Private Health Insurance: Impact of Premium Increases on Number of
Covered Individuals is Uncertain (Testimony, 06/11/99,
GAO/T-HEHS-99-147).

Pursuant to a congressional request, GAO discussed the impact of private
health insurance premium increases on the number of covered individuals,
focusing on: (1) the trends in employers' decisions to offer insurance
and employees' decisions to purchase it; (2) an assessment of recent
studies that have estimated the relationship between premium increases
and insurance coverage; and (3) conditions or factors that could affect
the impact of premium increases on insurance coverage.

GAO noted that: (1) from the late 1980s to the mid-1990s--a period of
rising health insurance premiums--the proportion of employees offered
coverage rose from about 72 percent to 75 percent, while the share that
accepted insurance fell from approximately 88 percent to 80 percent; (2)
the extent to which various factors contributed to the fall in the
acceptance rate is unclear; (3) it may have resulted from employees
being asked to pay a larger share of the premiums or other factors, such
as decreases in some workers' real income; (4) Medicaid-eligibility
expansions and changes in benefit levels also may have contributed to
the fall in the acceptance rate; (5) few studies have attempted to
estimate the effects of premium increases on insurance coverage, and no
study adequately estimates the coverage loss that might result from new
legislative mandates; (6) studies by the Lewin Group, for example,
suggest that 300,000 to 400,000 individuals might drop or lose insurance
coverage if premiums increased 1 percent; (7) however, these estimates
assume across-the-board premium increases; (8) the potential coverage
loss might be much lower if mandates primarily affect health maintenance
organization premiums and employers and employees can switch to
different types of coverage; (9) furthermore, serious data limitations
affect the precision of many of these studies' estimates; (10) many
factors can affect the impact that health insurance mandates have on the
number of individuals covered by private insurance; (11) for example, if
new mandates result in changes that individuals consider worthwhile,
they may be willing to pay higher premiums; and (12) the extent to which
employers pass on premium increases to employees, employees'
opportunities to switch to less expensive plans, and changes in economic
factors--such as income, or changes in public insurance program
eligibility requirements--can also affect the number of individuals with
private health insurance.

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

 REPORTNUM:  T-HEHS-99-147
     TITLE:  Private Health Insurance: Impact of Premium Increases on
	     Number of Covered Individuals is Uncertain
      DATE:  06/11/99
   SUBJECT:  Health insurance
	     Health insurance cost control
	     Health maintenance organizations
	     Projections
	     Insurance premiums
	     Employee medical benefits
	     Surveys
IDENTIFIER:  Civilian Health and Medical Program of the Uniformed
	     Services
	     Census Bureau Current Population Survey
	     Medicare Program
	     Medicaid Program
	     CHAMPUS

******************************************************************
** 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

Before the Subcommittee on Employer-Employee Relations, Committee on
Education and the Workforce, House of Representatives

For Release on Delivery
Expected at 9:30 a.m.
Friday, June 11, 1999

PRIVATE HEALTH INSURANCE - IMPACT
OF PREMIUM INCREASES ON NUMBER OF
COVERED INDIVIDUALS IS UNCERTAIN

Statement of William J.  Scanlon, Director
Health Financing and Public Health Issues
Health, Education, and Human Services Division

GAO/T-HEHS-99-147

GAO/HEHS-99-147T

(101859)

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

  CPS - Test
  HMO - Test
  PPO - Test

PRIVATE HEALTH INSURANCE:  IMPACT
OF PREMIUM INCREASES ON NUMBER OF
COVERED INDIVIDUALS IS UNCERTAIN
============================================================ Chapter 0

Mr.  Chairman and Members of the Subcommittee: 

Approximately 150 million individuals obtained health insurance
through the work place in 1996, either through their own employer or
the employer of a family member.  During the last several years, an
increasing number of these individuals have enrolled in some form of
managed care rather than in fee-for-service plans.  Recently,
concerns have grown regarding the ways in which some managed care
plans operate and the adequacy of the information that plans and
their providers share with members. 

In response to these concerns, several legislative proposals have
been made to require health insurance plans to adopt specified
operational practices.  The proposals apply to all types of plans but
would likely have their greatest impact on health maintenance
organizations (HMO).  Other types of plans--such as preferred
provider organizations (PPO) and indemnity, or fee-for-service,
plans--will likely be affected to a lesser degree.  Included in the
various proposals are requirements, for example, to disclose certain
information,\1 guarantee patient access to emergency and specialty
services, implement internal and external grievance policies, and
guarantee freedom of communication between providers and patients. 
Some lawmakers are concerned, however, that these types of mandates
could increase the cost of health insurance and have the unintended
consequence of reducing the number of individuals covered by private
health insurance. 

To help inform congressional consideration of these proposals, you
asked us to present the findings of a study we did for Senator
Jeffords last July that analyzed the relationship between private
insurance premium increases and changes in the number of covered
lives.\2 My remarks today are based on that study.  Specifically, I
will focus on (1) the trends in employers' decisions to offer
insurance and employees' decisions to purchase it, (2) an assessment
of recent studies that have estimated the relationship between
premium increases and insurance coverage, and (3) conditions or
factors that could affect the impact of premium increases on
insurance coverage. 

In summary, from the late 1980s to the mid-1990s--a period of rising
health insurance premiums--the proportion of employees offered
coverage rose from about 72 percent to 75 percent, while the share
that accepted insurance fell from approximately 88 percent to 80
percent.  The extent to which various factors contributed to the fall
in the acceptance rate is unclear.  It may have resulted from
employees being asked to pay a larger share of the premiums or other
factors, such as decreases in some workers' real income. 
Medicaid-eligibility expansions and changes in benefit levels also
may have contributed to the fall in the acceptance rate. 

Few studies have attempted to estimate the effects of premium
increases on insurance coverage, and no study adequately estimates
the coverage loss that might result from new legislative mandates. 
Studies by the Lewin Group, for example, suggest that 300,000 to
400,000 individuals might drop or lose insurance coverage if premiums
increased 1 percent.  However, these estimates assume
across-the-board premium increases.  The potential coverage loss
might be much lower if mandates primarily affect HMO premiums and
employers and employees can switch to different types of coverage. 
Furthermore, serious data limitations affect the precision of many of
these studies' estimates. 

Finally, many factors can affect the impact that health insurance
mandates have on the number of individuals covered by private
insurance.  For example, if new mandates result in changes that
individuals consider worthwhile, they may be willing to pay higher
premiums.  The extent to which employers pass on premium increases to
employees, employees' opportunities to switch to less expensive
plans, and changes in economic factors--such as income or changes in
public insurance program eligibility requirements--can also affect
the number of individuals with private health insurance. 

--------------------
\1 Legislative proposals would require each plan to disclose, for
example, information on appeals procedures, restrictions on
reimbursement for care received outside of the plan's network of
providers, and the location of plan providers and facilities. 

\2 Private Health Insurance:  Impact of Premium Increases on the
Number of Covered Individuals Is Uncertain (GAO/HEHS-98-203R, July 7,
1998). 

   BACKGROUND
---------------------------------------------------------- Chapter 0:1

Between 1995 and 1997, real health insurance premiums (adjusted for
inflation) remained nearly constant or fell slightly across all plan
types.  (See table 1.) This represents a sharp decline from the
previous 5 years--in 1990, inflation-adjusted growth was as high as
11.6 percent for indemnity plans and 10.6 percent for HMOs.  In 1998,
premiums increased for all insurance types, but the increase was much
lower than was experienced in the early 1990s. 

                          Table 1
          
            Percentage of Real Annual Growth in
           Premiums, by Type of Health Plan, 1991
                          to 1998

Plan type   1991  1992  1993  1994  1995  1996  1997  1998
----------  ----  ----  ----  ----  ----  ----  ----  ----
Indemnity    7.8   8.0   5.5   2.5     -     -   0.3   1.9
                                     0.1   1.8
PPO          5.9   7.6   5.2   0.6   0.7     -     -   2.3
                                           2.4   0.2
HMO          7.9   6.8   5.3   2.7     -     -     -   1.3
                                     2.4   3.4   0.3
----------------------------------------------------------
Sources:  GAO calculations based on data from KPMG Peat Marwick
(1991-98) and Bureau of Labor Statistics' Consumer Price Index. 
Includes employer and employee shares of premiums for workers in
private firms with at least 200 employees. 

In 1996, about 70 percent of the population under age 65 was covered
by health insurance purchased through an employer or union or
purchased privately as an individual, according to Current Population
Survey (CPS) data.  About 12 percent was covered by Medicare,
Medicaid, or the Civilian Health and Medical Program of the Uniformed
Services, and about 18 percent was uninsured.  From 1989 to 1996, the
percentage of the population covered by employer-sponsored,
union-sponsored, or individual insurance\3 decreased slightly, but
these options still remained a prominent source of coverage for
people under age 65.  (See fig.  1.) During the same period, the
proportion of the population covered by Medicaid and the proportion
without insurance both increased. 

   Figure 1:  Sources of Health
   Insurance for People Under Age
   65, 1989 and 1996

   (See figure in printed
   edition.)

Sources:  U.S.  Bureau of the Census, CPS (March 1989 to March 1997). 

--------------------
\3 Individual insurance is coverage that an individual purchases
directly from an insurer or through a broker. 

   MORE WORKERS WERE OFFERED
   INSURANCE, BUT FEWER ACCEPTED
   COVERAGE AS PREMIUMS INCREASED
---------------------------------------------------------- Chapter 0:2

Recent studies suggest that employers typically do not stop offering
health insurance when premiums increase.  Between 1988 and 1996,
health insurance premiums--unadjusted for inflation--increased by
about 8 percent per year on average.  During approximately the same
time period, one study found that the fraction of workers offered
insurance by their employers grew slightly, from 72.4 percent to 75.4
percent.\4

The proportion of workers who had access to employer-sponsored
insurance, either through their own employer or the employer of a
family member, remained essentially constant at about 82 percent. 
Another study reported that the fraction of small firms (those with
fewer than 200 employees) offering insurance coverage grew from 46
percent in 1989 to 49 percent in 1996.\5 The study also found that 99
percent of large firms offered insurance in 1996. 

Fewer workers, however, are choosing to accept employer-sponsored
coverage for themselves or their dependents.  In 1987, 88.3 percent
of workers accepted coverage when their employers offered it.  In
1996, only 80.1 percent of workers accepted coverage.  The fall in
the acceptance rate was relatively large for workers under age 25
(from 86.5 percent to 70.1 percent) and those making $7 per hour or
less (from 79.7 percent to 63.2 percent).  The fraction of workers
who accepted employer-sponsored insurance either through their own
job or that of a family member also declined, from 93.2 percent to
89.1 percent.  Consequently, even though a greater percentage of
employers offered insurance, a smaller proportion of workers was
covered by employer-sponsored insurance in 1996 compared with 1987. 

The fall in the acceptance rate may be attributable partly to
required increases in employees' insurance premium contributions. 
One study found that employees in small firms paid an average of 12
percent of single coverage premiums in 1988 and employees in large
firms paid 13 percent.\6 In 1996, the employee share had risen to 33
percent in small firms and 22 percent in large firms.  According to
the Lewin Group, the combined effect of the increase in premiums and
the increase in the employees' share of those premiums resulted in
workers paying 189 percent more in real terms for single coverage and
85 percent more in real terms for family coverage in 1996 compared
with 1988. 

Other factors also may have contributed to the drop in the acceptance
rate.  A decline in real wages for some workers may have made
coverage less affordable.  Expansions in Medicaid eligibility
provided a coverage alternative for some families and may have
decreased workers' willingness to accept employer-sponsored
insurance.  Furthermore, possible changes in benefit packages may
have made coverage less desirable. 

--------------------
\4 P.  Cooper and B.  Schone, More Offers, Fewer Takers for
Employment-Based Health Insurance:  1987 and 1996, Health Affairs
16(6) (Nov./Dec.  1997). 

\5 P.  Ginsburg and others, Tracking Small-Firm Coverage,
1989-1996, Health Affairs 17(1) (Jan./Feb.  1998). 

\6 J.  Gabel and others, Small Employers and Their Health Benefits,
1988-1996:  An Awkward Adolescence, Health Affairs 16(5) (Sept./Oct. 
1997). 

   STUDY DESIGNS LIMIT THE ABILITY
   TO PREDICT POTENTIAL COVERAGE
   LOSS FROM NEW HEALTH INSURANCE
   REQUIREMENTS
---------------------------------------------------------- Chapter 0:3

Relatively few studies have analyzed the relationship between an
increase in the cost of insurance and the change in the number of
individuals covered.  Several studies have examined the extent to
which insurance premium subsidies might affect employers' decisions
to offer insurance, but these results do not directly address the
question of how much coverage loss might arise from an increase in
premiums.  The relevance of two studies that attempted to answer this
question is limited because of implicit assumptions embedded in the
studies' designs and shortcomings in the available data. 

In November 1997, the Lewin Group estimated that 400,000 fewer people
might be covered by health insurance if new legislation caused
premiums to rise by 1 percent.  Its estimate was largely based on
studies of the effects of insurance premium subsidies on employers'
decisions to offer insurance.  These studies focused primarily on
small employers and varied widely both in their research questions
and their findings.  The Lewin Group selected a midpoint estimate
from a range of estimates it judged to be the best available.  It
then adjusted the estimate to account for the likelihood that
individuals might obtain insurance through working family members'
policies, the individual insurance market, or public insurance
programs if premiums rose on their employer-sponsored policies. 
However, the Lewin Group's estimate of potential coverage loss did
not consider the possibility that employers or employees might switch
to different types of insurance products if one type became
relatively more expensive.  Because many of the proposed federal
mandates are expected primarily to affect HMOs and have little or no
impact on PPOs and indemnity plans, Lewin's estimate may overstate
the potential coverage reduction. 

To correct for some shortcomings of its earlier study, the Lewin
Group performed its own data analysis and released the findings in
January 1998.  The results indicated a lower potential coverage loss
of 300,000 individuals for every 1 percent increase in premiums.\7
The findings of the 1998 Lewin Group's analysis, however, may have
been affected by data limitations.  Specifically, the analysis rests
on an accurate measure of health insurance premiums paid by
employees.  Because this information was unavailable, the Lewin Group
had to impute this amount.  In addition, two aspects of the study's
design limit its ability to predict insurance coverage reductions
that might result from new legislative mandates.  First, the coverage
loss estimate--just as in the first study--applies to situations
where all premiums increase by 1 percent.  A 1-percent increase in
HMO premiums would likely result in a smaller coverage reduction if
employers and employees switched to other types of health coverage. 
Second, the Lewin Group explicitly assumed that all observed coverage
changes resulted from employees' decisions to not accept coverage.\8
This assumption is broadly supported by findings from other studies. 
However, to the extent that some employers decided to no longer offer
insurance, the Lewin Group's estimate incorrectly predicts employees'
reactions to changes in premiums. 

--------------------
\7 The new estimate was based on the Lewin Group's statistical
analysis of the relationship between what employees paid for
insurance and the probability that they, their spouses, and their
dependent children would have employer-sponsored health insurance. 
Lewin used complex statistical models to estimate the proportion of
the population covered by employer-sponsored insurance grouped by a
number of demographic characteristics, including race, age, income,
full-time/part-time status, occupation, industry, firm size, and the
imputed employee share of the premium costs. 

\8 The data used in the Lewin study do not indicate whether observed
coverage losses are the result of employers' decisions not to offer
insurance or employees' decisions not to accept it. 

   MULTIPLE FACTORS AFFECT
   POTENTIAL IMPACT OF PREMIUM
   INCREASES ON NUMBER OF COVERED
   INDIVIDUALS
---------------------------------------------------------- Chapter 0:4

Insufficient information is currently available to accurately predict
the reduction in the number of individuals covered by private
insurance (referred to as coverage changes) that may result from
health insurance premium increases associated with new federal
mandates.  One problem is that estimates of the effects of mandates
on premiums have some uncertainty.  However, even if the premium
increase was known with certainty, previous research and economic
theory suggest that the impact on coverage depends on a number of
conditions.  Coverage changes will depend on the extent to which
premiums rise for employees and whether they can switch to insurance
plans less affected by the mandates.  The specific policy adopted
also can affect how employees respond to resulting premium increases. 
Finally, changes in many economic and other factors can cause
coverage changes that mask or exaggerate the impact of premium
increases.  The following list describes several conditions that
could affect observed changes in health insurance coverage if new
federal mandates increase insurance costs. 

  -- The percentage of premiums paid by employees and the amount of
     any premium increase that employers pass on to employees.  If,
     as recent evidence suggests, employees' decisions largely affect
     the extent of coverage, then the relevant price increase is the
     percentage increase in their contribution.  For example, about
     two-thirds of employees in small firms had to contribute toward
     premium costs in 1996.  Those employees paid about 50 percent of
     the total premium.  If total premiums rise by 1 percent and
     employers pass on the full increase to employees, then the
     employees' contribution would rise by 2 percent. 

  -- The extent to which additional benefits are valued by consumers. 
     If higher insurance premiums are the result of additional
     benefits that consumers value, then any coverage loss will be
     less than the coverage loss that might occur if premiums
     increased but benefits stayed the same (or the additional
     benefits had little consumer value).  In a November 1997 letter,
     the Lewin Group noted that its estimates of the number of
     persons losing coverage will differ depending upon the health
     policy being analyzed and pointed out that some proposals that
     increase premium costs are often associated with other
     provisions that may either lessen or intensify incentives for
     individuals to drop coverage.

  -- The extent to which employees can switch plans that have no or
     low premium increases.  Proposed new federal mandates are
     expected primarily to increase costs for HMOs.  Faced with a
     rise in HMO premiums, some employees may switch to PPOs or
     indemnity insurance rather than drop coverage entirely. 

  -- Changes in other insurance benefits.  Instead of raising
     premiums in response to new mandated benefits, insurance
     companies and employers may find ways to reduce other parts of
     the insurance package to keep premiums constant.  It is unknown
     how employees might respond to such changes in their insurance
     plans. 

  -- Changes in real wages and other factors.  Changes in economic
     conditions or eligibility for public insurance programs can also
     affect private insurance coverage.  For example, the Lewin Group
     estimated that a 1-percent rise in real income could increase
     private insurance coverage by nearly 0.37 percent (about 550,000
     workers and dependents).  Likewise, expansions in Medicaid
     eligibility could cause some workers to substitute public
     insurance for employer-sponsored family coverage. 

   CONCLUSIONS
---------------------------------------------------------- Chapter 0:5

The extent to which new legislative requirements for health insurance
providers could affect the number of individuals with insurance
coverage depends on the answers to several questions.  To what extent
will insurers raise premiums?  Will fewer employers offer coverage to
their employees, or will employers pass some or all of the increased
premium costs onto their employees?  How many employees will decline
offered coverage if they must pay higher premiums? 

The available studies offer only limited insights into these issues
and illustrate the difficulty of estimating how the number of
individuals covered by health insurance might be affected.  Many of
the studies we reviewed were hampered by incomplete data.  Moreover,
the design of the studies and the assumptions they incorporated limit
their applicability to the current issue.  Studies by the Lewin
Group, for example, estimate coverage loss that might result from an
across-the-board premium increase.  Legislation that affected some
types of insurance providers' costs more than others might have a
much smaller impact if beneficiaries can switch from plans with
larger premium increases to plans with smaller premium increases. 
Finally, a host of other factors--including, for example, the extent
to which individuals value the results of the specific mandates and
general economic conditions--will likely play a role in determining
the impact that legislative mandates have on the number of insured
individuals. 

-------------------------------------------------------- Chapter 0:5.1

Mr.  Chairman, this concludes my prepared statement.  I will be happy
to answer any questions. 

   GAO CONTACT AND ACKNOWLEDGMENTS
---------------------------------------------------------- Chapter 0:6

For future contacts regarding this testimony, please call William J. 
Scanlon at (202) 512-7114.  Key contributors to this testimony
include James C.  Cosgrove and Susanne M.  Seagrave. 

*** End of document. ***