Medicaid: Spending Pressures Drive States Toward Program Reinvention
(Chapter Report, 04/04/95, GAO/HEHS-95-122).
The $131 billion Medicaid program is at a crossroads. Between 1985 and
1993, Medicaid costs tripled and the number of beneficiaries rose by
more than 50 percent. Medicaid costs are projected to rise to $260
billion, according to the Congressional Budget Office. Despite federal
and state budgetary constraints, several states are pressuring to expand
the program and enroll hundreds of thousands of new beneficiaries. The
cost of expanded coverage, they believe, will be offset by the
reallocation of Medicaid funds and the wholesale movement of
beneficiaries into some type of managed care arrangement. This report
examines (1) federal and state Medicaid spending, (2) some states'
efforts to contain Medicaid costs and expand coverage through waiver of
federal requirements, and (3) the potential impact of these waivers on
federal spending and on Medicaid's program structure overall. The
Comptroller General summarized this report in testimony before Congress;
see: Medicaid: Spending Pressures Drive States Toward Program
Reinvention, by Charles A. Bowsher, Comptroller General of the United
States, before the House Committee on the Budget. GAO/T-HEHS-95-129,
Apr. 4 (six pages).
--------------------------- Indexing Terms -----------------------------
REPORTNUM: HEHS-95-122
TITLE: Medicaid: Spending Pressures Drive States Toward Program
Reinvention
DATE: 04/04/95
SUBJECT: Medicaid programs
Health care cost control
State-administered programs
Federal/state relations
Health maintenance organizations
Funds management
Disadvantaged persons
Future budget projections
Waivers
Quality assurance
IDENTIFIER: Florida
Hawaii
Oregon
Tennessee
Arizona
Ohio
Rhode Island
Kentucky
TennCare
**************************************************************************
* This file contains an ASCII representation of the text of a GAO *
* report. Delineations within the text indicating chapter titles, *
* headings, and bullets are preserved. Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are *
* identified by double and single lines. The numbers on the right end *
* of these lines indicate the position of each of the subsections in the *
* document outline. These numbers do NOT correspond with the page *
* numbers of the printed product. *
* *
* No attempt has been made to display graphic images, although figure *
* captions are reproduced. Tables are included, but may not resemble *
* those in the printed version. *
* *
* A printed copy of this report may be obtained from the GAO Document *
* Distribution Facility by calling (202) 512-6000, by faxing your *
* request to (301) 258-4066, or by writing to P.O. Box 6015, *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time. *
**************************************************************************
Cover
================================================================ COVER
Report to the Chairman, Committee on the Budget, House of
Representatives
April 1995
MEDICAID - SPENDING PRESSURES
DRIVE STATES TOWARD PROGRAM
REINVENTION
GAO/HEHS-95-122
Medicaid Spending Pressures
Abbreviations
=============================================================== ABBREV
AFDC - Aid to Families with Dependent Children
COBRA - Consolidated Omnibus Budget Reconciliation Act of 1985
DEFRA - Deficit Reduction Act of 1984
DSH - disproportionate share hospital
EPSDT - early and periodic screening, diagnostic, and treatment
FMAP - federal medical assistance percentage
FPL - federal poverty level
HCFA - Health Care Financing Administration
HHS - Department of Health and Human Services
HMO - health maintenance organization
IRCA - Immigration Reform and Control Act of 1986
MCCA - Medicare Catastrophic Coverage Act of 1988
NACHC - National Association of Community Health Centers
OBRA1986 - Omnibus Budget Reconciliation Act of 1986
OBRA1987 - Omnibus Budget Reconciliation Act of 1987
OBRA1989 - Omnibus Budget Reconciliation Act of 1989
OBRA1990 - Omnibus Budget Reconciliation Act of 1990
OBRA1993 - Omnibus Budget Reconciliation Act of 1993
OMB - Office of Management and Budget
SSI - Supplemental Security Income
Letter
=============================================================== LETTER
B-233299
April 4, 1995
The Honorable John R. Kasich
Chairman, Committee on the Budget
House of Representatives
Dear Mr. Chairman:
This report, prepared at your request, reviews federal and state
spending trends and states' efforts to restructure their Medicaid
programs.
We are sending copies of this report to the Secretary of Health and
Human Services; the Administrator, Health Care Financing
Administration; the Director, Office of Management and Budget; and
other congressional committees. Copies of this report will also be
made available to others on request. If you or your staff have any
questions, please call me at (202) 512-6806. Major contributors to
this report are listed in appendix V.
Sincerely yours,
Janet L. Shikles
Assistant Comptroller General
EXECUTIVE SUMMARY
============================================================ Chapter 0
PURPOSE
---------------------------------------------------------- Chapter 0:1
The $131 billion Medicaid program, a health care lifeline for over 33
million low-income Americans, is at a crossroads. Between 1985 and
1993 Medicaid costs tripled and the number of beneficiaries increased
by over 50 percent. Current projections suggest that program costs
will double over the next 5 to 7 years. While federal and state
budgetary constraints highlight the urgency of containing costs, a
number of states are pressuring to expand the program and enroll
hundreds of thousands of new beneficiaries. The cost of expanded
coverage, they believe, will be offset by the reallocation of certain
Medicaid funds and the wholesale movement of beneficiaries into some
type of managed care arrangement.
The House Budget Committee is examining Medicaid, among other
entitlement programs, looking for opportunities to increase program
efficiency and constrain spending growth. In light of this effort,
the Chairman of the Committee asked GAO to examine (1) federal and
state Medicaid spending, (2) some states' efforts to contain Medicaid
costs and expand coverage through waivers of certain federal
requirements, and (3) the potential impact of these waivers on
federal spending and on Medicaid's program structure overall.
BACKGROUND
---------------------------------------------------------- Chapter 0:2
Medicaid is not 1, but 56 separate programs (including the 50 states,
District of Columbia, Puerto Rico, and the U.S. territories).
Federal mandates call for common eligibility and benefit
requirements, but states have discretion in how they implement their
programs. As a result, the populations served and benefits provided
vary across states. For example, Nevada serves 284 Medicaid
beneficiaries for every 1,000 poor or near-poor individuals in the
state, whereas Rhode Island serves 913 per 1,000. Similarly,
Mississippi spends, on average, less than $2,400 per person on
Medicaid services, while New York spends an average of almost $7,300
per person.
State programs also vary in the percentage of program expenditures
that are covered by the federal government. The federal percentage
is determined by a formula based on a state's per capita income.
Low-income states receive higher percentages, though the range is
legislatively limited from a minimum of 50 percent to a maximum of 83
percent. The federal government matches what the state spends on
Medicaid by this percentage. In fiscal year 1993, 13 states,
including New York, received 50 cents for every dollar spent on their
respective Medicaid programs, whereas Mississippi, West Virginia, and
Utah each received more than 75 cents for every Medicaid dollar
spent. Yet because of the differences in prices and in the benefits
and services offered, New York receives $3,600 per beneficiary in
federal aid, while Mississippi receives $1,900.
Since the 1970s states have been experimenting with a variety of
managed care networks that limit which physicians and hospitals can
serve the state's Medicaid beneficiaries. In some cases, the managed
care networks are prepaid a fixed amount per enrollee. This
financing arrangement is known as capitation and has demonstrated the
ability to lower service utilization, which in turn can hold down
costs.
More recently, escalating program costs have persuaded some states to
move most or all of their Medicaid population into capitated managed
care as a way of controlling future cost growth. To do so, states
must obtain approval to waive certain federal Medicaid requirements.
The waiver authority that gives states the greatest flexibility in
implementing statewide managed care programs resides in section 1115
of the Social Security Act. Most waiver states are simultaneously
seeking authority to use Medicaid funds to provide health care
coverage to a portion of their low-income population that is
currently ineligible for Medicaid benefits. States anticipate that
savings from capitated managed care systems plus the redirection of
other Medicaid or state funds will finance the coverage of these
additional people. As part of the process of obtaining a section
1115 waiver, states must propose a financing plan that, over 5 years,
is intended not to require greater federal expenditures on their
Medicaid program than would have been the case without a waiver.
RESULTS IN BRIEF
---------------------------------------------------------- Chapter 0:3
Medicaid costs are projected to increase from about $131 billion to
$260 billion by the year 2000, according to the Congressional Budget
Office. Between 1985 and 1993, federal Medicaid expenditures grew
each year, on average, by 16 percent. Although the program's current
growth rate has moderated from that experienced in the early 1990s,
it is still much higher than the 3.8-percent rate at which the
overall federal budget has been growing. The continued phasing in of
mandated populations and the increase in the expensive aged and
disabled populations figure into estimates of Medicaid's future
growth rate.
In the mid-1980s, some states began using creative financing
mechanisms to leverage additional federal dollars. States collected
donations and taxes from specific hospitals and then returned a
portion of the collected funds to the same providers as special
Medicaid payments. Because these payments (called DSH payments) were
made to hospitals that served a disproportionate share of Medicaid
and low-income patients, they generated matching federal dollars.
The states' bookkeeping sleight-of-hand increased federal payments
without the states having to dip deeper into their own treasuries.
Thus, the states were able to effectively increase the share of
Medicaid funded by the federal government. This phenomenon helped
fuel annual federal spending increases to over 25
percent--representing billions of dollars--in both 1991 and 1992.
Although federal law has limited DSH payment growth since 1993, the
gaming of these payments in some states has both increased the level
and affected the distribution of current and future federal Medicaid
spending.
More recently, states began seeking section 1115 waivers designed to
contain the cost of their Medicaid programs through the use of
capitated managed care delivery systems and expand coverage to
uninsured individuals who would not normally qualify for Medicaid
benefits. Certain states' initial estimates of these newly eligible
individuals are high: 1.1 million in Florida, 395,000 in Ohio, and
500,000 in Tennessee. In addition to the 7 waivers approved since
1993, 15 states have either applied for waivers or made inquiries
about submitting waiver applications. If similar coverage expansions
are approved for additional and more densely populated states, the
federal government as well as the states could be supporting millions
more Medicaid beneficiaries.
GAO's analysis to date of four states with approved waivers shows
that Florida, Hawaii, and Oregon may obtain more federal funding than
they would have likely received under their original Medicaid
programs. For Tennessee, on the other hand, GAO estimates that
federal Medicaid spending over the life of the waiver may be less
than might have been spent had the waiver not been approved. GAO has
not analyzed the potential impact of the other waivers on federal
spending.
To date, the administration has portrayed all statewide 1115
demonstration waivers as budget neutral to the federal government.
Its approvals of waivers, however, may have state Medicaid programs
working at cross purposes with federal deficit reduction goals.
While these expansions will extend health care benefits to more
low-income individuals, the result could also be a heavier burden on
the federal budget.
PRINCIPAL FINDINGS
---------------------------------------------------------- Chapter 0:4
MEDICAID CONSUMES GROWING
SHARE OF THE FEDERAL BUDGET
-------------------------------------------------------- Chapter 0:4.1
Medicaid currently consumes about 6 percent of all federal outlays
(three times the share devoted to Food Stamps and five times the
share devoted to Aid to Families with Dependent Children). Moreover,
Medicaid's slice of the federal budget is growing faster than most
other major budget items. If Medicaid continues indefinitely at its
expected pace (10.7 percent), federal spending on the program will
double roughly every 5 to 7 years.
Creative financing mechanisms used by states to leverage additional
federal dollars contributed significantly to Medicaid's spending
growth of over 25 percent in both 1991 and 1992. (Fig. 1
illustrates recent Medicaid spending growth.) Part of these financing
mechanisms involved making payments to hospitals that served a
disproportionate share of Medicaid and other low-income patients.
These payments exploded from slightly less than $1 billion in 1990 to
over $17 billion in 1992. However, the Medicaid Voluntary
Contribution and Provider-Specific Tax Amendments of 1991 as well as
provisions in the Omnibus Budget Reconciliation Act of 1993 now
prevent these payments from rising faster than overall Medicaid
spending.
Figure 1: Medicaid Spending
More Than Tripled Since 1985
(See figure in printed
edition.)
Sources: Medicaid Source Book: Background Data and Analysis (A 1993
Update) (A Report Prepared by the Congressional Research Service for
the Use of the Subcommittee on Health and the Environment of the
Committee on Energy and Commerce, U.S. House of Representatives,
Jan. 1993), and HCFA.
Other factors have worked to increase Medicaid costs: medical
inflation, higher utilization of services, and more beneficiaries.
Although many of the new beneficiaries were pregnant women and
children made eligible by congressional mandates enacted since 1984,
the addition of this group played a less significant role in
increasing Medicaid costs because these individuals are relatively
inexpensive to serve. The pressure on Medicaid costs is expected to
continue. For example, the number of individuals with disabilities
receiving Medicaid benefits is growing faster than any other
beneficiary group. While only about 15 percent of all Medicaid
beneficiaries are disabled, this group accounts for a much larger
share of program costs.
WAIVERS USED TO CONTAIN
COSTS AND EXPAND PROGRAM
COVERAGE
-------------------------------------------------------- Chapter 0:4.2
Section 1115 waivers are intended to permit experimentation by
allowing states to run demonstration projects that further the goals
of the Medicaid program. As such, these projects must have an
evaluation component and are expected to run for a limited period of
time. The only prior use of section 1115 authority comparable to
recent statewide waiver applications was the 1982 initiation of a
managed care program in Arizona, a state that had not participated
previously in Medicaid. On a number of occasions, the Congress and
the executive branch have extended the Arizona demonstration
authority.
Section 1115 waivers are significant in two regards: they allow
states greater flexibility to test cost containment
strategies--namely alternative delivery arrangements such as
capitated managed care--and they allow states to expand program
eligibility beyond traditional Medicaid populations. Since 1993, the
Health Care Financing Administration (HCFA), which oversees the
Medicaid program, has approved seven statewide demonstration waivers
for implementation: Florida, Hawaii, Kentucky, Ohio, Oregon, Rhode
Island, and Tennessee. Florida and Ohio are awaiting state
legislative approval and have not yet implemented their programs. In
Kentucky, state legislators refused to authorize waiver
implementation because they doubted that managed care savings would
be sufficient to offset the costs associated with coverage of
additional groups. Another 12 states have applications pending, and
3 more states have held discussions with HCFA about statewide
demonstrations.
States are seeking waivers of managed care restrictions because they
believe capitated managed care can cut health care costs. As of June
1994, 15 percent of program beneficiaries were signed up with
capitated health plans, the fastest growing type of Medicaid managed
care enrollment. However, some of the managed care restrictions
being waived were partly a response to quality of care problems that
emerged from states' experiments with managed care during the 1970s.
When waiving these managed care restrictions through 1115
demonstration authority, HCFA has required states to implement
improved quality assurance systems, including the collection of
encounter data.
Because only two states have completed their first full year of
waiver implementation, it is too early to assess the impact of the
shift to managed care on beneficiary access to services and quality
of care. However, the experience of Tennessee is instructive
regarding the need to proceed cautiously when implementing a
capitated managed care program statewide. At the outset, Tennessee
start-up problems threatened beneficiary access to services. The
managed care networks were incomplete when beneficiaries were
required to choose a plan and even on "opening day," many physicians
had not yet determined which networks to join. During the first
months of implementation, some providers reported slow or no payments
for services. Also, a significant number of beneficiaries have
expressed dissatisfaction with Tennessee's managed care program. The
most recent beneficiary survey reported that 45 percent of enrollees
who had previously received care under Medicaid's fee-for-service
program were less than satisfied with TennCare.
SECTION 1115 DEMONSTRATIONS
COULD INCREASE FEDERAL
SPENDING
-------------------------------------------------------- Chapter 0:4.3
The administration asserts that all of the approved statewide 1115
waivers are "budget neutral"--that is, federal spending for the
demonstrations will not exceed what would have been spent if the
states continued to operate their previous Medicaid programs.
However, the administration has shown considerable flexibility in
establishing 1115 waiver spending limits. For example, the cost of
covering populations that could have been but were not
covered--so-called "hypotheticals"--under the predemonstration
program, have been considered budget neutral. Such flexibility has
made it easier for demonstrations to meet the administration's
redefined budget-neutrality test.
In contrast, GAO's initial test of budget neutrality involved a
comparison of the approved waiver spending limits to a current
services benchmark. This benchmark is defined by applying the
administration's national projection of Medicaid spending growth
during the life of the waiver to actual spending in the year prior to
waiver approval. This comparison indicates that federal spending has
the potential to be greater in Florida, Hawaii, and Oregon than it
would have been in the absence of waivers, whereas federal spending
over the life of Tennessee's waiver may be less than it otherwise
would have been. The potential net spending impact associated with
the four state waivers we analyzed is likely to be small relative to
overall federal Medicaid spending. However, the impact of section
1115 demonstrations could increase substantially if the waivers
pending or proposed by several additional states--or even a few large
states--are not budget neutral.
Because of the bilateral nature of the negotiations between states
and the administration, and because of the discretion granted to the
executive branch in approving 1115 waivers, the terms and conditions
for each state are unique. As a result, the federal government's
financial liability varies with each state, depending on the waiver
terms negotiated. Most states have agreed to limits on federal
spending per person, or per capita caps. Two have agreed to a total
spending limit, or aggregate cap. Although for some states federal
liability is potentially greater than would be the case without the
waiver, it is limited in that each waiver agreement ensures a ceiling
on federal spending and holds some financial risk for the states.
OBSERVATIONS ON MEDICAID'S
FUTURE
-------------------------------------------------------- Chapter 0:4.4
Over 33 million low-income women, children, elderly, blind, and
disabled Americans depend upon health care made possible by the
Medicaid program. However, the program's double-digit spending
growth rate imperils efforts to bring the federal deficit under
control. Consistent with the Committee's interest in constraining
federal spending, states believe they need the flexibility to manage
their respective programs. Requiring states to obtain waiver
approval in order to aggressively pursue managed care strategies may
hamper their cost containment efforts. Yet, because current program
restrictions on managed care were designed to reinforce quality
assurance, in the absence of these restrictions, continuous oversight
of managed care systems is required to protect both Medicaid
beneficiaries from inappropriate denials of care and federal dollars
from payment abuses. Finally, GAO believes that the potential for
increased federal spending under future statewide demonstrations
warrants close scrutiny of section 1115 waiver approvals.
AGENCY COMMENTS
---------------------------------------------------------- Chapter 0:5
GAO discussed its findings with officials of HCFA's Medicaid Bureau,
the Office of Research and Demonstrations, and the Office of State
Health Reform. In addition, GAO discussed its approach to measuring
budget neutrality with officials at the Office of Management and
Budget (OMB). The administration's technical comments have been
reflected, as appropriate, in this report.
Administration officials agreed with GAO that their waiver review
policies, which they describe as more flexible, differ from those of
previous administrations. In particular, the administration's
approach to defining budget neutrality is characterized by (1)
flexibility in determining appropriate baseline expenditures and (2)
assessment of budget neutrality over the life of the waiver rather
than year by year. The administration disagrees with GAO's efforts
to develop a consistent methodology for assessing budget neutrality
by using OMB's national estimate of projected Medicaid spending.
Instead, it suggests that a "state-specific approach" that
incorporates states' historical experience is necessary to reflect
the "dramatic variation" that exists in state Medicaid programs.
This report notes, however, that recent growth in Medicaid
expenditures was influenced by a number of state strategies, such as
DSH payment schemes and federal eligibility mandates, that are
unlikely to be repeated. GAO believes that using unique methods in
each state has created the potential for budget-neutrality decisions
to be based on the technique most favorable to a particular state.
If state-specific reasons exist for future Medicaid growth to differ
from the national average, they should be well documented.
The administration disputes GAO's contention that the comprehensive
scope of the waiver demonstrations makes it controversial to
characterize them as experimental. GAO's view remains that
regardless of how innovative the elements of these demonstrations
are, the extension of health care coverage to several hundred
thousand people will make it difficult to terminate a demonstration
that fails to achieve its objectives. (The complete text of the
administration's comments is in app. IV.)
INTRODUCTION
============================================================ Chapter 1
In 1993, Medicaid, the government's health financing program for
certain low-income populations, cost the federal and state
governments $131 billion--almost $100 billion more than just a decade
ago.\1 During that same period, the number of Medicaid beneficiaries
increased from 22 million to 33 million. Although about
three-quarters of the beneficiaries are poor children and their
parents, two-thirds of the expenditures are for poor elderly, blind,
and disabled individuals. Funding is a shared responsibility between
the federal and state governments, with respective federal and state
shares determined through a statutory matching formula. Federal law
mandates coverage of certain medical services and population groups.
It also includes coverage options, allowing states to choose whether
to cover additional services or low-income population groups.
--------------------
\1 Amounts include both health services and administrative costs.
NOT 1, BUT 56 SEPARATE MEDICAID
PROGRAMS
---------------------------------------------------------- Chapter 1:1
Dramatic differences in Medicaid programs across states reflect
spending priorities of the states, their ability to pay, and
incentives inherent in the federal matching formula.\2 States partly
determine the extent of their Medicaid program when they set
eligibility requirements for receiving cash assistance (primarily Aid
to Families with Dependent Children--AFDC).\3 Under federal law, AFDC
recipients are automatically eligible for Medicaid. States also
determine program scope by selecting which optional services or
groups to include in their programs. With some discretion to modify
eligibility and benefit provisions, the proportion of the poor (as
defined by the federal poverty level) and near-poor served by
Medicaid varies greatly by state.\4 For example, for every 1,000
people with incomes under 150 percent of the federal poverty level,
Rhode Island serves 913 beneficiaries, whereas Nevada serves 284.
Nationally, in 1994 Medicaid covered medical services for 58 percent
of people under age 65 living in poverty.
State Medicaid programs also vary in the levels of benefits provided.
For example, some states limit their Medicaid coverage of inpatient
days (a required service) to a certain number per person. States
also have the discretion to drop their coverage of optional services,
such as dental or optical care. The average Medicaid benefit (that
is, total medical spending divided by total beneficiaries) ranges
from $2,372 in Mississippi to $7,286 in New York. (See fig. 1.1 for
grouping of states by their average benefit level and table 1.1 for
state-by-state data.) States with higher Medicaid benefits tend to
cover a greater proportion of poor people, relative to states with
lower benefits. Thus, the differences among states in spending per
poor person are even greater than the differences among states in
spending per beneficiary.
Figure 1.1: Average Spending
per Medicaid Beneficiary Varies
Across States
(See figure in printed
edition.)
Source: Medicaid Statistics, Program and Financial Statistics, 1993,
U.S. Department of Health and Human Services, Health Care Financing
Administration.
Table 1.1
Medicaid Expenditures by State, Fiscal
Year 1993
Federa
Federa l
l and Federa Medica Medica
state l id id
Medica medica Medica spendi spendi
id l id ng per ng per
spendi assist spendi person person
ng\a ance ng per in in
(thous percen benefi povert povert
State ands) tage\b ciary y y
-------------------- ------ ------ ------ ------ ------
Alabama $1,637 71.45 $3,139 $2,258 $1,616
,242
Alaska 295,38 50.00 4,539 5,680 3,034
4
Arizona 1,365, 65.89 3,379 2,220 1,481
046
Arkansas 1,031, 74.41 3,038 2,130 1,587
148
California 13,538 50.00 2,801 2,333 1,168
,038
Colorado 1,091, 54.42 3,890 3,084 1,687
709
Connecticut 2,274, 50.00 6,817 8,212 4,118
592
Delaware 252,99 50.00 3,670 3,466 1,742
3
District of Columbia 686,71 50.00 5,710 4,346 2,176
9
Florida 4,948, 55.03 2,836 1,974 1,088
988
Georgia 2,798, 62.08 2,930 3,045 1,898
657
Hawaii 380,66 50.00 3,462 4,183 2,098
8
Idaho 293,67 71.20 2,951 1,958 1,397
4
Illinois 4,981, 50.00 3,569 3,113 1,561
454
Indiana 2,815, 63.21 4,984 3,999 2,532
525
Iowa 987,20 62.74 3,413 3,404 2,141
0
Kansas 889,66 58.18 3,663 2,721 1,587
6
Kentucky 1,863, 71.69 3,017 2,443 1,754
697
Louisiana 3,493, 73.71 4,651 3,122 2,303
823
Maine 855,86 61.81 5,070 4,367 2,704
0
Maryland 1,960, 50.00 4,409 4,093 2,054
419
Massachusetts 4,131, 50.00 5,402 6,446 3,228
904
Michigan 4,362, 55.84 3,724 2,958 1,655
644
Minnesota 2,167, 54.93 5,093 4,283 2,358
025
Mississippi 1,196, 79.01 2,372 1,872 1,480
475
Missouri 2,251, 60.26 3,695 2,706 1,635
606
Montana 323,27 70.92 3,631 2,545 1,822
1
Nebraska 564,16 61.32 3,426 3,338 2,052
9
Nevada 423,44 52.28 4,789 3,003 1,578
7
New Hampshire 417,62 50.00 5,264 3,729 1,874
7
New Jersey 4,706, 50.00 5,930 5,434 2,724
049
New Mexico 571,20 73.85 2,373 2,026 1,510
0
New York 19,980 50.00 7,286 6,703 3,360
,838
North Carolina 2,896, 65.92 3,224 2,998 1,982
330
North Dakota 269,67 72.21 4,343 3,852 2,795
5
Ohio 5,179, 60.25 3,474 3,545 2,139
121
Oklahoma 1,089, 69.67 2,819 1,646 1,153
730
Oregon 955,60 62.39 2,938 2,633 1,647
5
Pennsylvania 5,612, 55.48 4,589 3,512 1,953
714
Rhode Island 829,02 53.64 4,337 7,676 4,123
6
South Carolina 1,682, 71.28 3,576 2,481 1,772
379
South Dakota 266,29 70.27 3,826 2,611 1,859
4
Tennessee 2,675, 67.57 2,943 2,681 1,812
390
Texas 7,118, 64.44 3,084 2,241 1,448
558
Utah 477,62 75.29 3,224 2,353 1,773
4
Vermont 255,47 59.88 3,171 4,330 2,605
6
Virginia 1,791, 50.00 3,111 2,858 1,435
773
Washington 2,316, 55.02 3,657 3,654 2,017
480
West Virginia 1,200, 76.29 3,459 3,001 2,290
412
Wisconsin 2,114, 60.42 4,489 3,325 2,014
971
Wyoming 134,79 67.11 2,914 2,106 1,423
3
============================================================
Total $126,4
05,107
National average 57.34\ $3,870 $3,223 $1,846
c
------------------------------------------------------------
\a Includes payments made to disproportionate share hospitals (DSH).
Excludes federal and state administrative costs.
\b The FMAP is the federal government's share of each state's
payments for services.
\c Weighted average.
Source: Medicaid Statistics, Program and Financial Statistics, 1993,
and Bureau of the Census.
--------------------
\2 Participation in the Medicaid program by states is voluntary.
However, all 50 states plus the District of Columbia and American
Samoa, Guam, the Northern Mariana Islands, Puerto Rico, and the
Virgin Islands have elected to administer Medicaid programs. We
refer to all 56 entities as "states" for this report.
\3 In 1993, the qualifying level for AFDC varied across states from
17 to 93 percent of the federal poverty level, which in 1993 was
income of $11,890 for a family of three.
\4 States also have discretion in setting their own standards and
methods for reimbursement of Medicaid services.
SERVICES FOR ELDERLY AND
DISABLED CONSUME GREATEST SHARE
OF FUNDING
---------------------------------------------------------- Chapter 1:2
By far, the majority of Medicaid funds are paid on behalf of poor
elderly, blind, and disabled individuals. Representing only slightly
more than a fourth of all Medicaid beneficiaries, these groups
incurred about 66 percent ($82.2 billion) of Medicaid's 1993
expenditures, as shown in figure 1.2.\5 The average per person
spending for these beneficiaries exceeded $9,200--four times more
than was spent on other adults ($2,292) and seven times more than was
spent on other children ($1,360). The per person spending for all
Medicaid beneficiaries averaged nearly $3,900; in contrast, health
spending per person for the U.S. population averaged about $3,300.
Figure 1.2: Cost of Serving
Disabled and Elderly High
Compared With That of Serving
Other Children and Adults
(See figure in printed
edition.)
Source: The Kaiser Commission on the Future of Medicaid.
In 1993, about 35 percent of Medicaid spending was for long-term
care, including nursing home care payments. Medicaid pays about 52
cents of every nursing home care dollar. About 46 percent of all
Medicaid spending was for acute care. (See fig. 1.3 for a breakdown
of Medicaid spending.)
Figure 1.3: Long-Term Care
Accounts for One-Third of
Medicaid's Spending for
Services
(See figure in printed
edition.)
\a Payments to hospitals that serve a disproportionate number of
Medicaid and other low-income patients with special needs. There are
no federal restrictions on how hospitals use these payments.
Source: The Kaiser Commission on the Future of Medicaid.
--------------------
\5 These figures represent national averages--percentages for
specific states may be higher or lower.
FEDERAL MEDICAL ASSISTANCE
PAYMENTS BASED ON PER CAPITA
INCOME
---------------------------------------------------------- Chapter 1:3
The financing of Medicaid is shared by the federal government and the
states. The federal share paid to a state is based on the federal
medical assistance percentage (FMAP). Determined separately for each
state, the FMAP ranges from a minimum of 50 percent to a maximum of
83 percent of total Medicaid costs.
The formula for determining the FMAP is based on a state's per capita
income. It is designed to provide a higher federal match to states
with low per capita incomes. In fiscal year 1993, Mississippi, West
Virginia, and Utah had the highest federal match percentages; each
received more than 75 percent of its Medicaid program funding from
the federal government. Alternatively for that year, the federal
government paid 50 percent of Medicaid costs, the minimum match, in
12 states and the District of Columbia.
In spite of the formula's sliding match rate, federal spending per
Medicaid beneficiary tends to be highest in states with high per
capita incomes. Because high-income states tend to offer relatively
generous benefits, the absolute amount of federal spending per
beneficiary is high, even though the match rate is only 50 percent.\6
For example, New York--a high-income state with a 50-percent
match--receives $3,600 per beneficiary in federal aid, while
Mississippi--a low-income state with approximately an 80-percent
match--receives $1,900 per beneficiary in federal aid.\7
Reporting on this subject in 1983, 1990, and 1993, we noted that the
FMAP formula could be improved to meet federal goals.\8 The federal
assistance formula is designed to reduce differences among the states
in medical care coverage of the poor and distribute fairly among the
states the burden of financing program benefits. However, the
formula's reliance on a state's per capita income does not adequately
address these equity goals. In fiscal year 1991, for example, the
District of Columbia and Nevada received the same FMAP, despite the
District's greater proportion of people eligible for Medicaid. Our
prior work in this area concluded that the formula could better
adjust for states' different needs and varying abilities to fund
benefits by considering a state's number of persons in poverty and
more comprehensive measures of taxing capacity.
--------------------
\6 The top six states in terms of spending per beneficiary (New York,
Connecticut, New Jersey, District of Columbia, Massachusetts, and New
Hampshire) have high per capita incomes, relative to the national
average, and receive only a 50-percent federal share.
\7 To the extent that the cost of medical services varies with state
per capita income, the dollar differences somewhat overstate the
differences in real Medicaid services provided by states.
\8 Medicaid: Changing Medicaid Formula Can Improve Distribution of
Funds to States (GAO/GGD-83-27, Mar. 9, 1983), Medicaid Formula:
Fairness Could Be Improved (GAO/T-HRD-91-5, Dec. 7, 1990), and
Medicaid: Improving Funds Distribution (GAO/HRD-93-112FS, Aug. 20,
1993).
SECTION 1115 WAIVERS PERMIT
MAJOR PROGRAM CHANGES WITHOUT
CONGRESSIONAL DEBATE
---------------------------------------------------------- Chapter 1:4
In recent years, many states have been experimenting with health
financing reforms. Under Medicaid law, states proposing to implement
payment and service delivery innovations outside the range allowed by
federal rules must seek approval to waive certain federal
requirements. Until recently, though, Medicaid experiments have been
relatively limited.
However, states have been seeking approval for broader changes under
section 1115 waivers, so called because of the statutory provision
authorizing them,\9 which offer more open-ended opportunities to make
major changes in Medicaid than program waivers.\10 Changes proposed
in section 1115 waiver applications are considered demonstration
projects. Unlike the routine approvals for program waivers,
approvals for demonstration waivers are subject to much greater
discretion by the Secretary of the Department of Health and Human
Services (HHS). (The Health Care Financing Administration (HCFA) and
the Office of Management and Budget (OMB) have also been key players
in the waiver approval process.) Section 1115 waivers can exempt
states from compliance with normal Medicaid requirements and allow
them to receive federal funding for expenditures normally not
eligible for such assistance. HHS requires that waiver applications
include a formal research or experimental methodology, and
historically waivers have not been considered automatically
renewable. Section 1115 waivers typically involve limited research
projects, but since 1993 several statewide demonstration projects
have been approved.
Among other things, recent section 1115 waivers have allowed states
to mandate enrollment of all their beneficiaries in
Medicaid-participating managed care plans and to prevent
disenrollment for specific periods of time, usually 6 to 12 months.
States with recently approved waivers have proposed expanding
coverage of their indigent citizens by recycling savings anticipated
in changing from a fee-for-service to a managed care system and
redirecting other Medicaid and state funds. If these states'
demonstrations are implemented as planned, the states would be
covering some low-income individuals that would not otherwise qualify
for Medicaid. We discuss the significance of these waivers to the
restructuring of the Medicaid program in chapter 3 and their
potential impact on the federal budget in chapter 4.
--------------------
\9 The waiver authority resides in section 1115(a) of the Social
Security Act, 42 U.S.C. 1315(a), which gives the Secretary of HHS
broad authority to waive statutory requirements for Medicaid and
other programs, such as AFDC, to permit states to run experimental,
pilot, or demonstration projects "likely to assist in promoting the
[program's] objectives." States may be exempted from compliance with
program requirements, or receive federal funding for expenditures
normally not eligible for such assistance, or both.
\10 There are various types of waivers under Medicaid, and those
typically referred to as "program" waivers (approval to waive certain
requirements for a specific population or geographic area) are
designed for routine approval, as long as states meet certain
qualifications. States may apply for periodic renewals. In 1994 all
states, except for Arizona and the District of Columbia, operated one
or more innovations under program waivers.
OBJECTIVES, SCOPE, AND
METHODOLOGY
---------------------------------------------------------- Chapter 1:5
The House Budget Committee asked us to examine (1) the nature of
federal and state Medicaid spending in recent years, (2) some states'
efforts to contain Medicaid costs and expand coverage through waivers
of certain federal requirements, and (3) the potential impact of the
waivers on federal spending and on Medicaid's program structure
overall.
COLLECTING INFORMATION ON
MEDICAID SPENDING AND
WAIVERS
-------------------------------------------------------- Chapter 1:5.1
To address the issue of federal and state Medicaid spending, we
analyzed data contained in HCFA statistical reports and also reviewed
and synthesized information from research articles and major studies
that focused on Medicaid costs. To obtain information on states'
efforts to contain Medicaid costs and expand coverage through
waivers, we interviewed Medicaid program and other state agency
officials, providers, advocacy groups, and state legislators in five
of the eight states with approved waivers. We also interviewed HCFA
officials responsible for examining waiver proposals and reviewed
documentation for the approved and pending projects as well as
statistical and other reports on the Medicaid program.
EXAMINING THE POTENTIAL
IMPACT OF WAIVERS
-------------------------------------------------------- Chapter 1:5.2
We determined the potential impact of waivers on federal spending by
developing and applying a consistent framework that compares spending
limits approved in each waiver with a benchmark incorporating current
services budgeting concepts.\11 To eliminate the effect of
accelerated program growth from past years--acceleration due in part
to high medical inflation, large increases in state disproportionate
share hospital (DSH) payments, and mandated expansions in program
eligibility--we compared the waiver spending limits with a projection
of spending for each state's current Medicaid program over the life
of the waiver. To make these projections we used the
administration's estimated rate of growth of Medicaid current
services outlays for the nation as a whole. These estimated growth
rates are based only on expected increases to medical inflation and
program population growth--without any change in eligibility
requirements.
To make the current services projection for each state's Medicaid
program (what we call our budget-neutrality benchmark) we multiplied
each state's base-year estimates of Medicaid funding (for the segment
of the population to be covered by their waiver) by the
administration's estimates of national program growth made when HCFA
approved the waiver. We adjusted our baseline, when necessary, to
reflect constraints on the growth of DSH payments in each state's
Medicaid program. The actual process varied slightly, depending on
whether the state's waiver was subject to an aggregate or per capita
funding limit.
Once we had determined our administration-based estimate of current
services outlays, we compared it with the estimated program funding
from the waiver agreements. In states with aggregate caps, we
compared our estimates with actual program caps. In states with per
capita caps, we compared our estimates with state program funding
levels consistent with the enrollment figures found in the waiver
applications, recognizing that actual funding ceilings would be
higher or lower, depending on actual enrollment.
We subtracted our annual, budget-neutrality benchmark figures from
the waiver agreement figures and then discounted this amount to
determine the present value of potential annual savings or increased
funding. The difference--if positive--showed the amount of program
funds above our benchmark amount that might be needed to finance the
program and--if negative--showed the amount the program might save.
We did our work for this report and other related studies between
August 1994 and March 1995 in accordance with generally accepted
government auditing standards.
--------------------
\11 According to HCFA, each approved waiver is budget neutral, that
is, the waiver programs--including the expanded populations--will
cost no more to operate than the traditional Medicaid programs would
have in each state over the 5 years of the demonstrations. However,
in determining the theoretical funding level for continuing the
traditional Medicaid program, HCFA and the states made assumptions
that we do not believe are consistent with the concept of "budget
neutrality," such as including persons who states could make, but are
not currently, eligible (frequently referred to as "hypotheticals")
in the program baseline and allowing for program growth that more
closely approximates historically high levels than the lower
projection of future growth.
DESPITE A RECENT SLOWDOWN,
MEDICAID CONSUMES A GROWING SHARE
OF THE FEDERAL BUDGET
============================================================ Chapter 2
Medicaid has become one of the fastest growing components of federal
and state budgets. In part, this is the result of program
expansions. After remaining relatively stable for a decade, the
number of Medicaid beneficiaries began to climb rapidly after the
mid-1980s.\12 Since that time, the number of beneficiaries served
annually has increased by over 11 million, reaching 33.4 million in
1993. However, costs have grown even faster than the number of
beneficiaries. (See fig. 2.1.) Between 1985 and 1993, federal
Medicaid expenditures grew each year, on average, by 16 percent.
During this time, state Medicaid expenditures grew at a slightly
slower rate of 15 percent. In 5 years, Medicaid spending is expected
to exceed $260 billion annually. (See fig. 2.2.)
Figure 2.1: Medicaid Spending
Outpaced Rise in Beneficiary
Numbers from 1983 to 1993
(See figure in printed
edition.)
Sources: Medicaid Source Book: Background Data and Analysis (A 1993
Update) (A Report Prepared by the Congressional Research Service for
the Use of the Subcommittee on Health and the Environment of the
Committee on Energy and Commerce, U.S. House of Representatives,
Jan. 1993) and Medicaid Statistics, Program and Financial Statistics
for 1991-93.
Figure 2.2: Medicaid Expected
to Grow to $260 Billion by Year
2000
(See figure in printed
edition.)
Sources: Medicaid Source Book: Background Data and Analysis (A 1993
Update), Medicaid Statistics, Program and Financial Statistics for
1991-93, and Congressional Budget Office estimates.
--------------------
\12 Between 1975 and 1985 the total number of beneficiaries varied by
less than 1.5 million, ranging from 21.5 million to 22.8 million.
MEDICAID COVERAGE HAS EXPANDED
SINCE 1984
---------------------------------------------------------- Chapter 2:1
The federal government influences total Medicaid spending indirectly
by specifying the percentage of state Medicaid expenditures the
federal government will pay (higher matching rates encourage greater
state spending), and directly by specifying some eligibility and
benefit requirements. Between 1984 and 1990, the Congress expanded
Medicaid coverage by mandating that states serve certain low-income
groups and provide certain services, and allowing optional coverage
for other low-income groups and services.\13 (See app. I.)
The mandates extended coverage primarily to additional low-income
pregnant women, children, and Medicare beneficiaries. The mandates
also extended transitional Medicaid coverage for families leaving
AFDC for gainful employment, established a new categorically needy
coverage group for severely impaired individuals under the age of 65,
and added coverage of certain chronically ill or disabled
individuals.\14 The Kaiser Commission on the Future of Medicaid noted
that between 1990 and 1992, enrollment increased by about 6 million
individuals, with about one-third of this increase attributable to
pregnant women and children.\15 Since some states provided benefits
when coverage of these groups was optional, the mandates tended to
equalize benefits among states.
Because both the mandates and options tied coverage to the federal
poverty level, which is uniform across states, they greatly increased
the number of persons receiving Medicaid who were not eligible for
cash assistance. (See fig. 2.3.) This resulted in a decline in the
proportion of beneficiaries eligible for Medicaid through a
cash-assistance-related category. In 1993, about 60 percent of the
beneficiaries were eligible for Medicaid coverage under a
cash-assistance-related category, down from 68 percent in 1990.
Figure 2.3: Enrollment
Increases Highest for Those Not
Receiving Cash Assistance
(See figure in printed
edition.)
\a Individuals who have incomes too high to qualify for cash
assistance, but who have incurred medical expenses that depleted
their incomes and resources to levels that make them needy. States
are permitted, but not required, to cover medically needy
individuals.
Sources: Medicaid Source Book: Background Data and Analysis (A 1993
Update) and Medicaid Statistics, Program and Financial Statistics for
1991-93.
--------------------
\13 The Congress also required states to increase certain aspects of
service coverage.
\14 Access to Medicaid coverage for disabled children has increased
since 1990, when the Supreme Court ruled that disability standards
for children may be no more narrow than those applied to adults.
Sullivan v. Zebley, 493 U.S. 521 (1990). As a result, eligibility
criteria for children are based on the child's developmental delay
and limitations on the child's ability to engage in age-appropriate
activities of daily living. This has increased the number of
children classified as disabled. Prior to 1990, the same disability
criteria that applied to adults, which emphasize inability to engage
in employment, were also applied to children.
\15 Kaiser Commission Update: Medicaid Spending Growth, 1990 to
1992, The Kaiser Commission on the Future of Medicaid, Policy Brief,
No. 1 (Washington, D.C.: July 1991).
MEDICAID GREW FASTER THAN
MEDICAL INFLATION, OTHER
SPENDING
---------------------------------------------------------- Chapter 2:2
During most of the 1980s, spending for Medicaid, private health
insurance, and Medicare grew at an annual rate of 10 to 11 percent.
In 1989, however, Medicaid's spending growth began to rise sharply.
In 1991 Medicaid's 27-percent annual growth rate was over three times
greater than the growth rate for national health expenditures and
over four times greater than the rate for Medicare. (See fig. 2.4.)
In 1992 Medicaid's annual growth rate reached a record high of about
29 percent. Even though it fell to 11 percent in 1993, Medicaid's
annual growth rate continued to outpace the consumer price index and,
except for Medicare, other key indicators used to measure health care
spending. Medicaid consumes slightly more than 14 cents of every
personal health care dollar.
Figure 2.4: Dramatic Increases
in Medicaid Expenditures Slowed
in 1993
(See figure in printed
edition.)
Sources: Data for Medicaid are from Medicaid Source Book:
Background Data and Analysis (A 1993 Update). Data for Medicare are
from annual reports of the trust funds' trustees, 1989-93. Data for
national health expenditures are from Department of Health and Human
Services National Health Expenditures for 1993. Data for CPI and
Medical Care Index are from Bureau of Labor Statistics.
Figure 2.5 illustrates spending by Medicaid and Medicare since 1981.
Historically, Medicaid spending on health care services for the poor
has been significantly less than Medicare spending for the nation's
elderly. In the 1980s, the gap widened, and by 1990, Medicare
spending exceeded Medicaid's by over $37 billion. By 1992, however,
the difference had narrowed to $14 billion.
Figure 2.5: Medicaid Spending
Approached Spending for
Medicare in 1992
(See figure in printed
edition.)
Sources: Medicaid data are from Medicaid Source Book: Background
Data and Analysis (A 1993 Update). Medicare data are from annual
reports of the trust funds' trustees, 1981-93.
Medicaid spending has also increased sharply when compared with
spending for other federal social programs. (See fig. 2.6.) In
1993, Medicaid consumed 46 percent of all the funds spent on
means-tested programs, over three times more than its closest
competitor--Food Stamps. By the year 2000, federal spending for
Medicaid is expected to reach $149 billion, or about 51 percent of
the dollars spent on means-tested programs.
Figure 2.6: Medicaid Increased
More Than Other Social Programs
From 1985 to 1993
(See figure in printed
edition.)
Source: Budget of the United States Government, Historical Tables,
Fiscal Year 1996.
MEDICAID SPENDING GREW FASTER
THAN ENROLLMENT
---------------------------------------------------------- Chapter 2:3
Enrollment growth--over 11 million between 1983 and 1993--does not
fully explain the rise in Medicaid spending. For example, the Kaiser
Commission estimated that enrollment growth accounted for less than
half of the increase in Medicaid costs during the period 1990 to
1992. The federal mandates of the 1980s that added low-income
pregnant women and low-income children to the rolls were a minor
factor in increasing costs, accounting for about 6.5 percent of
spending growth. Although the mandates added a large number of
beneficiaries, the populations added--low-income pregnant women and
children--are relatively inexpensive to serve. In contrast,
low-income elderly, disabled, and blind beneficiaries were
responsible for about 25 percent of the total expenditure growth,
even though they accounted for only 18 percent of the enrollment
growth from 1990 to 1992.
In addition to enrollment increases, medical price inflation, higher
provider reimbursements (that, to some extent, have been fueled by
the Boren amendment\16 ), utilization growth, and creative financing
mechanisms have also contributed to the significant growth in
Medicaid expenditures. The importance of each of these factors in
explaining cost growth depends upon the specific time period
selected. For example, in 1993 the Kaiser Commission reported that
enrollment expansions, medical price inflation, and increases in
expenditures per beneficiary above inflation each accounted for about
one-third of the $37 billion increase in annual Medicaid expenditures
from 1988 through 1991.\17
--------------------
\16 Provider suits brought under the Boren Amendment, 42 U.S.C.
1396a(13), have been a major factor pressuring states to increase
payment rates. Enacted in the early 1980s, the Boren Amendment
modified federal requirements related to Medicaid payment rates to
hospitals and nursing homes. The intent was to provide states with
greater flexibility, which some states utilized, for example, to
initiate prospective payment systems. Specifically, the Boren
Amendment requires states to provide payments "which the state finds
. . . are reasonable and adequate to meet the costs which must be
incurred by efficiently and economically operated facilities in order
to provide care and services in conformity with applicable . . .
standards and to assure that individuals eligible for medical
assistance have reasonable access to inpatient hospital services of
adequate quality." Particularly in recent years, states have been
dogged by provider lawsuits forcing them to better justify or raise
their Medicaid payment rates to hospitals and nursing homes.
\17 The Medicaid Cost Explosion: Causes and Consequences (Baltimore,
Maryland: The Kaiser Commission on the Future of Medicaid, Feb.
1993).
DISPROPORTIONATE SHARE HOSPITAL
PAYMENTS RESPONSIBLE FOR HUGE
RISE IN 1991-92 SPENDING
---------------------------------------------------------- Chapter 2:4
Disproportionate share hospital (DSH) payments were the most
important cost driver in 1991 and 1992, when Medicaid spending grew
by 27 percent and 29 percent, respectively. These are supplemental
payments to hospitals that serve large numbers (a disproportionate
share) of Medicaid and other low-income patients. The payments are
intended to partially reimburse hospitals for the cost of providing
care not covered by public or private insurance. In 2 years, DSH
payments grew from slightly less than $1 billion in 1990 to $17.4
billion in 1992 and represented about $1 of every $7 Medicaid spent
on medical services.
DSH payments grew rapidly because they were part of creative
financing mechanisms that allowed states to gain additional federal
dollars and effectively increase the federal government's share of
Medicaid funding. Beginning in the mid-1980s, states were allowed to
use revenue raised from "provider-specific" taxes--that is, taxes
imposed on hospitals serving Medicaid patients--and "voluntary
contributions" (called donations), as part of the state share
eligible for federal matching funds. States then returned to
providers the funds collected from such taxes and donations along
with part of the matching federal payments. In some cases, a portion
of the federal matching funds was then redirected to state general
revenues and spent on nonhealth-care services. This swapping and
redirecting of revenues among providers, the state, and the federal
government resulted in increased federal spending, increased funds
for providers, and--in some cases--additional revenue for states'
treasuries. (See app. II for examples of how states used
provider-specific taxes and donations to obtain federal matching
dollars without actually spending state funds.)
DSH payments are now capped as a result of the Medicaid Voluntary
Contribution and Provider-Specific Tax Amendments of 1991 and will
not be a significant cost driver in the future. In 1993, DSH
payments fell slightly to about $16.7 billion. The 1991 amendments
also severely restricted states' ability to use provider-specific
taxes and donations (but not intergovernmental transfers) as a source
of matching funds, and capped DSH payments at 12 percent of the
Medicaid program.\18 States that relied heavily on this financing
mechanism are now seeking a new revenue source, cutting their
Medicaid programs, or dramatically restructuring their programs under
a section 1115 waiver (see ch. 3).
--------------------
\18 Legislation passed in 1993 placed additional restrictions on DSH
payments. (See app. II.)
MEDICAID'S SHARE OF FEDERAL
BUDGET EXPECTED TO GROW
---------------------------------------------------------- Chapter 2:5
Medicaid's tremendous spending growth of the late 1980s and early
1990s is not likely to recur. In 1994, Medicaid spending is expected
to have grown about 8 percent. The Congressional Budget Office
estimates that program spending will grow by 10.7 percent annually
over the next 5 years. As in the past, future growth will be
determined by medical inflation, provider reimbursement, utilization,
and enrollment increases. Although no new eligibility or benefit
mandates are foreseen, some of those already legislated are still
being phased in. Mandatory coverage of poor children under the age
of 19 will not be fully effective until 2002. Furthermore, the
number of disabled individuals covered by Medicaid, although
relatively small, is rising rapidly (12 percent between 1992 and
1993) and will increase costs.
SOME STATES USE MEDICAID WAIVERS
TO INCREASE MANAGED CARE AND
EXPAND COVERAGE
============================================================ Chapter 3
In the past, some states solved their health care financing problems
by capitalizing on the use of federal subsidies. Most notably, the
DSH payment mechanisms mentioned earlier were used to augment a
state's total Medicaid spending to obtain higher federal matching
payments: the higher the state's own spending on Medicaid, the
higher the federal match. However, since the limits on state DSH
payments that were enacted in 1991, continuing budget shortfalls have
pressured states to seek alternative financing measures. Now some
states are seeking waivers from federal requirements in order to make
greater use of alternatives such as prepaid, or HMO-type, managed
care.
Under the broad authority of section 1115 of the Social Security Act,
HHS has allowed some states to waive certain federal requirements in
exchange for the states' expansion of Medicaid coverage and agreement
to specific federal funding limits. The waiver authority, as
recently granted, permits states to mandate enrollment in managed
care health plans; subsidize health care coverage for people with
incomes above the federal poverty level; and use funding from other
sources, such as public and mental health programs, to finance these
efforts.
As of April 1995, 8 states have HHS-approved waivers, 12 have
applications pending, and 3 have made inquiries about submitting
waiver
applications.\19 \20 (See table 3.1 and fig. 3.1.) Together, these
23 states have about 50 percent of the Medicaid beneficiaries in the
country.
Table 3.1
Section 1115 Statewide Demonstration
Waivers Applied for Since 1991, by
Submission Date
State Submission Approval Implementation
------------- ------------- ------------- ---------------
Oregon Aug. 1991\a Mar. 1993 Feb. 1994
Kentucky Mar. 1993 Dec. 1993 \b
Hawaii Apr. 1993 July 1993 Aug. 1994
Tennessee June 1993 Nov. 1993 Jan. 1994
Rhode Island July 1993 Nov. 1993 Aug. 1994
Florida Feb. 1994 Sept. 1994 \b
Ohio Mar. 1994 Jan. 1995 \b
South Mar. 1994 \c
Carolina
Massachusetts Apr. 1994
New Hampshire June 1994
Missouri June 1994
Delaware July 1994
Minnesota July 1994
Illinois Sept. 1994
Louisiana Jan. 1995
Oklahoma Jan. 1995
Vermont Feb. 1995
New York Mar. 1995
Kansas Mar. 1995
------------------------------------------------------------
\a Oregon's initial proposal was denied in August 1992. The state
revised and resubmitted the proposal, which was approved in March
1993.
\b Awaiting state legislature approval.
\c HCFA has approved South Carolina's waiver proposal framework.
However, certain issues must be resolved before the state is allowed
to implement its demonstration program.
Source: HCFA.
Figure 3.1: HCFA Has Approved
Eight Statewide Demonstration
Projects
(See figure in printed
edition.)
Source: HCFA.
--------------------
\19 HCFA has approved seven statewide section 1115 waivers since
1993. The eighth state, Arizona, has operated its Medicaid program
under a section 1115 waiver since 1982.
\20 Texas, Utah, and Washington have inquired about submitting waiver
applications with HCFA.
WAIVERS EASE RESTRICTIONS ON
USE OF MANAGED CARE FOR
MEDICAID POPULATION
---------------------------------------------------------- Chapter 3:1
WAIVERS INCREASE MANAGED
CARE OPTIONS
-------------------------------------------------------- Chapter 3:1.1
To constrain rising health care costs, states are increasingly
turning to the mandatory enrollment of portions of their Medicaid
population into prepaid managed care delivery plans that limit
enrollees to certain provider networks as a way to control
utilization and costs. Because federal Medicaid law requires that
beneficiaries have freedom to choose among providers willing to
accept Medicaid reimbursement, states have had to seek approval to
waive this freedom-of-choice requirement. The Secretary of HHS may
waive this requirement under two sections of the Social Security Act.
Section 1915(b) allows states, among other things, to implement a
primary care case management system, to contract with HMOs, and to
selectively contract with a limited set of providers. Section 1115
waivers, which are more difficult to obtain, allow states greater
flexibility in the use of managed care options, particularly HMO-type
delivery. As described in chapter 1, under section 1115 waivers,
states can establish criteria to designate prepaid plans that serve
only Medicaid enrollees and can prohibit disenrollment from these
plans for a certain period of time--two provisions not generally
permitted by Medicaid law.
STATE EXPERIENCE INDICATES
NEED FOR CAREFUL PLANNING
WHEN IMPLEMENTING NEW
MANAGED CARE NETWORKS
-------------------------------------------------------- Chapter 3:1.2
States are seeking waivers of managed care requirements because they
believe managed care financing and delivery systems can cut health
care costs as well as improve access to and quality of care. In our
1993 studies of Medicaid managed care and private sector managed
care, we reported that a combination of factors rendered findings of
cost savings inconclusive.\21 Because the term managed care refers to
a range of financing arrangements--including traditional
fee-for-service plans with very few care management features, such as
prior approval of hospitalization--our studies of cost savings are
quick to distinguish among the various forms of managed care. For
the most service-restrictive form of managed care--the "capitated"
HMO--reported evidence of lower utilization is straightforward. This
promise of lower utilization has made HMO-type plans attractive to
the states.\22
As of June 1994, 15 percent of Medicaid beneficiaries were enrolled
in fully capitated health plans. In one year (June 1993 to June
1994), Medicaid full-risk capitation enrollment grew by 91 percent,
in part because of the statewide managed care demonstrations obtained
through the section 1115 waivers. In the past, the elderly and
disabled Medicaid population, which consumes the largest share of
state Medicaid budgets (about 66 percent), has rarely been covered by
managed care programs. Increasingly, however, states are including
or plan to include these populations as well.
When we and others reported in the 1970s on quality of care problems
and other abuses in some Medicaid managed care experiments, the
Congress placed restrictions on HMO enrollment for Medicaid
beneficiaries.\23 In general, these restrictions were designed to
ensure that HMOs provide Medicaid beneficiaries a standard of care
equal to that provided to private patients. Under section 1115
waivers, HCFA can waive these restrictions to help states enlist HMOs
to serve Medicaid beneficiaries. HCFA can waive the "75-25 rule,"
which specifies that HMOs serving Medicaid beneficiaries must have at
least 25 percent of their enrollment consist of private patients
(with certain limited exceptions), a rule that--in effect--prohibits
Medicaid-only HMOs.\24 HCFA can also waive a federal provision
permitting beneficiaries to terminate their enrollment in an HMO at
any time; this provision prevents the "lock-in" of beneficiaries that
are dissatisfied with a plan.\25 From the beneficiary's perspective,
these provisions provide some protection against the possibility of
being enrolled by an HMO seeking excessive profit at the expense of
quality. From the HMO's perspective, however, the provision allowing
beneficiaries to disenroll at will makes planning for financial
stability difficult and therefore makes the enrollment of Medicaid
patients less attractive.\26
In February 1995 we testified that, with the Medicare program (which
covers medical services for the nation's elderly and certain other
groups) HCFA has not aggressively enforced compliance with federal
HMO standards.\27 Similar concerns arise about enforcement of
standards for plans that enroll Medicaid beneficiaries, particularly
because large-scale enrollment in prepaid HMOs is a recent occurrence
and state Medicaid departments have limited experience in overseeing
such delivery systems. Advocacy groups in some states with waivers
have stated that while HMO-type managed care may potentially expand
beneficiaries' access to providers, quality of care and access
problems will go undetected without adequate beneficiary education
and vigorous state and federal oversight.\28
The recent experience of Tennessee's section 1115 demonstration
program illustrates the need to proceed cautiously when implementing
a capitated HMO program statewide.\29 In response to escalating
Medicaid costs, Tennessee sought a section 1115 waiver in 1993. The
state proposed "TennCare," a demonstration to enroll in managed care
organizations\30 all its Medicaid beneficiaries, in addition to its
medically uninsurable citizens and certain other uninsured
residents.\31
In fewer than 2 months after HCFA granted the waiver, Tennessee
implemented TennCare. In 1993, the year before the demonstration
began, about 2.7 percent of the state's Medicaid population was
enrolled in managed care, compared with about 12.4 percent
nationally. As of January 1, 1994--TennCare's "opening day"--plans
for contracting with provider networks were incomplete, and many of
the state's physicians had not yet determined which networks, if any,
they might join. Some beneficiaries were initially required to
choose a plan without knowing which ones would include their
physicians. Systems to process bills were not fully developed, and
some providers reported slow or no payments for services during the
first months of the waiver. The submission of encounter data to the
state, which allow officials to monitor access and quality, was not
complete even 1 year after the program began. The most recent
beneficiary satisfaction survey taken in 1994 reported that 45
percent of enrollees who had previously received care under
Medicaid's fee-for-service program were less than satisfied with
TennCare. Individuals who were previously uninsured, however, were
not as dissatisfied (8 percent).
--------------------
\21 Medicaid: States Turn to Managed Care to Improve Access and
Control Costs (GAO/HRD-93-46, Mar. 17, 1993) and Managed Health
Care: Effect on Employers' Costs Difficult to Measure
(GAO/HEHS-94-3, Oct. 19, 1993).
\22 States with approved section 1115 waivers are using a variety of
forms of managed care.
\23 Better Controls Needed for Health Maintenance Organizations Under
Medicaid in California (Sept. 1974, B-164031(3)), Deficiencies in
Determining Payments to Prepaid Health Plans Under California's
Medicaid Program (Aug. 1975, MWD-76-15), and Relationships Between
Nonprofit Prepaid Health Plans With California Medicaid Contracts and
For Profit Entities Affiliated With Them (Nov. 1976, HRD-77-4). For
a summary of the status of Medicaid HMO contracting as of early 1981
and the outstanding problems at that time, see Trieger, Sidney, Trudi
W. Galblum, and Gerald Riley, HMOs: Issues and Alternatives for
Medicare and Medicaid, HCFA Pub. No. 03107 (Baltimore: 1981).
\24 HCFA regulations allow waiver of the 75-25 rule in some cases
without a section 1115 waiver.
\25 States can opt to restrict disenrollments, without a waiver, for
up to 6 months for federally qualified HMOs.
\26 The longer the guaranteed enrollment per enrollee, the more
stability the HMO experiences in cash flow and the greater the
opportunity to adequately plan for meeting the health care needs of
its enrollees.
\27 Medicare: Opportunities Are Available to Apply Managed Care
Strategies (GAO/T-HEHS-95-81, Feb. 10, 1995).
\28 HCFA has required states, through its special terms and
conditions associated with approved section 1115 waivers, to
implement improved quality assurance systems and collect encounter
data. HCFA has also contracted for evaluations to examine cost,
quality, and access to care.
\29 Medicaid: Experience With State Waivers to Promote Cost Control
and Access to Care (GAO/T-HEHS-95-115, Mar. 23, 1995).
\30 These managed care organizations are prepaid plans, some of which
operate as HMOs with "gatekeepers" (providers designated to
coordinate the care for individual enrollees) and some as preferred
provider organizations, which do not use gatekeepers.
\31 These are residents of any income level without access to
employer-sponsored or other insurance, although with higher incomes,
greater cost-sharing is expected.
SECTION 1115 WAIVERS PERMIT
STATES TO EXPAND ELIGIBILITY
BEYOND TRADITIONAL MEDICAID
POPULATION
---------------------------------------------------------- Chapter 3:2
States are also using section 1115 waivers to expand Medicaid
coverage to more individuals. Most states' demonstrations expand
coverage to low-income individuals and families who were previously
ineligible for Medicaid (although income and categorical requirements
for these new eligibles vary considerably from state to state).\32 In
sheer numbers, Tennessee, Florida, and Ohio have the most ambitious
expansion programs. Oregon, Kentucky, Hawaii, and Rhode Island plan
more modest expansions. (See table 3.2.)
Table 3.2
Estimated Maximum Number of New
Eligibles Under Approved Statewide
Section 1115 Waivers, by State
New
eligibles\
State a Eligibility requirements
------------ ---------- ------------------------------------------------------
Florida 1,100,000 Individuals and families with incomes below 250% of
the federal poverty level (FPL) are eligible for
subsidized private insurance.
Individuals and families are eligible only if
uninsured for 12 months or recently disenrolled from
Medicaid.
Hawaii 80,000 Uninsured persons below 300% of FPL.
Kentucky 201,000 Individuals with incomes below FPL.
Ohio 395,000 Individuals and families with incomes below FPL.
Oregon 112,000\b Individuals and families with incomes below FPL.
Rhode Island 11,000 Pregnant women and children up to age 6 with family
incomes between 185% and 250% of FPL.
Extension of family planning services for women for 2
years after giving birth.
Tennessee 500,000\c All uninsured, regardless of employment or income
status, including individuals who cannot obtain
coverage because of a preexisting condition.
(Enrollment capped for newly entitled, not capped for
traditional Medicaid recipients. Eligibility
restricted to those uninsured prior to a date within
the last year.)
--------------------------------------------------------------------------------
\a Includes expansions to optional groups of Medicaid eligibles.
\b Actual new enrollment as of March 3, 1995.
\c In January 1995, Tennessee closed enrollment to the uninsured;
demonstration enrollment was 438,000 in February 1995.
Source: State waiver proposals and supporting documentation.
If all states were granted waivers that permitted eligibility
expansions, the number of individuals in Medicaid-supported programs
could be sizeable. In a recent study,\33 the Urban Institute
estimates that expanding coverage up to 200 percent of the federal
poverty level would increase the number of beneficiaries between 18
million and 45 million, depending on the extent to which eligible
people apply.\34 A more limited expansion covering individuals with
incomes up to the federal poverty level, as approved for Oregon and
Kentucky, would increase the number of beneficiaries between 8
million and 14 million.
--------------------
\32 Some states include in their demonstration individuals and
families who could have been covered at state option under the
regular Medicaid program.
\33 Increasing Insurance Coverage Through Medicaid Waiver Programs
(Washington, D.C.: Nov. 1994).
\34 These figures do not represent a net decrease in the number of
uninsured, since some of these new eligibles currently have private
insurance. Some states with generous program expansions, like
Florida, plan to limit the substitution of Medicaid for private
insurance by requiring a 12-month waiting period before new eligibles
can receive benefits.
AS GROWING NUMBER OF STATES
SEEK SECTION 1115 WAIVERS,
CONTROVERSY ENSUES
---------------------------------------------------------- Chapter 3:3
Section 1115 waivers are granted for research purposes to test
program improvements in keeping with Medicaid goals or examine issues
of interest to HCFA. Until Oregon's application was approved in
March 1993, only Arizona had been granted authority to run its
demonstration statewide.\35 Before that time, most section 1115
demonstrations were relatively small and affected only a small
geographic area, target population, or provider group.
Objecting to the section 1115 demonstrations, a legal challenge has
been mounted against the waivers. In June 1994, the National
Association of Community Health Centers (NACHC) filed suit in federal
court (NACHC v. Shalala). NACHC argues, among other things, that
the Secretary of HHS acted beyond her statutory powers in approving
these statewide section 1115 waivers. The suit challenges the award
of waivers granted before the suit was filed and seeks injunctive
relief against subsequent approval of similar waivers. Waiver
approvals for Hawaii, Kentucky, Oregon, Rhode Island, and Tennessee
preceded the suit. Since the suit was filed, Florida and Ohio have
also been granted section 1115 waivers.
The issue of expanded coverage may stall the implementation of
proposed section 1115 demonstrations in certain states. Florida,
Kentucky, and Ohio have waivers approved by HCFA but not by their own
state legislatures. Lawmakers in these states are concerned that
implementation of the waivers will increase, not decrease, state
costs. The Kentucky legislature has required that the state realize
savings from managed care before coverage is expanded.
--------------------
\35 Since its inception in 1982, Arizona's entire medical assistance
program for AFDC and SSI recipients has been operated as a section
1115 demonstration known as the Arizona Health Care Cost Containment
System.
SOME SECTION 1115 DEMONSTRATIONS
COULD INCREASE FEDERAL
EXPENDITURES
============================================================ Chapter 4
Some states with approved section 1115 waivers could be eligible for
more federal Medicaid funds than they might have received without the
waivers. Some states have proposed ambitious coverage expansions,
which, through the waivers, would be financed in part by the Medicaid
program. While estimating the demonstration projects' potential
impact on federal spending is difficult, the administration's
approval of the waivers as budget neutral depends on a number of
critical assumptions that could be subject to challenge.
Furthermore, it is not clear that the administration's rationale for
defining budget neutrality in the approved statewide waivers is
consistent across all waivers.
We have been evaluating the budget neutrality of approved waivers
using a consistent framework that compares approved budgets with a
benchmark using current services budgeting concepts. While more
extensive analysis is in progress, it appears that one of the four
waivers we analyzed may be budget neutral, but three others may not
be if fully implemented by the states.\36 In Tennessee, our analysis
indicates that federal spending on the Medicaid demonstration program
could be lower over the life of the waiver, compared with what might
have been spent without the waiver. However, the demonstration
programs in Hawaii, Florida, and Oregon may increase federal spending
on Medicaid acute care. Because the waiver agreements only specify
ceilings on federal spending, the total amount of new federal dollars
that could be required by the demonstrations depends upon several
factors, including future action by the states. For example, federal
spending for Florida will depend, partly, upon how many of the
planned maximum 1,100,000 newly eligible individuals that state will
actually cover.
The potential net spending impact associated with the four state
waivers we analyzed, while it could amount to hundreds of millions of
additional federal dollars, is small relative to overall federal
Medicaid spending. However, the impact of the section 1115
demonstrations could increase substantially if the waivers pending or
proposed by several additional states--or even a few large
states--are not budget neutral.
--------------------
\36 These results are derived by comparing the maximum potential
costs of the demonstrations (as outlined in the waiver applications)
with a "current services" baseline (the projected amount of federal
and state spending necessary to maintain the current level of
Medicaid services) calculated separately for each demonstration
project. Our ongoing analysis is examining ways to adjust the
current services baseline for differences in state experiences. The
results of this analysis will be reported at a later date.
BUDGET NEUTRALITY REDEFINED
---------------------------------------------------------- Chapter 4:1
Since the early 1980s, the administration has required that federal
spending on a state's section 1115 demonstration project be budget
neutral--that is, federal expenditures may not exceed what would have
been spent without the waiver. This budget-neutrality policy is
still in force. Since 1993, however, the administration has applied
the financial test in such a way as to allow states more flexibility
in showing that their proposed projects would be budget neutral.
Projecting future spending for the current program is a critical
element in determining budget neutrality since this amount represents
the baseline against which spending for the demonstration project is
compared. The two key steps in developing projections for the
current program are estimating (1) base, or initial-year, spending
and (2) future annual increases. From a state's perspective, higher
base costs or higher annual increases make it easier to claim budget
neutrality. That is, a higher baseline makes it possible for planned
spending to increase while still meeting the definition of budget
neutrality. States have used one or both of the following arguments
to secure access to increased federal funds while complying with the
current budget-neutrality policy.
Hypothetical Expansion Argument. Four states--Hawaii, Rhode
Island, Kentucky, and Ohio--successfully argued that their
baselines should include the cost of extending coverage to
optional eligibility groups that could have been, but were not
then, served by the state's Medicaid program. These
hypothetical expansions primarily covered women and children
(but may also include elderly and disabled individuals), whose
coverage was optional.\37 Although in Hawaii many of these
individuals received limited benefits under one of two
state-funded programs, these groups were not served at all in
the Medicaid programs. The states argued that since they could
have submitted a state plan amendment and thus received federal
medical assistance payments for providing health services to
these groups, an equivalent amount of federal funding for the
Medicaid demonstration project, which would include them, should
be considered budget neutral. Including the section 1902(r)(2)
population increased Hawaii's baseline by about $58 million over
5 years. Although the Kentucky legislature suspended
implementation of the state's waiver, inclusion of the section
1902(r)(2) populations that included the elderly and disabled
with incomes up to the federal poverty level would have added
about $524 million to the baseline over 5 years.\38
Historical Inflation Argument. The administration allowed some
states to project increases in their baseline costs at rates
higher than the administration's existing projections of
national Medicaid spending growth. These states successfully
argued that past years' spending increases exceeded average
national spending growth and thus justified a higher or faster
growing baseline. For example, in the first year of their
waivers, Florida and Tennessee received relatively large
increases (about 16 percent) compared with the previous year.
Florida also received large annual increases for the remaining
years of the waiver.
Because Medicaid costs in the late 1980s and the early 1990s grew for
reasons not expected to affect future spending, estimates of future
Medicaid expenditures that rely on historical cost growth may be
inflated. For example, although the states' use of provider taxes
and donations contributed to the rapid rise of DSH payments between
1989 and 1993, recent legislation strictly limits DSH growth. Also,
some of the past cost increases were due to the absorption of
mandated and optional populations and changes in benefits. Major new
mandates have not been added since 1991 and are not likely to be
added in the foreseeable future.\39 Furthermore, continued rapid
Medicaid cost increases may have been unsustainable because of the
great strain they placed on state budgets. Facing rapid cost
increases, states would have increased incentive to adopt
cost-containment measures. For example, Tennessee officials said the
state would have faced massive health care coverage reductions had
the waiver not been approved. In another example, after refusing to
approve the waiver, the Kentucky legislature reduced physician
reimbursements by $50 million.
Another part of the policy change since 1993 is to allow
demonstrations to show budget neutrality over the life of the waiver,
but not to require it year by year. HCFA agreed that states'
demonstration projects may involve temporary start-up costs and
permitted increased federal funding to help pay these initial costs,
as long as states expected to offset these costs by reduced federal
funding in later years. Oregon, and perhaps some of the other states
with waivers approved since 1993, expect that lower Medicaid cost
growth in later years will compensate the federal government for
higher expenses in the early years of the demonstrations.
--------------------
\37 Under section 1902(r)(2) of the Social Security Act, 42 U.S.C.
1396b(r)(2), states were essentially given the option to cover such
individuals by being permitted to employ less restrictive eligibility
methodologies. As a matter of policy, HCFA generally limits
arguments for hypothetical expansions to those individuals eligible
under 1902(r)(2) who will be included in the demonstration.
\38 Federal Medicaid expenditures could increase substantially if all
states provided eligibility for their 1902(r)(2) populations either
under a section 1115 demonstration or as part of their regular
program. In its November 1994 report, the Urban Institute estimates
that annual state and federal spending could increase by between $5.5
billion and $23 billion. The exact amount would depend on how many
eligible individuals applied for coverage and how many individuals
with private or employer-provided insurance migrated to the publicly
subsidized Medicaid program.
\39 As in the past, however, some states may attempt to shift
state-funded health and social service programs into Medicaid by
extending coverage to optional beneficiaries and thus obtain federal
matching payments.
PARTIAL CAPS MAY LIMIT FEDERAL
LIABILITY WHILE HEIGHTENING
STATE RISK
---------------------------------------------------------- Chapter 4:2
Each of the waiver agreements ensures some limits on federal spending
and holds some financial risk for the states. Because of the
bilateral nature of the negotiations between the states and HHS, and
because of the discretion granted to the Secretary of HHS in
approving the waivers, the terms and conditions for each state are
unique. Thus, the net effect on federal spending resulting from the
waiver states' Medicaid programs after the 5-year life of the
demonstrations may be different for each state. Under any of the
negotiated waivers, however, states are not to receive federal
assistance in paying for costs above the agreed-upon limits.
Hawaii, Kentucky, Ohio, Oregon, and Rhode Island have agreed to
federal spending limits called per capita caps, which are based on
Medicaid spending per person.\40 As shown in table 4.1, these caps
allow per person spending to increase by a fixed percentage (plus the
percentage change in medical prices in Hawaii and Kentucky). Thus,
if the number of eligibles rises or, in Hawaii and Kentucky, if
medical prices increase, the states can obtain additional funding.
However, if spending rises because of greater utilization of services
or a change in the mix of services provided, the states are
financially liable.
Under the waiver agreements for Tennessee and Florida, "aggregate
caps," or limits on total spending, limit the federal government's
financial exposure. In Tennessee, the state and the federal
government agreed to set a maximum for federal funding for each year
of the waiver. Under these terms, spending for Tennessee's Medicaid
program may increase from $2.7 billion in 1993 to $4.1 billion in
1998. Florida and the federal government also set a maximum for
annual federal Medicaid funding. Unlike Tennessee's agreement,
however, federal spending can exceed the specified amounts if the
number of traditional Medicaid beneficiaries exceeds the projected
number by more than 3 percent. This means the state's financial risk
is limited if the Medicaid rolls swell because of an economic
downturn, for example. The two caps also differ in that Tennessee's
covers long-term care expenditures, while Florida's does not.
Table 4.1
Type of Expenditure Cap and Annual
Adjustment, by State
State Type of cap Annual adjustment
------------------ ------------------ --------------------
Hawaii Per capita 4% plus medical
inflation\a
Tennessee Aggregate 5.1% to 8.3%\b
Florida Aggregate\c 14.3% to 15.7%\b
Kentucky Per capita 3% plus medical
inflation
Ohio Per capita 6.7% to 9.4%\d
Oregon Per capita 8.5%\
Rhode Island Per capita 4% to 8%\b
------------------------------------------------------------
\a Medical inflation is measured by the medical component of the
consumer price index.
\b Varies by year.
\c Federal funding is increased if the number of Medicaid eligibles
grows by more than 3 percent.
\d Varies by year and category of spending.
Some states may find funding their planned expansions difficult.
Hawaii is experiencing financial problems largely because more new
eligibles enrolled than expected. Oregon may also face difficulty,
due in part to higher than anticipated enrollment of new eligibles
and the continuing effects of a tax initiative passed in 1990 that
limits state funds available for Medicaid and other programs. Oregon
is considering reducing the capitation rate, changing eligibility
rules, implementing premiums and copayments, delaying full
implementation of mental health services, and reducing benefits, but
will need HCFA approval before doing so. Moreover, a state analysis
suggests the program will not be budget neutral without the cost
savings anticipated from an employer mandate, for which the state
must obtain congressional action.\41 Even with the necessary
congressional action, however, state political support for the
mandate has eroded because of business opposition and changes in the
makeup of the state legislature.
The impact of higher than anticipated costs on federal funding
depends on the specific waiver agreement and HCFA's enforcement. In
Florida's case, HCFA would help pay for the increased costs over the
life of the waiver, but expects to get paid back for the excess costs
at the end of the 5-year project. Ohio's agreement is similar to
Florida's, except that it specifies cumulative cost overrun ceilings.
If Ohio's cumulative spending exceeds the ceiling, the state must
submit a corrective action plan. In Tennessee, the budget-neutrality
agreement has a formula to prevent federal expenditures from
exceeding the yearly cap by more than a specified percentage.
Tennessee has agreed to take steps such as limiting enrollment when
this happens.
The total amount of federal funding required for the demonstrations
also depends, in part, on how HCFA defines and enforces its
methodology for determining the federal expenditure cap in states
with per capita limits. For example, even though the Oregon
demonstration has been operating for more than a year, HCFA and the
state have yet to agree on a specific methodology. The difference
between two possible methods for calculating the upper limit on
federal spending amounts to more than 5 percent of the state's
Medicaid demonstration budget.\42
--------------------
\40 In Hawaii, Kentucky, Ohio, and Rhode Island, federal spending
limits are based on the number of Medicaid eligibles (including the
hypothetical expansions) multiplied by historically based per person
fee-for-service costs, adjusted for actual inflation experience. In
Oregon, the federal spending limit is based on the number of Medicaid
eligibles plus a portion of otherwise ineligible individuals
multiplied by actuarially developed per person costs that are agreed
to by HCFA and the state.
\41 To mandate that employers provide health insurance to their
employees, states need an exception to provisions of the Employee
Retirement Income Security Act. To date, only Hawaii has obtained
such an exception, largely because its employer mandate predated the
passage of the act.
\42 Other implementation and enforcement issues may also affect
federal spending. For example, most state demonstrations change the
method for counting qualifying income from a net to a gross income
test. This change makes counting the number of Medicaid-eligible
individuals imprecise. As a result, the per capita cap states can
only estimate the numbers to determine total federal expenditures for
the demonstration. The estimated counts will likely be different
from actual counts using the net income test, but whether they will
be higher or lower is unknown.
OBSERVATIONS ON MEDICAID'S FUTURE
============================================================ Chapter 5
Millions of Americans--not only poor mothers and children, but also
poor elderly, blind, and disabled individuals--depend upon health
care made possible by the Medicaid program. Medicaid pays for both
acute care and long-term care services. The program underwrites the
deliveries of about one-third of the babies born in the United States
each year and finances about 50 cents of every dollar paid to nursing
homes.
Medicaid is an expensive program. It currently consumes about 6
percent of all federal outlays (3 times the share devoted to Food
Stamps and 5 times the share devoted to AFDC). Moreover, Medicaid's
slice of the federal budget is growing faster than most other major
budget items, including Medicare. If Medicaid continues indefinitely
at its forecast growth rate (10.7 percent), federal spending on the
program will double roughly every 5 to 7 years.
Success in containing cost growth in Medicaid, as for the health care
sector at large, has been elusive. Care for the elderly, blind, and
disabled populations--particularly long-term care--is expensive,
consuming more than $80 billion of the $130 billion cost of Medicaid
in 1993. For acute care services, however, the promise of lower
utilization has made the capitated payment features of managed care
plans attractive to the states.
Federal restrictions on the use of managed care reflect concerns for
quality, because capitated payments allow profits to be earned from
underservice. For example, the 75-25 rule seeks to ensure that
Medicaid managed care plans are of sufficient quality, as
demonstrated by their ability to attract private enrollees. At the
same time, states believe that restrictions like this hamper their
efforts to implement managed care effectively and that the need to
obtain section 1115 waiver approval poses unnecessary obstacles.
Requiring states to seek waivers before implementing these health
care delivery system reforms may be burdensome, especially if
alternatives for addressing quality of care concerns exist.
Nonetheless, continuous oversight of managed care systems is required
to protect Medicaid beneficiaries from inappropriate denial of care
and federal dollars from payment abuses.
Section 1115 waivers, while freeing states to implement managed care
cost-containment strategies, could in the long run undermine efforts
to contain federal expenditures. Our study of the section 1115
waivers approved to date raises the following concerns:
The administration is allowing states to apply the federal share of
Medicaid savings from managed care to finance coverage of
additional populations not included under Medicaid law. To meet
the budget-neutrality terms of section 1115 waivers, the
administration and states assume that the enrollment of the
Medicaid populations in HMOs will save states enough money to
cover additional low-income people in the state at no extra cost
to the federal government. Even if the proposed demonstrations
will not require new federal dollars, the administration's
approval of coverage expansions means that anticipated Medicaid
cost savings (from more aggressive use of capitated care) will
not be used to reduce federal spending. At issue is whether the
federal treasury should benefit from these savings. Also of
concern is whether eligibility should be made available for new
groups only after congressional debate and legislative action.
The administration's method for determining budget neutrality may
allow states access to more federal funding than they would have
received without the waiver. Our initial examination of four
states' proposed demonstrations suggests that claims of budget
neutrality for these states may not be sustainable in all cases.
While Tennessee's demonstration project may be budget neutral,
the demonstrations in Florida, Hawaii, and Oregon may require
increased financial commitment from the federal government.
Relative to overall federal Medicaid spending, the amount spent
in states with approved section 1115 waivers is small. However,
the methods used by the administration to assess the budget
neutrality of pending and future waiver proposals may greatly
affect federal Medicaid spending in the years to come.
The Congress may find it difficult to scale back section 1115
demonstrations if they prove more costly than forecast. A
demonstration waiver, granted for a limited period of time, may
be a shortsighted approach to reducing states' uninsured
populations. If at the end of 5 years the demonstrations have
cost much more than estimated, the Congress may face the choice
of increasing federal funding or relying on the states to reduce
benefits or deny coverage to hundreds of thousands of people
newly enrolled under the waivers.
For these reasons, we believe that the granting of additional section
1115 waivers merits close scrutiny.
AGENCY COMMENTS AND OUR EVALUATION
============================================================ Chapter 6
The administration's comments focused on two aspects of this report:
waiver review policy changes and our approach to analyzing the budget
neutrality of approved statewide section 1115 waivers, with which it
disagreed.\43 In brief, our approach begins with a uniform and
consistent budget "neutrality" benchmark--OMB's current services
projections of future Medicaid growth--for assessing the potential
budget impact of waivers. The administration contends that by using
OMB's national estimate of projected Medicaid spending growth, we do
not accurately capture the variation in state spending patterns.
Moreover, it maintains that current services is an inappropriate tool
for establishing budget neutrality.
--------------------
\43 Our methodology is described in chapter 1.
CHANGES IN WAIVER REVIEW POLICY
---------------------------------------------------------- Chapter 6:1
The administration's response notes that its approach both to
reviewing waivers and to determining budget neutrality represents a
departure from previous administrations' policies. Overall, it
characterized its new approach as "more flexible."
Although budget neutrality is not specifically defined by regulation
or statute, the administration's new approach is characterized by (1)
flexibility in determining appropriate baseline expenditures and (2)
assessment of budget neutrality over the life of the waiver rather
than year by year. Our analysis recognizes these key differences.
In fact, we conclude that the administration's new approach
represents a redefinition of budget neutrality.
THE ADMINISTRATION'S
BUDGET-NEUTRALITY METHODOLOGY
---------------------------------------------------------- Chapter 6:2
The administration characterizes its budget-neutrality policy as
"state-specific," reflecting the historically "dramatic variation in
state Medicaid programs." These variations, the administration notes,
have resulted in "radically different levels of expenditures and
growth rates across the states." Lacking a state-specific Medicaid
baseline, the administration compares its national estimate of growth
in Medicaid for the nation as a whole with state historical
expenditure growth in determining a budget-neutrality benchmark.
Furthermore, the administration points out that it relies on current
federal law in establishing a state-specific neutrality benchmark.
It notes that "under current federal law states are able to engage in
program expansions and contractions" and that, in establishing a
state-specific budget-neutrality baseline, some judgment about
states' behavior should be involved.
Our discussion of the administration's approach to budget neutrality
recognizes the state-specific nature of its methodology. Thus, in
chapter 4, we indicate that the waiver funding agreements for Florida
and Tennessee were based on an analysis of large annual growth rates
over the past 5 years. We conclude, however, that some of the
factors underlying Medicaid's very rapid and recent cost growth will
be much less significant in the future and that relying on historical
cost growth is inappropriate. For example, a major cause of
Medicaid's rapid growth in the past--DSH payments--has been strictly
limited by 1991 and 1993 legislation. In addition, major new
mandatory coverage expansions are not likely to be undertaken in the
foreseeable future.
DEMONSTRATION NATURE OF THE
SECTION 1115 WAIVERS
---------------------------------------------------------- Chapter 6:3
The administration disputes our contention that the comprehensive
scope of the waiver demonstrations makes it controversial to
characterize them as experimental. It notes that the waivers allow
states to experiment with innovative delivery and payments systems
for their Medicaid and low-income populations and that each
demonstration has unique components deemed worthy of evaluation. We
view the sheer magnitude of some of the waiver projects as what makes
it difficult to consider them experimental. Once coverage has been
extended to several hundred thousand people, we anticipate
significant difficulty in terminating a demonstration if the
innovations being tested fail to achieve the intended objectives.
KEY AREAS OF DISAGREEMENT
---------------------------------------------------------- Chapter 6:4
As suggested by the administration, in our initial analysis we have
applied a more rigid method for assessing budget neutrality. We take
a more uniform base--previous-year spending--and allow it to grow at
the rate of increase expected by OMB for the national program.
Because we recognize the significant differences in state Medicaid
programs' experiences, our preference would have been to use
consistently generated state forecasts of expected Medicaid spending.
However, neither OMB nor the Congressional Budget Office has
developed such forecasts. We believe that using states' historical
experiences without adjustments may not be a good guide for the
future. In our ongoing work, we are examining state-specific
information to determine what adjustments to the national forecasts
might be appropriate.
In contrast to our efforts to develop a consistent methodology, the
administration's budget-neutrality assessments appear not to be
consistent. Thus, in some states, aggregate spending trends over the
past 5 years appear to have been the dominant factor used to project
budget-neutral spending over the next 5 years. In other states,
trends in average costs per person were the major factor used to
define budget-neutral trends over the life of the waiver. By using
unique methods in each state, the administration has created the
potential for budget-neutrality decisions to be based on the
technique most favorable to a particular state. We continue to
believe that a consistent approach is more appropriate for budget
purposes.
If state-specific reasons exist for future Medicaid program growth
above (or below) national standards, they should be well documented.
However, in reviewing state waiver submissions, no special factors
were identified that would justify future growth above a national
benchmark.
Finally, the administration suggests that a current law baseline is
more appropriate than a current services baseline in assessing budget
neutrality. That is, the budget-neutrality assessment should be
based on "what states might do" under current law rather than "what
states are doing." However, the administration also notes that the
President's budget does not distinguish between current services and
current law. Although current law gives states the latitude to
expand coverage to optional populations, not all states have chosen
to do so. We continue to believe that "what states are doing" should
be the basis for determining budget neutrality.
The complete text of the administration's comments is in appendix IV.
MAJOR FEDERAL EXPANSIONS OF
MEDICAID ELIGIBILITY AND SERVICES
(1984-93)
=========================================================== Appendix I
Table I.1
Federal Medicaid Expansion to AFDC
Recipients and Related Populations
Population
affected Expansion Mandate/option
------------------ ------------------ --------------------
DEFRA (Deficit Reduction Act of 1984) (P.L. 98-369)
------------------------------------------------------------
Infants\a and Requires coverage Mandate
children of all children
born after 9/30/
83 who meet state
AFDC income and
resource
standards,
regardless of
family structure.
Pregnant women Requires coverage Mandate
from date of
medical
verification of
pregnancy,
providing the
mother would (1)
qualify for AFDC
once child was
born or (2)
qualify for AFDC-
UP\b once child
was born,
regardless of
whether state has
AFDC-UP program.
Infants Requires automatic Mandate
coverage for 1
year after birth
if mother already
is receiving
Medicaid and
remains eligible
and infant resides
with her.
AFDC families Requires limited Mandate\c
extension of
Medicaid coverage
if AFDC
eligibility is
lost as a result
of increased
earnings.
AFDC families Extends earned Mandate
income disregard\d
from 4 to 12
months.
COBRA (Consolidated Omnibus Budget Reconciliation Act of
1985) (P.L. 99-272)
------------------------------------------------------------
Pregnant women Requires coverage Mandate
if family income
and resources are
below state AFDC
levels, regardless
of family
structure.
Postpartum women Requires 60-day Mandate
extension of
coverage
postpartum if
eligibility was
pregnancy-
related.
Pregnant women Allows provision Option
of enhanced
benefits.
Infants and Allows extension Option
children of DEFRA coverage
up to age 5
immediately,
instead of
requiring phase-
in by birth date.
Adoptive and Requires coverage Mandate
foster children even if adoption/
foster agreement
was entered into
in another state.
OBRA 1986 (Omnibus Budget Reconciliation Act of 1986) (P.L.
99-509)
------------------------------------------------------------
Pregnant women and Creates new Option
infants optional
categorically
needy group for
those with incomes
below poverty
line. Women
receive pregnancy-
related services
only.
Pregnant women and Allows assets test Option
infants to be dropped for
this newly defined
category of
applicants.
Pregnant women Allows presumptive Option
eligibility for up
to 45 days to be
determined by
qualified
provider.
Pregnant women Allows guarantee Option
of continuous
eligibility
through postpartum
period.
Children Allows coverage up Option
to age 5 if family
income is below
poverty line
(phased in).
Infants and Requires Mandate
children continuation of
eligibility (for
those who
otherwise would
become ineligible)
if individuals are
hospital
inpatients when
age limit is
reached.
OBRA 1987 (Omnibus Budget Reconciliation Act of 1987) (P.L.
100-203)
------------------------------------------------------------
Pregnant women and Allows coverage if Option
infants family income is
below 185% of
poverty line.
Children Allows immediate Option
extension of OBRA
1986 coverage for
children up to age
5 in families with
incomes up to the
poverty line.
Children Clarifies that Option
states may provide
in-home services
for qualified
disabled children.
Children Allows coverage Option
for children aged
5-7 up to state
AFDC level (phased
in by age).
Children Allows coverage Option
for children below
age 9 in families
with incomes up to
the poverty line
(phased in by
age).
MCCA (Medicare Catastrophic Coverage Act of 1988) (P.L. 100-
360)
------------------------------------------------------------
Pregnant women and Makes mandatory Mandate
infants the OBRA 1986
option of coverage
up to the poverty
line (phased in by
% of poverty
line).
Family Support Act of 1988 (P.L. 100-485)
------------------------------------------------------------
AFDC families Increases required Mandate\e
period of Medicaid
coverage if AFDC
cash assistance is
lost as a result
of increased
earnings.
AFDC families with Requires coverage Mandate
unemployed parent if otherwise
(AFDC-UP) qualified.
OBRA 1989 (Omnibus Budget Reconciliation Act of 1989) (P.L.
101-239)
------------------------------------------------------------
Pregnant women and Requires coverage Mandate
infants if family income
is below 133% of
poverty line.
Children Requires coverage Mandate
up to age 6 if
family income is
below 133% of
poverty line.
Children Requires provision Mandate
of all Medicaid-
allowed treatment
to correct
problems
identified during
early and periodic
screening,
diagnostic, and
treatment (EPSDT),
even if treatment
is not covered
otherwise under
state's Medicaid
plan.
Children Requires Was an option, now
interperiodic\f mandated
screenings under
EPSDT when medical
problem is
suspected.
OBRA 1990 (Omnibus Budget Reconciliation Act of 1990) (P.L.
101-508)
------------------------------------------------------------
Children Requires coverage Mandate
up to age 18 if
family income is
below the poverty
line (phased in by
age).
Pregnant women Makes mandatory Mandate
the OBRA 1986
option of
continuous
eligibility
through postpartum
period.
Pregnant women Extends period of Mandate
presumptive
eligibility before
written
application must
be submitted.
Pregnant women and Requires states to Mandate
children receive and
process
applications at
convenient
outreach sites.
Infants Requires Mandate
continuous
eligibility if (1)
born to Medicaid-
eligible mother
who would remain
eligible if
pregnant and (2)
remaining in
mother's
household.
OBRA 1993 (Omnibus Budget Reconciliation Act of 1993) (P.L.
103-66)
------------------------------------------------------------
Mothers and Expands scope of Mandate
newborns required nurse-
midwife services
to include
services outside
the maternity
cycle that
midwives are
authorized to
perform under
state law.
Children Requires state Mandate
Medicaid programs
to establish a
program to
distribute
pediatric vaccines
furnished by the
federal
government.
------------------------------------------------------------
\a Infants are children up to age 1.
\b AFDC-UP allows coverage in two-parent families if principal
wage-earner is unemployed.
\c Mandate is for 9 months. State may opt to provide additional
6-month period of coverage.
\d Certain expenses associated with work are disregarded from income
in calculating AFDC eligibility.
\e Mandate is for 12 months. State may opt to provide additional
6-month period of coverage.
\f States establish a screening schedule: "Interperiodic" visits are
added to the standard schedule if a problem is suspected.
Table I.2
Federal Medicaid Expansion to the
Population Receiving SSI
Population
affected Expansion Mandate/option
------------------ ------------------ --------------------
DEFRA (Deficit Reduction Act of 1984) (P.L. 98-369)
------------------------------------------------------------
SSI recipients Increases Mandate
qualifying asset
limits for
applicants for
limited time
period (1984-89).
COBRA (Consolidated Omnibus Budget Reconciliation Act of
1985) (P.L. 99-272)
------------------------------------------------------------
Children with Requires coverage Mandate
special needs regardless of
income/resources
of adoptive/
foster parents.
OBRA 1986 (Omnibus Budget Reconciliation Act of 1986) (P.L.
99-509)
------------------------------------------------------------
Aged and disabled Creates new Option
optional
categorically
needy group for
those with incomes
below the poverty
line under certain
resource
constraints.
Option can be
exercised for this
group only if
exercised also for
pregnant women and
infants.
Aged and disabled Allows Medicare Option
buy-in\a up to the
poverty line for
qualified Medicare
beneficiaries
under certain
resource
constraints.
Severely impaired Establishes new Mandate
individuals mandatory
categorically
needy coverage
group for
qualified
individuals under
age 65.
Ventilator- Allows coverage of Option
dependent at-home
individuals respiratory care
services.
SSI recipients Makes permanent Mandate
the previous
temporary
provision
requiring coverage
of some former
disabled SSI
recipients who
have returned to
work.
Employment Opportunities for Disabled Americans Act of 1986
(P.L. 99-643)
------------------------------------------------------------
Disabled Makes permanent a Mandate
individuals previous
demonstration
program for
individuals able
to engage in
substantial
gainful activity
despite severe
medical
impairments.
OBRA 1987 (Omnibus Budget Reconciliation Act of 1987) (P.L.
100-203)
------------------------------------------------------------
Elderly Allows provision Option
of home and
community-based
services to those
who otherwise
would need nursing
home care.\b
Nursing home Requires states to Mandate
applicants establish
preadmission
screening programs
for mentally ill
and retarded
individuals.
Nursing home Requires Mandate
residents preadmission
screening and
annual resident
review for
mentally ill or
retarded
individuals.
MCCA (Medicare Catastrophic Coverage Act of 1988) (P.L. 100-
360)
------------------------------------------------------------
Elderly and Makes mandatory Mandate
disabled for qualified
individuals Medicare
beneficiaries the
OBRA 1986 option
of Medicare buy-
in for individuals
with incomes up to
the poverty line
(phased in by % of
poverty line).
OBRA 1990 (Omnibus Budget Reconciliation Act of 1990) (P.L.
101-508)
------------------------------------------------------------
Elderly and Extends the MCCA Mandate
disabled qualified Medicare
individuals beneficiary
provision to
individuals with
incomes up to 120%
of poverty line
(phased in by % of
poverty line).
Elderly and Allows limited Option
disabled program permitting
individuals states to provide
home and
community-based
services to
functionally
disabled
individuals, and
community-
supported living
arrangements to
mentally retarded/
developmentally
disabled
individuals.
OBRA 1993 (Omnibus Budget Reconciliation Act of 1993) (P.L.
103-66)
------------------------------------------------------------
SSI recipients Allows states to Option
offer Medicaid
coverage to TB-
infected
individuals who
meet the state's
income and
resource tests.
------------------------------------------------------------
\a Medicaid covers Medicare cost-sharing charges: premiums,
deductibles, and coinsurance.
\b This is not automatic. HCFA must grant a waiver to any state
wishing to provide these services.
Table I.3
Federal Medicaid Expansion to Other
Populations and Service Additions
Population
affected Expansion Mandate/option
------------------ ------------------ --------------------
COBRA (Consolidated Omnibus Budget Reconciliation Act of
1985) (P.L. 99-2
------------------------------------------------------------
Terminally ill Allows provision Option
individuals of hospice
services.
OBRA 1986 (Omnibus Budget Reconciliation Act of 1986) (P.L.
99-509)
------------------------------------------------------------
Aliens Requires provision Mandate
of emergency
services if
otherwise eligible
(financially and
categorically).
IRCA (Immigration Reform and Control Act of 1986) (P.L. 99-
603)
------------------------------------------------------------
Newly legalized Requires provision Mandate
aliens of emergency and
pregnancy-related
services if
otherwise
eligible. Also
requires full
coverage for
eligible
individuals under
18.
Anti-Drug Abuse Act of 1986 (P.L. 99-570)
------------------------------------------------------------
Homeless Requires state to Mandate
provide proof of
eligibility for
individuals
otherwise eligible
but having no
permanent address.
OBRA 1993 (Omnibus Budget Reconciliation Act of 1993) (P.L.
103-66)
------------------------------------------------------------
Medicaid Makes coverage of Option
beneficiaries personal care
services outside
the home an
optional rather
than a mandatory
service.
Aliens Clarifies that Mandate
Medicaid-covered
emergency services
for aliens do not
include care and
services related
to organ
transplant
procedures.
------------------------------------------------------------
APPLICATION OF THE
DISPROPORTIONATE SHARE HOSPITAL
PROGRAM
========================================================== Appendix II
OVERVIEW OF PROGRAM OPERATION
-------------------------------------------------------- Appendix II:1
The disproportionate share hospital (DSH) program was established in
1981 to enable states to provide additional payments to hospitals
with heavy caseloads of Medicaid and other low-income patients. The
supplemental payments are intended to help defray costs not covered
by Medicaid (because of low reimbursement rates) or private
insurance. Subject to certain minimums required by federal law, each
state decides on its own criteria for identifying DSHs and its
formula for making payments to these hospitals.
In the mid-1980s, some states began using recently relaxed rules on
raising state Medicaid funds, together with the DSH program, to
leverage additional federal dollars. Using this creative financing
scheme, these states were able to increase the share of the state
Medicaid program funded by the federal government. One common
variant of the financing mechanism could be described as follows. A
state would receive donations from or levy a tax on specific
hospitals. The state would then return the funds to the same
hospitals in the form of DSH payments. Because these payments would
trigger federal medical assistance payments, the hospitals would
receive more in DSH payments than they had provided to the state in
donations or taxes. In the end, the hospitals would have received
more money because of the scheme, all of it coming from the federal
government and none from the state. States could also set the DSH
payments so that the new federal funding flowed to the state, and the
hospitals received total DSH payments (federal and state combined)
just large enough to exactly compensate them for their taxes and
donations.
Our August 1994 report detailed how Michigan, Tennessee, and Texas
used various financing schemes to increase the federal share of
Medicaid expenditures without effectively committing their share of
state matching funds.\44
--------------------
\44 Medicaid: States Use Illusory Approaches to Shift Program Costs
to Federal Government (GAO/HEHS-94-133, Aug. 1, 1994).
FINANCING MECHANISMS USED IN
MICHIGAN
-------------------------------------------------------- Appendix II:2
In fiscal year 1993, Michigan used hospital donations to help raise
funds for its Medicaid program. Michigan made DSH payments of $458
million, including $256 million in federal matching funds, to 53
hospitals; however, the hospitals returned all but $6 million to the
state. As a result, the state received a $250 million net benefit
from the federal share of the DSH payments. Michigan stopped this
practice because 1991 federal legislation, which took effect for
Michigan on January 1, 1993, severely limited provider donations.
In response to these limitations, Michigan's 1994 DSH program
included $489 million for those hospitals that provided at least 6
percent of inpatient services to indigent patients in the state.
State officials determined that only one hospital would qualify--the
state-owned University of Michigan hospital.
On October 2, 1993, Michigan made a DSH payment of $489 million to
the University of Michigan hospital. This included $276 million in
federal matching funds and $213 million in state funds. Later that
day, the hospital returned the entire payment to the state, resulting
in a net benefit to the state of $276 million.
Beginning in 1995, the University of Michigan DSH payment will be
severely restricted by the Omnibus Budget Reconciliation Act of
1993.\45 State officials have calculated that the state can make a
1995 DSH payment to the University of Michigan hospital of $136.3
million. To make up for the shortfall from the restrictions on the
payments to the University of Michigan in 1995, the state has
proposed making payments of about $590 million, including federal
funds of $335 million, to 92 government-owned hospitals and community
health boards. According to the proposal, these governmental
entities will then transfer the funds back to the state, thereby
allowing the state to continue to benefit from federal Medicaid
matching funds.
Under each of the financing mechanisms, providers that received
Medicaid payments from the state in turn paid the state almost as
much as they received.\46 In effect, in fiscal year 1993 Michigan
increased the federal percentage share of total Medicaid spending
from 56 percent to 68 percent.
--------------------
\45 This legislation limits such payments for 1995 to 200 percent of
each qualifying hospital's costs for Medicaid and uninsured patients,
less the hospital's total Medicaid reimbursement and payments
received from the uninsured. In subsequent years, the act limits DSH
payments to the amount of a hospital's uncovered costs.
\46 HCFA advised us that other states are using similar financing
arrangements.
THE URBAN INSTITUTE'S DSH STUDY
-------------------------------------------------------- Appendix II:3
In addition to our work on this issue, a December 1994 Urban
Institute study reported on how 39 states developed special financing
programs.\47 States collected $5.8 billion through intergovernmental
transfers, and taxes and donations from private and county health
care providers. This money was augmented with $7.6 billion from
federal matching payments, for a total of about $13.4 billion. The
states used these revenues to make payments to private, county, and
state providers.
The study found that most of the DSH payments, about $8.5 billion,
were made to help needy private and county providers. Since these
providers contributed $5.8 billion, they realized a gain of $2.7
billion. Hospital officials stated that they used these funds to (1)
cover overall hospital operations, including uncompensated care and
losses on Medicaid patients; (2) provide AIDS services; (3) pay for
capital expenditures, such as opening a new clinic, purchasing an
ambulance, or replacing x-ray equipment; (4) secure a commercial bank
loan; (5) maintain day-to-day operations because Medicaid payments
were often in arrears; and (6) generate income that resulted from
placing the funds in interest bearing trusts. These officials
emphasized that although the DSH payments helped their hospitals,
they were not enough to cover total uncompensated care losses.
State hospitals received $4.8 billion in DSH payments. However,
hospital officials indicated that only a small share of the gains
were actually retained and available to pay for health care services,
such as uncompensated care. Instead, most of the gains were
transferred back to state general revenue accounts. Here, they were
mixed with other state funds and used to help balance overall state
budgets. In some cases, the extra funds were used to support
Medicaid, mental health, or general health and welfare spending.
Hospital officials believed that to the extent that the DSH programs
helped the overall state budget, the programs indirectly helped the
hospital's budget. State officials believed that these additional
funds prevented larger cuts in the Medicaid program.
--------------------
\47 Medicaid Disproportionate Share and Other Special Financing
Programs: A Fiscal Dilemma for States and the Federal Government,
The Urban Institute, (Dec. 1994). The study was sponsored by the
Kaiser Commission on the Future of Medicaid. It obtained information
from 39 state Medicaid agencies relating to special financing
programs and described the types of hospitals that participated in
the programs, how much they contributed, and how much they received
in DSH and related payments. To obtain detailed information about
the use of DSH funds at the state, county, and hospital levels, case
studies, using telephone interviews, were conducted in six states.
GAO PRODUCTS ON MEDICAID
========================================================= Appendix III
Medicaid: Restructuring Approaches Leave Many Questions
(GAO/HEHS-95-103, Apr. 4, 1995).
Uninsured and Children on Medicaid (GAO/HEHS-95-83R, Feb. 14, 1995).
Medicaid Long-Term Care: Successful State Efforts to Expand Home
Services While Limiting Costs (GAO/HEHS-94-167, Aug. 11, 1994).
Medicaid: Changes in Best Price for Outpatient Drugs Purchased by
HMOs and Hospitals (GAO/HEHS-94-194FS, Aug. 5, 1994).
Medicaid: States Use Illusory Approaches to Shift Program Costs to
Federal Government (GAO/HEHS-94-133, Aug. 1, 1994).
Health Care Reform: Potential Difficulties in Determining
Eligibility for Low-Income People (GAO/HEHS-94-176, July 11, 1994).
Medicaid Prenatal Care: States Improve Access and Enhance Services,
but Face New Challenges (GAO/HEHS-94-152BR, May 10, 1994).
Managed Health Care: Effect on Employers' Costs Difficult to Measure
(GAO/HRD-94-3, Oct. 19, 1993).
Medicaid Managed Care: Healthy Moms, Healthy Kids--A New Program for
Chicago (GAO/HRD-93-121, Sept. 7, 1993).
Medicaid: Improving Funds Distribution (GAO/HRD-93-112FS, Aug. 20,
1993).
Medicaid Drug Fraud: Federal Leadership Needed to Reduce Program
Vulnerabilities (GAO/HRD-93-118, Aug. 2, 1993).
Medicaid Estate Planning (GAO/HRD-93-29R, July 20, 1993).
Medicaid: Data Improvements Needed to Help Manage Health Care
Program (GAO/IMTEC-93-18, May 13, 1993).
Medicaid: HealthPASS--An Evaluation of a Managed Care Program for
Certain Philadelphia Recipients (GAO/HRD-93-67, May 7, 1993).
Medicaid: The Texas Disproportionate Share Program Favors Public
Hospitals (GAO/HRD-93-86, Apr. 30, 1993).
Medicaid Formula Alternatives (GAO/HRD-93-18R, Mar. 31, 1993).
Medicaid: Outpatient Drug Costs and Reimbursements for Selected
Pharmacies in Illinois and Maryland (GAO/HRD-93-55FS, Mar. 18,
1993).
Medicaid: States Turn to Managed Care to Improve Access and Control
Costs (GAO/HRD-93-46, Mar. 17, 1993).
Medicaid Formula Alternative (GAO/HRD-93-17R, Mar. 2, 1993).
Medicaid: Changes in Drug Prices Paid by HMOs and Hospitals Since
Enactment of Rebate Provisions (GAO/HRD-93-43, Jan. 15, 1993).
Medicaid: Disproportionate Share Policy (GAO/HRD-93-3R, Dec. 22,
1992).
Medicaid: Oregon's Managed Care Program and Implications for
Expansions (GAO/HRD-92-89, June 19, 1992).
Medicaid: Ensuring That Noncustodial Parents Provide Health
Insurance Can Save Costs (GAO/HRD-92-80, June 17, 1992).
Medicaid Third-Party Liability (GAO/HRD-92-21R, Mar. 3, 1992).
Medicaid: Changes in Drug Prices Paid by VA and DOD Since Enactment
of Rebate Provisions (GAO/HRD-91-139, Sept. 18, 1991).
Managed Care: Oregon Program Appears Successful But Expansion Should
Be Implemented Cautiously (GAO/T-HRD-91-48, Sept. 16, 1991).
Medicaid Expansions: Coverage Improves But State Fiscal Problems
Jeopardize Continued Progress (GAO/HRD-91-78, June 25, 1991).
Substance Abuse Treatment: Medicaid Allows Some Services But
Generally Limits Coverage (GAO/HRD-91-92, June 13, 1991).
Medicaid: Alternatives for Improving the Distribution of Funds
(GAO/HRD-91-66FS, May 20, 1991).
Medicaid: HCFA Needs Authority to Enforce Third-Party Requirements
on States (GAO/HRD-91-60, Apr. 11, 1991).
Medicaid: Legislation Needed to Improve Collections From Private
Insurers (GAO/HRD-91-25, Nov. 30, 1990).
Medicaid: Millions of Dollars Not Recovered From Michigan Blue
Cross/Blue Shield (GAO/HRD-91-12, Nov. 30, 1990).
Long-Term Care Insurance: Proposals to Link Private Insurance and
Medicaid Need Close Scrutiny (GAO/HRD-90-154, Sept. 10, 1990).
Nursing Homes: Admission Problems for Medicaid Recipients and
Attempts to Solve Them (GAO/HRD-90-135, Sept. 5, 1990).
Medicaid: Oversight of Health Maintenance Organizations in the
Chicago Area (GAO/HRD-90-81, Aug. 27, 1990).
Medicaid: States Expand Coverage for Pregnant Women, Infants, and
Children (GAO/HRD-89-90, Aug. 16, 1989).
Medicaid: Recoveries From Nursing Home Residents' Estates Could
Offset Program Costs (GAO/HRD-89-56, Mar. 7, 1989).
Medicaid: Some Recipients Neglect to Report U.S. Savings Bond
Holdings (GAO/HRD-89-43, Jan. 18, 1989).
Medicaid: Changing Medicaid Formula Can Improve Distribution of
Funds to States (GAO/GGD-83-27, Mar. 9, 1983).
(See figure in printed edition.)Appendix IV
COMMENTS FROM THE ADMINISTRATION
========================================================= Appendix III
(See figure in printed edition.)
(See figure in printed edition.)
(See figure in printed edition.)
(See figure in printed edition.)
MAJOR CONTRIBUTORS TO THIS REPORT
=========================================================== Appendix V
GAO CONTACTS
William J. Scanlon, Associate Director, (202) 512-4561
James C. Cosgrove, Assistant Director, (202) 512-7029
Alfred R. Schnupp, Evaluator-in-Charge, (202) 512-7159
ACKNOWLEDGMENTS
In addition to those named above, the following individuals made
important contributions to this report: Hannah F. Fein wrote major
sections of the draft; Michael Gutowski, Walter S. Ochinko, and
Cheryl A. Williams provided information and analysis on section 1115
Medicaid waivers; Richard N. Jensen provided advice on Medicaid
programmatic details; Paul T. Wagner, Jr. gathered data and
produced the graphics.