.

Ownership of Individual Retirement
Accounts (IRAs) and Policy Options for
Congress
John J. Topoleski
Analyst in Income Security
November 5, 2010
Congressional Research Service
7-5700
www.crs.gov
R41476
CRS Report for Congress
P
repared for Members and Committees of Congress

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Summary
Preparing for financial security in retirement continues to be a concern of working Americans and
policymakers. Although most Americans participate in the Social Security system, Social Security
is likely to be only a part of income in retirement. A recent Gallup poll indicated that while 34%
of working Americans expect Social Security to be a major source of retirement income, 45% of
those polled expect private retirement savings accounts to be a major source of retirement
income.
Since the 1920s, Congress has provided tax incentives to employers to sponsor pension plans for
their workers. Recognizing that workers for companies that did not offer pension plans did not
benefit from the incentives, in 1974, Congress permitted workers without pension plans to
establish Individual Retirement Accounts (IRAs). IRAs are tax-advantaged savings accounts to
encourage workers to save for retirement. Since 1974, eligibility for IRAs and the tax treatment
of contributions to IRAs have changed. Currently, workers without pensions and workers with
pensions whose income is under certain limits may make tax-deductible contributions to IRAs.
Other workers may make non-deductible contributions. Since 1998, some workers have been able
to make non-deductible contributions to Roth IRAs. Lump-sums from employer-sponsored
pensions may be rolled-over into either traditional or Roth IRAs. Most of the funds in IRAs
consist of rollover balances rather than the accumulation of contributions and investment
earnings.
IRAs may be falling short of their goal of encouraging workers to save for retirement. While
about one-half of working Americans participate in employer-sponsored pensions, only about
one-third of working households in the United States owned an IRA in 2007. Although IRAs were
originally intended for workers in employment without pension plans, IRA ownership rates are
higher among households in which the head or spouse participates in a pension plan at work.
Among households with pension coverage, 38.7% had an IRA in 2007. Among households
without pension coverage, 25.5% had an IRA in 2007.
Analysis of the 2007 Survey of Consumer Finances from the Federal Reserve indicates that
households that own IRAs tend to be older, wealthier, more educated, and have higher
propensities to save than households that do not own IRAs. Although one-third of working
households owned an IRA in 2007, relatively few households made a tax-deductible IRA
contribution in 2007. Statistics of Income data from the Internal Revenue Service indicate that
2.3% of tax returns reported a tax-deductible IRA contribution in 2007. The percentage of
taxpayers reporting tax-deductible IRA contributions was lower in 2007 than in any year since
1978. Both the percentage of taxpayers reporting tax-deductible IRA contributions in 2007 and
the average amount of that contribution increased as taxpayers’ reported income increased.
Although IRA ownership rates are lower among households without pension coverage, IRA
ownership patterns are similar among households with and without pension coverage. For
example, IRA ownership increases as households’ income increases and IRA ownership rates are
higher among households that own their homes compared with households than do not own their
homes. Some policy proposals that supporters argue will increase IRA ownership rates and
account balances include making the Retirement Savings Contributions Credit a refundable tax
credit, adopting Automatic IRA proposals, increasing household financial literacy, and making the
fees that financial institutions charge IRA owners more transparent.
Congressional Research Service

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Contents
Introduction ................................................................................................................................ 1
Traditional IRAs, Roth IRAs, and Rollover IRAs ........................................................................ 2
Traditional IRAs ................................................................................................................... 2
Roth IRAs............................................................................................................................. 3
Rollover IRAs....................................................................................................................... 3
Legislative History of IRAs......................................................................................................... 4
Annual Contribution Limits .................................................................................................. 4
Tax-Deduction for Traditional IRA Contributions .................................................................. 5
Tax-deductible IRA Contributions from 1974 to 2007.................................................................. 6
Contributions to Traditional IRAs in 2007 ................................................................................. 10
IRA Ownership and Account Balances in 2007.......................................................................... 11
IRAs and Pension Coverage ...................................................................................................... 13
IRA Ownership and Firm Size............................................................................................. 13
IRA Ownership and Industry of Workers ............................................................................. 14
IRA Ownership, Pension Coverage, and Household Financial Characteristics...................... 15
Demographic Characteristics and IRA Ownership ............................................................... 15
Income.......................................................................................................................... 16
Age............................................................................................................................... 16
Education...................................................................................................................... 16
Race and Ethnicity ........................................................................................................ 17
Homeownership ............................................................................................................ 17
Marital Status................................................................................................................ 17
Saver Characteristics and IRA Ownership ........................................................................... 19
Policy Options .......................................................................................................................... 20
Changing the Retirement Savings Contributions Credit ....................................................... 20
Adopting Automatic IRAs................................................................................................... 21
Improving Financial Literacy .............................................................................................. 22
Increasing Fee Transparency ............................................................................................... 22

Figures
Figure 1. Total Value of Tax-Deductible Contributions to Traditional IRAs.................................. 8
Figure 2. Average Value of Tax-Deductible Contributions to Traditional IRAs ............................. 9
Figure 3. Number of Tax Returns with Tax-Deductible Contributions to Traditional IRAs .......... 9
Figure 4. Percentage of Tax Returns with Tax-Deductible Contributions to Traditional
IRAs ...................................................................................................................................... 10

Tables
Table 1. Individual Retirement Accounts: Major Laws Affecting Eligibility and
Contributions ........................................................................................................................... 5
Congressional Research Service

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Table 2. Tax-Deductible IRA Contributions in 2007 .................................................................. 11
Table 3. IRA Ownership Among Working Households in 2007 .................................................. 12
Table 4. Pension Plan Coverage and IRA Ownership Among Working Households in
2007 ...................................................................................................................................... 13
Table 5. Pension Plan Coverage and IRA Ownership by Size of Employer, 2007 ....................... 14
Table 6. Pension Plan Coverage and IRA Ownership by Industry, 2007 ..................................... 14
Table 7. Financial Characteristics of Households by Pension Coverage and IRA
Ownership in 2007................................................................................................................. 15
Table 8. Demographic Characteristics of IRA Owners ............................................................... 18
Table 9. Saver Type Characteristics of IRA Owners, 2007 ......................................................... 19
Table 10. Retirement Savings Contributions Credit Use In 2007................................................ 21

Contacts
Author Contact Information ...................................................................................................... 23

Congressional Research Service

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Introduction
Income in retirement comes from several sources. Most workers participate in the Social Security
system, which pays retirement income as early as 62 years old. Retirement income security can
also be provided by non-government sources, such as employer-sponsored pensions and
participant-initiated retirement accounts. About half of the U.S. workforce participates in one or
both of two kinds of employer-sponsored pensions.
The two kinds of employer-sponsored pensions are defined benefit (DB) and defined contribution
(DC) pensions. DB pensions provide retirees with monthly benefits in which the amount is
calculated using a formula based on the number of years of employment and final salary. DC
pensions, of which the 401(k) plan is the most common, provide participants individual accounts
in which contributions from individual employees—and often their employers—are placed in tax-
advantaged accounts that accrue investment returns.1 The accumulated contributions and
investment earnings are then used for participants’ retirement income. Although employers
choose whether to offer DC plans, individual employees decide the contribution amounts (subject
to an annual limit) and investment allocations of the contributions.
Over the past 30 years, fewer employers have been providing DB plans and instead have offered
DC plans.2 One consequence of this shift is that workers have had increasing responsibility for
their retirement income security. Individual Retirement Accounts (IRAs) can play an important
role by providing workers with tax-advantaged savings accounts in which to accumulate
retirement savings.
Although enrollment in a DC or DB plan is contingent on employer sponsorship of such a plan,
most individuals (or married couples) who have taxable compensation can establish and
contribute to an IRA. Similar to 401(k) plans, IRAs are tax-advantaged accounts in which
contributions and investment earnings accumulate, which are then used for retirement income.
Depending on the type of IRA, contributions may be made on a pre- or post-tax basis and
investment earnings are either tax-deferred or tax-free. 3
This report provides an overview and a brief legislative history of IRAs, examines the tax-
deductible contributions to traditional IRAs in 2007, analyzes the characteristics of households
that have IRAs, and compares IRA owners who have pension coverage at work with those who
do not have pension coverage. The report concludes with a discussion of some of the policy
proposals that some suggest might increase IRA ownership rates among American households.

1 401(k) plans are offered by private sector employers. Other types of defined contribution retirement accounts include
403(b) plans offered by charitable and educational employers, 457 plans offered by state and local governments, and
the Thrift Savings Plan (TSP) offered by the Federal government.
2 For more information on this shift, see James Poterba, Steven Venti, and David A. Wise, “The Shift From Defined
Benefit Pensions to 401(k) Plans and the Pension Assets of the Baby Boom Cohort,” Economics of Health and
Mortality Special Feature: The Economics and Demography of Aging, Proc. Natl. Acad. Sci., vol. 104 no. 33, (August
14, 2007), pp. 13238-13243, available at http://www.pnas.org/content/104/33/13238.full.
3 More information on eligibility requirements for IRAs can be found in CRS Report RL34397, Traditional and Roth
Individual Retirement Accounts (IRAs): A Primer, by John J. Topoleski. For more information on current laws on
IRAs, see U.S. Congress, Joint Committee on Taxation, Present Law And Analysis Relating To Individual Retirement
Arrangements, 110th Cong., 2nd sess., June 24, 2008, JCX-53-08.
Congressional Research Service
1

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Traditional IRAs, Roth IRAs, and Rollover IRAs
The two types of IRAs available to households in which either the head of the household or the
spouse receives income from wages are traditional IRAs and Roth IRAs.4 Traditional and Roth
IRAs differ in their eligibility requirements and the tax treatment of contributions and
withdrawals. Individuals can also rollover lump-sum retirement assets from an employer-
sponsored pension or an IRA into a rollover IRA. Rollover IRAs may be either traditional IRAs,
in which investment earnings continue to accumulate tax-deferred, or Roth IRAs, in which
investment earnings continue to accumulate tax-free.
Traditional IRAs
Individuals with taxable compensation who are under the age of 70½ can contribute to traditional
IRAs. Traditional IRAs have at least two tax advantages over accounts that do not receive tax
advantages: (1) the tax-deductibility of contributions allows individuals to have larger yearly
contributions than otherwise would be the case; and (2) the earnings grow tax-deferred.
Individuals may annually contribute the lesser of $5,000 or their taxable compensation to an IRA.
Individuals aged 50 or older may make an additional $1,000 annual contribution.5 Contributions
to traditional IRAs are tax-deductible (up to the maximum contribution) for individuals who (1)
are not covered by an employer-sponsored pension plan or (2) are covered by an employer-
sponsored pension plan but have income under statutory limits.6 In 2010, the phase out of the
deduction begins at an Adjusted Gross Income (AGI) of $55,000 for single filers with pension
coverage; at an AGI of $89,000 for married filers in which the head of the household has pension
coverage; and at an AGI of $166,000 for married filers in which the head of the household does
not have pension coverage but in which the spouse is covered by a pension plan. No deduction is
allowed for AGI of $65,000 or more for single filers with pension coverage; $109,000 or more for
married filers in which the head of the household has pension coverage; or $176,000 or more for
married filers in which the head of the household does not have pension coverage but in which
the spouse is covered by a pension plan.7

4 Although individuals open traditional and Roth IRAs, employers may sponsor one of two kinds of IRA-based
retirement plans: Simplified Employee Pension (SEP-IRAs) or Savings Incentive Match Plan for Employees (SIMPLE-
IRA). Both kinds of employer-sponsored IRAs have higher contribution limits than traditional or Roth IRAs but are
otherwise subject to many of regulations that govern IRAs. SEP- and SIMPLE-IRAs were created to allow employers
to provide retirement accounts for their employees but without many of the regulatory requirements of 401(k) plans.
For more information, see U.S. Government Accountability Office, Individual Retirement Accounts: Government
Could Encourage More Employers to Offer IRAs to Employees, GAO-08-590, June 4, 2008.
5 The contribution limits are adjusted annually for inflation.
6 Individuals are considered covered by a pension plan if contributions have been made to a defined contribution plan
in the tax year or if they are eligible to participate in a defined benefit. Individuals are not considered covered by a
pension plan if they are only covered under Social Security, Railroad Retirement, or if the only reason for their
participation in a pension plan is because they are reservists or volunteer firefighters. See 26 U.S.C. 219(g) and IRS
Publication 590, Individual Retirement Arrangements (IRAs), available at http://www.irs.gov/publications/p590.
7 IRS Publication 590, Individual Retirement Arrangements (IRAs), has details on calculating the deductible amount of
IRA contributions. It is available at http://www.irs.gov/publications/p590/ch01.html#en_US_publink1000230433.
Congressional Research Service
2

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Withdrawals from traditional IRAs, except for that part of a withdrawal that represents after-tax
contributions, must be included in account holders’ taxable income.8 Congressional intent was
that IRAs were to provide financial security in retirement and not be used for either pre-
retirement expenses or as tax-free asset transfers to heirs. Most withdrawals from traditional IRAs
before the account holder reaches the age of 59½ are subject to an additional 10% penalty tax,
unless the reason for the withdrawal meets one of the exceptions found in 26 U.S.C. § 72(t).9 In
addition, account holders older than the age of 70½ must receive Required Minimum
Distributions (RMDs). RMDs are withdrawals of a specified percentage of the IRA balances each
year.10
Roth IRAs
Roth IRAs were authorized by the Taxpayer Relief Act of 1997 (P.L. 105-34). The key differences
between traditional and Roth IRAs are that (1) contributions to Roth IRAs are not tax-deductible
(thus the contributions are after-tax dollars) and (2) qualified distributions are not included in
taxable income. The annual contribution limit to Roth IRAs is the lesser of $5,000 or the
individual’s taxable compensation.11 Individuals aged 50 or older may make an additional $1,000
annual contribution. Contributions to Roth IRAs are permitted only by individuals with incomes
under prescribed statutory limits, currently at $120,000 for single filers and $176,000 for joint
filers. Because qualified distributions are tax-free, Roth IRAs have no RMD requirement.
Rollover IRAs
Rollovers are transfers of assets from one retirement plan to another retirement plan, often upon
separation from an employer. Generally, individuals may rollover account balances from
employer-sponsored pension plans into traditional or Roth IRAs upon separation from
employment. Individuals may also rollover traditional IRA account balances to Roth IRAs.
Rollovers not completed within 60 days are considered taxable distributions. Rollovers are
considered neither distributions nor contributions. The amount of the rollover is not included in
taxable income, is not subject to the contribution limits, nor subject to rules that govern early
distributions. Rollovers can come from traditional IRAs or from employers’ pension plans. Prior
to 2010, rollovers into Roth IRAs were limited to individuals with incomes under specified
limits12; these income limits were eliminated beginning in 2010.13

8 Withdrawals that represent non-deductible contributions do not have to be included in income because doing so
would subject the contributions to double taxation since income used to make non-deductable contributions is taxed
when the income is earned.
9 Some of the exceptions to the early withdrawal penalty include distributions for higher-education expenses, purchase
of a first home, or if the individual becomes disabled. For more information, see CRS Report R40192, Early
Withdrawals and Required Minimum Distributions in Retirement Accounts: Issues for Congress, by John J. Topoleski.
10 The percentage is based on the individual’s life expectancy according to IRS-approved mortality tables.
11 While individuals may contribute to traditional IRAs, Roth IRAs, or both, an individual’s total annual IRA
contribution to all IRAs cannot exceed the lesser of $5,000 ($6,000 if aged 50 or older) or total taxable compensation.
12 In 2009, individuals with Modified Adjusted Gross Income (for Roth IRA purposes) over $100,000 could not
rollover amounts to Roth IRAs.
13 These were eliminated by the Tax Increase Prevention and Reconciliation Act of 2005 (TIPRA, P.L. 109-222).
Congressional Research Service
3

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Legislative History of IRAs
IRAs were first authorized in 1974 by the Employee Retirement Income Security Act of 1974
(ERISA; P.L. 93-406) to encourage workers who did not have a pension plan at work to save for
retirement. Prior to ERISA, Congress allowed self-employed individuals and their employees to
make tax-deductible contributions to Keogh accounts for the purpose of providing for income in
retirement.14 Some Keogh plan features, such as imposing penalties for withdrawing funds before
the account owner reached the age of 59½ or for failing to take required withdrawals after the age
of 70½, were incorporated into the design of IRAs.
Table 1 lists the major IRA legislation since 1974 that has affected eligibility for and
contributions to IRAs.
From 1974 to 1981, only individuals who were not covered by an employer-sponsored pension
plan could establish and contribute to IRAs. From 1982 to 1986, all workers regardless of pension
coverage could make tax-deductible IRA contributions (up to the annual limit). Because of the
revenue loss to the U.S. Treasury, the Tax Reform Act of 1986 (P.L. 99-514) limited the tax
deduction on IRA contributions to non-covered workers and to covered workers with incomes
under $25,000.15 Since 1986, workers without pension coverage and workers with pension
coverage and who have incomes under prescribed limits can make tax-deductible IRA
contributions (up to the annual limit) while covered workers who have incomes greater than the
prescribed limits may make non-deductible contributions.
Annual Contribution Limits
The annual contribution limit was set at $1,500 in 1974 and was increased to $2,000 in 1982. The
limit was increased to $3,000 in 2002, $4,000 in 2005, and has been annually inflation-adjusted
for increases in the Consumer Price Index (CPI) since 2008. The contribution limit for 2010 is
$5,000. Beginning in 2002, individuals aged 50 and older could make an additional annual $500
“catch up” contribution to their IRAs. This was increased to $1,000 beginning in 2006. From
1982 to 1996, a non-working spouse could annually contribute $250 to an IRA. Since 1997, the
contribution limit for non-working spouses has been equal to the contribution limit for working
spouses.16

14 Keogh plans, also known as H.R. 10 plans, were established in 1962 by the Self-Employed Individuals Tax
Retirement Act (P.L. 87-792) to encourage the owners of small businesses to establish retirement plans through the use
of tax-deferred savings.
15 The Office of Management and Budget (OMB) estimated the tax expenditure for IRA contributions in 1986 to be
$19.6 billion ($39.1 billion in 2009 dollars). See Budget of the United States Government, Fiscal Year 1986, Office of
Management and Budget. OMB estimated the tax expenditure for IRA contributions in 2009 to be $12.1 billion. See
Budget of the United States, Table 16–1. Estimates of Total Income Tax Expenditures for Fiscal Years 2009 - 2015
available at http://www.whitehouse.gov/sites/default/files/omb/budget/fy2011/assets/receipts.pdf.
16 For example, a married household in which neither the head nor spouse has a pension plan at work could make a
total tax-deductible contribution of $10,000 to their IRAs in 2010.
Congressional Research Service
4

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Tax-Deduction for Traditional IRA Contributions
From 1974 to 1981, workers covered by a pension plan at work were prohibited from making
IRA contributions. From 1982 to 1986, workers covered by an employer-sponsored pension plan
could make tax-deductible contributions up to the annual limit of $2,000 to an IRA. Since 1987,
workers covered by a pension plan may deduct 100% of their IRA contribution only if their
incomes are under statutory limits. The percentage of the contribution that may be deducted is
gradually reduced from 100% to 0% for incomes that are $10,000 or $20,000 (depending on
filing status) greater than the statutory limits. No deduction is available for taxpayers with income
greater than the upper limit of the phase-out range.17
Since 1998, workers with taxable income under statutory limits could make contributions to Roth
IRAs. The maximum contribution to a Roth IRA is phased out for persons with income greater
than limits specified in statute.18
Table 1. Individual Retirement Accounts: Major Laws Affecting Eligibility
and Contributions
Law
Major Provisions Affecting IRAs
Employee Retiree Income Security Al owed workers without pensions to make up to $1,500 in annual deductible
Act of 1974 (ERISA, P.L. 93-406)
contributions. Workers with pension coverage were ineligible to open IRAs.
Economic Recovery Tax Act of
Extended IRA eligibility to individuals in covered employment.
1981 (P.L. 97-34)
Raised the annual contribution limit to $2,000.
Al owed non-working spouses to make annual IRA contributions up to $250.
Tax Reform Act of 1986
Began to phase out the tax-deduction for contributions for single filers who have
(P.L. 99-514)
pension coverage at work and who made more than $25,000 per year or married
filers who made more than $40,000.
Smal Business Job Protection
Raised the contribution limit for non-working spouses from $250 to $2,000.
Act of 1996 (P.L. 104-188)

17 See 26 U.S.C. § 219(g). The phase-out range is $10,000 for single filers with pension coverage and married filers in
which the head of the household does not have pension coverage but in which the spouse is covered by a pension. For
married filers in which the head has pension coverage and in which the spouse also has pension coverage, the phase-out
range was increased from $10,000 to $20,000 beginning in 2007. The threshold at which phase-out for the deduction
begins depends on the filing status and pension coverage of the taxpayer and the spouse’s pension coverage. For single
filers with pension coverage, the deduction phase-out began at $25,000 from 1987 to 1997; $30,000 in 1998; $31,000
in 1999; $32,000 in 2000; $33,000 in 2001; $34,000 in 2002; $40,000 in 2003; $45,000 in 2004; $50,000 in 2005 and
2006; $52,000 in 2007; $53,000 in 2008; $55,000 in 2009; and $56,000 in 2010. For married filers in which the head of
the household has pension coverage, the deduction phase-out began at $40,000 from 1987 to 1997; $50,000 in 1998;
$51,000 in 1999; $52,000 in 2000; $53,000 in 2001; $54,000 in 2002; $60,000 in 2003; $65,000 in 2004; $70,000 in
2005; $75,000 in 2006; $83,000 in 2007; $85,000 in 2008; and $89,000 in 2009 and 2010. For married filers in which
the head does not have pension coverage but in which the spouse has pension coverage, the deduction phase-out began
at $40,000 from 1987 to 1997; $150,000 from 1998 to 2006; $156,000 in 2007; $159,000 in 2008; $166,000 in 2009;
and $167,000 in 2010.
18 The phase-out range is $15,000 for single filers and $10,000 for taxpayers filing joint returns. For single files, the
phase began at $95,000 from 1998 to 2006; $99,000 in 2007; $101,000 in 2008; $105,000 in 2009; and is $105,000 in
2010. For taxpayers filing joint returns, the phase out began at $150,000 from 1998 to 2006; $156,000 in 2007;
$159,000 in 2008; $160,000 in 2009; and is $167,000 in 2010.
Congressional Research Service
5

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Law
Major Provisions Affecting IRAs
Taxpayer Relief Act of 1997
Raised the threshold at which phase-out of the tax deduction for contributions
(TRA,P.L. 105-34)
begins:
• Single filers who have pension coverage at work: $30,000 in 1998; $31,000 in
1999; $32,000 in 2000; $33,000 in 2001; $34,000 in 2002; $40,000 in 2003;
$45,000 in 2004; and $50,000 after 2004.
• Married filers in which the head of the household has pension coverage and in
which the spouse also has pension coverage: $50,000 in 1998; $51,000 in 1999;
$52,000 in 2000; $53,000 in 2001; $54,000 in 2002; $60,000 in 2003;
$65,000 in 2004; $70,000 in 2005; $75,000 in 2006; $80,000 after 2006.
• Married filers in which the head of the household does not have pension coverage
but in which the spouse has pension coverage: $150,000.
Increased the phase-out range for married filers who file a joint return from
$10,000 to $20,000.
Created the Roth IRA.
Economic Growth and Tax
Raised the contribution limit to $3,000 in 2002; $4,000 in 2005; and $5,000 in
Relief Reconciliation Act of 2001
2008.
(EGTRRA, P.L. 107-16)
Allowed “catch-up” contributions: individuals aged 50 and older could make an
additional yearly contribution of $500 from 2002 to 2005 and $1,000 in each year
thereafter.
Adjusted, based on increases in the Consumer Price Index (CPI) beginning in
2008, the fol owing: the IRA contribution limits; the income threshold above
which taxpayers may not claim a tax-deduction for traditional IRA contributions;
and the income threshold above which taxpayers may not make Roth IRA
contributions.
Established the Retirement Savings Contributions Credit to encourage lower
income workers to save for retirement.
Pension Protection Act of 2006
Made permanent EGTRRA provisions establishing the Retirement Savings
(P.L. 109-280)
Contributions Credit and the adjustments in contribution limits and income
thresholds for the Consumer Price Index.
Tax Increase Prevention and
Removed income restriction on converting traditional IRAs to Roth IRAs
Reconciliation Act of 2005
beginning in 2010.
(P.L. 109-222)
Source: Congressional Research Service (CRS).
Notes: Covered workers have either defined benefit or defined contribution pension plans provided through
their workplace.
Tax-deductible IRA Contributions from 1974 to 2007
Less than 3% of taxpayers made tax-deductible IRA contributions in 2007. As a percentage of all
tax returns filed, fewer tax returns had IRA deductions in 2006 and 2007 than in any year since
1977. Figure 1 through Figure 4 show historical tax-deductible IRA contributions.19 Figure 1

19 The data used in this report are from two sources: (1) Internal Revenue Service (IRS), Statistics of Income Division,
Individual Income Tax Returns and (2) The Federal Reserve Board, Survey of Consumer Finances. The IRS publishes
detailed tax data from Form 1040. IRS data for 1974 to 1983 is from IRS Publication 75, Individual Income Tax
Returns; data for 1985 to 2007 is from IRS publication 1304, Individual Income Tax Returns. Data for 1990 to 2007 is
(continued...)
Congressional Research Service
6

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

shows the inflation-adjusted yearly amount of tax-deductible contributions to traditional IRAs
from 1974 to 2007. Figure 2 shows the inflation-adjusted average tax-deductible contribution.
Figure 3 and Figure 4 show the number and percentage of taxpayers who made tax-deductible
IRA contributions from 1974 to 2007. The total amount of IRA contributions was nearly twice as
much in 1981 compared with 1975, although the average amount was more than 30% lower. The
large increase in 1982 both in the number and percentage of returns reporting IRA contributions
and the average amount of IRA contributions is a result of the Economic Recovery Tax Act of
1981 (P.L. 97-34), which extended eligibility for IRAs to workers with employer-sponsored
pensions. The Tax Reform Act of 1986 (P.L. 99-514) that limited the eligibility for tax-deductible
contributions beginning in 1987 sharply curtailed both the number of returns reporting IRA
contributions and the average IRA contribution. This suggests that tax-deductible IRA
contributions were largely made by workers with higher incomes and pension coverage.
The data suggest that since the mid-1990s some taxpayers have responded to changes in the
contribution limit. The total and average value of IRA contributions generally declined from 1990
to 2001. Total tax-deductible IRA contributions were 46% lower in 2001 compared with 1990 and
the average contribution was 16% lower. From 1987 to 2001, neither the income limit for
eligibility for tax-deductible contributions nor the contribution limit was increased, although in
real (inflation-adjusted) terms these limits were more than 25% lower in 2001 than in 1987. The
5.5% increase in average IRA contributions from 1996 to 1997 may be due to an increase in the
contribution limit for non-working spouses from $250 to $2,000.
The increases in the contribution limit from $2,000 to $3,000 in 2002 and from $3,000 to $4,000
in 2005 were accompanied by increases in the average contribution. Interestingly, there was not a
corresponding rise in the number of returns reporting an IRA contribution nor a rise in the
percentage of returns reporting a contribution. This suggests that the response to these changes
was among taxpayers who were already contributing near the contribution limit and increased the
amount of their contributions. This is consistent with research that has shown that many
individuals who make IRA contributions contribute close to the statutory limit.20
The introduction of Roth IRAs beginning in 1998 appears to have had no effect on the total
amount of the contributions to traditional IRAs nor on the average traditional IRA tax-deductible
contribution. Continuing the trend that began in 1990, both total and average IRA contributions
were lower from 1998 to 2001, the first four years of the existence of Roth IRAs.21 While it is
possible that the decline in traditional IRA contributions over this period is as a result of
taxpayers making Roth IRA contributions instead, a more likely explanation is that decline in
traditional IRA contributions is a result of the decreased real value of the IRA contribution limit.

(...continued)
available online from Table A: Selected Income and Tax Items for Selected Years (in Current and Constant Dollars)
available at[http://www.irs.gov/taxstats/indtaxstats/article/0,,id=134951,00.html#_tbla. The Survey of Consumer
Finances (SCF) is a triennial survey conducted on behalf of the Board of Governors of the Federal Reserve and
contains detailed information on U.S. household finances, such as the amount and types of assets owned, the amount
and types of debt owed, and detailed demographic information on the head of the household and spouse. Each survey
interviews different households to provide separate cross-sections of data. The SCF uses the household as the unit of
analysis as assets such as homes and bank accounts are often jointly owned by married couples. Further information on
the SCF is available in the codebook at http://www.federalreserve.gov/PUBS/oss/oss2/2007/codebk2007.txt.
20 See, for example, The American Economic Review, Vol. 84, No. 5 (Dec., 1994), “IRAs and Household Saving,” by
William G. Gale and John Karl Scholz.
21 The contribution limit for traditional IRAs remained unchanged over this period.
Congressional Research Service
7

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Roth IRA contributions were likely not a result of individuals shifting their contributions from
traditional IRAs to Roth IRAs.
Figure 1. Total Value of Tax-Deductible Contributions to Traditional IRAs
Calendar Years 1975 – 2007, Billions of 2007 Dollars
$80
$70
$60
$50
$40
$30
$20
$10
$0
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007

Source: CRS analysis of various years of Statistics of Income data from the Internal Revenue Service.
Congressional Research Service
8

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Figure 2. Average Value of Tax-Deductible Contributions to Traditional IRAs
Calendar Years 1975 – 2007, Constant 2007 Dollars
$5,500
$5,000
$4,500
$4,000
$3,500
$3,000
$2,500
$2,000
$1,500
$1,000
$500
$0
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007

Source: CRS analysis of various years of Statistics of Income data from Internal Revenue Service.
Figure 3. Number of Tax Returns with Tax-Deductible Contributions to
Traditional IRAs
Calendar Years 1975 – 2007, Millions of Returns
18
16
14
12
10
8
6
4
2
0
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007

Source: CRS analysis of various years of Statistics of Income data from Internal Revenue Service.
Congressional Research Service
9

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Figure 4 shows that the percentage of taxpayers who make IRA contributions was highest in
1985 at 15.9%; declined sharply after 1986 as a result of the limitations on tax-deductible
contributions by workers covered by a pension plan; and gradually declined from 1989 (4.7%) to
2007 (2.3%). Since 1999, less than 3% of tax returns each year reported tax-deductible
contributions to traditional IRAs.
Figure 4. Percentage of Tax Returns with Tax-Deductible Contributions to
Traditional IRAs
Calendar Years 1975 – 2007
16%
14%
12%
10%
8%
6%
4%
2%
0%
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007

Source: CRS analysis of various years of Statistics of Income data from Internal Revenue Service.
Contributions to Traditional IRAs in 2007
Although less than 3% of taxpayers made a tax-deductible contribution in 2007, the average
contribution was $3,902, which suggests that many households made close to the maximum
allowable contribution.22 Contributions to traditional IRAs were disproportionately made by
households with higher incomes. Table 2 indicates that although households with Adjusted Gross
Incomes of less than $50,000 in 2007 filed 65.6% of all federal income tax returns, these
households accounted for 41.1% of households claiming deductions for IRA contributions and
32.0% of the total dollar amount of tax-deductible IRA contributions. Higher-income households

22 Recent research indicates that the median contribution in 2008 was $3,300 to traditional IRAs and was $3,500 to
Roth IRAs. See the Investment Company Institute’s Research Fundamentals, The Role of IRAs in U.S. Households’
Savings for Retirement, 2009, by Sarah Holden, available at http://www.ici.org/pdf/fm-v19n1.pdf.
Congressional Research Service
10

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

have an incentive to make tax-deductible IRA contributions as they are in higher tax brackets and
receive a greater benefit for each tax-deductible dollar contributed to a traditional IRA.23 Both the
percentage of households that made tax-deductible IRA contributions and the average IRA
contribution amount increased with household income. Of the 93.8 million tax returns with
incomes below $50,000, approximately 1.4 million (1.4%) claimed the IRA deduction whereas of
the 29.7 million tax returns with incomes above $75,000, 1.2 million (4.1%) claimed the
deduction.24 The average IRA contribution among taxpayers with incomes above $100,000 was
$5,440, which was more than twice the average IRA contribution among households with
incomes below $25,000.
Table 2. Tax-Deductible IRA Contributions in 2007
Adjusted Gross Income
Less than
$25,000 to
$50,000 to
$75,000 to
$100,000

$25,000
$49,999
$74,999
$99,999
or more
Number of Returns Filed
58,893.5 34,896.9 19,450.7 11,744.1 17,993.5
(Thousands)
Number of Returns Reporting IRA
450.1 905.1 736.0 481.0 727.6
Contributions
(Thousands)
Percentage of Returns Reporting
0.8% 2.6% 3.8% 4.1% 4.0%
IRA Contributions
Total IRA Contributions
$1.2 $2.9 $2.9 $1.9 $4.0
(Billions)
Average IRA Deduction Per Return
$2,656
$3,225
$3,907
$4,010
$5,440
Source: CRS analysis of the Internal Revenue Service’s Statistics of Income data.
Notes: Table reflects tax-deductible contributions to traditional IRAs only. Contributions to Roth IRAs and
non-deductible contributions to traditional IRAs are not reported in the Statistics of Income data.
IRA Ownership and Account Balances in 2007
The Survey of Consumer Finances (SCF) indicates that 28.2 million (32.7%) working households
owned an IRA in 2007. Among working households that had IRAs, 77.0% had a single IRA and
23.0% had more than one IRA.
Table 3 indicates that more working households had traditional IRAs in 2007 (15.9 million) than
had Roth IRAs (10.2 million) or Rollover IRAs (9.4 million). 25 Roth IRAs have been popular.

23 Households under certain income limits (single filers with incomes up to $27,750 or married filers with incomes up
to $55,500) are eligible for the non-refundable Retirement Savings Contributions Credit of up to $1,000 per year.
Approximately 5.9 million households received the credit in 2007, although the IRS data does not break down the
numbers for contributions to IRAs versus contributions to DC plans. For more information, see CRS Report RS21795,
The Retirement Savings Tax Credit: A Fact Sheet, by John J. Topoleski
24 These percentages are calculated from data in Table 2. Approximately 93.8 million tax returns had incomes less than
$50,000 in 2007. Among tax returns with incomes below $50,000, 1.4 million tax returns claimed the IRA deduction.
Among tax returns with incomes of $75,000 or greater, 1.2 million tax returns claimed the IRA deduction.
25 The Investment Company Institute (ICI) estimates that 37.7 million U.S. households (32.5%) owned a traditional
(continued...)
Congressional Research Service
11

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Although Roth IRAs were authorized in 1997—23 years after traditional IRAs—Table 3 shows
that there were two-thirds as many Roth IRAs as traditional IRAs in 2007.26 Because Roth IRAs
have not had as long to accrue contributions and investment earnings, the median Roth IRA
account balance was $10,000, which was 63% lower than the median account balance of
traditional IRAs ($27,000).27
In 2007, IRA assets totaled $3.0 trillion. A plurality of IRA assets were held in Rollover IRAs.
Rollover IRAs held $1.4 trillion in assets in 2007, which was 46.3% of all IRA assets; traditional
IRAs held $1.3 trillion in assets in 2007, which was 43.7% of all IRA assets; and Roth IRAs held
$306 billion in assets in 2007, which was 10.1% of all IRA assets. The median rollover IRA
account balance was nearly twice as large as the median traditional IRA account balance and
more than five times as large as the median Roth IRA account balance. This is likely because
traditional and Roth IRAs consist mostly of yearly contributions whereas Rollover IRAs consist
of large sums transferred from other retirement accounts, like 401(k) plans.
Table 3. IRA Ownership Among Working Households in 2007

Traditional IRA
Roth IRA
Rollover IRA
Number of households (millions)
15.9
10.2
9.4
Total assets (billions) $1,327
$306
$1,406
Percentage of total IRA assets
43.7%
10.1%
46.3%
Average account balance
$83,661
$30,141
$149,490
Median account balance
$27,000
$10,000
$50,000
Source: CRS analysis of the 2007 Survey of Consumer Finances.
Note: Percentages do not add to 100% due to rounding.

(...continued)
IRA in 2007 and 17.3 million households (14.9%) owned Roth IRAs in 2007. See Appendix: Additional Data on IRA
Ownership, January 2010, available at http://www.ici.org/pdf/fm-v19n1_appendix.pdf. The ICI estimate does not
separately classify rollover IRAs, it includes rollover IRAs as either traditional or Roth IRAs. ICI includes households
in which neither the head of the household nor the spouse were working. This report focuses on IRA ownership among
households in which either the head or the spouse were working at the time of the survey. Households in which neither
the head of the household nor the spouse were working would not be able to contribute to either a traditional or Roth
IRA. The SCF indicates that in 2007, of the 116 million working and non-working households in the United States, 35
million (30.31%) had a traditional IRA, Roth IRA, or Rollover IRA. Households in which neither the head of the
households nor the spouse were working could still own IRAs as a result of contributions made in previous years.
26 The analysis of data in this report emphasizes medians rather means or simple averages. The median is the middle
value of a series of numbers, such that half the values are above the median and half are below the median. Many
financial variables are skewed; that is, some individuals or households report very large values. In such cases, median
values may be better representative of data series than average values; average values are influenced by a few very
large values.
27 Amounts in traditional and Roth accounts may not be directly comparable. Amounts withdrawn from traditional
IRAs generally must be included in taxable income, so traditional IRA account balances generally overstate the amount
available for spending after taxes are taken into account. Roth IRA withdrawals are not included in taxable income.
Most Rollover IRAs are likely in taxable accounts, although some Rollover IRAs may be after-tax.
Congressional Research Service
12

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Most households that had traditional IRAs did not make taxable contributions to their IRAs in
2007. Although 15.9 million households had a traditional IRA in 2007, only 3.3 million tax
returns reported tax-deductible contributions to traditional IRAs.28
IRAs and Pension Coverage
Although IRAs were authorized to encourage workers without pension coverage to save for
retirement, IRA ownership is greater among households with pensions than among households
without pensions. Table 4 indicates that 38.7% of households with pension coverage had IRAs in
2007, compared with 25.3% of households without pension coverage. With regard to current
policy, the data highlight that many individuals have not responded to the tax incentives designed
to encourage IRA ownership among households without pension coverage.29
Table 4. Pension Plan Coverage and IRA Ownership Among Working Households
in 2007
Household Has Pension
Household Does Not Have Pension
Number with IRA
Number with IRA
(millions of households)
Percentage
(millions of households)
Percentage
18.3 38.7% 9.9 25.3%
Source: CRS analysis the 2007 Survey of Consumer Finances.
Note: Data is for households in which either the head of the household, the spouse, or both worked.
IRA Ownership and Firm Size
As Table 5 shows, pension coverage increased with firm size in 2007. Among households with
pension coverage, IRA ownership rates were highest in workplaces that had fewer than 20
employees. Small professional offices (e.g., medical practices and law firms) are more likely to
offer pension plans than non-professional small businesses. These professionals (with high
incomes and education levels) are also more likely to have IRAs. Among households without
pension coverage (except for households that work in firms with 100—499 employees), about
one-quarter own an IRA.

28 IRS tax data are not available on the number of households that made non-deductible traditional IRA contributions,
Roth IRA contributions, or Rollover IRA contributions.
29 Individuals are considered covered by a pension plan if contributions have been made to a defined contribution plan
in the tax year or if they are eligible to participate in a defined benefit plan. Individuals are not considered covered by a
pension plan if they are only covered under Social Security, Railroad Retirement, or if the only reason for their
participation in a pension plan is because they are reservists or volunteer firefighters. See 26 U.S.C. 219(g) and IRS
Publication 590, Individual Retirement Arrangements (IRAs), available at http://www.irs.gov/publications/p590/
ch01.html#en_US_publink1000230449.
Congressional Research Service
13

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Table 5. Pension Plan Coverage and IRA Ownership by Size of Employer, 2007
Households with
Households without
Percentage of
Pension Coverage
Pension Coverage
Size of
Employer
Households
with Pension
Number with
Percentage
Number with
Percentage
Coverage
IRA (millions)
with IRA
IRA (millions)
with IRA
Less than 20
24.6% 2.8
48.1% 4.4 24.4%
employees
20 - 99
52.1% 2.3
35.0% 1.5 24.0%
employees
100 - 499
64.0% 2.6
33.5% 0.8 18.1%
employees
500 or more
72.2% 9.2
35.9% 2.3 23.5%
employees
Source: CRS analysis of the 2007 Survey of Consumer Finances.
Notes: Size of employer is for head of household or spouse, whoever is working. If both head and spouse work,
then size of employer is for person with larger IRA balance.
IRA Ownership and Industry of Workers
Table 6 shows that in 2007 pension plan coverage was highest among workers in the public
sector and manufacturing industries and was lowest among workers in agriculture, mining, or
construction industries. Among households without pension coverage, rates of IRA ownership
varied considerably by industry. Households with workers in agriculture, mining, and
construction industries had low rates of IRA ownership (11.3%) whereas about one-third (33.6%)
of households with workers who worked in financial services had IRAs.
Table 6. Pension Plan Coverage and IRA Ownership by Industry, 2007
Households with Pension
Households without Pension
Coverage
Coverage
Percentage of
Households
Number with
Number with
Industry of
with Pension
IRAs
Percentage
IRAs
Percentage
Employer
Coverage
(millions)
with IRAs
(millions)
with IRAs
Agriculture, mining,
34.2% 1.4 34.7% 0.9 11.3%
or construction
Trade
43.5% 1.9 33.2% 1.5 19.7%
Public
sector 87.5% 1.8 36.0% 0.1 20.4%
Manufacturing and
65.7% 9.6 36.5% 4.7 28.0%
transportation
Financial
services
49.3% 2.3 46.2% 1.7 33.6%
Source: CRS analysis of the 2007 Survey of Consumer Finances.
Notes: Industry of employer is for head of household or spouse, whoever is working. If both head and spouse
work, then industry is for person with larger IRA balance.
Congressional Research Service
14

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

The low IRA ownership rates among households without pension coverage and who worked in
agriculture, mining, or construction may reflect the transient nature of these occupations, high-
employee turnover, or lower education levels. The higher rates of IRA ownership for financial
services workers may be a result of greater education level or that these workers are likely to be
more knowledgeable about the benefits of IRAs than workers in other industry groups. This
suggests a potentially important role for financial literacy efforts. Increasing workers’ knowledge
of financial matters and helping workers improve financial decision-making skills may increase
IRA ownership.
IRA Ownership, Pension Coverage, and Household Financial
Characteristics
Table 7 shows financial characteristics (median dollar amounts) for households with and without
IRAs and with and without pension coverage. Households that had both pension coverage and
IRAs were the wealthiest group in the measures of median retirement assets, total assets, and net
worth. This group also had higher median debt than the other groups. Households with IRAs
(whether or not they had pensions) were wealthier than those without IRAs. For example, among
households with pension coverage, those with IRAs had median net worth that was four times
larger than households without IRAs ($411,000 versus $100,150). The gap was even larger
among households without pension coverage, as those with IRAs had median net worth that was
more than 16 times the net worth of households without IRAs ($339,540 versus $20,560).30
Table 7. Financial Characteristics of Households by Pension Coverage and IRA
Ownership in 2007
Median Dollar Amounts

Households with Pension Coverage
Households without Pension Coverage

Have IRA
Do not Have IRA
Have IRA
Do not Have IRA
Non-IRA
$40,000
$12,000
$0
$0
retirement assets
Total retirement
$105,000 $12,000 $38,000
$0
assets
Total assets
$602,800
$223,970
$484,300
$47,980
Debt
$135,000
$78,400
$72,000
$14,000
Net worth
$411,000
$100,150
$339,540
$20,560
Source: CRS analysis of the 2007 Survey of Consumer Finances.
Demographic Characteristics and IRA Ownership
Table 8 provides a demographic breakdown of IRA owners, grouped by their pension coverage.
In nearly every category, IRA ownership was higher among households with pension coverage.

30 Non-IRA retirement assets are mainly 401(k) and other thrift accounts. Total assets are financial assets and non-
financial assets. Financial assets are mainly saving and checking account, stocks, bonds, and savings bonds. Non-
financial assets are mainly housing, vehicles, and business interests.
Congressional Research Service
15

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

The exceptions were among households with incomes between $50,000 and $75,000; households
with incomes greater than $100,000; and households in which the head of the household is aged
70 or older.
Income
IRA ownership increased with income for both households with and without pension coverage
although households with pension coverage did not have higher IRA ownership rates in all
income groups. Within income groups, there is not a clear pattern of IRA ownership between
households with and without pension coverage. Households without pension coverage had higher
IRA ownership rates than households with pension coverage in two income groups: households
with incomes from $50,000 to $75,000 and households with incomes greater than $100,000. At
least two factors might explain the higher IRA ownership rate for households with incomes
greater than $100,000 and without pension coverage: (1) the positive correlation between IRA
ownership and income and (2) all of these households were eligible for a tax deduction for their
contributions because they did not have pension coverage and were likely in the highest income
tax brackets, which means they received larger per-dollar tax benefits for each dollar of IRA
contribution compared to lower income households.
Age
The 2007 Survey of Consumer Finances data show that for both households with and without
pension coverage IRA ownership increased with age until age 70. Households in which the head
was aged 70 or older had lower rates of IRA ownership than those aged 60 to 69.31 The difference
between rates of IRA ownership decreased as the age of the head of the household increased.
Among households in which the head was under the age of 30, the rate of IRA ownership was
10.8 percentage points lower for households without pension coverage (22.5% compared with
10.7%). By comparison, among households in which the head was aged 60 to 69, the rate of IRA
ownership was 1.9 percentage points lower for households without pensions (50.0% compared
with 48.1%).
Education
IRA ownership rates were lower among households in which the head of the household did not
have a college degree. The difference was particularly large among households without pensions:
the rate of IRA ownership for households without pension coverage and in which the head had
some college education was less than half the IRA ownership rate for households without pension
coverage but in which the head had a college degree (23.8% versus 47.9%). This may indicate a
willingness to trade current income for larger future income, as college students generally forgo at
least some current income while anticipating higher income in the future. Contributing to an IRA
is a similar decision, as IRA owners forgo current income in exchange for larger income in
retirement.

31 This difference may not be surprising given that the sample excludes households where neither the head nor spouse
were working. Most households in which the head is aged 70 or older were retired and therefore not included in
tabulations for this report.
Congressional Research Service
16

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Race and Ethnicity
The IRA ownership rate was 32.4% among households without pension coverage in which the
head of the household was white, 7.3% among households that did not have pension coverage and
in which the head was African-American, and 6.1% among households that did not have pension
coverage and in which the head was Hispanic. Some of these groups may be particularly
economically vulnerable, as they likely have few assets and may need to rely on Social Security
for most of their retirement income.32 The IRA ownership rate was 37.4% among households that
did not have pension coverage and in which the head was Asian, Pacific-Islander, American
Indian, or another race.33
Homeownership
IRA ownership rates were higher among homeowners than non-homeowners. Low IRA
ownership rates among non-homeowners (17.9% for non-homeowners with pensions and 10.6%
for non-homeowners without pensions) may reflect several factors. Homeownership rates
increase as household income and age increase. In addition, homeownership may be a proxy for
long-term planning or higher levels of financial literacy, both of which are associated with higher
IRA ownership rates.
Marital Status
Married households had higher IRA ownership rates than single households, although the
difference was not as pronounced among households without pension coverage. Among
households without pension coverage, 26.9% of married households had IRAs in 2007, while
23.5% of single male households had IRAs and 22.5% of single female households had IRAs. It
is interesting to note that IRA ownership rates were only one percentage point higher among
single male households compared with single female households.

32 Analysis of the SCF indicates that median net worth in 2007 was $163,000 among households in which the head of
the household was white, and $17,100 (89.5% lower) among households in which the head was African-American,
$21,010 (87.1% lower) among households in which the head was Hispanic, and $156,000 (4.3% lower) among
households in which the head was not white, African-American, or Hispanic.
33 For its public release, the SCF combines Asian, Pacific-Islander, American Indian, and other races into one category.
Congressional Research Service
17

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Table 8. Demographic Characteristics of IRA Owners
Households with Pensions
Households without Pensions
Number of
Number of
Households
Percentage
Households
Percentage

(millions)
with IRAs
(millions)
with IRAs
Income
< $50,000
2.2
17.6%
3.2
13.2%
$50,000 - 75,000
3.1
28.0%
2.2
32.8%
$75,001 - $100,000
3.7
45.1%
1.2
36.4%
> $100,000
9.4
59.4%
3.4
63.5%
Age
20 – 29
1.1
22.5%
0.9
10.7%
30 – 39
3.2
30.4%
1.5
17.0%
40 – 49
5.5
40.0%
2.1
23.6%
50 – 59
6.0
46.7%
2.2
34.9%
60 – 69
2.2
50.0%
2.3
48.1%
70 and older
0.3
32.2%
0.9
45.3%
Education
High school or less
3.1
20.5%
2.8
13.6%
Some col ege
4.1
34.9%
2.2
23.8%
Finished college
6.2
48.6%
2.8
47.9%
Graduate school
4.9
64.2%
2.1
57.3%
Head of household is
White 15.5
43.4%
8.6
32.4%
African-American 1.0
18.7%
0.4
7.3%
Hispanic 0.7
20.0%
0.4
6.1%
Asian, Pacific-Islander, American
1.1 40.1% 0.6 37.4%
Indian, or other
Household owns home
16.6 43.9% 8.2 36.2%
Household does not own home
1.7 17.9% 1.7 10.6%
Marital status of household
Married
14.1
41.3%
6.5
26.9%
Single male
1.8
35.5%
1.4
23.5%
Single female
2.4
29.6%
2.1
22.5%
Source: CRS analysis of the 2007 Survey of Consumer Finances.
Note: The Survey of Consumer Finances reports age, education, and racial / ethnic category for the head of the
household. Income is reported for the entire household.
Congressional Research Service
18

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Saver Characteristics and IRA Ownership
Individuals vary in their preferences for saving. Some individuals may be quite willing to give up
current consumption in order to receive higher consumption in the future. Other individuals may
be less willing to make this trade-off. Individuals who have a high preference for saving may be
more likely to own IRAs. Table 9 presents IRA ownership rates in combination with factors that
indicate attitudes toward saving. The Survey of Consumer Finances asks respondents a number of
questions to elicit their attitudes toward saving. The answers might provide insight into saving
propensities, which might be an important factor in determining which households own IRAs.
Three questions in the survey related to saver type: (1) the household’s most important reason for
saving, (2) the most important time period in planning the family’s spending and saving, and (3)
the amount of shopping around to find the best terms when making saving and investment
decisions.
Among both households with and without pension coverage, households that have higher
propensities to save have higher IRA ownership rates. Among households without pension
coverage, IRA ownership rates increase substantially as households increase their savings horizon
from less than one year to a savings horizon of two to 10 years (from 12.3% to 31.0%), and as
households increase the amount of shopping the household engages in when making saving
decisions from little shopping to a moderate amount of shopping (from 18.0% to 28.5%). This
suggests that efforts to increase financial literacy among less financially literate households may
lead to higher IRA ownership rates.
Table 9. Saver Type Characteristics of IRA Owners, 2007

Households with Pensions
Households without Pensions
Number of
Number of
Households
Percentage
Households
Percentage

(millions)
with IRAs
(millions)
with IRAs
Retirement is not primary reason for
7.3 29.9%
5.3 18.7%
saving
Retirement is primary reason for
11.0 48.0%
4.6 42.9%
saving





Savings horizon: 1 year or less
2.9
24.5%
1.8
12.3%
Savings horizon: 2 to 10 years
10.5
39.7%
6.2
31.0%
Savings horizon: More than 10 years
4.9
53.9%
1.8
44.9%





Household does little shopping when
4.6 32.0%
2.4 18.0%
making savings decisions
Household does moderate shopping
6.2 37.8%
3.8 28.5%
when making savings decisions
Household does a great deal of
7.5 45.2%
3.7 29.5%
shopping when making savings
decisions
Source: CRS analysis of the 2007 Survey of Consumer Finances.
Congressional Research Service
19

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Policy Options
Analysis of the data presented in this report suggests that pension plan coverage is an important
factor for determining whether a household owns an IRA. IRA ownership is not as widespread
among households without pension coverage, yet these are the households that IRAs are targeted
to help. In fact, IRA ownership rates are more than 50% higher among households with pension
coverage compared with households without any pension coverage. Households with neither
IRAs nor pensions are about 34% of all working households and disproportionately represent
economically vulnerable populations, such as lower-income, less-educated, and minority
households. Some policymakers have expressed concern that many workers have inadequate
savings for retirement. Half of the working households with neither pension coverage nor IRAs
have net worth less than $20,560. In some cases, these workers may have less wealth than they
anticipated to enjoy retirement. Some of these households may have to rely on public assistance
to meet their basic needs. To formulate more effective retirement savings policies, policymakers
need to be aware of how IRA ownership is affected by factors such as pension coverage,
education, income, and behavior (e.g., saver characteristics).
An issue for policymakers is how to encourage workers to better prepare for retirement,
particularly when they may have other, more immediate financial concerns. Retirement security is
often thought of as a three-legged stool: Social Security, private savings, and employer-provided
pensions. Although participation in Social Security is mandatory for most workers, the other legs
have traditionally been voluntary, though encouraged through tax incentives. For example,
employer contributions to pension plans are a tax-deductible expense; participants’ contributions
to defined contribution plans are not included in current taxable income; contributions to
traditional IRAs may be tax-deductible; and qualified distributions from Roth IRAs are not taxed.
Current tax incentives, however, may not be effective for some target populations. Several policy
proposals have been suggested to help workers better prepare for retirement.
Changing the Retirement Savings Contributions Credit
The Retirement Savings Contributions Credit (the Saver’s Credit) is an additional incentive to
encourage lower-income households to save for retirement.34 This non-refundable tax credit of up
to $1,000 was authorized by the Economic Growth and Tax Relief Reconciliation Act of 2001
(P.L. 107-16) for eligible individuals who contribute to an IRA or an employer-sponsored
retirement plan. The credit is available to single filers with incomes up to $27,750 and married
filers with incomes up to $55,500. IRS data indicates that 5.9 million households received the
Saver’s Credit in 2007 and that the average amount of the credit received was $167. Table 10
indicates that the credit may not be widely used: 5.9% of households with Adjusted Gross
Incomes under $50,000 claimed the Saver’s Credit in 2007.

34 For more information see CRS Report RS21795, The Retirement Savings Tax Credit: A Fact Sheet, by John J.
Topoleski.
Congressional Research Service
20

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Table 10. Retirement Savings Contributions Credit Use In 2007
Number of Tax Returns
Percentage of Tax
Adjusted Gross
With Saver’s Credit
Returns in AGI Group
Average (Mean) Saver’s
Income
(thousands)
With Saver’s Credit
Credit Amount
Less than $10,000
30.5
0.1%
$74
$10,000 to $19,999
945.6
4.1%
$162
$20,000 to $29,999
2,042.6
10.8%
$173
$30,000 to $39,999
1,427.0
9.7%
$168
$40,000 to $49,999
1,126.5
10.1%
$159
$50,000 to $74,999
290.0
1.5%
$171
Source: Internal Revenue Service, Statistics of Income data.
Note: No returns with Adjusted Gross Income greater than $75,000 claimed the Saver’s Credit in 2007.
Several factors may limit the effectiveness of the Saver’s Credit: (1) it is non-refundable, so
households with little or no tax liability receive little or no benefit; (2) it is not available to
individuals who file their taxes using form 1040EZ; and (3) the credit may be used by individuals
who would have made retirement plan contributions in the absence of the credit. Thus, not all
retirement plan contributions tied to the credit are “new” contributions.35
Adopting Automatic IRAs
Because current tax and other incentives have not substantially increased voluntary participation
in the retirement savings system, some policymakers believe that IRA ownership rates could be
raised through increased access to retirement accounts. One of the challenges is determining how
to comprehensively provide retirement accounts to workers who do not have employer-sponsored
pension plans. The success of automatic enrollment in 401(k) plans has prompted calls for an
automatic enrollment program for IRAs.36
The Automatic IRA (Auto IRA) is a proposal for increasing workers’ access to retirement
accounts.37 Under such proposals, employers above a certain size that have been in business a
certain number of years and that do not offer a pension plan would be required to establish IRAs
for their employees. The employer would direct a specified percentage of each employee’s pay
into the IRA. Eligible employees would automatically be enrolled by their employers in the Auto
IRA program, but could opt-out of participation or change the amount of their contribution.
Employers could, but would not be required to, match their employees’ contributions.

35 In the 111th Congress, Rep. Earl Pomeroy introduced H.R. 1961, the Savings for American Families’ Future Act of
2009, and Sen. Kirsten Gillibrand introduced S. 3090. These bills would expand eligibility for the Saver’s Credit and
make the credit refundable. In addition, H.R. 1961 would double the amount of the tax credit if the credit is paid
directly into a retirement account.
36 In automatic enrollment defined contribution plans, new employees are deemed to be participating in the pension
plan, so eligible participants must opt-out rather than opt-in to the plan. Participation may include employee and
employer contributions to the account.
37 In the 111th Congress, Sen. Jeff Bingaman introduced S. 3760, the Automatic IRA Act of 2010, which would provide
for Auto IRAs. Rep. Richard Neal introduced companion legislation, H.R. 6099. Further background on the Auto IRA
proposal can be found in Pursuing Universal Retirement Security Through Automatic IRAs, by J. Mark Irwy and David
C. John, available at http://www.brookings.edu/papers/2009/07_automatic_ira_iwry.aspx.
Congressional Research Service
21

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

Although approximately 39.1 million households work for employers that do not offer pension
plans (about 50% of U.S. working households), not all of these households would be covered by
Auto IRAs. Some households might work for employers that are small enough to be exempt from
offering Auto IRAs or some households might not meet job tenure requirements.38
An Auto IRA program for most workers without pension coverage could be costly to implement
and administer, as millions of individual accounts would likely need to be processed. One goal of
Auto IRA proponents is to minimize employers’ administrative burdens and fiduciary obligations.
For example, there may be questions about which default investments are appropriate for workers
that do not make active participation decisions. In addition, Auto IRAs would provide IRAs to
some workers who already have traditional or Roth IRAs. Among households that do not have an
employer-sponsored pension plan, and therefore would be potentially eligible for the Auto IRA,
25.3% already own IRAs. While individuals could opt out of the Auto IRA, the possibility exists
that some individuals may inadvertently exceed the yearly IRA contribution limit. The tax penalty
is 6% of amount of the excess contribution.39 The Treasury Inspector General for Tax
Administration noted that noncompliance with IRA contribution limits is an area of concern and
that the IRS has inadequate procedures to ensure compliance. It is estimated that 146,976
taxpayers exceeded the contribution limit in 2007.40
Improving Financial Literacy
Efforts to improve financial literacy may increase IRA contributions and ownership rates by
increasing knowledge of the benefits of IRAs, the Saver’s Credit, and the option to have federal
income tax refunds deposited directly to IRAs. Many government agencies, non-profit groups,
and for-profit companies have ongoing campaigns aimed at improving financial literacy (for
example, increasing awareness of credit card fees). In particular, some of these efforts focus on
the importance of saving for retirement.41
Increasing Fee Transparency
Small differences in the amount of fees that financial institutions charge for managing IRAs can
yield large differences in the account balances at retirement. Some policymakers have expressed
concern regarding the fees that financial institutions charge 401(k) plan participants. Legislation
has been introduced that would increase 401(k) plan sponsors’ and participants’ awareness of the
fees that they pay.42 However, legislation to increase fee transparency for IRA owners has not
been introduced. The Department of Labor recently issued an interim final rule that would require

38 The SCF indicates that 10.4 million households worked for an employer that had less than 10 employees and did not
offer a pension plan in 2007.
39 If the excess contribution is withdrawn by the due date of the tax return, the 6% tax penalty does not apply.
40 For more information, see Treasury Inspector General for Tax Administration, A Service-wide Strategy Is Needed to
Address Growing Noncompliance With Individual Retirement Account Contribution and Distribution Requirements,
available at http://www.treas.gov/tigta/auditreports/2010reports/201040043fr.pdf.
41 For more information, see NBER Working Paper No. 13824, Household Saving Behavior: The Role of Financial
Literacy, Information, and Financial Education Programs, by Annamaria Lusardi, available at http://www.nber.org/
papers/w13824.
42 More information is available in CRS Report RL34678, Fee Disclosure in Defined Contribution Retirement Plans:
Background and Legislation, by John J. Topoleski.
Congressional Research Service
22

---

.
Ownership of Individual Retirement Accounts and Policy Options for Congress

greater transparency of fees paid in 401(k) plans.43 The final rule specifically excludes IRAs,
noting that IRA owners are solely responsible for managing their accounts and that IRA fee
disclosure requirements could be quite costly. One difference between IRAs and 401(k) plans is
that IRA owners have complete control over the choice of the financial institution and
investments. However, individuals may find the decisions associated with opening and
maintaining IRAs complicated and perhaps overwhelming. This suggests that opportunities may
exist for financial institutions to charge higher fees than necessary. A direct comparison of IRA
and 401(k) fees is not possible because of the different structures of the plans. IRAs are likely
administratively costly, as each separate account requires separate record keeping, whereas
401(k) plan administrators can take advantage of economies of scale.

Author Contact Information

John J. Topoleski

Analyst in Income Security
jtopoleski@crs.loc.gov, 7-2290



43 See Department of Labor, Employee Benefits Security Administration, “Reasonable Contract or Arrangement Under
Section 408(b)(2)—Fee Disclosure,” Federal Register, vol. 75, no. 136 (July 16, 2010), pp. 41600-41638.
Congressional Research Service
23

---