Copyright © 2011 by the Cato Institute.

All rights reserved.

Cover design by Jon Meyers.

Printed in the United States of America.

Cato Institute
1000 Massachusetts Ave., N.W.
Washington, D.C., 20001
www.cato.org
 Contents

Executive Summary 1

Introduction 3
Medicaid: Programs, Coverage, and Financing 7
State Population Projections 8
State Medicaid Spending Projections 9
Enrollment Projections 11
Alternative Federal Cost-Sharing Scenarios 15

Conclusion 18

Appendix 19
Methodology for Projecting Texas Medicaid Expenditures under PPACA 19
A1. Methodology for Projecting Medicaid Expenditures in Texas 19
A2. Medicaid Eligibility Criteria 21
A3. Medicaid Eligibility, Enrollment, Recipiency, and Average Benefits per Enrollee 23

Notes 34
 Executive Summary

Unless it is repealed, the Patient Protection fornia, Florida, Illinois, New York, and Texas.
and Affordable Care Act of 2010 promises to State Medicaid spending is projected to increase
increase state government obligations for Med- considerably even without PPACA in California,
icaid by expanding Medicaid eligibility and in- Florida, and Texas, with smaller increases in Illi-
troducing an individual health insurance man- nois and New York. With PPACA, projected
date for all U.S. citizens and legal permanent spending is actually reduced for California,
residents. Once PPACA becomes fully effective while spending increases are positive and large
in 2014, the Medicaid benefits of those who for Florida and Texas. Both Illinois and New
become newly eligible and enroll into Medi- York have the potential for considerably higher
caid will be almost fully covered by the federal enrollments and increased expenditures.
government through 2019, with federal finan- My estimates of the states’ PPACA Medic-
cial support expected to be extended thereafter. aid burdens are considerably larger than those
But PPACA provides states with no additional reported elsewhere, such as in the Kaiser Fam-
federal financial support for new enrollees among ily Foundation’s study, which appears to have
those eligible for Medicaid under the old laws. That used fixed enrollment rates for new- and old-
makes increased state Medicaid spending from eligibles based on 2007 data. In this study, the
higher enrollments by “old-eligibles” virtually individual mandate’s impact depends on his-
certain as they enroll in Medicaid in response torical enrollment trends—stronger where en-
to the individual mandate to purchase health rollment rates were low or declining, weaker
insurance. where they were high and increasing. Thus,
This study estimates and compares poten- methodological differences may underlie the
tial increases in Medicaid expenditures from sizable differences in estimates of states’ addi-
PPACA by the five most populous states: Cali- tional costs from PPACA.

Jagadeesh Gokhale is a Cato Institute senior fellow and member of the Social Security Advisory Board. He works
on U.S. fiscal policy and entitlement reforms. Gokhale is the author of Social Security: A Fresh Look at
Reform Alternatives, published in 2010 by the University of Chicago Press.
 and aging effect by spurring Medicaid enroll- State
Introduction ments, thereby reinforcing upward pressure expenditures
on health care expenditures and transmit-
This study focuses on the effect of the ting downward pressure through the state on items such
Patient Protection and Affordable Care Act budget on other public services. as infrastructure
(PPACA) on the future growth of states’ In the absence of PPACA, Medicaid
General Revenue Medicaid spending ob- spending growth in Illinois and New York
and other public
ligations.1 Ever since Medicaid was intro- would be relatively slower than in the other services have
duced in the mid-1960s as a key element of three states. The reason, again, lies in their been constrained
President Lyndon Johnson’s Great Society much slower projected population growth
agenda, state expenditures on items such as during the next two decades. However, the by rapid growth
infrastructure, education, and other public introduction of PPACA will provide a much in state Medicaid
services to maintain economic competitive- stronger impulse for Medicaid expenditure obligations.
ness have been constrained by rapid growth growth in these two states. The main reason
in state Medicaid obligations. PPACA was is Illinois’s and New York’s low and declin-
enacted in March 2010 and expands states’ ing trends in enrollment rates among key
Medicaid funding burdens yet again. It re- groups that are eligible for Medicaid. The in-
quires states to maintain current eligibility troduction of PPACA’s individual mandate,
levels and also expands eligibility for Medi- combined with public awareness campaigns
caid benefits to additional categories of peo- to drive home the importance of comply-
ple and to those with incomes both above ing with the mandate and the availability of
and below the federal poverty level (FPL). subsidized health coverage under Medicaid,
The federal government will bear nearly all is likely to cause many more individuals to
of the cost of providing Medicaid coverage to sign up for Medicaid coverage rather than
these newly eligible individuals. The effect of remain uninsured or else purchase costlier
PPACA’s “individual” mandate, however, will private health insurance. Indeed, some peo-
be to induce additional enrollments among ple may also terminate their existing private
those who are already eligible for but not en- health coverage and enroll in Medicaid in
rolled in Medicaid, thereby increasing state response to such campaigns.
Medicaid spending. Finally, PPACA increases As described in the Appendix, the estima-
uncertainty about future escalations in state tion of Medicaid expenditure projections
Medicaid expenditures through the possibil- carries forward historical Medicaid eligibil-
ity that surging federal deficits and debt will ity, enrollment, recipiency, and per recipi-
force a reduction of federal financial support ent benefit rates into the future, separately
beyond 2019 for those made newly eligible for each state and for detailed demographic
for Medicaid. and special-eligibility population groups.
This study projects future Medicaid ex- The calculations are first implemented by
penditures for the five most populous states— excluding the effects of PPACA. If recent
California, Florida, Illinois, New York, and state-specific trends in population growth,
Texas2—both with and without PPACA, there- Medicaid eligibility, enrollments, benefit re-
by revealing the burden that the law imposes ceipt, and Medicaid benefits per beneficiary
on these state budgets. were to continue into the future:
The results differ across the five states.
They suggest that even without PPACA, ●● California’s general-revenue (GR) fund-
Medicaid expenditures will soar in Califor- ed Medicaid expenditures would almost
nia, Florida, and Texas, partly because their double from $19.4 billion in 2008 to
populations are projected to grow and age $35.2 billion by 2020. Medicaid expen-
rapidly. For these three states, PPACA is pro- ditures will continue to increase during
jected to compound the population growth the 2020s, amounting to almost $60 bil-

3
 lion per year by the end of that decade. lion in 2008 to $32.9 billion by 2020
From 2010 forward, the average annual and to $37.1 billion by 2030. The two-
(nominal) Medicaid expenditure growth decade projected spending growth
rate is projected to be almost 9 percent rate beginning in 2010 is 3.7 percent
through 2020, slowing to 5.5 percent per per year, slower than New York’s his-
year thereafter.3 The 2010–30 projected torical annual (nominal) GDP growth
Medicaid spending growth rate of 7.2 of 4.5 percent.
percent is considerably faster than Cali-
fornia’s average annual (nominal) GDP Thus, even if PPACA had not been enact-
growth rate of 5.2 percent per year.4 ed, projected growth in Medicaid spending in
●● Florida’s GR-funded Medicaid expen- California, Florida, and Texas would be on an
ditures would double from $6.3 bil- unsustainable trajectory—if judgment is based
lion in 2008 to $12.6 billion by 2020 on projected Medicaid expenditure growth
and would increase to $19.5 billion by relative to historical experience in state GDP
2030. Medicaid expenditures will grow growth rates. On that basis, Medicaid expen-
rapidly through 2020 at 7.9 percent per ditures projected without PPACA would be
year, and the growth rate will slow to sustainable in Illinois and New York.
PPACA’s health 4.5 percent per year during the 2020s. The results suggest a positive associa-
insurance The two-decade projected Medicaid tion between economic growth and growth
mandate implies spending growth rate of 6.2 percent per in Medicaid expenditures. States such as Il-
year is appreciably faster than Florida’s linois and New York with slower population
that Medicaid historical average annual GDP growth and economic growth rates experience slow-
spending rate of 5.8 percent per year. er growth in Medicaid expenditures, attrib-
●● For Texas, GR-funded annual Medi- utable to slower growth in Medicaid eligibil-
increases caid expenditures would grow from ity, enrollment, and benefit claiming rates
from the new $8.5 billion in 2008 to $18.0 billion and benefit amounts. States that experi-
law would by 2020 and to $32.5 billion by 2030. enced more rapid population and economic
Medicaid expenditures are projected growth since the mid-1990s and that appear
be especially to grow at 9.3 percent per year between likely to continue growing relatively faster
pronounced in 2010 and 2020. The two-decade pro- are likely to experience more rapid growth in
Illinois and jected annual Medicaid expenditure Medicaid expenditures, attributable to high
growth rate through 2030 equals 7.7 and rapid growth in Medicaid eligibility, en-
New York. percent—far exceeding historical an- rollments, and benefits.
nual (nominal) GDP growth in Texas Adding PPACA’s expansion of eligibility
of 5.9 percent. for Medicaid coverage will increase future
●● Of the five states considered here, Illi- Medicaid expenditures in all states. How-
nois has the smallest Medicaid expen- ever, PPACA’s health insurance mandate im-
ditures. It also has the lowest Medi- plies that Medicaid spending increases from
caid spending growth rates: Illinois’s the new law would be especially pronounced
Medicaid expenditures are projected in Illinois and New York—states with the
to increase from $5.8 billion in 2008 to smallest capacity to fund the increases be-
$6.9 billion by 2020, and to $7.6 billion cause these two states are likely to continue
by 2030. The two decade expenditure their slower historical growth experience in
growth rate is projected to be 3.0 per- the future as their populations remain stag-
cent per year, well within the historical nant or decline:
rate of annual (nominal) GDP growth
of 3.9 percent. ●● The projected number of new enroll-
●● New York’s Medicaid expenditures ees among old-eligibles from PPACA
are projected to grow from $23.8 bil- in 2020, calculated as a percentage

4
 of total projected enrollments with- ●● Percentage increases in projected GR
out PPACA in 2020, is 21.2 percent in Medicaid expenditures under PPACA
Illinois and16.8 percent New York, as compared with projected spending
compared with 1.9 percent California, levels without PPACA in the year 2030
16.3 percent Florida, and 13.4 percent are also striking: spending increases
in Texas. for Illinois (34.3 percent) and New
●● Estimates of enrollment increases York (31.5 percent) from introduc-
among old-eligibles in 2030 are even ing PPACA are much larger than for
more pronounced, with Illinois (23.3 California (–1.4 percent), Florida (22.3
percent) and New York (23.5 percent) percent), and Texas (7.9 percent).
projected to experience higher enroll-
ment increases than California (2.7 This result arises, in part, because the po-
percent), Florida (17.3 percent), and tential under PPACA for additional enroll-
Texas (11.1 percent). ments—relative to enrollments projected by
excluding PPACA—are almost exhausted by
These enrollment increases will directly the mid-2020s for California, Florida, and
lead to higher GR Medicaid expenditures, Texas. In Illinois and New York, however,
if new enrollees claim benefits at the same enrollments in key age and eligibility groups
rates as those projected to be enrolled in were stable or declining during 2000–08,
Medicaid irrespective of PPACA: which means the potential for increases in
enrollments driven by the health insurance
●● Percentage increases in projected GR mandate persists for much longer in these
Medicaid expenditures in the year two states.
2014 under PPACA compared with The potential for magnified state budget
spending projected without PPACA in pressures can be appreciated by comparing
the same year are sizable in four states: differences in cumulative Medicaid expendi-
22.2 percent for Illinois, 6.4 percent for tures over the first 10 years of the new law’s
New York, 9.0 percent for Florida, and implementation beginning in 2014. The
13.5 percent for Texas. Only in Cali- spending estimates, both with and without
fornia is the change negative—2.9 per- PPACA, are stated relative to a flat spend-
cent—because enrollment rates among ing baseline: projected spending for 2014
old-eligibles are already very high multiplied by 10 to produce the 10-year flat
whereas savings from uncompensated spending total for each state. The results are
care are estimated to be sizable.5 consistent with those summarized above: The 10-year
●● Projected GR Medicaid expenditures the 10-year GR baseline flat expenditures Medicaid
in the year 2020 (the seventh year of are $262.6 billion for California, $98.4 bil-
PPACA’s Medicaid mandate) in Cali- lion for Florida, $64.2 billion for Illinois, expenditures
fornia, Florida, and Texas are 0.2 per- $287.1 billion for New York, and $126.8 bil- projection
cent, 20.1 percent, and 13.3 percent lion for Texas.
larger, respectively, with PPACA than Without PPACA, the 10-year Medicaid
carries forward
Medicaid expenditures in 2020 pro- expenditures are projected to be larger than historical trends
jected without PPACA. the flat-spending baseline after 2014 in all of of generally
●● Percentage increases in projected GR the five states. This is because the projection
Medicaid expenditures under PPACA in carries forward historical trends of generally increasing
the year 2020 are 30.5 percent and 19.8 increasing eligibility, enrollments, benefit eligibility,
percent larger in Illinois and New York, claim rates, and increases in Medicaid expen- enrollments,
respectively, compared with Medicaid ditures per beneficiary:
spending projections for the same year and benefit
without PPACA. ●● Percentage increases in 10-year expen- claim rates.

5
 The federal ditures without PPACA relative to the growth rate of 3.9 percent. New York’s
budget is flat-spending baseline in California, annual Medicaid growth rate over the
Florida, and Texas are 26.2 percent, same period would increase from 3.7
already under 20.9 percent, and 31.7 percent, respec- percent to 5.2 percent, also above its
considerable tively. historical gross state product growth
●● Percentage increases in 10-year expen- rate of 4.5 percent per year.
strain without ditures without PPACA relative to the
additional flat-spending baseline are smaller for Finally, under PPACA, the federal govern-
burdens Illinois and New York, at 5.1 percent ment is to pay the full cost for those newly
and 10.8 percent, respectively. made eligible for Medicaid during the first
imposed by ●● With PPACA, the increase in 10-year three years (2014–16). Under the new law,
PPACA. expenditures relative to the flat- the marginal federal cost-sharing rate for newly
spending baseline is 25.4 percent for eligible Medicaid enrollees would be gradu-
California—a small decline compared ally reduced from 100 percent to 92.8 per-
with the increase in spending of 26.2 cent by 2019.6 The standard expectation (or
percent without PPACA beyond the assumption) among budget experts is that
10-year flat-spending level. the marginal cost-sharing rate will remain
●● GR Medicaid spending projected un- at 92.8 percent after 2019. However, the
der PPACA is higher compared with federal budget is already under consider-
flat spending by 41.6 percent for Flori- able strain, with unprecedented and unsus-
da and by 48.7 percent for Texas. These tainable budget deficits projected through
increases are sizable compared with 2019 and beyond. That puts all programs
those without PPACA. funded out of federal general revenues at
●● The with-PPACA increases, relative to risk, including Medicaid support for states.
flat spending, are 34.6 percent for Illi- To account for a possible further reduc-
nois and 29.1 percent for New York. tion in federal marginal cost-sharing for
These figures are not as large cumu- newly eligible Medicaid beneficiaries, Med-
latively as those for Florida and Tex- icaid’s spending-time profile for each state
as, but nonetheless are considerably is projected under alternative assumptions
larger compared with spending in- regarding federal financial participation
creases without PPACA. beyond 2019. For instance, assuming that
●● In dollar terms, the 10-year spending federal financial support for newly eligible
increase from PPACA (compared with Medicaid beneficiaries is gradually reduced
without it) is highest for New York, after 2019 at a rate consistent with making
primarily because of its high health it equal to the standard Federal Medical
care costs and Medicaid benefits per Assistance Percentage, after 10 years (after
enrollee. Additional enrollments of 2028) states’ GR Medicaid expenditures will
old-eligibles post-PPACA will cause a increase by even more:
rapid increase in total additional Medi-
caid spending. ●● By 2030, Illinois and New York will
●● Figure 1 shows that for both Illinois both spend about 45 percent more,
and New York, the growth of annual respectively, compared with expen-
(nominal) Medicaid expenditures dur- ditures projected without PPACA—
ing 2010–30 increases well above sus- much more than the 34 percent and
tainable rates as a result of PPACA. For 32 percent increases, respectively, from
Illinois, the two-decade expenditure introducing PPACA but maintaining
growth rate increases from 3.0 percent marginal federal cost-sharing at rates
to 4.5 percent—higher than its histori- as scheduled for 2019.
cal (1997–2008) gross state product ●● For California, Florida, and Texas, the

6
Figure 1
Medicaid Growth with and without PPACA and Historical Growth
in Gross State Products
Percent

Source: Author’s calculations based on data from the Bureau of Economic Analysis, Current Population Survey, and
Medicaid Statistical Information System.

spending increases from eliminating be sustained or rejected in court cases filed

marginal federal cost-sharing are 4.1 by several states to challenge the mandate’s
percent, 46.3 percent, and 20.9 percent, constitutionality. If it is rejected, large parts
respectively. These numbers, again, are of PPACA could be struck down with it, or
considerably larger than the spending will become unworkable and require a radi-
increases of −1.4 percent, 22.3 percent, cal overhaul. State lawmakers face consider-
and 7.9 percent, respectively, under able uncertainty about increased Medicaid
PPACA with marginal federal cost- spending commitments that they face be- With the
sharing maintained at their 2019 rates cause of PPACA—commitments that they
through the end of the projection hori- must either accept and prepare for, or take enactment of
zon of 2030. steps to avoid, perhaps by opting out of PPACA, concern
Medicaid. about runaway
With the enactment of PPACA, concern
about runaway Medicaid spending is mo- Medicaid: Programs, Coverage, Medicaid
tivating many state policymakers to find and Financing spending is
ways to restrain Medicaid expenditures. One Medicaid programs are a part of state wel-
method receiving serious consideration is to fare programs that provide subsidized health
mo­tivating
opt out of Medicaid altogether—an option care to low-income and medically needy in- many state
that has always existed under the Social Se- dividuals. Federal rules determine the basic policymakers
curity Act—and to provide basic health cov- coverage criteria that take account of house-
erage to low-income and medically needy hold income and asset thresholds, medical to find ways
groups financed exclusively out of GR funds. conditions, and eligibility for special groups to restrain
As of this writing, however, it is uncertain such as pregnant women, children, disabled
whether PPACA’s individual mandate to individuals, the aged, and so on. However,
Medicaid
purchase health insurance will ultimately state governments, at their discretion, may expenditures.

7
 Beginning extend coverage to broader groups—by speci- remain at their 2011 values when projecting
in 2011, fying higher income and asset eligibility future expenditures. Those values are also
thresholds than the federally mandated lev- applied for determining federal and state
FMAP rates els and by including additional groups based cost-sharing for beneficiaries who are eligi-
will revert on medical conditions, family resources, and ble for Medicaid under the old laws (without
so on. Many states cover children and preg- PPACA) and are projected to become new
close to their nant women even if their incomes are above enrollees in Medicaid in response to the in-
pre-recession the state’s eligibility levels but are deemed dividual mandate to purchase health insur-
values, which insufficient to meet the medical costs that ance.9
they face. Other groups covered under op- PPACA mandates new spending commit-
will require tional programs in many states include non- ments for state governments under Medi-
states to come up disabled children and their related caretak- caid. All five states examined here have con-
with additional ers, pregnant women, the aged, blind, dis- stitutional balanced-budget requirements,
abled, and others with medical expenditures either on the state legislature or on the gov-
financing. exceeding their incomes.7 ernor’s budget submission. Thus, increased
State Medicaid programs pay for a wide spending commitments from entitlements
range of health care services including physi- such as Medicaid that are difficult to re-
cian, hospital (in- and out-patient), lab, nurs- verse, and revenue losses during the recent
ing, home health care, pharmacy costs, and recession, are worsening pressures on other
more. Usually, the federal government pro- budget items. Many states, including the five
vides matching funds to share state Medic- evaluated in this study, are projecting persis-
aid costs. Federal cost-sharing is implement- tent budget gaps during 2011 and 2012 that
ed using the FMAP formula, which is based must be addressed by increasing revenues
on each state’s per capita income relative to and reducing state public services. Medicaid
that of the nation overall.8 The statutory benefits are unlikely to be spared as federal
minimum FMAP percentage for all states is financial assistance is reduced after 2010.
50 percent, the maximum being 83 percent.
The average FMAP value across all states is State Population Projections
about 59 percent. During 2009–10, FMAP For the five states under consideration,
rates were higher than normal because of the Table 1 shows population growth rates cal-
temporary FMAP enhancement enacted as culated based on projections of the U.S.
part of the American Recovery and Reinvest- Census Bureau. Through 2020, California,
ment Act of 2009. Florida, and Texas are projected to experi-
California, Illinois, and New York have ence significantly higher population growth
pre-ARRA FMAP rates set to 50 percent; they rates than Illinois and New York. The growth
are among the states with the highest in- rates in the former three states are larger in
comes per capita. Pre-ARRA FMAP rates for all age categories, and especially among their
Florida and Texas are higher, 55.45 percent retiree populations. Illinois and New York
and 60.56 percent, respectively. The ARRA- are projected to have declining populations
inclusive (year-end) values are about 10–15 among the working population—those aged
percentage points higher for all states during 19 through 64. Beyond 2020, population
2009 and 2010, implying a smaller Medicaid growth is projected to increase in the three
funding burden. Beginning in 2011, howev- already rapidly growing states, whereas it is
er, FMAP rates will revert close to their pre- expected to decline in the two slow-growing
recession values, which will require states to states. Indeed, New York’s overall population
come up with additional financing, as Med- growth is expected to be negative during this
icaid caseloads have continued to increase century’s third decade.
during 2009 and 2010. For all states, federal Examining each state’s demographic pro-
Medicaid cost-sharing rates are assumed to file and dynamics provides clues for under-

8
Table 1
Projected Population Growth (annualized, percent)

Age 2010–20 2020–30

Category CA FL IL NY TX CA FL IL NY TX
0–18 0.69 1.47 0.12 −0.15 1.33 0.64 1.70 0.02 −0.11 1.24
19–64 0.51 1.12 −0.13 −0.27 0.85 0.56 1.21 −0.30 −0.63 1.24
65+ 2.91 3.40 1.82 1.71 3.15 2.95 4.29 1.95 1.88 3.28
All 0.86 1.64 0.20 0.06 1.26 0.96 2.05 0.15 -0.05 1.53

Source: Author’s calculations based on data on population projections from the Census Bureau. As the text explains,
results for Illinois and New York (shaded) are qualitatively different from those for the other three states.

standing the results on state-specific Med- Population projections for Illinois and
icaid enrollment and expenditure changes, New York (shaded) exhibit considerably
both with and without PPACA, that are re- greater constancy, both in population size
ported later in this study. That is because each and age composition, and suggest that
state’s projected total Medicaid expenditures Medicaid expenditures will not increase as PPACA
are anchored by its population projections by rapidly in these two states compared with broadens
age and gender (see Appendix). State-specific the other three states with more robust pop- Medicaid
population projections are obtained directly ulation growth and faster aging population.
from the Census Bureau.10 These projections eligibility by
are based on the 2000 Census and use fertil- State Medicaid Spending Projections increasing
ity, mortality, and migration trends for each PPACA broadens Medicaid eligibility by
state to project their populations forward increasing income thresholds for children
income
through 2030. and adults. Children living in families with thresholds
The Census-projected population age dis- incomes less than 138 percent of the FPL for children
tributions for the five states are shown in Fig- (including PPACA’s new 5 percent income
ure 2. The figure indicates that demographic disregard) will now qualify for Medicaid. In and adults.
changes in terms of changes in the size and addition, adults with or without qualifying
age composition are occurring more rapidly children are also made eligible under the new
in California, Florida, and Texas. Florida FPL threshold. Expanded eligibility levels
stands out for its rapid increase in the num- under PPACA will increase state Medicaid
ber and proportion of elderly residents—as expenditures. But it will not significantly
expected, because the absence of income increase state GR Medicaid expenditures, at
taxes makes it a popular destination for re- least in the short term, because of the high
tirees. The California population profiles marginal cost-sharing provided by the federal
show substantial increases in the number of government for individuals made newly eligi-
children, young adults, and the elderly. Cali- ble for Medicaid. GR-funded state Medicaid
fornia’s younger populations are expected to expenditures would not increase by much
grow, partly because of continuing migration if enrollment rates among those eligible for
from the nation’s eastern and mid-western Medicaid under the old laws remain low.
regions. Similar to California, Texas’s popu- That possibility appears unlikely, however,
lation profile is growing throughout the age because the key goal of PPACA is to reduce
distribution, but the increase in the young- rates of non-insurance extensively and inten-
adult population is not as pronounced. The sively—that is, by expanding Medicaid eligi-
significant increases in these populations bility and by facilitating enrollment and con-
suggest growing Medicaid expenditures even ducting widespread enrollment drives that
without PPACA. induce non-enrolled old-eligibles to sign up

9
Figure 2
State Population Projections by Age, 2005–30

California New York

+ +

Florida Texas

+ +

Illinois

Source: United States’ Census Bureau.

10
Table 2
Enrollment Increases Induced by PPACA (thousands of people)
New Enrollees (Newly Eligible) New Enrollees (Old-Eligibles)
Year CA FL IL NY TX CA FL IL NY TX

2014 2,985 2,315 979 2,059 2,118 204 624 518 467 858
2020 3,078 2,434 925 2,001 2,329 257 739 620 677 820
2030 3,479 3,125 886 1,917 2,729 412 937 767 1,014 864
Percent
2014 24.9 55.4 36.0 42.2 40.8 1.7 14.9 19.0 12.3 16.5
2020 23.3 53.6 31.7 41.3 38.0 1.9 16.3 21.2 16.8 13.4
2030 22.6 57.6 26.9 38.6 34.9 2.7 17.3 23.3 23.5 11.1

Source: Author’s calculations based on the Medicaid Statistical Information System and the Current Population
Surveys. As the text explains, results for Illinois and New York (shaded) are qualitatively different from those for
the other three states.

for Medicaid. Moreover, the individual man- burden on beneficiaries compared with their
date, even when it does not result in a penalty current employer-provided or privately pur-
for non-insurance, will induce Americans to chased health insurance coverage. The added
sign up for health insurance—Medicaid if burden states bear will depend on the suc-
they are eligible for it—simply out of a desire cess of those promotional efforts.11
to comply with the nation’s laws. In calculating enrollments under PPACA,
Though it is uncertain whether the indi- it is assumed that enrollments by those new-
vidual mandate will survive court challeng- ly eligible will either follow the same enroll-
es from many states, this study’s Medicaid ment rates as those presently eligible or they
spending growth estimates are constructed will enroll at the rate of those with no other
under the assumption that it will. The man- health insurance, depending on which rate is
date and health insurance enrollment drives larger. A similar method is followed for those
will induce an increase in enrollment by who are eligible for Medicaid under the old
those who were eligible under the old laws laws but who are not enrolled in Medicaid.
but who were not enrolled in Medicaid or Applying these rules yields a sizable increase The man­date and
any other health insurance plan. Although in enrollments in 2014.
PPACA provides full federal support for Enrollment Projections. Even without health insurance
newly eligible Medicaid enrollees through PPACA, Medicaid enrollments are project- enrollment
2019, it provides zero additional support for ed to increase substantially in California,
new enrollees among old-eligibles. States will Florida, and Texas. The top panel of Table
drives will induce
bear the cost of these old-eligibles according 2 shows projected increases in Medicaid en- an increase in
to their (post-ARRA) FMAP rates. rollments as a result of PPACA in the five enrollment by
PPACA will be implemented with special states being considered. The five columns
efforts to advertise the availability of Medic- on the left show the number (in thousands) those who were
aid’s health care coverage options to newly of new enrollees that would result under eligible under the
eligible populations to increase the enroll- PPACA in selected future years (2014, 2020, old laws but who
ment rates among newly eligible children and 2030). These are counts of enrollees esti-
and adults. However, the enrollment facili- mated from among those made newly eli- were not enrolled
tation will also induce some old-eligibles gible for Medicaid benefits under PPACA’s in Medicaid or
to switch from non-Medicaid to Medicaid broader eligibility criteria. The lower panel
coverage because the latter is subsidized and (again of the five columns on the left) shows
any other health
imposes zero (or a much smaller) financial the increase that these new enrollees repre- insurance plan.

11
 PPACA sent as a percent of projected enrollees in the tered around 20 percent—in the various years
adds to states’ same years in a world without PPACA. The shown in Table 2. This occurs because these
increases in Medicaid caseloads are sizable— two states have smaller projected enrollment
fiscal burden, ranging from the mid-20 percents for Cali- rates without PPACA, leaving more scope for
primarily fornia to the mid-50 percents for Florida enrollment increases as the PPACA laws take
and New York. However, the costs of these effect in 2014. This means that in each future
by bringing increases in Medicaid caseloads will be paid year, Medicaid enrollments would be between
forward in time for almost entirely out of marginal federal 10 to 25 percent higher in all the states exam-
enrollment cost-sharing. ined except in California.
The five columns on the right of Table Table 3 shows GR Medicaid spending in-
increases that 2 show the number (in thousands) of new creases in the five states, comparing projec-
would likely have enrollees among old-eligibles who are pro- tions with and without PPACA in selected
occurred later. jected to enroll as a result of PPACA. Effec- future years. Spending increases are gener-
tively, applying for health insurance through ally larger than enrollment increases because
state-operated health insurance exchanges the historical increases in benefits per ben-
would reveal if the applicant is eligible for eficiary are projected forward in time and
Medicaid under the old eligibility rules augment expenditure growth from larger
(without PPACA). If so, that person would enrollments.
be directed or advised to sign up for Medi- The top panel of Table 3 shows GR Medi-
caid unless he or she desired an alternative caid expenditures (in billions of dollars) for
insurance source. The estimation procedure selected future years constructed with and
mentioned above—taking the larger of the without PPACA. The two rows in the middle
uninsured rate or the existing enrollment of Table 3 show the percentage increases in
rate observed for old-eligibles—produces the states’ GR Medicaid expenditures, separate-
estimates reported in Table 2.12 ly with and without PPACA, between 2014
The lower panel shows the increase that and 2020 and between 2014 and 2030:
new enrollees among old-eligibles represents
as a percent of Medicaid enrollees in the same ●● In California, Florida, and Texas, spend-
years projected without PPACA. The esti- ing growth is sizable between 2014 and
mates are much smaller compared with those 2020, even without PPACA. The in-
in the columns on the left. But the Medicaid creases range from about 28 percent to
benefits of these new enrollees, under the about 42 percent for these three states.
new law, would be paid for entirely out of ●● In Illinois and New York, cumulative
state budgets: PPACA provides zero federal changes through the year 2020 are
financial support on account of new enroll- much smaller, ranging between 7 per-
ees among old-eligibles. cent to about 15 percent.
It is noteworthy that California is projected ●● Through 2030, cumulative spending in-
to experience very small increases in new en- creases even without PPACA are, again,
rollments among old-eligibles into Medi-caid. much larger for California, Florida, and
That’s because California already has very Texas—doubling or more even without
high enrollment rates among Medicaid eligi- PPACA. For Illinois and Florida, the
bles and there is not much scope for expand- cumulative spending increases are rela-
ing enrollments among old-eligibles under tively modest—ranging between about
PPACA. Florida and Texas are projected to ex- 18 and 29 percent.
perience Medicaid enrollment increases in the
10 to 16 percent range in years beginning in Without PPACA, GR spending projec-
2014. However, Illinois and New York (shad- tions exhibit very different trends in the five
ed) are projected to gain Medicaid enrollments states, increasing much more rapidly for Cali-
among old-eligibles much more rapidly—clus- fornia, Florida, and Texas, and more slowly

12
Table 3
Projected General Revenue Medicaid Expenditures with and without PPACA (billions of dollars)

CA FL IL NY TX
Without With Without With Without With Without With Without With
Year PPACA PPACA PPACA PPACA PPACA PPACA PPACA PPACA PPACA PPACA
2008 19.4 - 6.3 - 5.8 - 23.8 - 8.5 -
2014 26.3 26.5 9.8 10.7 6.4 7.8 28.7 30.5 12.7 14.4
2020 35.2 36.8 12.6 15.1 6.9 8.9 32.9 39.4 18.0 20.3
2030 59.9 62.3 19.5 23.8 7.6 10.2 37.1 48.9 32.5 35.1
Percent Increase from Spending Projections Without PPACA for 2014
2020/2014 34.1 34.4 27.8 53.6 6.8 39.3 14.5 37.2 41.7 60.5
2030/2014 128.0 124.8 97.9 141.9 17.8 58.2 29.4 70.2 156.6 177.0
Percent Change due to PPACA
2014 -2.9 9.0 22.2 6.4 13.5
2020 0.2 20.1 30.5 19.8 13.3
2030 -1.4 22.3 34.3 31.5 7.9

Source: Author’s calculations based on the Medicaid Statistical Information System, Current Population Surveys, and CMS-64 reports. As the text
explains, results for Illinois and New York (shaded) are qualitatively different from those for the other three states.

for Illinois and New York. A salient reason ditures in California are negative—the
is that the former states are projected to ex- result of savings from uncompensated
perience growing populations among age- care (reductions in state Dispropor-
gender groups that have historically high tionate Share Hospital expenditures)
Medicaid eligibility and utilization rates, dominating those from increased Medi-
whereas the latter two states have stagnant or caid enrollments by old-eligibles.13
declining populations among those groups. ●● Medicaid expenditures in Florida and
With PPACA, however, spending increases Texas increase more rapidly as a result
are much larger, and the increases are espe- of PPACA, ranging between 8 percent
cially sizable for Illinois and New York. and 22 percent in the years shown.
Another way to measure PPACA’s effect ●● Illinois and New York (shaded) are pro-
on spending is shown in the bottom panel of jected to have higher spending increas-
Table 3. This panel shows the percent increase es as a result of PPACA—increasing by
in projected Medicaid expenditures due to up to the mid-30 percents by 2030.
PPACA in selected years. Here, the story is re-
versed compared with projected spending in- This result emerges because, consistent
creases over time: states with high spending with historical data, Illinois and New York are
increases over time without PPACA exhibit projected to have lower eligibility and enroll-
low projected spending increases from intro- ment rates than California, Florida, and Texas
ducing PPACA and states with low spending (without PPACA) and those rates are projected
growth over time without PPACA exhibit rel- to increase more slowly in both of those states.
atively more rapid spending increases from With the introduction of PPACA’s individual
introducing PPACA: mandate, however, many among the old-
eligibles who do not have health insurance will
●● PPACA’s effects on Medicaid expen- be induced to acquire it. If applying for health

13
Table 4
Projected General Revenue Medicaid Spending Growth by Age and Special Eligibility Categories without PPACA
(annualized, percent)

Age and Special 2010–20 2020–30

Eligibility Category CA FL IL NY TX CA FL IL NY TX
0–18* 9.0 11.1 7.4 7.0 12.0 4.1 7.4 2.8 2.6 7.8
19–64* 8.6 10.4 -3.1 5.5 5.9 4.6 7.2 -0.8 -2.3 3.5
65+* 12.5 4.6 6.3 4.0 5.5 8.4 -2.8 1.9 -3.3 2.3
Medically Needy 6.3 9.1 2.0 6.4 -0.7 5.0 6.6 -2.1 1.2 -0.9
Breast and Cervical
Cancer Act 16.3 20.7 22.7 15.6 16.8 9.8 11.7 14.2 8.4 10.1
Foster Care 10.3 9.4 2.4 1.5 13.6 7.2 5.5 2.1 5.2 8.5
Family Planning 13.4 -30.0 9.4 10.0 7.6 7.5
Disabled Adults 7.0 6.9 5.8 7.8 7.6 2.4 2.5 -0.1 2.4 3.8
Waivers 4.2 6.4 3.4 3.4 4.4 0.9 3.3 0.1 -0.7 1.4
All 8.9 7.9 5.0 6.3 9.3 5.5 4.5 1.0 1.2 6.1
*Excludes people eligible under other special categories. Spending growth from PPACA is shown without netting out state saving from uncompensated
care (DSH).
Source: Author’s calculations based on the Medicaid Statistical Information System, Current Population Surveys, and CMS-64 reports. As the text explains,
results for Illinois and New York (shaded) are qualitatively different from those for the other three states.

insurance at state-operated exchanges reveals Table 4 shows Medicaid spending growth

eligibility for Medicaid under the old eligibil- rates by age group and special eligibility cat-
ity rules, enrollment in Medicaid will rise, usu- egories. The growth rates by age categories
ally leading to benefit claims. In addition, the are calculated without PPACA after exclud-
health insurance mandate may cause some ing individuals that qualify for and claim
individuals to realize that obtaining health in- Medicaid benefits under special eligibility
surance through Medicaid is more cost effec- rules. The table shows that the annual aver-
tive than their existing private insurance and age Medicaid spending growth on account
may cause them to shift to Medicaid. Alterna- of children (not including expenditures of
tively, if employers choose to drop coverage, the Children’s Health Insurance Program)
many old-eligibles who are currently insured is projected to be quite high through the
in the private market may enroll in Medicaid. year 2020 in California, Florida, and Texas—
The message from Tables 2 and 3 is that ranging between 9 and 12 percent per year.
states with already high enrollments and rap- In Illinois and New York that growth rate
id growth in Medicaid enrollments would face is not as high. Between 2020 and 2030, the
serious budget problems from rising Medi- growth rate is lower in all states, but espe-
caid expenditures. PPACA adds to their fiscal cially in Illinois and New York (shaded).
burden, primarily by bringing forward in time The same remarks are applicable to Medi-
enrollment increases that would likely have caid spending growth on account of non-
occurred later. Moreover, the additional Medi- disabled adults aged 19–64. However, Illinois
caid spending from PPACA is relatively small. and New York (shaded) will experience low or
However, states with heretofore low Medicaid negative spending growth rates, probably be-
enrollments and slower growth in enrollments cause of the projected decline in their popula-
are likely to experience a substantial increase tions of working-age adults. Among retirees,
in Medicaid expenditures because of PPACA’s growth rates vary across states and over time:
health insurance mandate. rates are highest in California but negative

14
over the long term in Florida and New York. GDP—indicating the precarious condition
Those negative growth rates are probably the of federal finances—through the next 10
result of increasing use of prescription drugs, years. It means that the promised high mar-
payment for which was switched from Med- ginal federal cost sharing for new enrollees
icaid to Medicare during the early 2000s, and among those made newly eligible for Medi-
the consequent decline in the ratios of benefi- caid under PPACA could be reduced rather
ciaries to enrollees in these two states.14 than maintained at the 2019 value of 92.8
Among special-eligibility categories, spend- percent. Two alternative projections are con-
ing growth rates are highest among women structed for states’ Medicaid expenditures
with breast and cervical cancer, followed by assuming reduced marginal federal cost
foster-care children and blind/disabled adults. sharing. The first scenario implements a
Across all categories, projected annual (nomi- gradual reduction in marginal federal FMAP
nal) Medicaid spending growth rates are 8.9 support after 2019 and the second a more
percent for California, 7.9 percent for Florida, rapid reduction of the same. The results are
and 9.3 percent in Texas—much larger than shown in Table 5.
the 5 percent for Illinois and 6.3 percent for The top panel of Table 5 shows projected
New York. During the subsequent decade, an- Medicaid expenditures with PPACA and the
nual (nominal) Medicaid spending growth bottom panel shows the percentage increase
rates are slower, between 4.5 and 6.1 percent in expenditures caused by PPACA for select-
for California, Florida, and Texas, and they are ed years. In each state’s panel in Table 5, the
very small for Illinois (1.0 percent) and New first column shows the “Base Case,” taken
York (1.2 percent) (shaded). The slower spend- from the first panel of Table 3; the second
ing growth after the year 2020 in most cases is column, “Alternative 1,” projects the percent
explained by projected eligibility/enrollment/ increase in each state’s Medicaid spending
beneficiary rates, eventually attaining maxi- under the assumption that the marginal
mum values of 100 percent with no scope for FMAP cost-sharing rate under PPACA is
additional increases. gradually reduced by 1 percentage point per
Alternative Federal Cost-Sharing Scenarios. year until it reaches the standard FMAP rate
The Congressional Budget Office projects applicable for each state; and the third col-
unprecedented federal deficits as a share of umn, “Alternative 2,” projects the percent

Table 5
Medicaid Spending Increases Post-PPACA under Alternative Federal Match Policies

CA FL IL NY TX
Base Base Base Base Base
Year Case Alt 1 Alt 2 Case Alt 1 Alt 2 Case Alt 1 Alt 2 Case Alt 1 Alt 2 Case Alt 1 Alt 2
2014 25.5 25.5 25.5 10.7 10.7 10.7 7.8 7.8 7.8 30.5 30.5 30.5 14.4 14.4 14.4
2020 35.3 35.4 35.7 15.1 15.2 15.7 8.9 9.0 9.1 39.4 39.5 40.5 20.3 20.4 21.0
2030 59.0 60.2 62.3 23.8 25.4 28.5 10.2 10.4 10.9 48.9 50.6 53.9 35.1 36.6 39.3
Percent change over Medicaid spending projections without PPACA
2014 -2.9 -2.9 -2.9 9.0 9.0 9.0 22.2 22.2 22.2 6.4 6.4 6.4 13.5 13.5 13.5
2020 0.2 0.4 1.5 20.1 20.9 25.1 30.5 30.8 32.9 19.8 20.3 22.9 13.3 13.8 16.7
2030 -1.4 0.5 4.1 22.3 30.4 46.3 34.3 37.8 44.6 31.5 36.2 45.2 7.9 12.3 20.9
Notes: Alt 1: Marginal federal match for new-eligibles is reduced by one percentage point each year through 2030. Alt 2: Marginal federal match for
new-eligibles is eliminated by 2028.
Source: Author’s calculations from the Medicaid Statistical Information System, Current Population Surveys, and CMS-64 reports. As the text explains,
results for Illinois and New York (shaded) are qualitatively different from those for the other three states.

15
Table 6
Increases in 10-Year (2014–23) Medicaid Spending with and without PPACA, Relative to a Flat Spending Baseline

CA FL IL NY TX
1. 10-Year Flat-Spending Baseline (Based on 2014 Spending Without PPACA; billions of dollars) 262.6 98.4 64.2 287.1 126.8
2. 10-Year Change from Flat-Spending Baseline to without PPACA (%) 26.2 20.9 5.1 10.8 31.7
3. 10-Year Change from Flat-Spending Baseline to with PPACA (%) 25.4 41.6 34.6 29.1 48.7
4. 10-Year Change from Flat-Spending Baseline to Alternate Scenario 1 (%) 25.7 42.7 35.0 29.7 49.5
5. 10-Year Change from Flat-Spending Baseline to Alternate Scenario 2 (%) 26.8 46.7 36.6 31.8 52.5
6. Memo: 10-Year Flat Enrollment (millions of people) 119.7 41.8 27.2 48.8 51.9
7. Memo: Average State GR Medicaid Spending per Enrollee (dollars) 2,195 2,355 2,358 5,878 2,441

Source: Author’s calculations from the Medicaid Statistical Information System, Current Population Surveys, and CMS-64 reports. As the text explains,
results for Illinois and New York (shaded) are qualitatively different from those for the other three states.

increase under the assumption that federal The sixth row of the table shows flat-lined
marginal cost-sharing is reduced rapidly so enrollments—that is, assuming that enroll-
that it achieves the standard FMAP rate for ments are maintained at the 2014 level for
each state by 2028. Note that the terminal 10 years thereafter. Dividing the 10-year flat
year through which marginal federal cost- spending with the 10-year flat enrollment
sharing rates have been specified under yields the average spending per enrollee, as
PPACA is 2019. shown in the last row of the table. New York
The lower panel of Table 5 shows that is by far the most expensive Medicaid state,
PPACA would reduce California’s GR Medi- spending almost $6,000 per Medicaid en-
caid spending in the year 2030 by 1.4 percent rollee. All of the other four states included in
under the Base Case, but increase it margin- this study are projected to experience much
ally under both alternative scenarios—by smaller expenditures of a little more than
4.1 percent under Alternative 2, where the $2,000 per enrollee—only about one-third of
reduction of federal cost sharing is more New York’s average Medicaid spending per
rapid. For other states, however, spending enrollee.15
increases under the two alternative scenar- The second row of Table 6 shows the per-
ios are much larger. In Florida, for example, centage increase in projected Medicaid ex-
the additional Medicaid spending from penditures if future enrollments and bene-
PPACA would increase from 22.3 percent to fits per beneficiary continue to evolve along
30.4 percent under the first alternative scen- historical trends in each of the age and
ario and to 46.3 percent under the second. special-eligibility categories as described in
Significant increases in additional Medicaid the Appendix. This projection shows that
expenditures from PPACA also arise for Illi- Medicaid expenditures would increase sub-
nois, New York, and Texas, as Table 5 shows. stantially for California (26.2 percent), Flori-
Table 6 provides the overall picture of da (20.9 percent), and Texas (31.7 percent)
Medicaid spending growth during the 10 even without PPACA. Ten-year spending
years after the law is implemented in 2014. increases are projected to be quite low for
The first row of the table shows total 10-year Illi-nois (5.1 percent) and New York (10.8
spending under the assumption that Medi- percent) (shaded) because enrollment ratios
caid’s dollar spending out of state general in these two states have been historically low
revenues is frozen at the 2014 level. That and increased less steeply than in others.
spending is highest for New York, followed The third row of Table 6 shows that
by California, with Texas a distant third. PPACA reduces California’s 10-year Medi-

16
caid spending by a small amount; the in- trends in eligibility, enrollments, beneficiary The overall
crease from the flat-spending with PPACA ratios, and expenditures per beneficiary, as results show
is just 25.4 percent compared with the 26.2 well as PPACA’s Medicaid expansion. Among
percent increase without PPACA. Because the five states examined, PPACA causes the that California,
the ratio of enrollees to old-eligibles is al- steepest rise in spending in Illinois—by Florida, Illinois,
ready quite high in California, there is little almost 30 percentage points—whereas the
scope for PPACA to increase Medicaid expen- steepest cumulative increase (48.7 percent) is
New York,
ditures further on account of new enrollees projected for Texas. California’s cost increase and Texas are
among old-eligibles in California compared is already sizable, but much of its uninsured facing a 10-year
with Florida and Texas.16 Projections for the population would either be ineligible to
latter two states suggest a much larger scope enroll in Medicaid or be in categories that Medicaid
for spending increases under PPACA, as en- would impose higher costs on that state. spending
rollment rates have stagnated, especially These results are at odds with previous increase larger
among the largest category of non-disabled portrayals of how much state expenditures
adults aged 19–64. This is confirmed in would increase under PPACA. The debate than 30 percent.
Table 6, which shows that compared with appears to be informed largely by estimates
the flat-spending level, Florida’s 10-year reported by the Kaiser Family Foundation,
Medicaid costs would escalate by 41.6 per- which projects very small increases in total
cent because of PPACA rather than by just state expenditures during 2014–19. Kaiser’s
20.9 percent without it. In the case of Texas, estimates of state spending increases range
the 10-year total spending would increase by from 0.0 percent for New York to 3.0 percent
48.7 percent under PPACA rather than by for Texas—much smaller than the estimates
just 31.7 percent without it. Thus, the cumu- reported here.17 Kaiser’s methodology ap-
lative impact of carrying forward historical pears to have used future enrollment rates
trends in eligibility, enrollments, beneficiary for new-eligibles and old-eligibles that are
ratios, and benefits per beneficiary lead to not calibrated from historical trends, but
the largest escalation in Medicaid expen- instead are based on calibrations from a
ditures in Texas—both without and with given year (2007) and Congressional Bud-
PPACA—casting a bright spotlight on why get Office assumptions about future enroll-
Texas state policymakers are so highly con- ment rates that are assumed to remain fixed.
cerned with the implications of PPACA for These assumptions would tend to produce
that state’s budget. smaller estimates of enrollment and spend-
In Illinois and New York, the ratio of en- ing on old-eligibles compared with the as-
rollees to eligibles is low and, in some instanc- sumptions made here. For example, in those
es, it has historically declined. Again, this cases where historical trends show consis-
explains why PPACA would lead to sizable tently low or declining enrollments among
spending increases—the individual mandate old-eligibles, the enrollment-increasing ef-
will spur old-eligibles into enrolling under fect of the individual health insurance man-
Medicaid. Table 6 shows that compared date on non-enrolled old-eligibles may be
with the flat spending baseline, PPACA understated under Kaiser’s methodology.
would escalate Illinois’s 10-year expendi- This study does not assume that enrollment
tures by 34.6 percent, rather than by just 5.1 rates among old-eligibles would remain
percent without PPACA. And New York’s fixed at the levels observed in 2007, a year
spending increase would be by 29.1 percent, with no federal mandate to purchase health
not by just 10.8 percent. insurance or widespread public education
The overall results show that all of the five campaigns about the need to comply with
states are facing a 10-year Medicaid spend- the law by obtaining health insurance. This
ing increase larger than 30 percent, resulting study makes the alternative assumption
from a combination of already increasing that the individual mandate would raise

17
 enrollment rates among old-eligibles and, under PPACA than without it. Florida
therefore, tracks changes to historical trends and Texas, however, are projected to have
in eligibility, enrollment, benefit recipiency, larger populations of non-enrolled old-
and average benefits per beneficiary that eligibles: spending without PPACA is pro-
would arise from introducing PPACA. All jected to be 20.9 percent higher in Florida
of these estimates are anchored to nation- compared with holding it at the 2014 level
ally representative historical micro-data for 10 years. With PPACA, it would be 41.6
surveys and Medicaid’s state-wise adminis- percent higher compared with the flat base-
trative information. Thus, methodological line. The corresponding increases for Texas
differences between Kaiser’s and this study’s are 31.7 percent without PPACA and 48.7
calculations may underlie the sizable differ- percent with PPACA. Illinois and New York
ences in estimates of states’ additional costs exhibit historically stable or declining en-
from PPACA, especially the potential addi- rollment rates among old-eligibles, imply-
tions to state Medicaid expenditures from ing higher potential spending on these in-
new enrollments by old-eligibles. dividuals in 2014 and thereafter. Without
PPACA, Medicaid expenditures are project-
ed to increase from the flat-spending base-
The estimates Conclusion line by just 5.1 percent in Illinois and 10.8
of projected percent in New York. With it, however, the
Medicaid Detailed estimates of the effect of PPACA increases in the two states are 34.6 percent
on the budgets of the five most populous and 29.1 percent, respectively.
expenditures may states suggest the law will impose large un- It should be noted that the estimates pre-
be conservative funded mandates to expand Medicaid case- sented here of projected Medicaid expendi-
loads and increase state Medicaid outlays. tures in five states, both with and without
because the States will have to meet this increased finan- PPACA, are based on standard assumptions
effect of future cial burden either through cutbacks in other and methods for extending eligibility, enroll-
shifts from public services or higher tax burdens—both ment, benefit recipiency status, and benefit
of which will exert negative effects on states’ award rates into the future. The estimates
private coverage economies. may be conservative because the effect of
to Medicaid, The results suggest that even without future shifts from private coverage to Med-
post-PPACA, PPACA, Medicaid expenditures would in- icaid, post-PPACA, is not fully incorporated.
crease rapidly in California, Florida, and Texas, The results are also uncertain because future
is not fully each of which has growing populations across trajectories of all of these rates will be af-
incorporated. many Medicaid eligibility and enrollment fected by many factors not considered here:
groups. Medicaid spending increases project- the economic environment, the specific im-
ed without PPACA are relatively small in Illi- plementation of PPACA, the availability and
nois and New York, states whose populations cost of non-Medicaid health insurance for
are projected to remain generally stagnant or low-income individuals, and so on.
to decline during the next two decades. When Supporting such high Medicaid spending
PPACA’s effects on enrollment in Medicaid increases under PPACA, but also irrespec-
are included—especially enrollment by old- tive of PPACA, would require higher taxes
eligibles who would now be directed to enroll or reductions in other public services, both
as a consequence of the new law’s health in- of which appear to state policymakers as
surance mandate—Medicaid will impose large economically undesirable. Concerns about
financial burdens on all five states. runaway Medicaid expenditures are motivat-
The projected cumulative post-PPACA ing policymakers in many states to explore
Medicaid spending increase during 2014–23 ways to restrain Medicaid expenditures, in-
is negative only for California: spending is cluding opting out of Medicaid altogether—
projected to decline by a smaller amount an option that has always existed under the

18
Social Security Act—using alternative state categories of individuals such as children,
programs to provide basic health coverage pregnant women, aged, blind, disabled, and
to low-income and medically needy groups. medically needy individuals qualify for “cate-
Another possibility is to allow the quality of gorical coverage,” even though their incomes
Medicaid-covered health care services to de- and resources exceed federally mandated in-
teriorate in order to prevent the crowd-out come and asset qualification thresholds.
of private health coverage that has histori- First, the total population for the state
cally occurred after every expansion of the in question is calculated by gender, age cat-
Medicaid program. Prospects of such steep egory, income range (f) relative to the federal
increases in Medicaid expenditures probably poverty level (FPL), and year (t), based on
explain growing support among citizens and data from the Current Population Survey,
many state policymakers to alter PPACA, if CPS_STTPOPg,a,f,t.18 Because the CPS un-
not to repeal it. dercounts state populations relative to Cen-
sus Bureau counts for all states, the Census
population CEN_STTPOPg,a,t is also catego-
Appendix rized according to gender, age category, and
year cells, and the latter population is used
Methodology for Projecting Texas to rescale CPS population counts: for each
Medicaid Expenditures under PPACA demographic cell, the ratio of the two popu-
Medicaid spending projections for the lations
five states considered in this study are based CEN_STTPOPg,a,t
on various data sources, namely, the Medi- Ug,a,t =
Sf CPS_STTPOPg,a,f,t
caid Statistical Information System, the
Current Population Surveys, and the U.S. provides a measure of the cell-specific popu-
Census Bureau. Section A1 of this Appen- lation over- or under-counts in the CPS rela-
dix explains the general methodology and tive to the Census population.
Section A2 discusses the rules applied to de- Next, populations of Texas Medicaid
termine Medicaid eligibility for various age benefit-eligible individuals by demograph-
and eligibility categories, including differ- ic cells are calculated from the CPS:
ences in rules across the five states evaluated CPS_E_STTg,a,f,t. These cells are calculat-
here. Section A3 describes historical trends ed separately for specific income ranges
of Medicaid eligibility, enrollment, recipien- (f) relative to FPL values.19
cy, and average benefits per recipient sepa- For example, take a male aged a in 2008.
rately for various demographic groups and Adults qualify for Medicaid coverage if they
eligibility categories. have a covered child. In turn, the child is Prospects of
A1. Methodology for Projecting Medicaid Medicaid-eligible if the income of the child’s steep increases
Expenditures in Texas. The Medicaid Statis- family falls within the income threshold or
tical Information System State Data Mart the child qualifies based on non-income- in Medicaid
website provides administrative informa- related criteria such as disability and foster expenditures
tion on the number of Medicaid beneficia- care (for which income-eligibility limits are probably explain
ries by gender (g), age-category (a), and eli- different). Thus, the eligibility rate, e, for
gibility group (e) for the years 1999–2008. It adults aged a, of gender g, with FPL-relative growing support
also provides information on total Medicaid income f, and in year t, can be calculated among citizens
benefits awarded to state residents (B_STT) conditional on their children’s eligibility as
in those years, where the suffix, STT, stands
and many state
for the state in question. Ug,a,t x E_CPSg,a,f,t policymakers to
eg,a,f,t = .
In all states, residents qualify for Medi- Ug,a,t x CPS_STTPOPg,a,f,t alter PPACA, if
caid benefits based on a range of income-
and asset-related criteria. In addition, special Here, the numerator refers to the total num- not to repeal it.

19
Component rates, ber of state residents found to be Medicaid genders and aggregated to yield total (MSIS-
independently eligible in the CPS after applying the eligi- based) Medicaid expenditures for the year in
bility rules and the population adjustment question.
calculated and ratio Ug,a,t (described above). Total Medicaid expenditures derived
projected, capture Next, the enrollment rate, n, is calculated in this manner for the base year, 2008, are
as the number of Medicaid enrollees divided benchmarked to total (expended) Medicaid
different policy by the number of Medicaid eligibles: expenditures in 2008 as reported in the state
or environmental budget. This step takes account of Dispro-
N_MSISg,a,t
factors, each with ng,a,t = . portionate Share Hospital, Upper Payment
Ug,a,t x Sf E_CPSg,a,f,t Limit, and Medicaid administrative expen-
the potential to ditures that are not included in MSIS data.
exhibit its own Here, the numerator is the total number of Thus, these additional expenditures are im-
future trend. male state residents aged a, of gender g, in plicitly distributed across age, gender, and
year t, who are enrolled in Medicaid based eligibility categories in the same proportion
on data obtained from MSIS. One limita- as Texas Medicaid expenditures included in
tion of the data from MSIS is that they are MSIS data.
not decomposed by FPL-relative income The simplest way to project states’ Medi-
categories. Therefore, the average age- caid expenditures for future years is to repre-
gender enrollment rate is applied to all three sent total expenditures in earlier years by age,
FPL categories. Next, the recipiency rate, r, gender, and income, Mg,a,f,t, t=2001–2008, as
is calculated as the number of Medicaid re- above, and extrapolate each of the compo-
cipients (or beneficiaries) among Medicaid nent elements over future years. The prod-
enrollees: uct of those terms in future years provides
R_MSISg,a,t estimates of future Medicaid expenditures
rg,a,t = .
N_MSISg,a,t in the state for each particular gender, age,
and FPL category. Summing over all catego-
Again, data for the number of state resi- ries provides the future year’s total Medicaid
dents who received Medicaid benefits are expenditures.
obtained from MSIS. Finally, average Med- The reason for calculating and indepen-
icaid benefits per recipient, b, in the state in dently projecting each of these component
question are calculated from the MSIS as rates when constructing Medicaid’s expen-
diture projections is that those rates capture
B_MSISg,a,t
bg,a,t = , different policy or environmental factors,
R_MSISg,a,t each with the potential to exhibit its own
future trend. For example, while the Med-
where the numerator refers to total Medi- icaid eligibility rate for a particular popula-
caid benefits for this group. The average tion sub-group is determined by federal and
age-gender ratios rg,a,t and bg,a,t are applied state policies about which types of individu-
to those who are Medicaid eligible in each als should qualify for Medicaid benefits,
FPL-relative income category. Thus, total enrollment rates for different population
state Medicaid expenditures in 2008 on sub-groups may be determined by the avail-
males aged a, gender g, FPL category f, and ability and cost of alternative health insur-
year t, can be represented as ance coverage, individual perceptions about
their health care needs, the quality and out-
Mg,a,f,t = Ug,a,t x CPS_STTPOPg,a,f,t x of-pocket expenditures of Medicaid’s health
eg,a,f,t x ng,a,t x rg,a,t x bg,a,t . care provision, and public awareness about
the availability of Medicaid coverage for
This method of calculating the four rates people with similar demographic, econom-
can be applied to all age groups and both ic, and health characteristics.

20
 Furthermore, Medicaid recipiency rates a. Federal Poverty Level. Having a family in-
could be different among different popula- come below a specific FPL is one of the key
tion sub-groups by age, gender, and other eligibility criteria. Prior to the new health
characteristics, depending on their frequen- care law, Medicaid’s federal income eligibil-
cies of adverse health episodes and health ity threshold was 100 percent of the FPL
service needs. Finally, average benefit rates for children aged 6 through 18, with a state
would differ depending on the incidence option up to 133 percent and an option to
of chronic conditions: whether recipients extend the eligible age to 20 for those in
are elderly or disabled; the type, quality, school. For 1- to 5-year-olds the federal in-
and cost of health care treatments that are come eligibility threshold was 133 percent.
locally available, and so on. Basing projec- For newborns and pregnant women, the in-
tions on detailed historical information on come limit was at 133 percent of FPL, with
the group-specific trends of all four com- options for states to increase it up to 185
ponents separately: by age; gender; whether percent of FPL.
disabled; and income level (relative to the The new law establishes an income eli-
federal poverty level), whether medically gibility threshold for everyone (including
needy, unemployed, single- or dual-income childless adults) at 133 percent of FPL. It
family, number of children, etc.—provides also introduces an income disregard at 5 The new law
greater confidence that the rich variety of percent of family income. establishes an
independent influences of policies, environ- Other special deductions were applied income eli­gibility
mental conditions, and behavioral propen- before the new health care laws were enacted,
sities on Medicaid expenditures has been and varied by state. These generally includ- threshold for
adequately accounted for. ed items such as work-related ($90 a month) everyone at
PPACA changes eligibility rules for low- and dependent-care expenses ($175–$200 a
income individuals and mandates health month), child-support payments, earnings
133 percent
insurance coverage for all. In addition, it of children under age 19 and in school, all of the federal
envisions a vigorous public-awareness and income from Social Security’s Supplemen- poverty level.
enrollment facilitation drive that would in- tal Security Income program, other public
crease enrollment rates among both those assistance, and educational assistance.
people eligible under the old Medicaid laws Of the five most populous states, Califor-
and those newly eligible under PPACA. So nia, Illinois, and New York have elected to
state Medicaid expenditures under PPACA cover children aged 6 through 18 at the 133
are likely to be quite different (and consid- percent level. All of the states cover newborns
erably larger) compared with expenditures and pregnant women up to 185 percent of
under the old health care laws. FPL. Not only that, but each of this study’s
A2. Medicaid Eligibility Criteria. There states, except for Texas, has chosen to pay
are three key requirements to be eligible for pregnant women up to 200 percent from
for Medicaid in any state. This section will outside its federal Medicaid budgets.
briefly discuss each of these rules and how b. TANF/AFDC. The original federal wel-
they vary by state. The section will then cov- fare program, Aid to Families with Depen-
er a handful of other reasons someone may dent Children, was reformed in 1996, be-
be eligible for Medicaid. All of the rules de- coming the Temporary Assistance for Needy
scribed here for the three key eligibility crite- Families program. Originally, AFDC qualify-
ria are coded to determine eligibility to Medi- ing thresholds and conditions were used to
caid among the CPS sample populations by determine Medicaid eligibility, but today,
age, gender, FPL-relative income category, TANF thresholds and conditions are used. A
and those eligible under special rules for the family is eligible for AFDC/TANF, and there-
years spanning 2000 to 2008—the latest year fore Medicaid, if they are citizens with de-
for which CPS data are available. pendent children and have incomes less than

21
 the qualifying thresholds. In addition, for as also maintains separate TANF rules with
two-parent households, the primary earner a 1996-level income limit for a household
must either be unemployed (or disabled) or of three with one parent at $9,012, approxi-
earn less than the AFDC income threshold, mately 50 percent of FPL, and declining. If
or else be underemployed (as defined by each the family is currently receiving Medicaid,
state). they get a one-third earned income disre-
Run separately by each state, the TANF gard, bringing the effective FPL to 74 per-
eligibility rules vary widely between the five cent.
states examined in this report. In California c. Blind/Disabled and Elderly. Social Secu-
the income limits vary each year, with the rity Supplemental Security Income recipi-
2008 income limit at $12,960 for a three- ents are also eligible for Medicaid benefits
person household, or around 70 percent of under the Medicaid for Employed Persons
FPL. California also uses an income disre- with Disabilities program. The qualifying
gard (income not included when calculating rules consider unearned income (net of a
Medicaid’s income eligibility threshold) that monthly $20 exclusion) and earned income
equals 100 percent of FPL minus the 1996 (net of a monthly $65 exclusion and an an-
AFDC Maximum Aid Payment. For a three- nually determined student earned-income
person household this brings the effec- exclusion), the sum of which must be below
tive 2008 FPL to 125 percent. Unlike Califor- a specific annually indexed dollar threshold
nia, Florida’s income limits are fixed at the ($11,472 for a couple in 2009). Addition-
1996 AFDC standard of $3,636 for a three- ally, retirees and disabled individuals qual-
person household, or around 20 percent of ify for subsidies to pay for their Medicare
FPL, and are declining over time. Also, Flor- costs (premiums, co-pays, etc.) funded out
ida only has an income disregard for those of Medicaid through the Medicare Savings
who are already receiving Medicaid. For Program. These rules require individuals
these people it is the $200 and a one-half to be receiving Social Security or Railroad
earned income disregard, which pushes up Retirement benefits and have family income
the effective 2008 FPL to 166 percent. less than 200 percent of FPL for retired
Illinois’s income limits vary periodically persons and 135 percent of FPL for blind/
and by whether or not there is an adult in disabled individuals. Further, disabled work-
the household. The 2008 limit was $4,752 ers with earned income less than 250 percent
for a three-person household with a parent, of FPL qualify for the Medicaid buy-in pro-
or 26 percent of FPL. If a family is already gram. The blind/disabled and elderly eligi-
receiving Medicaid, they are eligible for the bility rules are the same across all states as
$30 and one-third earned income disregard, dictated by the Social Security Act.
or 55 percent of the FPL in 2008. New York d. Other Reasons. The federal government
income limits also vary over time but not requires that states’ Medicaid programs
necessarily every year. In 2008 the income cover the above groups of people in order
limit for a three-person household was to receive federal matching funds for Med-
$12,276, or around 67 percent of FPL. New icaid. States are not compelled to cover the
Run separately York also has the $30 and one-third earned following groups of people, although many
by each state, the income disregard for those already receiving states—including all five states examined in
Medicaid, bringing the effective 2008 FPL to this report—do extend at least some benefits
TANF eligibility 137 percent. Texas’s AFDC income limit is to these groups.
rules vary widely also stagnant at the 1996 level, which varies Foster-care children are covered under
between the five by the number of adults in the household. Medicaid if the household they came from
The income level for a single-parent, three- qualifies for AFDC/TANF or has an income
states examined person household is $16,668, which in 2008 below the federal poverty level. When a
in this report. was at 91 percent of FPL, and declining. Tex- foster-care child ages out of the system at age

22
18, the child continues to be fully covered lations span the years 2000–08, correspond- Women aged
through age 21 if that child has an income ing to the latest available data from the CPS 18–44 with
below a certain FPL. Under the new health (on eligibility rates) and MSIS (enrollment
care law, the age limit to receive Medicaid for and recipiency rates, average benefits per re- incomes below
aged-out foster-care children has risen to 25. cipient). a state-specific
Medically needy individuals can also be fully It should be noted that the methodology
covered by Medicaid if the state elects to. for projecting eligibility, enrollments, and
federal poverty
Medically needy individuals are determined other ratios as described above is different level qualify
by a combination of income thresholds and from those adopted in other studies, par- for Medicaid
medical expenses. ticularly the Kaiser study referenced earlier
Women aged 18–44 with incomes below (see note 16). As in my study, Medicaid eligi- family-planning
a state-specific FPL qualify for Medicaid bility rates are determined from the Current services.
family-planning services (of which preg- Population Survey’s Annual Social and Eco-
nant women receive full Medicaid benefits). nomic Supplement in the Kaiser study. The
States may also receive waivers to expand latter study fixes baseline eligibility rates ac-
family planning to more of their popula- cording to those observed in the 2007 data.
tions. Women between the ages of 18 and In my study, however, eligibility trends are
64 may also be eligible for breast or cervi- established between 2000 and 2008 and
cal cancer care under Medicaid if they are those trends are projected forward using
found with either of those cancers and their linear regressions—one for each eligibility
income is below a certain threshold as deter- group. This method is likely to more ro-
mined by the states. bustly establish and extrapolate the direc-
Last but not least, certain groups of peo- tion of change of group-specific population
ple may become eligible under the waiver eligibility rates projected, with and without
system. Specifically, the 1115 waiver allows PPACA’s eligibility rules. The same remark
states to write-off certain rules for a demon- applies to the determination of enrollment
stration or pilot project. New York has taken and beneficiary ratios and average benefits
this opportunity to allow childless individu- per beneficiary. Thus, my study tracks the
als below 100 percent of FPL to be eligible change in the trend in state GR Medicaid
for Medicaid. In most cases the 1115 waiver spending that PPACA would induce over
and the other waivers (1915(b) and 1915(c)) and above spending without PPACA, where-
are used to expand services or change the as the Kaiser study measures changes from
way services are conducted. fixed eligibility rates benchmarked to 2007.
A3. Medicaid Eligibility, Enrollment, Re- Also, the Kaiser study uses fixed, assumed
cipiency, and Average Benefits per Enrollee. rates of new enrollments post-PPACA for
This section describes information obtained old-eligibles and for those newly eligible un-
from calculating each of the four compo- der PPACA. Those rates are taken from aver-
nents for those with FPL less than 100 per- age enrollments over all groups reported by
cent as noted in Section A1 above—namely, the Congressional Budget Office. My study,
eligibility rates, enrollment rates, recipiency in contrast, allows the CPS and MSIS data
rates, and average benefits per recipient. Be- to inform the calibration of group-specific
cause eligibility conditions and health needs new enrollment rates for those newly eligi-
differ substantially by age and gender, the ble and those who are not enrolled (without
four items are calculated separately for vari- PPACA) among old-eligibles. For both types
ous age groups (see note 17), gender, and of new enrollees, group-specific enrollment
FPL-relative income levels. In addition, spe- rates are taken to be the larger of the enroll-
cial eligibility groups such as the medically ment rate without PPACA (for the entire
needy, foster-care children, family planning, group) or the non-insurance rate among
and others are treated separately. The calcu- those (old-eligibles or the newly eligible) not

23
 enrolled in Medicaid, again assuming the for men (as they are less likely to be in low-
absence of PPACA. earning families and are also less likely to
a. Eligibility, Enrollments, Recipiency, and have a Medicaid-eligible child as a depen-
Benefits among Children. In all states, Medic- dent, on average), and the pattern of eligi-
aid eligibility rates increased during the last bility by age-group is reversed compared
decade among children of all ages and both with females: older males have a higher like-
genders.20 The data indicate that more than lihood of qualifying for Medicaid, probably
80 percent of almost all child age and gender because a higher proportion of men work in
groups (except for newborns) were Medicaid- strenuous jobs and become disabled or un-
eligible by 2008. During the early 2000s, en- employed at older working ages.
rollment rates were much smaller for older Figures A3 (females) and A4 (males) show
children compared with younger ones. How- that Medicaid enrollment rates among eligi-
ever, enrollment rates for older children have bles is widely divergent across the five states
increased steadily so that, by 2008, more examined here. Enrollment rates are highest
than 60 percent of all eligible children are in California for both genders. Enrollment
enrolled in the program. Medicaid recipiency rates are low and/or declining in Illinois
rates were quite high during the early 2000s and New York, a fact that plays a key role in
In all states, and have increased consistently during the generating high spending increases in those
Medicaid last decade: at least 85 percent of all child two states when PPACA’s individual health
eligibility rates groups received Medicaid benefits during insurance mandate is included when mak-
2008—again, except for newborns, who have ing spending projections. Florida and Tex-
increased during the smallest recipiency rates. as have mixed enrollment rates across age
the last decade On the other hand, the data indicate that groups—high for those aged 19–20, but low
newborns incur the highest Medicaid expen- among many older age-gender groups.
among children ditures. Excluding newborns and those aged Figures A5 and A6 show Medicaid recipi-
of all ages and 1–5, Average Medicaid expenditures per re- ency rates above 80 percent for females in all
both genders. cipient are smaller for younger children and states. For males, recipiency rates in all states
they increase with age. However, average ex- except California were low at the turn of the
penditures for the oldest children are only century, but increased rapidly to reach the
about one-half of those for newborns. Aver- same levels as those for females by 2008. Cali-
age expenditures have trended upward dur- fornia’s recipiency rates, however, appear to
ing the last decade, reflecting the general be declining rapidly—a factor that explains
rapid increase in health care costs. the low impact of introducing PPACA when
b. Working-age Adults. Working-age adults making spending projections. Figures A7 and
are split between disabled adults, non- A8 show that older non-disabled adult ben-
disabled adults, and others, where the last eficiaries receive larger Medicaid benefits, on
category includes medically needy individu- average, compared with younger ones—ex-
als and women eligible for benefits from cept in California. Benefit awards per bene-
the Breast and Cervical Cancer Act under ficiary are stable or gradually increasing in
Medicaid. For non-disabled adults, eligibil- California, Florida, Illinois, and Texas. They
ity rates under Medicaid are distinctly dif- increase more rapidly in New York for both
ferent for males versus females. As Figure genders. Differences in benefit awards per
A1 shows, female eligibility rates among beneficiary by age-group are much larger for
the 0–100 percent FPL category were at 40 men compared with women.
percent or less and barely increased during c. Retirees. Among those aged 65 and old-
the last decade.21 That is not surprising, be- er in the 0–100 percent FPL range, eligibil-
cause women are more likely to be part of ity rates are highest in Texas—well above 80
a low-income family. Figure A2 shows that percent among younger retirees and close to
Medicaid eligibility rates are much smaller 100 percent among older ones.22 Given Tex-

24
as’s moderate historical enrollment rates, identification of the eligible populations for Medicaid
these conditions imply a strong spending- these groups.23 Hence, calculations are based recipiency rates
increasing effect from introducing PPA- on directly calculating the share of enrollees
CA. Eligibility rates are lowest for Illinois in the population based on MSIS data for were quite high
and average about 50 percent for the other foster-care children, BCCA women, family- in all of the five
three states. Enrollment rates are highest planning recipients, and medically needy in-
for California and Florida, with rates in the dividuals.
states considered
three other states varying from 30 percent For blind/disabled adults, however, eligi- here—about
in Texas to 80 percent in Illinois and New bility co-criteria based on income (including 80 percent
York. Enrollment rates have generally in- spousal income where applicable) are incor-
creased, more so among the oldest retirees. porated, again counting all eligible sources during the early
Medicaid recipiency rates were quite high in and net of applicable exemptions, deduc- years of the
all of the five states considered here—about tions, and income disregards. Medicaid 21st century.
80 percent during the early years of the 21st eligibility rates were higher for older blind/
century—but have declined since for both disabled adults compared with younger
genders, especially in Florida, New York, ones in most states, and they have generally
and Texas. The reason for this may be the increased during the early years of the 21st
expansion of Medicare Part B coverage and century across all age groups. Data show sta-
the shift of many retirees’ Medicaid cover- ble enrollment rates for most blind/disabled
age for prescription drugs to the Medicare women—higher in California and Florida
program. Finally, average Medicaid expen- than in the other three states. Enrollment
ditures per recipient increased across all rates have been higher, but stable overall, for
retiree age groups—Illinois being the excep- disabled/blind men compared with wom-
tion—and the increase was especially rapid en during the last decade—also highest in
for the oldest retirees. California and Florida. Medicaid recipiency
d. Other Groups. Calculations of eligibility, rates have been stable or declining in Flori-
enrollment, recipiency, and average benefits da and Texas for most disabled adults, and
per recipient are implemented separately for stable or increasing in the other three states.
foster-care children, medically needy individ- However, Medicaid expenditures per blind/
uals, women qualifying under the Breast and disabled recipient are among the highest
Cervical Cancer Act, the family-planning pro- among all population groups and have in-
gram, and blind/disabled adults. Except for creased consistently for both genders in all
blind/disabled adults, CPS data do not allow states except Illinois during the last decade.

25
Figure A1
Shares of Medicaid Eligibles in State Populations and Linear Trends for Female Non-Disabled Adults Aged 19–64;
0–100 Percent of FPL; 2000–08

California Florida

Share
Share

Illinois New York

Share

Share

Texas
Share

Source: Author’s calculations based on the Medicaid Statistical Information System, current population surveys, CMS-64 reports, and the U.S. Census
Bureau.

26
Figure A2
Shares of Medicaid Eligibles in State Populations and Linear Trends for Male Non-Disabled Adults aged 19–64;
0–100 percent of FPL; 2000–08

California Florida
Share

Share

Illinois New York

Share
Share

Texas
Share

Source: Author’s calculations based on the Medicaid Statistical Information System, current population surveys, CMS-64 reports, and the U.S. Census
Bureau.

27
Figure A3
Shares of Medicaid Enrollees among Eligibles and Linear Trends for Female Non-Disabled Adults Aged 19–64;
0–100 Percent of FPL; 2000–08

California Florida
Share

Share

Illinois New York

Share

Share

Texas
Share

Source: Author’s calculations based on the Medicaid Statistical Information System, current population surveys, CMS-64 reports, and the U.S. Census
Bureau.

28
Figure A4
Shares of Medicaid Enrollees among Eligibles and Linear Trends for Male Non-Disabled Adults Aged 19–64;
0–100 Percent of FPL; 2000–08

California Florida
Share

Share

Illinois New York

Share

Share

Texas
Share

Source: Author’s calculations based on the Medicaid Statistical Information System, current population surveys, CMS-64 reports, and the U.S. Census
Bureau.

29
Figure A5
Shares of Medicaid Beneficiaries among Enrollees and Linear Trends for Female Non-Disabled Adults Aged 19–64;
0–100 percent of FPL; 2000–08

California Florida

Share
Share

Illinois New York

Share

Share

Texas
Share

Source: Author’s calculations based on the Medicaid Statistical Information System, current population surveys, CMS-64 reports, and the U.S. Census
Bureau.

30
Figure A6
Shares of Medicaid Beneficiaries among Enrollees and Linear Trends for Male Non-Disabled Adults Aged 19–64;
0–100 percent of FPL; 2000–08

California Florida
Share

Share

Illinois New York

Share
Share

Texas
Share

Source: Author’s calculations based on the Medicaid Statistical Information System, current population surveys, CMS-64 reports, and the U.S. Census
Bureau.

31
Figure A7
Average Medicaid Expenditures per Beneficiary and Exponential Trends for Female Non-Disabled Adults Aged
19–64; 0–100 percent of FPL; 2000–08

California Florida

Thousands of dollars
Thousands of dollars

Illinois New York

Thousands of dollars

Thousands of dollars

Texas
Thousands of dollars

Source: Author’s calculations based on the Medicaid Statistical Information System, current population surveys, CMS-64 reports, and the U.S. Census
Bureau.

32
 Figure A8
Average Medicaid Expenditures per Beneficiary and Linear Trends for Male Non-Disabled Adults Aged 19–64;
0–100 percent of FPL; 2000–08

California Florida
Thousands of dollars

Thousands of dollars

Illinois New York

Thousands of dollars

Thousands of dollars

Texas
Thousands of dollars

Source: Author’s calculations based on the Medicaid Statistical Information System, current population surveys, CMS-64 reports, and the U.S. Census
Bureau.

33
 Notes much of California’s uncompensated cost saving
under PPACA is likely to arise from newly privately
This paper draws heavily from “The Effects of insured individuals. These costs savings appear to
ObamaCare on Texas’s Medicaid Expenditures’ dominate the state’s increased Medicaid spend-
Growth,” by Jagadeesh Gokhale, published in ing from the few new Medicaid enrollees among
the 2010 PolicyPerspective series of the Texas Pub- old eligibles and those made newly eligible under
lic Policy Foundation. Angela Erickson provided PPACA.
excellent research assistance.
6. There is an exception for states with waivers
1. State budget reports include expenditures already covering these populations, such as New
on a General Revenue basis and the All Funds York. These states will continue to pay Medicaid
basis, the former referring to expenditures out of benefits to these individuals for the first three
state revenue sources and the latter being inclu- years and then will have an increased Federal
sive of expenditures funded out of federal grants. Medical Assistance Percentage for the newly eli-
This study focuses only on states’ General Rev- gible that will match those for other states by
enue–funded Medicaid expenditures, projected 2019. For New York and other non-expansion
with and without PPACA. states this represents a cost savings per childless
adult enrollee.
2. Constructing these estimates is a laborious
and time-intensive process, and doing so for all 7. Qualification for Medically Needy Program
50 states and the District of Columbia was not benefits varies by state. In Texas, for example, an
feasible. applicant must be: (1) a pregnant woman with
no child eligible for the Temporary Assistance
3. The growth rate of state GR Medicaid ex- for Needy Families Program; (2) a child under
penditures is calculated beginning in the year 19 years of age; or (3) an adult caretaker whom
2010 and includes the effect of reduced federal the state’s Health and Human Services Commis-
support from the American Recovery and Rein- sion includes in the certified group, and who
vestment Act. However, see the note attached to ordinarily receives and manages the benefits for
Table 4. the certified group, except that the caretaker’s
countable income exceeds TANF limits, the care-
4. The slowdown in economic growth all five taker’s 60-month time-limited TANF benefits
states experienced during 2007–09 makes it very are exhausted, the caretaker chooses Medicaid-
difficult to project GDP for future years. Project- only benefits, or the caretaker is disqualified
ed Medicaid expenditure growth is, therefore, from TANF for a reason that is not applicable
compared with historical GDP growth calcu- to Medicaid; and (4) have countable income that
lated using data from the United States Bureau meets the applicable income limit. The income
of Economic Analysis. The earliest year of data limit is defined based on family size; for a family
available for historical calculations of annual of two people, it is $216 per month. Applicants
(nominal) GDP growth is 1997. whose income exceeds the limit may spend down
excess income to pay medical bills and qualify.
5. States partially cover “disproportionate share
hospital” (DSH) hospital costs of treating unin- 8. The formula equals 100 percent minus the
sured patients. Those include illegal immigrants, state’s share where the state’s share equals 0.45 x
old-eligibles who are not yet enrolled into Medi- (SPCI/USPCI)2, where SPCI is state’s per capita
caid, those who will become newly eligible to Med- income and USPCI is the United States’ per cap-
icaid under PPACA, and those who are not eligible ita income. A higher SPCI translates into a lower
to Medicaid currently and will not become eligible FMAP value.
under PPACA. As everyone except illegal immi-
grants acquires health insurance under PPACA’s 9. The 2012 FMAP value for Texas is expected
individual mandate, state payments to DSH hos- to be published by the federal department of
pitals for uncompensated care will decline. But Health and Human Services in November 2010.
some of the cost savings will be offset because
some of the previously uninsured (among both 10. Census Bureau state-wise population projec-
old-eligibles and those made newly eligible under tions are available online at: http://www.census.
PPACA) will enroll into Medicaid. However, other gov/population/www/projections/projections
newly insured individuals may enroll into private agesex.html.
health insurance—generating sizable state savings
of uncompensated care. State uncompensated 11. Increases in future Medicaid expenditures in
care costs will remain positive on account of illegal each state will also depend on how successful ef-
immigrants. Because California has very few po- forts are to repeal PPACA. Twenty-six states have
tential new Medicaid enrollees (especially among filed court cases to challenge the new health care
old-eligibles) but has a large number of uninsured, law on two grounds: (1) that mandating purchase

34
of health insurance by individuals (with failure enrollments in the base (historical) year, 2008.
punishable by a fine) is unconstitutional under
the Tenth Amendment and Article I of the U.S. 16. See Note 6. A further reason for the low
Constitution (the commerce clause); and (2) that spending increase in California from PPACA is
the new health care law increases states’ Medi- that the ratio of beneficiaries among enrollees
caid expenditures without recompense from the has historically declined, especially among the
federal government—that is, it constitutes an un- largest group of non-disabled adults—a feature
funded mandate. that would dampen future expenditures under
the methodology of this study.
12. Despite determining enrollment rates for
heretofore non-enrolled old-eligibles (as de- 17. Even when the incremental Medicaid spend-
scribed in the text), a valid question is whether ing totals under this study are restricted to the
it is appropriate to apply the same beneficiary/ 2014–19 period, spending increases among the
enrollment and benefits/beneficiary rates appli- four states, excluding California (for which the
cable to pre-PPACA Medicaid enrollees when increase is small), range from 17 percent for Flori-
making projections on a post-PPACA basis. da to 28 percent for Illinois. Kaiser’s estimates
There are several reasons that justify doing so: are available online at: http://www.statehealth
first, the uninsured do not have zero health care facts.org/comparereport.jsp?rep=68&cat=4.
costs; their emergency room visits are costly and
the costs are shifted to other patients who either 18. The age categories correspond to those of
pay out-of-pocket or through their insurers. Sec- the Medicaid State Information System’s age
ond, the uninsured have low measured health ranges: 0, 1–5, 6–12, 13–14, 15–18, 19–20, 21–44,
care costs because they are uninsured. Once they 45–64, 65–74, 75–84, and 85+.
are insured under Medicaid via PPACA and the
individual health insurance mandate, their utili- 19. The income ranges are defined according to
zation of health care goods and services is likely the applicable cutoffs before and under the new
to increase and match that of other Medicaid health care law. Those cutoffs are generally dif-
enrollees. Third, some of the uninsured may be ferent for population groups served by various
without insurance because they have pre-existing Medicaid programs in Texas.
conditions that are expensive to treat. Indeed,
after they enroll into Medicaid under PPACA, 20. Charts for selected child age/gender/FPL
those costs may surpass the average costs of those groups are available from the author upon request.
already insured. Fourth, historical trends in the
ratio of beneficiaries among enrollees are increas- 21. Although post-PPACA Medicaid eligibility is
ing, suggesting that future expenditures may be based only on an income test, pre-PPACA eligibil-
larger as more people need and claim health in- ity requires an asset test as well. The pre-PPACA
surance benefits. Finally, the old-eligibles newly asset test is incorporated based on information
enrolled under PPACA may be uninsured and from the 2007 Survey of Consumer Finances.
young, but their health care needs and costs will Calculations show that the population share of
grow over time. adult non-disabled household heads who fail the
Medicaid asset test and who do not receive Medi-
13. Uncompensated care savings are calculated caid is 13.2 percent for the nation as a whole. Un-
by distributing total state DSH spending among fortunately, the survey does not allow separate
the state’s uninsured population, and allocat- identification of Texas residents, nor of blind/
ing the cost to those who are legally obligated to disabled individuals. Therefore, an approximate
acquire health insurance under PPACA—all asset-based constraint is applied to pre-PPACA
uninsured except for illegal immigrants. Un- eligibility rates, to restrict Medicaid eligibility to
compensated care savings in future years are 86.8 percent (100 percent minus 13.2 percent)
anchored on the state’s population projections. of the income-based eligibility rate as calculated
from the Current Population Survey.
14. All five states exhibit stable or declining
ratios of beneficiaries-to-enrollees among non- 22. Charts for selected age/gender/FPL groups
disabled retirees. The declines are steepest in are available from the author upon request.
Florida and New York.
23. Medicaid eligibility criteria for children in
15. These results depend on projections of aver- foster care and younger than age 18 are the same
age benefits per recipient and enrollments on a pre- as those for non-foster-care children aged less
PPACA basis between 2008 (the latest year with full than 18. Eligibility under AFDC/TANF rules are
information available at the time of writing this based on the incomes of the household a child
study) and 2014. However, the results obtained comes from before foster care placement. For
for 2014 and beyond are consistent with those children older than age 18 there is the Medicaid
obtained by simply dividing total expenditures by for Transitioning Foster Care Youth program,

35
whereby the person must have aged out of fos- persons with breast and cervical cancer include
ter care, must be between the ages of 18 and the provision that the person must be diagnosed
20, must not be covered under another health with breast cancer (men and women) or cervi-
plan offering adequate benefits, and must have cal cancer (women only), must not have income
income at or below 400 percent of the Federal more than 200 percent of the FPL, and must not
Poverty Level. Medicaid eligibility criteria for have alternative medical insurance coverage.

36
 Cato Institute
Founded in 1977, the Cato Institute is a public policy research foundation dedicated to
broadening the parameters of policy debate to allow consideration of more options that are
consistent with the traditional American principles of limited government, individual liberty,
and peace. To that end, the Institute strives to achieve greater involvement of the intelligent,
concerned lay public in questions of policy and the proper role of government.
The Institute is named for Cato’s Letters, libertarian pamphlets that were widely read in
the American Colonies in the early 18th century and played a major role in laying the philo-
sophical foundation for the American Revolution.
Despite the achievement of the nation’s Founders, today virtually no aspect of life is free
from government encroachment. A pervasive intolerance for individual rights is shown by
government’s arbitrary intrusions into private economic transactions and its disregard for
civil liberties.
To counter that trend, the Cato Institute undertakes an extensive publications program that
addresses the complete spectrum of policy issues. Books, monographs, and shorter studies
are commissioned to examine the federal budget, Social Security, regulation, military spend-
ing, international trade, and myriad other issues. Major policy conferences are held through-
out the year, from which papers are published thrice yearly in the Cato Journal. The Institute
also publishes the quarterly magazine Regulation.
In order to maintain its independence, the Cato Institute accepts no government fund-
ing. Contributions are received from foundations, corporations, and individuals, and other
revenue is generated from the sale of publications. The Institute is a nonprofit, tax-exempt,
educational foundation under Section 501(c)3 of the Internal Revenue Code.

Cato Institute
1000 Massachusetts Ave., N.W.
Washington, D.C. 20001
www.cato.org
$10.00

OTHER PUBLICATIONS OF INTEREST

Social Security: A Fresh Look Bad Medicine: A Guide to the Real
at Policy Alternatives Costs and Consequences of
A highly detailed assess- the New Health
ment of how 6 prominent Care Law
reform packages—two liber-
al, two centrist, and two This incisive report pro-
conservative—would impact vides an authoritative
Social Security’s financial and revealing explana-
health and future. tion of the new health
care law’s provisions. Its
“ Innovative, interesting, and
important. Gokhale delivers
diagnosis is that the law
is bad medicine. It is like-
on the promise in the title, ly to make Americans
providing a new appraisal of less healthy, less able to
a variety of plans to reform direct their own health
Social Security that will care decisions, and
appeal to a wide range of places huge burdens
”
readers.
on our economy and
—DALE JORGENSON, national debt.
Harvard University

$35.00, HARDBACK $10.00, SOFTCOVER REPORT • ONLY AVAILABLE AT CATO.ORG.

Healthy Competition: What’s Crisis of Abundance:

Holding Back Health Care and Rethinking How We Pay for
How to Free It Health Care
With America’s health care Using basic economic con-
system spiraling downward, cepts, Kling demonstrates
the authors demonstrate how how a greater reliance on
increased competition and private savings and market
consumer choice provide innovation eliminates waste,
higher quality and contains health care costs,
lower prices. and improves the quality
of care.
“ Cannon and Tanner offer
proposals that would further
“ His book is clear, concise,
tap the power of markets to and eminently readable.
make health care more valu- I warmly recommend his
able and more affordable. book to general readers who
That makes Healthy Competition want to understand what
essential reading.
”
economics has to say about
health care.
”
—GEORGE P. SHULTZ,
Former Secretary of State
—ARNOLD S. RELMAN, M.D.,
Harvard Medical School, writing in the
$11.95, HARDBACK • $9.00, E-BOOK New England Journal of Medicine $16.95, HARDBACK • $9.98, PAPERBACK • $9.00, E-BOOK

AVAILABLE IN BOOKSTORES NATIONWIDE, ONLINE AT CATOSTORE.ORG, OR AT 800.767.1241

1000 Massachusetts Avenue, N.W. • Washington, D.C. 20001 • www.cato.org