Angela M. Antonelli
Research Professor and Executive Director
Policy Report 20-02
December 2020
What are the Potential Benets of
Universal Access to Retirement Savings?
An Analysis of National Options to Expand Coverage
In conjunction with
Support for this research was provided by
economics | strategy | insight
The Center for Retirement Initiatives (CRI) at Georgetown University is a research center of the McCourt School
of Public Policy, one of the top-ranked public policy programs in the nation. Through its academic reputation
and ability to engage with policymakers, business leaders, and other stakeholders, the McCourt School attracts
world-class scholars and students who have become leaders in the public, private, and nonprot sectors. The
CRI is dedicated to:
• Strengthening retirement security by expanding access and coverage for the private sector workforce;
• Providing thought leadership and developing innovative new approaches to retirement savings, investment,
and lifetime income;
• Serving as a trusted policy advisor to federal, state, and local policymakers and stakeholders.
600 New Jersey Avenue, NW, 4th Floor
Washington, D.C. 20001
202-687-4901 · https://cri.georgetown.edu/
About the Center for Retirement Initiatives (CRI)
Econsult Solutions, Inc. provides businesses and public policymakers with consulting services in urban
economics, real estate, transportation, public infrastructure, public policy and nance, community and
neighborhood development, planning, and thought leadership, as well as expert witness services for litigation
support. Our technical expertise ranges from Big Data analysis to GIS-based spatial analytics, sophisticated
benet-cost analysis, and pro forma-based project feasibility analysis.
ESI’s government and public policy practice combines rigorous analytical capabilities with a depth of experience
to help evaluate and design eective public policies and benchmark and recommend sound governance
practices. ESI has assisted policymakers at multiple levels of government to design and evaluate programs that
help citizens increase their economic security.
1435 Walnut Street, 4th Floor
Philadelphia, PA 19102
215-717-2777 · https://econsultsolutions.com/
About Econsult Solutions, Inc. (ESI)
The Berggruen Institute’s mission is to develop foundational ideas and shape political, economic, and social
institutions for the 21st century. By providing critical analysis using an outwardly expansive and purposeful
network, we bring together some of the best minds and most-authoritative voices from across cultural and
political boundaries to explore fundamental questions of our time. Our objective is to have an enduring impact
on the progress and direction of societies around the world. To date, projects inaugurated at the Berggruen
Institute have helped develop a youth jobs plan for Europe; fostered a more-open and constructive dialogue
between Chinese leadership and the West; strengthened the ballot initiative process in California; and launched
Noema, a new publication that brings thought leaders from around the world together to share ideas. In addition,
the Berggruen Prize, a $1 million award, is conferred annually by an independent jury to a thinker whose ideas
are shaping human self-understanding to advance humankind.
Bradbury Building
304 S. Broadway, Suite 500
Los Angeles, CA 90013
310-550-70983 · https://www.berggruen.org/
About the Berggruen Institute
© 2020, Georgetown University
All Rights Reserved
iii
What are the Potential Benets of Universal Access to Retirement Savings?© 2020 Georgetown University Center for Retirment Initiatives
Acknowledgments
The Georgetown University Center for Retirement Initiatives (CRI) is grateful to the Berggruen Institute for
the generous support that has made this report possible, and to Econsult Solutions, Inc. (ESI) for a research
collaboration that has allowed the Center’s vision for this report to become a reality. We are honored to partner
with these organizations to advance our shared mission of strengthening retirement security and promoting the
expansion of access to savings options for millions of American workers who currently lack such access.
The CRI also thanks Courtney Eccles, Yakov Feygin, J. Mark Iwry, David John, Michael Kreps, and David Morse
for their helpful consultations and feedback in the preparation of this report. The ndings and conclusions
expressed are solely those of the author and do not represent those of the Berggruen Institute, Econsult
Solutions, Inc., or the Center for Retirement Initiatives.
Suggested Report Citation
Antonelli (2020). What are the Potential Benets of Universal Access to Retirement Savings?
Georgetown University Center for Retirement Initiatives in conjunction with Econsult Solutions, Inc.
© 2020, Georgetown University
All Rights Reserved
iv
What are the Potential Benets of Universal Access to Retirement Savings? © 2020 Georgetown University Center for Retirment Initiatives
Contents
Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1. Closing the Signicant Gaps in Access to Retirement Savings . . . . . . . . . . . . . . . . 9
1.1 Signicant Gaps Remain in Access to Retirement Savings . . . . . . . . . . . . . . . . . 9
Gaps in Private Sector Access Disproportionately Impact Certain Groups . . . . . . . . . . 9
Too Many Have Little Saved for Retirement. . . . . . . . . . . . . . . . . . . . . . . 10
An Aging Population Increases the Urgency . . . . . . . . . . . . . . . . . . . . . . 11
1.2 Policy Approaches Taken to Close the Access Gap . . . . . . . . . . . . . . . . . . . 12
International Models Toward Universal Access . . . . . . . . . . . . . . . . . . . . . 13
US Eorts Have Fallen Short of Universal Access . . . . . . . . . . . . . . . . . . . . 14
National Proposals for Universal Access . . . . . . . . . . . . . . . . . . . . . . . . 17
Analyzing How Dierent Design Options Aect Access and Savings . . . . . . . . . . . . . 19
2. Analyzing the Potential Benets of National Universal Access to Retirement Savings Options . . . 21
2.1 Participation, Savings, and Assets under a Baseline Auto-IRA Scenario . . . . . . . . . . . 21
Starting Sooner and Saving Longer Signicantly Improves Retirement Outcomes . . . . . . . 21
How a Payroll Deduction Auto-IRA Expands Access and Builds Savings . . . . . . . . . . . 23
Protecting Savings is Critical to Asset Growth and Retirement Income . . . . . . . . . . . . 27
2.2 Policy Choices Have Impacts on Coverage and Savings . . . . . . . . . . . . . . . . . 29
Thresholds for Employer Participation Dictate the Remaining Access Gap . . . . . . . . . . 29
Voluntary Employer Contributions Can Increase Participant Savings . . . . . . . . . . . . . 31
Required Employer Contributions Can Increase Employee Returns — with Economic Trade-os . . 33
Analyzing How Policy Choices Aect Participation, Savings, and Asset Building . . . . . . . . 34
3. Long-Term National Impacts from Increased Savings . . . . . . . . . . . . . . . . . . . . 38
3.1 Increased Economic Growth and Tax Revenue . . . . . . . . . . . . . . . . . . . . . 38
Enhancing Economic Productivity and Accelerating Growth . . . . . . . . . . . . . . . . 38
Increased Tax Revenues from Economic Growth . . . . . . . . . . . . . . . . . . . . 40
3.2 Assessing the Impact on Federal Benet Programs . . . . . . . . . . . . . . . . . . . 41
Federal Benet Program Spending is Anticipated to Rise Signicantly . . . . . . . . . . . . 41
Increasing Retiree Incomes Can Reduce Program Expenditures . . . . . . . . . . . . . . 42
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Endnotes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
1
What are the Potential Benets of Universal Access to Retirement Savings?© 2020 Georgetown University Center for Retirment Initiatives
Part 1: Closing the Signicant Gaps in
Access to Retirement Savings
Workers in the United States are being asked to
take responsibility for their nancial well-being in
retirement now more than ever. What used to be
considered the foundation for building a secure
retirement — Social Security, employer-provided
pensions, and personal savings — has been
weakening for decades as traditional dened
benet (DB) pension plans have been replaced by
a dened contribution (DC) system of savings that
was originally meant to supplement, not replace,
traditional pensions.
Most employers today that have retirement plans
only oer DC options. This shift over time from
employer-provided pensions to DC plans has put
greater responsibility on workers to make complex
savings and investment decisions that will aect the
amount of money available in retirement. Americans
who have access to retirement savings accounts
often do not save enough to maintain their quality of
life in retirement.
Making matters worse, while employer-sponsored
retirement plans have become the primary way private
sector workers build retirement savings, employers in
the United States are not required to oer retirement
savings plans. Today, there are an estimated 57
million private sector workers (46%) who do not have
access to a plan through the workplace (see Figure
ES.1). These access gaps are inequitably distributed,
aecting more small businesses, and with larger gaps
among lower-income workers, younger workers,
minorities, and women.
For several years, there have been discussions and
proposals in the United States about how to expand
access to ways to save for retirement. If we look
internationally, there is usually little debate about
the value of universal access to retirement savings,
and several countries require employers to provide a
retirement savings option for their employees. With
all workers covered, dierences can be found in
the design of such options to achieve the levels of
savings needed to boost income in retirement.
Universal access to retirement savings options
would give all workers the opportunity to save, and
evidence from other countries, from individual states,
and from private sector plans suggests that many
would begin to do so, especially when encouraged
using default options, such as automatic enrollment.
Workers would benet from the increased savings
and the additional income in retirement. At the
same time, the economy benets from stronger
savings, investment, and economic growth, and the
nation benets from a reduction in scal pressures
to support an aging population lacking sucient
retirement income.
An Aging Population Increases the Urgency
This lack of access to employer-sponsored
retirement savings plans takes on greater urgency
due to the aging of the US population. Senior
households are growing signicantly in number and
as a share of the population, with the “dependency
ratio” projected to fall from its historic norms of
almost four working age households for each elderly
household to a ratio of closer to two to one (see
Executive Summary
67.3M57.3M
46%
GAP
Figure ES.1: More than 57 Million Employees
Lack Access to a Retirement Savings Plan in their
Workplace (2020)
ESI analysis of Census Bureau Current Population Survey
and BLS National Compensation Survey Data.
Access to coverage at work
Coverage access gap
124.6 M
Private Sector Employees
2
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
Figure ES.2). Since working age households are
the primary contributors to the tax base, this falling
dependency ratio creates greater scal pressures as
the demand for benet programs increases.
This shift in population composition also underscores
the importance of enabling younger generations
like millennials and Gen Z (which will cover the
prime working ages of 30–60 by 2040) to have
opportunities during their crucial savings years to
build resources to support their nancial futures.
Policy Approaches Taken to Close the Access
Gap
Federal policymakers in the United States have
developed and started to implement reforms
intended to close the gap in private sector retirement
savings access, encourage savings, and strengthen
the retirement readiness of workers. International and
state examples provide models to achieve universal
access that can do much more to expand coverage
and savings levels.
International Models Toward Universal Access
Eorts to expand access, participation, and savings
are not unique to the US. Many countries have
adopted a mix of public and private models to move
toward universal access, often requiring employer
participation and/or the automatic enrollment of
workers who can choose to opt out, that have resulted
in signicant coverage and savings levels. Established
programs in countries like Australia, New Zealand
and the United Kingdom have gained signicant scale
over time, demonstrating the sustainability of these
types of programs to help participants save more for
retirement (see Figure ES.3).
US Eorts Have Fallen Short of Universal Access
Several legislative proposals intended to achieve
national universal access, modeled on international
experience and the innovative design ideas of policy
experts, have been introduced in Congress for more
than a decade and as recently as 2019. To date,
these proposals have not had sucient support to
advance.
In the absence of national action, some states have
started to adopt innovative public-private partnership
models to expand access to their workers. A few
of these new state programs have adopted and
launched an Auto-IRA model, which requires
employers that do not already oer their workers
a retirement savings plan to automatically enroll
their workers in the state program to begin to save
unless the worker opts out. These state programs
are currently providing many employers and their
employees with new ways to save, and the number
of new accounts and assets is now growing at a
steady pace (see Figure ES.4).
1
Recent Congressional action, such as the SECURE
Act (P.L. 116-94), intended to expand the adoption
Figure ES.3: Employer-Based International
Savings Programs
Australia Superannuation Guarantee – 16.7 million
participants
Requires employers to contribute 9.5% of an eligible
employee’s earnings to a retirement savings account.
KiwiSaver – 3 million participants
Workers auto enrolled (can opt out) to contribute ≥
3% of earnings + 3% employer match and a tax credit
contribution.
UK NEST – 9 million participants
Uncovered workers auto enrolled (can opt out) at default
contribution levels of 5% employee + 3% employer.
Figure ES.2: Falling Ratios of Working-Age to
Elderly Households Create Fiscal Pressure
ESI analysis of US Census Bureau data
and University of Virginia Population Projections.
3.9
3.8
3.6
3.6
3.8
3.9
3.7
3.2
2.8
2.5
2.3
2.3
2.3
2040203520302025202020152010200520001995199019851980
3
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
and improve the design of dened contribution
plans, is another positive step.
2
While these
individual state programs and recent incremental
federal reforms are benecial, these initiatives are
unlikely to achieve a signicant national expansion
of coverage and savings.
Part 2: Analyzing the Potential Benets of
National Universal Access to Retirement
Savings Options
The experience of well-established international
programs and, more recently, the experience of
individual state retirement savings programs point
to the need for serious consideration of national
universal access to retirement savings options to
expand the number of employers who oer their
workers a way to save for retirement. Such options
would require certain employers to provide their
employees with access to a savings option, while
retaining the ability of employees to choose to opt
out of saving. A national, universal access approach
to retirement savings would substantially increase
participation and savings levels, particularly among
low- and middle-income workers.
Drawing on a range of state, national, and
international programs and proposals, this study
analyzes the potential impacts for access, savings,
and retirement security of a “baseline” national
universal access retirement savings option, and
the relative impacts on coverage and savings of
a number of potential policy variations from this
baseline.
The baseline model analyzed is an automatic
enrollment payroll deduction individual retirement
account (Auto-IRA) that is similar to a model
currently being implemented by some states and
included in legislative proposals introduced in
Congress. This streamlined and low-cost approach
uses automatic enrollment, default savings, and
auto-escalation mechanisms, which encourage
participation and savings while leaving participants
with full control over their participation and
contribution levels. In this model, all contributions
are made by the employee with no employer
contribution.
This model is used as the baseline because it is
comprehensive in requiring workplace access,
and simple in its structure and implementation.
Alternatives to this baseline are then analyzed by
adjusting several design features, including:
• Varying the type of savings account used
between a payroll deduction Roth IRA and
Roth 401(k), factoring in dierences in the
administrative requirements and the costs of
such accounts;
• Adding employer size and age thresholds,
exempting the smallest and youngest
businesses from the requirement to provide their
employees with access to retirement savings;
• Including a voluntary employer contribution,
as permitted in 401(k) accounts to give
businesses the discretion to contribute to
employee accounts; and
• Requiring an employer contribution by
adding a new requirement for employers to
provide contributions into an employee’s 401(k)
account, improving the return on investment
for savers but generating additional economic
implications for businesses and workers.
Figure ES.4: Recently Launched State Auto-IRA
Programs
OregonSaves – Launched 2017
Auto-IRA program required for all employers without an
existing qualied plan, 5% default employee contribution
with auto-escalation, and no employer match permitted.
Illinois Secure Choice – Launched 2018
Auto-IRA program required for employers with ≥ 25
employees without an existing qualied plan, 5% default
employee contribution, and no employer match permitted.
CalSavers – Launched 2019
Auto-IRA program required for employers with ≥ 5
employees without an existing qualied plan, 5% default
employee contribution with auto-escalation, and no
employer match permitted.
4
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings? © 2020 Georgetown University Center for Retirment Initiatives
These policy variations are applied to generate four
modeled scenarios (see Figure ES.5). Most policy
features are retained across the scenarios to isolate
the impact of only those features that have been
adjusted on access, savings, and retirement security
for workers currently lacking access. Modeling and
discussion within this analysis reects the trade-os
among these objectives, the potential challenges for
dierent groups (such as employers and employees),
and some of the technical considerations inherent
in policy eorts of this scale. Policy options are
analyzed through the year 2040, assuming adoption
of a policy in 2021 and a phased implementation
period from 2022–2026.
Analysis of these scenarios shows that national
universal workplace access scenarios could
reduce the access gap and expand retirement
savings coverage by 28 to 40 million workers
(depending on the chosen design features) by
the year 2040, with additional participation from
50 to 70% of private sector workers currently
lacking access. Because employees can choose to
opt out, no scenario will achieve 100% participation
by all eligible workers. Nevertheless, by starting
to save early in their careers, through simple,
automatic, and consistent contributions, and by
capitalizing on incentives to save and compounding
investment returns over an extended time horizon,
millions of additional private sector workers with
typical earnings levels will begin to save and build
substantial private savings that will increase their
retirement incomes.
Starting Sooner and Saving Longer Signicantly
Improves Retirement Outcomes
Because the scenarios analyzed examine the
impact on coverage and savings through the
year 2040, retirees within this time frame only
include the cohort of older savers who will begin
to access retirement savings. However, younger
workers from the millennial and Gen Z cohorts
who will not yet have reached retirement age
within the study period have greater opportunities
to build assets through continued contributions
and additional years of compounding growth.
As a result, future generations of Americans will
see far greater benets from savings than those
quantied as of 2040 within these estimates.
A simple illustration of the additional supplemental
lifetime income at age 65 for a young Roth Auto-IRA
Figure ES.5: Modeled Scenarios Isolate the Impact of Policy Variations on Access and Savings
√
√
√
√
√
√
√
√
√
√
√
√
√
√
5
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
saver demonstrates the long-term benets to the
youngest workers, who will not have yet reached
retirement age by 2040. A young (25-year-old) saver
with modest earnings levels of around $35,000
per year contributing at the default level (5% auto-
escalating up to a cap of 10%) envisioned in the
baseline scenario would make contributions of
about $110,000 over a 40-year period, and have
an account that grows to more than $262,000 in
assets. If this lump sum is used to purchase an
immediate xed annuity at the age of 65, it would
generate an annual supplemental income stream of
$14,320 over the remainder of the saver’s lifetime
(see Figure ES.6). The returns for this young saver
could be helped even more by making an enhanced,
refundable Saver’s Tax Credit (“Saver’s Credit”)
available that would boost savings to more than
$390,000 and generate an annual supplemental
income stream of $21,300 for the remainder of the
saver’s lifetime.
The benets of starting sooner and saving longer
can produce signicant improvements in retirement
income outcomes and long-term retirement security.
The passage of time and the power of compound
interest boost savings, because future market returns
apply not only to initial contributions, but also to the
market returns already achieved. This compounding
dynamic means that options that encourage savings
at a younger age can have signicant long-term
payos for participants, even in instances where
savers are not able to contribute to their accounts
throughout their entire careers, as balances built
up in early years continue to grow throughout the
duration of a saver’s working years.
Expanding Access to Retirement Savings
The ability to close the access gap and boost
savings will be aected by the design of the savings
option. The type of retirement savings accounts
(IRA and/or 401(k) structure), the employers required
to participate, and the default levels of employee
contributions and any employer contributions over
time are all factors that will drive access, savings,
asset growth, and retirement income.
The Auto-IRA model with no employer threshold
(“Baseline Auto-IRA”) would expand access to
workers at all private sector rms, increasing
participation by more than 40 million workers in
the year 2040 (with the remaining workers gaining
access but choosing to opt-out of saving).
Participation levels fall signicantly if employers
below a certain employee threshold are exempt.
Policy options, whether through an Auto-IRA or
401(k) approach, that exempt smaller and younger
rms from the requirement to provide access would
limit the degree to which the scenarios can close
the access gap. As an example, an exemption of
Figure ES.6: Supplemental Lifetime Income at age 65 for a Young Auto-IRA Saver
(With and Without a Refundable Saver’s Credit)
65
55
453525
Assets w/ Credit: $390,456
Annual Annuity: $21,300
Assets w/o Credit: $262,427
Annual Annuity: $14,320
Contributions: $110,122
$0
$100,000
$200,000
$300,000
$400,000
6
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
rms with fewer than 10 employees or in existence
less than two years would reduce participation by
an estimated 11 million workers by 2040 under an
Auto-IRA model (“Threshold Auto-IRA”), with modest
additional reductions in coverage under 401(k)
approaches because of anticipated variations in the
number of rms and workers likely to participate (see
Figure ES.7).
Encouraging Savings and Asset Growth
While participation levels are lower in models
exempting some employers, those participating
may have higher average contributions and
savings. These dierentials are due in part to the
characteristics of the covered population (with
average participant earnings increasing when
excluding rms below an employer size and age
threshold). Dierences also arise between types of
savings accounts, with a 401(k) model producing
higher average contributions and savings levels
relative to an IRA model for any given employer
threshold level.
Average savings levels increase with a 401(k) model
with discretionary employer contributions (“Voluntary
Employer Contribution 401(k)”) when compared to
the IRA models due to matching or supplemental
contributions from some employers, the eect of
the increased annual contribution limit on a small
sub-set of savers, and lower anticipated levels of
early withdrawals. Savings levels are estimated to
be slightly lower under a 401(k) approach requiring
employer contributions (“Mandatory Employer
Contribution 401(k)” relative to a voluntary employer
contribution 401(k) approach, due in part to the
constraint on wage growth for workers from the
required employer contribution.
Average account balances for participants reaching
age 65 in 2040 grow from $66,300 in the threshold
Auto-IRA scenario to $75,200 under the voluntary
employer contribution 401(k) approach. The tradeo
for the higher balances for some savers is that a
larger number of workers will remain uncovered. The
baseline Auto-IRA model generates a lower average
account balance for participants reaching 65 of
$60,600 due to participation of more low-income
workers, which decreases average balances.
However, the baseline Auto-IRA that covers
all employers has the highest overall level of
savings. While per-participant savings are higher
under alternative approaches, the expansion of
coverage anticipated under the baseline Auto-IRA
scenario with no employer threshold leads to the
largest increase in overall savings among the policy
options modeled.
Annual contributions to savings accounts are
estimated to total more than $130 billion by 2040
under the baseline Auto-IRA model, adding up to a
cumulative $1.89 trillion over the analysis period,*
with policy alternatives producing $1.4–$1.5 trillion
in cumulative contributions (see Figure ES.8). Among
options with an employer threshold, the voluntary
employer contribution 401(k) generates slightly
higher savings levels than threshold Auto-IRA model.
These results illustrate potential trade-os for
consideration between payroll deduction IRA and
401(k) options. When analyzed using equivalent
employer thresholds, an IRA model encourages
a higher level of participation by presenting the
lowest barriers to participation for businesses and
savers. However, a 401(k) approach can encourage
0
10
20
30
40
50
60
40.4
29.6
28.3
28.1
Access Gap Among Workers <65 (2040): 58.6M
Figure ES.7: Required Universal Access to
Savings Options Can Increase Participation
by 50 to 70% Among Workers Currently
Lacking Access
Active Program Participants, 2040 (M)
Baseline
Auto-IRA
Mandatory
Employer
Contribution
401(k)
Voluntary
Employer
Contribution
401(k)
Threshold
Auto-IRA
7
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
higher average levels of contributions and asset
accumulation over time among those who do
participate due to its provisions around contributions
and withdrawals.
If feasible, a voluntary employer contribution
401(k) approach without a threshold for required
participation (similar to the baseline Auto-IRA
scenario) or a mandatory employer contribution
401(k) approach with a more-aggressive employer
contribution level could produce higher levels of
savings than the baseline Auto-IRA model. However,
these approaches and requirements have impacts
on participating businesses and the broader
savings market, and federal 401(k) or IRA legislative
proposals to date have typically contemplated
an employer threshold out of consideration for
the implications for the smallest businesses. The
inclusion of the baseline Auto-IRA scenario is
intended to show how important the decision of
whether to include and where to draw an employer
participation threshold is to overall levels of access,
participation, and savings.
Part 3: Long-Term National Impacts from
Increased Savings
In addition to the impacts on participating savers,
enhancing access, and building retirement savings
would have “downstream” impacts on the broader
economy and the nation’s scal health.
Increased Economic Growth and Tax Revenue
Savings programs have implications for the
everyday decision-making of businesses, workers,
and families. These individual microeconomic
decisions around what job to take, whether to start
a business, and how to spend disposable income
aggregate together to have signicant impacts on
the economy. More-accessible savings options help
the competitiveness of small businesses and the
nancial security of workers, including the self-
employed, encouraging a more-dynamic economy,
while increased savings levels grow the income
that senior households have available to spend in
retirement. In addition to the returns they generate
for individuals, personal savings provide a source of
capital for business investment and growth.
Increased productivity growth from increased
savings and investment accelerates GDP growth.
Expected increases in the growth rate from the
scenarios analyzed would add $72–$96 billion
(depending on program design) to the national GDP
in the year 2040 (see Figure ES.9).
Increases are highest under the baseline Auto-IRA
approach, which generates the largest increase in the
personal savings rates through the highest coverage
levels, thus stimulating the greatest productivity
growth. Among scenarios with an employer
threshold, the voluntary employer contribution 401(k)
generates slightly more growth than the threshold
Auto-IRA, though in either case the tradeo is that
the overall coverage is signicantly lower than the
baseline scenario. Increased economic activity
would also grow the tax base, increasing federal tax
collections in the year 2040 by $11–$14 billion.
*A phased implementation period is assumed from 2022–2026 for a policy
enacted in 2021, with participation in early years consistent with
some voluntary early sign-ups by employers before a phased
implementation of coverage requirements by employer size.
$132
$1.89 T
$1.53 T
$1.48 T
$1.42 T
Cumulative
$0
$30
$60
$90
$120
$150
2040
2035
2030
2025
2022
Total Contributions ($Billions)
Figure ES.8: Cumulative Savings Contributions
are Highest Within the Baseline Auto-IRA Model,
Totaling $1.9 Trillion through 2040
Baseline
Auto-IRA
Threshold
Auto-IRA
Voluntary Employer
Contribution 401(k
)
Mandatory Employer
Contribution 401(k)
8
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
Reduced Benet Program Spending
Reduced demand for government benet programs
is another long-term impact of increasing retirement
security. Several federal programs provide a
range of support resources to elderly Americans
with demonstrated needs, including health care,
nutrition, housing, and supplemental income. Federal
spending on these programs already totals nearly
$100 billion per year and is often supplemented
by state funding. Federal expenditures on these
programs are anticipated to grow by $75 billion
over the next two decades (absent any change in
retirement income trends) as the composition of
the population changes, increasing the demand
from an elderly population and the tax burden on
proportionately smaller generations of future workers.
The modeled universal access scenarios are all
expected to diminish this rate of growth in program
expenditures for low-income seniors over time by
increasing savings and retiree resources. Federal
and state governments share in these savings,
due to the shared nature of many programs. Federal
savings in the year 2040 under the baseline Auto-
IRA scenario are estimated at $6.2 billion and state
savings at $2.5 billion, for a total of $8.7 billion,
while alternative scenarios generate an estimated
combined federal and state program savings of
approximately $7 billion in 2040.
Conclusion
Any eort to signicantly improve retirement
readiness must expand access to ways to save for
retirement to as many workers as possible. The
ability to close the access gap and boost savings
will be aected by the way a program is designed.
The type of retirement savings accounts (IRA and/
or 401(k) structure), the employers required to
participate, and the default levels of employee
contributions and any employer contributions over
time are all factors that will drive access, savings,
asset growth, and retirement income.
Regardless of the model selected, what is clear is
that the benets to savers, retirees, and the nation’s
scal and economic well-being can be enormous.
Depending on the design features, a national
approach to universal access to retirement savings
which would require some or all employers to oer
their workers either an IRA or 401(k) could:
• Increase the number of workers saving for
retirement in the year 2040 by 28–40 million,
with participation from about 50–70% of private
sector workers who currently lack access;
• Help a young worker with a modest income who
starts saving early and follows program defaults
for 40 years to save enough to generate as
much as $14,320 in additional annual income
for retirement, increasing to $21,300 in annual
income if eligible to take advantage of a
refundable Saver’s Credit;
• Increase cumulative total retirement savings by
$1.4–$1.9 trillion by the year 2040; and
• Accelerate economic growth, increasing
national GDP by $72–$96 billion in the year
2040.
Experiences from other countries and the early
evidence from states here in the US demonstrate
that increases in access can be achieved in a simple,
cost-eective way that supports and includes a
private market of providers ready and willing to
compete to provide such options for employers and
their workers.
Additional Annual US GDP ($Billions)
$96 B
$78 B
$76 B
$72 B
$0
$20
$40
$60
$80
$100
$120
Figure ES.9: Increased Savings and Investment
Boost GDP Growth by $72–$96 Billion in
the Year 2040
Baseline
Auto-IRA
Threshold
Auto-IRA
Voluntary Employer
Contribution 401(k
)
Mandatory Employer
Contribution 401(k)
2040
2035
2030
2025
2022
9
What are the Potential Benets of Universal Access to Retirement Savings?© 2020 Georgetown University Center for Retirment Initiatives
1.1 Signicant Gaps Remain in Access to
Retirement Savings
Workers in the United States are being asked to take
responsibility for their nancial well-being in retirement
now more than ever. What used to be considered the
foundation for building a secure retirement — Social
Security, employer-provided pensions, and personal
savings — has been weakening for decades as
traditional dened benet (DB) pension plans have
been replaced by a dened contribution (DC) system
of savings that was originally meant to supplement,
not replace, traditional pensions.
Most employers today that have retirement plans only
oer DC options. This shift over time from employer-
provided pensions to DC plans has put greater
responsibility on workers to make complex savings
and investment decisions that will aect the amount
of money available in retirement. Even Americans
who have access to retirement savings accounts
often do not save enough to maintain their quality of
life in retirement. Making this situation worse is the
reality that almost half of all private sector workers do
not have access to employer-sponsored retirement
savings plans to help them save.
A rapidly aging population and dierences across
generations increase the urgency to address
retirement savings shortfalls. As senior households
grow in both in number and as a share of the
population, there will be fewer working households
to support the needs of the elderly, non-working
population. This demographic shift makes the
ability of elderly households to maintain their living
standards in retirement an important economic and
quality of life issue for all US households.
Over the next decade, for example, the nal wave of
baby boomers will reach retirement age, Generation
X will approach retirement, and millennials and
increasingly Generation Z will be in their prime
working years. This shift in population composition
also underscores the importance of enabling younger
generations like millennials and Gen Z (which by 2040
will cover the prime working ages of 30–60) to have
opportunities during their crucial savings years to
build resources to support their nancial futures.
Gaps in Private Sector Access Disproportionately
Impact Certain Groups
Millions of private sector workers in the United States
lack access to an employer-sponsored retirement
savings plan. Estimates of the size of this “access
gap” range signicantly based on the data source
and method of analysis, ranging from 33% (about
40 million) to 64% (about 80 million) of the roughly
125 million private sector employees in the United
States.
3
Using a blend of data from the Current
Population Survey of the US Census Bureau and
the National Compensation Survey from the Bureau
of Labor Statistics, this analysis estimates that
46% of private sector workers lack access to an
employer-sponsored plan, representing about 57
million workers as of 2020 (see Figure 1.1).
4
This
gure is anticipated to grow to more than 64 million
by 2040 under the continuation of current trends.
Workers are much more likely to save for retirement
if they have access to an employer-sponsored
retirement savings plan. Although workers can
establish their own retirement savings accounts
if they lack such access, they rarely do so in
1. Closing the Signicant Gaps in Access to Retirement Savings
ESI analysis of Census Bureau Current Population Survey
and BLS National Compensation Survey Data.
Access to coverage at work
Coverage access gap
67.3M57.3M
46%
GAP
124.6 M
Private Sector Employees
Figure 1.1: More than 57 Million Employees Lack
Access to a Retirement Savings Plan in their
Workplace (2020)
10
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
practice, with workers 15 times more likely to
save for retirement if they have access to a payroll
deduction savings plan at work.
5
Workers at
rms that provide an employer-sponsored plan
are considered to have access to coverage,
although they may not choose to be participants.
For small businesses, the complexity, cost,
and perceived legal risk reduce the likelihood
they will oer a plan to their employees.
Programs that make access to savings easier
by connecting a worker to a savings account
and including design features such as automatic
enrollment and auto-escalation can signicantly
increase participation and savings levels.
6
The gaps in access to retirement savings plans
are greater among younger workers, women,
minorities, and lower income workers.
7
Access to
retirement savings plans also varies signicantly by
employer size and industry. Larger employers — for
example, those with more than 500 employees, and
in sectors paying higher wages — are more likely
to oer their workers retirement savings plans.
8
These dierences contribute to variations in access
among demographic groups and widen access
gaps among dierent segments of the population.
Too Many Have Little Saved for Retirement
These gaps in access have serious implications,
leaving many ill-prepared nancially for retirement.
While elderly Americans are supported by
Social Security, many elderly households fall
short of the income replacement standards
recommended to maintain the quality of life they
enjoyed during their working years. Even when
considering a generous measure of retirement
savings (net worth), more than three-quarters of
Americans fall short of conservative retirement
savings targets for their age and income level.
9
Putting Social Security in Context
Social Security is one of the key pillars of the
American retirement system, but was never
designed to meet all retirement income needs.
Social Security provides a basic retirement income
oor for retirees and should be supplemented
by employer-based and personal savings. In
2020, the average monthly Social Security retiree
benet was $1,503 per month for an individual,
equivalent to an annual income of just 1.4x the
Federal Poverty Level, or $2,531 for a couple.
10
Unfortunately, a signicant proportion of the retired
population in the US has come to rely on Social
Security for a material proportion, if not all, of their
retirement income. Among elderly Social Security
beneciaries, 70% of unmarried people receive half
or more of their income from Social Security, as do
50% of married couples. About 45% of unmarried
people rely on Social Security for 90% or more of
their income.
11
While a large share (42%) of the
baby boomer cohort expects to rely heavily on
Social Security as a source of income in retirement,
younger generations are expecting lower income
replacement from Social Security and to rely primarily
on self-funded savings for retirement income.
12
Shortfalls in Private Savings
The shift over time from employer-provided
pensions to dened contribution plans has put
greater responsibility on workers to ensure their
nancial well-being in retirement. However, even
Americans who have access to retirement savings
accounts often do not achieve sucient savings
levels to maintain their quality of life in retirement.
Researchers from the Center for Retirement
Research at Boston College report that median
account balances for 55- to 64-year-old working
households with incomes near the median are below
$100,000.
13
For lower income households who are
less likely to have access to retirement savings plans
through their employers, the retirement readiness
gap is even more stark. Among workers nearing
retirement with the lowest 20% of income, 79%
have no retirement account assets whatsoever.
14
Younger generations are also struggling to build
the foundational savings that will help support their
retirement readiness. Young savers face a range of
challenges, such as rising educational costs and
student loan debt burdens, challenges in securing
housing, and a cycle of economic challenges. Amidst
these challenges, two-thirds of working millennials
lack any retirement savings, raising concerns
about their long-term retirement readiness.
15
11
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
An Aging Population Increases the Urgency
Within this context, a rapidly aging population
and dierences between generations increase the
urgency to address retirement savings shortfalls.
Over the next decade, for example, the nal
wave of baby boomers will reach retirement
age, Generation X will approach retirement,
and millennials and increasingly Generation
Z will be in their prime working years.
The aging of the baby boomers continues a shift
that has been occurring for decades in the balance
between the retiree and working-age population.
The University of Virginia’s Weldon Cooper Center
projects the nation’s elderly population will increase
from 54 million in 2020 to 71 million by 2040, a
growth rate of 32%, about three times the rate of
the non-elderly population.
16
This also aects the
composition of US households, with households
headed by seniors anticipated to grow from 33
million in 2020 to 43 million in 2040, an increase of
more than three times the expected rate of growth
for working age households (see Figure 1.2).
17
As senior households grow in both in number
and as a share of the population, there will be
fewer working households to support the needs
of an elderly, non-working population. Census
Bureau data indicate that the “dependency ratio”
is currently falling rapidly from its historic norms
— from almost four working age households
for each elderly household in 2005 to a ratio of
closer to two to one by 2030 (see Figure 1.3).
Since working age households are the primary
contributors to the tax base, this falling dependency
ratio will create signicant scal pressures as
demand for benet programs increases. This
demographic shift makes the ability of elderly
households to maintain their living standards in
retirement an important economic and quality of
life issue for all US households. While considerable
focus has been placed on the future scal
solvency of Social Security and Medicare, several
means-tested programs like Medicaid and the
Supplemental Nutrition Assistance Program (SNAP)
also will see signicant increases in demand
if elderly households lack sucient income in
retirement. This also portends a lower economic
growth environment, with the workforce growing
at a slower rate than in prior generations.
Structural factors indicate that this shift in the
balance between retiree and younger households
is likely to reect a new normal. Increasing life
expectancy will help to grow the elderly population,
while younger generations show declining birth rates
and are having their rst children later in life (slowing
generational replacement cycles). Figure 1.4 shows
projected changes to the US “population pyramid”
by age and generation over the next two decades.
This shift in population composition underscores
the importance of enabling younger generations
Figure 1.3: ... While Falling Ratios of Working-Age
to Elderly Households Create Fiscal Pressure
ESI analysis of US Census Bureau data
and University of Virginia Population Projections.
3.9
3.8
3.6
3.6
3.8
3.9
3.7
3.2
2.8
2.5
2.3
2.3
2.3
2040203520302025202020152010200520001995199019851980
US Households (Millions)
2020
2040
7.2
44.9
48.4
43.4
6.3
40.0
45.9
32.8
65+45-6425-44<25
Age Group
+13%
+12%
+5%
+32%
Figure 1.2: Senior Households are
Growing as the Population Ages ...
ESI analysis of American Community Survey Data
and University of Virginia Population Projections.
12
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
like millennials and Gen Z to have opportunities to
build resources during their crucial savings years. By
2040, the millennial and Gen Z cohorts will occupy
the prime earnings years of ages 30 to 60 and will be
helping to supporting a larger retiree population than
ever before while trying to ensure their own nancial
futures. Enhancing the ability of these generations
to strengthen their nancial security is crucial to the
nation’s long-term economic health and prosperity.
1.2 Policy Approaches Taken to Close the
Access Gap
Policymakers in the United States have developed
and started to implement reforms intended to
close the gap in private sector retirement savings
access, encourage savings, and strengthen the
retirement readiness of workers. Such eorts are
not unique to the US, with other countries having
already adopted a mix of public and private
models to move toward universal access that
have resulted in signicant savings over time.
National universal access models have been
proposed by academics and policy experts, and
several legislative proposals have been introduced
in Congress over the past decade and as recently as
2019. In the absence of national action, several states
have adopted innovative public-private partnership
models to expand access requiring employers to
provide a retirement savings option for their workers.
A few of these new state programs have launched,
providing many employers and their employees with
new ways to save, and the number of new accounts
and assets are now growing at a steady pace.
18
At
the same time, recent Congressional action, such as
the SECURE Act (P.L. 116-94), intended to expand
the adoption and improve the design of dened
Millennials
and Gen Z
30 - 60
Years
Figure 1.4: Millennials and Gen Z will be in Prime Earnings and Savings Years by 2040
Post-Alpha Alpha Gen Z Millennials Gen X Boomers Silent
ESI analysis of American Community Survey Data and University of Virginia Population Projections
US Population (M)
Age Cohort
85+
80 to 84
75 to 79
70 to 74
65 to 69
60 to 64
55 to 59
50 to 54
45 to 49
40 to 44
35 to 39
30 to 34
25 to 29
20 to 24
15 to 19
10 to 14
5 to 9
0 to 4
2020 2040
13
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
contribution plans is another positive step.
19
However,
both of these initiatives are unlikely to achieve a
signicant national expansion of coverage and
savings.
International and state examples, as well as national
proposals advanced by legislators and policy experts,
provide several scenarios intended to achieve the
goal of universal access. This section reviews these
approaches and outlines a set of policy scenarios that
are modeled and analyzed in Part 2 of this study to
see how they expand access and boost retirement
income.
International Models Toward Universal Access
Several countries have launched programs to
provide universal workplace access to retirement
savings options. These programs feature a mix of
public and private structures for administration and
contributions. Common elements include automatic
features to help make enrollment and saving easier
for participants, which helps to build scale and
control costs.
Programs in Australia, New Zealand, and the United
Kingdom, for example, have gained signicant
scale over time, with millions of participants
and billions in assets under management (see
Figure 1.5). Their stories demonstrate the
sustainability of these types of programs and
their potential to build signicant wealth.
Australia: Superannuation Guarantee
Launched in 1992, the Superannuation Guarantee
in Australia requires employers to contribute
to a retirement savings account on behalf of
eligible employees. Employers are currently
required to contribute 9.5% of an employee’s
earnings to a superannuation (or “super”) fund
on behalf of workers above certain salary and
hours thresholds.
20
The guaranteed contribution
rate will rise to 12% by July 2025. Contributions
are not required for very-low-wage and part-
time workers. However, contributions made for
low- and middle-income workers are matched
by the national government up to a maximum
amount of $500 annually to help build assets.
21
Although most employees are free to determine
which fund they prefer their employers contribute to,
many allow default investment funds to be applied.
Funds can be organized by a nancial services
company, employer or industry group, or through
self-managed funds for ve people or fewer.
As of 2020, 16.7 million Australians held super-
accounts and super-fund assets totaled $2.9 trillion.
The average account balance of those with savings
in super-funds (non-zero balances) is approximately
$121,000 for women and $169,000 for men.
22
New Zealand: KiwiSaver
Launched in 2007, the KiwiSaver is a publicly
administered dened contribution system in New
Zealand. Participation is voluntary, but its auto-
enrollment feature requires that a worker must
opt out if they choose not to participate. Once an
account is created, it is portable among employers
and requires contributions from both employers
and employees. Employees set a contribution level
of 3% or higher of earnings, employers provide a
contribution of 3% of earnings, and the government
makes an additional “tax credit” contribution.
Early withdrawals are highly restricted before
the retirement age of 65, but employees may
be able to make early withdrawals of part (or
all) of their savings if they are buying a rst
home, moving overseas permanently, suering
signicant nancial hardship, or seriously ill. As
Figure 1.5: Employer-Based International
Savings Programs
Australia Superannuation Guarantee – 16.7 million
participants
Requires employers to contribute 9.5% of an eligible
employee’s earnings to a retirement savings account.
KiwiSaver – 3 million participants
Workers auto enrolled (can opt out) to contribute ≥
3% of earnings + 3% employer match and a tax credit
contribution.
UK NEST – 9 million participants
Uncovered workers auto enrolled (can opt out) at default
contribution levels of 5% employee + 3% employer.
14
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
14
of 2020, KiwiSaver has grown to more than 3
million participants and $62 billion in assets.
23
United Kingdom: National Employment Savings Trust
(NEST)
NEST is a dened contribution savings plan in the
United Kingdom that launched in 2012. It provides
individualized savings accounts to those who do
not have access to an employer-based plan. The
NEST program’s administration is funded through
fees on contributions, and program services
are contracted to private nancial providers by
the NEST board. Employers can participate in
private sector plans or use the NEST program,
which essentially functions as a public option.
NEST is required to take any employer, but the
self-employed are currently not covered.
Workers must be auto enrolled, and they can choose
from a set of investment options, including target
date funds, but there must be a default investment
option and fees are capped at 75 basis points.
Default contribution levels for the plan have grown
over time to 5% of earnings for the employee and
3% of earnings for the employer, totaling 8%. The
overall savings opt-out rate is about 10% and 99% of
participants stay with the default investment option.
As of March 2020, the program had grown to more
than 9 million participants, received $4.8 billion in
contributions, and had $9.5 billion in assets.
24
US Eorts Have Fallen Short of Universal Access
State Eorts to Enhance Savings
Due to the continued failure of Congress to take
action to close the access gap, several US states
are adopting simple, low-cost, easily accessible
ways for more private sector workers to save for
retirement. States are acting out of necessity.
They already understand that they face signicant
budgetary and economic consequences if their
residents retire with insucient retirement income.
As the population ages, states will be increasingly
pressed to deal with dramatic increases in the cost
of social service programs for seniors living at or
below the poverty line — namely, programs related
to healthcare, housing, food, and energy assistance.
ESI studies for task forces examining the issue
of insucient retirement savings in Pennsylvania
and Colorado have shown that the “cost of doing
nothing” for each of these states will amount
to several billion dollars in additional state
expenditures.
25
For a representative household
in Colorado, the study found that additional
savings of just over $100 a month over 30 years
could close the gap and achieve recommended
income replacement levels in retirement.
26
Recognizing the signicant costs of doing nothing,
states across the country have initiated a variety
of eorts aimed at helping private sector rms
overcome the barriers to oering retirement savings
options for their employees. Since 2012, at least 45
states have introduced legislation to either establish
a state-facilitated retirement program for private
sector workers or study the feasibility of establishing
one.
27
States are designing dierent models that
seek to address these issues (see Figure 1.6).
State-facilitated programs seek to establish the
program architecture and administration at a
statewide level, enabling employers to participate
with minimal eort. The state generally appoints a
board to develop program rules and contract with
administrative and investment managers. Program
funding is covered by fees to the participants, and
states do not subsidize the program ongoing or
assume nancial liability for investment outcomes.
There are currently three basic state models with
variations considered:
Auto-IRA: The most common approach for state-
facilitated programs has been a payroll deduction
“Auto-IRA” model. States facilitate a simple and
low-cost IRA program using automatic enrollment,
a voluntary enrollment mechanism where the saver
has complete control over participation in the
program and can opt out at any time or change
the default contribution level. All the employer
must do is provide basic employee information
to the program, and remit payroll deductions.
Employers who do not already have a plan of their
own would be required to facilitate the use of the
state program for their workers. These programs
seek to maximize participation and savings and
15
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
minimize fees through automatic features, simple
design in investment options, and scale. Active
programs in California (CalSavers), Illinois (Illinois
Secure Choice), and Oregon (OregonSaves)
follow this approach (see Figure 1.7).
Participation in the Auto-IRA programs is typically
required for private rms meeting certain criteria
(such as a size threshold) if they do not already
oer their employees a qualifying alternative.
However, a program also can be structured as a
voluntary payroll deduction option that is voluntary
for both employers and employees, requiring opt-
in. This approach has been adopted, but not yet
implemented, by New York and New Mexico.
Multiple Employer Plans (MEPs): A MEP is
essentially a 401(k) used by several businesses
that join together to oer a common plan to each
employer’s workforce, pooling their resources and
outsourcing plan management. Because 401(k) plans
are ERISA plans, participation by employers must
be voluntary.
28
This scenario is currently operating
in Massachusetts and will soon launch in Vermont.
Marketplace: A marketplace by design is
a voluntary platform that creates a single
“clearinghouse” for private sector providers to
oer plans. It enables small businesses to nd and
compare retirement savings plans in an apples-
to-apples manner. It presents a diverse array of
plans (IRAs and 401(k)s) pre-screened by the state
to ensure certain standards are met. This reduces
search costs for employers, allowing them to rely
on the state to establish certain standards for plans
oered and ensure that the oerings meet those
standards. This model is currently operating in
Washington.
Hybrid: In addition to these three basic models,
states also have considered combining these models
to create a “hybrid” version of a program. New
Mexico is the rst state to adopt a hybrid model
that includes both a voluntary payroll deduction IRA
and a marketplace. Other options considered but
not yet adopted include oering both an Auto-IRA
and a MEP or combining all three approaches.
To date, international experience with the KiwiSavers
program in New Zealand and early state experiences
with MEPs and marketplaces in the US suggest that
it is much more challenging for a voluntary program
to achieve signicant reductions in the access gap.
29
Figure 1.7: Recently Launched State Auto-IRA
Programs
OregonSaves – Launched 2017
Auto-IRA program required for all employers without an
existing qualied plan, 5% default employee contribution
with auto-escalation, and no employer match permitted.
Illinois Secure Choice – Launched 2018
Auto-IRA program required for employers with ≥ 25
employees without an existing qualied plan, 5% default
employee contribution, and no employer match permitted.
CalSavers – Launched 2019
Auto-IRA program required for employers with ≥ 5
employees without an existing qualied plan, 5% default
employee contribution with auto-escalation, and no
employer match permitted.
Figure 1.6: State-Facilitated Retirement Savings Models Adopted to Date
Auto-IRA
(Secure Choice)
Voluntary IRA
Voluntary
Marketplace
Voluntary Open Multiple
Employer Plan (MEP)
California (active)
Illinois (active)
Oregon (active)
Colorado
Connecticut
Maryland
New Jersey
New York Washington (active) Massachusetts (active)
Vermont
New Mexico (hybrid)
16
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
SECURE Act
In December 2019, Congress passed the Setting
Every Community Up for Retirement Enhancement
(SECURE) Act (P.L. 116-94) to allow small business-
es to provide retirement savings options for their
employees using Multiple-Employer Plan (MEP)
and Pooled Employer Plan (PEP) arrangements.
30
A MEP structure allows and makes it easier for
related businesses to join together in a single
retirement plan. MEPs are organized and run by
a sponsoring entity (which may or may not be
a dened membership organization, such as an
industry association) that is responsible for admin-
istrative duties and takes on duciary liability for the
plan. The SECURE Act made this structure easier to
establish and more appealing, by reducing auditing
requirements and eliminating the “one bad apple”
rule, where compliance failures of one employer
could disqualify an entire plan. A PEP structure,
introduced under the SECURE Act, allows unrelated
employers to band together to form a single retire-
ment plan through a 401(k) savings vehicle. This
structure envisions common administration through
a third-party administrator or bundled record-keeper.
These new provisions, eective as of 2021, give
small employers additional options that may reduce
costs to plan participants through increased scale.
A recent analysis by Morningstar found that “MEP
fees decrease and become more predictable as
plans grow,” with decreases in fees per participant as
assets grow outweighing smaller increases in admin-
istrative cost per participant as the number of par-
ticipants grows.
31
This suggests that expanded MEP
and PEP availability could provide better options for
rms that currently operate on single employer plans.
However, fees for MEP plans that fail to achieve
signicant scale often remain high and may lack
transparency due to limited reporting requirements
for smaller plans. While these provisions are con-
structive steps, MEPs and PEPs are unlikely to mate-
rially reduce the access gap if they remain voluntary.
The SECURE Act also includes provisions to help
savers plan for and manage their savings once they
retire, in the form of a monthly income throughout
their lifetimes. The Act requires, for the rst time,
that statements to plan participants include in-
formation about the monthly income their current
savings would generate in retirement. The SECURE
Act also makes it easier from a regulatory stand-
point for plan providers to oer lifetime income
solutions (annuities). These provisions reect an
increasing emphasis on improving dened contri-
bution plans as lifetime income-generating plans
to support a better quality of life in retirement.
In October 2020, House Ways and Means Committee
Chairman, Representative Richard Neal and Ranking
Member, Representative Kevin Brady introduced
the Securing a Strong Retirement Act of 2020 – a
“SECURE ACT 2.0.”
32
This bipartisan bill builds on
the goals of the SECURE Act, with a number of
additional measures to increase options and protec-
tions for savers and retirees. Among its provisions,
it would require certain newly created plans to
automatically enroll eligible employees at automat-
ically escalating contribution levels, with voluntary
employee opt-out of coverage. The legislation
includes nancial incentives for small businesses to
oer retirement plans and expands savings options
for nonprots. Other provisions in SECURE Act 2.0
increase exibility for savers over 60 as they near
retirement and extend the time individuals can save
by increasing the minimum distribution age to 75.
SECURE Act 2.0 also aims to support low-income
earners to save by enhancing the existing Saver’s
Credit — a federal tax credit for contributions to a
retirement plan. While these measures and those in
the original SECURE Act are steps toward improving
access and savings levels, they are not expected
to signicantly reduce the national access gap.
33
Saver’s Tax Credit
The Saver’s Tax Credit (“Saver’s Credit”) was cre-
ated by Congress in 2001 to encourage savings by
low- and moderate-income taxpayers. Structured
as a tax credit on federal income tax liability, the
Saver’s Credit provides an incentive to save through
its value as a “match” to lower-income savers’
retirement contributions. The amount of the current
credit is based on a taxpayer’s income level, with
the lowest-income earners eligible for a 50% match
to their savings contributions and credit amounts
falling to 20%, 10%, and 0% (above the highest
17
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
income threshold) as income rises. The maximum
credit amount is capped at $1,000 for an individual
or $2,000 for a married couple ling together.
34
Due to its current structure and administration, the
Saver’s Credit has been underused, limiting its po-
tential to aect the savings behavior of lower-income
households. According to a study by AARP, 9.3%
of returns were eligible for the credit in 2013, but
just 5% claimed it.
35
The credit is complex to apply
for, and credit amounts decrease sharply as income
increases, meaning a small increase in a ler’s in-
come can lead to a signicant decrease in their credit
amount. Importantly, because the credit is non-re-
fundable, households must have an income tax
liability to realize the gains, making many of the low-
est-income earners ineligible. According to a 2006
study from the Congressional Budget Oce, 18% of
lers met the income criteria for the credit but were
ineligible because they had no income tax liability.
36
Many policymakers have suggested enhancements
to the credit to increase its impact for lower-income
savers. The Brookings Institution has suggested
increasing matching rates, increasing eligibility
limits, and making the credit refundable.
37
AARP
has suggested making the credit a savings match
into retirement accounts, and restricting the ability
of savers to withdraw those funds, in addition to
simplifying the ling process and increasing the
eligibility limits and match amounts.
38
Broader reform
proposals have linked the saver’s credit to other
components of the tax code, such as the mortgage
interest deduction, proposing a at credit for all sav-
ings.
39
The proposed SECURE Act 2.0 bill includes
an expansion of the credit, including a higher max-
imum credit amount, increased maximum income
eligibility, and single credit rate rather than a tiered
rate structure.
40
The legislation does not, however,
structure the credit as refundable or institute it as a
deposit directly into retirement savings accounts.
Fintech Can Help Remove Barriers
In addition to policy innovations, the private mar-
ket is responding to changes in the landscape
with the creation of more data-driven technol-
ogy companies focused on providing improved
nancial engagement and performance.
Developing the right technology platforms and
the correct messages can help people under-
stand and use customized tools and products.
Surveys suggest that consumers, particularly
millennials and younger generations, are much
more comfortable with technology companies
as a vehicle for acquiring nancial products.
41
Entrepreneurial nancial technology (“ntech”) rms
deploy technology in innovative new ways that reach
all workers more eectively, including previously
underserved communities, to help them save and
invest for their futures. Advances in technology
focused on nancial applications represent another
potential path to lower cost and complexity and
increased retirement security. The best-known com-
ponent of this approach is through “robo advisors”
that use computerized algorithms to provide nancial
advice and manage portfolios. As these technologies
evolve, they have the potential to provide sound
advice about a broader set of nancial management
strategies, including decumulation, at low cost.
42
These approaches can help build on the initial
wave of digitization and behavioral nudges, such
as auto-enrollment, that have helped increase
quality and lower costs within retirement savings
plans. These eorts still face limitations in expand-
ing access, an uncertain regulatory environment,
and challenges in consumer comfort with these
technologies, with many providers moving to-
ward a hybrid robo and in-person approach.
Technology can be a tremendous asset when tailor-
ing and customizing plans to meet an individual’s
goals and long-term needs, but the rise of apps and
broader societal concerns about platform data col-
lection, and how that information is used or shared,
can potentially be an issue for privacy-minded users.
National Proposals for Universal Access
Congressional Legislation
Over the past several years, Congress has intro-
duced several legislative proposals to expand ac-
cess to retirement savings for private sector workers.
The leading proposals discussed in this report have
proposed requiring employers to make available
18
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
either Auto-IRAs or 401(k) plans to their employ-
ees if they do not already oer a qualied plan.
• The “Automatic IRA Act of 2019” (S. 2370)
introduced by Senator Whitehouse in the
116th Congress would require businesses with
10 or more employees that do not currently
have a plan to oer a payroll deduction IRA.
• This bill is similar to an earlier version
of an Auto-IRA proposal introduced by
current US House Ways and Means
Committee Chairman Richard Neal in the
115th Congress in 2017 (H.R. 3499) and
the 114th Congress in 2015 (H.R. 506).
• Chairman Neal also introduced, in the
115th Congress, the “Automatic Retirement
Plan Act of 2017” (H.R. 4523), which es-
tablishes a requirement that employers
that do not already have a qualied retire-
ment plan adopt a 401(k)-type plan.
While the type of savings account may dier,
these proposals are similar regarding requiring
employer participation for rms with 10 or more
employees, automatic enrollment, protecting cur-
rent state-level programs, and other features.
There have also been other legislative proposals
intended to expand access to retirement savings.
These approaches include automatically enroll-
ing uncovered or contract workers in a national
retirement savings plan modeled on the Thrift
Savings Plan oered to federal workers (“American
Savings Account Act”), establishing a universal
savings account plan for uncovered workers with
a mandatory employer contribution (“American
Savings Act”), establishing national retirement
fund options to be made automatically available to
employees of rms without qualifying plans (“USA
Retirement Funds Act”), and creating a portable
account at birth that would follow workers from
job to job throughout their careers (“Portable
Retirement and Investment Account Act”).
43
National Reform Proposals
Many of the Congressional proposals have been
based, in whole or in part, on proposals developed
by academic and policy experts. A few of the
more widely considered examples include:
• The concept of a national Auto-IRA was rst
proposed in 2006 by David John of the Heritage
Foundation and Mark Iwry of the Brookings
Institution.
44
This bipartisan eort gained the
support of both presidential campaigns in 2008,
maintains appeal to policymakers across the
ideological spectrum, and is currently the most
commonly adopted model at the state level.
• The Bipartisan Policy Center convened a
Commission on Retirement Security and
Personal Savings that issued a 2016 report
recommending that employers with 50 or more
employees that do not already oer a plan
should be required to oer their employees a
national Retirement Security Plan administered
by a third party or the federal myRA program,
which has since been terminated.
45
Smaller
employers would have the option to participate
in the national plan. The report also includes
additional recommendations to enhance
savings and increase retirement security.
• The Center for American Progress developed
a blueprint in 2016 for a “National Savings
Plan” modeled on the Thrift Savings Plan (a
401(k) plan open to federal employees and
members of Congress).
46
This plan featured
an initial default employee contribution of 3%
(with auto-escalation to higher levels) into
an appropriate life cycle fund, with optional
employee contributions at a universal dollar
amount or percentage for all employees.
• The Economic Policy Institute advocates for
a “Guaranteed Retirement Account” pro-
posed by New York University’s Economist
Theresa Ghilarducci and Blackstone President
Hamilton James.
47
This approach mandates
an employer and employee contribution
for all, regardless of current retirement
plan status, and extends the model to all
workers, including the self-employed.
For several years now, presidential campaigns
representing both political parties have also
19
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
made references in their platforms to either
Auto-IRAs or 401(k) national models, or models
similar to the federal Thrift Savings Plan (TSP).
Analyzing How Dierent Design Options Aect
Access and Savings
Drawing on the range of state, national, and in-
ternational programs and proposals, this study
analyzes a “baseline” universal access option, the
payroll deduction Auto-IRA, and then builds upon
this baseline scenario by examining a number of
alternative policy options. National impacts are
modeled over a 20-year time horizon, with results
shown as of 2040 (with all results reported in 2020
dollars to allow for easy comparison of impacts
through the analysis). Notably, this time frame does
not capture the full benet for younger workers,
who will derive the biggest gains from a career of
accumulating retirement savings but will realize
those gains in retirement beyond the year 2040.
Modeling a baseline universal access retirement
savings scenario and then examining the impact of
varying design features and options allows an explo-
ration of the potential implications of policy variations
for access; savings accumulations; retirement
income; and the longer-term impacts on the econ-
omy, federal expenditures, and tax revenues. This
can provide information helpful to decisions about
future US retirement system reforms focused on
expanding access and enhancing retirement security.
Baseline Universal Access Design Option:
The Payroll Deduction Auto-IRA
A “baseline” universal access scenario is dened
in this study by following the Auto-IRA model
adopted in several states and implemented on
a universal basis in Oregon. This report uses a
payroll deduction Auto-IRA structure requiring the
participation of employers of all sizes as a baseline
because it is comprehensive in expanding access
and simple in its structure and implementation.
The baseline Auto-IRA scenario is dened as:
• All rms required to provide cov-
erage to their employees;
• Automatic enrollment with an ability for
employees to opt-out;
• A Roth-IRA savings account, with contributions
made post-tax (no tax on qualied withdraw-
als of account contributions or earnings);
• 5% default initial employee contribution, with
an auto-escalation of 1% per year up to 10%;
• No employer contribution;
• Coverage requirements implemented in
three phases by employer size, starting
with the largest employers (100 or more
employees) two years after enactment
(assumed to be December 31, 2021) and
covering mid-size employers (20–99 em-
ployees) three years after enactment and
all rms within four years of enactment;
• Existing state Auto-IRA programs
“grandfathered in” (consistent with federal
legislative proposal);
• Enhancement of the Saver’s Credit to in-
corporate the higher income limits, credit
amount, and credit maximum reected in
the SECURE Act 2.0 legislative proposal,
as well as a refundable structure providing
for matching funds to be deposited directly
into savings accounts (a component not
included in the SECURE Act 2.0 proposal).
Section 2.1 details the potential impacts of this
baseline approach in terms of coverage and re-
tirement assets for individual savers and at the
national level. Analysis is included that isolates
the impact of the enhanced Saver’s Credit, which
is included in modeling across all scenarios, on
savings accumulations for representative savers.
Other Policy Design Options Modeled
Building from this baseline scenario, a number of
alternative policy options are modeled to under-
stand their impacts on participation, savings, and
government expenditures and revenues. The policy
options and features modeled broadly reect the
range of national legislative proposals discussed.
This analysis is not intended to designate a single
20
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
proposal as optimal, but rather to provide policy-
makers with insight into the dierential impacts
of policy variations on access and savings.
Alternative policy options are modeled by ad-
justing several design features, including:
• Varying the type of savings account used
between a payroll deduction IRA and 401(k),
factoring in dierences in the administrative
requirements and the costs of such accounts;
• Adding employer size and age thresholds,
exempting the smallest and youngest busi-
nesses from the requirement to provide their
employees with access to retirement savings;
• Including a voluntary employer contri-
bution, as permitted in 401(k) accounts
to give businesses the discretion to con-
tribute to employee accounts; and
• Requiring an employer contribution by
adding a new requirement for employers to
provide contributions into an employee’s 401(k)
account, improving the return on investment
for savers but generating additional economic
implications for businesses and workers.
These policy variations are applied in sequence,
producing four modeled scenarios (see Figure
1.8 above). Modeled results for each scenario
are compared to the baseline Auto-IRA scenario
outlined above to understand the implications of
dierent policy options for participation and savings
levels. Next, Section 3 details the long-term national
impacts of the baseline scenario and each policy
option for economic growth, tax revenues, and
government expenditures on benet programs.
This analysis assumes stable employment condi-
tions and a lower-growth economic environment
across all scenarios (consistent with projections
from the Congressional Budget Oce).
48
Continuity
is assumed in federal policy outside the specied
initiatives concerning retirement security. Employers
are assumed to build up to full compliance with the
requirements of the policy options over time.
49
Additional information about the modeling
approach used in this study is included in
a supporting Methodology Appendix.
50
Figure 1.8: Modeled Scenarios Isolate the Impact of Policy Variations on Access and Savings
√
√
√
√
√
√
√
√
√
√
√
√
√
√
21
What are the Potential Benets of Universal Access to Retirement Savings?© 2020 Georgetown University Center for Retirment Initiatives
2.1 Participation, Savings, and Assets
under a Baseline Auto-IRA Scenario
National universal access to retirement savings
for private sector workers means opportunities for
millions of lower- and middle-income households
to build wealth over time and signicantly
boost income in retirement through simple,
automatic, and consistent contributions.
By starting to save early in their careers, taking
advantage of available incentives to save, and
beneting from compounding investment returns
over an extended time horizon, low- and moderate-
income workers can generate meaningful account
balances by end of their working careers. Modeling
potential returns for representative participants
following the default contributions in the payroll
deduction Auto-IRA baseline scenario covering
all employers (“baseline Auto-IRA”) shows how
workers could generate meaningful assets to
supplement other sources of income, like Social
Security, to enhance their retirement security.
Design features in the baseline Auto-IRA scenario,
such as auto-enrollment and auto-escalation
of contribution levels, would expand access
and help participants build savings. Protecting
these savings by minimizing fees and leakage
and capitalizing on market returns is crucial to
growing the balances available to participants
as they approach retirement. Modeling of the
national impacts of the baseline Auto-IRA design,
accounting for mitigating factors like opt-outs,
discontinued accounts, fees, and early withdrawals,
still shows signicant potential increases in access,
savings, and asset levels among workers.
Starting Sooner and Saving Longer Signicantly
Improves Retirement Outcomes
Beginning to save as soon as possible and
saving consistently for as long as possible
makes an enormous dierence in a worker’s
ability to build savings and convert those savings
into retirement income when it is needed.
Since national impacts are modeled over a 20-year
time horizon (through 2040), younger workers from
the millennial and Gen Z cohorts who will not yet
have reached retirement age within the study period.
However, these younger workers will have greater
opportunities to build assets through continued
contributions and additional years of compounding
growth. Extending the analysis of individual savers
to a full career illustrates the potential benets of
universal access for future generations, and the
power of starting sooner and saving longer.
An examination of three savings scenarios
illustrates dierences in savings and retirement
income for participants based on the starting
age, years of participation, and employer
size. Using rates of contributions and returns
drawn from the Auto-IRA baseline scenario,
three representative savers are modeled:
• A “young saver” starting their account at
age 25 and earning the average salary at
a small employer over a 40-year career.
These workers could generate substantial
returns even with modest earnings, through
consistent contributions and compounding
returns over the span of their careers.
• A “mid-career saver” starting at age 35 and
earning the average salary for a mid-size
rm over the remaining 30 years of their
career. These workers have fewer years
of accumulation than those starting at the
beginning of their careers, but may be able
to support higher contribution levels to
generate assets to supplement other forms of
retirement income, such as Social Security.
• An “older saver” starting at age 45 and
earning the average salary for a small
employer over the remaining 20 years of
their career. These workers have missed
important accumulation years, but still
could accumulate material assets for their
retirement. These savings can still improve
retirement security and help delay the start of
drawing Social Security benets, which would
increase benet levels in later years of life.
51
2. Analyzing the Potential Benets of National Universal Access to
Retirement Savings Options
22
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
Figures 2.1 to 2.3 show the total contributions,
rate of return on those contributions, assets at
retirement, and annual annuity amounts supported
by these assets for the representative saving
scenarios, assuming an enhanced Saver’s Tax
Credit. For example, a young (25-year-old) saver
with modest earnings levels of around $35,000
per year contributing at the default level (5%,
auto-escalating up to a cap of 10%) envisioned
in the baseline Roth Auto-IRA design would make
contributions of about $110,000 over a 40-year
period, and see their account grow to $390,000
in assets. If they use this lump sum to purchase
an immediate xed annuity at the age of 65, it
would generate an annual supplemental income
stream of $21,300 per year over the remainder
of the saver’s lifetime. Parallel calculations are
shown for the mid-career and older savers.
Isolating the Impact of the Enhanced Saver’s Credit
A substantial component of the return on investment
from the perspective of a saver can come from
matching funds with their initial contributions. For a
many savers using a 401(k) model through a private
sector employer or a government-sponsored plan,
these supplements may come from an employer
match. Another means to enhance returns and
incentivize contributions is the Saver’s Tax Credit.
As currently designed, the Saver’s Credit osets
federal tax liabilities up to 50% of contribution levels
for qualifying low- and moderate-income savers,
but does not directly supplement retirement savings
accounts. This analysis includes a refundable credit
structured as a matching contribution into a sav-
ings account. This approach greatly enhances the
return on investment for savers — both immediately
through the match of up to 50% of contributions, and
to an even-greater degree over time through market
returns.
52
The impact for a representative household of the
Saver’s Credit under this enhanced design can be
isolated by estimating the asset accumulations for
the representative savers (under the baseline Auto-
IRA approach) modeled above, with and without this
enhancement:
• Returns to a young Roth Auto-IRA saver on
contributions of $110,000 across their career
grow from 138% to 255% due to the Saver’s
Credit, supporting an annual income stream
of $21,300 with the credit, compared to about
$14,300 without (see Figure 2.4).
• Similar proportional increases are observed in
the other savings examples, with the mid-ca-
reer saver experiencing an annual increase in
supplemental retirement income from $11,500
to $15,900 due to the enhanced Saver’s Credit,
and the older saver seeing an increase from
$5,500 to $7,800.
Figure 2.1, Figure 2.2, and Figure 2.3:
Supplemental Lifetime Income at Age 65 for an
Auto-IRA Saver with Enhanced Saver’s Tax Credit
Fig. 2.1
:
Returns to a Young Saver
$21,300 Annual Annuity
6555453525
$0
$100000
$200000
$300000
$400000
$390,456
Assets
$110,122
Contributions
255%
Return
Assets
Contributions
Fig. 2.2
:
Returns to a Mid-Career Saver
$15,930 Annual Annuity
6555453525
$0
$100000
$200000
$300000
$400000
$291,914
Assets
$113,871
Contributions
156%
Return
Fig. 2.3
:
Returns to an Older Saver
$7,780 Annual Annuity
Fig. 2.3
:
Returns to an Older Saver
$7,780 Annual Annuity
6555453525
$0
$100000
$200000
$300000
$400000
$142,615
Assets
$69,215
Contributions
106%
Return
$142,615
Assets
$69,215
Contributions
106%
Return
23
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
Isolating the Impact of Starting to Save Early
The passage of time and the power of compound
interest also boost savings. Savings are invested
and as the market grows, so do account balances.
Future market returns apply not only to initial
contributions, but also to the market returns already
achieved. This compounding dynamic means that
encouraging savings at a younger age can have
signicant long-term payos for participants.
This compounding eect holds even when savers are
not able to contribute to their accounts throughout
their entire careers. Many workers face growing
nancial pressures over time, leading them to
allocate their income to costs like childcare and
educational savings for their children, mortgage
payments, or support for aging family members.
Workers who cease contributions within their careers
for these or other reasons (such as a change in their
employment situation) still enjoy further market-
driven benets from their initial contributions as
balances built up in early years continue to grow
throughout the duration of a saver’s working years.
The impact of starting savings early for a
representative household is isolated by comparing
two households with identical earnings (around
$35,000) and identical contributions (totaling
around $54,000) made over 20 years out of a
40-year career, using the baseline Auto-IRA
savings model in this analysis (see Figure 2.5).
• The “early saver” contributes from ages 25–45,
and then ceases contributions but maintains
their account. Assets continue to grow to about
$282,000 by age 65, supporting an annual
income stream of around $15,400 in retirement.
• The “late saver” contributes from ages 45–65,
making an identical total contribution throughout
the overall period as the early saver. However,
due to the shortened period of compounding
growth, this saver achieves a balance of
around $117,000, supporting an annual income
stream of around $6,400 in retirement. This
represents less than half of the retirement
resources achieved by the young saver despite
an identical level of total contributions.
How a Payroll Deduction Auto-IRA Expands
Access and Builds Savings
Universal access would signicantly expand the
number of households saving for retirement.
Because employees can choose to opt out, no
scenario will achieve 100% participation by all
eligible workers. However, the design will have an
impact on levels of participation and savings, with
default settings playing a particularly important role.
Modeling the baseline Auto-IRA scenario generates
estimates of participation and contribution levels
among the population of workers currently
lacking access to a savings plan. In this baseline
scenario, only employers who do not already
have a qualied retirement plan would have
to meet the requirements to oer coverage,
and any existing state-level programs enacted
as of 2020 would continue uninterrupted.
$0
$75,000
$150,000
$225,000
$300,000
654525
Assets (Early Saver):
$281,603
Annual Annuity: $15,630
+ $164,775
Assets (Late Saver):
$116,828
Annual Annuity: $6,370
Contributions (Early Saver)
Contributions (Late Saver)
$53,809
Figure 2.5: Starting to Save Early Dramatically
Increases Account Balances
Figure 2.4: An Enhanced Saver’s Credit Increases
Lifetime Returns to a Young Auto-IRA Saver
6555453525
Assets w/ Credit: $390,456
Annual Annuity: $21,300
Assets w/o Credit: $262,427
Annual Annuity: $14,320
Contributions: $110,122
$0
$100,000
$200,000
$300,000
$400,000
24
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
Auto-Enrollment Signicantly Boosts Participation
Research shows that automatic enrollment of
employees (with the option to opt-out) produces
signicantly higher participation rates than models
that require active opt-ins.
53
Participation rates
among workers with access also vary signicant by
employee and rm type, with the lowest participation
among the youngest workers (those below age 25)
and higher participation among workers at larger
rms.
54
Recognizing that not all groups will behave
the same way, opt-out estimates are dierentiated
by age and employer size. The average opt-out
rate among employees gaining access is estimated
at 30% for the purposes of this analysis, based
on early participation data from state programs.
Scenarios are modeled to include a phased
implementation, with large rms (100+ employees)
required to provide coverage in 2024 (or two
years after the potential enactment), mid-size
rms (20–99 employees) in 2025, and all rms
by 2026. Full compliance with the requirement
is assumed after the initial phase-in period.
Based on these assumptions, the analysis illustrates:
• Signicant expansions in access and
participation. By the required implementation
year of 2026, more than 35 million workers
are expected to be saving. By the year 2040,
participation is expanded to more than 40
million, or nearly 70% of the population
of nearly 59 million private sector workers
under the age of 65 who would otherwise
lack access to coverage (see Figure 2.6).
• Growth in the number of workers with
retirement savings accounts. Total accounts
among workers under 65 will grow beyond
the implementation phase, as normal turnover
within the labor force results in a substantial
number of workers who have account
balances that continue to grow through
investment returns, even though they are not
contributing in a given year. Total accounts
are estimated to grow to around 71 million
by 2040 (of which about 40 million are active
contributors, and 31 million are accounts
growing through investment returns only).
• More workers with additional savings at
age 65. Each year, a subset of participants will
reach the age of 65 (the assumed retirement
age in this analysis).
55
This cohort totals an
estimated 9.6 million by the year 2040, about
two-thirds of the 14.5 million workers currently
lacking access who are expected to reach the
age of 65 over this time period (see Figure 2.7).
Workers who reach this age having participated
and saved will have additional resources to
support their quality of life in retirement.
Figure 2.7: … Allowing Millions of Workers to
Arrive at Age 65 with Additional Private Savings
0.0
2.5
5.0
7.5
10.0
12.5
15.0
9.6
14.5
Retirees (Millions)
Workers
w/o Access
Reaching 65
Participants
Reaching 65
2040
2039
2038
2037
2036
2035
2034
2033
2032
2031
2030
2029
2028
2027
2026
2025
2024
2022
2023
Figure 2.6: A Universal Access Auto-IRA Scenario
Signicantly Closes the Access Gap …
Access Gap
(Workers <65)
53.3
58.6
Active Program
Participants
Implementation
Period
54.3
35.4
40.4
Participants (Millions)
0
10
20
30
40
50
60
2040
2039
2038
2037
2036
2035
2034
2033
2032
2031
2030
2029
2028
2027
2026
2025
2024
2023
2022
25
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
While the analysis period ends in 2040, there
are nearly eight account holders under 65 for
each participant that has reached 65 as of that
year. This ratio is indicative of the substantial
benets that would be achieved among future
generations of retirees in the years beyond 2040.
Building Savings Account Contributions
Default plan features are similarly important to
encouraging contributions once workers are
enrolled. The default initial contribution levels in
state programs in Oregon, Illinois, and California
have been set to 5% of post-tax earnings (due to
the Roth IRA default structure of those programs),
and the majority of participants have used this
default contribution level as a starting point.
Notably, several state feasibility studies explored
potential negative eects of initial default rates on
participation, and states like California, Illinois, and
Connecticut concluded, through survey research and
academic literature, that a default rate of 5% or 6%
is unlikely to discourage participation relative to a
default rate of 3%.
56
This higher savings rate has a
signicant impact on how much the individual would
accumulate over their savings lifetime.
57
Design also typically includes the auto-escalation of
the contribution over time. These auto-escalations
typically take place annually as a percentage
increase in contributions as a share of income, and
are ideally aligned with increases in earnings over
time, so workers do not experience a decrease
in take-home pay. Rather, they see a small share
of pay increases devoted to additional savings.
Early experience with OregonSaves and CalSavers
suggests the vast majority of savers accept their
annual auto-escalation.
For purposes of analysis, the default initial employee
contribution level in the baseline Auto-IRA scenario
is modeled at 5%, with an auto-escalation of 1%
each year of participation, up to a total of 10%. A
degree of opt-outs is anticipated from each stage of
escalation, and new savers continuously start at the
initial contribution level, leading to a “blended rate”
among participants at any point in time that is below
the maximum auto-escalation rate of 10%.
Employee contribution amounts are a function of
both the contribution level (in percentage terms) and
the post-tax earnings of the participant. Based on
data from state programs, the set of participants
currently lacking coverage is anticipated to have
somewhat lower average incomes than the overall
population of private sector workers. Incomes are
modeled using expected patterns in participation
by age and employer size, with higher earnings for
older workers and those at larger employers. Pre-tax
incomes for program participants are estimated to be
approximately $38,000 on average, which translates
to post-tax earnings of around $31,700.
58
Contributions are assumed to made into a Roth
IRA account in the baseline scenario. Under a Roth
structure, contribution percentages are applied to
“post-tax” or “take-home” earnings, as opposed
to the “pre-tax” contributions in a traditional IRA
or 401(k) structure, which create tax implications
at the point of withdrawal. The Roth structure is
simpler from the perspective of the saver and has
been the preferred approach in most state Auto-IRA
programs.
59
Post-tax earnings are combined with the contribution
rate to estimate employee contributions in dollar
terms. Adjustments are made to account for the
small proportion of savers who would otherwise
exceed annual IRA contribution limits ($6,000 for
employees under 50 and $7,000 for employees
50 and older in 2020), based on their anticipated
earnings and contribution percentage.
Employee contributions are supplemented by annual
contributions through an enhanced, refundable
Saver’s Tax Credit. Saver’s Tax Credit amounts
are modeled based on the applicable share of
contributions (50% for most savers, up to a cap of
$3,000). Employee contribution and Saver’s Credit
amounts are summed into a total annual contribution
for participants.
Based on these assumptions, the analysis illustrates:
• Employee contribution growth over time
with auto-escalation. Average employee
contributions grow from $1,880 in 2026 (after
the phase-in period) to around $2,600 by 2030.
This average contribution stabilizes with the
26
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
cap in auto-escalation, job turnover, and the
assumption of stable earnings patterns in real
terms.
• Enhanced savings from the Saver’s Credit.
Average deposits from the Saver’s Credits
grow to around $650 per participant by 2040.
Through this enhancement, the $2,630 in
employee contributions grows to around $3,280
in total contributions for the average participant
in 2040 (see Figure 2.8).
• Signicant annual savings levels. Across the
population of participants, annual employee
contributions and Saver’s Credits are estimat-
ed to total $132 billion annually by 2040 (see
Figure 2.9).
Figure 2.8: Employee Contributions Augmented by Saver’s Credits Total Nearly $3,300 per Year for the
Average Participant under the Baseline Auto-IRA …
Average Annual Contributions
$0
$500
$1000
$1500
$2000
$2500
$3000
$3500
2040203920382037203620352034203320322031203020292028202720262025202420232022
$2,349
$465
$1,884
$3,276 Total Contributions
$646 Saver’s Credit
$2,629 Employee Contribution
$132.4 B Total Contributions
$26.1 B Saver’s Credit
$106.3 B Employee Contribution
Figure 2.9: … and More than $130 Billion per Year in Total
$0
$30
$60
$90
$120
$150
2040203920382037203620352034203320322031203020292028202720262025202420232022
Annual Deposits ($ Billions)
27
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
Protecting Savings is Critical to Asset Growth and
Retirement Income
If the long-term goal of expanding access and
participation is to boost savings to build retirement
income, design considerations also must encourage
and support consistent contributions, low fees, and
avoiding withdrawals. These components will allow
compounding returns over time to accumulate signif-
icant assets for the average participant. Minimizing
fees and leakage, and using default investment op-
tions, such as target date funds, can help to achieve
these objectives.
Minimizing Fees and Leakage
With universal access to retirement savings required,
savers would still pay fees for administration and in-
vestment management, just as savers typically do in
existing employer-sponsored retirement plans. Fees
eectively reduce investment returns, meaning that
minimizing fees can be as important as maximizing
gains to asset growth. Large-scale plans or programs
can provide signicantly lower fees than those avail-
able to small individual providers. Approaches that
oer straightforward design structures and limited
options are also able to limit costs more eectively.
60
Fees in the baseline universal access Auto-IRA
scenario are modeled to start at 0.90% of assets and
decline to 0.35% over time as assets grow, below
private sector benchmarks of more than 1.00%.
61
Some universal access structures rely largely on a
single provider, achieving scale to help control costs,
while others divide the market through various mech-
anisms. In that case, a combination of scale and
competition between providers has a similar eect in
achieving reductions in cost.
Leakage through early withdrawals is a risk that can
create challenges for building savings. The Roth
IRA investment vehicle envisioned in the baseline
scenario does make early withdrawals easier, when
compared to traditional IRAs and 401(k)s, by allowing
for the withdrawal of post-tax contributions and, in
some cases, earnings with limited penalties.
62
In some cases, a household may need to tap its
retirement savings account to cover unexpected
expenses that would otherwise have negative
consequences. The availability of such savings can
create a buer that allows households to avoid nan-
cially damaging options, such as taking unfavorable
loans or eroding their credit in response to short-
term nancial shocks, which may lead to greater
nancial stability over the long term. However, many
retirement experts have advocated for nding a way
to oer short-term liquidity in the form of a “sidecar”
savings account — a separate savings account —
available for emergencies to avoid withdrawals of
retirement funds.
Due to the modest income prole of workers cur-
rently lacking access, withdrawals have the potential
to erode a material share of the assets held by new
savers. Withdrawals are modeled as a share of
annual contributions in initial years as the baseline
Auto-IRA option gains scale, with higher relative
levels among younger savers, consistent with the
levels and patterns of withdrawals seen in the initial
years of state programs.
63
Over time, withdrawals
are modeled to represent 2.25% of account assets,
growing in dollar terms as account balances grow,
based on benchmarks among current savers.
64
Asset Growth through Compounding Market Returns
Retirement savings plans enable workers to see their
assets (net of fees and withdrawals) benet from
market returns. While results for any time period are
variable, market growth has been reliable over time in
growing the real value (above and beyond ination) of
contributed assets.
To navigate market variability, savers are typically
advised to pursue a more-aggressive mix of invest-
ments in their younger years, and then shift to a
more-conservative mix as they approach retirement
age. Target date funds implement this approach for
the saver over time by gradually shifting from higher
yield/higher risk assets for younger savers toward
lower yield/safer assets as savers near the targeted
retirement age. Expected returns in this analysis
are modeled to vary by age from 5.4% annually for
the youngest savers to 4.3% annually for the oldest
savers, based on data from the Employer Benets
Research Institute (EBRI) on the mix of asset types
held by savers at dierent ages, and benchmarks of
anticipated performance by asset type.
65
28
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
Account balances for participants are modeled
as a function of contributions and market returns,
net of fees and withdrawals. At the national
level, assets held by participants reaching the
assumed retirement age of 65 are treated as
withdrawn, because these participants exit their
saving years and use their accumulated assets
to supplement their income in retirement.
Based on these assumptions about fees,
withdrawals, and returns, the analysis illustrates:
• Growing account balances over time. Each
cohort of savers over the analysis period
reaches the age of 65 with additional years of
accumulation, increasing the average account
balance. Participants reaching 65 in the
year 2040 have an average account balance
of more than $60,000 (see Figure 2.10).
Average account balances for each cohort
reaching 65 exceed the average account balance
among all participants, because older savers have
higher average earnings and lower job turnover.
However, younger cohorts have additional years
of contributions and investment returns ahead
of them. This pattern will enable the trend of
increasing average balances as of age 65 to
continue for future retirees beyond 2040.
Supplementing Retirement Income through Private
Savings
Because the goal of any national universal access
proposal is to reach workers who previously lacked
access to savings through their employers, the
assets that they accumulate saving for retirement
serve to supplement other sources of income,
including Social Security, through their retirement
years.
Retirement experts are increasingly focused on
the range of “decumulation” strategies to help
retirees optimize their lifetime income and quality
of life. A 2019 CRI report, in conjunction with
nancial experts Willis Towers Watson, explores
the growing demand for lifetime income solutions,
and the range of models that can protect assets
and mitigate risk for retirees, depending on their
nancial situations.
66
That analysis compares the
trade-os of dierent income solutions to the
income generated by an immediate xed annuity.
To illustrate how a “lump sum” account balance
can translate into a stream of lifetime income
in retirement, this analysis uses the simplied
framework of the “immediate annuity.” This approach
is just one of many potential strategies for generating
lifetime income, and will not be appropriate for all
households, but nonetheless, represents a common
basis on which to understand the ongoing value of
accumulated assets to savers. Based on market
rates, an average account balance of around $60,000
would support an annual annuity of around $3,300
for the average saver for the remainder of their life.
67
Savings accumulations will vary for individual
workers and can be substantially higher even
at modest contribution levels if savers can
avoid early withdrawals, start saving early, and
accumulate returns over an extended period.
Results estimated as of 2040 by denition only
Figure 2.10: Average Account Balances Under an Auto-IRA Grow through Additional Years of Savings
$60,600
Account Holders Reaching 65
$26,400
Average Account Balance
$0
$10000
$20000
$30000
$40000
$50000
$60000
$70000
204020392038203720362035203420332032203120302029202820272026
All Accounts
29
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
include the cohort of older savers who will begin
to access retirement savings. Younger workers
from the millennials and Gen Z cohorts who will
not yet have reached retirement age within the
study period will have greater opportunities to
build assets through continued contributions and
additional years of compounding growth. As a
result, future generations of Americans will see far
greater benets from increased savings than those
quantied as of 2040 within these estimates.
2.2 Policy Choices Have Impacts on
Coverage and Savings
Universal access programs implemented
internationally and by US states have varied in their
design characteristics. These design dierences
have included the type of retirement savings account
(IRA and/or 401(k) structure), which employers
are subject to coverage requirements, whether
employer contributions are available, and whether
these employer contributions are voluntary or
required. Such dierences reect the consideration
of trade-os to achieve the shared policy goals of
expanding access and increasing savings. They
also include the consideration of the potential
adoption challenges for dierent groups (such
as employers and employees), and the technical
considerations inherent in policy eorts of this scale.
Modeling of the impact of dierent retirement
savings design options is intended to help inform
the conversation and consideration of these
dierences in approach and their trade-os.
The following scenarios are quantied and
compared:
68
• Auto-IRA baseline covering all employers
(“Baseline Auto-IRA”), with features such as
auto-enrollment and auto-escalation described
above;
• Auto-IRA with employer threshold (“Threshold
Auto-IRA”), which applies a threshold for
employer size and age below which rms are
exempted from the requirement to provide
access to coverage;
• 401(k) Voluntary Employer Contribution with
employer threshold (“Voluntary Employer
Contribution 401(k)”), which changes the sav-
ings account type from an IRA to a 401(k), which
— among other implications — gives employers
the discretion to make contributions; and
• 401(k) Mandatory Employer Contribution with
employer threshold (“Mandatory Employer
Contribution 401(k)”), in which employer contri-
butions are required rather than discretionary.
Policy variations are applied in sequence, retaining
most features from scenario to scenario to isolate
the impact of specic features on outcomes. In
practice, it may be possible for policymakers to “mix
and match” these design components, although
the scenarios dened in this analysis are broadly
reective of models enacted at the state levels and
those envisioned in national legislative proposals.
Thresholds for Employer Participation Dictate the
Remaining Access Gap
Results from voluntary programs, both internationally
and in US states, suggest a limited impact in closing
the access gap.
69
Eorts to achieve universal access
thus generally include a requirement for employers
to oer some type of retirement savings option to
their employees, often with a ne or penalty for non-
compliance as an enforcement mechanism. While
this approach maximizes access and participation,
concerns have been raised about the challenges
of compliance, especially for small businesses.
As a consequence, here in the US, state programs
often establish a “threshold” in rm size with an
additional consideration for how long a rm has
been in business, with rms below this threshold
exempted from the provisions to oer coverage
to their employees. State Auto-IRA programs in
California and Illinois each include a minimum
employer size for required employer participation,
which is set at ve employees in California and 25
employees in Illinois. Oregon’s program, by contrast,
does not have a threshold requirement, making it
“universal” in application to all full- and part-time
employees who have worked at least 60 days.
70
30
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
Federal Auto-IRA proposals introduced in the Senate
(S. 2370) and in the House (H.R. 3499) each set a
threshold that exempts employers with 10 or fewer
employees, and exempt businesses that have not
been in existence for two full calendar years.
71
The Automatic Retirement Plan Act (H.R. 4523),
which envisions a national 401(k) requirement,
exempts businesses with 10 or fewer employees
on a “typical business day,” as well as employers
that have not been “in existence for three years.”
72
This legislation also envisions a phase-in of the
coverage requirement, with small businesses
(under 100 employees) receiving additional time
to comply. This phased implementation has
also been built into state approaches and is
assumed in this analysis for all scenarios.
73
OregonSaves, which does not have a minimum
employer threshold, has a participation deadline
in January 2021 for businesses with fewer than
ve employees. Research conducted by the Pew
Charitable Trusts in July 2020 showed high levels of
satisfaction with the program among employers that
have participated to date, with nearly three-quarters
of employers expressing a positive or neutral
impression of the program and 79% reporting that
they had not experienced any out-of-pocket costs
associated with the program.
74
The experience of the
smallest employers as their participation becomes
required will be important to monitor in 2021.
The intent of these employer exemptions to the
coverage requirement is to avoid the imposition
of even a de minimis eort on the smallest and
newest businesses. These businesses are least
likely to have the administrative apparatus (such a
human resources department or manager, or payroll
provider) that makes it even easier to facilitate their
employees’ ability to save. However, the smallest
rms are currently also the least likely to oer
coverage to their employees. As a consequence,
these thresholds can perpetuate the access
gaps that universal programs seek to close.
Employees at Firms Required to Participate under
Potential Employer Thresholds
The inclusion of a threshold for required employer
participation, the level at which that threshold
is set, and who is eligible to participate have
signicant implications on the number of workers
gaining access to retirement savings through
their workplace. To estimate the number of
employees working at rms required to provide
access under dierent potential thresholds, data
sets from the US Census Bureau are combined to
develop a snapshot of the composition of private
sector employment by rm size and rm age as
of 2020.
75
Potential rm size thresholds of ve,
10, and 20 employees are considered, along with
potential rm age thresholds of 1, 2 and 3 years.
Through this approach, the analysis illustrates:
• The size of the workforce at rms required
to participate varies signicantly by
potential threshold. A threshold dened as
ve employees and one year in existence
would include 50 million out of the 57 million
private sector workers at businesses that do
not currently provide access. By contrast, a
threshold dened as 20 employees and three
years in existence would apply to rms with
32 million workers, with 25 million working
at exempted rms (see Figure 2.11).
• Firm size thresholds have a greater impact
than rm age thresholds. Each increment
in exempted rm size (from ve to 10 to
20 employees) exempts rms employing
several million workers nationally, while
each additional year of rm age (from one
to two to three years) exempts rms with
a total workforce closer to 1 million.
Consistent with federal legislative proposals that
have typically included an employer threshold of 10
employees, and the importance of the employer size
threshold on access, modeling of each of the policy
variations (beyond the baseline Auto-IRA scenario)
assumes an employer threshold of 10 workers and
two years in existence for the requirement to provide
access. If fully implemented among the current
private sector workforce (as of 2020), required
participation above this threshold would apply to
rms employing an estimated 40.8 million workers,
with 16.5 million workers at exempted rms.
31
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
Voluntary Employer Contributions Can Increase
Participant Savings
The Auto-IRA model implemented in several states
and modeled in this analysis (with and without
a threshold) does not allow for contributions
by employers to supplement the savings of
employees. This constraint could be addressed
through the use of a Roth 401(k) rather than Roth
IRA as the retirement savings account option,
which would enable employer contributions.
However, the use of the 401(k) brings additional
considerations into play for employers, notably
that 401(k)s are retirement plans covered by the
Employee Retirement Income Security Act (ERISA)
of 1974 and would have to meet the administrative
and regulatory requirements of this law. These
requirements must be considered, along with
potential hybrid and alternative approaches, in
contemplating approaches to allow voluntary
contributions and support greater levels of savings.
Changing the Type of Savings Account from an IRA
to a 401(k)
The most-recognizable savings option that combines
employee and employer contributions is the 401(k).
Data from Vanguard indicates that the contribution
levels within this model have been relatively stable
in recent years, with employee contributions
averaging around 7% of income and employer
contributions around 3.7% for a total of 10.7%.
76
The national approach envisioned by the Automatic
Retirement Plan Act of 2017 introduced by House
Ways and Means Chair Representative Richard
Neal envisions a requirement for all private sector
employers above a size and age threshold to enable
their employees to participate in a 401(k) plan, if
they do not currently oer a qualifying alternative.
77
Employer contributions in this approach would be
voluntary and left to the discretion of the employer.
If the employer chooses to make such contributions,
they are deciding to take on the additional costs
to do so. Firms that are required to enable their
employees to have access to a 401(k) vehicle may
nd it in their interest to oer a voluntary match
for competitive reasons, especially if other rms
are doing so. However, the group of employers
that are aected by a participation requirement
because they do not currently oer access has
revealed a preference to limit their expenses on
employee benets. Modeling of a voluntary employer
contribution scenario through a 401(k) assumes
that employers would contribute an average of 5
cents for each dollar contributed by employees.
78
Figure 2.11: Employer Thresholds Have Signicant Impacts on the Number of Workers at Firms Required
to Provide Access to Savings
ESI analysis of US Census Bureau Business Dynamic Statistics and Quarterly Workforce Indicators Data.
0
10
20
30
40
50
60
0
6.9
8.2
16.5 17.6
24.8
57.3
50.5
49.2
40.8
39.7
32.5
20+ Emp 3+ Yrs10+ Emp 3+ Yrs10+ Emp 2+ Yrs5+ Emp 2+Yrs5+ Emp 1+ YrsAll Ages All Sizes
8.2
16.5
17.6
Workers at Firms
Below Threshold
Workers at Firms
Required to
Participate
Total Workforce as of 2020 at Firms Required
to Provide Access to Coverage (Millions)
Firms Required to Participate
32
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
IRAs vs. 401(k)s: Dierences to Consider
ERISA sets forth consumer protection and coverage
standards that rms must adhere to when oering
retirement savings plans to their employees. These
regulations mean that employers assume a certain
degree of legal risk and associated cost in providing
retirement savings options for their employees.
States that have adopted an Auto-IRA model have
done so with condence that such a program
is not subject to ERISA, because employers
perform only “ministerial” functions, such as
making and remitting payroll deductions, and do
not have any responsibility or discretion in the
administration of the state-facilitated program.
79
By contrast, 401(k) models that enable employer
contributions are subject to ERISA protections.
Employers contributing to an auto-401(k) plan
exercise discretion and control over the plan’s
assets, typically assuming duciary responsibilities
for the plan’s investment “menu,” selection of service
providers, and operations. Thus, employers can be
liable for non-compliance with the basic standards
of conduct and subject to monetary penalties.
80
This additional complexity associated with
401(k) models adds to the administrative burden,
compliance costs, and risks associated with
participation from the perspective of a business.
However, it is important to note that the expansion of
alternative 401(k) arrangements in the SECURE Act,
such as Multiple Employer Plans (MEPs) and new
Pooled Employer Plans (PEPs), may help to reduce
these burdens by outsourcing administrative and
most duciary responsibilities relative to a traditional
standalone 401(k) plan.
81
In addition, many national
universal access proposals have included provisions
for tax credits to help businesses with one-time
start-up costs associated with providing access.
82
401(k) and IRA models also vary in terms of the
degree to which they increase access among the
youngest workers. Within a 401(k) plan, employers
have the discretion to set an age threshold (with
a maximum allowable age of 21) for employee
participation, and IRS data indicate that the
majority of 401(k) plans use this provision.
83
The
Automatic Retirement Plan Act of 2017 (which
uses a 401(k) option) requires employers to
provide access only to workers 21 and older,
while the Automatic IRA Act of 2019 requires
access once an employee reaches the age of 18.
This dierential in the starting age for required
coverage contributes to larger remaining access
gaps for a 401(k) model compared with an IRA.
Hybrid IRA and 401(k) Approach
Some alternative approaches have been suggested
that seek to capitalize on the asset-building
advantages of the 401(k) while minimizing
the cost and administrative burden on certain
businesses. Universal access to retirement savings
could be designed using a “hybrid” approach
that allows employers to choose whether they
adopt a payroll deduction IRA or a 401(k) option
to meet their requirements to oer access of
some kind. Employers could then decide, taking
into consideration the fundamental dierences
between the two types of accounts, which one they
prefer. Because of the simplicity of an Auto-IRA,
more employers are likely to choose it, but some
employers, especially mid-size rms that may not
yet have plans of their own, may choose a 401(k),
concluding that the benets of higher contribution
levels and the potential for an employer match is
important to help their businesses become more
competitive in attracting and retaining talent.
Notably, participation requirements in states with
Auto-IRA programs have led to some businesses
that previously did not provide coverage opting
to institute their own 401(k) plan. This dynamic
illustrates the potential for Auto-IRA models to
complement the robust existing 401(k) space.
It is important to keep in mind that any employer that
adopts an Auto-IRA can always choose later to move
to a 401(k) model through a private provider. This
is another advantage of having all rms oer their
employees a way to save. For newer businesses,
this approach helps to get them started, but then
gives them the experience to consider moving to a
401(k) as their company grows and resources allow.
33
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
“IRA Plus” Approach
Finally, the limitations discussed in the Auto-IRA
model are statutory and could be addressed through
policy changes. Congress could innovate by creating
another type of IRA. For example, Congress has
previously created a variant of the traditional IRA:
the Savings Incentive Match Plan for Employees of
Small Employers (SIMPLE) IRA, available to busi-
nesses with 100 rms or less. This option, which is
ERISA-regulated, features a simplied set-up and
administrative structure relative to 401(k) plans,
requires an employer contribution, and has a higher
annual contribution limit for employees ($13,500 in
2020) relative to a traditional IRA.
84
Similarly, an “IRA
Plus” approach could be developed that retains the
simplicity and minimal burden of the IRA structure,
while enabling voluntary employer contributions into
employee accounts.
Required Employer Contributions Can Increase
Employee Returns — with Economic Trade-os
Options that allow for voluntary employer contribu-
tions give businesses the choice to contribute to the
savings accounts of their employees, and the discre-
tion not to do so. Alternative proposals developed
and advanced by policy and industry experts would
require mandatory contributions from employers as
a means to build savings, through programs similar
to the federal Thrift Savings Program (TSP), a 401(k)
program for millions of federal workers (including
members of Congress).
This approach to universal access would enable
low- and moderate-income households to access
the benets from employer contributions that many
higher-income workers already enjoy, improving the
eective return on investment on employee contri-
butions. However, this approach would create a new
non-discretionary direct cost to employers, which
raises a set of challenges within the labor market
that are unique to this policy approach. From an
economic perspective, this required cost functions
similar to a tax, with the incidence either passed
back to the employee (through a reduction in wages
or other benets) or absorbed by businesses (limit-
ing their ability to invest and grow). This approach
would also have implications for the existing savings
marketplace, which would have to be brought in
alignment with the employer match requirement to
avoid creating an uneven playing eld between new
and existing employer retirement plans and grandfa-
thered state programs.
Constructing and Analyzing a Required Employer
Contribution
Employer contribution proposals envision a shared
responsibility for funding savings accounts between
the employer and the employee to reach equivalent
(or better) savings levels compared to employee-only
contribution models. Required contribution levels
would have to be carefully calibrated to limit the
increase in costs to businesses at implementation,
and to minimize disincentives for hiring over the long
term.
Modeling assumes an initial employee contribution of
4% and an initial employer contribution of 1%, equal-
ing the overall contribution level of 5% envisioned
under each approach. These levels are envisioned to
escalate slowly, with employee contributions growing
0.5% per year up to 7% (after Year 6) and employer
contributions growing 0.25% per year up to 3% (after
Year 8). This escalation schedule, which seeks to
minimize disruptive impacts on businesses, results in
a slightly slower acceleration of total contributions to
the long-term level of 10% than is envisioned under
models relying entirely on employee contributions.
Implications for Participating Businesses
While the direct expenditures from a required
contribution are made by employers, labor market
dynamics would dictate how the true economic cost
of the requirement is shared. A portion of costs are
likely to be passed back to employees in the form of
reductions in wages or other benets — undercutting
some of the benets from increased savings — and
a portion is likely to be borne by employers, limiting
future hiring, business investment, and growth. The
economic framework of cost incidence generally es-
tablishes that when a new cost is imposed, the party
(between buyers and sellers or between employers
and employees) that is more price-sensitive will be
least willing to absorb the additional costs, shifting
the majority of the incidence to the other party.
34
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
Empirical research suggests a high level of wage
sensitivity for low-income workers, who value
changes in wage earnings at the expense of
other aspects of employee compensation.
85
This
puts lower-income workers in a stronger position
relative to higher-income earners to minimize
wage reductions and oset increases in benets
(assuming stable economic conditions). In addition,
osetting reductions in fringe benets would
be relatively unlikely in this case, because the
employers that currently do not oer retirement
access are often those that have limited benets
packages for their employees in other areas, such as
healthcare coverage, from which they could recoup
these costs.
86
Due to this combination of factors,
employers would be expected to bear the majority
of cost incidence from an employer contribution
requirement, although workers would be expected
to see some reduction in earnings over time.
87
This additional cost would have a dampening
eect on business investment and growth over
time. Impacts would fall primarily on small and
mid-size rms, which often do not operate with
excess prots or reserves. On the margins,
increased cost will reduce demand for future
hiring, investment, and expansion among these
businesses. Market pressures will also threaten
the viability of rms in highly competitive markets
and rms that rely heavily on low-wage workers,
potentially accelerating existing trends toward
the market consolidation of larger employers.
Implications for Existing Savings Programs
In addition to the eects on the millions of workers
currently lacking access to a retirement savings
plan, this approach would necessitate changes to
the framework of existing private market and state
retirement programs to eectuate the aim of required
employer contributions. The universal access models
analyzed in this study are generally structured to
complement the existing savings framework, with
rms that already oer coverage options exempted
from new requirements. Employer contributions are
common but voluntary among private sector plans,
and state programs following the Auto-IRA model do
not enable these contributions. Absent any additional
changes, businesses that do not currently provide
access could (and probably would) procure a private
plan without an employer contribution component
or join state Auto-IRA programs if available. This
would, in eect, function as a “loophole” to the
employer contribution requirement, limiting its eect.
Therefore, it is envisioned that a mandatory
employer contribution approach would have to
extend the contribution requirement among all
qualifying employers, including those that oer
existing coverage to their employees and including
existing state eorts as well. Implementation
would be expected to be gradually phased in over
several years to allow for those employers that do
not currently make employer contributions to be
able to do so and to ensure consistent standards
across employers. The potential disruptive eects
of this approach on employers currently oering
coverage and on their employees are not directly
accounted for within this study, which limits
analysis to the population of rms and workers that
currently lack access, and does not capture the full
range of potential dynamic eects from changes
in policy aecting businesses with existing plans.
Analyzing How Policy Choices Aect
Participation, Savings, and Asset Building
Analysis of the policy options — the type of
savings account, thresholds for participation
requirements, and the ability or requirement for
employer contributions — through a consistent
modeling approach enables a comparison of
these policy variations to the baseline Auto-IRA
scenario, and to each other. This sequential
process enables estimates to be developed
of the incremental impacts on participation,
savings, and asset building of each policy option
in isolation. Comparative analysis shows results
for participation, contribution levels, and savings,
reecting both per-participant measures and
aggregate measures across all participants.
Analysis of participation by scenario illustrates:
• Universal access models increase the
number of workers saving for retirement
in 2040 by 28 to 40 million. Participation
through active contributions to a savings
35
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
account is anticipated from about 50
to 70% of private sector workers who
currently lack access (see Figure 2.12);
• Participation levels fall signicantly if em-
ployers below a certain employee threshold
are exempt. Policy options, that exempt smaller
and younger rms from the requirement to
provide access would limit the degree to which
the models can close the access gap (whether
within Auto-IRA or 401(k) approach). An exemp-
tion of rms with fewer than 10 employees or in
existence for less than two years would reduce
participation by an estimated 11 million by 2040
under an Auto-IRA model, with additional reduc-
tions in coverage under 401(k) approaches.
• Additional reductions in participation of 1.3
to 1.5 million workers are anticipated with
the 401(k) models under the same employer
thresholds. This is due to dierentials in
coverage requirements by age, as well as
lower anticipated levels of voluntary partici-
pation from rms exempt from the coverage
requirement under the more-challenging
administrative structure of the 401(k).
Dierentials between models in total account
holders (including accounts growing without
contributions) are larger still, with nearly 19 million
fewer accounts as of 2040 (a decrease of 26%)
under the employer threshold IRA model relative
to the baseline Auto-IRA, and additional declines
under the 401(k) approaches (Figure 2.13).
Analysis of savings and asset growth by scenario
illustrates that:
• While participation levels are lower in
models exempting some employers, those
participating may have higher average
contributions and savings. Policy alternatives
are likely to increase average contributions
and account balances among those who have
access and participate relative to the baseline
Auto-IRA model. These dierentials are due
in part to the characteristics of the covered
population (with average participant earnings
increasing when excluding rms below an
employer size and age threshold), and due
in part to the scenario characteristics.
• Average contributions and savings are
highest in the 401(k) scenarios. Average
savings levels increase with a voluntary
employer contribution 401(k) when
compared to the threshold Auto-IRA, due
to contributions from some employers, the
eect of the increased annual contribution
Figure 2.12: Required Universal Access Can
Increase Participation by 50 to 70% Among
Workers Currently Lacking Access
0
10
20
30
40
50
60
40.4
29.6
28.3
28.1
Mandatory
Employer
Contribution
401(k)
Voluntary
Employer
Contribution
401(k)
Threshold
Auto-IRA
Baseline Auto
IRA
Active Program Participants, 2040 (M
)
Access Gap Among Workers <65 (2040): 58.3M
Figure 2.13: Participation Falls with an Employer
Threshold, and Slightly Further with 401(k)
Approaches
Mandatory Employer
Contribution 401(k)
Voluntary Employer
Contribution 401(k)
Threshold
Auto-IRA
Baseline
Auto IRA
71.4
52.6
50.3
49.9 (0.4)
(2.3)
(18.8)
Total Account Holders (Millions)
36
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
limit on a small sub-set of savers, and lower
anticipated levels of early withdrawals.
• Savings levels are estimated to be slightly
lower under a mandatory employer
contribution 401(k) approach than a voluntary
employer contribution 401(k) approach,
due to the constraint on wage growth
from the required employer contribution,
as well as a slower pace of escalation in
contributions implemented to mitigate the
impacts on employers in this scenario.
• Average balances for participants reaching
age 65 in 2040 grow from $66,300 in the
threshold Auto-IRA scenario to $75,200
under the voluntary employer contribution
401(k) approach (see Figure 2.14).
• The baseline Auto-IRA model generates
a lower average account balance for
participants reaching 65 of $60,600 due to
participation of more low-income workers,
which decreases average balances.
These results illustrate potential trade-os for
consideration between payroll deduction Roth IRA
and Roth 401(k) options. When analyzed using
equivalent employer thresholds, an IRA model
encourages a higher level of participation by
presenting the lowest barriers to participation for
businesses and savers. However, a 401(k) approach
can encourage higher average levels of contributions
and asset accumulation over time among those
who do participate due to its provisions around
contributions and withdrawals.
Analysis of overall savings levels, accounting for
both participation and average participant outcomes,
illustrates that:
• Overall savings are highest in the baseline
Auto-IRA scenario, which covers all employ-
ers. While per-participant savings are higher
under alternative approaches, the expansion of
coverage anticipated under the baseline Auto-
IRA scenario with no employer threshold leads
to the largest increase in overall savings among
the policy options modeled.
Baseline
Auto-IRA
Threshold
Auto-IRA
Voluntary Employer
Contribution 401(k)
Mandatory Employer
Contribution 401(k)
$69,600
$66,300
$60,600
$0
$10,000
$20,000
$30,000
$40,000
$50,000
$60,000
$70,000
$80,000
204020392038203720362035203420332032203120302029202820272026
Average Account Balance
at Retirement
Figure 2.14: Average Account Balances at 65 Range from $60,000 - $75,000 by 2040 and are Largest
within the 401(k) Models
$75,000
37
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
• Annual contributions to savings accounts
are estimated to total more than $130
billion by 2040 under the baseline Auto-IRA
model, adding up to a cumulative $1.89
trillion over the analysis period, with policy
alternatives producing $1.4–$1.5 trillion in
cumulative contributions (see Figure 2.15).
• The voluntary employer contribution 401(k)
scenario generates the highest savings
levels among policy alternatives with an
employer threshold. Higher participation levels
under threshold Auto-IRA approach are bal-
anced by higher savings levels per participant in
the 401(k) models, with the voluntary employer
contribution 401(k) model producing slightly
higher total savings and the mandatory employ-
er contribution 401(k) producing slightly lower
total savings than the threshold Auto-IRA.
This analysis arms the value of the 401(k) as a
savings vehicle that maximizes the potential asset
accumulation among those who have access and
participate. If feasible, a voluntary employer con-
tribution 401(k) approach without a threshold for
required participation (similar to the baseline Auto-
IRA scenario) or a mandatory employer contribution
401(k) approach with a more-aggressive employer
contribution level could produce higher levels of
savings than the baseline Auto-IRA model. However,
these approaches and requirements have impacts
on participating businesses and the broader savings
market, and federal 401(k) or IRA proposals to date
have typically contemplated an employer threshold
out of consideration for the implications for the
smallest businesses. The inclusion of the baseline
Auto-IRA scenario illustrates the importance of
considering whether to include and where to draw an
employer participation threshold on overall levels of
access, participation, and savings.
Figure 2.15: Cumulative Savings Contributions are Highest Within the Baseline Auto-IRA Model, Totaling
$1.9 Trillion through 2040
Baseline
Auto-IRA
Threshold
Auto-IRA
Voluntary Employer
Contribution 401(k
)
Mandatory Employer
Contribution 401(k)
$0
$30
$60
$90
$120
$150
2040203920382037203620352034203320322031203020292028202720262025202420232022
$1
$4
$27
$52
$83
$96
$107
$116
$121
$125
$126
$127
$128 $128
$129
$130
$131
$132 $132
$1.89 T
$1.53 T
$1.48 T
$1.42 T
Cumulative
B
38
What are the Potential Benets of Universal Access to Retirement Savings? © 2020 Georgetown University Center for Retirment Initiatives
3.1 Increased Economic Growth and Tax
Revenue
In addition to the positive impacts on participating
workers, expanding coverage and increasing
retirement savings would create additional positive
“downstream” impacts for the nation’s economy.
More-accessible savings options would help
the competitiveness of small businesses and
the nancial security of workers, including the
self-employed, encouraging a more-dynamic
economy, while increased savings levels
will grow the income that senior households
have available to spend in retirement.
In addition to the returns they generate for
individuals, personal savings provide a source
of capital for business investment and growth.
This positive cycle produces stronger growth in
employment and activity across the economy,
and in turn generates additional tax revenue.
Established models of the relationship between
savings, investment, and growth are used to
translate the national increases in savings under
the savings scenarios into increases in GDP
growth. These relationships also inform analysis
of the impact on government tax revenue.
Enhancing Economic Productivity and
Accelerating Growth
Micro Eects on Small Businesses, Workers, and
Households
The design of savings options have implications
for the everyday decision-making of businesses,
workers, and families. These individual
microeconomic decisions about what job to
take, whether to start a business, and how to
spend disposable income aggregate together
to have signicant impacts on the economy.
Making Small Businesses More Competitive
Closing retirement savings access gaps has
the potential to increase business dynamism by
leveling the playing eld between small and large
businesses in employee recruitment and retention.
As smaller employers and providers adapt to new
requirements and coverage options are widely
adopted, the competitive advantages currently held
by larger businesses will diminish. Removing this
barrier would allow for better preference matching
between employees and employers, improving
overall productivity and job satisfaction.
88
This would
increase business dynamism by removing a growth
constraint on small businesses, the key engine
of growth and dynamism in the US economy.
Flexible Work Arrangements
Policy makers also need to wrestle with changes
in the relationship between businesses and
workers, as non-traditional work arrangements
become increasingly prevalent as workers look
to combine multiple approaches to contribute
to their total income. Analysts studying broad
trends about “the future of work” expect these
patterns to intensify over time, driven by forces
like technology (which allows for greater remote
work and exibility) and automation (which will
reshape traditional business models).
89
Increasing
the availability of retirement savings options will
help more workers to establish accounts and begin
saving. Crucially, modeling does not assume that
savers will stay at the same employer across their
career, but recognizes that more universal access
will enable workers to continue to build savings
across multiple jobs and work arrangements.
90
Facilitating this can enable workers to be more
exible and entrepreneurial in their career
choices, and adapt to changing conditions.
Increasing Disposable Income for Senior Households
As they grow in number, seniors are also an
increasingly important consumer segment with
their household spending power. Retirees with
insucient savings and a shortage of disposable
income are forced to cut back on their spending
patterns, forgoing purchasing goods, services, and
recreation to focus their spending on the essentials
like housing and food.
91
These cutbacks have
material impacts on the economy, reducing the
demand for a variety of goods and services that
are often sold and produced locally.
92
Initiatives to
enhance savings will help retirees better maintain
their established spending patterns, beneting the
economy and the businesses that serve them.
3. Long-Term National Impacts from Increased Savings
39
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
Macro Eects on GDP Growth
In addition to micro-level eects on individual house-
holds and businesses, universal retirement access
would aect the macroeconomy by increasing the
level of national savings and investment. When
individuals save a portion of their earnings, they
make an immediate trade-o between consumption
and savings. At the macro level, these savings do
not sit idle, but instead represent available capital for
business investments that improve the productivity
of workers and the economy. This eect accelerates
economic growth, increasing GDP per capita and the
overall standard of living.
Macroeconomic models describe the relationships
between the levels of consumption, savings and
investment and the rate of economic growth. This
analysis uses the well-established “Solow Growth
Model” to run simulations of the eect of the in-
crease in savings generated by the scenarios on the
growth path of the US economy over the next two
decades.
93
Impacts are calibrated to the slow growth
environment projected by the Congressional Budget
Oce (CBO), which anticipates that changes in the
demographic composition of the population will
limit workforce growth below past norms over the
2020–2040 period.
94
Since the standard of living is
directly tied to economic growth, these projections
highlight the growing importance of policies aimed to
improve productivity and output per worker.
The annual savings anticipated from the coverage
approaches studied would be expected to increase
the personal savings rate nationally from an existing
benchmark of around 7.5% to a range of 8.0%–8.2%
(varying by scenario). This in turn would translate to
increases in the total national savings rate and would
increase private business investment, stimulating
increases in productivity in the private sector.
Based on this framework, the analysis illustrates that:
• Increased productivity growth from
increased savings and investment
accelerates GDP growth. The annual rate
of real GDP growth is estimated to grow from
the 1.70% per year projected over the next
two decades by the CBO to 1.71%–1.72%
annually. While subtle, this rate increase applies
across the full US economy and compounds
each year, producing signicant incremental
growth over time. This increase in growth
rate would add $72–$96 billion to the national
GDP in the year 2040 (see Figure 3.1).
• Increases are highest under the baseline
Auto-IRA approach. National GDP is
estimated to be nearly $100 billion higher
in the year 2040 under this scenario,
which generates the largest increase in the
personal savings rates through the highest
participation levels, thus stimulating the
greatest productivity growth. Among scenarios
with an employer threshold, the voluntary
employer contribution 401(k) generates slightly
more growth than the threshold Auto-IRA.
This eect translates to growing wealth and
increasing living standards over time for all
Figure 3.1: Increased Savings and Investment Boost GDP Growth by $72–$96 Billion in the Year 2040
Baseline
Auto-IRA
Threshold
Auto-IRA
Voluntary Employer
Contribution 401(k
)
Mandatory Employer
Contribution 401(k)
$0
$4
$7
$11
$14
$22
$29
$37
$43
$49
$55
$61
$67
$73
$79
$85
$91
$96 B
$78 B
$76 B
$72 B
Additional Annual US GDP ($Billions)
$0
$20
$40
$60
$80
$100
$120
204020392038203720362035203420332032203120302029202820272026202520242023
2022
40
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
Americans, whether or not they participate
directly in a savings option. The increased rate
of growth results in a per capita GDP increase
of $240–$310 per US resident in the year 2040,
which fundamentally translates to higher earnings
levels and standard of living across the economy.
Increased Tax Revenues from Economic Growth
Accelerating rates of economic growth also
translates into increases in tax revenue collected
by governments at the federal, state, and local
levels. Current ratios of tax collections to economic
activity can be used to analyze the tax revenue
implications of additional economic growth
(assuming a continuation of existing tax policy).
Federal revenues largely result from taxes on
households, payroll, and corporate earnings
that are fundamentally income-generated.
Increasing the rate of GDP and national income
growth, therefore, translates directly to increased
tax revenues for the federal government.
Based on this framework, the analysis illustrates that:
• Economic growth will produce increased
tax revenue for the federal government.
The additional economic growth, stimulated
by the cycle of increased savings, investment,
and productivity, is estimated to generate
an increase of $11–$14 billion in federal tax
revenue in the year 2040 (see Figure 3.2).
• The largest overall increase is again
seen under the baseline Auto-IRA
scenario, and among scenarios with an
employer threshold, within the voluntary
employer contribution 401(k) model.
This level represents a new, higher base
of activity relative to current trends. These
annual increases would be expected to
continue, and to magnify as more participants
reach retirement, in subsequent years.
State and local governments have separate and
distinct tax bases from the federal government, and
often apply their own taxes to personal and business
income, as well as other assets (such as real estate)
that would benet from enhanced economic security
and growth. Tax rates and growth implications vary
by location; specifying the level and distribution of
these state and local benets would require a more-
granular assessment not addressed in this report.
Figure 3.2: Increased Economic Growth Leads to $11 - $14 Billion in Additional Federal Tax Revenues in
the Year 2040
Baseline
Auto-IRA
Threshold
Auto-IRA
Voluntary Employer
Contribution 401(k)
Mandatory Employer
Contribution 401(k)
$11.6 B
$11.2 B
$10.8 B
Additional Federal Tax Revenue ($Billions)
$0
$2
$4
$6
$8
$10
$12
$14
$16
204020392038203720362035203420332032203120302029202820272026202520242023
$14.3 B
2022
41
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
3.2 Assessing the Impact on Federal
Benet Programs
As the nation’s population ages and too many
have too little saved for retirement, the projected
cost of federal benet programs supporting
seniors is anticipated to increase substantially.
Helping future retirees increase their savings
and their resources in their retirement years
has the benet of reducing their need for these
programs, many of which are means-tested.
Program data and eligibility rules are used in
the analysis to dene the current relationship
between income and benet program expenditures
for the senior population. The modeled
universal access scenarios are all expected
to diminish the rate of growth in program
expenditures for low-income seniors over time
by increasing savings and retiree resources.
Federal Benet Program Spending is Anticipated
to Rise Signicantly
Federal and state governments operate a number
of benet programs that serve, in part or in total,
lower-income seniors. As the population ages,
demand for these programs is expected to grow
materially. Increasing the resources available
to seniors would help reduce the government
spending associated with these programs.
Federal programs provide a range of
support resources to elderly Americans with
demonstrated needs, including health care,
nutrition, housing, and supplemental income.
Federal spending on these programs already
totals nearly $100 billion per year and is often
supplemented by additional state funding.
Importantly, this gure does not include the
two largest senior-targeted programs: Social
Security and Medicare (for which government
expenditures are not directly tied to retiree
incomes), or generalized spending (such as
defense, infrastructure, etc.) that benets the
full population but is not targeted to seniors.
The largest means-tested program is Medicaid,
which represents $62 billion of the $96 billion
and funds supplemental health insurance and
long-term care for many low-income seniors.
95
An additional $19 billion is spent on low-
income subsidies within Medicare Part D, for
a total of $81 billion in healthcare costs.
Other key support programs fund services for the
elderly population, such as supplemental income ($6
billion in Supplemental Security Income), food ($6 billion
in the Supplemental Nutritional Assistance Program
and Nutrition Program for the Elderly), heating ($1 billion
in Low Income Home Energy Assistance), housing
($1 billion on Supportive Housing for the Elderly), and
additional supportive and caregiver services ($1 billion).
Collectively, these programs represent another $15
billion in support spending for the elderly population as
of 2020, for a total of $96 billion when combined with
healthcare costs.
Absent any changes in policy, benet program
spending on the elderly is anticipated to grow rapidly
over the next two decades. Much of this expected
growth is driven by both the aging of the population
and the addition of more than 10 million elderly
households over this period (an increase of 32%).
In addition to population change, medical care costs
are expected to continue to grow in real terms,
with the CBO projecting growth of 1.1–1.6% per
year in “excess medical costs” (beyond ination)
over the next three decades for the Medicare
and Medicaid programs, respectively.
96
Since
programs providing medical care represent the
majority of benet expenditures for the elderly,
this excess growth means that, in addition to
increases in demand, the cost per participant will
also rise to maintain the same level of services.
Under current trends in elderly incomes, the
combination of the growth in senior households and
increasing medical costs are anticipated to increase
federal spending on these senior support programs
by 78% over the next two decades (see Figure 3.3).
This represents an increase of $75 billion, of which
$69 billion is associated with healthcare programs
and $21 billion with other support programs.
The growing senior population and associated
expenditure needs are coupled with a decline
42
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
from historic norms in the ratio of working age
households to retiree households. This change
in composition, and the increased scal pressure
associated with it, are expected to endure beyond
the current generation of baby boomer retirees.
Increasing Retiree Incomes Can Reduce Program
Expenditures
Because the nancial security of future retirees is
directly linked to future demand for benet programs,
eorts that improve the income and nancial security
of future retirees will also contribute to slowing
the rate of growth in government spending.
To understand the contribution of supplemental
savings income to the well-being of the senior
population, current trends in retiree incomes
are extrapolated forward to 2040. Within this
scenario, the incomes of the elderly population
as of 2040 match those of the prior generation in
terms of income replacement levels achieved.
97
Since universal access scenarios analyzed in
this report are targeted to workers who currently
lack access to savings, additional savings
generated through these policy approaches
can be understood as supplemental to incomes
anticipated under the continuation of current trends.
Next, current and anticipated program expenditures
are estimated by household income level, using a
mix of program data from administering departments
and program eligibility rules.
98
On a per-household
basis, benet program expenditures fall signicantly
as incomes rise, particularly for the lowest-income
households that comprise the majority of program
expenditures. Using this relationship between elderly
incomes and government expenditures, potential
government savings can be estimated for each
of the policy approaches considered. The level
of possible expenditure savings associated with
each scenario is driven by the degree to which it
could increase the resources of future retirees.
Many of the benet programs reviewed in this
report have signicant state funding components in
addition to the federal expenditures. Most notably,
states provide an additional $0.55 for every dollar
of federal funding for Medicaid. While specic
rules and match rates vary by state, the ratio of
federal and state expenditure for each program
analyzed can be used to understand the order
of magnitude of the eect of increased retiree
resources on state benet program spending.
Based on this framework, the analysis illustrates that:
• Savings increases result in material
decreases in benet program spending.
Universal access scenarios that increase
savings are all expected to diminish the
rate of growth in program expenditures
for low-income seniors over time.
• Federal and state governments share in the
savings. Due to the shared nature of many of
these benet programs, both federal and state
governments yield savings from reduced need.
• The baseline Auto-IRA model produces
the largest benets through the broadest
expansion of coverage and savings. Federal
savings in the year 2040 are estimated at
$6.2 billion and state savings at $2.5 billion,
for a total of $8.7 billion under the baseline
Auto-IRA scenario, alternative models
generate combined program savings of
around $7 billion in 2040 (see Figure 3.4).
$0
$20
$40
$60
$80
$100
$120
$140
$160
$180
$200
20402020
$80.7
$150.1
$15.0
$20.7
$95.7
$170.8
Other Programs
Health Care
Figure 3.3: Federal Spending on Means-Tested
Benet Programs for the Elderly is Substantial
and Growing
Federal Spending on Selected Programs
($Billions)
43
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
Benets would grow substantially larger beyond
2040, as the population of participants reaching
retirement with additional income continues to
increase, and later cohorts arrive at retirement age
with more-substantial account balances each year
due to the additional years of savings and returns.
While the federal government would realize about
70% of the savings from these shared federal
and state programs, increasing retiree income
probably would also provide states with additional
savings on state-specic eorts that support
quality of life for seniors above and beyond state
contributions to the federal programs described
above. State-level studies conducted by ESI in
Pennsylvania and Colorado indicate signicant
potential savings for state governments associated
with increasing retiree incomes, including state-
level programs providing property tax relief,
transportation, and other support services in
addition to shared federal and state programs.
99
Within this analytical framework, increases in elderly
income translate to decreases in benet program
demand (holding program features constant). This
mechanism can, in some cases, represent a cost
shift from government to private households for the
same services (for example, if a household no longer
qualies for Medicaid due to income eligibility and
purchases a comparable private insurance plan).
As an alternative policy consideration, there has
been some discussion of whether some additional
retirement income would be excluded from
consideration for one or more of these programs,
such as SNAP. This approach would have the eect
of reducing the savings to the government and
transferring gains to households. The degree of this
transfer would vary depending on the expenditures
associated with the program, with medical insurance
programs representing the most material cost
drivers. From an economic standpoint, the net social
benet remains the same regardless of this transfer.
Government savings from enhanced revenues (from
increased economic growth) and expenditures
(from reduced benet program demand) can
be measured against the additional federal
expenditures associated with the enhanced
Saver’s Credit envisioned within the universal
access models studied in this report.
Annual deposits to savers through the enhanced
Saver’s Credit are estimated to total $20–$26 billion
under these models, growing about $18–$25 billion
on net from the current net scal impact of the credit
($1–$2 billion in foregone revenue). Federal revenue
increases of $11–$14 billion from economic growth
and federal expenditure savings of $5–$6 billion
from reduced program demand nearly oset the
entire cost increase associated with the enhanced
credit by 2040. The net scal eect is likely to be
positive in future years, as future generations of
participants reach retirement and scal benets grow
at a faster rate than Saver’s Credit expenditures.
$0
$1
$2
$3
$4
$5
$6
$7
$8
$9
$10
Mandatory
Employer
Contribution
401(k)
Voluntary
Employer
Contribution
401(k)
Threshold
Auto-IRA
Baseline
Auto-IRA
$6.2
$2.5
$8.7
$5.2
$2.1
$7.3
State Savings
Federal Savings
Expenditure Savings in 2040
($Billions)
$4.9
$1.9
$6.8
$4.8
$1.9
$6.7
Figure 3.4: Increased Retiree Resources Leads to
Government Program Savings of $7–$9 Billion in
the Year 2040
44
What are the Potential Benets of Universal Access to Retirement Savings? © 2020 Georgetown University Center for Retirment Initiatives
Any eort to signicantly improve retirement
readiness must expand access to ways to save
for retirement to as many workers as possible.
If we look at how this is done internationally,
there is usually little debate about the primary
goal of universal access, and several countries
require employers to provide a retirement savings
option for their employees. With all workers
covered, dierences can be found in aspects of
model design to achieve the levels of savings
needed to boost income in retirement.
Policymakers here in the US have been willing to
learn from experience of other countries, but have
stopped short of a fundamental redesign of the
system that would reshape the existing private
sector retirement market. As a consequence,
eorts to signicantly improve retirement
readiness must expand access to ways to save
for retirement within a payroll deduction IRA
and/or 401(k) structure. The type of retirement
savings accounts, the employers required to
participate, and the default levels of employee
contributions and any employer contributions
over time are all factors that will drive coverage,
savings, asset growth, and retirement income.
Of the scenarios analyzed in this report, the largest
reductions in the access gap and largest increases
in overall savings are achieved by the simplest
model that follows a payroll deduction Auto-IRA
approach, covering all employers (even without the
opportunity for employer contributions). Modifying
this approach to include an employer threshold,
exempting the smallest and newest businesses from
required participation to reduce the administrative
imposition on small employers, will leave a larger
access gap with fewer workers covered.
On the other hand, giving employers the ability
to make contributions would help to boost the
average levels of savings among those that
participate. A 401(k) option that allows for employer
contributions achieves modest increases in
average contribution levels, and further increases
in asset levels at retirement. However, it does
have some additional administrative burdens
and risks to employers, reducing likely coverage
levels relative to an Auto-IRA at any required
participation threshold. A 401(k) approach with a
mandatory employer contribution would increase
the return on investment from the standpoint of
the saver, but would have more-disruptive impacts
on existing businesses and savings plans.
Regardless of the model selected, what is clear
is that the benets to savers, retirees, and the
nation’s scal and economic well-being can be
enormous. Depending on the design features, a
national approach to universal access to retirement
savings which would require some or all employers
to oer their workers either an IRA or 401(k) could:
• Increase the number of workers saving for
retirement in the year 2040 by 28–40 million,
with participation from about 50–70% of private
sector workers who currently lack access;
• Help a young worker with a modest income
who starts saving early and follows savings
defaults for 40 years to save enough to
generate as much as $14,320 in additional
annual income for retirement, increasing to
$21,300 in annual income if eligible to take
advantage of a refundable Saver’s Credit;
• Increase cumulative total retirement
savings between $1.4 trillion and $1.9
trillion by the year 2040; and
• Accelerate economic growth,
increasing national GDP by $72 billion
to $96 billion in the year 2040.
Experiences from other countries and the
early evidence from states here in the US
demonstrate that increases in access can be
achieved in a simple, cost-eective way that
supports and includes a private market of
providers ready and willing to compete to provide
options for employers and their workers.
Conclusion
45
What are the Potential Benets of Universal Access to Retirement Savings?© 2020 Georgetown University Center for Retirment Initiatives
1 For more information about state initiatives to expand retirement access, see information maintained by the Georgetown University
Center for Retirement Initiatives: < https://cri.georgetown.edu/states/>
2 Setting Every Community Up for Retirement Enhancement (SECURE) Act of 2019, Pub.L. 116–94, <https://www.congress.gov/116/
plaws/publ94/PLAW-116publ94.pdf>
3 Private sector employment estimates are based on “covered employment” as dened by the Bureau of Labor Statistics, excluding
independent contract work. Low- and high-end estimates of retirement savings access are derived from the 2019 National Compensation
Survey of the Bureau of Labor Statistics (low end) <https://www.bls.gov/ncs/> and analysis of Current Population Survey data by Ghilarducci
and Papadopoulos (2020), Retirement Plan Coverage by Industry, Firm, and Worker Characteristics. Schwartz Center for Economic Policy
Analysis at the New School. <https://www.economicpolicyresearch.org/images/docs/research/retirement_security/Research_Note_1_2020_
Retirement_Plan_Coverage_by_Industry_Firm_and_Worker_Characteristics.pdf>
4 This estimate is derived by averaging the estimates of access to coverage among private sector workers based on two leading data
sources: the National Compensation Survey of the Bureau of Labor Statistics and the Current Population Survey of the US Census Bureau.
The Methodology Appendix to this report describes how this calculation is applied (as well as others throughout the analysis) at [https://cri.
georgetown.edu/research]
5 Information about worker contributions to retirement plans from Harvey (2017), Access to Workplace Retirement Plans by Race and
Ethnicity, AARP Policy Institute, citing analysis from the Employee Benet Research Institute of data from the Survey of Income and Program
Participation (SIPP). <https://www.aarp.org/content/dam/aarp/ppi/2017-01/Retirement%20Access%20Race%20Ethnicity.pdf>
6 See: Cohen, et al. (2019), DCIIA Fourth Biennial Plan Sponsor Survey — Auto Features Continue to Grow in Popularity. Dened
Contribution Institutional Investment Association. <https://cdn.ymaws.com/dciia.org/resource/collection/23D6FA15-31A6-4ABA-826B-
A8718DC03E59/DCIIA_Fourth_Biennial_Plan_sponsor_survey_8._FINAL.11.30.17.pdf>
7 Additional information about retirement access gaps from Ghilarducci and Papadopoulos (2020), Retirement Plan Coverage, and Scott,
John, et al. (2016), Who’s In, Who’s Out: A look at access to employer-based retirement plans and participation in the states. Pew Charitable
Trusts. <https://www.pewtrusts.org/~/media/assets/2016/01/retirement_savings_report_jan16.pdf>
8 Of workers at rms with fewer than 10 employees, 22% report access to a workplace savings plan or pension, whereas at rms with 500
or more employees, 74% of workers report access. See Scott (2016), Who’s In, Who’s Out.
9 Estimates of retirement adequacy in dierent cohorts of the population from Brown, Saad-Lessler, and Oakley (2018), Retirement
in America: Out of Reach for Working Americans? National Institute on Retirement Security. < https://www.nirsonline.org/wp-content/
uploads/2018/09/SavingsCrisis_Final.pdf>
10 See estimated average monthly Social Security benets for retired workers from Social Security Fact Sheet (2020), Social Security
Administration. <https://www.ssa.gov/news/press/factsheets/colafacts2020.pdf>
11 See Social Security Fact Sheet (2020) — Social Security data as of June 2020, Social Security Administration. <https://www.ssa.gov/
news/press/factsheets/basicfact-alt.pdf>
12 Collinson, Rowey, and Cho (2019), What Is “Retirement”? Three Generations Prepare for Older Age. Transamerica Center for Retirement
Studies. <https://transamericacenter.org/docs/default-source/retirement-survey-of-workers/tcrs2019_sr_what_is_retirement_by_generation.
pdf> and Munnell and Hou (2018), Will Millennials Be Ready for Retirement? Center for Retirement Research at Boston College. <https://crr.
bc.edu/wp-content/uploads/2018/01/IB_18-2.pdf>
13 Incomes near the median are based on households in the third income quintile (those with incomes between the 40
th
and 60
th
percentile).
Munnell and Chen (2020), 401(k)/IRA Holdings in 2019: An Update from the SCF. Center for Retirement Research at Boston College. <https://
crr.bc.edu/wp-content/uploads/2020/10/IB_20-14.pdf>
14 Ibid.
15 Information about relative preparedness for retirement of millennials, Generation X, and baby boomers from Brown (2018), Millennials
and Retirement: Already Falling Short. National Institute on Retirement Security. <https://www.nirsonline.org/wp-content/uploads/2018/02/
Millennials-Report-1.pdf> and Munnell and Hou (2018), Will Millennials Be Ready (see link above).
16 University of Virginia Weldon Cooper Center Demographics Research Group (2018), National Population Projections. <https://
demographics.coopercenter.org/national-population-projections>
17 Population projections from the University of Virginia are converted to household projections using American Community Survey (ACS)
data to calculate the current (2018) average household size for each age cohort by dividing the population by the number of “householders”
in each age bracket. This ratio (also known as the “headship rate”) is held constant for each age cohort across the analysis period and
applied to the population forecasts by age group to translate population estimates to household estimates (see Methodology Appendix for
further detail).
18 For more information about state initiatives to expand retirement access, see information maintained by the Center for Retirement
Initiatives at Georgetown University. < https://cri.georgetown.edu/states/>
19 Setting Every Community Up for Retirement Enhancement (SECURE) Act of 2019, Pub.L. 116–94. <https://www.congress.gov/116/
plaws/publ94/PLAW-116publ94.pdf>
Endnotes
Endnotes below reect sourcing for citations contained within this report. The accompanying Methodology
Appendix contains a complete explanation of the approach and calculations used to model the impact of univer-
sal access scenarios.
46
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
20 Information about the current and future superannuation guarantee percentages from the Australian Taxation Oce (2020), Table 21
– Super Guarantee Percentage. <https://www.ato.gov.au/Rates/Key-superannuation-rates-and-thresholds/?page=24#Super_guarantee_
percentage> Information about eligible employee thresholds from the Australian Taxation Oce (2020), Working Out if You Have to Pay
Super. <https://www.ato.gov.au/business/super-for-employers/working-out-if-you-have-to-pay-super/>
21 See Low Income Super Tax Oset (2020). Australian Taxation Oce. <https://www.ato.gov.au/Rates/Key-superannuation-rates-and-thres
holds/?anchor=Lowincomesuperannuationcontribution#Lowincomesuperannuationcontribution>
22 Information about super-fund accounts and assets based on data from the Australian Prudential Regulation Authority (APRA) and the
Australian Bureau of Statistics (ABS) as cited in Superannuation Statistics August 2020 (2020), Association of Superannuation Funds of
Australia. <https://www.superannuation.asn.au/ArticleDocuments/269/SuperStats-Aug2020.pdf.aspx?Embed=Y>
23 For more information about the KiwiSaver program, see KiwiSaver Annual Report 2020 (2020). Financial Markets Authority. <https://www.
fma.govt.nz/assets/Reports/Kiwisaver-Annual-Report-2020.pdf>
24 For more information about the UK NEST program, see UK NEST Quarterly Investment Report Q1 2020. <https://www.nestpensions.org.
uk/schemeweb/nest/aboutnest/investment-approach/other-fund-choices/fund-factsheets.html>
25 See The Fiscal Impacts of Insucient Retirement Savings in Colorado (2020), Econsult Solutions. <https://econsultsolutions.com/wp-
content/uploads/2020/02/The-Fiscal-Impacts-of-Insucient-Retirement-Savings-in-Colorado-Feb-2020.pdf> and The Impact of Insucient
Retirement Savings on the Commonwealth of Pennsylvania, Econsult Solutions (2018). <https://patreasury.gov/pdf/Impact-Insucient-
Retirement-Savings.pdf>
26 Annual savings required to close the income suciency gap for an average elderly household with income of less than $75,000, following
standard investment assumptions, a at annual savings level over 30 years, an investment return of approximately 6.5% annually, and an
annual drawdown on savings of 4.5%. See additional detail in The Fiscal Impacts of Insucient Retirement Savings in Colorado (2020),
Econsult Solutions.
27 See information maintained by the Center for Retirement Initiatives at Georgetown University. < https://cri.georgetown.edu/states/>
28 MEPs fall within the scope of plans regulated by the Employee Retirement Income Security Act (ERISA) of 1974. However, a state-
facilitated structure enables a state board or similar entity to serve as the named duciary of the a plan, limiting the administrative burden
and risk on the participating employers. For more information about the state-level MEP model, see Morse and Antonelli (2017), Multiple
Employer Plans: An Overview of Legal, Regulatory and Plan Design Considerations for the States. Center for Retirement Initiatives at
Georgetown University. <https://cri.georgetown.edu/wp-content/uploads/2017/08/CRI_MEP_PolicyReport17-2.pdf> Any federal 401(k)
universal access proposal would address the requirement for employers to oer a plan.
29 The Colorado Secure Plan Savings Board, for example, evaluated marketplace eorts at the federal and state levels (most notably in
Washington State) and concluded that “the trend data on coverage indicate that these programs have not led to a signicant expansion of
coverage on either a nationwide or state-wide basis” as part of its recommendation to pursue an alternative approach. Recommendations
to Increase Retirement Savings in Colorado (2020), Colorado Secure Plan Savings Board, p. 23. <https://www.colorado.gov/pacic/sites/
default/les/atoms/les/CSSP_Retirement%20Security%20in%20Colorado_02-28-2020.pdf>
30 Setting Every Community Up for Retirement Enhancement (SECURE) Act of 2019, Pub.L. 116–94. <https://www.congress.gov/116/
plaws/publ94/PLAW-116publ94.pdf>
31 Mitchell and Szapiro (2020), Paperwork or Panacea: As PEPs Come of Age, What Can Their Forebearers Tell Us About How They Will
Work? Morningstar Policy Research. <https://www.morningstar.com/lp/paperwork_or_panacea>
32 Information about provisions of SECURE Act 2.0 from The Securing a Strong Retirement Act of 2020 — Expanding Coverage and
Increasing Retirement Savings – Section-by-Section Summary (2020), House Committee on Ways & Means. <https://waysandmeans.house.
gov/sites/democrats.waysandmeans.house.gov/les/documents/2.0Sectionbysection_nal.pdf>
33 See: Iwry, John, and Gale (2020), The SECURE Act: A Good Start but Far More Is Needed. Brookings Institution. <https://www.brookings.
edu/blog/up-front/2020/01/08/the-secure-act-a-good-start-but-far-more-is-needed/>
34 Details about the structure of the existing Saver’s Credit from Retirement Savings Contributions Credit (2020). Internal Revenue Service.
<https://www.irs.gov/retirement-plans/plan-participant-employee/retirement-savings-contributions-savers-credit>
35 Brown and John (2017), Improving the Saver’s Credit for Low- and Moderate-Income Workers. AARP Public Policy Institute. <https://
www.aarp.org/content/dam/aarp/ppi/2017/09/improving-the-savers-credit-for-low-and-moderate-income-workers.pdf>
36 The Use of Tax Incentives for Retirement Saving in 2006 (2011). Congressional Budget Oce. <https://www.cbo.gov/sites/default/
les/112th-congress-2011-2012/reports/2011-10-14-TaxIncentives.pdf>
37 Gale, Iwry, and Orszag (2004), The Saver’s Credit: Issues and Options (2004). Brookings Institution. <https://www.brookings.edu/wp-
content/uploads/2016/06/20040503.pdf>
38 Brown and John, Improving the Saver’s Credit (2017).
39 For an overview of policy proposals, including the Simpson-Bowles plan, Sperling proposal, and others, see: Busette and Eizenga,
A Small Change to the Saver’s Credit Can Go a Long Way (2012). Center for American Progress. <https://www.americanprogress.org/
wp-content/uploads/issues/2012/01/pdf/small_change_savers_credit.pdf> For information about the Biden Campaign Plan, see: Biden
proposes 401(k) changes to give low-income savers bigger tax benets (2020), CNN. <https://www.cnn.com/2020/09/19/politics/biden-plan-
retirement-savings-tax-benets/index.html>
40 House Committee on Ways & Means (2020), Securing a Strong Retirement Act of 2020 Section-by-Section Summary.
41 For a summary, see: Rooney (2018), After the crisis, a new generation puts its trust in tech over traditional banks. CNBC. <https://www.
cnbc.com/2018/09/14/a-new-generation-puts-its-trust-in-tech-over-traditional-banks.html>
42 For an overview of research on the inuence of FinTech, see Agnew and Mitchell (2019), The Disruptive Impact of FinTech on Retirement
Systems. Pension Research Council, Wharton School, University of Pennsylvania. <https://pensionresearchcouncil.wharton.upenn.edu/wp-
content/uploads/2020/01/FinTech-Chapter-1-Agnew-and-Mitchell.pdf>
47
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
43 For more information about legislation and hearings related to expanding access to and strengthening retirement security for private
sector workers, see information maintained by the Center for Retirement Initiatives at Georgetown University. <https://cri.georgetown.edu/
federal-legislative-proposals/>
44 A comprehensive version of the proposal from 2007 is available through the Brookings Institution. Iwry and John (2007), Pursuing
Universal Retirement Security through Automatic IRAs. The Retirement Security Project. <https://www.brookings.edu/wp-content/
uploads/2016/06/04_universal_retirement_iwry_john.pdf>
45 Securing Our Financial Future: Report of the Commission on Retirement Security and Personal Savings (2016). Bipartisan Policy Center.
<https://bipartisanpolicy.org/report/retirement-security/>
46 Madland, Rowell, and Davis (2016), Improving Americans’ Retirement Savings Outcomes through the National Savings Plan. Center
for American Progress. <https://www.americanprogressaction.org/issues/economy/reports/2016/01/28/128146/improving-americans-
retirement-outcomes-through-the-national-savings-plan/>
47 Ghilarducci and James (2016), A Comprehensive Plan to Confront the Retirement Savings Crisis. <https://www.economicpolicyresearch.
org/images/Retirement_Project/Retirement_Security_Guaranteed_digital.pdf>
48 2020 Long-Term Budget Outlook (2020). Congressional Budget Oce. <https://www.cbo.gov/publication/56516>
49 Modeling assumes a degree of early adoption by employers before required participation dates and lesser participation levels (90–95%)
in the rst two years of required participation, followed by full compliance in subsequent years (see Methodology Appendix for further detail).
50 The Methodology Appendix can be accessed at: [https://cri.georgetown.edu/research]
51 For discussion of the potential impact of this nancial option, see: Auto-IRAs Could Help Retirees Boost Social Security Payments
(2018). Pew Charitable Trusts. <https://www.pewtrusts.org/-/media/assets/2018/03/auto_ira_accounts_and_social_security_brief.pdf>
52 Under the revised income limit, credit levels and qualied contribution limits envisioned in the SECURE Act 2.0 proposal, a single ler
earning up to $40,000 would be eligible for a 50% tax credit on qualied contributions of up to $3,000, with reductions in the qualied
contribution level on a sliding scale for higher incomes (see Methodology Appendix for further detail).
53 For discussion of the impact of automatic enrollment and other plan features on participation, see: Choi, et al. (2002), “Dened
Contribution Pensions: Plan Rules, Participant Choices, and the Path of Least Resistance,” Tax Policy and the Economy, vol. 16. National
Bureau of Economic Research Inc., p. 67–114. See also: Clark and Young (2018), Automatic Enrollment: The Power of the Default. Vanguard
Research. <https://institutional.vanguard.com/iam/pdf/CIRAE.pdf>
54 These estimates are derived from a combination of custom analysis of Current Population Survey (CPS) data on participation by age and
Social Security Administration analysis of Survey of Income and Program Participation (SIPP) data by rm size (see Methodology Appendix
for further detail).
55 It is recognized that in practice, retirement ages will vary by individual, and that trends in recent years have been toward higher labor
force participation among the elderly. Notably retirement decisions are not made independent of available assets, meaning that retirement
savings associated with the modeled program may, in practice, aect pre-existing patterns of retirement ages. Modeling should be
understood to reect the assets with which participants would arrive at age 65, whether or not they choose to exit the workforce at that time.
56 See ndings on the impact of various initial default contribution rates on participation from:
California Secure Choice – Market Analysis, Feasibility Study, and Program Design Consultant Services – Final Report to the California Secure
Choice Retirement Savings Investment Board (2016). Overture Financial. <https://cri.georgetown.edu/wp-content/uploads/2020/05/CA-Mar-
2016-MArket-Analysis-Feasibility-Study-nad-Program-Design-Report.pdf>
Report on Design of Connecticut’s Retirement Security Program (Appendix) (2015). Center for Retirement Research at Boston College.
<https://www.osc.ct.gov/crsb/docs/nalreport/Appendix_A_CRSB.pdf> Beshears, et al. (2009),
The Importance of Default Options for Retirement Saving Outcomes: Evidence from the United States. National Bureau of Economic
Research. <https://www.nber.org/system/les/working_papers/w12009/w12009.pdf>
57 For example, using sample “young saver” under the baseline Auto-IRA model described is this analysis would see their savings at age 65
reduced by more than 20% (reaching about $307,000 rather than $390,000) if they start savings at 3% and escalate to 8% over time, rather
than the 5% initial rate and auto-escalation to 10% assumed in the baseline Auto-IRA scenario.
58 This estimate is derived through analysis of contribution amounts and levels reported in the state program data from California and
Oregon, accounting for the eective tax rates in those states (see Methodology Appendix for further detail).
59 Each of the active state Auto-IRA programs (Oregon, Illinois, and California) uses a Roth IRA as the default investment options, although
traditional IRAs are available.
60 For a review of key determinants of fee levels, see: The Economics of Providing 401(k) Plans: Services, Fees, and Expenses, 2019 (July
2020). ICI Research Perspective. <https://www.ici.org/pdf/per26-05.pdf>
61 Ibid.
62 Roth IRAs are structured as post-tax contributions, meaning that funds can be withdrawn without tax implications once the saver is 59½
years old and the account is at least ve years old. In addition, earlier withdrawals can be made in some instances without penalty, if they are
for a qualied reason or are funded only through contributions and not account earnings.
63 Analysis of program data from CalSavers from Q3 2019–Q2 2020 shows that savers under age 45 (as indicated by Target Date Fund
selections) withdraw assets at about 1.5x the rate of contributions of savers over age 45 (see Methodology Appendix for further detail).
64 A 2019 Government Accountability Oce (GAO) analysis using administrative data from the Internal Revenue Service (IRS) and
Department of Labor (DOL) represents the most recent and comprehensive review identied of early withdrawal behavior among savers in
the US. This study and its implications for early withdrawals among universal access program participants are discussed at length in the
Methodology Appendix. Jeszeck, et al. (2019). Retirement Savings: Additional Data and Analysis Could Provide Insight into Early Withdrawals.
United States Government Accountability Oce. <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3389474>
48
© 2020 Georgetown University Center for Retirment InitiativesWhat are the Potential Benets of Universal Access to Retirement Savings?
65 Data on the mix of asset types are drawn from research published by the Employee Benets Research Institute (EBRI) on asset mixes by
age group, while anticipated returns for those assets are drawn from a mix of historic performance and market forecasts (see Methodology
Appendix for further detail). Vander Hei, et al. (2018). 401(k) Plan Asset Allocation, Account Balances, and Loan
Activity in 2016. Employee Benets Research Institute. <https://www.ebri.org/docs/default-source/ebri-issue-brief/ebri_ib_458_k-update-
10sept18.pdf?sfvrsn=bca4302f_6>
66 Antonelli, et al. (2019), Generating and Protecting Retirement Income in Dened Contribution Plans: An Analysis of How Dierent
Solutions Address Participant Needs. Georgetown University Center for Retirement Initiatives, in conjunction with WillisTowerWatson.
<https://cri.georgetown.edu/wp-content/uploads/2019/06/policy-report-19-02.pdf>
67 This calculation is based on market rates reected in the 2019 CRI analysis in conjunction with WillisTowerWatson, adjusted to reect the
declining value of a consistent nominal income stream in real terms (see Methodology Appendix for more detail).
68 These scenarios are described in more detail in Section 1.2 of this report, with complete detail on each found in the Methodology
Appendix.
69 See results in Recommendations to Increase Retirement Savings (2020). Colorado Secure Plan Savings Board, cited previously in
this report, which show limited impacts to coverage levels of a voluntary program such as a marketplace model, as well as limitations of
expanded MEP availability in expanding coverage as raised in Iwry, John, and Gale (2020), The SECURE Act: A Good Start.
70 See information on state program regulations on rm size and age thresholds from: CalSavers Retirement Savings Program (2020). State
of California Employment Development Department. <https://edd.ca.gov/employers/calsavers.htm> See also: Program Details – Facilitate
Oregon Saves (2020). Oregon Saves. <https://employer.oregonsaves.com/home/employers/program-details.html> As well as: Iekel (2019),
Mid-Size Employers Must Register for Illinois Secure Choice by July 1. National Association of Plan Advisors. <https://www.napa-net.org/
news-info/daily-news/mid-size-employers-must-register-illinois-secure-choice-july-1>
71 See: Automatic IRA Act of 2019, S.2370, 116th Cong. (2019). <https://www.congress.gov/bill/116th-congress/senate-bill/2370/text> and:
Automatic IRA Act of 2017, H.R.3499, 115th Cong. (2017). <https://www.congress.gov/bill/115th-congress/house-bill/3499/text>
72 See: Automatic Retirement Plan Act of 2017, H.R.4523, 115th Cong. (2017). <https://www.congress.gov/bill/115th-congress/house-
bill/4523/text>
73 For additional detail on the phase-in approach for state programs, see: Program Details – Facilitate Oregon Saves (2020). Oregon Saves.
See also: CalSavers History: From Pioneering Vision to Launch (2020). California State Treasurer’s Oce. <https://www.treasurer.ca.gov/
calsavers/history.asp> and CalSavers Retirement Savings Program (2020). State of California Employment Development Department. See
also: State-Facilitated Retirement Savings Programs for Private Sector Workers (2018). National Conference of State Legislatures. <https://
www.ncsl.org/research/scal-policy/state-facilitated-retirement-savings-programs-for-private-sector-workers.aspx> and Iekel (2019), Mid-
Size Employers Must Register for Illinois Secure Choice.
74 See Scott and Hines (2020), Employers Express Satisfaction with New Oregon Retirement Savings Program. Pew Charitable Trusts.
<https://www.pewtrusts.org/en/research-and-analysis/articles/2020/07/30/employers-express-satisfaction-with-new-oregon-retirement-
savings-program>
75 Estimates are derived from analysis of Census Bureau Business Dynamics Statistics on rm size and tenure, which are matched with
more-recent Census Bureau Quarterly Workforce Indicators data on employment by rm size (see Methodology Appendix for further detail).
76 Data from Vanguard reported in Munnell and Chen (2020), 401(k) / IRA Holdings in 2019.
77 Automatic Retirement Plan Act of 2017, H.R. 4523.
78 This employer contribution level should be understood as a blended rate, which most employers expected to decline to contribute
(eective rate of zero), and higher eective rates relative to employee contributions among those that choose to oer matching or
supplemental employer contributions.
79 The structure of CalSavers was armed as exempted from ERISA by a federal District Court in a March 2020 decision. See: Wille (2020),
California’s Auto-Retirement Program Upheld from ERISA Challenge. Bloomberg Law. <https://news.bloomberglaw.com/employee-benets/
californias-auto-retirement-program-upheld-from-erisa-challenge>
80 Fact Sheet: Adjusting ERISA Civil Monetary Penalties for Ination (2020). US Department of Labor. <https://www.dol.gov/sites/dolgov/
les/ebsa/about-ebsa/our-activities/resource-center/fact-sheets/adjusting-erisa-civil-monetary-penalties-for-ination.pdf>
81 Multiple Employer Plans (MEPs) and Pooled Employer Plans (PEPs) allow groups of businesses to join together to form a single
retirement plan through a 401(k) savings vehicle. MEP arrangements allow related businesses to join to form a plan, whereas PEP
arrangements allow unrelated businesses to join to form a plan. MEP and PEP structures allow businesses to pool their resources and
outsource plan management to a sponsoring entity that is responsible for administrative duties and takes on duciary liability for the plan. For
additional detail on the expansions to MEPs and PEPs in the SECURE Act, see: Setting Every Community Up for Retirement Enhancement
(SECURE) Act of 2019, Pub.L. 116–94.
82 Tax credit provisions to address start-up costs are included in both the Automatic IRA Act of 2019 (S. 2370) and the Automatic
Retirement Plan Act of 2017 (H.R. 4523).
83 IRS data indicate that 64% of employers set an age requirement of 21 for participation, with an additional 16% setting a requirement
of 18–20 years, and 20% setting no requirement. Section 401(k) Compliance Check Questionnaire: Interim Report (2012), Internal Revenue
Service, Employee Plans Compliance Unit, Figure 2. <https://www.irs.gov/pub/irs-tege/401k_interim_report.pdf>
84 See: Simple IRA Plans for Small Businesses (2019). Employee Benets Security Administration (EBSA). US Department of Labor and
Internal Revenue Service. <https://www.dol.gov/sites/dolgov/les/EBSA/about-ebsa/our-activities/resource-center/publications/simple-ira-
plans-for-small-businesses.pdf> See also: Retirement Topics - SIMPLE IRA Contribution Limits (2020). Internal Revenue Service. <https://
www.irs.gov/retirement-plans/plan-participant-employee/retirement-topics-simple-ira-contribution-limits>
49
© 2020 Georgetown University Center for Retirment Initiatives What are the Potential Benets of Universal Access to Retirement Savings?
85 See the Appendix for further discussion, as well as the following study, which evaluates the relative wage sensitivity of higher and lower
income workers, as well as wage sensitivity by gender: Smith and Toder (2011), Do Low-Income Workers Benet from 401(k) Plans? Urban
Institute. <https://www.urban.org/sites/default/les/publication/26771/412463-Do-Low-Income-Workers-Benet-from--k-Plans-Full-Report-.
PDF>
86 According to the Bureau of Labor Statistics’ National Compensation Survey, 52% of private rms with fewer than 50 employees oer
some form of retirement benets compared to 95% for rms with more than 100 employees. National Compensation Survey: Employee
Benets in the United States, March 2019 (2019). US Bureau of Labor Statistics, p. 179. <https://www.bls.gov/ncs/ebs/benets/2019/
employee-benets-in-the-united-states-march-2019.pdf>
87 Based on elasticities estimated by Smith and Toder (2011), adjusted to the characteristics of the uncovered employee population,
employers are anticipated to bear about 60% of the incidence of each marginal dollar of direct costs. See the Methodology Appendix for
further detail on this calculation and research.
88 Research shows that employees value retirement benets, making them more likely to stay at a rm that oers them. See for example:
Thriving in the New Work-Life World: MetLife’s 17th Annual US Employee Benets Trends Survey 2019 (2019). MetLife. <https://www.metlife.
com/content/dam/metlifecom/us/ebts/pdf/MetLife-Employee-Benet-Trends-Study-2019.pdf> as well as employer focus group research
such as Business Owners’ Perspectives on Workplace Retirement Plans and State Proposals to Boost Savings (2016). Pew Charitable Trusts.
<https://www.pewtrusts.org/en/research-and-analysis/issue-briefs/2016/09/business-owners-perspectives-on-workplace-retirement-plans-
and-state-proposals-to-boost-savings>
89 See: McKinsey & Company (2020), Future of Work. <https://www.mckinsey.com/featured-insights/future-of-work>
90 Estimates of enhanced participation in savings in this analysis include only traditional employees, meaning that enhanced savings among
these independent workers outside of periods of traditional employment would produce additional benets beyond those modeled.
91 Ebrahimi (2019), How Do Retirees’ Spending Patterns Change Over Time? Employee Benet Research Institute. <https://www.ebri.org/
docs/default-source/ebri-issue-brief/ebri_ib_492_spendovertime-3oct19.pdf?sfvrsn=9f503c2f_10>
92 A state-level analysis by ESI projected that insucient resources will reduce spending by retiree households in Pennsylvania by
more than $3 billion in 2030 under current trends. These reductions have further spillover eects on businesses, their supplies, and their
employees, reducing output in the state by around $4.3 billion and shrinking the economy by about 30,000 jobs. The Impact of Insucient
Savings in Pennsylvania (2018). Econsult Solutions.
93 This analysis relies on standard macroeconomic models of the relationship between national savings behavior and economic growth, as
dened in the “Solow Growth Model” developed by Nobel-laureate Economist Robert Solow. Notably, population and employment growth
estimates are held consistent across program scenarios, meaning that increased growth translates to an increased national standard of
living for the same pool of future residents. This framework and modeling approach is discussed at length in the Methodology Appendix.
Solow (1956), “A Contribution to the Theory of Economic Growth.” Quarterly Journal of Economics, vol. 70(1), pages 65–94. This framework
is consistent with approaches to modeling potential economic growth from national savings used by the Government Accountability Oce
(GAO) and Congressional Research Service (CRS). See: National Saving: Answers to Key Questions (2001). Government Accountability
Oce. < https://www.govinfo.gov/content/pkg/GAOREPORTS-GAO-01-591SP/pdf/GAOREPORTS-GAO-01-591SP.pdf> Saving in the
United States: How Has It Changed and Why Is It Important? (2003). Congressional Research Service. <https://www.everycrsreport.com/
les/20030117_RL30873_dcb36d191b110ddc22915ebca58c7126b629e30e.pdf>
94 2020 Long-Term Budget Outlook (2020). Congressional Budget Oce. <https://www.cbo.gov/publication/56516>
95 Annual total federal expenditures per program in 2020 are dened by extrapolating forward Federal Fiscal Year (FFY) 2018 budget
data presented in Federal Spending on Benets and Services for People with Low Income: FY2008–FY2018 Update (2020). Congressional
Research Service. Expenditures on the elderly population (65+) are isolated using budget and program data sources detailed by program in
the Methodology Appendix.
96 2019 Long-Term Budget Outlook (2019). Congressional Budget Oce. <https://www.cbo.gov/system/les/2019-06/55331-LTBO-2.pdf>
See Methodology Appendix for further discussion of this calculation.
97 This calculation is undertaken through a longitudinal comparison of changes to the income distribution of working age and retiree cohorts
over time, using data income date from the Current Population Survey and decennial Census (see Methodology Appendix for further detail).
98 In Section 3 of the Methodology Appendix, additional detail is provided on the approach and data sources used to allocate program
expenditures across elderly households of dierent income levels. See in particular, Figure 3.5, which outlines the specic data sets used for
each program.
99 The referenced state-level studies conducted by ESI in Pennsylvania and Colorado used a dierent modeling framework to quantify the
scal “cost of doing nothing.” These studies do not examine the impact of any potential policy or scenario, but rather, estimate the scal cost
of the total gap in retirement savings suciency in the state. For this reason, the order of magnitude of government expenditure impacts in
these studies cannot be compared in an apples-to-apples manner with the results of this study. See the full studies for additional detail: The
Fiscal Impacts of Insucient Savings in Colorado (2020). Econsult Solutions, and The Impact of Insucient Savings in Pennsylvania (2018).
Econsult Solutions.
Endnotes
The inside back cover TBD (if print, could be left intentionally blank;
digital edition could be last page of endnotes)
