June 26, 2009
Senator Kent Conrad recently introduced what he characterized as a compromise on the contentious issue of having a public plan under healthcare reform: a healthcare co-op. This interview with Milliman principal Jim O’Connor analyzes the co-op concept in more detail.
Q: What is a co-op and how does it differ from a public option?
A: A cooperative health plan, or co-op, is member-controlled rather than government-controlled, and is typically not a for-profit corporation. A co-op must negotiate its payment or reimbursement rates for hospitals and physicians in the market, and cannot dictate them through legislation and regulation like the government could under a public plan. With a co-op, the roles of government regulator and health plan providing coverage are kept separate. The co-op concept is oriented around member decision-making and control, distinguishing it from for-profit insurance carriers. Because it would compete against for-profit and other heath carriers, however, a co-op must offer products that are attractive to the public in a marketplace with choices and must employ prudent rating and enrollment practices if it is to be successful. Determining any requirements or restrictions on those practices, as they apply to all health plans, is one of the many important questions under reform.
Q: How is a co-op different from a commercial insurance company?
A: The governance of a co-op is fundamentally different from that of a for-profit insurance company, and also from that of a non-profit HMO or hospital and medical services corporation. Operationally, however, co-ops must follow policies and apply practices that are similar in many ways to those of any well-run health plan. A co-op might actually use an insurance carrier under contract as its administrator for certain functions. Or the co-op itself might house an insurance capability, as is the case with Group Health Cooperative in Seattle, an example cited by some as a successful co-op in the healthcare field.
A look at existing union plans may help to clarify the nature of a co-op plan. Some of the larger union plans, for example, basically act as insurance carriers. The union plans' operating policies, however, usually differ from traditional commercial carriers in terms of the procedures they follow and what they will and will not allow. Such differences reflect the fact that their obligation is to the union members for whom they provide coverage, as opposed to shareholders or others. But these plans still need to maintain adequate capital and surplus, or reserves, in order to guarantee rates, pay for future claims, and ensure ongoing solvency.
Senator Conrad has made the comparison of his idea to Group Health Cooperative in the Pacific Northwest. Group Health is a Milliman client, and rather than comment, we’ll quote its CEO, Scott Armstrong, who spoke to the Washington Post last week: "Because the co-op proposal in the Senate is still evolving, I can't comment specifically on how we compare to their model. My experience tells me that a proposal that incents the integration and coordination of care will be most effective—and that's a key feature of what Group Health is all about." (For more on Group Health, visit our blog.) Integration and coordination of care reflects the traditional objectives of HMOs. Well-run and market-savvy HMOs have had success in their ability to compete with traditional health insurance companies and in their acceptance by consumers. They are still typically local, although they may be part of a regional or national affiliation.
Q: Have we ever seen similar proposals in Congress?
A: The association health plan (AHP) concept may provide a meaningful point of comparison. This concept emerged from the idea that small employers should have access to at least some of the benefits enjoyed by larger self-insured employers through ERISA. State insurance regulation has often resulted in higher-premium rates for small employers because of various state rules, rating restrictions, and benefit mandates; these requirements and restrictions vary widely from state to state. So the idea emerged to form a national plan that would be subject to only one state’s regulations, which would streamline the ability of insurance carriers to offer smaller employers something less costly.
Any bona fide association (i.e., one that exists for a primary purpose other than offering insurance) could form an AHP. The legislation proposed did not provide for eligibility of individuals, only for employees and dependents of small employers. Presumably such a restriction would not apply to co-ops. The ability to vary rates by age, health status, and other characteristics—at least to the extent allowed by the state in which the AHP was formed—was an important feature of the AHP proposal. With age and health status as rating criteria, insurers would have had greater ability to select and manage risks over and above what carriers could do within certain other states. Even the simple ability to rate by age would have attracted younger groups from states that did not allow such rating, which was a significant objection to the AHP legislation by some in Congress.
Many critical policy details regarding rating and enrollment practices for the future, both overall and for co-ops specifically, remain to be worked out. Obviously, there are many complexities involved. The end result may or may not resemble what was proposed for AHPs.
Q: Can a co-op health plan achieve sufficient economies of scale to contain costs?
A: Health plan size can affect cost levels in two ways. First, administrative expense is a key factor: the larger the co-op is, up to a point of critical mass, the more broadly it can spread its fixed administrative expenses. This reflects classic economies of scale. Empirical evidence has shown, however, that such economies of scale for health carriers do not extend indefinitely, so that medium-sized regional or even sizeable local plans can compete.
Risk dispersion across the plan is a second consideration. Successfully spreading risk requires sufficient size to form a cross-section of risk (for which differences are not reflected in rates) and to help minimize the effects of random fluctuation that is due to insufficient numbers. Size alone, however, is not sufficient to spread the risk. The dynamics of members selecting a health plan and of the health plan accepting members must produce an unbiased mix of risks relative to the rating structures permitted (especially young versus old, and healthy versus otherwise) if these risks are to be effectively spread across a broad base of members.
In addition, sales and marketing is central to the subject of administrative expenses. If reform provides for the formation of a point of purchase "connector" and a health plan is marketed using the connector, then in essence "economies of scale" can be realized in this area of expense. The agent and broker approach to marketing healthcare coverage may well be largely eliminated, thus enabling a co-op or any other health plan to offer lower premiums to the extent that its overall marketing and sales costs were reduced. That might account for as much as an average 5%-8% reduction in current premium levels in the small group markets in many states.
Q: How would the co-op do in a state that has a high uninsured population?
A: A lot depends on how health plan coverages and rating are regulated. A co-op that didn't have to meet benefit mandates under current state insurance regulations, for example, should be able to offer somewhat lower rates than it can now, becoming more attractive to those without coverage. Since the uninsured tend to be younger than the average insured, the ability of the co-op to vary rates by age (if not precluded under reform) would enable it to offer more affordable rates to young uninsured persons. Obviously, government subsidies for lower-income families would help lower the net cost to uninsured individuals and thereby help to reduce the number of them, with co-ops poised to compete for these newly covered individuals.
Q: Who would regulate co-ops?
A: Details regarding Senator Conrad's proposal are unclear at this point. In the case of the AHP proposal, it called for a federal certification of association plans. The association plans could then choose to self-insure coverage themselves, offer fully insured coverage through health insurance carriers, or a combination of both. The AHP, subject to the laws and regulations of whatever situs state it chose, could offer coverage to small employers in every state. The regulation of co-ops could be similarly established and made subject to the situs state’s regulations, or it could be subject to a new set of federal rules that supersedes state requirements.
Q:How would co-ops be affected by the presence of an individual or employer mandate?
A: It is probably valuable to first clarify some language. An "individual mandate" is a requirement that everyone obtain health plan coverage. An "employer mandate" is a requirement imposed on employers. Neither should be confused with a "benefit mandate," which requires that health plans cover certain types or levels of benefits.
Individual and/or employer mandates, if constructed correctly, would likely bring a positive impact on the healthcare market overall, compared to a guaranteed issue environment without any such mandates. Contrary to popular opinion, a proportionately large number of the uninsured are relatively healthy. Bringing all of these uninsureds into the market would not only result in a positive impact on insurers overall in a guaranteed issue environment, but also help redirect people to more appropriate and less costly healthcare provider settings (e.g., uninsureds would no longer need to go through a hospital emergency room to get to see a doctor for routine and non-urgent care). However, it should be noted that although the uninsured utilize considerably fewer healthcare services (partially because they tend to be healthy, but also because they defer treatment as long as they can), when they get coverage they are likely to utilize an increased level of services, at least temporarily. This additional utilization would need to be anticipated.
If open enrollment periods were required, a sound mechanism to pool high risks of the unhealthy would help stabilize the market and financial operations of both co-ops and traditional commercial insurers. This could be done through high-risk pools or through a risk-adjuster mechanism similar to that used by the Medicare Advantage program. Depending on the effectiveness of the risk-adjuster method, health plans might not need to increase rates in anticipation of biased adverse selection. A mandatory high-risk pool might have a similar effect, especially if consistent government support were used to help fund the pool. By contrast, a voluntary high-risk pool, funded only by the insurers participating, would likely not be as successful in lowering rates.
Q:Finally, can co-ops satisfy consumers and begin to control costs?
A: Many critical details will need to be established before we can answer that question. Co-ops, by virtue of their structure, can provide for member direction in overall health plan policy and tradeoffs between coverage, benefits, and service on the one hand, and costs on the other.
If co-ops emphasize integration and coordination of care, and are successful at lowering utilization levels through management of the care provided, they should be in a position to lower—or at least stabilize—premiums accordingly. This is conditioned, of course, on good business management practices being followed, a reasonable cross-section of risks maintained, and realistic regulatory requirements applied consistently across all competitors.
James O'Connor is a principal and consulting actuary at Milliman.