Private Health Insurance: Continued Erosion of Coverage Linked to Cost
Pressures (Chapter Report, 07/24/97, GAO/HEHS-97-122).

Pursuant to a congressional request, GAO reviewed major trends in the
private health insurance market, during the 1980s and 1990s, focusing
on: (1) the decline in private health insurance coverage and factors
contributing to this decline; (2) trends in health insurance premiums
and reasons for these trends; and (3) employer's efforts to control
health benefits costs.

GAO noted that: (1) private health insurance coverage has slowly but
steadily declined; (2) between 1980 and 1995, the population under age
65 covered by private health insurance decreased from 79.5 percent to
70.5 percent; (3) coverage for children, early retirees, and near-poor
families has declined faster than that for the overall population; (4)
declining private health coverage has been accompanied by growth in the
uninsured population and Medicaid enrollment, which in turn increased
government health expenditures; (5) a major reason for declining private
health coverage is the rising cost of health insurance; (6) this has
absorbed a growing share of business and family incomes and has
influenced employers and employees' health insurance decisions; (7)
health insurance premiums, in contrast, have alternated between rapid
growth and relative stability over the past two decades; (8) several
reasons cited as contributing to the recent near-zero growth in health
plan premiums include the cyclical nature of health insurance premiums,
the expansion of managed care, the increasingly competitive market for
health insurance, and low overall inflation; (9) nonetheless, the high
level of health insurance costs and uncertainty about future increases
remain a concern of employers and individuals purchasing coverage; (10)
as employers have adopted various strategies to control the costs of
health benefits, the costs consumers bear have increased and the types
of health insurance products they receive has evolved; (11) since 1980,
employees are more likely to be required to pay a share of their health
plans' premiums, typically 20 to 30 percent; (12) in addition, most
Americans are now enrolled in a network-based plan; (13) a network-based
plan requires or encourages enrollees to use physicians and hospitals
affiliated with the plan; (14) however, network-based plans often have
lower deductibles, and most of these plans allow enrollees to use
nonaffiliated providers at a higher cost; (15) nearly 40 percent of
persons with private employer-based health insurance participate in a
self-funded plan; (16) in such plans, unlike conventional health
insurance, the employer assumes the risk for health claims and the plan
is exempt from state insurance regulation; (17) if at some point health
insurance costs start rising again, employers and insurers will face
heightened pressure to control costs; and (18) this may lead to increas*

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

 REPORTNUM:  HEHS-97-122
     TITLE:  Private Health Insurance: Continued Erosion of Coverage 
             Linked to Cost Pressures
      DATE:  07/24/97
   SUBJECT:  Health insurance
             Insurance premiums
             Health insurance cost control
             Fringe benefits
             Managed health care
             Health maintenance organizations
             Employee medical benefits
             Insurance regulation
             Health care programs
             Cost sharing (finance)
IDENTIFIER:  Medicaid Program
             New Jersey
             New York
             Vermont
             Census Bureau Current Population Survey
             
******************************************************************
** 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.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


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


Report to the Chairman, Committee on Labor and Human Resources, U.S. 
Senate

July 1997

PRIVATE HEALTH INSURANCE -
CONTINUED EROSION OF COVERAGE
LINKED TO COST PRESSURES

GAO/HEHS-97-122

Private Health Insurance Trends

(101513)


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

  AAHP - American Association of Health Plans
  BLS - Bureau of Labor Statistics
  CPS - Current Population Survey
  ERISA - Employee Retirement Income Security Act of 1974
  HCFA - Health Care Financing Administration
  HIAA - Health Insurance Association of America
  HIPAA - Health Insurance Portability and Accountability Act
  HMO - health maintenance organization
  IPA - independent practice association
  POS - point of service
  PPO - preferred provider organization

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


B-276643

July 24, 1997

The Honorable James M.  Jeffords
Chairman, Committee on Labor
 and Human Resources
United States Senate

Dear Mr.  Chairman: 

In response to your request, this report provides information on
major trends in the private health insurance market during the 1980s
and 1990s.  Specifically, the report discusses the decline in private
health insurance coverage and factors contributing to this decline,
trends in health insurance premiums and reasons for these trends, and
employers' efforts to control health benefits costs. 

We are sending copies of the report to interested congressional
committees and are making copies available to others on request. 

This report was prepared under the direction of Michael Gutowski,
Assistant Director, who may be reached at (202) 512-7128 if you or
your staff have any questions.  Other major contributors to this
report are listed in appendix III. 

Sincerely yours,

Jonathan Ratner
Associate Director, Health Financing
 and Systems Issues


EXECUTIVE SUMMARY
============================================================ Chapter 0


   PURPOSE
---------------------------------------------------------- Chapter 0:1

Spending on private health insurance represents one-third of all U.S. 
health expenditures--or nearly 5 cents of every dollar spent in the
United States.  This significant share of the U.S.  economy provides
health coverage for 7 of every 10 Americans.  During the past decade,
private health insurance has undergone fundamental changes in who is
covered, how much coverage costs, and the type of coverage Americans
receive.  To better understand the major trends in private health
insurance during the 1980s and 1990s, the Chairman of the Senate
Committee on Labor and Human Resources asked GAO to report on

  -- trends in the number of people covered by private health
     insurance and factors affecting health coverage rates and

  -- changes in health insurance premiums, including reasons for
     these changes and efforts to control the costs of providing
     health benefits. 


   BACKGROUND
---------------------------------------------------------- Chapter 0:2

Most Americans rely on private health insurance to help pay for
medical expenses.  More than 90 percent of people with private health
insurance coverage have access to group insurance through employment. 
However, some people, particularly those unable to get
employment-based health benefits, purchase health insurance directly
in the individual insurance market.  While the premium cost of
employment-based coverage is generally shared by the employer and the
employee, participants in the individual market must absorb the
entire premium costs out of pocket. 


   RESULTS IN BRIEF
---------------------------------------------------------- Chapter 0:3

Private health insurance coverage has slowly but steadily declined. 
Between 1980 and 1995, the population under age 65 covered by private
health insurance decreased from 79.5 percent to 70.5 percent.  (In
1995, about 164 million persons under age 65 were covered by private
health insurance.) This trend has continued even though the U.S. 
economy has been strong and employment has grown.  Coverage for
children, early retirees, and near-poor families has declined faster
than that for the overall population.  For example, just from 1989 to
1995 the level of private health coverage for children declined by 7
percentage points compared to a decline of 4.5 percentage points in
the overall population under 65.  Declining private health coverage
has been accompanied by growth in the uninsured population and
Medicaid enrollment, which in turn increased government health
expenditures.  The proportions of the population under age 65 that
are uninsured or are Medicaid recipients increased by 5.5 and 4.3
percentage points, respectively, between 1980 and 1995.  A major
reason for declining private health coverage is the rising cost of
health insurance.  This has absorbed a growing share of business and
family incomes and has influenced employers and employees' health
insurance decisions.  Other factors that contribute to declining
coverage include shifts in employment patterns, low growth in real
family income, and the indirect effects of expanded Medicaid
coverage. 

Health insurance premiums, in contrast, have alternated between rapid
growth and relative stability over the past two decades.  During the
late 1980s, employers faced sharply rising health insurance premiums. 
In contrast, premium growth rates have decelerated in the 1990s and
remained relatively stable in recent years.  It is unclear whether
this stability in premiums is likely to continue.  Several reasons
cited as contributing to the recent near-zero growth in health plan
premiums include the cyclical nature of health insurance premiums,
the expansion of managed care, the increasingly competitive market
for health insurance, and low overall inflation.  Nonetheless, the
high level of health insurance costs and uncertainty about future
increases remain a concern of employers and individuals purchasing
coverage.  For example, in 1996, average annual premiums for family
coverage ranged from $5,071 for a health maintenance organization
(HMO) plan to $5,388 for an indemnity plan.  Average premiums for
family coverage have more than doubled since 1988. 

As employers have adopted various strategies to control the costs of
health benefits, the costs consumers bear have increased and the
types of health insurance products they receive has evolved.  Since
1980, employees are more likely to be required to pay a share of
their health plans' premiums--typically 20 to 30 percent.  In
addition, most Americans are now enrolled in a network-based plan.  A
network-based plan requires or encourages enrollees to use physicians
and hospitals affiliated with the plan.  However, network-based plans
often have lower deductibles, and most of these plans allow enrollees
to use nonaffiliated providers at a higher cost.  Finally, nearly 40
percent of persons with private employer-based health insurance
participate in a self-funded plan.  In such plans, unlike
conventional health insurance, the employer assumes the risk for
health claims and the plan is exempt from state insurance regulation. 

If at some point health insurance costs start rising again, employers
and insurers will face heightened pressure to control costs.  This
may lead to increased cost shifting from employers to employees,
reduced benefits in employer-based health plans, and faster declines
in the number of employers offering health benefits. 


   PRINCIPAL FINDINGS
---------------------------------------------------------- Chapter 0:4


      PRIVATE HEALTH COVERAGE
      DECREASING ACROSS ALL
      SEGMENTS OF THE POPULATION
      BUT AT HIGHER RATES AMONG
      CHILDREN, EARLY RETIREES,
      AND THE NEAR POOR
-------------------------------------------------------- Chapter 0:4.1

Private health coverage gradually declined for many Americans between
1980 and 1995.  During this period, the population under 65 with
private health insurance fell by 9 percentage points (from 79.5
percent to 70.5 percent).  This decline was accompanied by an
increase in the numbers of nonelderly Americans who are uninsured or
covered by Medicaid.  The uninsured as a percentage of Americans
under age 65 increased from 11.8 percent in 1980 to 17.3 percent in
1995--over 40 million Americans in 1995.  Similarly, Medicaid
enrollment among the population under 65 grew from about 8.2 percent
to 12.5 in the same period. 

The erosion of private health insurance has affected people in nearly
all demographic and employment categories, but children, early
retirees, and near-poor families experienced the greatest decline in
coverage.  Health coverage for children decreased by 7 percentage
points, from about 73 percent to 66 percent between 1989 and 1995,
compared to a 4.5 percentage point decline for the overall nonelderly
population during this period.  If children's private health
insurance coverage had remained at the 1989 level, nearly 5 million
more children would have been covered by private health insurance in
1995.  Similarly, private health coverage rates for retirees under
age 65 dropped from nearly 76 percent in 1989 to 69 percent in 1995. 
While private coverage also decreased for families between 100 and
200 percent of the federal poverty level, some children and pregnant
women in this income range gained Medicaid coverage. 


      RISING HEALTH CARE COSTS A
      KEY FACTOR IN THE DECLINE IN
      HEALTH COVERAGE
-------------------------------------------------------- Chapter 0:4.2

A number of factors have contributed to the decline in private health
coverage for specific time periods.  Among the reasons researchers
cite are rising health care costs, shifts in employment patterns, low
rates of growth in real family incomes, and the indirect effects of
expanded Medicaid coverage.  Health care cost increases have affected
employers' decisions to limit or drop coverage for employees and to
increase employees' share of health insurance costs, as well as
employees' decisions to purchase health coverage.  No single study
has measured the effect these factors had throughout the 1980s and
1990s.  The extent to which Medicaid expansions have caused private
health coverage to decline as opposed to covering people who
otherwise would have been uninsured is still in dispute. 


      RAPID GROWTH IN HEALTH
      INSURANCE PREMIUMS DURING
      THE LATE 1980S FOLLOWED BY
      RECENT NEAR-ZERO GROWTH
-------------------------------------------------------- Chapter 0:4.3

Private health insurance premiums have exhibited a cyclical growth
pattern over the past two decades.  Premium growth rates increased
sharply during the late 1980s, but growth rates decelerated
throughout the 1990s.  For example, real premium growth rates of
indemnity, preferred provider organization (PPO), and HMO plans
peaked at 15.2, 13.2, and 11.2 percent, respectively, in 1989.  Real
premium growth rates slowed to an average of about 5 percent across
all health plan types in 1993.  Over the past 2 years, health
insurance premiums have grown at lower rates than the overall price
and medical price indexes. 

Health experts attribute the declining growth in health insurance
premiums to several factors, including cyclical trends in the health
insurance industry.  Since 1970, premiums have declined to near zero
growth in real terms in at least three periods.  Periods of very low
growth in private health insurance premiums--similar to the current
situation--occurred around 1980 and 1986.  Other factors often cited
as contributing to the slowdown in premium growth rates include the
low overall inflation rate of the past several years, the effects of
managed care, and the increasingly competitive market for health
insurance from purchasers' seeking lower prices and insurers' and
HMOs attempting to gain market share. 

Some health analysts maintain that recent declines show that health
cost growth has been tamed, but other analysts are beginning to
discuss early signs of potential premium increases.  For example,
Foster Higgins reported that many HMOs have raised premiums because
of sagging profits and that most of the savings from managed care
have already been achieved.  Thus, whether premium growth will rise
or remain low is unclear. 


      EMPLOYERS' RESPONSES TO
      RISING HEALTH BENEFIT COSTS
      ARE CHANGING THE NATURE OF
      PRIVATE HEALTH INSURANCE
-------------------------------------------------------- Chapter 0:4.4

Employers have adopted a variety of methods to address their high
health benefits costs, such as requiring employees to assume a
greater share of health plan costs, encouraging use of managed care
plans, and self-funding employee coverage.  Some of these methods are
changing the cost and nature of private health insurance coverage for
employees.  For example, the average share of premiums paid by
employees for single coverage more than doubled from 10 percent in
1988 to 22 percent in 1996.  Three-quarters of people with
employment-based health coverage are enrolled in a network-based
health plan.  Most of this growth has occurred in newer, hybrid types
of managed care plans, such as PPOs and point-of-service (POS) plans,
which are less restrictive than traditional HMOs. 

In addition, many employers have self-funded their health plans,
assuming much of the risk for health claims directly rather than
purchasing insurance from a third party.  In this way, the employer
gains greater control over the health plan, avoids costs associated
with state taxes and regulation, and can provide uniform benefits
across states.  However, the plan is not subject to state oversight. 


   CONCLUDING OBSERVATIONS
---------------------------------------------------------- Chapter 0:5

Rising health care costs and the erosion of private employment-based
health insurance coverage have contributed to the rapid
transformation of America's health insurance system.  People who
receive coverage through private sector employment are paying a
growing share of their health plan's premiums, but their
out-of-pocket expenses have only increased moderately.  In addition,
many people with employer-based coverage are more likely to have some
limitation in their choice of health plan or health provider. 
Further, more people depend on Medicaid coverage or are uninsured. 

The erosion of private health insurance coverage has continued even
during years of national economic growth.  Although health policy
experts are not anticipating the rapid increases in health costs of
the late 1980s, the extent of future changes in premium costs remain
uncertain.  Many of the underlying pressures for rising health care
costs, including an aging society and new medical technologies,
remain.  If private health coverage continues to erode, federal and
state policy decisionmakers will be called upon to tackle issues of
access, affordability, and quality of health insurance, particularly
for children, early retirees, and near-poor families. 


   RECOMMENDATIONS
---------------------------------------------------------- Chapter 0:6

GAO is making no recommendations in this report. 


   COMMENTS FROM OUTSIDE REVIEWERS
---------------------------------------------------------- Chapter 0:7

GAO obtained comments on a draft of this report from experts on
private health insurance.  The reviewers agreed with the report and
also provided suggestions that GAO included where appropriate. 


INTRODUCTION
============================================================ Chapter 1

The U.S.  health insurance market is often characterized as a
private, employer-based system.  Most Americans under age 65
participate in the health insurance market through employer-based
health plans, and contributions to private health insurance represent
the bulk of employer spending on health care.  Federal and state
policy initiatives such as the Employee Retirement Income Security
Act of 1974 (ERISA) and state legislation regulating the small
employer health insurance market have had major implications in the
development of the private health insurance market. 


   PRIVATE HEALTH INSURANCE:  A
   SIGNIFICANT SHARE OF U.S. 
   HEALTH SPENDING, PARTICULARLY
   FOR PRIVATE BUSINESSES
---------------------------------------------------------- Chapter 1:1

Of the $950 billion spent on health care in the United States in
1994, private health insurance accounts for about one-third ($313
billion).  Businesses paid for the majority of these private health
insurance costs, but individuals and federal, state, and local
governments paid for over 40 percent of total spending on private
health insurance.\1 (See fig.  1.1 for distributions of total health
spending by businesses, households, and government and the share
spent by each for private health insurance.) Moreover, about
one-third of what businesses paid for private health insurance--an
estimated $65 billion in 1996--is returned in the form of tax
subsidies resulting from the tax deductibility of employer health
insurance expenses. 

   Figure 1.1:  Businesses,
   Households, and Government All
   Pay for Private Health
   Insurance

   (See figure in printed
   edition.)

Note:  "Other" includes nonpatient revenues and research and
construction. 

Source:  HCFA, Office of the Actuary. 

Public health insurance, including Medicare and Medicaid, also
represents about one-third of total U.S.  health spending.  The
remaining 36 percent of health spending is not financed through
health insurance but represents out-of-pocket spending by consumers
for copayments, deductibles, and medical services not covered by
insurance.\2

Figure 1.2 shows the changes in health expenditures from 1980 to 1994
for different categories of spending.  Increases in health care
expenditures during this period stemmed from numerous factors,
including changes in medical technology, general price inflation,
rising health care prices, the aging of the population, and an
increase in the overall population. 

   Figure 1.2:  Private Health
   Insurance and National Health
   Expenditures, 1980-94

   (See figure in printed
   edition.)

Note:  "Other" includes out-of-pocket spending; other federal, state,
and local programs; nonpatient revenues; and research and
construction. 

Source:  HCFA, Office of the Actuary. 


--------------------
\1 Individuals pay for the employee share of employment-based private
health insurance and the entire cost of individually purchased
coverage.  Federal, state, and local governments contribute to the
premium costs of their employees' health benefits.  For more details
on health spending by sector, see Cathy A.  Cowan and others,
"Business, Households, and Government:  Health Spending, 1994,"
Health Care Financing Review, Vol.  17, No.  4 (summer 1996), pp. 
157-78. 

\2 The remainder also includes research and construction costs and
nonpatient revenues (such as charitable donations). 


   PUBLIC POLICY HAS SHAPED
   PRIVATE HEALTH INSURANCE
---------------------------------------------------------- Chapter 1:2

Both federal and state government policies have shaped the
development of private health insurance.  Two major federal laws have
influenced the employer provision of health benefits.  Since 1954,
the tax code has encouraged employment-based health coverage by
making employer health benefit payments tax-deductible and by
excluding employee-provided benefits from employees' taxable income. 
ERISA allows employers to offer uniform national health benefits by
preempting states from directly regulating employer benefit plans. 
As a result, states are unable to directly regulate self-funded
health plans but can regulate health insurers.\3 Between 1990 and
1994, 45 states enacted legislation to increase access and
affordability of health insurance for small employers; by 1995, 25
states had also enacted reforms in the individual market.\4 States
also have required insurers to provide coverage for specific
benefits, such as mental health care, mammography screening, and
services provided by chiropractors and optometrists.\5

Recent federal legislation, the Health Insurance Portability and
Accountability Act of 1996 (HIPAA), has expanded the federal role in
regulating private health insurance and may set precedents for
further amendments to ERISA and the tax deductibility of health
insurance. 


--------------------
\3 For information on the implications of ERISA for private health
insurance, see Employer-Based Health Plans:  Issues, Trends, and
Challenges Posed by ERISA (GAO/HEHS-95-167, July 25, 1995). 

\4 For more information on state reforms in the small employer and
individual health insurance markets, see Health Insurance Regulation: 
Variation in Recent State Small Employer Health Insurance Reforms
(GAO/HEHS-95-161FS, June 12, 1995) and Private Health Insurance: 
Millions Relying on Individual Market Face Cost and Coverage
Trade-Offs (GAO/HEHS-97-8, Nov.  25, 1996). 

\5 For information on mandated benefits and other state requirements
on health insurers, see Health Insurance Regulation:  Varying State
Requirements Affect Cost of Insurance (GAO/HEHS-96-161, Aug.  19,
1996). 


   SCOPE AND METHODOLOGY
---------------------------------------------------------- Chapter 1:3

This report focuses on key trends in the number of people covered by
private health insurance and the cost of health coverage since 1980. 
We also examined how efforts to control costs have resulted in
changes to the health coverage Americans receive.  To examine overall
changes in the number of people covered, we analyzed data reported by
the Bureau of the Census's Current Population Survey (CPS) from 1980
to 1995.  Because of changes in the CPS during this time, we adjusted
the data for some years to make them more comparable.  While these
adjustments enhance the comparability of data on health coverage for
the population under age 65, we did not attempt to adjust data to
make comparisons among regional, employment, and demographic
population groups.  Instead, to examine trends in coverage among
these segments of the population, we analyzed data from the CPS for
1989 and 1995--a period for which the data are comparable without
further adjustments. 

To develop trend data on costs of coverage and changes in
employer-sponsored health benefits, we used data from a variety of
sources.  We reviewed several employer surveys, including those
conducted by the Bureau of Labor Statistics (BLS), Foster Higgins,
and KPMG Peat Marwick to gain information on employment-based health
plans, including premiums and types of plans offered.  Data on
national health spending are from the Health Care Financing
Administration (HCFA). 

In addition, we reviewed literature on reasons for declining health
coverage, changes in health costs, and private sector initiatives to
control costs.  We also built on information from earlier GAO reports
on private health insurance.  A list of related GAO reports is
included at the end of this report.  Appendix I provides a more
detailed discussion of the data sources we used and our methodology
for making estimates.  We conducted our review from December 1996 to
May 1997 in accordance with generally accepted government auditing
standards. 


THE SHARE OF THE POPULATION UNDER
65 WITH PRIVATE HEALTH INSURANCE
COVERAGE HAS SLOWLY BUT STEADILY
DECLINED
============================================================ Chapter 2

During the 1980s and 1990s, private health insurance has covered a
declining share of the population under age 65.  The gradual decline
in private health coverage has been nearly universal, although
children, early retirees, and near-poor families have experienced
somewhat greater declines in coverage.  Changes in employment
patterns, low real growth in family incomes, and expanded Medicaid
coverage may have contributed to the decline in private coverage, but
the rapid increase in health insurance premiums during much of this
time period has made coverage less affordable for many families and
employers. 


   DECLINING PRIVATE HEALTH
   INSURANCE HAS BEEN ACCOMPANIED
   BY GROWTH IN THE UNINSURED
   POPULATION AND IN MEDICAID
   ENROLLMENT
---------------------------------------------------------- Chapter 2:1

The slow but steady erosion of private coverage has resulted in
significant long-term increases in the uninsured population and
increases in publicly provided insurance.  We estimate that between
1980 and 1995, the share of the nonelderly population covered by
private health insurance fell from 79.5 percent to 70.5 percent.\6

(See fig.  2.1.) While most Americans under 65 continue to have
private health insurance--164 million people in 1995--nearly 21
million more would have had private health insurance if coverage had
remained at the 1980 level. 

   Figure 2.1:  Estimates of
   Private Health Coverage and
   Uninsured Rates, 1980-95

   (See figure in printed
   edition.)

Note:  Data were adjusted to account for changes in survey
methodology.  See app.  I. 

Source:  GAO estimate based on Bureau of the Census data. 

An increase in the share of the population under 65 that is uninsured
or covered by Medicaid has accompanied this decline in private
coverage.  In 1995, 17.3 percent of the population under 65--more
than 40 million Americans--lacked any health insurance coverage,
compared to 11.8 percent in 1980.  Similarly, the share of the
population under 65 covered by Medicaid grew from an estimated 8.2
percent in 1980 to 12.5 percent in 1995.\7


--------------------
\6 Our estimates of health insurance coverage rates are based on data
from the Bureau of the Census's CPS for the period 1979 to 1996. 
However, changes in the survey methodology make comparisons of
coverage rates over time difficult.  Appendix I discusses how we
adjusted the CPS data prior to 1992 to make them more comparable with
the recent CPS data on health insurance coverage. 

\7 The remainder of the nonelderly population was covered by military
health care programs (3.5 percent) or Medicare (1.8 percent).  The
total of these categories exceeds 100 percent because some people
maintain more than one type of health coverage during a year. 


   PRIVATE HEALTH INSURANCE
   COVERAGE VARIES AMONG STATES
   AND DEMOGRAPHIC, INCOME, AND
   EMPLOYMENT GROUPS
---------------------------------------------------------- Chapter 2:2

The prevalence of private health insurance coverage varies among
segments of the population.  For example, people living in southern
states, low-income families, and young adults are less likely to have
private health coverage than the national average.  In addition,
persons employed by small firms, part time, and in industries such as
construction or agriculture are less likely to have private health
coverage.  (App.  II presents private health insurance rates in 1989
and 1995 by state and for several demographic, employment, and income
categories.)

Private health insurance coverage is most common among people living
in the midwestern and northeastern United States and least common
among those in the southwestern and south central United States. 
(See fig.  2.2.) For example, only half of the population under 65 in
New Mexico had private health insurance in 1995 compared to 82
percent of the population under 65 in Connecticut and Minnesota.  As
shown in figure 2.3, states with low rates of private health
insurance coverage tend to have high uninsured rates. 

   Figure 2.2:  Private Health
   Insurance Coverage by State,
   1995

   (See figure in printed
   edition.)

Source:  Bureau of the Census, March 1996 Current Population Survey. 

   Figure 2.3:  Uninsured Rates by
   State, 1995

   (See figure in printed
   edition.)

Source:  Bureau of the Census, March 1996 Current Population Survey. 

Private health insurance coverage is particularly low among young
adults aged 18 to 24 (60 percent), blacks (51 percent), and Americans
of Hispanic origin (43 percent).  These low levels are reflected in
high uninsured rates among these groups (28 percent, 22 percent, and
35 percent, respectively.) In addition, low-income families are much
less likely to have private health insurance than high-income
families.  Whereas less than one-third of families with incomes of
less than $20,000 had private health insurance in 1995, over 90
percent of families with incomes of at least $60,000 had private
insurance coverage. 

Several employment characteristics are important in private health
insurance coverage levels.  In 1995, most Americans (64.6 percent)
received private health insurance through employment--either as
workers or as dependents of workers.  Workers in small firms are less
likely to have employment-based health insurance than are workers in
large firms.  For example, 50 percent of workers in firms with fewer
than 10 employees had private employment-based health insurance in
1995 compared to 82 percent of workers in firms with at least 1,000
employees.\8 Recent research based on KPMG Peat Marwick and HIAA data
for 1989 to 1996 indicates that while more small firms are offering
coverage now compared to the past, a smaller percentage of employees
are covered now compared to 1989.\9

Contingent workers, including part-time and temporary workers, are
also less likely to have employer-sponsored coverage.  While 79
percent of full-time, full-year employees had employment-based health
insurance in 1995, only 59 percent of employees working part-time or
part-year had employment-based health insurance.  Employment-based
health insurance covers less than 60 percent of workers in the
agricultural, personal services, business and repair services, and
construction industries but at least 82 percent of workers in the
finance and insurance, mining, and public administration industries. 

Some people who are unable to get health benefits through their
employment purchase private health insurance directly.  In a previous
report, we estimated that more than 10 million Americans--4.5 percent
of the nonelderly U.S.  population--had individual health insurance
as their only source of health coverage in 1994.\10 Individual health
insurance is most common in some Mountain and Plains states and among
the self-employed and agricultural workers.  Other people who do not
have employment-based coverage and do not qualify for Medicaid may be
unable to purchase individual health insurance because of its cost or
a preexisting medical condition.  Some of these people may be able to
get private coverage through a state high-risk pool or other
programs, but others remain uninsured.\11


--------------------
\8 For all firm sizes, 72 percent of workers had employment-based
health insurance. 

\9 Analysis completed by Jon Gabel, KPMG Peat Marwick, and Paul
Ginsburg, Center for Studying Health System Change. 

\10 For information on the characteristics of individual insurance
enrollees, the structure of the individual insurance market, and
insurance reforms undertaken by states, see Private Health Insurance: 
Millions Relying on Individual Market Face Cost and Coverage
Trade-offs (GAO/HEHS-97-8, Nov.  25, 1996). 

\11 About 25 states have high-risk pools, covering about 100,000
persons nationwide. 


      PRIVATE HEALTH COVERAGE HAS
      DECLINED AMONG ALMOST ALL
      SEGMENTS OF THE POPULATION
-------------------------------------------------------- Chapter 2:2.1

The decline in private health insurance coverage has been nearly
universal in nearly all demographic, income, and employment groups. 
Between 1989 and 1995, private coverage declined in 43 of 50 states
(representing 93 percent of the U.S.  nonelderly population) and 10
of 12 industrial categories (representing 93 percent of all U.S. 
workers).\12 Similarly, between 1989 and 1995 private health coverage
also declined among firms of all size categories, among both
full-time and part-time workers, and among all income categories. 

While declining health insurance has been widespread, coverage has
declined more quickly for some population groups.  For children under
18 years old, private health insurance declined by 7 percentage
points between 1989 and 1995 (from 73.1 percent to 66.1 percent)
compared to a 4.5 percentage point decline for the nonelderly
population overall.  If children's level of private health insurance
coverage had remained at the 1989 level, nearly 5 million more
children would have been covered by private health insurance in
1995.\13

The level of private health insurance coverage has also declined more
rapidly among people younger than 65 who are retired (69.0 percent in
1995 compared to 75.9 percent in 1989).  This trend reflects the
decline in employers offering retiree health coverage.  Foster
Higgins reported that only 40 percent of large employers offered
medical coverage to retirees younger than 65 in 1996 compared to 46
percent in 1993.\14

Private health insurance coverage has also declined more quickly
among near-poor families.  While Medicaid is the primary source of
insurance coverage for persons at or below the federal poverty level,
private health insurance covered 55 percent of families between 100
and 200 percent of the federal poverty level in 1995.  This
represents a 6 percentage point decline since 1989.  Many of these
near-poor families losing private health coverage gained Medicaid
coverage.  For some children and pregnant women in families with
incomes between 100 and 200 percent of the federal poverty level,
Medicaid coverage grew from 6.5 percent in 1989 to 11.2 percent in
1995. 


--------------------
\12 We are using 1989 as a base for comparison because of
difficulties in comparing CPS data for earlier years, particularly
when examining subgroups of the population.  For 1989, the Bureau of
the Census released revised estimates (weighted to the 1990 decennial
census) that are comparable to the 1995 CPS data on private health
insurance coverage.  See app.  I. 

\13 For information on children's health insurance coverage, see
Health Insurance for Children:  Private Insurance Coverage Continues
to Deteriorate (GAO/HEHS-96-129, June 17, 1996) and an update in
Children's Health Insurance, 1995 (GAO/HEHS-97-68R, Feb.  19, 1997). 

\14 Foster Higgins, National Survey of Employer-Sponsored Health
Plans, 1996. 


   COST PRESSURES ARE ONE OF MANY
   FACTORS LINKED TO DECLINING
   EMPLOYER-BASED COVERAGE
---------------------------------------------------------- Chapter 2:3

The decline in private health insurance coverage during the 1980s and
the 1990s reflects decisions made by both employers who offer health
insurance and their employees.  The long-term growth of health
insurance premiums has made health insurance a more prominent
component of both employer and household budgets.  These cost
increases contributed to employer decisions to limit or drop coverage
for workers and to increase the share of insurance costs paid by the
worker. 

A decline in the number of employers offering health insurance is
prominently cited as a key contributor to declining coverage.\15
Particularly for small employers, costs are cited as a key factor in
their decision to drop coverage for their workers or to consider
offering it.\16 For those employing lower-wage workers, health
premiums represent a significant share of total compensation.  Even
firms that provide coverage for their workers often exclude part-time
or temporary workers--a rapidly growing component of the labor force. 

Not surprisingly, key indicators of the structure of the labor force
have been prominently cited in research that attempts to explain
declines in health coverage.  For example, coverage is higher in the
Midwest and in industries with higher wage rates.  But recent
employment growth has been concentrated in industries and areas of
the country where insurance coverage has not been as prominent. 
Several studies suggest that this change in employment composition
was an important factor in recent declines in coverage. 

Table 2.1 summarizes some of the key factors that these studies have
identified as contributing to the decline in private health coverage. 
While these studies embrace many of the same variables, their effect
is not always consistent because of differences in modeling, data
sources, the specific population covered, and time periods.  There is
clearly a complex interaction among a number of variables affecting
costs over time.  However, these studies acknowledge that much of the
decline in coverage remains unexplained by these factors. 



                               Table 2.1
                
                 Summary of Studies Identifying Factors
                   Contributing to Declining Coverage

                        Factor contributing to
                        declining coverage      Factor not
Author: period,         (percentage accounted   contributing to
population analyzed     for by factor)          declining coverage
----------------------  ----------------------  ----------------------
Gregory Acs: 1988-91,   Family income (67%),    Industry shifts,
nonelderly population   unemployment (15%)      location, firm size,
                                                demographics

Gregory Acs: 1988-91,   Family income (38%),    Location, firm size
workers                 industry shifts (17%)

Kronick: 1979-89, low-  Industry shifts (not
income workers          available)

Fronstin and Snider:    Wage rate (23%),        Demographics, firm
1988-93, workers        industry shifts (10%),  size, occupation,
                        part-time work (7%),    region
                        unionization (6%)

Long and Rodgers:       Industry shifts (<15%)  Part-time work
1980-87; workers
----------------------------------------------------------------------
Sources:  Gregory Acs, "Explaining Trends in Health Insurance
Coverage Between 1988 and 1991," Inquiry, Vol.  32 (spring 1995), pp. 
102-10; Richard Kronick, "Health Insurance, 1979-1989:  The Frayed
Connection Between Employment and Insurance," Inquiry, Vol.  28
(winter 1991), pp.  318-32; Paul Fronstin and Sarah C.  Snyder, "An
Examination of the Decline in Employment-Based Health Insurance
Between 1988 and 1993," Inquiry, Vol.  33 (winter 1996-97), pp. 
317-25; Stephen H.  Long and Jack Rodgers, "Do Shifts Toward Service
Industries, Part-time Work, and Self-Employment Explain the Rising
Uninsured Rate?" Inquiry, Vol.  32 (spring 1995), pp.  111-16. 

Moreover, the relative importance of each variable may change over
time.  For example, declines in real wages and family income
contributed to the declines in coverage in the 1980s, but whether
recent increases in real wages may have the opposite effect in the
next few years is unclear.  The analysis of most of the studies
covers a fairly short time period, which may not reflect many of the
potential interactions among these variables.  A key variable like
industry mix may have played a more prominent role in the 1980s as
service sector employment increased relative to manufacturing,
government, and transportation--industries in which health coverage
has traditionally been more prevalent.  But the service industry now
represents a more prominent share of total employment, and some of
its recent growth has been channeled into particular segments like
health care or computer services, where wage rates are higher and
coverage is more common. 

Factors affecting an employer's decision to offer coverage are
important determinants of the level of coverage.  But the growing
cost of health insurance has also affected decisions individuals
make.  Employees of firms that offer coverage are being asked to pay
a higher share of premiums.  Data from the National Health Interview
Survey indicate that over 62 percent of employed, uninsured family
heads report that the main reason for not being covered is that
health insurance is too expensive.\17 For those without
employer-based coverage, the rise in premiums for policies purchased
in the individual insurance market has been borne exclusively by
those individuals. 

Some analysts have also indicated that the expansion in Medicaid
enrollment during the late 1980s and 1990s has "crowded out" private
health insurance coverage--that is, some low-income families who
previously received private coverage are now replacing it with
publicly funded Medicaid.  One study concluded that as much as 15
percent of the decline in private health insurance between 1987 and
1992 could be attributed to the substitution of Medicaid
enrollment.\18 However, the extent to which the Medicaid expansions
have caused a decline in private health insurance coverage rather
than absorbed coverage that otherwise would have been lost remains in
dispute. 

The erosion of private health insurance coverage for many Americans
has continued through the mid-1990s, a time when health insurance
premiums have been relatively stable and the U.S.  economy has been
strong, with low unemployment and steady growth.  However, many of
the studies attempting to quantify the reasons for declining private
health coverage use data only through the early 1990s.  Researchers
have yet to examine the causes of the decline in coverage over these
past few years, taking into account the more recent economic trends. 


--------------------
\15 Paul Fronstin and Sarah C.  Snyder, "An Examination of the
Decline in Employment-Based Health Insurance Between 1988 and 1993,"
Inquiry, Vol.  33 (winter 1996-97), pp.  317-25. 

\16 See Access to Health Insurance:  State Efforts to Assist Small
Business (GAO/HRD-92-90 May 14, 1992), pp.  12-15. 

\17 See statement of Patrick J.  Purcell, Congressional Research
Service, before the U.S.  House of Representatives, Committee on Ways
and Means, Subcommitee on Health, April 8, 1997. 

\18 See David M.  Cutler and Jonathan Gruber, "Medicaid and Private
Insurance:  Evidence and Implications," Health Affairs, Vol.  16, No. 
1 (Jan.-Feb.  1997), pp.  194-200; Lisa Dubay and Genevieve Kenney,
"Did Medicaid Expansions for Pregnant Women Crowd Out Private
Coverage?" Health Affairs, Vol.  16, No.  1 (Jan.-Feb.  1997), pp. 
185-93. 


SHARP INCREASES IN PRIVATE HEALTH
INSURANCE PREMIUMS HAVE SUBSIDED,
BUT THE FUTURE REMAINS UNCERTAIN
============================================================ Chapter 3

Health insurance premiums for employer-sponsored plans increased
sharply during the late 1980s, but their growth rates have
decelerated substantially in the 1990s, with premium changes reaching
record lows in 1996.  Studies have attributed the downturn in premium
growth to various factors, including the effects of managed care and
cyclical patterns in the health insurance industry.  While some
health insurance experts are identifying signs of future premium
increases, it is too early to tell how much longer premium increases
will remain low and how much higher future growth rates will be. 
Furthermore, although trend data are not available for purchasers of
individual health insurance, premiums in this market are very
sensitive to the age, gender, and health of enrollees as well as to
state regulation. 


   TRENDS IN HEALTH INSURANCE
   PREMIUM GROWTH RATES DURING THE
   1980S AND 1990S
---------------------------------------------------------- Chapter 3:1

Increases in employer-based health insurance premiums far exceeded
the general price inflation rate in the late 1980s but premium growth
rates have declined in the 1990s.  For example, the average annual
premium for employer-based family health insurance coverage increased
by 111 percent, from $2,530 to $5,349 between 1988 and 1996, while
general prices rose by 33 percent during this period.  Similarly, the
average annual premium for employer-based single coverage increased
by 79 percent, from $1,153 to $2,059 between 1988 and 1996.\19

However, the real annual rate of increase in health insurance
premiums has slowed across all employer-based health plan types in
the past 7 years.\20 As figure 3.1 shows, real premium growth rates
of indemnity, preferred provider organization (PPO), and health
maintenance organization (HMO) plans peaked in 1989 at rates of 15,
13, and 11 percent, respectively.  Real premium growth rates slowed
to an average of about 5 percent across all plan types in 1993; over
the past 2 years, premiums experienced near-zero growth.  In 1996,
premiums increased at lower rates than the consumer price index and
the medical cost index.  As figure 3.1 also shows, the differences in
premium growth rates among various types of health plans narrow after
1990.  See also table 3.1. 

   Figure 3.1:  Real Growth in
   Premiums by Health Plan Type,
   1987-96

   (See figure in printed
   edition.)

Sources:  GAO calculations using data from KPMG Peat Marwick
(1991-96); HIAA (1987-90), and the BLS consumer price index. 
Includes employer and employee shares of premiums for workers in
private firms with at least 200 employees. 



                                     Table 3.1
                      
                         Real Annual Growth in Premiums Per
                                Health Plan, 1987-96

Pl
an    1987    1988    1989    1990    1991    1992    1993    1994    1995    1996
--  ------  ------  ------  ------  ------  ------  ------  ------  ------  ------
In    -0.2     6.9    15.2    11.6     7.8     8.0     5.5     2.5    -0.1    -1.8
 d
 e
 m
 n
 i
 t
 y
PP     0.9    12.9    13.2     9.6     5.9     7.6     5.2     0.6     0.7    -2.4
 O
HM     0.1     4.9    11.2    10.6     7.8     6.8     5.3     2.7    -2.4    -3.4
 O
----------------------------------------------------------------------------------
Sources:  GAO calculations based on data from KPMG Peat Marwick
(1991-96); HIAA (1987-90), and BLS consumer price index.  Includes
employer and employee shares of premiums for workers in private firms
with at least 200 employees. 


--------------------
\19 These average premiums are based on surveys from KPMG Peat
Marwick, Survey of Employer Sponsored Health Benefits, 1996, and
Health Insurance Association of America (HIAA), Employee Survey,
1988. 

\20 "Real premium growth rate" refers to a growth rate that has been
adjusted to take into account the effect of general price inflation. 


   MANAGED CARE AND CYCLICAL
   FACTORS IN THE HEALTH INSURANCE
   INDUSTRY AMONG MAJOR CAUSES OF
   SLOWDOWN IN PREMIUM GROWTH
   RATES
---------------------------------------------------------- Chapter 3:2

Health insurance researchers attribute the slowdown in annual premium
growth to several causes, including the increased use of managed care
and a downward trend in the health insurance underwriting cycle in
which premiums tend to decline when health insurers' profits are
high.  However, no studies have comprehensively examined the reasons
for the recent 7-year decline in premium increases.  While the
precise effect of managed care on premiums continues to be debated,
some studies contend that managed care has contributed to the
slowdown in premium increases because HMO plans generally cost less
than other health plans and many managed care organizations control
health care use.  Hence, the savings occur from moving consumers from
indemnity plans to HMO plans.  In contrast, other research attributes
the savings from managed care to changes in the operation of the
health insurance market when managed care penetration rates reach a
critical threshold. 

The increased HMO enrollment has contributed to a one-time reduction
in premium levels.  KPMG Peat Marwick's survey found that in 1996 HMO
premiums for employer-sponsored health insurance premiums averaged 6
to 10 percent less than indemnity plan premiums, whereas PPO and
point-of-service (POS) plan premiums were similar to or higher than
indemnity plan premiums.  (See table 3.2.) As noted above, however,
all employer-sponsored plan types have had similar growth patterns
since 1987, with HMO premium growth rates tending to be slightly
lower than indemnity plan premiums. 



                               Table 3.2
                
                Average Annual Health Insurance Premiums
                 by Employer-Sponsored Plan Type, 1996

                                            Single          Family
                                           coverage        coverage
                                        --------------  --------------
                                                Percen          Percen
                                                     t               t
                                        Averag  differ  Averag  differ
                                             e     ent       e     ent
                                        annual    from  annual    from
                                        premiu  indemn  premiu  indemn
Plan                                         m     ity       m     ity
--------------------------------------  ------  ------  ------  ------
Indemnity                               $2,090          $5,388
PPO                                     $2,173   +4.0%  $5,377   -0.2%
POS                                     $2,058    -1.5  $5,477    +1.7
HMO                                     $1,883    -9.9  $5,071    -5.9
----------------------------------------------------------------------
Source:  KPMG Peat Marwick Survey of Employer-Sponsored Health
Benefits, 1996.  This survey includes employers with at least 200
employees. 

Some studies indicate that a fundamental cause of the downturn in
premium increases lies in the nature of the health insurance
underwriting cycle.\21 Health insurer's profits and premiums are
inversely related.  Health insurance researchers have found that
historically health insurance premiums have increased about 2 years
after health insurers' profits declined and vice versa.  Industry
profits were high from 1993 to 1995, which suggests that low premium
increases for 1996 were predictable.  As shown in figure 3.2, 1974
and 1980 were other periods of near-zero real premium growth.  Our
review showed that the 7-year downward trend in premium growth rates
since 1989 is the longest period of declining rates of growth in
private health insurance premiums over the past two decades.  (Fig. 
3.2 shows changes in enrollee spending in private health insurance
from 1970 to 1995.)

   Figure 3.2:  Change in Private
   Health Insurance Real Spending
   Per Enrollee, 1970-95

   (See figure in printed
   edition.)

Source:  HCFA Office of Actuary and BLS, Consumer Price Index. 

Some analysts also attribute part of the duration of this downward
trend to market responses to the comprehensive federal health care
reforms proposed during the early 1990s and to increased competition
among health plans trying to build market share through favorable
pricing.  In addition, the private health insurance industry has been
undergoing consolidation and mergers.\22 However, there is limited
research on the effects of these mergers, and it is too early to
determine their effect on health insurance premiums. 

It is difficult to tell whether the low premium growth rates of the
past 2 years are temporary or a signal of price stability in the
private health insurance market.  Nonetheless, some health
researchers are beginning to identify signs of possible premium
increases in the next few years.  For example, according to Foster
Higgins' 1996 National Survey of Employer-Sponsored Health Plans,
premiums may increase because (1) many HMOs have raised their rates
for 1997 following sagging profits in 1996, (2) health care providers
are consolidating and gaining more bargaining clout with managed care
plans, (3) state and federal laws are requiring health insurance
plans to expand the benefits covered, and (4) there is little room
for additional savings in costs from shifting employees to managed
care plans, since more than three-fourths of active employees are
already in managed care plans.  In addition, with low unemployment,
the prospects of a tight labor market could reverse the low, overall
inflation rate of the past years. 


--------------------
\21 See Gail A.  Jensen and others, "The New Dominance of Managed
Care:  Insurance Trends in the 1990s," Health Affairs, Vol.  16, No. 
1 (Jan.-Feb.  1997), and Jon Gabel and others, "Tracing the Cycle of
Health Insurance," Health Affairs (winter 1991). 

\22 See Roger Feldman, Douglas Wholey, and Jon Christianson,
"Economic and Organizational Determinants of HMO Mergers and
Failures," Inquiry, Vol.  23 (summer 1996), pp.  118-32.  This study
discusses the implications of HMO mergers on premiums and the effects
of mergers, depending on the competitive nature of the HMO markets. 


   PREMIUMS FOR INDIVIDUAL HEALTH
   INSURANCE VARY WITH DEMOGRAPHIC
   TRAITS, HEALTH STATUS, AND
   STATE REGULATION
---------------------------------------------------------- Chapter 3:3

In most states, premiums for health coverage purchased in the
individual health insurance market primarily reflect the demographic
characteristics and health status of each applicant, unlike
employer-based health insurance premiums, which reflect the average
risk characteristics of the entire insured group.  Premium rates in
the individual market vary substantially among states and carriers,
affecting individual consumers differently, depending on their
particular circumstances.\23 The demographic characteristics that
carriers often considered in setting premium prices in the individual
insurance market include age, gender, geographic area, and family
composition.  For example, a plan in Arizona charges a healthy
55-year-old man more than three times the rate for a healthy
25-year-old man for the same coverage ($191 versus $57 per month). 
In addition, people with a preexisting condition, such as knee
injuries or diabetes, would not be able to purchase this plan or
would have an exclusion for that condition.\24

State regulation has also affected premium prices in the individual
market.  Some states, including New Jersey, New York, and Vermont,
have enacted legislation requiring community rating.  Under pure
community rating, the cost of insurance is spread equally among all
community members, regardless of demographic characteristics and
health status.\25 Although prices range widely depending on the plan,
all applicants are eligible for and may select from among any of the
plans that carriers provide.  For example, in New Jersey, the monthly
premium price for a fee-for-service plan ranges from $155 to $565. 


--------------------
\23 For details on premium price variation stemming from demographic
characteristics and health status, see Private Health Insurance: 
Millions Relying on Individual Market Face Cost and Coverage
Trade-Offs (GAO/HEHS-97-8, Nov.  25, 1996). 

\24 Effective July 1, 1997, HIPAA allows people with continuous group
coverage to purchase individual coverage without meeting preexisting
condition exclusions.  Some states have indicated that they may
develop alternative mechanisms to provide access to coverage for
people with preexisting conditions when they lose coverage, such as
high-risk pools. 

\25 Some states that have enacted community rating, such as New York,
allow for limited adjustments by geographic regions; some also allow
adjustments for age or gender. 


EMPLOYERS' EFFORTS TO CONTROL
COSTS HAVE CHANGED THE NATURE OF
PRIVATE HEALTH INSURANCE
============================================================ Chapter 4

The increasing costs of providing health benefits to employees during
the 1980s and 1990s have stimulated employers to become increasingly
aggressive in attempting to control their costs.  Between 1980 and
1996, the costs of providing private health coverage to employees
increased from 3.4 percent of total compensation to 5.6 percent.\26
(See fig.  4.1.) Many of the changes in employment-based health
coverage over the past two decades--notably requiring employees to
pay a greater share of premium costs, encouraging enrollment in
network-based managed care plans, and self-funding health
benefits--have fundamentally changed the nature of private health
insurance coverage for employees. 

   Figure 4.1:  Employer Private
   Health Insurance Spending as a
   Percentage of Total
   Compensation, 1980-94

   (See figure in printed
   edition.)

Source:  HCFA, Office of the Actuary. 


--------------------
\26 Total business health costs, including private health insurance
premiums as well as workers compensation costs, payments to the
Medicare Hospital Insurance Trust Fund, and on-site health services
rose, from 4.7 percent of total compensation in 1980 to 7.5 percent
of total compensation in 1994. 


   EMPLOYEES ARE REQUIRED TO PAY
   AN INCREASING SHARE OF PREMIUMS
   FOR THEIR COVERAGE
---------------------------------------------------------- Chapter 4:1

As private health insurance costs increased rapidly during the late
1980s and early 1990s, employers required their employees to pay an
increasing share of the premiums for employment-based health
insurance.  On the average, employees paid about 30 percent of the
premium for family coverage and 22 percent of the premium for single
coverage in 1996.  In comparison, in 1988 employees paid 26 percent
of family plan premiums and 10 percent of single plan premiums.  (See
fig.  4.2.)

   Figure 4.2:  Employee Share of
   Premiums, 1988-96

   (See figure in printed
   edition.)

Sources:  KPMG Peat Marwick, 1991-96; HIAA, 1988-90. 

The increasing share of premiums paid by employees represents a trend
away from employers paying the entire premium for employees'
coverage.  BLS reports that health benefits paid entirely by the
employer for single coverage were available to 72 percent of
employees in 1980.  By 1993, only half as many employees (37 percent)
had single coverage available at no cost from their employer.  For
family coverage, employees with health benefits paid entirely by the
employer fell from 51 percent in 1980 to 21 percent in 1993.\27

While the share of premiums paid by the employee has grown more for
single coverage than for family coverage, family plan premiums have
grown more quickly than single plan premiums.  As noted in chapter 3,
premiums for family coverage increased 111 percent compared to 79
percent for single coverage between 1988 and 1996.  In addition,
employers have increasingly adopted tier-rating practices that charge
employees different rates, depending on the number of people covered
in their family.  As a result, employees have had to pay more for
dependent coverage, particularly if they are covering several
dependents.\28 To some extent, this higher cost for dependent
coverage has contributed to the more rapid decline in private health
insurance coverage for children. 

While employees have paid an increasing share of premiums, their
out-of-pocket costs for coinsurance and deductibles have not changed
much.  According to the BLS survey, the median deductible for single
coverage in non-HMO plans offered by medium and large employers was
$200 in 1993.\29 Deductibles for family coverage are generally two or
three times deductibles for single coverage.  Most health plan
participants in medium and large firms had a $100 deductible in 1980,
which would have been $175 in real 1993 dollars.\30

Similarly, the most common coinsurance rate has remained at 80
percent. 

Cost sharing arrangements vary by plan type.  HMOs generally do not
require a deductible or coinsurance, instead charging a copayment
such as $10 per visit.  PPOs and POS plans typically vary the
coinsurance rates, depending on whether the patient visits a provider
within or outside their network.  For example, KPMG Peat Marwick
reports that in 1996 most PPOs paid 80 or 90 percent for preferred
providers, whereas they paid 70 or 80 percent for nonpreferred
providers. 


--------------------
\27 See Bureau of Labor Statistics, Employee Benefits in Medium and
Large Private Establishments, 1993 (Washington, D.C.:  Nov.  1994),
and Employee Benefits in Medium and Large Firms, 1988 (Washington,
D.C.:  Aug.  1989). 

\28 For a detailed discussion of changes in dependent coverage, see
Employment-Based Health Insurance:  Costs Increase and Family
Coverage Declines (GAO/HEHS-97-35, Feb.  24, 1997). 

\29 In contrast, in the individual insurance market there tends to be
a wider range of deductible levels, with deductibles commonly between
$250 and $2,500.  In a previous report, we found that for one insurer
an individual health plan with a $2,000 deductible was nearly half
the cost of a similar plan with a $250 deductible ($565 versus
$1,073) for a healthy 30-year-old man. 

\30 Smaller firms tend to have higher deductibles.  The 1993 Robert
Wood Johnson Foundation Employer Health Insurance Survey found that
the average annual deductible was $408 for firms with 1 to 4
employees compared to $202 for firms with 50 or more employees. 


   NETWORK-BASED HEALTH PLANS HAVE
   BECOME PREVALENT, BUT WIDE
   VARIATION IN PLAN TYPES REMAIN
---------------------------------------------------------- Chapter 4:2

Health plans available in the private market continue to evolve, with
greater reliance on management of care and medical services provided
through a defined network of physicians and hospitals.  Most insured
Americans under 65--over 100 million people--are enrolled in a
network-based health plan that requires or encourages them to use
hospitals and physicians affiliated with the plan.  This is a
fundamental change from 1980, when nearly all privately insured
Americans were enrolled in traditional indemnity plans without
restrictions or incentives to use particular providers.  However,
even with the rapid growth of network-based health plans, most people
have enrolled in plans that retain their ability to receive medical
care from physicians not affiliated with the plan at a higher cost to
the patient. 

Network-based health plans have often been categorized as "managed
care," but the methods they use vary widely and so does the extent to
which they manage enrollees' use of health care services.  Examples
of these plans include group and staff model HMOs, independent
practice association (IPA) HMOs, PPOs, and POS plans.  These plans
vary as to whether enrollees can use health care providers other than
those affiliated with the plans' network of providers, the breadth of
the network relative to the number of enrollees, and how physicians
and hospitals are reimbursed for serving the plans' enrollees.  Table
4.1 summarizes variation among the different types of network-based
health plans according to these characteristics. 



                         Table 4.1
          
            Characteristics of Health Insurance
                           Plans

              Choice of     Size of       Provider
Plan          provider      network       payments
------------  ------------  ------------  ----------------
Traditional   Unlimited     No network    Fee-for-service
indemnity

PPO           Financial     Narrow to     In-network
              incentives    broad         physicians,
              to use        network       discounted fee-
              preferred                   for-service;
              providers                   out-of-network
                                          physicians, fee-
                                          for-service

POS           Financial     Narrow to     In-network
              incentives    broad         physicians,
              to use        network       discounted fee-
              preferred                   for-service or
              providers;                  capitation; out-
              may have                    of-network
              primary care                physicians, fee-
              gatekeeper                  for-service

IPA HMO       Restricted    Narrow to     Capitation or
              to network    broad         fee-for-
              providers;    network       service, may
              generally                   include
              includes                    withholds or
              primary care                bonuses
              gatekeeper

Group and     Restricted    Narrow        Capitation or
staff HMO     to network    network       salary
              providers;
              primary care
              gatekeeper
----------------------------------------------------------
The most tightly managed plans are traditional group and staff HMOs;
enrollment in these types of plans has fallen since 1989.  (See fig. 
4.3 and table 4.2.) These HMOs contract with physicians who generally
serve the plan's enrollees exclusively and are paid either a per
member, per month payment (capitation) or salary.  Enrollees are
restricted to the providers affiliated with the plan and pay only a
small copayment, such as $10, per visit.  Generally enrollees are
assigned a primary care physician, sometimes called a "gatekeeper,"
who coordinates their care and approves the use of specialist
services. 

   Figure 4.3:  Enrollment in
   Network-Based Health Plans by
   Type, 1980-95

   (See figure in printed
   edition.)

Sources:  HMO enrollment from Interstudy and includes Medicare and
Medicaid.  PPO and POS enrollment GAO estimate based on data from
KPMG Peat Marwick. 



                         Table 4.2
          
          Enrollment in Network-Based Health Plans
                      by Type, 1980-95

                                  IPA/
                    Group/    network/
                 Staff HMO   mixed HMO       PPO       POS
--------------  ----------  ----------  --------  --------
1980                   7.4         1.7       0.0       0.0
1981                   7.8         2.4       0.0       0.0
1982                   7.6         3.3       0.0       0.0
1983                   8.4         4.1       0.0       0.0
1984                   8.7         6.4       1.3       0.0
1985                   9.2         9.7       5.8       0.0
1986                  10.2        13.5      16.5       0.0
1987                  10.4        18.2      17.9       0.0
1988                  12.3        19.1      17.5       0.0
1989                  13.2        19.3      28.0       0.0
1990                  13.1        20.6      20.0       7.7
1991                  12.9        22.2      28.9       4.1
1992                  11.9        25.3      39.4      12.1
1993                  11.7        28.1      30.4      13.8
1994                  10.8        31.5      37.2      22.3
1995                   9.9        36.3      30.2      24.7
----------------------------------------------------------
Sources:  HMO enrollment data are from Interstudy and includes
Medicare and Medicaid.  PPO and POS enrollment numbers are GAO
estimates based on data from KPMG Peat Marwick. 

Most Americans with private health insurance are enrolled in one of a
variety of newer network-based health plans that incorporate some of
the elements of both the traditional group and staff HMOs as well as
unrestricted fee-for-service plans.  These plans, including IPA HMOs,
PPOs, and POS plans, were nearly nonexistent in 1980.  Nearly all the
growth in HMO enrollment since 1980 has occurred in IPA and similar
HMOs rather than group and staff HMOs.  As with group and staff model
HMOs, IPA HMO enrollees receive coverage only for medical services
provided by providers affiliated with the HMO.  However, the plan
contracts with physicians in different practice settings who
generally serve patients from other health plans as well as the
contracting HMO.  Furthermore, physicians contracting with IPA HMOs
are less likely to be reimbursed through capitation or salary; often,
these physicians receive a fee-for-service payment with financial
incentives to control use.\31

The most common types of network-based health plans, enrolling at
least 55 million Americans, are PPOs and POS plans.\32 These plans
encourage enrollees to seek care from physicians affiliated with the
plan but allow enrollees to receive care from physicians outside the
plan's network at a higher out-of-pocket cost to the patient.  Thus,
they provide enrollees with more options in choosing a provider but
are less able to manage the enrollees' use of health services.  PPOs
generally resemble traditional indemnity plans, except that they give
financial incentives for enrollees to use physicians who have agreed
to accept discounted fee-for-service payments and other standards
established by the plan.  Often, POS plans developed from HMOs but
allow enrollees to choose a physician outside the panel of HMO
physicians at greater expense to the enrollee.  Increasingly,
however, the distinction between plan types has become blurred. 
Thus, some PPOs may have greater use controls than some IPA HMOs.\33


--------------------
\31 Some former staff and group model HMOs have expanded by
incorporating IPA contracts with physicians.  These are generally
categorized as mixed-model HMOs. 

\32 This may underestimate PPO and POS plan enrollment.  The American
Association of Health Plans (AAHP) reports that by year-end 1995 HMOs
enrolled 58.2 million people (including 10.8 million in HMOs with POS
options) and PPOs enrolled 91 million people.  However, the AAHP
survey may double count some PPO enrollees.  See app.  I for a
discussion of our methods in estimating enrollment in the different
network-based health plans. 

\33 For information on how managed care attempts to control the use
of health services, see Managed Health Care:  Effect on Employers'
Costs Difficult to Measure (GAO/HRD-94-3, Oct.  19, 1993). 


   EMPLOYERS HAVE INCREASINGLY
   SELF-FUNDED THEIR HEALTH
   BENEFITS
---------------------------------------------------------- Chapter 4:3

Increasingly, employers have assumed greater direct control over
their health benefits by self-funding their plans, paying the cost of
health claims directly from their revenues rather than contracting
with an insurer to assume the risk.  Many employers moderate the
level of the risk that they are self-funding by purchasing stop-loss
coverage to insure against health costs that exceed a set threshold. 
In addition, many contract with a third-party administrator to handle
claims payments, benefit design, and other administrative tasks
traditionally performed by an insurance company. 

Employers decide to self-fund their health plans for several reasons,
but a significant advantage to employers results from the fact that
states cannot apply their insurance requirements to self-funded
health plans.  ERISA preempts states from directly regulating
employer pension and welfare benefit plans, including health benefit
plans.  However, ERISA allows states to continue their traditional
role in regulating the terms and conditions of insurance.\34 Thus,
states require insurers to pay premium taxes, offer specific
benefits, meet solvency standards, and use specific practices for
setting premium prices, but self-funded health plans are preempted
from these state requirements. 

This exemption from state regulation allows employers who self-fund
their health plans to save costs associated with state regulation and
to offer a uniform, interstate health plan.  In an earlier report, we
examined the cost differences to self-funded and fully insured health
plans resulting from these state requirements.\35 We found that the
extent to which these requirements increased the costs of insured
health plans compared with self-funded health plans varies by state. 
On the average, premium taxes increased costs to commercial health
insurers by about 2 percent.  Although studies of the claims costs
associated with state-mandated benefits range from 5 to 22 percent,
most self-funded plans offer many of the benefits that are mandated
by states.  As a result, the additional costs to most self-funded
plans, if they were required to comply with state-mandated benefits,
would not be as high as these studies estimate.  In addition, most
insurers voluntarily exceed states' minimum financial solvency
standards, indicating that these standards have a limited potential
effect on their costs. 

The share of insured workers who get their health coverage through
self-funded plans has grown.  As shown in figure 4.4, this share has
tripled since 1980 for workers and dependents in medium and large
private establishments (from 15 percent in 1980 to 46 percent in
1993).  Data on self-funding among smaller firms are more limited. 
BLS reports that 31 percent of participants in establishments with
fewer than 100 employees self-funded their health plans in 1992. 
However, this overstates the extent of self-funding among small
employers because some self-funded larger employers may have multiple
establishments that are included in the survey.  Another 1993 survey,
by the Robert Wood Johnson Foundation in 10 states, found that only 6
to 10 percent of establishments with fewer than 50 employees
self-funded their health plans. 

   Figure 4.4:  Growing Share of
   Insured Workers and Dependents
   in Self-Funded Health Plans,
   1980-93

   (See figure in printed
   edition.)

Note:  Includes firms with at least 100 employees 1988-93 and at
least 50, 100, or 250 employees depending on industry 1980-86. 

Source:  BLS Employee Benefit Surveys, 1980-93. 

The growth of self-funding has occurred primarily among indemnity and
PPO plans, but some employers are beginning to develop arrangements
in which they self-fund their HMO plans as well.  In general, large
employers have self-funded their indemnity plans and contracted with
HMOs as fully insured products that are subject to state insurance
requirements.  However, KPMG Peat Marwick's survey indicates that as
much as 20 percent of HMO enrollment may be in self-funded
arrangements.  In addition, AAHP, an association that represents
HMOs, reports that more than half of HMOs offered self-insured
products to employers in 1995. 


--------------------
\34 For information on ERISA and its implications for health plans,
see Employer-Based Health Plans:  Issues, Trends, and Challenges
Posed by ERISA (GAO/HEHS-95-167, July 25, 1995). 

\35 For a detailed discussion of the costs associated with these and
other state insurance requirements, see Health Insurance Regulation: 
Varying State Requirements Affect Cost of Insurance (GAO/HEHS-96-161,
Aug.  19, 1996.)


CONCLUDING OBSERVATIONS
============================================================ Chapter 5

Although private health insurance premiums have stabilized in recent
years, the share of Americans covered by private health insurance
continues to gradually decline.  This trend has persisted despite
growth in employment since 1992.  Rising costs and declining coverage
have contributed to fundamental changes in many Americans' health
coverage.  More people have Medicaid coverage or are uninsured. 
Those with private coverage often have to pay a greater share of a
plan's premiums and are more likely to be in a health plan that
manages their use of health services. 

Although it is difficult to predict trends, at some point costs are
likely to start rising again.  Many of the underlying causes of
rising health care costs, such as an aging society and new medical
technologies, remain.  Furthermore, many of the easily obtained
savings from switching to managed care plans may have been achieved. 
If health insurance costs increase, employers and insurers will have
greater pressure to continue pursuing stronger cost control methods. 
Thus, employees may continue to be required to pay more for health
coverage, employers will likely encourage enrollment in tighter forms
of managed care, and some employers may even stop offering health
benefits. 

Public demand for federal and state reforms to address the
affordability of and access to private health coverage may increase
if private health coverage continues to erode.  Federal legislation,
the Health Insurance Portability and Accountability Act, has expanded
the federal role in regulating private health insurance.  Future
attention may focus on groups that have been most likely to lose
private coverage--children, early retirees, and near-poor families. 
In addition, the public sector may be called upon to react to changes
in the private sector, such as by examining managed care practices
and the consequences of increased self-funding of health benefits by
employers. 


MAJOR DATA SOURCES AND METHODOLOGY
FOR GAO ESTIMATES
=========================================================== Appendix I

There is no single, comprehensive source of data on private health
insurance.  Instead, a variety of surveys provide information based
on samples of people receiving private health insurance, employers
offering health insurance, or health plans providing health coverage. 
Each survey samples different populations, focuses on different
questions, and uses different methodologies.  In addition, few data
sources have consistently used a comparable methodology over a long
period of time, often requiring that long-term trends be identified
from a variety of data sources or adjusting the data to make them
more comparable.  This appendix briefly describes the major data
sources used for this report and the methodology we used when major
adjustments were necessary to make estimates for trends in private
health insurance. 


   DATA ON HEALTH INSURANCE
   COVERAGE FROM THE BUREAU OF THE
   CENSUS' CURRENT POPULATION
   SURVEY
--------------------------------------------------------- Appendix I:1

The Current Population Survey (CPS) provides data on the source of
health insurance coverage, or lack thereof, for the civilian
noninstitutional population of the United States.  The March 1996 CPS
surveyed 48,000 households with over 136,000 individuals and asked
about health insurance coverage during the previous year.  Thus, the
March 1996 CPS asks about health coverage in 1995.\36

The CPS contains data on health insurance coverage for the period
1979 to 1996, but several methodological changes make it difficult to
compare the data.  These include changes in the questions asked, the
way in which responses are weighted to be nationally representative,
and the method of interviewing and processing data.  While some of
these changes have had only minor effects on the comparability of the
data, in other cases we have adjusted the data reported by the CPS to
make them more comparable with more recent years. 

The most significant revisions occurred in March 1988 when the
questions regarding private health insurance coverage and children's
health coverage were changed.  In particular, these revisions
resulted in the identification of many children as having health
insurance who would have previously been identified as being
uninsured.  Following this change in the CPS questions, the CPS
reported a drop in the number of uninsured persons from 37.4 million
in 1986 to 31.0 million in 1987 while private health insurance
coverage went from 170.4 million to 182.2 million.  Census notes that
"most of the difference was a result of changes to the health
insurance questions" rather than an actual decline in the number of
uninsured.  Therefore, we adjusted the CPS data prior to 1987 to make
them more comparable with data since 1987.  Specifically, we assumed
that the rate of change in coverage between 1986 and 1987 was the
average rate of change for the 2 years before and after 1986-87.  In
effect, this increases the share of the nonelderly population with
private health insurance as reported by CPS prior to 1987 by about
3.6 percentage points. 

Another redesign of the CPS private health insurance questions
occurred in March 1995.  Census representatives report that this
change affects the distribution of respondents with
"employment-based" coverage and "other private" coverage, but the
number of people with private insurance broadly (combining both
"employment-based" and "other private" coverage) is not significantly
affected.  For this reason, this report discusses the combined
"private health coverage" rather than "employment-based" coverage
when comparing trends from the CPS before and after 1994.\37

We made a minor adjustment to the CPS data prior to 1992 to reflect
changes in the data resulting from the way Census weights CPS
responses to be nationally representative.  As reported by Census,
data prior to 1992 were weighted based on the 1980 decennial census,
whereas since 1992 data have been weighted based on the 1990
decennial census.  Census has reported data from the March 1990 and
March 1993 CPS with both the 1990 and 1980 weights.  Comparing data
for 1989 and 1992 with both sets of weights, we calculate that the
change in weights resulted in a shift in private health insurance
coverage rates of about -0.5 percentage points.\38 To avoid having a
one-time shift in coverage rates resulting from the change in
weights, we adjusted the CPS data to spread the effect of the
weighing change over multiple years.  Specifically, we revised the
data from earlier CPS years by a graduated proportion of this
difference resulting from the weights.  That is, the March 1989 CPS
data are decreased by an amount equal to 90 percent of the effect
attributed to changing the weights (90 percent of -0.5 percentage
points for private health insurance coverage), whereas the March 1981
CPS is decreased only by 10 percent of the effect attributed to
changing the weights.  Data from the March 1991 and March 1992 CPS
were adjusted for the full effect of the change in weights (-0.5
percentage points for private health insurance). 

Table I.1 presents the unadjusted CPS data on private health
insurance as reported by Census and our estimates based on the
adjustments discussed above; figure I.1 shows our estimates of health
coverage resulting from these adjustments. 



                         Table I.1
          
          GAO Estimates of Private Health Coverage
          Based on the Current Population Survey,
                          1979-95

                    Unadjusted      Adjusted
                       private       private    Net effect
                        health        health            of
                      coverage    coverage\a    adjustment
----------------  ------------  ------------  ------------
1979                     76.4%         80.0%           3.6
1980                       n/a        79.5\b           n/a
1981                      75.5          78.9           3.5
1982                      74.4          77.8           3.4
1983                      73.1          76.5           3.4
1984                      72.5          75.8           3.3
1985                      72.5          75.8           3.3
1986                      72.7          75.9           3.2
1987                      76.2          75.8          -0.5
1988                      75.5          75.1          -0.5
1989                   75.5,\c          75.0          -0.5
                        75.0\d
1990                      73.8          73.3          -0.5
1991                      72.7          72.2          -0.5
1992                    71.6\c          71.0          -0.5
                        71.0\d
1993                      70.8          70.8
1994                      70.6          70.6
1995                      70.5          70.5
----------------------------------------------------------
Note:  "n/a" is not available. 

\a CPS data adjusted to make trends in coverage more comparable as a
result of (1) a change in survey questions in the March 1988 CPS and
(2) changes in weights in the March 1993 survey. 

\b The 1981 CPS did not ask about nongroup private health insurance
coverage for 1980.  We estimated 1980 private health insurance
coverage as the average for 1979 and 1981. 

\c 1980 weights. 

\d 1990 weights. 

   Figure I.1:  Estimate of
   Private Health Coverage,
   Population Under Age 65,
   1979-95

   (See figure in printed
   edition.)

Source:  GAO estimate based on Bureau of the Census data. 

We used the March 1990 CPS (1989 data) and the March 1996 CPS (1995
data) to assess how private health insurance coverage has changed
among subgroups of the U.S.  population because the data from the two
years are comparable without making any adjustments.  In particular,
Census provided revised weights for the March 1990 CPS to reflect the
1990 decennial census.  This is the same weighing method used in
making the March 1996 CPS nationally representative.  Therefore, we
were able to compare changes in private coverage between 1989 and
1995 without any artificial data distortions resulting from changes
in the CPS weighing methods. 


--------------------
\36 Some experts, however, believe that most respondents provide
information on their health coverage at the time of the survey--that
is, March 1996. 

\37 Other revisions in the CPS since 1979 include changes in how
responses are weighted to be nationally represented and a shift from
paper and pencil to computer-assisted telephone interviews. 

\38 In addition, Census has provided revised weights for the March
1990 CPS.  For this reason, we use the March 1990 CPS as our base in
comparing changes in health coverage among segments of the
population. 


   HEALTH SPENDING DATA FROM THE
   HEALTH CARE FINANCING
   ADMINISTRATION
--------------------------------------------------------- Appendix I:2

The Health Care Financing Administration (HCFA) reports data on
national health care expenditures.\39 These national health
expenditure accounts include data by spending type (for example,
hospital, physician, and pharmaceutical), as well as funding sources
such as businesses, households, and government.  In this report, we
report on total national health care expenditures and spending for
private health insurance.  National health care expenditures include
all spending for personal health care services and
supplies--including hospitals, physicians, home health,
pharmaceuticals, nursing facilities, and other services and
supplies--as well as spending associated with health research and
construction.  Private health insurance expenditures include payments
by private insurers to health care providers as well as the net costs
of administration and profits.  The premiums paid to private health
insurers are divided into payments by businesses, households, and
governments. 


--------------------
\39 See Katharine R.  Levit and others, "National Health
Expenditures, 1995," Health Care Financing Review, Vol.  18, No.  1
(fall 1996), pp.  175-214, and Cathy A.  Cowan and others, "Business,
Households, and Government:  Health Spending, 1994," Health Care
Financing Review, Vol.  17, No.  4 (summer 1996), pp.  157-78. 


   EMPLOYER HEALTH BENEFIT SURVEYS
--------------------------------------------------------- Appendix I:3

Several organizations conduct nationally representative surveys based
on random samples of employer health plans.  In general, these
surveys ask employers about the types of health plans provided to
employees, premiums paid, and features of the benefit design.  Key
features of the employer surveys used in this report are summarized
below.\40


--------------------
\40 For a detailed summary of various surveys of employer health
plans, see Pamela Farley Short, Data Sources for Studies of
Self-Insured Health Plans (Washington, D.C.:  RAND, Nov.  1995). 


      BLS EMPLOYEE BENEFIT SURVEYS
------------------------------------------------------- Appendix I:3.1

The Bureau of Labor Statistics (BLS) annually surveys a nationally
representative sample of employers regarding their employee benefit
plans, including health benefits.  Between 1979 and 1986, BLS
conducted employee benefit surveys of full-time employees in medium
and large establishments, defined then as having at least 100, or 250
or more, employees, depending on industry.  In 1988 and in
odd-numbered years since 1989, BLS has surveyed medium and large
private establishments, currently defined as employing at least 100
workers.  BLS has surveyed small private establishments (employing
fewer than 100 workers) and state and local governments during
even-numbered years since 1990.\41

The 1993 medium and large employer survey included 2,325 responding
establishments; the 1994 small employer survey included 2,135
responding establishments. 


--------------------
\41 See BLS, Employee Benefits in Medium and Large Private
Establishments, 1993 (Washington, D.C.:  Nov.  1994); Employee
Benefits in Small Private Establishments, 1994 (April 1996); and
Employee Benefits in State and Local Governments, 1994 (1996). 


      KPMG PEAT MARWICK'S SURVEY
      OF EMPLOYER-SPONSORED HEALTH
      BENEFITS
------------------------------------------------------- Appendix I:3.2

Since 1991, KPMG Peat Marwick has annually surveyed a nationally
representative sample of private and public employers with 200 or
more workers regarding the health benefits provided to their
employees.\42 Because the KPMG survey questions are similar to
employer surveys conducted by the Health Insurance Association of
America (HIAA) from 1987 to 1993, we use data from both the HIAA and
KPMG surveys to report on trends in private health insurance since
1987.  Although the KPMG annual survey is limited to employers with
at least 200 employees, KPMG and Wayne State University conducted
similar surveys of employers with fewer than 200 employees in 1993,
1995, and 1996.  We combined information from these surveys to
provide data for employers of all sizes.\43


--------------------
\42 In 1996, the KPMG survey included 1,151 firms.  See KPMG Peat
Marwick, Health Benefits in 1996 (Tysons Corner, Va., and San
Francisco, Calif.:  Oct.  1996). 

\43 See Gail Jensen and others, "The New Dimension of Managed Care
Insurance Trends in the 1990s," Health Affairs, Vol.  16, No.  1
(Jan.-Feb.  1997). 


      FOSTER HIGGINS
------------------------------------------------------- Appendix I:3.3

Foster Higgins also conducts an annual survey on health benefits
offered by private employers.\44 Since 1993, the survey has been
based on a random sample of employers with at least 10 employees. 
Between 1986 and 1992, the survey was not randomly selected and
focused on large employers (typically with at least 500 employees). 


--------------------
\44 See Foster Higgins, National Survey of Employer-sponsored Health
Plans (1996). 


      ROBERT WOOD JOHNSON
      FOUNDATION
------------------------------------------------------- Appendix I:3.4

In 1993, the Robert Wood Johnson Foundation commissioned a survey of
2,000 employers regarding their health benefits in 10 states: 
Colorado, Florida, Minnesota, New Mexico, New York, North Dakota,
Oklahoma, Oregon, Vermont, and Washington.\45 According to this
study, the survey, which included firms of all sizes, generally
reflected the national pattern of employers health benefits. 


--------------------
\45 See Joel C.  Cantor, Stephen H.  Long, and M.  Susan Marquis,
"Private Employment-Based Health Insurance in Ten States," Health
Affairs (summer 1995), pp.  199-211. 


   GAO ESTIMATES OF MANAGED CARE
   ENROLLMENT
--------------------------------------------------------- Appendix I:4

We estimated enrollment in managed care health plans using several
sources.  We used data from Interstudy's annual survey of U.S. 
health maintenance organizations (HMO) for HMO enrollment since
1980.\46 We estimated enrollment in preferred provider organizations
(PPO) and point-of-service (POS) plans since 1987 using data from the
KPMG Peat Marwick and HIAA survey discussed above.  The KPMG-HIAA
survey provides the distribution of enrollment in the various health
plan types for employees in firms with at least 200 employees.  We
multiply the ratio of PPO and POS enrollees to HMO enrollees to the
number of private HMO enrollees (not including Medicaid and Medicare
enrollment) to estimate the number of enrollees in PPO and POS plans. 

Our method may underestimate PPO and POS plan enrollment if the ratio
of PPO and POS enrollees to HMO enrollees is higher in smaller firms
(those with fewer than 200 employees).  Although smaller firms are in
general less likely to use managed care, this fact does not affect
our estimates of managed care enrollment.  However, if smaller firms
tend to have more enrollment in PPO and POS plans than in HMO plans
relative to larger firms, then we may underestimate PPO and POS
enrollment.  According to the 1995 survey of firms of all sizes
conducted by KPMG and Wayne State University, the ratio of PPO and
POS to HMO enrollment is somewhat greater if small firms are
included.  If we had used this ratio of all firm sizes rather than
the ratio for firms with more than 200 employees, our estimate of PPO
and POS enrollment would have been 10.4 million higher (65.3 million
rather than 54.9 million).  However, comparable data on managed care
enrollment among smaller firms are not available for earlier years. 

The American Association of Health Plans (AAHP) reports that 58.2
million people were enrolled in HMOs by October 1, 1995, including
10.8 million in HMOs with a POS option.  Thus, if HMO enrollees with
a POS option are excluded, the 1995 AAHP survey reports a similar
number of Americans enrolled in an HMO plan (47.4 million) as the
1995 Interstudy survey (46.2 million). 

The AAHP survey finds that 1995 enrollment in PPO plans is much
higher (91.0 million) than our estimate (54.9 million, including POS
enrollment).  However, AAHP's survey may double count some PPO
enrollment.  Thus, while our method may underestimate PPO and POS
enrollment, the AAHP survey may overestimate PPO enrollment.  In
addition, AAHP reports on PPO enrollment only since 1992.  In order
to be able to report on managed care enrollment since 1980, we used a
consistent method to estimate PPO enrollment since 1987.  We used
data from HIAA to estimate PPO enrollment between 1984 and 1986. 
Prior to 1984, fewer than 1 million Americans were enrolled in PPOs. 


--------------------
\46 See Interstudy Publications, The InterStudy Competitive Edge: 
HMO Industry Report 6.2 (Minneapolis:  1996). 


PRIVATE HEALTH INSURANCE COVERAGE
WITHIN POPULATION CATEGORIES, 1989
AND 1995
========================================================== Appendix II



                               Table II.1
                
                 Share of Population Under Age 65 With
                  Private Health Insurance Coverage by
                          State, 1989 and 1995

                                                                Differ
                                                Privat  Privat    ence
                                                     e       e      in
                                                health  health  covera
                                                covera  covera     ge,
                                                   ge,     ge,   1989-
State                                             1989    1995      95
----------------------------------------------  ------  ------  ------
U.S.                                             75.0%   70.5%    -4.5
New England                                       83.2    78.6    -4.6
Connecticut                                       87.7    81.7    -6.0
Maine                                             80.5    76.2    -4.3
Massachusetts                                     81.9    77.7    -4.2
New Hampshire                                     81.3    81.2    -0.1
Rhode Island                                      81.5    76.0    -5.5
Vermont                                           83.6    75.5    -8.1
Middle Atlantic                                   78.3    72.8    -5.5
New Jersey                                        81.9    76.6    -5.3
New York                                          74.5    68.1    -6.4
Pennsylvania                                      82.0    77.6    -4.4
East North Central                                80.6    78.0    -2.6
Illinois                                          79.1    76.3    -2.8
Indiana                                           78.8    79.9     1.1
Michigan                                          79.3    79.0    -0.3
Ohio                                              82.9    75.9    -7.0
Wisconsin                                         83.5    82.0    -1.5
West North Central                                81.5    78.2    -3.3
Iowa                                              85.9    80.8    -5.1
Kansas                                            82.0    72.0   -10.0
Minnesota                                         84.5    82.1    -2.4
Missouri                                          77.5    75.4    -2.1
Nebraska                                          79.4    79.9     0.5
North Dakota                                      82.7    80.6    -2.1
South Dakota                                      76.9    78.3     1.4
South Atlantic                                    74.5    68.7    -5.8
Delaware                                          73.8    75.2     1.4
District of Columbia                              61.8    60.7    -1.1
Florida                                           70.4    65.1    -5.3
Georgia                                           73.9    68.1    -5.8
Maryland                                          81.2    72.6    -8.6
North Carolina                                    76.1    70.5    -5.6
South Carolina                                    75.8    68.5    -7.3
Virginia                                          77.1    72.8    -4.3
West Virginia                                     72.9    65.9    -7.0
East South Central                                72.0    68.1    -3.9
Alabama                                           71.5    73.0     1.5
Kentucky                                          74.8    68.0    -6.8
Mississippi                                       64.2    61.3    -2.9
Tennessee                                         74.7    67.7    -7.0
West South Central                                66.6    62.0    -4.6
Arkansas                                          69.9    66.5    -3.4
Louisiana                                         66.8    59.1    -7.7
Oklahoma                                          67.9    63.3    -4.6
Texas                                             66.0    61.9    -4.1
Mountain                                          74.2    69.8    -4.4
Arizona                                           72.1    63.2    -8.9
Colorado                                          74.2    77.7     3.5
Idaho                                             77.6    74.4    -3.2
Montana                                           73.6    69.6    -4.0
Nevada                                            75.7    72.5    -3.2
New Mexico                                        64.2    50.5   -13.7
Utah                                              83.6    80.8    -2.8
Wyoming                                           79.0    71.6    -7.4
Pacific                                           68.8    64.4    -4.4
Alaska                                            65.8    68.3     2.5
California                                        66.2    61.5    -4.7
Hawaii                                            78.7    72.4    -6.3
Oregon                                            78.3    73.9    -4.4
Washington                                        77.6    74.1    -3.5
----------------------------------------------------------------------
Source:  U.S.  Bureau of the Census, Current Population Survey (Mar. 
1989 and Mar.  1996). 



                               Table II.2
                
                Share of Population Under Age 65 Covered
                 by Private Health Insurance, 1989 and
                                  1995

                                                Private health
                                                   coverage
                                                --------------
                                                                Differ
                                                                  ence
                                                                    in
                                                                covera
                                                                   ge,
                                                                 1989-
Age                                               1989    1995      95
----------------------------------------------  ------  ------  ------
Under 18                                         73.1%   66.1%    -7.0
18 to 24                                          65.1    60.2    -4.9
25 to 34                                          73.5    68.3    -5.2
35 to 44                                          81.1    76.2    -4.9
45 to 54                                          81.4    80.0    -1.4
55 to 64                                          78.8    76.5    -2.3
======================================================================
U.S. average (0 to 64)                           75.0%   70.5%    -4.5
----------------------------------------------------------------------
Source:  U.S.  Bureau of the Census, Current Population Survey (Mar. 
1989 and Mar.  1996). 



                               Table II.3
                
                Share of Population Under Age 65 Covered
                 by Private Health Insurance by Race or
                      Ethnic Group, 1989 and 1995

                                                Private health
                                                   coverage
                                                --------------
                                                                Differ
                                                                  ence
                                                                    in
                                                                covera
                                                                   ge,
                                                                 1989-
Race or ethnic group                              1989    1995      95
----------------------------------------------  ------  ------  ------
White\a                                          82.2%   79.1%    -3.1
Black