Employment-Based Health Insurance: Costs Increase and Family Coverage
Decreases (Letter Report, 02/24/97, GAO/HEHS-97-35).

Pursuant to a congressional request, GAO provided information on the
decline in employment-based health insurance, focusing on: (1) recent
trends in employment-based private health insurance, particularly for
family coverage; (2) any corresponding changes in the number of adults
and children with private insurance coverage as dependents; and (3) the
potential effect of these changes, if any, on public costs for health
care coverage.

GAO found that: (1) eroding employer financial support for providing
health insurance to employees' families has contributed to the overall
decline in private health insurance coverage; (2) each year between the
late 1980s and 1994, increases in employers' costs to provide health
insurance to their employees and their employees' families outpaced
inflation, with cost growth of 18 percent one year; (3) as health
insurance reached 10 percent of employees' payroll costs, many employers
began to reconsider the amount of support they would provide to
employees, particularly for family coverage; (4) acquiring or
maintaining health insurance has become more difficult for some families
because of changes that some employers made to their firms' health
coverage; (5) some employers, particularly smaller employers, dropped
coverage altogether; (6) in 1993, over 29 million employees, almost
one-fourth of the workforce, were employed by firms that did not offer
group health insurance for employees' families; (7) most employers
continued to offer coverage, but many raised employees' premium
contributions significantly, especially for family coverage; (8) in
1993, 16 percent of employees in large private firms paid $150 or more
per month for family health insurance premiums and 36 percent of state
and local government employees paid as much in 1992; (9) some employers
have used other mechanisms that could discourage employees from
two-worker families from purchasing family coverage from them; (10) as
these changes occurred, the percentage of Americans under 65 years old
with private health insurance coverage decreased from 75 percent in 1989
to about 71 percent in 1995; (11) of this general decline, about 70 to
90 percent was due to fewer working-age adults and children being
covered as dependents; (12) between 1989 and 1995, the percentage of
working-age adults, 18 to 64 year olds, with private insurance coverage
decreased from 76 percent to 73 percent; (13) over these 6 years, the
percentage of children under 18 years old with private health insurance
decreased from more than 73 percent to 66 percent; (14) declines in
employment-based dependent coverage can increase the number of uninsured
Americans and shift a greater burden for health care onto public payers;
(15) between 1994 and 1996, health insurance premium costs have been
relatively stable, which may help slow the erosion of private coverage;*

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

 REPORTNUM:  HEHS-97-35
     TITLE:  Employment-Based Health Insurance: Costs Increase and 
             Family Coverage Decreases
      DATE:  02/24/97
   SUBJECT:  Health insurance
             Employee medical benefits
             Dependents
             Health insurance cost control
             Health care programs
             Insurance premiums
             Inflation
             Employee benefit plans
             Managed health care
IDENTIFIER:  Medicaid Program
             Census Bureau Current Population Survey
             Aid to Families with Dependent Children Program
             AFDC
             Supplemental Security Income Program
             HHS Temporary Assistance for Needy Families Program
             
******************************************************************
** 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 Ranking Minority Member, Subcommittee on Children and
Families, Committee on Labor and Human Resources, U.S.  Senate

February 1997

EMPLOYMENT-BASED HEALTH INSURANCE
- COSTS INCREASE AND FAMILY
COVERAGE DECREASES

GAO/HEHS-97-35

Family Health Insurance

(101396)


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

  AFDC - Aid to Families With Dependent Children
  CHAMPUS - Civilian Health and Medical Program of the Uniformed
     Services
  CPS - Current Population Survey
  DOL - Department of Labor
  HMO - health maintenance organization
  POS - point of service
  PPO - preferred provider organization
  SSI - Supplemental Security Income program
  TANF - Temporary Assistance for Needy Families

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


B-271082

February 24, 1997

The Honorable Christopher J.  Dodd
Ranking Minority Member
Subcommittee on Children and Families
Committee on Labor and Human Resources
United States Senate

Dear Senator Dodd: 

Nearly two-thirds of Americans under 65 years old--some 150 million
people--have employment-based private health insurance.  Although
many employers remain committed to providing employee and family
coverage, the percentage of people with private coverage is
declining.  At the same time, the percentage of Americans who are
uninsured or rely on Medicaid--particularly children--continues to
increase.  The effect of being uninsured on the health of families
can be significant.  For example, uninsured children are less likely
than insured children to receive primary care, immunizations, and
treatment for injuries.  The lack of such care can lead to health
conditions and disabilities that require more costly and long-term
care. 

Concerned about the decline in employment-based health insurance
coverage, you asked us to (1) identify any recent trends in
employment-based private health insurance, particularly for family
coverage; (2) determine any corresponding changes in the number of
adults and children with private insurance coverage as dependents;
and (3) identify the potential effect of these changes, if any, on
public costs for health care coverage.  To answer these questions, we
analyzed surveys of health insurance coverage conducted by the
Department of Labor (DOL) and by private benefits consultants, such
as KPMG Peat Marwick and Hewitt Associates.  We also analyzed the
Bureau of the Census' Current Population Surveys (CPS) on health
insurance coverage for 1989, 1991, 1993, and 1995.  In addition, we
discussed trends with experts, insurance company executives, and
benefits consultants to determine how employer practices may have
changed in the past several years.  We also reviewed research reports
on private insurance and the health insurance marketplace. 

Because more limited information is available on benefit practices at
small firms (fewer than 100 employees), our report primarily focuses
on large firms (100 or more employees) and major firms (over 1,000
employees).  For this report, family generally refers to a group of
people whom insurers would consider a family for the purposes of
health insurance coverage--typically adults related by marriage,
parents, and their children under 18 years old.\1 (See app.  I for
more details of our methodology.)


--------------------
\1 Other adults who could be included as adult dependents in our CPS
analysis include young unmarried adults under 19 years old or in
college who are covered through their parents' health insurance
policies and married spouses who have separated from the primary
family policyholder.  In addition, some employers extend health
insurance coverage and other benefits to unmarried partners of
employees as dependents--either gays or unmarried heterosexuals--and
they could also be included as adult dependents. 


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

Eroding employer financial support for providing health insurance to
employees' families has contributed to the overall decline in private
health insurance coverage.\2 Each year between the late 1980s and
1994, increases in employers' costs to provide health insurance to
their employees and their employees' families outpaced
inflation--with cost growth of 18 percent one year.  As health
insurance reached 10 percent of employees' payroll costs, many
employers began to reconsider the amount of support they would
provide to employees, particularly for family coverage. 

Acquiring or maintaining health insurance has become more difficult
for some families because of changes that some employers made to
their firms' health coverage.  Some employers--particularly smaller
employers--dropped coverage altogether.  In 1993, over 29 million
employees--almost one- fourth of the workforce--were employed by
firms that did not offer group health insurance for employees'
families.  Most employers continued to offer coverage, but many
raised employees' premium contributions significantly--especially for
family coverage.  In 1993, 16 percent of employees in large private
firms paid $150 or more per month for family health insurance
premiums; 36 percent of state and local government employees paid as
much in 1992.  Some employers have used other mechanisms, such as
financial incentives, that could discourage employees from two-worker
families from purchasing family coverage from them. 

As these changes occurred, the percentage of Americans under 65 years
old with private health insurance coverage decreased from 75 percent
in 1989 to about 71 percent in 1995.  Of this general decline, about
70 to 90 percent was due to fewer working-age adults and children
being covered as dependents.  Between 1989 and 1995, the percentage
of working-age adults (18 to 64 years old) with private insurance
coverage decreased from 76 percent to 73 percent.  If the same
percentage of working-age adults had been covered in 1995 as in 1989,
about 5 million more adults would have had private insurance. 
However, children experienced the greatest loss of private coverage. 
Over these 6 years, the percentage of children under 18 years old
with private health insurance decreased from more than 73 percent to
66 percent.  If private coverage levels had not decreased, about 5
million more children would have private insurance. 

Declines in employment-based dependent coverage can increase the
number of uninsured Americans and shift a greater burden for health
care onto public payers.  Between 1994 and 1996, health insurance
premium costs have been relatively stable, which may help slow the
erosion of private coverage.  However, unless the decline in
employment-based insurance coverage abates, public payers could face
increased costs for health care--either for uncompensated care or for
public insurance. 


--------------------
\2 Most people under 65 years old with private coverage obtain their
health insurance through employment-based plans.  Private insurance
purchased directly by individuals covers about 5 percent of the
population under 65 years old as their only source of health
insurance coverage.  For more information on the structure of the
private market for individual coverage, see Private Health Insurance: 
Millions Relying on the Individual Market Face Cost and Coverage
Trade-Offs (GAO/HEHS-97-8, Nov.  25, 1996). 


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

Support for employment-based health insurance by employers
contributes to the health and financial security of employees and
their families.  U.S.  employers traditionally have provided private
group health insurance as an employment benefit for their employees
and their employees' spouses and children.  Beginning with World War
II--when wages were frozen and employers wanted to attract good
employees--employment-based insurance became a more common fringe
benefit.  Today, private health insurance offered through employment
is the main source of health insurance coverage in the United
States--in 1995, more than 90 percent of people under 65 years old
with private insurance--150 million people--were insured through
their employment.  The majority of working adults 18 to 64 years old
with private insurance (74 percent) work for private companies, but
17 percent work for the federal, state, or local governments.  Most
of the remainder are self-employed. 

Employment-based insurance--where employers pay part or all of the
costs--can be advantageous for employees and many employers.\3 For
employees, such health insurance is generally more affordable because
they receive group rates for coverage, which are typically lower than
those for individual coverage.  In addition, employees do not pay
taxes on contributions that their employers make toward their
employment-based coverage--an advantage employees would lose if they
were to receive additional cash income and to purchase individual
coverage.  For employers, offering affordable health insurance is an
attractive benefit that helps them promote the health and
productivity of their work force and remain competitive in recruiting
new employees.  Employers' contributions to employee benefits are
deductible from their companies' gross income and thus reduce their
companies' tax liability.\4 In the United States, the system of
private health insurance based on employment is entirely voluntary. 
Employers are not required to provide insurance, nor are employees
required to purchase it. 

In addition to private insurance, the federal, state, and some local
governments provide public funding for health insurance, primarily
through Medicaid, Medicare, state health insurance plans, and the
Civilian Health and Medical Program of the Uniformed Services
(CHAMPUS).  Medicaid is the largest public source of health insurance
coverage for children and working-age adults, covering 29 million
people under 65 years old in 1995.  Enacted in 1965, Medicaid was
designed to provide health care coverage for populations whose
incomes and resources were insufficient to meet the costs of needed
medical care, including adults and children receiving Aid to Families
With Dependent Children (AFDC) and aged, blind, or permanently and
totally disabled individuals.  Although Medicaid has expanded
eligibility beyond these groups, it still limits eligibility to
specific populations of lower-income adults and children.  Medicare
provides health care coverage for over 37 million people--most of
them people 65 years old or older.  Medicare also covers people
entitled to disability benefits for 24 months or more, people with
end-stage renal disease requiring dialysis or kidney transplant, and
certain others who elect to buy into the program through premium
payments.  CHAMPUS provides medical care for active-duty or retired
military families, as well as to the immediate families of deceased
active-duty or retired military personnel. 

Although these private and public health insurance systems provide
coverage for many Americans, many remain uninsured.  In 1995, more
than 40 million people under 65 years old had no health insurance for
the entire year, including many employees and their families. 


--------------------
\3 For employers of low-wage or part-time employees, the advantage
for the employer is not as great.  When employers consider salary and
benefit costs together as an employee's total compensation, benefits
represent a much larger share of a low-wage employee's total
compensation than a high-wage employee's total compensation. 

\4 However, employers would get the same deduction if they paid any
other legitimate business expense, such as cash wages. 


   EMPLOYERS REEXAMINED THEIR ROLE
   AS PREMIUM COSTS INCREASED
------------------------------------------------------------ Letter :3

From the late 1980s to the early 1994, the cost of health insurance
premiums\5 rose rapidly, especially for family coverage.  With these
increases in costs, many employers began to reexamine their role and
the benefits they offered to employees, and some employers began to
question the extent of their responsibility to finance coverage for
families. 

In the late 1980s, the cost of employment-based health insurance
premiums significantly outpaced inflation.  Between 1988 and 1989,
employer costs for health insurance rose 18 percent in one year.  By
contrast, general inflation was under 5 percent.  Health insurance
premium costs began to stabilize recently.  (See fig.  1.) However,
health insurance continues to be a major portion of employers' total
compensation to employees--7.3 percent of payroll costs in
1993,compared with 4.4 percent in 1980. 

   Figure 1:  Increases in Health
   Insurance Premiums, 1991 to
   1996

   (See figure in printed
   edition.)

Source:  KPMG Peat Marwick, Health Benefits in 1996.  This was a
survey of about 1,000 randomly selected public and private employers
with 200 or more employees. 

Between 1989 and 1996, cost increases for family premiums were 13 to
23 percent higher than cost increases for employee-only premiums,
depending on the type of health plan.  For example, since 1989,
premium costs for health maintenance organization (HMO)\6 coverage
for families increased 59 percent, while premium costs for
employee-only HMO coverage increased only 36 percent.  (See fig.  2
and table II.1)

   Figure 2:  Percent Increases in
   Health Insurance Premiums for
   Employee-Only and Family
   Coverage, 1989-96

   (See figure in printed
   edition.)

Source:  Health Insurance Association of America and KPMG Peat
Marwick.  The Health Insurance Association of America survey was of
about 2,600 public and private employers with at least 2 employees
(for 1989) and the KPMG Peat Marwick surveys were of about 1,000
randomly selected public and private employers with 200 or more
employees (all subsequent years). 


--------------------
\5 Many large companies self-insure, so that while their employees
generally contribute to the cost of their health coverage, they are
not paying an insurance premium.  However, for simplicity, we will
refer to all employee contributions for their health coverage that
function similar to a premium payment as premiums. 

\6 HMOs are organized health care systems that are responsible for
both the financing and delivery of a broad range of comprehensive
health services to an enrolled population. 


      SOME EMPLOYERS QUESTION
      THEIR ROLE IN PROVIDING
      FAMILY COVERAGE
---------------------------------------------------------- Letter :3.1

With the surge in health insurance premium costs, some companies
began to reevaluate their obligation to provide coverage to employees
and especially their dependents.  A recent survey of 601 businesses
found that 40 percent would prefer to pay 50 percent or less of
employee insurance premiums and only a minority believed that they
should continue to pay the full cost of employee-only premiums.  Of
those who thought businesses should be required or encouraged to
provide insurance to employees' families, nearly half agreed that
employers should contribute an even smaller share for family coverage
than employee coverage.  According to the study, employers viewed
their role in providing coverage to employees and their dependents as
diminishing.\7

Some firms--particularly those with fewer than 25 employees or that
primarily employ low-wage employees--do not offer health insurance. 
In a 1994 survey of over 22,000 establishments in 10 states, 42
percent did not offer health insurance benefits; most were
establishments with 1 to 4 employees or with a higher-than-average
percentage of low-wage or part-time employees.\8

Historically, large employers and certain types of businesses--such
as manufacturing and other highly unionized industries--have provided
insurance packages with generous benefits for employees and their
families.  Yet by doing so, these large employers in essence
subsidize other employers who do not cover their employees or offer a
less comprehensive package.\9 Offering an attractive and costly
benefits package can put the firm at a competitive disadvantage with
firms who do not pay as much for benefits, and do not attract their
employees' families to enroll. 

According to several benefits consultants, some employers no longer
want to subsidize families to the extent that they have because they
prefer to more closely link total employee compensation to work
contribution.  Employers that provide generous family health
insurance packages, in effect, pay employees with family coverage
more than they pay employees without family coverage--considering the
value of benefits.  Some employers are concerned that this is not
equitable.  For example, a company that pays 100 percent of the cost
of employee health insurance premiums could provide a benefit that is
worth, on average, more than twice as much to the employee who
chooses family coverage over employee-only coverage--$5,000 versus
$2,000. 


--------------------
\7 Jack A.  Meyer, Diane H.  Naughton, and Michael J.  Perry,
Assessing Business Attitudes on Health Care (Washington, D.C.: 
Economic and Social Research Institute, 1996). 

\8 Joel Cantor, Stephen Long, and M.  Susan Marquis, "Private
Employment-Based Health Insurance in Ten States," Health Affairs,
Vol.  14, No.  2 (1995), pp.  199-211. 

\9 Deborah Chollet, "Employer-based Health Insurance in a Changing
Work Force," Health Affairs, Vol.  13, No.  1 (1994), pp.  315-26. 


   EMPLOYERS RAISE EMPLOYEES'
   CONTRIBUTIONS TO PREMIUMS,
   ESPECIALLY FOR FAMILIES, AND
   DISCOURAGE FAMILY COVERAGE IN
   OTHER WAYS
------------------------------------------------------------ Letter :4

Employers have responded to the increases in health insurance costs
in several ways.  Some employers stopped offering health insurance
coverage altogether.  Many who retained coverage have switched to
managed care plans in an attempt to control premium costs.  Many
employers also increased the amounts employees had to pay toward
their premiums, with growth in premium contributions by employees for
family coverage outstripping growth in premium contributions for
employee-only coverage.  Some employers used other strategies to
encourage employees not to choose family health insurance coverage,
including paying incentives to those who choose employee-only
coverage. 

These changes can provide significant savings for companies.  A
benefits consultant reported to us that certain companies have saved
15 to 20 percent in costs associated with their health plans by
increasing family health insurance premium costs or by otherwise
discouraging employees from choosing family coverage. 


      TO OFFSET INCREASES, SOME
      EMPLOYERS DROPPED
      COVERAGE--OTHERS SWITCHED TO
      MANAGED CARE PLANS
---------------------------------------------------------- Letter :4.1

A small percentage of employees may have lost coverage because their
employer dropped health insurance or because they began working for a
firm that did not offer coverage.  Overall, 78.4 percent of employees
reported that their employers sponsored health insurance plans in
1993, compared with 79.3 percent in 1988.\10 Smaller firms were more
likely to stop offering health insurance--13 percent fewer people
working in firms with under 10 employees reported that their
employers' offered coverage in 1993, compared to 1988.  Larger
firms--under 250 employees--also dropped coverage, but at much lower
rates. 

By 1993, more than 29 million employees--almost one-fourth of the
workforce--could not get employment-based health insurance for their
families.  Eighteen percent of these employees worked for firms that
did not offer health insurance; about 5 percent worked for firms that
offered employee-only health insurance but no coverage for other
family members.\11

Other employers reacted to health insurance premium increases by
encouraging their employees to enroll in managed care.  From 1984 to
1993, the number of employees in large firms who were enrolled in
managed care plans, such as HMOs, increased from 5 percent to 50
percent.  HMO premiums were generally lower than premiums for
fee-for-service plans, so employers could lower premium costs by
switching the types of plans they offered.  This is coupled with
fewer employers offering indemnity plans.  KPMG Peat Marwick reported
that 89 percent of employees with employment-based coverage could
choose a conventional indemnity plan in 1989; that percentage dropped
to 57 percent by 1996. 


--------------------
\10 The question asked if employers' had a plan, whether or not the
employee was eligible to participate.  See Employment-Based Health
Benefits:  Analysis of the April 1993 Current Population Survey,
Employee Benefit Research Institute Special Report SR-24 and Issue
Brief No.  152 (Washington, D.C.:  Employee Benefit Research
Institute, 1994). 

\11 Employment-Based Health Benefits:  Analysis of the April 1993
Current Population Survey. 


      EMPLOYERS INCREASED
      EMPLOYEES' CONTRIBUTIONS FOR
      HEALTH INSURANCE,
      PARTICULARLY FOR FAMILY
      COVERAGE
---------------------------------------------------------- Letter :4.2

To offset increases in health insurance costs, some employers opted
to have employees share in the costs of their insurance premium or
increased their share of these costs.  According to DOL, less than
one-half the employees in large firms contributed to employee-only
health insurance premiums in 1988.  By 1993, more than 60 percent
did.\12 In addition, more than 75 percent of employers required
employees to share in the costs of family premiums in 1993. 
Employees' share of premium costs are higher for family coverage--30
percent for family coverage in 1996, compared to 22 percent for
employee-only coverage, according to Peat Marwick.  Since 1989,
employees' share of premiums increased more rapidly for employee-only
coverage, as fewer firms offered coverage at no cost--which is more
common for employee-only coverage.  Hewitt Associates also found that
fewer major employers provided health insurance plans at no cost to
their employees or employees' families in 1995 than the same
companies did in 1990.  (See fig.  3.)

   Figure 3:  Percent of Major
   Firms Offering Health Insurance
   With No Employee Premium
   Contribution, 1990 and 1995

   (See figure in printed
   edition.)

Notes:  Major firms have 1,000 or more employees.  Preferred provider
organization (PPO) plans provide financial incentives for patients to
get care from a selected network of doctors and hospitals by charging
additional fees if patients go to providers outside the preferred
network.  Indemnity plans refer to traditional insurance plans, which
reimburse providers and patients on a fee-for-service basis.  HMOs
require patients to have services delivered by providers affiliated
with them, except for emergency treatment.  HMOs also typically
require patients to select a primary care physician to coordinate the
patient's care, especially for services involving referrals to
specialists and hospital care.  Point-of-service (POS) plans are
similar to PPO plans, in that they encourage enrollees to use a
selected network of doctors and hospitals, but allow patients to see
providers out of the network if the patient pays additional fees for
that care.  Like HMOs, POS plans have enrollees select a primary care
physician who coordinates care for the patient, including care
requiring referrals to specialists.  The Hewitt Associates report did
not provide comparative information about HMOs and POS-type plans for
1990--only information for 1995. 

Source:  Hewitt Associates, Salaried Employee Benefits Provided by
Major U.S.  Employers in 1990 and 1995:  A Comparison Study, 1996. 

Employees' average monthly contribution also increased significantly
between 1988 and 1993.  Increases generally were greater for
employees with family coverage than for those with employee-only
coverage.  According to DOL, in large firms average monthly
contributions for family coverage increased 79 percent, compared to
64 percent for employee-only coverage between 1988 and 1993.  (See
table 1.)



                                Table 1
                
                 Average Monthly Premium Contributions
                Paid by Employees in Large Firms, 1988,
                         1989, 1991, and 1993\a

                                                                Percen
                                                                     t
                                                                increa
                                                                    se
                                                                 1988-
Average monthly contribution\b    1988    1989    1991    1993      93
------------------------------  ------  ------  ------  ------  ------
Employee-only coverage             $19     $25     $27     $32      64
Family coverage                     60      72      97     107      79
----------------------------------------------------------------------
Note:  Large firms have 100 or more employees.  Percent increase may
not calculate exactly from the premium costs in this table due to
rounding. 

\a Full-time employees only in medium and large establishments (100
or more employees).  DOL also surveys small establishments (fewer
than 100 employees). 

\b Based only on those employees who contribute to the cost of their
employment-based premium. 

Source:  Bureau of Labor Statistics. 

Similarly, a Hewitt Associates study comparing benefits offered by
the same set of major U.S.  firms in 1990 and 1995 showed that for
given types of health insurance plans, employees' median monthly
premium contributions increased more for family than for
employee-only coverage.  (See fig.  4.)

   Figure 4:  Median Monthly
   Premium Contributions by
   Employees for Indemnity and PPO
   Plans in Major Firms 1990 and
   1995

   (See figure in printed
   edition.)

Source:  Hewitt Associates, Salaried Employee Benefits Provided by
Major U.S.  Employers in 1990 and 1995:  A Comparison Study, 1996. 


--------------------
\12 DOL surveys medium and large establishments (with 100 or more
employees) and small establishments (fewer than 100 employees)
separately every other year.  References in this report to DOL's
information on large firms includes DOL's surveys of medium and large
establishments only. 


      HEALTH INSURANCE IS
      EXPENSIVE FOR SOME FAMILIES
---------------------------------------------------------- Letter :4.3

For some families, the overall rise in health insurance costs has
made the current price of employment-based health insurance difficult
to afford.  DOL reported that in 1993, 16 percent of employees paid
$150 or more per month for family health insurance premiums.  The
percentage of employees in state and local governments who spend $150
or more per month for family health insurance is even greater than in
private industry.  In 1992, 36 percent of state and local employees
paid $150 or more per month for family health insurance.  In
addition, state and local employees' average monthly premium
contribution for families was $139.23--almost five times the $28.97
average premium contribution for employee-only coverage. 

For low-income families, high premiums may make health insurance
unaffordable.  For example, premiums of $150 per month represent 9
percent of gross income for a family with annual income of $20,000. 
Lower-income and part-time employees are less likely than
higher-income and full-time employees to have employment-based
insurance, in part because it is less affordable for them.  (See fig. 
5.)

   Figure 5:  Percentage of People
   Under 65 Years Old and
   Employees 18 to 64 Years Old
   With Private Health Insurance
   Coverage, by Federal Poverty
   Level, 1995

   (See figure in printed
   edition.)

Note:  The federal poverty level shows the relation of family size
and income to the Federal Poverty Income Guidelines.  In 1996, a
family of three with income at or below $12,980 would be considered
poor--or with income under 100 percent of the federal poverty level. 
A family of three with income between 101 and 200 percent of the
federal poverty level would have income between $12,980 and $25,960. 

Source:  Analysis of March 1996 Current Population Survey. 

A recent study found that 64 percent of the uninsured people
interviewed did not have insurance because they felt that they could
not afford to purchase it, while only 8 percent did not have
insurance because they did not want or need it.\13 In the same study,
36 percent of uninsured people reported problems paying prior medical
bills.  Of both insured and uninsured people having problems paying
their medical bills, 49 percent paid more than $1,000 in
out-of-pocket medical expenses in the previous year and 8 percent
paid more than $5,000. 


--------------------
\13 Karen Donelan, and others, "Whatever Happened to the Health
Insurance Crisis in the United States?  Voices From a National
Survey," The Journal of the American Medical Association, Vol.  276,
No.  16 (1996), pp.  1346-50.  An earlier study also showed cost was
a major issue for the uninsured:  See David U.  Himmelstein, and
Steffie Woolhandler, "Care Denied:  U.S.  Residents Who Are Unable to
Obtain Needed Medical Services," American Journal of Public Health,
Vol.  85, No.  3 (1995), pp.  341-44. 


      EMPLOYERS USE OTHER
      STRATEGIES TO REDUCE COSTS,
      ESPECIALLY THOSE ASSOCIATED
      WITH FAMILY COVERAGE
---------------------------------------------------------- Letter :4.4

Some changes that employers have made to their benefits packages may
discourage employees from choosing family coverage.  These benefit
changes include introducing flexible benefit plans and establishing
premium rates based on family size. 

Some firms have designed their benefit plans in ways that encourage
employees in dual-income families to purchase health insurance
coverage from their spouses' employers.  This, coupled with increases
in cost, can eliminate duplicate coverage for dual-income families,
which provides savings for employers.  It may also result in some
employees dropping coverage for their spouses or other family
members. 


         FLEXIBLE BENEFIT PLANS
         ENCOURAGE SUBSTITUTION OF
         OTHER BENEFITS FOR FAMILY
         COVERAGE
-------------------------------------------------------- Letter :4.4.1

To control benefit-related cost increases and to broaden employees'
choice of benefits, large firms increasingly are offering flexible or
cafeteria-type benefit plans and flexible spending accounts.  In
1995, Hewitt Associates reported that 88 percent of the major firms
in its database offered at least one of these options.  Flexible
plans and accounts allow employees to select the benefits they want
from a menu of benefits, thus allowing them to maximize the value of
their benefits by selecting the ones they need most.  Generally,
firms designate a portion of employee salaries as per-year credits;
employees then allocate these credits among available benefits,
including health insurance.  The amount of employer-provided flexible
benefit credits is typically set each year with reference to some
target--such as the change in current cost of one health insurance
option--plus enough to cover certain other benefits.  However, the
increased flexible benefit credits may not cover employees' increased
costs.  If employees choose health insurance whose cost, along with
other benefits, exceeds the employer-provided credits, employees must
pay the difference.  Some firms allow employees to designate an
additional portion of their salary to increase their flexible
benefits plan or set up a flexible spending account. 

By eliminating the direct link between their contributions and the
cost of health insurance, firms can use flexible benefit plans to
control their benefit costs.  Over time, they can make other changes
to shift more cost to employees with families, such as expanding the
benefits for employees using employee-only coverage and restraining
benefits for families.  For example, one major high-technology
manufacturing firm gives a $1,000 credit to an employee's flexible
benefits account if the employee chooses to get health insurance
coverage through his or her spouse.  The employee can use this $1,000
credit for other benefits.  Even with the credit, the firm saves over
$2,000 per year, per employee, when the employee chooses not to elect
family coverage. 


         TIER RATING OFFERS
         GREATER DIFFERENTIATION
         OF PREMIUM RATES
-------------------------------------------------------- Letter :4.4.2

Some employers restructured employee premium payments to base them on
the number of dependents covered.  This tier structure can reduce
health insurance costs for smaller families but raise them for larger
families.  Without these changes, smaller families are subsidizing
larger ones.  Lower premiums for a family composed of a single parent
with one child could encourage such families to purchase coverage for
dependents.  But a higher premium for the larger family could
discourage such families from purchasing coverage for their
dependents. 

A simple two-tier structure would include one price for employee-only
coverage and another price for employee and dependent coverage. 
According to some of the benefit consultants we spoke with,
increasing the number of tiers beyond the two-tier structure is
becoming more common for large firms.  The majority of major firms
have three or more tiers.  (See fig.  6.) An example of a
three-tiered plan would be one that has separate premium prices for
employee-only coverage, for employee plus one dependent, and for
employee with two or more dependents. 

   Figure 6:  Coverage Tiers for
   Major Firms in 1995

   (See figure in printed
   edition.)

Source:  Hewitt Associates, Salaried Employee Benefits Provided by
Major U.S.  Employers in 1995, 1996. 


         EMPLOYERS IMPLEMENT
         STRATEGIES TO SHIFT
         BURDEN OF COVERAGE ONTO
         WORKING SPOUSES
-------------------------------------------------------- Letter :4.4.3

According to benefits consultants, some firms design their benefit
plans to encourage employees with working spouses to get their
insurance from their spouses' company.  These strategies include

  -- refusing to cover a spouse if the spouse has other health
     insurance coverage,

  -- providing incentive payments to employees who refuse family
     coverage,

  -- imposing a surcharge for working spouses covered as dependents,
     and

  -- refusing to provide dependent coverage unless the employee is
     the family's primary wage-earner. 

For example, one major manufacturing firm offers a policy that
supplements major medical for employees' families--covering costs
that other policies do not--if the employees use their spouses'
health insurance as the primary coverage.  This policy covers 100
percent of the first $1,000 of eligible expenses for the
employee--thus allowing the employee to avoid any deductible on the
primary policy--and then pays an additional 25 percent of covered
expenses, with an out-of-pocket maximum of $1,500. 

How often these strategies are used is not known.  In addition, some
of these strategies are difficult to implement without the
cooperation of employees because they depend on self-reporting of
other coverage by working spouses. 

Whether or not such strategies lead to a loss of coverage may depend
on family circumstances.  Where dual-income families have more than
one source of coverage and can absorb any increase in costs, the
effect on coverage might be minimal.  A 1992 survey showed that only
30 percent of major companies allowed their employees to opt out of
health insurance coverage without at least a sworn statement that the
employees had other coverage.  However, even firms that require that
the employees have coverage may not require their employees'
dependents to have coverage. 


   LOSS OF DEPENDENT COVERAGE
   ACCOUNTS FOR MOST OF THE RECENT
   LOSS IN PRIVATE COVERAGE
------------------------------------------------------------ Letter :5

As employers dropped coverage or raised the cost of coverage for
employees and families, the percentage of people with private health
insurance coverage declined.  In 1989, 75 percent of people under 65
years old had private health insurance; by 1995, this number dropped
to just under 71 percent.  Most of this decline was among
dependents.\14

Changing or losing jobs leads to breaks in coverage, but even when
working steadily on the same job, employees and their families can
lose their insurance. 

Children and working-age adults both lost health insurance coverage
and losses were greatest among lower-income people.  In 1989, 76
percent of working-age adults and almost 74 percent of children under
18 years old had private insurance.  In 1995, almost 73 percent of
working-age adults had private health insurance compared with almost
66 percent of children.  If the same percentage of children and
working-age adults had been privately covered in 1995 as had been
covered in 1989, about 5 million more children and about 5 million
more adults would have had private insurance. 

Between 1989 and 1995, a larger percentage of people under 65 years
old whose incomes were at or below 200 percent of the federal poverty
level lost private insurance than those whose incomes were above the
200 percent level.  For the poorer group, coverage dropped from about
45 percent in 1989 to less than 40 percent in 1995.  Private coverage
for those with incomes above 200 percent of the federal poverty level
also dropped but by less--from about 89 percent to 87 percent. 

Most of the overall decline in private coverage is due to the loss of
coverage for dependents--between 69 and 91 percent, depending on
which years are compared.\15 Both children and adults lost coverage
as dependents.  (See fig.  7.) In addition to the loss of children's
dependent coverage cited above, 24 percent of adults had private
dependent coverage in 1989, which dropped to 21 percent by
1995--almost all of the drop in adult coverage in those years.  In
contrast to the loss of adult dependent coverage, the percentage of
working-age adults as the primary holder of private health insurance
was similar in 1989 and 1995--51.6 percent compared with 51.4
percent. 

   Figure 7:  Percentage of People
   Under 65 Years Old With Private
   Dependent Health Insurance
   Coverage, 1989-95

   (See figure in printed
   edition.)

Source:  GAO analysis of the March 1990, 1992, 1994, and 1996 Current
Population Survey. 

Job change is not the only reason for loss of coverage.  One study
showed that between February 1991 and September 1993, 36 percent of
adults and children who lost insurance for at least 1 month were
dependents of a employee who remained on the same job.\16 Another 25
percent of adults and children experienced breaks in their insurance
coverage when a family member changed jobs or occupations and 21
percent lost insurance at the same time an employed family member
lost his or her job.  Some children lost insurance because they
became too old to be covered under their parent's policy, while other
adults and children lost insurance because of the death of or divorce
from an employed family member. 

Some of the loss of adult dependent coverage probably represents
shifting among adults in their coverage status.  It is likely that
some dual-income families found it less costly to have each earner
covered under his or her own employer's policy--these families may
now have two policyholders.  However, other families might have an
employee who dropped or lost coverage entirely.  For example, women
were and continue to be more likely to be covered as dependents on
others' health insurance policies than men, but the percentage of
women as policyholders has increased.  In 1989, almost 55 percent of
women 18 to 64 years old with private health insurance were
policyholders; by 1994, this number was almost 60 percent. 
Similarly, comparing 1989 with 1995, the percentage of people who
were married without children increased as policy holders, which
helped offset their decrease in dependent coverage. 


--------------------
\14 Employment-based coverage is over 90 percent of all private
coverage.  We discuss trends in private coverage in this section
because changes in the CPS have made data on employment-based
coverage in 1994 incompatible with previous years' data on
employment-based coverage.  (See app.  I.)

\15 Sixty-nine percent of the decline was due to loss of dependent
insurance comparing 1989 with 1993.  Comparing 1989 and 1995,
dependent coverage becomes an even greater percentage of the loss in
private coverage--over 90 percent.  However, changes in the CPS for
March 1995 may have affected our analysis for 1995, so we are
reporting a range of estimates.  (See app.  I.) For a different
analysis of the CPS, which came to a similar conclusion, see John
Sheils, and Lisa Alecxih, Recent Trends in Employer Health Insurance
Coverage and Benefits, prepared by The Lewin Group, Inc., for the
American Hospital Association (Washington, D.C.:  American Hospital
Association, 1996).  These researchers chose to adjust earlier years'
CPS data so that they could compare employment-based coverage in 1994
with earlier years. 

\16 Sheils and Alecxih, Recent Trends in Employer Health Insurance
Coverage and Benefits. 


   REDUCTIONS IN PRIVATE COVERAGE
   MAY SHIFT MORE BURDEN FOR
   HEALTH CARE TO PUBLIC PAYERS
------------------------------------------------------------ Letter :6

Families that do not have individual or employment-based private
health insurance basically have one of two options:  they can remain
uninsured or they can seek health insurance through public
assistance.  Part of the burden to pay for health care for
individuals without private insurance then falls onto taxpayers
through directly subsidized health providers, such as public
hospitals or community clinics, or through publicly subsidized
insurance.  Medicaid, the main public health insurance program for
children and working-age adults, has greatly expanded its enrollment
in recent years.  Evidence is mixed on the extent to which Medicaid
expansion served to dampen the effects of deteriorating private
coverage or exacerbated losses in private coverage by encouraging
some low-income people to drop private insurance. 

Welfare reform efforts may decrease Medicaid enrollment and increase
the percentage of uninsured Americans.  Eligibility rules have
changed for some groups, and states will be moving welfare recipients
into the workforce.  However, low-income adults moving into the
workforce may not gain access to private insurance, while losing
Medicaid coverage. 


      BEING UNINSURED HAS SERIOUS
      HEALTH AND FINANCIAL
      CONSEQUENCES
---------------------------------------------------------- Letter :6.1

Being uninsured can have serious health and financial consequences. 
According to a recent survey funded by the Kaiser Family Foundation,
45 percent of uninsured adults had problems getting health care and
most reported having serious financial and health consequences as a
result.  Thirty-six percent of uninsured adults reported having
trouble paying health care bills.\17 Moreover, people without health
insurance tend to forego health care more than those with health
insurance.  Therefore, when the uninsured seek care, their condition
often is more advanced and, thus, more expensive to treat.\18
Compared with all adults, uninsured adults who had trouble getting
care or paying their health care bills were more likely to be in fair
or poor health (34 percent compared with 19 percent), to be disabled
(38 percent compared with 14 percent), or to have been hospitalized
in the previous year (21 percent compared with 12 percent).\19

People without health insurance are more likely to seek care in
public clinics and hospital emergency rooms--increasing the burden on
these facilities.  Covering the expenses of treating the uninsured
has become increasingly difficult for hospitals.  Due to more
aggressive contracting by insurers and managed care companies, these
payers are less likely to pay full hospital charges.  Public payments
to hospitals through Medicaid and Medicare have helped hospitals
cover the cost of caring for the uninsured, although Medicare and
Medicaid traditionally paid less than full charges for hospital
costs.  But in the past, private payers helped to subsidize the
difference.  Increased use of managed care in the public sector in
some areas of the country may also be shifting patients away from the
hospitals that primarily serve the uninsured. 


--------------------
\17 Twenty-eight percent had problems with getting care and paying
bills, 17 percent only had problems getting care, 8 percent only had
problems paying bills, and 47 percent reported neither problem. 

\18 See David U.  Himmelstein, and Steffie Woolhandler, "Care Denied: 
U.S.  Residents Who Are Unable to Obtain Needed Medical Services."

\19 See Donelan, and others, "Whatever Happened to the Health
Insurance Crisis in the United States?  Voices from a National
Survey."


      AS PRIVATE COVERAGE ERODED,
      MEDICAID EXPANDED--BUT SO
      DID THE UNINSURED
---------------------------------------------------------- Letter :6.2

While the percentage of people with private insurance declined, the
percentage of people with Medicaid coverage increased.  In 1987,
about 18 million people under 65 years old had public insurance
through Medicaid; by 1995, this number escalated to almost 29
million.  More families sought assistance through AFDC and
Supplemental Security Income (SSI)--which entitled them to Medicaid
coverage as well--and Medicaid eligibility was expanded to include
pregnant women and children.  In addition, several states, through
federal waivers of Social Security law, expanded Medicaid coverage to
low-income populations not previously eligible.  While expanding
Medicaid helped to stabilize the percentage of insured people in
families with incomes below 200 percent of the federal poverty level
between 1989 and 1994, a greater percentage of people with family
incomes above 200 percent of the federal poverty level were uninsured
in 1994. 

According to two studies, expanded government insurance programs may
have encouraged some families to drop private health insurance
coverage in favor of Medicaid.\20 The extent to which this happens is
unclear, however, because three other studies found no effect.\21
Cutler and Gruber estimate that between 1987 and 1992, 17 percent of
the decrease in employment-based insurance was due to Medicaid
expansions for pregnant women and children.  They attribute the other
83 percent to changes in employer behavior unrelated to Medicaid
generosity, changes in the demographic mix of the population, and
economic conditions at the time.  Dubay and Kenney, using somewhat
different assumptions, estimate that 12 to 18 percent of children's
increase in Medicaid coverage was linked to a reduction in
employment-based health insurance coverage.  They also state that
children above federal poverty levels were more likely to have
displaced private insurance with Medicaid than poor children. 

Unless the decline in private insurance coverage abates, public
payers may be facing increased costs for health care--either for
uncompensated care or public insurance.  If employment-based
insurance continues to decline, the number of people who are
uninsured will likely increase.  The Lewin Group, Inc., has estimated
that the number of uninsured Americans will increase from about 40.0
million Americans in 1995 to 45.6 million Americans in 2002. 


--------------------
\20 See David M.  Cutler, and Jonathan Gruber, Does Public Insurance
Crowd Out Private Insurance?  National Bureau of Economic Research,
working paper No.  5082 (Cambridge, Mass.:  1995), and Lisa Dubay,
and Genevieve Kenney, Revisiting the Issues:  The Effects of Medicaid
Expansions on Insurance Coverage of Children (Washington, D.C.:  The
Urban Institute, 1995). 

\21 See Lara D.  Shore-Sheppard, "Stemming the Tide?  The Effect of
Expanding Medicaid Eligibility on Health Insurance Coverage"
(unpublished draft, Nov.  1995), Lara D.  Shore-Sheppard, "The Effect
of Expanding Medicaid Eligibility on the Distribution of Children's
Health Insurance Coverage" (unpublished draft presented at the
Cornell/Princeton conference on Reforming Social Insurance Programs,
May 1996), and Esel Y.  Yazici, "Medicaid Expansions and the Crowding
Out of Private Health Insurance" (paper presented at the 18th Annual
Research Conference of the Association for Public Policy Analysis and
Management, Pittsburgh, Penn., Nov.  2, 1996). 


      LOW-INCOME FAMILIES MAY LOSE
      COVERAGE THROUGH WELFARE
      REFORM
---------------------------------------------------------- Letter :6.3

Whether Medicaid will continue to expand its enrollment and help hold
down growth in the number of uninsured people over the next few years
is unclear.  Under welfare reform, eligibility rules have changed in
ways that could affect Medicaid coverage for low-income children and
adults. 

Before reform, over 60 percent of the children and adults receiving
assistance through Medicaid were automatically enrolled under AFDC or
SSI.  Under the new welfare program--Temporary Assistance for Needy
Families (TANF)--families receiving cash assistance will not be
automatically enrolled in Medicaid unless a state chooses to do this. 
Past studies of the Medicaid population have shown that automatic
enrollment in Medicaid for AFDC recipients led to higher coverage
levels than among people who had to apply separately for Medicaid. 
If states develop separate application processes for TANF and
Medicaid, such processes may raise barriers to Medicaid enrollment. 

Generally, families who would have qualified for AFDC-Medicaid
coverage will still qualify for Medicaid.  The law extends
eligibility to people who would have been eligible under AFDC rules
in effect as of July 16, 1996.  However, states can roll back the
July 16, 1996, AFDC income standards to May 1, 1988, levels.  States
also can raise these levels for inflation and can use more liberal
methodologies to determine countable income and resources.  The new
law allows states to terminate Medicaid eligibility for any adult who
is terminated from TANF because of failure to work, but their minor
children cannot be terminated from Medicaid on that basis.  The law
also extends Medicaid for up to a year for those who either become
employed or have increased earnings and received Medicaid under the
prewelfare reform AFDC eligibility criteria in 3 of the preceding 6
months. 

There are no changes in current eligibility rules for pregnant women
and children based on age and income.  However, SSI eligibility for
children with disabilities is tighter under the new law, which could
reduce Medicaid enrollment of SSI children.  Some children who lose
eligibility based on SSI or because their families exhaust
transitional Medicaid may be able to gain eligibility based on age
and income. 

The new law limits eligibility in a significant way--based on
citizenship status.  Before the new law, all legal immigrants and
permanent residents who qualified based on other eligibility criteria
were eligible for Medicaid.  Under the new law, qualified aliens
currently residing in the United States will be eligible for Medicaid
only at the state's option, unless a qualified alien is a member of
one of the excepted groups whose Medicaid eligibility is mandated by
law.\22 Qualified aliens who enter the country in the future will be
banned from Medicaid coverage for 5 years from their date of entry,
except for treatment of emergency medical conditions. 


--------------------
\22 A qualified alien is a person lawfully admitted for permanent
residence, asylees, refugees, persons paroled into the United States
for at least 1 year, persons whose deportation has been withheld, and
persons granted conditional entry.  The excepted groups of qualified
aliens are legal permanent residents with 40 qualifying quarters of
work; and for 5 years from entry to the U.S.--refugees, asylees, and
persons whose deportation has been withheld; and certain veterans and
active duty military and their families. 


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

As the cost of health insurance escalated, many employers
restructured their benefits.  Some employers dropped health insurance
coverage entirely--particularly small employers--shifting the burden
entirely to employees.  But more commonly, employers increased the
amount employees had to pay to gain coverage, particularly for family
coverage. 

As this occurred, coverage became less available and less affordable
for many Americans.  The percentage of Americans under 65 years old
with private health insurance decreased.  Thus, many Americans who
are unable to purchase health insurance for themselves and their
families have trouble getting health care.  In particular, some
children and working adults who earn low wages are being squeezed out
of the private insurance market.  At the same time, many of these
Americans are not eligible for public medical assistance.  This slow
erosion of private coverage contributed to a loss of
coverage--leaving more than 40 million Americans under 65 years old
uninsured. 

Public pressure to increase publicly funded care may intensify if the
number of Americans who lose private insurance coverage continues to
rise.  However, state and federal efforts to reform welfare may
decrease the number of people covered through Medicaid.  In addition,
policymakers have become concerned that increasing public coverage
will encourage employers or families to drop private coverage in
favor of public coverage.  If the availability of both public and
private coverage continues to erode, the number of uninsured will
inevitably continue to grow. 

Through welfare reform, states will be trying to move families off
assistance and into the private sector.  Ideally, as welfare
recipients begin working, they will gain access to private insurance. 
However, former welfare recipients tend to land low-wage jobs, which
often do not offer health insurance coverage or may not offer
insurance which they can afford.  Even after a year in the workforce,
many former welfare recipients may still not be able to access or
afford private coverage for their families.  This suggests that
low-income working families may continue to need subsidized health
insurance if they are to have health insurance coverage. 


   AGENCY COMMENTS
------------------------------------------------------------ Letter :8

We sought comments on a draft of this report from experts on private
health insurance issues and from the Health Care Financing
Administration on the section of the draft report that dealt with
Medicaid and welfare reform.  The reviewers generally agreed with our
report, but provided technical suggestions that we included where
appropriate. 


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

As agreed with your office, we plan no further distribution of this
report for 30 days.  At that time, we will make copies available on
request.  Please contact me at (202) 512-7114 or Michael Gutowski at
(202) 512-7128 if you or your staff have any further questions.  This
report was prepared by Michael Gutowski, Sheila Avruch, Paula Bonin,
and Karen Sloan. 

Sincerely yours,

Jonathan Ratner
Associate Director, Health Financing
 and Systems Issues


INFORMATION SOURCES AND
METHODOLOGY
=========================================================== Appendix I

This report used several sources of information, including the Bureau
of the Census' Current Population Survey.  This appendix discusses
the sources of information and selected information on the CPS and
how we analyzed it. 


   SOURCES OF INFORMATION FOR THIS
   REPORT
--------------------------------------------------------- Appendix I:1

Several sources of information can be used to track trends in health
insurance coverage.  Each of the sources we used provides different
information in different ways for different years.  In general, less
information is available on small businesses than on large and major
firms.  We define large firms as those with at least 100 employees
and major firms as those with over 1,000 employees.  Therefore, we
focused on large and major firms.  Consequently, we may have
overstated support for employer coverage of health insurance because
larger firms are more likely to provide coverage to employees and
families than smaller firms. 

We used the following sources of information: 

DOL's Bureau of Labor Statistics.  The Bureau's surveys of employee
benefits provide representative data for 34 million employees in
medium and large private establishments (places of work that employ
100 or more people) in 1988, 1989, 1991, and 1993 and for 49 million
employees in small private establishments (places of work that employ
fewer than 100 people) in 1990, 1992, and 1994.  The Bureau surveys
establishments of different size in alternate years.  We reported
information from the surveys on medium and large establishments
(which fit our definition of large firms) and state and local
governments. 

The Health Insurance Association of America.  The Association began a
survey in 1987 that was continued for several years. 

KPMG Peat Marwick.  Since 1991, KPMG Peat Marwick has conducted a
nationwide telephone survey of about 1,000 randomly selected private
and public employers with 200 or more employees on the health
benefits they provide.  The survey instrument and sample design is
similar to the Health Insurance Association of America surveys and,
thus, can be compared. 

Hewitt Associates.  This company collects information on benefits
provided by major U.S.  employers (many with over 5,000 employees)
and has published reports on major companies' benefits. 

The Robert Wood Johnson Foundation.  The Foundation studied health
insurance coverage in 10 states by surveying over 22,000
establishments. 

The Bureau of the Census' CPS.  The CPS is a nationally
representative survey that is the official source of government
statistics on employment and unemployment.  Every March the Bureau
collects additional information on health insurance coverage.  We
used the CPS to measure private insurance coverage and private
dependent insurance coverage in 1989, 1991, 1993, and 1995.  Because
of certain methodological changes implemented in March 1995 and
continued in March 1996 (which affected the 1994 and 1995 data),
including changes in the questionnaire, we considered it more
appropriate to compare private insurance coverage, rather than make
such comparisons for employment-based insurance coverage in 1995 to
earlier years. 

We did not independently verify data from these sources.  The private
surveys are proprietary, and DOL and the Bureau of the Census conduct
their own validity and reliability checks of their data.  We checked
some of our CPS analyses against published Census data and have
consulted with the Bureau of the Census to ensure accurate analyses
of its data files. 


   METHODOLOGICAL ISSUES IN THE
   CPS ANALYSIS
--------------------------------------------------------- Appendix I:2

For the March 1995 CPS, the Bureau of the Census implemented a number
of changes in an effort to improve the accuracy and ease of
administering the survey.  These changes include moving to a
computer-assisted telephone interviewing system and reordering and
rewording survey questions on health insurance.  The earlier
questionnaire asked people (1) if they had private insurance, (2) if
they were the policyholder, (3) if the insurance was obtained through
their employment, and (4) who else was covered as a dependent.  The
new questionnaire asks (1) if they have private insurance through an
employer or union; (2) who is the policyholder; (3) who else is
covered; and (4) if they have purchased individual health insurance
and, if so, additional people who are covered by this policy. 

These changes appear to affect how people answer questions about
their insurance coverage, which can affect estimates of insurance
coverage.  These changes also can affect the comparability of 1994
and subsequent years' estimates with previous years' estimates.  When
the 1994 data were released, officials at Census stated that the 1994
estimate of overall private insurance agreed well with previous
years' estimates.  However, the number of people who report that
their private insurance coverage comes from their employer or union
increased, while the number who reported that their private insurance
was individually purchased decreased.  Therefore, because these
apparent differences may be due to the questionnaire change--rather
than actual changes in the composition of private health insurance
coverage--comparisons of employment-based insurance coverage in 1994
and 1995 with previous years may not be appropriate to understand
trends in coverage.  In particular, dependents appeared to be
misclassified as not having employment-based insurance in past
surveys.  This is why our CPS analysis compares private insurance,
rather than employment-based insurance. 

Because we were concerned that changes in the questionnaire would
affect estimates of private dependent health insurance coverage, we
also partially analyzed the March 1994 CPS (1993 data), to compare
dependent coverage in 1993.  However, the change in the questionnaire
did not appear to affect the estimates of dependent coverage. 

Our work was conducted between February and November 1996 in
accordance with generally accepted government auditing standards. 


AVERAGE MONTHLY HEALTH INSURANCE
PREMIUMS FOR EMPLOYER-SPONSORED,
EMPLOYEE-ONLY, AND FAMILY
COVERAGE, 1989-96
========================================================== Appendix II

                                                                          Percen
                                                                               t
                                                                          increa
                                                                              se
                                                                           1989-
Plan type         1989\a    1991    1992    1993    1994    1995    1996      96
----------------  ------  ------  ------  ------  ------  ------  ------  ------
Conventional
--------------------------------------------------------------------------------
Employee-only       $119    $145    $154    $170    $181    $166    $174      46
Family               268     355     384     441     463     433     449      68

HMO
--------------------------------------------------------------------------------
Employee-only        116     139     148     158     166     160     157      35
Family               267     350     377     422     450     423     423      58

Preferred provider organization
--------------------------------------------------------------------------------
Employee-only        119     150     157     181     177     174     181      52
Family               271     376     412     454     453     433     448      65
--------------------------------------------------------------------------------
\a The 1989 survey included smaller firms.  Smaller firms generally
had higher premium costs.  We used the average cost by type of
insurance and, for HMOs, calculated a weighted average by type of
HMO.  This made baseline insurance costs in 1989 a little higher for
indemnity plans and HMOs and lower for PPOs than if we had only
reported costs for firms with 1,000 or more employees. 

Source:  Health Insurance Association of America and KPMG Peat
Marwick.  The Health Insurance Association of America survey was of
about 2,600 public and private employers with at least 2 employees
(for 1989) and the KPMG Peat Marwick surveys were of about 1,000
randomly selected public and private employers with 200 or more
employees (all subsequent years). 



RELATED GAO PRODUCTS
=========================================================== Appendix 0

Private Health Insurance:  Millions Relying on Individual Market Face
Cost and Coverage Trade-Offs (GAO/HEHS-97-8, Nov.  25, 1996). 

Medicaid and Uninsured Children, 1994 (GAO/HEHS-96-174R, July 9,
1996). 

Health Insurance for Children:  Private Insurance Coverage Continues
to Deteriorate (GAO/HEHS-96-129, June 17, 1996). 

Medicaid:  Spending Pressures Spur States Toward Program
Restructuring (Testimony, GAO/T-HEHS-96-75, Jan.  18, 1996). 

Health Insurance for Children:  State and Private Programs Create New
Strategies to Insure Children (GAO/HEHS-96-35, Jan.  18, 1996). 

Medicaid and Children's Insurance (GAO/HEHS-96-50R, Oct.  20, 1995). 

Health Insurance for Children:  Many Remain Uninsured Despite
Medicaid Expansion (GAO/HEHS-95-175, July 19, 1995). 

Medicaid:  Spending Pressures Drive States Toward Program Reinvention
(GAO/HEHS-95-122, Apr.  4, 1995). 

Medicaid:  Restructuring Approaches Leave Many Questions
(GAO/HEHS-95-103, Apr.  4, 1995). 

Medicaid:  Experience With State Waivers to Promote Cost Control and
Access Care (GAO/HEHS-95-115, Mar.  23, 1995). 

Uninsured and Children on Medicaid (GAO/HEHS-95-83R, Feb.  14, 1995). 

Employer-Based Health Insurance:  High Costs, Wide Variation Threaten
System (GAO/HRD-92-125, Sept.  22, 1992). 

Access to Health Insurance:  State Efforts to Assist Small Businesses
(GAO/HRD-92-90, May 14, 1992). 


*** End of document. ***