Private Health Insurance: Cooperatives Offer Small Employers Plan Choice
and Market Prices (Letter Report, 03/31/2000, GAO/HEHS-00-49).

Pursuant to a congressional request, GAO reviewed small employer health
insurance purchasing cooperatives, focusing on: (1) a bill that would
establish Healthmarts; (2) studies of small employer purchasing
cooperatives and the small group market; (3) GAO's interviews with
health policy association officials and experts; and (4) experiences of
five small employer health purchasing cooperatives.

GAO noted that: (1) small employer purchasing cooperatives have been an
important component of several states' efforts during the 1990s to
improve small groups' health insurance options; (2) purchasing
cooperatives aim to provide small employers with some of the same
advantages larger employers have in offering health insurance; (3) the
five small employer purchasing cooperatives GAO examined have
demonstrated that these cooperatives can offer two of these
advantages--administrative services and a range of benefit options to
participating small employers; (4) by participating in a cooperative,
small employers have a single point of entry to multiple insurers' plans
with standardized benefit packages; (5) furthermore, since cooperatives
typically offer plans sponsored by a variety of insurance carriers with
different benefit levels and managed care features, participating small
employers can offer their employees a choice of multiple health
insurance products; (6) these advantages have led some small employers
that previously had not been offering health insurance to join a
cooperative; (7) however, even the larger cooperatives cover only a
small fraction of the small group health insurance market in their
states; (8) the experiences to date of small employer purchasing
cooperatives typically have not resulted in a third advantage--leverage
in negotiating lower premiums; (9) the cooperatives' potential to reduce
overall premiums is limited because: (a) they lack sufficient leverage
as a result of their limited market share; (b) the cooperatives have not
been able to produce administrative cost savings for insurers; or (c)
their state laws and regulations already restrict to differing degrees
the amount insurers can vary the premiums charged different groups
purchasing the same health plan; (10) cooperatives can potentially offer
lower premiums for firms with high-risk, high-cost individuals if they
restrict the premiums insurers may charge individual firms more than
restrictions already imposed by state statutes and regulations for all
small groups; (11) to ensure their viability in the small group market,
the purchasing cooperatives GAO examined have taken several steps to
maintain a sufficient number of participating insurers and employers;
(12) recognizing and countering insurers' perceptions that high-risk
individuals and groups are likely to enroll through the cooperative is
key to gaining insurer participation; and (13) furthermore, to increase
employer participation, the cooperatives have maintained insurance
agents' principal role in guiding employers to the cooperative.

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

 REPORTNUM:  HEHS-00-49
     TITLE:  Private Health Insurance: Cooperatives Offer Small
	     Employers Plan Choice and Market Prices
      DATE:  03/31/2000
   SUBJECT:  Employee medical benefits
	     Small business
	     Proposed legislation
	     Insurance premiums
	     Health insurance
	     Health insurance cost control
IDENTIFIER:  California
	     Connecticut
	     Florida
	     North Carolina
	     Texas

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** 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.                                         **
**                                                              **
******************************************************************

GAO/HEHS-00-49

Health, Education, and
Human Services Division

B-282868

March 31, 2000

The Honorable Thomas J. Bliley, Jr.
Chairman, Committee on Commerce
House of Representatives

Dear Mr. Chairman:

Nearly two-thirds of nonelderly Americans rely on employer-sponsored
coverage for their health insurance. Yet, despite low rates of unemployment,
the number of uninsured individuals has increased during the 1990s. More
than 70 percent of the working-age uninsured in 1998 were employed, and many
uninsured employees worked for a small firm. Of the 44 million uninsured
individuals in that year, nearly 16 million (more than one-third) were in
families where the family head was employed by a firm with fewer than 100
employees.1 Small firms are much less likely than other employers to offer
health insurance to their employees, and when offering coverage, they
usually offer only a single plan.

Concerned about the increasing number of uninsured, policymakers have sought
ways to improve the accessibility and affordability of health insurance for
individuals working for small employers. One approach that attempts to
create better access and choice as well as lower costs for employees of
small firms is to facilitate cooperatives and other pooled purchasing
arrangements that employers can join to purchase health insurance.2 Recent
congressional proposals would create a new type of pooled purchasing
arrangement for small employers known as a "Healthmart." According to
proposed legislation, Healthmarts would be nonprofit entities offering a
choice of health insurance plans to employers with 2 to 50 employees. To
assist the Congress as it considers proposals to facilitate the development
of such purchasing arrangements, you asked us to examine the experiences of
small employer health insurance purchasing cooperatives, in particular,
those sharing some key design features with the proposed "Healthmarts."
Specifically, we answered the following questions:

1. What advantages do health insurance purchasing cooperatives offer to
small employers, and to what extent have these advantages been effective in
attracting employers?

2. How successful have cooperatives' strategies been to obtain premium
reductions?

3. How do health purchasing cooperatives maintain their viability in the
small group market?

To address these questions, we (1) reviewed a bill that would establish
Healthmarts,3 (2) reviewed studies of small employer purchasing cooperatives
and the small group market, (3) interviewed health policy association
officials and experts, and (4) examined the experiences of five small
employer health purchasing cooperatives. These five cooperatives were drawn
from a list of small employer purchasing cooperatives identified by the
Institute for Health Policy Solutions.4 Similar to the proposed Healthmarts,
all five of the cooperatives we examined provide coverage to firms with 50
or fewer employees and offer at least two fully insured coverage options.
Three that we examined--the Pacific Health Advantage in California,5 the
CBIA Health Connections in Connecticut, and Florida's Community Health
Purchasing Alliance6--are among the largest small employer cooperatives in
the nation. We also examined two smaller cooperatives--North Carolina's
Caroliance and the Texas Insurance Purchasing Alliance. In our review of the
cooperatives, we interviewed their officials and selected participating
employers and insurers. In addition, we reviewed documents from the
cooperatives and the insurers, as well as state laws regulating premiums in
these small group markets. While these five cooperatives share many features
with the proposed Healthmarts, their experiences may not be fully
generalizable to other types of purchasing arrangements or those operating
under other state insurance regulations or in different health insurance
markets. We conducted our review between May 1999 and January 2000 in
accordance with generally accepted government auditing standards.

Small employer purchasing cooperatives have been an important component of
several states' efforts during the 1990s to improve small groups' health
insurance options. Established either by state legislation or private
employer associations, purchasing cooperatives aim to provide small
employers with some of the same advantages larger employers have in offering
health insurance, such as administrative simplicity, choice of multiple
insurers and benefit packages, and leverage in negotiating lower premiums.
The five small employer purchasing cooperatives we examined have
demonstrated that these cooperatives can offer two of these
advantages--administrative services and a range of benefit options to
participating small employers. By participating in a cooperative, small
employers have a single point of entry to multiple insurers' plans with
standardized benefit packages that can be compared easily instead of having
to individually identify insurers and their agents, review widely varying
benefit options, and determine price and terms of coverage. Furthermore,
since cooperatives typically offer plans sponsored by a variety of insurance
carriers with different benefit levels and managed care features,
participating small employers can offer their employees a choice of multiple
health insurance products--a sharp contrast to the single-plan option
offered by most small and many large employers. These advantages have led
some small employers that previously had not been offering health insurance
to join a cooperative. However, even the largest cooperatives cover only a
small fraction (typically about 5 percent or less) of the small group health
insurance market in their states.

The experiences to date of small employer purchasing cooperatives typically
have not resulted in a third advantage, which is available to large
employers: leverage in negotiating lower premiums. Officials of the
purchasing cooperatives and participating insurers as well as several recent
studies reported that cooperatives typically offer plans at market prices
for plans with similar benefits offered to small employers outside the
cooperative. This similarity in premiums is also reflected by rate
quotations we obtained from several insurers. The cooperatives' potential to
reduce overall premiums is limited because (1) they lack sufficient leverage
as a result of their limited market share; (2) the cooperatives have not
been able to produce administrative cost savings for insurers; or (3) their
state laws and regulations already restrict to differing degrees the amount
insurers can vary the premiums charged different groups purchasing the same
health plan. Cooperatives can potentially offer lower premiums for firms
with high-risk, high-cost individuals if they restrict the premiums insurers
may charge individual firms more than restrictions already imposed by state
statutes and regulations for all small groups. The Texas cooperative
initially imposed significantly more restrictive rating criteria than were
required by the state and as a result attracted a disproportionate share of
high-risk groups. Concerned about this trend, and the defection of insurers,
the cooperative attempted to make premiums more comparable to those
available in the small group market by eliminating restrictions and allowing
participating insurers to set premiums under the same terms allowed by the
state for the small group market.

Finally, to ensure their viability in the small group market, the purchasing
cooperatives we examined have had to take several steps to maintain a
sufficient number of participating insurers and employers. Recognizing and
countering insurers' perceptions that high-risk individuals and groups are
likely to enroll through the cooperative is key to gaining insurer
participation. Therefore, the cooperatives have attempted to manage
potential risk selection through their design of standard benefit packages
and other approaches. Furthermore, to increase employer participation, the
cooperatives have maintained insurance agents' principal role in guiding
employers to the cooperative. The cooperatives have had mixed success in
maintaining their viability. While the Connecticut and California
cooperatives continue to enjoy relatively stable employer participation,
other cooperatives have faced declining participation, and a few, including
the one in Texas, have disbanded.

While nearly two-thirds of nonelderly Americans rely on employer-sponsored
coverage for their health insurance, only about one-half of very small firms
(those with three to nine employees) offered health insurance in 1998,
compared with more than 95 percent of those with 50 to 199 employees. In
addition, a recent employer survey found that, compared with plans offered
by large employers, very small firms paid premiums that averaged about 10
percent higher for plans that covered fewer benefits, and required
deductibles twice as high.7 Furthermore, employees of small firms frequently
have lower incomes than those in larger firms. Lower-wage employees are less
likely to accept coverage if they must pay part of the premium.8

Studies in the early 1990s consistently pointed to the high and rising cost
of insurance as the key factor preventing small employers from offering
coverage to their workers. In addition, some insurance practices exacerbated
the problem by substantially increasing the costs or denying coverage for
some higher-risk firms or workers. Consequently, in the early and mid-1990s,
most states adopted some type of insurance market reform designed to improve
access and affordability for small employers. Reforms passed in many states
included measures to ensure that

ï¿½ employers who want health insurance coverage for their employees will be
accepted and renewed by insurers for at least one plan ("guaranteed issue");

ï¿½ premiums charged different employers purchasing coverage cannot vary by
more than a specified percentage (rate restrictions); and

ï¿½ small employer purchasing cooperatives can be established to improve
firms' access to multiple insurers and to seek lower premiums potentially
available to larger firms.

The Health Insurance Portability and Accountability Act of 1996 (HIPAA) set
federal standards for certain aspects of private health insurance that apply
to all small employers.9 One such standard is a requirement that insurers
that provide coverage in a group market must accept all small employers that
apply for any of the plans they offer (guaranteed issue). The extent to
which HIPAA's guaranteed issue provision improved market access for small
employers in a given state, however, is largely dependent on the extent of
state reforms preceding HIPAA. Further, while HIPAA may have improved the
choices of products available to small employers, the cost of coverage,
especially for high-risk groups, may still be unaffordable.

Various pooled purchasing arrangements, including large employer purchasing
cooperatives,10 multiple employer trust or welfare arrangements,11 trade or
other associations,12 and small employer health insurance purchasing
cooperatives, have been used as strategies for improving the provision of
employer-sponsored health insurance. The pooling arrangements may vary
significantly in terms of their coverage options, who can participate, and
the geographic area in which they operate.

More than 20 states have adopted legislation allowing for the establishment
of small employer purchasing cooperatives. Many of the small employer
purchasing cooperatives started in the 1990s were based upon the National
Association of Insurance Commissioners' (NAIC) purchasing alliance model
acts and differ from previous pooled purchasing arrangements in several
ways. Some tenets of the NAIC model acts that are also reflected, in varying
degrees, in the five cooperatives we reviewed include the following:13

ï¿½ accepting any small employer choosing to join the cooperative;14

ï¿½ establishing a governing board that is dominated by participating employer
and employee representatives;

ï¿½ prohibiting any financial risk for the cost or provision of health
services to be borne by the cooperative, but instead contracting with at
least three unaffiliated insurers to make a variety of fully insured plans
available to its members;15

ï¿½ obtaining insurers' participation by contracting with qualified group
carriers meeting objective criteria established by the cooperative16 through
a fair, competitive process;

ï¿½ negotiating the administrative portion of premiums with participating
insurers to reflect any cost savings the insurer experiences in the coverage
it offers through the cooperative; and

ï¿½ allowing employees to enroll with any insurer or plan offered through the
cooperative.17

See table 1 for more specific characteristics of the five cooperatives
examined in our study.

Table 1: Characteristics of Five Small Employer Health Purchasing
Cooperatives

            Name of                                         Number of
            small        Date       Size of      Average    employer
 State      employer     coverage   employer     employer   groups and  How sponsored or
            purchasing   available  eligible to  size       enrollees,  established
            cooperative             participate             1999
                                                            8,216
            Pacific                                         groups,
 California Health       1993       2 to 50      10                     State-established
            Advantage                                       144,424
                                                            lives
                                                            3,500
                                                            groups,
 ConnecticutCBIA Health  1995       3 to 50      8                      Privately
            Connections                                                 established
                                                            55,000
                                                            lives
            Florida                                         18,000
            Community                                       groups,
 Florida    Health       1994       1 to 50      2                      State-established
            Purchasing                                      75,000
            Alliance                                        livesa

 North                                                      920 groups,
 Carolina   Caroliance   1994       1 to 50      4          2,900       State-established
                                                            livesb
            Texas                                           Ceased
 Texas      Insurance    1994-July  2 to 50      6          operations  State-established
            Purchasing   1999                               in July
            Alliance                                        1999c

aFlorida's cooperative reached a peak enrollment of 24,000 groups and 92,000
lives in 1998.

bNorth Carolina's cooperative had a peak enrollment of 1,200 groups and
5,000 lives in 1995.

c Texas' cooperative had a peak enrollment of 13,000 lives in 1997.

To create an additional option for small employers seeking health insurance
coverage, recent congressional proposals would create a new pooled
purchasing arrangement for small employers known as a Healthmart. As
proposed, Healthmarts share many design characteristics with existing
purchasing cooperatives. Unlike most existing small employer purchasing
cooperatives, however, Healthmarts would be operated jointly by employers,
providers, insurers, and employees; would be exempt from state-mandated
benefits18; and could operate in more than one state. Table 2 displays some
key features of purchasing cooperatives and how they compare with the
proposed Healthmarts.

Table 2: Comparison of Healthmart Proposal and Typical Existing Small
Employer Purchasing Cooperatives

 Design feature      Healthmarts                  Small employer purchasing
                                                  cooperatives
 Choice of           Would offer at least 1       Offer more than 2
 insurers/plan       insurer and 2 coverage       insurers and coverage
 options             options                      options
                     Would offer fully insured    Offer fully insured
 Risk of pooled      health plans; Healthmarts    health plans;
 arrangement         would bear no financial      cooperatives bear no
                     risk                         financial risk
 Subject to state    Exempt from state-mandated
 insurance           benefits, otherwise subject  Subject to state
 regulations         to state regulation          regulation
                                                  Open to all firms with 50
                                                  or fewer employees. While
                                                  some exclude firms with
 Size of eligible    Open to all firms with 2 to  fewer than 2 or 3
 group               50 employees                 employees, others include
                                                  self-employed individuals
                                                  or employers with more
                                                  than 50 employees.
                     Equal representation from    Typically represent
 Governing board     employers, employees,        employer and employee
                     providers, and insurers      purchasers
                     Nonprofit entity operated
 How sponsored or    jointly by employer,         Varies. Some established
 established         employee, provider, and      by state, some by private
                     insurer representatives      associations of employers

Source: Information for Healthmarts is based on the Healthmarts proposal as
approved by the House of Representatives in H.R. 2990, The Quality Care for
the Uninsured Act of 1999, Oct. 1999. Information for the small employer
purchasing cooperatives is based on a summary of 13 small employer
purchasing cooperatives by the Institute for Health Policy Solutions. See
http://www.ihps.org/chpglist.html (Nov. 2, 1999) and Jack Meyer and others,
Small Employer Health Insurance Purchasing Arrangements: Can They Expand
Coverage? (Economic and Social Research Institute, May 1999).

but Represent a Small Portion of the Small Group Market

Small employer cooperatives offer administrative services to employers
seeking coverage by preselecting a group of insurers, standardizing benefit
packages, and obtaining rates on behalf of the small employer purchasers.
Employers participating in the cooperatives typically offer their employees
a wider range of plan choices than do nonparticipating small employers. The
five small employer purchasing cooperatives we reviewed also enrolled some
very small employers as well as employers that had not previously offered
health insurance coverage. Nevertheless, the cooperatives remain a very
small fraction of the small group market in their states.

Insurance Coverage

Small employers seeking health insurance for their employees outside of a
purchasing cooperative can be faced with significant administrative burdens.
Cooperatives seek to facilitate the purchase and choice of insurance
available to small employers and their employees by providing them with a
single point of entry to a choice of plans offered by multiple insurers. In
addition, cooperatives simplify the selection of coverage by collecting and
publishing premiums for sets of standardized benefits. Specifically, each of
the five cooperatives we reviewed offered at least two standard sets of
benefits and copayment levels. According to the literature and cooperative
officials we interviewed, having standardized benefits not only helps
consumers evaluate the costs and benefits of each plan option but forces
insurers to distinguish themselves through other features such as premiums
and provider networks rather than differences in covered services. Further,
another benefit to a firm's participating in a cooperative is assistance in
providing information to the firm's employees about the insurer's benefits
and, in some cases, helping to resolve conflicts that may arise between an
insurer and the firm's employees.

Cooperatives also facilitate access to coverage for newly formed or very
small businesses because they impose fewer barriers to enrollment. Both
officials of the cooperatives and an insurer reported that some insurers
require more extensive documentation, for example, to show that an employer
has been in business a sufficient period of time prior to offering the
employer coverage. For example, a Florida cooperative official reported that
new employers can demonstrate their eligibility for coverage by simply
submitting their estimate of taxes, while outside the cooperative, insurers
may request additional documentation, such as a copy of an employer's tax
returns. Although this makes it easier for new businesses to purchase
insurance through cooperatives, one Florida insurer cautioned that this
practice resulted in some fictitious groups gaining coverage, such as a
self-employed individual who adds a family member or friend to the rolls of
the firm solely for the purpose of purchasing group health insurance.19

Options

By participating in a cooperative, small employers have more opportunity to
offer their employees multiple plan choices. Choice within a cooperative
also provides collateral benefits to employers and employees. For example, a
greater choice of insurers inside the cooperative may assist participating
employers in their recruitment and retention efforts, and creates a better
likelihood that their employees are able to select a plan that includes the
provider of their choice. Employees do not always have access to all plans
offered through the cooperative, however, depending on their location or
employer.

Employers in each of the five cooperatives we reviewed offered their
employees a greater choice of health plans than did small employers outside
of the cooperative. When small firms sponsor health insurance outside of a
cooperative, they usually make only one plan available to their employees.
Results of a 1999 survey of nearly 2,000 employers showed that only a single
plan choice was available to more than 90 percent of covered workers in
firms with 3 to 49 employees. In contrast, at least three plan choices were
available to nearly 84 percent of covered employees in firms with more than
5,000 employees. See figure 1 for the number of plan choices by firm size.20

Figure 1: Percentage of Covered Employees and Choice of Health Plans, by
Firm Size, 1999

Source: Kaiser Family Foundation and Health Research and Educational Trust,
Employer Health Benefits 1999 Annual Survey.

Furthermore, each of the cooperatives we examined offered multiple managed
care plans, insurers, benefit packages, and premium options. Most
cooperatives offered at least two types of managed care plans, usually
several health maintenance organizations (HMO) and either a point-of-service
(POS) plan or a preferred provider organization (PPO) option. Among the
cooperatives, the California cooperative currently has the largest number of
plan options and the North Carolina cooperative has the fewest (see table
3). In choosing among different plans, an individual could see variation in
total premiums of as little as 28 percent in the Florida cooperative to more
than 100 percent in the California cooperative.

Table 3: Plan Types and Premiums Available Through Four Cooperatives, 1999

                                                                    Monthly
                                                     Available      premiums
                                                                    (employee-only
 State/cooperative         Available   Available     standardized   coverage for a
                           insurers    plan typesa   benefit
                                                     packages       30- to
                                                                    39-year-old
                                                                    enrollee)
                                                                    $89.22 to
 California−Pacific                                           $204.02b
 Health Advantage          18          HMO, POS      3

 Connecticut−CBIA:                                            $130.81 to
 Health Connections        4           HMO, POS      2              $233.08
                                                                    $89.51 to
 Floridac                  8           HMO, HMO      2              $147.89e
                                       open-accessd

 North                                 HMO, PPO,                    $81.29 to
 Carolina−Caroliance 3                         2              $160.64
                                       indemnity

Note: The Texas cooperative is not included in this table because it ceased
operations in 1999 and thus information comparable to that for the other
cooperatives was not available.

a All plan types may not be available in all locations.

bRates are for July 1, 1998, through June 30, 1999.

c Includes information only for region 6 (Central Florida) of Florida's
Community Health Purchasing Alliance.

dAn open-access HMO permits enrollees to visit a specialist without a
referral from a primary care doctor or other gatekeeper.

eRates for a 30-year-old male in Hillsborough County (Central Florida)
available between October 1999 and December 1999.

Even though cooperatives attempt to offer choice to their participants, not
all plans are available in all areas served by each cooperative, and
individual employers using some cooperatives may limit the choice of plans
their employees can select. For example, the Connecticut cooperative made 16
plan options available to employers, but only one-half of surveyed employers
offered all of the plans to their employees. Similarly, the Texas
cooperative offered as many as eight plans in urban areas but only two plans
in rural areas.

The five small employer purchasing cooperatives we reviewed have attracted
some very small employers as well as varying proportions of employers that
had not previously offered health insurance coverage. However, small
employers participating in the cooperatives represent only a very small
share of their states' small group health insurance market. Furthermore, the
HIPAA requirement that insurers must issue health insurance to any small
employer willing to purchase it, regardless of the health condition or
anticipated costs of the employees, may have reduced the incentive for some
small employers to join a cooperative. At the same time, it may have
enhanced cooperatives' viability by reducing the differences between
purchasing through a cooperative and in the general small group market.

The purchasing cooperatives that we examined appeared to have enrolled
primarily very small employers, but they attracted very different
proportions of employers who previously did not offer coverage. For example,
the average employee group size in the cooperatives ranged from 2 in Florida
to 10 in California. In addition, cooperative officials reported that about
half of the newly participating employers in the Florida, North Carolina,
and Texas cooperatives had not sponsored coverage in the prior year. In
contrast, just over one-quarter of the newly participating employers in the
California cooperative and fewer than an estimated 10 percent of employers
newly participating in the Connecticut cooperative had not sponsored
coverage in the prior year. However, these newly participating employers
might have offered health insurance even in the absence of the
cooperative.21

None of the purchasing cooperatives we reviewed had a large enough market
share to create bargaining leverage and therefore had a limited ability to
significantly increase the percentage of small employers offering coverage
in their states. For example, the Pacific Health Advantage, with 144,000
covered lives, is one of the largest small employer purchasing cooperatives
in the nation, but it accounts for only about 2 percent of the small group
health insurance market in California. Except for the Connecticut
cooperative, which accounted for 5 to 8 percent of the small group market in
the state, the other purchasing cooperatives we reviewed accounted for less
than 5 percent of their state's market.

Despite efforts to negotiate lower premiums, cooperatives have only been
able to offer premiums that are comparable to those in the general small
group market. The cooperatives we reviewed typically did not obtain overall
premium reductions because (1) their market share provided insufficient
leverage, (2) they could not produce administrative savings for insurers, or
(3) premium variation is already restricted by state laws and regulations.
While one cooperative attempted to impose more stringent restrictions on the
premiums that could be charged for high-risk individuals than the
restrictions already imposed by the state, insurers withdrew from the
cooperative because they were concerned they would receive a large
proportion of high-risk, high-cost individuals. As a result, the cooperative
abandoned this practice.

Premiums

The cooperatives we reviewed used a number of strategies to obtain lower
premiums from insurers that could benefit either the cooperative as a whole
or, in some cases, employers who were likely to face higher premiums
outside. In one strategy, the cooperatives attempted to use their pooled
market strength to generate competition among insurers and negotiate
favorable rates for all participants comparable with the rates of a large
purchaser. To participate in the California cooperative, insurers are
precluded from offering a plan outside of the cooperative at a lower price
if it is equal to or greater in actuarial value than the one offered inside
the cooperative. The California and Connecticut cooperatives also encourage
participating insurers to resubmit bids that are at the high end of rates
allowed by state insurance rules. Also, since its second year of operation,
the California cooperative has restricted the entry of new insurers to those
that offer lower-priced plans than those currently available or those whose
plans otherwise improve the cooperative's selection of available providers
or product lines.

Another strategy cooperatives used was attempting to obtain lower premiums
from insurers by assuming administrative responsibility for activities such
as billing and enrollment that are currently performed by the insurers. One
Connecticut insurer reported that since the cooperative was responsible for
billing and collecting premiums, this insurer was able to set up one account
to receive payments from the cooperative instead of managing individual
accounts for each employer. In addition, in an effort to achieve
administrative efficiencies for participating insurers, the Florida
cooperative's new third-party administrator intends to apply new
technologies to electronically scan enrollment applications and transmit
billing data to insurers and agents.

To make insurance more affordable for employers with high-risk individuals,
one cooperative employed a third strategy--restricting how much insurers
could vary premiums for different firms. This meant that some employers
could potentially find more affordable insurance through the cooperative
than in the general small group market.

to Other Small Employers

Officials of the cooperatives and selected participating insurers reported
that premiums available through the cooperatives were typically similar to
those available in the market, and might be slightly higher or lower in
certain instances. As shown in table 4, premium quotations we obtained from
selected insurers participating in several cooperatives indicated that
prices of plans offered through the cooperative were approximately the same
as those for similar plans available to nonparticipating small employers.

Table 4: Selected Insurers' Premiums for a Plan Offered Through a
Cooperative and a Similar Plan Outside the Cooperative

                     25-year-old                55-year-old
                                     Similar                    Similar
 State        Plan   Cooperative's   plan's     Cooperative's   plan's
              type   monthly premium monthly    monthly premium monthly
                                     premium                    premium
 California   HMO    $108            $101       $231            $216
 Connecticuta HMO    138             147        277             296
              POS    187             169        375             338
 Floridab     HMO    127             127        357             357

Note: The premium quotations are based on two age groups of 10 individuals
each--employees aged 20 to 25, and employees aged 55 to 59. The premium
quotations we received from the Connecticut insurer for these two groups are
for individuals less than 30 years of age and individuals 50 to 59 years of
age, respectively. All premium quotations are preliminary and could change
on the basis of additional information received by the insurer.

aPremium quotations are for Hartford, Connecticut.

bPremium quotations are for the cooperative's region 6, which includes
Central Florida.

Similarly, a national survey of over 21,000 employers also found premiums
within pooled purchasing arrangements to be comparable to those outside.
Specifically, this survey found that 1997 monthly single (self-only)
premiums for small employers participating in any pooled purchasing group,
including cooperatives, multiple employer welfare arrangements, and
association plans, were $180, compared with $172 for nonparticipants.22

By initially imposing greater restrictions on insurers' ability to vary
premiums than the restrictions imposed by state law for insurers in the
small group market, the Texas cooperative was able to temporarily offer
significantly lower rates to employers with high-risk employees.
Specifically, participating insurers were not allowed to vary premiums on
the basis of group health status, group size, or industry, as they could
outside of the cooperative. As a result, a cooperative official reported
that less healthy individuals could obtain premiums about 30 percent lower
through the cooperative; however, healthy individuals could pay up to 30
percent more for coverage within the cooperative. Because this practice
caused the cooperative to attract a disproportionate share of high-risk
groups, beginning in January 1997 it revised its rating practices to become
more comparable to rates available in the small group market.

Three factors appear to explain why cooperatives have not typically achieved
greater overall premium reductions. These factors are a small market share,
minimal administrative savings, and state rating requirements.

First, despite their efforts to leverage market share, the cooperatives we
reviewed were limited in their ability to do so. In part, this was a
function of their relatively small size. Three of the
cooperatives--California, North Carolina, and Texas--each represented 2
percent or less of the small group market in their states, while Florida and
Connecticut represented about 5 and 10 percent, respectively, of market
share. In some cases, a cooperative's bargaining leverage was impeded by
legal and practical constraints that largely prevented it from selectively
contracting with only a few insurers, thereby fragmenting an already limited
pooled market strength. For example, some of the cooperatives we examined
experienced difficulty in leveraging their market share because they were
required to contract with all qualified health plans, as in Florida, or they
had difficulty attracting and retaining insurers, as in Texas. To avoid a
threatened boycott by agents and brokers, the California cooperative
initially accepted all plans that met its terms; however, the cooperative
has subsequently been able to be more selective in admitting new plans. In
contrast, the Connecticut cooperative--a privately sponsored entity--faced
neither legal nor business impediments and therefore had more flexibility in
limiting participation to four insurers. It cannot, however, pass any
savings it may achieve to its members in the form of lower premiums because
the state Department of Insurance interprets the state's community rating
statute to prohibit insurers from adjusting their rates as a result of
administrative savings.

Insurers' perceptions of risk also limit a cooperative's abilities to use
its pool of small employers to obtain lower premiums. Specifically,
actuaries and insurers reported that the risk of insuring a large employer
is perceived to be more favorable than that of a small employer, even when
the small employer participates in a purchasing cooperative. Insurers
reportedly anticipate a greater chance for adverse selection within a
purchasing pool composed of several small employers than for a comparably
sized pool created by one large employer. In addition, some insurers believe
that some small business owners are inclined to purchase health insurance
only when they or an employee imminently needs coverage.

Second, while the cooperatives tried to obtain premium reductions by
assuming some of the administrative responsibilities of insurers, the
anticipated administrative savings either never materialized or were not
valued by insurers. According to some cooperative officials, administrative
savings are inherently limited because cooperatives can relieve insurers of
only a fraction of their costs. Moreover, when assuming responsibility for
an administrative task, such as marketing, a cooperative generates its own
costs that must be covered, typically by a fee to members. Administrative
costs can be higher for insurers selling insurance to a cooperative of small
employers than for a single large employer. This is due, in part, to
insurers facing an increased need to create name recognition and product
differentiation when employees in cooperatives can choose among competing
plans. One insurer also reported that its administrative costs actually
increased because it had to modify its computer programs in order to
accommodate the information requirements for business obtained through the
cooperative.

Third, cooperatives were limited in their ability to offer significantly
lower premiums to firms with high-risk employees because state laws, to
varying degrees, already restrict premium variation allowed in the small
group market. For example, the premiums within the Florida cooperative are
essentially based on the same rating factors that can be used outside--age,
gender, family composition, geography, and tobacco use. As a result, most
groups could obtain similar premium prices from an insurer whether they
purchased a plan through this cooperative or on their own. The California
cooperative based its premiums on three rating factors--age, geography, and
family size. Outside of the cooperative, state insurance laws and regulation
allowed insurers to vary premium rates by plus or minus 10 percent from
their base rate for additional factors such as health status. Therefore, if
insurers applied these additional underwriting standards,23 premiums outside
the cooperative could be as much as 10 percent higher or lower for some
groups.

In addition, cooperatives were concerned that offering lower premiums to
firms with high-risk employees could attract too many high-risk individuals,
a phenomenon called "adverse selection."As mentioned earlier, the Texas
cooperative was able to temporarily offer significantly lower rates to
high-risk employers by imposing greater restrictions on insurers' ability to
set premiums than the restrictions allowed by state laws and regulations for
all insurers in the small group market. By doing so, however, the
cooperative produced the classic "death spiral" by charging relatively
higher rates to those who were healthier, leading to an increase in average
premiums as more high-cost individuals enrolled in the cooperative and the
low-cost individuals left. While the cooperative eventually adopted rating
practices similar to those used by insurers outside, insurers withdrew their
participation anyway and the cooperative ceased operation in July 1999.

To remain viable, cooperatives have to attract a sufficient number of
insurers and employers interested in participating. Key to this effort is
maintaining both the perception and the reality that the participating
employer groups do not include disproportionate numbers of high-risk
individuals. To accomplish this, cooperatives used a variety of strategies
to protect individual insurers from receiving a disproportionate share of
high-risk enrollees, including establishing standardized benefit packages
and formal risk-adjustment mechanisms. Furthermore, the cooperatives have
learned that they need to maintain close working relationships with
insurance agents in the small group market to increase participation of
employers.

Small employer cooperatives have had varied success in maintaining their
viability. The cooperatives in Connecticut and California have generally
enjoyed stable--even growing--employer participation. However, participating
PPO plans withdrew from the California cooperative because of adverse
selection by high-risk groups. In contrast, cooperatives in Texas, Iowa, and
Kentucky have discontinued operation. Other cooperatives, including those in
Florida and North Carolina, have struggled with declining participation by
insurers and employers. Several insurers have withdrawn from both the
Florida and the North Carolina cooperatives, some citing high administrative
costs and the risk perceived to be associated with the cooperatives.

Insurers

The cooperatives we examined used various approaches to manage potential
risk selection--adverse or favorable--among insurers, including establishing
standard benefit packages, requiring all participating insurers to offer
both an HMO and a POS option, establishing employer participaton
requirements, and establishing formal mechanisms to adjust payments for
instances where adverse selection had occurred. Because cooperatives allow
employees to choose directly among competing plans, the potential for
selection bias may be compounded, since employees have more information
about their own likelihood of using services than do their insurers or
employers. To preserve and enhance a broad choice of plans for their
members, cooperatives instituted approaches to mitigate the potential for
adverse selection affecting insurers that offered out-of-network options
such as PPOs and POSs that might be attractive to high-risk individuals.

Each of the cooperatives we reviewed made standardized benefit packages
available to limit the potential for risk selection. By requiring all
participating insurers to offer similar benefit packages, the cooperatives
try to minimize attempts by insurers to attract only lower-risk groups by
excluding benefits particularly valued by high-risk individuals. This was
echoed by a cooperative official and an insurer we interviewed who indicated
that the cooperatives designed their benefit packages to closely match those
offered in the small group market so that they would not attract an undue
proportion of low- or high-risk groups seeking coverage.

Beyond standardizing benefits, all participating insurers in the Connecticut
cooperative offered the same two types of managed care plans--HMO and POS.
The cooperative deliberately designed its plan offering to reduce risk
selection that might develop for insurers if they offered only a POS option.
The goal was to minimize the likelihood that any one participating insurer
would find itself with an undue proportion of higher- or lower-risk
individuals because of the type of plan that it offered.

The California cooperative implemented a risk-adjustment system that
retrospectively redistributes funds to insurers who attract a
disproportionately sick population. California's risk-adjustment model is
designed to redistribute funds from plans that enroll a population with
expected costs at least 5 percent less than the cooperative's average to
plans with costs expected to be at least 5 percent greater. Plans with
lower-risk enrollees pay a portion of their revenues to plans with
higher-risk enrollees.24 Even with this risk-adjustment mechanism, however,
insurers no longer offer a PPO option, considered too high-cost, through the
California cooperative.

The North Carolina cooperative explored two alternative risk-sharing
mechanisms but did not implement either. Officials of the cooperative
reported that they considered a risk-adjustment system and a reinsurance
mechanism providing stop-loss protection to insurers for individual claims
exceeding $150,000. The cooperative explored financing these mechanisms with
funds contributed to a pool by insurers or with funds from the state for the
purpose of creating a high-risk pool. The cooperative did not implement
either mechanism because the expected costs would have obviated potential
premium reductions sought by the cooperative from participating insurers.

Cooperatives

Officials of the cooperatives we examined indicated that working
collaboratively with insurance agents and obtaining their support is
essential to success. Typically, small employers rely on agents to assist
them in procuring health insurance. Specifically, agents assist small
employers in identifying plans, completing applications, and obtaining
premium quotations. Without the cooperation of agents, a cooperative is
likely to face difficulty identifying small employers interested in offering
health insurance and in enrolling those employers, and may experience
adverse selection.

Cooperatives fulfill many of the functions that insurance agents
traditionally have performed for small employers. In recognizing this, the
California cooperative reported taking actions that made optional the use of
agents and, for those using agents, limited the commissions paid to agents,
thereby alienating many of them. These actions included paying agents a set
fee per employer instead of a percentage of premiums; listing the amount of
the agent's commission on the employer's bill; and enrolling employers
directly, thus bypassing agents. As a result, agents were less likely to
market the cooperative to small employers who initially contacted agents.
The state board overseeing the cooperative discontinued these practices
because it realized agents are needed to enroll small employers; currently,
77 percent of employers in the California cooperative use agents or brokers.
A North Carolina cooperative official stated that the cooperative attempted
to work more collaboratively with agents as a result of hearing about the
California experience.

The commissions paid to the agent by the cooperative and insurer, and the
agent's relationship with insurers, can also influence which groups an agent
will enroll in the cooperative. When the cooperatives and insurers set
different agent commissions for the same groups, agents have a financial
incentive to enroll their groups where the commission is highest. While
cooperatives offering higher commissions than insurers outside the
cooperative may increase employer referrals by agents, it does not guarantee
that agents will not tend to direct higher-risk groups to the cooperative.
Cooperative officials reported that agents are sometimes reluctant to enroll
higher-risk groups directly with insurers in order to protect existing
financial relationships. Instead, such groups may be steered toward the
cooperative, since some cooperatives are sometimes viewed as de facto
high-risk pools. For example, after raising commissions to 10 percent to
enhance its competitive position and broaden its agent base, the Texas
cooperative reported that agents still tended to enroll groups with
high-risk individuals in the cooperative, rather than enroll them directly
with insurers. In this way, agents could receive a relatively higher
commission for high-risk groups while maintaining positive relations with
insurers that provide the bulk of their compensation by continuing to enroll
low-risk groups with them outside of the cooperative.

The experience of existing cooperatives--based on our analysis of 5
cooperatives, as well as other recent research--demonstrates that they can
provide employees of small employers with an enhanced choice of health plans
offering standardized benefits. In addition, a cooperative can offer
employers fewer administrative hurdles to obtaining health insurance.
However, in general, existing cooperatives have not realized any potential
to significantly reduce premiums for employers or employees. Additionally,
not all of the cooperatives have been successful in attracting and retaining
insurers willing to participate within the constraints of the cooperative or
in avoiding enrolling a disproportionate share of firms with higher-risk
individuals.

Key to the cooperatives' ongoing viability is how practices such as benefit
offerings, premiums, and requirements for insurers affect the distribution
of risk among enrollees--that is, the extent to which the cooperative and
the insurers within it disproportionately attract and retain healthy,
low-cost enrollees or less healthy, higher-cost enrollees. Several
cooperatives established in the early to mid-1990s have either ceased
operation or faced declining participation by insurers and employers, and
even the largest small employer cooperatives have attained only a small
share of the health insurance market for small employers. As new forms of
pooled purchasing arrangements designed for small employers are proposed
that operate under different regulatory and market conditions, the
experiences of existing small employer purchasing cooperatives can provide
meaningful insights. However, differences in the design and regulatory
environment of the proposed Healthmarts and the existing cooperatives mean
that their experiences may not be fully generalizable for new pooled
purchasing arrangements.

Officials from the cooperatives we examined as well as expert reviewers from
two organizations provided comments on a draft of this report. In general,
they concurred with our findings and made technical comments, which we
incorporated where appropriate. In their comments, several reviewers
highlighted specific design features that they believe could enhance or
impede cooperatives' viability as more effective purchasers for small
employers. In particular, they emphasized the importance of the ability to
negotiate rates, selectively contract with insurers, and maintain
comparability in benefits and rating practices with the small group market.
As discussed in our report, the cooperatives we examined demonstrated these
characteristics to varying degrees but were constrained in their ability to
selectively contract and negotiate rates. All of them eventually established
benefits and rating practices that were generally comparable to the small
group market, and the existing ones remain a small part of that market. One
reviewer also noted that the design and operation of some more recently
established, privately sponsored cooperatives--such as those in Colorado,
New York, Oregon, and Washington--reflect these design features. Because
these cooperatives have only recently come into existence, have very few
participants, serve employers with more than 50 employees, or operate in
limited geographic areas, we did not include them in our review and cannot
comment on their operations.

As agreed with your office, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 30 days after
its issue date. We will then make copies available to appropriate
congressional committees and others upon request.

Please call me at (202) 512-7118 if you have any questions about this
report. The information presented here was developed by N. Rotimi Adebonojo,
JoAnne Bailey, and Mark Vinkenes, under the direction of John Dicken.

Sincerely yours,
Kathryn G. Allen
Associate Director, Health Financing and
Public Health Issues

(101831)

Table 1: Characteristics of Five Small Employer Health Purchasing
Cooperatives 11

Table 2: Comparison of Healthmart Proposal and Typical Existing
Small Employer Purchasing Cooperatives 12

Table 3: Plan Types and Premiums Available Through Four
Cooperatives, 1999 17

Table 4: Selected Insurers' Premiums for a Plan Offered Through a
Cooperative and a Similar Plan Outside the Cooperative 20

Figure 1: Percentage of Covered Employees and Choice of Health
Plans, by Firm Size, 1999 16

Table 1: Characteristics of Five Small Employer Health Purchasing
Cooperatives 11

Table 2: Comparison of Healthmart Proposal and Typical Existing
Small Employer Purchasing Cooperatives 12

Table 3: Plan Types and Premiums Available Through Four
Cooperatives, 1999 17

Table 4: Selected Insurers' Premiums for a Plan Offered Through a
Cooperative and a Similar Plan Outside the Cooperative 20

Figure 1: Percentage of Covered Employees and Choice of Health
Plans, by Firm Size, 1999 16
  

1. See Paul Fronstin, "Both Job-Based Health Coverage and Uninsured Continue
to Rise, CPS Shows," EBRI Notes (Nov. 1999), based on an analysis of the
March 1999 Current Population Survey by the Employee Benefit Research
Institute.

2. These small employer purchasing cooperatives are also sometimes referred
to as "alliances."

3. Our analysis was based on the Healthmarts proposal in H.R. 2990, The
Quality Care for the Uninsured Act of 1999 (Oct. 1999).

4. As of March 2000, the Institute for Health Policy Solutions identified 15
health purchasing cooperatives. See Institute for Health Policy Solutions'
Internet listing of "Consumer-Choice Health Purchasing Groups" at
http://www.ihps.org/ (Nov. 11, 1999). The Blue Cross and Blue Shield
Association identifies 21 purchasing cooperatives, some of which also serve
larger employers or government employees. See State Legislative Health Care
and Insurance Issues, 1998 Survey of Plans (Blue Cross Blue Shield
Association, Dec. 1998).

5. As of July 1, 1999, Pacific Business Group on Health assumed
administrative responsibilities for the cooperative, formerly known as the
Health Insurance Plan of California. Our review primarily represents the
experiences of the Health Insurance Plan of California, which was
administered by the state of California's Managed Risk Medical Insurance
Board.

6. Florida's Community Health Purchasing Alliance is based on a regional
governance structure. Each region is governed by its own board, which enjoys
wide autonomy in operation and decision-making. For the purposes of this
study we interviewed representatives of region 6, located in central
Florida, including the Greater Tampa Area.

7. See Kaiser Family Foundation and Health Research and Educational Trust,
Employer Health Benefits 1999 Annual Survey.

8. Philip F. Cooper and Barbara Steinberg Schone, "More Offers, Fewer Takers
for Employment-Based Health Insurance: 1987 And 1996," Health Affairs, Vol.
16, No. 6 (Nov.-Dec. 1997), pp. 142−9.

9. For additional information on HIPAA, see Private Health Insurance:
Progress and Challenges in Implementing 1996 Federal Standards
(GAO/HEHS-99-100, May 12, 1999 ).

10. See Access to Health Insurance: Public and Private Employers'
Experiences with Purchasing Cooperatives (GAO/HEHS-94-142, May 31, 1994 ).

11. See Employee Benefits: States Need Labor's Help Regulating Multiple
Employer Welfare Arrangements (GAO/HRD-92-40, Mar. 10, 1992 ).

12. See Employer Association Health Plans (GAO/HEHS-96-59R, Dec. 6, 1995).

13. See the "Single Health Care Voluntary Purchasing Alliance Model Act,"
and "Regional Health Care Voluntary Purchasing Alliance Model Act," NAIC,
Model Regulation Service, Oct. 1996.

14. Employers must agree to pay membership fees and the coverage premium as
well as follow the bylaws and rules of the cooperative. Additionally,
cooperatives may require employers to pay a minimum share of the total
premium and that all or a certain percentage of employees eligible to
purchase the employer sponsored coverage do so through the cooperative.

15. The state's commissioner of insurance may, upon a showing of good cause,
waive the requirement for three unaffiliated insurers.

16. Objective criteria may include requiring the insurer to obtain
certification from the state insurance commissioner that the insurer is
licensed in the small group market, satisfies state financial requirements,
and is in good standing.

17. Employees can be limited to those plans that provide coverage where he
or she lives. In addition, an alternative allows the employer to limit
employee choice to at least three insurers, of which one must provide
out-of-network coverage and one a managed care plan, if available.

18. Our 1996 report summarized studies in six states showing that mandated
benefits represented between 5 and 22 percent of total claims costs. This
report also found that most employers voluntarily offered commonly mandated
benefits, such as obstetrical care, mental health benefits, and mammography
screening, even if they were not required to do so. Therefore, to the extent
that employers typically offer these or similar benefits, the potential
premium savings from preempting mandated benefits may be less than their
share of claims costs. See Health Insurance Regulation: Varying State
Requirements Affect Cost of Insurance (GAO/HEHS-96-161, Aug. 19, 1996 ).

19. Insurers are concerned about fictitious groups because these individuals
may be obtaining coverage only when they expect to incur medical expenses.
By seeking group, rather than individual, coverage, they are guaranteed
coverage and may be able to obtain lower rates than in the individual
market.

20. See Kaiser Family Foundation and Health Research and Educational Trust,
Employer Health Benefits 1999 Annual Survey .

21. Many of the employers that were newly offering health insurance through
the cooperative may represent businesses that are just starting up as well
as other employers that might have begun offering coverage even in the
absence of the cooperative. Also, an employee of a firm that does not
sponsor insurance might still have coverage through another source, such as
individually purchased insurance or employer-sponsored coverage through a
spouse.

22. See Long and Marquis, "Pooled Purchasing: Who Are the Players?" Health
Affairs (July/Aug. 1999). This study did not report premiums for specific
small employer purchasing cooperatives. Nonetheless, it found that
individuals participating in the California, Connecticut, and Florida
cooperatives were no different for selected risk factors--age, sex, and
earnings--than those who purchased insurance outside of the cooperative.

23. The American Academy of Actuaries defines underwriting as "the process
of identifying and classifying the potential degree of risk represented by a
proposed insured or group of insured. Medical underwriting is sometimes used
to identify risks which are expected to incur high medical costs."

24. As discussed in John Bertko, Health Based Payments--What Do We Know
About Risk Adjusted Payments? (Jan. 1998) and Jill Yegian and others, Health
Insurance Purchasing Alliances for Small Firms: Lessons From the California
Experience (May 1998). The California cooperative's risk adjustment process
involves identifying individual enrollees who have been hospitalized in a
previous period with a "marker diagnosis"--that is, one of approximately 120
diagnoses with high costs, all requiring inpatient admission. Each
individual with one of these diagnoses is assigned a weight based on average
costs derived from California health care experience in managed care plans
during the period 1992 to 1994. All other enrollees receive an "average
weight." An individual insurer's average weight for all enrollees then
determines a risk assessment score that is used to calculate proposed
risk-adjustment amounts. In 1996, the plan that received the most favorable
selection paid $11.80 per contract per month into the pool, and the plan
that received the most adverse selection received $46.04 per contract per
month. In 1997, four plans paid out a maximum of $8 per contract per month,
and one PPO received $16 per contract per month. In 1997, 1998, and 1999,
the percentage of premiums transferred was 1.14 percent, 0.04 percent, and
0.11 percent, respectively. Pacific Health Advantage is in the process of
converting to a new risk-adjustment method.
*** End of document. ***