874772ANN The Annals of The American AcademyWhat does medicaid do, and can we do it better?

research-article2019

This article provides an overview of the Medicaid pro-

gram and summarizes the evidence about its effective-
ness in terms of providing insurance and promoting
health. The evidence shows that Medicaid has
improved the lives of low-income people since its crea-
tion in 1965. Expansions in Medicaid have led to
increases in coverage and access to medical care,
reductions in medical debt, and improvements in
health outcomes with little evidence of significant
reductions in beneficiaries’ labor supply. Yet there are
Medicaid: What concerns around the program’s high and increasing
costs, the quality of care provided, and difficulty
Does It Do, accessing Medicaid-funded care. We discuss some of
the proposed Medicaid reforms meant to address

and Can We Do these problems in light of previous evidence.

Keywords: Medicaid; health insurance; health care;

It Better? medical care; health policy

M edicaid—publicly funded health insurance

for low-income people—is one of the larg-
est U.S. social programs, currently providing
By health insurance coverage to 73 million
Janet Currie Americans,1 at a cost of $565.5 billion in 2016.2
and These costs accounted for 17 percent of total
Valentina Duque national health expenditures and for about 3 per-
cent of the U.S. gross domestic product (GDP).3
The program serves four distinct populations:
lower-income children and their mothers; the

Janet Currie is the Henry Putnam Professor of Economics

and Public Affairs at Princeton University and codirects
Princeton’s Center for Health and Wellbeing and the
NBER Children’s program. She is a member of the
National Academy of Sciences and of the National
Academy of Medicine and a fellow of the AAPSS.
Valentina Duque is an assistant professor of economics
at the University of Sydney, Australia.
NOTE: We would like to thank Robert Moffitt and James
Ziliak for helpful comments, as well as Karen Dynan,
Amitabh Chandra, Nicole Maestas, and participants at
The ANNALS “Entitlement Reform” Conference. Owen
Smith provided excellent research assistance.

Correspondence: [Link]@[Link]
DOI: 10.1177/0002716219874772

148 ANNALS, AAPSS, 686, November 2019

What does medicaid do, and can we do it better? 149

disabled; elderly people in nursing homes; and, since 2014, low-income nondisa-
bled adults in states that have adopted Medicaid expansions that were allowed
under the Affordable Care Act (ACA). This heterogeneity in the populations
served, combined with the fact that each state runs its own Medicaid program
subject to federal government regulations, makes it difficult to draw general con-
clusions about the functioning and impacts of the Medicaid program and the best
options for reforming it.
In this article, we first provide a very brief overview of the program and its
history and then summarize evidence about the effectiveness of Medicaid in
terms of providing insurance and promoting health. There is considerable evi-
dence that Medicaid has improved the lives of low-income people. Yet there is
also dissatisfaction about the high cost of Medicaid, concerns about the availabil-
ity of Medicaid-funded care and the quality of care provided, and lingering con-
cerns about whether eligibility for Medicaid creates perverse incentives that
discourage self-sufficiency in able bodied people who can work. Moreover, many
Americans have incomes too high to qualify for Medicaid but too low to comfort-
ably afford private health insurance, notwithstanding the health insurance
exchanges created by the ACA (which we discuss further). The final part of this
article discusses some proposed Medicaid reforms meant to address these
problems.

Background
Medicaid was adopted in 1965, in an era when poor families had little access to
private insurance or charity care and many people lacked health care. For
instance, Goodman-Bacon (2018) shows that before Medicaid, only 40 to 50
percent of poor children had any doctor visits in a year. Medicaid greatly
increased public spending on health care for the poor, mandating a defined set
of services that were to be covered without requiring the patient to pay.
Figure 1 shows the number of Medicaid beneficiaries each year from 1975 to
2018, by eligibility category. Children are the largest single group of beneficiar-
ies. While enrollment was initially flat, the growth in the coverage of children
since the late 1980s was intentional. States were first allowed and then required
to raise income eligibility limits for pregnant women and children. These expan-
sions broke the link between welfare receipt and Medicaid coverage and allowed
children in families with higher incomes to be covered. By 2010, 48 percent of
all births in the United States were covered by Medicaid (Markus et al. 2013).
The introduction of the state Child Health Insurance Program (CHIP or
SCHIP) in 1998 further expanded coverage of low-income children.4 Between
1986 and 2005, the share of children eligible for Medicaid/CHIP increased from
between 15 and 20 percent of U.S. children, to between 40 and 50 percent of
children, depending on the age group (Currie, Decker, and Lin 2008).
Coverage of low-income, nondisabled adults and the disabled also increased
over time, especially beginning in the late 1990s, when some states raised income
150 THE ANNALS OF THE AMERICAN ACADEMY

Figure 1
Medicaid (and CHIP) Beneficiaries, By Eligibility Group, FY1975–2018 (thousands)

SOURCE: We used several sources of data to construct this figure: (1) the “MACStats:
Medicaid and CHIP Data Book” provided information on Medicaid enrollment by group from
1975 to 2013 ([Link]
[Link]), (2) the “CMS – Brief Summaries of Medicare & Medicaid” provided informa-
tion from 2014 to 2016 ([Link]
Statistics-Trends-and-Reports/MedicareProgramRatesStats/SummaryMedicareMedicaid.
html), and (3) the “CMS – Fast Facts” provided statistics in 2017 and 2018 (link: [Link]
[Link]/fastfacts/). We used the “CMS – Statistics Reference Booklet” that provided CHIP
enrollment rates from 1998 to 2016 ([Link]
Systems/Statistics-Trends-and-Reports/CMS-Statistics-Reference-Booklet/[Link]) and the
“CMS – Fast Facts” for CHIP enrollment in 2017 ([Link] CHIP
enrollment for 2018 was not available in the CMS Fast Facts.
NOTE: The numbers represent “person-year equivalents” or total individuals ever enrolled in
Medicaid/CHIP in a fiscal year. CHIP is the Children’s Health Insurance Program that was
first implemented in 1998. For Medicaid and CHIP, data are reported by individual states and
are representative of the unduplicated number of children ever enrolled in Medicaid and
CHIP. The dotted line represents a change in the source of data.

cutoffs for these eligibility groups, and again after the passage of the ACA in
2014.
Coverage of the elderly has been remarkably constant, in part because most
medical expenditures for the aged are covered under a separate federal program,
Medicare. Medicaid primarily covers low-income elderly people in nursing
homes, who have spent down their assets so that they are poor enough to qualify.
Medicaid now covers six out of ten nursing home residents.5
What does medicaid do, and can we do it better? 151

Figure 2
Real Spending Per Beneficiary by Eligibility Group, FY1975–2016

SOURCE: We used two data sources to construct this figure: (1) the CMS Tables (2012) from
1975 to 2011 ([Link]
and-Reports/MedicareMedicaidStatSupp/Downloads/2013_Section13.pdf#table13.11) and (2)
the “CMS – Brief Summaries of Medicare & Medicaid” from 2012 to 2016 ([Link]
g o v / R e s e a r c h - S t a t i s t i c s - D a t a - a n d - S y s t e m s / S t a t i s t i c s - Tr e n d s - a n d - R e p o r t s /
MedicareProgramRatesStats/[Link]).
NOTE: Amounts are in 2010 dollars. The dotted line represents a change in the data source.

Figure 1, which shows the number of Medicaid-eligible individuals, should be

contrasted with Figure 2, which shows spending per beneficiary by eligibility
category. Although the majority of recipients are children, annual spending per
child is low, at approximately $3,000 (2010 U.S. dollars). Low-income, nondisa-
bled adults are also relatively inexpensive to cover. Figure 2 shows clearly that the
disabled and the elderly are the high-cost groups, with costs more than four times
the average per capita cost for a child or nondisabled adult. Thus, although the
majority of recipients are children, the elderly and disabled account for about 60
percent of Medicaid’s costs.
Since Medicaid is run at the state level under federal guidelines, these aver-
ages mask considerably the variability in the generosity of the programs and in
the eligibility rules across states. Income cutoffs for the elderly and disabled
range from 74 to 100 percent of the federal poverty rate, while cutoffs for preg-
nant women varied from 145 to 375 percent of poverty as of April 1, 2019.6 These
differences in generosity across states mean that similar families will be treated
quite differently in terms of health care coverage, depending on where they live.
In constructing these figures, we were forced to rely on different sources of
data before and after 2012.7 The Center for Medicare and Medicaid Services
152 THE ANNALS OF THE AMERICAN ACADEMY

Figure 3
Real Federal and State Medicaid Expenditures (Millions), Fiscal Years 1966 to 2017

SOURCE: We used two sources of data to construct this figure: (1) the CMS, Office of the
Actuary, National Health Statistics Group: “National Health Expenditures by Type of
Service and Source of Funds: Calendar Years 1960 to 2016” from 1966 to 2016 (https://
[Link]/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/
NationalHealthExpendData/[Link]) and (2) the
“MACStats: Medicaid and CHIP Data Book” for 2017 ([Link]
tion/macstats-medicaid-and-chip-data-book-2/).
NOTE: Amounts are in 2010 dollars.

(CMS) is working on a new database (Transformed Medicaid Statistical

Information System, or TMSIS), which has caused delays in the public release of
information on enrollments and per capita spending levels. These data series
changes mean that it may not be possible to determine the effects of the ACA
(whose main provisions came into effect in 2014) on Medicaid with absolute
certainty.8
Figure 3 shows trends in state and federal expenditures on Medicaid and
CHIP. State governments run Medicaid and receive a federal “match” of state
expenditures. These matching funds now account for over half of the program’s
total costs. Figure 3 shows the relief that the federal government provided to
states during the Great Recession (2007–2009), when economic hardship caused
Medicaid coverage to increase at a time when state budgets were already under
severe strain from a downturn in tax receipts.9 Without this increase, states would
have had to implement painful cuts in Medicaid or other important areas such as
education.
Figure 4 shows that despite increases in federal aid, Medicaid has consumed
a growing share of state budgets over time. Much of the increase is because infla-
tion in health care costs has generally outstripped overall inflation. Despite the
What does medicaid do, and can we do it better? 153

Figure 4
Medicaid as a Share of State Budgets, Including and Excluding Federal Funds,
FY1990–2016

SOURCE: Medicaid and CHIP Payment and Access Commission. These data are taken from
the following source: [Link]
ets-including-and-excluding-federal-funds/.

growth in costs, however, total Medicaid spending in 2019 is expected to be 20 to

30 percent lower than the actuaries’ initial estimates, suggesting that Medicaid
provides health care in a relatively efficient way (Glied and Tavenner 2019).

The Effects of Medicaid on Health, Well-Being,

and Behavior
Whether Medicaid affects people’s health is a first-order question. But to affect
health, Medicaid must impact other aspects of people’s lives, such as health insur-
ance coverage, quantity and/or quality of care received, or financial well-being.
Medicaid may also affect people’s economic behavior: most obviously, healthy
people are more likely to be able to work and will likely be able to work to older
ages than unhealthy people. To the extent that Medicaid makes people better off
financially, they may also choose to consume some of this “wealth” by working
fewer hours. In the United States, most privately insured people obtain health
insurance through their employers. Consequently, Medicaid could allow some
people with employer health insurance to drop out of the labor force or reduce
154 THE ANNALS OF THE AMERICAN ACADEMY

their work hours. Medicaid’s sharp income cutoffs might discourage people from
earning more than these cutoffs if increasing their earnings means that they
would lose their Medicaid coverage.
Medicaid program rules can also impact the behavior of health care providers.
There are many reports of providers limiting the number of Medicaid patients in
their practice (or refusing to treat Medicaid patients) because Medicaid typically
reimburses providers less than other insurances. Attempts to correct this prob-
lem by giving hospitals that treat a “disproportionate share” of Medicaid patients
additional lump sum payments can create their own perverse incentives. For
example, they can encourage providers to “cream skim” relatively healthy
Medicaid patients. The literature on provider responses is ably summarized in
Bitler and Zavodny (2017) and Buchmueller, Ham, and Shore-Sheppard (2016),
so we will not further delve into it here. By and large, providers respond in the
expected way to incentives created by low Medicaid reimbursement. Hence, dif-
ferential treatment of Medicaid patients by providers would likely be reduced if
Medicaid reimbursed providers at market rates, an idea that we discuss more
here.
In what follows, we give a brief overview of some of the main findings from
decades of Medicaid research, followed by a section on new evidence on effects
since the implementation of the ACA. Given the sheer volume of work, we have
not attempted to be comprehensive but rather to highlight important points.
Research on Medicaid tends to emphasize different outcomes for different eligi-
bility groups, in part because there are differences in the types of services that
are covered. Hence, we have opted to have separate discussions of low-income
women and children; elderly people in nursing homes; low-income, nondisabled
adults; and the disabled.

Low-income women and children

Pregnant women and children are the majority of Medicaid recipients. The
staggered implementation of the Medicaid expansions of the late 1980s and
1990s provided an opportunity to examine the effect of eligibility for public
health insurance in these groups. Some studies focus on pregnant women and
infant health outcomes, while other studies examine the combined effects of
coverage in the in utero period and early childhood on later child outcomes.
Effects of the expansions for pregnant women on infant health outcomes were
significant and immediate. For example, Currie and Gruber (1996a) show that
the 30 percent increase in the eligibility of pregnant women that took place was
associated with an 8.5 percent decline in the infant mortality rate. The Medicaid
expansions to pregnant women had little effect on fertility (Zavodny and Bitler
2010; DeLeire, Lopoo, and Simon 2011). However, expansions of Medicaid
funding for family planning services increased contraceptive use and reduced
births (Kearney and Levine 2009; Zavodny and Bitler 2010).
The roughly 15 percent increase in Medicaid eligibility for children that
occurred over the same period reduced the probability that a child went without
any doctor visits during the year by 9.6 percent and reduced preventable
What does medicaid do, and can we do it better? 155

hospitalizations (Currie and Gruber 1996b; Kaestner, Joyce, and Racine 2001;
Aizer, Currie, and Moretti 2007; Dafny and Gruber 2005; Howell and Kenney
2012). These results suggest that as many as 6 million children received increased
access to basic preventive care. Being eligible for Medicaid conferred gains in
access to care even without formal Medicaid coverage, as providers became
aware that most low-income children were eligible and that they could be reim-
bursed for services rendered ex post.
These expansions in public health insurance for pregnant women and children
dramatically reduced mortality among poor children and especially among poor
African American children. The result is that inequality in mortality has been
falling among children since 1990, even while it has increased for adults (Currie
and Schwandt 2016). In fact, the child mortality rate in the United States con-
verged toward the lower Canadian rate between 1990 and 2010 (M. Baker,
Currie, and Schwandt 2017).
Goodman-Bacon (2018) uses historical data about the rollout of Medicaid
across states in the late 1960s and finds that mortality fell more rapidly among
infants and children in states with bigger Medicaid expansions. Among nonwhite
children, who were poorer on average and therefore most likely to be affected,
mortality fell by 20 percent. Thus, the historical results mesh well with those
from the more recent Medicaid expansions.
Many recent studies look at the long-term effects of Medicaid coverage in
early childhood on adult outcomes. Goodman-Bacon (2016) finds that the rollout
of Medicaid in the 1960s reduced adult disability and increased labor supply
measured up to 50 years later, while Boudreaux, Golberstein, and McAlpine
(2016) find that access to Medicaid in early childhood reduces high blood pres-
sure, diabetes, heart disease, and obesity.
Several recent papers look at the long-term effects of the expansions of mater-
nal and child Medicaid eligibility in the late 1980s and early 1990s (D. Brown,
Kowalski and Lurie 2015; Cohodes et al. 2016; Miller and Wherry 2019; Wherry
and Meyer 2016; Wherry et al. 2018). These studies follow cohorts who received
Medicaid coverage in utero or early childhood over time. Levine and Schanzenbach
(2009) find long-run effects of Medicaid on child educational attainment and
achievement test scores, while Currie, Decker, and Lin (2008) find that they are
in better health in adolescence. As young adults, covered cohorts are more likely
to work, have higher earnings, have more education, and are in better health
(using self-reported health, mortality, and hospitalization rates as outcomes) than
cohorts who were not covered by the Medicaid/CHIP expansions. East et al.
(2017) find that the effects of the Medicaid expansion are being felt intergenera-
tionally: women who were covered by Medicaid as infants because of the expan-
sions in the late 1980s and early 1990s are giving birth to healthier children today.
Finally, Buchmueller, Ham, and Shore-Sheppard (2016) provide an in-depth
overview of an earlier body of literature examining the effects of Medicaid expan-
sions in the 1980s and 1990s on the work effort of mothers with young children
and of the disabled. They conclude that there is little evidence of any effect.
In summary, there is overwhelming evidence that expansions of Medicaid
coverage to low-income pregnant women and children have had dramatic
156 THE ANNALS OF THE AMERICAN ACADEMY

positive effects on their health and well-being and have improved the quality of
the labor force as the children enter adulthood. Recent cuts to Medicaid cover-
age of children threaten these gains. For example, recent changes to the rules
governing when immigrants are deemed a “public charge” threaten health care
coverage for children who are U.S. citizens but have undocumented family
members.10

Elderly people in nursing homes

About a fifth of Medicaid expenditures are for the elderly—primarily through
spending on long-term care services such as nursing homes.11 Eligibility require-
ments require elderly people to “spend down” their assets and have low enough
incomes to qualify for Medicaid coverage of their nursing home care.12 In most
states with an income cap, individuals may not have monthly incomes higher than
$2,205. In states that use “medically needy” criterion for eligibility, an individual
must spend all her or his income on medical expenses except for a small personal
allowance of $50 to $100 per month to qualify. Single individuals applying for
Medicaid’s Long-Term Care coverage are allowed to have a maximum of $2,000
in “countable” assets. Income and asset cutoffs are slightly higher for married
couples.
Research into the role of Medicaid in the nursing home market has focused
on four main areas: (1) whether increases in Medicaid generosity affect nursing
home utilization; (2) whether there are shortages of nursing home beds for
Medicaid patients; (3) whether the quality of nursing home care for Medicaid
recipients is worse than the care received by private-pay elderly; and (4) the
relationship between Medicaid reimbursement levels, service quality, and patient
health.
The demand for nursing home care seems to be unresponsive to Medicaid’s
generosity, which suggests that people who use nursing homes have few viable
alternatives and an urgent need for care (Grabowski and Gruber 2007). In the
past, Medicaid patients have faced shortages of nursing home beds perhaps
because Medicaid’s payment rate for nursing home care tends to be significantly
below the private market rate or the Medicare rate (although Medicare does not
pay for long-term care, it does pay for short-term rehabilitative stay) (Reschovsky
1996). However, nursing home occupancy rates have declined nationwide from
93 percent in 1977 to 82.3 percent in 2014 (Laes-Kushner 2018). States such as
Oregon are struggling with occupancy rates of only 52 percent. As the cost of
maintaining empty beds tends to be reflected in overall nursing home costs,
states have experimented with policies such as “bed buybacks” to encourage pro-
viders to voluntarily retire licensed beds and increase occupancy rates. However,
the oldest baby boomers are only in their 70s, and as this large cohort enters
extreme old age, demand for nursing homes may rise.
Once Medicaid patients receive a bed, Medicaid-funded residents in nursing
homes receive care that is of similar quality compared to private-pay patients
(Grabowski 2001, 2004; Grabowski, Gruber, and Angelelli 2008). Moreover,
reductions in Medicaid reimbursements to nursing homes affect service quality
What does medicaid do, and can we do it better? 157

and patients’ health for all patients. Mor et al. (2011) show that a cut in the
Medicaid reimbursement rate of around 10 percent led to a functional decline of
almost 10 percent in nursing home residents (that is, a decline in their ability to
walk or use the restroom by themselves). Reductions in Medicaid reimbursement
rates raise the odds that patients will be in persistent pain and that they will have
bedsores (Grabowski, Gruber, and Mor 2017; Grabowski and Angelelli 2004).
Looking after nursing home patients is a labor-intensive job, and so having
enough staff and having high-quality staff are critical to providing good care.
Hence, lower Medicaid reimbursement rates can affect patient’s health if nursing
homes react by reducing staff. Stevens et al. (2015) suggest that one reason that
recessions appear to improve the health of older people is that by lowering labor
costs, recessions allow for improved levels of staffing in nursing homes.
Policy discussions about nursing homes have been dominated by the issue of
escalating costs. Some estimate that the average cost of nursing home care is
around $125,000 per year (The Association for Long Term Care Planning
2019).13 State governments pay approximately half of Medicaid costs, and in
many states, the rising cost of nursing home care threatens to crowd out other
necessary spending, such as spending on education. At the same time, when
there are shortages of nursing home beds, or when the quality of nursing home
care is poor, it can result in people being hospitalized, which is even more
expensive.
One-quarter of nursing home residents are hospitalized each year, and the
daily cost of caring for them more than quadruples when they move to the hos-
pital. Research shows that a reduction in nursing home reimbursements of
around 10 percent leads to a 5 percent rise in the odds that residents will be
hospitalized (Intrator et al. 2007). However, to the extent that care in hospitals is
paid for by Medicare, which does not require state cost-sharing, it may be in the
fiscal interest of state governments to allow people to be pushed into hospitals
rather than to have them stay in nursing homes.
The main approaches that states have employed to lower nursing home costs
are reducing Medicaid payment rates and directly limiting the number of nursing
home beds via certificate-of-need laws and construction moratoria. These laws
are effective in limiting nursing home construction and state Medicaid expendi-
tures (Grabowski, Ohsfeldt, and Morrisey 2003) but reduce nursing home com-
petition and lower both quality of care and access to services for Medicaid
residents (Grabowski, Angelelli, and Mor 2004; Grabowski et al. 2004). Lowering
costs by denying nursing home care to needy people does not seem like good
public policy. Not surprisingly, the laws do little to control hospital costs (Lanning,
Morrisey, and Ohsfeldt 1991; Conover and Sloan 1998; Salkever 2000).
Capitated payments are another policy measure that has been implemented to
contain costs. In managed care plans, providers typically receive a fixed amount
for each enrolled individual. In principle, this system should encourage lower
utilization of nursing home services and increase the use of cheaper home- or
community-based services. However, health care quality may also decline as pro-
viders are no longer compensated for providing additional services (Grabowski
2004).
158 THE ANNALS OF THE AMERICAN ACADEMY

Another type of policy aimed at reducing the public cost of nursing home care
involves encouraging people to purchase private long-term care (LTC) insurance
policies. Currently, only 10 to 12 percent of the U.S. population has LTC insur-
ance (Johnson 2016),14 a fraction that is declining as the cost of LTC insurance
has risen. Goda (2011) evaluates a tax-incentive policy that aimed to encourage
purchases of LTC plans and shows that it did increase private LTC insurance
coverage (by 2.8 percentage points) but only among high-income or wealthy
individuals who were already more likely to have the option to purchase LTC
policies from their employers (and were less likely to be covered by Medicaid).

Low-income, nondisabled adults

Until quite recently, low-income adults who were not mothers of young chil-
dren and who were not disabled were seldom eligible for Medicaid. Hence, it is
not surprising that there has been less research on the impact of Medicaid on
low-income, nondisabled adults than on other groups. Most research on this
subject prior to the ACA focuses on one-off state-level expansions of Medicaid,
especially the expansion in Oregon, which is discussed further below. Research
on this group has focused on (1) insurance coverage; (2) financial security; (3)
health care use; (4) health outcomes; and (5) measures of work effort including
participation, employment, and job mobility.
Perhaps the least controversial finding from this literature is that Medicaid
expansions increase health insurance coverage. There is a large body of literature
examining crowd out of private health insurance by public health insurance,
beginning with Cutler and Gruber (1996). See Buchmueller, Ham, and Shore-
Sheppard (2016) and Bitler and Zavodny (2017) for excellent recent summaries
of this work.
Medicaid coverage also improves financial security and reduces medical debt.
The evidence from both experimental evaluations and natural experiments sug-
gests that health insurance reduces out-of-pocket medical spending and, in par-
ticular, reduces catastrophic out-of-pocket expenditures that might impose huge
consumption costs on individuals (Clemans-Cope et al. 2013; Hu et al. 2016;
Mazumder and Miller 2016; Michalopolous et al. 2011). Gross and Notowidigdo
(2011) examine the effect of the Medicaid expansions between 1992 and 2004 on
bankruptcies and find that a 10 percent increase in eligibility for Medicaid
reduces bankruptcy by 8 percent. Since catastrophic medical bills can push peo-
ple into poverty, these results suggest that Medicaid provides a safety net that
may reduce poverty.15
The evidence also suggests that Medicaid improves access to health services
and increases the consumption of health care among low-income able-bodied
adults. Sommers et al. (2017) find a 41-percentage-point increase in the probabil-
ity of having a regular source of care among uninsured people gaining coverage.
Measures such as having a usual source of care, receiving an annual checkup, and
self-reported ability to pay for needed treatment also increased (Sommers et al.
2016, 2017; Miller and Wherry 2017; Courtemanche et al. 2018), while the
What does medicaid do, and can we do it better? 159

probability of delaying care because of cost fell (Sommers, Baicker, and Epstein
2012).
The Oregon Health Insurance Experiment was a social experiment designed
to examine the effects of providing Medicaid coverage to low-income, uninsured
adults. The state received federal permission to expand Medicaid coverage of this
group but did not have funds to cover all low-income adults. Therefore, adults
were randomly chosen from those who applied. The results regarding effects on
insurance coverage, utilization of care, and financial security are consistent with
those from the nonexperimental studies discussed here. There was no crowd out
of private health insurance coverage. Medicaid coverage increased health care
use across all dimensions including outpatient care, preventive care, prescription
drugs, hospital admissions, and emergency room visits. For instance, the average
person in the control group visited an emergency room 1.02 times, but people in
the treatment group increased ER use by 0.41 visits per person, or 40 percent
(Finkelstein et al. 2012; Baicker et al. 2013; Taubman et al. 2014; Baicker et al.
2014; Finkelstein et al. 2016). Receiving Medicaid also reduced the risk of cata-
strophic out-of-pocket medical expenditures, defined as expenditures of 30 per-
cent of household income or greater (Finkelstein, Mahoney, and Notowidigdo
2018).
Findings regarding the health effects of Medicaid coverage are more hetero-
geneous. Sommers, Baicker, and Epstein (2012) employ a difference-in-­
difference strategy comparing changes in county-level mortality rates in New
York, Maine, and Arizona—three states that expanded their Medicaid programs
in the early 2000s—to those in neighboring states that did not implement such
expansions, before and after the policy change. They find that Medicaid expan-
sions were associated with a significant 6.1 percent reduction in mortality, con-
centrated among adults 35 to 64, nonwhites, and residents of poorer counties.
These expansions also improved self-reports of “excellent” or “very good” health
(by 3.4 percent).
The Oregon Health Insurance Experiment finds improvements in self-
reported health and especially in mental health but does not find any significant
effects on physical health measures (such as blood pressure) or on mortality,
although an important caveat to both the Oregon study and the earlier one is that
one might not expect to see a large mortality effect within a few years of a change
in health insurance status. It is possible that the differing findings across studies
reflect heterogeneity across groups in addition to statistical uncertainty. Sommers
et al. (2015) find that the largest improvements in coverage and in access to
medical care occurred among racial and ethnic minorities, a finding that echoes
the result that the earlier Medicaid expansions to children had their largest
effects on minority children. Sommers, Baicker, and Epstein (2012) show that
the declines in mortality were particularly concentrated among older adults, non-
whites, and residents of poorer counties. It may be that the health impacts are
also concentrated in these groups.
The most controversial issue associated with extending Medicaid to low-
income adults who are not disabled concerns possible effects on the incentive to
work. Income cutoffs for the program may cause people to lower their incomes
160 THE ANNALS OF THE AMERICAN ACADEMY

by reducing work hours to maintain their Medicaid coverage. In addition, if pub-

lic health insurance is available, it may remove an incentive to try to find a job
that offers private health insurance.
The Oregon Health Insurance Experiment finds no statistically significant
effect of Medicaid coverage on employment or earnings (Baicker et al. 2013).
Dague, DeLeire, and Leininger (2017) examine a policy change in Wisconsin,
which had expanded Medicaid coverage of low-income adults but then placed a
sudden cap on new enrollments in the Medicaid program in October 2009. Using
state administrative data from 2005 to 2011, they compare people who enrolled
before the cap to those who had applied but were waitlisted after the announce-
ment of the cap. They find that Medicaid coverage reduced employment in the
covered group by about 12 percent relative to the waitlisted group.
At the other end of the spectrum, Garthwaite, Gross, and Notowidigdo (2014)
compare childless adults to other adults in Tennessee and in other southern
states before and after a Medicaid policy change in Tennessee. Beginning in July
2005, Tennessee disenrolled around 170,000 people from Medicaid, and the
authors find that this disenrollment led to a 60 percent increase in the labor sup-
ply of affected adults. These large estimates have been critiqued, largely on the
basis that the dataset they use, the March Current Population Survey, yields a
relatively small sample with an imprecise measure of Medicaid status. A study of
the effect of this same policy change using administrative data does not find sta-
tistically significant effects (Gouskova 2015). Similarly, Ham and Ueda (forth-
coming) find little effect of the policy change on labor supply using alternative
data sources. In any case, the Tennessee estimates, published as they were on the
eve of implementation of the ACA, have proven influential in part because they
suggested that the new law might have a large effect on employment. But it is
important to note that these results are an outlier relative to the other studies of
this issue.

The disabled
Although research on Medicaid has mostly focused on low-income children
and their mothers and on the institutionalized elderly, Medicaid spends more on
disabled adults than on any other group. In 2011, although the disabled only
represented 12 percent of the total Medicaid population, they accounted for 36
percent of the cost (Musumeci and Young 2017). Their cost per capita is substan-
tially higher than for other groups, given their often complex health care needs,
the persistence of their health conditions, and their often low employment rates.
Studies about Medicaid and the disabled focus on four key areas: (1) health
care coverage, (2) the quality of care, (3) interactions with other programs, and
(4) effects on labor-market outcomes. Surprisingly, there are few studies about
the effects of Medicaid on the health of the disabled, which may be due to data
limitations. The National Health Interview Survey has had questions about limi-
tations on activities of daily living since the 1990s, but it has little information
about health per se. Questions about activity limitations were added to the
What does medicaid do, and can we do it better? 161

American Community Survey and the Current Population Surveys in 2008, but
again, these datasets have little information about underlying health status.
States are required to give Medicaid coverage to disabled people who qualify
for federal Supplemental Security Income (SSI). SSI is available to those who
have low income and assets and a disability that impairs their ability to work. The
SSI federal benefit rate is generally less than the federal poverty level. Disabled
nonelderly people who have a substantial work history may be eligible for Social
Security Disability Insurance (SSDI), which is a completely different and sepa-
rate program with benefit levels geared to a person’s earnings history. Those who
become eligible for SSDI can receive Medicare coverage after two years, reflect-
ing a more generous package than Medicaid.
When the federal SSI program replaced state-only programs for the disabled
in 1972, states were given the option to retain some of their pre-1972 rules for
Medicaid eligibility in place of the federal rules. Currently, eleven states have
retained such rules, which typically make their programs less generous than the
federal program.16
Many disabled people are unable to access Medicaid via the SSI program, which
is not surprising given relatively onerous requirements for proving eligibility.
Benitez-Silva, Buchinsky, and Rust (2004) estimate that up to 60 percent of those
whose claims are denied are actually disabled and that the adjudication of claims can
take years. Hence, general increases in the generosity of public health insurance
coverage can also increase coverage of the disabled. For example, the Massachusetts
health care reform reduced the probability that working age people with disabilities
lacked health insurance by about 50 percent and also reduced the probability of
going without care due to cost (Gettens, Mitra, and Himmelstein 2011).
Research has also focused on how Medicaid coverage affects labor force par-
ticipation and employment among the disabled. There is a large branch of litera-
ture suggesting that increasing the generosity of disability programs (which
generally include public health insurance coverage) can discourage people from
working (see, for example, Autor et al. 2016; Deshpande 2016). Here, we con-
sider the narrower question of how changes in Medicaid eligibility of the disabled
can affect labor market outcomes and participation in disability insurance
programs.
It is difficult and time-consuming to qualify for SSI, and SSI recipients must
maintain a low income to remain qualified. People on SSI may be discouraged
from working because if their incomes exceed the cutoff, they lose SSI eligibility
and, consequently, health insurance coverage. Baicker et al. (2014) show, how-
ever, that the Oregon Health Experiment had no effect on enrollees’ take-up of
disability insurance. Similarly, Maestas, Muellen, and Strand (2014) find no evi-
dence that the Massachusetts health care reform led to a change in the number
of SSI applications. Hence, there is little evidence so far that disabled adults are
changing their work effort to qualify for health insurance.
In addition to the possible interactions between Medicaid generosity and SSI,
there may be trade-offs between what the government spends on the disabled
under Medicaid and what it spends under Medicare. Autor, Chandra, and
Duggan (2011) show that there is substantial variation in spending under these
162 THE ANNALS OF THE AMERICAN ACADEMY

two programs across geographic regions and that areas with high Medicaid
spending tend to have lower Medicare spending on the disabled. The variation in
Medicaid spending on the disabled tends to be driven by variation in the intensity
of care. These findings suggest that some of the cost of increasing the level of
Medicaid expenditures on the working-aged disabled today may be offset by
lower Medicare costs on the disabled elderly tomorrow. Given the high cost of
this part of the program, further investigation of the reasons for variations in
spending on the disabled under Medicaid and possible trade-offs with expendi-
tures on Medicare is warranted.

Medicaid and the ACA

The ACA is the most important health care reform legislation enacted in the
United States since the creation of Medicare and Medicaid in 1965. The primary
aim of the ACA was to reduce the number of Americans without health insurance
and to improve the accessibility, affordability, and quality of health care for con-
sumers (Obama 2016). For example, the ACA required most insurance plans to
cover ten classes of benefits, including treatment of mental illness, which histori-
cally had been covered less generously than other health care services (Garfield,
Lave, and Donohue 2010).
The ACA contained few provisions that were expected to directly address
health care prices and, therefore, had little hope of reducing costs (Gruber 2008).
For instance, the ACA tried to address the high cost of long-term care by creating
a voluntary system of LTC insurance (i.e., the CLASS Act), but due to high cost,
this system never actually got off the ground (James, Gellad, and Hughes 2017).
As enacted in March 2010, the ACA created subsidized marketplaces
(exchanges) where uninsured individuals with incomes less than 400 percent of
the federal poverty line could purchase private health insurance. The ACA ini-
tially required all states to extend eligibility for Medicaid to adults in families with
incomes less than 138 percent of the federal poverty line, as these people were
viewed as too poor to be able to purchase health insurance even in the subsidized
exchanges. The ACA also changed the rules regarding the determination of gross
income and required every state to increase income eligibility for Medicaid by 5
percentage points above the March 2010 level, effective January 1, 2014.
But in 2012, the Supreme Court ruled that requiring states to expand their
Medicaid programs was unconstitutional. As of December 2018, thirty-six states
and the District of Columbia had chosen to expand Medicaid, with the remainder
of the states refusing to do so, even with the offer of large federal subsidies to
help pay for newly eligible individuals. However, Frean, Gruber, and Sommers
(2017) find that even in states that did not increase their income limits, more
individuals who were eligible under existing criteria enrolled in Medicaid, a phe-
nomenon they dub a “welcome mat effect.” Still, states that have not expanded
Medicaid have ended up with a large number of people with incomes too high to
qualify for Medicaid but too low to benefit from the exchanges.
What does medicaid do, and can we do it better? 163

There is no doubt that the ACA was successful in increasing the fraction of
Americans with health insurance (Gruber and Sommers 2019; Frean, Gruber,
and Sommers 2017; Courtemanche et al. 2017; Kaestner et al. 2017; Wherry and
Meyer 2016). For example, Courtemanche et al. (2017) show that the ACA
increased the proportion of residents with insurance by 5.9 percentage points
compared to 3.0 percentage points in states that did not expand Medicaid.

Effects of the ACA on health and health care

These increases have had positive effects on the use of health care and on
some measures of health. Sommers et al. (2017) use survey data collected
through the end of 2016 from Kentucky and Arkansas, which both expanded
coverage, compared to Texas, which did not. They find a 23-percentage-point
increase in “excellent” self-reported health in Kentucky and Arkansas relative to
Texas. Other studies have shown positive effects of Medicaid expansions follow-
ing from the ACA on health measures such as better self-assessed health, lower
incidence of depression, and smoking cessation (Simon, Soni, and Cawley 2017;
Winkelman and Chang 2018; Koma et al. 2017).
Ghosh, Simon, and Sommers (2017) show that within the first 15 months of
the ACA expansion, Medicaid-paid prescriptions increased by 19 percent, which
represented approximately seven additional prescriptions per year per newly
enrolled beneficiary.17 MacLean et al. (2017) show that Medicaid expansions
between 2011 and 2017 increased the use of psychotropic prescriptions by 22.3
percent relative to nonexpansion states. These effects were experienced across
most major mental illness categories and across states with different levels of
patient need, system capacity, and scope for expansion. The study, however, finds
no evidence that Medicaid expansions reduced suicide, a proxy for the most seri-
ous mental illnesses.
A recent study by Black et al. (2018) also finds no evidence that ACA expan-
sions have changed mortality for nonelderly adults. They comment, however, that
their estimated confidence intervals are large and that even the national natural
experiment of the ACA is underpowered to find effects on mortality of the size
that one might expect given past research on the relationship between medical
care and shorter-term mortality changes in nondisabled adults.
One caveat to the generally improved outlook for the health and health care
of patients receiving Medicaid because of the ACA is that the sudden influx of
newly insured people into the system increased waiting times for appointments,
reflecting a shortage of primary care physicians available to treat Medicaid
patients (Miller and Wherry 2017).

Effects of the ACA on work

Because the ACA Medicaid expansions primarily cover able-bodied working
age adults, much of the attention has been focused on changes in their work
effort. It is worth considering that any work reductions experienced by childless
164 THE ANNALS OF THE AMERICAN ACADEMY

adults who suddenly were covered, would imply that people were working only
because they needed health insurance. It is questionable whether as a social
policy, one wants to force people to work to get health care, if they have health
problems and have difficulty working.
In any case, Gooptu et al. (2016) directly examine the effects of the ACA on
labor market outcomes among adults with incomes below 138 percent of the
federal poverty level. They compare states that expanded Medicaid and those
that did not, looking at groups who were eligible and groups who were not eligi-
ble for Medicaid before and after the ACA (2005 to 2013 versus 2014 to 2015),
and find no statistically significant effect of Medicaid coverage on employment.
Similarly, Leung and Mas (2018) find no effect of the ACA on employment.
Overall then, with the sole exception of the Garthwaite, Gross, and Notowidigdo’s
(2014) pre-ACA study where there may be some issues of data quality, the litera-
ture about the labor supply effects of Medicaid on nondisabled adults suggests
that these effects are modest.
Turning to disabled adults, a small-scale survey using difference-in-difference
methods finds that adults with disabilities living in states that expanded eligibility
for Medicaid under the ACA were significantly more likely to be employed than
those in nonexpansion states (38.0 percent versus 31.9 percent) (J. Hall et al.
2017).
Several studies have examined the impact of the ACA on young adults, a group
with historically low rates of health insurance coverage, on education and
employment. The ACA allowed parents to cover children up to age 26 under
their parents’ plans. However, Slusky (2017), using standard data sources such as
the Current Population Surveys, shows that difference-in-differences estimates
for this group are unreliable due to small sample sizes.

Possible Reforms to the Medicaid Program

In this section, we first explore some of the most persistent complaints about the
Medicaid program and then consider reform options that have been suggested.

Cost and access to care: The main complaints about Medicaid

One common complaint is that even in states that expanded Medicaid after
the ACA, there are still people who lack health insurance coverage because they
cannot afford even the subsidized insurance available on the exchanges. In states
that did not expand Medicaid, there are a large number of people whose incomes
are too high for Medicaid but too low to qualify them to participate in the federal
or state health insurance exchanges. This is not a complaint about Medicaid per
se, but about the way that free or subsidized public insurance is means-tested.
One of the most persistent complaints about Medicaid itself is its high and
rising cost to government. Rising health care costs are likely to put a strain on
both state and federal budgets for the foreseeable future. When considering
What does medicaid do, and can we do it better? 165

costs, it is important to consider that both the quantity of health care demanded
and its price affect total cost. As a society, we have deliberately chosen to expand
Medicaid coverage to millions of previously uninsured low-income pregnant
women, children, disabled people, and now to nondisabled adults. We should not
be surprised that people who previously had no health insurance consume more
care when care becomes free for them. That is the intention of the policy.
Compared to most European countries, Americans pay more for specific
health services as well as for prescription drugs (Kamal and Cox 2018). That is,
high prices are the main aspect of health care costs that have not been implicitly
chosen through public policy. Consistent with the higher prices, Americans typi-
cally consume less health care (e.g., fewer doctor visits and fewer hospital days)
than residents of other wealthy countries, while paying twice as much per capita
for our care. Rising drug prices are a particular problem for state Medicaid pro-
grams. For some diseases, like severe mental illnesses or cystic fibrosis, most
patients are considered to be disabled and so are covered by Medicaid. Drugs for
these conditions are expensive and getting more so over time. Cooper et al.
(2019) show that there are vast and unexplained differences in the prices of simi-
lar services across U.S. hospitals, suggesting that there is little underlying market
rationale for the high prices we pay. Instead, high prices may reflect a noncom-
petitive and highly concentrated market for health care (National Academy of
Social Insurance 2015).
This discussion suggests that any policy that reduced the cost of Medicaid
without greatly increasing the number of uninsured would have to tackle the
systemic problem of high U.S. health care prices. So far, there has been little
political will to do this. The leading options for controlling prices include meas-
ures to encourage price transparency, vigorously enforcing existing antitrust laws,
and allowing government agencies to negotiate with providers over prices. To be
clear, state governments do set the prices paid for services under Medicaid. But
in the absence of any negotiation about prices in Medicare, this simply creates an
incentive for providers to prefer Medicare to Medicaid patients.
A second common complaint about the Medicaid program involves difficulty
in accessing care. For example, Alexander and Currie (2017) show that children
with asthma (the most common reason for childhood hospitalization) were less
likely to be hospitalized after presenting at the emergency room if they had pub-
lic health insurance (though this had no measurable effect on their health out-
comes). As discussed above, some providers do not participate in Medicaid
because of low reimbursements. Just as eligibility varies a great deal across states,
so does provider reimbursement, so access issues are more acute in some states
than in others. Access is a particular problem for specialist care. In addition to
low reimbursements, providers cite administrative difficulties with reimburse-
ment as a reason for limiting their participation in the program.

Medicaid managed care: The dominant reform idea to date

One attempted solution to both the high cost of Medicaid and the access prob-
lem has been for state Medicaid agencies to contract with managed
166 THE ANNALS OF THE AMERICAN ACADEMY

care organizations. Increasingly, the same third-party companies administer both

private and public health insurance plans. From 1991 to 2009, the fraction of
Medicaid recipients who shifted from the traditional fee-for-service model to
Medicaid managed care (MMC) increased from 11 percent to 71 percent
(Duggan and Hayford 2013).
MMC has not proven to be a panacea, however. Using data on state and
local level MMC mandates and detailed data from CMS on state Medicaid
expenditures, Duggan and Hayford (2013) show that the shift to MMC
increased Medicaid spending, a finding that echoes Duggan (2004). However,
the effects of the shift to MMC on Medicaid spending varied significantly
across states as a function of the generosity of the state’s baseline Medicaid
provider reimbursement rates. MMC can increase costs to the state when
insurance companies negotiate reimbursement formulas that turn out to be
more advantageous to them than the state presumably foresaw. MMC can also
increase costs by negotiating a reimbursement rate based on average Medicaid
costs and then systematically encouraging only low-cost individuals to enroll in
their plans (called “cream skimming”).
Studies examining the estimated effects of MMC on access to care have shown
mixed results. Bindman et al. (2005) study the welfare population and find reduc-
tions in hospitalization for ambulatory care sensitive conditions, which were
larger for minority patients. They interpret this as evidence of improved access
to preventive care, though it could also reflect decreased access to hospital care.
L. Baker and Afendulis (2005) find that MMC did shift care away from emer-
gency rooms and toward outpatient visits, but it also led to increases in the
reporting of unmet medical needs and decreases in having a usual source of care.
Herring and Adams (2011) also find reductions in access to care.
Studies of MMC implementation in Tennessee and North Carolina (Conover,
Rankin, and Sloan 2001) and California (Aizer, Currie, and Moretti 2007) find
that compulsory managed care had a negative impact on the use of prenatal care
and on birth outcomes. In California, this may be because Medicaid managed
care contracts “carved out” care for sick newborns—that is, high-cost newborns
were covered by a state fund rather than by the managed care companies, so that
companies had little incentive to act to prevent poor birth outcomes. The exam-
ple suggests that the contracts governing these plans have important effects on
the quality of care provided.
The switch to managed care may also affect who receives services if MMC
organizations are actively trying to select low-risk patients. Kuziemko, Meckel,
and Rossin-Slater (2018) find that managed care companies in Texas were less
likely to enroll African American women than other women, leading to lower use
of prenatal care and poorer birth outcomes among this population. Currie and
Fahr (2005) report similar findings for California. Kaestner, Dubay, and Kenney
(2005) also find that compulsory MMC for pregnant women was associated with
increases in the incidence of low birth weight and prematurity.
On the other hand, Chorniy, Currie, and Sonchak (2018) find that in South
Carolina, a move to managed care increased screening and treatment of chronic
conditions in children, in part because providers received higher
What does medicaid do, and can we do it better? 167

reimbursements for children with such conditions. Layton et al. (2019) also find
that switching the disabled Medicaid caseload to managed care increased spend-
ing and improved health outcomes. One mechanism for these effects was that the
Texas Medicaid program released the prescription drug cap when privatization
was introduced, leading to an increase in drug spending across a range of medica-
tions used for management of acute and chronic conditions (e.g., asthma, antide-
pressants, etc.), and a reduction in mental health–related hospitalizations.
There is also mixed evidence on how the transition to managed care affected
the disabled. On one hand, using the National Health Interview Survey,
Coughlin, Long, and Graves (2008) find some evidence of improved access to
care under MMC relative to fee-for-service (FFS) Medicaid, with most of the
gains limited to beneficiaries in urban areas. Their findings suggest that under
MMC, the disabled experienced increases in having a usual source of care for
preventative care, having received some type of health care in the past 12
months, and having had contact with both general practitioners and specialists.
On the other hand, LoSasso and Freund (2000) look at two California counties
and find that increases in MMC enrollments among the disabled were associated
with increases in the use of ambulatory services and prescription drug expendi-
tures but also with increases in emergency department and hospital admissions.
Burns (2009a) shows that disabled adults in MMCs still face large out-of-pocket
expenses: 82 percent of enrollees reported incurring an out-of-pocket health care
expense during the year, mostly for prescription medicines. Burns (2009b) shows
that compared to disabled FFS enrollees, MMC enrollees are 24.9 percent more
likely to wait more than 30 minutes to see a provider, 32 percent more likely to
report a problem accessing a specialist, and 10 percent less likely to receive a flu
shot within the past year.
The trend toward managed care for Medicaid patients seems unlikely to be
reversed, so it is not an active policy debate. However, we have discussed the
research on it in some detail because it offers a cautionary example of how a
“private-public” partnership can turn out for state governments. Private health
insurance companies have proven adept at outmaneuvering the state when it
comes to negotiating contracts, so Medicaid costs have often gone up rather than
down under MMC. Additionally, state governments have often proven relatively
ineffectual in requiring that MMC provide improved access to care.
The examples suggest that the main issue with MMC has to do with the structure
of provider reimbursements and the incentives that they provide to serve (or not
serve) particular types of clients. That is, the specific details of the contracts matter.
At the state level, greater cooperation among states in terms of sharing contract lan-
guage and experiences might be helpful. At the federal level, CMS might consider
offering technical assistance and “best practices” for managing these contracts.

Additional possible reforms

In this subsection, we first discuss some proposed changes to the Medicaid
program itself and then briefly consider some proposals that would subsume
Medicaid in a reform of the United States’ larger health insurance system.
168 THE ANNALS OF THE AMERICAN ACADEMY

Block grants. One perennial proposal for Medicaid reform would involve
making it an entirely state-run program that receives a “block grant” from the
federal government. The idea is similar to the 1996 welfare reform that trans-
formed the old Aid for Families with Dependent Children (AFDC) program,
which was jointly funded by the federal government and the state, into the
­state-only Temporary Assistance for Needy Families Program (TANF). Under
AFDC, states were entitled to unlimited matching funds to support needy fami-
lies but set their own benefit levels and administrative rules under federal guide-
lines. Under TANF, states received block grants that initially were well above the
matching grants that they had been receiving for AFDC and were given much
more freedom to change program rules. Critics of this change argue that states
yielded to a powerful temptation to use block grants for other purposes and that
harsh new program rules (e.g., imposing mandatory job search at application,
cutting time limits from five years to two, etc.) have resulted in shrunken pro-
grams that were of little use during the economic downturn that began in 2008
(Dilger and Boyd 2014).
In some ways, the argument for block-granting Medicaid is at least as strong
as that for block-granting welfare, given that regulating health insurance is con-
stitutionally a function of state governments. However, states have been wary
about embracing this idea, largely because the rate of growth of health care costs
has outstripped the overall inflation rate for decades. Therefore, even a generous
block grant indexed to the rate of overall inflation would likely prove insufficient
to cover future health care costs. Currently, Tennessee is considering petitioning
the federal government to replace its Medicaid program with a block grant. The
state is attempting to build in indexing to inflation in health care costs to protect
against future erosion of state health care financing.
Even if this newer approach to block-granting Medicaid goes forward, it does
not address a fundamental critique of state-level control over Medicaid pro-
grams. Many people believe that all Americans should have access to at least a
core set of covered services, regardless of which state they live in, which implies
a more limited role, rather than a broader role, for states in rule-making about
what Medicaid covers and who should be eligible. Block-granting Medicaid
would likely reduce Medicaid costs mainly by limiting the amount that some
states could or would spend, which would lead to cuts in coverage and in covered
services and greater inequities in health care access across states.

Work requirements. A second proposal that has gained currency recently is to

impose work requirements on “able-bodied,” working-age, adult Medicaid recip-
ients. The best-case scenario is that work requirements would encourage
Medicaid recipients to find jobs that included private health insurance and that
these jobs would provide a steppingstone to a better life. However, even if such
jobs are out of reach for many Medicaid recipients, some observers want to
impose work requirements because they find it objectionable to offer benefits to
people who are able to work if they are not working.
In January 2018, CMS invited states to solicit proposals for imposing work
requirements. As of December 2018, six state proposals had been approved by
What does medicaid do, and can we do it better? 169

CMS and another nine states had submitted applications. Work requirements
exempt children, pregnant women, and the disabled, and may or may not exempt
primary caregivers for young children or the disabled. Typically, they require
nonexempt individuals to work, volunteer, or go to school for 80 hours a month.
Some proposals also count hours of job search toward the 80-hour total. Unlike
work requirement programs for TANF and SNAP (the Supplemental Nutrition
Assistance Program, or Food Stamps), the Medicaid work requirement programs
do not build in any funds for job training, job search assistance, child care assis-
tance, or other types of assistance that support employment.
Medicaid work requirements have been controversial and subject to legal chal-
lenge. They aim to fix a problem that does not exist: the research literature reviewed
above suggests that only one study has ever found large effects of Medicaid on
employment. The Oregon Health Insurance Experiment, which has arguably the
best research design, found no effect on employment. Most able-bodied welfare
recipients work already but have low-wage jobs that do not offer private health
insurance coverage. Thus, work requirements in Medicaid have more to do with
making a political point than with addressing a real social problem.
While they are highly unlikely to lead to large employment gains, the evidence
indicates that Medicaid work requirements lead to large losses in health insur-
ance coverage. This should not be surprising given a great deal of evidence that
imposing administrative hurdles (such as proving that one has work or is exempt
from work requirements) generally reduces coverage in any public program
(Currie and Gahvari 2008).
However, there is also a specific example available to policy-makers. Arkansas’
work-requirement program began in June 2018 and is continuing in the face of a
legal challenge. In the first six months of Arkansas’ program, seventeen thousand
people lost coverage out of about sixty-five thousand people who needed to fulfill
the work requirements (Scott 2018). Under the Arkansas rules, once people lose
coverage, they are ineligible to reapply for Medicaid until the next calendar year.
Many of the people who lost coverage in Arkansas may actually have been in
compliance with the new requirements but simply failed to report. They may
have been unaware of the new requirement, been unable to navigate the state’s
complicated web interface for reporting (use of the web interface is compulsory),
or simply lacked Internet access. Arkansas’ experience suggests that work
requirements will reduce Medicaid costs, primarily by disenrolling previously
eligible low-income recipients. However, keeping in mind that the elderly and
the disabled account for 60 percent of Medicaid costs, the cost savings obtained
by disenrolling nondisabled adults may prove less than what legislators expected
when imposing these measures.

Long-term care. There are few reforms under discussion that would address
the large fraction of Medicaid costs consumed by elderly and disabled people in
nursing homes. In principle, Medicaid funding could be used to encourage aging
in place by providing home health care.18 One demonstration, the “Community
Aging in Place—Advancing Better Living for Elders” (CAPABLE), provides
home visits by teams of clinicians (i.e., occupational therapists, nurses, and
170 THE ANNALS OF THE AMERICAN ACADEMY

“handy workers”) combined with adaptation of the home (i.e., home repairs,
adaptive modifications, or installation of assistive devices valued up to $1,300 in
2013 USD). The intervention led to significant reductions in expenditures and
other medical service use in this population compared to a matched control over
the first 17 months (Szanton et al. 2018). The problem with expanding these
types of programs is that they would likely draw many more people into Medicaid
by making long-term care arrangement more appealing. Lieber and Lockwood
(2019) suggest that the increase in demand for Medicaid services could be very
large, placing additional financial pressure on the program. That is, for every dol-
lar saved by allowing an existing nursing home resident to stay in his or her home,
more than a dollar would likely be spent on new entrants to the Medicaid pro-
gram who would be attracted by more generous home health benefits.
Attempts to encourage people to purchase private long-term care insurance
have shown little success, as we discussed. Another potential option would be to
use the Social Security system to provide a form of long-term health insurance,
allowing new Social Security recipients to choose lower income payments in
exchange for using insurance for LTC (see J. Brown and Dynan 2017). This pro-
posal would in effect allow people to use their Social Security benefits to pur-
chase LTC insurance, but it is unclear at this point how many people would take
up such an option, especially if publicly funded nursing home care continues to
be provided to those who have not “purchased” such coverage.
Another proposal for long-term care involves ending Medicaid coverage of nurs-
ing home care and instead creating a new, universal Medicare benefit similar to the
way the government created a new Medicare prescription drug benefit (Medicare
Part D). At present, elderly people must spend down their assets to qualify for
Medicaid coverage of long-term nursing home care and Medicare currently pays for
only 100 days of rehabilitative nursing home care. Presumably, states would embrace
a proposal to transfer the responsibility for covering nursing home care from
Medicaid to Medicare as it would make funding nursing home care a solely federal
responsibility. Moreover, elderly people would no longer have to become destitute to
receive assistance with nursing home costs. However, such a proposal would greatly
increase rather than decrease overall costs. The Trump administration has been mov-
ing in the opposite direction, with proposals for large cuts in Medicaid, which would
necessitate reductions in state expenditures on nursing home care, and proposals to
shift more of the financial burden of LTC to family members (Rau 2017).

Reimbursement rates. Leaving aside the issue of long-term care, another pro-
posal to reform Medicaid involves raising Medicaid reimbursement rates to
match Medicare rates but leaving the structure of the program otherwise
unchanged. Such a change would increase provider participation in Medicaid
and overall quality but would greatly increase costs given that Medicaid reim-
burses about 70 percent of what Medicare pays on average.19 To make such a
proposal viable might require a fundamental rethinking of the way Medicare sets
fees. Unlike in other rich countries, CMS is forbidden by law from negotiating
Medicare fees with private providers, resulting in the government being unable
to use its market power to lower provider prices.
What does medicaid do, and can we do it better? 171

Single-payer schemes. A more radical proposal that has been gaining traction
would be to eliminate Medicaid and offer “Medicare for All.” This proposal is not
equivalent to a “single-payer” system because Medicare currently offers the option
to select “Medicare Advantage” plans offered by private insurers. Medicare’s drug
benefit (Medicare Part D) is also offered through private health insurers. Medicare
for All would be an expensive proposal, especially given that many employers
would likely stop offering full-service private insurance if high-quality public insur-
ance were freely available to people younger than 65. (Private insurers might con-
tinue to offer “Medi-gap”-like policies to pay for services that Medicare did not
cover.) Once again, making this proposal feasible would likely require a commit-
ment to negotiate Medicare fees with providers and private insurers, as well as
higher taxes to pay for the program. Also, given that Medicare currently only pays
for 100 days of nursing home care, Medicare for All would need to include a new
benefit covering longer-term care currently paid for by Medicaid.
Another alternative would be a single-payer health insurance system like
Canada’s. This proposal is often regarded as politically unfeasible in the United
States, where private insurers are large and powerful companies, and many con-
sumers are suspicious of attempts to standardize care and restrict consumer
choice. Canada’s single-payer system is an outlier among rich countries. Most
developed countries have systems that mix privately and publicly funded care
(Stabile and Thompson 2014). Even the United Kingdom’s National Health
Service has moved in this direction in recent years. Hence, it seems likely that
the United States will continue to operate some type of hybrid public/private
system of health insurance rather than moving to a truly single-payer system.
In fact, one could argue that the ACA moved the United States much closer
to many European models of health insurance coverage by patching together a
mix of public and private health insurance sources that has come close to achiev-
ing full coverage even in the face of considerable political resistance. While far
from perfect, the most pragmatic and feasible policy option is likely to be to work
within the ACA framework to strengthen regulation of health care providers
(including some consideration of some form of price controls), continue to
expand and reform Medicaid, and strengthen health care exchanges. The increas-
ing prevalence of Medicaid-managed care plans operated by private companies
may offer one mechanism for eventually allowing people who are not low income
the option of purchasing Medicaid coverage.

Conclusion
Medicaid directly affects not only poor women and children but many middle-
class people who become disabled, spend down their assets, and require nursing
home care. In fact, although the majority of Medicaid recipients are women and
children, the majority of the spending is on the elderly and disabled. Because of
its size and impact on state and federal budgets, Medicaid also has indirect
effects on the health care system and on other public spending. Medicaid has
172 THE ANNALS OF THE AMERICAN ACADEMY

improved the health of low-income pregnant women and children, and the
elderly and disabled increasingly depend upon it. There is less evidence about
the longer-term effects of Medicaid for nondisabled adults, given that wide-
spread Medicaid coverage of this group is quite new. However, the available
evidence suggests that the program increases well-being and reduces financial
uncertainty without having much effect on labor-force outcomes in any group. At
the same time, the program is extremely expensive, and costs continue to grow
along with overall U.S. health care costs.
Some of the proposals for reforming Medicaid include transforming Medicaid
into a “block grant” to states rather than a federal program or allowing states to
impose work requirements on recipients. Given our discussion of this reform
option, we do not feel that these proposals would achieve the goal of providing
minimally acceptable health care coverage for all Americans. Other proposals
focus on replacing Medicaid with a universal program such as Medicare for All.
This may prove a heavy lift, given that policy-makers have tried and failed to
implement such a program before. Still other proposals would keep Medicaid’s
basic framework but make major changes, such as finding an alternative mecha-
nism for funding universal LTC (perhaps through Medicare) or making Medicaid
payments to health care providers more comparable to those of Medicare and
private payers. While such changes would not be easy, we feel they are worth
pursuing. Finally, basic reforms that would affect Medicaid and all other health
insurers include strengthening antitrust regulation of health care providers,
adopting some form of price regulation, encouraging price transparency, and
strengthening health insurance exchanges. These smaller changes likely have the
best chance of success and should be pursued vigorously.
As the political tug-of-war over expanding or contracting Medicaid continues,
it is worthwhile to ask how much spending on medical care is too much from a
social perspective? R. Hall and Jones (2007) argue that as a society becomes
wealthier and people have more things, the marginal value of consumption falls
relative to the marginal value of having additional healthy life years to enjoy
them. Cutler, Rosen, and Vijan (2006) argue that the increased spending on
medical care in the United States since 1960 has been “worth it” given any rea-
sonable valuation of the improvements in health and life expectancy. From this
perspective, the most important question is how we can improve both the equity
and the efficiency of health care provision in the United States, rather than how
we can reduce spending. Although expansions of the Medicaid program have
improved the equity of health care provision, targeting both equity and efficiency
may require more fundamental reforms of the U.S. health care system.

Notes
1. [Link]
report-highlights/[Link].
2. [Link]
healthexpenddata/[Link].
What does medicaid do, and can we do it better? 173

3. Medicaid is often confused with Medicare, a public health insurance program that covers all elderly
people and is run by the federal government. In 2016, Medicare accounted for 20 percent of national
health expenditures.
4. Medicaid targets children in the poorest households, those living in families with incomes up to 133
percent of the federal poverty level (FPL) (for children 0–6 years of age) or up to 100 percent of the FPL
(for children ages 6–18). CHIP covers children starting at the point at which a state’s Medicaid eligibility
threshold ends, providing coverage to children in families with income up to 200 percent of the FPL or
higher.
5. [Link]
6. [Link]
[Link].
7. For instance, MACStats: Medicaid and CHIP Data Book (2017, 141) says, “One consequence of the
transition from the MSIS to the T-MSIS is that there is now a gap in available data from many states.
Several states began the transition to the T-MSIS in 2014 and do not have complete information for fiscal
year (FY) 2014 available in the MSIS. Although many of these states have submitted data for these missing
months through the T-MSIS, the data are still being validated by CMS and are not available for publication
at this time. As a result, MACPAC was not able to fully update several exhibits that provide enrollment
and spending data by eligibility group.”
8. While Figure 1 shows a gentle increase in participation among the aged after 2012, Figure 2 shows
a marked decline in spending per beneficiary on this group. Since there is no major reason for this decline
in real spending among the elderly, we attribute this pattern to the change in data source.
9. Bitler, Hoynes, and Kuka (2016) examine the role of the safety net (including Medicaid) in mitigat-
ing the adverse effects of the Great Recession on child poverty.
10. According to the American Immigration Council (2018), roughly 8 million U.S.-citizen children
have a family member who is undocumented.
11. Long-term care spending includes spending on nursing home and continuing care retirement com-
munities, home health care, and residential and personal care. The major form of insurance for nursing
home utilization is Medicaid, which provides some support to approximately 65 percent of nursing home
residents and pays for 45 percent of the total nursing home bill. Medicare does not pay for long-term care,
but the program covers up to 100 days of skilled nursing home care per benefit period, and Medicare pays
for 17 percent of nursing home care (Georgetown University, Long-Term Care Financing Project,
“National Spending for Long-Term Care,” fact sheet, 2007).
12. See [Link] for a useful summary of the
rules pertaining to income and asset limits.
13. [Link]
14. [Link]
15. The flip side of increasing financial security is that Medicaid eligibility may reduce savings. Gruber
and Yelowitz (1999) find that the Medicaid expansions of the late 1980s and early 1990s reduced asset
holdings, although the sensitivity of their results has been critiqued (Gittleman 2011). Maynard and Qiu
(2009) find effects on savings for households in the middle of the income distribution, which makes sense
given that poor families have little savings in any case.
16. For example, states may have lower income cut offs or tougher criteria for determining eligibility.
For more details about these programs see Watts, Cornachione, and Musumeci (2016) and [Link]
[Link]/report/understanding-medicaid-home-and-community-services-primer-2010-edition/state-
209b-option. The law requires states to continue to offer Medicaid coverage to “medically needy” recipi-
ents whose SSI benefits place their income above the income cut off.
17. Medication classes that experienced relatively large increases included contraceptives (a 22 percent
increase) and cardiovascular drugs (a 21 percent increase). Several drug classes more consistent with acute
conditions such as allergies and infections experienced significantly smaller increases.
18. Like nursing home care, home health care can be paid for either by Medicaid, if it is long-term or
by Medicare if it is up to 100 days. Even though the statutory eligibility criteria have varied little, changes
in the interpretations of the criteria and in the licensing of home health care agencies have led to large
swings in the availability and cost of home health care over time (Davitt and Choi 2008).
19. The Kaiser Family Foundation creates a Medicaid-to-Medicare fee index, which is available at:
[Link]
174 THE ANNALS OF THE AMERICAN ACADEMY

Model=%7B%22colId%22:%22Location%22,%22sort%22:%22asc%22%7D. The index shows a great

deal of variation across states and is generally lower for primary care than for specialist care.

References
Aizer, Anna, Janet Currie, and Enrico Moretti. 2007. Does managed care hurt health? Evidence from
Medicaid mothers. Review of Economics and Statistics 89 (3): 385–99.
Alexander, Diane, and Janet Currie. 2017. Are publicly insured children less likely to be admitted to hos-
pital than the privately insured (and does it matter)? Economics and Human Biology 25:33–51.
American Immigration Council (AIC). 23 May 2018. US citizen children impacted by immigration
enforcement. Fact Sheet. Washington, DC: AIC.
Association for Long Term Care Planning. 2019. Costs of long term care across states. Fort Myers, FL:
Association for Long Term Care Planning. Available from [Link]
term-care-cost.
Autor, David, Amitabh Chandra, and Mark Duggan. 2011. Public health expenditures on the working age
disabled: Assessing Medicare and Medicaid utilization of SSDI and SSI recipients. MIT Working
Paper, Massachusetts Institute of Technology, Cambridge, MA.
Autor, David, Mark Duggan, Kyle Greenberg, and David S. Lyle. 2016. The impact of disability benefits
on labor supply: Evidence from the VA's disability compensation program. American Economic
Journal: Applied Economics 8 (3): 31–68.
Baicker, Katherine, Amy Finkelstein, Jae Song, and Sarah Taubman. 2014. The impact of health insurance
expansions on other social safety net programs. American Economic Review: Papers and Proceedings
104 (5): 322–28.
Baicker, Katherine, Sarah L. Taubman, Heidi L. Allen, Mira Bernstein, Jonathan H. Gruber, Joseph P.
Newhouse, Eric C. Schneider, Bill J. Wright, Alan Zaslavsky, and Amy N. Finkelstein. 2013. The
Oregon experiment — Effects of Medicaid on clinical outcomes. New England Journal of Medicine
368:1713–22.
Baker, Lawrence, and Christopher Afendulis. 2005. Medicaid managed care and health care for children.
Health Services Research 40 (5): 1466–88.
Baker, Michael, Janet Currie, and Hannes Schwandt. 2017. Mortality inequality in Canada and the US:
Convergent or divergent trends. NBER Working Paper 23514, National Bureau of Economic
Research, Cambridge, MA.
Benitez-Silva, Hugo, Moshe Buchinsky, and John Rust. 2004. How large are the classification errors in the
Social Security Disability award process? NBER Working Paper 10219, National Bureau of Economic
Research, Cambridge, MA.
Bindman, Andrew, Arpita Chattopadhyay, Denis Osmond, William Huen, and Peter Bacchetti. 2005. The
impact of Medicaid care on hospitalizations for ambulatory care sensitive conditions. Health Services
Research 40 (1): 19–38.
Bitler, Marianne, Hillary Hoynes, and Emily Kuka. 2016. Child poverty, the Great Recession and the social
safety net in the United States. Journal of Policy Analysis and Management 36 (2): 358–89.
Bitler, Marianne, and Madelyn Zavodny. 2017. Medicaid: A review of the literature. In Handbook in law
and economics federalism, ed. Jonathan Klick, 183–213. Cheltenham: Edward Elgar.
Black, Bernard, Alex Hollingsworth, Leticia Nunes, and Kosali Simon. 2018. The effect of health insurance
on mortality: What can we learn from the Affordable Care Act coverage expansions? Institute for Policy
Research Working Paper, Washington, DC. Available from [Link]
insurance_mortality_2018-[Link].
Boudreaux, Michel H., Ezra Golberstein, and Donna D. McAlpine. 2016. The long-term impacts of
Medicaid exposure in early childhood: Evidence from the program’s origin. Journal of Health
Economics 45:161–75.
Brown, David W., Amanda Kowalski, and Ithai Z. Lurie. 2015. Medicaid as an investment in children:
What is the long-term impact on tax receipts? NBER Working Paper 20835, National Bureau of
Economic Research, Cambridge, MA.
What does medicaid do, and can we do it better? 175

Brown, Jason, and Karen Dynan. 2017. Increasing the economic security of older women. The Hamilton
Project, Policy Proposal 2017-11. Available from [Link]
increasing_economic_security_older_women _BrownDynan.pdf.
Buchmueller, Thomas, John Ham, and Lara Shore-Sheppard. 2016. The Medicaid program. In Economics
of means-tested transfer programs in the United States, ed. Robert Moffitt, 21–136. Chicago, IL:
National Bureau of Economic Research.
Burns, Marguerite E. 2009a. Medicaid managed care and cost containment for adult beneficiaries with
disabilities. Medical Care 47 (10): 1076–96.
Burns, Marguerite E. 2009b. Medicaid managed care and health care access for adult beneficiaries with
disabilities. Health Services Research 45 (5): 1521–41.
Chorniy, Anna, Janet Currie, and Lydmila Sonchak. 2018. Exploding asthma and ADHD caseloads: The
role of Medicaid managed care. Journal of Health Economics 60:1–15.
Clemans-Cope, Lisa, Sharon K. Long, Teresa Coughlin, Alshadye Yemane, and Dean Resnick. 2013. The
expansion of Medicaid coverage under the ACA: Implications for health care access use and spending
for vulnerable low-income adults. Inquiry 50 (2): 135–49.
Cohodes, Sarah, Daniel Grossman, Samuel Kleiner, and Michale Lovenheim. 2016. The effect of child
health insurance access on schooling: Evidence from public insurance expansions. Journal of Human
Resources 51 (3): 727–59.
Conover, Chistopher J., Peter J. Rankin, and Frank Sloan. 2001. Effects of Tennessee Medicaid managed
care on obstetrical care and birth outcomes. Journal of Health Politics Policy and Law 26 (6): 1291–
1324.
Conover, Chris J., and Frank A. Sloan. 1998. Does removing certificate-of-need regulations lead to a surge
in health care spending? Journal of Health Politics, Policy and Law 23 (3): 455–81.
Cooper, Zachary, Stuart V. Craig, Martin Gaynor, and John Van Reenen. 2019. The price ain’t right?
Hospital prices and health spending on the privately insured. Quarterly Journal of Economics 134 (1):
51–107.
Coughlin, Teresa, Sharon Long, and John Graves. 2008. Does managed care improve access to care for
Medicaid beneficiaries with disabilities? A national study. Inquiry 45:395–407.
Courtemanche, Charles, James Marton, Benjamin Ukert, Aaron Yelowitz, and Daniela Zapata. 2018. Early
effects of the Affordable Care Act on health care access risky health behaviors and self-Assessed health.
Southern Economic Journal 84 (3): 660–91.
Courtemanche, Charles, Benjamin Ukert, Daniela Zapata, James Marton, and Aaron Yelowitz. 2017.
Impacts of the Affordable Care Act on health insurance coverage in Medicaid expansion and non-
expansion states. Journal of Policy Analysis and Management 36 (1): 178–210.
Currie, Janet, Sandra Decker, and Wanchuan Lin. 2008. Has public health insurance for older children
reduced disparities in access to care and health outcomes? Journal of Health Economics 27 (6):
1407–1652.
Currie, Janet, and John Fahr. 2005. Medicaid managed care: Effects on children's Medicaid coverage and
utilization of care. Journal of Public Economics 89 (1): 85–108.
Currie, Janet, and Fahrooz Gahvari. 2008. Transfers in cash and in-kind: Theory meets the data. Journal
of Economic Literature 46 (2): 333–83.
Currie, Janet, and Jonathan Gruber. 1996a. Health insurance eligibility utilization of medical care and
child health. Quarterly Journal of Economics 111 (2): 431–66.
Currie, Janet, and Jonathan Gruber. 1996b. Saving babies: The efficacy and cost of recent changes in the
Medicaid eligibility of pregnant women. Journal of Political Economy 104 (6): 1263–96.
Currie, Janet, and Hannes Schwandt. 2016. Mortality inequality: The good news from a county-level
approach. Journal of Economic Perspectives 30 (2): 29–52.
Cutler, David, and Jonathan Gruber. 1996. Does public insurance crowd out private insurance? Quarterly
Journal of Economics 111 (2): 391–430.
Cutler, David, Allison Rosen, and Sandeep Vijan. 2006. The value of medical spending in the United States
1960–2000. New England Journal of Medicine 355:920–27.
Dafny, Leemore, and Jonathan Gruber. 2005. Public insurance and child hospitalizations: Access and
efficiency effects. Journal of Public Economics 89 (1): 109–29.
Dague, Laura, Thomas DeLeire, and Lindsey Leininger. 2017. The effect of public insurance coverage for
childless adults on labor supply. American Economic Journal: Economic Policy 9 (2): 124–54.
176 THE ANNALS OF THE AMERICAN ACADEMY

Davitt, Joan K., and Sunha Choi. 2008. Tracing the history of Medicare home health care: The impact of
policy on benefit use. Journal of Sociology & Social Welfare: 35 (1): Article 12.
DeLeire, Thomas, Leonard M. Lopoo, and Kosali Simon. 2011. Medicaid expansions and fertility in the
United States. Demography 48 (2): 725–47.
Deshpande, Manasi. 2016. Does welfare inhibit success? The long-term effects of removing low- income
youth from the disability rolls. American Economic Review 106 (11): 3300–3330.
Dilger, Robert, and Eugene Boyd. 15 July 2014. Block grants: Perspectives and controversies. Washington,
DC: Congressional Research Service. Available from [Link]
Duggan, Mark. 2004. Does contracting out increase the efficiency of government programs? Evidence
from Medicaid HMOs. Journal of Public Economics 88:2549–72.
Duggan, Mark, and Tamara Hayford. 2013. Has the shift to managed care reduced Medicaid spending?
Evidence from state and local-level mandates. Journal of Policy Analysis and Management 32 (3):
505–35.
East, Chloe, Sarah Miller, Marianne Page, and Laura Wherry. 2017. Multi-generational impacts of child-
hood access to the safety net: Early life exposure to Medicaid and the next generation’s health. NBER
Working Paper 23810, National Bureau of Economic Research, Cambridge, MA.
Finkelstein, Amy, Neale Mahoney, and Matthew J. Notowidigdo. 2018. What does (formal) health insur-
ance do, and for whom? Annual Review of Economics 10 (1): 261–86.
Finkelstein, Amy N., Sarah L. Taubman, Heidi L. Allen, Bill J. Wright, and Katherine Baicker. 2016. Effect
of Medicaid coverage on ED use — Further evidence from Oregon’s experiment. New England
Journal of Medicine 375:1505–7.
Finkelstein, Amy, Sarah Taubman, Bill Wright, Mira Bernstein, Jonathan Gruber, Joseph Newhouse, Heidi
Allen, and Katherine Baicker, and The Oregon Health Study Group. 2012. The Oregon health insur-
ance experiment: Evidence from the first year. Quarterly Journal of Economics 127 (3): 1057–1106.
Frean, Michael, Jonathan Gruber, and Benjamin Sommers. 2017. Premium subsidies, the mandate and
Medicaid expansion: Coverage effects of the Affordable Care Act. Journal of Health Economics
53:72–86.
Garfield, Rachel I., Judith R. Lave, and Julie M. Donohue. 2010. Health reform and the scope of benefits
for mental health and substance use disorder services. Psychiatric Services 61 (11): 1081–86.
Garthwaite, Craig, Tal Gross, and Matthew Notowidigdo. 2014. Public health insurance labor supply and
employment lock. Quarterly Journal of Economics 129 (2): 653–96.
Gettens, J. M., Henry A. D. Mitra, and J. Himmelstein. 2011. Have working-age people with disabilities
shared in the gains of Massachusetts health reform? Inquiry: The Journal of Health Care Organization
Provision and Financing 48 (3): 183–96.
Ghosh, Ausmita, Kosali Simon, and Benjamin D. Sommers. 2017. The effect of state Medicaid expansions
on prescription drug use: Evidence from the Affordable Care Act. NBER Working Paper No. 23044,
National Bureau of Economic Research, Cambridge, MA.
Gittleman, Maury. 2011. Medicaid and wealth: A re-examination. B.E. Journal of Economic Analysis 11
(1). Available from [Link]/10.2202/1935-1682.3029.
Glied, Sherry, and Marilyn Tavenner. 2019. Medicaid through the crystal ball of historical CMS projec-
tions. Health Affairs. doi 10.1377/hblog20190226.922368.
Goda, Gopi S. 2011. The impact of state tax subsidies for private long-term care insurance on coverage on
Medicaid expenditures. Journal of Public Economics 95 (7): 744–57.
Goodman-Bacon, Andrew. 2016. The long-run effects of childhood insurance coverage: Medicaid imple-
mentation adult health and labor market outcomes. NBER Working Paper 22899, National Bureau of
Economic Research, Cambridge, MA.
Goodman-Bacon, Andrew. 2018. Public insurance and mortality: Evidence from Medicaid implementa-
tion. Journal of Political Economy 126 (1): 216–62.
Gooptu, Angshuman, Asako Moriya, Kosali Simon, and Benjamin Sommers. 2016. Medicaid expansion did
not result in significant employment changes or job reductions in 2014. Health Affairs 35 (1): 111–18.
Gouskova, Elena. 2015. Public health insurance labor supply and employment lock: Effects or data arti-
facts? Available from [Link]
Grabowski, David. 2001. Medicaid reimbursement and the quality of nursing home care. Journal of Health
Economics 20 (4): 549–69.
What does medicaid do, and can we do it better? 177

Grabowski, David. 2004. A longitudinal study of Medicaid payment private-pay price and nursing home
quality. International Journal of Health Care Finance and Economics 4 (1): 5–26.
Grabowski, David, and Joseph J. Angelelli. 2004. The relationship of Medicaid payment rates bed con-
straint policies and risk-adjusted pressure ulcers. Health Service Research 39 (4): 793–812.
Grabowski, David, Joseph J. Angelelli, and Vincent Mor. 2004. Medicaid payment and risk-adjusted nurs-
ing home quality measures. Health Affairs 23 (5): 243–52.
Grabowski, David, Zhanlian Fend, Orna Intrator, and Vincent Mor. 2004. Recent trends in state nursing
home payment and the role of provider taxes. Medical Care Research and Review 61 (1): 89–115.
Grabowski, David, and Jonathan Gruber. 2007. Moral hazard in nursing home use. Journal of Health
Economics 26 (3): 560–77.
Grabowski, David, Jonathan Gruber, and Joseph J. Angelelli. 2008. Nursing home quality as a public good.
Review of Economics and Statistics 90 (4): 754–64.
Grabowski, David, Jonathan Gruber, and Vincent Mor. 13 June 2017. You’re probably going to need
Medicaid. New York Times.
Grabowski, David, Robert Ohsfeldt, and Michael Morrisey. 2003. The effects of CON repeal on Medicaid
nursing and long term care expenditures. Inquiry 40 (2): 146–57.
Gross, Tal, and Matthew Notowidigdo. 2011. Health insurance and the consumer bankruptcy decision:
Evidence from expansions of Medicaid. Journal of Public Economics 95 (7): 767–78.
Gruber, Jonathan. 2008. Covering the uninsured in the United States. Journal of Economic Literature 46
(3): 571–606.
Gruber, Jonathan, and Benjamin Sommers. 2019. The Affordable Care Act’s effects on patients, providers,
and the economy: What we know so far? NBER Working Paper 25932, National Bureau of Economic
Research, Cambridge, MA.
Gruber, Jonathan, and Aaron Yelowitz. 1999. Public insurance and private savings. Journal of Political
Economy 107 (6): 1249–74.
Hall, Jean P., Adele Shartzer, Noelle K. Kurth, and Kathleen C. Thomas. 2017. Effect of Medicaid expan-
sion on workforce participation for people with disabilities. American Journal of Public Health 107 (2):
262–64.
Hall, Robert, and Charles Jones. 2007. The value of life and the rise in health spending. Quarterly Journal
of Economics 122 (1): 39–72.
Ham, John, and Kenneth Ueda. Forthcoming. Comparing the Conley-Taber and the standard approaches
to inference in difference-in-difference models based on small policy variation: The case of TennCare.
Journal of Labor Economics.
Herring, Bradley, and Esther K. Adams. 2011. Using HMOs to serve the Medicaid population: What are
the effects on utilization and does the type of HMO matter? Health Economics 20 (4): 446–60.
Howell, Embry, and Genevieve Kenney. 2012. The impact of the Medicaid/CHIP expansions on children:
A synthesis of the evidence. Medical Care Research and Review 69 (4): 372–96.
Hu, Luojia, Robert Kaestner, Bashkar Mazumder, Sarah Miller, and Ashley Wong. 2016. The effects of the
Patient Protection and Affordable Care Act Medicaid expansions on financial well-being. NBER
Working paper 22170, National Bureau of Economic Research, Cambridge, MA.
Intrator, Orna, David C. Grabowski, Jacqueline Zinn, Mark Schleinitz, Zhanlian Feng, Susan Miller, and
Vince Mor. 2007. Hospitalization of nursing home residents: The effects of states’ Medicaid payment
and bed-hold policies. Health Service Research 42 (4): 1651–71.
James, Everette, Walid Gellad, and Meredith Hughes. 2017. In this next phase of health reform we cannot
overlook long term care. Health Affairs Blog. Available from 10.1377/hblog20170316.059218.
Johnson, Richard W. 2 August 2016. Who is covered by private long-term care insurance? Washington,
DC: The Urban Institute.
Kaestner, Robert, Lisa Dubay, and Genevieve Kenney. 2005. Managed care and infant health: An evalua-
tion of Medicaid in the US. Social Science & Medicine 60:1815–33.
Kaestner, Robert, Bowen Garrett, Jiajia Chen, Anuj Gangopadhyaya, and Caitlyn Fleming. 2017. Effects
of ACA Medicaid expansions on health insurance coverage and labor supply. Journal of Policy Analysis
and Management 36 (3): 608–42.
Kaestner, Robert, Theodore Joyce, and Andrew Racine. 2001. Medicaid eligibility and the incidence of
ambulatory care sensitive hospitalizations for children. Social Science & Medicine 52 (2): 305–13.
178 THE ANNALS OF THE AMERICAN ACADEMY

Kamal, Rabah, and Cynthia Cox. 8 May 2018. How do healthcare prices and use in the US compare to other
countries? San Francisco, CA: The Kaiser Family Foundation.
Kearney, Melissa, and Phillip Levine. 2009. Subsidized contraception fertility and sexual behavior. Review
of Economics and Statistics 91:137–51.
Koma, Jonathan W., Julie M. Donohue, Colleen L. Barry, Haiden A. Huskamp, and Marian Jarlenski.
2017. Medicaid coverage expansions and cigarette smoking cessation among low-income adults.
Medical Care 55 (12): 1023–29.
Kuziemko, Ilyana, Katherine Meckel, and Maya Rossin-Slater. 2018. Does managed care widen infant
health disparities? Evidence from Texas Medicaid. American Economic Journal: Economic Policy 10
(3): 255–83.
Laes-Kushner, Rebecca. 27 March 2018. Skilled nursing facilities: Too many beds. Commonwealth
Medicine. Available from [Link]
Lanning, Joyce A., Michael Morrisey, and Robert Ohsfeldt. 1991. Endogenous hospital regulation and its
effects on hospital and non-hospital expenditures. Journal of Regulatory Economics 3 (2): 137–54.
Layton, Timothy J., Nicole Maestas, Daniel Prinz, and Boris Vabson. 2019. Private vs. public provision of
social insurance: Evidence from Medicaid. NBER Working Paper, National Bureau of Economic
Statistics, Cambridge, MA.
Leung, Pauline, and Alexandre Mas. 2018. Employment effects of the Affordable Care Act Medicaid
expansions. Industrial Relations 57 (2): 206–34.
Levine, Phillip, and Diane Schanzenbach. 2009. The impact of children’s public health insurance expan-
sions on educational outcomes. NBER Working Paper, National Bureau of Economic Statistics,
Cambridge, MA.
Lieber, Ethan, and Lee M. Lockwood. 2019. Targeting with in-kind transfers: Evidence from Medicaid
home care. American Economic Review 109 (4): 1461–85.
LoSasso, Anthony, and Deborah Freund. 2000. A longitudinal evaluation of the effect of Medi-Cal man-
aged care on Supplemental Security Income and Aid to Families with Dependent Children enrollees
in two California counties. Medical Care 38 (9): 937–47.
MacLean, Johanna C., Benjamin Cook, Nicholas Carson, and Michael Pesko. 2017. Public insurance and
psychotropic prescription medications for mental illness. NBER Working Paper 23760, National
Bureau of Economic Statistics, Cambridge, MA.
Maestas, Nicole, Kathleen J. Muellen, and Alexander Strand. 2014. Disability insurance and health insur-
ance reform: Evidence from Massachusetts. American Economic Review 104 (5): 329–35.
Markus, Anna R., Ellie Andres, Kristina D. West, Nicole Garro, and Cynthia Pellegrini. 2013. Medicaid
covered births 2008 through 2010 in the context of the implementation of health reform. Women’s
Health Issues 23 (5): e273–80.
Maynard, Alex, and Jiaping Qiu. 2009. Public insurance and private savings: Who is affected and by how
much? Journal of Applied Econometrics 24 (2): 282–308.
Mazumder, Bhashkar, and Sarah Miller. 2016. The effects of the Massachusetts health reform on house-
hold financial distress. American Economic Journal: Economic Policy 8 (3): 284–313.
Medicaid and CHIP Payment and Access Commission. 2017. MACStats: Medicaid and CHIP data book
2017. Available from [Link]
[Link].
Michalopolous, Charles, David Wittenburg, Dina Israel, Jennifer Schore, Anne Warren, Aparajita Zutshi,
Stephen Freedman, and Lisa Schwartz. 2011. The accelerated benefits demonstration and evaluation
project: Impacts on health and employment at 12 months. New York, NT: MDRC.
Miller, Sarah, and Laura Wherry. 2017. Health and access to care during the first 2 years of the ACA
Medicaid expansions. New England Journal of Medicine 376 (10): 947–56.
Miller, Sarah, and Laura Wherry. 2019. The long-term effects of early life Medicaid coverage. Journal of
Human Resources 54 (3): 785–824.
Mor, Vincent, Andrea Grunerir, Zhanlian Feng, David Grabowski, Orna Intrator, and Jacqueline Zinn.
2011. The effect of state policies on nursing home resident outcomes. Journal of American Geriatric
Association 59 (1): 3–9.
Musumeci, MaryBeth, and Katherine Young. 1 May 2017. State variation in Medicaid per enrollee spend-
ing for seniors and people with disabilities. San Francisco, CA: Kaiser Family Foundation.
What does medicaid do, and can we do it better? 179

National Academy of Social Insurance (NASI). April 2015. Addressing pricing power in health care mar-
kets. Washington DC: NASI.
Obama, Barack. 2016. United States health care reform progress to date and next steps. JAMA 316 (5):
525–32.
Rau, J. 24 June 2017. Medicaid cuts may force retirees out of nursing homes. New York Times.
Reschovsky, James. 1996. Demand for and access to institutional long-term care: The role of Medicaid in
nursing home markets. Inquiry 33 (1): 15–29.
Salkever, David. 2000. Regulation of prices and investment in hospitals in the United States. Handbook of
Health Economics 1 (Part B): 1489–1535.
Scott, Dylan. 18 December 2018. 16932 people have lost Medicaid coverage under Arkansas’s work
requirements. Vox.
Simon, Kosali, Aparna Soni, and John Cawley. 2017. The impact of health insurance on preventive care
and health behaviors: Evidence from the first two years of the ACA Medicaid Expansions. Journal of
Policy Analysis and Management 36 (2): 390–417.
Slusky, David. 2017. Significant placebo results in difference-in-differences analysis: The case of the ACA’s
parental mandate. Eastern Economic Journal 43 (4): 580–603.
Sommers, Benjamin, Katherine Baicker, and Andrew Epstein. 2012. Mortality and access to care among
adults after state Medicaid expansions. New England Journal of Medicine 367:1025–34.
Sommers, Benjamin, Robert J. Blendon, John Orav, and Andrew Epstein. 2016. Changes in utilization and
health among low-income adults after Medicaid expansion or expanded private insurance. JAMA
Internal Medicine 176 (10): 1501–9.
Sommers, Benjamin, Munira Z. Gunja, Kenneth Finegold, and Thomas Musco. 2015. Changes in self-
reported insurance coverage access to care and health under the Affordable Care Act. JAMA 314 (4):
366–74.
Sommers, Benjamin, Bethany Maylone, Robert J. Blendon, E. John Orav, and Arnold M. Epstein. 2017.
Three-year impacts of the Affordable Care Act: Improved medical care and health among low-income
adults. Health Affairs 36 (6). Available from [Link]
Stabile, Mark, and Sarah Thompson. 2014. The changing role of government in financing health care: An
international perspective. Journal of Economic Literature 52 (2): 480–518.
Stevens, Ann H., Douglas L. Miller, Marianne E. Page, and Mateusz Filipski. 2015. The best of times, the
worst of times: Understanding pro-cyclical mortality. American Economic Journal: Economic Policy 7
(4): 279–311.
Szanton, Sarah L., Y. Natalia Alfonso, Bruce Leff, Jack Guralnik, Jennifer L. Wolff, Laura N. Gitlin, and
David Bishai. 2018. Medicaid cost savings of a preventive home visit program for disabled older adults.
Journal of the American Geriatrics Society 66 (3): 614–20.
Taubman, Sarah L., Heidi L. Allen, Bill J. Wright, Katherine Baicker, and Amy N. Finkelstein. 2014.
Medicaid increases emergency-department use: Evidence from Oregon’s health insurance experiment.
Science 343 (6168): 263–68.
Watts, Molly O., Elizabeth Cornachione, and MaryBeth Musumeci. 2016. Medicaid financial eligibility for
seniors and people with disabilities in 2015. Available from [Link]
medicaid-financial-eligibility-for-seniors-and-people- with-disabilities-in-2015-report/.
Wherry, Laura, and Bruce Meyer. 2016. Saving teens: Using a policy discontinuity to estimate the effects
of Medicaid eligibility. Journal of Human Resources 51 (3): 556–88.
Wherry, Laura, Sarah Miller, Robert Kaestner, and Bruce Meyer. 2018. Childhood Medicaid coverage and
later life health care utilization. Review of Economics and Statistics 100 (2): 287–302.
Winkelman, Tyler N. A., and Virginia W. Chang. 2018. Medicaid expansion mental health and access to
care among childless adults with and without chronic conditions. Journal of General Internal Medicine
33 (3): 376–83.
Zavodny, Madelyn, and Marianne Bitler. 2010. The effect of Medicaid eligibility expansions on fertility.
Social Science & Medicine 71 (5): 918–24.