Study

Elliot K. Wicks, Ph.D.
Jack A. Meyer, Ph.D.
With assistance from Sharon Silow-Carroll, M.B.A., M.S.W.
New Directions for Policy
May 1999

TABLE OF CONTENTS

EXECUTIVE SUMMARY
INTRODUCTION
THE HISTORICAL CONTEXT
THE REFORMS
SMALL-GROUP INSURANCE REFORM: SPREADING RISK
COLLECTIVE PURCHASING ARRANGEMENTS
ANALYSIS OF HEALTH MARTS AND ASSOCIATION HEALTH PLANS
ACHIEVING OBJECTIVES
AVOIDING THREATS TO PAST REFORM PROGRESS
FOSTERING COLLECTIVE PURCHASING ARRANGEMENTS
COMPARING HEALTH MARTS AND ASSOCIATION HEALTH PLANS TO MORE ENCOMPASSING REFORMS
INCENTIVES, SUBSIDIES, AND BUY-INS
EMPLOYER MANDATES
INDIVIDUAL MANDATES
CONCLUSION
ENDNOTES

EXECUTIVE SUMMARY

Workers employed by small firms make up a disproportionate share of the uninsured, and small employers have an especially difficult time buying reasonably priced health insurance. Small employers have thus been a focus of numerous health insurance reforms over the past decade. Congress is now considering legislation designed to make health insurance more affordable for small employers, and thus to reduce the number of uninsured. This report examines the advantages and disadvantages of two reform vehicles included in proposed legislation-Health Marts and Association Health Plans-in the context of: (1) the unique difficulties small firms encounter in trying to buy and retain health insurance; (2) existing insurance laws and market reforms that aim to limit medical underwriting and risk selection; (3) recent experience in the health care marketplace with small business health insurance purchasing arrangements.

Health Marts, as outlined in a bill passed by the U.S. House of Representatives in July 1998, bear many similarities to existing Health Purchasing Cooperatives (HPCs). HPCs are collective purchasing arrangements for small employers that have been initiated in a number of states. Like HPCs, Health Marts would be administrative entities. They would offer multiple health plans (at least two), accept all small employers, and have to conform to state laws governing the difference in premiums among small firms. But Health Marts would differ in some important respects from HPCs. Most notably, Health Marts would be free of state mandates that require coverage of certain benefits and providers – allowing them, in theory, to offer a less costly benefit package. In addition, they would be governed by boards of employers, employees, health plans, health insurers, and health care providers. Thus, unlike HPCs, they would not be agents of the purchasers of health care but of all interested stakeholders.

Association Health Plans – as outlined in last year’s House bill and in newly proposed legislation this year – would be small business purchasing entities. National and local trade organizations would be permitted to form Association Health Plans. Like Health Marts, Association Health Plans would not have to conform to state laws mandating coverage of certain benefits and providers. In addition, they would be able, under certain circumstances, to offer self-insured health benefit plans. This would permit some small employers in Association Health Plans to avoid state insurance regulations and premium taxes – an advantage now accruing mostly to larger employers. As with Health Marts, the aim and expectation is that Association Health Plans would be able to offer less expensive health coverage. In addition, if an Association Health Plan selectively enrolled employers whose workers were younger and healthier than average, its premiums would be lower than the rest of the market.

The central conclusion of this study is that while Health Marts and Association Health Plans will offer advantages to some small firms and may somewhat reduce the deterioration in health insurance coverage in the U.S., they will not by themselves solve the problem of the uninsured. That is primarily because, on balance, neither Health Marts nor Association Health Plans are likely to reduce health costs enough to significantly entice most small firms not now offering coverage to buy health insurance. In addition, benefit packages that are significantly less comprehensive than typical do not seem to have broad appeal, and may still be too costly for most small businesses.

Health Marts and Association Health Plans could also undermine existing state reforms. Specifically, Association Health Plans that sought out or attracted small employers with low-risk workers would weaken the equitable small business “risk pools” that states have spent years trying to build. Small firms judged to be above-average risks could find their insurance rates climbing steeply as low-risk small firms joined Association Health Plans. With respect to Health Marts, the proposed legislation would allow them to operate in non-contiguous counties. This could permit them to engage in “red-lining” – choosing to operate where they can avoid small firms with high-risk employees.

The framers of the legislative proposals have included safeguards to limit the potential for risk selection. For example, Association Health Plans cannot be formed solely for the purpose of buying insurance. A new trade association or professional group must wait three years before offering an Association Health Plan. And Association Health Plans cannot refuse coverage to any group that is an association member. But these safeguards may be inadequate to prevent lower-risk groups from separating themselves out.

Health Marts and Association Plans must be gauged against the backdrop of reforms in the small-group market already enacted by the states and Congress to solve problems associated with risk selection and medical underwriting. Previously enacted laws require health insurers to sell coverage to all small employers seeking it, guarantee portability of coverage as people move from employer to employer, and limit variation in premium rates within defined ranges. Most people agree that these reforms have been essential, though not always sufficient, to make health insurance more accessible to higher-risk small employers that can afford reasonably priced coverage. Future reforms must be crafted to ensure that they do not jeopardize these gains.

Health Marts and Association Health Plans also should be assessed in light of what they add to existing small business collective purchasing arrangements. For example, existing HPCs must meet state mandates. Allowing Health Marts to avoid such mandates (which include such things as coverage for mental health and chiropractic services) would yield less expensive coverage. Such a step might be worth it if a significant number of uninsured people were to gain coverage. But if Health Marts don’t offer enough advantages over HPCs or attract enough uninsured people, they would lead to a shrinkage in coverage for some workers with health insurance without the offsetting gains for those who currently lack insurance.

The proposed legislation, while making new forms of collective purchasing arrangements available, does little to promote them. To encourage small firms to offer coverage and help them afford it, the federal government could provide funds to finance feasibility studies and start-up activities, as was done to encourage HMOs in the 1970s. Both federal and state government could provide various forms of financial incentives for employers to participate in collective purchasing arrangements, such as offering tax credits. A more regulatory approach would be to require small employers that choose to purchase health insurance to do so through a collective purchasing entity.

Unfortunately, even if all of these steps were taken, it is unlikely that the result would be a large reduction in the number of uninsured in the United States. Many small employers will not offer coverage even if they could buy it at the prices paid by today’s large self-insured companies. Even the most optimistic estimates of the impact of eliminating state mandated benefits or implementing Association Health Plans suggest that between 80 percent and 90 percent of the 43 million Americans who are uninsured today would remain uninsured.

Even so, small business health insurance purchasing arrangements are an idea still worth pursuing. They have the potential to be an element of a more comprehensive plan to expand health insurance coverage over time. That is why they have been a central component of several major health reform proposals during the last decade. As has been known for some time, solving the uninsured problem in a comprehensive way will almost certainly require some combination of government subsidies for low-income families, effective health care cost constraints, and mandated purchase of health insurance.

INTRODUCTION

In 1998, the U.S. House of Representatives passed two pieces of legislation that were designed to make the purchase of health insurance more affordable for small employers. One would have allowed the formation of “Health Marts,” and the other would have set the rules for the formation of “Association Health Plans.” Although the proposals did not appear in the final version of the budget bill, the ideas on which they are based have been around for some time and are likely to reappear in subsequent legislative proposals. The purpose of this report is to examine the advantages and disadvantages of these two approaches to facilitating the purchase of health coverage for small employers¹ and to show how they fit with other efforts to improve the performance of the health system.

THE HISTORICAL CONTEXT

It is useful to begin with some historical context. Small employers have been a focus of numerous insurance reform efforts for at least the last ten years, for three reasons. First, they represent a significant portion of the work force and a disproportionate share of uninsured workers. In 1997, workers in firms with fewer than 100 employees represented 32 percent of all workers ages 18 to 64. Sixty percent of these 42.6 million workers obtain health coverage through their employer or their spouse’s employer, but 28 percent are uninsured. These uninsured employees in small firms account for 49 percent of all uninsured workers.² (See Figure 1.)

Third, small firms, because of their size, are much more likely than large firms to face high premiums and to experience large premium fluctuations (mostly increases) from year to year. To understand why this is so, it is useful to recall the fundamental principle of insurance.

The purpose of insurance is to spread risk. Individuals cannot predict when they will experience a serious medical problem and thus have to bear a high cost. So they enter into an arrangement with other similarly situated individuals. Each agrees to trade the certainty of incurring a relatively small loss-the payment of a premium into an insurance pool-to avoid the chance of incurring a large unpredictable loss resulting from a serious episode of illness. In effect, the people who happen to be healthy in any year subsidize the people who happen to experience medical problems and require care. Eliminating risk in this way is possible because of the law of large numbers. What is unpredictable for one individual becomes predictable for a large enough group. An individual cannot accurately predict if he or she will have a heart attack, but actuaries can quite accurately predict what proportion of a thousand people will have a heart attack.

Large businesses have a large enough work force to be able to benefit from the law of large numbers. The proportion of people experiencing serious illnesses is likely to be relatively constant from year to year, so they can self-insure-that is, assume the risk themselves-or, if they buy insurance, their year to year premiums are likely to be relatively stable. Small firms, those with fewer than 50 workers, do not have these advantages. They cannot safely self-insure, so they have to buy coverage. And the insurers’ costs of marketing and selling coverage to small firms is relatively high because of the diseconomies of small scale-agents’ commissions, high advertising costs, the high turnover from year to year, etc. More importantly, the variation of risk from employer to employer is likely to be quite large, and the change in risk from year to year for any particular small group is also likely to be substantial. These characteristics create very strong incentives for insurers in the small-group market to try to insure low-risk groups and avoid insuring high-risk groups-by selecting low-risk groups initially, and by refusing to renew their coverage if they become a high-risk group.

To see how strong this incentive to risk select can be, it is useful to recall the rule of thumb that, in any year, about 10 percent of any large population group accounts for about 70 percent of the medical expenses for the group. Oversimplifying, that means that if an insurer that is contemplating insuring 100 (representative) people can avoid covering just one of the 10 people who will be sickest, the insurer can save approximately 7 percent in total medical claims. If the insurer can identify and exclude half of the people who will be very sick (that is, 5 people), the insurer can reduce claims costs by about 35 percent.³ In short, by far the easiest way for insurers to keep premiums down, attract new business, and make profits is to avoid insuring high-risk individuals and high-risk groups. And, of course, small groups with below-average risk have very strong incentives to seek out an insurer that will group them with other below-average-risk small groups and charge them a lower premium. The problem is that when these people leave the larger insurance pool, the rates for those remaining in the pool rise. Although the remaining people in the risk pool will have a higher average risk, there will still be differences, and those with lower risk have incentives to separate themselves out. This process of segmenting risk continues until the principle of spreading risk has essentially been destroyed.

Before the days of small-group insurance reform, which limited insurers’ ability to segment risk, this was what happened. Insurers had strong incentives to use resources to do medical “underwriting,” that is, to employ strategies to distinguish between high-risk and low-risk groups and charge premiums and make decisions to accept, reject, or terminate the group based on an assessment of expected claims experience. Insurers refused to sell coverage at all or charged very high rates to high-risk groups. They might cover the group but only if high-risk individuals were excluded, or they would cover the individual with a potential medical problem but exclude coverage for any pre-existing condition that might prove to be expensive. Rates were unstable; a group that had previously had low rates but experienced high claims in a subsequent year might find its rates raised dramatically, or it might be denied coverage entirely. Premiums were prohibitively high for some small groups and very unstable for most groups (since a group that underwriting identifies as a low-risk group often becomes a high-risk group later with the “aging” of the group). The system seemed arbitrary, capricious, and unfair: even small groups that could afford reasonably priced insurance and had been paying for coverage for years might, through no fault of their own, find themselves unable to afford coverage because one or two members of the group became seriously ill.

It is easy to blame insurers for employing the tactics that produced these problems. And it is true that some were especially aggressive and creative in devising techniques to avoid higher-risk groups. But once the process starts, it is very difficult for insurers to stay in business unless they, too, adopt similar techniques. Nobody can survive by taking on a seriously disproportionate number of high-risk people.

The problems of the small-group health insurance market prior to reform can be summarized as follows:

Higher-risk groups could not get coverage or could do so only by paying very high premiums.
The system seemed unfair because people were denied coverage or had to pay very high insurance rates as a result of circumstances largely beyond their control, in particular, a deterioration in health status.
Insurers and health plans were rewarded for using resources to become efficient at selecting low-risk groups rather than for being efficient at managing the delivery of medical services.
Premiums were unstable and unpredictable from year to year.
Premiums were high because of the high marketing and administrative costs insurers incurred in serving the small-group market.

THE REFORMS

Small-Group Insurance Reform: Spreading Risk

The reform efforts of the last decade have been aimed at correcting the problems of the small-group market. The most pressing need was to restore the risk pool-to curtail risk segmentation and reunite the low-risk and high-risk groups. Achieving this objective has been the primary intent of the laws passed by states over the last ten years or so, and subsequently by the federal government, to reform the small-group health insurance laws. In general, these laws have included the following elements:Guaranteed issue – Insurers operating in the small-group market are not permitted to deny coverage to any small employer regardless of the risk associated with the group seeking coverage.Guaranteed portability – Once an individual is covered and so long as he or she remains insured, the individual cannot be denied coverage for a condition, no matter how serious, if he or she changes employers.Limits on exclusion for pre-existing conditions – An insurer can exclude coverage for a particular condition for a newly insured person, but the period of exclusion is limited to a reasonable period of time – one year or less. Once the person satisfies the pre-existing condition “waiting period,” the exclusion cannot be re-imposed even if the person selects new coverage (for example, because he or she takes a new job).Fair marketing rules – Insurers are required to market their products in a way that does not discriminate against high-risk groups and systematically attract a disproportionate number of low-risk groups.Rate reform – In setting premiums, insurers are limited to using only certain factors, and the amount of variation in premiums is limited. In particular, the variation in rates based on health status or past medical claims experience is limited. The extreme form of this reform is community rating, which requires insurers to charge the same rates to all small groups, although modified community rating is more common. Modified community rating typically allows some rate variation for age and perhaps gender or geographic location. (It is important to note that without rating reform, a guaranteed issue requirement is largely meaningless, since an insurer could charge a high-risk group such a prohibitively high rate that the group could not buy coverage even though it was technically available.)

The culmination of this legislative effort was passage in 1996 of the federal Health Insurance Portability and Accountability Act (HIPAA), which began to take effect in mid-1997. It establishes certain minimums which states must meet with respect to guaranteeing access to and portability of coverage. It does not address limits on variation in premiums, leaving that up to individual states.

Insurance reform is probably the most important ingredient in a strategy to make health coverage more accessible and more affordable for small businesses. Without the regulations that severely constrain health plans’ ability to segment groups by risk, including limits on variation in premium rates, the market cannot continue to work properly because higher-risk groups will be shut out. People disagree about the extent to which groups should be pooled. For some, the ideal is complete risk spreading: all small employers should pay the same regardless of their need for or utilization of medical services. Others favor allowing various degrees of grouping by risk, with premium rates varying with the amount of risk. The argument reflects both philosophical differences about the degree to which the healthy should subsidize the less-healthy and disagreements about practical issues relating to who will be left uncovered.

Whatever one’s position on that, it is clear that expanding access to health insurance and making coverage affordable for higher-risk small business groups is not possible without significant limits on insurers’ ability to separate groups by risk. Limits on price variation are also essential.⁴ It is important to keep this in mind when considering other reforms, including health purchasing cooperatives, Health Marts, and Association Health Plans. If these strategies for aggregated purchasing create opportunities for lower-risk groups to segregate themselves from the larger pool of small employers, the spiral of risk segmentation, widening rate variation, and market breakdown is very likely to recur. In seeking additional ways to improve coverage for small business, we need to make sure we preserve the gains already made.Collective Purchasing Arrangements

Long before states enacted small-group insurance reforms, small employers were trying to find ways to decrease their health insurance costs. A common approach was to have many small employers aggregate their purchase of insurance by forming some kind of association to buy as a group. In general, the intent was to give small employers, acting together, the purchasing advantages that large employers enjoy.

One objective of purchasing collectively was to pool risks, but this was not likely to be sustainable over long periods of time, for reasons already noted: the lower-risk members of the association had an incentive to find less expensive insurance elsewhere, perhaps by forming a separate association of their own. Besides, today’s small-group insurance reform essentially achieves the objective of pooling risks without the need for forming collective purchasing arrangements. But there are other potential benefits of collective purchasing:

Leverage in negotiating with health plans – A buyer that commands a large market share has more clout in bargaining with a supplier, in this case the health plan, than does a small buyer. The seller does not want to lose such a large customer to a competitor and will act to avoid that outcome. By purchasing collectively, small employers may be able to negotiate a lower premium or cajole a health plan into making changes to become more efficient, improve quality, or provide better service
Economies of scale – Administrative costs related to the sale of coverage and after-sale service might be reduced. It is less work and less costly to sell coverage once to a single entity that represents 500 small employers than to sell coverage to those 500 employers separately. The same is true for administration and servicing of the contract (if the association handles these tasks through a centralized process).
Employee choice of health plans – Few small employers can afford to offer their employees a choice of health plans because the administrative expense and burdens are too high. But choice becomes more important as managed care plans come to dominate the health plan market. Being in one plan rather than another can affect an employee’s choice of providers, style and level of service, and quality of care. Being able to offer employees a choice of plans is seen as a significant benefit by both employers and employees. Small employers can cost-effectively offer such choice if they buy coverage through a collective purchasing arrangement because the purchasing entity realizes economies of scale in having contracts with a number of health plans.

Collective purchasing arrangements have taken a variety of forms to date. These include, most prominently, “multiple-employer welfare arrangements” (MEWAs), Health Purchasing Cooperatives (HPCs), and association plans. The proposed Health Marts and Association Health Plans are both conceived in this tradition. We discuss the evolution of these arrangements below.Associations and MEWAs

As noted, employers have long been attracted to the idea of banding together to buy health insurance, as well as to provide other benefits. Existing trade or business associations that were formed for other purposes, such as Chambers of Commerce or professional associations, commonly undertook this additional function to serve their existing membership. In other instances, associations were formed solely for the purpose of providing insurance to a particular group of employers (though some states have laws prohibiting such “fictitious groups” from forming just to buy insurance).

One motivation for the establishment of such association plans was economies of scale. A second motivation for some groups-namely, those composed of employers with below-average health risk profiles-was being able to buy lower-priced coverage reflecting their lower risk. A third motivation for forming self-insured associations is freedom from regulation by states as an insurer, including “avoidance of state benefit mandates, premium taxes, contributions to risk pools and guaranty funds, rating requirements, and solvency standards.”⁵ (More will be said later about problems related to self-insured associations.)

It is difficult to determine how many employers buy coverage through association-like organizations. One reason is that they can take many forms: associations may include in their membership both large and small employers and even individuals. They may be fully insured-that is, they may buy insurance coverage from a risk-bearing insurance carrier that is licensed and regulated as such by the state-or they may be self-insured or self-funded-that is, take on the risk internally (though an insurer may administer the plan). The consequence is that no one knows for sure how many employers buy association-based coverage. A 1993 survey asked employers if they had considered buying coverage through “a local business association, trade group, employer coalition, or some other groups of small purchasers.” Of those providing health coverage, 59 percent indicated they that had considered this option, but only 17 percent said they currently purchased coverage through such an organization.⁶ Several state insurance regulators who were informally polled about the extent of association coverage indicated that from 5 percent to over 50 percent of the small-employers purchase coverage from associations.⁷ (The wide variation reflects, in part, different definitions of association plans.)

The controversies surrounding association plans have resulted from ambiguities about states’ authority to regulate them. State insurance regulators regulate licensed insurers to ensure that they can deliver on their promise to pay health care claims and not go bankrupt and leave insured enrollees with no coverage and no way of paying already incurred claims. Among other things, regulators make sure that insurers collect enough in premiums to cover medical claims and other costs, and as an additional safeguard, require insurers to maintain certain levels of reserves and/or buy reinsurance to cover unexpectedly high medical claims. And, of course, they monitor insurers’ activities to prevent fraud and other forms of financial misconduct.

In general, state insurance regulators have powers to protect members of associations when the associations buy health coverage from risk-bearing insurers, since states regulate insurers. The problems arise when association plans are self-insured. Under court interpretations of the Employee Retirement Income Security Act of 1974 (ERISA), state insurance officials cannot regulate health coverage provided by self-insured employers. This regulatory loophole created many problems with association plans. As noted earlier, there are numerous reasons why associations would be motivated to self-insure, and many chose this approach. But the consequence was that without the discipline required by state insurance regulation, many association plans became insolvent in the 1970s and early 1980s, leaving hundreds of thousands of people stranded without coverage and providers with unpaid fees. Some went under because of bad management and financial miscalculations. Others were started by unscrupulous entrepreneurs whose only incentive was to make a “quick buck” and get out without any concern about the plight of those covered under the association plan.⁸

The rash of failures led to a 1983 amendment to ERISA which gave states authority to regulate self-insured MEWAs. These were defined by the amendment as an employee welfare benefit plan or similar arrangement, including a health plan, established or maintained for the purpose of offering or providing benefits (other than pensions) to employees of two or more employers.⁹ Only self-insured plans established or maintained by a union or a single employer remained exempt from most state insurance regulation. In general, this means that associations that are both self-insured (that is, take on risk) and serve multiple employers are defined as MEWAs and are thus subject to state insurance regulation.¹⁰

Although the increased regulatory authority states have over MEWAs (and thus self-insured associations) has reduced the problems, the problems have not evaporated. States have not adopted uniform ways of regulating MEWAs, in part, because of continuing uncertainties about the extent of their regulatory powers. Some mismanaged and fraudulent associations continue to operate. Many association plans probably insure a variety of groups without the knowledge of state or federal regulators. Some associations try to escape state regulation by setting up a sham union or a sham employer association, self-insuring, and then claiming that they are not a MEWA and are thus not subject to insurance regulation. The consequences are sometimes disastrous for the people covered by these bogus schemes.

There are other problems. Some states have permitted association plans to be exempt from small-group market reforms, thereby creating incentives for low-risk businesses to band together to get lower-cost insurance coverage. Insurance regulators are sometimes uncertain whether an association licensed in one state but operating in other states is subject to the small-group reform laws in each state in which it sells coverage or need conform to only the laws of the state in which it is “domiciled.”¹¹

To summarize, association plans raise two general categories of problems: (1) If they bring together people who have below-average risk and exclude others and are not subject to state small-group rating rules, they draw off people from the larger insurance pool, thereby raising premiums for those who remain in the larger pool. (2) If they are not subject to appropriate insurance regulation to prevent fraud and ensure solvency and long-run financial viability, they may leave enrollees with unpaid medical claims and no coverage for future medical expenses.

Of course, many association plans have operated well, without financial difficulties or attempts to circumvent the law, and perform an important function for their members. Some, like COSE (the Council of Smaller Enterprises) in Cleveland, which enrolls 12,000 small and mid-sized businesses and insures 190,000 employees and dependents, have opened their membership to small businesses in general and thus bear a close relationship to the next category of collective purchasing arrangement, the HPC.¹²Health Insurance Purchasing Cooperatives and Coalitions

The Concept

In the late 1980s and into the 1990s, a number of initiatives evolved to form health purchasing cooperatives or health purchasing coalitions (HPCs). Like association plans, HPCs were designed to realize the advantages of collective purchasing. But they differed from previous collective arrangements by contracting with health plans to provide coverage to any small-employer applicant. The HPCs did not themselves take on any risk; they offered only fully-insured health plans and products. Unlike previous collective purchasing efforts, essentially the only criteria for eligibility to participate was being below the firm size limit (generally 50) and being willing to pay the premium.

Small-group insurance reform made such unrestricted eligibility requirements possible. A HPC could open up coverage to any small employer, regardless of the group’s risk profile, without extensive underwriting, and at premiums that vary within a relatively narrow band. But they could do this only because health plans selling outside the HPC were required to accept all applicants and limit the premium differences to the same rate range. In other words, the rules were essentially the same for all. Without these rules and restrictions, the insurers would drain off all the lower-risk firms, leaving the HPC with just the expensive, high-risk employers. HPCs cannot afford to be significantly less restrictive in their eligibility and rating practices than insurers outside the HPC. Market reforms are a prerequisite to establishment of HPCs of the type described here.

HPCs sought to achieve other objectives by structuring themselves in particular ways. To foster price competition among health plans, most HPCs offered only a few standardized benefit packages that all contracting health plans were required to offer. Because benefits were identical, people choosing among plans could concentrate on comparing levels of service, convenience, accessibility, and quality. The price-value comparisons were easier for consumers to make, and as a result, health plans were under greater pressure to compete by becoming more efficient. Standardization of benefits also prevented health plans from structuring benefits to try to attract primarily lower-risk people into their plans.

Another key feature of most HPCs was employee-choice. Employees rather than employers decide which health plan they will enroll in. The concept is straightforward: employees are best suited to decide which plan matches their needs and preferences. Moreover, if they change employers, and the new employer also participates in the HPC, the employee can continue enrollment in the same health plan, without a disruptive change in providers. HPC supporters felt the employee-choice approach also created the right kind of competitive incentives: health plans that met the needs of employees (not employers) would be rewarded with higher enrollment.

Virtually all of the HPCs saw themselves as representing the purchasers-employers and employees, the people who pay for coverage-but they differed with respect to their position on negotiating with health plans about price. Some, like the Community Health Purchasing Alliances (CHPAs) in Florida, accept any health plan willing to meet the specifications for coverage, leaving it up to the health plan to offer a competitive price. Others, most notably the Health Insurance Plan of California (HIPC), aggressively seek price discounts from health plans, just as many large employers do. Some supporters of rate limitations and some insurance commissioners have reservations about giving HPCs this power. They fear that allowing this exception to the general rule that all small employer coverage should be priced according to rating reform rules could threaten the principle of rating reform. The supporters of HPCs argue that this practice poses no threat to the principle of modified community rating so long as the HPC is open to any small employer. Any price advantage a HPC gains through negotiation would then be due not to favorable risk selection but to discounts related to bulk purchasing or other forms of efficiency. They would be employing the same kind of bargaining leverage that is available to large employers; in effect, the aggregation of small employees can be seen as a large group. Not everyone was persuaded by this argument. Some states, such as Florida, continue to prohibit HPCs from negotiating on price. Others, like Colorado, permit variation only for the administrative component and not the medical component of the premium.The Controversy

If measured in terms of visibility and controversy, the peak of the HPC idea was its inclusion in various proposals put forth in the early 1990s to reform the health care system. HPCs (by several different names) were in fact a central component of many comprehensive health reforms proposals in that period, and especially those expounded by advocates of “managed competition.” Professor Alain Enthoven at Stanford, for example, included the HPC concept-though he used the term “sponsors”-in various widely read and highly influential articles and speeches promoting a set of principles to make competition more robust in health care markets. Others elaborated on the concept. At the time, there was much controversy about several key HPC issues.

One issue was whether small-employer participation in a HPC should be voluntary or compulsory. Those who favored requiring all small employers who offer coverage to buy it through a HPC argued that this approach would ensure economies of scale, guarantee that health plans would participate, give the HPC greater bargaining power, and greatly reduce the potential for risk segmentation. The opponents generally argued that compulsion was inherently undesirable and in this case unnecessary.

A second and related issue was whether there should be only one HPC for each market or several competing HPCs. The arguments for a single HPC were essentially the same as those for making participation compulsory-economies of scale, reduced opportunities for risk segmentation, and bargaining power with health plans. The counter argument was that having only a single supplier with no competition would result in bad performance-poor levels of service, inadequate attention to customer needs and preferences, and so forth.

A third issue was the size of firms that would be permitted-or required-to purchase coverage through the HPC. The argument for making it exclusive to small firms, usually defined as those with fewer than 50 or 100 workers, was that these were the firms that the current insurance market served poorly. Moreover, requiring large firms to participate would stifle the experiments that a number of large employers were sponsoring as part of their efforts to cut costs and improve value. The arguments for including larger firms were again bargaining power, economies of scale, and other efficiencies.

A fourth issue was whether the HPC could negotiate over price or whether it would be a “price taker,” simply accepting all willing health plans and letting the consumers-the employees or employers-decide which plans offered the best values. Those who favored giving the HPC bargaining power wanted HPCs to be selective in choosing plans and to be purchasers in the way large employers were already purchasers, using that power to force health plans to offer better value. The “price takers” feared such bigness and thought the choices of individual consumers would provide sufficient competitive impetus to force health plans to improve performance.

The HPC idea received greatest attention in the debate over President Clinton’s health care reform proposal in 1993-1994. The architects of that plan made the collective purchasing idea the foundation of their proposal. In fact, initially, they would have required all firms with fewer than 5,000 employees to purchase coverage through a HPC-like organization in each state. These new organizations would have been given a number of health care budgeting and regulatory responsibilities. The controversy aroused by this proposal was intense. Alternative, less sweeping HPC proposals were offered by various members of Congress. Although the administration seemed open to more moderate versions of comprehensive reform so long as they incorporated the basic principles of their proposal, the initial iteration raised so much opposition and created so many opportunities for opponents to attack the proposal, that the public support for fundamental reform dissipated.Current Practice

But the HPC idea did not die when health reform died. Many efforts were made, and continue to be made, to implement that concept of group purchasing open to all small employers.

HPCs have been both publicly and privately sponsored. The Health Insurance Plan of California (HIPC) has operated as part of a public agency (though it will soon be transferred to a private entity, the Pacific Business Group on Health). HPCs in several states, such as Florida and North Carolina, were established by state law, though they were organized as non-profit entities outside of government, often initially with state start-up funds. In other instances, such as the HPC of the Connecticut Industry and Business Association, the HPC operates without any HPC-specific state legislation and is entirely private.

The actual experience with HPCs has been mixed. Some have been moderately successful if measured in terms of enrollment. In California, HIPC covers about 148,000 people¹³ and the HPC of the private Connecticut Business and Industry Association covers about 56,000 lives.¹⁴ California’s HIPC reports that about 20 percent of its enrollees were uninsured immediately prior to enrolling. Other HPCs have floundered, finding few health plans willing to participate and attracting only a small number of enrollees. Even the most successful do not account for a large share of the small-group market. In California, for example, the state-sponsored HIPC has about 2 percent of small-group employers in the state. Altogether, there are probably only about 20 to 25 HPCs operating or near the operational stage. Although it is difficult to pin down the number of people enrolled in HPCs because of widely varied definitions of what constitutes a HPC, at least 540,000 people are covered by HPC-like organizations.¹⁵

The limited success of HPCs can be attributed to several factors. Early proponents often thought that one way to realize economies was to reduce or eliminate the role of agents, reasoning that the HPC itself could perform the marketing and sales functions and thereby avoid the considerable cost of agent commissions. Hindsight makes it clear that alienating agents was a mistake because without agent participation it is extremely difficult to reach small businesses. Small employers depend on agents for information and advice; few small businesses seem to take the initiative to buy directly from a HPC. Finding ways to market the HPC idea and attract small employers has been a major challenge.

Getting health plans to participate has also proved more difficult than expected. Instead of seeing the HPCs as a new source of business, many health plans have viewed them as a threat to their existing small-employer business. They do not see any advantage to putting themselves in the kind of direct head-to-head competition that the HPC involves. And because the HPC does not command a large market share, the plans are able to decide not to participate. This is a basic dilemma for the HPCs. Without a significant market share, they cannot bargain effectively to get lower prices. And without lower prices, they cannot attract enough employers to have an effective market share.

Other problems for HPCs include state laws that are not hospitable to effective HPCs, lack of any group willing to initiate and finance HPC startup, and ineffective organizational structures.

In summary then, a number of HPCs exist and are functioning successfully. But many HPCs have struggled or been only moderately successful and most places in the country lack a HPC.Health Marts

Health Marts, as outlined in HR 4250, the House-passed version of the Patient Protection Bill of 1998, represent a similar concept but with some important differences. (Future iterations of the legislation may, of course, be different.)

In general, the Health Mart concept is less narrowly defined than a HPC, presumably to encourage more efforts to form such collective purchasing arrangements.

Health Marts would have to offer at least two insured options, such as a PPO and an HMO (though perhaps through just one insurer) to all small employers within a defined geographic area (no smaller than a county). The Health Mart itself may not assume risk. Rate variation would be permitted only within the bounds determined by state law. Health Marts would provide administrative services, such as billing, enrollment, etc., and they must provide information about benefits, satisfaction, and quality to enrollees. In these respects, then, Health Marts are similar to classic HPCs: they can contract with insurers, do not assume risk, can presumably be selective in accepting or rejecting health plans, must accept all small employers, and must conform to state small-group insurance reforms, including rating constraints.

But Health Marts differ from classic HPCs in several important respects. Governance would be conducted by a board of equal numbers of representatives of employers, employees, health plans and insurers, and health care providers. Thus the Health Mart is not solely an agent of the purchasers, that is, the employers and employees. Since the legislation is silent on the issue, Health Marts would not necessarily adopt the employee-choice model and would not necessarily offer standardized benefits.

The proposed legislation would permit Health Marts to be free of state mandates that require insurers to cover certain benefits or providers. States would also not be permitted to preclude the formation of Health Marts. “Fictitious group” laws, which prohibit formation of groups solely for the purchase of buying health insurance, would be superceded by the federal Health Marts law.

On the other hand, the law would not alter states’ powers with respect to regulating insurers’ regarding licensure, solvency, premium taxes, payments to guaranty funds or high-risk pools, fair marketing practices, or premium rating. State patient protection laws would presumably also apply to health coverage available through the Health Marts.

General administrative oversight and the responsibility for issuing implementing regulations would be assigned to a new Health Care Marketplace Division of the U.S. Department of Health and Human Services. Health Marts would be required to submit information demonstrating compliance with the law to this agency or, apparently at the option of this agency, to state regulatory authorities.Association Health Plans

As outlined in HR 4250, the House-passed version of the Patient Protection Bill of 1998, and in HR 1496, a bill submitted in April 1999 by Rep. James M. Talent (R-Mo.), Association Health Plans (AHPs) aim to achieve the same ends as HPCs, existing association plans and Health Marts. But under both pieces of legislation, AHPs represent an additional step. They would be able to self-insure and thus be exempt from many state insurance regulations.¹⁶ This would put them on a very different plane and path relative to existing HPCs. Both laws would amend the Employee Retirement Income Security Act of 1974 (ERISA) to permit AHPs to self-insure and obtain the same flexibility currently enjoyed by large self-insured employers.

Although AHPs would be required to offer an insured plan where available, they could offer self-insured products as well under certain circumstances. Thus, small firms buying through an AHP would not be subject to state mandated benefits and would not have to pay state premium taxes (since, like self-insured employers under ERISA, they would generally not be subject to regulation by state insurance commissioners). This would allow AHPs to offer less expensive health insurance for four reasons. (1) The benefits could be less comprehensive than is typical. (2) There would be no charge to cover premium taxes. (3) Many of the costs associated with complying with state insurance regulation would be eliminated. (4) Any savings associated with not having to pay an insurer to take on risk could be passed on to Association members.¹⁷ Some have claimed potential savings of as much as 30 percent.¹⁸

There is a fifth possible way in which an AHP could offer lower premiums: by developing strategies to confine its membership to lower-risk groups. This prospect may be the reason why some groups support AHPs as proposed in the legislation. It is what those who want to preserve pooling of risk in the small-group market most fear.

ANALYSIS OF HEALTH MARTS AND ASSOCIATION HEALTH PLANS

In assessing the utility of these new entities to facilitate the purchase of health coverage by small businesses, we apply two tests: (1) Can they achieve their intended objective (most notably to attract small firms that do not offer health insurance)? (2) Do they jeopardize past progress toward making insurance more affordable and more equitably accessible for small businesses (the “do no harm” test)? We consider these in turn, but first, we need to note something about this analysis.

We do not try to analyze every detail and nuance of these proposals as they appear in HR 4250 or HR 1496, for two reasons. First, this legislation is likely to be further modified in the future. Second, it would be impossible in a paper of reasonable length to analyze the complexities and implications of every provision that might have important consequence, depending on the exact way it is written and interpreted. What we seek to do, instead, is to point out the important issues that need to be considered in assessing any proposal that is based on the same design foundation as these two. We want readers to be forewarned about the potential areas of danger as well as to be able to understand what benefits might result. (Figure 2 is a comparison of some of the key features and conditions of HPCs, Health Marts, and Association Health Plans.)Achieving Objectives

It appears that the proposals for both Health Marts and AHPs seek to remove barriers to collective purchasing and to make coverage less expensive. The expectation is that the establishment of these organizational structures would lure more small groups into collective purchasing arrangements, and that they could then buy coverage more economically. Lower costs would result in a greater demand for coverage and, presumably, fewer uninsured small groups.Figure 2: Comparison of Features and Conditions of Health Purchasing Cooperatives and Legislatively Proposed Health Marts, and Association Health Plans

Feature or Condition	Association Plans	Health Marts	HPCs
Board represents purchasers	Presumably, since association is sponsor	No-purchasers (employers and employees), providers, insurers	Yes
Must accept all willing insurers and health plans	No	Presumably not	Varies
Can negotiate with plans over premiums, etc.	Presumably	Seems unlikely given board structure	Yes, for the most part
Offers multiple health plans (not just plan-type options)	Presumably not required, but must offer at least 1 insured option even if offers self-insured plan	Not required	Yes
Employee choice	Presumably not	Not required	Normally
Standardized benefits	Presumably not	Not required	Normally
Subject to state mandated-benefits laws	No	No	Yes
What group size limits?	Presumably none; determined by association membership	2-50	Varies, usually 50 and under, but same as size applicable to state insurance reform laws
Must take all small-groups that apply, regardless of health status and without underwriting based on health status	Within association membership only. Non-members excluded.	Yes	Yes
Subject to state rating requirements	No	Yes (except perhaps in non-home states)	Yes (though can sometimes negotiate rates)
Subject to other small-group insurance reforms	Only HIPAA, not state laws.	Apparently (but exempt from prohibition on fictitious groups)	Yes
Rates based on health status or claims experience	No	No	No
Accepts only groups or individuals also	Groups but also the self-employed.	Groups only.	Groups only.
Can define geographic service area	Yes. Presumably the same as association membership, often multi-state	Yes (though must be by county, though not contiguous)	Usually whole state
Take on risk, that is, self-insure	Yes	No	No
Must meet reserve requirements and other protections to ensure coverage is not lost if financial difficulties develop	Yes, as specified by federal law.	No (because not self-insured, do not take on risk)	No (because not self-insured, do not take on risk)

Covering the Insured or the Uninsured

It is important to make a distinction here between success in attracting customers who are already insured and drawing in small groups that would otherwise be uninsured. An argument can be made that these proposals should be judged primarily on their capacity to reduce the number of uninsured.

Both employers who currently offer health insurance and those who do not are likely to purchase coverage from a new collective purchasing organization primarily because it offers a price advantage over existing sources of coverage. The price advantage can have three sources: greater efficiency, less comprehensive coverage, or coverage of below-average-risk groups (favorable risk selection). If a collective purchasing organization attracts customers because of greater efficiency, that is a clearly socially desirable outcome, whether or not those new members are currently insured. However, if groups buy coverage from a new entity because they do not have to pay for mandated benefits, the change would be hard to justify unless the new members would generally have previously been uninsured. Otherwise, there would be no real reason to override state mandated benefit provisions, especially for just some groups. If groups buy coverage from the new entity because its price advantage results from attracting primarily below-average-risk groups, that is not a desirable social outcome, given the purposes of previous reforms.

Let us consider how various collective purchasing arrangements achieve cost advantages:

HPCs can offer greater efficiency related only to economies of scale; they have to cover all mandated benefits, and they cannot risk select because they accept all small businesses regardless of risk. Even if they attract primarily people who already have coverage, a desirable social end is served.
Health Marts can go beyond HPCs by offering less comprehensive coverage (by not covering mandated benefits). If Health Marts attract primarily insured groups, the same social efficiency-related objective is achieved as with HPCs but at what some would say is the cost of overturning mandated benefit laws. If they do not attract uninsured groups that HPCs would not attract, Health Marts offer no advantages over HPCs and entail the social cost of overturning mandated benefit laws in an inequitable way, that is, for just some groups.
Association Health Plans go further than HPCs or Health Marts by offering some additional possible savings that result from self-insurance. (Whether these savings-such as not being subject to the same degree of insurance regulation-represent real social efficiencies is debatable. Moreover, some research indicates that the administrative costs of self-insured plans are higher than for conventional plans.)¹⁹ But they also have the potential to include only below-average risk employers because, unlike HPCs and Health Marts, they are not required to provide guaranteed-issue coverage to all small employers but only to those that qualify for membership in the association. If an Association Health Plan does not attract the uninsured but instead covers mostly insured people who would not otherwise choose a HPC, the likely reason is that the Association Health Plan allows them to escape benefit mandates or that it is attracting groups of below-average risk. The first outcome, as already observed, is not seen by everybody as having social utility. The latter outcome raises the price of insurance for all those remaining in the risk pool and jeopardizes the good results already achieved through small-group market reform.

If this reasoning is correct, Health Marts and AHPs should be judged primarily on their capacity to attract currently uninsured firms. Unless they can attract large numbers of uninsured groups that would not be attracted to HPCs, it is hard to justify them as achieving any desired social objective that HPCs cannot achieve with fewer potential dangers.The Prospect for Covering the Uninsured

In assessing their ability to attract either the currently insured or those without coverage, the first thing to note is that small employers will not necessarily flock to such new organizational structures, even if they could potentially achieve savings. The experience with HPCs is instructive: no HPC accounts for a large market share. The California HIPC, for example, was one of the first, is perhaps the most attractive and successful in the nation, and at least initially offered lower rates than were otherwise available. Yet it accounts for only about 2 percent of the small-group market in a state where people are used to innovative health care financing arrangements. As noted earlier, one of the dilemmas that these collective purchasing arrangements face is that they can achieve cost savings-by realizing economies of scale and by using negotiating leverage-only if they are large. But they cannot attract enough groups to become large unless they can offer a significantly lower price.A Different Governing Structure

The designers of the new organizational arrangements have tried to make them more attractive than existing collective purchasing efforts in several ways. Health Marts require the governing board to include all the stakeholders-employers, employees, providers, and insurers. The apparent reasoning behind this tack is that cooperative efforts may be more likely to be successful than confrontational efforts. Current HPCs normally represent only purchasers and bargain to get better prices from insurers and thus, ultimately, providers.

But past experience suggests involving a broader number of stakeholders may not work. Federal legislation in the mid-1970s mandated that states set up mechanisms for health resources planning. State and local planning agencies were responsible for assessing their needs. A primary tool they could use to implement those plans was the power to approve or disapprove proposals for significant capital expenditures, primarily hospital equipment and buildings, by granting or denying a “certificate of need.” The planning agencies were governed by (or in the case of the state agency, advised by) boards made up of all the major stakeholders-consumers, employers, providers, insurers, state agency officials, relevant trade association representatives, etc. The results were disappointing if measured in terms of creating real changes in the health system. Even though consumers were in the majority, the providers and others with professional staff to support them were often able to exert a decisive influence to block actions they disliked. The experience showed that a consensual model that includes all the stakeholders does not produce very far-reaching results; there are too many conflicts of interest and too many ways to thwart controversial policies. To give an obvious example of the problem relevant to Health Marts, why would providers and insurers on the board support an aggressive bargaining posture in seeking to get a good price on coverage for a Health Mart when the result would be lower income for providers and insurers?No Mandated Benefits

Another way to make collective purchasing more attractive is to offer a lower-priced product. Legislative proposals would allow Health Marts and AHPs to sell at a lower price, in part, by not requiring that coverage include the benefits mandated by state law (although they would be required to cover mandated diseases). If the purchasing organizations offered, and small employers bought, benefit packages that offered less comprehensive benefits, premiums could be reduced compared to the cost of benefit packages now available to these groups. These new collective purchasing arrangements would be free to offer virtually any benefit package they wished, including perhaps one that was essentially a catastrophic plan, with high deductibles and cost sharing.

Organized small business has long objected to mandated benefits. They argue that small employers, just like large self-insured businesses, should have the freedom to decide which services they will pay for. What is fair for these large employers should be fair for small employers, they argue. Speaking from the workers’ perspective, other critics contend that employees-who ultimately pay for the additional coverage in the form of foregone wages-should be able to choose the mix of wages, health benefits, and other employee benefits unencumbered by rules that establish minimum levels of coverage.

But there is a more important argument for freeing small employers from mandates. Many small businesses are on the margin of being able to afford coverage. The opponents of mandated benefits believe that if such firms had the option of deciding to buy coverage without paying for mandated services, more would be able to buy coverage initially and fewer would be forced to drop coverage when premiums rise. Of course, some firms would choose to buy coverage that includes many of the mandated benefits, but for others the option would be to forego mandated benefits or to forgo coverage altogether. The small-business critics of mandated benefits argue that no one is served when, because of mandated benefits, a business has to decline coverage that it could otherwise afford. It is better to have limited coverage than no coverage.

This criticism of mandates is supported by some research. One recent study, using data from the Current Population Survey, concluded that eliminating all mandated benefits would cause the number of uninsured adults to drop by slightly more than 20 percent.²⁰ Another soon-to-be-published study that looked at employers with fewer than 50 workers, estimated that 18 percent of uninsured small firms would be insured if there were no mandated benefits.²¹

Critics of mandated benefits, contend, moreover, that many mandated benefits are on the books for reasons which have little to do with what is critical to patients. Many services are mandated by states because a variety of self-interested provider groups have been able to effectively pressure legislatures to include the services in insurance coverage. Others reflect the political potency of disease-specific advocacy groups. It would be hard to show, the critics believe, that all mandated benefits are necessarily those that are most cost-effective or those that consumers believe are most essential.

While there is intuitive appeal to this line of reasoning, the expectation that lean benefit packages will entice many new businesses to buy coverage may prove unrealistic. Past experience shows that small employers do not find “bare bones” benefit packages attractive. They simply do not sell. Many states that adopted small-group market reforms required that insurers offer on a guaranteed-issue basis “standard” and “basic” plans. Although the defined set of benefits differed from state to state, the packages generally did not include all state-mandated benefits and were less comprehensive than typical plans otherwise available. The experience is consistent: the “bare-bones” basic plans sold very poorly, and even the more generous and slightly more popular standard plans usually accounted for only a tiny portion of the total market. In some instances, this was true even when the standard plan was not significantly different from the average “street” plan, but the perception was that coverage was inferior. Both products were typically bought only by groups that could not get normal coverage because they were at high risk²² (before HIPAA required that all small-group products be sold on a guaranteed-issue basis). An exception to the poor sales record is Colorado. In that state, the standard plan sold well, but the benefits were essentially the same as those typically offered by HMOs. (Most states continue to offer their standard and basic plans even though all small-group products are now provided on a guaranteed-issue basis because of HIPAA requirements. Thus insurance that is exempt from many of the mandated benefits requirements is already available in many states, and it still is not selling well.)

Another indication that substantially slimmed-down products may not appeal to employers comes from an assessment of the packages that current self-insured employers offer. In general, benefits do not appear to be significantly different from those offered by insured indemnity and PPO plans, though they are somewhat less generous than typical HMO coverage.²³ Of course, the comprehensive coverage among self-insured employers may reflect the fact that many are larger and thus find comprehensive coverage more affordable and necessary to recruit a skilled workforce.

This evidence from state efforts to offer scaled-down benefit plans, along with the experience of self-insured employers, suggests that most small employers are not willing to buy coverage that is significantly less comprehensive than what is generally available, even if there is a price advantage to doing so. More importantly, it may be that the savings are not sufficient to draw into the market many firms that find normal coverage to be unaffordable. Thus freeing Health Marts and AHPs from the obligation to cover mandated benefits may not attract many additional groups into the health insurance market.

Although there is little controversy about finding ways to lower insurance costs by improving efficiency, a policy to lower costs by excluding coverage for mandated benefits is more controversial. Some people object because they think that legislatures mandate benefits for a good reason, namely, that people should not be without coverage for certain key services, like mental health or substance abuse services, which they may undervalue until they need the service. Another concern is that such an exclusion from mandates creates an unfair competitive environment when some insurers must cover mandated benefits but others need not do so, which clearly gives the latter a government-mandated price advantage not related to any form of efficiency. (It is worth noting, however, that only about a quarter of the population in a state with mandates is subject to them;²⁴ others get coverage through public programs or ERISA plans that are not subject to mandates.)Self-Insurance

The 1998 House-approved proposal for AHPs sought to gain market share for collective purchasing by allowing organizations that already have significant small-business membership to offer members health coverage that is more attractive than what they could get elsewhere. The key provision that makes it more likely that associations would offer such plans is the possibility for self-insurance, a lower-cost option that is not available now. Since many small employers already have a relationship to associations, an AHP could be thought to have an advantage in marketing to them. But the question is whether many of those who take up the offer would otherwise not be covered. Do many small employers that belong to associations not have coverage at present, and of those who fall in this category, would they buy coverage if it became available? If most of the firms that buy newly available association-sponsored coverage would be covered anyway, nothing much is accomplished in terms of reducing the number of uninsured, although the buyers would benefit from lower premiums.Impact on the Uninsured

On balance, it is not clear that Health Marts or AHPs would do much to entice small groups not now covered into the insurance market. Inertia is a powerful force, and a rush to collective purchasing is unlikely unless the advantages over present arrangements are large and visible. Apart from the perhaps significant prospect of reducing the price somewhat by overriding state mandated benefits laws, Health Marts do almost nothing to encourage collective purchasing. In fact, except in states that prohibit the formation of “fictitious groups” (groups formed solely for the purpose of buying insurance), a Health Mart could be started now (though coverage would, of course, have to include mandated benefits). The legislative proposals provides no additional financial incentive, start up funds, technical assistance, or other inducements beyond what already exists. Since Health Marts must be non-profit organizations, they would not attract classic entrepreneurs. It is not clear who with the appropriate skills, talents, and needed start-up financial resources would be motivated to organize a Health Mart.

The likelihood that an AHP could be attractive seems greater. But, as noted, the crucial question is whether the buyers would be now-uninsured employers. Existing associations already have an organizational structure and a membership that are potential buyers, so developing a health coverage product and attracting customers are less formidable obstacles. The price advantage is also likely to be somewhat larger. But unless that price advantage includes a cost savings realized because the association covers below-average risk employers, it may not be large enough to attract currently uninsured groups qualified to belong to the association.

Although an estimate of the market share that Health Marts and AHPs might draw is beyond the scope of this study, some research on this issue has been done. The Urban Institute has a research project underway to make such estimates. These results should be available in the near future. The CONSAD Research Corporation issued a paper in July 1998 that reports on their econometric analysis of the impact of implementing the Association Health Plan proposal. They estimate that between 1.1 million and 4.4 million employees and 1.0 million to 4.1 million dependents would be newly covered as a result of the formation of Association Health Plans.²⁵Avoiding Threats to Past Reform Progress

The critics of Health Marts and AHPs worry that these purchasing arrangements could undermine the progress that has already been made toward making the small-group market function more effectively. There are several provisions that cause concern.

The new entities would create new opportunities for risk selection, potentially allowing them to draw away low-risk groups from the small-group insurance pool. This would re-start the vicious cycle of ever-increasing risk segmentation. In particular, the concern is that any savings that these proposed entities might generate would not be sufficient to attract the uninsured but would be sufficient to encourage risk selection among already insured small businesses.

In thinking about the potential danger, it is useful to remember a few key points: Anyone involved in selling to small groups-whether insurer, agent, or third-party administrator (TPA)-makes money by attracting customers. The best way to attract customers is to offer a lower price. The cheapest and quickest way to reduce costs so that it is possible to offer a better price is to avoid insuring sick people. The motivations to engage in risk selection are great because the financial rewards for being successful are high. Past experience shows that where there are loopholes, they will be used.

We consider the possible problems with the two proposals separately.Health Marts

The proposal for Health Marts tries to address the risk segmentation issue by making them subject to all state and federal small-group insurance reform laws:

They would have to sell to all small employers (those employing 2 to 50 employees) who seek coverage within a geographic area and on the same competitive footing as all other insurers in the state.
They could not exclude high-risk groups, nor charge them higher rates.
The Health Mart’s geographic area must include at least an entire county (or equivalent).

Although these features provide significant protection against risk selection, the critics fear that there are loopholes related to the possibility of “red-lining.” If a Health Mart chooses to operate in more than one county, the law does not require that the counties be contiguous. Thus a Health Mart could choose to operate in only counties (whether one or more) that include a disproportionate number of low-risk groups-for example, areas where most small businesses were professionals, software companies, etc. This would give them a clear price advantage over competitors that have to provide their product on a guaranteed-issue basis over a larger area. On the other hand, it is probably reasonable to allow Health Marts to operate over a limited geographic area, especially as they are first being formed. Other competing insurers, especially HMO and PPO health plans, also frequently have a limited service area.

Health Marts can operate in more than one state, but the states must be contiguous. This further widens the loophole just mentioned, but perhaps creates another as well. The law seems to be unclear about whether a multi-state Health Mart would have to obey the insurance reform laws of each state or only those of the home-office state. If the latter is the case, a Health Mart might establish itself in a state with permissive rating rules-that is, rules that permit substantial rate variation related to health status, claims experience, industry, etc.-and then sell its products in a state with restrictive rules. It would then have an advantage over competitors in this more restrictive state by being able to charge higher rates to high-risk groups and lower rates to low-risk groups. The potential problem could, of course, be eliminated by requiring Health Marts to conform to the small-group reform laws of each of the states within which they sell coverage.Association Health Plans

AHPs pose a more serious threat to the preservation of a large risk pool, primarily because they can self-insure and are not required to sell coverage to any small employer seeking it but only to members of the sponsoring association. Even many existing associations will not represent a cross section of risk. If a high proportion of those with members who are relatively healthy decide to offer AHPs-perhaps with the help of an enterprising consultant-that alone could cause a significant dilution of the small-group risk pool. And unless the law precludes doing so, lower-risk groups will have a strong incentive to join some existing or newly formed association for the purpose of getting less expensive health coverage for themselves by excluding from membership groups that exhibit higher risk profiles. If there were no restrictions on who could form an association for purposes of getting health coverage, we would surely see a proliferation of associations that bring together just lower-risk employers who then self-insure.

Those who have crafted proposals for AHPs have tried to be responsive to this concern and have included the following provisions to limit the possibilities for risk selection.

An association offering an AHP cannot be formed solely for the purpose of buying insurance; it must perform some other substantial legitimate functions for its members.
An association that is not already in existence at the time of the law’s implementation must have been in existence for three years before it becomes eligible to offer an AHP.
An association offering an AHP cannot deny membership to any group or vary the membership fee for the group because of the group’s health status or risk profile.
An AHP cannot base premium rates on the health status, claims experience, or industry of the applicant employer.
No individual employee in a group can be excluded from coverage by either the employer or the AHP.
During a five-year pilot program, an AHP is permitted to self-insure only if (1) it offers self-insurance on the date of the law’s enactment, (2) the employers in the plan represent a broad cross-section of the trade, or (3) the employers represent trades that have above-average risk profiles.
The AHP must be actively marketed to all members of the association.
Since an AHP would be required to meet the tests of an ERISA “group health plan,” it would be subject to the preexisting condition, portability, nondiscrimination, and renewability provisions of HIPAA.

These provisions would certainly help to limit the formation of new AHPs by entrepreneurs, insurers, agents, and third-party administrators trying to find a way to sell coverage to low-risk small employers. But some loopholes remain:

It may be very difficult to enforce the provision that an association has some legitimate function in addition to providing insurance. For example, would a new organization for health club employees (including aerobic instructors) that published a monthly newsletter and held an annual conference (even if almost nobody came) be a legitimate association? What would be the test to determine whether the non-insurance function was significant enough to permit it to offer health coverage?
If the financial incentives are strong enough-that is, the potential rewards from risk selection are great enough-the three-year waiting period may also not be much of a barrier to starting a new association. A three-year time horizon is not excessive for some insurers or entrepreneurs.
The provisions which prevent using health status, claims experience, or industry as a basis for setting premiums are laudable and important. But the industry restriction may be of little real value. Although it does require cross-industry associations (for example, a Chamber of Commerce) to accept all groups regardless of industry, it would not affect the many associations that are formed around industries nor prevent associations from organizing on an industry basis.
The provisions requiring a broad cross-section of trades or coverage of above-average-risk groups might be very difficult to monitor and enforce. It also would be difficult to establish tests that are tough enough but not unreasonable or arbitrary.

On balance, the provisions to limit risk selection would probably prevent a massive initial shift from other kinds of coverage to AHPs. But over the longer run, there may be reason for concern. In assessing the potential danger, it is useful to think about the motivation for starting an AHP. At one time, before HIPAA and state insurance reform, some association insurance plans were formed for high-risk industries where firms simply could not get coverage through normal channels. But the combination of HIPAA’s access reforms and states’ rate reforms means that small businesses in most states are now guaranteed access to all small-group products at the adjusted community rate or within some constrained rate range. So nobody can be refused coverage. What kind of association, then, would still find a significant price advantage to offering health coverage to its members? Primarily those with members whose risk profile is below average, that is, those who would benefit by getting out of the risk pool. Of course, some AHPs might be formed because their members want to buy coverage that is scaled down by not covering mandated benefits. But, as discussed above, a limited benefit package is not likely by itself to attract large numbers of groups to an AHP.

There is another potential problem related to an AHP’s ability to self-insure. As noted earlier, past experience with inadequately regulated association arrangements documents that unless there are stringent requirements to insure the financial viability of the self-insuring entity, people who have paid their premiums can be left without coverage and any way to pay incurred medical expenses. The insolvency of many MEWAs, brought on by mismanagement or fraud, left hundreds of thousands of people in just this position in the past.²⁶

Since self-insured AHPs would be risk-bearing entities, protections have to be in place to ensure that claims get paid even if the association underestimates its liability to pay medical claims and other expenses. Traditionally, the safeguard has been to require insurers to have reserves, to buy reinsurance, and to contribute to some state pool that could be used to pay off medical claims if an insurer does go bankrupt. The AHP proposals address these issues. But some believe that they do so inadequately. The proposal identifies the required protections against insolvency and allows states to enforce these provisions. (It also gives states greater authority to regulate existing MEWAs if they choose not to be certified as AHPs.) If a state chooses not to regulate the new entities, the Department of Labor would do so. It is questionable whether the federal government is prepared to take on a unfamiliar regulatory burden that could be quite large and difficult.

FOSTERING COLLECTIVE PURCHASING ARRANGEMENTS

Collective purchasing of health coverage clearly has advantages for small employers. But the idea has not taken hold in most areas of the country. Simply permitting certain forms of collective purchasing, as the Health Marts and AHP proposals do, might not be enough to induce people to establish these or similar vehicles, such as HPCs, or to motivate large numbers of small employers to buy coverage from them. In this section, we address the challenge of promoting a collective purchasing strategy for small employers without taking the politically controversial steps of establishing new government subsidies or mandating that employers finance or individuals obtain coverage. In the next section, we relax these constraints and briefly revisit the debate over the advantages and disadvantages of new subsidies and mandates.

Without new funding from government or business, a collective purchasing strategy per se can go only so far in expanding coverage among employees of small enterprises. But certain design features could be built into this approach that would stretch its potential as far as possible.

First, Congress should follow the rule of “do no harm.” In considering Health Marts, AHPs and similar proposals, lawmakers should be certain that provisions are included to ensure these entities do not re-introduce risk selection in the small-group market. If they were implemented in a way that allowed them to risk-select, they would threaten the good results of past rate reforms as well as jeopardize existing collective purchasing arrangements.

Second, HIPAA could be strengthened to make collectively purchasing arrangements more feasible in some states. A few states have not passed small-group reform laws, and some states that have done so have lenient rating rules that allow health plans to vary premiums by large amounts from group to group, depending upon employee health status and other differences among employers. It is difficult for small-group collective purchasing arrangements to operate in such an environment. One of the advantages of HPCs and similar arrangements is that employers and employees can easily compare health plans in terms of price (given that benefits are usually standardized for all health plans). But when prices can vary so substantially from group to group, depending on the group characteristics, premium quotations cannot be provided until each group is medically underwritten (that is, assessed to determine the risk associated with the group).²⁷ This adds time, complexity, cost, and confusion to an approach whose appeal is supposed to be, in part, its simplicity. Moreover, because of the underwriting problems, it is more difficult to implement the feature which permits employees in a group to choose different plans, one of the main attractions of HPCs and similar arrangements. Under such circumstances, some health plans may be willing to participate only if all employees in a group are required to enroll in the same plan.²⁸

The federal government could solve this problem by amending HIPAA to include restrictions on rate variation in the small-group market, requiring some form of modified community rating. Congress explicitly chose not to address rate restrictions when it originally passed HPAA, leaving that issue up to individual states.

Third, the government may have to go beyond simply permitting certain forms of collective purchasing, as the Health Marts and AHP proposals do, to actively encourage their formation, development and use. This active intervention (still stopping short of mandates) may be needed to induce people to establish AHPs, health marts, or HPCs, as well as to motivate large numbers of small employers to buy coverage from them.

This active encouragement could take two forms. First, the federal government could provide funds to finance feasibility studies and start-up activities, as was done to encourage HMOs in the 1970s. Congress could authorize a limited amount of funds (only a fraction of what would be spent for new subsidies) to establish a grant program to seed innovative collective purchasing arrangements, perhaps through the Health Care Financing Administration or the Agency for Health Care Policy and Research. Under one variant of this approach, one of these agencies could work with grant-making foundations to assist groups with their initial development costs and to evaluate their results-including the impact on cost, quality, and coverage. Under a different variant, the Commerce Department (for example, the Small Business Administration) could make these grants. This would reflect more of a seed-capital approach, with less emphasis on research and evaluation. As noted earlier, without such stimulus, few people may be willing or able to initiate such collective purchasing for small employers.

Under a second approach, both federal and state government could provide various forms of financial incentives for employers to participate in collective purchasing arrangements once the option is available, such as offering premium tax credits for a year or two. (Presumably, after gaining a year or so of experience with collective purchasing, employers recognize the advantages, and no further external inducements would be necessary.) Some states have used temporary tax credits for employers newly offering health coverage. President Clinton’s FY 2000 budget includes a proposal to provide a temporary tax break for businesses with 50 or fewer employers who newly sponsor health coverage if they purchase health insurance through not-for-profit purchasing coalitions. Firms would receive a tax credit of 10 percent of their contributions to health plans, capped at $500 per family policy and $200 per individual policy.

The advantage of this approach is that it uses a carrot rather than a stick to broaden employer coverage, which may lessen adverse side effects and defuse business opposition. Tax subsidies that are temporary and capped would be less costly than more open-ended government spending. The disadvantages include the history of very limited use by employers of various tax subsidies designed to alter their compensation and hiring policies, as well as the potential for windfall gains on the part of employer groups that might have engaged in collective purchasing arrangements without tax inducements. On the other hand, some inequities are introduced because some small firms get subsidies while others in essentially equal circumstances that have already joined collective purchasing entities are ineligible and get no subsidies.

A different approach falls somewhere between voluntary inducements and mandates for employers to provide and finance coverage. Employers could still voluntarily decide whether to offer and finance health coverage. But if they did so, they would be required to purchase it from one of several forms of collective purchasing mechanisms, including HPCs, Health Marts and AHPs.²⁹ Getting employers to take the initial step of participating in current collective purchasing options is difficult. Mandating participation by those who offer coverage would clearly solve that problem as well as provide an inducement for entrepreneurs to organize mechanisms to offer such coverage. Of course, a requirement to purchase collectively was part of past reform proposals, including the 1993 Clinton health reform plan. Any mandate on the business community will generate strong opposition. But this approach would be much more modest in scope than the Clinton plan, since only small firms would participate in collective purchasing. And this mandate would presumably lower rather than increase business spending on health insurance.

Implementing any of these ideas would require real vigilance to make sure that the policy was having the desired effect and not promoting risk segmentation or creating opportunities for unscrupulous operators to bilk small employers. Implementation would probably also raise a number of other thorny issues, such as how to develop a risk-adjustment and premium transfer mechanism to offset windfall losses or gains. Such loses or gains would be created when health plans enroll a disproportionate number of low-risk groups while others find themselves with a predominance of high-risk groups.

COMPARING HEALTH MARTS AND ASSOCIATION HEALTH PLANS TO MORE ENCOMPASSING REFORMS

Although this analysis does not estimate the size of enrollment that Health Marts and Association Plans might attract, it is questionable, as noted earlier, whether these new collective purchasing arrangements would be able to recruit large numbers of currently uninsured groups. The cost savings, from whatever sources, may not be sufficient to bring in large numbers of employers that have currently decided that they cannot afford coverage. On the other hand, if they were able to entice large numbers of currently insured employers, a major reason would likely be that they offer especially attractive prices because they can risk select. If that proved to be the case, it would not represent an incremental improvement toward a more effective and equitable health insurance system in the U.S.

If the main purpose of these proposals is to reduce the number of uninsured, that purpose may not be realized. Many small employers would not offer coverage even if they could buy coverage at the price paid by large self-insured employers. Either the coverage would still be unaffordable for the firm or its employees, or they would make the judgment that the cost of paying for it would not be justified by the benefits. As noted earlier, even the most optimistic estimates of the impact of eliminating mandated benefits or implementing Association Plans suggest that perhaps 10 percent to 20 percent of uninsured people would become covered. But that would still leave between 80 percent and 90 percent of the currently 43 million uninsured in that same status they are now. The fact is that solving the uninsured problem in a comprehensive way will almost certainly require some combination of subsidies for low-income Americans, effective health care cost control, and mandated purchase of coverage.

Without effective cost control, neither mandated purchase nor subsidies will be viable. For the last five years or so, attention has not been focused on cost control because premiums were rising slowly. Recent reports indicate, however, that health insurance premiums in 1999 are expected to increase at the highest rates since the early 1990s, and a growing number of employers are expected to transfer even more of the costs to employees by the year 2000. ³⁰ Lower-income workers and those in small firms will be hardest hit. Among employees in businesses of fewer than 200 workers, the average worker family already contributes 44 percent of the premium.³¹

The advantages of Health Marts and Association Health Plans, such as greater negotiating clout, exemption from state benefit mandates, and resulting reduction in premiums, would also not reach the 16 million Americans who are purchasing individual coverage. In fact, the individual market involves such intractable problems related to self-selection based on health status that even state insurance reforms have not made much of a dent in the problem. Nor would these benefits be available to the millions of uninsured Americans who are not currently tied to the workforce.

Bolder reforms will be needed to make a significant impact on expanding health coverage. A radical change such as scrapping our private, employment-based system and replacing it with a universal access model something like the Canadian system is not under consideration now. There are, however, other types of reforms that would guarantee access to virtually all Americans, while maintaining many aspects of our current system. These include variations on employer mandates and individual mandates. Many of these reform proposals also attempt to control health care costs through tax reform, spending limits, and other means. Short of mandates, strong incentives to extend coverage through public subsidies, tax credits, and buy-in opportunities could go a long way toward reducing the number of uninsured.

Importantly, many of these alternative approaches utilize collective purchasing arrangements and competition among health plans (the core of the Health Mart and AHP proposals) as the vehicle for efficient purchasing. They are coupled, however, with stronger incentives or requirements that ensure their use.Incentives, Subsidies, and Buy-Ins

Short of mandating employers or individuals to obtain health coverage, strong incentives such as heavy government subsidization and opportunities to “buy in” to existing programs could help expand coverage to large numbers of uninsured Americans.

One example would be a subsidized buy-in to the Federal Employees Health Benefits Program (FEHBP). Initiated in 1959, FEHBP now covers about ten million federal employees and their families nationwide, including members of Congress, postal workers, retirees, and the President. FEHBP is based on the principles of consumer choice and competition. It contracts with over 400 competing plans, from traditional fee-for-service carriers, to HMOs, to employee organization plans. FEHBP is not subject to state mandated benefits laws or state premium taxes, and it enjoys the low administrative costs, economies of scale, and shared risk associated with collective purchasing arrangements. Since FEHBP already exists nationwide, it would be feasible to open the program to additional groups, such as low-income families or uninsured workers. It could also be opened to small businesses.

Like HPCs, Health Marts, and AHPs, FEHBP has the advantage of being a more efficient vehicle for purchasing coverage for small employers. But the basic problem of affordability remains. Significant subsidies or other financial incentives would be necessary to make FEHBP affordable for many and a real force in attracting enrollees. For needy individuals, subsidies could take the form of tax credits or vouchers. If opened to small businesses, tax credits or temporary “matching contributions” could encourage employers to buy in.

Another example of this incremental approach is to expand existing subsidy programs to include additional vulnerable populations. An initial step in this process was inclusion in the Balanced Budget Act of 1997 of the Children’s Health Insurance Program, which dedicated $48 billion over 10 years to expanding children’s health coverage. About half of the 48 approved state proposals (as of January 1999) extend coverage to children in families with incomes up to 200 percent of poverty. Almost 1 million children have already gained coverage under CHIP according to recently released government data.

President Clinton has also proposed a program to extend Medicare coverage to early retirees. Participants would be expected to pay most or all the premium. Proponents see this as a valuable step to assist older workers who lose their jobs and have health problems that make health coverage difficult to afford. Critics believe that costs under this program will be driven up by risk selection as early retirees with health problems enroll, requiring a growing governmental subsidy.

An advantage to these approaches is that they use existing structures; no new large, complicated bureaucracies would need to be built. FEHBP, for example, has only one-fifth as many staff per covered life as Medicare. This is because FEHBP negotiates contracts with health plans but does not regulate prices of specific health care services. To the extent that subsidies are successful in expanding insurance, however, they would also be expensive.Employer Mandates

A more sweeping group of reform proposals involves mandating employers to provide coverage to their employees and contribute some minimum toward the cost. The proponents of this approach note that about two-thirds of the uninsured live in the households of a full-time worker, and many of these will not be covered unless employers are required to provide coverage. An obvious danger of employer mandates is that employers will cut back on cash wages and other benefits in order to afford the new health expenditure. (Economists argue that, at least in the long-run, employees pay the full costs of health coverage, since the decision to hire a worker is made on the basis of the full compensation cost, not the mix of elements.)32 Among firms paying at (or near) the minimum wage, employers cannot legally make this substitution of lower cash wages for health premium payments. The result has to be a reduction in profits, fewer new hires, or layoffs. Similarly, the additional burden on low-wage workers of paying their share of premiums can be onerous. Employer mandate proposals try to address these problems by including public subsidies to small firms and/or individuals.

Requiring employers to purchase insurance is criticized as a regressive financing method, since the premium is a higher proportion of total compensation for low-income workers than for higher-income workers (although the same is true of voluntary employer contributions). Alternatively, requiring employers to contribute through payroll tax financing is somewhat less regressive, since a uniform-rate payroll tax rate requires higher income employees to contribute more in absolute dollars. A proposal introduced in 1992 by California’s Insurance Commissioner John Garamendi combined payroll tax financing with the formation of large, regional purchasing cooperatives. Garamendi’s state reform proposal received national attention because of its unique combination of market and regulatory aspects and possible broader applicability.Individual Mandates

Another group of proposals geared toward access expansion would mandate that individuals, rather than employers, obtain health insurance. These proposals generally include tax subsidies or refunds to help low-income individuals purchase coverage. They try to maximize consumer choice of health plans, placing the focus of responsibility on individuals for selecting the coverage that provides the best value. The expectation is that when consumers make the choice and pay directly, they will be more cost conscious than they are when the employer foots much of the bill. In the more radical proposals, employers would lose their role in both selecting and financing health care; other proposals retain employment as a vehicle for pooling risk, achieving market clout, and attaining economies of scale.An example of the latter type is the Responsible National Health Insurance (RNHI), proposed by Mark Pauly and colleagues.33 This plan has the following features:A requirement that each citizen obtain at least minimum adequate coverage against catastrophic medical expenses, with required coverage increasing as family income declines.Refundable fixed-dollar tax credits or adjustments in total federal taxes that cover the full cost of coverage for very poor families, with a decline as income increases.Tax reform that would change the provisions of the tax code that allow employees to exclude from taxable income the portion of the premium that their employers pay for them; employer payments would be included in employees’ income that is subject to income and payroll tax.Preemption of state mandated benefit laws.RNHI relies on regulation (mandate with public subsidies) to expand access, and market incentives (tax reform) to promote cost-conscious consumer choices. One important criticism of RNHI is that it does not encourage collective purchasing and in fact would likely lead to a shift away from group coverage, toward individual coverage-with its higher administrative and marketing costs, diseconomies of scale, and other inefficiencies. Moreover, some employers who currently provide coverage may drop it altogether – although if they operate in competitive labor markets, they would have to increase other forms of compensation to maintain the same quality of work force.Clearly, mandates do little, if anything, to make coverage more affordable. For an individual mandate to be fair and feasible, adequate subsidies are required so that basic coverage is truly affordable for everyone. The greater the subsidies, the higher the public cost, which would have to be borne through income taxes, payroll taxes, elimination of existing tax exclusions, or some other type of public financing.

CONCLUSION

In addition to the financial obstacles and potential negative side effects mentioned above, all of the alternative approaches discussed in this report face significant political obstacles. All are controversial, and as with any change to the status quo, there would be “winners” and “losers.” The lobbying battles, as we have seen in the past, could be fierce.

Nevertheless, consideration of the strengths and weaknesses of these alternatives is important because the proposals represent examples that go beyond incremental changes that may or may not make a dent in the access problem. They take bolder, more aggressive steps that are more likely to get the job done.

Health Marts and AHPs may have fewer unintended consequences than more sweeping reform plans. But they are also likely to do less good. Some powerful forces have been driving up the number of uninsured Americans in the U.S. year after year-despite the fact that in the last several years the economy has been strong and unemployment has dropped sharply. These trends include shifts in the labor market toward industries where health coverage is less prevalent and toward part-time, temporary, and contract workers. Some employers have stopped contributing to dependent coverage. Others are using fewer older workers who are not yet eligible for Medicare but who are expensive to cover under an employer-sponsored plan. Increasing numbers of workers are turning down employer-sponsored coverage because of the difficulty they face in paying their share of the cost. There are also some “free riders” in the system who could afford coverage but choose not to take it.

Against this backdrop, Health Marts and AHPs may only moderate the continuing deterioration in coverage or perhaps achieve a slight reduction. They are also likely to have little net positive effect on health care spending increases, although they may save some workers money at the expense of others.

As health care spending accelerates and the number of uninsured rises, a renewed debate over more sweeping reform plans becomes inevitable. The mix of market forces and regulation will vary from one plan to another, and each will have some adverse side effects. The debate should focus on whether the side effects are justified by the substantial gains. Regardless of the approach we choose, however, we have to accept the reality that we cannot substantially reduce the number of uninsured without increasing subsidies to people who currently find insurance unaffordable. The evidence indicates that people increasingly value health insurance and believe that it is essential to have coverage. When they go without health insurance, it is usually because they believe they cannot afford it. Providing coverage that is available through more efficient and thus less costly purchasing vehicles, such as HPCs, Health Marts, and Association Health Plans will help. It will not solve the larger problem.

ENDNOTES

This report does not address the difficult problems of finding better ways to cover people purchasing individual health coverage, that is, people who do not buy coverage as a result of being a member of an employed group.
Employee Benefits Research Institute, “Sources of Health Insurance and Characteristics of the Uninsured: Analysis of the March 1998 Current Population Survey,” EBRI Issue Brief #204, December 1998, p. 9.
Of course, this example oversimplifies the situation, since even among the sickest 10 percent, some individuals will experience much higher costs than others. Moreover, insurers have not been particularly skillful at predicting which individuals are likely to incur very high expenses, as shown by the fact that when insurers are allowed to assign people to state reinsurance pools, the pools frequently do not to pay out extraordinary amounts for those individuals.
A research project under the direction of Mark Hall at Wake Forest University (in which one of the authors participated) examined the effect of small-group insurance reform in 12 states. The study documents general agreement among health insurance executives, insurance regulators, and insurance agents that small-group reform has been essential to creating a viable insurance market for small businesses. Reports based on this research are available at the following internet site: www.phs.wfubmc.edu/insure.
Karl Polzer, “Preempting State Authority to Regulate Association Plans: Where Might It Take Us?,” Issue Brief of the National Health Policy Forum, No. 707, October 1997, p. 3.
Michael A. Morrisey, Gail A. Jensen, and Robert J. Morlock, “Small Employers and the Health Insurance Market,” Health Affairs Vol. 13, No. 5, Winter 1994, pp. 149-161.
From a December 6, 1995, letter from the General Accounting Office (Mark V. Nadel, Associate Director, National and Public Health Issues) to Harris W. Fawell, Chairman, Subcommittee on Employer-Employee Relations, Committee on Economic and Educational Opportunities, House of Representatives, U. S. Congress.
Polzer, p. 3.
Polzer, p. 4
The law which pertains to what is or is not a MEWA is very complicated and somewhat murky, and many of the designations have to made on a case-by-case basis. Thus it is not possible to make a hard-and-fast statement that can be used to determine whether any particular association arrangement is a MEWA.
GAO letter to Representative Harris Fawell.
Prior to the passage of HIPAA, COSE did medical underwriting, rejecting between 12 percent and 13 percent of applicants with high risk profiles. Under HIPAA, of course, they must provide coverage on a guaranteed-issue basis. COSE does offer choice of plans from Medical Mutual (formerly Blue Cross and Blue Shield of Northern Ohio) and Kaiser Permanente, although Kaiser has only a 10 percent to 15 percent share. Personal correspondence from Mark Hall, Wake Forest University, January 1999.
From the following internet site: www.mrmib.ca.gov/MRMIB/HIPCRptSum.html.
Unpublished data from the Institute for Health Policy Solutions, Washington, D.C., 1998.
This estimate is based on unpublished 1998 data from the Institute for Health Policy Solutions (IHPS), using a restrictive definition of a HPC, including the criteria that the HPC must offer employees a choice of health plans. However, this estimate includes enrollment in COSE, which is not included in the IHPS data.
A relevant question is whether Association Plans would be subject to any or all provisions in the many state patient protection acts that have been passed in the last several year. It seems likely that Association Plans would be freed from having to conform to some but not all provisions. Congress is also expected to pass some form of patient protection legislation in the near future. That legislation would probably determine how federal provisions would apply to Association Plans.
Economists would generally argue that if the insurance industry is a competitive market, the real savings the Association could realize by taking on risk itself rather than hiring an insurer to do so would be minimal. In competitive markets, “pure” profits-profits beyond a reasonable return on investment-largely disappear, so that buyers are getting the product for the total costs of production. Unless they can be more efficient by producing the product themselves, which is doubtful, they save nothing by not buying the product in the market.
Estimates of the U.S. Chamber of Commerce and the National Federation of Independent Business, reported in Karl Polzer,”Preempting State Authority to Regulate Association Plans: Where Might It Take Us?,” Issue Brief of the National Health Policy Forum, No. 707, October 1997, p. 5.
Gail Jensen, Roger Feldman, and Bryan Dowd, “Corporate Benefit Policies and Health Insurance Costs,” Journal of Health Economics, Vol. 3, No. 3, 1984, pp. 274-296.
Frank A. Sloan and Christoper J. Conover, “Effects of State Reforms on Health Insurance Coverage of Adults,” Inquiry, Vol. 35, 1998, pp. 280-293.
Gail A. Jensen and Michael A. Morrisey, “Small Group Reform and Insurance Provision by Small Firms, 1989-1995,” Inquiry, forthcoming, 1999.
Based on interviews of one of the authors with insurers, agents, and state insurance regulators for a project assessing the impact of small-group market reform. (The project was directed by Mark Hall of Wake Forest University.)
Polzer, p. 6.
Gail A. Jensen, and Michael A. Morrisey, “Mandated Benefit Laws and Employer Sponsored Health Insurance, unpublished manuscript, December 1998, p. 9.
The Projected Impacts of the Expanded Portability and Health Insurance Coverage Act on Health Insurance Coverage, Pittsburgh: CONSAD Research Corporation, July 10, 1998.
Polzer, p. 4.
As noted earlier, a HPC or similar entity cannot adopt standards for determining premiums that are substantially more lenient than those used by health plans selling in the outside market. If they do, they will be inundated with high-risk groups and soon be forced out of business.
The authors are indebted to Rick Curtis of the Institute for Health Policy Solutions for help in understanding the complexities of implementing employee choice when medical underwriting is necessary.
The authors thank Karl Polzer for suggesting the idea that a mandated system could give employers the option of purchasing coverage from a variety of collective purchasing sponsors.
Hewitt Associates LLC, “Employers Face Significant Health Care Hikes in 1999 Following Modest ’98 Increases,” December 7, 1998, website: www.hewitt.com/news/pressrel/1998/12-07-98.htm.
Jon Gabel, Paul Ginsburg, and Kelley Hunt, “Small Employers and Their Health Benefits, 1988-96: An Awkward Adolescence,” Health Affairs, September/October 1997, Vol. 16, No. 5, pp. 103-110.
Mark V. Pauly. Health Benefits at Work: An Economic and Political Analysis of Employment-Based Health Insurance. Ann Arbor: University of Michigan Press, January 1998.
Mark V. Pauly, Patricia Danzon, Paul J. Feldstein, and John Hoff. Responsible National Health Insurance AEI Press, Washington, D.C. 1992.