Ryan White Care Act of 1990: Opportunities to Enhance Funding Equity
(Letter Report, 11/13/95, GAO/HEHS-96-26).

Pursuant to a congressional request, GAO reviewed the funding formulas
established under the Ryan White Care Act, focusing on: (1) whether the
existing formulas distribute funds equitably to states and eligible
metropolitan areas (EMA); (2) the factors that inhibit greater funding
equity; and (3) formula changes that are needed to improve funding
equity.

GAO found that: (1) although Ryan White Care Act funding formulas
include factors used in equity-based formulas, they result in per-case
funding discrepancies because EMA cases are double counted; (2) states
without EMA do not benefit from double counting and receive
significantly less funding than states with EMA; (3) the indicators used
to target funds to needy states and EMA fail to take geographic cost
differences into consideration; (4) EMA funding levels are based on the
cumulative number of reported acquired immunodeficiency syndrome (AIDS)
cases, resulting in the oldest EMA receiving the most funding; (5)
better cost indicators could be used to target more funds to states and
EMA where resources are the most strained by AIDS; and (6) funding
equity could be improved by eliminating the inappropriate double
counting of AIDS cases and by using more appropriate measures of EMA and
state funding needs.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  HEHS-96-26
     TITLE:  Ryan White Care Act of 1990: Opportunities to Enhance 
             Funding Equity
      DATE:  11/13/95
   SUBJECT:  Acquired immunodeficiency syndrome
             Funds management
             Health care services
             Health care costs
             Public health legislation
             Formula grants
             Health care programs
             Federal aid to states
             Intergovernmental fiscal relations
             Health resources utilization
IDENTIFIER:  Medicare Hospital Wage Cost Index
             Dallas (TX)
             Oakland (CA)
             New York (NY)
             San Francisco (CA)
             Miami (FL)
             AIDS
             
**************************************************************************
* This file contains an ASCII representation of the text of a GAO        *
* report.  Delineations within the text indicating chapter titles,       *
* headings, and bullets are preserved.  Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are           *
* identified by double and single lines.  The numbers on the right end   *
* of these lines indicate the position of each of the subsections in the *
* document outline.  These numbers do NOT correspond with the page       *
* numbers of the printed product.                                        *
*                                                                        *
* No attempt has been made to display graphic images, although figure    *
* captions are reproduced. Tables are included, but may not resemble     *
* those in the printed version.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


Cover
================================================================ COVER


Report to Congressional Requesters

November 1995

RYAN WHITE CARE ACT OF 1990 -
OPPORTUNITIES TO ENHANCE FUNDING
EQUITY

GAO/HEHS-96-26

Ryan White Funding Formulas

(118109)


Abbreviations
=============================================================== ABBREV

  AIDS - acquired immunodeficiency syndrome
  CARE - Comprehensive AIDS Resources Emergency
  CDC - Centers for Disease Control and Prevention
  EMA - eligible metropolitan area
  GSP - Gross State Product
  HHS - Department of Health and Human Services
  HIV - human immunodeficiency virus
  HCFA - Health Care Financing Administration
  MHWC - Medicare Hospital Wage Cost (Index)
  PCI - per capita income
  PPS - prospective payment system
  SPNS - Special Projects of National Significance
  TTR - total taxable resources

Letter
=============================================================== LETTER


B-265841

November 13, 1995

The Honorable Nancy Kassebaum
Chairman, Committee on Labor
 and Human Resources
United States Senate

The Honorable Hank Brown
United States Senate

The human immunodeficiency virus (HIV) epidemic has become one of the
most serious health threats to the American public.  The HIV
infection rate is estimated to be as high as 1 in every 250 persons
nationwide.  Metropolitan areas are especially affected by HIV with
rates as high as 1 in 25.  By December 1994, nearly 450,000 people
with HIV had been reported to have progressed to acquired
immunodeficiency syndrome (AIDS), and more than 270,000 had been
reported to have died of the disease.  Currently, only one federal
program provides funds specifically for medical and support services
to individuals with AIDS and HIV.  Established by the Ryan White
Comprehensive AIDS Resources Emergency (CARE) Act of 1990, the
program distributed more than $579 million in fiscal year 1994 to
eligible metropolitan areas (EMA) through title I of the act and to
states through title II. 

Citing examples of disparities in per-case funding, you expressed
concerns that the existing title I and II funding formulas may lead
to an inequitable distribution of funds to states and EMAs.  For this
reason, you asked us to determine (1) how equitably the existing
formulas are distributing title I and II funds, (2) which factors
inhibit the formulas from achieving greater equity, and (3) what
formula changes could improve equity. 

To assess the title I and II formulas, we reviewed the enacting
legislation and conducted interviews to examine the basic rationale
for the factors used in the current CARE Act formulas.  We compared
the funding distributions resulting from the existing formulas
against two widely recognized equity criteria. 

The first criterion--beneficiary equity--considers the degree to
which a formula allocates funds to ensure that EMAs and states are
able to purchase a comparable level of services for their HIV
populations.  Under this criterion, dollars would be distributed
according to two indicators:  (1) the potential number of people with
AIDS (that is, caseload) and (2) the cost of providing services.  The
second criterion--taxpayer equity--considers the degree to which EMAs
and states are able to finance a comparable level of services with
comparable burdens on their taxpayers.  This second standard is
broader than the first one.  In addition to including the two
indicators used in the first standard (caseload and cost), it uses a
measure of each EMA's and state's capacity to fund AIDS and HIV
services from its own resources.\1

On the basis of interviews with experts and a review of available
literature, we identified data with which to apply these standards to
assess the equity of the title I and II formulas.\2

We used Centers for Disease Control and Prevention (CDC) data to
develop a proxy measure of people living with AIDS, Health Care
Financing Administration (HCFA) data to measure differences in
service costs, and Department of Treasury data to determine fiscal
capacities.  We used regression analysis to determine how closely the
distribution of CARE Act funds reflected our equity standards. 

We conducted our work from February 1994 through October 1995 in
accordance with generally accepted government auditing standards. 


--------------------
\1 See the bibliography for other studies that describe these
criteria. 

\2 The experts we interviewed were officials of the Agency for Health
Care Policy and Research, the Centers for Disease Control and
Prevention, the Health Care Financing Administration, and the Health
Resources and Services Administration. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

Although the title I and II funding formulas currently include some
factors used in equity-based formulas, they result in wide variations
in per-case funding for several reasons.  The most important is that
EMAs' AIDS cases are double counted; once when calculating funding
for EMA medical services under title I and again when calculating
funding for state medical and other services under title II.  A state
without an EMA does not benefit from double counting and receives
substantially less funding than a state with an EMA. 

In addition to double counting, the indicators used to target funds
to high-need states and EMAs are inaccurate and could be improved. 
For example, we found evidence of large geographic differences in the
cost of serving AIDS patients, yet neither the title I nor the title
II formula takes these differences into account.  Also, the title I
formula targets funding to EMAs on the basis of their cumulative
number of reported AIDS cases, yet over 60 percent of these reported
cases have since died.  As a result, the oldest EMAs receive the most
generous funding, and newly emerging EMAs with more recent growth in
AIDS cases receive less funding. 

Finally, both title I and title II attempt to target more funds to
states and EMAs where resources are most strained by the epidemic. 
However, the indicators used in both formulas are inadequate.  The
indicator used in the title I formula (AIDS incidence rate) does not
directly take into account an EMA's tax base, which determines the
area's own capacity to provide services to people with HIV.  The
opposite is true for the title II formula.  A tax base indicator is
used (state per capita income), but the effect of a high AIDS
incidence rate on the tax base is ignored. 

Funding equity can be improved by changing the structure of the two
formulas to eliminate the inappropriate double counting of AIDS cases
and by using more appropriate measures of EMA and state funding
needs. 


   BACKGROUND
------------------------------------------------------------ Letter :2

The Congress enacted the Ryan White CARE Act on August 18, 1990, to
"improve the quality and availability of care for individuals and
families with HIV disease." The CARE Act makes funds available
through four titles to states, EMAs, and nonprofit entities for
developing, organizing, coordinating, and operating more effective
and cost-efficient service delivery systems.  The Health Resources
and Services Administration, part of the Department of Health and
Human Services' U.S.  Public Health Service, administers the program. 

Over $579 million in CARE Act funds were appropriated in fiscal year
1994 for services to people with AIDS and HIV.  About $326 million
(56 percent) of these funds were appropriated for title I, which
provides "emergency assistance" to EMAs--metropolitan areas
disproportionately affected by the HIV epidemic.  Half of title I
funds are distributed by formula, and half are distributed
competitively.  To be eligible, a metropolitan area must have a
cumulative count of more than 2,000 cases of AIDS since reporting
began in 1981 or a cumulative count of AIDS cases that exceeds
one-quarter of 1 percent of its population.  In fiscal year 1994,
there were a total of 34 EMAs in 17 states, the District of Columbia,
and Puerto Rico.  Since fiscal year 1991, the number of EMAs has more
than doubled. 

For title II, $184 million (32 percent of total CARE Act funds) were
appropriated in fiscal year 1994.  Title II provides funds to states
to improve the quality, availability, and organization of health care
and support services for people with HIV.  Of the title II funds
distributed to the states in fiscal year 1994, 90 percent were
distributed by formula, and 10 percent were distributed competitively
through Special Projects of National Significance.\3

The remaining titles--titles IIIb and IV--were funded at about $48
million (8 percent) and $22 million (4 percent), respectively, in
fiscal year 1994.  Title IIIb funds are intended for early
intervention programs, and title IV funds are intended for pediatric
AIDS programs.  Under both of these titles, funds are awarded
competitively. 


--------------------
\3 Special Projects of National Significance (SPNS) are competitive
grants awarded to public and nonprofit organizations to advance
knowledge and skills in the delivery of health and support services
to people with HIV. 


   THE EXISTING FORMULAS MEET
   NEITHER EQUITY CRITERION
------------------------------------------------------------ Letter :3

Our examination of the existing title I and II formulas indicates
that neither formula meets the beneficiary and taxpayer equity
criteria.  Per-case funding is not systematically related to either
EMA or state service costs or their fiscal capacity.  (See app.  II
for details of our analysis.)


      THE TITLE I FORMULA DOES NOT
      MEET EITHER EQUITY CRITERION
---------------------------------------------------------- Letter :3.1

The title I formula does not meet the beneficiary equity criterion
because per-case funding is not systematically related to the cost of
treating people with HIV.  Specifically, our analysis of fiscal year
1994 funding for EMAs showed that per-case funding ranged from $818
to $2,663--a difference of over 200 percent.  However, only 10
percent of this variation was related to cost differences\4 --though
the cost differences themselves were significant.  As an
illustration, the Dallas and Oakland EMAs each received title I
allocations of approximately $1,200 per person with AIDS, but the
cost of providing health care services in Oakland is about 37 percent
higher than in Dallas. 

The title I formula also does not meet the taxpayer equity criterion
because, in addition to not being systematically related to cost
differences, EMA grant amounts are not highly related to the EMAs'
fiscal capacity.  Our analysis of fiscal year 1994 funding for all
EMAs showed that more than 40 percent of the variation in EMAs'
per-case funding was unrelated to differences in cost and fiscal
capacity.  For example, the Dallas and Oakland EMAs received about
the same per-case funding, but Oakland's funding capacity when
measured in terms of its tax base, costs, and concentration of AIDS
cases is about 17 percent lower than that of Dallas. 


--------------------
\4 The two EMAs located in Puerto Rico--Ponce and San Juan--were
excluded from this analysis.  With these EMAs included in the
analyses, cost differences account for only 2 percent of the
variation in per-case funding. 


      COMBINED TITLE I AND II
      FUNDING MEETS NEITHER EQUITY
      CRITERION
---------------------------------------------------------- Letter :3.2

The distribution of combined title I and II funds across states does
not meet either the beneficiary or the taxpayer equity criterion.\5
Total per-case funding for California and New York is about 20
percent and 30 percent above the national average, respectively,
while Hawaii, Ohio, and Vermont have total per-case funding levels
about 50 percent below the national average.  These funding
differences are not strongly related to differences in states' costs
and fiscal capacity to provide services.  Our statistical analysis
found that differences in service costs and fiscal capacity account
for 33 percent of these differences in per-case funding.\6 That is,
67 percent of the variation in state funding per AIDS case is
unrelated to states' funding needs.  (See app.  II for details.)


--------------------
\5 We compared the total amount of title I and II funds within each
state to assess interstate funding equity.  We did not conduct a
separate assessment of title II funds because of the difference in
the purposes of these funds between states with and without EMAs. 

\6 To develop a more valid estimate, we excluded from our analysis
those states that received the minimum title II grant amount of
$100,000. 


   FUNDING INEQUITIES RESULT FROM
   THE STRUCTURE AND COMPONENTS OF
   THE FORMULAS
------------------------------------------------------------ Letter :4

Several features of the title I and II formulas contribute to the
funding inequities we have identified.  Specifically, inequities
occur because EMA cases are counted in both the title I and II
formulas, an inappropriate caseload measure is included in the title
I formula, an inappropriate measure of EMAs' and states' fiscal
capacity is included in both formulas, and neither formula includes a
measure of EMAs' and states' service costs.  (See appendixes for
details.)


      DOUBLE COUNTING EMA CASES
---------------------------------------------------------- Letter :4.1

Our analysis of differences in states' per-case funding amounts
indicates that about half of the variation is due to the double
counting of EMA cases in both the title I and II formulas rather than
differences in funding needs (that is, cost or fiscal capacity
differences).  States where most cases live in EMAs receive the
largest amounts per case, since larger proportions of their caseloads
are double counted.  For example, per-case funding was about $1,100
in states without an EMA, $1,700 in states where less than half the
state's caseload lived in an EMA, and $2,200 in states where more
than half of the caseload lived in an EMA (see fig.  1).  Thus, most
of the variation in per-case funding can be explained by the extent
to which a state's caseload is double counted rather than by the
state's funding needs. 

   Figure 1:  State Funding by
   Proportion of AIDS Cases
   Residing in an EMA

   (See figure in printed
   edition.)


      INAPPROPRIATE TITLE I
      CASELOAD MEASURE
---------------------------------------------------------- Letter :4.2

The title I caseload measure is based on the cumulative number of
people with AIDS that EMAs reported to CDC since 1981 when reporting
began.  By the end of 1993, however, two-thirds of these people had
been reported to have died and were, therefore, no longer using
services funded by title I. 

Because the formula includes deceased persons, the EMAs that
experienced the first outbreak of AIDS receive substantially more
per-case funding than do newer EMAs.  For example, in fiscal year
1994, the 18 EMAs that were eligible to receive title I funds in the
first 2 years of eligibility--1991 and 1992--were funded at about
$1,500 per case.  In contrast, the 16 EMAs that became eligible in
1993 and 1994 were funded at only about $1,000 per case--one-third
less than the older EMAs (see fig.  2). 

   Figure 2:  Fiscal Year 1994
   Funding per Case

   (See figure in printed
   edition.)

Note:  Early EMAs are the 18 that entered the program during 1991 and
1992; later EMAs are the 16 that entered in 1993 and 1994. 


      ABSENCE OF A COST MEASURE
---------------------------------------------------------- Letter :4.3

While the cost of providing AIDS and HIV services varies among EMAs
and states, neither the title I nor title II formula includes a
factor to measure those differences.  Information on the actual costs
of providing health and support services to people with AIDS and HIV
within different geographic areas is not available.  However, most of
the delivery costs appear to be associated with the personnel who
provide the labor-intensive outpatient health, support, and case
management services titles I and II primarily fund.  A proxy measure
for these labor costs is available through the Medicare Hospital Wage
Cost Index.\7

Using this index for title I cities, we estimated that the cost of
providing medical services was about 30 percent above the national
average in the New York, Oakland, and San Francisco EMAs and about 10
percent below the national average in the Miami EMA--a difference of
about 40 percent.  This suggests that the New York, Oakland, and San
Francisco EMAs must spend much more than the Miami EMA to provide a
comparable level of services to their patients.  Similarly, under
title II, we estimated that the cost of providing medical services
was more than 15 percent above the national average in the states of
Alaska, California, and New York, about 15 percent below the national
average in Alabama and Arkansas, and about 20 percent below the
national average in Mississippi. 


--------------------
\7 The Medicare Hospital Wage Cost Index was designed to reflect
personnel costs in hospitals subject to the Medicare prospective
payment system; it was derived from hospital salary surveys. 


      INAPPROPRIATE FISCAL
      CAPACITY MEASURES
---------------------------------------------------------- Letter :4.4

State and EMA fiscal capacity depends on the size of the tax base and
the service demands placed on that tax base.  The current title I
formula measures the demand for services through the use of an AIDS
incidence rate factor, but the strength of each EMA's tax base is not
included.  As a result, the title I formula does not adequately
adjust EMAs' allocations to target those with smaller tax bases and
fewer resources to draw upon to meet the needs of the cases they must
serve. 

The title II formula does measure the strength of each state's tax
base through the use of per capita personal income.  However, it does
not consider the demand for services that is placed on state tax
bases.  As a result, the title II formula does not adequately adjust
state allocations to target states with tax bases that are burdened
by a heavy demand for services. 


   GREATER FUNDING EQUITY CAN BE
   ACHIEVED
------------------------------------------------------------ Letter :5

Greater funding equity can be achieved by changing the formulas'
structure and components.  The formulas can be modified to make their
funding distribution meet either the beneficiary equity criterion or
the taxpayer equity criterion.  Alternatively, although no formula
can completely satisfy both criteria simultaneously, the formulas
could be modified to partly meet both criteria, emphasizing
beneficiary equity over taxpayer equity or vice versa.  Regardless of
which criterion is emphasized, however, the following changes could
make the title I and II formulas more equitable.  (See appendixes for
details.)


      INEQUITIES CREATED BY DOUBLE
      COUNTING COULD BE AVOIDED
---------------------------------------------------------- Letter :5.1

The current title I and II structure could be revised to avoid
inequities created by counting EMA cases in both formulas. 
Presently, funding for titles I and II does not always reflect the
division of service responsibilities between EMAs and state
governments.  Through title I, EMAs provide medical and support
services to people who reside in their areas of coverage.  Through
title II, states provide medical and support services to people
living outside EMAs and commonly provide these services to people
living in EMAs as well.  In addition, through title II, states
administer services such as medication assistance and insurance
continuation statewide for cases both within and outside of their
EMAs.\8 Nonetheless, while EMAs typically provide the bulk of medical
services to people living within their areas, title II provides
funding as if states were providing both medical and statewide
services to the EMA cases.  This results in a higher level of
per-case funding for states with EMAs because the EMA cases are
double counted. 

A more equitable structure would, in effect, double count all cases. 
Cases would be counted once for the statewide services such as
medication assistance and insurance continuation, and again for
medical services that are jointly provided by states and EMAs. 

One means for achieving this would be to make separate appropriations
for the major activities funded by the CARE Act.  One appropriation
would be made for services that state governments provide statewide,
and a second appropriation would be made for medical services that
are jointly provided by states and their EMAs (see fig.  3).\9

   Figure 3:  Proposed Structure

   (See figure in printed
   edition.)

Funding for statewide services would be allocated to state
governments on the basis of each state's total AIDS caseload. 
Funding for medical services would be divided into two separate
allocations for state governments and EMAs.  The allocation to state
governments would be based on AIDS cases living outside a state's
EMAs.  The allocation to EMAs would be based on AIDS cases living in
their service delivery areas.  With this method, each state's entire
caseload is counted twice:  once for funding statewide services and
again for funding state-EMA medical services. 

The approach would only be a means of allocating federal funds to the
entities responsible for delivering services and would not change the
latitude currently afforded local governments and states in deciding
how to best use those funds.  Consequently, this approach should have
only a minimal effect on existing service delivery structures because
it leaves EMA and state responsibilities essentially unchanged. 


--------------------
\8 In some cases, EMAs provide funds to the state for the statewide
medication assistance program. 

\9 A third appropriation could be made for discretionary purposes
(that is, for competitive grants, evaluations, and technical
assistance). 


      BETTER INDICATORS OF
      CASELOAD, COST, AND CAPACITY
      WOULD IMPROVE EQUITY
---------------------------------------------------------- Letter :5.2

In addition to changing the structure of the formulas, funding equity
could be improved by changing the formulas' components. 
Specifically, funding equity could be improved by modifying the
existing caseload and fiscal capacity measures, and by including a
cost measure. 

First, funding equity could be improved by including a caseload
measure that better reflects the number of people living with AIDS
and excludes deceased persons.  We have developed a proxy measure of
people living with AIDS from existing CDC data. 

Funding equity could also be improved by including a cost measure,
such as the Medicare Hospital Wage Cost Index.  Use of such a measure
would better compensate the EMAs and states that must pay more to
provide services to their patients because of their higher private
sector health care costs. 

Finally, to increase resources in states and EMAs with poorer fiscal
capacity, the current fiscal capacity factors could be revised to
better measure the EMAs' and states' AIDS incidence rates and tax
bases.  Currently, the title I fiscal capacity factor lacks a measure
of EMAs' tax bases, and the title II factor lacks a measure of
states' AIDS incidence rates.  By having more complete measures of
EMA and state fiscal capacities, the formulas could adjust grants on
the basis of both the demand for services and the strength of tax
bases.  In addition, using total taxable resources (TTR) in the state
formula instead of personal income could result in a more
comprehensive measure of state tax bases.  (For the effects of these
changes on specific state and EMA grants, see app.  V.)


   CONCLUSIONS
------------------------------------------------------------ Letter :6

Our analysis of the existing formulas demonstrates that federal
funding under titles I and II of the CARE Act can be made more
equitable.  An important purpose of the Ryan White CARE Act was to
target emergency funding to areas of greatest need.  At the time the
law was enacted, high incidences of HIV were found in fewer areas of
the country, service delivery networks were just beginning to form,
and these service delivery systems had to rely primarily on private
and volunteer resources.  In the past 5 years, however, the HIV
epidemic has become more widespread and less localized.  Hence, areas
where the AIDS caseload has burgeoned recently need per-case funding
levels comparable to those in areas where AIDS was initially
concentrated. 


   RECOMMENDATIONS TO THE CONGRESS
------------------------------------------------------------ Letter :7

To achieve greater equity in the distribution of funds, we recommend
that the Congress modify the funding formulas to

  reduce the double counting of EMA cases so that comparable medical
     services funding is available for people with AIDS, regardless
     of where they live,

  adopt a caseload indicator that better reflects the number of
     people living with AIDS who are in need of services, and

  include an indicator that reflects the relative differences across
     states and EMAs in the cost of serving people with AIDS. 


   MATTERS FOR CONGRESSIONAL
   CONSIDERATION
------------------------------------------------------------ Letter :8

If the Congress wishes to target more aid to states and EMAs with
limited fiscal capacity, then it may consider adopting an indicator
that reflects the relative strength of local tax bases and
concentrations of people with AIDS.  Alternatively, the Congress may
wish to discontinue the use of AIDS incidence rates in the title I
formula and per capita income in the title II formula because of the
funding inequities that these components produce. 

Finally, modifying the formulas to achieve a more equitable
distribution of funds will involve significant changes in grants to
some EMAs and states.  To avoid possible disruption of service
delivery, the Congress may wish to consider phasing in formula
modifications.  This should minimize, if not avoid, disruption for
the service delivery networks the CARE Act has made possible over the
last 5 years. 


---------------------------------------------------------- Letter :8.1

If you or your staff have any questions regarding this report, please
contact me on (202) 512-7119 or Jerry Fastrup, Assistant Director, on
(202) 512-7211.  Major contributors to this report are listed in
appendix VII. 

William J.  Scanlon
Associate Director,
Health Financing Issues


ASSESSING THE EQUITY OF THE
FUNDING FORMULAS
=========================================================== Appendix I

To determine how equitably title I and II funds are distributed, we
examined the existing formulas, applied two widely recognized equity
criteria, and determined whether the existing or alternative formula
factors would best allocate funds according to these standards. 


   EXISTING FORMULAS
--------------------------------------------------------- Appendix I:1

Title I funds are distributed on the basis of the cumulative number
of AIDS cases EMAs report and their cumulative AIDS incidence rate. 



where



In this formula,

Casesi = the cumulative number of AIDS cases in the ith EMA,

REMA = the per capita incidence of cumulative AIDS cases in an EMA,

RAll EMAs = the per capita incidence of cumulative AIDS cases in all
EMAs. 

Note:  i ranges over all EMAs. 

Title II funds are distributed to states on the basis of the number
of AIDS cases they reported in the 2 most recent fiscal years and
their per capita income. 



where



In this formula,

Casesi = the number of cases reported by the ith state in the 2 most
recent fiscal years,

PCI = the average per capita income of the ith state/the United
States. 

Note:  i ranges over all states, the District of Columbia, and
territories. 


   EQUITY CRITERIA
--------------------------------------------------------- Appendix I:2

The two standards of equity that we applied were the beneficiary and
taxpayer equity criteria.  To meet the beneficiary equity criterion,
funding should be distributed in a way that enables EMAs and states
to purchase comparable levels of AIDS and HIV medical and support
services.  In other words, per-case funding should be about the same
in each of the EMAs and states after adjusting for cost differences. 

The formula for producing a funding distribution that meets the
beneficiary equity criterion is



In this formula,

Casesi = the number of people in need of services in the ith EMA or
state,

Cost Indexi = an index measuring relative differences in the per-case
cost of serving recipients in the ith EMA or state,

A = the total amount of funds to be allocated. 

Note:  i ranges over all EMAs/the District of Columbia, all states
and territories. 

To meet the taxpayer equity criterion, funding should be distributed
in a way that enables EMAs and states to purchase comparable levels
of AIDS and HIV services with comparable burdens on their taxpayers. 
Therefore, under this criterion, per-case funding should be about the
same in each of the EMAs and states, once adjusted for differences in
their service costs and fiscal capacities.  Per-case funding should
only differ to the extent that costs and fiscal capacities do.  The
formula for producing a funding distribution that meets the taxpayer
equity criterion is



In this formula, cases and costs are the same as in the beneficiary
equity formula and represent an EMA's or state's funding need.  The
federal percentage represents the share of an EMA's or state's
funding need that will be counted in the formula and varies with
EMAs' and states' fiscal capacity according to the following
formula:\10



The fiscal capacity index represents the ability of grantees to fund
services from state and local resources.  We applied a weight of 0.20
to this index because that is the weight implicitly applied to fiscal
capacity through the AIDS incidence rate found in the existing title
I formula.\11

As shown in the preceding figures, to meet the beneficiary equity
standard, the funding formula would base its allocation on states' or
EMAs' cases and costs, and to meet the taxpayer equity standard, the
formula would also include a fiscal capacity factor.  Hence, in
determining whether the formulas distribute title I and II funds in
accordance with the beneficiary and taxpayer equity criteria, we
sought indicators that were reflective of these three factors and
were appropriate for use in grant allocation formulas. 


--------------------
\10 Equalizing taxpayer burdens requires the mathematical form shown
here.  See Maternal and Child Health:  Block Grant Funds Should Be
Distributed More Equitably (GAO/HRD-92-5, Apr.  2, 1992), pp.  55-62,
for a more detailed discussion and derivation of this formula. 

\11 The existing title II fiscal capacity factor does not adjust
states' per capita income by their AIDS incidence, which results in a
perverse allocation of title II funds.  Because of this, we did not
use the weight associated with the existing title II fiscal capacity
factor. 


   CASELOADS
--------------------------------------------------------- Appendix I:3

We considered four approaches to estimating the number of people
living with AIDS in each of the EMAs and states: 

  cumulative AIDS cases,

  AIDS cases less reported deaths,

  AIDS cases reported in the 2 most recent years, and

  weighted AIDS cases. 

The first approach--cumulative AIDS cases--is the caseload measure
found in the current title I formula.  In the context of our equity
criteria, this approach assumes that the number of people currently
living with AIDS can be estimated by using the cumulative number of
AIDS cases reported since 1981.  About 66 percent of these AIDS cases
are no longer living, however, and the likelihood of death increases
substantially the longer one has AIDS.  As a result, this measure
would direct funds more to where the epidemic occurred initially
rather than to where it appeared more recently. 

The second approach--AIDS cases less reported deaths--subtracts each
state's and EMA's total reported deaths from their total reported
AIDS cases for the 10 most recent years.  The total number of living
cases is then determined by adding each year's surviving cases. 

While this approach appears to potentially provide a reasonable
estimate of the number of people living with AIDS, it is not an
appropriate caseload measure for allocating funds.  Our interviews
with experts and our review of the literature indicated that this
estimate would be biased because AIDS-related deaths are more
extensively and quickly reported in some states and EMAs than in
others, and this results in measurement errors.  Furthermore, since
funds are based on the number of people living with AIDS, those
states and EMAs that underreport AIDS-related deaths would be
rewarded, while others with more reliable reporting would, in effect,
be penalized.  Many of the experts we interviewed expressed concerns
that this method could introduce incentives to purposely underreport
deaths.  Consequently, states and EMAs might delay or not even report
these deaths, which could lead to another bias to the caseload
measure and result in less reliable information on the lifespans of
people with the disease. 

The third approach that we considered uses the number of AIDS cases
reported in the most recent 2 years to estimate the number of living
AIDS cases.  This is the caseload measure currently used in the title
II formula, and it appears to reasonably estimate the number of
living cases.  However, because this measure consists of cases from a
narrow time frame, we believe it could be too sensitive to sudden
caseload changes and disrupt the continuity of funding over time. 
Also, the expected lifespans for people with AIDS could increase over
time.  If this occurs in the future, the cases reported in a 2-year
interval may not accurately reflect the number of people living with
the disease. 

The final approach--weighted AIDS cases--is a proxy measure of living
AIDS cases.  This approach estimates the number of AIDS cases living
in an EMA or state on the basis of on the number of AIDS cases
reported to CDC for each of the most recent 10 years and national
average survival rates since a case was first reported. 
Specifically, the number of AIDS cases that an EMA or a state had
reported for each of these 10 years would be weighted by the national
percentage of cases estimated to be living as of the first day of the
most recent year of that period.\12 These percentages would be
estimated from national data on the number of people reported to have
AIDS during a 10-year period who had not been reported to have died
of the disease. 

Table I.1 shows the cumulative survival rates for each of 10 years as
of fiscal year 1992.  According to these data, 88 percent of the
cases reported in 1992 were estimated to have survived at least 1 day
in that year, and 57 percent of the prior year's cases were estimated
to still be alive as of that date. 



                         Table I.1
          
          Cumulative AIDS Survival Rates for a 10-
                        Year Period

                                                Proportion
                                                 surviving
Year reported with AIDS                       into FY 1992
------------------------------------  --------------------
FY 1992                                                .88
FY 1991                                                .57
FY 1990                                                .37
FY 1989                                                .24
FY 1988                                                .16
FY 1987                                                .10
FY 1986                                                .08
FY 1985                                                .06
FY 1984                                                .06
FY 1983                                                .06
----------------------------------------------------------
This approach appears to be the most appropriate.  Unlike the
cumulative AIDS cases measure, it has been adjusted to account for
people with AIDS who are no longer living and thus better reflects
the intended service population.  In contrast to the second approach,
this one averages out differences in reporting mortality and avoids
incentives to underreport deaths.  Specifically, since the algorithm
for estimating living cases would be based on national data, any
uniqueness in how states and EMAs report mortality would not affect
the amount of funds that they would receive.  Finally, this measure
applies differential weights to cases from a wide time frame.  As a
result, sudden caseload changes should not significantly disrupt
funding continuity over time.  Also, this measure can be adjusted to
recognize changes in AIDS mortality. 


--------------------
\12 Basing our estimate on the number of people who survived as of
this date means that it includes the largest number of people that
could potentially need services during the last year of that period. 
In contrast, an estimate based on the number of people who had
survived the entire last year of that period would tend to
underestimate the population in need of services.  This is because
people who had survived part, but not all, of the last year would be
omitted from the population even though they had potentially used
services sometime during that last year. 


   CURRENT CASELOAD MEASURES COULD
   BE IMPROVED
--------------------------------------------------------- Appendix I:4

Table I.2 compares the proxy measure of people living with AIDS based
on weighted-AIDS cases and the proxy measure based on cumulative AIDS
cases--the existing title I case measure--for each of the EMAs as of
December 1993.  Also shown is each EMA's caseload share based on
these measures.  Since funds are distributed on the basis of caseload
shares rather than number of cases, the former is actually the more
relevant measure from a formula perspective. 

The extent to which the cumulative case measure distorts EMAs' demand
for services is shown by the differences in weighted and cumulative
caseload shares.  For example, the cumulative case measure
overestimates caseload shares for the New York City, San Francisco,
Newark, and Jersey City EMAs from 7.85 to 15.52 percent.  Conversely,
the cumulative case measure underestimates demand for services in
EMAs such as Riverside-San Bernardino, Orlando, St.  Louis, Tampa-St. 
Petersburg, and Phoenix from 13.22 to 18.39 percent.  These
differences reflect the distortions created by the large number of
deceased persons in the case counts for the older and larger EMAs. 



                         Table I.2
          
                  Title I Caseload Counts


                        Coun  Shar  Coun  Shar  Percentage
                           t     e     t     e  difference
EMA                      (#)   (%)   (#)   (%)          \a
----------------------  ----  ----  ----  ----  ----------
New York, NY            23,4  21.4  58,8  23.2        8.12
                          84     6    96
San Francisco, CA       7,23  6.61  18,1  7.13        7.85
                           8          07
Los Angeles, CA         9,51  8.69  22,7  8.95        2.93
                           2          10
Miami, FL               5,11  4.67  10,9  4.32       -7.51
                           3          70
Newark, NJ              3,24  2.97  8,20  3.23        8.98
                           6           7
Houston, TX             3,94  3.61  9,53  3.76        4.17
                           8           9
Washington, DC          4,37  4.00  10,1  3.99       -0.24
                           6          27
Chicago, IL             4,27  3.91  9,78  3.85       -1.36
                           6           4
Atlanta, GA             3,09  2.83  7,18  2.83        0.17
                           2           4
Fort Lauderdale, FL     2,33  2.13  5,53  2.18        2.25
                           2           0
Philadelphia, PA        3,72  3.40  7,93  3.12       -8.18
                           5           3
Dallas, TX              2,70  2.47  6,08  2.40       -3.09
                           8           8
Boston, MA              3,28  3.01  7,06  2.78       -7.36
                           9           8
Jersey City, NJ         1,19  1.09  3,20  1.26       15.52
                           4           0
San Diego, CA           2,44  2.23  5,29  2.09       -6.60
                           5           8
Oakland, CA             1,89  1.73  4,32  1.70       -1.84
                           9           3
Baltimore, MD           2,51  2.29  5,02  1.98      -13.77
                           0           1
New Orleans, LA         1,28  1.17  3,10  1.22        4.44
                           2           7
Seattle, WA             1,64  1.50  3,70  1.46       -2.67
                           0           2
Tampa-St. Petersburg,   2,04  1.86  4,06  1.60      -13.99
 FL                        0           9
Orange County, CA       1,28  1.17  2,96  1.17       -0.51
                           5           6
Nassau-Suffolk, NY      1,45  1.33  3,36  1.32       -0.24
                           3           3
Detroit, MI             1,80  1.65  3,65  1.44      -12.71
                           3           0
West Palm Beach, FL     1,39  1.28  3,18  1.26       -1.59
                           7           8
Denver, CO              1,47  1.35  3,09  1.22       -9.46
                           5           8
Orlando, FL             1,25  1.15  2,39  0.94      -18.09
                           8           1
Riverside-San           1,58  1.45  3,00  1.18      -18.39
 Bernardino, CA            7           4
Bergen-Passaic, NJ      1,11  1.02  2,64  1.04        2.27
                           3           1
Kansas City, MO         1,08  0.99  2,29  0.90       -8.77
                           3           2
New Haven, CT           1,35  1.24  2,85  1.12       -9.08
                           4           5
Phoenix, AZ             1,12  1.03  2,27  0.89      -13.22
                           9           2
St. Louis, MO           1,23  1.13  2,39  0.94      -16.18
                           3           8
San Juan, PR            3,24  2.97  7,29  2.87       -3.23
                           8           1
Ponce, PR                676  0.62  1,59  0.63        1.79
                                       5
----------------------------------------------------------
\a While the data in the table have been rounded to two decimal
positions, the percentage differences were computed based on
additional decimal positions. 

Table I.3 compares our proxy measure of people living with AIDS that
is based on weighted living cases and the existing title II case
measure--2 years of cases--for each of the states and territories as
of December 1993.  The table also shows the caseload shares based on
these two measures.  Once again, an examination of caseload shares
under these two methods demonstrates the distortions that result from
using only 2 years of cases to estimate the number of people living
with AIDS.  The measure overestimates caseload shares for Delaware
and South Dakota from 9.61 to 17.94 percent and underestimates
caseload shares for Hawaii, Montana, and New Jersey from 5.22 to 7.18
percent. 



                         Table I.3
          
                  Title II Caseload Counts


                        Case  Shar  Case  Shar  Percentage
                           s     e     s     e  difference
State                    (#)   (%)   (#)   (%)          \a
----------------------  ----  ----  ----  ----  ----------
Alabama                 1,16  0.73  1,17  0.76        4.82
                           4           3
Alaska                    90  0.06    87  0.06        0.89
Arizona                 1,58  0.99  1,62  1.05        6.05
                           9           1
Arkansas                 664  0.41   684  0.44        7.16
California              28,2  17.6  27,3  17.7        0.49
                          99     7    46     6
Colorado                1,77  1.11  1,73  1.12        1.61
                           3           2
Connecticut             2,37  1.48  2,40  1.56        5.20
                           5           3
Delaware                 483  0.30   509  0.33        9.61
District of Columbia    2,42  1.52  2,28  1.49       -2.07
                           9           8
Florida                 16,5  10.3  16,0  10.4        0.43
                          74     5    07     0
Georgia                 4,36  2.73  4,12  2.68       -1.84
                           9           4
Hawaii                   556  0.35   496  0.32       -7.18
Idaho                    111  0.07   112  0.07        4.55
Illinois                4,96  3.10  4,86  3.16        2.05
                           0           8
Indiana                 1,34  0.84  1,35  0.88        4.73
                           3           3
Iowa                     305  0.19   313  0.20        6.85
Kansas                   524  0.33   546  0.35        8.26
Kentucky                 548  0.34   537  0.35        1.99
Louisiana               2,35  1.47  2,21  1.44       -1.88
                           1           8
Maine                    207  0.13   193  0.13       -3.27
Maryland                3,73  2.33  3,73  2.42        3.84
                           5           0
Massachusetts           3,69  2.31  3,57  2.32        0.64
                           2           3
Michigan                2,56  1.60  2,56  1.67        4.14
                           0           4
Minnesota                897  0.56   877  0.57        1.71
Mississippi              739  0.46   722  0.47        1.58
Missouri                2,46  1.54  2,45  1.59        3.72
                           1           5
Montana                   58  0.04    53  0.03       -5.55
Nebraska                 242  0.15   240  0.16        3.12
Nevada                   896  0.56   887  0.58        2.95
New Hampshire            184  0.11   171  0.11       -3.13
New Jersey              8,19  5.12  7,46  4.85       -5.22
                           5           9
New Mexico               414  0.26   401  0.26        0.80
New York                27,5  17.2  25,8  16.7       -2.62
                          93     3    39     8
North Carolina          2,01  1.26  1,95  1.27        0.42
                           9           0
North Dakota              20  0.01    20  0.01        1.85
Ohio                    2,39  1.49  2,33  1.52        1.57
                           1           6
Oklahoma                 962  0.60   989  0.64        6.93
Oregon                  1,09  0.69  1,06  0.69        1.00
                           9           7
Pennsylvania            4,66  2.91  4,53  2.95        1.25
                           0           7
Rhode Island             452  0.28   454  0.29        4.44
South Carolina          1,81  1.13  1,86  1.21        6.99
                           7           9
South Dakota              31  0.02    35  0.02       17.94
Tennessee               1,58  0.99  1,60  1.04        5.32
                           9           9
Texas                   11,0  6.90  10,4  6.79       -1.68
                          55          52
Utah                     408  0.25   399  0.26        1.72
Vermont                   96  0.06   100  0.06        8.11
Virginia                2,43  1.52  2,40  1.56        2.60
                           9           6
Washington              2,25  1.41  2,12  1.38       -2.20
                           5           1
West Virginia            175  0.11   160  0.10       -5.13
Wisconsin                947  0.59   960  0.62        5.37
Wyoming                   49  0.03    46  0.03       -1.76
Puerto Rico             5,20  3.25  4,81  3.13       -3.88
                           6           2
Virgin Islands            70  0.04    66  0.04       -2.55
----------------------------------------------------------
\a While the data in the table have been rounded to two decimal
positions, the percentage differences were computed based on
additional decimal positions. 


   COSTS
--------------------------------------------------------- Appendix I:5

Neither the title I nor title II formula includes a factor that
reflects differences in the cost of serving AIDS cases.  We were not
able to locate existing information on the actual cost of providing
health and support services to people with AIDS and HIV within
different geographic areas.  As a result, we constructed a proxy for
the cost of serving AIDS cases. 

The major factors that typically affect service costs are the
personnel who supply the service, capital costs such as office rent,
and supply costs such as for medications.  Titles I and II primarily
fund outpatient health, support, and case management services, which
are labor-intensive.  Hence, most of the service delivery costs for
services funded by titles I and II would be associated with the
personnel who provide the services. 

Furthermore, from our discussions with experts, we determined that an
existing measure of health labor costs--the Medicare Hospital Wage
Cost (MHWC) Index--might be an appropriate indicator of differences
in labor costs among EMAs and states.  This wage index was derived by
HCFA from hospital salary surveys and was designed to reflect
personnel costs in hospitals subject to the Medicare prospective
payment system (PPS).  Accordingly, the index is based on the
salaries of nurses, therapists, technicians, physicians, and
administrative staff.  In addition to being used for PPS, the MHWC
Index has been used to estimate cost variation for ambulatory service
centers, home health care providers, and skilled nursing facilities. 

An underlying assumption in our using the MHWC Index to estimate
costs for the personnel who deliver services funded by the CARE Act
is that the relative differences in these costs should mirror the
relative differences in costs of hospital personnel.  That is, in
places where hospital personnel costs are high, costs should also be
high for the personnel who provide services funded by the CARE Act. 
Likewise, in places where hospital personnel costs are low, the costs
for the personnel providing services funded by the CARE Act should
also be low. 

HCFA collects nationwide data on hospitals participating in PPS, so
cost data are readily available for each of the EMAs and non-EMA
areas.  HCFA publishes these data for metropolitan areas, and using
HCFA's automated MHWC database, we were able to construct a wage
index for each of the states. 

We were unable to locate existing data on the second major cost
category--capital costs.  High-cost areas, however, tend to have high
costs both for salaries and capital (for example, rent for office
space).  In our view, therefore, the MHWC Index would appear to be a
reasonable proxy for differences in both personnel and capital costs. 

The third major cost category--the cost of supplies such as
medications--is assumed not to systematically vary by location.  This
is because the amount an EMA or state pays for supplies like
medications is determined by a number of factors, including the price
that they are able to negotiate with suppliers. 

For our analysis, we constructed a cost index assuming 30 percent of
costs do not systematically vary by location and 70 percent do.  The
MHWC Index served as a proxy for the variation of these costs. 



We applied a weight of 30 percent for costs that do not
systematically vary because that is the approximate percentage of
title II funds typically expended on medications. 

Tables I.4 and I.5 display our estimated service costs for each of
the EMAs and states.  As shown in these tables, costs can vary by as
much as 100 percent.  For example, service costs in Oakland are twice
those in Ponce and San Juan, and 48 percent higher than in Miami. 
Similarly, service costs in Alaska are over 50 percent higher than in
Mississippi. 



                         Table I.4
          
                    Title I Cost Factor

                                        GAO cost index (.3
                            MHWC Index        + .7 * MHWC)
EMA                   (average = 1.00)    (average = 1.00)
------------------  ------------------  ------------------
Oakland, CA                       1.47                1.33
New York, NY                      1.41                1.29
San Francisco, CA                 1.41                1.29
Nassau-Suffolk, NY                1.31                1.21
Los Angeles, CA                   1.25                1.18
New Haven, CT                     1.23                1.16
San Diego, CA                     1.21                1.14
Boston, MA                        1.17                1.12
Riverside-San                     1.17                1.12
 Bernardino, CA
Bergen-Passaic, NJ                1.15                1.10
Philadelphia, PA                  1.11                1.08
Newark, NJ                        1.11                1.08
Jersey City, NJ                   1.11                1.08
Washington, DC                    1.11                1.08
Seattle, WA                       1.10                1.07
Detroit, MI                       1.09                1.06
Chicago, IL                       1.07                1.05
Fort Lauderdale,                  1.06                1.04
 FL
Denver, CO                        1.06                1.04
Atlanta, GA                       1.03                1.02
Phoenix, AZ                       1.01                1.01
West Palm Beach,                  1.00                1.00
 FL
Orange County, CA                 1.00                1.00
Houston, TX                       0.99                0.99
Baltimore, MD                     0.99                0.99
Dallas, TX                        0.95                0.97
New Orleans, LA                   0.95                0.96
Kansas City, MO                   0.95                0.96
Orlando, FL                       0.94                0.96
Tampa-St.                         0.94                0.96
 Petersburg, FL
St. Louis, MO                     0.91                0.94
Miami, FL                         0.86                0.90
Ponce, PR                         0.45                0.62
San Juan, PR                      0.44                0.61
----------------------------------------------------------


                         Table I.5
          
                    Title II Cost Factor

                                        GAO cost index (.3
                            MHWC Index        + .7 * MHWC)
State                 (average = 1.00)    (average = 1.00)
------------------  ------------------  ------------------
Alaska                            1.29                1.20
California                        1.25                1.18
New York                          1.24                1.17
District of                       1.21                1.15
 Columbia
Connecticut                       1.21                1.15
Massachusetts                     1.16                1.11
Hawaii                            1.13                1.09
New Jersey                        1.12                1.08
Nevada                            1.11                1.08
Rhode Island                      1.09                1.06
Washington                        1.05                1.04
Michigan                          1.05                1.04
Oregon                            1.05                1.03
Delaware                          1.03                1.02
New Hampshire                     1.02                1.01
Pennsylvania                      1.01                1.01
Minnesota                         1.00                1.00
Maryland                          1.00                1.00
Illinois                          0.98                0.99
Arizona                           0.98                0.98
Colorado                          0.97                0.98
Utah                              0.95                0.96
Florida                           0.95                0.96
Ohio                              0.94                0.96
Vermont                           0.93                0.95
Indiana                           0.92                0.94
New Mexico                        0.91                0.94
Maine                             0.91                0.94
Georgia                           0.91                0.94
Texas                             0.91                0.93
Wisconsin                         0.90                0.93
Virginia                          0.89                0.92
Nebraska                          0.89                0.92
Missouri                          0.88                0.92
North Carolina                    0.88                0.92
Kansas                            0.88                0.91
Louisiana                         0.87                0.91
Tennessee                         0.86                0.91
South Carolina                    0.86                0.90
Idaho                             0.85                0.90
Montana                           0.84                0.89
Kentucky                          0.84                0.89
North Dakota                      0.84                0.89
West Virginia                     0.83                0.88
Iowa                              0.83                0.88
Alabama                           0.80                0.86
Wyoming                           0.80                0.86
Oklahoma                          0.80                0.86
South Dakota                      0.78                0.84
Arkansas                          0.76                0.83
Mississippi                       0.70                0.79
Puerto Rico                       0.44                0.61
Virgin Islands                     N/A                 N/A
----------------------------------------------------------
Note:  N/A = Not applicable. 


   FISCAL CAPACITY
--------------------------------------------------------- Appendix I:6

A comprehensive indicator of an EMA's or state's fiscal capacity to
provide AIDS and HIV health and support services is one that includes
both a measure of the resource base (that is, tax base) and the
potential demand placed on these resources to fund AIDS and HIV
services. 

For the title I formula, we used per capita income (PCI) as the proxy
measure for EMA resources.  PCI data are compiled by the Department
of Commerce and are used to measure the income received by a
jurisdiction's residents, including wages and salaries, rents,
dividends, interest earnings, and income from nonresident corporate
business.  PCI also includes an adjustment for the rental value of
owner-occupied housing on the grounds that such ownership is similar
to the interest income earned from alternative financial investments. 
While PCI does not measure all taxable income, it is the most
comprehensive measure of EMA residents' income currently available. 

As a proxy for the level of demand placed on EMAs' resources, we used
AIDS incidence rates based on our estimate of living AIDS cases. 
AIDS incidence indicates the proportion of each EMA's population that
has been reported to have AIDS.  As such, AIDS incidence considers
the relative rather than the absolute demand placed on an EMA's
resources.  Those EMAs with larger proportions of their populations
having the disease are expected to have greater demands on their
resources than are EMAs with smaller proportions of their populations
infected. 

A complete title I fiscal capacity measure was constructed by first
producing cost-adjusted income amounts for each EMA through dividing
their PCI by their MHWC Index values.  This adjustment ensured that
EMAs were compared in terms of income that was of comparable
purchasing power.  Next, we divided these cost-adjusted values by
each EMA's AIDS incidence rate. 



We followed similar steps in constructing the title II fiscal
capacity measure, with the exception of using total taxable resources
(TTR)\13 to measure income.  TTR is a broader measure of income than
PCI because it considers all income potentially subject to a state's
taxing authority.  TTR is an average of PCI and per capita Gross
State Product (GSP).  GSP measures all income produced or received
within a state, whether received by residents, nonresidents, or
retained by business corporations.  Below is the the title II fiscal
capacity measure. 




--------------------
\13 TTR data are only available at the state level, so they cannot be
used to estimate the fiscal capacity of EMAs. 


   CURRENT FISCAL CAPACITY
   MEASURES COULD BE IMPROVED
--------------------------------------------------------- Appendix I:7

Under the current formulas, fiscal capacity is incompletely measured. 
The title I formula includes a measure of EMAs' AIDS incidence but
omits a measure of their resources, which creates a bias against
those EMAs with relatively low tax bases.  In table I.6, we show
EMAs' fiscal capacity as measured by the complete indicator that we
constructed--real PCI per weighted case--and by the existing
measure--AIDS incidence.  This table also shows the percentage
difference or disparity between these two measures. 

As shown in this table, fiscal capacity for the Riverside-San
Bernardino EMA is estimated to be 147 percent of the EMA average when
measured with a complete indicator--PCI per weighted case.  When only
AIDS incidence is considered, however, the EMA's fiscal capacity is
estimated to be 245 percent of the average.  Hence, when demand for
services is considered relative to available resources, Riverside-San
Bernardino's fiscal capacity is estimated to be 67 percent lower than
what is estimated under the existing formula.  Conversely, when
measured with a complete indicator, San Francisco's fiscal capacity
is estimated to be about 21 percent higher than is estimated under
the current formula.  This occurs because of the EMA's relatively
high tax base. 



                         Table I.6
          
              Title I Fiscal Capacity Factors

                                     Current
                    GAO's real  title I AIDS
                  PCI per wtd.     incidence
                    case index       index\a
                    (average =    (average =    Percentage
EMA                       100)          100)  difference\b
----------------  ------------  ------------  ------------
New York, NY                41            38         -8.46
San Francisco,              29            23        -20.68
 CA
Los Angeles, CA             94           104         10.67
Miami, FL                   40            48         17.95
Newark, NJ                  83            61        -26.16
Houston, TX                106            96         -8.93
Washington, DC             134           112        -16.46
Chicago, IL                218           201         -7.82
Atlanta, GA                118           114         -3.42
Fort Lauderdale,            67            61         -8.36
 FL
Philadelphia, PA           156           162          4.26
Dallas, TX                 129           120         -7.68
Boston, MA                 200           209          4.21
Jersey City, NJ             50            45         -9.31
San Diego, CA              102           128         24.75
Oakland, CA                112           129         15.12
Baltimore, MD              119           126          6.40
New Orleans, LA            103           109          6.03
Seattle, WA                168           149        -11.34
Tampa-St.                  113           135         19.20
 Petersburg, FL
Orange County,             258           218        -15.33
 CA
Nassau-Suffolk,            226           204         -9.59
 NY
Detroit, MI                265           307         16.10
West Palm Beach,           108            74        -31.71
 FL
Denver, CO                 138           144          4.26
Orlando, FL                108           142         30.96
Riverside-San              147           245         66.79
 Bernardino, CA
Bergen-Passaic,            169           127        -24.53
 NJ
Kansas City, MO            176           184          4.61
New Haven, CT              167           149        -11.08
Phoenix, AZ                211           267         26.45
St. Louis, MO              255           273          7.01
San Juan, PR                33            67        103.54
Ponce, PR                   17            56        236.48
----------------------------------------------------------
Note:  The average values are based on the average for all EMAs. 

\a A high AIDS incidence rate value reflects a lower fiscal capacity
while a high PCI value reflects a higher fiscal capacity. 
Consequently, we inverted the AIDS incidence rate to make it
comparable with the other indicator. 

\b While the data in the table have been rounded to two decimal
positions, the percentage differences were computed on the basis of
additional decimal positions. 

In contrast to the title I formula, the title II formula measures
states' income and omits their AIDS incidence rates.  This omission
creates a bias against those states with relatively high service
demands on their resources.  Table I.7 shows states' fiscal capacity
when measured by a complete indicator--real TTR per weighted
case--followed by the existing measure--nominal per capita income. 
Also, the table shows the percentage difference or disparity between
these two measures. 



                         Table I.7
          
              Title II Fiscal Capacity Factors

                                     Current
                    GAO's real      title II
                  TTR per wtd.   nominal PCI
                    case index         index
                    (average =    (average =    Percentage
State                     100)          100)  difference\a
----------------  ------------  ------------  ------------
Alabama                    208            82        -60.68
Alaska                     544           109        -80.00
Arizona                    129            86        -33.05
Arkansas                   209            78        -62.72
California                  64           107         68.81
Colorado                   124           101        -18.43
Connecticut                101           136         34.43
Delaware                   106           108          1.87
District of                 18           133        618.07
 Columbia
Florida                     48            98        103.41
Georgia                     96            91         -5.23
Hawaii                     134           107        -19.97
Idaho                      545            81        -85.13
Illinois                   162           109        -32.86
Indiana                    257            91        -64.66
Iowa                       610            92        -84.88
Kansas                     314            98        -68.95
Kentucky                   405            83        -79.40
Louisiana                  111            79        -28.89
Maine                      354            92        -74.01
Maryland                    90           116         28.24
Massachusetts              109           121         11.03
Michigan                   213            98        -53.74
Minnesota                  321           101        -68.54
Mississippi                198            71        -64.18
Missouri                   135            95        -29.74
Montana                    777            81        -89.58
Nebraska                   431            96        -77.67
Nevada                      92           102         11.09
New Hampshire              397           116        -70.90
New Jersey                  71           133         87.34
New Mexico                 199            77        -61.06
New York                    42           119        182.77
North Carolina             214            89        -58.48
North Dakota             1,910            85        -95.55
Ohio                       283            94        -66.87
Oklahoma                   199            82        -58.89
Oregon                     147            92        -37.39
Pennsylvania               157           102        -34.97
Rhode Island               129           102        -20.47
South Carolina             113            81        -28.43
South Dakota             1,461            83        -94.29
Tennessee                  193            87        -54.66
Texas                      101            90        -11.05
Utah                       226            77        -65.80
Vermont                    362            95        -73.78
Virginia                   185           104        -43.83
Washington                 141           103        -26.84
West Virginia              548            76        -86.15
Wisconsin                  337            94        -71.99
Wyoming                    754            88        -88.35
Puerto Rico                N/A           N/A           N/A
Virgin Islands             N/A           N/A           N/A
----------------------------------------------------------
Notes:  The average values are based on the average for the United
States.

N/A = Not applicable. 

\a While the data in the table have been rounded to two decimal
positions, the percentage differences were computed on the basis of
additional decimal positions. 

As shown in table I.7, Kentucky's fiscal capacity is estimated to be
17 percent below the national average when only PCI is considered. 
When income is adjusted by AIDS incidence rates, however, the state's
fiscal capacity is estimated to be more than four times the national
average.  This occurs because of the relatively low AIDS incidence
rate in Kentucky as compared with the state's available resources. 
Conversely, while the District of Columbia has a relatively large
resource base (33 percent above the national average), its AIDS
incidence is also relatively high.  Consequently, when measured with
a complete indicator, the District of Columbia's fiscal capacity is
found to be 72 percent below the national average. 


TITLE I FUNDING AND COMBINED
TITLES I AND II FUNDING MEET
NEITHER EQUITY CRITERION
========================================================== Appendix II

We compared the distribution of title I funding and the combined
title I and II funding against the beneficiary and taxpayer equity
criteria.  These comparisons indicated that the current formulas do
not distribute funding in accordance with either of the equity
criteria. 


   EQUITY OF TITLE I FUNDING
-------------------------------------------------------- Appendix II:1

Under the beneficiary equity standard, the size of the grant award
depends on two factors:  the number of cases and the cost of
services.  If the grant is expressed on a per-case basis, this
implies that per-case funding should vary only with differences in
the cost of services.  To determine how well the current distribution
of title I funds meets the beneficiary equity standard, we performed
a regression analysis\14 to determine the extent to which cost
differences can account for differences in nominal per-case
funding.\15

If the current distribution of title I funds reflected the
beneficiary equity standard, then a substantial share of the
variation in per-case funding could be explained by cost differences. 
Our statistical analysis, however, indicates the current distribution
of title I funds bears little relation to the variation in costs. 
The strength of a relationship is commonly measured by a statistic
known as R\2 .  In this case, the R\2 is 0.10, meaning that
differences in cost account for only 10 percent of the variation in
per-case funding for EMAs.  Hence, 90 percent of the variation is
related to other factors. 

Examples of the inequities that result from this low R\2 are
displayed in figure II.1.  If differences in per-case funding and
costs were perfectly correlated, all EMAs would fall along the
straight line shown in this figure.  The wide scatter around the
line, however, demonstrates that per-case funding and costs are not
systematically related.  For example, the service costs for the
Oakland and San Francisco EMAs are about the same; yet, per-case
funding for San Francisco is about twice that for Oakland.  If
per-case funding and costs were more strongly correlated, both EMAs
would be positioned closer to the straight line.  Furthermore,
Oakland's per-case funding would even be slightly higher than San
Francisco's rather than vice versa. 

   Figure II.1:  Nominal Title I
   Funding Per Case and Cost

   (See figure in printed
   edition.)

Note:  The two EMAs located in Puerto Rico--Ponce and San Juan--are
excluded from this figure. 

This relationship is also illustrated by other pairs of EMAs.  For
example, New York City's service costs are about 20 percent higher
than Jersey City's, but their per-case funding is about the
same--about 30 percent above the EMA average.  Consequently, at their
current funding levels, the Jersey City EMA can purchase more
services for its patients than can New York City. 

From the perspective of beneficiary equity, therefore, the current
per-case funding distribution is inequitable.  However, if these
differences can be accounted for by differences in fiscal capacity,
then the grant distribution may reflect our taxpayer equity
criterion. 

The taxpayer equity standard implies that per-case funding should be
related to both cost differences and differences in fiscal capacity. 
Thus, to determine if the formula meets the taxpayer equity standard,
we performed a regression analysis to determine the extent to which
differences in per-case funding can be explained by both cost and
fiscal capacity differences.  We used per-case funding (measured in
nominal dollars) as the dependent variable and used both cost and
fiscal capacity (also measured in nominal dollars) as independent
variables. 

Our statistical results show that the current distribution of title I
funds does more closely reflect the taxpayer equity standard.  The
R\2 is 0.56, meaning that differences in cost and fiscal capacity
account for 56 percent of the variation in EMAs' per-case funding. 
Nonetheless, a considerable amount of unexplained variation in
per-case funding--44 percent--
remains. 

The tendency to allocate more aid to low-capacity EMAs is shown
graphically in figure II.2.  San Francisco's funding capacity is the
lowest of all the EMAs and its cost-adjusted per-case real funding is
high--nearly 40 percent above the average.  West Palm Beach has
approximately average funding capacity and receives an average
per-case real funding amount.  Detroit, Orange County, and St.  Louis
have high funding capacities and receive relatively lower per-case
funding amounts.  Thus, there is a clear tendency to target more aid
to EMAs with lower funding capacities. 

   Figure II.2:  Real Title I
   Funding Per Case and Tax Base

   (See figure in printed
   edition.)

Note:  The two EMAs located in Puerto Rico--Ponce and San Juan--are
excluded from this figure. 

However, there are many exceptions to this general tendency.  For
example, while the San Francisco and New York City EMAs' funding
capacities are comparable, they receive very different per-case real
funding amounts.  Real per-case funding is 40 percent above the
average for San Francisco and only about average for New York City. 
Since the grant amounts have already been adjusted for cost
differences, we would conclude that the New York City EMA is
underfunded compared to San Francisco.  Similarly, both West Palm
Beach and Tampa have average funding capacities, but West Palm Beach
receives about 25 percent more title I funds than Tampa.  Based on
examples like these and our regression results, we conclude that
while title I funding demonstrates a tendency to target more aid to
low-capacity EMAs, substantial inequities exist. 


--------------------
\14 Regression analysis is a statistical technique used to measure
the degree to which variation in a variable can be explained by
variation in other variables. 

\15 When we refer to nominal dollars or funds, we mean an amount that
has not been adjusted for service cost differences among EMAs and
states.  In contrast, when we refer to real dollars or funds, we mean
an amount that has been cost-adjusted. 


   EQUITY OF COMBINED TITLE I AND
   II FUNDING
-------------------------------------------------------- Appendix II:2

The beneficiary equity standard for the combined distribution of
title I and II funds uses caseload and cost measures that encompass
the entire state rather than just an EMA.  Using state rather than
EMA data, we estimated the same regression models to assess the
equity of the combined title I and II funding.\16

These results show that the current distribution of title I and II
funds, in combination, does not meet either equity standard.  If
beneficiary equity were fully realized, cost differences would
account for all of the variation in per-case funding, but costs
explain only 14 percent of this variation.  To express this another
way, under the beneficiary equity model, states with the same
relative cost of services should receive equal funding on a per-case
basis.  As shown in figure II.3, however, states with dramatically
different service costs received comparable per-case funding amounts. 
For example, service costs for both Georgia and New Mexico are about
average; but Georgia's per-case funding is 69 percent higher than New
Mexico's.  A similar situation exists for Ohio and Texas.  Both the
regression results and these examples indicate that title I and II
funds, in combination, are not distributed in a way that meets the
beneficiary equity criterion. 

   Figure II.3:  Nominal Title I
   and II Per-Case Funding and
   Cost

   (See figure in printed
   edition.)

Note:  States receiving the minimum title II grant amount of $100,000
and the territories are excluded from this figure. 

Under the taxpayer equity model, cost and fiscal capacity should
account for 100 percent of the variation in per-case funding.  Our
regression analysis indicates these two factors account for only 33
percent of the variation in per-case funding.  By implication, about
67 percent of the variation in per-case funding is unrelated to need
as reflected by differences in cost and fiscal capacity. 

Specific examples of the inequities are illustrated in figure II.4. 
For example, Massachusetts and New Hampshire receive comparable
per-case funding amounts, but New Hampshire's tax base is about four
times that of Massachusetts.  Similar situations exist for the states
of Connecticut and Kentucky and for Delaware and Maine.  In contrast,
Hawaii receives a grant that is about half the amount that Missouri
receives; yet, the states' tax bases are comparable.  This is also
the case for the states of Georgia and Nevada and for Illinois and
Oregon.  On the basis of this analysis, we conclude that the combined
title I and II funding does not meet the taxpayer equity criterion. 

   Figure II.4:  Real Title I and
   II Per-Case Funding and Tax
   Base

   (See figure in printed
   edition.)

Note:  States receiving the minimum title II grant amount of $100,000
and the territories are excluded from this figure. 


--------------------
\16 States receiving the minimum title II grant amount of $100,000
are excluded from these analyses because their funds are not
distributed by formula and cannot be related to an equity analysis. 


FUNDING INEQUITIES RESULT FROM THE
STRUCTURE OF THE TITLE I AND II
FORMULAS
========================================================= Appendix III

When funding for titles I and II is considered jointly, the major
cause of funding inequities is that EMA cases are counted in both
formulas but cases outside EMAs are not.  As a result, states with
few or no cases in an EMA receive disproportionately less per-case
funding than do states with large proportions of their caseloads in
EMAs. 

The following two-state example demonstrates how the current
structure produces funding inequities between a state with an EMA and
one without an EMA.  For this example, we will assume that $1,000 has
been appropriated for each of titles I and II.  Also, the two states
are assumed to be alike in terms of their costs and funding capacity;
however, they differ in the number of cases they must serve and
whether these cases live in an EMA. 

State A has 200 cases, all living in an EMA while State B has 100
cases and no EMA.  Hence, State A has two-thirds and State B has
one-third of the total cases.  Since title I funds are allocated
based on each state's share of EMA cases, the entire $1,000 would be
distributed to the EMA in State A, and none of the funds would be
distributed in State B (see fig.  III.1). 

   Figure III.1:  Calculation of
   Title I Grant

   (See figure in printed
   edition.)

Title II funding is allocated in proportion to each state's total
caseload.  Since State A has two-thirds of all cases, it would
receive two-thirds ($667) of the title II appropriation.  State B
would receive one-third ($333) of the appropriation. 

Each state's total grant is then determined by summing their title I
and II grants (see fig.  III.2). 

   Figure III.2:  Total Funding

   (See figure in printed
   edition.)

To determine the states' per-case funding amounts, their total grant
amounts are divided by their total caseloads (see fig.  III.3). 

   Figure III.3:  Per-Case Funding

   (See figure in printed
   edition.)

In this example, the current title structure produced differences in
per-case funding for these two states that amounts to about 150
percent.  Moreover, this difference is unrelated to the states'
funding needs and occurs solely because of the existing title
structure. 


   DOUBLE COUNTING EMA CASES
   ACCOUNTS FOR MOST OF THE
   FUNDING INEQUITIES
------------------------------------------------------- Appendix III:1

To determine the extent to which the current structure accounts for
per-case funding differences, we compared two regression models.  The
first model was our earlier one that examined the effects of
differences in cost and fiscal capacity on states' combined title I
and II per-case funding amounts.  For the second model, we examined
these effects along with the effect of the percentage of AIDS cases
in an EMA on states' combined title I and II per-case funding
amounts. 

As discussed in appendix II, the cost and fiscal capacity model
explains only 33 percent of the variation in nominal title I and II
per-case funding.  In contrast, however, the model that also includes
the percentage of EMA AIDS cases as a factor explains 85 percent of
this variation. 

The relationship between the percentage of EMA AIDS cases and
per-case funding is displayed in figure III.4.  As shown in this
figure, the states with fewer cases in EMAs, for example, Hawaii,
Ohio, and West Virginia receive the smallest grants, and the states
with larger percentages of cases in EMAs, for example, California,
the District of Columbia, and New York, tend to receive the largest
grants.  As demonstrated by the regression analysis and this figure,
the funding differences result from the structure of the formulas
rather than funding needs as measured by cases, costs, and fiscal
capacity.  Consequently, these differences are inequitable. 

   Figure III.4:  Nominal Combined
   Title I and II Per-Case Funding
   and Percentage of EMA Cases

   (See figure in printed
   edition.)

Notes:  States receiving the minimum title II grant amount of
$100,000 and the territories are excluded from this figure.

Of the states without an EMA, only the two that received the smallest
and largest grants are displayed. 


STRUCTURAL CHANGES CAN IMPROVE
EQUITY
========================================================== Appendix IV

While improvements in funding equity can be achieved by adopting
better indicators of caseload, cost, and funding capacity in the
allocation formulas, greater improvement could result from changing
the allocation structures to avoid inequities created by counting EMA
cases in both formulas.  A variety of approaches could be used that
vary in terms of how they would affect existing service delivery
responsibilities and structures. 


   CONSOLIDATED GRANTS
-------------------------------------------------------- Appendix IV:1

The simplest way to improve funding equity is to consolidate titles I
and II into a single grant and distribute funds to the state
governments through an equity-based formula.  State governments would
be the political entity responsible for using the aid provided for
serving those in need.  The funds would be allocated based on each
state's total cases, thus avoiding any double counting of EMA cases. 

Two implications of the consolidated grant approach, however, are the
potential infringement on the autonomy currently afforded EMAs in
delivering services and changes in existing service responsibilities
and structures.  Currently, EMAs are responsible for delivering
services within their areas, and they have service delivery networks
already in place.  Under a consolidated grant, all funds would be
distributed to the states.  Hence, a state could potentially assume
total responsibility for service delivery in EMAs or continue to
allow the EMAs to administer the programs they now operate, funding
them from its consolidated grant. 


   GEOGRAPHIC DIVISION OF SERVICE
   RESPONSIBILITIES
-------------------------------------------------------- Appendix IV:2

A second corrective approach maintains the two distinct titles--title
I for EMAs and title II for states--however, the EMAs would become
responsible for all services in their areas, including those services
currently under the purview of the states through title II.  Hence,
EMAs and states would continue to be funded under separate titles,
but the services funded under these titles would be identical.  Both
title I and II funds would be allocated through equity-based
formulas.  Title I funds would be distributed to EMAs on the basis of
their respective shares of cases, and title II funds would be
distributed to states on the basis of their respective shares of the
non-EMA cases. 

Like the previous approach, this one avoids the inequities currently
caused by counting EMA cases in both formulas.  Furthermore, as do
the existing formulas, this approach maintains the EMAs' autonomy in
the delivery of services.  In addition, this approach allows
comparable per-case funding levels among the states and EMAs. 
However, this approach would lead to significant changes in service
delivery responsibilities.  Because EMAs are not currently
responsible for providing services such as insurance continuation and
medication assistance, they would have to develop the capacity to
administer these services in addition to the medical and support
services they now provide. 


   FUNDING BY TYPES OF SERVICE
-------------------------------------------------------- Appendix IV:3

A third corrective approach involves allocating funds for medical and
support services separately from those services that states provide
statewide.  This approach avoids inequities produced by double
counting EMA cases, continues the existing autonomy afforded EMAs,
and requires no changes in existing service delivery
responsibilities.  Furthermore, the approach ensures that comparable
per-case funding is available across EMAs and states and between EMA
and non-EMA areas. 

The following example, using the same two states from the example in
appendix III, illustrates how this approach would improve funding
equity.  The two states are alike in terms of their costs and funding
capacity, but they differ in the number of cases they serve and
whether the cases live in an EMA. 

In this example, a total of $2,000 is appropriated:  $1,500 for
medical services and $500 for statewide services.  As will be shown
through this example, dividing funds in this way would result in
funding amounts for title I and II activities that are comparable to
what was found in the earlier example in which both titles each had a
$1,000 appropriation.  That is, $1,000 would still be available for
the EMA, and $1,000 would still be available for the states because
state funding would be determined by adding together the $500 for
non-EMA medical services and $500 for statewide services. 

State A has 200 cases, all living in an EMA, and State B has 100
cases and no EMA.  Hence, once again, State A has 100 percent of the
EMA cases and two-thirds of total cases.  State B has 100 percent of
the non-EMA cases and one-third of total cases.  Expressed
differently, two-thirds of all cases live in an EMA and one-third of
all cases do not. 

Under this approach, the medical services appropriation would be
divided between EMA and non-EMA areas on the basis of their
respective caseloads.  Since two-thirds of all cases live in an EMA,
two-thirds of the medical services appropriation ($1,000) would be
set aside for the EMA.  The remaining one-third of the medical
services appropriation ($500) would be set aside for distribution to
states on the basis of the number of cases living outside an EMA. 

EMA medical services funds are allocated based on the shares of cases
living in an EMA.  Since State A contains all EMA cases, all of these
funds ($1,000) would be allocated within State A.  None of these
funds would be allocated within State B as it has no EMA.  Non-EMA
medical services funds would be allocated based on states' shares of
cases living outside an EMA.  Since State A has no non-EMA cases, it
would receive none of these funds.  State B would receive the entire
$500 of non-EMA medical services funds because it contains all
non-EMA cases. 

The statewide services appropriation ($500) would be allocated based
on each state's share of total cases.  Since State A has two-thirds
of the total cases, it would receive $334; State B would receive the
remaining one-third of funds ($167). 

The states' total grants would be the sum of their EMA medical
services grant, non-EMA medical services grant, and statewide
services grant.  In this case, State A would receive a total of
$1,334 and State B would receive a total of $667 (see fig.  IV.1). 

   Figure IV.1:  Total Grants

   (See figure in printed
   edition.)

As before, the states' per-case funding amounts would be obtained by
dividing their total grant amounts by their total caseloads.  Figure
IV.2 shows the per-case funding amounts for the two states. 

   Figure IV.2:  Per-Case Funding

   (See figure in printed
   edition.)

Under our proposed approach, each state would receive identical
per-case funding.  This contrasts significantly with the current
approach, which produces highly unequal per-case funding that is
unrelated to either costs or funding capacity and is therefore
inequitable. 


COMPARISON OF EMA AND STATE CARE
ACT GRANTS UNDER CURRENT AND
EQUITY-BASED FORMULAS
=========================================================== Appendix V

In this appendix, we describe how title I and II funding would be
distributed if the formulas were changed to meet either the
beneficiary or taxpayer equity criterion.  Both the beneficiary and
taxpayer equity formulas were described in greater detail in appendix
I, which also provided more detailed discussion of the caseload,
cost, and fiscal capacity factors used in these formulas. 

Depending on the amount of title I and II funds appropriated or the
use of funding-loss mechanisms such as hold-harmless provisions,
formula modifications could decrease funding to some EMAs and states
and increase funding to others.  Whether and how funding losses
should be prevented would be the decision of the Congress; however,
in this appendix, we show the effects of formula changes when title I
and II appropriations remain constant and no funding-loss mechanisms
are employed. 


   BENEFICIARY EQUITY FORMULAS
--------------------------------------------------------- Appendix V:1

Table V.1 displays each EMA's title I fiscal year 1995 funding under
both the existing and the beneficiary equity formulas, along with the
difference in funding that would be received under these formulas. 
Relative to the existing formula, changes in EMAs' allocations under
the beneficiary equity formula would range from a decrease of 33.57
percent to an increase of 58.72. 



                         Table V.1
          
             Title I FY 1995 Funding Under the
          Existing and Beneficiary Equity Formulas


                            Beneficiar
                  Existing    y equity
                formula FY  formula FY
                      1995        1995            Percenta
EMA             allocation  allocation   Dollars        ge
--------------  ----------  ----------  --------  --------
Atlanta, GA     $4,007,435  $4,081,413   $73,978      1.85
Austin, TX       1,085,663   1,288,596   202,933     18.69
Baltimore, MD    2,691,832   3,377,671   685,839     25.48
Bergen-          1,452,105   1,605,315   153,210     10.55
 Passaic, NJ
Boston, MA       3,456,473   4,830,099  1,373,62     39.74
                                               5
Chicago, IL      4,924,568   6,270,043  1,345,47     27.32
                                               5
Dallas, TX       3,385,351   3,472,214    86,863      2.57
Denver, CO       1,668,174   2,043,282   375,107     22.49
Detroit, MI      1,716,243   2,377,827   661,584     38.55
Dutchess           359,357     391,211    31,854      8.86
 County, NY
Fort             3,635,539   3,318,715         -     -8.71
 Lauderdale,                             316,825
 FL
Houston, TX      5,803,257   5,715,983   -87,273     -1.50
Jacksonville,    1,214,884   1,307,390    92,506      7.61
 FL
Jersey City,     2,406,293   1,759,633         -    -26.87
 NJ                                      646,659
Kansas City,     1,145,290   1,235,732    90,442      7.90
 MO
Los Angeles,    12,998,478  15,329,003  2,330,52     17.93
 CA                                            5
Miami, FL        8,079,775   6,585,215         -    -18.50
                                        1,494,56
                                               0
Nassau-          1,676,365   2,441,033   764,668     45.61
 Suffolk, NY
New Haven, CT    1,484,228   2,143,285   659,057     44.40
New Orleans,     1,798,493   1,728,625   -69,868     -3.88
 LA
New York, NY    48,636,026  42,538,803         -    -12.54
                                        6,097,22
                                               3
Newark, NJ       5,559,872   4,883,141         -    -12.17
                                         676,731
Oakland, CA      2,321,637   3,197,753   876,117     37.74
Orange County,   1,490,021   1,841,089   351,068     23.56
 CA
Orlando, FL      1,286,590   1,522,576   235,986     18.34
Philadelphia,    4,124,036   5,528,896  1,404,86     34.07
 PA                                            0
Phoenix, AZ      1,096,350   1,332,987   236,637     21.58
Portland, OR       986,510   1,278,214   291,704     29.57
Riverside-San    1,485,035   2,357,024   871,988     58.72
 Bernardino,
 CA
St. Louis, MO    1,137,857   1,385,224   247,367     21.74
San Antonio,       960,778   1,119,605   158,827     16.53
 TX
San Diego, CA    2,861,916   3,702,743   840,827     29.38
San Francisco,  19,126,679  12,705,987         -    -33.57
 CA                                     6,420,69
                                               2
Santa Rosa-        574,580     745,241   170,661     29.70
 Petaluma, CA
Seattle, WA      1,920,227   2,430,507   510,280     26.57
Tampa-St.        2,172,534   2,538,261   365,727     16.83
 Petersburg,
 FL
Vineland, NJ       197,896     207,866     9,970      5.04
Washington, DC   5,623,294   6,411,647   788,354     14.02
West Palm        1,961,600   1,849,331         -     -5.72
 Beach, FL                               112,269
Caguas, PR         489,261     529,640    40,379      8.25
Ponce, PR        1,020,387     885,680         -    -13.20
                                         134,707
San Juan, PR     4,662,110   4,390,498         -     -5.83
                                         271,612
==========================================================
Total           $174,685,0  $174,685,0
                        00          00
----------------------------------------------------------
Table V.2 displays the distribution of title II fiscal year 1995
funding under both the existing and the beneficiary equity formulas. 
Relative to the existing formula, changes in states' allocations
under the beneficiary equity formula would range from a decrease of
69.84 percent to an increase of 247.33 percent.\17



                         Table V.2
          
             Title II FY 1995 Funding Under the
          Existing and Beneficiary Equity Formulas


                  Existing  Beneficiar
                formula FY    y equity
State or              1995     FY 1995            Percenta
territory       allocation  allocation   Dollars        ge
--------------  ----------  ----------  --------  --------
Alabama         $1,349,942  $3,175,846  $1,825,9    135.26
                                              04
Alaska             100,000     337,517   237,517    237.52
Arizona          1,759,313   1,552,919         -    -11.73
                                         206,394
Arkansas           753,038   1,649,752   896,714    119.08
California      27,867,193  21,249,330         -    -23.75
                                        6,617,86
                                               3
Colorado         1,980,699   1,381,805         -    -30.24
                                         598,894
Connecticut      2,404,858   4,096,894  1,692,03     70.36
                                               6
Delaware           585,604   1,646,242  1,060,63    181.12
                                               8
District of      2,532,524     763,706         -    -69.84
 Columbia                               1,768,81
                                               8
Florida         17,780,752  13,052,493         -    -26.59
                                        4,728,25
                                               9
Georgia          4,731,696   4,778,646    46,950      0.99
Hawaii             499,350   1,734,379  1,235,02    247.33
                                               9
Idaho              138,867     324,283   185,416    133.52
Illinois         5,577,650   2,981,750         -    -46.54
                                        2,595,90
                                               0
Indiana          1,536,770   3,873,203  2,336,43    152.04
                                               3
Iowa               333,360     792,428   459,068    137.71
Kansas             568,263     903,092   334,829     58.92
Kentucky           643,697   1,618,407   974,710    151.42
Louisiana        2,785,044   3,389,561   604,517     21.71
Maine              228,492     599,407   370,915    162.33
Maryland         4,684,012   1,984,893         -    -57.62
                                        2,699,11
                                               9
Massachusetts    3,776,077   2,777,381         -    -26.45
                                         998,696
Michigan         2,675,943   2,904,689   228,746      8.55
Minnesota          973,550   2,649,145  1,675,59    172.11
                                               5
Mississippi        954,192   1,909,469   955,277    100.11
Missouri         2,504,335   1,680,127         -    -32.91
                                         824,208
Montana            100,000     158,978    58,978     58.98
Nebraska           267,083     636,498   369,415    138.31
Nevada             964,174   2,910,751  1,946,57    201.89
                                               7
New Hampshire      175,763     210,839    35,076     19.96
New Jersey       8,958,831   9,072,667   113,836      1.27
New Mexico         479,074   1,126,763   647,689    135.20
New York        29,093,044  16,378,152         -    -43.70
                                        12,714,8
                                              92
North Carolina   2,414,668   6,292,042  3,877,37    160.58
                                               4
North Dakota       100,000      89,715   -10,285    -10.28
Ohio             2,623,138   7,170,372  4,547,23    173.35
                                               4
Oklahoma         1,050,786   2,242,488  1,191,70    113.41
                                               2
Oregon           1,300,587   1,126,733         -    -13.37
                                         173,854
Pennsylvania     5,177,510   5,825,048   647,538     12.51
Rhode Island       554,753   1,618,244  1,063,49    191.71
                                               1
South Carolina   2,679,771   5,747,815  3,068,04    114.49
                                               4
South Dakota       100,000      89,241   -10,759    -10.76
Tennessee        1,846,877   4,406,616  2,559,73    138.60
                                               9
Texas           12,636,414  10,282,829         -    -18.63
                                        2,353,58
                                               5
Utah               428,266   1,094,580   666,314    155.58
Vermont            103,727     272,789   169,062    162.99
Virginia         2,642,609   4,722,057  2,079,44     78.69
                                               8
Washington       2,310,797   2,180,920         -     -5.62
                                         129,877
West Virginia      184,768     418,183   233,415    126.33
Wisconsin        1,063,650   2,510,890  1,447,24    136.06
                                               0
Wyoming            100,000     123,698    23,698     23.70
Guam                 2,902      10,411     7,509    258.76
Puerto Rico      7,682,087   3,962,806         -    -48.41
                                        3,719,28
                                               1
Virgin Islands           0     277,015   277,015       N/A
==========================================================
Total           $174,766,5  $174,766,5
                        00          00
----------------------------------------------------------
Note:  N/A = Not applicable. 


--------------------
\17 This range excludes Guam. 


   TAXPAYER EQUITY FORMULAS
--------------------------------------------------------- Appendix V:2

Table V.3 displays title I fiscal year 1995 funding under both the
existing and the taxpayer equity formulas.  Relative to the existing
formula, changes in EMAs' title I allocations under the taxpayer
equity formula would range from a decrease of 37.18 percent to an
increase of 38.90 percent. 



                         Table V.3
          
             Title I FY 1995 Funding Under the
           Existing and Taxpayer Equity Formulas


                  Existing    Taxpayer
                formula FY   equity FY
                      1995        1995            Percenta
EMA             allocation  allocation   Dollars        ge
--------------  ----------  ----------  --------  --------
Atlanta, GA     $4,007,435  $3,875,802         -     -3.28
                                        $131,633
Austin, TX       1,085,663   1,316,251   230,588     21.24
Baltimore, MD    2,691,832   3,222,078   530,246     19.70
Bergen-          1,452,105   1,296,500         -    -10.72
 Passaic, NJ                             155,605
Boston, MA       3,456,473   3,487,398    30,925      0.89
Chicago, IL      4,924,568   4,520,122         -     -8.21
                                         404,446
Dallas, TX       3,385,351   3,219,764         -     -4.89
                                         165,587
Denver, CO       1,668,174   1,852,549   184,375     11.05
Detroit, MI      1,716,243   1,223,021         -    -28.74
                                         493,222
Dutchess           359,357     380,235    20,879      5.81
 County, NY
Fort             3,635,539   3,612,284   -23,256     -0.64
 Lauderdale,
 FL
Houston, TX      5,803,257   5,798,947    -4,309     -0.07
Jacksonville,    1,214,884   1,307,433    92,549      7.62
 FL
Jersey City,     2,406,293   1,997,181         -    -17.00
 NJ                                      409,111
Kansas City,     1,145,290     926,488         -    -19.10
 MO                                      218,801
Los Angeles,    12,998,478  15,608,417  2,609,93     20.08
 CA                                            8
Miami, FL        8,079,775   7,618,124         -     -5.71
                                         461,651
Nassau-          1,676,365   1,669,138    -7,226     -0.43
 Suffolk, NY
New Haven, CT    1,484,228   1,783,837   299,609     20.19
New Orleans,     1,798,493   1,739,789   -58,704     -3.26
 LA
New York, NY    48,636,026  49,293,321   657,295      1.35
Newark, NJ       5,559,872   5,153,909         -     -7.30
                                         405,962
Oakland, CA      2,321,637   3,041,615   719,978     31.01
Orange County,   1,490,021   1,145,238         -    -23.14
 CA                                      344,783
Orlando, FL      1,286,590   1,469,051   182,461     14.18
Philadelphia,    4,124,036   4,793,757   669,721     16.24
 PA
Phoenix, AZ      1,096,350     851,132         -    -22.37
                                         245,218
Portland, OR       986,510     997,508    10,999      1.11
Riverside-San    1,485,035   2,062,696   577,660     38.90
 Bernardino,
 CA
St. Louis, MO    1,137,857     714,755         -    -37.18
                                         423,102
San Antonio,       960,778     974,770    13,992      1.46
 TX
San Diego, CA    2,861,916   3,659,542   797,626     27.87
San Francisco,  19,126,679  15,050,088         -    -21.31
 CA                                     4,076,59
                                               1
Santa Rosa-        574,580     765,980   191,399     33.31
 Petaluma, CA
Seattle, WA      1,920,227   2,065,044   144,817      7.54
Tampa-St.        2,172,534   2,432,673   260,139     11.97
 Petersburg,
 FL
Vineland, NJ       197,896     212,618    14,722      7.44
Washington, DC   5,623,294   5,885,978   262,684      4.67
West Palm        1,961,600   1,805,503         -     -7.96
 Beach, FL                               156,098
Caguas, PR         489,261     534,078    44,817      9.16
Ponce, PR        1,020,387     893,101         -    -12.47
                                         127,286
San Juan, PR     4,662,110   4,427,284         -     -5.04
                                         234,826
==========================================================
Total           $174,685,0  $174,685,0
                        00          00
----------------------------------------------------------
Title II funding for fiscal year 1995 under the existing and taxpayer
equity formulas is shown in table V.4.  Relative to the existing
formula, changes in states' allocations under the taxpayer equity
formula would range from a decrease of 77.14 percent to an increase
of 290.91 percent. 



                         Table V.4
          
             Title II FY 1995 Funding Under the
           Existing and Taxpayer Equity Formulas


                              Taxpayer
                  Existing      equity
                formula FY  formula FY
State or              1995        1995            Percenta
territory       allocation  allocation   Dollars        ge
--------------  ----------  ----------  --------  --------
Alabama         $1,349,942  $3,419,706  $2,069,7    153.32
                                              64
Alaska             100,000     270,771   170,771    170.77
Arizona          1,759,313   1,394,264         -    -20.75
                                         365,049
Arkansas           753,038   1,761,699  1,008,66    133.95
                                               1
California      27,867,193  20,848,641         -    -25.19
                                        7,018,55
                                               2
Colorado         1,980,699   1,131,187         -    -42.89
                                         849,512
Connecticut      2,404,858   4,397,664  1,992,80     82.87
                                               6
Delaware           585,604   1,926,550  1,340,94    228.99
                                               6
District of      2,532,524     915,525         -    -63.85
 Columbia                               1,616,99
                                               9
Florida         17,780,752  14,238,085         -    -19.92
                                        3,542,66
                                               7
Georgia          4,731,696   4,912,075   180,379      3.81
Hawaii             499,350   1,952,008  1,452,65    290.91
                                               8
Idaho              138,867     260,849   121,982     87.84
Illinois         5,577,650   1,582,396         -    -71.63
                                        3,995,25
                                               4
Indiana          1,536,770   3,990,018  2,453,24    159.64
                                               8
Iowa               333,360     588,194   254,834     76.44
Kansas             568,263     724,849   156,586     27.56
Kentucky           643,697   1,505,770   862,073    133.93
Louisiana        2,785,044   3,601,976   816,932     29.33
Maine              228,492     568,826   340,334    148.95
Maryland         4,684,012   1,561,736         -    -66.66
                                        3,122,27
                                               6
Massachusetts    3,776,077   2,273,689         -    -39.79
                                        1,502,38
                                               8
Michigan         2,675,943   2,126,388         -    -20.54
                                         549,555
Minnesota          973,550   2,547,146  1,573,59    161.63
                                               6
Mississippi        954,192   2,085,536  1,131,34    118.57
                                               4
Missouri         2,504,335   1,279,436         -    -48.91
                                        1,224,89
                                               9
Montana            100,000      96,946    -3,054     -3.05
Nebraska           267,083     534,883   267,800    100.27
Nevada             964,174   3,408,013  2,443,83    253.46
                                               9
New Hampshire      175,763      67,697         -    -61.48
                                         108,066
New Jersey       8,958,831   9,696,998   738,167      8.24
New Mexico         479,074   1,195,562   716,488    149.56
New York        29,093,044  16,787,811  12,305,2    -42.30
                                              33
North Carolina   2,414,668   6,826,133  4,411,46    182.69
                                               5
North Dakota       100,000      33,846   -66,154    -66.15
Ohio             2,623,138   7,253,179  4,630,04    176.51
                                               1
Oklahoma         1,050,786   2,373,374  1,322,58    125.87
                                               8
Oregon           1,300,587     950,165         -    -26.94
                                         350,422
Pennsylvania     5,177,510   5,283,271   105,761      2.04
Rhode Island       554,753   1,868,964  1,314,21    236.90
                                               1
South Carolina   2,679,771   6,725,031  4,045,26    150.96
                                               0
South Dakota       100,000      22,856   -77,144    -77.14
Tennessee        1,846,877   4,778,344  2,931,46    158.73
                                               7
Texas           12,636,414   9,992,376         -    -20.92
                                        2,644,03
                                               8
Utah               428,266   1,129,611   701,345    163.76
Vermont            103,727     254,367   150,640    145.23
Virginia         2,642,609   4,808,147  2,165,53     81.95
                                               8
Washington       2,310,797   1,815,450         -    -21.44
                                         495,347
West Virginia      184,768     301,237   116,469     63.04
Wisconsin        1,063,650   2,358,029  1,294,37    121.69
                                               9
Wyoming            100,000      76,242   -23,758    -23.76
Guam                 2,902      10,475     7,573    260.97
Puerto Rico      7,682,087   3,973,781  3,708,30    -48.27
                                               6
Virgin Islands           0     278,726   278,726       N/A
==========================================================
Total           $174,766,5  $174,766,5
                        00          00
----------------------------------------------------------
Note:  N/A = Not applicable. 

As demonstrated by these tables, changing the existing title I and II
formulas would redistribute funds across EMAs and states.  Compared
with the existing formulas, either the beneficiary or the taxpayer
equity formula would increase funding for more EMAs and states than
it would decrease it (see table V.5). 



                         Table V.5
          
          Number of Grantees Experiencing Funding
                  Increases and Decreases

                                 Increased       Decreased
Type of formula                    funding         funding
--------------------------  --------------  --------------
Title I beneficiary equity              31              11
Title II beneficiary                    37              16
 equity
Title I taxpayer equity                 22              20
Title II taxpayer equity                33              20
----------------------------------------------------------
Nonetheless, under either formula, funding would decrease for some
EMAs and states if appropriations remained stable.  A number of
mechanisms could be employed, however, to avoid EMA and state funding
losses when funding equity is improved.  Appropriations could be made
to a level that obviates funding losses, hold-harmless provisions
could be applied, or a limit could be placed on the amount of funds
eligible for redistribution.  For example, funding losses could be
avoided by only redistributing funds that were appropriated in excess
of a previous year's amount. 


COMMENTS FROM THE DEPARTMENT OF
HEALTH AND HUMAN SERVICES
========================================================== Appendix VI

We solicited comments on our report from the Department of Health and
Human Services (HHS) through the Director, Office of HIV/AIDS Policy,
Health Education and Human Services.  He provided us his comments,
along with those of the Division of HIV Services, Health Resources
Services Administration, which administers titles I and II of the
CARE Act.  We also received comments from officials of the Centers
for Disease Control and Prevention (CDC). 

In their general comments, the HHS officials stated that moving to a
more equitable funding formula could cause significant funding
changes and potential disruption to service delivery structures.  We
share these concerns and have discussed methods to avoid these kinds
of difficulties in our report. 

The HHS officials also raised questions about the appropriateness of
the Medicare Hospital Wage Cost (MHWC) Index as a proxy for
estimating labor costs for AIDS and HIV services among EMAs and
states.  While we believe a wage index that is more closely related
to these services would be preferable to the MHWC Index, we were
unable to locate such an index.  On the basis of our discussions with
experts, however, we determined that the MHWC Index would be an
appropriate alternative to a wage index that is specific to AIDS and
HIV services. 

In addition, the HHS officials expressed concerns about the adequacy
of both the level of funding and the health care infrastructure for
AIDS and HIV services.  While problems may exist with regard to
funding and infrastructure, these issues were not within the scope of
our study.  The HHS officials provided specific comments about our
report, which have been incorporated as appropriate. 

The CDC officials indicated that they agreed that a caseload
indicator based on an estimate of living cases was preferable to the
existing measure; however, they recommended the use of the number of
AIDS cases reported during the previous 2 years rather than our
proposed measure of weighted cases.  These officials stated that our
caseload measure would require annual revision, would serve as an
incentive to states to underreport AIDS-related mortality, would be
technically difficult to compute, and was not a standard method for
estimating living AIDS cases. 

We agree that our caseload measure might periodically need revision,
as would any such measure, in accordance with significant changes in
AIDS mortality.  However, our caseload measure, when adjusted over
time, would more appropriately reflect the impact of changes in AIDS
mortality on the number of people living with AIDS than would a
measure based on cases reported in the previous 2 years.  We do not
believe our proposed measure would serve as an incentive to
underreport AIDS-related mortality because states' funding would not
be directly affected by their reported mortality data.  As discussed
in appendix I, we propose the use of weighted AIDS cases as a proxy
measure rather than an actual estimate of living AIDS cases to avoid
this potential incentive.  Finally, we do not believe that our
proposed measure would be technically difficult to compute. 


GAO CONTACTS AND STAFF
ACKNOWLEDGMENTS
========================================================= Appendix VII

GAO CONTACTS

Jerry Fastrup, Assistant Director, (202) 512-7211
Greg Dybalski, Senior Economist, (202) 512-7210
Mark Vinkenes, Senior Social Science Analyst, (202) 512-6841

ACKNOWLEDGMENTS

In addition to those named above, the following individuals made
important contributions to this report:  David Bieritz, Evaluator;
Leslie Albin, Reports Analyst; Ann McDermott, Publishing Adviser. 


BIBLIOGRAPHY
============================================================ Chapter 0

Federal Security Agency, Social Security Administration.  The
Principle of Equalization Applied to the Allocation of Grants-in-Aid. 
Memorandum No.  66.  September 1947. 

Institute for Health and Aging, University of California, San
Francisco.  Review and Evaluation of Alcohol, Drug Abuse, and Mental
Health Services Block Grant Allotment Formulas, Final Report.  1986. 

Office of Education Research and Improvement, U.S.  Department of
Education.  Poverty, Achievement, and the Distribution of
Compensatory Education Services.  1986. 

Office of State and Local Finance, Department of the Treasury. 
Federal-State-Local Fiscal Relations:  Report to the President and
Congress.  September 1985. 

U.S.  General Accounting Office.  Older Americans Act:  Funding
Formula Could Better Reflect State Needs (GAO/HEHS-04-41, May 12,
1994). 

U.S.  General Accounting Office.  Maternal and Child Health:  Block
Grant Funds Should Be Distributed More Equitably (GAO/HRD-92-5, Apr. 
2, 1992). 

U.S.  General Accounting Office.  Highway Funding:  Federal
Distribution Formulas Should Be Changed (GAO/RCED-86-114, Mar.  31,
1986). 

U.S.  General Accounting Office.  Changing Medicaid Formula Can
Improve Distribution of Funds to States (GAO/GGD-83-27, Mar.  9,
1983). 


*** End of document. ***