Report | Retirement

Retirement Inequality Chartbook The State of American Retirement: How 401(k)s have failed most American workers

Retirement Inequality Chartbook

Download PDF

Press release

Overview

Today, many Americans rely on savings in 401(k)-type accounts to supplement Social Security in retirement. This is a pronounced shift from a few decades ago, when many retirees could count on predictable, constant streams of income from traditional pensions (see “Types of retirement plans,” below). This chartbook assesses the impact of the shift from pensions to individual savings by examining disparities in retirement preparedness and outcomes by income, race, ethnicity, education, gender, and marital status.

The first section of the chartbook looks at retirement-plan participation and retirement account savings of working-age families. The charts in this section focus on families headed by someone age 32–61, a 30-year period before the Social Security early eligibility age of 62 when most families should be accumulating pension benefits and retirement savings. The second section looks at income sources for seniors. Since many workers transition to retirement between Social Security’s early eligibility age and the program’s normal retirement age (currently 66, formerly 65), the charts in the second section focus on retirement outcomes of people age 65 and older.

Jump to the charts

Types of retirement plans

401(k) and similar plans are referred to as defined-contribution (DC) plans because employer contributions, rather than retirement benefits, are determined in advance and employers incur no long-term liabilities. Participants in these plans are responsible for making investment decisions and shoulder investment and other risks. In contrast, in traditional defined-benefit (DB) plans (pension plans, in layman’s terms), employers are responsible for funding promised benefits, making up the difference if the contributions are insufficient due to lower-than-expected investment returns, for example.

401(k)s are an accident of history. In 1980, a benefit consultant working on revamping a bank’s cash bonus plan had the idea of adding an employer matching contribution and taking advantage of an obscure provision in the tax code passed two years earlier clarifying the tax treatment of deferred compensation. Though 401(k)s took off in the early 1980s, Congress did not intend for them to replace traditional pensions as a primary retirement vehicle, and 401(k)s are poorly designed for this role (Sahadi 2001; Tong 2013).

The term “defined-contribution” is somewhat misleading because employers may not contribute anything to these plans, and employer contributions most often take the form of matching contributions contingent on employee contributions. In contrast, under traditional defined-benefit plans in the private sector, employers are generally responsible for the entire cost, though public-sector workers often share in pension costs.

Because they are employer-sponsored plans, defined-contribution plans are usually differentiated from Individual Retirement Accounts (IRAs). However, the line between employer-sponsored and individual plans is blurry because employers are not required to contribute anything to employee 401(k) accounts, because most funds in IRAs are rolled over from 401(k)s, and because employers do contribute to some types of IRAs.

Like defined-benefit plans, defined-contribution plans and IRAs receive preferential tax treatment intended to encourage employers to provide retirement benefits and help individuals to save for retirement. However, tax incentives for retirement savings are poorly targeted and ineffectual, as most of the subsidies go to high-income taxpayers who steer savings to tax-favored accounts rather than increase the amount they save (see, for example, Chetty et al. 2014).

Throughout the chartbook, we use “retirement account savings” to refer to savings in defined-contribution plans (such as 401(k)s), IRAs, and other plans in which participants accrue account balances, such as Keogh plans used in the past by self-employed workers. We reserve the word “pension” for benefits that take the form of income streams starting at retirement and ending when beneficiaries die. While some 401(k) participants may opt to convert account balances to life annuities, and some pension beneficiaries opt to withdraw lump sums at retirement, neither is the normal payout option for these plans.

Two of the charts refer to a family’s “participation” in an employer-based retirement plan, which means that at least one worker in the family (the survey respondent, spouse, or both) currently has access to and is signed up for a plan, not necessarily that the family has accumulated any benefits or balances in the plan. Conversely, in charts showing the share of families with retirement account savings, respondents and their spouses may or may not be currently participating in a plan—account holdings could be from past participation. The phrase “active participants” is used, when appropriate, to exclude retirees.

Part One: How will working families fare in retirement?

The first section of the chartbook looks at the retirement prospects of working-age families, focusing especially on retirement account savings. Except for one, all charts in this section are based on data from the Federal Reserve Board’s Survey of Consumer Finances (the first chart uses the Federal Reserve Board’s Financial Accounts data). In the Survey of Consumer Finances, a family consists of an “economically dominant” single person or couple, whether married or living together as partners, and all other persons in a household who are financially interdependent with that person or couple. The family’s age and education level are based on the age and education of the male in a mixed-sex couple or older spouse in a same-sex couple (Bricker et al. 2014).

Most of the charts focus on retirement account savings, a measure that includes savings in 401(k)-style defined-contribution (DC) plans, IRAs, and Keogh plans for self-employed people and small-business owners. The measure excludes assets held by defined-benefit pension funds, which are not account-type plans.

In addition to other demographic factors, the charts show trends in retirement preparedness by six-year age group or birth cohort from 1989 to 2013. Six-year groups were chosen because the Survey of Consumer Finances is conducted every three years, but six-year groups produce larger sample sizes. All charts use inflation-adjusted dollars and, where possible, are shown on comparable scales. Dollar amounts in charts may reflect rounding by survey respondents.

Key findings of the following charts include:

Retirement wealth has not grown fast enough to keep pace with an aging population and other changes. The first chart offers what at first appears to be an encouraging picture, the growth since 1989 in retirement wealth—assets in pension funds plus savings in retirement accounts—relative to income. Unlike other charts in this section, this measure is for the entire population, not just working-age families. As Figure 1 shows, retirement wealth more than kept pace with incomes over the past quarter century, growing faster than income in the 1990s and rebounding after two stock market downturns in the 2000s. Retirement wealth nearly doubled as a share of personal disposable income between 1989 and 2013, with retirement account savings exceeding pension fund assets after 2012 (and briefly in the late 1990s and mid-2000s). The shift in wealth from pension funds to retirement accounts occurred years after participation in defined-contribution plans surpassed that in defined-benefit plans (Figure 2).

What Figure 1 does not show is that retirement wealth should have increased more to keep pace with an aging population, offset Social Security cuts, and serve as a hedge against the increased longevity risks and investment risks brought on by a shift from traditional pensions to individual savings. Retirement account savings increased before the Great Recession as the large baby boomer cohort approached retirement. However, retirement account savings by age group stagnated or declined in the new millennium even as traditional pension coverage continued to decline (Figures 2-5). Meanwhile, Social Security benefits are replacing a declining share of pre-retirement earnings due to benefit cuts passed in 1983 that are gradually taking effect (Reno, Bethell, and Walker 2011). The change in plan type should have been accompanied by an increase in retirement assets to account for the diminishing use of pooled pension funds, which benefit from economies of scale and risk pooling and are thus more cost-effective than individual accounts. In other words, in a retirement savings account system, people need to set aside more, because these accounts are not as efficient as pensions.

The shift from traditional pensions to individual savings has widened retirement gaps. In addition to retirement wealth not growing fast enough, retirement disparities have grown with the shift from traditional pensions to retirement savings accounts. These disparities are the main focus of this chartbook. As Figure 6 shows, high-income, white, college-educated, and married workers participate in defined-benefit pensions at a higher rate than other workers, but participation gaps are much larger under defined-contribution plans. The distribution of savings in retirement accounts is even more unequal than participation in these plans (Figure 7). There are large differences between mean and median retirement savings because mean savings are skewed by large balances for a few families (Figure 8). For many groups—lower-income, black, Hispanic, non-college-educated, and unmarried Americans—the typical working-age family or individual has no savings at all in retirement accounts, and for those that do have savings, the median balances in retirement accounts are very low (Figures 9–15).

In theory, the shift from defined-benefit to defined-contribution plans could have broadened access to retirement benefits by making it easier and cheaper for employers to offer benefits. However, participation in all employer-based retirement plans has declined in the new millennium (Figure 2). Retirement inequality has grown because most 401(k) participants are required to contribute to these plans in order to participate, whereas workers are automatically enrolled in traditional pensions and, in the private sector, are not required to contribute to these plans. Thus, higher-income workers (with their greater capacity to make contributions) are more likely to participate in defined-contribution plans.

In addition to their greater disposable income, higher-income workers have a higher investment-risk tolerance, receive larger tax breaks for saving, and are more likely to work for employers that offer plans and provide generous matches (CBO 2013; Morrissey 2009). 401(k) and IRA contribution limits, a Saver’s Credit targeted at low- and moderate-income families, and other attempts to ensure that tax subsidies for retirement do not disproportionately flow to high-income families have proven ineffective at leveling the playing field.

Workers’ retirement prospects are increasingly affected by economic downturns. Much of the 401(k) era coincided with rising stock and housing prices that propped up family wealth measures even as the savings rate declined. This house of cards collapsed in 2000–2001 and again in 2007–2009. In 2013 most families still had not recovered their losses from the financial crisis and Great Recession, let alone accumulated additional savings for retirement (Figures 5, 7, 8).

Successive generations should be saving more in defined-contribution plans due to declining defined-benefit pension coverage, Social Security benefit cuts, and higher incomes. While the retirement account savings of families approaching retirement grew before the financial crisis and Great Recession, those of younger families stayed flat. But families approaching retirement had more to lose—and did—when asset prices collapsed in 2007–2009 (Figures 3–5, 16). Recessions can be very damaging to workers nearing retirement, since they have less time to make up losses and their retirement outcomes are influenced more by investment returns than new contributions. In addition, many older workers who lose jobs tap retirement savings.

Families with large account balances saw large dollar declines after 2007. But in percentage terms, the financial crisis and ensuing recession had a greater impact on families with small balances. While the median (50th percentile) family saw its meager retirement account savings drop by more than half between 2007 and 2010, the 90th percentile family experienced a 5 percent decline. Thus, retirement inequality continued to grow in the aftermath of the Great Recession (Figure 7).

Other forms of saving, including home equity, may be tapped to pay for retirement. But family net worth took an even bigger hit than retirement savings following the collapse of the housing bubble and ensuing recession (Figure 17). Like retirement savings, overall wealth has grown more unequal in recent decades, as all but the top 20 percent of working-age families have seen declines in net worth (Figure 18).

The growth in retirement inequality has not been random—the rich have gotten richer and the poor poorer. Participation in retirement savings plans is highly unequal across income groups. In 2013, nearly nine in 10 families in the top income fifth had retirement account savings, compared with fewer than one in 10 families in the bottom income fifth (Figure 9). This disparity has grown in the new millennium as the share of working-age families with retirement account savings declined for all except the top income group. While it is normal for higher-income families to have more savings, the fact that most families in the bottom half of the income distribution have no retirement account savings at all is a serious policy failure.

Income inequality and differences between younger and older families explain some, but not all, of the inequality in retirement savings. In 2013, families in the top income fifth accounted for 63 percent of total income, but 74 percent of total savings in retirement accounts (Figure 19). Even within age groups, retirement account balances are more unequally distributed than income (for example, Figure 20).

The shift from defined-benefit to defined-contribution plans has exacerbated racial and ethnic disparities. Black workers’ participation in employer-based retirement plans used to be similar to that of white workers, but black workers began lagging behind white workers in the 401(k) era, while Hispanic workers fell even further behind (Morrissey 2016; Morrissey and Sabadish 2013). Only 41 percent of black families and 26 percent of Hispanic families had retirement account savings in 2013, compared with 65 percent of white non-Hispanic families (Figure 10). Even among families nearing retirement (age 56–61), the majority of black and Hispanic families have no retirement account savings (not shown on chart). Racial and ethnic gaps in retirement account balances are even larger than participation gaps—and growing. For families with retirement account savings, the median amount is $22,000 for black and Hispanic families, compared with $73,000 for white non-Hispanic families (Figure 11).

Retirement readiness gaps have widened between workers with and without a college education. The share of families with retirement account savings increased across education groups in the 1990s and declined across education groups in the 2000s. In 2013, only families headed by someone with some college experience were more likely than not to have retirement account savings (Figure 12). Even among families approaching retirement (age 56–61), the typical family headed by someone with a high school education or less had no savings in retirement accounts (not shown on chart). For families with retirement savings (Figure 13), median account balances grew much more for families headed by someone with a college degree than for other families before the Great Recession, and have been fairly flat across the board since.

Single people and women face particular challenges. Single people tend to be less prepared for retirement than their married counterparts. Even among those approaching retirement (age 56–61), most single men and women do not have any retirement account savings (not shown on chart). In the past, married women were less likely to be covered through their own employers than were single women, but married women’s participation increased as their earnings grew (Morrissey and Sabadish 2013). In the new millennium, the biggest change has been the decline in the share of single men with retirement account savings and in the amount single men have saved in these accounts (Figures 14–15).

Though declines for single men closed savings gaps between single men and women, women generally remain more vulnerable than men because they live longer and are more likely to outlive their savings—and, in the case of married women, their spouses. Women can expect to live around 2.3 years longer than men in retirement (Social Security Trustees 2015). Because women earn less and accumulate less retirement savings and benefits than men, never-married, divorced, and widowed women are at greater risk of experiencing hardship in retirement than their male counterparts.

Retirement wealth has grown nearly twice as fast as income: Assets in retirement plans as a percent of personal disposable income by type, 1989–2014

1
Defined-benefit pension funds Defined-contribution plans and IRAs
1989 71% 31%
1990 72% 32%
1991 73% 37%
1992 76% 38%
1993 77% 42%
1994 77% 43%
1995 78% 50%
1996 77% 54%
1997 77% 64%
1998 77% 72%
1999 78% 82%
2000 76% 75%
2001 78% 68%
2002 79% 61%
2003 79% 71%
2004 81% 74%
2005 82% 76%
2006 81% 83%
2007 81% 87%
2008 80% 63%
2009 84% 77%
2010 89% 85%
2011 88% 82%
2012 87% 88%
2013 89% 104%
2014 88% 106%
ChartData Download data

The data below can be saved or copied directly into Excel.

Source: EPI analysis of  Federal Reserve Financial Accounts of the United States (December 10, 2015 release).

With an aging population, aggregate retirement wealth (assets in pension funds plus savings in retirement accounts) nearly doubled as a share of personal disposable income between 1989 and 2014, even as rising inequality worsened retirement insecurity for most families. Retirement account savings have exceeded pension fund assets since 2012, as well as briefly in the late 1990s and mid-2000s. Assets in retirement accounts are more affected by economic downturns than pooled pensions since contributions to these plans are voluntary and funds may be withdrawn in hard times. In addition, individual retirement account investments are less diversified and investment returns more volatile.

Share this chart:

Copy the code below to embed this chart on your website.

Retirement plan participation has declined even as baby boomers have approached retirement: Share of families age 32–61 participating in retirement plans by type, 1989–2013

2
Defined-benefit plan Defined-contribution plan Both types of plans Any plan
1989 41% 35% 17% 58%
1992 33% 35% 13% 54%
1995 28% 41% 12% 56%
1998 26% 46% 13% 59%
2001 28%  47%  15%  60% 
2004 25% 44% 11% 57%
2007 25%  47%  14%  57% 
2010 22% 42% 11% 53%
2013 21%  43%  11%  53% 
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Since DC and DB shares include families with both kinds of plans, the share with both types is subtracted from the total to produce the share with any plan. Shares indicate whether either the respondent or his or her spouse participated in such a plan or plans on a current job (individual participation rates are lower).

Source: EPI analysis of Survey of Consumer Finance data, 2013.

Participation in retirement plans has declined in the new millennium, with a steeper decline for workers in defined-benefit plans than in defined-contribution plans. For families headed by working-age workers (age 32–61), participation in any type of plan fell from 60 percent in 2001 to 53 percent in 2013. We would have expected participation to increase in the new millennium as the large baby boomer cohort entered their 50s and 60s, when participation rates tend to be high.

Share this chart:

Copy the code below to embed this chart on your website.

The share of families with retirement savings grew in the 1990s but declined after the Great Recession: Share of families age 32–61 with retirement account savings by age, 1989–2013

3
Working-age (32–61) 32–37 38–43 44–49 50–55 56–61
1989 49% 44% 51% 49% 53% 52%
1992 50% 42% 54% 51% 52% 52%
1995 55% 52% 56% 57% 57% 54%
1998 60% 56% 61% 59% 61% 62%
2001 62% 56% 62% 63% 64% 64%
2004 58% 53% 57% 60% 59% 63%
2007 62% 53% 58% 63% 67% 66%
2010 57% 51% 53% 60% 61% 57%
2013 57% 51% 56% 57% 57% 61%
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Retirement account savings include 401(k)s, IRAs, and Keogh plans.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

The share of working-age families with retirement account savings rose in the 1990s as employers replaced traditional pensions with 401(k)s. But it contracted after the 2001 and 2007–2009 recessions, and remains 5 percentage points below the 2001 and 2007 peaks. This drop-off reflects the fact that retirement account contributions are voluntary and funds may be tapped before retirement, making retirement savings more vulnerable than traditional pension benefits to economic downturns. The post–Great Recession drop is particularly worrisome for older workers who will have less time to make up losses.

Share this chart:

Copy the code below to embed this chart on your website.

Retirement savings have stagnated in the new millennium: Mean retirement account savings of families by age, 1989–2013 (2013 dollars)

4
    32–37     38–43     44–49     50–55     56–61 Working-age (32–61)
1989 $13,862 $30,741 $40,463 $52,778 $58,585 $35,492
1992 $12,552 $22,920 $44,749 $75,678 $68,060 $39,257
1995 $20,145 $33,510 $60,882 $85,352 $80,255 $50,475
1998 $29,458 $48,084 $64,166 $91,373 $114,907 $64,982
2001 $28,880  $52,843  $91,243  $129,938  $155,371  $86,187 
2004 $27,100 $54,205 $81,818 $128,848 $158,827 $87,818
2007 $27,145  $54,527  $91,237  $135,384  $211,885  $101,548 
2010 $25,864 $48,472 $85,454 $138,713 $175,696 $96,882
2013 $31,644  $67,270  $81,347  $124,831  $163,577  $95,776  
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Retirement account savings include 401(k)s, IRAs, and Keogh plans.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

While average (mean) retirement account savings grew somewhat between 2001 and 2013, this was due to the aging of the large baby boomer cohort, as older families have had more time to accumulate savings. The results are mixed when age is taken into account. Workers in their late 40s and early 50s are slightly behind their counterparts in 2001, while other age groups are slightly ahead. Rather than stagnation, we should be seeing rising 401(k) and IRA account balances at all ages to offset declines in defined-benefit pension coverage and Social Security cuts.

Share this chart:

Copy the code below to embed this chart on your website.

Most families—even those approaching retirement—have little or no retirement savings: Median retirement account savings of families by age, 1989–2013 (2013 dollars)

5
    32–37     38–43     44–49     50–55     56–61
1989 $0 $542 $0 $1,808 $1,808
1992 $0 $812 $812 $2,275 $1,024
1995 $379 $2,429 $3,795 $7,589 $4,857
1998 $2,144 $7,148 $9,006 $8,291 $12,866
2001 $1,313 $6,566 $12,607 $14,445 $15,759
2004 $1,233 $3,699 $9,372 $12,331 $21,580
2007 $1,123  $5,951  $15,158  $26,386  $35,929 
2010 $289 $1,929 $10,181 $13,932 $7,609
2013 $480  $4,200  $6,200  $8,000  $17,000 
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Scale changed for visibility. Retirement account savings include 401(k)s, IRAs, and Keogh plans.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

Nearly half of families have no retirement account savings at all. That makes median (50th percentile) values low for all age groups, ranging from $480 for families in their mid-30s to $17,000 for families approaching retirement in 2013. For most age groups, median account balances in 2013 were less than half their pre-recession peak and lower than at the start of the new millennium.

Share this chart:

Copy the code below to embed this chart on your website.

More people have 401(k)s, but participation in traditional pensions is more equal: Retirement plan participation of families age 32–61 by family income, race and ethnicity, education, gender, and marital status, 2013

6
Characteristic Defined-benefit plan  Defined-contribution plan 
All 21% 43%
1st (bottom) quintile 6% 4%
2nd (lower-middle) quintile 12% 21%
3rd (middle) quintile 19% 38%
4th (upper-middle) quintile 30% 54%
5th (top) quintile 27% 68%
Hispanic 9% 20%
Black 20% 32%
White non-Hispanic 24% 49%
No high school diploma/GED 6% 15%
High school diploma/GED 21% 34%
Some college 22% 39%
College degree or more 24% 57%
Single female 12% 32%
Single male 16% 26%
Married or living with partner 25% 50%
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: "College degree" includes associate degrees.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

Roughly twice as many families have defined-contribution plans as defined-benefit pensions, but participation in pensions is more equal across education, race, and income groups. Thanks in large part to unionized workers, who place a high value on pensions, the share of high-school graduates with pensions (21 percent) is almost as high as the share of college graduates (24 percent); and the share of blacks with pensions (20 percent) is almost as high as the share of non-Hispanic whites (24 percent). However, the participation gap for single women is wider for pensions than for defined-contribution plans.

Share this chart:

Copy the code below to embed this chart on your website.

The gap between the retirement ‘haves’ and ‘have-nots’ has grown since the recession: Retirement account savings of families age 32–61 by savings percentile, 1989–2013 (2013 dollars)

7
50th (median) 60th 70th 80th 90th
1989 $0 $5,423 $14,461 $32,536 $90,379
1992 $0 $4,874 $16,248 $39,384 $90,987
1995 $2,277 $9,866 $24,286 $47,054 $113,841
1998 $6,004 $17,440 $38,597 $73,763 $160,106
2001 $7,879 $23,638 $48,326 $92,818 $223,247
2004 $6,166 $19,730 $49,326 $102,351 $246,628
2007 $11,228 $30,315 $61,754 $123,508 $258,243
2010 $5,358 $19,291 $42,868 $96,453 $246,490
2013 $5,000 $20,100  $50,000  $116,000  $274,000 
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Retirement account savings include 401(k)s, IRAs, and Keogh plans. Scale changed to accommodate larger values.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

Nearly half of working-age families have nothing saved in retirement accounts, and the median working-age family had only $5,000 saved in 2013. Meanwhile, the 90th percentile family had $274,000, and the top 1 percent of families had $1,080,000 or more (not shown on chart). These huge disparities reflect a growing gap between haves and have-nots since the Great Recession as accounts with smaller balances have stagnated while larger ones rebounded.

Share this chart:

Copy the code below to embed this chart on your website.

Retirement account savings are inadequate and unequal: Retirement account savings of families age 32–61, 1989–2013 (2013 dollars)

8
Mean savings, all families Median savings, families with retirement savings Median savings, all families
1989 $35,492.41 $22,413.88 $0.00
1992 $39,256.91 $24,858.88 $0.00
1995 $50,474.67 $28,384.37 $2,276.82
1998 $64,982.50 $38,597.09 $6,003.99
2001 $86,187.11 $45,962.57  $7,879.30 
2004 $87,818.02 $49,325.68 $6,165.71
2007 $101,548.36  $57,262.57  $11,227.96 
2010 $96,881.71 $50,369.70 $5,358.48
2013 $95,775.93  $60,000.00  $5,000.00 
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Since DC and DB shares include families with both kinds of plans, the share with both types is subtracted from the total to produce the share with any plan. Shares indicate whether either the respondent or his or her spouse participated in such a plan or plans on a current job (individual participation rates are lower). Retirement account savings include 401(k)s, IRAs, and Keogh plans.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

Since nearly half of all working-age families have zero retirement account savings, it is not surprising that the median (50th percentile) family had only $5,000 saved in these accounts in 2013. Even families with retirement savings have inadequate savings in these accounts—the median for families with savings was $60,000. The large gap between mean retirement savings ($95,776) and median retirement savings ($5,000) indicates inequality—that the large account balances of families with the most savings are driving up the average for all families. As some families with small balances drained their savings in the wake of the Great Recession, mean and median savings declined after 2007 while the median for families with savings rose slightly.

Share this chart:

Copy the code below to embed this chart on your website.

High-income families are 10 times as likely to have retirement accounts as low-income families: Share of families age 32–61 with retirement account savings by income quintile, 1995–2013

9
1st (Bottom) 2nd (Lower-middle) 3rd (Middle) 4th (Upper-middle) 5th (Top)
1995 10% 33% 50% 68% 79%
1998 15% 38% 57% 69% 83%
2001 17%   40%   56%   77%   89%  
2004 9% 37% 53% 74% 84%
2007 12%  34%  59%  76%  90% 
2010 12% 36% 53% 68% 86%
2013 8%  30%  52%  69%  88% 
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Retirement account savings include 401(k)s, IRAs, and Keogh plans. Family-income quintiles are based on "normal income," a measure that ignores temporary fluctuations and is not available for years prior to 1995.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

Almost nine in 10 families in the top income fifth had savings in retirement accounts in 2013, compared with less than one in 10 families in the bottom income fifth. This reflects a growing disparity in the new millennium as the share of families with retirement account savings declined significantly for all except the top income group.

Share this chart:

Copy the code below to embed this chart on your website.

Most black and Hispanic families have no retirement account savings: Share of families age 32–61 with retirement account savings by race, 1989–2013

10
White non-Hispanic Black Hispanic
1989 57% 26% 22%
1992 57% 31% 26%
1995 60% 37% 40%
1998 65% 44% 29%
2001 67%  47%  39% 
2004 66% 39% 33%
2007 67%  47%  38% 
2010 65% 39% 31%
2013 65%  41%  26% 
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Retirement account savings include 401(k)s, IRAs, and Keogh plans.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

The share of Hispanic families with retirement account savings plummeted in the wake of the Great Recession, from 38 percent in 2007 to 26 percent in 2013, while the share of black families with retirement savings fell from 47 to 41 percent. In contrast, almost two-thirds (65 percent) of white non-Hispanic families had retirement savings in 2013, a share only slightly below the 2007 peak (67 percent).

Share this chart:

Copy the code below to embed this chart on your website.

Racial and ethnic gaps are large even among families with retirement savings: Median savings for families age 32–61 with retirement account savings by race, 1989–2013 (2013 dollars)

11
White non-Hispanic Black Hispanic
1989 $25,306 $14,461 $9,038
1992 $27,621 $12,998 $11,373
1995 $30,358 $15,179 $18,215
1998 $42,886 $20,013 $20,013
2001 $53,842  $13,132  $18,385 
2004 $58,944 $22,197 $20,963
2007 $69,613  $32,561  $27,508 
2010 $62,726 $23,577 $26,792
2013 $73,000  $22,000  $22,000 
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Retirement account savings include 401(k)s, IRAs, and Keogh plans. Scale changed for visibility.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

Large racial and ethnic disparities are evident even among families with retirement account savings. In 2013, the median white non-Hispanic family with retirement savings had over three times as much saved in a retirement account ($73,000) as the median black or Hispanic family with savings ($22,000 in both cases).

Share this chart:

Copy the code below to embed this chart on your website.

College-educated families are much more likely to have retirement savings: Share of families age 32–61 with retirement account savings by education, 1989–2013

12
No high school diploma/GED High school diploma/GED Some college College degree or more
1989 22% 43% 49% 69%
1992 18% 42% 50% 67%
1995 23% 49% 57% 71%
1998 30% 52% 62% 74%
2001 24%  54%  57%  82% 
2004 19% 51% 57% 75%
2007 27%  52%  61%  79% 
2010 18% 46% 57% 75%
2013 18%  43%  53%  76% 
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Retirement account savings include 401(k)s, IRAs, and Keogh plans. "College degree" includes associate degrees.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

The share of families with retirement account savings increased across education groups in the 1990s and declined across groups in the 2000s. However, the decline was steeper for less-educated groups. Over three-fourths (76 percent) of families headed by someone with a college degree or more education had savings in retirement accounts in 2013, compared with 43 percent and 18 percent, respectively, of families headed by someone with and without a high school diploma or GED.

Share this chart:

Copy the code below to embed this chart on your website.

College-educated families have much larger retirement account balances: Median savings for families age 32–61 with retirement savings by education, 1989–2013 (2013 dollars)

13
No high school diploma/GED High school diploma/GED Some college College degree or more
1989 $19,883 $14,461 $21,691 $36,151
1992 $9,749 $15,598 $19,497 $41,756
1995 $7,589 $22,768 $25,804 $39,465
1998 $20,013 $24,302 $32,879 $58,324
2001 $10,768 $30,204 $34,669 $73,212
2004 $14,798 $27,129 $28,362 $91,253
2007 $15,719 $35,929 $42,666 $94,315
2010 $13,932 $27,200 $31,079 $85,736
2013 $14,700 $30,000 $46,900 $95,000
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Retirement account savings include 401(k)s, IRAs, and Keogh plans. "College degree" includes associate degrees.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

Only families with at least some college experience are more likely than not to have retirement account savings. But even among families with retirement account savings, there are large differences in account holdings by education. The typical family with retirement savings headed by someone with a college degree or more education had more than three times as much ($95,000) as the typical family headed by someone with no more than a high school diploma or GED ($30,000), which in turn had twice as much as the typical family headed by someone without a high school diploma or GED ($14,700) in 2013.

Share this chart:

Copy the code below to embed this chart on your website.

Single people are less likely to have retirement savings: Share of families age 32–61 with retirement account savings by gender and marital status, 1989–2013

14
Married or living with partner Single men Single women 
1989 56% 40% 33%
1992 57% 36% 36%
1995 64% 37% 41%
1998 66% 49% 46%
2001 69%  55%  44% 
2004 67% 51% 39%
2007 69%  50%  47% 
2010 65% 41% 45%
2013 65%  43%  42% 
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Retirement account savings include 401(k)s, IRAs, and Keogh plans.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

Nearly two-thirds (65 percent) of married couples had retirement account savings in 2013, compared with 43 percent of single men and 42 percent of single women. Though all groups saw declines in the new millennium, the share of single men with retirement savings declined the most. Single women remain more vulnerable in retirement than single men due to lower lifetime earnings and longer lifespans.

Share this chart:

Copy the code below to embed this chart on your website.

Single people have less, but retirement savings are too low across the board: Median savings for families age 32–61 with retirement account savings, by gender and marital status, 1989–2013 (2013 dollars)

15
Married or living with partner Single men  Single women 
1989 $27,114 $18,076 $12,653
1992 $31,683 $17,872 $9,749
1995 $32,331 $24,286 $16,697
1998 $44,315 $31,449 $20,299
2001 $57,782 $31,517 $19,698
2004 $66,590 $35,761 $24,170
2007 $70,736 $51,649 $35,929
2010 $65,373 $34,294 $24,649
2013 $78,000 $34,000  $30,000 
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Retirement account savings include 401(k)s, IRAs, and Keogh plans.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

Single people are less likely to have retirement account savings than couples. Among those with savings, the typical single man ($34,000) and single woman ($30,000) had lower balances than the typical married couple ($78,000) in 2013. However, much if not all of this difference reflects family size and income. Thus, the problem is primarily one of lower participation for single people and low account balances across the board. In addition, women should be saving more than men because they live longer.

Share this chart:

Copy the code below to embed this chart on your website.

Recessions are most damaging to workers nearing retirement: Mean retirement account savings of families by birth cohort, 1989–2013 (2013 dollars)

16
1940–1945 1946–1951 1952–1957 1958–1963 1964–1969 1970–1975 1976–1981
1989 $40,463 $30,741 $13,862
1992 $63,567 $33,565 $16,360 $12,143
1995 $85,352 $60,882 $33,510 $20,145
1998 $95,371 $76,741 $57,068 $34,844 $28,935
2001 $155,371 $129,938 $91,243 $52,843 $28,880
2004 $149,481 $163,774 $98,686 $67,816 $36,734 $26,832
2007 $211,885 $135,384 $91,237 $54,527 $27,145
2010 $179,310 $165,089 $108,545 $66,408 $34,707 $23,151
2013 $163,577 $124,831 $81,347 $67,270 $31,644
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Retirement account savings include 401(k)s, IRAs, and Keogh plans.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

This chart shows savings trajectories by birth cohort over their working-age years (age 32–61). War babies (born 1940–1945) and early baby boomers (born 1946–1951) had the misfortune to be nearing retirement when the 2001 and 2007–2009 recessions hit. Older savers are more affected by market downturns because investment returns outweigh new contributions. Another factor that may explain why early boomers were more affected than middle boomers (born 1952–1957) by the Great Recession is that many older workers who lose jobs tap retirement savings.

Share this chart:

Copy the code below to embed this chart on your website.

Family finances still have not recovered from the collapse of the housing bubble: Median net worth of families age 32–61, 1989–2013 (2013 dollars)

17
    32–37     38–43     44–49     50–55     56–61 All 32–61
1989 $47,611 $104,026 $153,228 $191,422 $159,157 $111,640
1992 $39,823 $84,926 $106,633 $167,854 $203,420 $94,886
1995 $47,054 $80,190 $121,355 $167,270 $170,003 $99,345
1998 $54,179 $101,739 $136,033 $177,561 $179,834 $120,208
2001 $46,160 $112,897 $154,040 $213,529 $262,361 $135,787
2004 $50,066 $113,079 $155,253 $199,300 $314,328 $135,017
2007 $56,005 $117,894 $163,928 $239,043 $322,523 $156,069
2010 $21,005 $52,320 $93,773 $151,538 $165,791 $80,270
2013 $27,500 $56,640 $69,700 $121,100 $164,000 $78,660
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Scale changed to accommodate larger values.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

The typical family has more home equity than retirement account savings, if they have either. Thus it is no surprise that family finances were devastated by the collapse of the housing bubble. Working-age families’ median wealth, or net worth, fell by almost half during the Great Recession and its immediate aftermath and stagnated between 2010 and 2013 despite rebounds in stock and housing prices. Declines in the net worth of older families since 2010 are especially worrisome since they have less opportunity to make up losses before retirement.

Share this chart:

Copy the code below to embed this chart on your website.

The recession did not halt the decades-long growth in wealth inequality: Net worth of families age 32–61, by net worth percentile, 1989–2013 (2013 dollars)

18
10th  20th 30th  40th 50th (median) 60th 70th 80th 90th
1989 $54 $10,845 $35,645 $66,808 $111,640 $169,442 $240,606 $362,780 $688,594
1992 $536 $10,561 $29,408 $58,345 $94,886 $142,638 $208,782 $306,430 $603,437
1995 $425 $12,189 $35,367 $63,751 $99,345 $142,332 $204,383 $305,701 $627,188
1998 $829 $13,438 $39,583 $73,363 $120,208 $171,157 $269,465 $416,448 $752,743
2001 $1,445 $15,629 $41,826 $77,375 $135,787 $210,968 $319,519 $515,792 $1,021,420
2004 $1,603 $13,121 $42,753 $79,908 $135,017 $210,251 $344,306 $575,384 $1,043,115
2007 $1,123 $14,372 $44,912 $88,701 $156,069 $248,138 $366,256 $591,646 $1,031,400
2010 -$1,811 $5,358 $17,726 $43,618 $80,270 $140,928 $232,955 $410,674 $951,238
2013 -$2,090 $4,640 $15,950 $39,000 $78,660 $140,870 $233,500 $420,400 $920,900
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Scale changed to accommodate larger values.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

Net worth declined across the board after the Great Recession, leaving the bottom 70 percent of working-age families with less wealth in 2013 than their counterparts had in 1989—a devastating setback. The bottom 10 percent had negative net worth in 2010 and 2013 (not visible in the chart). While wealthy families also saw declines after 2007, these declines reversed only a small part of the gains they experienced in the 1990s and early 2000s. Thus, wealth inequality grew substantially in recent decades, even after the Great Recession.

Share this chart:

Copy the code below to embed this chart on your website.

401(k)s magnify inequality: Share of total retirement account savings and total income for families age 32–61 by income quintile, 2013

19
Retirement account savings Income
1st (Bottom) 0% 2%
2nd (Lower-middle) 1% 5%
3rd (Middle) 5% 10%
4th (Upper-middle) 20% 19%
5th (Top) 74% 63%
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Based on "normal income," which may differ from actual income if a family's income in the past year was unusually high or low. Retirement account savings include 401(k)s, IRAs, and Keogh plans.

Source: EPI analysis of  Survey of Consumer Finance data, 2013.

Despite rules intended to ensure that high-income families do not disproportionately benefit from tax subsidies for retirement saving, our savings-based retirement system does not simply reflect, but also magnifies, inequality. The bottom 60 percent of working-age families receive 17 percent of total income but hold 7 percent of retirement account balances. Meanwhile, the top 20 percent receive 63 percent of income and hold 74 percent of retirement account balances (numbers in chart may not add up to totals due to rounding).

Share this chart:

Copy the code below to embed this chart on your website.

Retirement inequality is greater than income inequality even in peak earning years: Share of total retirement account savings and total income for families in peak earning years (age 50–55) by income quintile, 2013

20
Retirement account savings Income
1st (Bottom) 0% 4%
2nd (Lower-middle) 4% 7%
3rd (Middle) 10% 12%
4th (Upper-middle) 23% 19%
5th (Top) 63% 59%
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Based on "normal income," which may differ from actual income if a family's income in the past year was unusually high or low. Retirement account savings include 401(k)s, IRAs, and Keogh plans.

Source: EPI analysis of Survey of Consumer Finance data, 2013.

The fact that retirement savings are more unequal than incomes in part reflects the fact that older workers tend to earn more and also have had longer to accumulate savings. However, upper-income families hold a disproportionate share of retirement account balances even within specific age groups, such as workers in their peak earning years (age 50–55). The bottom 60 percent of families in this age group receive 22 percent of total income but hold only 14 percent of account balances (numbers in chart may not add up to totals due to rounding).

Share this chart:

Copy the code below to embed this chart on your website.

Part Two: How are older Americans faring in retirement?

Charts in this section focus on the income of people 65 and older and are based on data from the U.S. Census Bureau’s Current Population Survey Annual Social and Economic Supplement (CPS-ASEC) (Flood et al. 2015). Unlike the Survey of Consumer Finances, the CPS focuses primarily on individuals rather than families. Though the goal is to assess retirement outcomes, some people in this age group are still working.

Until recently, the CPS did a poor job of capturing distributions from retirement accounts and other types of asset income, making it a problematic source for assessing retirement income. A 2014 survey redesign to correct this problem resulted in large percentage increases in these income measures (DeNavas-Walt and Proctor 2015; Semega and Welniak 2015). But as charts in this section will show, income from retirement accounts remains modest in dollar terms.

Because many families withdraw retirement account savings in lump sums, the size of these distributions for any family in a given year, whether a large sum or nothing at all, does not tell us how important this income source is for that family. But the mean value of these distributions across families does give a sense of the importance of retirement account savings relative to other sources of income. Going forward, it will be possible to assess how much these distributions are affected by economic conditions—for example, people tapping retirement funds when they lose jobs in recessions.

Key findings of the following charts include:

Less educated, minority, single, and female seniors are most likely to have low incomes in retirement. Seniors with a college degree or more education have median annual incomes more than twice as high as those with a high-school diploma or GED. The median income for seniors without a high-school diploma is barely above the official poverty threshold, which was $11,354 for single seniors in 2014 (DeNavas-Walt and Proctor 2015). Racial and ethnic disparities are also significant, with more than half of Hispanic seniors and nearly half of black seniors living on less than full-time minimum-wage earnings ($15,080 per year). Unmarried older women, including widows, have lower incomes than unmarried men and a lower standard of living than married couples, taking into account the fact that couples’ living expenses are less than twice those of a person living alone (Figure 21).

Social Security is the most important and evenly distributed source of retirement income. Social Security contributes 35 percent of total income for people age 65 and older (see “Sources of income,” below, for a description of this and other income categories). Social Security benefits are not only the most important income source but also the most evenly distributed. While benefits are somewhat smaller in dollar terms for low-income, less-educated, black, Hispanic, and female seniors, they are larger as a share of income for these groups than for other seniors (Figures 22–28).

Sources of income

  • “Earnings” include wage and salary income, nonfarm business income, and farm income.
  • “Social Security” includes Old Age and Survivor’s Insurance (OASI) and Social Security Disability Insurance (SSDI) benefits, the latter of which are converted to OASI benefits at the normal retirement age (currently 66). This measure does not include Supplemental Security Income (SSI) benefits (see “Other”).
  • “Private pensions” include company, union, and U.S. Railroad Retirement pensions and survivor pensions.
  • “Public pensions” include federal government retirement and survivor pensions; U.S. military retirement and survivor pensions; and state or local government retirement and survivor pensions.
  • “Asset income” includes payments from annuities or paid-up insurance policies, payments from estates and trusts, interest, dividends, and rent.
  • “401(k)s, IRAs, etc.” includes regular and lump-sum distributions from these and similar retirement savings accounts, including Keogh plans.
  • “Other” includes income from SSI, welfare benefits, unemployment benefits, workers’ compensation, veterans benefits, disability benefits (other than SSDI), educational assistance, child support, financial assistance from friends or relatives not living in the same household, and black lung survivor pensions. It also includes personal income from unspecified sources, possibly including some income from pensions, retirement accounts, and assets because only two types of “retirement” or “survivor” income are identified for each survey respondent.

Many seniors are still working. Earned income, including farm and business income, is the second most important income source for seniors after Social Security, providing 29 percent of total income. Earnings are not as evenly distributed as Social Security benefits, disproportionately accruing to younger, higher-income, college-educated, Hispanic, and male seniors. Earnings exceed Social Security benefits for seniors in their mid- to late 60s, those in the top income fifth, those with college degrees, and married men in this age group (Figures 22–27). Earned income has increased in importance as Americans increasingly delay retirement. The share of people 65 and older who are employed is now higher than in over half a century (author’s analysis of Bureau of Labor Statistics data).

A diverse group of retirees relies on pension benefits. Nearly three in 10 seniors (29 percent) receive public- or private-sector defined-benefit pension benefits, and the median benefit received is $13,200 (Figure 29). Public and private pension benefits together constitute the third-largest source of income for seniors after Social Security and earnings, providing 17 percent of total income (Figure 23).

Pension benefits are important to seniors across demographic groups, but are a somewhat larger share of income for middle- and upper-middle-income, college-educated, black, and male seniors than for other seniors (Figures 24–27). Women are almost as likely as men to receive public-sector pension benefits, which constitute 8 percent of seniors’ total income (Figure 27 and 30). Seniors with high-school diplomas are almost as likely as those with college degrees to receive private-sector pension benefits, which constitute 9 percent of seniors’ total income (Figures 26 and 31).

401(k)s and IRAs are not an important source of retirement income. Retirement account distributions account for less than 3 percent of seniors’ total income (Figure 23). They constitute a somewhat larger share of income for higher-income, white non-Hispanic, college-educated, and male seniors than for other seniors (Figures 24–27). Non-Hispanic white seniors are three times as likely as black or Hispanic seniors to receive retirement account distributions (Figure 32). Among seniors with income from these accounts, the median amount is higher among Hispanic ($6,000) than black ($3,000) seniors, perhaps because Hispanics are less likely than blacks to participate in defined-benefit pensions (Figure 29).

One possible explanation for why retirement account distributions are less important than pension benefits is that today’s seniors, when they were working, were more likely to participate in defined-benefit pensions than workers today. However, 401(k)s have been around long enough that they should be contributing more to senior incomes. In the private sector, the number of active participants in defined-contribution plans surpassed the number in defined-benefit plans a long time ago, in 1984 (U.S. Department of Labor 2015).

Retirement account distributions are a much smaller share of total income than pension benefits even for a subset of seniors who almost certainly participated in 401(k)s and other defined-contribution plans longer than they participated in defined-benefit pensions—those age 65–69 (Figure 23) (author’s analysis of SCF microdata for 1989–2013 and U.S. Department of Labor 2015). Though this partly reflects the fact that many 65- to 69-year-olds are still working and distributions from retirement account plans are not required until age 70½, the poor showing for defined-contribution plans and IRAs nevertheless bodes poorly for future retirees with less pension income to fall back on.

Older, minority, unmarried, female, and less-educated seniors are most vulnerable in retirement: Median income of people age 65 and older by age, race and ethnicity, education, gender, and marital status, 2014

21
Characteristic
All $21,265
Age 65–69 $25,337
Age 70–79 $20,650
Age 80+ $18,084
Hispanic $13,199
Black $16,356
White, non-Hispanic $23,244
Less than high school $12,167
High school diploma/GED $18,628
Some college $23,876
Bachelor’s degree or more $40,000
Unmarried woman $18,216
Unmarried man $25,260
Married woman $14,000
Married man $32,100

 

ChartData Download data

The data below can be saved or copied directly into Excel.

Source: EPI analysis of U.S. Census Bureau's Current Population Survey Annual Social and Economic Supplement microdata.

Incomes decline as seniors age, primarily due to lower earnings. Incomes vary even more by race, ethnicity, and education. More than half of Hispanic seniors and nearly half of black seniors live on less than what a full-time minimum-wage worker earns ($15,080 annually). The median income of seniors with a college degree or more education is more than twice as high as that of seniors with high-school diplomas. Unmarried women age 65 and older have lower incomes than unmarried men that age and a lower standard of living than married couples, taking into account married couples’ combined incomes and the fact that their living expenses are less than twice those of a person living alone.

Share this chart:

Copy the code below to embed this chart on your website.

Social Security is the most important source of income for seniors: Annual income of people age 65 and older by source, 2014

22
Measure Mean Median of people with specified income Share with income (right axis)
Earnings  $10,119 $29,900 22%
Social Security  $12,232 $14,400 82%
Public pensions  $2,840 $16,800 12%
Private pensions  $3,131 $9,00 22%
 401(k)s, IRAs, etc.  $966 $5,400 8%
Asset income  $4,019 $325 61%
 Other  $1,590 $4,800 17%

 

ChartData Download data

The data below can be saved or copied directly into Excel.

Note: See box in text defining income sources.

Source: EPI analysis of U.S. Census Bureau's Current Population Survey Annual Social and Economic Supplement microdata.

Social Security is the most evenly distributed source of retirement income, with 82 percent of people age 65 and older receiving benefits. Among senior beneficiaries, the median benefit is $14,400. Though 61 percent of seniors receive interest or other asset income—the next most common source of income—amounts are too small to matter much for most seniors. Earned income is a major source of income, but only for the 22 percent of seniors with earnings. Public and private pensions are a much more important source of income than distributions from retirement accounts.

Share this chart:

Copy the code below to embed this chart on your website.

Earnings and overall income decline with age: Mean annual income of people age 65 and older by source and by age, 2014

23
Earnings  Social Security  Public pensions  Private pensions  401(k)s, IRAs, etc.  Asset income  Other 
All $10,119 $12,232 $2,840 $3,131 $966 $4,019 $1,590
65–69 $18,272 $10,736 $2,714 $2,922 $899 $3,998 $1,821
70–79 $7,632 $13,147 $3,005 $3,393 $1,154 $4,119 $1,475
80+ $2,842 $12,769 $2,733 $2,970 $735 $3,875 $1,463
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: See box in text defining income sources.

Source: EPI analysis of U.S. Census Bureau's Current Population Survey Annual Social and Economic Supplement microdata

Most of the income differences between age groups can be explained by declining earnings as seniors transition into retirement, though younger generations also had higher earnings during their careers. Social Security income is lowest for seniors in their late 60s, many of whom are still working. Defined-benefit pension benefits are lower for seniors in their late 60s than for those in their 70s because younger cohorts were more likely to participate in defined-contribution plans during their working years. Nevertheless, retirement account distributions remain modest even for these younger retirees.

Share this chart:

Copy the code below to embed this chart on your website.

Low-income seniors are almost entirely reliant on Social Security, whereas earnings matter most for high-income seniors: Mean annual income of people age 65 and older by source and by family income quintile, 2014

24
 Quintile Earnings Social Security Public pensions  Private pensions  401(k)s, IRAs, etc. Asset income  Other Total
All $10,119 $12,232 $2,840 $3,131 $966 $4,019 $1,590 $34,897
1st (Bottom) $296 $8,229 $113 $200 $54 $265 $799 $9,956
2nd (Lower-middle) $1,376 $12,969 $743 $1,245 $195 $865 $820 $18,213
3rd (Middle) $3,820 $13,675 $1,755 $3,067 $583 $1,764 $1,381 $26,045
4th (Upper-middle) $9,621 $13,514 $4,299 $4,475 $1,078 $3,528 $2,256 $38,771
5th    (Top) $35,500 $12,772 $7,292 $6,668 $2,921 $13,680 $2,697 $81,529
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Scale changed to accommodate larger values. See box in text defining income sources.

Source: EPI analysis of U.S. Census Bureau's Current Population Survey Annual Social and Economic Supplement microdata.

Social Security benefits are relatively equal across the income distribution, while earned income and asset income are much greater for the top income group. Public and private pension benefits are larger in dollar terms for seniors in the top income group, but constitute a larger share of income for seniors in the middle- and upper-middle income groups. Distributions from retirement savings accounts in the top income group are triple the average amount, yet still constitute a small share of total income.

Share this chart:

Copy the code below to embed this chart on your website.

Social Security is important to all racial and ethnic groups, while pensions make biggest difference for blacks: Mean annual income of people age 65 and older by race and ethnicity, 2014

25
 Race/ Ethnicity Earnings  Social Security  Public pensions  Private pensions  401(k)s, IRAs, etc.  Asset income  Other  Total
All $10,119 $12,232 $2,840 $3,131 $966 $4,019 $1,590 $34,897
Hispanic $7,146 $8,937 $1,321 $1,281 $281 $1,491 $1,496 $21,954
Black non-Hispanic $6,594 $10,552 $2,732 $2,622 $136 $1,424 $1,919 $25,978
White non-Hispanic $10,816 $12,958 $3,065 $3,457 $1,162 $4,653 $1,549 $37,659
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: See box in text defining income sources.

Source: EPI analysis of U.S. Census Bureau's Current Population Survey Annual Social and Economic Supplement microdata

Social Security benefits constitute over a third of income for white, black, and Hispanic seniors alike, but the share is greater for blacks and Hispanics (both 41 percent) than non-Hispanic whites (34 percent). More than one-fifth (21 percent) of black seniors’ income comes from public and private pensions, compared with 17 percent for non-Hispanic whites and 12 percent for Hispanics. Hispanics have the highest share (33 percent) of earned income. Retirement account distributions and asset income are much less important for minority seniors than white non-Hispanic seniors.

Share this chart:

Copy the code below to embed this chart on your website.

Social Security is important to all education groups (and 401(k)s and IRA distributions are not): Mean annual income of people 65 and older by source and education, 2014

26
 Education level Earnings  Social Security  Public pensions Private pensions  401(k)s, IRAs, etc.  Asset income  Other 
All $10,119 $12,232 $2,840 $3,131 $966 $4,019 $1,590
Less than high school $2,740 $9,937 $453 $1,024 $94 $952 $1,436
High school diploma/GED $5,395 $12,390 $1,522 $2,471 $488 $2,073 $1,316
Some college $9,223 $12,643 $2,423 $3,029 $871 $3,247 $1,872
Bachelor’s degree or more $21,360 $13,018 $6,312 $5,311 $2,181 $9,013 $1,795 
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: See box in text defining income sources.

Source: EPI analysis of U.S. Census Bureau's Current Population Survey Annual Social and Economic Supplement microdata

Seniors with a college degree have more than twice as much earned income as the average senior. Social Security benefits are fairly equally distributed in dollar terms but constitute a much larger share of income for the least educated group. Public and private pension benefits are a similar share of income (16–20 percent) for all workers with at least a high school degree. Retirement account distributions are much less important than pension benefits for all education groups.

Share this chart:

Copy the code below to embed this chart on your website.

Single women rely most on Social Security: Mean annual income of people 65 and older by source, gender, and marital status, 2014

27
Earnings  Social Security  Public pensions  Private pensions  401(k)s, IRAs, etc.  Asset income  Other 
All $10,119 $12,232 $2,840 $3,131 $966 $4,019 $1,590
Unmarried woman $4,901 $12,219 $2,209 $2,187 $507 $2,803 $1,243
Unmarried man $9,325 $13,526 $3,578 $4,240 $1,178 $4,701 $2,790
Married woman $6,825 $9,304 $1,896 $1,411 $464 $3,510 $542
Married man $17,855 $14,153 $3,905 $4,975 $1,714 $5,279 $2,309

 

ChartData Download data

The data below can be saved or copied directly into Excel.

Note: See box in text defining income sources.

Source: EPI analysis of U.S. Census Bureau's Current Population Survey Annual Social and Economic Supplement microdata

Single women age 65 and older derive nearly half their income from Social Security benefits, while married couples (combining individual incomes) and single men in that age group derive about a third of their income from Social Security benefits. Women, especially single women, have less earned income and lower overall incomes than men age 65 and older.

Share this chart:

Copy the code below to embed this chart on your website.

Social Security is important to all retirees: Social Security benefits of people age 65 and older by family income, race and ethnicity, education, gender, and marital status, 2014

28
Characteristics Percent with benefits (bottom axis)  Median of those with benefits (top axis)
All 82% $14,400
1st (bottom) quintile 75% $11,184
2nd (lower-middle) quintile 90% $14,424
3rd (middle) quintile 89% $15,600
4th (upper-middle) quintile 83% $15,659
5th (top) quintile 74% $16,859
Hispanic 70% $11,459
Black non-Hispanic 78% $12,320
White non-Hispanic 85% $14,939
Less than high school 80% $11,903
High school diploma/GED 86% $13,992
Some college 84% $15,000
Bachelor’s degree or more 77% $16,800
Unmarried women 84% $14,220
Unmarried men 84% $15,936
Married women 81% $10,631
Married men 81% $17,723
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Totals do not include Supplemental Security Income.

Source: EPI analysis of U.S. Census Bureau's Current Population Survey Annual Social and Economic Supplement microdata

Social Security is by far the most important and equitably distributed retirement income source, with more than four out of five seniors receiving benefits. Hispanic and more-educated seniors are more likely to work at older ages and have lower Social Security take-up. Hispanics’ lower take-up also reflects lower participation during working-age years due to immigration, work in the informal sector, and other reasons.

Share this chart:

Copy the code below to embed this chart on your website.

Pensions remain an important source of retirement income: Pension benefits (public or private) of people age 65 and older by family income, race and ethnicity, education, gender, and marital status, 2014

29
Characteristics Percent with benefits (bottom axis)  Median of those with benefits (top axis)
All 29% $13,200
1st (bottom) quintile 8% $2,472
2nd (lower-middle) quintile 25% $6,804
3rd (middle) quintile 35% $12,000
4th (upper-middle) quintile 40% $18,000
5th (top) quintile 38% $28,000
Hispanic 14% $13,000
Black non-Hispanic 25% $14,400
White non-Hispanic 32% $13,200
Less than high school 16% $6,000
High school diploma/GED 28% $10,140
Some college 31% $12,000
Bachelor’s degree or more 37% $23,500
Unmarried women 30% $9,996
Unmarried men 33% $15,600
Married women 19% $10,972
Married men 35% $16,800

 

ChartData Download data

The data below can be saved or copied directly into Excel.

Source: EPI analysis of U.S. Census Bureau's Current Population Survey Annual Social and Economic Supplement microdata

Nearly three in 10 seniors receive income from defined-benefit pensions, and for these seniors the amounts rival Social Security benefits in importance. A diverse group of seniors receives pension benefits, though Hispanics, married women, and seniors without a high-school degree are less likely to receive benefits. Middle-class and upper-income seniors are more likely to receive pensions, partly because pensions push many seniors into these income groups.

Share this chart:

Copy the code below to embed this chart on your website.

Public pensions are a lifeline for black and female retirees: Public pension benefits of people age 65 and older by family income, race and ethnicity, education, gender, and marital status, 2014

30
Characteristics Percent with benefits (bottom axis)  Median of those with benefits (top axis)
All 12% $16,800
1st (bottom) quintile 2% $4,272
2nd (lower-middle) quintile 7% $9,300
3rd (middle) quintile 13% $12,000
4th (upper-middle) quintile 17% $21,600
5th (top) quintile 21% $26,976
Hispanic 6% $18,000
Black non-Hispanic 12% $17,664
White non-Hispanic 13% $16,800
Less than high school 4% $11,184
High school diploma/GED 9% $12,000
Some college 12% $15,000
Bachelor’s degree or more 21% $23,400
Unmarried women 13% $12,300
Unmarried men 13% $21,600
Married women 10% $13,188
Married men 13% $22,740
ChartData Download data

The data below can be saved or copied directly into Excel.

Source: EPI analysis of U.S. Census Bureau's Current Population Survey Annual Social and Economic Supplement microdata

Teaching and other public-sector jobs were a ticket to the middle class for many black and female workers, who gravitated toward jobs with secure pensions even if they paid for these benefits directly or indirectly through lower salaries. As a result, the share of black seniors receiving public-sector pension benefits is nearly as high as the share of non-Hispanic white seniors, and the share of female seniors receiving these benefits is almost as high as the share of male seniors. Female retirees, however, receive smaller pensions than male retirees, reflecting their lower earnings and shorter careers.

Share this chart:

Copy the code below to embed this chart on your website.

Private-sector pensions are an important income source for many workers without college degrees: Private pension benefits of people age 65 and older by family income, race and ethnicity, education, gender, and marital status, 2014

31
Characteristics Percent with benefits (bottom axis) Median of those with benefits (top axis)
All 22% $9,000
1st (bottom) quintile 7% $2,040
2nd (lower-middle) quintile 19% $5,520
3rd (middle) quintile 28% $9,348
4th (upper-middle) quintile 30% $12,456
5th (top) quintile 26% $18,000
Hispanic 10% $7,350
Black non-Hispanic 18% $10,800
White non-Hispanic 24% $9,000
Less than high school 14% $4,800
High school diploma/GED 23% $7,332
Some college 24% $9,000
Bachelor’s degree or more 24% $15,120
Unmarried women 22% $6,696
Unmarried men 25% $11,040
Married women 13% $6,804
Married men 28% $12,000
ChartData Download data

The data below can be saved or copied directly into Excel.

Source: EPI analysis of U.S. Census Bureau's Current Population Survey Annual Social and Economic Supplement microdata

Nearly one in four seniors with a high-school degree but no college education receives a private-sector defined-benefit pension. This is almost as high as the share of college-educated seniors who receive these pensions, reflecting the high priority unions place on secure retirement benefits. However, Hispanic seniors, seniors lacking high-school degrees, and married women are less likely to receive private-sector pension benefits.

Share this chart:

Copy the code below to embed this chart on your website.

Retirement account savings are still not a significant source of income for seniors: 401(k), IRA, and Keogh plan income of people age 65 and older by family income, race and ethnicity, education, gender, and marital status, 2014

32
Characteristics Percent with benefits (bottom axis) Median of those with benefits (top axis)
All 8% $5,400
1st (bottom) quintile 2% $1,500
2nd (lower-middle) quintile 5% $2,500
3rd (middle) quintile 9% $4,800
4th (upper-middle) quintile 11% $5,988
5th (top) quintile 12% $10,000
Hispanic 3% $6,000
Black non-Hispanic 3% $3,000
White non-Hispanic 9% $5,500
Less than high school 2% $2,400
High school diploma/GED 7% $4,000
Some college 9% $5,000
Bachelor’s degree or more 12% $9,000
Unmarried women 6% $3,792
Unmarried men 8% $8,000
Married women 7% $3,000
Married men 10% $8,000
ChartData Download data

The data below can be saved or copied directly into Excel.

Source: EPI analysis of U.S. Census Bureau's Current Population Survey Annual Social and Economic Supplement microdata

Fewer than one in 10 seniors receive retirement account distributions, and the median amount is $5,400. Since some retirees withdraw all their savings at once, the small share receiving distributions in any given year is not by itself a good indicator of the importance of this income source. However, this fact combined with the modest amount shows that three decades into the 401(k) revolution, defined-contribution plans are still not a significant source of retirement income.

Share this chart:

Copy the code below to embed this chart on your website.

Conclusion

The trends exhibited in these figures paint a picture of increasingly inadequate savings and retirement income for successive generations of Americans—and growing disparities by income, race, ethnicity, education, and marital status. Women, who by some measures are narrowing gaps with men, remain much more vulnerable in retirement due to lower lifetime earnings and longer life expectancies.

Decades after the number of active participants in 401(k)-style plans edged out those in traditional pensions, 401(k)s are not delivering substantial income in retirement, and that income is not equally shared.

Retirement security has also been affected by changes in Social Security, notably the gradual increase in the normal retirement age and other benefit cuts implemented in 1983; by broader income and wealth trends, such as growing earnings inequality and the collapse of the stock and housing bubbles; and by other factors, such as trends in out-of-pocket medical costs. A description of these broader trends is outside the scope of this chartbook. However, we can assume that as the value of employer-based retirement plans is declining and retirement savings are growing more unequal, retirement security is declining and growing more unequal, since there is little evidence of countervailing trends.

The shift from pensions to account-type savings plans has been a disaster for lower-income, black, Hispanic, non-college-educated, and single workers, who together add up to a majority of the American population. But even among upper-income white college-educated married couples, many do not have adequate retirement savings or benefits. The evidence presented in this chartbook—that the retirement system does not work for most workers—underscores the importance of preserving and expanding Social Security, defending defined-benefit pensions for workers who have them, and seeking new solutions for those who do not.

— The author thanks Tanyell Cooke for research assistance.

References

Board of Governors of the Federal Reserve System. 2015. Financial Accounts of the United States. December 10 release and historical data.

Board of Governors of the Federal Reserve System. 2010. Survey of Consumer Finances full public microdata.

Bricker, Jesse, Lisa J. Dettling, Alice Henriques, Joanne W. Hsu, Kevin B. Moore, John Sabelhaus, Jeffrey Thompson, and Richard A. Windle. 2014. “Changes in U.S. Family Finances from 2010 to 2013: Evidence from the Survey of Consumer Finances. Federal Reserve Bulletin, vol. 100, no. 4.

Chetty, Raj, John Friedman, Soren Leth-Petersen, Torben Nielsen, and Tore Olsen. 2014. “Active vs. Passive Decisions and Crowd-out in Retirement Savings Accounts: Evidence from Denmark.” Quarterly Journal of Economics vol. 129, no. 3, 1141–1219.

Congressional Budget Office (CBO). 2013. The Distribution of Major Tax Expenditures in the Individual Income Tax System. CBO Publication No. 4308.

Current Population Survey basic monthly microdata. Various years. Survey conducted by the Bureau of the Census for the Bureau of Labor Statistics [machine-readable microdata file]. Washington, D.C.: U.S. Census Bureau.

DeNavas-Walt, Carmen, and Bernadette D. Proctor. 2015. Income and Poverty in the United States: 2014. United States Census Bureau.

Flood, Sarah, Miriam King, Steven Ruggles, and J. Robert Warren. 2015. Integrated Public Use Microdata Series, Current Population Survey: Version 4.0. [Machine-readable database]. Minneapolis: University of Minnesota.

Morrissey, Monique. 2009. Toward a Universal, Secure, and Adequate Retirement System. Retirement USA Conference Report.

Morrissey, Monique. 2016. “White Workers Have Nearly Five Times as Much Wealth in Retirement Accounts as Black Workers.” Economic Policy Institute Snapshot, February 18.

Morrissey, Monique, and Natalie Sabadish. 2013. The Retirement Inequality Chartbook. Economic Policy Institute Report.

Reno, Virginia P., Thomas N. Bethell, and Elisa A. Walker. 2011. “Social Security Beneficiaries Face 19% Cut; New Revenue Can Restore Balance.” National Academy of Social Insurance.

Sahadi, Jeanne. 2001. “The 401(k) Turns 20.” CNN Money. January 4.

Semega, Jessica, and Edward Welniak, Jr. 2015. “Expert Meeting on Income, Poverty, and Health Insurance (Presentation).” United States Census Bureau.

Social Security Trustees. 2015. The 2015 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Federal Disability Insurance Trust Funds.

Tong, Scott. 2013. “Father of Modern 401(k) Says it Fails Many Americans.” Marketplace, American Public Media. June 13.

U.S. Department of Labor. 2015. Private Pension Plan Bulletin Historical Tables and Graphs 1975–2013.


See related work on Retirement

See more work by Monique Morrissey