This is the accessible text file for GAO report number GAO-08-87 
entitled 'Medicaid Demonstration Waivers: Recent HHS Approvals Continue 
to Raise Cost and Oversight Concerns' which was released on March 3, 
2008. 

This text file was formatted by the U.S. Government Accountability 
Office (GAO) to be accessible to users with visual impairments, as part 
of a longer term project to improve GAO products' accessibility. Every 
attempt has been made to maintain the structural and data integrity of 
the original printed product. Accessibility features, such as text 
descriptions of tables, consecutively numbered footnotes placed at the 
end of the file, and the text of agency comment letters, are provided 
but may not exactly duplicate the presentation or format of the printed 
version. The portable document format (PDF) file is an exact electronic 
replica of the printed version. We welcome your feedback. Please E-mail 
your comments regarding the contents or accessibility features of this 
document to Webmaster@gao.gov. 

This is a work of the U.S. government and is not subject to copyright 
protection in the United States. It may be reproduced and distributed 
in its entirety without further permission from GAO. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately. 

Report to Congressional Requesters: 

United States Government Accountability Office: 

GAO: 

January 2008: 

Medicaid Demonstration Waivers: 

Recent HHS Approvals Continue to Raise Cost and Oversight Concerns: 

Medicaid Demonstration Waivers: 

GAO-08-87: 

GAO Highlights: 

Highlights of GAO-08-87, a report to congressional requesters 

Why GAO Did This Study: 

Medicaid, a joint federal and state program, finances health care for 
60 million low-income people. Section 1115 of the Social Security Act 
authorizes the Secretary of Health and Human Services to waive certain 
federal Medicaid requirements and allow demonstration projects that are 
likely to promote Medicaid objectives. Under federal policy, states 
must show that federal spending for proposed demonstrations will be no 
greater than if the state’s existing Medicaid program were continued. 

GAO examined the extent to which HHS ensured that recent comprehensive 
1115 demonstrations—affecting a broad range of services for 
beneficiaries statewide—will (1) be budget neutral to the federal 
government and (2) maintain Medicaid’s fiscal integrity. For 
demonstrations approved in 2005 (Florida and Vermont), GAO obtained 
information from federal and state officials and also relied on past 
reviews of other demonstrations. 

What GAO Found: 

HHS did not adequately ensure that Florida’s and Vermont’s Medicaid 
demonstrations will be budget neutral to the federal government before 
approving them. HHS approved spending limits that were higher than the 
limits that would have been granted if HHS had held the states to 
limits based on benchmark growth rates, that is, the lower of the 
state’s historical spending growth or nationwide estimates of Medicaid 
growth. Although HHS allows states to deviate from these benchmarks if 
states can show that using them would not provide accurate projections, 
HHS’s basis for approving the higher spending limits was not fully 
supported by documentation. In Florida, HHS approved a $52.6 billion 
spending limit for the 5 year demonstration— $6.9 billion more than the 
documentation supported. In Vermont, HHS approved a $4.7 billion 
spending limit—$246 million higher than supported. 

HHS also did not ensure that the two demonstrations maintain Medicaid’s 
fiscal integrity. In Florida, HHS allowed the state to establish a 
spending limit using a historical spending base that included payments 
HHS had previously identified as problematic. In 2005, an HHS review 
found several problems with the payment arrangement—problems that 
potentially resulted in inflated and inaccurate payments. In Vermont, 
where the state proposed operating a managed care organization, HHS 
agreed to an administrative reimbursement rate higher than what the 
state received prior to the demonstration. Under this arrangement, the 
state can use excess revenues to pay for health-related programs that 
were previously funded by the state and that do not exclusively benefit 
Medicaid beneficiaries, such as a grant to the University of Vermont 
medical school. A July 2007 GAO letter to the Secretary discussed 
concerns about this approval’s consistency with federal law and 
recommended that the Secretary reexamine Vermont’s demonstration and, 
where appropriate, either modify its terms or seek statutory authority 
for it to continue in its current form. 

Concerns about HHS’s demonstration approval process in this report are 
consistent with those GAO has raised in past reviews of other states’ 
demonstration proposals. In 2002 and 2004, GAO recommended that HHS 
take steps to strengthen its fiscal oversight of Medicaid by improving 
the Medicaid demonstration review and approval process, in part by (1) 
clarifying criteria for reviewing and approving states’ demonstration 
spending limits, (2) better ensuring that valid methods are used to 
demonstrate budget neutrality and (3) documenting and making public 
material explaining the basis for any approvals. HHS has not taken 
action on these recommendations and maintains that its process is 
sufficient. Because HHS continues to disagree with these 
recommendations and with the need to reexamine the Vermont 
demonstration, GAO is elevating these issues to the Congress for 
consideration. 

What GAO Recommends: 

GAO recommends that the Congress consider (1) requiring HHS to improve 
the demonstration review and approval process and (2) addressing HHS’s 
authority to approve demonstrations such as Vermont’s. GAO recommends 
that HHS reexamine Florida’s spending limit. In its comments, HHS 
stated that its process was sufficient. GAO believes that the limit 
allows spending that should not be allowed. 

To view the full product, including the scope and methodology, click on 
[hyperlink, http://www.GAO-08-87]. For more information, contact 
Marjorie Kanof at (202) 512-7114 or kanofm@gao.gov. 

[End of section] 

Contents: 

Letter: 

Results in Brief: 

Background: 

HHS Did Not Adequately Ensure the Budget Neutrality of Medicaid 
Demonstrations in Florida and Vermont before Approving Them: 

HHS Has Not Ensured That Demonstrations in Florida and Vermont Maintain 
the Fiscal Integrity of the Medicaid Program: 

Conclusions: 

Matters for Congressional Consideration: 

Recommendation for Executive Action: 

Agency and State Comments and Our Evaluation: 

Appendix I: Scope and Methodology: 

Appendix II: Comments from the Department of Health and Human Services: 

Appendix III: Comments from the State of Florida: 

Appendix IV: Comments from the State of Vermont: 

Appendix V: GAO Contact and Staff Acknowledgments: 

Related GAO Products: 

Tables: 

Table 1: Spending Limit for Florida's Medicaid Demonstration as 
Proposed, Approved, and Calculated under HHS's Benchmark Policy and 
Supported by HHS's Explanations: 

Table 2: Comparison of Florida's Per Person Growth Rates as Proposed 
and Approved under HHS's Benchmark Policy and Supported by HHS's 
Explanations: 

Table 3: Spending Limit for Vermont's Medicaid Demonstration as 
Proposed, Approved, and Calculated under HHS's Benchmark Policy and 
Supported by HHS's Explanations: 

Table 4: Comparison of Vermont's Beneficiary Enrollment Growth Rates as 
Proposed and Approved under HHS's Benchmark Policy and Supported by 
HHS's Explanations: 

Table 5: HHS-Approved Distribution of Florida's Annual Low-Income Pool 
of Federal, State, and Local Funds, Demonstration Year One: 

Table 6: Examples of Vermont's Use of Excess Medicaid Revenues under 
Its Demonstration: 

Figures: 

Figure 1: Overview of Process for Projecting the Future Cost of a 
State's Existing Medicaid Program: 

Figure 2: Vermont Health Access Program Surpluses and Deficits: 

Abbreviations: 

CMS: Centers for Medicare & Medicaid Services: 
HHS: Department of Health and Human Services: 
HIFA: Health Insurance Flexibility and Accountability: 
OMB: Office of Management and Budget: 
PAS: Provider Access System: 
SCHIP: State Children's Health Insurance Program: 
UPL: Upper Payment Limit: 

United States Government Accountability Office: 

Washington, DC 20548: 

January 31, 2008: 

The Honorable Henry A. Waxman: 
Chairman: 
Committee on Oversight and Government Reform: 
House of Representatives: 

The Honorable John D. Dingell: 
Chairman: 
Committee on Energy and Commerce: 
House of Representatives: 

The Honorable Frank J. Pallone, Jr.: 
Chairman: 
Subcommittee on Health: 
Committee on Energy and Commerce: 
House of Representatives: 

The Honorable Sherrod Brown: 
United States Senate: 

Medicaid is a federal-state program that finances health care services 
for about 60 million low-income individuals, including children and 
aged or disabled adults. Established in 1965 under title XIX of the 
Social Security Act, Medicaid consists of more than 50 distinct state- 
based programs that cost the federal government and states an estimated 
$317 billion in fiscal year 2005.[Footnote 1] Each state administers 
its Medicaid program within federal requirements established in statute 
and regulations, and the federal government shares in the cost of each 
state's program by paying an established share of reported 
expenditures.[Footnote 2] Under section 1115 of the Social Security 
Act, however, the Secretary of Health and Human Services may waive 
certain federal requirements for demonstrations that the Secretary 
deems likely to promote Medicaid objectives--allowing states to test 
and evaluate new approaches for delivering Medicaid services. 

In the early 1980s, the Department of Health and Human Services 
(HHS)[Footnote 3] adopted a policy that required states to document 
that their proposed demonstrations would be budget neutral[Footnote 4] 
to the federal government, that is, the federal government will spend 
no more with the demonstrations than without them. Each demonstration 
operates under a negotiated budget neutrality agreement that places 
limits on federal Medicaid spending over the life of the demonstration. 
A spending limit governing a demonstration is based on the projected 
costs of the existing Medicaid program without the demonstration. 
States estimate the cost of continuing their existing Medicaid programs 
by projecting growth in per person costs and beneficiary enrollment 
over the 5-year standard demonstration period. HHS policy guidance 
states that spending limits are based on estimates of growth (growth 
rates) that are the lower of (1) the state's historical growth for 
Medicaid in recent years or (2) Medicaid growth rates projected for the 
nation. These estimates are termed benchmark rates in this 
report.[Footnote 5] 

In 1995, 2002, and 2004, we reported that HHS had not adequately 
ensured that approved Medicaid demonstrations would be budget neutral 
to the federal government.[Footnote 6] The core of our findings 
included that (1) HHS approved spending limits that were based on 
projections of growth that exceeded state-specific and nationwide 
benchmarks, (2) HHS approved spending limits that included costs that 
were impermissible or inappropriate, and (3) the basis for HHS's 
approval of states' demonstration spending limits was unclear and the 
process by which this was done was largely undocumented. We have also 
reported numerous times since the early 1990s about some states' 
financing arrangements that took advantage of the flexibility in the 
Medicaid program to boost the federal support they received for the 
program at little or no cost to states. Under one such financing 
arrangement, for example, states made illusory Medicaid payments to 
certain government-owned providers--payments in excess of standard 
Medicaid reimbursement rates, otherwise known as supplemental payments-
-to obtain the federal share on the supplemental payments, then 
required providers to return most or all of the payments to the 
state.[Footnote 7] We concluded that such practices undermined the 
fiscal integrity of the program, and HHS in recent years has sought to 
curtail them. Both of these issues--lack of budget neutrality and 
concerns about fiscal integrity--have contributed to our designating 
Medicaid as a high-risk program since 2003.[Footnote 8] 

You expressed interest in the costs of Medicaid section 1115 
demonstrations to the federal government and asked us to examine recent 
demonstration proposals approved by HHS. We selected demonstrations 
based on when they were approved--we selected demonstrations approved 
from July 2004 through December 2006--and whether they were 
comprehensive and accounted for the majority of the state's Medicaid 
expenditures.[Footnote 9] Two section 1115 demonstrations that HHS 
approved--for Florida and Vermont in 2005--met these criteria. Both 
demonstrations involve expanding the use of managed care to deliver 
services to Medicaid beneficiaries: in Florida, by requiring certain 
Medicaid beneficiaries to enroll in competing state-approved managed 
care plans; in Vermont, by creating a single state-run managed care 
organization. We have already reported to you on certain aspects of 
HHS's approval of demonstrations in Florida and Vermont. In July 2007, 
we reported that demonstrations in Florida and Vermont have mixed 
implications for beneficiaries and that opportunities for public input 
to HHS during the approval process were limited.[Footnote 10] Also in a 
July 2007 letter to the Secretary of HHS, we reported concerns about 
the consistency of the Florida and Vermont demonstrations with federal 
law: in Florida, about HHS allowing limits on covered benefits and cost 
sharing in excess of statutory limits without addressing statutory 
restrictions on its authority to do so; in Vermont, about HHS allowing 
the state to operate its own Medicaid managed care organization through 
a contract between two related state agencies and, through this 
arrangement, to apply Medicaid funds to programs previously funded by 
the state.[Footnote 11] This report addresses the extent to which the 
Secretary of HHS ensured, before approving them, that the Florida and 
Vermont demonstrations will (1) be budget neutral to the federal 
government and (2) maintain the fiscal integrity of the Medicaid 
program. 

To determine the extent to which HHS ensured that the demonstrations in 
Florida and Vermont will be budget neutral, we examined each state's 
projection of the combined federal and state spending needed to 
continue its existing Medicaid program. We compared the assumptions 
about cost and beneficiary enrollment growth used to develop the 
demonstration spending limits approved by HHS against our estimates of 
spending limits had HHS's benchmarks been used. We asked officials of 
the state Medicaid agencies, HHS, and the Office of Management and 
Budget (OMB) for explanations and quantitative support for spending 
projections that used growth assumptions exceeding HHS benchmarks and 
estimated the spending limits supported by these explanations and 
documentation.[Footnote 12] 

To determine the extent to which HHS ensured that the two 
demonstrations will maintain the fiscal integrity of the Medicaid 
program, we evaluated HHS's process for reviewing section 1115 
demonstration proposals and whether the demonstrations held potential 
for inappropriately leveraging federal Medicaid funds. We reviewed 
Florida and Vermont demonstration-related materials, including the 
demonstration proposals and supporting documentation, correspondence 
between HHS and the two states, and the special terms and conditions 
that govern implementation, operation, and evaluation of approved 
demonstrations. We also reviewed HHS documentation related to the 
states' Medicaid financing methods and supplemental payment 
arrangements. And we met with state and federal officials, including 
officials from HHS's Office of the Actuary, the Division of Family and 
Children's Health Programs Group (the division within the Centers for 
Medicare & Medicaid Services (CMS) that reviews section 1115 
demonstration proposals), the Division of Reimbursement and State 
Financing (the division within CMS that monitors the appropriateness of 
state financing arrangements), and from OMB. Appendix I more fully 
discusses our scope and methodology. We conducted our work from June 
2006 through January 2008 in accordance with generally accepted 
government auditing standards. 

Results in Brief: 

HHS did not adequately ensure that Florida's and Vermont's Medicaid 
demonstrations will be budget neutral to the federal government before 
approving them. The spending limits that HHS approved for the two 
demonstrations were higher than the limits that would have been granted 
if HHS had held the states to limits based on HHS's benchmark growth 
rates. Although HHS allows states to deviate from these benchmarks if 
states can show that using them would not provide accurate projections, 
HHS's basis for approving the higher spending limits was not fully 
supported by documentation. HHS provided support for part of the 
increase but not for the entire amount. 

* For Florida, HHS approved a 5-year spending limit for the 
demonstration estimated at $52.6 billion,[Footnote 13] an amount $6.9 
billion higher than supported. HHS approved demonstration spending 
limits based on projections of cost growth that exceeded HHS's 
benchmarks of the state's own recent history of Medicaid program growth 
and estimates of Medicaid growth nationwide. Specifically, HHS approved 
cost growth rates for one group of beneficiaries based on a selected 
period of unusually high growth from Florida's historical experience 
and cost growth rates for another group of beneficiaries based on 
unsupported increases to nationwide estimates. 

* For Vermont, HHS approved a 5-year spending limit for the 
demonstration of $4.7 billion, an amount $246 million higher than 
supported. Similar to what it did for Florida, although to a lesser 
degree, HHS approved a spending limit using rates for projecting 
enrollment growth that were higher than the state's historical growth 
and projections of Medicaid growth nationwide and did not fully support 
its reasons. HHS also allowed Vermont to boost its spending limit by 
allowing the state to include projections of spending that were 
"hypothetical" in the state's $4.7 billion spending limit, specifically 
funds that the state could have spent on a previous Medicaid section 
1115 demonstration but did not spend. 

HHS also did not ensure that demonstrations in Florida and Vermont 
maintained the fiscal integrity of the Medicaid program prior to 
approving them. 

* In Florida, HHS allowed the state to use spending from a supplemental 
payment arrangement that the state had in place prior to the 
demonstration as the basis for allowed spending under the 
demonstration. A 2005 HHS financial management review found several 
problems with this earlier arrangement, which involved supplemental 
payments to certain hospitals. The review found that Florida had 
incorrectly calculated the level of supplemental payments for which 
federal Medicaid funds were obtained, potentially resulting in inflated 
and inaccurate payments. Without correcting these problems, HHS allowed 
Florida to use the spending under the prior hospital supplemental 
payment arrangement as the basis for the spending allowed under the 
demonstration. To address problems with inaccurate methods and data for 
calculating allowable supplemental payment amounts used by other 
states, we had, in 2004, recommended that HHS establish appropriate 
methods for states' calculations of supplemental payments, but HHS has 
not implemented this recommendation for any state. 

* In Vermont, HHS allowed the state to operate its own managed care 
organization and to claim federal matching funds for payments to the 
organization, and agreed to reimburse Vermont for the administration of 
its public managed care organization at a rate higher than that 
typically paid to state Medicaid agencies. Specifically, the financing 
arrangement allows the state to pay its managed care organization for 
administrative costs at a rate that--although typical for private 
managed care organizations--is higher than the rate paid to Vermont 
before the demonstration and that of most state Medicaid agencies. 
Under this arrangement, HHS allowed the state to retain excess 
revenues[Footnote 14] and use these funds to support health-related 
programs that were previously funded by the state and that do not 
exclusively benefit Medicaid beneficiaries, such as a grant to the 
University of Vermont medical school. In July 2007, we raised concerns 
about this approval's consistency with federal law.[Footnote 15] In 
particular, the approval of the Vermont demonstration raised the 
question whether the Vermont Medicaid agency could enter into a managed 
care contract with one of its own offices and receive federal matching 
funds for lump-sum payments to that office rather than for payments 
based on actual costs. 

The findings in this report are consistent with certain findings in our 
earlier reports and indicate that action is still needed to ensure the 
transparency of added costs to the federal government associated with 
section 1115 demonstrations and to maintain Medicaid's fiscal 
integrity. HHS has not addressed long-standing concerns with the 
demonstration approval process, including the validity of its methods 
for determining budget neutrality, the basis for its approval of 
states' spending limits, and the transparency of the review process. 
Further, HHS continues to maintain that its approval of Vermont's 
demonstration is consistent with federal law, but has not addressed the 
concerns raised in our July 2007 letter. Consequently, this report 
includes two matters for congressional consideration: 

* The Congress should consider requiring increased attention to fiscal 
responsibility in the approval of section 1115 Medicaid demonstrations 
by requiring the Secretary of HHS to improve the demonstration review 
process through steps such as (1) clarifying criteria for reviewing and 
approving states' proposed spending limits, (2) ensuring that valid 
methods are used to demonstrate budget neutrality, and (3) documenting 
the basis for any approvals. 

* The Congress should consider addressing whether demonstrations that 
allow states to operate public managed care organizations and retain 
excess revenue to support programs previously funded by the state-- 
including the Vermont demonstration--are within the scope of the 
Secretary of HHS's authority under section 1115 of the Social Security 
Act. 

In this report we are also recommending that the Secretary of HHS 
calculate the level of supplemental payments for which Florida could 
have obtained federal Medicaid funds in the absence of the proposed 
demonstration, using appropriate methods and accurate data sources, and 
adjust Florida's spending limit accordingly. 

In commenting on a draft of this report, HHS strongly disagreed with 
our findings, conclusions, and recommendation. HHS commented that the 
draft report did not accurately characterize the demonstration programs 
or HHS's budget neutrality policies. HHS also noted that the draft did 
not adequately account for the likelihood of differences in 
professional interpretation of quantifiable analyses or adequately 
acknowledge HHS's efforts to ensure Medicaid compliance and fiscal 
integrity. HHS emphasized that the demonstrations are approved at the 
discretion of the Secretary of HHS and that HHS's review of 
demonstration proposals includes both budgetary and programmatic 
elements. With regard to HHS's approval of the Vermont demonstration, 
HHS disagreed with our concerns and earlier recommendation to reexamine 
the demonstration and, where appropriate, either modify the 
demonstration's terms or seek statutory authority for it to continue in 
its current form. HHS maintained that issues of legal authority were 
adequately and appropriately addressed in the information provided to 
us during the course of our fieldwork. 

We believe our findings, conclusions, and recommendation remain valid. 
Our characterizations of the programs and policies were based on 
documentation obtained from HHS and states and discussions with federal 
and state officials, and we believe we have captured them accurately. 
We acknowledge that the Secretary has some discretion when approving 
section 1115 demonstrations. As noted in this report, budget neutrality 
is a long-standing HHS policy, but is not required by law. We maintain, 
however, that to provide accountability and transparency in federal 
spending for the Medicaid program, the Secretary's approvals should be 
based on clearly articulated policies and spending limits that are 
consistent with these policies. Whenever HHS's decisions and spending 
estimates met these tests, we accepted them. We did not, however, 
accept estimates when program officials could not document or clearly 
articulate the reasoning they had used, demonstrate how this reasoning 
was consistent with budget neutrality and fiscal integrity principles, 
and explain how the resulting spending limits were derived. Because our 
findings in Florida and Vermont are consistent with findings from our 
earlier work, we believe that actions to improve the demonstration 
approval process--including the criteria used, the methods allowed to 
determine budget neutrality, and the documentation to support the final 
approved limits--are needed. Given HHS's opposition to taking 
recommended actions, we believe that elevating certain long-standing 
recommendations for congressional consideration is a necessary step. 

We also provided a draft of this report to Florida and Vermont. Florida 
stated that during the negotiations of the demonstration proposal, 
state officials worked closely with HHS to ensure that all data and 
documentation were provided in a timely and accurate manner to support 
the proposal. Vermont indicated that the state had assumed an 
unprecedented amount of risk related to program expenditures in 
exchange for the flexibility granted by the Secretary and that state 
and federal staff had engaged in extensive discussion and analysis of 
Vermont's historical expenditures, cost and caseload trends, and 
program policies in arriving at the final spending limit. On the basis 
of our review of available documentation, we agree that the states 
provided data and documentation to HHS supporting their demonstration 
proposals. Our concern remains, however, with the lack of 
documentation--from the states or HHS--showing how the final spending 
limits were derived, particularly since they were based on assumptions 
about cost and enrollment growth that were higher than HHS's 
benchmarks. In addition, our legal concerns about the Vermont 
demonstration remain. Consequently, we have elevated this matter to the 
Congress for consideration. 

Background: 

Medicaid is one of the largest programs in federal and state budgets. 
In fiscal year 2005, Medicaid expenditures totaled an estimated $317 
billion. States pay qualified health providers for a broad range of 
covered services provided to eligible beneficiaries. The federal 
government reimburses states for a share of these expenditures. The 
federal matching share of each state's Medicaid expenditures for 
services is determined by a formula defined under federal law and can 
range from 50 to 83 percent.[Footnote 16] 

Each state administers its Medicaid program in accordance with a state 
Medicaid plan that must be approved by HHS.[Footnote 17] Traditional 
Medicaid programs represent an open-ended entitlement, meaning the 
state will enroll all eligible individuals who apply for Medicaid, and 
both the state and federal government will pay their shares of 
expenditures for individuals covered under a state's approved Medicaid 
plan. States have considerable flexibility in designing their Medicaid 
programs, but under federal Medicaid law, states generally must meet 
certain requirements for what benefits are provided and who is eligible 
for the program. 

Medicaid demonstrations provide a way for states to innovate outside of 
many of Medicaid's usual requirements. Under section 1115 of the Social 
Security Act, the Secretary has authority to waive certain federal 
Medicaid requirements and authorize otherwise unallowable expenditures 
for "experimental, pilot, or demonstration projects" that are likely to 
promote Medicaid objectives.[Footnote 18] States have used the 
flexibility granted through section 1115 to implement major changes to 
existing state Medicaid programs. For example, some states used 
Medicaid section 1115 demonstrations in the 1980s and 1990s to 
introduce mandatory managed care for their Medicaid beneficiaries. 

Since the early 1980s, HHS has required that states show that their 
proposed section 1115 demonstrations will be budget neutral to the 
federal government--that is, federal expenditures under a state's 
demonstration will not be greater than if the state had continued its 
existing Medicaid program. HHS requires states to show that proposed 
demonstrations are budget neutral by preparing 5-year projections of 
spending (1) under the current Medicaid program and (2) under the 
proposed demonstration. HHS policy states that for a demonstration to 
be considered budget neutral, the federal share of projected Medicaid 
expenditures under the demonstration can be no greater than the federal 
share of projected Medicaid expenditures based on continuing the 
existing Medicaid program. 

Budget Neutrality Is Based on the Projected Cost of the Existing 
Medicaid Program: 

Because HHS bases spending limits for proposed demonstrations on the 
projected cost of continuing an existing Medicaid program, a state has 
an incentive to maximize its projected costs. HHS policy states that 
the federal share of spending on demonstrations will be limited by 
spending limits calculated from two components: 

* Spending base. States select a recently completed fiscal year that 
establishes base levels of funding for services and programs affected 
by the proposed demonstration--a state's "spending base." States also 
identify beneficiary groups for inclusion in the proposed 
demonstration. These beneficiary groups can, at the Secretary's 
discretion, include individuals enrolled in other demonstrations a 
state may be operating and beneficiaries from the State Children's 
Health Insurance Program (SCHIP), which provides health coverage to 
children in families whose incomes, while low, are above Medicaid's 
eligibility requirements.[Footnote 19] 

* Growth rates. States should submit to HHS 5 years of historical data 
for per person costs and beneficiary enrollment in their existing 
Medicaid programs, including quantified explanations for anomalies in 
their historical trends. HHS policy says that spending limits should be 
based on growth rates that are the lower of state-specific history or 
estimates of nationwide growth for the beneficiary groups included in 
the demonstration (referred to in this report as benchmark growth 
rates). HHS's guidance is specific to per person cost growth rates and 
does not explicitly address the application of enrollment growth rates; 
however, HHS refers to state historical and nationwide enrollment 
growth rates in considering the spending limits. Nationwide estimates 
of cost and beneficiary enrollment growth are developed by CMS 
actuaries to assist OMB in preparing the President's Budget. 

To project the costs of continuing a state's existing Medicaid program, 
HHS policy calls for applying the benchmark growth rates to the state's 
spending base over a 5-year period to establish total projected costs 
absent the demonstration.[Footnote 20] HHS sets spending limits for 
proposed demonstrations based in part on these total projected costs 
(see fig. 1). 

Figure 1: Overview of Process for Projecting the Future Cost of a 
State's Existing Medicaid Program: 

This figure is a flowchart showing an overview of process for 
projecting the future cost of a state's existing medicaid program. 

[See PDF for image] 

Source: GAO analysis of HHS information. 

[A] A state may propose, for example, that certain groups of 
beneficiaries, such as aged and disabled beneficiaries, will operate 
under the terms of the state's approved state Medicaid plan rather than 
under the terms of the demonstration. In the Florida and Vermont 
demonstrations, major Medicaid beneficiary groups--aged, blind, and 
disabled beneficiaries and families and children--were included in the 
demonstrations. 

[End of figure] 

HHS allows states to use higher-than-benchmark growth rates if they can 
establish that historical or nationwide data do not accurately depict 
anticipated growth in the state Medicaid program. HHS considers 
spending limits to be a product of negotiations that are informed by 
HHS's policy to consider the state's historical experience and 
projections of growth in the President's Budget. In addition, HHS's 
policy indicates that states, in providing HHS with state-specific 
historical growth rates, must quantify any anomalies in the trends. 

Recent Demonstrations Approved in Florida and Vermont Allow Significant 
Changes in How These States Operate Their Medicaid Programs: 

Recently approved section 1115 Medicaid demonstrations in Florida and 
Vermont significantly change the operation of the two states' Medicaid 
programs. Both demonstrations expand the use of managed care by 
requiring most Medicaid beneficiaries to enroll in managed care 
plans:[Footnote 21] Florida through state contracts with multiple 
managed care plans to provide services and Vermont by creating a single 
managed care organization operated by an office within the state 
Medicaid agency.[Footnote 22] 

Florida: Approved by HHS in October 2005 and launched in July 2006, 
Florida's demonstration is designed to give Medicaid beneficiaries 
different options for health care plans and benefits through increased 
use of managed care plans to provide Medicaid coverage to 
beneficiaries, in a competitive environment.[Footnote 23] In the 
initial phase of the demonstration, certain Medicaid beneficiaries in 
two counties are required to enroll in state-approved managed care 
plans. Managed care plans compete for Medicaid beneficiaries by 
offering different coverage options, including customized benefits and 
cost sharing, subject to certain limitations. Unlike many other 
previous Medicaid managed care systems, managed care plans in Florida 
have the authority to design benefit packages subject to approval by 
the state. Initially implemented in a two-county area, the managed care 
components of the demonstration are planned for statewide 
implementation by June 2010.[Footnote 24] Another key component of 
Florida's demonstration was the establishment of a pool of funds to 
finance supplemental payments--payments above the state's usual payment 
rate--to certain types of Florida health care providers. Known as the 
low-income pool, this component of the demonstration was designed in 
part to continue funding for a supplemental payment program for 
hospitals that the state had in place prior to the demonstration. 
Payments from the $5 billion low-income pool ($1 billion annually) are 
authorized for selected Medicaid providers statewide to help offset the 
cost of providing care to Medicaid beneficiaries and underinsured and 
uninsured individuals.[Footnote 25] 

Vermont: Approved by HHS in September 2005 and launched the following 
month, Vermont's demonstration is designed to contain costs and, by 
potentially delivering services to Medicaid beneficiaries for less and 
reinvesting excess revenue, to allow the state to serve more of its 
uninsured population. Under the demonstration, Vermont created a 
single, state-operated managed care organization to cover virtually all 
of the state's Medicaid population.[Footnote 26] HHS approved a managed 
care arrangement whereby the state Medicaid agency contracts with one 
of its own components (the Office of Vermont Health Access) to operate 
as a managed care organization.[Footnote 27] The Office of Vermont 
Health Access receives monthly actuarially certified lump-sum payments 
from the state Medicaid agency, which in turn receives the federal 
share of these lump-sum payments. The monthly payment is intended to 
cover the medical costs and administrative expenses of serving enrolled 
beneficiaries. Vermont also received authority to retain "savings," 
that is, any excess revenue generated by the state managed care 
organization, and apply them to programs that meet certain agreed-upon 
health objectives, such as increasing health insurance 
coverage.[Footnote 28] 

Concerns about Budget Neutrality of Medicaid Demonstrations and Certain 
Excessive Supplemental Payments Are Long-standing: 

On several occasions since the mid-1990s, we have reported concerns 
that HHS had approved Medicaid demonstrations that were not budget 
neutral to the federal government. 

* In 1995 we reported that HHS applied new, more flexible budget 
neutrality guidance allowing three states to consider "new 
methodologies" for determining budget neutrality of proposed 
demonstrations. Based in part on these new methodologies, HHS had 
approved spending limits for these demonstrations that were not budget 
neutral and could increase federal Medicaid expenditures.[Footnote 29] 

* In 2002, we reported that HHS approved spending limits for 
demonstrations that were not budget neutral to the federal government 
by allowing two states to include inappropriate or impermissible costs 
in their spending projections. We recommended that HHS ensure that 
valid methods are used to demonstrate budget neutrality by developing 
and implementing consistent criteria for reviewing and approving 
states' budget neutrality analyses. HHS disagreed with the 
recommendation, stating that its methods were valid.[Footnote 30] 

* In 2004, we reported that HHS approved spending limits for section 
1115 demonstrations in four states that were not budget neutral to the 
federal government. These states projected the costs of their Medicaid 
programs at rates of growth exceeding state-specific and nationwide 
benchmarks for Medicaid cost and enrollment growth without documenting 
the rationale for the higher growth rates. We recommended that HHS (1) 
clarify criteria for reviewing and approving spending limits of states' 
proposed demonstrations and (2) reconsider the spending limits of 
recently approved demonstrations. We also recommended that HHS document 
and make public the basis for any section 1115 demonstration approvals, 
including the basis for cost and enrollment growth rates used to set 
spending limits, and ensure that states comply with reporting and 
evaluation requirements.[Footnote 31] HHS concurred with our 
recommendations to make public the basis for its approvals, but did not 
concur with our recommendations on clarifying approval criteria and 
reconsidering recently approved demonstrations using these 
criteria.[Footnote 32] 

Our past work also includes reports addressing concerns with aspects of 
HHS's oversight of certain state supplemental payment arrangements that 
threatened the fiscal integrity of Medicaid's federal-state 
partnership. States, with HHS approval, can make supplemental Medicaid 
payments--payments above the state's usual Medicaid payment rates for 
certain services, such as nursing home care--and they often do so for 
appropriate reasons. For example, states may make supplemental Medicaid 
payments to certain safety net providers that serve a large share of 
high-cost Medicaid beneficiaries. However, our work since the early 
1990s examining some of these arrangements found that many states were, 
in essence, finding ways through the arrangements to inappropriately 
increase the federal share of Medicaid spending at little or no cost to 
the state.[Footnote 33] In February 2004, for example, we reported that 
states were taking advantage of Medicaid upper payment limit (UPL) 
provisions, resulting in excess federal payments. The UPL is the upper 
bound on what the federal government will pay as its share of Medicaid 
costs for different classes of covered services, and this limit often 
exceeds what states actually pay providers for services.[Footnote 34] 
This difference creates a "gap" between what states typically pay for 
services and the UPL. Some states took advantage of this gap between 
their usual payment rates and what Medicaid could pay under the UPL by 
making large supplemental payments to government providers, acquiring a 
federal share of those payments, and subsequently requiring the 
providers to return most or all of the supplemental payments to the 
state. These states have collected billions of excessive federal 
dollars in past years and often used these returned payments and the 
accompanying federal funds to finance their own share of the Medicaid 
program. We have reported on, and HHS has attempted to curb, such 
recycling of federal Medicaid funds. 

HHS Did Not Adequately Ensure the Budget Neutrality of Medicaid 
Demonstrations in Florida and Vermont before Approving Them: 

HHS approved 5-year demonstration spending limits for Florida and 
Vermont based on projections of cost and beneficiary enrollment growth 
rates that exceeded HHS's own benchmarks--that is, the lower of the 
state's recent historical experience or estimates of Medicaid growth 
nationwide--without adequate support for these deviations. For each 
state, HHS provided support for some, but not all, of the increase 
above these benchmark levels. For Florida, the unsupported difference 
totals about $6.9 billion of the $52.6 billion in projected spending 
over the 5-year demonstration. For Vermont, the unsupported difference 
totals about $246 million over its 5-year demonstration. HHS approved 
higher-than-benchmark growth rates in calculating spending limits for 
the demonstrations and, in the case of Vermont, allowed the state to 
include hypothetical projected expenses inappropriately. In particular, 
HHS allowed Vermont to include projected costs in its spending limit 
based on costs that had been budgeted for and allowed under a previous 
1115 demonstration but that had not been spent. Although HHS provided 
some documentation to justify the deviations from its benchmarks that 
it approved, HHS did not justify all the deviations. In some cases, HHS 
officials told us that the higher growth rates were the results of 
negotiations. However, such negotiations were not always fully 
documented. 

Spending Limit for Florida's Demonstration Not Fully Supported: 

HHS approved a spending limit for Florida's demonstration that exceeded 
the amount HHS could have approved under its benchmark policy, but did 
not fully support the additional spending. Florida's spending limit has 
two primary components, beneficiary services and supplemental payments 
to safety net hospitals. Beneficiary services account for the bulk of 
Medicaid spending and include the medical costs of demonstration 
enrollees. For beneficiary services, HHS established annual per person 
limits on federal funds for medical services to groups of 
beneficiaries.[Footnote 35] Projected over the 5 years of the 
demonstration, these per person limits would result in estimated 
Medicaid spending of about $47.6 billion. However, under HHS's 
benchmark policy of limiting projected growth to the lower of a state's 
recent historical experience or nationwide estimates of Medicaid 
growth, the maximum spending allowed would have been a projected $38.6 
billion, or about $9 billion less. HHS supported deviations from its 
benchmarks that would allow spending projected at an estimated $40.7 
billion--$2.1 billion above the level the benchmarks would have 
allowed, but still $6.9 billion less than what it approved. 

Table 1 shows the spending limit originally proposed by Florida and 
agreed to by HHS. It also shows the spending limit that would have 
resulted using the benchmark growth rates of the lower of the state's 
historical growth or the growth projected in the President's Budget, as 
well as the estimated spending limit that HHS and state officials 
supported through explanations and documentation. 

Table 1: Spending Limit for Florida's Medicaid Demonstration as 
Proposed, Approved, and Calculated under HHS's Benchmark Policy and 
Supported by HHS's Explanations: 

Beneficiary services[A]; 
Spending limit as proposed and approved: Proposed by Florida: $47.1; 
Spending limit as proposed and approved: Approved by HHS: $47.6; 
GAO estimate of spending limit calculated using HHS benchmarks and as 
supported by HHS explanations: Spending limit using HHS benchmarks: 
$38.6; 
GAO estimate of spending limit calculated using HHS benchmarks and as 
supported by HHS explanations: Spending limit supported by HHS 
explanations: $40.7. 

Supplemental payments; 
Spending limit as proposed and approved: Proposed by Florida: 5.9; 
Spending limit as proposed and approved: Approved by HHS: 5.0; 
GAO estimate of spending limit calculated using HHS benchmarks and as 
supported by HHS explanations: Spending limit using HHS benchmarks: 
5.0[B]; 
GAO estimate of spending limit calculated using HHS benchmarks and as 
supported by HHS explanations: Spending limit supported by HHS 
explanations: 5.0. 

Total; 
Spending limit as proposed and approved: Proposed by Florida: $53.0; 
Spending limit as proposed and approved: Approved by HHS: $52.6; 
GAO estimate of spending limit calculated using HHS benchmarks and as 
supported by HHS explanations: Spending limit using HHS benchmarks: 
$43.6; 
GAO estimate of spending limit calculated using HHS benchmarks and as 
supported by HHS explanations: Spending limit supported by HHS 
explanations: $45.7. 

Source: GAO analysis of data from HHS and Florida. 

[A] Amounts are based on applying annual per person allowed spending to 
expected Medicaid enrollment for two different eligibility groups: (1) 
aged, blind, and disabled beneficiaries and (2) families and children. 
Amounts are for total federal and state spending. The federal 
government matched Florida expenditures at a rate of 58.76 percent in 
2007. 

[B] HHS budget neutrality policy does not establish benchmarks for 
estimating growth of supplemental payment arrangements such as the one 
Florida proposed for its demonstration. The $5.0 billion projected 
amount for supplemental payments that HHS approved is consistent with 
the projected amount using the state's reported growth. A further 
discussion of the basis for HHS's approval of this amount is found in 
the next section of this report. 

[End of table] 

A further discussion of HHS's explanations of its approvals follows. 

Higher Spending Limit in Florida Is Based on Assumptions of Cost Growth 
That Exceed Benchmarks: 

Projected spending on medical services to beneficiary groups in Florida 
is based on assumptions of per person cost and beneficiary enrollment 
growth rates for two primary groups of Medicaid beneficiaries: (1) the 
aged, blind, and disabled and (2) children and families. Florida 
submitted to HHS 5 years of historical data and calculations of cost 
growth rates over this period for the two groups. HHS, in turn, 
compared Florida's state-specific history to estimates of Medicaid 
growth for these beneficiary groups nationwide. However, neither 
Florida's proposed nor HHS's approved spending limit is based on 
projected spending using the lower of the state-specific or nationwide 
benchmarks consistent with HHS policy. Instead, according to HHS and 
state officials, Florida proposed--and HHS approved--higher per person 
cost growth based on adjustments to the growth rates that were made 
during negotiations between HHS and state officials that were not 
documented. 

The cost growth rates HHS accepted in negotiations were substantially 
above those that would have been allowed under HHS's benchmark 
policy.[Footnote 36] As table 2 shows, HHS approved cost growth rates 
of 8 percent for both aged, blind, and disabled beneficiaries, and for 
families and children--lower than Florida was proposing for the first 
group, and slightly higher than Florida was proposing for the second 
group.[Footnote 37] Under HHS's benchmark policy, which calls for 
basing spending limits on projections of growth at the lower of state- 
specific history or estimates of Medicaid growth nationwide, the 
approved cost growth rates would have been 4.80 percent for aged, 
blind, and disabled beneficiaries and 3.11 percent for children and 
families. 

Table 2: Comparison of Florida's Per Person Growth Rates as Proposed 
and Approved under HHS's Benchmark Policy and Supported by HHS's 
Explanations: 

Rates in percentages. 

Rates in percentages: Aged, blind, and disabled; 
Cost growth rates as proposed and approved: Proposed by Florida: 8.73; 
Cost growth rates as proposed and approved: Approved by HHS: 8.00; 
HHS benchmark rates and rates supported by explanations: HHS 
benchmark[A]: 4.80 (N); 
HHS benchmark rates and rates supported by explanations: Rate as 
supported by HHS explanations: 6.45. 

Children and families; 
Cost growth rates as roposed and approved: Proposed by Florida: 7.96; 
Cost growth rates as proposed and approved: Approved by HHS: 8.00; 
HHS benchmark rates and rates supported by explanations: HHS 
benchmark[A]: 3.11 (S); 
HHS benchmark rates and rates supported by explanations: Rate as 
supported by HHS explanations: 3.77. 

Source: GAO analysis of data from HHS and Florida. 

[A] Lower of state historical spending (S) or estimates of Medicaid 
growth nationwide (N). 

[End of table] 

Part, but Not All, of Increase above HHS's Benchmarks Supported: 

HHS officials allow states to use higher cost growth rates if state 
officials can establish that state-specific or nationwide data do not 
accurately depict expected growth in the state Medicaid program. HHS's 
policy indicates that states are to provide quantified explanations of 
growth rate anomalies. For the Florida demonstration, HHS officials 
explained and provided public or internal documents to support part, 
but not all, of the increases to benchmark cost growth rates that HHS 
approved. 

1. For aged, blind, and disabled beneficiaries, HHS allowed adjustments 
to nationwide estimates of cost growth to account for effects of the 
Medicare prescription drug benefit, but quantified explanations for 
only part of the approved increase. HHS officials explained that the 
implementation of the Medicare Part D prescription drug benefit, which 
would have the effect of shifting the cost of many prescription drugs 
out of the Medicaid program and into Medicare, caused a sharp decrease 
in estimated costs for aged, blind, and disabled Medicaid beneficiaries 
nationwide in 2006. This decrease in the cost growth rate for that 
unusual year lowered the nationwide benchmark growth rate for these 
beneficiaries. HHS officials adjusted for the effects of the Medicare 
prescription drug benefit by removing drug-related expenditures from 
the nationwide estimates over the time period under review and 
recalculating estimated nationwide cost growth. As a result of these 
adjustments, HHS provided quantified explanations to the deviation from 
its benchmark and supported projected cost growth of 6.45 percent for 
aged, blind, and disabled beneficiaries--higher than the 4.80 percent 
allowed under its benchmark policy. HHS officials indicated that the 
remaining deviation from the benchmark was attributable to adjustments 
they made to the growth rate to account for an expected increase in 
enrollment of low-cost beneficiaries.[Footnote 38] HHS officials, 
however, did not identify and correct anomalies in the nationwide 
enrollment data to support the additional increase in cost growth from 
6.45 percent to the 8 percent HHS approved for the demonstration. 

2. For families and children, HHS claimed that higher growth over a 
selected time period more accurately reflected cost growth. According 
to HHS's guidance, state-specific growth rates are based on 5 years of 
historical data.[Footnote 39] HHS, however, allowed Florida to 
calculate a cost growth rate for its families and children based on 
data from a truncated period of higher growth of 3 years and 9 months. 
Cost growth over this shortened period of time was 5.88 percent, as 
compared to cost growth of 3.11 percent when using data from the full 5 
years leading up to Florida's base year. HHS officials explained that 
they allowed Florida to selectively use the higher years of data to 
calculate its growth rates for two reasons: (1) that the shortened time 
period replaced earlier years of unusually low cost growth with more 
recent data and (2) that HHS had recently approved a higher cost growth 
rate for a subset of this particular beneficiary group in renewing an 
ongoing Medicaid managed care demonstration. HHS officials did not, 
however, identify and correct an anomaly in the state's earlier data, 
nor did HHS document and explain why the state was allowed to establish 
its spending limits using growth rates that were based on anticipated 
higher growth under the demonstration. In particular, the state, in its 
application, stated that it anticipated higher cost growth under the 
demonstration due to greater use of managed care by Medicaid children 
and families. A review of Florida's historical cost growth that 
includes more recent data cited by HHS officials, and that uses data 
from a 5-year period as indicated by HHS policy, supports a cost growth 
rate of 3.77 percent. 

HHS officials maintained that HHS's methods for ensuring budget 
neutrality are valid and indicated that the department's budget 
neutrality decisions are, to some extent, the product of negotiations. 
HHS officials also said that HHS can assign growth rates that vary from 
the results of analysis of historical data and the President's Budget 
projections if officials are convinced that the trends are merited. 

Spending Limit for Vermont's Demonstration Also Not Fully Supported: 

To a lesser degree than Florida, HHS approved a spending limit for 
Vermont's demonstration based on assumptions of beneficiary enrollment 
growth that exceeded HHS's benchmarks. HHS approved a spending limit of 
$4.7 billion for Vermont's 5-year demonstration. This spending limit, 
however, is $180 million above the maximum supported by HHS and the 
state in explanations and documentation. HHS also allowed Vermont to 
include in its spending limit funds that were "hypothetical," that is, 
$67 million in funds that had been approved as budget neutral for a 
prior section 1115 demonstration but that the state had not actually 
spent. 

Table 3 shows the spending limit as originally proposed by Vermont and 
as agreed to by HHS. It also shows the limit that would apply if HHS 
benchmarks had been used and the limit that HHS and state officials 
explained and supported in documentation. Although HHS reduced 
Vermont's proposed spending limit by over $1.4 billion, this reduction 
resulted from an agreement that the state not include the financing of 
three major programs in the demonstration. These three programs that 
were removed from the state's initial proposal--a long-term care 
demonstration, payments to hospitals under the Disproportionate Share 
Hospital program, and SCHIP--account for most of the $1.4 billion and 
will continue to be operated and reimbursed apart from the Vermont 
demonstration, thus having no affect on the budget neutrality of the 
demonstration.[Footnote 40] 

Table 3: Spending Limit for Vermont's Medicaid Demonstration as 
Proposed, Approved, and Calculated under HHS's Benchmark Policy and 
Supported by HHS's Explanations: 

Dollars in billions. 

Beneficiary services[A]; 
Spending limit as proposed and approved: Proposed by Vermont: $6.0[B]; 
Spending limit as proposed and approved: Approved by HHS: $4.2; 
GAO estimate of spending limit calculated using HHS benchmarks and as 
supported by HHS explanations: Spending limit using HHS benchmarks: 
$3.9; 
GAO estimate of spending limit calculated using HHS benchmarks and as 
supported by HHS explanations: Spending limit supported by HHS 
explanations: $4.0. 

Other programs; 
Spending limit as proposed and approved: Proposed by Vermont: 0.2[C]; 
Spending limit as proposed and approved: Approved by HHS: 0.5[D]; 
GAO estimate of spending limit calculated using HHS benchmarks and as 
supported by HHS explanations: Spending limit using HHS benchmarks: 
0.5[E]; 
GAO estimate of spending limit calculated using HHS benchmarks and as 
supported by HHS explanations: Spending limit supported by HHS 
explanations: 0.5. 

Total; 
Spending limit as proposed and approved: Proposed by Vermont: $6.2; 
Spending limit as proposed and approved: Approved by HHS: $4.7[F]; 
GAO estimate of spending limit calculated using HHS benchmarks and as 
supported by HHS explanations: Spending limit using HHS benchmarks: 
$4.4; 
GAO estimate of spending limit calculated using HHS benchmarks and as 
supported by HHS explanations: Spending limit supported by HHS 
explanations: $4.5. 

Source: GAO analysis of data from HHS and Vermont. 

[A] Amounts for beneficiary services as approved were based on 
projecting cost and enrollment growth for four different eligibility 
groups: (1) aged, blind, and disabled beneficiaries; (2) children and 
families; (3) optional populations from an ongoing section 1115 
demonstration; and (4) beneficiaries from a separate developmental 
services demonstration. Amounts are for total federal and state 
spending. The federal government matched Vermont expenditures at a rate 
of 58.93 percent for fiscal year 2007. 

[B] Includes Vermont's SCHIP ($21 million) and costs associated with 
beneficiaries in a long-term care demonstration ($1.2 billion). 

[C] Disproportionate Share Hospital payments ($183 million). 

[D] Includes costs to administer Medicaid programs ($405 million), 
unspent funds from an ongoing demonstration ($67 million), and funds to 
replace the services of the Vermont State Hospital ($54 million). 

[E] HHS's budget neutrality policy does not establish benchmarks for 
estimating growth of administrative costs, for carrying forward 
surpluses from one demonstration to another, or for funding replacement 
services for state hospitals. 

[F] The $1.4 billion in projected costs for SCHIP, disproportionate 
share hospital payments, and a long-term care demonstration was omitted 
from the original proposal, but Vermont will continue to receive 
federal funding for these programs independent of the demonstration. 

[End of table] 

Higher Spending Limits in Vermont Are Based in Part on Assumptions 
about Beneficiary Enrollment Growth That Exceeded Benchmarks: 

Vermont submitted 5 years of historical cost and beneficiary enrollment 
data for assessment by HHS.[Footnote 41] For per person costs, HHS 
required Vermont to hold growth rates in line with the department's 
benchmark policy. For beneficiary enrollment, however, HHS approved 
growth rates that were higher than benchmark levels. For example, the 
benchmark rate for enrollment growth of aged, blind, and disabled 
beneficiaries was 1.52 percent per year; HHS approved a rate of 2.52 
percent. For the largest group of beneficiaries under the 
demonstration--families and children--HHS approved an enrollment growth 
rate of 1.99 percent, higher than the 1.05 percent nationwide benchmark 
for this group (see table 4). 

Table 4: Comparison of Vermont's Beneficiary Enrollment Growth Rates as 
Proposed and Approved under HHS's Benchmark Policy and Supported by 
HHS's Explanations: 

Rates in percentages. 

Aged, blind, and disabled beneficiaries[B]; 
Enrollment growth rates as proposed and approved: Proposed by Vermont: 
2.52; 
Enrollment growth rates as proposed and approved: Approved by HHS: 
2.52; 
HHS benchmark rates and rates as supported by explanations: HHS 
benchmark[A]: 1.52 (S); 
HHS benchmark rates and rates as supported by explanations: Rate as 
supported by HHS explanations: 1.52. 

Children and families; 
Enrollment growth rates as proposed and approved: Proposed by Vermont: 
1.99; 
Enrollment growth rates as proposed and approved: Approved by HHS: 
1.99; 
HHS benchmark rates and rates as supported by explanations: HHS 
benchmark[A]: 1.05 (N); 
HHS benchmark rates and rates as supported by explanations: Rate as 
supported by HHS explanations: 1.05. 

Certain populations from an ongoing section 1115 demonstration; 
Enrollment growth rates as proposed and approved: Proposed by Vermont: 
6.43; 
Enrollment growth rates as proposed and approved: Approved by HHS: 
6.43; 
HHS benchmark rates and rates as supported by explanations: HHS 
benchmark[A]: 1.10 (N); 
HHS benchmark rates and rates as supported by explanations: Rate as 
supported by HHS explanations: 4.93. 

Beneficiaries from an ongoing developmental services demonstration; 
Enrollment growth rates as proposed and approved: Proposed by Vermont: 
6.00; 
Enrollment growth rates as proposed and approved: Approved by HHS: 
6.00; 
HHS benchmark rates and rates as supported by explanations: HHS 
benchmark[A]: 1.52 (S); 
HHS benchmark rates and rates as supported by explanations: Rate as 
supported by HHS explanations: 4.25. 

Source: GAO analysis of data from Vermont and HHS. 

[A] Lower of state historical beneficiary enrollment (S) or estimates 
of Medicaid growth nationwide (N). 

[B] This beneficiary group consists of several small beneficiary 
groups: two main subgroups are those receiving services through 
Vermont's state Medicaid plan and those receiving services through an 
ongoing section 1115 demonstration. HHS approved a 3.52 percent 
enrollment growth rate for the former and a 2.52 percent growth rate 
for the latter. 

[End of table] 

Part, but Not All, of Increase Allowed above Benchmarks Supported: 

Similar to Florida, but to a lesser degree, a portion of Vermont's 
spending limit that exceeds HHS benchmarks was not supported. HHS used 
beneficiary cost and enrollment growth rates to provide the basis for 
Vermont's aggregate spending limit. Although HHS held Vermont's 
beneficiary cost growth rates to the lower of HHS's benchmarks, HHS did 
not do so with regard to enrollment growth rates. HHS approved 
enrollment growth rates that were proposed by Vermont but were higher 
than the benchmarks without adequate documentation for the higher 
growth rate (see table 4). HHS officials told us the higher growth 
rates were the results of negotiations. However, such negotiations were 
not well documented. For one group of beneficiaries--families and 
children--HHS approved an enrollment growth rate of 1.99 percent, 
consistent with the state's historical growth but almost twice the 1.05 
percent nationwide benchmark for this group, without explanation. For 
another group--beneficiaries from an ongoing developmental services 
demonstration--HHS allowed the state to exceed the benchmark growth 
rate based in part on a state management plan to cover more people in 
the future by removing them from a waiting list for developmental 
services. For this same group, Vermont supported part, but not all, of 
the spending limit in excess of the benchmarks by presenting a more 
narrowly focused analysis of its historical enrollment data. 
Specifically, for this subset of the state's aged, blind, and disabled 
beneficiary group, Vermont officials identified enrollment growth of 
4.25 percent, higher than the benchmark level for aged, blind, and 
disabled beneficiaries. HHS officials told us that they considered both 
state-specific and nationwide benchmarks before approving Vermont's 
requested enrollment growth rates. We estimate that the higher-than- 
benchmark enrollment growth rates approved by HHS that were not 
supported by explanation and documentation increased Vermont's spending 
ceiling by about $180 million.[Footnote 42] 

HHS also allowed Vermont to include in its spending limit the projected 
costs for "hypothetical" expenditures, that is, expenditures the state 
could have made but did not make. Specifically, HHS allowed Vermont to 
include in its spending limit nearly $67 million that the state was 
authorized to have spent under an ongoing Medicaid section 1115 
demonstration, but that was unspent under the program. At the time of 
its proposal, Vermont had an ongoing 1115 demonstration called the 
Vermont Health Access Program that began in 1996 and was later 
extended. While program expenditures for the Vermont Health Access 
Program were well under that demonstration's spending limit in the 
early years of the demonstration, in 2004 the program began operating 
at a deficit as expenditures exceeded its annual spending targets (see 
fig. 2). Vermont ended this demonstration early because the state could 
no longer afford to incur the deficits. Because of its early 
completion, $67 million under the spending limit for the demonstration 
was unspent. 

Figure 2: Vermont Health Access Program Surpluses and Deficits: 

This figure is a combination line graph showing Vermont Health Access 
Program surpluses and deficits. The X axis represents the year, and the 
Y axis represents the dollars in millions. 

[See PDF for image] 

Source: GAO analysis of information from Vermont. 

[End of figure] 

HHS policy requires a state to capture actual expenditures in its 
spending base, and the $67 million allowed was a hypothetical 
expenditure that did not represent true expenditures of the state under 
its program. We have previously reported a concern about HHS's allowing 
states to include hypothetical costs in their spending limits. For 
example, in 2002, we reported that HHS had approved an inflated 
spending limit for one state by allowing the state to include projected 
costs of covering a population that the state had not actually covered 
under its program.[Footnote 43] 

HHS Has Not Ensured That Demonstrations in Florida and Vermont Maintain 
the Fiscal Integrity of the Medicaid Program: 

HHS has not ensured that demonstrations in Florida and Vermont maintain 
the fiscal integrity of the Medicaid program. In Florida, HHS approved 
the state's use of spending from a problematic supplemental payment 
arrangement that the state had in place prior to the demonstration as 
the basis for allowed spending under the demonstration, without 
correcting all identified problems. A 2005 HHS financial management 
review found several problems with the earlier financing arrangement 
that involved supplemental payments to certain hospitals and other 
health care providers. Among the problems, the HHS review found that 
Florida had incorrectly calculated the level of supplemental payments 
for which federal Medicaid funds could be obtained, resulting in 
inflated payments under the arrangement. Without taking corrective 
action, HHS allowed Florida to use the prior financing arrangement as 
the basis for allowed spending in a $1 billion per year low-income pool 
under its current demonstration. We had, in 2004, recommended that HHS 
establish methods for states' calculations of supplemental payments, 
but HHS has not implemented this recommendation. In Vermont, HHS 
allowed the state to operate a managed care organization and, through 
this arrangement, retain excess revenue from payments to the 
organization for previously state-funded programs. In July 2007, we 
raised concerns about this demonstration's consistency with federal 
law. We recommended that the Secretary reexamine Vermont's 
demonstration and, where appropriate, either modify the terms of the 
demonstration or seek statutory authorization for the state to continue 
the demonstration in its present form. 

Florida Allowed to Use Spending under a Problematic Supplemental 
Payment Arrangement as the Basis for Spending under the Demonstration: 

A key component of Florida's demonstration is a pool of federal, state, 
and local money to supplement payments to the state, or to hospitals, 
clinics, or other providers (see table 5). Florida agreed to 
discontinue its supplemental payment program under the terms of the 
demonstration, but to ensure continued funding for providers that had 
been receiving supplemental payments under the former program, 
requested HHS approval to make supplemental payments through a low- 
income pool. HHS approved a $5 billion low-income pool that allows 
Florida to spend $1 billion per year for the 5 years of the 
demonstration for uncompensated medical care costs to the uninsured and 
underinsured, Medicaid costs above standard Medicaid reimbursement 
rates, health insurance premiums, and insurance products for such 
services provided to otherwise uninsured individuals.[Footnote 44] 

Table 5: HHS-Approved Distribution of Florida's Annual Low-Income Pool 
of Federal, State, and Local Funds, Demonstration Year One: 

Dollars in millions. 

Categories of eligible provider access systems: Hospitals that received 
supplemental payments under the former UPL program; 
Approved distribution of annual low-income pool funds[A]: $141.1; 
Description: Hospitals that serve a significant portion of Florida's 
Medicaid, uninsured, and underinsured population, including those in 
rural areas, with emergency services, inpatient hospital care, 
specialty pediatric care, and primary care. 

Categories of eligible provider access systems: Public, non-state-owned 
hospitals; 
Approved distribution of annual low- income pool funds[A]: 578.0; 
Description: Funds are distributed among four tiers of public hospitals 
depending on whether they receive local tax support and how much 
service they provide to Medicaid beneficiaries and those who lack 
adequate health insurance. 

Categories of eligible provider access systems: Health care providers 
in communities in which the local government provides more than $1 
million to support care for individuals who lack adequate health care 
coverage; 
Approved distribution of annual low- income pool funds[A]: 180.0; 
Description: The local community for providers in this category must 
provide more than $1 million in financial support for hospitals within 
its boundaries to fund care for the uninsured and underinsured. 

Categories of eligible provider access systems: Hospitals that do not 
receive local government support for the uninsured or underinsured or 
whose local governments provide $1 million or less to support care for 
individuals who lack adequate health care coverage; 
Approved distribution of annual low-income pool funds[A]: 80.5; 
Description: These health care providers must devote at least 10 
percent of their care to Medicaid patients and those who lack adequate 
health insurance. 

Categories of eligible provider access systems: Hospitals that operate 
poison control programs; 
Approved distribution of annual low-income pool funds[A]: 3.2; 
Description: Regional poison control centers affiliated with accredited 
medical schools or colleges of pharmacy in Tampa, Jacksonville, and 
Miami, as well as a data center in Jacksonville. 

Categories of eligible provider access systems: Federal Qualified 
Health Centers; 
Approved distribution of annual low- income pool funds[A]: 15.3; 
Description: The state proposed distributing $7.3 million to centers 
that qualify for state funds, but did not determine how to distribute 
the other $8.0 million prior to HHS approval. 

Categories of eligible provider access systems: County health 
initiatives to expand primary care services; 
Approved distribution of annual low-income pool funds[A]: 2.0; 
Description: Funds to expand primary care services in rural areas to 
Medicaid beneficiaries, underinsured, and other low-income uninsured 
individuals who do not qualify for Medicaid. 

Source: GAO analysis of information from Florida. 

[A] Distributions from the low-income pool are subject to authorization 
by the Florida State Legislature. 

[End of table] 

At the time the demonstration was approved, however, HHS also had 
indications that Florida made excessive supplemental payments through 
the existing supplemental payment arrangement, and these concerns had 
not been resolved as of the time the demonstration was approved. A 
September 2005 HHS review of Florida's financing arrangements[Footnote 
45] found that the methods and data used to calculate the amount of 
supplemental payments eligible for federal matching funds were 
unreliable. The reviewer, for example, found that Florida established 
Medicaid UPLs--which cannot exceed what Medicare would pay for the same 
services, and which determine the maximum amount of federal matching 
funds the state could obtain for its supplemental payment program-- 
without making adjustments to account for the fact that Medicare 
beneficiaries are typically older and more expensive to treat. For 
example, the review found that the state's estimate of what Medicare 
would pay for hospital services was nearly three times what Medicaid 
would typically pay, which the reviewer questioned. Not adjusting for 
the higher cost of treating Medicare patients inflates the state's 
calculation of allowable payments under the program. HHS's review also 
found that the data used to calculate the supplemental payment levels 
under Medicaid's UPL contained errors and did not provide a reliable 
basis for determining the appropriate payment levels. HHS's 2005 review 
did not estimate the actual allowable payments under the program or the 
extent that the prior supplemental payment arrangement was considered 
excessive or inflated. 

HHS required the state to correct one issue the review had identified 
with the source of the state's own funding for the supplemental 
payments it was making as a condition of approving the demonstration. 
However, HHS did not require Florida to address the problems with the 
methodology and data used to determine the amount of supplemental 
payments eligible for federal matching funds before projecting allowed 
spending under the demonstration. HHS's required terms of the 
demonstration, however, did allow for future adjustments to the 
spending limit under certain circumstances.[Footnote 46] In November 
2006, HHS officials said that problems identified with the improper 
calculation of states' allowed supplemental payment amounts would be 
corrected at a later date. 

Vermont Allowed to Use Medicaid Funds to Supplant State Funding for 
Certain Purposes: 

Under Vermont's demonstration, HHS authorized the state to operate its 
own managed care organization and, through this arrangement, to apply 
federal Medicaid matching funds to programs that were previously funded 
by the state and that do not exclusively benefit those eligible for 
Medicaid. Under this approach, the state's Medicaid agency--the Agency 
of Human Services--makes actuarially certified monthly lump-sum 
payments to one of its own offices. That office, the Office of Vermont 
Health Access, serves as the managed care organization for the Medicaid 
program. In state fiscal year 2006, for example, the state Medicaid 
agency made lump-sum payments to its Office of Vermont Health Access of 
$65.4 million per month. The Agency of Human Services, in turn, 
receives federal Medicaid matching funds on these monthly payments. If 
Vermont can operate its public managed care organization and provide 
services to Medicaid beneficiaries for less than $65.4 million per 
month, HHS allows the state, under the demonstration, to spend excess 
revenues on programs that meet any of four broad health care 
objectives: (1) increase health insurance coverage, (2) increase access 
to quality health care for Medicaid enrollees and those lacking 
adequate insurance, (3) improve health outcomes and quality of life for 
Medicaid-eligible individuals, and, (4) encourage public-private health 
care partnerships. 

In fiscal year 2006--the first full year of the demonstration--the 
Vermont-operated managed care organization generated $56.5 million in 
excess revenues and invested $43 million of the funds into various 
programs. HHS allowed Vermont to invest the remainder of the excess 
revenues generated by the managed care organization in a reserve fund 
for future use. Over the 5 years of the demonstration, Vermont 
estimates that it will accumulate $300 million in excess revenues. The 
state plans to use these excess revenues to supplant state funding for 
a number of programs that do not exclusively benefit Medicaid-eligible 
individuals (see table 6). For example, Vermont plans to use excess 
revenues from the demonstration to fund a grant for the University of 
Vermont and to provide loan forgiveness for doctors and dentists. 
Vermont officials indicated that state funds--freed up by investment of 
excess revenues from the demonstration--could then be used to reduce 
Vermont's budgetary constraints, projected in the demonstration's 
proposal as a $656.8 million 5-year shortfall in state funds to pay for 
the state's own share of Medicaid expenditures. 

Table 6: Examples of Vermont's Use of Excess Medicaid Revenues under 
Its Demonstration: 

Dollars in millions. 

State-funded health care program: Residential care for youth/substitute 
care; 
Examples of state fiscal year 2007 investments in health-related 
programs: $10.54; 
Description: Funds for residential care for youth in need of intensive 
behavioral health services. 

State-funded health care program: Mental health programs; 
Examples of state fiscal year 2007 investments in health- related 
programs: 8.25; 
Description: Funds to support access to mental health care and 
treatment services for children and adults. 

State-funded health care program: Department of Education school heath 
services[A]; 
Examples of state fiscal year 2007 investments in health-related 
programs: 6.40; 
Description: School health services include the professional services 
of nurses, occupational therapists, physical therapists, mental health 
counselors, certified mental health workers, psychologists, personal 
care aides, and other medical professionals. 

State-funded health care program: University of Vermont-Vermont 
physician training; 
Examples of state fiscal year 2007 investments in health-related 
programs: 3.87; 
Description: A grant to train medical professionals. Funding is used to 
support training of medical professionals and provide services to 
Medicaid-eligible, uninsured, and underinsured Vermonters. 

State-funded health care program: Department of Corrections programs; 
Examples of state fiscal year 2007 investments in health-related 
programs: 2.95; 
Description: Funds to promote community- based and residential 
treatment services for former inmates. 

State-funded health care program: Substance abuse treatment; 
Examples of state fiscal year 2007 investments in health- related 
programs: 2.80; 
Description: Funds for a program providing treatment services for 
individuals who lack health care coverage. Substance abuse treatment 
includes outpatient, intensive outpatient, residential, detoxification, 
and pharmacological treatment services. 

State-funded health care program: Aid to Aged, Blind, and Disabled 
Community Care Level III; 
Examples of state fiscal year 2007 investments in health-related 
programs: 2.62; 
Description: Funds to support payments to community care level III 
aged, blind, and disabled recipients. By law, eligible recipients 
receive a subsistence amount compatible with decency and health 
standards. 

State-funded health care program: Blueprint for Health Program; 
Examples of state fiscal year 2007 investments in health-related 
programs: 1.98; 
Description: Funds for a statewide program intended to advance 
innovative solutions and provide support to help doctors and patients 
effectively manage chronic disease. 

State-funded health care program: Health laboratory; 
Examples of state fiscal year 2007 investments in health- related 
programs: 1.91; 
Description: Funds to cover the nonfederal costs of running the public 
health laboratory, which identifies disease-causing agents in specimens 
from human, animal, and environmental sources. 

State-funded health care program: Tobacco Cessation Program; 
Examples of state fiscal year 2007 investments in health-related 
programs: 1.65; 
Description: Funds to reduce the use of tobacco among Vermonters, with 
an emphasis on discouraging young people from starting to smoke. 

State-funded health care program: WIC Coverage- Special Supplemental 
Nutrition Program for Women, Infants, and Children; 
Examples of state fiscal year 2007 investments in health- related 
programs: 1.17; 
Description: Funds to improve health by informing families about good 
health practices and by providing nutritious foods to eligible 
recipients. The program offers women, infants, and children health 
screenings and nutrition and health education. 

State-funded health care program: Flexible Family/ Respite Funding; 
Examples of state fiscal year 2007 investments in health-related 
programs: 1.14; 
Description: Funds to support eligible families with children or adult 
family members with developmental disabilities, to enhance their 
ability to live together. 

Source: GAO analysis of information from Vermont. 

Note: In state fiscal year 2007, Vermont also used excess revenues to 
invest in other health-related programs, including the state's Health 
Care Authority, Veterans Home, Essential Person Program, and Civil 
Union. In state fiscal year 2007, excess revenue investments totaled 
more than $46.5 million in these and other health-related programs. 
This total does not include 2007 spending on Department of Education 
school health services, which was pending as of November 2007. 

[A] Vermont officials indicated that spending on Department of 
Education school health services represents state fiscal year 2006 
expenditures. State fiscal year 2007 expenditures for this program were 
pending as of November 2007. 

[End of table] 

The Vermont arrangement generates excess administrative reimbursement 
in two ways. First, HHS allowed the Agency of Human Services to pay its 
Office of Vermont Health Access at a rate that, while typical for 
private managed care organizations, is higher than the rate Vermont had 
been paid prior to the demonstration and higher than average Medicaid 
agency administrative costs. Second, because the payments for which 
matching funds are provided are lump-sum payments that include the 
managed care organization's administrative costs, HHS pays a higher 
portion of the administrative costs associated with the managed care 
organization than it pays for administrative costs in the rest of 
Vermont's Medicaid program or in Medicaid programs in other states. At 
the state's historical rate and in a proportion consistent with other 
states' administrative costs, Vermont would have received an estimated 
$71 million less.[Footnote 47] 

The reimbursement attributable to administrative costs could help 
ensure that the state has excess revenues for the state's purposes, 
including supplanting state funding for non-Medicaid programs. A 
September 2005 independent review and risk analysis conducted by a 
consultant to Vermont concluded that the likelihood that there would be 
savings under the demonstration available to be used for programs 
formerly funded with state dollars was very high for two reasons. 
First, the spending limit and corresponding premium structure of the 
managed care organization assumed a 9 percent administrative cost 
component, which is typical for private managed care organizations but 
nearly double the average state Medicaid agency administrative costs. 
According to the consultant, such costs typically run in the 3 to 5 
percent range. According to HHS data, administrative costs averaged 4.6 
percent nationwide in fiscal year 2005. Second, unlike the situation 
where the state contracts with a private managed care organization, 
there is an incentive to pay the state-operated managed care 
organization on the high end of the actuarial range approved for the 
managed care premium because any excess payments can be used for state- 
funded programs.[Footnote 48] 

This financing arrangement allows Vermont to increase federal Medicaid 
payments to the state without a commensurate increase in state Medicaid 
spending. The state agency, by making a payment to itself in excess of 
the cost of providing Medicaid services, generates federal matching 
funds, which can be used to supplant state spending on certain 
programs. This supplanted state money, in turn, can be used to reduce 
Vermont's projected $656.8 million 5-year shortfall in state funds for 
Medicaid, thus generating even more federal matching funds in a process 
known as recycling. Curtailing practices that allow states to reduce 
the proportion of Medicaid spending for which they are responsible has 
been part of the ongoing congressional scrutiny of Medicaid 
programs.[Footnote 49] 

In a letter to the Secretary of HHS, we raised concerns about the 
Vermont program's consistency with federal law. These concerns stemmed 
from HHS's decision to allow the state to operate its own managed care 
organization and, through this arrangement, to apply federal Medicaid 
matching funds to programs previously funded by the state. The approval 
of the Vermont program raised the question whether the Vermont Medicaid 
agency could enter into a managed care contract with one of its own 
offices and receive federal matching funds for lump-sum payments to 
that office rather than for payments based on actual costs. The letter 
also noted that in connection with its managed care regulations, HHS 
has expressed concerns about states obtaining federal matching funds 
through managed care contracts for state-funded services for which such 
funds would not ordinarily be available. Given our concerns, we 
recommended that the Secretary of HHS reexamine the demonstration and, 
where appropriate, either modify its terms or seek statutory 
authorization for it to continue in its current form.[Footnote 50] 

Conclusions: 

After examining HHS's approvals of demonstrations in Florida and 
Vermont, our long-standing cost and oversight concerns related to HHS 
approvals of comprehensive Medicaid demonstration proposals remain. In 
determining the budget neutrality of proposed demonstrations, HHS 
approved spending limits for Florida and Vermont that exceeded its own 
benchmarks without adequately supporting the basis for the deviations. 
Our findings in Florida and Vermont are similar to the concerns we 
raised in our earlier reports--during its budget neutrality process, 
HHS did not adequately support the deviations from benchmark rates that 
it allowed in the development of states' spending limits, or clearly 
document and make public the basis for the approved limits. When 
combined, the spending limits approved for Florida and Vermont are 
nearly $7.2 billion more than what the documentation and explanations 
support for the demonstrations. Given the significant federal 
expenditures for these demonstrations, improved accountability and 
transparency in HHS's budget neutrality process, including in the 
approval of states' spending limits, is warranted. 

HHS's approvals in Florida and Vermont also raise concerns about 
precedents they establish that affect the federal and state partnership 
and fiscal integrity of the Medicaid program. By allowing Florida to 
use spending from a prior supplemental payment arrangement as the basis 
for new spending without correcting known problems, and by allowing 
Vermont to create its own state-run managed care organization and use 
excess revenue to fund other state programs, HHS has not taken the 
steps needed to ensure that Medicaid funds are used for Medicaid 
purposes. HHS has not corrected the problems it found with historical 
spending under Florida's supplemental payment arrangement--historical 
spending that was used to set the spending limit under the 
demonstration--and reexamined the level of Florida's spending limit 
accordingly. We believe a related recommendation from our 2004 report 
on the fiscal integrity of state Medicaid supplemental payment 
arrangements remains valid: that the department establish uniform 
guidance to states setting forth acceptable methods for calculating 
supplemental payment arrangements, such as the one that served as the 
basis for Florida's low-income pool. Such guidance could help ensure 
that payments under ongoing supplemental payment arrangements, and any 
related demonstration proposals, are appropriate in the future. HHS 
agreed to implement this recommendation in responding to our 2004 
report, but as of December 2007 had not done so. 

Our concerns about HHS approvals extend beyond those related to costs 
and oversight. The Secretary's approval of the Vermont demonstration 
establishes a precedent for future proposals, but raises legal 
concerns. As of January 2008, HHS had no plans to implement our July 
2007 recommendation to address concerns with the demonstration's 
consistency with federal law. Because HHS disagrees with this 
recommendation--and other recommendations we have made to improve the 
demonstration review process--we are elevating this and other 
recommendations to the Congress for its consideration. 

Matters for Congressional Consideration: 

The Congress should consider requiring increased attention to fiscal 
responsibility in the approval of section 1115 Medicaid demonstrations 
by requiring the Secretary of HHS to improve the demonstration review 
process through steps such as (1) clarifying criteria for reviewing and 
approving states' proposed spending limits, (2) better ensuring that 
valid methods are used to demonstrate budget neutrality, and (3) 
documenting and making public material explaining the basis for any 
approvals. 

The Congress should consider addressing whether demonstrations that 
allow states to operate public managed care organizations and retain 
excess revenue to support programs previously funded by the state-- 
including the Vermont demonstration--are within the scope of the 
Secretary of HHS's authority under section 1115 of the Social Security 
Act. 

Recommendation for Executive Action: 

To help ensure that the Florida demonstration will maintain the fiscal 
integrity of the Medicaid program, we recommend that the Secretary of 
HHS ensure that the level of supplemental payments for which the state 
could have obtained federal Medicaid funds in the absence of the 
proposed demonstration is calculated using appropriate methods and 
accurate data sources, and adjust the approved spending limit 
appropriately. 

Agency and State Comments and Our Evaluation: 

We provided a draft of this report for comment to HHS, Florida, and 
Vermont. All three provided written comments which we summarize and 
evaluate below. The full text of HHS's comments is reprinted in 
appendix II along with our response to certain comments. Florida's and 
Vermont's comments are reprinted in appendixes III and IV, 
respectively. HHS and each state also provided technical comments, 
which we incorporated as appropriate. 

HHS Comments and Our Evaluation: 

In commenting on a draft of this report, HHS strongly disagreed with 
our findings, conclusions, and recommendation, stating that the draft 
report mischaracterized the nature of the approved demonstration 
programs and HHS's budget neutrality policies. We based our 
characterizations of HHS programs and policies on documentation 
obtained from HHS and states and interviews with HHS and state 
officials; we believe we have captured and reported them accurately. In 
its comments, HHS also said that our analysis did not adequately 
account for the likelihood of differences in professional 
interpretation in quantifiable analyses. HHS emphasized that the 
demonstrations are approved at the discretion of the Secretary of HHS 
and that the review of demonstration proposals includes both budgetary 
and programmatic elements. We recognize that the Secretary has some 
discretion in approving demonstrations and in establishing policies and 
processes for doing so. But we believe that to maintain accountability 
and transparency in the Medicaid program, of which section 1115 
demonstrations are a major component, the Secretary has the 
responsibility to approve demonstrations based on clearly articulated 
policies and spending limits that are consistent with these policies. 
In conducting our work and preparing the draft report, we accepted 
HHS's explanations for spending limit amounts that deviated from HHS's 
benchmarks when they were clearly articulated and documented. Our draft 
report acknowledged these explanations in noting that some of the 
deviations from the benchmarks were explained. We did not, however, 
accept estimates when program officials could not clearly articulate 
the reasoning they had used, demonstrate how this reasoning was 
consistent with budget neutrality and fiscal integrity principles, and 
explain how the resulting spending limits were derived. 

HHS Comments Related to Past GAO Reviews on 1115 Demonstration and 
Fiscal Integrity Issues: 

HHS commented that we unnecessarily cite points from prior reviews 
regarding section 1115 demonstrations. We cite our earlier work to 
provide a broader perspective and context for our discussion about 
individual states. We also use our prior work as a basis to highlight 
actions that we have recommended that HHS take and that relate to 
problems we identified in this review, but that HHS has not acted upon. 
We believe it is an important part of our work to underscore recurring 
problems as well as areas where HHS has made significant progress. 

HHS also said that we had not given the agency sufficient credit for 
the steps it has taken to ensure fiscal integrity within the Medicaid 
program, stating that we overlooked and understated the progress HHS 
has made since the early 1990s to curtail improper financing 
arrangements. HHS said that the draft report inappropriately focused on 
our 2004 report that did not address related issues and omitted mention 
of our other relevant reports, including those that had recognized 
HHS's efforts. The reports we have cited were those that focused on 
areas relevant to the scope of this work. We have acknowledged in 
earlier reports that the agency has taken a number of steps in recent 
years to strengthen Medicaid's financial management, but in the 
particular areas of concern here--the demonstration criteria, methods, 
and documentation for agreed-upon spending limits--HHS has chosen not 
to make changes that would better ensure accountability and 
transparency. In 2002 and 2004, we recommended that HHS undertake these 
changes. Because it has not, we now raise these as a matter for 
congressional consideration. 

HHS Comments Related to the Application of Budget Neutrality Policy: 

HHS stated that there are multiple methods of establishing that a 
project is budget neutral and that each agreement must be considered as 
part of a larger picture, and suggested that we inappropriately 
characterized HHS's internal guideline as a "benchmark policy" and then 
criticized HHS for making minor adjustments for real-world factors that 
could affect a state's spending. We presented the information on HHS's 
policy in the draft report as found in written HHS guidelines on its 
Web site in March 2007 and as told to us by HHS officials. As noted in 
the draft report, HHS's policy for reviewing and approving 
demonstration proposals and their spending limits lacks transparency. 
HHS's complete policy should be clearly identifiable, in writing, and 
publicly available. Furthermore, we disagree that adjustments that 
account for billions of dollars in federal spending, without 
documentation and explanation, are of a minor nature. Agreements that 
commit the federal government to reimbursing states tens of billions of 
dollars should be documented and include explanations of the basic 
reasoning behind the final spending limits, including the adjustments 
to benchmarks that have been approved. 

HHS's Comments Related to Approvals of Florida and Vermont 
Demonstrations: 

HHS noted that one of its most significant concerns about the draft was 
that it failed to acknowledge that HHS had capped Medicaid program 
growth in Florida, which had averaged 13 percent in recent years. We 
disagree. As noted in the draft report, HHS approved a per person 
spending limit for Florida's demonstration; however, there is no 
aggregate cap on spending in Florida similar to that in Vermont, where 
HHS placed a cap on total spending. 

HHS also strongly disagreed with our recommendation that it recalculate 
the Florida spending limit using appropriate methods and data sources 
and adjust the spending limit accordingly. HHS indicated that Florida's 
data and methods for calculating payments for its supplemental payment 
program were irrelevant to the development of the Florida 
demonstration. We disagree that Florida's calculations were not 
relevant to the Florida demonstration, since Florida's historical 
payments were used as a basis for the low-income pool spending limit 
under the demonstration, and as a result, the spending limit allows for 
continuation of spending that a HHS review suggests should not have 
been allowed. 

With regard to HHS's approval of the Vermont demonstration, HHS 
disagreed with our concerns and prior recommendation to reexamine the 
terms of the demonstration and, where appropriate, to either modify its 
terms or seek statutory authority for the demonstration to continue in 
its current form. HHS maintained that issues of legal authority were 
adequately and appropriately addressed in the information provided to 
us during the course of our fieldwork. We disagree and note that HHS 
has not addressed the concerns raised in our July 2007 letter. 

HHS also commented that our concern regarding excessive reimbursement 
for administrative expenditures for the public managed care 
organization in Vermont was unwarranted because all demonstration 
revenue must be spent for demonstration purposes and costs matched by 
federal funds would be clearly identified. Our concern remains that the 
broad scope of costs identified as for "demonstration purposes"--for 
example, funding the state public health laboratory--can allow Vermont 
to shift costs to the federal government that were previously funded by 
the state and that do not exclusively benefit individuals eligible for 
Medicaid. 

Comments from Florida and Vermont and Our Evaluation: 

We provided a draft of this report to Florida and Vermont. Florida 
stated that during the negotiations over the demonstration waiver, 
state officials worked closely with HHS to ensure that all data and 
documentation were provided in a timely and accurate manner to support 
the waiver application. Vermont indicated that the state had assumed an 
unprecedented amount of risk related to program expenditures in 
exchange for the flexibility granted by the Secretary and that state 
and federal staff had engaged in extensive discussion and analysis of 
Vermont's historical expenditures, cost and caseload trends, and 
program policies in arriving at the final budget neutrality spending 
limit. Vermont also questioned our finding that HHS agreed to reimburse 
the state's administrative expenditures under the demonstration at a 
rate higher than prior to the demonstration, indicating that an 
independent actuary relied on Vermont's historical administrative 
expenditures in developing this component of the capitation rate. 

We agree that the states provided data and documentation to HHS to show 
the basis for their demonstration proposals. Our concern remains, 
however, with the lack of sufficient documentation showing how the 
final spending limits were derived, particularly since they were 
different from the proposals and were based on assumptions about cost 
and enrollment growth that were higher than HHS's benchmarks. Finally, 
we base our finding that HHS agreed to reimburse Vermont at a rate 
higher than what the state received prior to its demonstration in part 
on our review of the independent actuary's report. 

As arranged with your offices, unless you publicly announce the 
contents of this report earlier, we plan no further distribution until 
30 days after its issuance date. At that time, we will send copies of 
this report to the Secretary of Health and Human Services, the 
Administrator of the Centers for Medicare & Medicaid Services, and 
other interested parties. We will also make copies available to others 
upon request. In addition, the report will be available at no charge on 
the GAO Web site at [hyperlink, http://www.gao.gov]. 

If you or your staff members have any questions, please contact me at 
(202) 512-7114 or kanofm@gao.gov. Contact points for our Offices of 
Congressional Relations and Public Affairs may be found on the last 
page of this report. Major contributors to this report are acknowledged 
in appendix V. 

Signed by: 

Marjorie E. Kanof: 

Managing Director, Health Care Issues: 

[End of section] 

Appendix I: Scope and Methodology: 

Section 1115 of the Social Security Act provides the Secretary of 
Health and Human Services the authority to approve demonstration 
projects that test policy innovations likely to further the objectives 
of certain programs, including Medicaid. Under section 1115, the 
Secretary has authority to waive provisions of the Social Security Act, 
allowing states to operate demonstrations, and to provide federal 
Medicaid matching funds for states' costs that otherwise cannot be 
matched under federal law. 

Section 1115 demonstrations vary in scope, from targeted demonstrations 
limited to specific services or populations, to comprehensive 
demonstrations affecting Medicaid populations and services throughout a 
state and including most of a state's Medicaid expenditures. For 
example, a section 1115 demonstration in Virginia that the Department 
of Health and Human Services (HHS) approved in July 2002 affects 
limited Medicaid services--family planning services--for about 8,300 
beneficiaries. A section 1115 demonstration in New York that HHS 
approved in July 1997, on the other hand, changes the delivery of a 
broad range of Medicaid benefits for over 2.5 million beneficiaries 
from fee-for-service to managed care. 

Our review addressed the budget neutrality[Footnote 51] and fiscal 
integrity of recently approved, comprehensive section 1115 
demonstrations. We selected demonstrations to include in this review 
based on when they were approved and whether they were comprehensive 
and accounted for a major portion of the state's Medicaid program. 
Specifically, we selected demonstrations based on the following: 

1. Approval by HHS from July 2004 (when we last reviewed HHS-approved 
section 1115 demonstrations[Footnote 52]) through December 2006. 

2. Meeting HHS's definition of comprehensive, that is, those that 
affect a broad range of services for Medicaid populations statewide. 

3. The demonstration accounted for greater than 50 percent of the 
state's Medicaid expenditures. 

We used a two-step process to identify demonstrations that met our 
criteria. First, to identify comprehensive section 1115 demonstrations 
approved by HHS from July 2004 through December 2006, we reviewed a 
Centers for Medicare & Medicaid Services (CMS) report that listed all 
section 1115 demonstrations approved through February 2006 and updated 
the list though discussions with agency officials.[Footnote 53] Four 
comprehensive demonstrations met these criteria: the California Medi- 
Cal Hospital Uninsured Care program; the Florida Medicaid Reform 
program; the IowaCare program, and the Vermont Global Commitment to 
Health program. 

Second, to identify which of these four demonstrations met our third 
criterion that expenditures under the demonstration account for a 
majority of state Medicaid spending, we compared estimated first-year 
spending under the demonstration to 2004 total Medicaid spending in 
each state. First-year spending in two of the four states, California 
and Iowa, was less than 5 percent of total 2004 Medicaid spending, so 
we did not include these two states in our study. First-year 
demonstration spending in Florida and Vermont was projected to account 
for 59.9 and 117.2 percent, respectively, of 2004 Medicaid spending, so 
we included the demonstrations in these two states for further review 
in our study. 

To determine the extent to which the Secretary of HHS ensured that 
Medicaid section 1115 demonstrations would be budget neutral to the 
federal government prior to approving them, we reviewed HHS's policies 
for determining budget neutrality as documented on HHS's Web site and 
in information provided by HHS officials. We examined each state's 
projection of the total spending needed to maintain its existing 
Medicaid program in the absence of the proposed demonstrations. 
Specifically, we assessed the extent to which each state's assumptions 
about per person cost and beneficiary enrollment growth conform to 
HHS's policy that these growth rates are the lower of state-specific or 
nationwide benchmarks of Medicaid growth.[Footnote 54] In instances 
where per person and beneficiary enrollment growth rates exceeded the 
lower of these two benchmarks, we asked HHS and state officials for 
explanations and documentation to support the higher growth rates HHS 
approved. We also compared spending limits for the demonstrations-- 
based on the per person cost and beneficiary enrollment growth rates 
HHS approved--against (1) our estimates of demonstration spending 
limits had HHS required Florida and Vermont to have spending limits 
consistent with benchmarks and (2) our estimates of the spending limits 
had HHS held per person cost and beneficiary enrollment growth in each 
state to levels we determined that HHS and state officials had 
explained with quantified support. HHS's policy states that adjustments 
to benchmark growth rates should address anomalies in the underlying 
data. 

To determine the extent to which HHS ensured that the Florida and 
Vermont demonstrations maintain the fiscal integrity of the Medicaid 
federal-state financial partnership, we evaluated HHS's process for 
reviewing section 1115 demonstration proposals and reviewed related 
financial management reports. We interviewed HHS officials from the 
Center for Medicaid and State Operations that has direct oversight 
responsibilities for these demonstrations, including officials from the 
Division of Reimbursement and State Financing who reviewed funding of 
the demonstrations in Florida and Vermont to ensure consistency and 
compliance with federal requirements. We also interviewed state 
officials to gain their understanding of the waiver authorities HHS 
granted each state by approving its demonstration, as well as their 
understanding of the special terms and conditions that govern each 
demonstration. We also relied on the work conducted for an earlier 
study that reviewed the consistency of the Florida and Vermont 
demonstrations with federal law. 

Our findings concerning HHS's approval of these two states' 
demonstrations cannot be generalized to HHS's approval of other states' 
demonstrations. We used the selection criteria discussed above for 
purposes of assessing HHS's process as it was applied in these 
particular cases of importance. We considered these cases to be 
important because they allowed significant changes in the states' 
Medicaid programs and the majority of the states' Medicaid spending was 
governed by the terms of the demonstrations. 

To assess the reliability of the data submitted by states to HHS to 
calculate historical state spending and enrollment growth rates, we 
reviewed the steps HHS takes to ensure the accuracy of spending data 
compiled in states' automated Medicaid information systems. We obtained 
the data states' submitted to HHS and reviewed them for anomalies and 
missing information. We also interviewed HHS and state officials 
knowledgeable about the data. We discussed limitations of the automated 
Medicaid data, such as potentially incomplete data and states' ability 
to revise data for up to 2 years, with HHS officials. Because the data 
used to establish spending limits were for a time frame for which the 
states' data should have been largely completed and finalized, we 
concluded that states' Medicaid spending and enrollment data are 
sufficiently reliable for the purposes of this report. 

We conducted our work from June 2006 through January 2008 in accordance 
with generally accepted government auditing standards. 

[End of section] 

Appendix II: Comments from the Department of Health and Human Services: 

Note: GAO comments supplementing those in the report text appear at the 
end of this appendix. 

Department Of Health & Human Services: 
Centers for Medicare & Medicaid Services: 

Office of the Administrator: 
Washington. DC 20201: 

Date: December 21, 2007: 

Subject: Government Ac our ability Office (GAO) Draft Report: "Medicaid 
Demonstration Waivers: Recent HHS Approvals Continue to Raise Cost and 
Oversight Concerns" (GAO-08-87) 

Overview:  

We appreciate the opportunity to comment on the GAO draft report, 
"Medicaid Demonstration Waivers: Recent HHS Approvals Continue to Raise 
Cost and Oversight Concerns" (GAO-08- 87). Section 1115 demonstrations 
are an important tool for use at the Secretary's discretion to test 
innovations in health care delivery and coverage and by longstanding 
Department policy, must be budget neutral to the Federal Government. 
These demonstrations test important new ideas while extending health 
care coverage to thousands of low-income Americans who would otherwise 
lack that coverage. 

The draft report discusses two approved demonstration projects, and 
contains only one substantive recommendation. There is no 
recommendation for executive action for the Vermont section 11 15 
Global Commitment to Health Demonstration. With respect to the Florida 
section 1115 Demonstration, the draft report recommends that the 
Secretary of the Department of Health and Human Services (HHS) "ensure 
that the level of supplemental payments for which the State could have 
obtained Federal Medicaid funds in the absence of the proposed 
demonstration is calculated using appropriate methods and accurate data 
sources" (p. 34). It further urges the Secretary to adjust the approved 
spending limit accordingly. We disagree with this proposed 
recommendation. for reasons set forth in the discussion below. 

As we have noted through repeated meetings over the course of more than 
a year, involving hundreds of staff hours, we also disagree strongly 
with many of the characterizations and findings in the draft report in 
addition to that recommendation. The draft report makes numerous 
inaccurate assertions and characterizations with regard to both budget 
neutrality and Medicaid fiscal integrity. For example, the new draft 
report mischaracterizes the nature of these Secretarially-approved 
demonstration programs and the Secretarially-defined policies 
surrounding their approval, and does not adequately account for the 
likelihood of differences in professional interpretation in 
quantifiable analyses. Moreover, the draft report does not consider 
HHS' policies to focus limited federal resources to ensure Medicaid 
compliance and fiscal integrity in payments under ongoing State plans 
and demonstrations on a going-forward basis. As a result, the draft 
report seriously distorts HHS' efforts with respect to our fiscal 
integrity activities. 

This Administration's unparalleled commitment to correcting financing 
issues within the Medicaid program has been extremely successful in 
ensuring that claimed Medicaid expenditures are based on appropriately 
financed expenditures that are not redirected for other purposes. 
Moreover, this Administration has approved significant demonstration 
projects that would expand Medicaid eligibility, coverage and access to 
providers. These are initiatives in which the Federal Government 
obtains a tangible result such as the creation of Provider Access 
Systems in Florida, coverage initiatives in California, and the 
universal coverage commitment in Massachusetts. These program 
components of section 1115 demonstrations have a serious real- world 
impact on the lives of many low-income people. 

Restatement of Past HHS Responses to GAO on 1 115 Demonstration Issues: 

In this draft report, GAO unnecessarily reasserts points made in prior 
reviews with regard to 1115 demonstrations, With regard to these 
points, HHS continues to stand by the responses we provided to the May 
8, 2007 draft document, "Medicaid Demonstration Waivers: Lack of 
Opportunity for Public Input During Federal Approval Process Still a 
Concern" (GAO-07-694R). That response detailed the extensive efforts 
HHS has undertaken to make sure that the 1115 demonstration review 
process is transparent and to provide opportunities, as well as 
documenting the significant opportunities for public input during State 
legislative and administrative processes. 

Also, GAO raised technical and legal questions pertaining to waivers 
granted to allow the implementation of the Florida and Vermont 
demonstrations, in a policy letter sent to Daniel Meron, General 
Counsel of HHS. We believe that the issues of legal authority, program 
effects on the beneficiary, cost-sharing, and the HHS evaluation 
process were all adequately and appropriately addressed in the April 
26, 2007, response from Dennis Smith to Lynn H. Gibson, Managing 
Associate General Counsel of GAO. 

Medicaid Fiscal Integrity: 

Disappointingly, the report is significantly lacking in its discussions 
pertaining to Medicaid fiscal integrity. The criticism of the Agency's 
efforts in the section of the GAO's draft report related to CMS' fiscal 
oversight of State financing is not supported by actual evidence. HHS 
believes GAO has overlooked and understated the progress that CMS has 
made since the early 1990s to curtail improper financing arrangements. 
Also, GAO has overstated its involvement, indicating that GAO has 
reported improper financing arrangements numerous times, but only 
citing one such report (from 2004). This is unfair and severely 
understates the Agency's significant efforts over the past 4 years to 
end impermissible financing arrangements. We note further that our 
efforts to end such arrangements have been recently highlighted by the 
GAO itself in a March 2007 report. 

This draft report continually references a February 2004 GAO report 
(MEDICAID: Improved Federal Oversight of State Financing Schemes is 
Needed GAO-04-228) to support an allegation that CMS has allowed 
improper State financing in the approval of both the Vermont and 
Florida 1115 demonstrations and to intentionally imply that CMS has not 
made efforts to ensure proper State financing since the release of 
GAO's February 2004 report. However, the primary scope of the February 
2004 GAO report related to CMS' implementation of-nursing home upper 
payment limit (UUPL) transition periods, the significant majority of 
which have expired. According to the precise language of the referenced 
2004 report, "GAO was asked to examine CMS' oversight of nursing home 
UPL arrangements. including the status of and the basis for transition 
period decisions. None of GAO's recommendations in the 2004 report 
addressed State financing, as is suggested by the following statement 
on page 7 in this report: 

(See comment 1.): 

"Additionally, an earlier recommendation from our 2004 report on 
Medicaid fiscal integrity issues—a recommendation that 1111S agreed to 
implement hut thus far has not that HHS establish uniform guidance to 
states that set forth acceptable methods related to state financing 
arrangements remains valid." 

Because the 2004 GAO report did not address financing arrangements, 
this statement in the draft report is misleading. Moreover, this draft 
report erroneously suggests that Medicaid payment limit necessarily 
result in proper sources of State financing for the Medicaid program. 
This is inaccurate. A UPL test that complies with Federal requirements 
does not ensure that the financing of the non-Federal share of the 
Medicaid payments is proper. Instead, UPL demonstrations under the 
Medicaid State plan (i.e., Medicaid payment limits) and sources of the 
non-Federal share of Medicaid payments (i.e., State financing) are 
separate issues, governed by separate statutory and regulatory 
authority. (See Attachment 1 for further information related to CMS' 
position regarding the above-referenced recommendation in the February 
2004 GAO report.) 

Additionally, this draft report omits mention of a March 2007 GAO 
report (Medicaid Financing: Federal Oversight Initiative is Consistent 
with Medicaid Payment Principles but Needs Greater Transparency GAO-07-
214) in which GAO acknowledged significant fiscal oversight efforts on 
the part of CMS including the fact that CMS had ended improper 
financing arrangements in 29 States. Equally remarkable in this draft 
report is the omission of mention of the June 2006 GAO report (Medicaid 
Financial Management: Steps Taken to Improve Federal Oversight but 
Other Actions Needed to Sustain Efforts GAO-06-705), in which GAO 
acknowledged nine specific areas of improvement to CMS' Federal 
oversight efforts. including improved efforts to oversee State claims 
for Federal reimbursement, enhanced ability to address high-risk Slate 
funding practices, and the creation of goals to reduce inappropriate 
Federal reimbursement. The exclusive and continual reference to the 
GAO's February 2004 report to support current findings is troublesome. 

(See comment 2.): 

This draft report inaccurately states that CMS acceptance of the data 
sources historically utilized by Florida to perform an inpatient 
hospital UPL demonstration under the Medicaid State plan reflects a 
lack of commitment on the part of CMS to the fiscal integrity of the 
Medicaid program. As described in detail below, Florida's historical 
use of certain data sources was largely irrelevant to the development 
of the Low-Income Pool under Florida's 1115 demonstration. In addition, 
the treatment of Florida was consistent with the treatment accorded 
other States. Consistent treatment of all States in the assessment of 
financing was a key recommendation of GAO's March 2007 report. As CMS 
indicated in its February 16, 2007 response to the GAO's draft 2007 
report, CMS has instituted processes to ensure consistent treatment to 
all States, which included a series of standardized questions posed to 
all States under the Medicaid State plan amendment review process and a 
prospective termination of State financing arrangements that were 
inconsistent with Federal requirements. It appears the GAO omitted any 
reference to the March 2007 report because, by doing so, it would 
expose the illogical finding included in this report. 

This draft also inaccurately faults CMS for allowing the Vermont 
managed care organization to be reimbursed in an amount greater than 
cost. Contrary to that assertion, the Special Terms and Conditions for 
Vermont specify that the managed care organization must spend all its 
revenue under the demonstration for demonstration purposes. GAO 
suggests that CMS allowed the State to retain excess revenue and use 
the funds to support health-related programs previously funded by the 
State and that do not exclusively benefit Medicaid beneficiaries. 
However, under the Vermont demonstration, CMS has ensured the 
transparent identification of costs for which Federal matching funds 
are available by clearly identifying the expenditures to which Vermont 
must adhere with the revenue received by the managed care organization. 

(See comment 3.): 

Budget Neutrality—Unsubstantiated Allegations about Application of 
Policy 

The HHS staff has spent significant time educating GAO staff on the 
subject of budget neutrality, in addition to the equally lengthy 
discussions on fiscal integrity noted above. As you may know, budget 
neutrality has been a longstanding Departmental requirement of health 
care reform projects that the HHS has approved since 1994, and across 
two different administrations. Budget neutrality is, by its nature, an 
important requirement, for it ensures that the Federal Government does 
not spend more under a demonstration project than it would have absent 
that demonstration. There are multiple methods of establishing that a 
project is budget neutral, and each budget neutrality agreement must be 
considered as part of a larger picture of whether a State's proposal, 
in the Secretary's estimation, promotes the objectives of the relevant 
title of the Social Security Act (in these instances, Title XIX). These 
projects are approved at the discretion of the Secretary of HHS, and 
this discretionary review includes both budgetary and programmatic 
elements. This, however, is not to say that there is no guidance for 
external audiences or guidelines that are used in negotiations with 
States--or that HHS' application of these self determined budget or 
program parameters have been arbitrary or inconsistent. Nor is it to 
say that "everything is negotiable" about budget agreements. HHS spends 
many staff hours with States providing technical assistance and 
education about budget neutrality. 

The HHS has, as consistent with policy under this Administration, used 
as a starting point for discussions with States, a policy of using the 
lower of State historical spending or the projections from the 
President's Budget, sometimes referred to as the "lower-of' guideline. 

States are required to provide historic experience with their 
proposals. This begins a review performed by a Federal team, and 
analyses are performed using the "lower-of' guideline provided above. 
What follows are discussions with States to determine if any of the 
historical experience was affected by unusual events that render the 
data suspect for trend development, and therefore the prediction of 
future spending. 

The GAO has throughout its discussion documents and draft report, 
characterized our Agency's own internally-developed guideline as a 
"benchmark policy"-- and then criticized HI-IS for making relatively 
minor adjustments for real-world factors that could affect a State's 
spending. A few such examples include eligibility changes, benefit 
coverage changes (such as the implementation of Medicare Part D), 
claims processing abnormalities, and the like. In such instances the 
cost trend analysis can then be adjusted using techniques that are 
acceptable in statistical analysis and evaluations in order to yield a 
more "robust" analysis. GAO's characterization of our longstanding 
"lower-of' guideline as a "benchmark policy" misrepresents the 
guideline as being an inflexible standard. Instead it is applied as an 
initial position that is subject to adjustment based on actual 
expenditures experience based on our professional judgment. These 
adjustments, and the selection of the most appropriate "benchmark" are 
issues on which there can be differences in professional 
interpretation, as noted above. The draft report thus mischaracterizes 
the Department's policy. 

(See comment 4.): 

We also believe that the application of GAO's apparent interpretation 
would severely limit the negotiating strength of an HHS Secretary under 
any Administration in promoting and testing new ideas in the health 
care marketplace. 

(See comment 4.): 

State-Specific Comments and Issues: 

In the sections below we will address State-specific issues raised in 
this report, with respect to budget neutrality, program questions, and 
fiscal integrity. CMS officials spent countless hours meeting with GAO 
staff to provide explanation of the efforts put forth by CMS under the 
Medicaid State plan amendment financing initiative to ensure that 
Florida's and Vermont's Medicaid programs were properly funded. 
Unfortunately, the substance of those meetings has been omitted from 
this draft report, and we do not believe the information included in 
this report is accurately represented. 

Florida-Specific Issues: 

The GAO contends that HHS did not adequately ensure that Florida's 
Medicaid demonstration would be budget neutral to the Federal 
Government before approving it, and also that HHS' basis for approving 
the higher spending limits was not fully supported by the documentation 
(page 6). HHS believes that the adjustments to the State's historic 
expenditure growth rates were both appropriate and consistent with HHS 
policy to develop rates that are projected to be most reflective of 
projected future without waiver Medicaid expenditures. The established 
agreement reflected a careful analysis of current spending and 
specifically limited how it could grow in the future. One of our most 
significant issues with regard to the draft report is that GAO does not 
acknowledge in any way that HHS succeeded in capping Medicaid program 
growth in Florida which had averaged 13 percent in recent years. 

(See comment 5.): 

The GAO further contends that "HHS allowed Florida to use incorrectly 
calculated levels of supplemental payments as a basis for spending 
allowed under the demonstration" (page 5). Despite GAO's assertion that 
Florida's supplemental payment program is problematic. GAO does not 
provide the detail and financial data to support its position or an 
estimate of how the approved $1 billion annual Low Income Pool (LIP) 
should have been adjusted as a result of the "deficient" supplemental 
payment program. Because of the lack of explanation by GAO, this seems 
to be in conflict with our principle of using actual rather than 
calculated numbers or estimates for without waiver trend line 
development. Additionally, as noted above, as a condition of the 
Florida health reform demonstration, HHS required the State to 
discontinue its inpatient supplemental program, effectively freezing 
the growth of supplemental payments beyond the S1 billion annual 
threshold Moreover, HHS required the State to establish new Letters of 
Agreement between the Florida Agency for Health Care Administration and 
the "Funding Sources" that support the LIP. This ensures that the 
sources of funding are appropriate; an assurance never before available 
to the Federal Government and Federal taxpayers. 

(See comment 6.): 

Alarming also is the GAO position that HHS did not ensure the 
demonstration in Florida maintained the fiscal integrity of the 
Medicaid program prior to approval. As evidenced by materials provided 
by CMS to the GAO, Florida's financing of Medicaid institutional 
service payments, including supplemental payments, was initially 
reviewed under the same State plan fiscal oversight initiative that was 
applied to all States and that was documented in the GAO's March 2007 
report. During that review, Florida provided CMS with the assurance 
that the source of the non-Federal share of Medicaid supplemental 
payments was derived from local tax revenue and that hospitals were 
able to retain 100 percent of the total computable expenditures. 

(See comment 7.): 

While CMS accepted that written assurance from a Florida official, CMS 
continued to review Florida's financing through the Agency's financial 
management review (FMR) process, a process that was specifically 
applied to all States that agreed to terminate particular State 
financing arrangements and to certain States that provided CMS initial 
assurance of proper State financing. CMS initiated the FMR process in 
the affected States to ensure the States were meeting the commitments 
made to end certain financing arrangements and to confirm assurances 
provided by certain States that State financing arrangements were 
consistent with Federal requirements. 

By way of background, the basis for and scope of this type of FMR was 
to ensure nationally consistent policy application and to sustain CMS' 
legal position to challenge future claims; specifically, in the event a 
State continued to make claims for Federal matching funds in State 
fiscal year 2006 with (i) the same source of State financing they 
agreed to terminate at the end of their State fiscal year 2005; or (ii) 
with a source of State financing that was inconsistent with the 
assurance provided by the State under the Medicaid State plan amendment 
review process. CMS developed "internal-only" reports from the results 
of these FMRs while we monitored State claims during all affected 
States' fiscal year 2006. None of these types of FMRs were ever 
released to any of the affected States. 

In the case of Florida, CMS performed an FMR specific to the State 
financing utilized by Florida to make supplemental Medicaid payment to 
hospitals. The scope of this review as delineated in the Atlanta 
regional office's 2005 financial management work plan, was to examine 
whether or not hospitals were returning any of their supplemental 
Medicaid payments to the local Government. As part of the internal-only 
report developed as a result of the FMR, the Atlanta Regional Office 
also incorporated concerns with the sources of data utilized by Florida 
to perform the inpatient hospital UPL calculations based on information 
that was included in the 2005 funding specialist profile.[Footnote 56] 

While the FMR did not reveal that hospitals were returning their 
supplemental Medicaid payments to the State, the FMR did raise concerns 
with the manner in which the local financing was utilized by the State 
as the non-Federal share of the Medicaid supplemental payments. CMS 
used the findings of the FMR to require Florida to make prospective 
changes to their Medicaid program. 

We also take serious issue with GAO's characterization of the LIP 
associated with the Medicaid 1115 demonstration. Although this draft 
report accurately identifies the LIP as a key component of Florida's 
1115 demonstration, the GAO's characterization of the LIP's design is 
inaccurate and misleading. Most importantly, the LIP was not designed 
to continue a supplemental payment program as alleged on page 13 of 
this draft report. "I he GAO appears to concede elsewhere in the draft 
report that such characterization is inaccurate in its expressed 
acknowledgement that "Florida agreed to discontinue its supplemental 
payment program under the terms of the demonstration." (page 26). The 
LIP was designed to utilize funding that had historically been used by 
hospitals to subsidize their care of the uninsured to expand the scope 
of health care services available to Medicaid and uninsured individuals 
and to document such costs in a manner that was transparent to the 
Federal taxpayer. 

The GAO points to State law referencing the UPL to support the position 
that the LIP was merely an extension of the State's UPL program. 
However, it is irrelevant that "Florida law made the State's authority 
for pursuing the proposed demonstration contingent in part on Federal 
approval to preserve the upper payment limit funding mechanism," as is 
footnoted on page 13. It is not unusual for States to require this type 
of funding contingency for a program such as this demonstration before 
the State moves forward. The LIP, as designed in this demonstration, is 
based on historical State spending and is limited to documented costs 
to the State for providing health care. To the extent such costs exceed 
historical spending under the Medicaid supplemental payment program. 
Federal matching funds will not be available. Again. HHS is not 
credited for this noteworthy achievement. 

The draft report questions the data sources utilized under the State of 
Florida's historical inpatient hospital UPL demonstration and, in doing 
so, implies the LIP is inflated. CMS and Florida developed the LIP by 
determining the actual Medicaid supplemental payments that were made in 
excess of actual Medicaid costs. This portion of supplemental spending 
was moved into the LIP pool and formed the LIP ceiling. The State was 
required to remove all supplemental payments from the Medicaid State 
plan and could not make Medicaid payments to hospitals above Medicaid 
cost. Since the initial budget neutrality calculation was based on 
historical spending below the UPL and above Medicaid cost and then 
trended forward, this spending is clearly controlled. 

Budget neutrality has historically been limited to a State's actual 
current law Medicaid spending, meaning spending authorized under its 
approved Medicaid State plan. This spending would include approved 
supplemental payments that were under the State's UPL. 

It should be noted that Florida's UPL demonstration under the State 
plan resulted in approximately $1.6 billion of available spending 
(i.e., "gap"). However. Florida historically made Medicaid supplemental 
payments to hospitals in an amount that was approximately 58 percent of 
that total available spending under the UPL. Hence, the reference to an 
inflated UPL is misleading in that Florida's spending did not 
"maximize" Medicaid supplemental payment spending under the UPL. To 
have limited the State's spending under their UPL program, the 
questioned data would have needed to impact the UPL by inflating it 
nearly 50 percent. The UPL demonstration under the State plan 
successfully caps the State at a much lower amount. 

(See comment 9.): 

The CMS acknowledges that the Florida funding specialist profile 
initially questioned the accuracy of the data sources used by the State 
of Florida in calculating the UPL for inpatient hospital Medicaid 
supplemental payments. CMS does not have an historical enforcement 
practice of challenging the data used by States to perform the UPL 
demonstration under the State plan. As explained earlier, CMS reviews 
UPL demonstrations under the Medicaid State plan amendment review 
process and such review applies to the "methods" by which States 
calculate 

UPLs, not the data sources utilized by States. Under this review 
process, CMS is bound by regulatory timeframes and requires States to 
demonstrate proper UPL "methodologies" supported by readily available 
data to States. This process of reviewing UPL demonstrations under the 
Medicaid State plan amendment process allows CMS to ensure the 
application of uniform guidance to States by setting forth acceptable 
methods to calculate their UPLs, which happens to be the precise 
direction of the continually referenced recommendation of the February 
2004 GAO report. 

Of even greater significance is the GAO's apparent lack of 
understanding related to the structure of and limitations to the LIP. 
On page 28 of the draft report, the GAO incorrectly indicates that CMS 
did not require Florida to address the problems with the UPL 
methodology and the data used to determine the amount of supplemental 
payments when determining the amount available in the LIP under the 
demonstration. On the contrary, the special terms and conditions 
associated with the section 1115 demonstration require Florida to limit 
both Medicaid State plan spending and LIP spending to cost and to 
document allowable costs in a manner consistent with Medicare cost 
principles, including the use of the Medicare 2552-96 hospital cost 
report along with hospital accounting records and audited financial 
statements. Consequently, the design of Florida's LIP under the 1115 
demonstration ensures that Federal matching funds will be limited to 
the "actual" cost of providing inpatient hospital services, not a 
reasonable estimate of what Medicare would pay (i.e., the UPL). 

(See comment 10.): 

The CMS did not address Florida's UPL program because the State has 
discontinued it and has elected to limit its spending to cost. For 
example, if hospitals only incur $600 million of costs that are 
allowable under the Medicaid State plan or the LIP, Federal matching 
funds will only be available as a percentage of those actual allowable 
costs. Moreover, the GAO properly recognizes that the special terms and 
conditions allow CMS to adjust the spending limit under several 
circumstances including impermissible provider payments. 

(See comment 10.): 

Based on the structure of and limitation to the LIP, suggesting a 
retroactive adjustment to historical Medicaid supplemental spending is 
both an inappropriate and unnecessary step to ensure the fiscal 
integrity of Florida's 1115 waiver demonstration. 

(See comment 11.): 

Vermont-Specific Issues: 

As noted above, there was no recommendation for executive action for 
the Vermont section 11 15 Global Commitment to Health Demonstration. In 
its general overview, the GAO speculates that in the approval of the 
Vermont section 1115 Global Commitment to Health Demonstration, I HHS 
did not adequately ensure budget neutrality and the maintenance of 
fiscal integrity of the Medicaid program similar to its assertion with 
respect to Florida above. From what HHS can gamer from the report, 
these contentions are based primarily on four issues; Vermont's 
publicly- run managed care organization (MCO), the use of a higher 
trend rate for enrollment growth, the use of "hypothetical 
expenditures" from the previous Vermont Health Access Plan (VI IA P).

Demonstration to build a base ceiling, and the arrangement to pay 
administrative costs at a rate higher than previously paid. In prior 
efforts to provide the GAO with an expedient and comprehensive 
explanation HHS has already responded to many of these critiques. 
However, additional clarification is provided to further illuminate the 
Vermont case.

In 2004-2005, faced with the rising cost of health care and a State 
budget crisis Vermont approached HHS seeking approval under section 1 
115 authority to revamp its health care system to improve quality and 
access, while decreasing cost. In 2005, a new section 1115 
demonstration, The Global Commitment to Health, was approved. The new 
demonstration incorporated the current Vermont section 1115 VHAP 
Demonstration, currently costing significantly more than originally 
projected, and transitioned the State Medicaid program into a publicly 
run MCO. This publicly-run MCO for the State of Vermont is similar in 
concept to the Health Insuring Organizations (HIOs)—also known as 
County-Organized Health Systems--in several counties in the State of 
California where a single publicly-run entity arranges and manages the 
care for Medi-Cal beneficiaries. We note that the HIOs were established 
by Congress in the mid-1980s and are exempt from certain Medicaid 
requirements, and cover many lives in that State (approximately 500,000 
versus the Vermont Global Demonstration's approximately 140,000).

To begin, it is unclear why GAO characterizes accumulated savings from 
the expired VHAP Demonstration as "hypothetical" when credited to 
VHAP's successor, Vermont Global (pages 21 and 25 of the report and 
elsewhere). The Global Commitment to Health Demonstration was a 
continuation and expansion of VHAP, covering the same populations, 
including expansion populations. It has long been HHS policy to 
determine budget neutrality over the life of a demonstration, and there 
is no reason not to include a clearly related predecessor 
demonstration. Indeed, to make such an exclusion would be a significant 
departure from past policy on 1115 demonstrations across two 
Administrations. Furthermore, any "hypothetical" costs not spent 
secondary to the transition from the prior VHAP Demonstration to the 
Global Commitment to Health Demonstration certainly would have been 
expended if the current V HAP Demonstration had continued in its 
original form.

(See comment 12.): 

Under the Global Commitment to Health section 1115 Demonstration, 
Vermont is permitted to generate savings in the same manner as a 
privately-managed Medicaid MCO receiving Federal funds. However, per 
the special terms and conditions, Vermont is held to additional 
requirements placed on the publicly run MCO that ensure any savings 
generated are used to fund services and programs designed to serve the 
Medicaid and other low-income populations; meanwhile, the payments via 
the rate-setting are required to be compliant with actuarial 
certification requirements, as are all other Medicaid MCOs. 
Furthermore, the use of savings generated from traditionally 
categorically eligible Medicaid populations to expand services and 
programs to other low-income populations is a fundamental tenet of 1115 
health care reform demonstrations. This, again, is longstanding HHS 
policy across administrations. The authority to approve this 
demonstration rests with section 1 115 and 1903(m)(6)(A), as well as 
the ability of States to operate public MCOs under section 1915 and 
1932. 

See comment 13.): 

Regarding enrollment projections related to budget neutrality for the 
Global Commitment to Health, it appears that GAO contradicts itself 
While at some points in the document it states that HHS does not 
establish benchmarks for enrollment growth trends (page 10, for 
example), at other times it claims that the enrollment trends granted 
violate HI-IS benchmark policy (page 22). Unlike most approved section 
11 15 demonstrations, both enrollment and administrative costs are 
subject to the $4.7 billion Global budget neutrality ceiling (or cap). 
This arrangement places the State at risk for ensuring both enrollment 
and administrative costs remain reasonable and provides the State with 
an incentive to reduce administrative costs and carefully monitor 
enrollment. This is, in fact, a more stringent requirement than those 
in place in most section 1 115 demonstrations, where in fact, 
administrative costs are not usually subject to the budget neutrality 
ceiling. Moreover, as GAO itself highlights in the report (page 31), 
the projected administrative costs closely resemble that of the private 
industry. 

See comment 14.): 

Additional Technical Corrections and Other Comments: 

We also offer the following comments of a more technical nature. 

* The text of the report suggests that HI-IS policy requires that all 5 
years of historical data must be used for the historical trend 
analysis. This misimpression is addressed, but it is buried in a 
footnote (page 20, footnote 40). The Health Insurance Flexibility and 
Accountability Guidance that GAO emphasizes as documentation for the 
Ill IS Benchmark policy requires the "states to submit five years of 
historical data for assessment by CMS, with quantified explanations of 
trend anomalies," with the purpose of selecting the most appropriate 
data and time period to use in the analysis. 

* Page 7, a goal of the Florida demonstration is to provide 
beneficiaries with access that is broader than conventional managed 
care plans; it is also to include Provider Access Systems (PASs). We 
believe GAO should replace "state-approved managed care benefit plans" 
with "state-approved Managed Care Organizations (MCOs) and Provider 
Access Systems (PASs)".

* Page 10, CMS considers the GAO cited benchmarks to be wrong; 
especially the use of the 4.8 percent President's Budget trend rate for 
Aged, Blind and Disabled does not match-up with the expenditure data. 
The cost data provided by the State excluded Part D expenditures and 
did not include the expenditure and caseload experience of Medicare 
Savings Groups. 

* Footnote 32 (page 15) is confusing and conflates information on two 
separate 1115 demonstrations for Wisconsin-- it does not clearly 
identify which demonstration is being discussed. Wisconsin was required 
to return $10.2 million for exceeding spending limits for BadgerCare, 
while the prior 2004 GAO report referenced in the new draft report had 
criticized CMS' approach to budget neutrality for SeniorCare (a 
demonstration that focused on providing drug coverage to low-income 
seniors and which was implemented before the advent of Medicare Part 
D). HHS appreciates the notation on the 2007 return of $10.2 million to 
the Federal Government because the State exceeded spending limits for 
BadgerCare; this appropriately acknowledges that HHS takes seriously 
its fiduciary responsibility to the Federal taxpayer. We further note 
that with respect to SeniorCare, HHS was prepared to terminate it for 
failing to remain within approved spending limits. Despite the HHS 
position, Congress elected to extend the demonstration via legislative 
action. 

* Also in footnote 32, it is important to note that the collection of 
funds from Wisconsin for BadgerCare pre-dates the 2006 Program 
Assessment Rating Tool activity. Wisconsin had its second renewal 
during 2007; the collection of funds was an HHS-imposed requirement of 
the previous renewal. 

Attachment 1: 

In response to the draft 2004 report, CMS concurred with the 
recommendation that "CMS establish uniform guidance for states, which 
would set forth acceptable methods to calculate their UPLs." However, 
CMS specifically indicated in our response that we "did not necessarily 
agree with the GAO's definition of a reasonable estimate of the UPL" 
and we did not "believe that an extensive laundry list of acceptable 
methods could be compiled that would address every payment methodology 
to every provider in every State." Moreover, the GAO's recommendation 
in the February 2004 report addressed the need for guidance on the 
"methodologies" to calculate UPLs, not on the data sources utilized by 
States. Technical comments included in CMS' response to the 2004 draft 
report, which are conveniently "not printed" by the GAO in the February 
2004 report, specifically challenged the GAO's suggestion that a 
methodology requiring a demonstration of a "reasonable estimate" should 
rely on precise data sources. When calculating UPLs, States arc 
estimating what Medicare would pay for similar services provided to 
Medicaid patients and States are required to perform such 
demonstrations on a prospective basis. The precision to which the GAO 
suggested States should calculate their UPLs was merely based on the 
ability of the GAO to spend a year and a half reviewing calculations 
from prior periods. This luxury does not exist for CMS and the States 
due to the State plan amendment effective date requirements and State 
budgeting processes. 

The CMS has addressed proper UPL methodological demonstrations in 2 
ways under the Medicaid State plan amendment review process, processes 
of which were recognized by the GAO in the February 2004 report. 
Specifically, CMS requires (as part of the five standard funding 
questions) States to provide a UPL demonstration from all States 
proposing to increase Medicaid reimbursements. During the State plan 
amendment review process, CMS evaluates the reasonableness of a States 
UPL methodology but CMS does not validate the data sources utilized by 
the States. It is unreasonable to suggest that CMS validate the data 
sources of a prospective State estimate. CMS has not generally 
challenged data sources utilized by States to perform UPL 
demonstrations and the GAO's February 2004 report does not specifically 
recommend such a requirement. CMS also requires States to identify the 
sources of the non- Federal share of Medicaid payments under the 
Medicaid State plan amendment review process, which has significantly 
reduced State interest in maximizing UPL demonstrations. Specifically, 
many States making supplemental Medicaid payments were not allowing the 
health care providers to retain the Medicaid supplemental payments to 
which they were entitled. Therefore, States enjoyed a significant 
incentive to maximize the UPL demonstrations. However, as CMS required 
States to terminate such financing practices. States have demonstrated 
little interest in pursuing aggressive UPLs. 

Again, CMS appreciates the opportunity to review and comment on the 
draft report. 

The following are GAO's comments to certain concerns raised in HHS's 
letter dated December 21, 2007. 

GAO Comments: 

1. We refer to our 2004 report because of the significant role 
supplemental payments play in Florida's demonstration and our concern 
that HHS did not require the state to correct known problems with these 
supplemental payments before establishing a spending limit on the basis 
of historical payments. HHS stated that we did not address financing 
arrangements in our 2004 report, but we disagree. An objective of the 
2004 report was to determine if HHS's continuing oversight of 
supplemental payment arrangements was sufficient to ensure that claims 
submitted by states were calculated appropriately and complied with 
Medicaid requirements. Although we noted that HHS had taken a number of 
steps to strengthen its oversight of these payment arrangements, we 
found that HHS had not issued guidance for states' use on appropriate 
methods for calculating their Medicaid Upper Payment Limit (UPL). We 
recommended that HHS establish uniform guidance that would set forth to 
states acceptable methods for calculating the UPL. Our concern in this 
report is that HHS approved a spending limit for Florida's low income 
pool based on the state's UPL without first requiring Florida to 
address problems HHS identified in Florida's methodology for 
calculating this UPL. 

We disagree that the draft of this report erroneously suggested that 
Medicaid payment limits necessarily result in proper sources of state 
financing. As noted in the draft, we are concerned that HHS approved a 
spending limit for the low-income pool based on potentially inflated 
historical payments. Our draft report credited HHS for requiring 
Florida to correct the issue the department had identified with the 
source of the state's financing. 

2. We disagree with HHS's characterization of the findings from our 
2006 and 2007 reports. Although our 2006 and 2007 reports addressed HHS 
oversight of Medicaid and discussed agency actions to strengthen 
oversight, certain of the findings of these earlier reports resonate 
with our current findings. In 2007, for example, we found HHS review 
and approval of state plan amendments to be marked by a lack of 
transparency and clear guidance. And in 2006, although we noted recent 
improvements in the financial management processes HHS uses in its 
oversight of states, we found it too soon to assess their impact, and 
further noted additional weaknesses that HHS had not addressed. 

3. As discussed in the draft report, we are concerned that HHS allowed 
Vermont to seek reimbursement for administrative costs higher than that 
of other public health entities, and that HHS agreed to reimburse 
Vermont for a larger portion of these administrative costs than 
typically afforded other states. We disagree that HHS transparently 
identified costs for which federal reimbursement of excess revenues 
from the public managed care organization are available. As noted in 
the report, the purposes for which Vermont may spend these excess 
revenues are governed only by a set of broad health objectives. For 
example, HHS allowed Vermont to spend excess revenues on expenditures 
that increase access to quality health care for Medicaid enrollees and 
those lacking adequate insurance and that improve health outcomes and 
quality of life for Medicaid-eligible individuals. 

4. Our draft report recognized HHS's discretion in making adjustments 
in the spending limits. For example, we accepted a projected $2.1 
billion in adjustments to the spending limits for Florida and Vermont 
because these adjustments were supported by quantified explanations. As 
noted in the draft, seemingly small changes to per person cost growth 
rates are amplified by the high volume of beneficiaries that access 
Medicaid services and the number of years across which these cost 
growth rates are applied. As noted in the draft, the seemingly small 
changes to the growth rates in Florida and Vermont resulted in nearly 
$7.2 billion that we identified as not budget neutral. 

5. As noted in the draft report, the established agreement did limit 
how Medicaid spending in Florida could grow in the future. Our concern, 
however, is that these spending limits are not budget neutral. 
Furthermore, we disagree with HHS's assertion that Florida's spending 
limits reflect projected future growth in the absence of the 
demonstration. For example, over $5.5 billion in projected spending we 
identified as not budget neutral stems from an adjustment to reflect in 
part what the state projected would be higher anticipated costs of 
delivering Medicaid services in a managed care environment. These costs 
would not be incurred absent the demonstration, and as noted in the 
draft report, absent evidence supporting the approved changes to 
benchmark amounts, HHS should not have allowed them as a consideration 
in establishing a higher spending limit for the demonstration. 

We also disagree that HHS capped Medicaid program growth in Florida. 
The agency approved per person spending limits rather than a total 
limit on programmatic spending. Thus Medicaid spending in Florida may 
grow by more or less than 13 percent per year depending on enrollment 
in the program. 

6. We believe the spending limit HHS approved for Florida's low-income 
pool was problematic because HHS did not require Florida to correct 
known deficiencies in the state's method for calculating historical 
supplemental payments that served as the basis for the spending limit. 
We did not estimate how the low-income pool should have been adjusted 
because HHS's September 2005 review--which identified the problems with 
Florida's calculation of its financing arrangement--did not estimate 
the actual allowable payments under the program or the extent that the 
prior supplemental payment arrangement was considered excessive or 
inflated. Consequently, we did not have the information available to us 
that would allow a detailed estimate of how the low-income pool 
spending limit should be adjusted. We believe that the concerns raised 
by HHS's own review should have been addressed prior to establishing a 
spending limit based on historical spending. As noted in the draft 
report, HHS should ensure that the level of supplemental payments for 
which Florida could have obtained federal Medicaid funds in the absence 
of the demonstration is calculated accurately, and adjust the approved 
spending limit accordingly. 

7. By not requiring Florida to correct known deficiencies in the 
state's historical spending, we believe HHS did not ensure the fiscal 
integrity of Florida's low-income pool. In the draft report we credited 
HHS for requiring Florida to correct a problem the department 
identified with the manner in which Florida used local financing as the 
nonfederal share of its supplemental payments. Our concern remains, 
however, that HHS did not require Florida to correct a separate problem 
the department identified in the methods and data by which the state 
calculated the amount of supplemental payments eligible for federal 
matching funds under its program. By not requiring Florida to correct 
its method and data sources as a condition of approving the 
demonstration, HHS approved a spending limit for the low-income pool 
based on potentially inflated historical spending. 

8. We believe that we accurately characterized Florida's low-income 
pool in this report. We agree that Florida discontinued its inpatient 
supplemental payment UPL program as a condition of the demonstration, 
and have clarified the language to indicate that HHS allowed Florida to 
develop the low-income pool in order to continue funding for a program 
of supplemental payments to providers. 

9. HHS's estimation of the maximum amount Florida could have spent 
under its UPL is irrelevant to the discussion of the appropriate 
spending limit for the low-income pool. As we stated in the draft 
report, consistent with HHS policy, spending limits should be based on 
actual historical spending and quantified explanations for trend 
anomalies. We believe that, in the absence of reliable historical data, 
spending limits should be based on transparent, clearly articulated 
methodologies. Our concern is that HHS allowed Florida to base this 
spending limit on potentially inflated historical payments as a result 
of the state's flawed methodology for calculating its UPL. 

10. HHS's efforts to limit and document Florida's Medicaid state plan 
spending and low-income pool spending under the demonstration do not 
speak to the extent to which HHS ensured the fiscal integrity and 
budget neutrality of the state's proposed demonstration prior to 
approving it. We are concerned that HHS approved growth rates for the 
demonstration without adequate support and did not require Florida to 
correct problems the department identified in the state's methodology 
for calculating its UPL. 

11. To ensure that the spending limit on Florida's low-income pool is 
budget neutral and based on allowable historical spending, we believe 
that HHS should require Florida to correct problems the agency 
identified in Florida's methodology for calculating its UPL and adjust 
the spending limit for future payments made under the low-income pool 
accordingly. We are not suggesting a retroactive adjustment to the 
spending under the supplemental payment program. This recommendation is 
consistent with our long-standing conclusions that spending limits for 
proposed demonstrations should be based on valid methods. 

12. We characterize accumulated savings from an expired demonstration 
in Vermont as hypothetical because they do not represent actual 
expenditures incurred during the historical period HHS reviewed in 
approving a new demonstration in Vermont. We do not object to 
consideration of actual expenditures from a predecessor demonstration 
in determining a spending limit for a new demonstration. But according 
to HHS's written budget neutrality guidance, surpluses generated early 
in the life of the expired demonstration would not have been available 
to Vermont in the absence of a new demonstration. 

13. During the course of our review of the Vermont demonstration, we 
considered the statutory provisions cited by HHS, and nonetheless, as 
indicated in the July 2007 letter,[Footnote 55] had concerns about the 
consistency of the Vermont demonstration with federal law. 

14. Our discussion and use of enrollment growth benchmarks reflect 
HHS's description of its policy, as written in guidance and as 
described by officials. During the course of our work, HHS officials 
told us that they considered benchmarks of enrollment growth in 
determining an aggregate spending limit for the Vermont demonstration. 
Yet HHS approved enrollment growth rates for the demonstration equal 
to, and in some cases exceeding, the highest rates HHS considered. Our 
main concern is that HHS's basis for approving these enrollment growth 
rates was not well documented. We believe that to maintain 
accountability and transparency in the Medicaid program, the 
Secretary's approvals should be based on clearly articulated policies 
and spending limits that are consistent with these policies. 

[End of section] 

Appendix III: Comments from the State of Florida: 

Florida Medicaid: 
2727 Mahan Drive, MS #8: 
Tallahasse, Florida 32308: 
Visit AHCA online at [hyperlink, http://ahca.myflorida.com]: 

Charlie Crist: 
Governor: 

Andrew C. Agwunobi, M.D.: 
Secretary: 

November 30, 2007: 

Dr. Marjorie Kanof: 
Health Care Managing Director: 
United States Government Accountability Office: 
441 G Street, NorthWest: 
Washington, DC 20548: 

Dear Dr. Kanof:

Thank you for providing the Agency for Health Care Administration, the 
single state agency for administering the Florida Medicaid program, 
with the opportunity to comment on the draft report entitled Medicaid 
Demonstration Waivers: Recent HHS Approvals Continue to Raise Cost and 
Oversight Concerns (GAO-08-87). 

As provided under Section 1115 of the Social Security Act, the 
Secretary for Health and Human Services has broad authority to grant 
waivers of statutory provisions to implement experimental, pilot, or 
other demonstration projects with the goal to assist in promoting the 
objectives of the Medicaid statute. Florida was granted such a waiver 
in order to implement our state legislated reform project in October 
2005. The draft report focuses on recent waivers approved in Florida 
and Vermont, and examines the extent to which the Department of Health 
and Human Services ensured, before approving them, that the 
comprehensive 1115 demonstration waivers will: (a) be budget neutral to 
the federal government; and (b) maintain Medicaid's fiscal integrity. 

We support the Government Accountability Office's (GAO's) efforts to 
evaluate Florida's Medicaid Reform effort and analyze the above issues. 
Florida understands the need to carefully monitor the impact of our 
demonstration in meeting the established goals. Find below our comments 
on the draft report. 

General Comment: 

* During the negotiations of the demonstration waiver, Florida worked 
closely with the Centers for Medicare and Medicaid Services (CMS) to 
ensure that all data and documentation were provided in a timely and 
accurate manner to support our waiver application. 

Budget Neutrality: 

* On page 12, Florida recommends the term "maintenance-of-effort 
requirement" be clarified in footnote #23 to avoid the misperception 
that something was missing from the special terms and conditions of the 
1115 demonstration waiver. Per discussions with CMS, Florida was not 
expanding the population subject to managed care and therefore was not 
required to include this requirement as a part of the special terms and 
conditions. Through previous managed care efforts, Florida satisfied 
the maintenance- of-effort requirement.

Budget Neutrality: 

* On page 18, we recommend reformatting 'b' in alignment with the table 
on page 17 in order to provide a better flow of the table and the 
references. 

Low Income Pool: 

* On page 28, the report states, "In November 2006, HHS officials said 
that the problems identified with the improper calculation of states' 
allowed supplemental payment amounts would be corrected at a later 
date." This statement is linked to footnote #48 which is located on 
page 29. The footnote indicates that correcting this problem in the 
case of Florida was unnecessary. Prior to and during negotiations for 
the demonstration waiver, Florida worked closely with CMS to resolve 
any concerns. 

Again, we appreciate the opportunity to provide comments on your draft 
report. Should you have any questions about our comments, please 
contact me at (850) 488-3560. 

Sincerely,

Signed by: 

Thomas W. Arnold
Deputy Secretary for Medicaid TWA/lam: 

cc: Mr. Mark Thomas, Chief of Staff Mr. Clint Fuhrman, Deputy Secretary 
for Communications and Legislative Affairs:

[End of section] 

Appendix IV Comments from the State of Vermont: 

State of Vermont: 
Agency of Human Services: 
Office of the Secretary: 
[phone] 802-241-2220: 
[fax] 802-241-2979: 
103 South Main Street: 
Waterbury, VT 05671-0204: 
[hyperlink, http://www.ahs.state.vt.us]: 

November 20, 2007: 

Marjorie Kanof, MD
Managing Director, Health Care: 
United States Government Accountability Office: 
441 G Street, NW: 
Room 5A21: 
Washington, DC 20548: 

Dear Dr. Kanof: 

Thank you for the opportunity to review a copy of your proposed report 
entitled Medicaid Demonstration Waivers: Recent HHS Approvals Continue 
to Raise Cost and Oversight Concerns (GAO-08-87). 

As you are aware, the nation faces a massive challenge related to 
health care. Public spending for health care continues to escalate; 
meanwhile, 47 million Americans lack basic health care coverage. The 
debate regarding how to address the health care crisis will continue at 
both the state and federal levels. 

In Vermont, our Governor and Legislature recognized that the 
traditional approach for funding health care services was not 
financially sustainable. Absent meaningful reform, Vermont would 
continue to spend a greater percentage of its budget on health care 
with little or no opportunity to enhance access to health care or 
improve the Vermont health care delivery system. 

Section 1115 of the Social Security Act provides the Secretary of the 
US Department of Health and Human Services with authority to grant 
state Medicaid programs the flexibility to pursue alternative 
approaches to meet the objectives of the Medicaid program. In light of 
challenges related to Medicaid at the federal level, we appreciate the 
Secretary's willingness to partner with Vermont by granting us the 
authority to pursue meaningful reform. 

In exchange for the flexibility granted by the Secretary, the State of 
Vermont assumed an unprecedented amount of risk related to program 
expenditures. Federal spending for the vast majority of Vermont's 
Medicaid program is capped at an aggregate level over the course of the 
five-year demonstration. 

The State of Vermont's Administration and Legislature undertook an 
extensive review and analysis of spending projections and the proposed 
five-year spending limit for Global Commitment. Based on historical 
Medicaid growth rates in Vermont, Vermont program expenditures would 
exceed the aggregate spending cap established for the Global Commitment 
Demonstration. In the end, Vermont policymakers concluded that 
"business as usual" was not financially sustainable. Policymakers 
determined that the flexibility granted under the Demonstration would 
permit Vermont to engage in true reform and control program growth. By 
controlling program growth, Vermont's program would become financially 
sustainable and remain within the Demonstration's aggregate spending 
limit. 

Vermont's Demonstration, the Global Commitment to Health, is premised 
on three basic objectives: 1) enhance access to health care services; 
2) improve quality of health care; and 3) contain program costs. We 
believe these objectives are consistent with federal Medicaid policy 
and the goals of many other states. We believe that Vermont's 
experience will be a valuable resource for evaluating strategies for 
program reform throughout the country. 

We have reviewed the draft report and understand that GAO has two 
primary findings related to the Vermont demonstration. GAO asserts that 
HHS did not adequately ensure that Vermont's Demonstration: a) will be 
budget neutral to the federal government; and b) maintains the fiscal 
integrity of the Medicaid program. 

While we do not believe we are in a position to comment on whether HHS 
adhered to its policies and procedures, we would like to briefly share 
our experience in working with HHS and its federal partners. 

State staff and federal staff engaged in extensive discussion and 
analysis of Vermont's historical expenditures, cost trends, caseload 
trends and program policies over the course of several months. These 
activities lead to the final budget neutrality limit established by HHS 
and accepted by Vermont. The process was both iterative and rigorous. 
Many of these discussions included staff from CMS Central Office, CMS 
Region I and the Office of Management and Budget. The discussions 
included detailed reviews of financial models, ad hoc claims analyses, 
trends analyses, and program spending reports.

As indicated previously in this letter, Vermont's decision to accept 
the aggregate budget neutrality limit was not an easy one and carried 
with it significant risk. The budget neutrality limit requires Vermont 
to implement reform strategies that enable it to stay within the 
specified spending limit. 

Additionally, the draft GAO report acknowledges that the rates paid by 
the Single State Agency to the public MCO. must be actuarially 
certified. The independent actuary established payment rates using the 
same actuarial principles used to establish payment rates for Medicaid 
MCOs throughout the country. The rates are based on historical program 
expenditures and include only those services that are eligible for 
Medicaid coverage. We believe this requirement in the Demonstration's 
Special Terms and Conditions is one example of HHS's approach to 
maintaining the fiscal integrity of the Medicaid program.

We identified technical issues related to the draft report that you may 
want to consider as you finalize the report. These issues are listed 
below: 

1. The opening section of the "Highlights" indicates that GAO found 
that "HHS did not adequately ensure that Florida's and Vermont's 
demonstrations will be budget neutral to the federal government." If 
GAO's finding is that HHS did not adhere to its internal policies for 
determining budget neutrality, it does not follow that the HHS did not 
ensure that the program will be budget neutral to the federal 
government. Absent entering into a demonstration agreement, spending 
growth is not limited to "benchmark" growth rates. Arguably, state-
specific growth rates are a more accurate predictor of future growth 
trends within a particular state. If growth rates used to establish the 
Demonstration's expenditure ceiling are below state-specific experience 
(even if these growth rates are above the "benchmarks"), aggregate 
spending likely has been held to a lower level than would have been 
spent absent the demonstration. 

2. The "Highlights" section also indicates that "HHS agreed to 
reimburse Vermont for the administration of its public managed care 
program at a rate higher than what the state received prior to its 
demonstration." We do not believe that this is accurate. The 
independent actuary relied on Vermont Medicaid's historical 
administrative expenditures in developing this component of the 
capitation rate. Second, the federal matching rate for payments to MCOs 
and the historical matching rates for administrative expenses in 
Vermont are very similar (approximately 60 percent). It also is worth 
pointing out that the Special Terms and Conditions require the public 
MCO to adhere to nearly all of the federal requirements that apply to 
private Medicaid MCOs. This requirement has resulted in the need to 
invest additional resources in the public MCO in order to meet federal 
managed care requirements. (This finding is repeated on Pages 6 and 30 
and we believe it should be revised.) 

3. On Page 28 of the draft report, GAO indicates that the demonstration 
enables the state to use federal matching funds to pay for programs 
previously funded by the state. We believe this section should be 
clarified to indicate that Federal matching funds are used to make 
capitation payments to the public MCO. Under the Global Commitment 
Demonstration, the public MCO receives an actuarially-certified, 
capitated payment in exchange for its commitment to provide all covered 
services to enrolled members. If the MCO incurs expenses in excess of 
its capitation revenues, then state-only dollars or MCO reserves will 
be used to reimburse covered services. Conversely, if the public MCO is 
able to manage service utilization through its program reform efforts, 
it will have excess revenues available to put into reserves or invest 
in the health care system. Unlike private Medicaid MCOs, the public MCO 
is restricted in how it can use excess capitation revenues. Pursuant to 
the Special Terms and Conditions, excess capitation revenues may only 
be used for specific, broad health objectives, as delineated on Page 29 
of the draft GAO report. Essentially, the public MCO is required to re- 
invest savings into the Vermont health care system, benefiting 
Demonstration enrollees, uninsured Vermonters and the overall health 
care delivery system. 

In summary, we believe the Vermont Global Commitment to Health 
Demonstration design is innovative and intended to address the 
challenges of public health care financing. We understand that true 
reform requires us to examine and evaluate program objectives broadly 
and outside the framework of traditional program policies and 
management methods. We believe the Demonstration's flexibility will 
enable Vermont to strengthen its health care delivery system while 
controlling program expenditures. 

We appreciate the opportunity to provide comment regarding the draft 
report. Please feel free to contact me if you would like to discuss any 
of our comments. 

Sincerely, 

Signed by: 

Cynthia D. LaWare, Secretary: 
Agency of Human Services: 

cc: Joshua Slen, Director, OVHA: 
Suzanne Santarcangelo, AHS: 

[End of section] 

Appendix V: GAO Contact and Staff Acknowledgments: 

GAO Contact: 

Marjorie E. Kanof, (202) 512-7114 or kanofm@gao.gov: 

Acknowledgments: 

In addition to the contact named above, Katherine M. Iritani, Assistant 
Director; Kathryn Allen; Ted Burik; Tim Bushfield; Helen Desaulniers; 
Tom Moscovitch; Hemi Tewarson; Terry Saiki; Stan Stenersen; and 
Jennifer Whitworth made key contributions to this report. 

[End of section] 

Related GAO Products: 

Medicaid Demonstration Waivers: Lack of Opportunity for Public Input 
during Federal Approval Process Still a Concern. GAO-07-694R. 
Washington, D.C.: July 24, 2007. 

Medicaid Demonstration Projects in Florida and Vermont Approved Under 
Section 1115 of the Social Security Act. B-309734. Washington, D.C.: 
July 24, 2007. 

High-Risk Series: An Update. GAO-07-310. Washington, D.C.: January 
2007. 

Medicaid Financial Management: Steps Taken to Improve Federal Oversight 
but Other Actions Needed to Sustain Efforts. GAO-06-705. Washington, 
D.C.: June 22, 2006. 

Medicaid: States' Efforts to Maximize Federal Reimbursements Highlight 
Need for Improved Federal Oversight. GAO-05-836T. Washington, D.C.: 
June 28, 2005. 

Medicaid Waivers: HHS Approvals of Pharmacy Plus Demonstrations 
Continue to Raise Cost and Oversight Concerns. GAO-04-480. Washington, 
D.C.: June 30, 2004. 

Medicaid: Improved Federal Oversight of State Financing Schemes Is 
Needed. GAO-04-228. Washington, D.C.: February 13, 2004. 

SCHIP: HHS Continues to Approve Waivers That Are Inconsistent with 
Program Goals. GAO-04-166R. Washington, D.C.: January 5, 2004. 

Medicaid and SCHIP: Recent HHS Approvals of Demonstration Waiver 
Projects Raise Concerns. GAO-02-817. Washington, D.C.: July 12, 2002. 

Medicare and Medicaid: Implementing State Demonstrations for Dual 
Eligibles Has Proven Challenging. GAO/HEHS-00-94. Washington, D.C.: 
August 18, 2000. 

Medicaid Section 1115 Waivers: Flexible Approach to Approving 
Demonstrations Could Increase Federal Costs. GAO/HEHS-96-44. 
Washington, D.C.: November 8, 1995. 

Medicaid: Statewide Section 1115 Demonstrations' Impact on Eligibility, 
Service Delivery, and Program Cost. GAO/T-HEHS-95-182. Washington, 
D.C.: June 21, 1995. 

[End of section] 

Footnotes: 

[1] Each of the 50 states and the District of Columbia, Puerto Rico, 
and four U.S. territories have Medicaid programs. 

[2] Under federal law, the states and federal government share in 
Medicaid expenditures according to a formula that provides a more 
generous federal match for states in which per capita income is lower. 
See Social Security Act § 1905(b) (codified, as amended, at 42 U.S.C. § 
1396d(b) (2000)). 

[3] Although the Secretary of HHS has delegated the administration of 
the Medicaid program, including the approval of section 1115 
demonstrations, to the Centers for Medicare & Medicaid Services (CMS), 
we refer to HHS throughout this report because section 1115 
demonstration authority ultimately resides with the Secretary and, 
accordingly, other HHS offices and agencies are involved in the review 
and approval of these demonstrations. A federal review team, including 
officials from the Office of Management and Budget, HHS's Assistant 
Secretary of Planning and Evaluation, and HHS's Health Resources and 
Services Administration, assists CMS by reviewing and commenting on 
states' proposed demonstrations. 

[4] Budget neutrality is a requirement in place through HHS policy but 
is not a statutory requirement for Medicaid demonstrations. 

[5] HHS officials consider spending limits to be a product of 
negotiations that are informed by HHS's policy to consider the state's 
historical experience and projections of growth in the President's 
Budget. HHS's benchmarks are contained in HHS's budget neutrality 
policy as described for section 1115 demonstration proposals under the 
Health Insurance Flexibility and Accountability (HIFA) initiative. HHS 
considers this policy to be applicable for all section 1115 
demonstration proposals. 

[6] A list of related GAO reports appears at the end of this report. 

[7] For more information on inappropriate supplemental payment 
arrangements, see GAO, Medicaid: Improved Federal Oversight of State 
Financing Schemes Is Needed, GAO-04-228 (Washington, D.C.: Feb. 13, 
2004). 

[8] GAO, High-Risk Series: An Update, GAO-07-310 (Washington, D.C.: 
January 2007). 

[9] Section 1115 demonstrations vary in scope, from targeted 
demonstrations limited to specific services or populations, to 
comprehensive demonstrations affecting Medicaid populations and 
services throughout a state and including most of a state's Medicaid 
expenditures. We reviewed demonstrations that were comprehensive in 
scope. For the purpose of our work, we defined comprehensive 1115 
Medicaid demonstrations as those that (1) affect beneficiaries 
statewide, (2) cover a broad range of services, and (3) include most of 
a state's Medicaid expenditures. We did not consider to be 
comprehensive demonstrations that target specific populations or 
account for a small portion of a state's total Medicaid spending. We 
also limited our review to demonstrations approved from July 2004 (when 
we last reviewed HHS-approved section 1115 demonstrations) through 
December 2006. 

[10] For example, certain beneficiaries in Florida have more options in 
selecting health care plans and benefits, but bear increased 
responsibility for ensuring that their chosen plans maintain the 
benefits and services that meet their needs; beneficiaries in Vermont 
may have their covered benefit packages increased or decreased, which 
the state can do within certain limits without prior approval from HHS. 
HHS did not provide for public input to the demonstration proposals at 
the federal level (to HHS) before approving them. See GAO, Medicaid 
Demonstration Waivers: Lack of Opportunity for Public Input during 
Federal Approval Process Still a Concern, GAO-07-694R (Washington, 
D.C.: July 24, 2007). 

[11] GAO, Medicaid Demonstration Projects in Florida and Vermont 
Approved Under Section 1115 of the Social Security Act, B-309734 
(Washington, D.C.: July 24, 2007). 

[12] We considered an explanation to be quantified in support of higher 
growth if the explanation corrected a verifiable anomaly in either the 
state's historical data or nationwide estimates of Medicaid growth. 

[13] Florida estimated this amount based on applying the agreed-upon 
limit for per person spending under the demonstration to the state's 
projected enrollment over the duration of the demonstration. 

[14] We use the term excess revenue to refer to funds remaining after 
the managed care organization has paid necessary medical and 
administrative expenses; excess revenue has also been referred to as 
savings in past reports. 

[15] See B-309734, July 24, 2007. 

[16] See Social Security Act §§ 1903(a)(1), 1905(b) (codified, as 
amended, at 42 U.S.C. §§ 1396b(a)(1), 1396d(b)). States with lower per 
capita income typically receive higher federal matching shares. 

[17] A state Medicaid plan details the populations a state's program 
serves; the amount, scope, and duration of the mandatory and optional 
services the program covers; and the rates of and methods for 
calculating payments to providers. 

[18] Section 1115 allows waivers of requirements in Medicaid and 
several other programs authorized under the Social Security Act. See 
Social Security Act § 1115 (codified, as amended, at 42 U.S.C. § 1315); 
see also section 2107(e) of the act (codified, as amended, at 42 U.S.C. 
§ 1397gg(e)) regarding the applicability of section 1115 to the State 
Children's Health Insurance Program. 

[19] HHS policy identifies three general categories of beneficiaries 
that can be included in section 1115 demonstrations: (1) mandatory 
populations, referring to beneficiaries who must be covered under a 
Medicaid state plan; (2) optional populations, referring to 
beneficiaries who can be covered under a state plan, regardless of 
whether they are covered at the time the demonstration is approved; and 
(3) expansion populations, referring to beneficiaries who cannot be 
covered under a Medicaid or SCHIP state plan, and who can only be 
covered through the Secretary's authority under section 1115. 

[20] HHS considers this guidance to apply to proposed section 1115 
demonstrations. The guidance was originally published for section 1115 
demonstrations under the HIFA initiative. HIFA section 1115 
demonstrations are intended to utilize private health insurance 
coverage options to provide coverage to more people. According to HIFA 
guidance, HHS policy is to apply the lower of state-specific 
experience--using 5 years of Medicaid data--or the President's Budget 
Medicaid baseline for the eligible groups covered by the demonstration. 

[21] Managed care organizations can reduce costs by relying on a 
primary care physician who acts as a gatekeeper for obtaining other 
health services, such as hospital or specialty medical care. 

[22] The special terms and conditions for the approved demonstrations 
in Florida and Vermont do not include a specific maintenance-of-effort 
requirement. For certain 1115 demonstration initiatives such as HIFA, 
HHS policy has not permitted states to receive additional federal 
matching payments for previously state-only heath service programs 
under the demonstrations. Federal financial participation could not be 
claimed for any existing state-funded program because, for example, the 
expectation was that the state would expand benefits through the 
demonstration. This requirement was called maintenance of effort. 

[23] Florida's demonstration also includes (1) choice counselors to 
help beneficiaries make informed decisions when selecting a Medicaid 
reform health plan, (2) a program of enhanced benefits to promote and 
reward healthy behaviors, and (3) a provision allowing Medicaid 
beneficiaries to voluntarily "opt out" of Medicaid coverage altogether 
and use a state-paid Medicaid premium toward their costs to enroll in 
an employer-sponsored insurance plan or in a commercial insurance plan 
if they are self-employed. 

[24] According to Florida officials, statewide implementation of the 
demonstration is subject to approval by the Florida State Legislature. 

[25] Florida law made the state's authority for pursuing the proposed 
demonstration contingent in part on federal approval to preserve the 
upper payment limit funding mechanism (the supplemental payment 
arrangement). Fla. Stat. Ann. § 409.91211(1)(b). 

[26] Populations not covered by the state managed care organization 
include individuals enrolled in the state's long-term care 
demonstration and SCHIP. 

[27] To receive the federal share for capitation payments made to a 
managed care organization, a state is required to enter into a contract 
with a managed care organization. See Social Security Act § 
1903(m)(2)(A)(i)(codified, as amended, at 42 U.S.C. § 
1396b(m)(2)(A)(i)). HHS approved the agreement between the Vermont 
Medicaid agency and its Office of Vermont Health Access as such a 
contract. 

[28] Vermont is also allowed to change the covered benefit package 
offered to certain groups of beneficiaries, such as nonmandatory groups 
that previously received Medicaid coverage at the state's option, 
without additional HHS approval as long as the changes result in no 
more than a 5 percent increase or decrease each year from the prior 
year's total Medicaid expenditures. 

[29] GAO, Medicaid Section 1115 Waivers: Flexible Approach to Approving 
Demonstrations Could Increase Federal Costs, GAO/HEHS-96-44 
(Washington, D.C.: Nov. 8, 1995). 

[30] GAO, Medicaid and SCHIP: Recent HHS Approvals of Demonstration 
Waiver Projects Raise Concerns, GAO-02-817 (Washington, D.C.: July 12, 
2002). 

[31] In 2006, HHS developed a plan intended to mitigate federal risk 
with respect to budget neutrality of section 1115 demonstrations, once 
approved. This plan includes HHS reviews of spending under the approved 
demonstrations to ensure that the spending complies with the 
predetermined spending ceilings that HHS approved. OMB incorporated 
these reviews as a performance metric in its 2006 assessment of 
Medicaid under the Program Assessment Rating Tool. In 2007, HHS 
required Wisconsin to return $10.2 million to the federal government 
after notifying the state that it had exceeded spending limits for its 
previously approved section 1115 demonstration. 

[32] GAO, Medicaid Waivers: HHS Approvals of Pharmacy Plus 
Demonstrations Continue to Raise Cost and Oversight Concerns GAO-04-480 
(Washington, D.C.: June 30, 2004). 

[33] See GAO-04-228. 

[34] The UPL is based on how much Medicare, the federal government's 
health care program for seniors and some disabled people, would pay for 
comparable services. 

[35] In Florida, HHS approved annual per person limits--or "per capita 
caps"--on federal Medicaid funds for services to groups of 
beneficiaries, rather than an aggregate spending limit such as 
Vermont's. Under this type of limit, spending is limited for costs per 
person, but a state does not have to accept financial risk for 
unexpected growth in enrollment. For purposes of demonstrating budget 
neutrality, Florida calculated its estimated overall spending limit by 
multiplying its per person limits by projected enrollment. 

[36] Seemingly small changes to per person cost growth rates are 
amplified by the high volume of beneficiaries that access Medicaid 
services and the number of years across which these cost growth rates 
are applied. For example, a 1 percent decrease in the per person cost 
growth rate for Florida's aged, blind, and disabled beneficiaries 
reduces projected spending under the demonstration by $879 million over 
5 years; a 1 percent decrease in the per person cost growth rate for 
families and children reduces projected spending under the 
demonstration by $1.4 billion over 5 years. 

[37] Aged, blind, and disabled beneficiaries make up a much smaller 
percentage of Florida's Medicaid population than do Medicaid-eligible 
families and children. Data provided to HHS by Florida indicate that in 
state fiscal year 2004, for example, Florida attributed about 16 
percent of Medicaid enrollment among those groups included in the 
proposed demonstration to aged, blind, and disabled beneficiaries and 
about 84 percent to families and children. 

[38] Specifically, HHS officials removed annual growth for one year in 
which per person costs decreased (-1.4 percent growth), and replaced it 
with annual growth based on the average growth from two surrounding 
years (7.9 percent growth). 

[39] Specifically, HHS's guidance says that states, in supporting 
budget neutrality, should submit to HHS 5 years of historical data and 
that spending limits are based on growth rates that are the lower of 
state-specific history or estimates of nationwide growth for the 
beneficiary groups in the demonstration. However, HHS's guidance does 
not specifically define state-specific historical rates as calculated 
over 5 years. 

[40] In addition, HHS required Vermont to remove the projected costs 
for certain populations from an ongoing section 1115 demonstration. 
Vermont will be at risk for the cost of these populations. 

[41] The state initially proposed including in the demonstration all 
beneficiaries statewide, including beneficiary groups from several 
other demonstrations and SCHIP. HHS required Vermont to continue to 
operate a long-term care demonstration and SCHIP independent of the 
proposed demonstration, and to remove those beneficiary groups from its 
spending base. On the other hand, HHS allowed Vermont to include 
beneficiaries from a demonstration that provided services to people 
with developmental disabilities as well as certain beneficiary groups 
from an ongoing section 1115 demonstration. 

[42] As with Florida, changes in growth rates are amplified by the 
number of beneficiaries accessing Medicaid services and the number of 
years across which these growth rates are applied. For example, 
reducing the enrollment growth rate for families and children from 1.99 
percent to the benchmark level of 1.05 percent reduces Vermont's 
spending limit by more $32 million over 5 years. 

[43] See GAO-02-817. 

[44] Under the demonstration, funds from the low-income pool may be 
used for health care expenditures (medical care costs or premiums) 
incurred by "the State, by hospitals, clinics, or by other provider 
types for uncompensated medical care costs of medical services for the 
uninsured, Medicaid shortfall. . . and may include premium payments, 
payments for provider access systems (PAS) and insurance products for 
such services provided to otherwise uninsured individuals, as agreed 
upon by the State and CMS." 

[45] These reviews were conducted by HHS funding specialists hired in 
2004 and 2005 and resulted in a product known as a state funding 
profile. The funding profile documents the state Medicaid program's 
organizational structure, programmatic structure, and budget process 
and is made available to HHS staff for oversight and informational 
purposes. According to HHS, the funding specialists are hired using 
Health Care Fraud and Abuse Control Program funds. 

[46] Section 118 of the Special Terms and Conditions governing 
Florida's demonstration allows HHS to adjust the spending limit under 
several circumstances. Specifically, "The CMS reserves the right to 
adjust the budget neutrality ceiling to be consistent with enforcement 
of impermissible provider payments, health care related taxes, new 
Federal statutes, or policy interpretations implemented through SMD 
[state Medicaid director] letters, other memoranda on or regulations." 

[47] Over the course of the 5-year demonstration, we estimated that 
excess administrative reimbursement could total up to $71 million in 
part because Vermont was allowed under the demonstration to receive a 
higher-than-traditional administrative reimbursement level. Under 
traditional administrative claiming, Vermont would receive about $157 
million based on $4.3 billion in payments to the managed care 
organization (assuming 7.3 percent administrative claims matched at 50 
percent). Under the terms of its demonstration, Vermont could claim as 
much as $228 million (assuming 9 percent administrative claims matched 
at 58.93 percent, its 2007 federal matching rate). 

[48] See Health Management Associates, Vermont Global Commitment 
Independent Review and Risk Analysis, September 2005. 

[49] For more information on state recycling of federal Medicaid funds, 
see GAO-04-228. 

[50] B-309734, July 24, 2007. 

[51] Budget neutrality is a requirement in place through HHS policy but 
is not a statutory requirement for Medicaid demonstrations. 

[52] GAO-04-480. 

[53] Department of Health and Human Services, Centers for Medicare & 
Medicaid Services, Section 1115 Demonstrations, State Profiles: 
Approvals through February 28, 2006 (Washington, D.C.: 2006). 

[54] We accessed HHS guidance on budget neutrality on March 23, 2007, 
from CMS's Web site: [hyperlink, 
http://www.cms.hhs.gov/HIFA/02_Guidelines.asp]. 

[55] B-309734, July 24, 2007. 

[56] The funding specialist profile was established in 2005 and 
designed as an educational tool to assist the newly hired funding 
specialists in understanding the Medicaid State programs to which they 
were assigned. This tool included identification of budget, program, 
and financing. All funding specialists were required to complete these 
profiles, but such profiles were never considered to represent a formal 
Agency position and they were never formally released to States. 

GAO's Mission: 

The Government Accountability Office, the audit, evaluation and 
investigative arm of Congress, exists to support Congress in meeting 
its constitutional responsibilities and to help improve the performance 
and accountability of the federal government for the American people. 
GAO examines the use of public funds; evaluates federal programs and 
policies; and provides analyses, recommendations, and other assistance 
to help Congress make informed oversight, policy, and funding 
decisions. GAO's commitment to good government is reflected in its core 
values of accountability, integrity, and reliability.  

Obtaining Copies of GAO Reports and Testimony: 

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each 
weekday, GAO posts newly released reports, testimony, and 
correspondence on its Web site. To have GAO e-mail you a list of newly 
posted products every afternoon, go to [hyperlink, http://www.gao.gov] 
and select "Subscribe to Updates."  

Order by Mail or Phone: 

The first copy of each printed report is free. Additional copies are $2 
each. A check or money order should be made out to the Superintendent 
of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or 
more copies mailed to a single address are discounted 25 percent. 
Orders should be sent to:  

U.S. Government Accountability Office: 
441 G Street NW, Room LM: 
Washington, D.C. 20548:  

To order by Phone: 
Voice: (202) 512-6000: 
TDD: (202) 512-2537: 
Fax: (202) 512-6061:  

To Report Fraud, Waste, and Abuse in Federal Programs:  

Contact:  

Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: 
E-mail: fraudnet@gao.gov: 
Automated answering system: (800) 424-5454 or (202) 512-7470:  

Congressional Relations:  

Ralph Dawn, Managing Director, dawnr@gao.gov: 
(202) 512-4400: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7125: 
Washington, D.C. 20548:  

Public Affairs: 

Chuck Young, Managing Director, youngc1@gao.gov: 
(202) 512-4800: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7149: 
Washington, D.C. 20548: