America's history of organizing health insurance coverage on a group basis through large employers leads many to view pooling as the solution to the lack of coverage often experienced by small business employees. If enough small businesses can be aggregated together into a pool, their logic goes, the group should be able to collectively purchase coverage at the same rates as a large employer. The mixed track record of small business pooling arrangements, however, indicates that size is not the only factor essential to successful pooling. Two other features of large employer groups--stability and randomness--must also be replicated. By combining sound program design with other market-based reforms, state policymakers can create pooling arrangements that increase the number of insured workers and improve the overall health care system.

The Appeal of Pooling

Despite the inherent difficulties of pooling and some disappointing experiences, the prospect of creating health insurance pooling arrangements for small businesses and individuals still attracts policymakers, particularly at the state level. Four factors account for this continued interest.

First, the steady erosion of employer-sponsored health insurance is most acute in the small business sector and is a major contributor to the growing uninsured population. Today, only 60 percent of workers are covered by employer-sponsored insurance, and among those working in firms of 10 or fewer employees, the share has declined to 48 percent.[1] With each passing year, the notion that small employers should, like big businesses, separately sponsor and manage group health plans for their employees becomes increasingly problematic.

Second, as America continues to evolve a more flexible and mobile post-industrial information economy, the old notion of attaching fringe benefits to the employer, as opposed to the worker, is also becoming increasingly problematic--especially in the small business sector. The infrastructure of new retirement savings vehicles created by the federal government, such as IRAs and 401(k) plans, has met the needs of both businesses and workers for more portable and self-directed retirement planning. In the same fashion, employers and workers are now looking to policymakers for reforms to make health insurance benefits personal and portable.

Third, as with the shift from defined-benefit to defined-contribution retirement plans, policymakers are discovering that new arrangements can offer advantages beyond meeting the needs that originally drove the policy changes. In the case of health care, a system of portable coverage could begin to reduce the number of uninsured by helping individuals keep coverage through job and life changes while minimizing demands on tax-funded health care programs.

Fourth, if new pooling arrangements are combined with other insurance market reforms designed to transition from an employer-owned and seller-driven market to a consumer-owned and buyer-driven market, the result would be to inject fresh incentives for both controlling costs and increasing quality into the broader health system. Insurers and providers would be forced to find ways to meet consumer demands for better results at better prices.

The Key Ingredients: Size, Stability, and Randomness

Any grouping of small businesses or individuals is likely to be much less stable than the workforce of a single large employer. Not only are the businesses themselves less stable over time, but their workers tend to change jobs and employers more frequently than those in large firms. This instability increases the complexity and expense of administering coverage and also makes it more difficult for actuaries to project future claims costs when calculating premiums. This uncertainty, in turn, induces insurers to add a "margin of safety" to the premiums they charge--making the coverage more expensive and thus repelling some potential customers.

The situation is further complicated by differences in the two kinds of groups' compositions that affect randomness assumptions. A group consisting of one large firm's employees and their dependents is much more likely to be representative of the general population in a given area than a similarly sized group composed of the employees of multiple small businesses. Furthermore, pooling small employers creates many more opportunities for "selection effects" than is the case with large firms. Individuals with poor health status naturally prefer jobs that come with employer health coverage, but securing a job with a large employer is generally a lengthier and more difficult process than with a small firm--even if the large firm does not discriminate on the basis of job applicants' health status.

The track record of small business pooling arrangements illustrates the difficulties in addressing the factors of size, stability, and randomness. However, experience does show that large pool size, while not alone sufficient to assure success, is the critical prerequisite for addressing the other issues. A successful pool must be large enough to induce insurers to offer competitive benefit packages and prices. But to attract a large membership, the pool must offer insurance plans with competitive benefit packages and prices. This is a classic "chicken or egg" problem.

The more successful pooling arrangements are large. For example, the Connecticut Business and Industry Association's pool, Health Connections, and the Cleveland Council of Smaller Enterprises' (COSE) pool both captured considerable small business market share--estimated at over 10 percent for Health Connections and between 60 and 80 percent for COSE.[2] Health Connections' market share has been attributed to its close relationship with brokers; COSE's, to its ability to negotiate competitive prices.[3]

Other pooling arrangements have failed largely due to inadequate market share and declining insurer participation. California's PacAdvantage, North Carolina's Caroliance, the Texas Insurance Purchasing Alliance (TIPA), and Florida's Community Health Purchasing Alliance (CHPA) each struggled to obtain market share and retain insurers, ultimately resulting in their demises. PacAdvantage, Caroliance, TIPA, and CHPA captured less than 5 percent, 0.8 percent, 1 percent, and 5 percent of their small business markets, respectively, at their peaks.[4] Prior to disbanding, the number of participating insurers decreased from 20 to 2 in PacAdvantage, 6 to 1 in Caroliance, 20 to 1 in TIPA, and 35 to 6 in CHPA.[5]

Steps for State Policymakers

In designing alternative pooling arrangements, policymakers should learn from past experiences and consider some new tools to achieve their objectives. The following are some considerations for state policymakers interested in developing alternative pooling arrangements.

First, start with a core group. The best way to resolve the "chicken or egg" dilemma of pooling and to help ensure that a pool is large, stable, and random is to start the pool with a core group that meets those three characteristics. State lawmakers have such a group at hand: the state government's own workforce. Making the state government the first employer to join the new pool would have several positive effects:

1. Such a large, stable, and random group of employees and dependents would give actuaries a basis on which to price coverage offerings and provide a strong market incentive for insurers to offer attractive benefit packages at attractive premiums. That, in turn, would make the pool attractive to small businesses and their workers.

2. In most states, government workers would gain a wider choice of coverage. This, in turn, would make it easier for the state to extend coverage to its part-time and contract workers.

3. Public employees are, on average, older than other workers, while employees of small businesses tend to be younger than other workers. Combining the two groups in one pool could lower average premiums for state workers.

Second, ensure real plan choice. Research on small business health insurance cooperatives indicates that one of their principal attractions is the opportunity for individual employees to select their health plans. Employee choice of coverage is available in Connecticut's Health Connections and in New York's Health Pass, each of which offer plans from four insurance companies.[6] Employers in COSE may offer up to 3 of 19 health plans, though 90 percent of enrollment is through one insurer.[7]

Attempts to lower premiums by contracting with only one insurer for coverage or by limiting competition to two or three plans offering essentially the same benefit design are not likely to result in savings but will almost certainly make pooling arrangements less attractive to small businesses and their employees.

Third, limit selection effects. People usually choose what is in their perceived best interest, but in the health insurance system, their choices can sometimes produce disruptive results. Policymakers can act to prevent the most destabilizing selection effects while still preserving the important elements of consumer choice. The key is to ensure that when individuals or employers have a choice of two different health insurance markets, the same basic rules are applied to both markets. For example, if health status is permitted as a rating factor in one market but not in another, then the healthy will gravitate toward the first market while those in poorer health will join the second. At least three failed pooling arrangements--PacAdvantage, Caroliance, and TIPA--had to compete with external markets using health rating but were prohibited from using it themselves.[8]For pools to not be disadvantaged from the start, the same basic insurance rules must apply to coverage offered through the pool and coverage available outside of the pool.

Fourth, reduce obstacles to insurer and agent participation. Policymakers must be mindful that health insurance pooling arrangements change the basic business model for insurers and agents. For pooling to be successful, insurers and agents need to see the proposed design as an opportunity as much as a threat. For example, some small business purchasing pools initially thought they could save money by not paying commissions to agents, and that encouraged agents to steer clients away from the pools. Working with brokers and agents can help pools gain market share, as shown in Connecticut. When brokers see cooperatives as threats, such as in California, Texas, and Florida, attracting small businesses becomes difficult.[9]

Similarly, insurers and agents face higher expenses in selling or servicing health insurance for one hundred groups of ten individuals than ten groups of one hundred individuals. Beyond paying agents commissions, policymakers can take other steps to make pools more attractive. One important step is to design pooling arrangements such that they bring in low-risk but difficult to reach customers, such as young, healthy individuals and part-time workers. Another way is to structure pools to allow them to handle additional administrative tasks, such as collecting and aggregating premium payments, from multiple sources.

Fifth, make complementary health care reforms. While alternative health insurance pools can be a building block of state health care reform, creating a true system of portable, consumer-centered health insurance will also require additional changes. Other steps states should consider in conjunction with new pooling arrangements include: combining separately regulated small-group and individual insurance markets into a single market with uniform rules; chartering a health insurance exchange to more efficiently handle the administrative functions of enrollment, plan selection, open season, and payment collection; shifting public spending on health programs to a premium support model; and establishing health insurance risk-transfer mechanisms that allow insurers to proportionately redistribute the expenses associated with high-cost claims and individuals.

Conclusion

Rising prices and the burdens associated with obtaining and managing coverage on a group basis continue to erode employer-sponsored health insurance in the small business sector. However, the prospect of reducing insurance rates with pooling arrangements is made difficult by the lower levels of stability and randomness in small groupings of businesses and individuals.

State policymakers can solve the "chicken or egg" problem by making the state government the first employer to join the new pool. Applying lessons from past experiences and passing other market-based health care reforms will increase the chances for success. If designed properly, alternative health insurance pooling arrangements can not only expand the ranks of the insured but also accelerate the drive toward a higher-quality, consumer-driven health care system.

Edmund F. Haislmaier is Senior Research Fellow in the Center for Health Policy Studies at The Heritage Foundation. Michelle Bucci, a Heritage Health Policy Fellow from the University of Virginia, contributed research to this paper.