Saving for Retirement: Household Decisionmaking and Policy Options

July 2, 2020 R46441

Saving for Retirement: Household
July 2, 2020
Decisionmaking and Policy Options
Cheryl R. Cooper
Retirement saving and financial planning have become more important for achieving financial
Analyst in Financial
security during retirement. After three decades of growth, employer-sponsored defined
Economics
contribution (DC) plans and tax-favored Individual Retirement Accounts (IRAs) have become

two important retirement income sources for many Americans. About 52% of U.S. households
Zhe Li
had retirement assets in DC plans or IRAs in 2016. In the past few decades, the major
Analyst in Social Policy
responsibility of retirement investing and planning has shifted from the employer to the

American worker. With DC plans and IRAs, people typically need to make decisions about how
much to contribute each year, how to invest their retirement wealth over a lifetime, and how to
For a copy of the ful report,
withdraw their funds in retirement without outliving their assets. Making these decisions may
please cal 7-5700 or visit
require complicated calculations involving uncertainty about future conditions, such as
www.crs.gov.
expectations about inflation and market returns, estimates about health care expenditures, and
projections about family members’ longevity. Some people may find it difficult to navigate the various options and decisions.
American families with different income levels and different access to retirement savings vehicles may have different
retirement savings needs.
 Households with adequate savings in retirement accounts or on track for adequate retirement income may
include employees with high incomes and existing retirement plans, such as those who have employer-
sponsored pensions or who diligently fund their retirement plans. Retirement savings policies for those
families may focus on protecting their investment wealth by regulating investment disclosure and
management, facilitating investment choices, and regulating to ensure appropriate investment advice.
 Some middle-income families with little or no retirement savings or retirement income beyond Social
Security may not be on track for adequate retirement income. Some of those families may find saving for
retirement difficult for financial reasons. Others may not be eligible to participate in an employer-
sponsored retirement savings plan, making saving for retirement more challenging. Families who may have
difficulty maintaining their preretirement living standards during retirement may benefit from policies to
increase participation and contributions in retirement plans.
 For some lowest-income families, Social Security and Supplemental Security Income (SSI) generally
replace a substantial proportion of preretirement earned income. Income from DC plans and IRAs is
typically a small proportion of their retirement income. If these families are not on track for adequate
retirement income, policy changes in SSI and other government income support programs may be an
appropriate way to help these families meet basic needs in retirement.
Behavioral research suggests that humans tend to have biases in rather predictable patterns. For example, the number, order,
and structure of options, as well as the process around the choice, can change decisions for many people. Although
consumers might not be aware of these biases when making financial decisions, research suggests that these biases can be
used to encourage consumers to save more for retirement and make better retirement decisions.
Using best practices from behavioral research, policymakers may consider how to best structure retirement accounts and
retirement planning decisions to help more people achieve retirement security. Common policy options around saving for
retirement tend to relate to (1) retirement account features, (2) investment policies, and (3) financial literacy. For example,
some retirement account features that may help consumers save more for retirement include making payroll-deductible
retirement savings accounts available to every citizen, automatic enrollment of participants into retirement savings accounts,
and automatic escalation of participants’ contributions. In addition, how retirement account information and investment
options are disclosed may impact consumer decisionmaking. People may also find retirement planning challenging because
of limited financial management, knowledge, and skills (or financial literacy).
In 2019, Congress passed the Setting Every Community Up for Retirement Enhancement Act (SECURE Act; P.L. 116-94,
Division O). One of the act’s goals was to encourage retirement savings, for example, by expanding Americans’ DC plan
access through employers and making changes to DC plans to try to make retirement decisionmaking easier.
Congressional Research Service

Saving for Retirement: Household Decisionmaking and Policy Options

Introduction
Many Americans have more responsibility for their retirement planning today than in the recent
past. Over the past three decades, the number of employees covered by employer-sponsored
defined benefit (DB) pension plans has fal en, while an increasing percentage of workers are
covered by employer-sponsored defined contribution (DC) pension plans.1 From the early 1990s
(1992-1993) to 2019, the percentage of private-sector workers who were covered by DB plans
decreased from 32% to 12%, and the percentage of those workers who participated in DC plans
increased from 35% to 47%.2
DB pension plans usual y offer a lifetime annuity that is typical y based on employees’ tenure and
wages and is funded solely by employers.3 In contrast, employer-sponsored DC plans, such as a
401(k), are retirement savings accounts funded through tax-deductible contributions by the
worker (frequently matched in part or fully by the employer).4 Employees with self-directed DC
plans need to decide how much income to contribute to retirement accounts each year and how to
invest these contributions to maximize their retirement wealth. The shift from DB to DC
retirement plans means, among other things, that individuals and families over time have had to
assume more responsibility for managing their retirement finances.
Additional y, individuals or married couples who have employment earnings can also establish
Individual Retirement Accounts (IRAs) to accumulate funds for retirement on a tax-advantaged
basis.5 Over time, IRA account balances may contain both direct contributions and rollovers of
certain DC accounts from previous employers.
In 2016, about 30% of households had IRAs.6 Similar to those with DC plans, households with
IRA plans need to decide how much income to contribute to these accounts and how to manage
their investments.

1 See Sebastian Devlin-Foltz, Alice M. Henriques, and John Sabelhaus, The Evolution of Retirement Wealth, Board of
Governors of the Federal Reserve System (Federal Reserve), 2015, at https://www.federalreserve.gov/econresdata/feds/
2015/files/2015009pap.pdf.
2 See U.S. Bureau of Labor Statistics (BLS), National Compensation Survey (NCS), “T able 2. Retirement Benefits:
Access, Participation, and T ake-up Rates, Private Industry Workers,” March 2019, at https://www.bls.gov/ncs/ebs/
benefits/2019/ownership/private/table02a.pdf (hereinafter NCS, “ T able 2. Retirement Benefits: Private Industry
Workers,” 2019); and Stephanie L. Costo, “T rends in Retirement Plan Coverage over the Last Decade,” Monthly Labor
Review, February 2006, at https://www.bls.gov/opub/mlr/2006/02/art5full.pdf. T his report focuses on private-sector
workers. For federal employees and retirement savings, see CRS Report RL30387, Federal Em ployees’ Retirem ent
System : The Role of the Thrift Savings Plan , by Katelin P. Isaacs. Among state and local government employees, about
76% participated in defined benefit (DB) plans, compared with 17% in defined contribution (DC) plans in 2019.
T herefore, DB plans might be more relevant to the retirement income for state and local government workers. See BLS,
NCS, “T able 2. Retirement Benefits: Access, Participation, and T ake-up Rates, State and Local Government Workers,”
March 2019, at https://www.bls.gov/ncs/ebs/benefits/2019/ownership/govt/table02a.pdf.
3 In DB plans, workers receive monthly benefits in retirement (called annuities) based on an employee’s income, age,
and/or length of service. For DB plans, typically only the employer contributes to the plan on behalf of the employee.
In DB plans, the worker has a claim to the monthly benefit payable at retirement and n ot to any assets.
4 In DC plans, contributions and returns generally accumulate independently of job tenure, though employer
contributions may require the employee to meet a minimum vesting requirement.
5 Contributions in Individual Retirement Accounts (IRAs) may be made on a pre-tax basis in traditional IRAs or on a
post -tax basis in Roth IRAs. CRS Report RL34397, Traditional and Roth Individual Retirem ent Accounts (IRAs): A
Prim er, by Elizabeth A. Myers.
6 CRS analysis of the 2016 Survey of Consumer Finances (SCF). T he SCF is a triennial survey conducted on behalf of
the Federal Reserve. It contains detailed information on U.S. household finances, such as the amount and types of
Congressional Research Service
1

Saving for Retirement: Household Decisionmaking and Policy Options

Financial decisions related to retirement accounts have become more complex. The tax
implications of different retirement vehicles can be complicated. In addition, individuals
general y have access to more investment product offerings within retirement accounts, further
complicating these decisions.7
Americans are also on average living longer than in the past. Life expectancy at age 65 has
increased over the past century or so, from 11.9 years in around 1900 (on average, living to about
77 years old) to 19.2 years in 2010 (on average, living to about 84 years old).8 Moreover, the
likelihood of a 65 year old living past 90 years old rose from 25% for women and 10% for men in
1965 to 34% for women and 22% for men in 2015.9 While some Americans may be physical y
able to and choose to work longer and retire at an older age, this increase in life expectancy
means that many individuals need to plan for a longer period in retirement than in the past,
making saving and financial planning more important for maintaining their preretirement
standard of living.
For these reasons, household financial decisionmaking relating to retirement is both difficult and
important. Currently, retirement planning includes many decisions relating to saving, investing,
and decumulating (or spending down) retirement wealth over a person’s lifetime. Making these
decisions successfully may require complicated calculations involving uncertainty about the
future—for example, how high (or low) inflation and market returns wil be, how much future
health care wil cost, and how long family members may be able to work and wil live. Given the
variety of options and decisions to make, researchers claim that many households may find
planning for retirement to be overwhelming or intimidating.10
People may also find retirement planning chal enging because of limited financial management
experience, knowledge, and skil s (or financial literacy).11 Insufficient financial literacy can
reduce investment returns and is associated with less retirement planning and less wealth
accumulation.12 Retirement general y happens only once for a household. Unlike other financial
domains, such as auto loans, for which a family might shop every few years, many people do not
have prior experience to inform their own retirement needs. For these reasons, planning for
retirement can be particularly chal enging.
Given this retirement savings context, in 2019, Congress passed the Setting Every Community
Up for Retirement Enhancement Act (SECURE Act; P.L. 116-94, Division O). One of the goals of

assets owned, the amount and types of debt owed, and detailed demographic information on the head of the household
and spouse. For more information, see Federal Reserve, SCF, at https://www.federalreserve.gov/econres/scfindex.htm.
7 Annamaria Lusardi, Financial Literacy and Financial Decision-Making in Older Adults, American Society on Aging,
2012, at https://www.asaging.org/blog/financial-literacy-and-financial-decision-making-older-adults.
8 Loraine A. West et al., 65+ in the United States: 2010, Census Bureau, Current Population Reports, June 2014, p. 25,
at https://www.census.gov/content/dam/Census/library/publications/2014/demo/p23-212.pdf.
9 T he Hamilton Project, Brookings Institute, Probability of a 65-Year-Old Living to a Given Age, by Sex and Year, June
23, 2015, at https://www.hamiltonproject.org/charts/
probability_of_a_65_year_old_living_to_a_given_age_by_sex_and_year. Although life expectancy has generally been
increasing over time in the United States, researchers have long documented that it is lower for individuals with lower
socioeconomic status (SES) compared with individuals with higher SES. See CRS Report R44846, The Growing Gap
in Life Expectancy by Incom e: Recent Evidence and Im plications for the Social Security Retirem ent Age , by Katelin P.
Isaacs and Sharmila Choudhury.
10 Sheena S. Iyengar, Wei Jiang, and Gur Huberman, How Much Choice is Too Much?: Contributions to 401(k)
Retirem ent Plans, Pension Research Council, Working Paper, 2003 (hereinafter Iyengar, Jiang, and Huberman, 2003).
11 Annamaria Lusardi and Olivia Mitchell, “ T he Economic Importance of Financial Literacy: T heory and Evidence,”
Journal of Econom ic Literature, vol. 52, no. 1 (2014), p. 34 (hereinafter Lusardi and Mitchell, 2014).
12 Lusardi and Mitchell, 2014, pp. 24-44.
Congressional Research Service
2

Saving for Retirement: Household Decisionmaking and Policy Options

this act was to encourage saving for retirement—for example, through expanding Americans’ DC
plan access through employers and making changes to DC plans to try to make retirement
decisionmaking easier.13
This report focuses on household decisions related to saving for retirement in DC plans and IRAs.
First, it discusses household finances and retirement savings. Then, it summarizes relevant
consumer decisionmaking research. Lastly, it discusses policy options to encourage plan sponsors
to modify retirement savings accounts so that people may increase their participation and savings
in those accounts. Related policy issues discussed in this report include the following:
 availability, automatic enrollment, and automatic escalation in retirement saving
accounts;
 matching of individuals’ contributions in retirement accounts;
 financial literacy of the working-age population;
 retirement account information disclosure and reporting;
 investment options and management of retirement accounts;
 fiduciary and investment advice related to retirement savings;
 maximum age al owed for contributions to retirement accounts;
 leakage and early withdrawal rules for retirement accounts; and
 converting savings into retirement income.
Retirement Savings and Household Finances
Individuals and households frequently save for retirement in DC plans, IRAs, or both. About 52%
of U.S. households had retirement assets in DC plans or IRAs in 2016.14 Other parts of a
household’s finances, such as debts, emergency savings, and employment outcomes, may also
affect retirement security over time. Research has offered mixed conclusions regarding whether
Americans overal have enough savings for retirement.15 Some research suggests that certain
subgroups of the population may need to be saving more to fund retirement.16
DC Pension Plans and IRAs17
People may have retirement savings in a variety of tax-advantaged plans, such as DC plans or
IRAs. An individual or household also may simultaneously have DC plans and IRAs.
DC plans are available only to workers whose employers sponsor such plans.18 For those plans,
employees contribute a percentage of their wages, on a pre-tax basis, to a retirement savings

13 See CRS In Focus IF11174, The SECURE Act and the Retirement Enhancement and Savings Act Tax Proposals
(H.R. 1994 and S. 972), by Jane G. Gravelle.
14 CRS analysis of the 2016 SCF.
15 See Vickie L. Bajtelsmit and Anna Rappaport, “Retirement Adequacy in the United States,” Journal of Financial
Service Professionals, vol. 72, no. 6 (November 2018), pp. 71 -86 (hereinafter Bajtelsmit and Rappaport, 2018). This
journal article summarizes main findings in related studies.
16 Bajtelsmit and Rappaport, 2018.
17 For more information about DC plans, see CRS Report R40707, 401(k) Plans and Retirement Savings: Issues for
Congress, by John J. T opoleski; for more information about IRAs, see CRS Report RL34397, Traditional and Roth
Individual Retirem ent Accounts (IRAs): A Prim er, by Elizabeth A. Myers.
18 T he most common types of DC plans are 401(k), 403(b), 457(b), the T hrift Savings Plan (T SP), Savings Incentive
Congressional Research Service
3

Saving for Retirement: Household Decisionmaking and Policy Options

account established by the employer. Employers may also make a contribution to the plan equal
to some or al of the worker’s contribution (cal ed a “match”).19 Both contributions and gains are
tax deferred until withdrawn. Beginning in 2006, employers with 401(k) and 403(b) sponsored
plans were able to create qualified retirement accounts where employees could make
contributions using after-tax income, which are usual y referred to as Roth 401(k) and Roth
403(b).20 Withdrawals from a Roth-type plan are general y tax-free if the taxpayer meets the
applicable distribution requirements.
IRAs are tax-advantaged accounts that eligible individuals (or married couples) can establish to
accumulate funds for retirement.21 IRAs are available to any worker with taxable compensation,
as defined by the Internal Revenue Service (IRS).22 As with 401(k)s, there are two types of IRAs:
traditional and Roth. For traditional IRAs, contributions may be tax-deductible, depending on
whether the contributor or the contributor’s spouse (if present) has participated in a pension plan
at his or her place of employment and their income level.23 In retirement, withdrawals from
traditional IRAs are general y included in the IRA owner’s taxable income. Any individual who
has taxable compensation for the purpose of an IRA and whose income is under specified limits
may establish and contribute to a Roth IRA.24 Contributions to Roth IRAs are not tax-deductible,
and qualified distributions are not taxable.25
IRAs are often used to receive the balance of a DC retirement account as part of a rollover. A
rollover is the transfer of assets from one retirement plan to another (typical y to an IRA) upon
separation from the original employer, either at job change or at retirement.26

Match Plan for Employees (SIMPLE), and Simplified Employee Pension (SEP). 401(k), 403(b) and 457(b) plans are
named for the sections of the Internal Revenue Code that authorize the plans. Priv ate-sector employers sponsor 401(k)
plans, public school systems and nonprofit organizations sponsor 403(b) plans, and state and local governments
sponsor 457(b) plans. T he federal government sponsors the T SP. SIMPLE and SEP plans are for small business
employees.
19 An employee may contribute up to $19,500 ($26,000 if 50 or older) per year to a 401(k) plan in 2020. T he
contribution limit is adjusted annually for increases in the national wage. For more information, see Internal Revenue
Service (IRS), 2020 Lim itations Adjusted As Provided in Section 415 (d), etc., Notice 2019-59, at https://www.irs.gov/
pub/irs-drop/n-19-59.pdf.
20 Roth 401(k) and Roth 403(b) are authorized by the Economic Growth and T ax Relief Reconciliation Act of 2001
(P.L. 107-16).
21 Individuals may contribute up to $6,000 ($7,000 if 50 or older) in 2020 to an IRA. T he contribution limit is adjusted
annually for increases in the national wage. For more information, see CRS Report RL34397, Traditional and Roth
Individual Retirem ent Accounts (IRAs): A Prim er, by Elizabeth A. Myers.
22 Compensation for purposes of an IRA may include wages and salaries, commissions, self -employment income,
alimony and separate maintenance pay, and nontaxable combat pay. For more information, see IRS, Contributions to
Individual Retirem ent Arrangem ents (IRAs), Publication 590-A, p.6, at https://www.irs.gov/pub/irs-pdf/p590a.pdf.
23 For more information, see IRS, “IRA Deduction Limits,” December 20, 2019, at https://www.irs.gov/retirement-
plans/ira-deduction-limits.
24 No contributions are allowed in Roth IRAs in 2020 if income is greater than $206,000 for married filing jointly or
qualifying widow(er); $139,000 for single, head of household, or married filing separately and who did not live with a
spouse at any time during the year; and $10,000 for married filing separately and who lived with a spouse at any time
during the year.
25 Qualified distributions from a Roth IRA must satisfy both of the following conditions: (1) they are made after the
five-year period beginning with the first taxable year for which a Roth IRA contribution was made, and (2) they are
made on or after the age of 59½, because of disability, to a beneficiary or estate after death, or to purchase, build, or
rebuild a first home up to a $10,000 lifetime limit.
26 Within traditional IRAs, more funds flowed in through rollovers from employer -sponsored pension plans than from
regular contributions. For example, in 2016 (the latest year for which such data are available) funds from rollovers were
$430.8 billion, whereas funds from contributions were only $18.3 billion. In contrast, funds flowing in to Roth IRAs
Congressional Research Service
4

link to page 9 Saving for Retirement: Household Decisionmaking and Policy Options

Household Retirement Assets, 1989-2016
In response to the shift in retirement savings vehicles from DB to DC plans, household retirement
assets, such as funds in DC plans and IRAs, have grown over time. Figure 1 shows the
percentage of U.S. households that had assets in retirement savings accounts (excluding future
benefit estimates in DB plans or Social Security) and median retirement assets among those
households from 1989 to 2016.27 In 1989, about 37% of U.S. households had retirement assets; by
2016, this had increased to 52%.28 The median retirement assets among households participating
in DC plans and IRAs also increased during this time, from $20,524 in 1989 (inflation adjusted)
to $60,000 in 2016.29

from rollovers were $8.6 billion in 2016, whereas funds from contributions were $22.2 billion. For statistical tables of
IRA plans, see IRS, Statist ics of Income (SOI) Division, “ SOI T ax Stats – Accumulation and Distribution of Individual
Retirement Arrangements,” at https://www.irs.gov/statistics/soi-tax-stats-accumulation-and-distribution-of-individual-
retirement -arrangements#tables.
27 T he SCF includes the following tax-advantaged accounts for retirement assets: DC plans and IRAs, Profit Sharing
Plans, Supplementary Retirement Annuities, Cash Balance Plans, Portable Cash Option Plans, etc. Importantly,
estimates of future income from DB pensions or from Social Security are not included in the calculations of retirement
assets. For more information on the SCF, see Federal Reserve, Codebook for 2016 Survey of Consum er Finances, at
https://www.federalreserve.gov/econres/files/codebk2016.txt . T he definitions of retirement assets and other summary
statistics used by the Federal Reserve are available at https://www.federalreserve.gov/econres/files/bulletin.macro.txt .
28 For comparison, in 2000, there were $5.0 trillion in DB plan assets and $5.6 trillion in DC plans and IRAs. In the
first quarter of 2020, there were about $9.1 trillion assets in DB plans and $17.4 trillion in DC plans and IRAs. See
Investment Company Institute (ICI), Defined Contribution Plan Participants’ Activities, First Quarter 2020, May
2020, at https://www.ici.org/pressroom/news/20_recordkeeper_1q20 (hereinafter ICI, DC Plan Participants’ Activities,
First Quarter, 2020).
29 T he median lies at the middle of the retirement assets distribution. Half of the households have higher retirement
assets, and half have lower retirement assets. T he measure of average retirement assets is generally higher than the
median amount because a relatively small percentage of households have very high retirement assets. From 1989 to
2016, the average measure of retirement assets increased from $70,991 to $228,725 based on the SCF (adjusted to 2016
dollars based on the Consumer Price Index for All Urban Consumers [CPI-U]). T he median is generally considered to
be a more informative measure of retirement assets.
Congressional Research Service
5

Saving for Retirement: Household Decisionmaking and Policy Options

Figure 1. Percentage of Households with Retirement Assets and Median Retirement
Assets, 1989-2016

Source: Congressional Research Service analysis of the Survey of Consumer Finances (1989-2016).
Notes: Retirement assets include account balances in defined contribution plans, Individual Retirement
Accounts, and tax-advantaged accounts. Estimates of future income from defined benefit pensions or from Social
Security are not included in the calculations of retirement assets. Median retirement assets are for households
with retirement assets. Values have been adjusted to 2016 dol ars based on the Consumer Price Index for All
Urban Consumers.
Annual contributions to DC plans are general y much larger than for IRAs. In 2016, about 38.8%
of working taxpayers made contributions to DC plans, with an average annual amount of $5,204
among those who chose to contribute;30 whereas only 3.0% of taxpayers made contributions to
traditional IRAs and 4.5% to Roth IRAs, with an average annual contribution of $4,184 and
$3,441, respectively.31 The annual contribution limit for DC plans is general y higher than that for
IRAs. In 2020, the annual contribution limit for 401(k)s is $19,500 (with a $6,500 catch-up
contribution for those aged 50 and older), and the combined limit for traditional IRAs or Roth
IRAs is $6,000 (with an additional $1,000 catch-up contribution for those aged 50 and older).
However, because a large number of DC plan holders roll over their accounts into an IRA at job
separation, the overal assets in IRAs are larger than in DC plans ($11.0 tril ion compared to $8.9
tril ion at the end of 2019).32
Retirement Income and Saving Adequacy
Whether people have adequate retirement savings to achieve financial security in retirement has
been the subject of debate and research over the last few decades. Retirement researchers and
financial planners general y agree that there are two primary ways to assess income adequacy in
retirement—whether income can meet basic needs (such as being above the poverty line) and

30 IRS, SOI Division, “ SOI T ax Stats – Individual Information Return Form W2 Statistics,” at https://www.irs.gov/
statistics/soi-tax-stats-individual-information-return-form-w2-statistics.
31 IRS, SOI Division, “ SOI T ax Stats – Accumulation and Distribution of Individual Retirement Arrangements (IRA),”
at https://www.irs.gov/statistics/soi-tax-stats-accumulation-and-distribution-of-individual-retirement-
arrangements#tables.
32 ICI, DC Plan Participants’ Activities, First Quarter, 2020.
Congressional Research Service
6

link to page 7 Saving for Retirement: Household Decisionmaking and Policy Options

whether income is sufficient to maintain a preretirement standard of living.33 Programs such as
Supplemental Security Income (SSI) and other public assistance are designed to ensure that
eligible individuals have a minimum level of resources.34 In contrast, programs such as Social
Security and DB or DC plans are designed to help workers accumulate enough resources to
maintain their preretirement standard of living. Social Security benefits are calculated to replace a
greater share of career-average earnings for low-paid workers than for high-paid workers.35
Therefore, Social Security benefits general y comprise a much higher share of retirement income
for low earners than for high earners who have substantial amounts of other assets, mainly DB or
DC retirement plans or private pensions.36
Some studies suggest that more than half of U.S. households are on track to be adequately
financial y prepared for retirement.37 These households may include employees with high income
and those who have DB pensions or participated in employer-sponsored DC plans, as wel as
those who are diligently funding their retirement plans and taking advantage of employer matches
throughout their career. Retirement savings policies for such families may focus on protecting
their investment wealth by regulating investment disclosure and management, facilitating
investment choices, and regulating to ensure appropriate investment advice and distribution
information.
For lower-income households, SSI and Social Security may provide most of the retirement
resources a family has during retirement. One study using administrative-linked survey data
shows that, on average, for elderly families in the lowest 20% of the income distribution, about
90% of their retirement income is from Social Security and SSI, and less than 1% is from DC
plans or IRAs.38 This may be because some low-income families do not have income to save after
meeting regular daily expenses during their working years. Another study shows that it may be
optimal for very low earners not to engage in retirement savings because Social Security replaces
a sufficient portion of their preretirement earnings.39 Policy changes in Social Security or SSI
might better target these families to ensure they are able to meet basic needs in retirement.40
People who are not sufficiently preparing for retirement may include certain middle-income
families with little or no retirement savings beyond Social Security. Some of those families may
find saving for retirement difficult for financial reasons; others may not be eligible to participate
in an employer-sponsored retirement saving plan,41 making saving for retirement more

33 U.S. Congressional Budget Office, Measuring the Adequacy of Retirement Income: A Primer, October 2017, at
https://www.cbo.gov/system/files/115th-congress-2017-2018/reports/53191-retirementadequacy.pdf.
34 Other public assistance mainly includes housing subsidies, the Supplemental Nutrition Assistance Program,
refundable tax credits, and the Low Income Home Energy Assistance Program. For more information about
Supplemental Security Income (SSI), see CRS In Focus IF10482, Supplem ental Security Incom e (SSI), by William R.
Morton; and for information regarding how different resources affect the poverty rate among elderly Americans, see
CRS Report R45791, Poverty Am ong Am ericans Aged 65 and Older, by Zhe Li and Joseph Dalaker.
35 For more information on Social Security, see CRS Report R42035, Social Security Primer, by Barry F. Huston.
36 See ICI, “Who Participates in Retirement Plans, 2016,” ICI Research Perspective, vol. 25, no. 6 (August 2019), at
https://www.ici.org/pdf/per25-06.pdf.
37 For summaries of retirement adequacy studies, see Bajtelsmit and Rappaport, 2018, pp. 71 -86.
38 Adam Bee and Joshua Mitchell, Do Older Americans Have More Income Than We Think?, Census Bureau, SESHD,
Working Paper no. 2017-39, July 2017, at https://www.census.gov/content/dam/Census/library/working-papers/2017/
demo/SEHSD-WP2017-39.pdf.
39 Jason S. Scott et al., Can Low Retirement Savings be Rationalized?, National Bureau of Economic Research
(NBER), Working Paper no. 26784, February 20 20.
40 See CRS Report R45791, Poverty Among Americans Aged 65 and Older, by Zhe Li and Joseph Dalaker.
41 IRAs are available to any worker with taxable wages. See footnote 22.
Congressional Research Service
7

Saving for Retirement: Household Decisionmaking and Policy Options

chal enging. Families who may have difficulty maintaining their preretirement living standards in
retirement could benefit from policies to increase participation and contributions in retirement
plans.
Other Influences on Households’ Finances That Could Impact
Saving for Retirement
Families often have many financial goals during their working years, in addition to preparing for
retirement.42 For example, they may want to pay down debt and save for emergencies, home
purchases, a child’s education, or other large expenses. People also often invest in their education
or in a smal business to increase their earnings potential over time. These financial decisions can
affect financial wel -being both now and in the future.
This section of the report discusses three notable areas of many households’ finances which could
make saving for retirement more chal enging: household debt, a lack of emergency savings, and
employment and health uncertainties among older workers.
Household Debt
In addition to retirement savings, other parts of a household’s finances may affect their income
adequacy in retirement. For example, a family with a substantial amount of outstanding mortgage
debt near or in retirement may need to use retirement savings to pay mortgage expenses, thus
resulting in insufficient income to meet regular expenses or health-related costs. Data suggest that
more households are holding debt into retirement in recent years, particularly mortgage debt.43
Elderly people with large outstanding mortgages or other loans may need to work longer or face
financial insecurity during their retirement years.
Emergency Savings
A lack of emergency savings may be another potential risk to retirement planning. U.S.
households on average tend to have relatively low levels of liquid wealth, such as money in a
savings account, and a relatively high incidence of credit card borrowing.44 In 2017, 42% of
households did not set aside any money that year for emergency expenses.45 Therefore, a sizable
portion of the adult population reports they would have difficulty meeting an unexpected expense.
If faced with a $400 unexpected expense, 39% of adults say they would borrow, sel something,
or not be able to cover the expense.46
People with little emergency savings may borrow or withdraw from retirement savings accounts
to fund emergencies, thus reducing the retirement funds available to grow for the future. As

42 For more information about consumer and household finance, see CRS Report R45813, An Overview of Consumer
Finance and Policy Issues, by Cheryl R. Cooper.
43 See CRS Report R45911, Household Debt Among Older Americans, 1989-2016, by Zhe Li.
44 John Beshears et al., Behavioral Household Finance, NBER, Working Paper no. 24854, July 2018, p. 4 (hereinafter
Beshears et al., 2018).
45 Gerald Apaam et al., FDIC National Survey of Unbanked and Underbanked Households, October 2018, p. 43, at
https://www.fdic.gov/householdsurvey/2017/2017report.pdf.
46 Federal Reserve, Report on the Economic Well-Being of U.S. Households in 2018, May 2019, p. 2, at
https://www.federalreserve.gov/publications/files/2018-report -economic-well-being-us-households-201905.pdf.
Congressional Research Service
8

link to page 26 Saving for Retirement: Household Decisionmaking and Policy Options

described in more detail later in this report,47 one source for borrowing funds may be 401(k)
loans.48 These loans al ow individuals to borrow funds from their 401(k) accounts, but consumers
may face the risk of defaulting on the loan. Data suggest that close to 37% of active plan
participants borrow from their 401(k) in a five-year period.49 Early withdrawals from retirement
accounts, including cash-outs at job separation, are also common in the United States. Despite the
higher tax penalties from early retirement account withdrawals, research estimates at least 26% of
households take money from their accounts for nonretirement spending needs.50 Household
events, such as unexpected reductions in income or marital changes, seem to increase the
likelihood of early withdrawals.51
Employment and Health Uncertainties Among Older Workers
Other risks in retirement planning are future uncertainties, such as unexpected job loss or health
conditions, which can reduce earnings and thus retirement wealth accumulation. Over the
working years of a family, the decade or two right before retirement can be critical. Earlier in
one’s working years, many families are focused on other financial goals, such as paying off
student loan debt, owning a home, and maintaining expenses associated with raising children.
Therefore, later working years are often when people focus more on retirement planning, perhaps
catching up with savings or paying off a home. Yet workers who are 50 years old or older may
face increased uncertainty about their work future during this critical time.
Layoffs and il nesses in the family are common for those over age 50. One study estimates that
66% of older workers experience an unexpected job loss or health condition hindering
employment before turning 65 years old.52 For those who lose their jobs, older workers are likely
to experience larger wage losses than their younger counterparts upon reemployment.53 In
addition, research suggests that people with health issues may overestimate how long they can
work, and often their health worsens before the age at which they plan to retire.54 Evidence
suggests that financial losses from losing a job or a health condition impeding work can impact

47 For more details, see the “ Hardship Withdrawals and Loans in Retirement Plans” section of this report.
48 In 2018, more than 58 million American workers were active 401(k) participants. DC plans may (but are not required
to) allow individuals to withdraw funds for a financial difficulty, referred to as a hardship distribution. For more
information, see CRS In Focus IF11472, Withdrawals and Loans from Retirem ent Accounts for COVID-19 Expenses,
by John J. T opoleski and Elizabeth A. Myers. For the number of 401(k) participants, see “ Frequently Asked Questions
About 401(k) Plan Research,” April 2020, at https://www.ici.org/policy/retirement/plan/401k/faqs_401k.
49 NBER, “ Borrowing from 401(k)s,” Bulletin on Aging and Health, no. 2 (2015), at https://www.nber.org/aginghealth/
2015no2/w21102.html.
50 Matt Fellowes and Katy Willemin, The Retirement Breach in Defined Contribution Plans: Size, Causes, and
Solutions, HelloWallet, January 2013, at https://www.scribd.com/document/222463705/Retirement-Breach-In-Defined-
Contribution-Plans-from-HelloWallet.
51 Robert Argento, Victoria Bryant, and John Sabelhaus, “Early Withdrawals from Retirement Accounts during the
Great Recession,” Contemporary Economic Policy, vol. 33, no. 1 (January 2015), pp. 14-18.
52 Richard Johnson and Peter Gosselin, How Secure is Employment at Older Ages?, Urban Institute, December 2018,
pp. 7-8, at https://www.urban.org/sites/default/files/publication/99570/
how_secure_is_employment_at_older_ages_2.pdf (hereinafter Johnson and Gosselin, 2018).
53 Richard W. Johnson and Corina Mommaerts, Age Differences in Job Loss, Job Search, and Reemployment, Urban
Institute, January 2011, at https://www.urban.org/sites/default/files/publication/27086/412284-Age-Differences-in-Job-
Loss-Job-Search-and-Reemployment.PDF.
54 Alicia H. Munnell, Matthew S. Rutledge, and Geoffrey T . Sanzenbacher, “Retiring Earlier than Planned: What
Matters Most?,” Center for Retirem ent Research at Boston College, February 2019, Working Paper no. 19-3, at
https://crr.bc.edu/wp-content/uploads/2019/01/IB_19-3.pdf.
Congressional Research Service
9

Saving for Retirement: Household Decisionmaking and Policy Options

households for years after the initial employment loss.55 These disruptions in earnings can impact
future retirement security, potential y reducing a family’s ability to save and reducing future
Social Security benefits. These employment and health disruptions can also lead people to tap
into existing retirement savings to make ends meet between jobs or to retire earlier than expected.
For example, several studies have found that negative health shocks among older workers are
likely to lead to early retirement.56
Household Decisionmaking in Retirement Savings
As discussed earlier in this report, household decisionmaking related to retirement has become
more important over time. The shift from DB to DC retirement plans requires families to assume
more responsibility for managing their retirement and making decisions about retirement account
contributions and investments, as wel as making decisions about how to draw down these funds
in retirement. For this reason, understanding household decisionmaking in retirement planning is
important, particularly when considering retirement savings policy issues and the impact of
different policy options on retirement security.
Economic Decisionmaking and the Life-Cycle Model
The life-cycle model is a prevalent economic hypothesis that assumes households usual y want to
keep consumption levels stable over time.57 For example, severely reducing consumption one
month may be more painful for people than the pleasure of a much higher household
consumption level in another month. Therefore, people save and invest during their careers in
order to afford a stable income across their lives, including in retirement. This model suggests
that wealth should increase as people age, which general y fits household financial data in the
United States.58 In this theory, households adjust their savings rate during their working years
rational y, based on interest rates, investment returns, life expectancy, Social Security or pension
benefits, and other relevant factors. Evidence exists that some households adjust their retirement
planning based on these types of factors.59 However, in the United States, income and

55 Johnson and Gosselin, 2018, p. 15.
56 See Debra S. Dwyer and Jianting Hu, “Retirement Expectations and Realizations: T he Role of Health Shocks and
Economic Factors,” in Forecasting Retirem ent Needs and Retirem ent Wealth , eds. Olivia S. Mitchell, P. Brett
Hammond, and Anna M. Rappaport (Philadelphia, PA: University of Pennsylvania Press, 2000), pp. 274-287; Ronald
Hagan, Andrew M. Jones, and Nigel Rice, “ Health and Retirement in Europe,” International Journal of Environm ental
Research and Public Health, vol. 6, no. 10 (October 2009), pp. 2676 -2695; and Richard Disney, Carl Emmerson, and
Matthew Wakefield, “ Ill Health and Retirement in Britain: A Panel Data-Based Analysis,” Journal of Health
Econom ics, vol. 25, no. 4 (July 2006), pp. 621-649. One study also finds that American men and African Americans are
more likely than other groups to exit labor market due to health shocks. See Dalton Conley and Jason T hompson, “ The
Effects of Health and Wealth Shocks on Retirement Decisions,” Federal Reserve Bank of St. Louis Review, vol. 95, no.
5 (September/October 2013), pp. 389-404.
57 Franco Modigliani, “T he Life Cycle Hypothesis of Saving, the Demand for Wealth and the Supply of Capital,”
Social Research, vol. 33, no. 2 (Summer 1966), pp. 160 -217.
58 Jesse Bricker et al., “ Changes in U.S. Family Finances from 2013 to 2016: Evidence from the Survey of Consumer
Finances,” Federal Reserve Bulletin, vol. 103, no. 3 (September 2017), p. 13, at https://www.federalreserve.gov/
publications/files/scf17.pdf.
59 For example, many research papers find some households save more for retirement when pension or social security
benefits decrease. See Orazio P. Attanasio and Agar Brugiavini, “Social Security and Households’ Saving,” The
Quarterly Journal of Econom ics, vol. 118, no. 3 (August 2003), pp. 1075 -1119; Gary V. Engelhardt and Anil Kumar,
“Pensions and Household Wealth Accumulation,” The Journal of Human Resources, vol. 46, no. 1 (January 2011), pp.
203-236; and Orazio P. Attanasio and Susann Rohwedder, “ Pension Wealth and Household Saving: Evidence from
Pension Reforms in the United Kingdom,” The Am erican Econom ic Review, vol. 93, no. 5 (December 2003), pp. 1499 -
Congressional Research Service
10

Saving for Retirement: Household Decisionmaking and Policy Options

consumption move together more closely than the life-cycle model would predict, suggesting
some households may not save enough for their retirement needs or other lower-income periods.60
Mainstream economic theory asserts that competitive free markets general y lead to efficient
distributions of goods and services to maximize value for society.61 If certain conditions hold,
policy interventions cannot improve on the financial decisions that consumers make based on
their unique situations and preferences. For this reason, some policymakers are hesitant to disrupt
free markets, based on the theory that prices determined by market forces lead to efficient
outcomes without intervention. However, in these theoretical frameworks, a free market may
become inefficient due to departures from standard economic assumptions, which includes
assuming that consumers and firms act rational y with perfect information. When these
assumptions do not hold, it may cause a reduction in economic efficiency and consumer welfare.
In these cases, government policy can potential y bring the market to a more efficient outcome,
maximizing social welfare. Yet, policymakers often find it chal enging to determine whether a
policy intervention wil help or harm a particular market to reach its efficient outcome.
The following section discusses behavioral biases, which are a specific departure from the
rational decisionmaking condition associated with theoretical economic efficiency. This departure
is particularly important for understanding people’s decisionmaking in saving for retirement and
investment markets. When people act with predictable biases, markets may become less efficient,
and government policy—such as consumer disclosures or other plan design requirements—may
be appropriate. However, these policies may also lead to unintended outcomes, which should be
taken into account.
Behavioral Research and Retirement Savings Decisionmaking
Behavioral research suggests that people tend to have biases in rather predictable patterns.62 This
research suggests that the human brain has evolved to quickly make judgments in bounded,
rational ways, using heuristics—or mental shortcuts—to make decisions. These heuristics
general y help people make appropriate decisions quickly and easily, but they can sometimes
result in choices that make the decisionmaker worse off financial y. For example, the number,
order, and structure of options, as wel as the process around the choice, can change decisions for
many people.
A few of these biases tend to be particularly important for understanding retirement planning
decisionmaking:63
Choice Architecture. Research suggests that how financial decisions are framed can affect
consumer decisionmaking. Framing can affect decisions in many ways.

1521.
60 Beshears et al., 2018, p. 4.
61 N. Gregory Mankiw, “Chapter 7,” in Principles of Microeconomics, 7th ed. (South-Western College Pub, 2014).
62 Daniel Kahneman, Thinking Fast and Slow (New York: Farrar, Straus and Giroux, 2011); and Dan Ariely,
Predictably Irrational: The Hidden Forces that Shape our Decisions (New York: HarperCollins Publishers, 2008).
63 For a more comprehensive overview of behavioral research related to retirement saving decisions, see Melissa Knoll,
“T he Role of Behavioral Economics and Behavioral Decision Making in Americans’ Retirement Savings Decisions,”
Social Security, Office of Retirem ent and Disability Policy, Social S ecurity Bulletin, vol. 70, no. 4 (2010), at
https://www.ssa.gov/policy/docs/ssb/v70n4/v70n4p1.html.
Congressional Research Service
11

Saving for Retirement: Household Decisionmaking and Policy Options

 Anchoring. People can be influenced, or anchored, by an initial number, even if
it is unrelated to their next choice.64 In one il ustration of this concept,
researchers had subjects spin a wheel of fortune with numbers between 0 and
100, then asked them the percentage of African countries in the United Nations.
The random number generated in the first stage subconsciously affected subjects’
guesses in the second stage, even though they were not related. Therefore,
without the anchor, people’s estimates likely would have been different. In the
retirement savings context, the automatic contribution rate in 401(k)s and the
percent of salary at which employers provide maximum matches may be anchors
that influence how much a person decides to put toward retirement savings.65
 Defaults. People can also be influenced by defaults established in how a decision
is framed.66 For example, employees are more likely to be enrolled in a 401(k)
plan if an employer defaults them into it than if they actively need to make a
choice to participate.67
 Choice Overload. When making decisions, people often find it difficult to
navigate complexity, such as many choices to choose from or items to consider.
In the retirement savings context, this means that more investment fund options
in retirement savings plans can sometimes lead to procrastination or failure to
make a decision.68 Choice overload can also lead to poor decisionmaking, as
some research suggests that fewer choices in retirement savings plans might lead
to better retirement investment decisions. 69
 Asset Allocation and Diversification. People tend to naively make
diversification choices when making al ocation decisions. For example, in the
retirement context, when making decisions about how much to invest in a
collection of funds, some people choose to spread their investments evenly across
available funds (whether financial y appropriate for their situation or not).70
Biases Toward the Future. Research suggests that common cognitive biases towards the future
can also affect consumer decisionmaking.
 Present Bias. When people tend to put more value on having something now,
rather than in the future—even when there is a large benefit for waiting—this

64 Amos T versky and Daniel Kahneman, “Judgment under Uncertainty: Heuristics and Biases,” Science, vol. 185, no.
4157 (September 27, 1974), pp. 1124 -1131.
65 John Beshears et al., “T he Importance of Default Options for Retirement Savings Outcomes: Evidence from the
United States,” in Lessons from Pension Reform in the Americas, ed. Stephen J. Kay and T apen Sinha (New York:
Oxford University Press, 2008), pp. 59 -87 (hereinafter Beshears et al., 2008).
66 Richard T haler and Cass Sunstein, Nudge: Improving Decisions about Health, Wealth, and Happiness (New York:
Penguin Books [USA] Inc., 2008).
67 Brigitte C. Madrian and Dennis F. Shea, “T he Power of Suggestion: Inertia in 401(k) Participation and Savings
Behavior,” Quarterly Journal of Economics, vol. 116, no. 4 (November 2001), pp. 1149-1187 (hereinafter Madrian and
Shea, 2001); and Robert L. Clark and Denis Pelletier, Im pact of Defaults in Retirem ent Saving Plans: Public Em ployee
Plans, NBER, Working Paper no. 2623, September 2019.
68 Iyengar, Jiang, and Huberman, 2003.
69 Donald B. Keim and Olivia S. Mitchell, Simplifying Choices in Defined Contribution Retirement Plan Design,
NBER, Working Paper no. 21854, January 2016.
70 Shlomo Benartzi and Richard H. T haler, “Naive Diversification Strategies in Defined Contribution Saving Plans,”
Am erican Econom ic Review, vol. 91, no. 1 (March 2001).
Congressional Research Service
12

Saving for Retirement: Household Decisionmaking and Policy Options

behavior is cal ed present bias.71 For example, in the retirement context, people
tend to have a preference for lump sums over annuities, independent of risk
considerations. Research suggests that people with more present bias tend to save
less for retirement when controlling for other factors.72
 Self-Control. Even when people decide they should do something, such as
saving for the future or choosing a retirement plan, self-control and
procrastination may prevent them from following their intentions.73 These human
biases might lead consumers to make financial decisions that are not optimal,
such as undersaving.74
Although consumers might not be aware of these biases when making financial decisions, firms
may take advantage of them to attract consumers. For example, choice architecture biases might
influence how marketing materials are developed, emphasizing certain terms—such as high past
investment return rate—to make a financial product seem more desirable to consumers. In
addition, product features may be developed to take advantage of people’s present bias or self-
control mistakes. Less knowledgeable retirement savers’ decisionmaking might be more sensitive
to choice architecture biases.75
Biases can also be used to encourage people to save more for retirement and make better
retirement decisions. For example, some research suggests that choice architecture environments
can make retirement more salient (e.g., annual consumer disclosures that project future retirement
income may lead to more retirement savings).76 Moreover, how saving and investment options are
framed may help some people make better retirement decisions. For example, some research
suggests that preference checklists, which list factors—such as perceived health, life expectancy,
and risk of outliving one’s resources—that people should consider when making a retirement
decision, may improve retirement decisionmaking.77
Although these techniques can be used to encourage social y beneficial goals, such as planning
and saving more for retirement, changing the choice environment can also sometimes have
perverse impacts. For example, defaulting people at a fixed savings rate can increase participation
in retirement plans on average but may discourage some people from making an active decision
when they start a new job to increase the contribution rate from the default to a higher level. For
these people, the lower contribution rate may lead to less retirement savings over time. Likewise,
defaulting people into life-cycle retirement investment plans may lead to more appropriate long-

71 Richard T haler, “Some Empirical Evidence on Dynamic Inconsistency,” Economic Letters, vol. 8 (1981), pp. 201-
207.
72 Gopi Shah Goda et al., The Role of Time Preferences and Exponential-Growth Bias in Retirement Savings, NBER,
Working Paper no. 21482, August 2015.
73 Kungl. Vetenskapsakademien: T he Royal Swedish Academy of Sciences, Richard H. Thaler: Integrating Economics
with Psychology, Scientific Background on the Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred
Nobel 2017, October 3, 2017, pp. 10-14, at https://www.nobelprize.org/uploads/2018/06/advanced-
economicsciences2017-1.pdf.
74 H.M. Shefrin and Richard T haler, “An Economic T heory of Self-Control,” Journal of Political Economy, vol. 89, no.
2 (April 1981), pp. 392-406.
75 Anders Anderson and David T . Robinson, Who Feels the Nudge? Knowledge, Self-Awareness and Retirement
Savings Decisions, NBER, Working Paper no. 25061, September 2018.
76 Mathias Dolls et al., Do Savings Increase in Response to Salient Information about Retirement and Expected
Pensions?, NBER, Working Paper no. 22684, September 2016.
77 Eric Johnson et al., Preference Checklists: Selective and Effective Choice Architecture for Retirement Decisions,
T IAA Institute, Research Dialogue Issue no. 127, June 2016, at https://www.tiaainstitute.org/sites/default/files/
presentations/2017-02/rd_selective_effective_choice_architecture_for_retirement_decisions_1.pdf .
Congressional Research Service
13

Saving for Retirement: Household Decisionmaking and Policy Options

term investment decisions on average, but the investment default also may encourage fewer
people to make active decisions or put them in a plan that may conflict with other savings
vehicles.78 Moreover, although defaulting people into 401(k)s can increase the number of people
who save for retirement, it may also lead to increased consumer debt79 without large impacts on
household net worth over time.80
Policy Options Related to Household Retirement
Savings
Some Americans are on track for retirement security, but others may not be saving enough or
might fal off track before reaching retirement. Public policy may be able to encourage more
retirement saving for those not saving enough. Moreover, for households on track with saving for
retirement, public policy may be important to help these families protect their investment wealth.
Taking into consideration the behavioral aspects around plan design discussed above, this section
discusses policy options that may enhance retirement savings among certain population
subgroups. Evaluating how each policy option affects retirement planning and consumer
decisionmaking is important for understanding the policy’s ultimate impact on household
retirement security.
Availability, Automatic Enrollment, and Automatic Escalation
Three common policy tools to encourage more retirement savings are (1) increasing the
availability of retirement accounts, (2) automatic enrollment into retirement accounts when
available, and (3) automatic escalation of contributions.
Availability
Because of the growing importance of retirement savings, many researchers and policymakers
believe U.S. workers general y should have easy access to a payroll deduction-based retirement
savings plan. In March 2019, about 64% of private-sector workers had access to DC retirement
plans.81 IRAs are available to al earners, but unlike a DC plan, they general y do not link
automatical y with a payroll deduction. Two approaches have been developed to expand the

78 Gopi Shah Goda et al., “Do Defaults have Spillover Effects? T he Effect of the Default Asset on Retirement P lan
Contributions,” paper presented at the 20th Annual Joint Meeting of the Retirement Research Consortium, Washington,
DC, August 2018, at http://www.nber.org/2018rrc/summaries/1.3%20-
%20Goda,%20Levy,%20Manchester,%20Sojourner,%20Tasoff.pdf .
79 T his research study found that when the U.S. Army moved to automatic enrollment in the T SP at a 3% default
contribution rate, both the number of employees contributing and the average contribution rate increased; but the move
also significantly increased auto loan and mortgage balances. See John Beshears et al., Borrowing to Save? The Im pact
of Autom atic Enrollm ent on Debt, NBER, Working Paper no. 25876, May 2019 (hereinafter Beshears et al., 2019).
80 Using data from 34 U.S. 401(k) plans, this research study estimated that, on average, although automatic enrollment
into 401(k) plans increases savings in the plan in the short run, employees tend to r espond by saving less in the future;
so the long-term impact of automatic enrollment on retirement savings is not significant on average and significant only
for the lowest lifetime earnings groups. See T aha Choukhmane, “Default Options and Retirement Saving Dynamics,”
June 10, 2019, at https://cepr.org/sites/default/files/Choukhmane%20%282019%29%20-%20Default%20Options.pdf
(hereinafter Choukhmane, 2019).
81 NCS, “T able 2. Retirement Benefits: Private Industry Workers,” 2019.
Congressional Research Service
14

Saving for Retirement: Household Decisionmaking and Policy Options

availability of retirement savings plans: (1) extending DC plans to more employees and (2)
establishing mandatory employer-provided payroll deduction-based IRAs.
Part-time workers and workers in smal businesses are less likely to have access to employer-
sponsored DC plans than other workers. About 73% of full-time private-sector workers had
access to DC plans in 2019, compared with 35% of part-time workers. Additional y, about 82% of
workers in companies with 500 or more employees had access to DC plans, compared with 48%
of those in companies with 49 or fewer employees.82 In order to increase DC plan participation
for part-time and smal -business workers, the SECURE Act requires employers to offer DC plans
to long-term part-time employees who work more than 500 hours per year over three consecutive
years, while increasing the startup retirement plan tax credit for smal businesses from $500 to
$5,000 per year for three years.83
Another approach is to make payroll deductions to IRAs mandatory for employers who do not
offer a retirement plan.84 Some states have passed legislation to implement auto-IRAs, which
require employers who do not offer a retirement plan to automatical y enroll their workers in an
IRA-based savings program sponsored by the state.85 In most cases, workers can opt out. S. 2370,
the Automatic IRA Act of 2019, proposes a nationwide auto-IRA for workers of employers who
do not offer a retirement plan; H.R. 2120 and S. 1053 introduced in the 116th Congress, the
Saving for the Future Act, would set up a universal personal savings account for workers who do
not have an employer-sponsored DC or IRA-type plan. The Obama Administration also proposed
a similar universal auto-IRA program.86
Automatic Enrollment
Accessibility to payroll deduction-based DC or IRA plans may not guarantee consumer
participation. In some retirement savings plans, participants must make an active decision to
enroll, choose a percentage of their income to save, and choose how to invest these funds. In
March 2019, among the private-sector workers who had access to DC plans, 74% of them were
enrolled in a payroll deduction-based plan.87
Automatic enrollment—in which employees are automatical y signed up unless they opt out—can
be successful in overcoming consumer biases that can impede signing up, leading to much higher

82 NCS, “T able 2. Retirement Benefits: Private Industry Workers,” 2019.
83 T his new rule goes into effect beginning in 2021. Employers will need to make necessary modifications to their
administrative systems by no later than January 1, 2021, to account for the potential future eligibility of those long-term
part-time employees. Because of the long-term requirement of three consecutive years, the mandated participation of
those employees under this provision will be delayed until 2024.
84 Some differences between a 401(k) plan and an IRA include the following: 401(k) plans are covered by the
Employee Retirement Income Security Act of 1974 (ERISA; P.L. 93-406), but IRAs are generally exempt; establishing
and administering 401(k)s may be more complicated than IRAs; employer contributions can be available in 401(k)s but
not in IRAs; loans subject to certain limits are available in 401(k)s but not in IRAs; and contribution limits to IRAs are
lower than to 401(k)s.
85 California, Connecticut, Illinois, Maryland, New Jersey, and Oregon, plus Seattle, WA, have created similar state -run
auto-IRA programs. See Georgetown University McCourt School of Public Policy, Center for Retirement Initiatives,
“State-Facilitated Retirement Savings Programs: A Snapshot of Program Design Features,” State Brief 20 -02, February
29, 2020, at https://cri.georgetown.edu/wp-content/uploads/2018/12/CRI-State-Brief-States_SnapShotPlanDesign-Feb-
20-2020.pdf.
86 White House, Office of the Press Secretary, “Fact Sheet: Opportunity for All: Securing a Dignified Retirement for all
Americans,” January 29, 2014, at https://obamawhitehouse.archives.gov/the-press-office/2014/01/29/fact-sheet-
opportunity-all-securing-dignified-retirement -all-americans.
87 NCS, “T able 2. Retirement Benefits: Private Industry Workers,” 2019.
Congressional Research Service
15

Saving for Retirement: Household Decisionmaking and Policy Options

enrollment in DC plans.88 The Pension Protection Act of 2006 (PPA; P.L. 109-280) included a
provision to encourage firms to adopt automatic enrollment. After PPA’s enactment, evidence
suggests that the percentage of U.S. employers automatical y enrolling employees into 401(k)
plans has increased considerably. In 2005, before the PPA’s enactment, about 19% of employers
with 401(k) plans automatical y enrolled their employees,89 compared with 58% in 2009 and 61%
in 2017.90 Another study, using restricted microdata from the National Compensation Survey,
found that the share of workers with automatic enrollment 401(k) plans increased from 3.9% in
2002 (before PPA) to 32.3% in 2012 (after PPA).91 To further encourage automatic enrollment
plans, the SECURE Act provides a $500 tax credit to smal businesses that adopt automatic
enrollment. In addition, proposed nationwide auto-IRA programs also would general y include an
automatic enrollment requirement (such as S. 2370, H.R. 2120, and S. 1053 introduced in the
116th Congress).
Research general y finds that IRAs with automatic enrollment boost retirement contributions
among low- and moderate-wage individuals.92 A similar effect is found in the expansion of certain
DC plans.93 However, automatic enrollment in retirement savings plans may increase savings in
retirement accounts by increasing debt or reducing saving for other purposes. One recent study
found that automatic enrollment in retirement accounts may cause increases in auto loans and first
lien mortgage balances.94 Another recent study found that automatic enrollment may not
necessarily have large impacts on household net worth over time.95
Automatic Escalation
About three-quarters of automatic enrollment plans use an initial savings rate of 3% of earnings.96
Research shows that when offered the default rate, many participants wil passively accept it;
however, if they had been forced to choose their savings rate, some would have selected a higher
rate.97 Because participants may passively accept the initial savings rate and are less likely to

88 Shlomo Benartzi and Roger Lewin, Save More Tomorrow: Practical Behavioral Finance Solutions to Improve
401(k) Plans (New York: Penguin, 2012); and Madrian and Shea, 2001, pp. 1149 -1187.
89 Shlomo Benartzi and Richard H. T haler, “Behavioral Economics and the Retirement Savings Crisis,” Science, vol.
339 (March 2013), pp. 1152-1153 (hereinafter Benartzi and T haler, 2013).
90 Plan Sponsor Council of America, 61st Annual Survey, T able 111, 2018.
91 Barbara A. Butrica, Keenan Dworak-Fisher, and Pamela Perun, Pension Plan Structures before and after the Pension
Protection Act of 2006, T he Urban Institute, September 2015.
92 See U.S. Government Accountability Office (GAO), Automatic IRAs: Lower-Earning Households Could Realize
Increases in Retirem ent Incom e, GAO-13-699, August 2013, at https://www.gao.gov/products/GAO-13-699; Barbara
A. Butrica and Richard W. Johnson, How Much Might Autom atic IRAs Im prove Retirem ent Security for Low- and
Moderate-Wage Workers? T he Urban Institute, July 2011, at https://www.urban.org/sites/default/files/publication/
27386/412360-How-Much-Might -Automatic-IRAs-Improve-Retirement-Security-for-Low-and-Moderate-Wage-
Workers-.PDF; and Jack VanDerhei, “ Auto-IRAs: How Much Would T hey Increase the Probability of ‘Successful’
Retirements and Decrease Retirement Deficits? Preliminary Evidence from EBRI’s Retirement Security Projection
Model®,” Employee Benefit Research Institute (EBRI), EBRI Notes, vol. 36, no. 6 (June 2015), at
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2620846.
93 Jack VanDerhei, What if OregonSaves Went National: A Look at the Impact on Retirement Income Adequacy ,
October 2019, EBRI, Issue Brief no. 494, at https://www.ebri.org/docs/default-source/ebri-issue-brief/
ebri_ib_494_oregonsaves-31oct19.pdf?sfvrsn=8bd43c2f_6.
94 Beshears et al., 2019.
95 Choukhmane, 2019.
96 Vanguard Group, How America Saves 2012, 2012.
97 Beshears et al., 2008, pp. 59-87; and Madrian and Shea, 2001, pp. 1149-1187.
Congressional Research Service
16

Saving for Retirement: Household Decisionmaking and Policy Options

consider increasing the rate after enrolling into an account, some researchers have indicated that
these participants may insufficiently save for retirement and have suggested automatic escalation
based on behavioral research.98 A typical approach to automatic escalation is to ask employees to
commit to a future increase in their retirement savings rate, often linked with pay raises, until a
preset limit is reached, or the employee chooses to opt out. One research study estimated that
automatic escalation would substantial y boost annual savings.99 In 2017, about a quarter of DC
plans provided an automatic escalation to some or al plan participants.100 In addition, many
legislative proposals that would require automatic enrollment include an automatic escalation
provision (e.g., H.R. 2120 and S. 1053 introduced in the 116th Congress).
Matching of Retirement Contributions
The vast majority of employer-sponsored retirement savings plans include an employer match to
incentivize employees to save for retirement and to save at higher levels.101 A typical structure of
the employer match provides a dollar-for-dollar match (i.e., 100%) up to a certain threshold (such
as 4% of salary).102 Several recent legislative proposals (e.g., H.R. 2120 and S. 1053 introduced
in the 116th Congress) include a matching provision either from the employer or the federal
government.103
Although most studies suggest that matching contributions could increase the participation and
contributions in saving plans, the impact is often less significant than nonfinancial approaches,
such as automatic enrollment.104 For individuals already enrolled in a retirement plan, matching
has a smal effect on contributions.105 However, the match threshold tends to serve as a strong
anchor when individuals decide how much to save.106 Therefore, some researchers suggest that a
lower match rate with a higher match threshold may be a more effective way to increase
contributions in saving plans.107
In evaluating how matching affects retirement savings among low-income individuals, the
empirical evidence is less decisive. Some researchers find that a higher match increases
contributions for low-income individuals;108 others find that the match-based savings structure

98 Benartzi and T haler, 2013, pp. 1152-1153.
99 Benartzi and T haler, 2013. T he study estimates that automatic escalation could increase annual savings by $7.4
billion if the average annual compensation is assumed to be $60,000.
100 Plan Sponsor Council of America, 61st Annual Survey, 2018, T able 121.
101 Plan Sponsor Council of America, 61st Annual Survey, 2018, T able 42. In 2017, more than 96% of DC-type plans
made some matching contributions.
102 Plan Sponsor Council of America, 61st Annual Survey, 2018, T ables 42 and 47. Based on the survey, in 2017,
employers on average contributed 4% of eligible participants’ annual payroll into 401(k) plans.
103 H.R. 2120 and S. 1053 in the 116th Congress (Saving for the Future Act) would also require minimum employer
contributions in universal personal savings accounts.
104 Brigitte C. Madrian, Matching Contributions and Savings Outcome: A Behavioral Economics Perspective, NBER,
Working Paper no. 18220, July 2012 (hereinafter Madrian, 2012).
105 Madrian, 2012.
106 James J. Choi, David Laibson, and Brigitte C. Madrian, “$100 Bills on the Sidewalk: Violations of No -Arbitrage in
401(k) Accounts,” The Review of Economics and Statistics, vol. 93, no. 3 (2011), pp. 748-763, at
https://dash.harvard.edu/handle/1/9647368 (hereinafter Choi, Laibson, and Madrian, 2011).
107 Choi, Laibson, and Madrian, 2011.
108 Gur Huberman, Sheena S. Iyengar, and Wei Jiang, “Defined Contribution Pension Plans: Determinants of
Participation and Contribution Rates,” Journal of Financial Services Research, vol. 31, no. 1 (2007), pp. 1-32.
Congressional Research Service
17

Saving for Retirement: Household Decisionmaking and Policy Options

has no significant effect on the net worth of very low-income families.109 Given the relatively
higher cost of matching strategies, other policies may be more cost-effective at boosting savings
contributions among low-income families.
Financial Literacy
Consumers general y have to make a variety of financial decisions when planning for retirement,
including which tax-advantaged accounts to use, how much to save each year (in each account),
and which investments to choose. Research suggests that higher financial literacy is strongly
correlated with saving and retirement planning.110 In general, however, financial literacy—the
“knowledge and understanding of financial concepts and choices”111—is relatively low in the
United States.112 Those with low levels of financial literacy often get lower returns in financial
markets (e.g., by shunning the stock market or choosing investment options with higher fees).113
Moreover, a lack of emergency savings, which may be due to inadequate financial literacy or
financial management, may lead to retirement fund leakage and hardship loans. Due to these
reasons, in part, research suggests that financial literacy may have an impact on household wealth
accumulation, independent from educational attainment.114 Also for these reasons, some
policymakers are interested in increasing Americans’ financial literacy related to retirement
savings and planning.
The Financial Literacy and Education Commission (FLEC) coordinates financial literacy and
education efforts across relevant federal government agencies.115 Within agencies that participate
in the FLEC, the Department of Labor’s Employee Benefits Security Administration
(DOL/EBSA) is the primary agency educating the public on workplace retirement savings plans,
and the Securities and Exchange Commission (SEC) is the primary agency responsible for

109 Gregory Mills et al., “Effects of Individual Development Accounts on Asset Purchases and Saving Behavior:
Evidence from a Controlled Experiment,” Journal of Public Economics, vol. 92, nos. 5-6 (2008), pp. 1509-1530.
110 Lusardi and Mitchell, 2014, pp. 22, 25-26.
111 Consumer Financial Protection Bureau, Financial Literacy Annual Report, December 2018, p. 5, at
https://files.consumerfinance.gov/f/documents/bcfp_financial-literacy_annual-report_2018.pdf.
112 Lusardi and Mitchell, 2014, p. 34. Financial literacy related to investment decisions is often measured by standard
questions related to compound interest, inflation, and risk diversification:
1) Suppose you had $100 in a savings account and the interest rate was 2% per year. After 5 years,
how much do you think you would have in the account if you left the money to grow: [more than
$102, exactly $102, less than $102? Do not know, refuse to answer.]
2) Imagine that the interest rate on your savings account was 1% per year and inflation was 2% per
year. After 1 year, would you be able to buy: [more t han, exactly the same as, or less than today
with the money in this account? Do not know; refuse to answer.]
3) Do you think that the following statement is true or false? “ Buying a single company stock
usually provides a safer return than a stock mutual fund.” [T rue, False, Do not know; refuse to
answer.]
For more information, see Lusardi and Mitchell, 2014, p. 9.
113 Lusardi and Mitchell, 2014, p. 24.
114 T he relationship between financial literacy and wealth accumulation is not due to educational attainmen t generally.
When education level is controlled for, financial literacy still has a significant impact on wealth accumulation. See
Lusardi and Mitchell, 2014, pp. 23-24.
115 Recently, the T reasury Department released a report with recommended actions to impr ove the operations of the
Financial Literacy and Education Commission (FLEC). See U.S. Department of the T reasury, Federal Financial
Literacy Reform : Coordinating and Im proving Financial Literacy Efforts, July 2019, at https://home.treasury.gov/
system/files/136/FFLRCoordinatingImprovingFinancialLiteracyEfforts.pdf (hereinafter T reasury, Federal Financial
Literacy Reform , 2019).
Congressional Research Service
18

link to page 14 Saving for Retirement: Household Decisionmaking and Policy Options

investment information.116 The government’s goals relating to financial literacy include
supporting the development of core financial competencies in consumers through access to
effective information, financial education resources, and programs so consumers can make
informed financial decisions.117 Congress continues to show interest in improving financial
literacy around retirement planning.118 For example, S. 975 and H.R. 2005 introduced in the 116th
Congress, the Women’s Retirement Protection Act, include financial literacy education grants for
retirement planning programs and a link to the Consumer Financial Protection Bureau’s website
with retirement planning information on mandated consumer disclosures.
Information Disclosure and Reporting
As described in the retirement plan design section of this report, choice architecture, or the way
financial decisions are framed, can affect consumer decisionmaking.119 For this reason, retirement
account disclosures and reporting to consumers are important, because they can influence
consumer decisionmaking, impacting retirement security in the future. Given low financial
literacy, retirement plan disclosures may suggest to uninformed consumers what information is
important or whether most people make similar decisions. For this reason, the federal government
mandates particular information related to retirement accounts to be disclosed regularly, such as
the value of investments, fees and expenses, and other retirement plan terms.120
Congress continues to show interest in improving retirement account disclosures and reporting
requirements to help consumers make better retirement decisions, such as including design
features to better al ow consumers to compare investments, fees, and other important retirement
decisions. For example, the SECURE Act requires plans to provide participants with a lifetime
retirement income projection based on their current DC plan account balance to make it easier for
people to determine whether they are saving enough for retirement. In addition, S. 1431
introduced in the 116th Congress would consolidate retirement plan disclosures and benchmark
disclosures for investment funds to make these disclosures more usable and facilitate easier
comparisons of investment products. Lastly, S. 975 and H.R. 2005 introduced in the 116th
Congress (the Women’s Retirement Protection Act) would create additional targeted notices (e.g.,
additional spousal disclosures and consent requirements so people are aware of retirement
decisions that may have perverse or unintended consequences for their spouse’s retirement
security, such as beneficiary or retirement plan changes).

116 For more information on the major agency efforts around retirement planning and financial literacy, see T reasury,
Federal Financial Literacy Reform , 2019, pp. 30-37.
117 FLEC, Promoting Financial Success in the United States: National Strategy for Financial Literacy, 2011, pp. 8-9, at
https://www.treasury.gov/resource-center/financial-education/Documents/NationalStrategyBook_12310%20(2).pdf;
and FLEC, Prom oting Financial Success in the United States: National Strategy for Financial Literacy 2016 Update ,
2016, at https://www.treasury.gov/resource-center/financial-education/Documents/
National%20Strategy%202016%20Update.pdf.
118 For example, see U.S. Congress, Senate Committee on Health, Education, Labor, and Pensions, Subcommittee on
Primary Health and Retirement Security, Financial Literacy: the Starting Point for a Secure Retirem ent, 115th Cong.,
2nd sess., August 21, 2018.
119 For more detail, see the “ Behavioral Research and Retirement Savings Decisionmaking” section of this report.
120 For example, a guide from the Department of Labor (DOL) on reporting and disclosure requirements for employee
benefit plans can be found at DOL, Employee Benefits Security Administration (EBSA) , Reporting and Disclosure
Guide for Em ployee Benefit Plans, September 2017, at https://www.dol.gov/sites/dolgov/files/EBSA/about -ebsa/our-
activities/resource-center/publications/reporting-and-disclosure-guide-for-employee-benefit -plans.pdf.
Congressional Research Service
19

Saving for Retirement: Household Decisionmaking and Policy Options

Investment Options and Management
In addition to how retirement decisions are framed, choice overload and diversification biases can
affect consumer decisionmaking. Limiting or curating investment options can sometimes help
people make appropriate retirement decisions.121 Moreover, annuity options122 in retirement plans
to cover the risk of outliving one’s retirement assets can sometimes improve retirees’ outcomes.123
In addition, simplifying the management or lowering costs can make retirement plans more
accessible for consumers. For these reasons, Congress continues to show interest in improving
retirement investment options and management to improve consumer decisionmaking and,
ultimately, retirement security for consumers.
The SECURE Act now makes it easier for 401(k) plans to offer annuities.124 The act also aims to
reduce employer costs for providing retirement accounts to their workers, particularly for smal
businesses (e.g., through tax incentives and multiple employer plans), which al ow businesses to
pool administrative responsibilities when offering retirement plans to employees. In addition,
H.R. 2120 and S. 1053 introduced in the 116th Congress would make consumer management of
retirement accounts easier by providing retirement accounts for self-employed workers or
employees without a retirement account option at their workplace.
Fiduciary and Investment Advice
As described in this report, retirement planning may be complex, and individuals sometimes rely
on financial services professionals to assist them with their investment choices, tax planning, or
other decisions. Sometimes, financial services professionals are compensated by commissions
from investment funds or other incentives, which may lead to conflicts of interest or
recommendations that are not in the best interest of their clients. For these reasons, the federal
government sets certain standards for financial services professionals, depending on their roles
and actions.
To protect the interests of pension plan participants and beneficiaries, Congress passed the
Employee Retirement Income Security Act of 1974 (ERISA; P.L. 93-406). ERISA set the
standards that pension plans (such as employer-sponsored DC plans) must follow, including
requiring that those who oversee pension plans (such as plan sponsors and administrators) have a
fiduciary duty to operate these plans prudently and in the sole interests of plan participants.125

121 Donald B. Keim and Olivia S. Mitchell, Simplifying Choices in Defined Contribution Retirement Plan Design,
NBER, Working Paper no. 21854, January 2016.
122 Annuity products provide consumers a set monthly income for life after a particular age, which can mitigate the risk
of retirees outliving their money. T he financial product is often a combination of insurance and a managed investment
vehicle.
123 Vanya Horneff, Raimond Maurer, and Olivia S. Mitchell, Putting the Pension Back in 401(k) Plans: Optimal Versus
Default Longevity Incom e Annuities, NBER, Working Paper no. 22717, October 2016.
124 T he Setting Every Community Up for Retirement Enhancement Act (SECURE Act; P.L. 116-94, Division O)
provides a fiduciary safe harbor for plan sponsors to select lifetime income providers. In 401(k) and other DC plans,
employers who comply with ERISA’s fiduciary safe harbor rules are free of liability when selecting an insurance
company to provide the annuity. T he provision responds to employers’ concerns about liability exp osure if the selected
insurance company proves unable, years later, to meet its obligations under the annuity contracts. In this way, the
SECURE Act makes it easier to offer annuities in DC plans by removing ambiguity about the applicable fiduciary
standard.
125 What constitutes investment advice within pension and retirement plans has been controversial. In 2016, DOL
finalized a rule that redefined and expanded the term, therefore expanding who has a fiduciary duty. Under the prior
regulation, securities brokers and dealers who provided services to retirement plans and who were not fiduciaries were
Congressional Research Service
20

Saving for Retirement: Household Decisionmaking and Policy Options

During the late 1980s and early 1990s, the landscape for the delivery of investment advice began
to shift as broker-dealers increasingly offered financial advisory services somewhat akin to
investment advisers, including investment and retirement planning. In 2019, the SEC finalized a
best interest standard for broker-dealer investment professionals, requiring them to act in the best
interest of a customer when making recommendations.126 This standard is widely perceived to be
more rigorous than the prior suitability standard for broker-dealers but less rigorous than a
fiduciary rule standard.127 This new standard continues to be controversial, as some argue that the
higher fiduciary standard is necessary to protect consumers when receiving retirement advice,128
and others propose setting ERISA’s investment advice standards closer to the SEC’s best interest
standard.129
Age Requirements for Retirement Contributions and Withdrawals
The SECURE Act increased some of the age limits for retirement contributions and withdrawals.
Before 2020, only workers younger than 70½ years old at the end of the calendar year were
eligible to contribute to a traditional IRA. The SECURE Act eliminated this maximum age
contribution requirement.130 In addition, before 2020, traditional IRAs required an account holder
to take required minimum distributions (RMDs) starting on April 1 in the calendar year after
turning 70½ years old. The IRS also required that account holders start receiving RMDs from
their 401(k) plans after turning 70½ years old, or if stil employed, the year the employee retirees.
The SECURE Act increased the age at which RMDs begin to 72.131

not required to act in the sole interests of plan participants. Rather, their recommendations had to meet a suitability
standard, which requires that recommendations be suitable for the plan participant, given factors such as an individual’s
income, risk tolerance, and investment objectives. Under DOL’s 2016 regulation, brokers and dealers are generally
considered to be fiduciaries when they provide recommendatio ns to participants in retirement plans. Although some
argued that retirement investment advice should meet a high standard because of the importance of retirement security
and the rise of DC funds, others argued that a higher standard could discourage financial services professionals from
providing advice due to a higher compliance requirement and possible increased litigation. Currently, the fiduciary rule
has been struck down by the Fifth Circuit Court of Appeals. T he ruling is available at http://www.ca5.uscourts.gov/
opinions/pub/17/17-10238-CV0.pdf. For more information on DOL’s 2016 rule, see CRS Report R44884, Departm ent
of Labor’s 2016 Fiduciary Rule: Background and Issues, by John J. T opoleski and Gary Shorter.
126 For more information on the Securities and Exchange Commission’s (SEC’s) best interest proposal, see CRS In
Focus IF11073, The SEC’s Best Interest Proposal for Advice Given by Broker-Dealers, by Gary Shorter; and CRS
Report R46115, Regulation Best Interest (Reg BI): The SEC’s Rule for Broker-Dealers, by Gary Shorter.
127 In addition, the SEC finalized a new consumer disclosure, “Form CRS Relationship Summary,” to help inform retail
investors of the nature of their relationship with the financial professional. For more information, see SEC, “ SEC
Adopts Rules and Interpretations to Enhance Protections and Preserve Choice for Retail Investors in T heir
Relationships With Financial Professionals,” press release, June 5, 2019, at https://www.sec.gov/news/press-release/
2019-89.
128 Letter from Ranking Member of the House Financial Services Committee Maxine Waters et al., to SEC Chairman
Jay Clayton, September 12, 2018, at https://financialservices.house.gov/uploadedfiles/cmw_-_reg_bi_to_sec_-
_9.12.2018.pdf.
129 DOL proposed a rule allowing new exemptions for investment advice fiduciaries to align these standards more
closely with the SEC’s advice standards. See DOL, “U.S. Department of Labor Proposes to Improve Investment
Advice and Enhance Financial Choices for Workers and Retirees,” press release, June 29, 2020, at
https://www.dol.gov/newsroom/releases/ebsa/ebsa20200629.
130 Under current law, no maximum contribution age exists for Roth IRAs and 401(k) plans.
131 Under current law, there is no minimum distribution requirement for Roth IRAs. §2203 of the Coronavirus Aid,
Relief, and Economic Security Act (CARES Act; P.L. 116-136) suspends required minimum distributions (RMDs) for
2020. A special rule applies the RMD suspension to individuals who took their first RMD from January 1, 2020, to
April 1, 2020. Individuals who received their RMDs in 2020 —prior to the enactment of the CARES Act —may be able
to roll over these amounts to IRAs or other retirement plans, if rollover rules are followed. Among other requirements,
Congressional Research Service
21

Saving for Retirement: Household Decisionmaking and Policy Options

Congress changed these age limits for retirement savings accounts due to increases in life
expectancy and the rise of people working longer.132 During 1969 to 1971, the time before IRAs
were first authorized in 1974,133 the average life expectancy at age 65 was about 15.0 years (13.0
years for men and 16.8 years for women); by 2017, it had increased to 19.4 years (18.0 years for
men and 20.6 years for women).134 Americans are also working longer. In the early 1980s, about
10% of Americans aged 70 and older worked at some time during the year (17% of men and 6%
of women); in 2018, the share increased to almost 16% (21% of men and 12% of women).135
Because withdrawals from DC plans are taxable, such withdrawals may push workers into a
higher income tax bracket, thus discouraging older adults from working.136 Therefore, increasing
the required distribution age in these plans might incentivize certain eligible employees to work
longer.
Retirement Account Leakage
Retirement savings accounts are designed to al ow people to set aside wages for use during
retirement. Preretirement withdrawals or loans from retirement savings accounts, sometimes
described as “leakages,” may reduce long-term retirement savings. To discourage withdrawing
retirement wealth early, preretirement distributions are general y subject to a 10% early
withdrawal tax penalty unless the plan participant is over 59½ years old, or the funds are used for
specific purposes designated in law.137 Leakages from retirement plans can take a variety of
forms, including cashing out plan assets upon separation from an employer and leakages due to
households’ financial needs—“hardship” withdrawals from the plan prior to retirement or
borrowing against plan assets.

rollovers must be completed within 60 days of the distribution. For more information, see CRS In Focus IF11472,
Withdrawals and Loans from Retirem ent Accounts for COVID-19 Expenses, by John J. T opoleski and Elizabeth A.
Myers.
132 U.S. House Committee on Ways and Means, Chairman Richard E. Neal, The Setting Every Community Up For
Retirem ent Enhancem ent Act of 2019, at https://waysandmeans.house.gov/sites/democrats.waysandmeans.house.gov/
files/documents/SECURE%20Act%20section%20by%20section.pdf.
133 ERISA (P.L. 93-406).
134 Elizabeth Arias and Jiaquan Xu, “United States Life T able, 2017,” National Vital Statistics Reports, vol. 68, no. 7
(June 24, 2019), T able 21, at https://www.cdc.gov/nchs/data/nvsr/nvsr68/nvsr68_07-508.pdf. According to the National
Vital Statistics System, 15.0 years was the average life expectancy for men and women aged 65 during 1969 to 1971
(i.e., men and women aged 65 would, on average, be expected to live another 15 years, or to age 80); and 19.4 years in
2017 (i.e., men and women aged 65 would, on average, be expected to live another 19.4 years, or to age 84.4).
135 CRS analysis of data from the U.S. Census Bureau, Current Population Survey, 1974 -2019 Annual Social and
Economic Supplements. T he share of working men aged 70 and older was higher in the 1970s than in the 1980s. In
1974, 23% of men aged 70 and older worked some time during the year.
136 Damir Cosic and Richard W. Johnson, How Much Does Work Pay at Older Ages?, T he Urban Institute, November
2019.
137 T he exceptions to the 10% additional tax are listed at 26 U.S.C. §72(t)—examples include disability, death, and
medical expenses that exceed 7.5% of adjusted gross income; some instances related to domestic relations orders; and
calls to active duty. §2202 of the CARES Act exempts qualified individuals affected by Coronavirus Disease 2019
(COVID-19) from the 10% early withdrawal penalty for distributions (1) up to $100,000 , and (2) taken from January 1,
2020, through December 31, 2020. Qualified individuals are individuals (1) who tested positive for COVID-19 or those
with a spouse or dependent who tested positive for COVID-19; (2) facing financial difficulties due to being
quarantined, furloughed, laid off, or unable to work due to lack of child care or reduced work h ours as a result of
COVID-19; or (3) whose business closed or reduced hours as a result of COVID-19. Plan administrators may rely on
employees’ certifications as proof that they are qualified individuals. For more information, see CRS In Focus IF11472,
Withdrawals and Loans from Retirem ent Accounts for COVID-19 Expenses, by John J. T opoleski and Elizabeth A.
Myers.
Congressional Research Service
22

Saving for Retirement: Household Decisionmaking and Policy Options

Cash-Outs at Job Separation
Among al types of retirement account leakages, cash-outs from DC plans at job separation can
result in the largest amounts of leakage and the greatest proportional loss in retirement savings.138
Retirement plan participants may keep their accounts in a former employer’s plan if the balance is
greater than $5,000,139 roll over the account into a new employers’ retirement plan or an IRA, or
receive the account balance directly as a distribution from the plan (usual y subject to a tax
penalty). Current law requires that employers transfer retirement account balances above $1,000
but below $5,000 to an IRA when a job change occurs, absent an election from the participant.140
Individuals may keep the balance in this IRA account or take the balance out.
Certain policies aim to reduce cash-outs. The PPA requires employers to provide notice to
departing employees about the consequences of cashing out their retirement savings. Some
policies recommend employers provide separating participants with projections of their account
balance under different scenarios, ranging from keeping al assets in their tax-deferred retirement
account to cashing out the entire account as a lump sum. These types of disclosures aim to
discourage consumers from choosing to cash out their retirement plan when they leave their job.
Hardship Withdrawals and Loans in Retirement Plans
Other ways to access retirement assets early include hardship withdrawals and loans against
retirement assets. These types of retirement account leakage are usual y due to household
financial needs, such as major purchases or emergency expenses. Policies limiting or prohibiting
these types of leakage could help to preserve retirement assets, but they might also induce
workers to save less in DC plans or IRAs, to shift their savings to other types of assets, or to
borrow more to meet consumption needs.
Hardship withdrawal requirements differ between DC plans and IRAs.141 For 401(k) plans,
hardship distributions are subject to a tax penalty and can include expenses such as first-time
home purchase costs (excluding mortgage payments), certain postsecondary tuition expenses,
certain medical expenses, payments to prevent eviction or foreclosure, and expenses related to
damages from a federal y declared disaster.142 Withdrawals from IRAs have fewer restrictions
than 401(k) plans and can be cashed out for any purpose.143 In addition, the tax penalty can be
avoided for first-time home purchases (up to $10,000) and postsecondary education. The share of

138 Leakages have remained relatively steady based on Survey of Income and Program Participa tion data collected in
1998, 2003, and 2006. Estimates based those data show that more than 70% of total leakage amounts reported by
401(k) participants are from cash-outs at job separation. See GAO, 401(k) Plans: Policy Changes Could Reduce the
Long-term Effects of Leakage on Workers Retirem ent Savings, GAO-09-715, August 2009, at https://www.gao.gov/
products/GAO-09-715.
139 Internal Revenue Code of 1986 (P.L. 99-514), §411(a)(11).
140 T he Economic Growth and T ax Relief Reconciliation Act of 2001 ( P.L. 107-16).
141 See CRS In Focus IF11369, Early Withdrawals from Individual Retirement Accounts (IRAs) and 401(k) Plans, by
Elizabeth A. Myers.
142 For regulations regarding deemed hardship distributions, see 26 C.F.R §1.401(k)-1(d)(3)(B). Prior to 2019,
participants were unable to (1) make contributions for a six -month period following a hardship distribution or (2) take a
hardship distribution before requesting a plan loan. §§41113 and 41114 of the Bipartisan Budget Act of 2018 (P.L.
115-123) eliminated these requirements. P.L. 115-123 also expanded the types of contributions and earnings that can be
received as hardship distributions to include employee contributions, qualified nonelective and matching contributions,
and related earnings.
143 Early withdrawals from Roth IRAs are generally subject to different rules. See CRS In Focus IF11369, Early
Withdrawals from Individual Retirem ent Accounts (IRAs) and 401(k) Plans, by Elizabeth A. Myers.
Congressional Research Service
23

Saving for Retirement: Household Decisionmaking and Policy Options

DC plan participants, including 401(k) plan participants, who take hardship withdrawals is
relatively low—in 2018, about 1.6% of DC plan participants took hardship withdrawals.144 IRA
withdrawals are more common—based on data from the IRS, about 12% of taxpayers who had
traditional IRA accounts and were below age 60 took distributions in 2016.145 Research has found
that those with low incomes, few assets, and who were younger were more likely to take early
withdrawals from retirement plans.146
Policymakers debate whether hardship withdrawal rules should be expanded or tightened. Some
proposals expand qualified withdrawals for certain occasions when many households have special
financial needs. For example, the SECURE Act al ows certain penalty-free withdrawals from
retirement plans if a child is born or adopted. However, others argue that the hardship distribution
qualifications for 401(k) plans are too broad and give participants access to money for
circumstances that are both voluntary and foreseeable, such as first-time home purchase costs.
Policies that al ow more qualifying withdrawals may encourage more retirement savings but also
may reduce long-term retirement savings accumulations.
Another type of leakage is borrowing against plan assets. The Internal Revenue Code al ows
participants in employer-sponsored retirement plans to borrow against their accounts, but plans
are not required to al ow such loans. Loans are not permitted against IRAs.147 According to the
Investment Company Institute, the share of eligible DC plan participants who had loans
outstanding increased from 15.3% in 2008 to 18.5% in 2011, likely due to financial stress during
the 2007-2009 recession. Loan rates fel to 16.7% at the end of 2018.148 The amount of loan
defaults in 401(k) plans is relatively smal compared with the asset base.149 Unlike hardship
distributions, loans are not subject to income taxes or the early withdrawal penalty if repayments
continue on schedule.150 Loan defaults in 401(k)s are general y treated as a distribution from the
plan, subject to income tax and possibly an early distribution penalty. Starting in 2018, if a
participant were to leave his or her job with an outstanding 401(k) loan, the loan would be treated

144 ICI, “Defined Contribution Plan Participants’ Activities, 2018,” May 2019, at https://www.ici.org/pdf/
ppr_18_rec_survey.pdf (hereinafter ICI, “ DC Plan Participants’ Activities, 2018”).
145 T he amount of distributions accounted for 2% of the total balance in those traditional IRAs. See IRS, SOI Division,
Accum ulation and Distribution of Individual Retirem ent Arrangem ents (IRA) , at https://www.irs.gov/statistics/soi-tax-
stats-accumulation-and-distribution-of-individual-retirement-arrangements#tables.
146 Barbara A. Butrica, Sheila R. Zedlewski, and Philip Issa, Understanding Early Withdrawals from Retirement
Accounts, T he Urban Institute, May 2010, at https://www.urban.org/sites/default/files/publication/28706/412107-
Understanding-Early-Withdrawals-from-Retirement-Accounts.PDF.
147 Money in an IRA can, in effect, be “borrowed” for 60 days (once per 12 months) because the law states that any
distribution from an IRA that is not deposited in the same or another IRA within 60 days is a taxable distribution (26
U.S.C. §408(d)).
148 ICI, “DC Plan Participants’ Activities, 2018.”
149 In 2016, outstanding participant loans made up only 1.4% ($66.6 billion) of the $4.7 trillion in 401(k) plan assets. A
study estimates that $6 billion on average annually flows out of 401(k) plans as a result of loan defaults. See DOL,
EBSA, Private Pension Plan Bulletin: Abstract of 2016 Form 5500 Annual Reports, December 2018, at
https://www.dol.gov/sites/dolgov/files/ebsa/researchers/statistics/retirement -bulletins/private-pension-plan-bulletins-
abstract -2016.pdf; and T imothy Lu et al., Borrowing From The Future: 401(k) Plan Loans and Loan Defaults, NBER,
Working Paper no. 21102, April 2015.
150 Other reasons that borrowing from a 401(k) plan is usually preferable to taking a hardship with drawal include the
following: with a loan, the account balance is not permanently reduced because the loan will be repaid into the account,
generally within five years; and unlike a hardship distribution, after which employee contributions must be suspende d
for six months, the participant can continue to contribute to the plan while the loan is outstanding.
Congressional Research Service
24

link to page 25 Saving for Retirement: Household Decisionmaking and Policy Options

as an early withdrawal (possibly subject to tax penalty) if it is not repaid by the due date of the tax
return for the year when the participant leaves the job.151
Research shows that the existence of a loan feature may encourage workers to enroll in the
retirement plan and to al ocate more of their salary into the plan.152 But concerns also exist that by
using loans, people risk taking on too much debt, defaulting on their loans, and harming their
future retirement security. For example, the SECURE Act contains a provision to prohibit
qualified employer plans from making loans through credit cards and other similar
arrangements.153
Converting Savings into Retirement Income
At retirement, account holders—such as those with 401(k) plans and IRAs—general y need to
decide how to convert account balances into income. Withdrawals mostly are taken in the form of
lump-sum distributions or periodic withdrawals. Households may face the risk of either outliving
their resources by withdrawing too much during earlier years or consuming too little by spending
too conservatively. People may also have to make investment decisions on the remaining assets in
their retirement accounts.
One option for drawing down retirement assets is to purchase an annuity using part of retirement
account balances. Annuities may take different forms. A life annuity—also cal ed an immediate
annuity—is an insurance contract that provides guaranteed income payments for life in return for
an initial lump-sum premium. A life annuity pays income to the purchaser for as long as he or she
lives and, in the case of a joint and survivor annuity, for as long as the surviving spouse lives.
Another annuity product, the Advanced Life Deferred Annuity, can be purchased at retirement
with a smal er share of accumulated assets and payments beginning at a later age, such as 85.
People can also purchase an inflation-adjusted annuity that provides annual increases in
income.154 Annuities can be used to reduce the likelihood that people may outlive their resources
and to al eviate some post-retirement investment risks.155 One recent study shows that annuity

151 An Act to provide for reconciliation pursuant to titles II and V of the concurrent resolution on the budget for fiscal
year 2018 (P.L. 115-97).
152 Research on the impact of the availability of a loan feature on 401(k) plan participation and contributions has found
either positive effects or no significant impact. For example, one study found that adding a loan option increases 401(k)
plan participation but did not find a conclusive impact on contributions. See John Beshears et al., The Im pact of 401(k)
Loans on Saving, NBER, Retirement Research Center Paper no. NB 09-05, September 2010. Another study found no
impact of loans on participation rates, but concludes that the loan option increases the contribution rate by 10 % among
nonhighly paid participants. See Olivia Mitchell, Stephen P. Utkus, and T ongxuan Yang, “ T urning Workers into
Savers? Incentives, Liquidity, and Choice in 401(k) Plan Design,” National Tax Journal, vol. 60 (2007), pp. 469-489.
A 1997 report from GAO shows that the availability of loans in employer provisions slightly increases particip ation.
See GAO, 401(k) Pension Plans: Loan Provisions Enhance Participation But May Affect Incom e Security for Som e ,
GAO/HEHS-98-5, October 1997, at https://www.gao.gov/archive/1998/he98005.pdf.
153 §2203 of the CARES Act modifies rules governing DC plan loans for qualified individuals during 2020 by
increasing the maximum possible loan amount and extending debt repayment dates. Qualified individuals are defined in
the same way as footnote 136. For more information, see CRS In Focus IF11472, Withdrawals and Loans from
Retirem ent Accounts for COVID-19 Expenses, by John J. T opoleski and Elizabeth A. Myers.
154 In the inflation-adjusted annuity, the annual increases usually are paid for by a lower initial monthly annuity income
or a higher premium.
155 Some annuities allow the purchaser to share in investment gains from growth in equity markets as a way to offset
the effects of inflation. T hese annuities also require the purchaser to share in the investment losses if markets fall.
Congressional Research Service
25

Saving for Retirement: Household Decisionmaking and Policy Options

strategies can be used to delay Social Security claiming,156 thus resulting in a higher lifetime
income.157
Despite the potential advantages of annuities, the market for life annuities remains relatively
smal in the United States compared with other retirement financial products.158 Several reasons
may explain why the demand for annuities is low. For example, some people may already feel
they have a sufficient amount of annuitized income from Social Security. In addition, about a
third of people aged 65 and older have annuity income from DB pensions.159 Another reason may
be that the relatively high fees and expenses charged by insurance companies deter people’s
wil ingness to purchase life annuities.160 Further, annuity contracts are not easily canceled, and
many individuals fear that after purchasing an annuity, they may later need a large sum of money
to pay for unexpected expenses (e.g., long-term care or health expenses).161 In addition, some
people might be reluctant to purchase an annuity because they desire to leave a bequest to their
heirs.162
Disclosures can be designed to help retirement account holders make retirement account
withdrawal decisions. As described earlier in this report, the SECURE Act requires plans to
provide participants with a lifetime retirement income projection in at least one statement per
year. Such statements for retirees can provide an estimated withdrawal anchor to help retirees
make decisions. However, the lifetime income disclosure is an estimate and probably could not
take personalized post-retirement investment risks into account.
Plan sponsors can be encouraged to offer more annuity options to participants. Currently,
relatively few 401(k) plans provide the opportunity for retiring workers to convert al or part of
their 401(k) accounts into life annuities at retirement.163 As discussed earlier in this report, the
SECURE Act makes it easier for 401(k) plans to offer annuities. However, some researchers
suggest that defaulting some portion of 401(k) plan assets into an annuity would likely result in
more account holders using an annuity investment product, rather than offering annuities as an
option.164 In terms of the timing of annuity decisions, one recent study finds that the take-up rate

156 Delaying Social Security claiming from age 62 to 70 can increase potential Social Security monthly benefits. See
CRS Report R44670, The Social Security Retirem ent Age, by Zhe Li.
157 Alicia H. Munnell, Gal Wettstein, and Wenliang Hou, How Best to Annuitize Defined Contribution Assets?, Center
for Retirement Research, Boston College, Working Paper CRR WP 2019-13, October 2019 (hereinafter Munnell,
Wettstein, and Hou, 2019).
158 In 2018, sales of single premium immediate annuities amounted to only $9.7 billion. In compariso n, total long-term
care expenditures for the elderly amounted to roughly $150 billion. See Munnell, Wettstein, and Hou, 2019.
159 Edward N. Wolff, The Decline of African-American and Hispanic Wealth since the Great Recession, NBER,
Working Paper no. 25198, 2008.
160 See David McCarthy and Olivia S. Mitchell, “Estimating International Adverse Selection in Annuities,” North
Am erican Actuarial Journal, vol. 6, no. 4 (October 2002), pp. 38 -54; and Jeffrey R. Brown et al., The Role of Annuity
Markets in Financing Retirem ent, (Cambridge, MA: T he Massachusetts Institute of T echnology Press, 2001).
161 See James M. Poterba, Steven Venti, and David Wise, “T he Composition and Drawdown of Wealth in Retirement,”
Journal of Econom ic Perspectives, vol. 25, no. 4 (2011), pp. 95-118; and Society of Actuaries, “ 2017 Risks and
Process of Retirement Survey,” 2017, at https://www.soa.org/globalassets/assets/files/resources/research-report/2018/
risk-process-retirement.pdf.
162 Lee M. Lockwood, “Bequest Motives and the Annuity Puzzle,” Review of Economic Dynamics, vol. 15, no. 2
(2012), pp. 226-243.
163 See CRS Report R40707, 401(k) Plans and Retirement Savings: Issues for Congress, by John J. T opoleski.
164 Munnell, Wettstein, and Hou, 2019.
Congressional Research Service
26

Saving for Retirement: Household Decisionmaking and Policy Options

would be higher if participants were required to make annuity decisions a decade before
retirement.165
Conclusion
Due to the growth in DC plans and IRAs, it follows that Americans are more responsible for
managing their retirement, from a financial perspective, than in the recent past. This change has
required people to take on more responsibility for making informed decisions about retirement
planning and has put more risk on individuals and households.
New policies, employer initiatives, or financial products could make retirement planning
decisions less difficult or aim to reduce major risk exposures for households. For example, as
described in this report, information disclosure, financial literacy, and high-quality investment
advice and products can help people navigate retirement decisions. In general, a better
understanding of consumer decisionmaking in shaping a household’s finances during its working
years may be important for supporting a secure retirement.

Author Information

Cheryl R. Cooper
Zhe Li
Analyst in Financial Economics
Analyst in Social Policy

Disclaimer
This document was prepared by the Congressional Research Service (CRS). CRS serves as nonpartisan
shared staff to congressional committees and Members of Congress. It operates solely at the behest of and
under the direction of Congress. Information in a CRS Report should n ot be relied upon for purposes other
than public understanding of information that has been provided by CRS to Members of Congress in
connection with CRS’s institutional role. CRS Reports, as a work of the United States Government, are not
subject to copyright protection in the United States. Any CRS Report may be reproduced and distributed in
its entirety without permission from CRS. However, as a CRS Report may include copyrighted images or
material from a third party, you may need to obtain the permission of the copyright holder if you wish to
copy or otherwise use copyrighted material.

165 Anran Chen, Steven Haberman, and Stephen T homas, “T he implication of the hyperbolic discount model for the
annuitisation decisions,” Journal of Pension Economics & Finance, vol. 19, no. 3 (July 2020), pp. 372–391.
Congressional Research Service
R46441 · VERSION 1 · NEW
27

Download PDF

Download EPUB

Revision History

Jul. 2, 2020

HTML · PDF

Metadata

Topic areas

Domestic Social Policy

Economic Policy

Report Type: CRS Report

Source: CRSReports.Congress.gov

Raw Metadata: JSON