Individual Account Retirement Plans: An Analysis of the 2010 Survey of Consumer Finances

Transcription

1 September 2012 No. 375 Individual Account Retirement Plans: An Analysis of the 2010 Survey of Consumer Finances By Craig Copeland, Ph.D., Employee Benefit Research Institute A T A G L A N C E The share of American families with a member in any employment-based retirement plan from a current employer increased steadily from 38.8 percent in 1992 to 40.6 percent in 2007, before declining in 2010 to 37.9 percent. Ownership of 401(k)-type plans among families participating in a retirement plan more than doubled from 31.6 percent in 1992 to 79.5 percent in 2007, and increased again in 2010 to 82.1 percent. While overall retirement plan participation by families declined from , the percentage of family heads who were eligible for defined contribution plans and chose to participate held essentially stable at 78.2 percent in The percentage of families owning an individual retirement account (IRA) or Keogh plan declined from 30.6 percent in 2007 to 28.0 percent in In addition, the percentage of families with a retirement plan from a current employer, a previous employer s defined contribution plan, or an IRA/Keogh declined from 66.2 percent in 2007 to 63.8 percent in Rollover IRAs account for 43.2 percent of all IRA and Keogh assets, with regular IRA assets following very closely behind at 42.7 percent, while Roth IRAs account for 11.1 percent and Keoghs 3.0 percent. Among total IRA assets, rollover IRAs account for 44.5 percent of assets, regular IRAs 44.1 percent, and Roth IRAs 11.4 percent. Therefore, rollover IRAs account for a larger share of assets than regular IRAs, while the two together account for just under 90 percent of the IRA assets. Defined contribution retirement plan balances accounted for 58.1 percent of families total financial assets in 2007, and that share grew to 61.4 percent in Defined contribution and/or IRA/Keogh balances increased their share as well, from 64.1 percent of total family financial assets in 2007 to 65.7 percent in While regular IRAs account for the largest percentage of IRA ownership, rollover IRAs had a larger share of assets than regular IRAs in This and other EBRI studies show that many Americans are facing the likelihood of not having sufficient income in retirement unless they increase their savings, work longer, or significantly decrease their expenditures in retirement if they hope to make ends meet. A monthly research report from the EBRI Education and Research Fund 2012 Employee Benefit Research Institute

2 Craig Copeland is a senior research associate at the Employee Benefit Research Institute (EBRI). This Issue Brief was written with assistance from the Institute s research and editorial staffs. Any views expressed in this report are those of the author and should not be ascribed to the officers, trustees, or other sponsors of EBRI, EBRI-ERF, or their staffs. Neither EBRI nor EBRI-ERF lobbies or takes positions on specific policy proposals. EBRI invites comment on this research. Copyright Information: This report is copyrighted by the Employee Benefit Research Institute (EBRI). It may be used without permission but citation of the source is required. Recommended Citation: Craig Copeland, Individual Account Retirement Plans: An Analysis of the 2010 Survey of Consumer Finances, EBRI Issue Brief, no. 375, September Report availability: This report is available on the Internet at Table of Contents Introduction... 4 Employment-Based Retirement Plan Participation.. 5 Defined Contribution Plan Participation Rates of Family Heads. 6 IRA/Keogh Participation All Families.. 9 Retirement Plan Participation From Any Source All Families Individual Account Plan Balances.. 11 Median Values of IRA/Keogh Plans Median Value of Defined Contribution Plans Among Current Job Participants All Families Percentage of Financial Assets from Defined Contribution and IRA/Keogh Plans Relative Importance of IRAs vs. Employment-Based Individual Account Plans The Impact of Rollovers on the Relative Values of IRAs Conclusion..20 References. 23 Endnotes. 24 Figures Figure 1, U.S. Private Sector Retirement Plan and IRA Assets, Figure 2, Percentage of All Families With an Employment-Based Retirement Plan, by Various Demographic Categories, 1992, 2007, and Figure 3, Percentage of Families With Head Under Age 65 and a Worker With an Employment-Based Retirement Plan, by Various Demographic Categories, 1992, 2007, and Figure 4, Participation Rates of Family Heads Eligible for an Employment-Based Defined Contribution Plan, 1995, 2004, 2007, and Figure 5, Percentage of Families With an IRA/Keogh by Various Demographic Categories, 1992, 2004, 2007, and Figure 6, Percentage of All Families With a Retirement Plan From a Current or Previous Employer or an IRA/Keogh Plan, 2004, 2007, and ebri.org Issue Brief September 2012 No

3 Figure 7, Median Defined Contribution Plan Balances for Families With a Current Employer Defined Contribution Plan, by Various Demographic Categories, 1992, 2004, 2007, and Figure 8, Median Percentage of Financial Assets in Employment-Based Defined Contribution Plans and IRAs for All Families With the Specified Asset by Various Demographic/Economic Categories, 2004, 2007, and Figure 9, Average Balances for All Families With an Individual Account Plan, by Various Demographic Categories, Figure 10, Percentage of Families' IRA Ownership, by IRA Type or Combination of IRA Types, Figure 11, Percentage of IRA Assets Owned by Families, by IRA Type or Combination of IRA Types, Figure 12a, Percentage of Families With an IRA/Keogh and Percentage With Certain Types of IRAs, by Various Demographic Categories, Figure 12b, Average Balance of Families' IRA/Keoghs and Certain Types of IRAs and Ratio of Rollover IRA Balances to Nonrollover Balances, by Various Demographic Categories, Figure 13a, Percentage of Total IRA and Keogh Assets, by Keogh and IRA Type, Figure 13b, Percentage of Total IRA Assets, by IRA Type, Figure 14, Median and Mean Net Worth and Home Equity Percentage of Net Worth for Families With and Without an Individual Account Retirement Plan, by Family Income and Family Head Age, ebri.org Issue Brief September 2012 No

4 Individual Account Retirement Plans: An Analysis of the 2010 Survey of Consumer Finances By Craig Copeland, Ph.D., Employee Benefit Research Institute Introduction Individual account retirement plans continue to grow in importance as a source of retirement income for future retirees in both the public and private sectors. Individual account plans include employment-based retirement savings plans financed by employer contributions and workers own contributions (most notably, defined contribution plans such as 401(k) plans), as well as non-employment-based plans like individual retirement accounts (IRAs) and Keogh plans for the self-employed. Among public-sector employers, defined benefit (DB) pension plans remain the predominant type of retirement plan, although defined contribution (DC) 401(k)-type options are increasing. Among private-sector employers, the number of DB plans has been declining for many years, as DC plans have increased. Total private-sector DB pension assets were overtaken by defined contribution assets in 1996 and by IRAs in 1998 (Figure 1). Furthermore, many major corporations that still have DB plans have frozen them (for newly hired workers, and, in some cases, for existing workers as well), a trend that makes it ever-more important for most private-sector workers to build their retirement wealth through individual-account savings plans. 1 Consequently, the amount of assets accumulated in these accounts provides an indication of how prepared or unprepared most workers finances will be to supplement the Social Security benefits they will receive in retirement. This Issue Brief assesses the status of American families' accumulations in individual account plans, both through the incidence of ownership and the average amounts accumulated. The Survey of Consumer Finances (SCF), the Federal Reserve Board s triennial survey of wealth, is the basis for this study. SCF is a leading source of data on Americans ebri.org Issue Brief September 2012 No

5 wealth, as it provides detailed information on the incidence of retirement plan ownership and account balances that families have accumulated. 2 Building on the Employee Benefit Research Institute s (EBRI) previous research using prior SCF surveys (i.e., Copeland 2009; Copeland 2006b; Copeland 2003; Copeland and VanDerhei 2000), this study focuses specifically on retirement plan assets, concentrating on the amount that American families have accumulated in individual account retirement plans. 3 Using results from the prior studies, this report shows the changes in assets in these accounts as well as changes in the occurrence of these individual accounts, both inside and outside of employment-based arrangements. Particular attention is paid to ownership of individual retirement accounts, as the questions in SCF allow for estimates of the sources of IRA funds between rollovers and contributions and to determine whether the IRA assets are in traditional IRAs or Roth IRAs. 4 The 2010 SCF shows that the median (midpoint) net worth of American families decreased by 38.8 percent from 2007 to 2010, and the median value of family income also decreased from 2007 to 2010, although at a much smaller rate of 7.7 percent. In contrast, American families median net worth increased 17.7 percent in the prior study period (2004 to 2007), following increases of 1.5 percent from , 10.4 percent from , and 17.6 percent from In the 2010 data, the debt level was virtually unchanged from the prior survey period ( ), leading to only a slight increase in the relative debt payments-to-income ratio for American families. 5 This Issue Brief investigates the percentage of families that participate in various types of retirement plans, including IRAs. Next, it provides both median and mean (average) estimates of the value of the assets in these accounts, as well as the proportion of total financial assets represented and their relative percentages within the retirement plan universe. It then focuses on the value of employment-based rollovers as part of the total IRA market, in order to glean a sense of the full contribution that the employment-based retirement plan system makes to total retirement assets. Overall, the results for the employment-based retirement system in 2010 follow that of the overall decline in the value of families net worth during the recent economic recession, as the percentage of families with an employment-based plan declined and the median account balance decreased. However, the share of assets that these accounts represent held steady or increased from 2007 to While asset accumulation is a vital component of assessing retirement preparedness, it is not the only factor that will determine financial security in retirement. A second vital component is the use of accumulated funds such that retirees do not outlive their assets. Even for workers with DB plans, how and when these assets are spent remains an important decision especially when the lump-sum option is chosen. 6 Because of the growing prevalence of lump-sum distributions from employment-based retirement plans, increasing numbers of workers and retired workers will have the responsibility of managing their assets themselves. Thus, although this Issue Brie focuses on participation in all plans that are considered retirement income sources and the estimation of the median and average amounts of assets in these plans, it must be stressed that this is only an indicator of potential financial security because individuals financial security in retirement will ultimately be determined by the source and amount of retirement resources, how distributions are taken from these sources, how they are invested in the interim, and how fast assets are spent, along with individuals health status and life span. Employment-Based Retirement Plan Participation Employment-based plans are generally categorized as either defined benefit plans (pensions traditional or cash balance) or defined contribution plans (401(k)-type plans). Generally, traditional defined benefit plans provide benefits according to a formula based on the worker s tenure and salary history, and are not directly affected by the changes in the investment returns of the plan assets. Contributions to these plans are generally made by the employer and in some cases (most notably in the public sector) also by the individual participant. So-called hybrid individual account defined benefit plans, most commonly cash balance plans, provide benefits that are generally based on contributions by the sponsor and a credit rate set by the plan. The benefits to the individual cannot be reduced because of negative market returns. 7 ebri.org Issue Brief September 2012 No

6 By contrast, defined contribution plans provide benefits that are determined by the level of contributions (both from the worker and the employer) and any asset returns on these contributions. Workers not eligible for a plan through employment, and in some cases workers wanting to augment employment-based plans, as well as nonworking spouses, can set up an individual retirement account (IRA); and many self-employed workers can establish a Keogh plan to save for retirement. All Families In the 2010 SCF, 37.9 percent of all families included a participant in an employment-based retirement plan (either a defined benefit or defined contribution plan) from a current employer (Figure 2). This is down from 40.6 percent in 2007, and below the 1992 level of 38.8 percent. 8,9 Among those families with a participant in a plan, a significant shift occurred from 1992 to 2010: The percentage having only a defined benefit plan decreased (from 40.0 percent to 17.9 percent), while the percentage of those families participating in only a defined contribution plan surged (from 37.5 percent in 1992 to 61.3 percent in 2010). The percentage of families with participation in both types of plans decreased slightly from 22.5 percent in 1992 to 20.8 percent in Families with a member participating in an employment-based plan were more likely to do so via a 401(k)-type plan: 82.1 percent in 2010, up from 79.5 percent in 2007 and 31.6 percent in However, the percentage of those with both a 401(k)-type plan and a defined benefit plan decreased to 22.2 percent in 2010, compared with 26.8 percent in 2007 and 39.2 percent in Participation in an employment-based retirement plan was strongly linked to family income and the family head s educational level and race. In 2010, only 4.6 percent of families with annual income less than $10,000 had a participant in a plan from a current job, compared with 71.0 percent of the families with income of $100,000 or more. Among families with a head who had not attained a high school diploma, 12.8 percent had a participant in a plan. In contrast, among families with a head who had obtained a college degree or higher, 51.1 percent had a participant in a plan. These differences also existed in 1992 and Furthermore, families with a white, non-hispanic head were more likely to have a plan participant than families with a nonwhite head. In terms of net worth, families within the highest 10 percent of net worth were most likely to have a retirement plan participant in 2010, while the two net worth percentile breaks just below the highest had similar levels of participation to that of the highest net worth families. This was not the case in 1992 and 2007, when families in the lower levels of percentile of net worth were more likely to have a participant than those in the highest level. Families With Family Head Under Age 65 and a Worker The percentage of families with both a family head under age 65 and a worker who participated in an employment-based retirement plan from a current employer decreased from 56.6 percent in 2007 to 52.2 percent in This 2010 percentage is also lower than the 57.1 percent level in 1992 (Figure 3). Over the period, the level of participation for this family type held at 57 percent, except for declines in 2004 and As for changes within the various demographic categories, the results were qualitatively nearly identical with the results for all families participation. Participation levels across the various categories are higher, because families with a member in the workforce were more likely to have an employment-based retirement plan participant than those without a worker. Defined Contribution Plan Participation Rates of Family Heads Overall in 2010, 78.2 percent of defined contribution plan-eligible family heads chose to participate in the plan (conversely, nearly a quarter of eligible family heads chose not to participate). This was essentially unchanged from 78.8 percent in 2007 (Figure 4). 14,15 ebri.org Issue Brief September 2012 No

7 Figure 2 Percentage of All Families With an Employment-Based Retirement Plan From a Currernt Employer, by Various Demographic Categories, 1992, 2007, and Of Those Participating in a Plan Of Those Participating in a Plan Of Those Participating in a Plan Both Of those with Both Of those with Both Of those with Defined Defined defined 401(k)- 401(k)-type Defined Defined Defined 401(k)- 401(k)-type Defined Defined Defined 401(k)- 401(k)-type Any benefit contribution benefit and type plan with Any benefit contribution benefit and type plan with Any benefit contribution benefit and type plan with plan only only dontribution plan a DB Plan only only dontribution plan a DB Plan only only dontribution plan a DB Total 38.8% 40.0% 37.5% 22.5% 31.6% 39.2% 40.6% 17.4% 60.3% 22.3% 79.5% 26.8% 37.9% 17.9% 61.3% 20.8% 82.1% 22.2% Family Income Less than $10, a 16.4 a a 69.0 a a 53.4 a $10,000 $24, a $25,000 $49, $50,000 $99, $100,000 or more Age of Head < Education of Head Below HS diploma HS diploma Some college College degree Race White non-hispanic Nonwhite Working Status of Family Head Someone else Self-employed Retired Other nonwork Net Worth Percentile Bottom 25% Top 10% Source: Employee Benefit Research Institute estimates of the 1992, 2007, and 2010 Survey of Consumer Finances. a No families in this group were found to have these types of plans. Note: All income values are in 2010 $s. The 2007 and 2010 participation levels are not directly comparable to 1992 because of changes in the survey that began in ebri.org Issue Brief September 2012 No

8 Figure 3 Percentage of Families With Head Under Age 65 and a Worker With an Employment-Based Retirement Plan From a Current Employer, by Various Demographic Categories, 1992, 2007, and Of Those Participating in a Plan Of Those Participating in a Plan Of Those Participating in a Plan Both Of those with Both Of those with Both Of those with Defined Defined Defined 401(k)- 401(k)-type Defined Defined Defined 401(k)- 401(k)-type Defined Defined Defined 401(k)- 401(k)-type Any benefit contribution benefit and type plan with Any benefit contribution benefit and type plan with Any benefit contribution benefit and type plan with plan only only dontribution plan a DB plan only only dontribution plan a DB plan only only dontribution plan a DB Total 57.1% 39.6% 37.6% 22.8% 31.9% 39.6% 56.6% 17.3% 59.6% 23.1% 79.5% 27.8% 52.2% 17.7% 61.0% 21.4% 82.4% 22.8% Family Income Less than $10, a 16.4 a a 68.4 a a 59.8 a $10,000 $24, a a $25,000 $49, $50,000 $99, $100,000 or more Age of Head < Education of Head Below HS diploma HS diploma Some college College degree Race White non-hispanic Nonwhite Net Worth Percentile Bottom 25% Top 10% Source: Employee Benefit Research Institute estimates of the 1992, 2007, and 2010 Survey of Consumer Finances. a No families in this group were found to have these types of plans. Note: All income values are in 2010 $s. The 2007 and 2010 participation levels are not directly comparable to 1992 because of changes in the 2004 survey. ebri.org Issue Brief September 2012 No

9 Figure 4 Participation Rates of Family Heads Eligible for an Employment-Based Defined Contribution Plan, 1995, 2004, 2007, and Total 73.8% 75.7% 78.8% 78.2% Family Income Less than $10, $10,000 $24, $25,000 $49, $50,000 $99, $100,000 or more Age of Head < Education of Head Below HS diploma HS diploma Some college College degree Race White non-hispanic Nonwhite Net Worth Percentile Bottom 25% Top 10% Source: Employee Benefit Research Institute estimates of the 1995, 2004, 2007, and 2010 Survey of Consumer Finances. Note: All income values are in 2010 $s. The percentage of family heads who were eligible but not participating in a defined contribution plan, but who were participating in a defined benefit plan, increased in 2010 to 19.7 percent. This percentage had decreased from 45.0 percent in 1995 to 21.8 percent in 2004 and to 14.9 percent in A number of demographic differences have persisted over the four survey years: the increased likelihood of plan participation with higher levels of family income, net worth, and educational attainment. For example, in 2010, the participation rate was just 41.5 percent of family heads with family income of $10,000 $24,999 a year, compared with 91.8 percent for those with family income of $100,000 or more. Additionally, racial disparities were noted: White family heads were more likely to participate when eligible than nonwhite family heads. In 2010, 82.5 percent of white family heads who were eligible participated, compared with 68.6 percent for nonwhite family heads. The age of family head was not a strong factor among those of prime working ages (35 64), but for family heads younger or older than that, the likelihood of participating was much lower. IRA/Keogh Participation All Families The percentage of families that owned either an IRA or a Keogh plan decreased in 2010 to 28.0 percent from 30.6 percent in 2007, though this is still higher than the 26.1 percent level reported in 1992 (Figure 5). Ownership of an IRA increased with family income, the family head s educational level, and the family s net worth. Of families with less than $10,000 a year in income, just 8.4 percent owned an IRA/Keogh in 2010, compared with 59.6 percent of families with income of $100,000 or more. Not surprisingly, the percentage owning an IRA/Keogh increased even more substantially when measured by net worth: Only 5.6 percent of those in the lowest 25 th percentile of net worth owned an IRA/Keogh, compared with 76.9 of those in the top 10 percent. These differences were consistent over the years of the study. The ownership of IRA/Keoghs also increased with the family head s age through age 64, but families with the oldest heads had a lower likelihood of owning an IRA/Keogh than those whose heads were ages Families with a white family head were significantly more likely to own an IRA/Keogh in 2010 than those with nonwhite heads (35.4 percent versus 12.8 percent), and have been since ebri.org Issue Brief September 2012 No

10 Figure 5 Percentage of Families With an IRA/Keogh by Various Demographic Categories, 1992, 2004, 2007, and 2010 Percentage With IRA/Keogh Median Value of Those With an IRA/Keogh Total 26.1% 29.1% 30.6% 28.0% $19,446 $34,521 $34,574 $40,000 Family Income Less than $10, ,964 5,754 31,431 48,000 $10,000 $24, ,427 14,959 12,572 20,000 $25,000 $49, ,910 19,562 26,193 24,000 $50,000 $99, ,557 24,740 31,431 30,000 $100,000 or more ,300 70,193 54,480 99,000 Age of Head < ,734 9,206 8,696 10, ,682 23,014 25,145 19, ,539 40,274 37,717 40, ,881 69,041 68,101 60, ,331 86,302 69,148 85, ,464 34,521 36,670 53,000 Education of Head Below HS diploma ,132 14,038 26,193 32,000 HS diploma ,715 20,137 25,669 24,000 Some college ,198 21,863 26,193 30,000 College degree ,298 49,480 41,908 55,100 Race White, non-hispanic ,748 36,822 40,860 46,000 Nonwhite ,891 14,959 17,811 26,000 Working Status of Head Someone else ,198 24,164 26,193 26,000 Self-employed ,364 57,534 70,196 75,000 Retired ,781 50,630 50,290 65,000 Other nonwork ,648 26,466 31,431 20,000 Net Worth Percentile Bottom 25% ,488 2,877 3,143 5, ,066 8,745 11,525 9, ,166 17,260 26,821 22, ,298 51,781 44,003 60,000 Top 10% , , , ,000 Source: Employee Benefit Research Institute estimates of the 1992, 2004, 2007, and 2010 Survey of Consumer Finances. Note: All income and asset values are in 2010 $s. Retirement Plan Participation From Any Source All Families In 2010, 63.8 percent of families had a participant in a current or previous employer s retirement plan or an IRA/Keogh (Figure 6). The percentage participating in these plans increased with family income, net worth, and educational level of the family head. Families with a white family head were more likely to have participated in one of these plans. Less than 50 percent of families with a head under age 35 participated in one of these plans, though this increased to 73.9 percent for families with a head ages 55 64, before decreasing to 70.3 percent of families with a head age 75 or older. 16 In 2007, 66.2 percent of all families had a participant in a current or previous employer s retirement plan or an IRA/Keogh. The same differences across demographic groups discussed above for 2010 were also present in the 2007 SCF. In most cases, the 2007 levels of participation in these plans were higher than they were in ebri.org Issue Brief September 2012 No

11 Individual Account Plan Balances Median Values of IRA/Keogh Plans Among families with an IRA/Keogh plan, the median value of their account was $40,000 in 2010, a 16 percent increase from $34,574 in 2007 (Figure 5). 17 From , the median IRA/Keogh balance increased 78 percent, from $19,446 in Figure 6 Percentage of All Families With a Retirement Plan From a Current or Previous Employer or an IRA/Keogh Plan, 2004, 2007, and Total 65.4% 66.2% 63.8% Family Income Less than $10, $10,000 $24, $25,000 $49, $50,000 $99, $100,000 or more Age of Head < Education of Head Below HS diploma HS diploma Some college College degree Race White non-hispanic Nonwhite Working Status of Head Someone else Self-employed Retired Other nonwork Net Worth Percentile Bottom 25% Top 10% Source: Employee Benefit Research Institute estimates of the 2004, 2007, and 2010 Survey of Consumer Finances. Note: All income values are in 2010 $s. The factors related to higher median IRA/Keogh balances were the same as those linked with higher defined contribution plan balances, but with a few exceptions. A higher educational level of the family head was associated with higher median IRA/Keogh balances, but no clear pattern was evidenced in the median defined contribution plan balance until the jump for the most educated family heads. Median Value of Defined Contribution Plans Among Current Job Participants All Families Among all families with a defined contribution plan in 2010, the median (mid-point) plan balance was $29,000 (Figure 7). 18, 19 This is a 124 percent increase from the 1992 value of $12,964 (expressed in constant 2010 dollars) but an 8 percent decrease from the 2007 value of $31,431 (2010 dollars). 20 The median balance in 2010 increased with family income, family head age, and family net worth a pattern that held true in 2004 and Families with a white family head had a higher median balance: In 2010, the median balance for defined contribution plans of families with a white family head was $33,000, compared with $20,000 for families with a nonwhite head. The median defined contribution plan balance increased in 2010 for families headed by an individual with no high school diploma, while families headed by individuals in the other education categories enjoyed higher median balances in 2010 but experienced a decrease from A noteworthy finding when examining balances across the types of employment-based retirement plans is that families with both a defined benefit (pension) plan and a 401(k)-type plan had a higher median balance in the 401(k)-type plan than families with only a 401(k)-type plan. In 2010, the median balance in a 401(k)-type plan for families with a defined benefit plan and a 401(k)-type plan was $41,000, while the median balance for families with only a 401(k)-type plan was $25,000. This difference also occurred in the three prior survey years and was found in all the demographic breakdowns in 2010, except for heads of families with less than high school diplomas and families in the percentile of net worth. This suggests that families that participate in both types of plans (a defined benefit and a 401(k)-type plan) tend to either: (1) be more highly compensated and are therefore more able to save, (2) have a more generous defined contribution plan, and/or (3) have higher aptitudes for saving. ebri.org Issue Brief September 2012 No

12 401(k)- 401(k)- 401(k)- 401(k)- 401(k)- 401(k)- 401(k)- 401(k)- 401(k)- type plan type 401(k)- type plan type 401(k)- type plan type 401(k)- type plan type type With Defined plan type With Defined plan type With Defined plan type With Defined plan plan Benefit Plan Only plan Benefit Plan Only plan Benefit Plan Only plan Benefit Plan Only Total $12,964 $23,336 $9,075 $25,315 $41,425 $20,713 $31,431 $58,671 $24,097 $29,000 $41,000 $25,000 Family Income $10,000 $24,999 2,593 a 2,593 2,301 1,036 2,301 2,829 a 2,829 2,600 a 2,600 $25,000 $49,999 4,149 7,779 3,241 7,940 17,261 6,444 6,286 12,572 5,867 7,000 10,000 7,000 $50,000 $99,999 12,964 15,557 10,371 19,562 23,129 17,261 26,193 36,670 23,049 24,000 26,000 22,500 $100,000 or more 46,671 51,857 35,003 78,248 93,207 70,193 89, ,627 80, , , ,000 Age of Head <35 7,583 12,132 6,066 10,356 14,959 8,745 8,696 15,192 7,439 10,000 13,000 8, ,616 18,198 8,341 27,617 41,425 19,562 30,383 41,908 24,097 29,000 30,000 27, ,464 54,596 24,265 57,534 86,302 46,028 52,385 76,482 47,147 50,000 65,000 47, ,364 57,629 14,711 69,041 93,206 57,534 84, ,247 80,673 58,000 64,000 53,000 Education of Head Below HS diploma 4,550 a 3,943 16,110 40,274 10,356 10,477 28,812 10,477 11,000 7,500 11,000 HS diploma 9,099 12,132 7,583 19,562 34,521 16,110 24,830 31,431 20,954 18,000 30,000 14,800 Some college 19,715 21,231 9,554 14,959 25,315 12,658 22,002 38,660 13,851 16,000 35,000 15,000 College degree 21,232 40,947 12,132 46,028 80,548 36,822 52,385 83,816 48,194 46,300 53,000 44,000 Race White Non-Hispanic 15,166 27,298 10,616 33,370 57,534 23,014 44,003 73,339 31,431 33,000 45,000 29,000 Nonwhite 14,862 30,331 8,796 17,260 25,315 14,959 23,048 38,660 15,716 20,000 30,000 17,000 Net Worth Percentile Bottom 25% 1,516 a 1,517 3,567 6,904 2,877 2,619 10,058 2,410 5,000 8,600 4, ,583 9,099 6,066 12,658 19,562 11,507 13,620 15,716 12,572 10,000 15,000 10, ,924 19,715 15,166 34,521 43,726 27,617 47,147 57,624 41,908 38,000 45,000 36, ,464 42,464 30,331 97, ,672 80, , , , , , ,000 Top 10% 121, , , , , , , , , , , ,000 Source: Employee Benefit Research Institute estimates of the 1992, 2004, 2007, and 2010 Survey of Consumer Finances. a Sample size not sufficient for a reliable estimate. Note: All income and asset values are in 2010 $s. Figure 7 Median Defined Contribution Plan Balances for Families With a Current Employer Defined Contribution Plan, by Various Demographic Categories, 1992, 2004, 2007, and ebri.org Issue Brief September 2012 No

13 Percentage of Financial Assets from Defined Contribution and IRA/Keogh Plans The importance of individual account plans to the total financial assets of families with these plans can be measured by the percentage of financial assets that the plans represent. In 2010, defined contribution plan assets comprised 61.4 percent of total financial assets for the median family that participated in such a plan a 3.3 percentage point increase from 2007, when they represented 58.1 percent of the families total financial assets (Figure 8). Among families with defined contribution and/or IRA/Keogh plans in 2010, these accounts represented a median of 65.7 percent of the families total financial assets also an increase from the 2007 median percentage of 64.1 percent. Groups deriving a larger percentage of their financial assets from defined contribution plans or IRA/Keoghs include those that have a lower likelihood of having any retirement plan, such as families with a family head with lower educational attainment, nonwhite, or between the ages of 35 44; or with lower net worth. Phrased another way, if families with low education or are relatively young had a defined contribution plan or IRA/Keogh at all, the assets in these plans were likely to constitute a large share of their total assets. For example, in 2010, defined contribution assets represented a median of 84.0 percent of total financial assets for those with defined contribution plans among those families with a head without a high school diploma, compared with 48.2 percent for families with a head with a college degree. Furthermore, among families with heads ages 35 44, 75.3 percent of their financial assets are from defined contribution plans or IRA/Keoghs if they have such assets, whereas among families with family heads ages this figure was 52.9 percent. Relative Importance of IRAs vs. Employment-Based Individual Account Plans Figure 9 provides evidence of the relative importance of IRAs vs. employment-based individual account plans (such as a 401(k) plan) for Average values are reported in Figure 9, as opposed to the median (mid-point) values in Figures 5 and 7, because in many of the groups analyzed, less than 50 percent of the families hold an account balance in one or more of the components. 21 This is the case because the families need only have a balance in one of the accounts, rather than in each account, to be included for analysis in some columns in Figure 9. Figure 9 provides information on average balances for all families with a defined contribution retirement plan (current or previous employer) and/or an IRA/Keogh for The first three columns provide average balances for families with a positive account balance for that type of plan. For example, among families with either an IRA or Keogh plan, the average account balance in 2010 was $148,711. For the relatively small percentage of families with account balances in a previous employer s defined contribution plan, the average balance was $126,668. The current-employer defined contribution plan balance was $121, From a public policy perspective, it may be more useful to consider a family s entire portfolio of retirement assets from individual account plans. Columns four through seven in Figure 9 show similar information as the first three columns, although in this case averages are computed for any family with a positive combined account balance in these sources. Column four provides the sum of defined contribution plans with a current employer, any account balances that may have been left in a defined contribution plan of a previous employer, or any of the family s IRA/Keogh assets. The total average retirement portfolio account balance for all families with any type of retirement account in 2010 was $173,232. The final three columns in Figure 9 show the relative importance for each of the individual account components. Among families owning an individual account plan of any type in 2010, more than half (51.2 percent) were attributable to defined contribution plans with their current employer, the overall average share of assets attributable to IRA/Keogh balances was 41.6 percent, and account balances remaining in a previous employer s defined contribution plan accounted for just over 7 percent of the total. Further exploration of the relative share of individual account balances maintained in IRAs/Keoghs reveals some clear demographic and economic influences. First, there was a marked tendency for lower-income families to have larger percentages of their assets in IRAs/Keoghs: Families with incomes between $10,000 $24,999 had most (59.4 percent) of their assets in IRAs/Keoghs in 2010, whereas families with incomes of $100,000 or more had 40.1 percent of their assets in IRAs/Keoghs. Families headed by a white, non-hispanic had a larger percentage of their individual account plan assets in IRAs/Keoghs than their nonwhite counterparts (45.7 percent vs percent) in Those families ebri.org Issue Brief September 2012 No

14 Figure 8 Median Percentage of Financial Assets in Employment-Based Defined Contribution Plans and IRAs for All Families With the Specified Asset by Various Demographic/ Economic Categories, 2004, 2007, and Of Those Of Those Of Those Of Those Of Those Of Those With a DC With a DC With a DC With a DC With a DC With a DC Balance or IRA Balance Balance or IRA Balance Balance or IRA Balance Percentage Percentage Percentage Percentage Percentage Percentage of financial of financial of financial of financial of financial of financial assets DC assets DC & assets DC assets DC & assets DC assets DC & balance IRA balances balance IRA balances balance IRA balances represents* represent* represents* represent* represents* represent* Total 52.9% 59.6% 58.1% 64.1% 61.4% 65.7% Family Income Less than $10, $10,000 $24, $25,000 $49, $50,000 $99, $100,000 or more Age of Head < Education of Head Below HS diploma HS diploma Some college College degree Race White non-hispanic Nonwhite Working Status of Head Someone else Self-employed Retired Other nonwork Net Worth Percentile Bottom 25% Top 10% Source: Employee Benefit Research Institute estimates of the 2004, 2007, and 2010 Survey of Consumer Finances. * Includes DC balances with both current and previous employers. IRA balances include Keogh assets. Note: All income values are in 2010 $s ebri.org Issue Brief September 2012 No

15 Figure 9 Average Balances for All Families With an Individual Account Plan, by Various Demographic Categories, 2010 Defined Defined With Defined Defined Defined Defined contribution, contribution, any type contribution, contribution, contribution, contribution, IRA/ previous current of account previous current IRA/ current previous IRA/ Keogh employer employer balance employer employer Keogh employer employer Keogh Total $148,711 $126,668 $121,387 $173,232 $77,416 $9,439 $86, % 7.2% 41.6% Family Income Less than $10,000 a a a a a a a a a a $10,000 $24,999 55,468 49,797 11,209 46,661 2,952 3,662 40, $25,000 $49,999 99,849 68,725 22,805 69,071 13,243 5,073 50, $50,000 $99, ,118 87,334 53,564 91,850 34,941 5,456 51, $100,000 or more 230, , , , ,932 18, , Age of Head <35 19,628 18,621 24,284 27,419 19,478 1,018 6, ,070 39,509 65,888 85,156 54,710 2,888 27, ,122 85, , , ,148 9,396 63, , , , , ,941 15, , , , , ,199 58,084 24, , , ,586 33, ,636 2,072 5, , Education of Head Below HS diploma 43,353 24,918 29,418 35,266 18, , HS diploma 75,498 69,116 56,304 75,668 37,394 3,469 34, Some college 84,553 73,597 60,773 86,201 39,493 4,921 41, College degree 192, , , , ,953 14, , Race White non-hispanic 157, , , ,591 84,982 10, , Nonwhite 95,594 74,294 67,387 93,975 50,596 5,855 37, Working Status of Head Someone else 99,415 78, , ,487 85,288 6,018 45, Self-employed 224, , , , ,733 18, , Retired 208, ,712 92, ,777 13,888 16, , Other nonwork 86,407 89,989 45,822 88,536 18,420 10,644 59, Net Worth Percentile Bottom 25% 8,334 8,106 11,098 11,321 8, , ,595 14,240 16,896 17,993 12, , ,730 34,301 52,809 56,353 34,826 2,442 19, , , , ,120 77,303 11,073 79, Top 10% 387, , , , ,967 34, , Source: Employee Benefit Research Institute estimates of the 2010 Survey of Consumer Finances. a Sample size not sufficient for a reliable estimate. Note: All income and asset values are in 2010 $s. Of Those Participating in Specific Plan Of Those Participating in Any DC Plan or IRA Composition of Personal Account Plans ebri.org Issue Brief September 2012 No

16 headed by an older individual were more likely to have a larger percentage of assets from IRAs/Keoghs than those headed by younger individuals. For example, families with a head ages had a median percentage of assets of 50.2 percent, compared with 23.7 percent for those with a head under 35 years old. Families with lower net worth were more likely to have a larger share of assets in defined contribution plans from a current employer than those with more net worth. The Impact of Rollovers on the Relative Values of IRAs An important issue for policymakers and others is determining how much money will be available from the various types of IRAs Roth, rollover, and regular IRAs. 23 Measuring the amount of IRAs attributable to rollovers is particularly important in ascertaining the full impact of wealth generated within the employment-based retirement plan system, because rollover IRAs are primarily funded by assets generated in other types of retirement plans (notably pensions or 401(k)s). The 2004 SCF improved the ability to determine the source of IRA dollars and provides the first dataset of the distribution of assets among various IRA types held by families. The same format was retained in the 2007 and 2010 surveys, permitting a longitudinal comparison. The analysis starts by examining family ownership of an IRA, but not a Keogh plan, 24 to determine what type of plans or combination of plans they own. 25 The most commonly owned IRA type was the regular IRA: 37.8 percent of the family heads who owned an IRA owned only that type (Figure 10). The next-most commonly owned IRA type was the rollover, at 20.4 percent, and the third-most common type was the Roth IRA, at 18.2 percent. In terms of combined holdings, Roth and regular IRAs owned together accounted for 8.0 percent of IRA-type assets held by families; rollover and regular IRAs accounted for 7.5 percent; Roth and rollover IRAs, 5.4 percent; and rollover, regular, and Roth IRAs, 2.8 percent. When the breakdown of IRA types is done by the amount of assets, the relative percentages differ significantly from the ownership percentages. The regular IRA-only type still has the largest percentage, but on an asset basis it is relatively lower, at 27.9 percent (Figure 11). The rollover IRAs-only percentage increases to 22.2 percent and the percentage for rollover and regular IRAs grows tremendously, relative to ownership, to 18.2 percent on an asset basis. 26 The Roth IRAs-only have an analogous decline in asset basis, making up only 6.3 percent of assets, compared with the 18.2 percent of ownership. Further investigation of families who own either an IRA or a Keogh indicates that 98.1 percent of those owning either of these arrangements own only an IRA (Figure 12a). This percentage is very close across all the demographic groups examined, with no group having a percentage lower than 96.0 percent. Furthermore, 76.3 percent of families that have an IRA own just one type of IRA. Again, this is consistent over the various demographic groups, with no group having a percentage below 67.5 percent. The most commonly owned IRA is the regular IRA (56.1 percent), followed by the rollover (36.1 percent) and the Roth (34.3 percent). Regular IRAs for the most part have the highest ownership percentage of IRA types across all demographic groups. However, Roth IRAs emerge as the most prevalent for those families with heads younger than age 45 or with the lowest family income. These patterns hold true for the ownership of these IRA types singularly. The average IRA/Keogh balance for families who own either an IRA or a Keogh plan was $148,759 in 2010, while for those who own only an IRA the average balance was $143,817 (Figure 12b). 27 For those owning only an IRA and only one type of IRA, the average balance was $107,076, compared with $267,429 for those who own a combination of IRAs. Those who had a rollover IRA with or without another IRA had the highest average balance at $225,935. Those having a rollover IRA had the highest average balances across all the demographic groups, except for heads of families who were under age 35 or older than 75 and for families in the net-worth percentile. Average balances for each category follow the expected patterns across demographic groups increasing with age (peaking at age 74), education, and net worth, and higher for families with heads who are white, non-hispanic. The average IRA balance for families in the $50,000 $99,999 family-income group was below the levels of those families with incomes above and below this amount for certain IRA types, but above that of those in the income groups below for other IRA types. Not surprisingly, the balances were always below that of upper-income families. ebri.org Issue Brief September 2012 No

17 Figure 10 Percentage of Families' IRA Ownership, by IRA Type or Combination of IRA Types, 2010 Rollover and Regular IRAs 7.5% Rollover, Regular, and Roth IRAs 2.8% Roth and Regular IRAs 8.0% Roth and Rollover IRAs 5.4% Regular IRA Only 37.8% Roth IRA Only 18.2% Rollover IRA Only 20.4% Source: Employee Benefit Research Institute estimates of the 2010 Survey of Consumer Finances. Figure 11 Percentage of IRA Assets Owned by Families, by IRA Type or Combination of IRA Types, 2010 Rollover, Regular, and Roth IRAs 9.7% Rollover and Regular IRAs 18.2% Regular IRA Only 27.9% Roth and Regular IRAs 9.6% Roth and Rollover IRAs 6.2% Roth IRA Only 6.3% Rollover IRA Only 22.2% Source: Employee Benefit Research Institute estimates of the 2010 Survey of Consumer Finances. ebri.org Issue Brief September 2012 No

18 Figure 12a Percentage of Families With an IRA/Keogh and Percentage With Certain Types of IRAs, by Various Demographic Categories, 2010 Of Those With IRA/Keogh Of Those With IRAs-No Keogh Both Only one Combination Regular Rollover Roth Any IRA IRA/ IRA- Keogh- IRA and type of of Regular Rollover Roth IRA IRA IRA except Keogh no Keogh no IRA Keogh IRA IRAs IRA IRA IRA only only only rollover Total 28.0% 98.1% 0.5% 1.4% 76.3% 23.7% 56.1% 36.1% 34.3% 37.8% 20.4% 18.2% 63.9% Family Income Less than $10, $10,000 $24, a $25,000 $49, $50,000 $99, $100,000 or more Age of Head < Education of Head Below HS diploma a a HS diploma Some College College Degree Race White non-hispanic Nonwhite Net Worth Percentile Bottom 25% a Top 10% Source: Employee Benefit Research Institute estimates of the 2010 Survey of Consumer Finances. a No families in this group were found to have these types of plans. ebri.org Issue Brief September 2012 No

19 Figure 12b Average Balance of Families' IRA/Keoghs and Certain Types of IRAs and Ratio of Rollover IRA Balances to Nonrollover Balances, by Various Demographic Categories, 2010 Of Those With IRA/Keogh Of Those With IRAs Only-No Keogh Both Only One Combination Regular Rollover Roth Any IRA Ratio of rollover IRA/ IRA- Keogh- IRA and Type of Of Regular Rollover Roth IRA IRA IRA except rollover to only to Keogh no Keogh no IRA Keogh IRA IRAs IRA IRA IRA only only only rollover no rollover no rollover Total $148,759 $143,817 $217,627 $478,230 $107,076 $267,429 $169,053 $225,935 $134,226 $106,975 $157,777 $50,353 $99, Family Income Less than $10,000 a a a a a a a a a a a a a a a $10,000 $24,999 55,469 55,980 a 27,543 36, ,915 66,869 82,273 63,511 39,806 42,620 14,142 42, $25,000 $49,999 99,851 98, , ,784 74, , , ,834 98,444 82, ,259 16,416 69, $50,000 $99, , ,917 88, ,418 68, , , , ,318 70,474 95,305 34,615 68, $100,000 or more 230, , , , , , , , , , ,253 68, , Age of Head <35 19,628 19,045 16,000 98,755 14,829 37,693 28,886 26,014 19,905 20,067 12,021 13,008 17, ,116 64, , ,000 55,637 90,201 73,262 91,712 63,630 40,653 89,628 41,144 45, , ,269 89, ,220 84, , , , ,193 70, ,275 48,010 74, , ,613 37, , , , , , , , ,713 86, , , , , , , , , , , , , , , , , , , , , , , , , ,301 92, , Education of Head Below HS diploma 43,360 43,360 a a 32,327 99,147 44,304 58,997 77,041 29,149 39,140 4,500 33, HS diploma 75,502 73,843 a a 54, ,138 94, ,902 87,288 59,561 65,915 29,857 55, Some college 84,563 85,564 14,793 a 69, ,747 89, ,669 82,032 70,902 95,795 36,898 68, College degree 192, ,715 a 502, , , , , , , ,121 60, , Race White non-hispanic 158, , , , , , , , , , ,076 48, , Nonwhite 95,610 94, , ,838 73, , , , ,714 88,241 63,950 60,334 75, Net Worth Percentile Bottom 25% 8,334 8,173 17,784 a 6,747 15,248 8,184 10,122 9,513 6,254 5,874 8,259 7, ,595 14,332 60,000 24,965 12,900 22,529 17,720 15,382 13,086 16,156 11,103 10,259 13, ,730 35,865 12,740 28,091 33,430 47,804 38,981 44,836 29,230 34,917 42,518 21,700 30, , , , ,638 86, , , , ,373 82, ,190 55,717 81, Top 10% 387, , , , , , , , , , , , , Source: Employee Benefit Research Institute estimates of the 2010 Survey of Consumer Finances. a Sample size not sufficient for a reliable estimate. Ratio of ebri.org Issue Brief September 2012 No

20 To demonstrate the influence of the rollover on the average IRA balance of a family, the last two columns of Figure 12b show the ratio of the any-rollover average balance to the no-rollover average balance, and ratio of the rollover-only average balance to the no-rollover average balance. The impact is substantial, as the rollover average balance is 2.27 times as large as the no-rollover average balance, while if just the rollover-only accounts are included eliminating the combination of a rollover with another IRA type the ratio for the rollover-only average balance is 1.59 times as large as the no-rollover average balance. However, this impact varies significantly over demographic groups. For example, the ratio of the rollover-only to no-rollover balance for those families in the percentile of net worth is 1.53, while the ratio is 0.83 for those in the lowest quartile of net worth. Furthermore, families with the least-educated family heads had a ratio of 1.16, while families with heads having at least a college degree had a ratio of The previous sections broke down IRA ownership and assets based upon family demographics. However, this does not provide the breakdown of the total assets in each of the IRA types, due to the grouping of IRA types within the families. Figure 13a shows the share of total IRA and Keogh assets by type, and Figure 13b shows the share of total IRA assets by type. Rollover IRAs account for 43.2 percent of all IRA and Keogh assets, with regular IRA assets following very closely behind at 42.7 percent, while Roth IRAs account for 11.1 percent and Keoghs 3.0 percent (Figure 13a). Among total IRA assets, rollover IRAs account for 44.5 percent of assets, regular IRAs 44.1 percent, and Roth IRAs 11.4 percent (Figure 13b). Therefore, rollover IRAs account for a larger share of assets than regular IRAs, while the two together account for just under 90 percent of the IRA assets. 28 Conclusion This analysis of the 1992, 2007, and 2010 versions of the SCF found that the percentage of all families with a participant in an employment-based retirement plan from a current employer increased from 38.8 percent in 1992 to 40.6 percent in 2007, before declining in 2010 to 37.9 percent. Furthermore, the ownership of 401(k)-type plans among families participating in a retirement plan was shown to have increased again in 2010 to 82.1 percent, even after more than doubling, from 31.6 percent in 1992 to 79.5 percent in While overall retirement plan participation by families declined from , the percentage of family heads who were eligible for defined contribution plans and chose to participate held essentially stable at 78.8 percent in 2007 to 78.2 percent in The percentage of families owning an IRA or Keogh plan declined from 30.6 percent in 2007 to 28.0 percent in In addition, the percentage of families with a retirement plan from a current employer, a previous employer s defined contribution plan, or an IRA/Keogh declined from 66.2 percent in 2007 to 63.8 percent in While overall participation in employment-based plans and IRAs decreased in 2010, the median account balance of those families owning an IRA increased in However, the median defined contribution plan balance decreased in 2010 after increases from 1992 through The value was $12,964 in 1992, reached $31,431 in 2007, and fell to $29,000 in The median IRA/Keogh balance increased to $40,000 in 2010 from $34,574 in The median percentage of families total financial assets comprised by defined contribution plan assets and/or IRA/Keogh assets (assuming the family had any) increased from 2007 to 2010, and accounted for a clear majority of these assets: Defined contribution plan balances accounted for 58.1 percent of families total financial assets in 2007, and that share grew to 61.4 percent in Defined contribution and/or IRA/Keogh balances increased their share as well, from 64.1 percent of total family financial assets in 2007 to 65.7 percent in Across all demographic groups, these assets account for a very large share of total financial assets for those who own these accounts. Lastly, a breakdown of IRA ownership by families was examined to determine the relative importance of rollover IRAs. While regular IRAs account for the largest percentage of IRA ownership, rollover IRAs had a larger share of assets than regular IRAs in Rollover IRAs had the largest average balance, and, when compared with the IRAs that did not include a rollover, the ratio of rollover assets to nonrollover assets ranged from 1.59 to 2.27, depending on whether the rollover combination owners are included. ebri.org Issue Brief September 2012 No