Health Care Alliances: Issues Relating To Geographic Boundaries (Letter
Report, 04/08/94, GAO/HEHS-94-139).

A common feature of many health reform bills is the creation of public
or private health purchasing groups, known as alliances. These entities
have been proposed mainly as a way to broaden coverage, pool risks, give
consumers a choice of health care plans, and disseminate information on
the costs and quality of plans.  The major health reform proposals
relying on alliances, however, have boundary provisions that raise
concerns.  These concerns include the potential for gerrymandering,
changing the provision and receipt of health care, segmenting high-risk
groups, and isolating underserved areas.

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

 REPORTNUM:  HEHS-94-139
     TITLE:  Health Care Alliances: Issues Relating To Geographic 
             Boundaries
      DATE:  04/08/94
   SUBJECT:  Health care planning
             Health insurance cost control
             Health services administration
             Health care programs
             State-administered programs
             Health care services
             Insurance premiums
             Proposed legislation
             Program evaluation
IDENTIFIER:  Health Security Act
             Managed Competition Act of 1993
             Health Equity and Access Reform Today Act of 1993
             CalPERS Health Benefits Program (CA)
             CalPERS PERSCare Plan (CA)
             CalPERS PERSChoice Plan (CA)
             California
             Clinton Health Care Plan
             National Health Care Reform Initiative
             Michigan
             Florida
             Washington
             
**************************************************************************
* This file contains an ASCII representation of the text of a GAO        *
* report.  Delineations within the text indicating chapter titles,       *
* headings, and bullets are preserved.  Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are           *
* identified by double and single lines.  The numbers on the right end   *
* of these lines indicate the position of each of the subsections in the *
* document outline.  These numbers do NOT correspond with the page       *
* numbers of the printed product.                                        *
*                                                                        *
* No attempt has been made to display graphic images, although figure    *
* captions are reproduced. Tables are included, but may not resemble     *
* those in the printed version.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


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


Report to the Chairman, Committee on Finance, U.S.  Senate

April 1994

HEALTH CARE ALLIANCES - ISSUES
RELATING TO GEOGRAPHIC BOUNDARIES

GAO/HEHS-94-139

Health Alliance Boundaries


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

  CDBG - Community Development Block Grant program
  CMSA - Consolidated Metropolitan Statistical Area
  HCFA - Health Care Financing Administration
  HMO - health maintenance organization
  HRS - Health and Rehabilitative Services
  MSA - Metropolitan Statistical Area
  OMB - Office of Management and Budget
  PMSA - Primary Metropolitan Statistical Area

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


B-256940

April 8, 1994

The Honorable Daniel Patrick Moynihan
Chairman, Committee on Finance
United States Senate

Dear Mr.  Chairman: 

A common feature of many health reform bills is the creation of
public or private health purchasing groups, commonly called
alliances.\1 These entities have been proposed primarily as a means
for broadening coverage, pooling risks, providing consumers with a
choice of health care plans, and disseminating information on the
costs and quality of plans.  Decisions about the number, size, and
characteristics of people within each alliance will significantly
affect the price of insurance and access to health care.  Three major
legislative proposals incorporate alliances:  the Clinton bill, the
Chafee/Dole bill, and the Cooper/Breaux bill.\2 Though differing in
both major and minor ways, alliances in these and more recent bills,
such as the Stark proposal, embody the same basic concept of pooled
purchasing.  Moreover, a growing number of states are turning to
statewide cooperatives as a way of increasing the overall level of
insurance, especially among those who work for small businesses. 

On December 23, 1993, you requested a study on issues related to
geographic boundaries for proposed regional health alliances amidst
concerns about the impact of how alliance boundaries might be drawn. 
All of these proposals leave the tough choices of alliance boundaries
to the states.  Providers, insurers, and consumers are already
lobbying at the federal and state levels about real or perceived
problems relating to how alliances may be structured.  These concerns
include the potential for establishing boundaries that (1) unfairly
advantage particular geographic and socioeconomic groups regarding
the cost and quality of health care, (2) change the way consumers
receive care and the markets in which health providers operate, (3)
segment various high-risk groups, and (4) isolate underserved areas,
such as rural regions and urban centers.  On February 24, 1994, we
testified before your committee on the preliminary results of our
work. 

This report expands on the information contained in our testimony. 
Specifically, we discuss the (1) provisions of major health reform
bills concerning the configuration of alliance boundaries; (2)
features and procedures for establishing Metropolitan Statistical
Areas (MSAs), which are important geographic units contained in each
proposal that influence how boundaries are to be drawn; (3)
experiences of two states that have established entities similar to
alliances; and (4) issues relating to the potential effects of
alliance boundaries on existing health markets, access to health
care, and distribution of health care costs within a state. 


--------------------
\1 Two proposals refer to purchasing cooperatives or purchasing
groups.  For our discussion, we will refer to these entities as
alliances. 

\2 The formal names for the three major reform bills are as follows: 
(1) the Clinton bill, the Health Security Act (S.  1757/H.R.  3600);
(2) the Cooper/Breaux bill, the Managed Competition Act of 1993
(S.  1579/H.R.  3222); and (3) the Chafee/Dole bill, the Health
Equity and Access Reform Today Act of 1993 (S.  1770/H.R.  3704). 


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

In each of the three major health reform proposals, decisions on
alliance boundaries are left to the states except for provisions in
all three bills that require that MSAs remain intact, primarily as a
means to prevent discrimination against high-risk populations by
health plans.  The bills vary on whether alliance boundaries can
cross state lines, the minimum population size requirement for an
alliance area, and the number of alliances that can operate in each
coverage area.  All bills permit health plans to operate across state
lines or alliance boundaries. 

Because MSAs are important to drawing alliance boundaries in each
bill, concerns have been raised about how they are defined.  The
Office of Management and Budget (OMB) defines MSAs using statistical
information furnished by the Bureau of the Census and occasionally
considers the views of local officials.  Direct congressional action
has also led in a few instances to changes in MSA definitions.  In
the future, if changes in MSA definitions require states to
reconfigure their alliance boundaries, the implications for health
plans and health delivery may be substantial. 

Florida and Washington, two states that have already begun the
process of implementing health care reform, illustrate the intensity
of decisions related to the size, number, and boundaries of
alliances.  For example, Florida legislators drew upon existing
health planning districts in a political compromise to create 11
alliance areas for the state, ranging in population from about
500,000 to over 2 million.  Moreover, legislation establishing four
alliances in Washington outlined broad parameters for
boundaries--namely that these determinations be based on population,
geography, and other factors--but left the ultimate decision to the
state's Health Services Commission. 

The number of alliances that states would ultimately create and the
placement of the alliance boundaries have raised questions for
consumers, employers, and providers.  These concerns relate to
whether or by how much the creation of alliance boundaries will
impact the provision of care in existing health markets, segment and
limit access to care for disadvantaged or high-risk populations, and
redistribute health care costs among different geographic and
socioeconomic groups. 


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

Health alliances use pooled buying power of a large group of
individuals and employers to develop a more affordable insurance
product by spreading the risk over a larger population.  Small groups
and individuals are particularly disadvantaged in today's insurance
market.  Some cannot obtain insurance at any price because of their
actual or perceived health status.  Through the larger risk pooling
created by alliances, small employers and individuals can gain or
increase bargaining clout with health insurers, plans, and providers
and allow individuals a greater choice of health plans. 

The health alliance in the Clinton administration's proposal, the
health plan purchasing cooperative in the Cooper/Breaux bill, and the
purchasing group in the Chafee/Dole bill all draw their basic
structure from the managed competition approach to health care
reform.  While the bills differ on several key aspects, they all
serve as an organization through which employers or individuals
purchase their health insurance.  These alliances generally help
administer subsidies for low-income members, provide members with
information on the costs and quality of plans, and allocate collected
premiums to health plans. 

Each proposal is different in such areas as whether alliances can
negotiate premiums, whether the purchase of insurance through the
alliance is required, whether employers have to contribute to
premiums, and what segments of the population can be covered by
alliances.  Nonetheless, a substantial share of the population is
eligible to obtain its insurance coverage through these alliances. 
Because all three proposals may place enrollees in the alliance that
covers the area they live in, there are concerns that the geographic
boundaries defined by the states could affect access to particular
providers and the price of health insurance. 

To gain perspective on the potential issues that could arise because
of a state's choice of alliance boundaries, we reviewed the
legislation on geographic boundary limits in each proposal as well as
the literature and positions of interest groups on geographic
boundary issues.  We also made site visits to Florida and Washington,
where some decisions regarding the location of alliance boundaries
have already been made within the context of state reform efforts. 
We also drew upon our previous work and current efforts in assessing
existing public and private alliances that have been in operation for
some time.\3

Several geographic issues discussed in the following sections do not
specifically pertain to the provisions in any of the health reform
proposals.  These include concerns regarding regional differences in
the adequacy, availability, and choices of health care providers in
underserved rural and central city areas.  While some provisions of
the various health reform proposals affect these concerns, where or
how a geographic boundary is drawn probably cannot correct problems
of access to health services for all citizens in a defined alliance
area. 

To understand the process for establishing MSAs, we reviewed various
OMB documents that described the definitions of and guidelines for
establishing these areas as well as annual bulletins and other
reports that provided the rationale for changes in metropolitan area
definitions.  In addition, we interviewed OMB officials to obtain
their positions on various aspects of the agency's process for
changing definitions and on the nature of recent changes to
metropolitan areas. 

Our work was performed between December 1993 and March 1994 in
accordance with generally accepted government auditing standards. 


--------------------
\3 See Health Insurance:  California Public Employees' Alliance Has
Reduced Recent Premium Growth (GAO/HRD-94-40, Nov.  22, 1993). 


   GEOGRAPHIC BOUNDARY PROVISIONS
   CONTAINED IN REFORM BILLS
------------------------------------------------------------ Letter :3

Each of the three health reform proposals we examined gives the
states responsibility for and flexibility in establishing alliance
boundaries, with only a few constraints (see table 1). 



                           Table 1
           
               Geographic Provisions of Health
                   Proposals for Alliances

                              Cooper/
               Clinton Plan   Breaux Plan    Chafee/Dole
               (S. 1757/      (S. 1579/      Plan (S. 1770/
               H.R. 3600)     H.R. 3222)     H.R. 3704)
-------------  -------------  -------------  ---------------
Alliance can   No             No             No
subdivide an
MSA

Number of      One            One            None, one, or
alliances                                    more than one
that operate
in each
coverage area

Alliance can   No             Yes            Yes
cross state
lines

Minimum size   None--         Minimum        Minimum 250,000
requirement    National       250,000        individuals
for alliance   Health Board   eligible       residing in
area           reviews for    individuals    alliance area
               sufficient     residing in
               market size    alliance
                              area\a
------------------------------------------------------------
\a Individuals, and their families, who are unemployed,
self-employed, employed in firms of fewer than 101 workers, or
Medicaid-eligible are generally considered to be eligible for
coverage through an alliance. 

In all three legislative proposals, alliance boundaries are not
permitted to subdivide an MSA\4 or, in effect, a Primary Metropolitan
Statistical Area (PMSA).\5 Both the Chafee/Dole and Cooper/Breaux
bills require that designated alliance areas have a minimum
population base of 250,000.  While the Clinton plan does not specify
a number, it does require that the alliance area include a population
sufficiently large to provide the alliance with bargaining power to
promote competition among plans. 

Both the Clinton and Cooper/Breaux plans specify that a single
alliance will operate in each area.  The Chafee/Dole plan only
requires that the state designate health care coverage area
boundaries; if one (or more) alliance forms, then it must serve the
entire coverage area.\6 The Clinton plan does not permit alliance
boundaries to cross state lines; however, both the Cooper/Breaux and
Chafee/Dole plans permit alliance boundaries to cross state lines. 
All bills permit health plans to operate across state lines or
alliance boundaries. 


--------------------
\4 A metropolitan area consists of a large population center and
adjacent communities that have a high level of economic and social
integration with that population center.  Metropolitan areas are
classified as a Metropolitan Statistical Area (MSA) or a Consolidated
Metropolitan Statistical Area (CMSA).  CMSAs, which contain 1 million
or more people, consist of at least two separate statistical areas
called Primary Metropolitan Statistical Areas (PMSA) (see app.  I). 

\5 In the Clinton proposal, an alliance can subdivide an MSA or a
PMSA if that area crosses state lines. 

\6 Unlike the other two bills, the Chafee/Dole bill permits the
creation of competing alliances.  A single alliance may operate in
more than one coverage area. 


   MAINTAINING METROPOLITAN AREAS
   CENTRAL TO THE THREE PROPOSALS
------------------------------------------------------------ Letter :4

Each health care proposal requires states to keep MSAs intact when
defining alliance boundaries, primarily to prevent discrimination of
disadvantaged or high-risk groups by health plans.  While some of the
largest disparities in income distribution are found between inner
city and suburban areas within MSAs, there may also be differences in
income and other characteristics among contiguous MSAs and between
metropolitan communities and rural areas.  While the requirement that
MSAs remain intact may prevent some redlining that isolates areas
with high-risk populations, potential gerrymandering in defining
alliance boundaries could be a problem. 

Future issues may emerge if changes in MSA definitions require states
to reconfigure their alliance boundaries.  Over the past decade,
changes in MSA definitions have generally affected only a few areas
of the country.  Changes were based primarily on a yearly evaluation
of statistical data in relationship to criteria established each
decade by OMB.  However, in selected cases such decisions have also
been based on local opinion or congressional action.  Given the
potential importance of health alliance boundaries, there are
concerns that a change in the definition of an MSA by OMB may require
states to reconfigure their alliance boundaries. 

Many are concerned that future issues may emerge if changes in MSA
definitions require states to reconfigure their alliance boundaries. 
Other concerns center around the extent to which political
influences, namely the opinion of local officials and congressional
action, may affect OMB's definition of MSAs given the potential
importance of health alliance boundaries (see app.  I for more
information on metropolitan areas). 


   RECENT EXPERIENCES IN FLORIDA
   AND WASHINGTON ILLUSTRATE THE
   POLITICAL PROCESS INVOLVED IN
   DETERMINING ALLIANCE BOUNDARIES
------------------------------------------------------------ Letter :5

Florida and Washington have already faced the difficult decisions
required in defining boundaries for alliance-like structures as part
of their health reform legislation.  Their experiences may be
instructive as to the different points of view regarding the size,
number, and boundaries of alliances. 

Using the existing geographical structure of its Health and
Rehabilitative Services (HRS) planning districts, Florida legislators
divided the state into 11 separate alliance areas, ranging in
population from about 500,000 to over 2 million (see fig.  1). 

   Figure 1:  Metropolitan Areas
   and Health Alliances in Florida

   (See figure in printed
   edition.)

   Note:  Shaded regions represent
   metropolitan areas.

   (See figure in printed
   edition.)

Initial legislative proposals anticipated five to six alliances based
on health market areas, but market areas are not well defined and
local leaders could not agree on their specific boundaries.  Thus,
the Florida Legislature compromised by relying on existing HRS
planning districts.  However, the legislators provided for the option
of future mergers of up to three contiguous alliances that are not
primarily urban into a single alliance. 

Florida's alliance boundaries generally conform to the proposed
requirements of the national health reform bills.  However, portions
of the Tampa-St.  Petersburg-Clearwater MSA are included in three
separate alliances.  Also, the smaller alliances in the Florida
panhandle may not meet the Cooper bill requirement of a minimum
250,000 eligible individuals. 

Alliance boundaries established under the Washington Health Services
Act of 1993 also reflected political compromise.  The legislation
authorized the creation of four alliances and left to the state's
Health Services Commission the decision on specific boundaries for
these alliances.  The legislation also requires that the decision be
based on population, geographic factors, market conditions, and other
factors deemed appropriate by the commission.  The legislation
specified only that the population covered by an alliance should be
at least 150,000, which is smaller than the minimum size required
under the Cooper and Chafee plans.  The Washington Senate would have
preferred two alliance areas; the Washington House was concerned
about the potential power of larger alliances and wanted 10 areas. 


   STATES' PLACEMENT OF BOUNDARIES
   RAISES SOME CONCERNS
------------------------------------------------------------ Letter :6

Consumers, employers, and providers have raised questions about the
number of alliances that states will create and how boundary lines
will be drawn.  While these concerns cover a broad spectrum,
questions have surfaced in the key areas of how the creation of
alliance boundaries will affect the provision of care in existing
health markets, segment and limit access to care for disadvantaged or
high-risk populations, and redistribute health care costs among
different geographic or socioeconomic groups. 


      POTENTIAL IMPACT ON THE
      PROVISION OF CARE
---------------------------------------------------------- Letter :6.1

Individuals seeking insurance through the alliance that includes the
area they live in may have concerns about whether they will still be
able to use physicians, hospitals, and other health care facilities
that may be located outside the boundaries of their alliance. 
Similarly, physicians, hospitals, and other providers may also have
concerns as to whether they will be able to maintain the part of
their patient base that is located in another alliance area.  Whether
these concerns are justified depends more on the service areas
covered and provider networks and coordination mechanisms developed
by health plans than the geographic boundaries of alliances. 

Perhaps the more important issue is whether the structure of the
alliances will make coordination across areas and development of
broad-ranging networks by health plans easier or more difficult.  On
the one hand, the creation of a standard benefits package and the
broader coverage expected under these plans could make coordination
easier.  On the other hand, coordination could be more difficult if
states or alliances have different requirements for the collection
and dissemination of provider data.  This could result in health
plans not seeking certification, and thus the permission, to operate
in multiple alliances or states.  Similarly, if alliance fee
schedules are not roughly comparable, providers may avoid serving
patients from neighboring alliances. 

Obviously, the larger the number of alliances established, the more
coordination there will have to be, and, possibly, the higher the
administrative costs.  Ultimately, plans will have to assess whether
the benefits of operating in a different alliance area outweigh the
costs incurred in terms of meeting any additional requirements. 

Coordination could be most critical in areas where alliance
boundaries separate existing health markets.  This may be likely in
the 41 metropolitan areas that span state boundaries such as in the
Washington, D.C., and Philadelphia metropolitan areas. 

Administration officials contend that coordination should be no more
difficult than it is today, when plans operate across state lines. 
While the necessary coordination is anticipated under reform, no
provisions in the Clinton bill explicitly provide mechanisms or
incentives for this coordination. 

The Cooper/Breaux and Chafee/Dole bills also contain stipulations to
minimize the impact of alliance boundaries on the provision of care
for individuals and providers.  As with the Clinton proposal, they
permit plans to operate in multiple alliances or states and allow
states to coordinate their plan requirements.  Further, to keep
health markets that span state lines intact, the Cooper/Breaux and
Chafee/Dole bills allow multistate alliances.  Interstate cooperation
would be needed to create these alliances, and additional issues
could arise, such as the creation of an adequate oversight mechanism
for and the inclusion of Medicaid-eligible populations in multistate
alliances.  Neither of these bills specifies the mechanisms or
incentives to do so. 


      POTENTIAL RISK SEGMENTATION
      AND LIMITED ACCESS TO CARE
---------------------------------------------------------- Letter :6.2

Other concerns center around whether some alliances within a state
will have a disproportionate share of a state's high-risk population. 
Such alliances could have greater difficulty attracting a sufficient
number of health plans that would offer consumers an adequate choice
of plans.  The extent to which boundaries could cause this to happen
depends on factors like the number of alliances in a state and
whether states have metropolitan areas with markedly different
demographic profiles.  For example, some isolation of high-risk
communities could occur if states created a number of geographically
smaller alliances, such as one alliance for each metropolitan area. 
Such risk segmentation could occur in areas with specific
characteristics, such as unusual industrial, environmental, or
epidemiological conditions (for example, the West Virginia coal
mining region or areas with large concentrations of Acquired Immune
Deficiency Syndrome cases).  Moreover, risk segmentation could also
exist when two adjacent MSAs have different proportions of Medicaid
populations, as in the case of two primary metropolitan areas in
southern Florida.  For example, 16 percent of the population in the
Miami PMSA is eligible for Medicaid compared with only 8 percent for
the neighboring Ft.  Lauderdale PMSA. 

Isolation of rural areas depends largely on whether states choose to
separate rural areas in establishing alliance boundaries.  Because
the MSA rule has little relevance to rural areas, states could
establish boundary lines to segment rural populations that are
potentially high-risk or underserved.  The Cooper/Breaux and
Chafee/Dole requirements that alliance areas have a population of at
least 250,000 and the Clinton requirement that alliance population
size be sufficiently large to promote competition among plans make
segregation of rural areas difficult or unlikely. 

Further, risk segmentation may also occur on the plan level if plans
are not required to provide services throughout an alliance or
metropolitan area.  The Clinton bill contains a provision that allows
states to require a health plan to cover all or selected portions of
an entire alliance area.  The Chafee/Dole bill requires every
alliance to service an entire coverage area.  However, as with the
Cooper/Breaux bill, the Chafee/Dole proposal apparently has no
provisions regarding health plan service areas.  Minnesota is
attempting to address this problem in its reform initiative by
dividing the state into 20 health service areas.  Any plan operating
in a particular service area must demonstrate that it provides a
reasonable level of access to care for those in all geographic areas
within that health service area. 

Providing adequate care in rural areas has long been a challenge, and
doubts have been expressed about whether the managed competition
concept even has applicability to such areas.  For example, the
California Public Employees' Retirement System (CalPERs) health
alliance serving state and local workers throughout California
illustrates the limited choices that can exist in rural areas.  While
CalPERs offers a fee-for-service plan and over 20 health maintenance
organizations (HMO) plans to its members, few HMOs operate in the
more rural and remote areas of the state.  Thus, rural residents tend
to choose the more expensive fee-for-service plan under CalPERs in
large measure because their choice is restricted. 


      REDISTRIBUTION OF HEALTH
      CARE PREMIUMS
---------------------------------------------------------- Letter :6.3

Another question that has been asked about alliance boundaries is
whether boundaries will be drawn in such a way as to redistribute
health costs among different groups.  Under each proposal some people
may pay more for insurance than they do now, and those extra payments
will indirectly subsidize other people who will pay less than before. 
In general, however, such redistribution is less a consequence of new
health alliances than of health insurance reform.\7 Currently, most
individual firms pay premiums that reflect the health status and
medical costs of their workers.  Firms with a few high-risk workers
may be unable to get insurance unless they exclude those workers. 
Since a major goal of health care reform is to provide guaranteed
access to affordable insurance, covering these high-risk people will
necessarily entail that some of their costs will be paid by others. 

While cost redistribution is inevitable under reform, alliance
boundaries could affect whose premiums change and by how much. 
Larger alliances would provide greater risk sharing among a state's
population, but this could result in some persons paying higher
premiums.  Because premiums will be community-rated, persons living
in lower-cost areas would pay more and persons in higher-cost areas
would pay less if health plans attempt to serve the entire alliance
area.  For example, persons in Flint or Saginaw, Michigan, would pay
more if their alliance included Detroit.  At present, average net
health insurance claims costs in the Detroit area are about 20
percent higher than costs in Flint and nearly one-third higher than
in the Saginaw area. 

On the other hand, creation of smaller alliances within a state could
also result in higher premiums for some persons, as disproportionate
shares of high-risk persons are concentrated in some alliances. 
Citizens in those alliances would pay more because of the greater
costs of these high-risk persons. 


--------------------
\7 The demographics of redistribution can take many forms, for
example between high- and low-income groups, between rural and urban
populations, between easy and hard-to-serve areas, or between young
and old individuals.  Exactly which groups are affected by, and the
extent of, the redistribution will likely vary across regions
according to the representation of the different groups within each
region. 


   CONCLUSION
------------------------------------------------------------ Letter :7

Alliances have been proposed as a means for broadening coverage,
pooling risks, providing consumers with a choice of health care
plans, and disseminating information on the costs and quality of
plans.  However, the major health reform proposals relying on
alliances have various boundary provisions that raise concerns. 
These concerns include the potential for gerrymandering, changing the
provision and receipt of health care, segmenting high-risk groups,
and isolating underserved areas. 


---------------------------------------------------------- Letter :7.1

Although we did not obtain official agency comments on this report,
we discussed the information contained in this report on metropolitan
areas with OMB officials.  They generally agreed with our treatment
of the subject, and we incorporated their comments where appropriate. 

We are sending copies of this report to the Secretary of Health and
Human Services and the Director of OMB and other interested parties. 
Please call me on (202) 512-7119 if you or your staff have any
questions concerning this report.  Other major contributors to this
report are listed in
appendix II. 

Sincerely yours,

Sarah F.  Jaggar
Director, Health Financing
 and Policy Issues


FEATURES OF AND PROCEDURES FOR
ESTABLISHING METROPOLITAN AREAS
=========================================================== Appendix I

Each health care proposal requires states to keep Metropolitan
Statistical Areas intact when defining alliance boundaries as a means
to prevent discrimination of disadvantaged or high-risk groups by
health plans.  Concerns have been raised that procedures for defining
MSAs and alliance boundaries could become political decisions that
might affect existing health markets. 

The Office of Management and Budget establishes definitions for MSAs
using statistical information furnished by the Bureau of the Census. 
In addition, changes to MSA definitions have been occasionally
influenced by the views of local officials and congressional action. 
In this appendix, we discuss (1) the concept of metropolitan areas,
(2) how metropolitan areas are used in federal funding programs, (3)
the process used by OMB to define and change MSAs, and (4) how local
opinion and congressional actions have affected various changes to
MSA definitions. 


   THE METROPOLITAN AREA CONCEPT
--------------------------------------------------------- Appendix I:1

In general, a metropolitan area consists of a large population center
and adjacent communities that have common economic and social
characteristics.  Current OMB standards define an MSA as an area
including one city with 50,000 or more inhabitants or an urbanized
area of at least 50,000 inhabitants and a total metropolitan
population of at least 100,000 (75,000 in New England).  Moreover,
counties that contain the largest city are the central county, along
with any adjacent counties that have at least 50 percent of their
population in the urbanized area surrounding the largest city. 
Additional outlying counties are included in the MSA if they meet
various requirements of commuting to the central counties and possess
other metropolitan features (such as population density and percent
urban).  In New England, MSAs are defined in terms of cities and
towns rather than counties.\1

A metropolitan area that has more than 1 million population and meets
certain other requirements of the OMB standards may be classified as
a Consolidated Metropolitan Statistical Area.  Each CMSA consists of
two or more major components recognized as Primary Metropolitan
Statistical Area. 

As of June 1993, OMB recognized 253 MSAs, 76 PMSAs, and 19 CMSAs.\2

The number of metropolitan areas contained in a state can vary
widely; 4 states have only 1 metropolitan area, while 10 states have
over 10.  A sizable number of MSAs and PMSAs--41--cross state lines
(see table I.1). 



                                    Table I.1
                     
                        Metropolitan Areas Crossing State
                                     Borders

Metropolitan statistical areas           Primary metropolitan statistical areas
---------------------------------------  ---------------------------------------
Augusta-Aiken, GA-SC                     Boston, MA-NH

Charlotte-Gastonia-Rock Hill, NC-SC      Lawrence, MA-NH

Chattanooga, TN-GA                       Lowell, MA-NH

Clarksville-Hopkinsville, TN-KY          Portsmouth-Rochester, NH-ME

Columbus, GA-AL                          Worcester, MA-CT

Cumberland, MD-WV                        Cincinnati, OH-KY-IN

Davenport-Moline-Rock Island, IA-IL      Newburgh, NY-PA

Duluth-Superior, MN-WI                   Philadelphia, PA-NJ

Evansville-Henderson, IN-KY              Wilmington-Newark, DE-MD

Fargo-Moorhead, ND-MN                    Portland-Vancouver, OR-WA

Fort Smith, AR-OK                        Washington, DC-MD-VA-WV

Grand Forks, ND-MN

Huntington-Ashland, WV-KY-OH

Johnson City-Kingsport-Bristol, TN-VA

Kansas City, MO-KS

La Crosse, WI-MN

Las Vegas, NV-AZ

Louisville, KY-IN

Memphis, TN-AR-MS

Minneapolis-St. Paul, MN-WI

New London-Norwich, CT-RI

Norfolk-Virginia Beach-Newport News,
VA-NC

Omaha, NE-IA

Parkersburg-Marietta, WV-OH

Providence-Fall River-Warwick, RI-MA

St. Louis, MO-IL

Sioux City, IA-NE

Steubenville-Weirton, OH-WV

Texarkana, TX-Texarkana, AR

Wheeling, WV-OH
--------------------------------------------------------------------------------
The metropolitan area concept is intended to provide a nationally
consistent set of definitions for collecting, tabulating, and
publishing federal statistics.  The federal government uses these
definitions for purposes such as Census Bureau statistics on
population, housing, industry, trade, as well as in Bureau of Labor
Statistics data on employment, payroll, and labor markets.  The
private sector also uses the metropolitan area definitions for
marketing research. 


--------------------
\1 Because of this unique city/town configuration, OMB has a special
definition for this sector:  New England County Metropolitan Areas
(NECMAs).  However, NECMAs do not replace New England metropolitan
areas as the standard area. 

\2 These totals include 3 MSA's, 3 PMSAs, and 1 CMSA in Puerto Rico. 


      PROGRAMMATIC USES OF
      METROPOLITAN AREA
      DEFINITIONS
------------------------------------------------------- Appendix I:1.1

Although initially established for statistical purposes, metropolitan
area definitions are used by various federal agencies to structure
the geographic basis for allocating federal funds.  Some examples
include the following: 

  The Farmers Home Administration makes rural housing loans in towns
     of 10,000 to 20,000 population only if they are located outside
     of metropolitan areas. 

  The Community Development Block Grant program (CDBG) improves the
     housing environment and economic opportunities of low and
     moderate income persons.  The Department of Housing and Urban
     Development targets 70 percent of CDBG funds to so-called
     entitlement communities (cities of 50,000 or more, or central
     cities of metropolitan areas and metropolitan counties of more
     than 200,000 population, excluding the entitlement cities). 
     Thirty percent of the funds go to nonentitlement communities,
     which may be located either within or outside a metropolitan
     area.  The CDBG program uses various data at the metropolitan
     area level in formulas designed to determine funding levels. 

  The Health Care Financing Administration (HCFA) uses metropolitan
     areas for Medicare payment purposes, along with other factors. 
     HCFA's Medicare payments for inpatient hospital services are
     partially based on whether a hospital is located within a
     metropolitan area.  Medicare reimbursements are higher for
     hospitals located in metropolitan areas than those in
     nonmetropolitan areas.  HCFA uses statistical data for
     metropolitan areas in its payment formula, which includes a
     hospital wage index relating the amount hospitals pay to treat
     particular illnesses.  In addition, HCFA establishes cost levels
     for reimbursing home health agencies and cost limits for routine
     service in skilled nursing facilities based on metropolitan
     areas. 

As previously noted, metropolitan areas are important in determining
the allocation of federal funds.  The provision of major health
reform proposals that states may not separate MSAs when drawing
alliance boundaries would represent another important application of
the metropolitan area concept. 


   OMB'S PROCESS FOR DEFINING
   METROPOLITAN AREAS
--------------------------------------------------------- Appendix I:2

As authorized by the Paperwork Reduction Act of 1980 (44 U.S.C. 
3504), OMB establishes the standards for defining metropolitan areas. 
These standards are developed and published in the Federal Register
before each decennial census.  In developing the standards, OMB
considers comments received directly from the public and during a
public hearing, and recommendations from a 15-member Federal
interagency committee on metropolitan areas.  In general, OMB
standards outline baseline statistical criteria that must be met for
classification as a particular type of metropolitan area.  Moreover,
the standards require that, for selected cases, OMB solicit and
consider local opinion, which is a reflection of public views on the
application of the standards. 

Major revisions to metropolitan area definitions are made after each
decennial census, when the Census Bureau provides OMB with population
and commuting data.  A 15-member federal interagency committee on
metropolitan areas applies the standards to the census data, and
considers local opinion in some instances, to develop revised
definitions.  The committee then submits its recommendations to OMB
for final approval.  The OMB Director makes the final decision on all
changes to MSA definitions. 

Minor changes to metropolitan area definitions occur in June of each
year between the decennial census.  These intercensal changes are
largely based on the Census Bureau's annual population estimates,
which identify areas that are close to meeting the specific
statistical thresholds for revision.  Since the mid-1980s, these
changes have consisted chiefly of adding or deleting metropolitan
areas based on population changes. 


      DECENNIAL CENSUS PROVIDES
      BASELINE STATISTICAL DATA
      FOR DEFINING METROPOLITAN
      AREAS
------------------------------------------------------- Appendix I:2.1

OMB standards for defining metropolitan areas require that an area
meet various statistical thresholds for classification as a
particular type of metropolitan area.  In addition, the standards
outline the requirements for designating the central city or county
within metropolitan areas.  For determining metropolitan areas and
their central cities or counties, data from the decennial census are
applied to the standards.  Although extensive, these statistical
requirements primarily relate to population density as well as
commuting patterns of employed persons.  For example, for
classification as one of possibly several central cities within an
MSA, census data must show that the city has either the largest
population or at least 250,000 people. 


      LOCAL OPINION CONSIDERED IN
      SELECTED CASES
------------------------------------------------------- Appendix I:2.2

OMB recognizes that the statistical requirements may not always be
sufficient for determining changes to metropolitan area definitions. 
Therefore, in selected instances, OMB considers local opinion in
making final decisions.  Local opinion is considered for

  combining two adjacent metropolitan areas (of fewer than 1 million
     people) whose central cities are within 25 miles of each other;

  identifying PMSAs within CMSAs;

  assigning a county or place that, based on commuting, is eligible
     for inclusion in more than one metropolitan area; and

  titling PMSAs and CMSAs. 

OMB solicits local opinion on a matter through the appropriate
congressional delegation.  In this situation, OMB sends letters to
each member of the congressional delegation urging him or her to
contact a wide range of groups in his or her communities, including
business and other leaders, the Chamber of Commerce, planning
commissions, and local officials.  The letter contains the various
options OMB is considering for the matter, which are derived from an
application of the statistical criteria.  For example, from a review
of statistical data, local officials for one MSA could be asked
whether they would prefer to remain a separate entity or become
combined with another MSA.  After the views of local officials and
other citizens are obtained, the congressional member indicates the
consensus by placing a mark next to the appropriate option. 

Once OMB receives local opinion on a matter, the interagency
committee considers all comments it received along with relevant
statistical data.  Based on its review, the committee makes its
recommendation to OMB for final decision.  After deciding on the
matter, OMB will not again request local opinion on the same matter
until after the next decennial census. 


   CHANGES TO METROPOLITAN AREA
   DEFINITIONS BASED ON POLITICAL
   FACTORS
--------------------------------------------------------- Appendix I:3

Over the past decade, some changes to definitions of metropolitan
areas have been shaped by political factors that extend beyond the
statistical criteria established by OMB.  The views of local
officials have played a role in OMB decisions, as in a 1993 decision
affecting the Nassau-Suffolk,
New York, PMSA.  Furthermore, in other instances, Members of the
Congress have adopted changes in metropolitan area definitions
through legislation.  A discussion of these issues follows. 


      LOCAL OPINION A FACTOR IN
      OMB'S RECENT DECISION ON
      NASSAU-SUFFOLK, NEW YORK,
      PMSA
------------------------------------------------------- Appendix I:3.1

Local opinion has played a role in some changes to metropolitan area
definitions since 1980.  Specifically, in 1993, Census population
data showed that the two-county PMSA of Nassau-Suffolk, New York, was
eligible for separation.  OMB then asked the appropriate
congressional members to solicit local views on whether the
jurisdictions should remain a two-county PMSA, or split up so that
Suffolk County would be recognized as a one-county PMSA, and Nassau
County would be included in the PMSA with New York City.  Receiving
no response from the congressional delegation, OMB ruled to include
Nassau-Suffolk in the New York PMSA.  When OMB announced this
definition, Nassau County officials protested to OMB about the change
and were later given another opportunity to respond to the issue. 
Their subsequent responses resulted in OMB maintaining the two-county
Nassau-Suffolk PMSA. 

The controversy over the Nassau-Suffolk redesignation centered around
HCFA's Medicare reimbursements to hospitals.  According to letters
from members of the New Jersey congressional delegation, New Jersey
stood to gain financially because the change in PMSA had potential to
provide the state with $200 million in federal Medicare
reimbursements.  New Jersey hospitals allegedly had been losing
millions of dollars on Medicare reimbursements.  On the other hand,
New York localities would have lost substantial Medicare
reimbursements.  For example, members of the New York congressional
delegation alleged that hospitals in New York City stood to lose
approximately $121 million in reimbursements. 


      CONGRESSIONALLY MANDATED
      CHANGES TO METROPOLITAN AREA
      DEFINITIONS
------------------------------------------------------- Appendix I:3.2

Other changes to metropolitan area definitions have resulted from
congressional action.  During the 1980s five changes in metropolitan
area definitions resulted from legislative mandates.  Two examples of
such changes are the following: 

  In 1988, Congress passed a law for the exclusive purpose of
     creating the Decatur, Alabama, MSA, which merged the counties of
     Morgan and Lawrence, Alabama, into a single MSA.  According to
     the legislative history, if designated as an MSA, the two
     counties stood to receive additional funding from federal
     programs, such as higher Medicare hospital reimbursements and
     CDBG grants.  Furthermore, MSA status would have allowed these
     communities to address the issue of the decline in certain
     industries by facilitating their joint activities to leverage
     local efforts to develop and expand their economies.  OMB had
     originally ruled that the area fell short of certain population
     density thresholds for MSA classification.  Congressional action
     on this matter resulted because OMB's administrative process did
     not address the unique nature of this situation.  Constituents
     argued that the two counties should not have been denied MSA
     status because OMB's formula for computing population density
     considered an unpopulated area of federal forestland occupying a
     portion of Lawrence County.  The two counties would have
     qualified for MSA designation if the unpopulated forest area
     were excluded from OMB's formula. 

  In the Deficit Reduction Act of 1984, a provision was added to
     create the Kansas City, Missouri-Kansas MSA, which merged Kansas
     City, Missouri, and Kansas City, Kansas, into a single MSA. 
     According to the legislative history, this provision was
     designed to ensure that the two cities and their surrounding
     counties were treated as one MSA for purposes of federal taxes
     and other programs.  Previously, the two cities and associated
     counties had been treated as separate PMSAs. 

The three other legislative mandates to metropolitan area definitions
were enacted as part of continuing resolutions.\3


--------------------
\3 These three changes are as follows:  (1) In 1984, P.L.  98-473
created the St.  Louis, MO-IL MSA.  This change rescinded OMB's
previous designation of the St.  Louis-East St.  Louis-Alton, MO-IL
CMSA, which contained three PMSAs:  St.  Louis, MO; Alton-Granite
City, Illinois; East St, Louis-Belleville, Illinois; (2) In 1986,
P.L.  99-500 added Harvey County, Kansas to the Wichita, Kansas MSA,
and (3) in 1988, P.L.  100-202 added part of Sullivan City in
Crawford County, Missouri, to the St.  Louis, MO-IL MSA. 


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================== Appendix II

Mark V.  Nadel, Associate Director, National and Public Health Issues
Michael Gutowski, Assistant Director, (202) 512-7128
Glenn G.  Davis, Assignment Manager
Rolfe A.  Forland, Evaluator-in-Charge

