What If Low-Income American Workers Had Access to Wealth

What If Low-Income American

Workers Had Access to Wealth-

Building Vehicles Like the Federal

Employees’ Thrift Savings Plan?

Teresa Ghilarducci

The New School

Kevin A. Hassett

The Lindsey Group and

the Hoover Institution

1. Authors are listed alphabetically.

PUBLISHED March 25, 2021

INCLUSIVE WEALTH BUILDING INITIATIVE

Economic Innovation Group Inclusive Wealth Building Initiative

Table of Contents

I. Introduction

II. Policies to Stimulate Savings

III. The Thrift Savings Plan for Federal Employees

IV. TSP Serves as A Promising Model for

Building Wealth

V. Simulating the Wealth Generation Eﬀects

of the Proposal for Low-Income Individuals

VI. Conclusion

VII. References

INCLUSIVE WEALTH BUILDING INITIATIVE

Economic Innovation Group Inclusive Wealth Building Initiative

SECTION I

Although America’s market economy has been the global leader in wealth creation, a large share of

Americans nevertheless have no direct stake in rising national growth and prosperity. Speciﬁcally, the

bottom 50 percent of households own only 1.5 percent of total U.S. wealth (World Inequality Database

n.d.). The situation is particularly troubling for families at the bottom of the wealth distribution: The

bottom 25 percent of all Americans have a median net worth of merely $310, leaving them without the

security and well-being that wealth bestows (see table 1). Thus, it is a small wonder that sentiments

regarding capitalism are so mixed given how little capital a large share of the population owns.

The COVID-19 recession has laid bare the reality

and consequences of the lack of wealth among

millions of workers in the bottom half of the income

distribution, who have disproportionately lost their

jobs with no meaningful assets to fall back on absent

emergency aid.

Table 1 Distribution of Wealth and Income by Wealth Percentiles, 2019 (thousands)

Bottom 25%

25-49.9%

50-74.9%

75-89.9%

90-100%

Wealth percentiles

Median net

worth

$0.31

$57.60

$224.20

$653.10

$2,598.40

Source: Board of Governors of the Federal Reserve System 2019b.

Mean net

worth

-$13.63

$58.18

$236.28

$703.59

$5,710.34

Median

income

$29.53

$46.83

$70.25

$101.81

$262.20

Mean

income

$37.91

$58.21

$80.57

$129.94

$425.85

Economic Innovation Group Inclusive Wealth Building Initiative

2. By the time workers are in their 50s, many have found it nearly impossible to have saved enough for their looming retirement (Ghilarducci, Papadopoulos, and Webb 2018). Among workers nearing

retirement,15 percent of high-income workers (those who earn $115,000 plus per year), 20 percent of middle-class workers, and 50 percent of lower-income workers have no retirement savings and will rely only

on Social Security (Ghilarducci, Papadopoulos, and Webb 2017). And many older workers won’t be able to work longer. They are already being pushed out to a too early retirement. Sixty percent of older workers

leaving the labor force in April say they were retired, up from 53 percent pre-pandemic (Miller 2020).

3. TSP participants without a high school degree on average contribute 5.1 percent of their income into their TSP plan, while those in the bottom one-third contribute 4.6 percent (Falk and Karamcheva 2019).

SECTION I

A major component of this lack of wealth among middle- and low-income families is that millions

of Americans have little to no savings set aside for retirement. For the bottom 50 percent of the wealth

distribution, the median retirement savings account balance is $0 (Board of Governors of the Federal

Reserve System 2019b). The United States continues to have one of the highest elder poverty rates among

rich nations in the world (OECD n.d.). Furthermore, the retirement system exacerbates embedded gender

and racial wealth disparities: Women are 80 percent more likely than men to be impoverished at age

65 and older (Brown et al. 2016). Moreover, only 41 percent of Black families and 35 percent of Hispanic

families, respectively, had any amount of retirement account savings in 2016 (Morrissey 2019). According

to some estimates, the retirement plan coverage rate among American workers in 2019 was between

35 and 49 percent (Radpour, Papadopolous, and Ghilarducci 2021). Even some high-income earners—

through no fault of their own—lack access to a dependable, convenient, and aﬀordable way to practically

save for retirement.

For the most part, tax policies regarding retirement savings hinge on granting tax beneﬁts through

deductions, as opposed to credits or direct matches. The deduction approach is regressive because

those with the highest incomes and the highest marginal tax rate gain the most. Two workers saving the

maximum amount allowed in their retirement accounts have wildly diﬀerent levels of ﬁnancial help

from the government. The minimum wage worker whose income is so low that they do not deduct or pay

income taxes gets no additional beneﬁt for saving. By contrast, a worker earning enough to pay the top

marginal rate can shelter large amounts of money (since they put the money away pre-tax) and receives

an implicit 37 percent subsidy to do so. A dollar saved today is a dollar that does not face taxation.

Federal, state, and local governments have been successful in providing eﬃcient and adequate savings

vehicles for their workers, as have some private ﬁrms. One program stands out as remarkably successful

in stimulating wealth creation for low- and middle-income individuals: The federal employees’ Thrift

Savings Plan (TSP) system is designed with the latest scientiﬁc research about savings behavior and

insurance demand in mind. The TSP’s superior bargaining power yields low expense ratios that allow for

higher expected returns than in the commercial 401(k) and IRA market. TSP participation jumped when

matching and automatic enrollment, respectively, were introduced, suggesting that these features could

have similar eﬀects on all workers, especially those with a high school degree or less and those who are in

the bottom one-third of the earnings distribution.

The positive impact of the plan has also been shown to

be high for African Americans and women.

In this policy brief, we (1) summarize the academic literature that explores the impact of diﬀerent

policies on the savings behavior of low-income individuals, (2) relate the lessons from this literature to

evidence regarding the eﬀectiveness of the TSP program, (3) sketch a policy that makes TSP available to

all low-income individuals, and (4) simulate the wealth creation that such a policy could generate. We

ﬁnd that this policy could signiﬁcantly increase wealth for those at the bottom.

For the bottom 50 percent of the wealth distribution, the median

retirement savings account balance is $0.

(Board of Governors of the Federal Reserve System 2019b)

Economic Innovation Group Inclusive Wealth Building Initiative

The question of whether ﬁscal policy can be used to address savings inadequacy has been one of

the key focuses of the economics profession for decades. The research has had an additional urgency

because, beginning with the seminal work of Feldstein and Horioka (1980), individual and household

savings behavior has been documented as being crucial for national-level capital formation and growth.

The link between tax policy and savings, thus, is one of the key links relating government policy to the

welfare of nations. Throughout that time, policymakers have given academics an enormous panoply of

policies to study, and the research has exploded in numerous directions.

An important and path-breaking early corner of the literature involved the impact of employer

deﬁned-beneﬁt plans and wealth. By the 1970s, a substantial portion of the private and public sector

workers were covered by employer retirement plans. Though not liquid, these plans provided substantial

retirement wealth to workers who likely would not have otherwise had long-term savings.

In the 1980s, Individual Retirement Accounts (IRAs) made savings for retirement tax deductible and

thus, in theory, provided a strong government incentive for increasing savings. However, there were

theoretical and empirical reasons to suspect that the eﬀects of IRAs on the overall level of savings in

the economy could be ambiguous. At the aggregate level, a government incentive to increase household

savings could have an oﬀsetting eﬀect: The resulting reduced tax revenue and increased ﬁscal deﬁcit

could oﬀset some of the increased private household savings. Also, adding to the ambiguity, at the

household level, saving into an IRA might simply be transferring existing savings from a taxable to a

non-taxable account, and thus, a tax deduction for IRAs might not increase savings in the aggregate.

A series of papers documented that tax-favored accounts, such as an IRA, do not increase aggregate

savings (e.g., Engen, Gale, and Scholz 1996), while another set of papers (e.g., Poterba, Venti, and Wise

1996) documented that they do. In addition to the ambiguous aggregate impact of savings incentives, the

impact on individuals from this ﬁrst round of literature is also uncertain. But in any case, to the extent

that there was an impact on household savings, the distributional eﬀects were regressive. The action in

the accounts reviewed in this literature was at the middle to higher-end of the income distribution, where

households are already more likely to be active savers. As such, a proposal that boosts savings among

low-income households—who are statistically unlikely to save otherwise—will likely increase savings at

the aggregate and household level, respectively. More recent evidence, discussed below, suggests that

modern policy innovations have had more success at generating positive net savings.

A second related body of research explored the impact of policies that incorporate behavioral

insights. An important early contribution is attributable to Laibson (1997), who shows that an individual’s

SECTION II

Policies to Stimulate Savings

A proposal that boosts savings among low-income households—who

are statistically unlikely to save otherwise—will likely increase savings

at the aggregate and household level, respectively.

Economic Innovation Group Inclusive Wealth Building Initiative

shortsightedness and bias towards the present may undermine savings decisions, and thus make policies

to stimulate savings ineﬀective if they ignore behavioral eﬀects. Laibson’s original insight set oﬀ an

explosion of work that suggested that savings behavior is modiﬁable by policy. For our purposes, there are

two main strands of that literature that are impactful.

First, because people might be especially focused on their short-run well-being, as Laibson suggested,

it could be that increasing the immediate beneﬁt of long-term investing has a positive impact on savings.

If you put $1.00 away today and only get $1.10 tomorrow, maybe your desire for immediate gratiﬁcation

makes that option unattractive. But if you put $1.00 away today and get an immediate match from your

employer, perhaps you would respond diﬀerently, even if the matching amount is the same in present

value as the long-run tax savings from a more traditional vehicle.

Empirical studies were able to investigate such scenarios beginning in the early 1990s (e.g., Andrews

1992) because employers began to sponsor deﬁned contribution savings plans that contained employer

matches. Focus only on today and you get the satisfaction of consuming today, but you do not get the

generous match that the employer provides if you place the money in a retirement account. An explosion

of papers (e.g., Papke and Poterba 1995; Clark and Schieber 1998; Mitchell, Utkus, and Yang 2007)

demonstrated a strong link between employer matches and increases in savings. Perhaps the canonical

and most decisive paper in this area was produced by Duﬂo et al. (2006), who analyzed a controlled

experiment. They oﬀered some participants a chance to use their tax refund to create a retirement

account with a match, while others were randomly assigned the same opportunity, but without a match.

The percentage of participants opening an IRA increased dramatically with the level of the match.

The second strand of savings research explored another implication of the original Laibson (1997)

observation. If people have limited attention, maybe individuals will be more willing to contribute to

retirement accounts if they are automatically enrolled in accounts and can only not save if they opt out

of the default program. The ﬁrst and pioneering study of this approach was performed by Madrian and

Shea (2001), who found that participation in retirement plans increased by 50 percent when employers

automatically enrolled new employees. Madrian (2015) reviews the now ample literature that followed,

and the evidence is clear that automatic enrollment has a large eﬀect. Madrian (2015) also ﬁnds that the

parameters of the employee’s retirement plan—such as the percentage of income contributed and the

portfolio choice—are highly responsive to the default parameters chosen by the employers. Other work

has shown that the lack of retirement savings could be motivated by lack of wage growth and that fear,

more than inertia, prevented voluntary savings (Ghilarducci, Saad-Lessler, and Reznik 2018).

SECTION II

Economic Innovation Group Inclusive Wealth Building Initiative

The Thrift Savings Plan (TSP), created in 1987, is a deﬁned contribution savings plan available

to federal employees and members of the uniformed services. Several elements of the TSP (e.g., the

level of the employer match) have evolved signiﬁcantly over time. The recent implementation of a

higher automatic match is relevant for our argument that lower-income workers need an automatic

and relatively generous match. For employees hired after October 1st, 2020, 5 percent of their pay is

automatically contributed to the TSP account with a government match. For those who began their

federal service between August 1st, 2010, and September 30th, 2020, their agencies automatically enrolled

them at a rate of 3 percent (TSP n.d.). Contributions into the accounts are matched by the employers, with

the ﬁrst 3 percent of contributions matched 100 percent, and the next 2 percent of contributions matched

at a 50 percent rate. The matches stop for contributions above 5 percent. Employees’ own contributions to

the accounts are vested immediately, whereas the matching contributions are gradually vested over time.

The TSP funds can be invested in what plan administrators refer to as ﬁve “core” funds, or in a

“lifecycle” fund that varies the mix of the ﬁve core funds depending on the expected retirement date of

the plan participant.

Employees that are automatically enrolled have their funds placed in a diversiﬁed

lifecycle fund that is appropriate for the age group, and thus, will usually have their assets spread out

among the various funds. For example, most younger participants are in the lifecycle fund, and most

older participants are in the “G fund,” which is guaranteed by the federal government not to lose money

and only available to federal employees in the TSP (TSP 2020). Participants are able to roll over or transfer

qualiﬁed money from other individual or employer-sponsored retirement accounts into the TSP, but

without the match.

The TSP funds are distinct from other investments because of their enormous scale, with their

combined assets easily making up the largest employer-based deﬁned contribution plan in the country.

As such, plan managers are able to keep administrative costs exceptionally low, with the latest estimate

for expenses being only 0.042 percent, compared to an expense ratio of 0.91 percent for similar funds

(The Vanguard Group n.d.). Second, the plan’s “G fund” is an unusually attractive cash-like vehicle that

oﬀers investors exposure to risk, but pays an interest rate that matches the average yield on intermediate

term treasuries.

The TSP system is designed to incorporate the following best practices:

• Low expense ratios;

• Automatic enrollment;

• Employer match up to 5 percent;

• Simple asset allocation option for automatically enrolled;

• Allows roll-overs to help retirees consolidate all retirement accounts;

• Non-conﬂicted managers and not for proﬁt administration; and

• Bargaining power is used to get the best funds at the lowest prices.

The Thrift Savings Plan for Federal Employees

4. The five core funds include the “G” fund ($243.8 billion under management as of 12/31/2019), which invests in government securities, the “F” fund ($33.5 billion under management as of 12/31/2019),

which invests in fixed income securities that track the Bloomberg Barclays U.S. aggregate bond index, the “C” fund ($226.9 billion under management as of 12/31/2019), which invests in common stocks that

are expected to match the performance of the S&P 500, the “S” fund ($74.1 billion under management as of 12/31/2019), which invests in smaller capitalization stocks, and the “I” fund ($54.3 billion under

management as of 12/31/2019), which invests in international stocks.

SECTION III

Economic Innovation Group Inclusive Wealth Building Initiative

Put it all together, and extending the TSP model would have four major advantages for

the American worker, especially those in the lower end of the earnings distribution,

relative to current law:

1. The eﬃciency of a large fund would yield higher, risk-adjusted beneﬁts compared

to plain IRAs or 401(k) plans because of lower fees and professionally-managed

diversiﬁed portfolios;

2. The availability of non-commercial government annuities would yield cost-eﬀective

beneﬁts for life;

3. Participants would gain peace of mind from non-commercial, transparent governance;

and furthermore

4. Contributions could be consistent during times of unemployment, informal work, and

unpaid care work (i.e., the expanded TSP plan could accept money from people who do

not have a regular employer).

Its low expense ratios promise higher expected returns. Automatic enrollment has been shown to be

highly eﬀective at stimulating plan participation. The matching by the employer at a relatively high share

of income (now 5 percent) signals to workers that such a contribution is “correct,” so even if the impact

of matching is itself not so large, it is likely that automatically enrolled individuals would “anchor” at the

matching limit and save that amount. The matching of the contribution up to 5 percent of income makes

the default contribution ﬁnancially attractive as well, while the IRA-like structure (which can take either

standard or Roth treatment) provides a tax incentive and subsidy.

The simplicity of the program, wherein the automatically enrolled individual is given a carefully

documented and predetermined asset allocation option from a very simple set of assets, makes the

system more transparent and trustworthy. The TSP also helps people consolidate all their retirement

accounts because the TSP allows retirees to transfer qualiﬁed money (meaning retirement accounts) from

other individual or employer-sponsored retirement accounts into the TSP. Research shows that trust

matters in a participant’s desire to invest. Key to the trust that participants have in the TSP system is that

they know their employer—the federal government—is working with the asset managers to eliminate

conﬂicted advisors and remove the proﬁt motive from the administration of the fund (Turner, Klein, and

Stein 2016). Finally, the federal government uses its size and expertise to bargain for the best funds at the

lowest prices.

SECTION III

Economic Innovation Group Inclusive Wealth Building Initiative

The TSP’s success, as the largest deﬁned contribution plan in the country, is documented by a 2019

Congressional Budget Oﬃce (CBO) study, which showed that employee participation rates increased by

22 percentage points after the introduction of the employer match and by 19 percentage points after the

introduction of automatic enrollment (Falk and Karamcheva 2019).

The positive impact was especially prominent for those in the bottom one-third of the earnings

distribution, many of whom are nonwhite and have lower educational attainment. Among federal

employees who have a high school degree or less, the participation rate in the TSP is 64 percent for

those who were not oﬀered the match and 86 percent for those who were oﬀered (Falk and Karamcheva

2019). The participation rate for employees with some college experience increased from 64 percent to

90 percent when the government match increased. Low-income workers seem especially responsive

to employer matches, with only 55 percent of those in the bottom one-third of the income distribution

participating in the program before the match, and fully 85 percent participating after the match (Falk

and Karamcheva 2019).

As for the impact of automatic enrollment, it also appears to signiﬁcantly boost participation among

traditionally marginalized workers (e.g., workers with less educational attainment and lower-wage

workers). Fully ﬁve years after the TSP automatic enrollment began, those with a high school degree or

less increased their participation from 79 percent to 95 percent, while those in the bottom one-third of

earnings increased their participation from 74 percent to 95 percent. Furthermore, these changes were

also associated with large changes in average contribution rates, jumping from 4.7 to 5.1 percent for

those with a high school degree or less, and from 4.0 to 4.6 percent for those in the bottom one-third of

earnings. Higher participation and higher contributions lead to larger TSP balances relative to pay, with

these increasing by 3 percentage points for workers with a high school degree or less, and 4 percentage

points for workers in the bottom one-third of earnings (Falk and Karamcheva 2019).

Finally, the simplicity of the program, wherein the automatically enrolled individual is given a

carefully documented and predetermined asset allocation and picks from a very simple set of assets,

is also likely a positive. Indeed, Choi, Laibson, and Madrian (2009) studied a speciﬁc policy “Quick

Enrollment,” which to some extent mimicked this feature of TSP, and found that simplicity signiﬁcantly

increased participation. Accordingly, a consumer of the literature might expect that the plan would be

quite successful at inducing federal employees to save, just as the evidence suggests.

TSP Serves as A Promising Model for Building Wealth

SECTION IV

A TSP program study showed employee participation rates

increased by 22 percentage points after the introduction of

the employer match and by 19 percentage points after the

introduction of automatic enrollment.

(Falk and Karamcheva 2019)

Economic Innovation Group Inclusive Wealth Building Initiative

Would a TSP-like program for lower-income Americans work? Lessons f

rom other

policy experiments

The federal workforce eligible for TSP skews higher-income than the population in

general, but recent evidence suggests that the behavioral economic factors driving the

success of the TSP are also impactful for people in the left tail of the income distribution.

McKernan et al. (2020), for example, reviewed evidence of a randomized evaluation

of a program called “Assets for Independence” (AFI), which divided a sample of low-

income individuals up into treatment and control groups, with the treatment group

receiving a match (up to a limit) of the savings they place in an account. The authors

concluded that the AFI program increased savings, reduced ﬁnancial hardships, increased

homeownership among renters, and even increased access to health care. The authors

review the broader literature on similar programs and ﬁnd many promising results;

for example, they ﬁnd that a program called “SaveUSA,” which provided matches to

individuals who agreed to put a portion of their tax refunds into a savings account,

induced most participants to open

an account and increase their savings (Azurdia et al,

2014).

SECTION IV

This evidence suggests that if a plan like TSP, combined with government matches, were available to

everyone, then households at the middle and the low end of the income and wealth distribution would

have signiﬁcantly more ﬁnancial wealth. The TSP system takes advantage of best practices identiﬁed

in the literature discussed in the previous section. Finally, it is important to add that these conclusions

follow only under the assumption that the savings accumulated in the TSP account do not disqualify

savers from receiving other government beneﬁts that are sometimes subject to asset tests. This would be

an important factor to include in any prospective policy.

Economic Innovation Group Inclusive Wealth Building Initiative

How might a more widely available TSP model impact the ﬁnances and wealth of low-income

Americans? In this section, we perform straightforward simulations of asset growth and wealth

accumulation assuming that people participate in the matching program. The second half of this section

presents a similar simulation that allows for some “leakage” from retirement savings (i.e., for some

money to be taken out prior to retirement to pay for an emergency expense). Given the uncertainties over

participation and leakage, these estimates are designed to provide reasonable bounds of the possible

impact of this policy on wealth accumulation.

For the sake of determining the match percent, we assume that a household with median wealth also

has median income, and that the income is composed of eligible earnings. This allows us to, based on the

tax ﬁles at OSPC.org, simulate the likely savings going forward of low-income workers should the policy

be adopted. We also assume that all of the income is match-eligible earnings. This could be improved

with further analysis of the microdata. We consider two alternative rate-of-return scenarios, one in which

the nominal return on ﬁnancial assets is 3 percent, and one in which it is 7 percent.

Based on the latest wealth data, the median household in the bottom 25 percent of the wealth

distribution starts with $1,380 of ﬁnancial assets and $29,530 of income. Under the 3 percent nominal

return scenario, the household would have $1,850 of ﬁnancial assets after 10 years with no match and

no additional savings, $22,170 with 3 percent of earnings matched, $28,940 with 4 percent matched, and

$35,710 with 5 percent matched (see table 2). A key target of this policy, of course, would be young earners

starting out their working lives, so a look at longer time horizons is instructive as well. After 40 years,

the household would have $4,500 with no match and no additional savings, $138,100 with 3 percent of

earnings matched, $182,630 with 4 percent matched, and $227,160 with 5 percent matched.

Simulating the Wealth Generation Effects of the Proposal

for Low-Income Individuals

SECTION V

After 40 years, the household would have $20,660 with no

match and no additional savings, $374,380 with 3 percent of

earnings matched, $492,280 with 4 percent matched, and

$610,190 with 5 percent matched.

Economic Innovation Group Inclusive Wealth Building Initiative

Under the 7 percent return scenario, the household would have $2,710 of ﬁnancial assets after 10

years with no match, $27,190 with 3 percent of earnings matched, $35,350 with 4 percent matched, and

$43,510 with 5 percent matched. After 40 years, the household would have $20,660 with no match and no

additional savings, $374,380 with 3 percent of earnings matched, $492,280 with 4 percent matched, and

$610,190 with 5 percent matched. These ﬁgures are pre-tax and exclude the minimal fees a TSP-inspired

model would levy.

A key factor to consider as we simulate possible wealth creation is that individuals, especially lower-

income individuals, tend to need to access the wealth accumulated in the accounts as emergencies arise,

so that the wealth “leaks” from the accounts over time at a rate of up to 40 percent (Argento, Bryant,

and Sabelhaus 2015). Table 3 below presents ﬁnancial asset projections for the median household from

the bottom quarter of the wealth distribution with a 40 percent annual leakage from savings. Again, the

household starts with $1,380 of ﬁnancial assets and $29,530 of income. Under the 3 percent nominal

return scenario, the household would have $1,850 of ﬁnancial assets after 10 years with no match and

no additional savings, $14,040 with 3 percent of earnings matched, $18,100 with 4 percent matched, and

$22,170 with 5 percent matched. After 40 years, the household would have $4,500 with no match and no

additional savings, $84,660 with 3 percent of earnings matched, $111,380 with 4 percent matched, and

$138,100 with 5 percent matched. Thus, even when there is leakage, the assets accumulated under this

plan would be signiﬁcant.

SECTION V

Table 2 Financial asset projections: median family in the bottom

25 percent of the wealth distribution (thousands)

Year

3 percent nominal return 7 percent nominal return

match

$1.38

$1.60

$1.85

$2.15

$2.49

$2.89

$3.35

$3.88

$4.50

match

$1.38

$17.28

$35.71

$57.07

$81.84

$110.55

$143.84

$182.43

$227.16

match

$1.38

$15.52

$35.35

$63.17

$102.19

$156.91

$233.66

$341.30

$492.28

match

$1.38

$11.01

$22.17

$35.10

$50.10

$67.49

$87.64

$111.01

$138.10

match

$1.38

$1.94

$2.71

$3.81

$5.34

$7.49

$10.50

$14.73

$20.66

match

$1.38

$18.92

$43.51

$78.01

$126.40

$194.26

$289.45

$422.95

$610.19

match

$1.38

$14.14

$28.94

$46.09

$65.97

$89.02

$115.74

$146.72

$182.63

match

$1.38

$12.12

$27.19

$48.33

$77.98

$119.55

$177.87

$259.66

$374.38

Economic Innovation Group Inclusive Wealth Building Initiative

That point is even more dramatic, of course, if the rates of return are higher. Under the 7 percent

return scenario, the household would have $2,710 of ﬁnancial assets after 10 years with no match,

$17,400 with 3 percent of earnings matched, $22,300 with 4 percent matched, and $27,190 with 5 percent

matched. After 40 years, the household would have $29,660 with no match and no additional savings,

$232,890 with 3 percent of earnings matched, $303,640 with 4 percent matched, and $374,380 with 5

percent matched.

In addition, the beneﬁts of wealth creation go beyond ﬁnancial, as individuals with access to wealth

have a higher sense of well-being and are able to avoid emergencies such as cut-oﬀ utilities (D’Ambrosio,

Jäntti, and Lepinteur 2020). Almost half of American survey respondents said they would not have

enough emergency cash available to respond to a $400 emergency (Board of Governors of the Federal

Reserve System 2019a). Access to a retirement security plan like TSP could change that substantially if

it were designed in a manner that included emergency savings features or minimized early withdrawal

penalties for the exigencies of life. A smart design that balances long-term wealth-building and access to

emergency savings promises to dramatically alter the wealth distribution in the United States.

Table 3 Financial asset projections with leakage: median family in the bottom

25 percent of the wealth distribution (thousands)

Year

3 percent nominal return 7 percent nominal return

match

$1.38

$1.60

$1.85

$2.15

$2.49

$2.89

$3.35

$3.88

$4.50

match

$1.38

$11.01

$22.17

$35.10

$50.10

$67.49

$87.64

$111.01

$138.10

match

$1.38

$10.09

$22.30

$39.43

$63.45

$97.14

$144.40

$210.68

$303.64

match

$1.38

$7.24

$14.04

$21.92

$31.06

$41.65

$53.93

$68.16

$84.66

match

$1.38

$1.94

$2.71

$3.81

$5.34

$7.49

$10.50

$14.73

$20.66

match

$1.38

$12.12

$27.19

$48.33

$77.98

$119.55

$177.87

$259.66

$374.38

match

$1.38

$9.13

$18.10

$28.51

$40.58

$54.57

$70.78

$89.58

$111.38

match

$1.38

$8.05

$17.40

$30.52

$48.92

$74.73

$110.92

$161.69

$232.89

SECTION V

Economic Innovation Group Inclusive Wealth Building Initiative

SECTION V

A ﬁnal consideration would, of course, be the possible budgetary costs of such a policy. In work in

progress, the Economic Innovation Group is developing interactive tools to allow policymakers to modify

the parameters of the TSP proposal and estimate the budget costs. It is beyond the scope of our paper to

describe these estimates in detail, but assuming that the take up rates match existing CBO estimates, and

to foreshadow the results, a match capped at 3 percent for individuals making up to median household

earnings of around $52,000 would cost roughly $60 billion in tax year 2021, an amount that would rise

to $100 billion should the policy match rise to 5 percent. These cost estimates will, of course, change

considerably as the income thresholds and match levels vary.

Would creating a widely-available TSP-style program add to an exploding government deﬁcit? Well,

yes, and no. Many of the plan’s would-be beneﬁciaries pay little in federal income taxes today, so the

cost in foregone taxes is smaller than one might think. The government match does cost more, but the

government currently borrows at close to a zero percent interest rate, money which would be given to

workers to invest in private capital that has a higher return. In the short run, the government spending

increases the deﬁcit, but so long as the private return on capital is higher than the Treasury Department’s

borrowing cost, the policy is a positive for American well-being as a whole, and promises signiﬁcant

positive revenue feedback in the future. When that hypothetical 25-year-old starts to take her $600,000

out of her TSP account when she retires, she will pay tax on the withdrawal, as well.

Economic Innovation Group Inclusive Wealth Building Initiative

The unfortunate fact that most Americans lack adequate wealth and retirement savings has been

a consistent reality in the U.S. economy for decades. One major driver of this phenomenon is that the

United States does not have eﬀective retirement savings programs for low-income individuals. Another

part of the dynamic is that even for those workers who do have access to a plan, those savings programs

are voluntary, ad hoc, and usually initiated by employers. As such, reforms to provide more retirement

plan access to less well-oﬀ families would not only increase the retirement security of millions of low-

income Americans, but would also boost their overall wealth and allow them to pass it along to future

generations, thereby helping close the wealth gap in the process.

To achieve the goal of more equitable wealth distribution and retirement security, we view a plan

modeled oﬀ of the Thrift Savings Program (TSP) with a government match and the potential for a private

employer match as a hopeful way to expand retirement coverage to low- and middle-income workers. Our

estimates suggest that for the average household in the poorest 25 percent of the wealth distribution, 40

years of participation in the TSP could provide them with retirement account balances anywhere between

$138,000 and $610,000 before fees and taxes, depending on the level of the government match and the

rate of return on investment. These estimates, while wide ranging, would mean a signiﬁcant infusion of

wealth to millions of Americans. Policymakers should choose liberally from menus of options to ensure

high participation rates and maximize savings. A widely available program modeled after the TSP has the

potential to signiﬁcantly increase ﬁnancial well-being for low-income workers.

SECTION VI

Conclusion

Reforms to provide more retirement plan access to less well-off

families would not only increase the retirement security of millions

f low-income Americans, but would also boost their overall wealth

and allow them to pass it along to future generations, thereby

helping close the wealth gap in the process.

Economic Innovation Group Inclusive Wealth Building Initiative

Andrews, Emily. 1992. “The Growth and Distribution of 401(k) Plans.” In Trends in Pensions, edited by John Turner and Daniel Beller,

149–76. Washington, DC: U.S. Department of Labor, Pension, and Welfare Beneﬁts Administration.

Argento, Robert, Victoria L. Bryant, and John Selhaus. 2015. “Early withdrawals from Retirement accounts during the Great

Recession.” Contemporary Economic Policy 33 (1): 1–16.

Azurdia, Gilda, Stephen Freedman, Gayle Hamilton, and Caroline Schultz. 2014. Encouraging Low- and Moderate- Income Tax Filers

to Save: Implementation and Interim Impact Findings from the SaveUSA Evaluation. New York, NY: MDRC.

Board of Governors of the Federal Reserve System. 2019a. Report on the Economic Well-being of U.S. Households in 2018. Washington,

DC: Federal Reserve Board.

Board of Governors of the Federal Reserve System. 2019b. “Survey of Consumer Finances.” Database, Federal Reserve Board,

Washington, DC.

Brown, Jennifer Erin, Nari Rhee, Joelle Saad-Lessler, and Diane Oakley. 2016. “Shortchanged in Retirement: Continuing Challenges to

Women’s Financial Future.” Washington, DC: National Institute on Retirement Security.

Choi, James, David Laibson, and Brigitte C. Madrian. 2009. “Reducing the Complexity Costs of 401(k) Participation through Quick

Enrollment.” In Developments in the Economics of Aging, edited by David A. Wise, 57–82. Chicago, IL: University of Chicago Press.

Clark, Robert, and Sylvester Schieber. 1998. “Factors Aﬀecting Participation Rates and Contribution Levels in 401(k) Plans.” In Living

with Deﬁned Contribution Plans, edited by Olivia Mitchell and Sylvester Schieber, 69–97. Philadelphia, PA: University of

Pennsylvania Press.

D’Ambrosio, Conchita, Markus Jäntti, and Anthony Lepinteur. 2020. “Money and Happiness: Income, Wealth and Subjective Well-

Being.” Social Indicators Research 148: 47–66.

Duﬂo, Esther, William Gale, Liebman, J., Peter Orszag, and Emmanuel Saez. 2006. “Savings Incentives for Low- and Middle-Income

Families: Evidence from a Field Experiment with H&R Block.” Quarterly Journal of Economics 121 (4): 1311–46.

Engen, Eric, William Gale, and John Karl Scholz. 1996. “The Eﬀects of Tax-Based Saving Incentives on Saving and Wealth.” NBER

Working Paper 5759, National Bureau of Economic Research, Cambridge, MA.

Falk, Justin and Nadia Karamcheva. 2019. “The Eﬀect of the Employer Match and Defaults on Federal Workers’ Savings Behavior in the

Thrift Savings Plan.” Working Paper 2019-06, Congressional Budget Oﬃce, Washington, DC.

Feldstein, Martin, and Charles Horioka. 1980. “Domestic Saving and International Capital Flows.” The Economic Journal 90: 314–29.

Gale, William, Hilary Gelfond, Jason Fichtner, and Benjamin Harris. 2020. “The Wealth of Generations, With Special Attention to the

Millennials.” NBER Working Paper 27123, National Bureau of Economic Research, Cambridge, MA.

Ghilarducci, Teresa, Joelle Saad-Lessler, and Gayle Reznik. 2018. “Earnings Volatility and 401(k) Contributions.” Journal of Pension

Economics and Finance 17 (7): 554–75.

Ghilarducci, Teresa, Michael Papadopoulos, and Anthony Webb. 2017. “Inadequate Retirement Savings for Workers Nearing

Retirement”. Policy Note, Schwartz Center for Economic Policy Analysis and Department of Economics, The New School for Social

Research, New York, NY.

Ghilarducci, Teresa, Michael Papadopoulos, and Anthony Webb. 2018. “The Impact of Guaranteed Retirement Accounts on the

Retirement Crisis.” Policy Note, Schwartz Center for Economic Policy Analysis and Department of Economics, The New School for

Social Research, New York, NY.

Laibson, David. 1997. “Golden Eggs and Hyperbolic Discounting.” Quarterly Journal of Economics 112 (2): 443–78.

Madrian, Brigitte, and Dennis Shea. 2001. “The Power of Suggestion: Inertia in 401(k) Participation and Savings Behavior.” Quarterly

Journal of Economics 116 (4): 1149–87.

Madrian, Brigitte. 2015. “Matching Contributions And Savings Outcomes: A Behavioral Economics Perspective.” NBER Working Paper

18220, National Bureau of Economic Research, Cambridge, MA.

REFERENCES

Economic Innovation Group Inclusive Wealth Building Initiative

McKernan, Signe-Mary, Gregory Mills, Caroline Ratcliﬀe, William J. Congdon, Michael Pergamit, Breno Braga, and Kassandra

Martinchek. 2020. “Building Savings, Ownership, and Financial Well-Being First- and Third-Year Assets for Independence

Program Randomized Evaluation Findings in Context.” OPRE Report #2020-30, The Urban Institute, Washington, DC.

Miller, Mark. 2020, June 26. “A Pandemic Problem for Older Workers: Will They Have to Retire Sooner?” New York Times.

Mitchell, Olivia, Stephen Utkus, and Tongxuan Stella Yang. 2007. “Turning Workers into Savers? Incentives, Liquidity, and Choice in

401(k) Plan Design.” National Tax Journal 60 (3): 469–89.

Morrissey, Monique. 2019. “The State of American Retirement Savings.” Report, Economic Policy Institute, Washington, DC.

Organisation for Economic Co-operation and Development (OECD). n.d. “Pensions at a Glance: Income and Poverty of Older People.”

Database, OECD.Stat, Organisation for Economic Co-operation and Development, Paris, France.

Papke, Leslie, and James Poterba. 1995. “Survey Evidence on Employer Match Rates and Employee Saving Behavior in 401(k) Plans.”

Economics Letters 49 (3): 313–17.

Poterba, James, Steven Venti, and David Wise. 1996. “Personal Retirement Saving Programs and Asset Accumulation: Reconciling the

Evidence.” NBER Working Paper 5599, National Bureau of Economic Research, Cambridge, MA.

Radpour, Siavash, Michael Papadopoulos, and Teresa Ghilarducci. 2021. “Trends in Employer Sponsored Retirement Plan Access and

Participation Rates: Reconciling Diﬀerent Data Sources.” Research Note Series 2021-01, Schwartz Center for Economic Policy

Analysis and Department of Economics, The New School for Social Research, New York, NY.

The Vanguard Group. n.d. “Vanguard 500 Index Fund Admiral Shares: Expenses.” The Vanguard Group, Valley Forge, PA.

Thrift Savings Plan (TSP). n.d. “How the TSP Fits into Your Retirement.” Thrift Savings Plan, Federal Retirement Thrift Investment

Board, Washington, DC.

Thrift Savings Plan. 2020. “Annual Report of the Thrift Savings Plan, Required by §105 of the TSP Enhancement Act of 2009.” Report,

Thrift Savings Plan, Federal Retirement Thrift Investment Board, Washington, DC.

Turner, John, Bruce Klein, and Norman Stein. 20916. “Financial Illiteracy Meets Conﬂicted Advice: The Case of Thrift Savings Plan

Rollovers.” Journal of Retirement 3 (4): 47–65.

World Inequality Database. n.d. “Wealth inequality, USA, 1962-2019.” World Inequality Database.

REFERENCES

Our Mission—

EIG brings together leading entrepreneurs,

investors, economists, and policymakers from

across the political spectrum to address America’s

economic challenges.

The Inclusive Wealth Building Initiative

was made possible through the

generous support of philanthropist

Alexander von Fürstenberg.

PUBLISHED March 25, 2021

INCLUSIVE WEALTH BUILDING INITIATIVE