i
The Hamilton Project • Brookings
Lauren Bauer, Diane Whitmore Schanzenbach, and Jay Shambaugh
ECONOMIC ANALYSIS | OCTOBER 2018
Work Requirements and Safety Net Programs
ii
The Hamilton Project • Brookings
The Hamilton Project seeks to advance America’s promise
of opportunity, prosperity, and growth.
We believe that today’s increasingly competitive global economy
demands public policy ideas commensurate with the challenges
of the 21
st
Century. The Project’s economic strategy reflects a
judgment that long-term prosperity is best achieved by fostering
economic growth and broad participation in that growth, by
enhancing individual economic security, and by embracing a role
for effective government in making needed public investments.
Our strategy calls for combining public investment, a secure social
safety net, and fiscal discipline. In that framework, the Project
puts forward innovative proposals from leading economic thinkers
— based on credible evidence and experience, not ideology or
doctrine — to introduce new and effective policy options into the
national debate.
The Project is named after Alexander Hamilton, the nation’s
first Treasury Secretary, who laid the foundation for the modern
American economy. Hamilton stood for sound fiscal policy,
believed that broad-based opportunity for advancement would
drive American economic growth, and recognized that “prudent
aids and encouragements on the part of government” are
necessary to enhance and guide market forces. The guiding
principles of the Project remain consistent with these views.
MISSION STATEMENT
We thank reviewers John Coglianese, Jason Furman, Heather
Hahn, Kriston McIntosh, Ryan Nunn, and the Center on Budget
and Policy Priorities. We also thank the following who provided
research assistance: Patrick Liu, Jimmy O’Donnell, Jana Parsons,
Becca Portman, and Areeb Siddiqui.
ACKNOWLEDGMENTS
1
The Hamilton Project • Brookings
Work Requirements and Safety Net Programs
Lauren Bauer
The Hamilton Project and the Brookings Institution
Diane Whitmore Schanzenbach
The Hamilton Project, the Brookings Institution, and Northwestern University
Jay Shambaugh
The Hamilton Project, the Brookings Institution, and The George Washington University
OCTOBER 2018
2
The Hamilton Project • Brookings
receive disability income, people with young dependents, or
students; but, accurately exempting all those who are eligible
can be challenging and is likely to result in terminating
coverage for many people with health conditions or caregiving
responsibilities that fall outside of states’ narrow denitions.
Proponents of work requirements would ideally only like
to sanction individuals who are able to work, but choose
not to. But in practice strict enforcement of proposed work
requirements will sanction many groups, including: those
who are unable to work, those who are able to work but who
do not nd work, those who are working but not consistently
above an hourly threshold, and those who are meeting
work or exemption requirements but fail to provide proper
documentation. Evidence suggests that the vast majority of
those exposed to proposed work requirements for SNAP and
Medicaid fall into these groups.
In this paper, we analyze those who would be impacted by an
expansion of work requirements in SNAP and an introduction
of work requirements into Medicaid. Our principal
contribution is to characterize the types of individuals
who would face work requirements, describe what their
work experiences are over a two-year period, and identify
the reasons why they are not working if they experience a
period of unemployment or labor force nonparticipation. We
nd that most of those who fail the new work requirements
are either those who are in the labor force already but who
experience unstable employment, or those who might be
eligible for hardship exemptions, such as those with health
problems who are not already receiving disability income.
e compositional and labor market analyses reported below
suggest that the proposed work requirements will put at risk
access to food assistance and health care for millions who are
working, trying to work, or face barriers to working.
Introduction
Basic assistance programs such as the Supplemental Nutrition
Assistance Program (SNAP, formerly the Food Stamp
Program) and Medicaid ensure families have access to food
and medical care when they are low-income. ese programs
li millions out of poverty while reducing food insecurity and
increasing access to medical care. ey also support work,
and increase health and economic security among families in
the short term as well as economic self-suciency in the long
term.
Today, some policymakers at the federal and state levels intend
to add new work requirements in order for beneciaries to
receive SNAP benets and participate in the Medicaid health
insurance program. In general, those exposed to a work
requirement would be required to prove that they are working
or participating in a training program for at least 20 hours per
week each month. Failure to prove that they have met the work
requirement or are eligible for an exemption would mean that
a program participant would lose food assistance benets or
health insurance for a time, or until they met the standard.
Work requirements are meant to force work-ready individuals
to increase their work eort and maintain that work eort
every month by threatening to withhold and subsequently
withholding food assistance or health coverage if a person
is not working a set number of hours. e strategy presumes
that the reasons that many low-income individuals are not
working or meeting an hourly threshold every month is
either due to their own lack of eort or to work disincentives
theoretically inherent to means-tested programs. It is clear
that some people face barriers to working outside the home
and as such, many work requirements exempt people that
Abstract
Basic assistance programs such as the Supplemental Nutrition Assistance Program (SNAP, formerly the Food Stamp Program) and
Medicaid ensure families have access to food and medical care when they are low-income. Some policymakers at the federal and
state levels intend to add new work requirements to SNAP and Medicaid. In this paper, we analyze those who would be impacted
by an expansion of work requirements in SNAP and an introduction of work requirements into Medicaid. We characterize the
types of individuals who would face work requirements, describe their labor force experience over 24 consecutive months, and
identify the reasons why they are not working if they experience a period of unemployment or labor force nonparticipation.
We nd that the majority of SNAP and Medicaid participants who would be exposed to work requirements are attached to the
labor force, but that a substantial share would fail to consistently meet a 20 hours per week–threshold. Among persistent labor
force nonparticipants, health issues are the predominant reason given for not working. ere may be some subset of SNAP
and Medicaid participants who could work, are not working, and might work if they were threatened with the loss of benets.
is paper adds evidence to a growing body of research that shows that this group is very small relative to those who would be
sanctioned under the proposed policies who are already working or are legitimately unable to work.
3
The Hamilton Project • Brookings
limits or work requirements for a set of individuals as a
condition for program eligibility.
Such work requirements can undermine the insurance value
of the programs, though, if people who are not working either
cannot work due to individual limitations or are unable to
nd steady work due to economic uctuations. Evaluating
whether work requirements are an appropriate policy lever—
as opposed to addressing work disincentives through other
means—thus depends on the goals of the program overall,
the characteristics of the target population, the design of the
work requirements, the cost of administering the program,
the likelihood of erroneously limiting access, and the strength
of the incentive eects.
Work requirement policies oen have diculty distinguishing
between those who are able to work and those who are unable
to work, because both groups can be hard to observe and
verify. As a result, strict enforcement of work requirements
will sanction those who are unable to work, as well as those
who could work but do not obtain employment in response to
the requirements. ey may also sanction some who are able
to work but who are not able to nd work, as well as those who
are working but fail to provide proper documentation.
In order to evaluate whether a work requirement is in
keeping with the purpose of a means-tested program, there
are a number of dimensions by which a proposal should be
evaluated. One would want to exempt those whom society
does not feel should be forced to work, accommodate changes
in the business cycle that make work more dicult to nd,
and have a system of verication and exemption that does not
raise barriers to entry or remove program participants who
should maintain access. But, one would have to ensure that
work requirements do not punish those who cannot obtain a
job due to economic conditions in their area, penalize those
who are actually working but have temporarily lost hours,
limit access to programs for an extended period of time aer
failing a work requirement, or, compromise the insurance
goals of the program in question. ese parameters can be
quite dicult to meet and they set the criterion by which
policymakers can determine whether work requirements are
inappropriate for the program in question.
ere is an extensive literature on whether work requirements
can in fact push people into the labor force, principally
studying the impacts of the 1996 Temporary Assistance for
Needy Families (TANF) reform (see Blank 2002 and Ziliak
2016 for reviews). e labor supply of the TANF population
did in fact rise, but this took place amidst a strong economy
and support from the Earned Income Tax Credt (EITC)
expansion as well (Schanzenbach 2018). For example, Fang
and Keane (2004) nd that while work requirements were the
most important factor driving the decline in participation in
welfare programs, the EITC expansion and macroeconomic
Adding explicit work requirements to assistance programs
must be analyzed in the context of program goals and from
many angles. Who would be impacted by an expansion of
work requirements? What are the administrative costs and
challenges of managing the work requirements? How do the
requirements interact with the realities of the low-wage work
experience? And how would the requirements impact the
health and economic benets to program participation? For
example, removing Medicaid coverage may have little positive
work-incentive eect for the currently healthy but may
undermine public health goals and reduce the labor supply of
those who do encounter health problems and have lost their
coverage. Removing SNAP benets from working-age adults
may impact resources available not just to them, but also to
any seniors and dependents in the household. Finally, tight
work requirements can undermine the automatic stabilizer
aspect of these programs. Instead of SNAP expanding as the
unemployment rate rises, the work requirements would cause
the program to contract, resulting in more people losing
benets when work becomes dicult for them to nd.
ere may be some subset of individuals who could work, are
not working, and might work if they were threatened with the
loss of benets. is paper adds evidence to a growing body
of research that shows that this group is very small relative to
those who would be sanctioned under the proposed policies
who are already working or are legitimately unable to work
(Bauer and Schanzenbach 2018a, 2018b; Gareld et al. 2018;
Goldman et al. 2018).
e goals of safety net programs are to provide insurance
protection to those who are experiencing poor economic
outcomes and to support those who are trying to improve
their situation. Our analysis suggests that work requirements
will harm more individuals and families than they would help
the small share who might increase their labor supply.
SNAP, Medicaid, and Incentives to
Work
e social safety net is intended to provide insurance against
bad outcomes. But, for means-tested benet programs,
economic theory suggests it may reduce the incentive to
work because (1) individuals are only eligible for a program
when their income remains below a given threshold and
(2) participants stand to lose benets as income increases
or reaches the eligibility threshold. In addition, any time
someone receives unearned income of sucient size, it may
theoretically reduce the amount of work that an individual
wants to supply to the market. In some cases, worries about
work disincentives have led to the implementation of time
4
The Hamilton Project • Brookings
factors were more important in driving the increase in work
participation (they nd work requirements had a positive
impact as well, but the contribution was smaller). Work
requirements oen come with a variety of supports and involve
dierent enforcement mechanisms and levels of stringency.
See Hamilton et al. (2001) for a detailed review as part of the
National Evaluation of Welfare-to-Work Strategies. Many of
the work requirement programs that have generated positive
results also had substantial education and skills training
components (Pavetti and Schott 2016). Other studies, such
as Meyer and Rosenbaum (2001) and Grogger (2004) suggest
a smaller or negligible role for the TANF reforms compared
with other factors, especially the EITC expansion.
In this analysis, we focus more on the people who would be
impacted by new work requirements and the reasons why
they are not working, as opposed to the question of the
labor supply response. Given the extent to which the labor
market conditions—in particular for potentially impacted
populations—are dierent than those in the 1990s (Black,
Schanzenbach, Breitwieser 2017; Butcher and Schanzenbach
2018), it is helpful to consider specically what types of
individuals would be aected by proposed work requirements
and why they are not currently working to better understand
the possible impacts of expanded work requirements. In
this section we describe the SNAP and Medicaid programs,
the structure of their work incentives, and evidence of the
programs’ incentive eects on labor supply.
SNAP
Since the 1960s SNAP has provided resources to purchase
food for millions of low-income households. e goal of the
program is to provide beneciaries with resources to raise
their food purchasing power and, as a result, improve their
health and nutrition. Households are eligible for SNAP if
they meet an asset and income threshold, or if they receive
assistance from programs like Supplemental Security Income.
SNAP benet levels are targeted based on a given household’s
income and expenses.
SNAP currently addresses work disincentives in a variety of
ways. Similar to the EITC, SNAP addresses work disincentives
through an earnings disregard of 20 percent and a gradual
benet reduction schedule. is means that the size of the
earnings disregard increases as income increases and that
those with earned income receive larger SNAP benets than
those with no earned income (Wolkomir and Cai 2018).
When a person moves from being a labor force nonparticipant
to working while on SNAP, total household resources will
increase; as a beneciary’s earnings approach the eligibility
threshold, total household resources continue to increase.
e combination of the earnings disregard and a gradual
phase-out schedule—that states have the option to further
extend and smooth—ameliorate but do not eliminate work
disincentives.
States have had the option to impose work requirements on
certain beneciaries since the 1980s. Most SNAP participants
between the ages of 18 and 59 without dependents under 6
are required to register for work, accept a job if one is oered
to them, and not reduce their work eort. States are required
to operate an employment and training program, and
may require some SNAP recipients to participate or suer
sanctions. See Rosenbaum (2013) and Bolen et al. (2018) for
a detailed description of SNAP work requirements. Aer
1996, SNAP work requirements and benet time limits were
imposed on individuals aged 18–49 without dependents under
the age of 18, requiring them to register for work and accept
a job if one is oered to them. If they work or participate in
a training program for at least 20 hours per week, they can
maintain access to the program. is population is allowed
to receive 3 months of benets out of 36 months if they do
not work or participate in a training program. States are
permitted to exempt a share of individuals and apply to
the U.S. Department of Agriculture (USDA) for a waiver to
the time limit provisions, an essential capacity for SNAP’s
function as an automatic stabilizer. Studies show that when
SNAP payments increase to a local area in response to an
economic downturn, they serve as an eective scal stimulus
to the local area (Blinder and Zandi 2015; Keith-Jennings
and Rosenbaum 2015). Among other changes, the proposed
work requirements would make these regional waivers more
dicult to obtain.
SNAP improves health and economic outcomes in both the
near and long terms (see Hoynes and Schanzenbach 2016
for a review), but had a negative eect on employment in
the past. During the Food Stamp Program’s introduction in
the 1960s and 1970s, reductions in employment and hours
worked were observed, particularly among female-headed
households (Hoynes and Schanzenbach 2012). Whether work
requirements could oset this disincentive would depend on
their targeting and whether those who are not working could
readily increase their labor supply.
MEDICAID
Since 1965, the Medicaid program has been administered in
partnership between federal and state governments to provide
medical assistance to eligible individuals. e core goal of the
program is to provide health services and to cover health-care
costs in order to improve health. Under the Patient Protection
and Aordable Care Act (ACA), the eligible population
expanded to include low-income adults under the age of 65
who previously did not qualify.
Although some SNAP beneciaries have been subject to work
requirements since the 1980s, Medicaid work requirements
are being rolled out for the rst time in certain states. e ACA
does not allow work requirements to be imposed as a condition
for program participation in Medicaid, but states may apply
5
The Hamilton Project • Brookings
for a waiver under Section 1115 of the Social Security Act to
introduce work requirements if the Department of Health
and Human Services determines doing so advances program
objectives. ough the Obama administration and the U.S.
District Court for the District of Columbia (which rejected
Kentucky’s proposal for work requirements in Medicaid)
did not view work requirements as supporting core program
goals, the Trump administration has expressed its conviction
that work requirements are allowable (Centers for Medicare
& Medicaid Services 2018; Gareld, Rudowitz, and Damico
2018; Stewart v. Azar).
In the case of Medicaid, there are societal costs to taking
health insurance away from an otherwise eligible person
due to work requirements. For example, since there are
rules requiring hospitals to provide medical care to those
experiencing life-threatening emergencies regardless of the
individual’s ability to pay, those without insurance will in
many cases seek and receive treatment in ways that are more
BOX 1.
Trends in Prime-Age Labor Force Participation
For a number of decades labor force participation in the United States rose. is was especially true for prime-age (25–54) workers,
whose participation rose from 65 percent in the middle of the 20
th
century to a peak of 84 percent in 1999. is persistent trend
obscured an osetting force: Prime-age men were steadily working less while prime-age women were working more. In 1949 97
percent of prime-age men were in the labor force, but only 36 percent of women were. By 1999 those gures were 92 percent for men
and 77 percent for women.
Although women’s labor force participation rose in the 1980s and early 1990s, policymakers were concerned about the low labor
force participation for single women with children, which remained relatively at over that period. But for the past 20 years single
women who head households with children have participated in the labor market at nearly the same rate as single women without
children or married women without children. In fact, for the rst time, in 2017 the labor force participation rate of single women
with children was higher (79.09 percent) than single women without dependents (79.06 percent.) Married women with children are
still more likely to be out of the labor force (box gure 1). More recently, overall labor force participation has declined, in part due
to the aging population. Older working-age Americans (55–64) are less likely to work, with a labor force participation rate in 2017
around 72 percent for those aged 55–59 and 57 percent for those aged 60–64, compared to the current 82 percent for those aged
25–54.
ese trends provide context for who is not currently working that society might prefer to work. Most prime-age men work, though
nearly 10 percent do not. Most unmarried prime-age women with children also work. A much smaller share of older Americans
work.
BOX FIGURE 1.
Prime-Age Women’s Labor Force Participation, by Marital Status and Presence of
Children under Age 18
Married with children
Married, no children
Single with children
Single, no children
40
50
60
70
80
90
100
1977 1982 1987 1992 1997 2002 2007 2012 2017
Labor force participation rate
Source: Current Population Survey Annual Social and Economic Supplement (ASEC) (Bureau of Labor Statistics [BLS] 1977–2017); authors’ calculations.
6
The Hamilton Project • Brookings
expensive for society (Institute of Medicine 2003). Second,
care delivered via insurance may include preventive care,
check-ups, and other care that is more ecient than delaying
care until a medical problem becomes severe enough to be
treated in an emergency room. us, denying insurance may
not reduce costs for society. Finally, evidence suggests that
health insurance is valued by participants at less than its cost,
making proposed work requirements less eective at raising
employment (Finkelstein, Hendren, and Luttmer 2015).
Evidence of the eect of Medicaid participation on
employment for childless adults is decidedly mixed, with
population dierences and prevailing economic conditions
as potential explanations for why studies have shown positive,
negative, and no eects on employment (Buchmueller, Ham,
and Shore-Sheppard 2016). Nevertheless, in the years since
Medicaid expansion through the ACA, the preponderance
of evidence suggests that Medicaid receipt has had little or
positive eects on labor supply (Baicker et al. 2014; Duggan,
Goda, and Jackson 2017; Garthwaite, Gross, and Notowidigdo
2014; Gooptu et al. 2016; Kaestner et al. 2017), with notable
exceptions (e.g., Dague, DeLeire, and Leininger 2017).
While there is no research evidence regarding the eect
of work requirements in Medicaid, last month, as the rst
state to implement a plan, Arkansas disenrolled program
participants for failing to comply with work requirements.
Arkansas terminated coverage for 4,353 citizens for failing
to qualify for an exemption or to meet work requirements,
while an additional 1,218 reported 20 hours per week of work
activities and 2,247 reported an exemption in the month of
August (Rudowitz and Musumeci 2018).
For these programs to accomplish their goals, eligible people
should not be dissuaded from applying for or improperly
prevented from receiving those benets. Evidence suggests
that, under a variety of scenarios, the vast majority of those
losing access to Medicaid would not lose access because they
failed to meet a work requirement, but because they failed to
successfully report their work/training activity or exemption
(Gareld, Rudowitz, and Musumeci 2018; Goldman et al.
2018). For example, in Arkansas, the only state currently
implementing a work requirement in Medicaid, beneciaries
are required to report through an online portal, Access
Arkansas (Arkansas Department of Human Services n.d.),
despite a large number of program-eligible Arkansans who
lack internet access (Gangopadhyaya et al. 2018).
Characteristics of Those Who
Would Face New Work
Requirements
Potential loss of access to SNAP and Medicaid on the basis
of a work requirement is a function of whether the person is
qualied for and veried as exempt from working and, if not,
whether the person works sucient hours each month to meet
the requirement. ose who have a categorical exemption
from work requirements—students, for example—are not
required to work unless their status changes. Exemptions
from work requirements can be applied individually for a
variety of reasons, including temporary health problems, or,
more broadly, when the unemployment rate for a location
BOX 2.
Proposed Expansion of Work Requirements
In April 2018 President Trump issued an executive order requiring that all means-tested programs be reviewed for the presence
of current work requirements, the current state of enforcement and exemption, and, for those programs without current work
requirements, whether such requirements could be added (White House 2018).
is executive order builds on executive action to implement work requirements in Medicaid for the rst time. In letters to governors
(Price and Verma 2017) and state Medicaid directors (Neale 2018), the U.S. Department of Health and Human Services (HHS)
has oered guidance for states considering submitting a waiver request to apply work requirements for those receiving Medicaid.
Since the Centers for Medicare & Medicaid Services oered guidance to the states with regard to Medicaid in 2017, 14 states have
submitted work requirement proposals to HHS. HHS has approved four states’ plans, though Kentucky’s plan was vacated. e
state of Arkansas has begun to enforce work requirements (Urban Institute 2018). State proposals vary in terms of the age range and
household composition of exposure, who is exempt, and the hours required for work or approved activities.
Additionally, in reauthorizing the Farm Bill, in June 2018 the House voted to expand the scope of who is required to work in order
to receive SNAP benets to include adults 18–59 with dependent children aged 6–18 as well as those aged 50–59 without dependents
under the age of 6. As of publication, the conference committee is considering this proposal.
7
The Hamilton Project • Brookings
FIGURE 1.
Exposure to Work Requirements among Adult SNAP Participants, 2017
Source: ASEC (BLS 2018); authors’ calculations.
is high. Certain educational or training activities can also
qualify for meeting hourly thresholds.
To highlight one diculty in designing a work requirement
policy, consider the group of SNAP and Medicaid participants
who usually are not working. Many individuals in this group
are not expected to work, including the elderly, disabled,
children, students, caregivers, and the inrm. In fact, nearly
two thirds of individuals who participate in SNAP are elderly,
disabled, or children (USDA 2017a).
Some of these characteristics are straightforward to observe
and verify, such as age, school enrollment, and receipt of
disability benets. Other characteristics are dicult to
observe and costly to verify, such as those with temporary
medical conditions that make it impossible for them to work,
those who have a chronic health condition but do not meet the
high standard set for disability benets (or have not applied
for disability benets), and those who do not have the skills,
childcare, or transportation to obtain a job in their local
economy at present. Another share of this group might be
capable of employment but not willing to work; in that case
the work requirements might or might not provide enough
incentive for them to get jobs.
Using data from the Current Population Survey Annual Social
and Economic Supplement (ASEC), we quantify exposure
to work requirements in 2017 based on broad demographic
characteristics. To do so, we separate those who would likely
qualify for a categorical exemption from those who would
be required to work or who would qualify for a waiver to
maintain eligibility. To be clear, while we model who is
eligible for a categorical exemption, evidence suggests that not
everyone in these groups will successfully navigate the system
and obtain the exemption; in fact, estimates suggest that most
people who lose coverage under this policy will be eligible for
an exemption or already be working. For SNAP we followed
the federal guidelines for categorical exemption; for Medicaid
we created a composite from among the dierent plans put
forth by the states based on how frequently such groups are
exempt.
For SNAP, minors, those who are older than 59 years, students,
those receiving disability benets, and those with a child
under the age of 6 are exempt from both current and new,
proposed work requirements. e samples are further limited
to U.S. citizens and nonactive military. For simplication,
we describe those aged 18-49 without dependents as being
currently exposed to work requirements and those aged 18–
59 with a dependent between the ages of 6 and 17 (inclusive)
as well as those between the ages of 50 and 59 with no
dependents under the age of 6 as newly required to meet
work requirements or to participate in a training program in
order to receive SNAP benets. For the current group, some
Age 18−49,
no dependents
11.5%
More than 59 years old
24.0%
Dependent
less than 6 years old
23.6%
Disability income
13.6%
Students
5.7%
Age 18−49,
dependent age 6–17
13.1%
Age 50−59,
no dependents
less than 6 years old
8.5%
Current
Exempt New
8
The Hamilton Project • Brookings
may live in places exempt from work requirements or have an
unobserved good-cause exemption.
How many adult SNAP participants are—or would be—
exposed to work requirements? Figure 1 shows the entire adult
population (18 or older) who reported SNAP participation in
2017. Each rectangle represents a share of the total population
and whether the individuals in that share were eligible for a
categorical exemption to work requirements (teal), were in a
population currently exposed to a work requirement (green),
or would be newly exposed to work requirements under
the House proposal (purple). e shaded rectangles sum to
100percent, the total adult SNAP participant population.
Under the House bill parameters (described in box 2),
combined with current work requirements, one third of all
adults who reported receiving SNAP benets during 2017
would be exposed to work requirements, though a portion of
those impacted could apply for exemptions based on veried
health- or work-related concerns. Some already face work
requirements, but 22 percent of all participants would be
newly exposed to work requirements under the House bill
(purple).
Figure 1 also shows the reasons some participants would be
exempt from new requirements. e majority (67 percent)
of adults currently receiving SNAP benets would still be
exempt from work requirements based on age, having a
dependent under the age of 6, or having student or disability
status. Some would be exempt for multiple reasons; we group
them rst by age, then by the presence of dependents, and
then by student or disability status. For example, while gure
1 shows just 14percent exempt due to disability, 24percent of
all adult SNAP recipients report receipt of disability benets.
In 2017, 2.2million people who reported SNAP benet receipt
were exposed to work requirements during the year based on
their demographic characteristics. Under the House proposal
and based on 2017 numbers, this would more than double
with 2.5million adults aged 18–49 with dependent children
aged 6–17 and 1.6million adults aged 50–59 who would be
exposed to work requirements nationally for the rst time.
In any household, there may be others who rely on the benets,
and not just the individual facing work requirements. e
solution to concerns for other individuals in the household
has typically been to waive work requirements for those
who likely cannot work or who reside with those for whom
shielding from benet loss is a priority. Any reduction in
SNAP benets to adults would reduce the total amount of
resources available to them to purchase food, including food
for children. ere are 3.5 million children and 710,000
FIGURE 2.
Exposure to Work Requirements among Adult Medicaid Participants, 2017
Source: ASEC (BLS 2018); authors’ calculations.
More than 64 years old
or Medicare
11.9%
Dependent
less than 6 years old
22.1%
Disability
income
12.9%
Students
6.3%
Age 18−49,
no dependents less than 6 years old
30.9%
Age 50−64,
no dependents
less than 6 years old
15.8%
Exempt
New
9
The Hamilton Project • Brookings
seniors in these households that would be exposed to possible
benet loss due to work requirements.
We perform the same exercise to show the share of Medicaid
beneciaries who are targeted by the policy based on potential
new rules (gure 2). Minors, seniors (those over the age of
64), students, those receiving disability benets or Medicare,
and those with a child under the age of 6 are those who are
generally eligible to be exempt from work requirements based
on the plans that states submitted, though there is variation
across states. We apply these categories to the entire adult
Medicaid population, acknowledging that not every state has
submitted a work requirement proposal and that the aected
population varies by state plans. A nationwide expansion of
these rules would target 22.4million Americans for a possible
loss of Medicaid coverage.
Almost half of all adult Medicaid beneciaries would be
targeted by work requirements if the composite rules were
applied nationwide. e largest share of those exempt
from work requirements are parents with young children
(22 percent) followed by those reporting disability income
(13 percent) and Medicare/Medicaid dual enrollees
(12 percent). About 6 percent of Medicaid participants are
students.
Volatility in the Low-Wage Labor
Market
e decline in labor force participation—especially among
prime-age males—has drawn extensive attention in academic
and policy circles (e.g., Abraham and Kearney 2018; Council
of Economic Advisers [CEA] 2016; Juhn 1992). Some recent
academic work has emphasized the fact that participation may
be declining in part because an increasing number of labor
force participants cycle in and out of the labor force: a pattern
with direct relevance to proposed work requirements. e
most comprehensive look at the behavior of people cycling
through the labor force is Coglianese (2018). He documents
that, among men, this group—which he refers to as “in-and-
outs”—takes short breaks between jobs, returns to the labor
force fairly quickly (within six months), and, crucially, is no
more likely than a typical worker to take another break out of
the labor force. See also Joint Economic Committee (2018) for
a discussion of the in-and-out behavior of nonworking prime-
age men and reasons for their nonemployment.
SNAP or Medicaid participants who are employed but who
work in jobs with volatile employment and hours would be at
risk of failing work requirements. is group includes those
who lose their job; for example, the House bill sanctions
participants for months they are not working or in training for
at least 20 hours per week, even if they were recently employed
and are searching for a new job. Similarly, those who work in
jobs with volatile hours would be sanctioned in the months
that their average hours fell below 20 hours per week, whether
due to illness, lack of hours oered by the employer, or too few
hours worked by the participant if they fail to receive a good-
cause waiver.
Low-wage workers in seasonal industries such as tourism
would potentially be eligible for SNAP in the months when
they are working, but not in the months without employment
opportunities. In other words, while benets are most
needed when an individual cannot nd adequate work,
under proposed work requirements these are the times that
benets would be unavailable. Disenrollment could make
it more dicult for an individual to return to work—for
example, if a person with chronic health conditions is unable
to access needed care while they are between jobs. Any work
requirement that banned individuals from participation for
a considerable amount of time aer failing the requirements
would be even more problematic for those facing churn in the
labor market.
In a set of analyses, Bauer and Schanzenbach (2018a, 2018b)
found that although many SNAP beneciaries work on
average more than 20 hours a week every month, they
frequently switch between working more than 20 hours and
a dierent employment status over a longer time horizon.
Using the ASEC, those authors found that, over the course
of 16 months between 2016 and 2018, about 20 percent of
individuals aged 18–59 without a dependent child under age
6 switched between working more than 20 hours a week and
working fewer than 20 hours per week, seeking employment,
or being out of the labor force.
In this economic analysis we examine labor force status
transitions and the reasons given for not working among those
targeted for work requirements over 24 consecutive months,
January 2013–December 2014, using the rst two waves of the
Survey of Income and Program Participation (SIPP).
1
By using
a dataset that allows us to track workers over time, we identify
the share of program participants who are consistently out
of the labor force, the share who would consistently meet
a work requirement, and the share who would be at risk of
losing benets based on failing to meet a work requirement
threshold.
We assume that to comply with a program’s work requirement,
beneciaries would have to prove each month that they are
working for at least 20 hours per week averaged over the
month, which is the typical minimum weekly requirement
among the SNAP and Medicaid work requirement proposals.
Looking rst at SNAP and then at Medicaid, we calculate
the share of program participants who would be exposed
to benet loss because they are not working sucient hours
10
The Hamilton Project • Brookings
over the course of 24 consecutive months. Among those who
would be exposed to benet loss and who experienced a gap in
employment, we describe the reasons given for not working to
help quantify potential waiver eligibility.
We remove from the analysis all those who have a categorical
exemption. For SNAP and Medicaid, we exclude those outside
the targeted age range, those with children under 6, full- or
part-time students, and those reporting disability income.
ose receiving Medicare are additionally excluded from the
Medicaid analysis. As an instructive example, the labeled
group “18–49, no dependents” is additionally exclusive of
students and those reporting disability income. Program
participants are those who reported receiving SNAP or
Medicaid at any point between January 1, 2013, and December
31, 2014.
We categorize each individual in each month into one of
four categories: (1) employed and worked more than 20
hours a week on average, (2) employed and worked less than
20 hours a week on average, (3) unemployed and seeking
employment, or (4) not in the labor force. If a worker was
employed at variable weekly hours but maintained hours
above the monthly threshold (80 hours for a four-week month
and 120 hours for a ve-week month), then we categorize
them as “employed and worked more than 20 hours a week
for that month.” Individuals are considered to have a stable
employment status if they do not change categories over two
years, and are considered to have made an employment status
transition if they switched between any of these categories at
least once. ere is no employment status transition when a
worker changes jobs but works more than 20 hours a week at
each job.
EXPOSURE TO PROPOSED WORK REQUIREMENTS IN
SNAP
Among working-age adults, SNAP and Medicaid serve a
mix of the unemployed, low-income workers, and those who
are not in the labor force (USDA 2017b). Figure 3 describes
employment status by those groups who are currently exposed
to work requirements and who would be newly subject to work
requirements under the House proposal.
During the Great Recession, waivers to work requirements
were implemented nationwide. During the time period
covered by the SIPP (2013–14), 8 states stopped implementing
these waivers fully, and 10 states partially (Silberman
2013).
2
For analytic purposes, we look at employment status
transitions among 18 to 49 year-olds without dependents as the
demographic group currently exposed to work requirements,
regardless of whether they lived in state in which waivers
were implemented during 2013 and 2014. ose receiving
SNAP benets who are in the demographic group currently
exposed to work requirements—adults aged 18–49 with no
dependents—generally participate in the labor market, with
just 25percent consistently not in the labor force (discussed
FIGURE 3.
Employment Status over Two Years, SNAP Participants
Source: Survey of Income and Program Participation (SIPP) (U.S. Census Bureau 2013–14); authors’ calculations.
0
25
50
75
100
Age 18−49,
no dependents
Age 18−49,
dependent 6−17
Age 50−59,
no dependent under 6
Percent
Stable
Transitioned
Employed <20 hours
Between 20+ hours
and unemployment
or not in the labor force
Between 20+ hours
and <20 hours
Employed 20+ hours
Unemployed
Not in labor force
Other transition
Currently
exposed
Newly
exposed
11
The Hamilton Project • Brookings
below). While 58percent worked at least 20 hours per week
in at least one month over two years, 25percent were over the
threshold at some point but fell below the 20-hour threshold
during at least one month over two years. Very few are always
working less than 20 hours a week or always unemployed (less
than 2 percent in either case), and 14 percent move across
these categories.
ose aged 18–49 who are not subject to the three-month
time limit because they have a dependent aged 6–17 but
who would face it under the House proposal demonstrate a
similar distribution of employment status as those without a
dependent, but they are more likely to work. ere are fewer
individuals who are always out of the labor force (14percent)
and more that consistently work 20 hours a week or more
(46 percent).
3
ere is also substantial month-to-month
churn (16 percent) between working above 20 hours per
week and less than 20 hours per week and churn (12percent)
between working above 20 hours per week and being either
unemployed or not in the labor force. is highlights the
number who are actively in the workforce and meeting the 20-
hour threshold in at least one month, but who might fail new
work requirements from time to time.
Older SNAP participants (aged 50–59 without dependents
under age 6) who would also be newly exposed to work
requirements and time limits have a distinct employment
status pattern from those aged 18–49. Almost half were
permanently out of the labor force in large part due to their
health. While 23 percent worked consistently above the
threshold of 20 hours a week, nearly as many (18 percent)
worked above the threshold at some point but also below the
threshold at some point, meaning they would fail the work
requirement despite having sometimes met the threshold.
ere is a meaningful portion of SNAP participants in
the labor force and working, but not all are working above
the monthly work requirement threshold consistently.
Coglianese’s (2018) nding that workers who are in and out of
the labor force are not more likely to take another break later
on suggests it is unclear how much more consistently work
requirements would attach these people to the labor force.
We next examine the reasons given for not working over the
two-year period, rst for those aged 18–49 with a dependent
between the ages of 6 and 17, and second for those 50 to 59
without a dependent under age 6 (gures 4a and 4b). e
FIGURE 4A.
Most-Frequent Reason for Not Working for Pay, SNAP Participants Aged 18–49
with Dependents Age 6–17
Source: SIPP (U.S. Census Bureau 2013–14); authors’ calculations.
Caregiving
3.8%
Health or
disability
5.1%
Other
4.7%
Student
2.5%
Work-related
34.6%
Caregiving
1.5%
Health or
Disability
7.3%
Student
2.7%
Work-
related
2.5%
Not in
Labor
Force
Labor Force Participant
Stable work, never asked
35.0%
Not interested
in working
0.3%
12
The Hamilton Project • Brookings
FIGURE 4B.
Most-Frequent Reason for Not Working for Pay, SNAP Participants Aged 50–59 with No
Dependent under Age 6
Source: SIPP (U.S. Census Bureau 2013–14); authors’ calculations.
Not in Labor Force
Labor Force Participant
Stable work, never asked
19.2%
Caregiving
1.7%
Health or
disability
13.6%
Work-related
16.5%
Health or
Disability
39.6%
Work-related
3.8%
Not interested
in working
0.7%
Early
retirement
0.7%
Student
0.2%
Other
1.8%
Caregiving
1.1%
Early retirement
0.7%
Not interested
in working
0.2%
Other
0.1%
green crosshatch shows the share of the population that
did not experience a gap in employment over the two-year
period, and thus were never asked why they were not working.
Among those who were asked why they were not working
for pay during at least one week, we report the reason for not
working in months they were not working. ose in solid
shades of green were in the labor force but experienced at least
one spell of unemployment or labor force nonparticipation.
ose in the blue were out of the labor force for the entire two-
year period. Each person is assigned one reason—their most
frequent reason—for not working.
Among those aged 18–49 with dependents aged 6–17 who are
newly exposed to work requirements (gure 4a), 86 percent
were in the labor force at some point over two years but not all
worked stably. Among those who did not work for pay for at
least one week but were in the labor force, the overwhelming
majority gave work-related reasons (68 percent), such as
temporary loss of job, temporary loss of hours (e.g., weather-
related, not getting enough shis, etc.), or a company shutting
down a plant or location. Other large groups include those
who are caregivers and those with health concerns. In a
program with extensive good-cause waivers, it appears the
bulk of these workers would not lose benets if waivers were
implemented with delity; but the administrative burden
required to sort those with work-related problems from those
who choose to not work could be quite high.
Among those out of the labor force for the entire two-year
period, more than half cite health reasons for being out of
the labor force. In total, 0.3percent of those aged 18–49 who
would be newly exposed to work requirements and who were
labor force nonparticipants said that they were not interested
in working.
Among individuals aged 50–59 (gure 4b), far more are out
of the labor force consistently and far fewer have stable work.
Overall, health (87 percent) and work-related (8 percent)
issues dominate. e prevalence of health problems is striking
considering we have already limited the sample to those not
receiving disability payments. Fewer than 1 percent were
retired or not interested in working.
e share of older SNAP participants listing caregiving as
a reason for being not in the labor force is notably smaller
than the share of the younger SNAP participant population.
13
The Hamilton Project • Brookings
Roughly 11 percent of SNAP participants aged 18–49 with
a dependent 6–17 that were out of the labor force for the
entire 24-month period list caregiving as a reason for not
being in the labor force. However, even 11 percent is smaller
than many might expect. Many caregivers who are not in
the labor force are in two-adult households where the other
adult is working. In addition, many are in households with
dependents aged 0–5, and those households are exempt from
work requirements.
In summary, based on 2013–14 data, 5.5million adult SNAP
participants would be newly exposed to work requirements
with 3.8million who would have failed them at some point
in this two-year window. Notable among those who were
asked about a spell of not working, 2.1million report health
or disability issues and 1.5million report work-related issues.
Only about 90,000 list a lack of interest or early retirement as
their reason for not working.
EXPOSURE TO PROPOSED WORK REQUIREMENTS IN
MEDICAID
We study the work participation of Medicaid beneciaries in a
similar manner. Unlike SNAP, there is no current population
of participants who face work requirements across the country
to use as a comparison group. As noted above, previous
administrations and the courts have not viewed Medicaid
work requirements as supporting core program goals; there
are substantive doubts about whether work requirements for
health insurance are appropriate. Nevertheless, we consider
the employment status of Medicaid beneciaries to illuminate
how such requirements would function.
Since Medicaid beneciaries do not currently face work
requirements, we do not separately examine the population
aged 18–49 without dependents. It is instructive to
dierentiate the work status transitions of younger (aged
18–49) and older (aged 50–64) Medicaid beneciaries,
restricted to those who either have a dependent 6–17 or no
dependents, i.e. no dependents under the age of 6. We identify
employment status transitions and the reasons given for not
working among those targeted for work requirements over 24
consecutive months (January 2013–December 2014).
Figure 5 shows that over two years (2013 and 2014), 80percent
of Medicaid beneciaries aged 18–49 without a dependent
child under age 6 were in the labor force at some point.
While about 40percent consistently worked over the 20-hour
threshold, 25 percent worked more than 20 hours at some
point but would potentially lose benets for falling below the
20-hour threshold for a month at another point.
e picture is quite dierent for older Medicaid beneciaries
(50 to 64) who would be exposed to work requirements. Of
that population, 44percent were out of the labor force for all
24 months. About 29percent worked consistently more than
20 hours a week and about 17percent worked more than 20
hours at least once but failed to do so every month. e reasons
given among working-age adult Medicaid beneciaries not
working for pay suggest that labor market reasons dominate
FIGURE 5.
Employment Status over Two Years, Medicaid Participants
SIPP (U.S. Census Bureau 2013–14); authors’ calculations.
0
25
50
75
100
Age 18−49,
no dependent under 6
Age 50−64,
no dependent under 6
Percent
Stable
Transitioned
Employed <20 hours
Between 20+ hours
and unemployment
or not in the labor force
Between 20+ hours
and <20 hours
Employed 20+ hours
Unemployed
Not in labor force
Other transition
14
The Hamilton Project • Brookings
FIGURE 6A.
Most-Frequent Reason for Not Working for Pay, Medicaid Participants Aged 18–49 with No
Dependents under Age 6
Source: SIPP (U.S. Census Bureau 2013–14); authors’ calculations.
Caregiving
4.6%
Health or
disability
7.3%
Other
4.5%
Student
3.9%
Work-related
25.1%
Health or
Disability
14.5%
Student
1.6%
Work-
related
2.0%
Not in
Labor
Force
Labor Force Participant
Stable work, never asked
33.8%
Caregiving
1.5%
Not interested
in working
1.1%
FIGURE 6B.
Most-Frequent Reason for Not Working for Pay, Medicaid Participants Aged 50–64 with No
Dependents under Age 6
Not in Labor Force
Labor Force Participant
Stable work, never asked
28.0%
Health or disability
8.8%
Work-related
12.5%
Health or Disability
35.0%
Work-related
3.7%
Early
retirement
1.9%
Other
2.3%
Caregiving
1.7%
Early retirement
3.1%
Not interested
in working
0.5%
Not interested
in working
0.9%
Care-
giving
1.4%
15
The Hamilton Project • Brookings
among labor force participants and health reasons dominate
among labor force nonparticipants (gures 6a and 6b). Once
again, only a small number of labor force nonparticipants are
not interested in work or are retired.
Among older participants of Medicaid (aged 50–64 without a
dependent under age 6, the population making up 37percent
of the sample population), 35 percent of those with Medicaid
coverage are out of the labor force for health reasons; this
group represents 79 percent of those who were not in the
labor force for the full two years. It is worth noting that
work requirements for this group would necessitate either
lax requirements with a very large portion of the population
getting waivers, or an administratively burdensome process
to determine which individual’s health concerns truly limit
them from work.
Work Status in a Snapshot vs.
Two Years
In its report on work requirements, the Council of Economic
Advisers (CEA 2018) looked at employment among adult
program participants for the month of December 2013 using
the SIPP and found that about three in ve participants
worked fewer than 20 hours per month. e CEA concludes
that this level of work—or lack thereof—“suggest[s] that
legislative changes requiring them to work and supporting
their transition into the labor market, similar to the approach
in TANF, would aect a large share of adult beneciaries and
their children in these non-cash programs” (1–2).
A critical empirical takeaway from the analysis presented
herein is that frequent movement between labor status
categories over time increases the number of people exposed
to losing benets for failing to consistently meet a work
requirement, and decreases the number of people who are
entirely out of the labor market. We now examine how the
analysis of work experiences diers when we compare a
snapshot in time—one month—with analysis that includes
transitions across status over two years. When we compare the
one month of SIPP data cited in the CEA report (December
2013) against 24 months, we nd that fewer program
participants are labor force nonparticipants and fewer meet
the work requirement threshold.
Figure 7 demonstrates how observed employment status is
dierent in one month versus two years. e rst two bars
show employment status categories for the full population
aged 18–59 without dependents aged 0–5, disability
payments, or status as students. e second two bars show
employment status categories in one month and two years
for SNAP participants aged 18–59 with no dependents aged
0–5, disability payments, or status as students. An “other”
transition during a one-month period are those who report
FIG U R E 7.
Employment Status in One Month vs. Two Years, SNAP
SIPP (U.S. Census Bureau 2013–14); authors’ calculations.
Month Two year Month Two year
0
25
50
75
100
Percent
Overall SNAP
Stable
Transitioned
Employed <20 hours
Between 20+ hours
and unemployment
or not in the labor force
Between 20+ hours
and <20 hours
Employed 20+ hours
Unemployed
Not in labor force
Other transition
16
The Hamilton Project • Brookings
FIGURE 8.
Employment Status in One Month vs. Two Years, Medicaid
SIPP (U.S. Census Bureau 2013–14); authors’ calculations.
One month Two years One month Two years
0
25
50
75
100
Percent
Stable
Transitioned
Employed <20 hours
Between 20+ hours
and unemployment
or not in the labor force
Between 20+ hours
and <20 hours
Employed 20+ hours
Unemployed
Not in labor force
Other transition
Overall Medicaid
being unemployed and a labor force nonparticipant during
dierent weeks within December 2013.
e rst feature that jumps out of the data is that far fewer
people are out of the labor force than is generally assumed.
While a one-month snapshot shows that 20 percent of the
overall population is not working (either out of the labor
force or unemployed), over the course of two years more
than 90percent of the overall population is employed at some
point. Many people are not truly on the sidelines as much as
they are cycling in and out of the game. Furthermore, fewer
people are solidly in the 20 hours plus–workforce. e share of
the overall population that stably works more than 20 hours
per week falls from 76percent in the one-month snapshot to
69percent over two years.
Looking only at those who participated in SNAP at any point
during the two-year period, the one-month snapshot is also
dierent from the two-year, both in terms of the number
of participants out of the labor force and the number who
would retain benets under the work requirement proposal.
Instead of 42percent being out of the labor force and roughly
11percent unemployed in the one-month snapshot—leading
to more than half of the group being labeled “not working”
in the one-month snapshot—roughly 29 percent are out of
the labor force and just 1percent are persistently unemployed
over two years, meaning fewer than one third are not working
consistently. Recall that the higher “not working” rate among
SNAP beneciaries is largely driven by those aged 50–59.
SNAP recipients aged 18–49 without dependents have a
“not working” rate of 25 percent over two years, and those
with dependents aged 6–17 have a “not working” rate of just
14percent. Almost a quarter of SNAP participants would fail
the work requirements some months and pass them in others,
with the majority giving work-related reasons for their change
in status.
A similar pattern holds for Medicaid beneciaries: the
monthly snapshot overstates the number of labor force
nonparticipants and understates those who would meet a
work requirement. ere is a 10 percentage point–reduction
in the share of those not working over one month (39percent)
versus two years (29percent). Forty-twopercent would meet
the work requirement in one month, but only 36percent do
over two years. In addition, in the two-year sample 22percent
of participants work over 20 hours in at least one month in the
sample but fail to in other months (gure 8).
Conclusion
e combination of a strong labor market, work requirements
to receive cash benets through TANF, and work incentives
generated by the EITC raised labor force participation rates
among single mothers in the mid-1990s (Ziliak 2016), leading
some to believe that further participation gains could be
obtained by extending only the work requirement component
to other programs (Haskins 2018; CEA 2018).
17
The Hamilton Project • Brookings
Work requirements are intended to counter any work
disincentives that come from a social safety net and to
ensure that society is not unnecessarily supporting people
who could otherwise support themselves. At the same time,
such work requirements add administrative complexity to
social programs and risk keeping benets from parts of the
population that should be receiving them. is economic
analysis establishes a set of facts that are relevant when
considering the expansion of work requirements.
What types of populations will face these new work
requirements? How many would fail to meet the requirements?
Do program participants appear to already be in the labor
force facing work-related constraints on hours or do they
choose not to work? And how many would in theory be
eligible for waivers relative to those individuals that society
would like to push toward work?
A large number of SNAP and Medicaid participants who
would face new work requirements cycle in and out of the labor
force and would thus lose benets at certain times. Among
those who are in the labor force, spells of unemployment are
either due to job-related concerns or health issues. Very few
reported that they were not working due to lack of interest.
Among those out of the labor force for the entire two-year
period, health concerns are the overriding reason for not
working, even aer removing those who receive disability
benets from the sample. e older portion of the population
newly exposed to work requirements is more likely to be out of
the labor force for extended periods of time. Among this group,
again, health reasons are the overriding factor in not working.
Work requirements for this group might push more onto
disability rolls, make the disability adjudication even more
consequential, and require a separate health investigation to
settle all the necessary waivers. Failure to receive a waiver
would result in disenrollment; losing access to these programs
would reduce resources available to purchase food and health
insurance among otherwise eligible households.
For those who qualify for exemptions, satisfy waiver
requirements, or work enough to meet the requirements, there
are still signicant informational and administrative barriers
to compliance. Program participants must understand how
the work requirement policy relates to them, obtain and
submit documentation, and do so at the frequency prescribed
by the state (Wagner and Solomon 2018). Frequent exposure to
verication processes, such as the monthly reporting periods
prescribed in the Agricultural Act of 2014 (the Farm Bill) and
many states’ Medicaid proposals, increases the administrative
burden on participants and enforcers, the likelihood of error,
and cost (Bauer and Schanzenbach 2018b). ese continuing
roadblocks to participation, with attendant informational and
transactional costs, are likely to result in lower take-up among
the eligible population and disenrollment (Finkelstein and
Notowidigdo 2018).
Looking at snapshots of work experience, such as a single
month, inates both the number of SNAP and Medicaid
participants who are out of the labor force and the number of
people who work sucient hours to satisfy work requirements.
Over 24 consecutive months the number of SNAP and
Medicaid program beneciaries not working or seeking work
as well as those working consistently above 20 hours fall
substantially.
ere are safety net levers that can be used to pull those out
of the labor force into work. Steps such as increasing the EITC
might be a very eective way to increase work participation
in this group without the same administrative burdens and
negative spillovers to vulnerable populations. (See Hoynes,
Rothstein, and Runi 2017 for a specic proposal along these
lines.) at proposal is estimated to increase participation by
600,000 people. Raising the returns to work via the EITC or
other measures, creating training or educational opportunities
that can increase individuals’ human capital, and providing
child care or improved treatment and medical care to reduce
health barriers to work could make full attachment to the
labor force more viable for many individuals.
18
The Hamilton Project • Brookings
Endnotes
1. See technical appendix tables 1 and 2 for additional work status transition
statistics.
2. e states not implementing able-bodied adult without dependents
waivers at some point during 2013–14 are: Delaware, Guam, Iowa, Kansas,
Nebraska, Oklahoma, Utah, Virginia, and Wyoming. States implementing a
partial waiver (partial referring to dierent parts of the state or only part of
the year): Colorado, Minnesota, New Hampshire, New York, North Dakota,
Ohio, South Dakota, Texas, Vermont, Wisconsin.
3. ose who meet the 20-hour threshold monthly hours variable include both
those who meet the threshold every week and those whose hours varied
each week but averaged to 20 hours per week each month. e volatility of
their hours may suggest they are more likely to fail the work requirement
threshold but they did not do so over the two-year window.
References
Abraham, Katharine G., and Melissa S. Kearney. 2018. “Explaining
the Decline in the U.S. Employment-to-Population Ratio:
A Review of the Evidence.” Working Paper 24333, National
Bureau of Economic Research, Cambridge, MA.
Agricultural Act of 2014 (Farm Bill), Pub.L. 113-79 (2014).
Arkansas Department of Human Services. n.d. “Access Arkansas.”
(https://access.arkansas.gov.)
Baicker, Katherine, Amy Finkelstein, Jae Song, and Sarah Taubman.
2014. “e Impact of Medicaid on Labor Market Activity
and Program Participation: Evidence from the Oregon
Health Insurance Experiment.” American Economic Review
104 (5): 322–28.
Bauer, Lauren, and Diane Whitmore Schanzenbach. 2018a, July
27. “Employment Status Changes Put Millions at Risk
of Losing SNAP Benets for Years.” Blog. e Hamilton
Project, Brookings Institution, Washington, DC.
———. 2018b, August 9. “Who Loses SNAP Benets If Additional
Work Requirements Are Imposed? Workers.” Blog. e
Hamilton Project, Brookings Institution, Washington, DC.
Black, Sandra E., Diane W. Schanzenbach, and Audrey Breitwieser.
2017. “e Recent Decline in Women’s Labor Force
Participation.” In Driving Growth through Women’s
Economic Participation edited by Diane W. Schanzenbach
and Ryan Nunn, 5–17. Washington, DC: e Hamilton
Project.
Blank, Rebecca. 2002.“Evaluating Welfare Reform in the United
States.” Journal of Economic Literature 40 (4): 1105–66.
Blinder, Alan S., and Mark Zandi. 2015. “e Financial Crisis:
Lessons for the Next One.” Policy Futures, Center on
Budget and Policy Priorities, Washington, DC.
Bolen, Ed, Lexin Cai, Stacy Dean, Brynne Keith-Jennings, Catlin
Nchako, Dottie Rosenbaum, and Elizabeth Wolkomir.
2018. “House Farm Bill Would Increase Food Insecurity
and Hardship.” Center on Budget and Policy Priorities,
Washington, DC.
Buchmueller, omas, John C. Ham, and Lara D. Shore-Sheppard.
2016. “e Medicaid Program.” In Economics of Means-
Tested Transfer Programs in the United States, Vol. I, edited
by Robert A. Mott, 21–136. Chicago, IL: University of
Chicago Press.
Bureau of Labor Statistics (BLS). 1977–2017. “Current Population
Survey Annual Social and Economic Supplement.” Bureau
of Labor Statistics, U.S. Department of Labor, Washington,
DC. Retrieved from IPUMS-CPS, University of Minnesota.
———. 2018. “Current Population Survey Annual Social and
Economic Supplement.” Bureau of Labor Statistics, U.S.
Department of Labor, Washington, DC. Retrieved from
IPUMS-CPS, University of Minnesota.
Butcher, Kristin F., and Diane Whitmore Schanzenbach. 2018.
“Most Workers in Low-Wage Labor Market Work
Substantial Hours, in Volatile Jobs.” Policy Futures, Center
on Budget and Policy Priorities, Washington, DC.
e Centers for Medicare & Medicaid Services. 2018. “Remarks
by Administrator Seema Verma at the 2018 Medicaid
Managed Care Summit.” As prepared for delivery:
September 27, 2018. e Centers for Medicare & Medicaid
Services, Washington, DC.
Coglianese, John. 2018, February. “e Rise of In-and-Outs:
Declining Labor Force Participation of Prime Age Men.”
Working Paper, Harvard University, Cambridge, MA.
Council of Economic Advisers (CEA). 2016. “e Long-Term
Decline in Prime-Age Male Labor Force Participation.”
Council of Economic Advisers, White House, Washington,
DC.
———. 2018. “Expanding Work Requirements in Non-Cash
Welfare Programs.” Council of Economic Advisers, White
House, Washington, DC.
Dague, Laura, omas DeLeire, and Lindsey Leininger. 2017. “e
Eect of Public Insurance Coverage for Childless Adults
on Labor Supply.” American Economic Journal: Economic
Policy 9 (2): 124–54.
Duggan, Mark, Gopi Shah Goda, and Emilie Jackson. 2017. “e
Eects of the Aordable Care Act on Health Insurance
Coverage and Labor Market Outcomes.” Working Paper
23607, National Bureau of Economic Research, Cambridge,
MA.
Fang, Hanming, and Michael P. Keane. 2004. “Assessing the Impact
of Welfare Reform on Single Mothers.” Brookings Papers on
Economic Activity (1): 1-95.
19
The Hamilton Project • Brookings
Finkelstein, Amy, Nathaniel Hendren, and Erzo F. P. Luttmer. 2015.
“e Value of Medicaid: Interpreting Results from the
Oregon Health Insurance Experiment.” Working Paper
21308, National Bureau of Economic Research, Cambridge,
MA.
Finkelstein, Amy, and Matthew J. Notowidigdo. 2018. “Take-up and
Targeting: Experimental Evidence from SNAP.” Working
Paper 24652, National Bureau of Economic Research,
Cambridge, MA.
Gangopadhyaya, Anuj, Genevieve M. Kenney, Rachel A. Burton,
and Jeremy Marks. 2018. “Medicaid Work Requirements in
Arkansas: Who Could Be Aected, and What Do We Know
about em?” Urban Institute, Washington, DC.
Gareld, Rachel, Robin Rudowitz, and Anthony Damico. 2018.
“Understanding the Intersection of Medicaid and Work.”
Henry J. Kaiser Family Foundation, San Francisco, CA.
Gareld, Rachel, Robin Rudowitz, and MaryBeth Musumeci. 2018.
“Implications of a Medicaid Work Requirement: National
Estimates of Potential Coverage Losses.” Henry J. Kaiser
Family Foundation, San Francisco, CA.
Gareld, Rachel, Robin Rudowitz, MaryBeth Musumeci,
and Anthony Damico. 2018. “Implications of Work
Requirements in Medicaid: What Does the Data Say?”
Henry J. Kaiser Family Foundation, San Francisco, CA.
Garthwaite, Craig, Tal Gross, and Matthew J. Notowidigdo. 2014.
“Public Health Insurance, Labor Supply, and Employment
Lock.” Quarterly Journal of Economics 129 (2): 653–96.
Goldman, Anna L., Stee Woolhandler, David U. Himmelstein,
David H. Bor, and Danny McCormick. 2018. “Analysis of
Work Requirement Exemptions and Medicaid Spending.”
JAMA Internal Medicine online, September 10.
Gooptu, Angshuman, Asako S. Moriya, Kosali I. Simon, and
Benjamin D. Sommers. 2016. “Medicaid Expansion Did
Not Result in Signicant Employment Changes or Job
Reductions in 2014.” Heath Aairs 35 (1): 111–18.
Grogger, Jerey. 2004. “Welfare Transitions in the 1990s: e
Economy, Welfare Policy, and the EITC.”Journal of Policy
Analysis and Management23 (4): 671–95.
Hamilton, Gayle, Stephen Freedman, Lisa Gennetian, Charles
Michalopoulos, Johanna Walter, Dianna Adams-Ciardullo,
and Anna Gassman-Pines. “National Evaluation of
Welfare-to-Work Strategies.”2001. Prepared by MDRC,
Washington, DC for the U.S. Department of Health and
Human Services and U.S. Department of Education.
Haskins, Ron. 2018, July 25. “Trump’s Work Requirements Have
Been Tested Before. ey Succeeded.” e Washington Post.
Hoynes, Hilary, Jesse Rothstein, and Krista Runi. 2017. “Making
Work Pay Better through an Expanded Earned Income Tax
Credit.” Policy Proposal 2017-09, e Hamilton Project,
Brookings Institution, Washington, DC.
Hoynes, Hilary Williamson, and Schanzenbach, Diane Whitmore.
2012. “Work Incentives and the Food Stamp Program.”
Journal of Public Economics 96 (1): 151–62.
———. 2016, January. “e Safety Net as an Investment.” Working
Paper, University of California, Berkeley.
Institute of Medicine. 2003. “Spending on Health Care for
Uninsured Americans: How Much, and Who Pays?”
In Hidden Costs, Values Lost: Uninsurance in America,
edited by Institute of Medicine (US) Committee on the
Consequences of Uninsurance. Washington, DC: National
Academies Press.
Joint Economic Committee. 2018. “Inactive, Disconnected, and
Ailing: A Portrait of Prime-Age Men Out of the Labor
Force.” SCP Report 2-18, Joint Economic Committee,
Washington, DC.
Juhn, Chinhui. 1992. “Decline of Male Labor Market Participation:
e Role of Declining Market Opportunities.” e
Quarterly Journal of Economics 107 (1): 79–121.
Kaestner, Robert, Bowen Garrett, Jiajia Chen, Anuj
Gangopadhyaya, and Caitlyn Fleming. 2017. “Eects
of ACA Medicaid Expansions on Health Insurance
Coverage and Labor Supply.” Journal of Policy Analysis and
Management 36 (3): 608–42.
Keith-Jennings, Brynne, and Dottie Rosenbaum. 2015. “SNAP
Benet Boost in 2009 Recovery Act Provided Economic
Stimulus and Reduced Hardship.” Center on Budget and
Policy Priorities, Washington, DC.
Meyer, Bruce D., and Dan T. Rosenbaum. 2001. “Welfare, e
Earned Income Tax Credit, and e Labor Supply of Single
Mothers.”e Quarterly Journal of Economics116 (3): 1063-
1114.
Neale, Brian. 2018. “RE: Opportunities to Promote Work and
Community Engagement among Medicaid Beneciaries.”
SMD Letter 18-002, Center for Medicare & Medicaid
Services, U.S. Department of Health and Human Services,
Washington, DC.
Patient Protection and Aordable Care Act (ACA), Pub. L. 111-148,
124 Stat. 119 through 124 Stat. 1025 (2010).
Pavetti, LaDonna, and Liz Schott. 2016. “TANF at 20: “TANF at 20:
Time to Create a Program that Supports Work and Helps
Families Meet eir Basic Needs.” Center on Budget and
Policy Priorities, Washington, DC.
20
The Hamilton Project • Brookings
Price, omas E., and Seema Verma. 2017. “Letter to the Nation’s
Governors.” U.S. Department of Health and Human
Services, Washington, DC.
Rosenbaum, Dorothy. 2013. “e Relationship between SNAP and
Work among Low-Income Households.” Center on Budget
and Policy Priorities, Washington, DC.
Rudowitz, Robin, and MaryBeth Musumeci. 2018. “An Early
Look at State Data for Medicaid Work Requirements
in Arkansas.” Henry J. Kaiser Family Foundation, San
Francisco, CA.
Schanzenbach, Diane W. 2018. “Can Benets and Incentives
Promote Work?” Journal of Policy Analysis and
Management 37 (4): 903-11.
Silberman, Lizbeth. 2013. “Supplemental Nutirition Assistance
Program (SNAP): Able Bodied Adults Without Dependents
Waivers for Fiscal Year 2014.” Policy Memorandum,
Augiust 2. U.S. Department of Agriculture, Food and
Nutrition Service. (https://fns-prod.azureedge.net/sites/
default/les/FY_2014_ABAWD_Waiver_Status.pdf.)
Stewart v. Azar (Ronnie Maurice Stewart et al. v. Alex M. Azar II et
al.). Civil Action No. 18-152 (JEB) (D.D.C. Apr. 10, 2018).
Urban Institute. 2018, October. “Work Requirements Tracker.”
Urban Institute, Washington, DC.
U.S. Census Bureau. 2013–14. “Survey of Income and Program
Participation.” U.S. Census Bureau, U.S. Department of
Commerce, Suitland, MD.
U.S. Department of Agriculture (USDA). 2017a. “Characteristics
of USDA Supplemental Nutrition Assistance Program
Households: Fiscal Year 2016 (Summary).” U.S.
Department of Agriculture, Washington, DC.
———. 2017b. “Supplemental Nutrition Assistance Program
(SNAP): Facts About SNAP.” U.S. Department of
Agriculture, Washington, DC.
Wagner, Jennifer, and Judith Solomon. 2018. “States’ Complex
Medicaid Waivers Will Create Costly Bureaucracy and
Harm Eligible Beneciaries.” Center on Budget and Policy
Priorities, Washington, DC.
White House. 2018. “Executive Order Reducing Poverty in America
by Promoting Opportunity and Economic Mobility.” White
House, Washington, DC.
Wolkomir, Elizabeth, and Lexin Cai. 2018. “e Supplemental
Nutrition Assistance Program Includes Earnings
Incentives.” Center on Budget and Policy Priorities,
Washington, DC.
Ziliak, James P. 2016. “Temporary Assistance for Needy Families.”
In Economics of Means-Tested Transfer Programs in the
United States, Vol. I, edited by Robert A. Mott, 303–93.
Chicago, IL: University of Chicago Press.
21
The Hamilton Project • Brookings
Appendix
Stable
(not in
labor
force)
Stable
(unemployed)
Stable
(employed
20+ hours)
Stable
(employed
<20 hours)
Transitioned
between 20+
hours and <20
hours
Transitioned
between 20+
hours and
unemployment
or not in labor
force
Other
transition
Age 18–49, no dependents
2013 34.3% 5.5% 33.3% 4.1% 7.9% 4.9% 10.0%
2014 32.6% 5.5% 37.4% 3.5% 9.1% 7.2% 4.7%
2013–14 24.6% 1.7% 32.6% 1.7% 16.0% 9.3% 14.1%
Age 18–49, dependent 6–17
2013 20.4% 4.9% 49.9% 2.4% 8.9% 6.0% 7.5%
2014 21.0% 4.2% 50.2% 2.4% 8.6% 9.9% 3.8%
2013–14 14.0% 0.7% 45.6% 0.4% 15.9% 12.3% 11.3%
Age 50–59, no dependent under 6
2013 50.4% 4.6% 25.8% 2.6% 5.7% 3.9% 7.0%
2014 53.3% 3.5% 26.1% 2.5% 5.9% 5.1% 3.6%
2013–14 45.7% 1.3% 23.0% 1.4% 10.1% 7.9% 10.7%
APPENDIX TABLE 1.
Employment Status, SNAP Participants
Source: SIPP (U.S. Census Bureau 2013–14); authors’ calculations.
Note: The sample is limited to U.S. citizens, nonactive military, who reported receiving SNAP benefits at any point between January 2013
and December 2014. Only respondents with 24 months of data were included. Those with children under age 6, full- or part-time students,
and those who reported receiving disability benefits were excluded from the sample based on categorical work requirement exclusions.
Those who were assigned to “stable” categories were observed as not in the labor force, unemployed, above the 20-hour threshold,
or below the 20-hour threshold per week. Those who were stable and employed more than 20 hours a week were assigned either by
meeting the threshold every week or because the monthly hours total averaged to above 20 hours per week. Regardless of the number of
transitions made, each person who was observed as switching between work statuses was assigned to one group in the following order:
first, transitioned between more than and less than 80 hours per month; second, transitioned between more than 80 hours per month
and unemployment or labor force nonparticipation; third, other. “Other” includes those who transitioned between less than 80 hours
per month and unemployment or labor force nonparticipation as well as those who transitioned between unemployment and labor force
nonparticipation.
22
The Hamilton Project • Brookings
Stable
(not in
labor
force)
Stable
(unemployed)
Stable
(employed
20+ hours)
Stable
(employed
<20 hours)
Transitioned
between 20+
hours and <20
hours
Transitioned
between 20+
hours and
unemployment
or not in labor
force
Other
transition
Age 18–49, no dependent under 6
2013 27.7% 3.8% 42.6% 3.6% 8.1% 4.1% 10.0%
2014 26.4% 4.2% 46.1% 3.3% 7.3% 7.6% 5.1%
2013–14 19.6% 1.1% 39.6% 1.1% 14.8% 10.9% 12.8%
Age 50–64, no dependent under 6
2013 48.4% 3.2% 32.9% 3.8% 5.5% 2.2% 4.0%
2014 51.2% 2.7% 29.9% 3.5% 5.0% 4.6% 3.1%
2013–14 44.1% 0.8% 28.5% 1.8% 11.7% 5.0% 8.2%
APPENDIX TABLE 2.
Employment Status, Medicaid Participants
Source: SIPP (U.S. Census Bureau 2013–14); authors’ calculations.
Note: The sample is limited to U.S. citizens, nonactive military, who reported receiving Medicaid benefits at any point between January 2013
and December 2014. Only respondents with 24 months of data were included. Those with children under age 6, full- or part-time students,
those who reported receiving Medicare, and those who reported receiving disability benefits were excluded from the sample based on
categorical work requirement exclusions. Those who were stable labor force nonparticipants are contrasted with those who were in the labor
force (working or seeking work) at least once during the two-year period. Those who were assigned to “stable” categories were observed as
not in the labor force, unemployed, working above the 20-hour threshold, or working below the 20-hour threshold per week. Those who were
stable and employed more than 80 hours per week were assigned either by meeting the 20 hours per week threshold every week or because
the monthly hours total averaged above 20 hours per week. Regardless of the number of transitions made, each person who was observed
as switching between work statuses was assigned to one group in the following order: first, transitioned between more than and less than
80 hours per month; second, transitioned between more than 80 hours per month and unemployment or labor force nonparticipation; third,
other. “Other” includes those who transitioned between less than 80 hours per month and unemployment or labor force nonparticipation as
well as those who transitioned between unemployment and labor force nonparticipation.
23
The Hamilton Project • Brookings
Technical Appendix
Box Figure 1. Prime-Age Women’s Labor Force
Participation, by Marital Status and Presence of
Children under Age 18
Source: Current Population Survey Annual Social and
Economic Supplement (ASEC) (Bureau of Labor Statistics
[BLS] 1977–2017); authors’ calculations.
Note: “Prime-age” indicates ages 25 to 54, inclusive. “Married”
is dened by women who have a spouse in the household or
not in the household. “Single” is dened as all other women,
including divorced and widowed women. “With children” is
dened as having at least one child in the household under the
age of 18. “No children” is dened as having no children in the
household under the age of 18. Population counts calculated
using the Annual Social and Economic Supplement weight.
Figure 1. Exposure to Work Requirements among
Adult SNAP Participants, 2017
Source: ASEC (BLS 2018); authors’ calculations.
Notes ose who would be exempt from work requirements if
the House bill work requirements were passed include those
over the age of 59, those with a dependent under the age of 6,
full- or part-time students, and those who receive disability
benets. While in some states work requirements are
waived for those aged 18–49 with no dependents, state-level
dierences are not accounted for in identifying those who are
currently exposed to work requirements. Population counts
calculated using the Annual Social and Economic Supplement
weight among U.S. citizens over the age of 18 who reported
receiving SNAP benets at some point during 2017.
Figure 2. Exposure to Work Requirements among
Adult Medicaid Participants, 2017
Source: ASEC (BLS 2018); authors’ calculations.
Note: States applying for waivers to add work requirements
to Medicaid have identied dierent categorical exemptions
and conditions for waivers. For this exercise, we identied
the most frequent categorical exemptions and applied those
rules nationally. ose who are over the age of 64 or who are
dual Medicare enrollees are exempt, those receiving disability
income are exempt, those with a dependent under the age
of 6 are exempt, and full- or part-time students are exempt.
Population counts are calculated using the Annual Social and
Economic Supplement weight among U.S. citizens over the
age of 18 who reported receiving Medicaid benets at some
point during 2017.
Figure 3. Employment Status over Two Years, SNAP
Participants
Source: Survey of Income and Program Participation (SIPP)
(U.S. Census Bureau 2013–14); authors’ calculations.
Note: e sample is limited to U.S. citizens, nonactive
military, aged 18–59 who reported receiving SNAP benets
at any point between January 2013 and December 2014. Only
respondents with 24 months of data were included. ose
with children under age 6, full- or part-time students, and
those who reported receiving disability benets were excluded
from the sample based on categorical work requirement
exclusions. ose who were assigned to “stable” categories
were observed as not in the labor force, unemployed, working
above the 20-hour threshold, or working below the 20-hour
threshold per week. ose who were stable and employed
more than 20 hours a week were assigned either by meeting
the threshold every week or because the monthly hours
total averaged to above 20 hours per week. Regardless of the
number of transitions made, each person who was observed as
switching between work statuses was assigned to one group in
the following order: rst, transitioned between more than and
less than 80 hours per month; second, transitioned between
more than 80 hours per month and unemployment or labor
force nonparticipation; third, other. “Other” includes those
who transitioned between less than 80 hours per month and
unemployment or labor force nonparticipation as well as
those who transitioned between unemployment and labor
force nonparticipation.
Figures 4A and 4B. Most Frequent Reason for Not
Working for Pay, SNAP Participants
Source: SIPP (U.S. Census Bureau 2013–14); authors’
calculations.
Notes e sample is limited to U.S. citizens, nonactive
military, aged 18–59 who reported receiving SNAP benets
at any point between January 2013 and December 2014. Only
respondents with 24 months of data were included. ose with
24
The Hamilton Project • Brookings
children under age 6, full- or part-time students, and those
who reported receiving disability benets were excluded from
the sample based on categorical work requirement exclusions.
Figure 4a is further restricted to those between the ages of 18
and 49 with a dependent between the ages of 6 and 17 while
gure 4b is limited to those between the ages of 50 and 59 with
no dependents under the age of 6. Each person’s most frequent
response for why they were not working was used to calculate
the distribution; ties were assigned in descending order by
work-related, health or disability, caregiving, student, early
retirement, not interested in working, and other. e “stable
work, not asked” group indicates that the respondent was
never asked this survey question because they were working
for pay every week. “Work-related” includes not being able to
nd work, being laid o, or working for more than 15 hours
for no pay at a family business or farm. “Health or disability”
includes being unable to work because of an injury, illness, or
chronic health condition or disability. “Caregiving” includes
those not working due to pregnancy or recent childbirth, or
taking care of children or other persons. Students included
in the sample are those who did not report that they were
enrolled full- or part-time but reported not working because
they were going to school.
Figure 5. Employment Status over Two Years,
Medicaid Participants
Source: SIPP (U.S. Census Bureau 2013–14); authors’
calculations.
Note: e sample is limited to U.S. citizens, nonactive
military, aged 18–64 who reported receiving Medicaid
benets at any point between January 2013 and December
2014. Only respondents with 24 months of data were included.
ose with children under age 6, full- or part-time students,
those who reported receiving Medicare, and those who
reported receiving disability benets were excluded from the
sample based on categorical work requirement exclusions.
See technical appendix entry for gure 3 with regard to
employment status assignment.
Figures 6A and 6B. Most Frequent Reason for Not
Working for Pay, Medicaid Participants
Source: SIPP (U.S. Census Bureau 2013–14); authors’
calculations.
Note: e sample is limited to U.S. citizens, nonactive military,
aged 18–64 who reported receiving Medicaid benets at
any point between January 2013 and December 2014. Only
respondents with 24 months of data were included. ose
with children under age 6, full- or part-time students, those
who reported receiving Medicare, and those who reported
receiving disability benets were excluded from the sample
based on categorical work requirement exclusions. ose
who were stable labor force nonparticipants are contrasted
with those who were in the labor force (working or seeking
work) at least once during the two-year period. Figure 6a
is further restricted to those between the ages of 18 and 49
with a dependent between the ages of 6 and 17, whereas gure
6b is limited to those between the ages of 50 and 64 with no
dependents under the age of 6. See technical appendix entry
for gures 4a and 4b with regard to reason assignment.
Figure 7. Employment Status in One Month vs. Two
Years, SNAP
Source: SIPP (U.S. Census Bureau 2013–14); authors’
calculations.
Note: e sample is limited to U.S. citizens, nonactive military,
aged 18–59. Only respondents with 24 months of data were
included. ose currently exposed to work requirements,
those with children under age 6, full- or part-time students,
and those who reported receiving disability benets were
excluded from the sample. e one-month and two-year
samples dier by reported SNAP benet receipt. In the one-
month sample, “other” refers to those who switched between
labor force nonparticipation and unemployment during the
month of December 2013, the month chosen in the SIPP by
CEA for its report on work requirements.
Figure 8. Employment Status in One Month vs. Two
Years, Medicaid
Source: SIPP (U.S. Census Bureau 2013–14); authors’
calculations.
Note: e sample is limited to U.S. citizens, nonactive military,
aged 18–64. Only respondents with 24 months of data were
included. ose with children under age 6, full- or part-time
students, those who reported receiving Medicare, and those
who reported receiving disability benets were excluded
from the sample based on categorical work requirement
exclusions. e one-month and two-year samples dier by
reported Medicaid benet receipt. In the one-month sample,
“other” refers to those who switched between labor force
nonparticipation and unemployment during the month of
December 2013.
25
The Hamilton Project • Brookings
ADVISORY COUNCIL
GEORGE A. AKERLOF
University Professor
Georgetown University
ROGER C. ALTMAN
Founder & Senior Chairman, Evercore
KAREN L. ANDERSON
Senior Director of Policy & Communications
Becker Friedman Institute for
Research in Economics
The University of Chicago
ALAN S. BLINDER
Gordon S. Rentschler Memorial Professor of
Economics & Public Affairs
Princeton University
Nonresident Senior Fellow
The Brookings Institution
ROBERT CUMBY
Professor of Economics
Georgetown University
STEVEN A. DENNING
Chairman, General Atlantic
JOHN M. DEUTCH
Institute Professor
Massachusetts Institute of Technology
CHRISTOPHER EDLEY, JR.
Co-President & Co-Founder
The Opportunity Institute
BLAIR W. EFFRON
Partner, Centerview Partners LLC
DOUGLAS W. ELMENDORF
Dean & Don K. Price Professor
of Public Policy
Harvard Kennedy School
JUDY FEDER
Professor & Former Dean
McCourt School of Public Policy
Georgetown University
ROLAND FRYER
Henry Lee Professor of Economics
Harvard University
JASON FURMAN
Professor of the Practice of
Economic Policy
Harvard Kennedy School
Senior Counselor
The Hamilton Project
MARK T. GALLOGLY
Cofounder & Managing Principal
Centerbridge Partners
TED GAYER
Executive Vice President
Joseph A. Pechman Senior Fellow,
Economic Studies
The Brookings Institution
TIMOTHY F. GEITHNER
President, Warburg Pincus
RICHARD GEPHARDT
President & Chief Executive Officer
Gephardt Group Government Affairs
JOHN GRAY
President & Chief Operating Officer
Blackstone
ROBERT GREENSTEIN
Founder & President
Center on Budget and Policy Priorities
MICHAEL GREENSTONE
Milton Friedman Professor in Economics & the
College
Director of the Becker Friedman Institute for
Research in Economics
Director of the Energy Policy Institute
University of Chicago
GLENN H. HUTCHINS
Co-founder, North Island
JAMES A. JOHNSON
Chairman, Johnson Capital Partners
LAWRENCE F. KATZ
Elisabeth Allison Professor of Economics
Harvard University
MELISSA S. KEARNEY
Professor of Economics
University of Maryland
Nonresident Senior Fellow
The Brookings Institution
LILI LYNTON
Founding Partner
Boulud Restaurant Group
HOWARD S. MARKS
Co-Chairman
Oaktree Capital Management, L.P.
MARK MCKINNON
Former Advisor to George W. Bush
Co-Founder, No Labels
ERIC MINDICH
Chief Executive Officer & Founder
Eton Park Capital Management
ALEX NAVAB
Former Head of Americas Private Equity
KKR
Founder, Navab Holdings
SUZANNE NORA JOHNSON
Former Vice Chairman
Goldman Sachs Group, Inc.
PETER ORSZAG
Vice Chairman & Global Co-Head of Healthcare,
Lazard
Nonresident Senior Fellow
The Brookings Institution
RICHARD PERRY
Managing Partner & Chief Executive Officer
Perry Capital
PENNY PRITZKER
Chairman, PSP Partners
MEEGHAN PRUNTY
Managing Director, Blue Meridian Partners
Edna McConnell Clark Foundation
ROBERT D. REISCHAUER
Distinguished Institute Fellow&
President Emeritus
Urban Institute
ALICE M. RIVLIN
Senior Fellow, Economic Studies
Center for Health Policy
The Brookings Institution
DAVID M. RUBENSTEIN
Co-Founder & Co-Executive Chairman
The Carlyle Group
ROBERT E. RUBIN
Former U.S. Treasury Secretary
Co-Chair Emeritus
Council on Foreign Relations
LESLIE B. SAMUELS
Senior Counsel
Cleary Gottlieb Steen & Hamilton LLP
SHERYL SANDBERG
Chief Operating Officer, Facebook
DIANE WHITMORE SCHANZENBACH
Margaret Walker Alexander Professor
Director
The Institute for Policy Research
Northwestern University
Nonresident Senior Fellow
The Brookings Institution
STEPHEN SCHERR
Chief Executive Officer
Goldman Sachs Bank USA
RALPH L. SCHLOSSTEIN
President & Chief Executive Officer, Evercore
ERIC SCHMIDT
Technical Advisor, Alphabet Inc.
ERIC SCHWARTZ
Chairman & CEO, 76 West Holdings
THOMAS F. STEYER
Business Leader & Philanthropist
LAWRENCE H. SUMMERS
Charles W. Eliot University Professor
Harvard University
LAURA D’ANDREA TYSON
Professor of Business Administration &
Economics
Director
Institute for Business & Social Impact
Berkeley-Haas School of Business
JAY SHAMBAUGH
Director
26
The Hamilton Project • Brookings
WWW.HAMILTONPROJECT.ORG
1775 Massachusetts Ave., NW
Washington, DC 20036
(202) 797-6484
Printed on recycled paper.
Abstract
Basic assistance programs such as the Supplemental Nutrition Assistance Program (SNAP, formerly the Food Stamps Program)
and Medicaid ensure families have access to food and medical care when they are low-income. Some policymakers at the federal and
state levels intend to add new work requirements to SNAP and Medicaid. In this paper, we analyze those who would be impacted
by an expansion of work requirements in SNAP and an introduction of work requirements into Medicaid. We characterize the
types of individuals who would face work requirements, describe their labor force experience over 24 consecutive months, and
identify the reasons why they are not working if they experience a period of unemployment or labor force nonparticipation.
We nd that the majority of SNAP and Medicaid participants who would be exposed to work requirements are attached to the
labor force, but that a substantial share would fail to consistently meet a 20 hours per week–threshold. Among persistent labor
force nonparticipants, health issues are the predominant reason given for not working. ere may be some subset of SNAP
and Medicaid participants who could work, are not working, and might work if they were threatened with the loss of benets.
is paper adds evidence to a growing body of research that shows that this group is very small relative to those who would be
sanctioned under the proposed policies who are already working or are legitimately unable to work.
FIGURE 3.
Employment Status over Two Years, SNAP Participants
Source: Survey of Income and Program Participation (SIPP) (U.S. Census Bureau 2013–14); authors’ calculations.
0
25
50
75
100
Age 18−49,
no dependents
Age 18−49,
dependent 6−17
Age 50−59,
no dependent under 6
Percent
Stable
Transitioned
Employed <20 hours
Between 20+ hours
and unemployment
or not in the labor force
Between 20+ hours
and <20 hours
Employed 20+ hours
Unemployed
Not in labor force
Other transition
Currently
exposed
Newly
exposed
