor

income at age 55. In contrast, median Social Security income varied less across ventiles than

either labor income or retirement income. As a result, Social Security benefits represented a

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

higher share of total income at age 72 for those with lower age 55-59 income while retirement

plan distributions represented a higher share of total income at age 72 for those with moderate

to moderately high age 55-59 income.

Retirement plan distributions were common across the income distribution. At age 72, the

share of the population who received retirement income either directly or through a spouse was

greater than 60 percent for the top 16 income ventiles (that is, the top 80 percent of the age 55–59

income distribution), above 80 percent for the top 12 ventiles (that is, the top 60 percent), and 90

percent or more for the top 8 ventiles (that is, the top 40 percent). For those with the income,

median per capita retirement income at age 72 ranged from under $4,000 for the three lowest

income ventiles to just over $39,000 for the highest income ventile.

5. Conclusion

In this study, we use administrative tax data to build a unique panel dataset that follows

the 1945 birth-year cohort from 2000 (when aged 55) through 2017 (when aged 72). These data

allow us to observe changes in the amount and composition of individuals’ income from before

they are eligible to claim Social Security retirement benefits until after they are eligible for

maximum Social Security benefits and generally must begin taking distributions from their

IRAs and DC plans.

We find that the typical individual maintained more than 90 percent of their age 55-59

spendable income—that is, the income available to spend after paying taxes and saving for

retirement—through age 72. Those with lower age 55-59 income typically had higher

spendable-income replacement rates. At age 72, the median spendable-income replacement rate

was greater than 100 for the bottom 25 percent of the age 55-59 income distribution, between 90

and 95 percent near the middle, and only less than 80 percent for the top 10 percent.

Over the period studied, most individuals received nearly all their income from three

sources: labor income (wage and salary, self-employment earnings, and unemployment

compensation), Social Security income (disability and retirement benefits), and retirement

income (IRA distributions and income from pensions and annuities). Although most transition

from predominately relying on labor income at younger ages to predominately relying on Social

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

Security and retirement income at older ages, that transition often takes place over a number of

years.

We find much higher incidence of (non-Social-Security) retirement income than is typically

reported in household survey data. By age 72, 67 percent receive retirement income directly and

75 percent receive the income directly or through a spouse. Retirement income is common

across the age 55-59 income distribution, with own or spouse incidence above 80 percent at age

72 for the middle 20 percent and 90 percent or more for the top 40 percent.

The median share of income from Social Security was 47 percent at age 72, with 25 percent

of the panel getting 75 percent or more of their income from Social Security, and 25 percent

getting 29 percent or less. Reflecting the design of the US Social Security system, those with

lower age 55-59 income tended to rely more on Social Security benefits in retirement while those

with higher age 55-59 income tended to rely more on retirement plan distributions.

The results of this study challenge two widely held beliefs about the US retirement system.

The data show that the US retirement system has allowed most workers to replace a substantial

share of their spendable income in retirement, with those with lower pre-retirement income

typically replacing a higher percentage. The data also show that most retirees rely on a

combination of Social Security benefits and retirement plan distributions in retirement.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

References

Bee, Adam C. and Joshua Mitchell. 2017. “Do Older Americans Have More Income Than We

Think?” SEHSD Working Paper No. 2017-39. Washington, DC: US Census Bureau Social,

Economic, and Housing Statistics Division. Available at

https://www.census.gov/content/dam/Census/library/working-papers/2017/demo/SEHSDWP2017-39.pdf.

Brady, Peter J., 2010. "Measuring retirement resource adequacy." Journal of Pension Economics and

Finance, 9 (2), 235–262.

Brady, Peter J., 2016. How America Supports Retirement: Challenging the Conventional Wisdom on

Who Benefits. Washington, DC: Investment Company Institute. Available at

https://www.ici.org/pdf/rpt_16_america_supports_retirement.pdf.

Brady, Peter J. and Steven Bass. 2020a. “Reconciling Form 1040 and Form 1099-R Data.” SOI

Working Paper. Washington, DC: Internal Revenue Service. Available at

https://www.irs.gov/pub/irs-soi/20rpreconciling10401099R.pdf.

Brady, Peter J. and Steven Bass. 2021. “Comparing the Current Population Survey to Income

Tax Data.” SOI Working Paper. Washington, DC: Internal Revenue Service. Available at

https://www.irs.gov/pub/irs-soi/21rpcomparingcpstoincometaxdata.pdf.

Brady, Peter J. and Steven Bass. 2023. “Imagine All the People: Constructing a Representative

Sample of the US Population from Tax Data.” SOI Working Paper. Washington, DC: Internal

Revenue Service.

Brady, Peter, Steven Bass, Jessica Holland, and Kevin Pierce. 2017. “Using Panel Tax Data to

Examine the Transition to Retirement.” SOI Working Paper. Washington, DC: Internal Revenue

Service. Available at https://www.irs.gov/pub/irs-soi/17rptransitionretirement.pdf.

Brady, Peter and Kevin Pierce. 2012. “The Promise and Potential Pitfalls of Using

Administrative Tax Data: A Case Study.” Unpublished paper, Investment Company Institute

(April).

Engen, Eric M., William G. Gale, and Cori E. Uccello. 2005. “Lifetime Earnings, Social Security

Benefits, and the Adequacy of Retirement Wealth Accumulation.” Social Security Bulletin, 66 (1),

38–57. Available at https://www.ssa.gov/policy/docs/ssb/v66n1/v66n1p38.html.

Federal Reserve Economic Data (FRED). St. Louis: Federal Reserve Bank of St. Louis.

Hurd, Michael D. and Susann Rohwedder. 2015. "Measuring Economic Preparation for

Retirement: Income Versus Consumption." Michigan Retirement Research Center Research Paper

no. WP 2015-332. Available at https://mrdrc.isr.umich.edu/publications/papers/pdf/wp332.pdf.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

Hurst, Eric. 2008. “Understanding Consumption in Retirement: Recent Developments.” In

Recalibrating Retirement Spending and Saving, edited by John Ameriks and Olivia S. Mitchell, 29–

45. New York: Oxford University Press.

Internal Revenue Service. 2018a. U.S. Tax Guide for Aliens: For Use in Preparing 2017 Returns,

Publication 519. Washington, DC: Internal Revenue Service. Available at

https://www.irs.gov/pub/irs-prior/p519--2017.pdf.

Internal Revenue Service. 2018b. Tax Guide for Seniors: For Use in Preparing 2017 Returns,

Publication 554. Washington, DC: Internal Revenue Service. Available at

https://www.irs.gov/pub/irs-prior/p554--2017.pdf.

Internal Revenue Service. 2018c. Tax Guide for Individuals with Income from U.S. Possessions for Use

in Preparing 2017 Returns, Publication 570. Washington, DC: Internal Revenue Service. Available

at https://www.irs.gov/pub/irs-prior/p570--2017.pdf.

Internal Revenue Service. 2017a. Exemptions, Standard Deduction, and Filing Information: For Use

in Preparing 2017 Returns, Publication 501. Washington, DC: Internal Revenue Service. Available

at https://www.irs.gov/pub/irs-prior/p501--2017.pdf.

Internal Revenue Service. 2017b. 1040 Instructions 2017. Washington, DC: Internal Revenue

Service. Available at https://www.irs.gov/pub/irs-prior/i1040gi--2017.pdf.

Meyer, Bruce D., Wallace K. C. Mok, and James X. Sullivan. 2009. “The Under-Reporting of

Transfers in Household Surveys: Its Nature and Consequences.” NBER Working Paper no.

15181 (July). Cambridge, MA: National Bureau of Economic Research. Available at

https://www.nber.org/papers/w15181.

Munnell, Alicia H. and Dina Bleckman. 2014. "Is Pension Coverage a Problem in the Private

Sector?" (April). Center for Retirement Research at Boston College Issue in Brief, no. 14-7. Available

at https://crr.bc.edu/wp-content/uploads/2014/04/IB_14-7-508.pdf.

Munnell, Alicia H., Anqi Chen, and Robert L. Siliciano. 2021. "The National Retirement Risk

Index: An Update from the 2019 SCF." Center for Retirement Research at Boston College Issue Brief,

no. 21-2 (January). Available at https://crr.bc.edu/wp-content/uploads/2021/01/IB_21-2.pdf.

Office of Personnel Management. 2011. Information for FERS Annuitants. Washington, DC: Office

of Personnel Management. Available at https://www.opm.gov/retirement-center/publicationsforms/pamphlets/ri90-8.pdf.

Rector, Robert E., Kirk A. Johnson, and Sarah E. Youssef. 1999. “The Extent of Material

Hardship and Poverty in the United States.” Review of Social Economy, 57 (3): 351–58.

Schieber, Sylvester J. 1995. “Why Do Pension Benefits Seem So Small?” Benefits Quarterly 11(4):

57–70.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

Scholz, John Karl and Ananth Seshadri. 2007. "Children and Household Wealth" (October).

Michigan Retirement Research Center Research Paper no. WP 2007-158. Available at

https://mrdrc.isr.umich.edu/publications/papers/pdf/wp158.pdf.

Scholz, John Karl, Ananth Seshadri, and Surachai Khitatrakun. 2006. “Are Americans Saving

‘Optimally’ for Retirement.” Journal of Political Economy, 114 (4): 607–643.

Social Security Administration. 2016. “Income of the Population 55 or Older, 2014.” SSA

Publication No. 13-11871. Washington, DC: Social Security Administration. Available at

https://www.ssa.gov/policy/docs/statcomps/income_pop55/2014/incpop14.pdf.

VanDerhei, Jack. 2019. "Retirement Savings Shortfalls: Evidence from EBRI’s 2019 Retirement

Security Projection Model®." EBRI Issue Brief, no. 475 (March). Available at

https://www.ebri.org/docs/default-source/ebri-issue-brief/ebri_ib_475_rspm-7mar19.pdf.

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϭ

WĂŶĞůŽŶƐŝƐƚƐŽĨ/ŶĚŝǀŝĚƵĂůƐtŚŽ^ƵƌǀŝǀĞƚŽŐĞϳϮ

^ŚĂƌĞŽĨƉŽƉƵůĂƚŝŽŶďŽƌŶŝŶϭϵϰϱĂŶĚĂůŝǀĞŽŶ:ĂŶƵĂƌǇϭ͕ϮϬϬϬďǇĂŐĞͬǇĞĂƌ;ƉĞƌĐĞŶƚĂŐĞͿ

ĞĂĚĂƚǇĞĂƌͲĞŶĚ

ůŝǀĞĂƚǇĞĂƌͲĞŶĚ

ϭϬϬ

ϭϵ

ϴϬ

ϲϬ

ϵϵ

ϴϭ

ϰϬ

ϮϬ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

&ŝŐƵƌĞϮ

^ŚĂƌĞŽĨƚŚĞWĂŶĞůtŚŽ&ŝůĞĂZĞƚƵƌŶ&ĂůůƐǁŝƚŚŐĞ

ŽŵƉŽƐŝƚŝŽŶŽĨƉĂŶĞů;ĂŐĞϱϱŝŶϮϬϬϬĂŶĚĂůŝǀĞĂƚĞŶĚŽĨϮϬϭϳͿďǇƐĂŵƉůĞĐŽŵƉŽŶĞŶƚĂŶĚĂŐĞͬǇĞĂƌ;ƉĞƌĐĞŶƚĂŐĞͿ

EŽŶͲĚĞƉĞŶĚĞŶƚŶŽŶĨŝůĞƌ

ĞƉĞŶĚĞŶƚŶŽŶĨŝůĞƌ

&ŝůĞƌ

ϭϬϬ

ϮϮ

ϴϬ

ϲϬ

ϵϰ

ϰϬ

ϳϲ

ϮϬ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϯ

DŽƐƚŝŶWĂŶĞůDĂƌƌŝĞĚƵƚDĂƌƌŝĞĚ^ŚĂƌĞ&ĂůůƐǁŝƚŚŐĞ

ŽŵƉŽƐŝƚŝŽŶŽĨƉĂŶĞů;ĂŐĞϱϱŝŶϮϬϬϬĂŶĚĂůŝǀĞĂƚĞŶĚŽĨϮϬϭϳͿďǇŵĂƌŝƚĂůƐƚĂƚƵƐ;ƉĞƌĐĞŶƚĂŐĞͿ

EŽƚŵĂƌƌŝĞĚ

DĂƌƌŝĞĚ

ϭϬϬ

ϴϬ

ϯϭ

ϯϳ

ϲϵ

ϲϯ

ϲϬ

ϰϬ

ϮϬ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

EŽƚĞ͗/ŶĂŐŝǀĞŶǇĞĂƌ͕ŶŽŶĨŝůĞƌƐĂƌĞĐůĂƐƐŝĨŝĞĚĂƐΗŶŽƚŵĂƌƌŝĞĚΗƵŶůĞƐƐƚŚĞǇǁĞƌĞŽďƐĞƌǀĞĚĨŝůŝŶŐĂũŽŝŶƚϭϬϰϬŝŶĂ

ƐĞƉĂƌĂƚĞǇĞĂƌǁŝƚŚĂƐƉŽƵƐĞǁŚŽŝƐƐƚŝůůĂůŝǀĞŝŶƚŚĞǇĞĂƌŝŶĚŝĐĂƚĞĚ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϰ

DŽƐƚ/ŶĚŝǀŝĚƵĂůƐ,ĂǀĞ/ŶĐŽŵĞdŚƌŽƵŐŚŽƵƚƚŚĞWĂŶĞů

^ŚĂƌĞŽĨƉĂŶĞůǁŝƚŚŽǁŶŽƌƐƉŽƵƐĞƚŽƚĂůŝŶĐŽŵĞďǇĂŐĞͬǇĞĂƌ;ƉĞƌĐĞŶƚĂŐĞͿ

ϭϬϬ

ϵϵ͘ϭ

ϵϵ

ϵϴ͘ϱ

ϵϴ

ϵϳ

ϵϲ

ϵϱ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘ĐĐŽƌĚŝŶŐƚŽƚŚĞEZ ƐƵƐŝŶĞƐƐǇĐůĞĂƚŝŶŐŽŵŵŝƚƚĞĞ͕ƌĞĐĞƐƐŝŽŶƐ

ŽĐĐƵƌƌĞĚĨƌŽŵDĂƌĐŚƚŽEŽǀĞŵďĞƌϮϬϬϭĂŶĚĨƌŽŵĞĐĞŵďĞƌϮϬϬϳƚŽ:ƵŶĞϮϬϬϵ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

&ŝŐƵƌĞϱ

^ƉĞŶĚĂďůĞ/ŶĐŽŵĞĞĐůŝŶĞƐďǇ>ĞƐƐdŚĂŶdŽƚĂů/ŶĐŽŵĞĞĐĂƵƐĞdĂǆĞƐĞĐůŝŶĞ

DĞĚŝĂŶƉĞƌĐĂƉŝƚĂŝŶĨůĂƚŝŽŶͲĂĚũƵƐƚĞĚĂŵŽƵŶƚƐĨŽƌƚŚŽƐĞǁŝƚŚƚŽƚĂůŝŶĐŽŵĞďǇĂŐĞͬǇĞĂƌ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

dŽƚĂůŝŶĐŽŵĞ

^ƉĞŶĚĂďůĞŝŶĐŽŵĞ

dŽƚĂůĨĞĚĞƌĂůƚĂǆ

ϱϬ

ϰϰ͘ϭ

ϰϬ

ϯϰ͘ϴ

ϯϯ͘ϰ

ϯϲ͘ϴ

ϯϬ

ϮϬ

ϭϬ

ϳ͘ϭ

ϭ͘ϭ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ΎDĞĚŝĂŶƐĂƌĞĂƌĞĐĂůĐƵůĂƚĞĚĨŽƌŝŶĚŝǀŝĚƵĂůƐǁŝƚŚŶŽŶͲǌĞƌŽƚŽƚĂůŝŶĐŽŵĞŝŶĂŐŝǀĞŶǇĞĂƌ͕ƌĞŐĂƌĚůĞƐƐŽĨǁŚĞƚŚĞƌƚŚĞǇ

ŚĂǀĞƐƉĞŶĚĂďůĞŝŶĐŽŵĞŽƌƚĂǆĞƐ͘

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϲ

&ĞĚĞƌĂůdĂǆĞƐ^ƵďƐƚĂŶƚŝĂůůǇ>ŽǁĞƌŝŶZĞƚŝƌĞŵĞŶƚ

&ĞĚĞƌĂůƚĂǆƌĂƚĞƐΎĨŽƌƚŚŽƐĞǁŝƚŚƚŽƚĂůŝŶĐŽŵĞďǇĂŐĞͬǇĞĂƌ;ƉĞƌĐĞŶƚͿ

ϳϱƚŚƉĞƌĐĞŶƚŝůĞ

DĞĚŝĂŶ

ϮϱƚŚƉĞƌĐĞŶƚŝůĞ

Ϯϰ

ϮϬ

dŽƚĂůĨĞĚĞƌĂůƚĂǆƌĂƚĞ

Ϯϭ͘Ϯ

ϭϳ͘ϭ

ϭϲ

ϭϮ ϭϭ͘ϭ

ϭϬ͘Ϯ

ϯ͘ϯ

Ϭ͘Ϭ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ϭϴ

ϭϱ

&ĞĚĞƌĂůŝŶĐŽŵĞƚĂǆƌĂƚĞ

ϭϰ͘ϵ

ϭϮ

ϭϬ͘ϱ

ϱ͘ϵ

ϵ͘Ϭ

Ϯ͘ϭ

Ϭ͘Ϭ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

WĂǇƌŽůůƚĂǆƌĂƚĞ

ϴ ϳ͘ϳ

ϲ͘ϲ

ϯ͘Ϯ

Ϭ͘ϵ

Ϭ͘Ϭ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ΎdĂǆƌĂƚĞƐĂƌĞĐĂůĐƵůĂƚĞĚĨŽƌĞĂĐŚŝŶĚŝǀŝĚƵĂůĂůŝǀĞĂƚǇĞĂƌͲĞŶĚϮϬϭϳĂƐƚŚĞŝƌĂǀĞƌĂŐĞĞĨĨĞĐƚŝǀĞƚĂǆƌĂƚĞ ƚŚĂƚŝƐ͕ƚŚĞ

ƌĂƚĞƐĂƌĞĐĂůĐƵůĂƚĞĚĂƐƚĂǆĞƐƉĂŝĚĚŝǀŝĚĞĚďǇƚŽƚĂůŝŶĐŽŵĞ͘DĞĚŝĂŶƐĂƌĞƚŚĞŵĞĚŝĂŶƌĂƚĞĂĐƌŽƐƐŝŶĚŝǀŝĚƵĂůƐǁŝƚŚ

ŶŽŶͲǌĞƌŽƚŽƚĂůŝŶĐŽŵĞ͕ƌĞŐĂƌĚůĞƐƐŽĨǁŚĞƚŚĞƌƚŚĞǇŚĂǀĞĨĞĚĞƌĂůŽƌƉĂǇƌŽůůƚĂǆĞƐ͘

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϳ

,ŝŐŚĞƌ/ŶĐŽŵĞ/ŶĚŝǀŝĚƵĂůƐDŽƌĞ>ŝŬĞůǇƚŽ^ƵƌǀŝǀĞƚŽŐĞϳϮ

^ŚĂƌĞŽĨƉĂŶĞůĂŶĚƉĞƌĐĞŶƚĂŐĞŽĨŝŶĚŝǀŝĚƵĂůƐƐƵƌǀŝǀŝŶŐƚŽĂŐĞϳϮďǇŝŶĐŽŵĞǀĞŶƚŝůĞΎ

^ŚĂƌĞŽĨƉĂŶĞů;ůĞĨƚĂǆŝƐͿ

ŐĞϳϮƐƵƌǀŝǀĂůƌĂƚĞ;ƌŝŐŚƚĂǆŝƐͿ

ϲ͘Ϭ

ϰ͘ϱ

ϵϮ͘ϵ

ϳϰ͘ϴ

ϭϬϬ͘Ϭ

ϳϱ͘Ϭ

ϯ͘Ϭ

ϱ͘ϱ

ϱϬ͘Ϭ

ϰ͘ϰ

ϭ͘ϱ

Ϭ͘Ϭ

Ϯϱ͘Ϭ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ϭϴ

ϭϵ

ϮϬ

Ϭ͘Ϭ

/ŶĐŽŵĞǀĞŶƚŝůĞΎ

Ύ/ŶĚŝǀŝĚƵĂůƐĂůŝǀĞĂƚĂŐĞϱϵĂƌĞƌĂŶŬĞĚďǇƚŚĞŝƌĂǀĞƌĂŐĞƚŽƚĂůŝŶĐŽŵĞďĞƚǁĞĞŶĂŐĞƐϱϱĂŶĚϱϵ͘/ŶĚŝǀŝĚƵĂůƐǁŝƚŚƉŽƐŝƚŝǀĞ

ĂǀĞƌĂŐĞƚŽƚĂůŝŶĐŽŵĞĂƌĞƐƉůŝƚŝŶƚŽϮϬĞƋƵĂůůǇƐŝǌĞĚŐƌŽƵƉƐŽƌΗǀĞŶƚŝůĞƐΗǁŝƚŚǀĞŶƚŝůĞϭŚĂǀŝŶŐƚŚĞůŽǁĞƐƚŝŶĐŽŵĞĂŶĚ

ǀĞŶƚŝůĞϮϬŚĂǀŝŶŐƚŚĞŚŝŐŚĞƐƚŝŶĐŽŵĞ͘dŚĞϭ͘ϬƉĞƌĐĞŶƚŽĨƚŚĞƉĂŶĞůǁŝƚŚǌĞƌŽŽƌŶĞŐĂƚŝǀĞĂǀĞƌĂŐĞƚŽƚĂůŝŶĐŽŵĞ

ďĞƚǁĞĞŶĂŐĞƐϱϱĂŶĚϱϵĂƌĞŝŶĐůƵĚĞĚŝŶƚŚĞƚŽƚĂůƐďƵƚŶŽƚƉƌĞƐĞŶƚĞĚƐĞƉĂƌĂƚĞůǇŚĞƌĞ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϴ

dŽƚĂů/ŶĐŽŵĞĞĐůŝŶĞƐƚŚĞDŽƐƚtŝƚŚŐĞ&Žƌ,ŝŐŚĞƐƚ/ŶĐŽŵĞ

DĞĚŝĂŶƉĞƌĐĂƉŝƚĂŝŶĨůĂƚŝŽŶͲĂĚũƵƐƚĞĚƚŽƚĂůŝŶĐŽŵĞΎďǇĂŐĞͬǇĞĂƌĂŶĚŝŶĐŽŵĞǀĞŶƚŝůĞ

;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

/ŶĐŽŵĞǀĞŶƚŝůĞ͗

ϮϱϬ

ϭϬ

ϭϭ

ϭϱ

ϮϬ

ϮϬϬ

ϭϱϬ

ϭϬϬ

ϱϬ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ΎDĞĚŝĂŶƐĂƌĞĂŵŽŶŐŝŶĚŝǀŝĚƵĂůƐǁŝƚŚƚŽƚĂůŝŶĐŽŵĞ͘

&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞŝŶĐŽŵĞǀĞŶƚŝůĞƐ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϳ͘

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϵĂ

dĂǆĞƐ&ĂůůƚŚĞDŽƐƚŝŶƚŚĞDŝĚĚůĞŽĨƚŚĞ/ŶĐŽŵĞŝƐƚƌŝďƵƚŝŽŶ

dŽƚĂůĨĞĚĞƌĂůƚĂǆƌĂƚĞƐΎďǇĂŐĞͬǇĞĂƌĂŶĚŝŶĐŽŵĞǀĞŶƚŝůĞ ;ƉĞƌĐĞŶƚĂŐĞͿ

ϳϱƚŚƉĞƌĐĞŶƚŝůĞ

DĞĚŝĂŶ

ϮϱƚŚƉĞƌĐĞŶƚŝůĞ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ŐĞ

zĞĂƌ

ϭϯ

ϭϳ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϭϬ

ϭϰ

ϭϴ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϭϭ

ϭϱ

ϭϵ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϯϮ

Ϯϰ

ϭϲ

ϭϮ

ϭϲ

ϮϬ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

Ύ&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞƚĂǆƌĂƚĞĐĂůĐƵůĂƚŝŽŶ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϲ͘DĞĚŝĂŶƐĂŶĚƉĞƌĐĞŶƚŝůĞƐĂƌĞĂŵŽŶŐŝŶĚŝǀŝĚƵĂůƐǁŝƚŚƚŽƚĂůŝŶĐŽŵĞ͘

&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞŝŶĐŽŵĞǀĞŶƚŝůĞƐ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϳ͘

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϵď

dĂǆĞƐ&ĂůůƚŚĞDŽƐƚŝŶƚŚĞDŝĚĚůĞŽĨƚŚĞ/ŶĐŽŵĞŝƐƚƌŝďƵƚŝŽŶ

&ĞĚĞƌĂůŝŶĐŽŵĞƚĂǆƌĂƚĞƐΎďǇĂŐĞͬǇĞĂƌĂŶĚŝŶĐŽŵĞǀĞŶƚŝůĞ ;ƉĞƌĐĞŶƚĂŐĞͿ

ϳϱƚŚƉĞƌĐĞŶƚŝůĞ

DĞĚŝĂŶ

ϮϱƚŚƉĞƌĐĞŶƚŝůĞ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

ϭϯ

ϭϳ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

ϭϬ

ϭϰ

ϭϴ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

ϭϭ

ϭϱ

ϭϵ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

ϭϮ

ϭϲ

ϮϬ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

ŐĞ ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

Ϯϴ

Ϯϭ

ϭϰ

Ͳϳ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

zĞĂƌ

&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞŝŶĐŽŵĞǀĞŶƚŝůĞƐ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϳ͘

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϵĐ

dĂǆĞƐ&ĂůůƚŚĞDŽƐƚŝŶƚŚĞDŝĚĚůĞŽĨƚŚĞ/ŶĐŽŵĞŝƐƚƌŝďƵƚŝŽŶ

WĂǇƌŽůůƚĂǆƌĂƚĞƐΎďǇĂŐĞͬǇĞĂƌĂŶĚŝŶĐŽŵĞǀĞŶƚŝůĞ ;ƉĞƌĐĞŶƚĂŐĞͿ

ϳϱƚŚƉĞƌĐĞŶƚŝůĞ

DĞĚŝĂŶ

ϮϱƚŚƉĞƌĐĞŶƚŝůĞ

ŐĞ

zĞĂƌ

ϭϯ

ϭϳ

ϭϬ

ϭϰ

ϭϴ

ϭϭ

ϭϱ

ϭϵ

ϭϮ

ϭϲ

ϮϬ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞŝŶĐŽŵĞǀĞŶƚŝůĞƐ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϳ͘

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϭϬ

>ĂƌŐĞƐƚ^ƉĞŶĚĂďůĞ/ŶĐŽŵĞĞĐůŝŶĞƐŵŽŶŐ,ŝŐŚĞƐƚ/ŶĐŽŵĞĞƐƉŝƚĞZĞĚƵĐƚŝŽŶŝŶdĂǆĞƐ

DĞĚŝĂŶƉĞƌĐĂƉŝƚĂŝŶĨůĂƚŝŽŶͲĂĚũƵƐƚĞĚŝŶĐŽŵĞΎďǇĂŐĞͬǇĞĂƌĂŶĚŝŶĐŽŵĞǀĞŶƚŝůĞ ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

dŽƚĂůŝŶĐŽŵĞ

^ƉĞŶĚĂďůĞŝŶĐŽŵĞ

ϭϮ

Ϯϰ

ϭϴ

ϭϮ

ϰϬ

ϯϬ

ϮϬ

ϭϬ

ϱϮ

ϯϵ

Ϯϲ

ϭϯ

ϴϬ

ϲϬ

ϰϬ

ϮϬ

ŐĞ

zĞĂƌ

ϭϯ

ϭϳ

ϭϮ

Ϯϴ

Ϯϭ

ϭϰ

ϰϬ

ϯϬ

ϮϬ

ϭϬ

ϱϲ

ϰϮ

Ϯϴ

ϭϰ

ϵϮ

ϲϵ

ϰϲ

Ϯϯ

ϭϬ

ϭϰ

ϭϴ

ϭϲ

ϭϮ

ϯϮ

Ϯϰ

ϭϲ

ϰϰ

ϯϯ

ϮϮ

ϭϭ

ϲϰ

ϰϴ

ϯϮ

ϭϲ

ϭϮϬ

ϵϬ

ϲϬ

ϯϬ

ϭϭ

ϭϱ

ϭϵ

ϮϬ

ϭϱ

ϭϬ

ϯϲ

Ϯϳ

ϭϴ

ϰϴ

ϯϲ

Ϯϰ

ϭϮ

ϳϮ

ϱϰ

ϯϲ

ϭϴ

ϮϮϰ

ϭϲϴ

ϭϭϮ

ϱϲ

ϭϮ

ϭϲ

ϮϬ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

ΎDĞĚŝĂŶƐĂƌĞĂŵŽŶŐŝŶĚŝǀŝĚƵĂůƐǁŝƚŚƚŽƚĂůŝŶĐŽŵĞ͘

&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞŝŶĐŽŵĞǀĞŶƚŝůĞƐ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϳ͘

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϭϭ

dǇƉŝĐĂů/ŶĚŝǀŝĚƵĂůZĞƉůĂĐĞƐDŽƌĞdŚĂŶϵϬWĞƌĐĞŶƚŽĨ^ƉĞŶĚĂďůĞ/ŶĐŽŵĞ

^ƉĞŶĚĂďůĞŝŶĐŽŵĞƌĞƉůĂĐĞŵĞŶƚƌĂƚĞΎďǇĂŐĞͬǇĞĂƌ;ƉĞƌĐĞŶƚĂŐĞͿ

ϳϱƚŚƉĞƌĐĞŶƚŝůĞ

DĞĚŝĂŶ

ϮϱƚŚƉĞƌĐĞŶƚŝůĞ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϰϬ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ΎdŚĞƐƉĞŶĚĂďůĞŝŶĐŽŵĞƌĞƉůĂĐĞŵĞŶƚƌĂƚĞŝƐŝŶĨůĂƚŝŽŶͲĂĚũƵƐƚĞĚƐƉĞŶĚĂďůĞŝŶĐŽŵĞĂƚĂŐŝǀĞŶĂŐĞĂƐĂƉĞƌĐĞŶƚĂŐĞŽĨ

ĂǀĞƌĂŐĞŝŶĨůĂƚŝŽŶͲĂĚũƵƐƚĞĚƐƉĞŶĚĂďůĞŝŶĐŽŵĞďĞƚǁĞĞŶĂŐĞϱϱĂŶĚϱϵ͘/ŶĚŝǀŝĚƵĂůƐǁŝƚŚǌĞƌŽŽƌŶĞŐĂƚŝǀĞĂǀĞƌĂŐĞĂŐĞ

ϱϱ ϱϵŝŶĐŽŵĞĂƌĞĞǆĐůƵĚĞĚĨƌŽŵƚŚĞƌĞƐƵůƚƐƉƌĞƐĞŶƚĞĚŚĞƌĞ͘

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϭϮ

>ŽǁĞƌ/ŶĐŽŵĞdǇƉŝĐĂůůǇDĂŝŶƚĂŝŶ,ŝŐŚĞƌ^ŚĂƌĞŽĨ^ƉĞŶĚĂďůĞ/ŶĐŽŵĞ

^ƉĞŶĚĂďůĞŝŶĐŽŵĞƌĞƉůĂĐĞŵĞŶƚƌĂƚĞΎďǇĂŐĞͬǇĞĂƌĂŶĚŝŶĐŽŵĞǀĞŶƚŝůĞ ;ƉĞƌĐĞŶƚĂŐĞͿ

ϳϱƚŚƉĞƌĐĞŶƚŝůĞ

DĞĚŝĂŶ

ϮϱƚŚƉĞƌĐĞŶƚŝůĞ

ϲϰϬ

ϰϴϬ

ϯϮϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϰϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϰϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϰϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϭϯ

ϭϳ

ϮϬϬ

ϭϱϬ

ϭϬϬ

ϱϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϰϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϰϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϰϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϭϬ

ϭϰ

ϭϴ

ϮϬϬ

ϭϱϬ

ϭϬϬ

ϱϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϰϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϰϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϰϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϭϭ

ϭϱ

ϭϵ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϰϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϰϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϰϬ

ϭϲϬ

ϭϯϬ

ϭϬϬ

ϳϬ

ϰϬ

ϭϴϬ

ϭϰϬ

ϭϬϬ

ϲϬ

ϭϮ

ϭϲ

ϮϬ

ϰϬ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ϰϬ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ϰϬ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ϮϬ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

ŐĞ

zĞĂƌ

Ύ&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞƐƉĞŶĚĂďůĞŝŶĐŽŵĞƌĞƉůĂĐĞŵĞŶƚƌĂƚĞ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϭϭ͘

&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞŝŶĐŽŵĞǀĞŶƚŝůĞƐ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϳ͘

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϭϯ

,ŝŐŚĞƐƚ/ŶĐŽŵĞǆƉĞƌŝĞŶĐĞ^ŚĂƌƉĞƐƚĞĐůŝŶĞƐŝŶ^ƉĞŶĚĂďůĞ/ŶĐŽŵĞ

^ƉĞŶĚĂďůĞŝŶĐŽŵĞƌĞƉůĂĐĞŵĞŶƚƌĂƚĞΎĂƚĂŐĞϳϮďǇŝŶĐŽŵĞǀĞŶƚŝůĞ ;ƉĞƌĐĞŶƚĂŐĞͿ

ϮϱƚŚƉĞƌĐĞŶƚŝůĞ

DĞĚŝĂŶ

ϳϱƚŚƉĞƌĐĞŶƚŝůĞ

ϮϱϬ

ϮϬϬ

ϭϱϬ

ϭϬϬ

ϱϬ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ϭϴ

ϭϵ

ϮϬ

/ŶĐŽŵĞǀĞŶƚŝůĞ

Ύ&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞƐƉĞŶĚĂďůĞŝŶĐŽŵĞƌĞƉůĂĐĞŵĞŶƚƌĂƚĞ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϭϭ͘dŚĞϳϱƚŚƉĞƌĐĞŶƚŝůĞ

ƌĞƉůĂĐĞŵĞŶƚƌĂƚĞĨŽƌŝŶĚŝǀŝĚƵĂůƐŝŶƚŚĞĨŝƌƐƚŝŶĐŽŵĞǀĞŶƚŝůĞǁĂƐϲϬϰƉĞƌĐĞŶƚ͘

&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞŝŶĐŽŵĞǀĞŶƚŝůĞƐ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϳ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϭϰ

dǇƉŝĐĂů/ŶĚŝǀŝĚƵĂů'ĞƚƐƵůŬŽĨ/ŶĐŽŵĞĨƌŽŵ>ĂďŽƌ͕^ŽĐŝĂů^ĞĐƵƌŝƚǇ͕ĂŶĚZĞƚŝƌĞŵĞŶƚ

WĂŶĞůŝŶĐŝĚĞŶĐĞĂŶĚĐŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶŝŶĨůĂƚŝŽŶͲĂĚũƵƐƚĞĚŝŶĐŽŵĞďǇĂŐĞͬǇĞĂƌ

dŽƚĂůŝŶĐŽŵĞ

>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ

EŽŶͲ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ

ϭϬϬ

KǁŶŽƌƐƉŽƵƐĞŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

ϵϵ

ϵϴ

ϵϱ

ϴϬ

ϴϮ

ϲϴ

ϲϬ

ϰϬ

ϮϬ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ϱϬ

ϰϬ

DĞĚŝĂŶƉĞƌĐĂƉŝƚĂŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ϰϰ

ϰϯ

ϯϱ

ϯϬ

ϯϮ

ϮϬ

ϭϬ

Ϭ͘ϳϯ

Ϭ͘ϲϮ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϭϱ

ZĞůŝĂŶĐĞŽŶ^ŽĐŝĂů^ĞĐƵƌŝƚǇĂŶĚZĞƚŝƌĞŵĞŶƚ/ŶĐŽŵĞ/ŶĐƌĞĂƐĞƐƵƌŝŶŐdƌĂŶƐŝƚŝŽŶŝŶƚŽZĞƚŝƌĞŵĞŶƚ

WĂŶĞůŝŶĐŝĚĞŶĐĞĂŶĚĐŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶŝŶĨůĂƚŝŽŶͲĂĚũƵƐƚĞĚŝŶĐŽŵĞďǇĂŐĞͬǇĞĂƌ

KǁŶ

KǁŶŽƌƐƉŽƵƐĞ

KǁŶ

>ĂďŽƌŝŶĐŽŵĞŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

ϭϬϬ

ϵϬ

ϴϬϳϴ

ϱϬ

ϰϰ

ϰϬϰϬ

ϲϬ

ϯϬ

ϰϬ

ϮϬ

ϯϰ

Ϯϯ

ϮϬ

WĞƌĐĂƉŝƚĂ

ŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶůĂďŽƌŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ϭϰ

ϭϭ

ϭϬ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

KǁŶ

ϭϬϬ

KǁŶŽƌƐƉŽƵƐĞ

^ŽĐŝĂů^ĞĐƵƌŝƚǇŝŶĐŽŵĞŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

KǁŶ

ϵϳ

ϵϲ

Ϯϱ

ϴϬ

ϮϬ

ϲϬ

ϭϱ

ϭϮ

ϭϬ ϵ

ϰϬ

ϮϬϭϬ

WĞƌĐĂƉŝƚĂ

ŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶ^ŽĐŝĂů^ĞĐƵƌŝƚǇŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ϭϳ

ϭϲ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

KǁŶ

ϭϬϬ

KǁŶŽƌƐƉŽƵƐĞ

KǁŶ

ZĞƚŝƌĞŵĞŶƚŝŶĐŽŵĞŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

ϴϬ

Ϯϱ

ϮϬ

ϳϱ

ϲϳ

ϲϬ

ϰϬ

Ϯϳ

ϮϬϭϳ

WĞƌĐĂƉŝƚĂ

ŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶƌĞƚŝƌĞŵĞŶƚŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ϭϲ

ϭϱ

ϭϱϭϯ

ϭϭ

ϭϬ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϭϲ

ůĂŝŵŝŶŐŽĨ^ŽĐŝĂů^ĞĐƵƌŝƚǇŶŽƚůǁĂǇƐƐƐŽĐŝĂƚĞĚǁŝƚŚ^ƚŽƉƉŝŶŐtŽƌŬ

WĂŶĞůŝŶĐŝĚĞŶĐĞŽĨ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞĂŶĚĐŽŵƉŽŶĞŶƚƐďǇĂŐĞͬǇĞĂƌ;ƉĞƌĐĞŶƚĂŐĞͿ

>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ

ϭϬϬ

>ĂďŽƌŝŶĐŽŵĞ

KǁŶŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

^ŽĐŝĂů^ĞĐƵƌŝƚǇ/ŶĐŽŵĞ

ϵϴ

ϭϬϬ

ϵϲ

ϴϰ

ϴϬ

ϵϴ

ϵϱ

ϵϳ

ϵϬ

ϴϬ

ϳϴ

ZĞƚŝƌĞŵĞŶƚŝŶĐŽŵĞ

KǁŶŽƌƐƉŽƵƐĞŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

ϳϱ

ϲϳ

ϲϬ

ϰϬ

Ϯϳ

ϮϬ ϭϳ

ϮϬ

Ϯϯ

ϯϰ

ϭϬ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

>ĂďŽƌŽŶůǇ

ϭϬϬ

>ĂďŽƌƉůƵƐƌĞƚŝƌĞŵĞŶƚ

ZĞƚŝƌĞŵĞŶƚŽŶůǇ

>ĂďŽƌƉůƵƐ^^ƉůƵƐƌĞƚŝƌĞŵĞŶƚ

KǁŶŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

ϯ ϯ ϯ ϰ ϰ

ϯ ϰ

ϱ ϲ

ϭϯ ϭϰ

ϭϱ ϭϱ

ϭϲ

ϲϬ

ϴϬ

ϭϬϬ

ϭϬ ϭϭ

ϭϴ ϭϳ ϭϱ ϭϰ ϭϭ

ϭϱ

ϭϴ ϮϮ

ϱ ϭϭ

ϰϳ ϰϵ

Ϯϳ ϯϬ

ϰϬ

ϭϰ ϯ ϭϭ

ϯϯ ϯϲ ϰϮ

ϭϰ

ϲϰ ϲϭ ϱϴ

ϭϭ ϯ

ϱϱ ϱϮ

ϭϬ ϯ

ϰϱ ϰϮ

ϮϬ

ϭϬ ϭϵ

ϯϮ Ϯϲ

ϭϴ ϭϲ

ϭϱ ϭϲ ϭϳ ϭϱ

Ϯϭ ϭϱ

ŐĞ ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

zĞĂƌ

ϮϬ

ϰ ϰ ϳ ϴ ϭϬ ϭϭ

ϲ ϳ

ϱ ϲ ϭϭ ϭϯ ϭϰ ϭϰ

ϴϬ

ϭϬ

ϮϬ

ϭϮ ϭϬ ϭϰ

ϮϮ

ϭϳ Ϯϭ

Ϯϯ

ϲϬ

Ϯϯ

ϮϮ

Ϯϱ

Ϯϰ

Ϯϲ

Ϯϳ

ϰϬ

ϯϬ

ϭϲ

ϲϯ ϱϵ

ϱϱ ϱϭ

ϭϮ

ϰϳ

ϭϬ

ϰϬ ϯϲ

ϮϬ

Ϯϳ Ϯϭ

ϭϲ ϭϭ

ŐĞ ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ

ϮϬ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

>ĂďŽƌƉůƵƐ^^

^^ŽŶůǇ

KǁŶŽƌƐƉŽƵƐĞŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

Ϯϰ Ϯϰ

ϭϯ ϭϲ ϭϵ ϮϬ Ϯϰ Ϯϲ Ϯϳ Ϯϵ Ϯϳ

ϯ ϳ

ϭϬ ϭϬ

^^ƉůƵƐƌĞƚŝƌĞŵĞŶƚ

zĞĂƌ

ϭϮ ϭϯ ϭϰ ϭϱ ϭϱ ϭϱ ϭϲ

ϭϴ ϭϳ ϭϱ ϭϰ ϭϭ

Ϯϱ Ϯϵ

ϯϯ ϯϳ ϰϭ ϰϱ ϰϴ

ϯϰ

ϯϯ ϯϭ

Ϯϴ Ϯϵ Ϯϴ Ϯϲ

ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϭϳ

DŽƐƚZĞĐĞŝǀĞ>ĂďŽƌн^^нZĞƚŝƌĞ/ŶĐŽŵĞdŚƌŽƵŐŚŽƵƚWĞƌŝŽĚ^ƚƵĚŝĞĚ

WĂŶĞůŝŶĐŝĚĞŶĐĞĂŶĚĐŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶŝŶĨůĂƚŝŽŶͲĂĚũƵƐƚĞĚŝŶĐŽŵĞĨƌŽŵƚŚĞĐŽŵďŝŶĂƚŝŽŶŽĨůĂďŽƌ͕^ŽĐŝĂů^ĞĐƵƌŝƚǇ͕ĂŶĚƌĞƚŝƌĞŵĞŶƚŝŶĐŽŵĞ

KǁŶ

KǁŶŽƌƐƉŽƵƐĞ

>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

ϭϬϬϵϱ

ϴϰ

ϴϬ

KǁŶ

>ĂďŽƌн^^нZĞƚŝƌĞĐŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ϵϴ

ϱϬ

ϰϰ

ϰϯ

ϰϬ

ϲϬ

ϯϬ

ϰϬ

ϮϬ

ϭϬ

WĞƌĐĂƉŝƚĂ

ϯϮ

ϯϬ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϭϴ

,ŝŐŚĞƌͲ/ŶĐŽŵĞtŽƌŬ>ŽŶŐĞƌĂŶĚůĂŝŵ>ĂƚĞƌ

WĂŶĞůŝŶĐŝĚĞŶĐĞŽĨŝŶĐŽŵĞďǇĂŐĞͬǇĞĂƌĂŶĚŝŶĐŽŵĞǀĞŶƚŝůĞΎ

ŐĞϱϱ

ŐĞϲϮ

ŐĞϲϲ

ŐĞϳϮ

KǁŶŝŶĐŽŵĞŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

ϭϬϬ

KǁŶŽƌƐƉŽƵƐĞŝŶĐŽŵĞŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

>ĂďŽƌ

ϭϬϬ

ϴϬ

ϲϬ

ϰϬ

ϮϬ

ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ ϭϴ ϭϵ ϮϬ

>ĂďŽƌ

/ŶĐŽŵĞǀĞŶƚŝůĞΎ

ϭϬϬ

^ŽĐŝĂů^ĞĐƵƌŝƚǇ

ϭϬϬ

ϴϬ

ϲϬ

ϰϬ

ϮϬ

ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ ϭϴ ϭϵ ϮϬ

ZĞƚŝƌĞŵĞŶƚ

ϭϬϬ

ϴϬ

ϲϬ

ϰϬ

ϮϬ

ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ ϭϴ ϭϵ ϮϬ

/ŶĐŽŵĞǀĞŶƚŝůĞΎ

ϴϬ

ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ ϭϴ ϭϵ ϮϬ

^ŽĐŝĂů^ĞĐƵƌŝƚǇ

/ŶĐŽŵĞǀĞŶƚŝůĞΎ

ϭϬϬ

/ŶĐŽŵĞǀĞŶƚŝůĞΎ

ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ ϭϴ ϭϵ ϮϬ

/ŶĐŽŵĞǀĞŶƚŝůĞΎ

ZĞƚŝƌĞŵĞŶƚ

ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ ϭϴ ϭϵ ϮϬ

/ŶĐŽŵĞǀĞŶƚŝůĞΎ

Ύ&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞŝŶĐŽŵĞǀĞŶƚŝůĞƐ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϳ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϭϵ

DĞĚŝĂŶ>ĂďŽƌ/ŶĐŽŵĞ/ƐdǇƉŝĐĂůůǇ>ŽǁĞƌĨŽƌdŚŽƐĞǁŚŽŽŶƚŝŶƵĞƚŽtŽƌŬ

ŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶŝŶĨůĂƚŝŽŶͲĂĚũƵƐƚĞĚůĂďŽƌŝŶĐŽŵĞďǇĂŐĞͬǇĞĂƌĂŶĚŝŶĐŽŵĞǀĞŶƚŝůĞΎ

/ŶĐŽŵĞǀĞŶƚŝůĞ͗Ύ

ϮϬϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

KǁŶůĂďŽƌŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ϮϬϬ

ϭϲϬ

ϭϮϬ

ϴϬ

ϰϬ

ϭϲ

ϭϳ

ϭϴ

ϭϵ

ϮϬ

WĞƌĐĂƉŝƚĂůĂďŽƌŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

/ŶĐŽŵĞǀĞŶƚŝůĞ͗Ύ

ϰϬ

KǁŶůĂďŽƌŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ϰϬ

ϯϬ

ϮϬ

ϭϬ

WĞƌĐĂƉŝƚĂůĂďŽƌŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

Ύ&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞŝŶĐŽŵĞǀĞŶƚŝůĞƐ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϳ͘

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϮϬ

ŵŽƵŶƚŽĨŐĞϳϮZĞƚŝƌĞŵĞŶƚ/ŶĐŽŵĞsĂƌŝĞƐDŽƌĞĐƌŽƐƐsĞŶƚŝůĞƐdŚĂŶ^ŽĐŝĂů^ĞĐƵƌŝƚǇ

ŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶŝŶĨůĂƚŝŽŶͲĂĚũƵƐƚĞĚŝŶĐŽŵĞďǇĂŐĞͬǇĞĂƌĂŶĚŝŶĐŽŵĞǀĞŶƚŝůĞΎ

/ŶĐŽŵĞǀĞŶƚŝůĞ͗Ύ

ϱϬ

ϭϬ

KǁŶ^ŽĐŝĂů^ĞĐƵƌŝƚǇŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ϱϬ

ϰϬ

ϯϬ

ϮϬ

ϭϬ

ϭϭ

ϭϱ

ϮϬ

KǁŶƌĞƚŝƌĞŵĞŶƚŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

ϱϬ

WĞƌĐĂƉŝƚĂ^ŽĐŝĂů^ĞĐƵƌŝƚǇŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ϱϬ

ϰϬ

ϯϬ

ϮϬ

ϭϬ

WĞƌĐĂƉŝƚĂƌĞƚŝƌĞŵĞŶƚŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

Ύ&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞŝŶĐŽŵĞǀĞŶƚŝůĞƐ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϳ͘

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϮϭ

DŽƐƚ,ĂǀĞ>ĂďŽƌн^^нZĞƚŝƌĞ/ŶĐŽŵĞďǇŐĞϲϲ

WĂŶĞůŝŶĐŝĚĞŶĐĞŽĨ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞďǇĂŐĞͬǇĞĂƌĂŶĚŝŶĐŽŵĞǀĞŶƚŝůĞΎ

ŐĞϱϱ

ŐĞϲϭ

ŐĞϲϮ

ŐĞϲϲ

ŐĞϳϮ

ϭϬϬ

KǁŶŝŶĐŝĚĞŶĐĞŽĨ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ;ƉĞƌĐĞŶƚĂŐĞͿ

ϴϬ

ϲϬ

ϰϬ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ϭϴ

ϭϵ

ϮϬ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ϭϴ

ϭϵ

ϮϬ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ϭϴ

ϭϵ

ϮϬ

/ŶĐŽŵĞǀĞŶƚŝůĞΎ

ϮϬ

^ƉŽƵƐĞŽŶůǇŝŶĐŝĚĞŶĐĞŽĨ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ;ƉĞƌĐĞŶƚĂŐĞͿ

ϭϬ

ϭϭ

ϭϮ

/ŶĐŽŵĞǀĞŶƚŝůĞΎ

ϭϬϬ

KǁŶŽƌƐƉŽƵƐĞŝŶĐŝĚĞŶĐĞŽĨ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ;ƉĞƌĐĞŶƚĂŐĞͿ

ϴϬ

ϲϬ

ϰϬ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

/ŶĐŽŵĞǀĞŶƚŝůĞΎ

Ύ&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞŝŶĐŽŵĞǀĞŶƚŝůĞƐ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϳ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϮϮ

ŝĨĨĞƌĞŶĐĞƐŝŶ>ĂďŽƌн^^нZĞƚŝƌĞ/ŶĐŽŵĞĐƌŽƐƐsĞŶƚŝůĞƐĞĐůŝŶĞǁŝƚŚŐĞ

ŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶŝŶĨůĂƚŝŽŶͲĂĚũƵƐƚĞĚ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞďǇĂŐĞͬǇĞĂƌĂŶĚŝŶĐŽŵĞǀĞŶƚŝůĞΎ

/ŶĐŽŵĞǀĞŶƚŝůĞ͗Ύ

ϮϬϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

KǁŶ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ϮϬϬ

ϭϲϬ

ϭϮϬ

ϴϬ

ϰϬ

ϭϲ

ϭϳ

ϭϴ

ϭϵ

ϮϬ

WĞƌĐĂƉŝƚĂ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

/ŶĐŽŵĞǀĞŶƚŝůĞ͗Ύ

ϰϬ

KǁŶ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ϰϬ

ϯϬ

ϮϬ

ϭϬ

WĞƌĐĂƉŝƚĂ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

ŐĞ

ϱϱ ϱϲ ϱϳ ϱϴ ϱϵ ϲϬ ϲϭ ϲϮ ϲϯ ϲϰ ϲϱ ϲϲ ϲϳ ϲϴ ϲϵ ϳϬ ϳϭ ϳϮ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

zĞĂƌ

ϬϬ Ϭϭ ϬϮ Ϭϯ Ϭϰ Ϭϱ Ϭϲ Ϭϳ Ϭϴ Ϭϵ ϭϬ ϭϭ ϭϮ ϭϯ ϭϰ ϭϱ ϭϲ ϭϳ

Ύ&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞŝŶĐŽŵĞǀĞŶƚŝůĞƐ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϳ͘

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϮϯ

DĂŶǇ,ĂǀĞEŽŶͲ>ĂďŽƌн^^нZĞƚŝƌĞ/ŶĐŽŵĞƵƚŵŽƵŶƚƐdǇƉŝĐĂůůǇDŽĚĞƐƚĂŶĚǇĐůŝĐĂů

WĂŶĞůŝŶĐŝĚĞŶĐĞ͕ĐŽŶĚŝƚŝŽŶĂůŝŶĨůĂƚŝŽŶͲĂĚũƵƐƚĞĚŵĞĚŝĂŶ͕ĂŶĚŵĂƌŬĞƚƌĞƚƵƌŶƐďǇĂŐĞͬǇĞĂƌ

EŽŶͲ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ

dĂǆĂďůĞŝŶƚĞƌĞƐƚŝŶĐŽŵĞ

KƚŚĞƌŶŽŶͲ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ

ϭϬϬ

KǁŶŽƌƐƉŽƵƐĞŶŽŶͲ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

ϴϮ

ϴϬ

ϳϯ

ϲϯ

ϲϴ

ϲϬ

ϱϰ

ϱϯ

ϰϬ

ϮϬ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ϭ͕ϱϬϬ

ŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶƉĞƌĐĂƉŝƚĂŶŽŶͲ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ;ϮϬϭϳĚŽůůĂƌƐͿ

ϭ͕ϮϱϬ

ϭ͕ϬϬϬ

ϳϱϬ

ϱϬϬ

ϮϱϬ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ǀĞƌĂŐĞĨĞĚĞƌĂůĨƵŶĚƐƌĂƚĞ;ůĞĨƚĂǆŝƐͿ

ŶŶƵĂůĂǀĞƌĂŐĞƉĞƌĐĞŶƚĐŚĂŶŐĞŝŶƚŚĞtŝůƐŚŝƌĞϱϬϬϬƚŽƚĂůƌĞƚƵƌŶŝŶĚĞǆ;ƌŝŐŚƚĂǆŝƐͿ

ŶŶƵĂůŵĂƌŬĞƚƌĞƚƵƌŶƐ;ƉĞƌĐĞŶƚĂŐĞͿ

ϯϬ

ϭϱ

Ͳϰ

Ͳϭϱ

Ͳϴ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ͲϯϬ

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌĐĞƐ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂĂŶĚ&ĞĚĞƌĂůZĞƐĞƌǀĞĂŶŬŽĨ^ƚ͘>ŽƵŝƐ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϮϰ

EŽŶͲ>ĂďŽƌн^^нZĞƚŝƌĞ/ŶĐŝĚĞŶĐĞ&ĂůůƐDŽƌĞĨŽƌ>ŽǁĞƌ/ŶĐŽŵĞƵƚdĂǆĂďůĞ/ŶƚĞƌĞƐƚĞĐůŝŶĞƐtŝĚĞƐƉƌĞĂĚ

WĂŶĞůŝŶĐŝĚĞŶĐĞŽĨŶŽŶͲ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞďǇĂŐĞͬǇĞĂƌĂŶĚŝŶĐŽŵĞǀĞŶƚŝůĞΎ

ŐĞϱϱ

ŐĞϲϭ

ŐĞϲϮ

ŐĞϲϲ

ŐĞϳϮ

ϭϬϬ

KǁŶŽƌƐƉŽƵƐĞŶŽŶͲ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

ϴϬ

ϲϬ

ϰϬ

ϮϬ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ϭϴ

ϭϵ

ϮϬ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ϭϴ

ϭϵ

ϮϬ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ϭϴ

ϭϵ

ϮϬ

/ŶĐŽŵĞǀĞŶƚŝůĞ

ϭϬϬ

KǁŶŽƌƐƉŽƵƐĞƚĂǆĂďůĞŝŶƚĞƌĞƐƚŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

ϴϬ

ϲϬ

ϰϬ

ϮϬ

ϭϬ

ϭϭ

/ŶĐŽŵĞǀĞŶƚŝůĞ

ϭϬϬ

KǁŶŽƌƐƉŽƵƐĞŽƚŚĞƌŶŽŶͲ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŝĚĞŶĐĞ;ƉĞƌĐĞŶƚĂŐĞͿ

ϴϬ

ϲϬ

ϰϬ

ϮϬ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

/ŶĐŽŵĞǀĞŶƚŝůĞ

Ύ&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞŝŶĐŽŵĞǀĞŶƚŝůĞƐ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϳ͘

^ŽƵƌĐĞ͗ƵƚŚŽƌƐ ƚĂďƵůĂƚŝŽŶŽĨ/Z^ĚĂƚĂ

Brady and Bass

May 10, 2023

When I’m 64 (or Thereabouts): Changes in Income from Middle Age to Old Age

&ŝŐƵƌĞϮϱ

EŽŶͲ>ĂďŽƌн^^нZĞƚŝƌĞ/ŶĐŽŵĞŵŽƵŶƚƐ^ƵďƐƚĂŶƚŝĂůKŶůǇĨŽƌdŚŽƐĞǁŝƚŚ,ŝŐŚĞƐƚŐĞϱϱͲϱϵ/ŶĐŽŵĞ

ŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶŝŶĨůĂƚŝŽŶͲĂĚũƵƐƚĞĚŶŽŶͲ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞďǇĂŐĞͬǇĞĂƌĂŶĚŝŶĐŽŵĞǀĞŶƚŝůĞΎ

;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

/ŶĐŽŵĞǀĞŶƚŝůĞ͗Ύ

ϱϬ

ϭϬ

ϭϭ

ϭϱ

ϮϬ

ŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶƉĞƌĐĂƉŝƚĂŶŽŶͲ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ϰϬ

ϯϬ

ϮϬ

ϭϬ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶƉĞƌĐĂƉŝƚĂƚĂǆĂďůĞŝŶƚĞƌĞƐƚŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

ϰϬ

ŽŶĚŝƚŝŽŶĂůŵĞĚŝĂŶƉĞƌĐĂƉŝƚĂŽƚŚĞƌŶŽŶͲ>ĂďŽƌн^^нZĞƚŝƌĞŝŶĐŽŵĞ;ƚŚŽƵƐĂŶĚƐŽĨϮϬϭϳĚŽůůĂƌƐͿ

ϯϮ

Ϯϰ

ϭϲ

ŐĞ

ϱϱ

ϱϲ

ϱϳ

ϱϴ

ϱϵ

ϲϬ

ϲϭ

ϲϮ

ϲϯ

ϲϰ

ϲϱ

ϲϲ

ϲϳ

ϲϴ

ϲϵ

ϳϬ

ϳϭ

ϳϮ

zĞĂƌ

ϬϬ

Ϭϭ

ϬϮ

Ϭϯ

Ϭϰ

Ϭϱ

Ϭϲ

Ϭϳ

Ϭϴ

Ϭϵ

ϭϬ

ϭϭ

ϭϮ

ϭϯ

ϭϰ

ϭϱ

ϭϲ

ϭϳ

Ύ&ŽƌĂĚĞƐĐƌŝƉƚŝŽŶŽĨƚŚĞŝŶĐŽŵĞǀĞŶƚŝůĞƐ͕ƐĞĞƚŚĞŶŽƚĞŽŶ&ŝŐƵƌĞϳ͘

EŽƚĞ͗'ƌĂǇƐŚĂĚŝŶŐŝŶĚŝĐĂƚĞƐĂƌĞĐĞƐƐŝŽŶ͘^ĞĞŶŽƚĞŽŶ&ŝŐƵƌĞϰ͘

^ŽƵƌ

This text is long and has been trimmed here. Open the source document for the complete record.

This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.