TASK FORCE ON THE FUTURE OF HEALTH INSURANCE Issue Brief NOVEMBER 2002 Health Insurance Purchasing Cooperatives Elliot K.Wicks Economic and Social Research Institute S The Commonwealth Fund is a mall employers1 have particular disadvantages as purchasers of private foundation supporting independent research on health health insurance. They often are not well informed about the and social issues. insurance options available to them because, unlike large employ- ers, they do not have specialized staff to manage their employee benefit For more information, please contact: programs. Small business owners, who usually make decisions about insur- Cathy Schoen ance for their firms, typically have little time or expertise to bring to the Vice President for Health Policy, task and often must rely on an agent’s advice. Even when small employers Research, and Evaluation The Commonwealth Fund make informed choices, their health insurance premiums are likely to One East 75th Street be substantially higher than what large employers pay for comparable New York, NY 10021-2692 coverage. Tel 212.606.3800 Fax 212.606.3500 One idea for overcoming these disadvantages continues to generate E-mail cs@ cmwf.org widespread interest—the health insurance purchasing cooperative. People attracted to the idea reason that small employers who join together to pur- chase health coverage collectively should be able to strike a better deal than they would by acting separately. Acting jointly, it seems that they should be able realize the advantages that large employers enjoy because of their size and bargaining power. If, by aggregating their purchasing power, small employers were able to buy coverage at lower cost, firms not previ- ously offering health coverage might be encouraged to do so, thus reduc- ing the numbers of uninsured. The collective purchasing idea appeals to Additional copies of this (#567) and other Commonwealth Fund people of divergent philosophical perspectives because putting it in place publications are available online at does not require major institutional changes or government regulation and www.cmwf.org does not seem to be very costly. Publications can also be ordered by calling 1.888.777.2744. To learn about new Fund Elliot K. Wicks, Ph.D., is a senior fellow at the Economic and Social Research Institute and a publications when they appear, visit senior consultant with Health Management Associates. As a health economist and policy analyst the Fund’s website and register to with more than 25 years of experience, he has expertise in the areas of small-employer purchasing, receive e-mail alerts. health insurance reform, cost containment, and policies to extend health coverage to the uninsured. 2 The Commonwealth Fund This brief compares the expectations of physicians. Employee choice has particular value in health insurance purchasing cooperatives for small an insurance market dominated by managed care employers with the actual experiences of different plans, each with different treatment philosophies, co-ops and draws lessons about the potential for expectations of enrollees, and panels of providers. similar future purchasing efforts. Because of the administrative burdens associated with having health insurance contracts with multiple The Case for Co-ops health plans, few small employers are able to offer The theory of purchasing cooperatives is straight- individual employee choice except when they par- forward. Acting as a group, small employers should ticipate in a collective purchasing arrangement. be able to wield the kind of purchasing clout that Another advantage often attributed to pur- large employers enjoy in their negotiations with chasing cooperatives is risk-pooling. People fre- health plans. If they bring many prospective clients quently make the argument that, by joining to the table, the cooperative should be able to together, small employers can spread risk by pool- negotiate more favorable prices and to persuade ing firms of normal risk, higher-than-average risk, health plans to cater to their particular needs. and lower-than-average risk. (In this context, Collective purchasing might also be expected to “risk” refers to the probability that members of an reduce administrative diseconomies and thereby employer’s workforce and their insured dependents reduce the costs of coverage. When insurers serve will incur medical expenses for which the insurer hundreds of small employers, they incur higher will have to pay.) The expectation is that coverage administrative costs than when they serve a single would become more affordable for higher-risk large employer with the same number of employ- groups. Unfortunately, this advantage can seldom ees. It is expensive to send insurance agents to each be realized in practice. If the legal environment small employer to explain coverage, handle sales, permits insurers to use risk-rating to set premiums and service accounts once the coverage is in place. for firms that buy coverage outside of the co-op— Because each insurer and health plan serving this as is the case in many states—lower-risk employers market has to establish mechanisms and pay people will typically find it more advantageous to buy to perform these functions, there is much duplica- coverage in that outside market, since they will be tion. If a single purchasing cooperative served small offered a price that reflects their lower risk.2 If that employers, it could presumably consolidate many happens, the average level of risk of the employers of the marketing, educating, billing, and servicing remaining in the purchasing cooperative will rise, tasks under a single roof, thereby realizing economies and the cost of coverage will rise in turn. This will of scale and lowering the costs of coverage. set off a chain reaction of spiraling prices and The purchasing cooperative approach could retreating employers that will destroy the pool’s potentially bring another advantage to small viability. With respect to pooling risks, a purchasing employers—a choice of health plans for individual cooperative has to follow essentially the same prac- employees. Cooperatives typically allow individual tices as the outside market when dealing with employees to choose from the array of plans with employers with different levels of risk. If the out- which the cooperative has negotiated contracts. side market is permitted to charge higher-risk Individual employees, rather than their employers, employers higher premiums, the purchasing coop- are thus able to decide what kind of health plan erative has to do the same to survive—greatly best meets their particular needs. If workers leave complicating the co-op’s pricing and associated one job for another position at a firm that buys administrative tasks. A cooperative can use various through the cooperative, they can avoid having to forms of community rating—which prohibit bas- change health plans and therefore avoid changing ing premium rates on enrollee’s health status—only Health Insurance Purchasing Cooperatives 3 if the law requires insurers to follow the same York Business Group on Health collaborated with community rating rules in pricing all of their New York City to establish a purchasing coopera- products sold in the small-group market. (Currently, tive in late 1999, which has a current enrollment states vary in their rules on this matter.) But in of about 7,000 and is growing. that case, the cooperative is not the risk-pooling A significant proportion of co-ops have failed. agent; risk-pooling occurs because the law requires The failures have included co-ops that initially insurers to use community rating whether they sell seemed to be quite successful as well as others that inside or outside the pool. never really became viable. The demise of the Florida Community Health Purchasing Alliances Measuring Expectations Against Reality was perhaps the most notable failure because the In practice, purchasing cooperatives have not co-op had once seemed quite successful. Originally always met expectations. Although there are cer- a state-created consortium of 11 separate alliances, tainly successful models, there have also been some the cooperatives enrolled 92,000 people when notable failures.3 enrollment peaked in 1998. But over the years, the One of the first successful collective pur- Florida alliances had increasing difficulty attracting chasing co-ops was created in Cleveland, Ohio— any but the smallest employers and gradually found the Council of Smaller Enterprises, commonly themselves losing health plans. As a consequence, known as COSE. Today, COSE dominates the enrollment also fell, and the purchasing alliances small-group market in the Cleveland area with an ceased operations in 2000. enrollment of about 200,000 people. COSE is not The Texas Insurance Purchasing Alliance, a prototypical purchasing co-op because, for virtu- begun in 1994, never reached the enrollment levels ally all of its history, a single health plan has of the Florida effort, covering only about 1,000 accounted for nearly all of its sales. Although firms and 13,000 people at its height. Difficulty in COSE now offers a choice of two independent attracting employers led to the withdrawal of health plans and a number of plan types, the health plans, and the Alliance governing board employers, rather than individual employees, choose ultimately decided that the operation was not the health plan. Nevertheless, because it dominates viable and closed it down. The North Carolina the small-group market in its area, COSE has Purchasing Alliances, which opened for enrollment proven that small employers banding together can in 1995, were patterned after the Florida model, be effective purchasers of health coverage. but they struggled to attract employers throughout The best-known purchasing cooperative is their existence, and the leaders finally admitted the one in California known in its early years as defeat in 2000. the Health Insurance Plan of California (HIPC). The Alliance in Colorado was the most Originally a creation of the state and operated by recent failure. Established in 1995, the Alliance a state agency, the California co-op is now run by closed in the summer of 2002 after one of its three the Pacific Business Group on Health, an organiza- health plans withdrew from the state small-group tion that during most of its history represented market, a second capped enrollment, and the last only large employers. Now called PacAdvantage, decided to stop participating. the co-op provides coverage for approximately A number of conclusions can be drawn 147,000 people and offers about a dozen health from the experiences of both the successful and plans (not including different plan models offered unsuccessful purchasing co-ops. The most impor- by the same parent company). Another successful tant is that collective purchasing arrangements are co-op is the Connecticut Business and Industry unlikely to succeed unless they can attract large Association, which enrolls about 10,000. The New numbers of employers, and not just the smallest of Continued on page 4 4 The Commonwealth Fund employers with only two or three employees. It has also become clear that some signifi- Without large numbers or substantial market share, cant diseconomies of scale are inherent in serving co-ops cannot exert purchasing power,4 they can- small employers and that these costs cannot be not achieve economies of scale, and they cannot eliminated by centralizing the administration. It attract and retain health plans. Health plans have will always be more costly to serve 5,000 employers often been hostile to the purchasing co-op model with 10 employees each than to serve a single for several reasons. First, they are understandably employer with 50,000 employees. wary of the model because it gives their customers Early proponents of purchasing co-ops bargaining clout. Second, they do not like the believed that administrative costs could be lowered individual-choice feature of co-ops because it pro- by reducing or eliminating the role of insurance vides enrollees with a ready way of switching to a agents and having the sales activities administered different health plan during every open enrollment centrally.8 Proponents sometimes assumed that if period. Third, they believe that their chances of a co-op offered a high-quality, high-value product, getting and keeping all of the employees in an it would more or less sell itself. That assumption employer group—which brings in more revenue proved to be incorrect. Selling health insurance of and helps to spread risk—are much better when any kind in the small-group market is extremely they market to that group outside of the purchas- difficult without the cooperation and even enthu- ing co-op.5 Therefore, unless a co-op commands a siastic support of insurance agents and brokers. significant market share that health plans cannot Early efforts to save the cost of commissions by afford to pass up, plans are unlikely to be eager to diminishing agents’ roles or eliminating them alto- participate.This creates a “chicken-or-egg” dilemma: gether backfired. Insurance agents not only did not without large numbers of enrollees, it is difficult sell purchasing co-op plans, they also became for purchasing cooperatives to attract health plans; strong and effective opponents of the concept. but without the ability to offer substantial choice Today, co-ops assiduously cultivate the good will among well-known health plans, it is difficult for of agents as necessary allies. co-ops to attract enrollees, who are drawn to Early proponents of cooperatives also hoped co-ops in part because of their ability to offer that these new organizations could offer prices such choice.6 somewhat lower than were generally available in Size is also a prerequisite for purchasing co- the market. This hope was not fulfilled. With very ops to achieve economies of scale and to reduce few exceptions, premiums for employers buying administrative costs. As long as their non–co-op through co-ops have not been lower than those business accounts for the bulk of their revenue, available to small employers elsewhere.9 This failure health plans must maintain their existing adminis- to realize the expected price advantage is attributa- trative structures to handle that business. The ble to several factors. Co-ops have not been able to health plans contend that any savings they might reduce administrative costs. They have not had realize as a result of the co-op’s assumption of enough market share to bargain for discounts. And some administrative functions for the relatively in many instances state laws have prohibited insur- small number of co-op enrollees is more than off- ers from offering co-ops premiums lower than set by the extra cost the plans incur because they those they charge to employers outside the coop, have to change their administrative systems to even if the insurers’ costs are lower for co-ops.10 accommodate the administrative structure of the Many supporters hoped that purchasing co- co-op.7 In short, health plans believe that dealing ops would attract a large number of employers with the co-op adds to, rather than reduces, their who had not previously offered coverage. The administrative costs. prospects for success in this area were dimmed by Health Insurance Purchasing Cooperatives 5 co-ops’ inability to offer lower premiums. Even if ipate and continue participating, a co-op must co-ops had realized price reductions, however, have a significant market share. But without most uninsured small employers would still not the participation of a variety of highly reputable have been induced to offer coverage to their plans, it will be difficult for co-ops to attract employees. The research evidence shows that even the number of employers that would yield a a 30 percent reduction in premiums—far more significant market share. Furthermore, co-ops than co-ops could be expected to produce—would do not sell themselves. Without the support of cause only 15 percent of currently uninsured small health plans and insurance agents, small employ- employers to offer coverage.11 ers will not seek out co-op coverage. But Though they have failed to generate signifi- health plans and agents have often been hostile cant savings, co-ops have succeeded in one impor- or, at best, indifferent, to the co-op model. tant respect: participating employers have been able to offer their employees a choice among health 4. Even if co-ops could offer lower premi- plans. This unique feature has proven to be attrac- ums, they could not substantially reduce tive to employers, but it alone has been insufficient the number of uninsured because the to induce a significant number of employers who premium reductions would not be big had not previously offered coverage to do so. enough to induce large numbers of unin- The available evidence, though limited, suggests sured employers and uninsured workers that purchasing co-ops attracted about the same to opt for coverage. proportion of newly insuring employers as the non–co-op market.12 5. Co-ops cannot be the vehicle for pooling high-risk, low-risk, and medium-risk Summary employers. If co-ops follow premium rating An analysis of the efforts to implement the purchas- rules or rules for accepting applicants that are ing cooperative model yields the following lessons: significantly more permissive than those that apply in the outside market, they will suffer 1. The principal advantage that current from adverse selection and ultimately fail. co-ops offer to small employers is not lower premiums but the opportunity for 6. Co-ops are likely to become an impor- individual employees to select different tant source of health coverage only if health plans from the variety the co-op some significant change makes them the offers. favored or perhaps the sole source of coverage for particular groups. This could 2. In the future, co-ops might be able to happen, for example, if employers adopting a offer more attractive prices, but that defined-benefits approach to health coverage would depend on reaching “critical mass” channeled employees to co-ops, or if govern- size. To offer attractive prices, a co-op has to ment offered co-ops as the source of coverage be able to realize administrative savings and/or for individuals who receive certain kinds of have bargaining leverage with health plans. subsidies. Without a change of this sort, pur- Both these conditions require that co-ops chasing cooperatives are unlikely to become a control significant market shares. major feature on the health care landscape. 3. Achieving critical mass size is difficult. To persuade a number of health plans to partic- NOTES 1 For the purposes of this brief, the term “small 6 Elliot K. Wicks and Mark A. Hall, “Purchasing employer” refers to firms employing roughly 100 or Cooperatives for Small Employers: Performance and fewer workers. Most purchasing cooperatives have Prospects,” Milbank Quarterly, Vol. 78, No. 4, 2000, limited membership to firms with 50 or fewer work- p. 534. ers because state insurance laws have typically defined 7 General Accounting Office, March 2000, p. 22. small groups using this cutoff point. 8 Jill Yegian et al., May 1998, p. 8. 2 Jill Yegian et al., Health Insurance Purchasing Alliances for Small Firms: Lessons from the California Experience, 9 General Accounting Office, March 2000, p. 20. California Health Care Foundation, May 1998, p. 7; 10 General Accounting Office, March 2000, pp. 21–22. U.S. General Accounting Office, Private Health Insurance: Cooperatives Offer Small Employers Plan 11 James D. Reschovsky and Jack Hadley, “Employer Choice and Market Prices, GAO-HEHS-00-49, March Health Insurance Premium Subsidies Unlikely to 2000, p. 20. Enhance Coverage Significantly,” Issue Brief: Findings 3 from HSC, No. 46, Center for Studying Health For a detailed discussion of six purchasing coopera- Systems Change, December 2001; see also M. Susan tives, including both successful and unsuccessful Marquis and Stephen H. Long, “To Offer or Not to efforts, see Elliot K. Wicks, Mark A. Hall, and Jack A. Offer: The Role of Price in Employers’ Health Meyer, Purchasing Health Coverage for Small Employers: Insurance Decisions,” HSR: Health Services Research, Barriers to Small-Group Purchasing Cooperatives, Vol. 36, No. 5, October 2001, p. 946. Economic and Social Research Institute, March 2000. 12 4 Elliot K. Wicks and Mark A. Hall, 2000, pp. 517–18. General Accounting Office, March 2000, p. 18. 5 Elliot K. Wicks, Mark A. Hall, and Jack A. Meyer, March 2000, pp. 116–17.