Ten Problems of Managed Competition

1. Problem No. 1: Health plans would compete to avoid sicker people and attract healthier people.

   As we have already discussed, each health plan would have an incentive to attract healthy people and avoid those likely to generate high health care costs. Since the plans would not be allowed to compete on the basis of price, they would try to compete in other ways - primarily by offering services that attract healthy people rather than sick ones. The general tech niques for engaging in this type of "biased risk selection" are well known. They have been discussed both in Jackson Hole meetings and in the professional literature.68 But while health economists have a general understanding of the problem, their knowledge is far from exhaustive. As a result, there has been no credible proposal for preventing biased risk selection through regulation (including various risk adjustment mechanisms). The issue is discussed in greater detail below.

2. Problem No. 2: The more fiercely the plans compete to avoid bad risks and attract good ones, the more unstable the market would become.

   Health plans could not afford to be known for providing the "best" care " especially for expensive-to-treat diseases " for the reasons given above. In the attempt to avoid sick people " like a game of musical chairs69 " some would be more successful than others.70 The less successful would have higher costs, which would require higher premiums, which would result in fewer customers. Eventually, the less successful insurers would leave the market. Some criticize the current system for allowing insurers to "skim the cream" and to compete based on their "ability to select good risks." Indeed, advocates of managed competition have been at the forefront in advancing these criticisms. However, under managed competition the results would be worse. Under community-rated artificial premiums, there would be far more cream and the pressures to skim it would be far more intense.

   "Competition would force plans to underprovide to the sick and overprovide to the healthy."

3. Problem No. 3: Competition would force the plans to underprovide services to the sick and overprovide services to the healthy.

   Health plans could underprovide services to the sick vis-à-vis the healthy in numerous ways.71 These include (1) making facilities less accessible (e.g., locating in healthier areas and making access inconvenient for the disabled), (2) making patients wait for service, (3) withholding information and (4) providing substandard care.72 Health economists " both those who favor and those who oppose managed competition " generally acknowledge these techniques.

4. Problem No. 4: Buyers would tend to choose among health plans based on medical services rather than insurance services.

   As noted above, the system of managed competition provides insurance against the loss of personal assets due to illness; however, individual health plans offer not insurance but medical services. It is inevitable that both the sick and the healthy will select plans based on the quality of the services they intend to use.

5. Problem No. 5: Because a successful competitive strategy requires limiting expensive services to the sick, fee-for-service plans would be unable to survive.

   By the nature of fee-for-service plans, patients are free to select a physician and physicians are free to practice medicine according to their conscience and their knowledge of medical procedures. These freedoms would make it virtually impossible for fee-for-service plans to avoid the sick if they are in competition with HMOs. Fee-for-service plans, therefore, are unlikely to survive.73 Proponents of managed competition are well aware of this. Jim Cooper, for example, says, "My guess is that fee-for-service medicine will be discouraged and mostly die out."74 One way such a plan might try to survive is by imposing a fee schedule - limiting the amount it pays for different services. But if the plans were allowed to underpay for, say, the treatment of chronic conditions they would use this ability to make their plans more attractive to the healthy than to the sick. According to Alain Enthoven, this cannot be allowed: What about indemnity insurance as one competitor? I rather doubt that this can be reconciled with managed competition. Indemnity insurance means no contract between a carrier and providers, and that means no defined provider group for quality accountability. And without a contract with providers, the insurance may not be paid in full. ... It is difficult to design properly for risk management with indemnity insurance.75

6. Problem No. 6: Health plans would have very weak incentives to find quality-enhancing and cost-reducing ways of caring for the sick.

   That there would be little advantage in finding innovative ways of curing the ill should be clear. The incentives would be to reduce spending in this area, not increase it. But suppose that an innovation succeeded in reducing the cost of care. In today’s market, a facility that discovered cost-reducing innovations in the treatment of a disease such as AIDS would have an incentive to advertise for patients and profit from its discovery. Under managed competition, the most efficient provider for AIDS patients would have an incentive to avoid those patients, as would every other health plan.76 Incentives to such innovation would, therefore, be very weak.

   "Each buyer would tend to get medical services that cost no more than the premium paid."

7. Problem No. 7: If competition were not constrained, health plans would naturally tend to supply each buyer with medical services which cost exactly the same as the community-rated premium.

   Economists have employed the concept of "equilibrium" to describe competitive markets. The concept can also be applied to managed competition. But to our knowledge, no one has ever constructed a mathematical model of a managed competition market. As a result, the existence and stability of an equilibrium has never been analyzed. Intuition suggests that there probably is no equilibrium under pure managed competition, for the reasons given above. If there is equilibrium, it would be very different than the one that characterizes competitive markets.

   As Table I shows, in competitive markets competition tends to cause the price to change until it equals average cost. Thus, to the extent that price is a measure of the value consumers place on a good or service, the marginal benefit people receive tends to equal the cost of producing that benefit.

   The same tendencies exist under managed competition.77 Yet because prices (community-rated premiums) are artificial, cost must change to equalize price and average cost. Take those patients whom we have identified as "unprofitable." If premiums are free to rise for those people, insurers will compete them up to the level of the cost of their care. But if the premiums are artificially constrained at a lower level, insurers will tend to compete the cost of care down to the level of the artificial premium.78 As Table I shows, the reverse pressures exist for those people we have identified as "profitable." If the artificial premiums cannot be competed down to the level of average cost, the tendency will be to compete cost up to the level of the artificial premium.

   These conclusions follow from well-known principles of the economics of regulation. In the United States, we have had decades of experience with regulated markets. [See the sidebar on why price controls won’t work.] For example, under regulations imposed by the Civil Aeronautics Board (CAB) for most of the post-World War II period, the government dictated airline fares. Unable to compete on price, the airlines competed by offering more flights, flights at more convenient times, more spacious seating and other amenities. Price regulation imposed by the CAB was similar to cartel pricing and had the potential to allow the airlines to earn supra-normal profits. However, these profits were competed away as airlines increased their costs by making passenger-pleasing adjustments.79

   "Unable to compete price up to the level of costs, health plans would reduce quality to compete cost down to the level of price."

   The reverse tendency emerges when prices are kept artificially low. Under rent control laws, landlords are prohibited from raising their rents to the level of average cost. Since rents cannot rise, quality tends to fall. Landlords tend to allow housing quality to deteriorate until housing costs equal the government-controlled rent.80 A similar phenomenon is evident in the Medicare and Medicaid programs, where below-cost levels of reimbursement have caused a lowering of the quality of care. [See the sidebar on controls on payments to providers].

   A different way of appreciating this result is to consider it in terms of a basic principle taught in all introductory economics courses: when firms are maximizing profits, marginal revenue must equal marginal cost. Under managed competition, marginal revenue (the amount of premium each enrollee brings to a plan) must be the same for every enrollee. That means that marginal cost (the amount the plan spends on health care) must also be the same for every enrollee.

   FiguresIIA and IIB are simplifications of Figure I. In the diagrams, the "cost-of-care" line shows what would be spent on patients given current standards of medical practice. The artificial premium line is based on the average cost of care for all patients under community rating. The figure illustrates a condition under which healthy people are subsidizing sick people. Clearly, this is what many proponents of managed competition believe equilibrium would look like for each health plan under their scheme. But simple analysis shows that the diagram in Figure IIA cannot be an equilibrium and that it must give way to something else.

   Roughly speaking, an equilibrium exists if no health plan can adjust to become more profitable.81 However, the plan represented in Figure IIA can easily become more profitable if it can lower the cost of caring for its sicker customers. As long as these customers stay in the plan, it will have the same premium income and lower costs. If they shift to another plan, this is advantageous since the sicker customers are unprofitable by definition. On the other hand, the fact that healthier customers are being overcharged " the cost of care is below the premium they are paying " means that other health plans can lure these customers away by providing higher benefit levels for the same premium.

   The illustration in Figure IIA, therefore, cannot represent equilibrium for any health plan under managed competition. Instead, competitive pressures exist to change medical practice until the cost-of-care line coincides with the artificial premium line.82 [See Figure IIB.] This means that health plans would be forced to underprovide services to the sick and overprovide services to the healthy until each person receives health services whose cost of production is equal to the community-rated premium they are paying.

   Problem No. 8: If premiums were risk adjusted, each customer’s cost of medical services would tend to equal that customer’s risk-adjusted premium.

   As we shall see below, most advocates of managed competition favor risk adjustment mechanisms. Under these schemes, the amount of premium paid would be determined by community rating. But the amount of premium a health plan receives would be determined by the expected health costs of the plan’s enrollees. Premiums could be adjusted prospectively by a government agency such as a HIPC or alliance or retrospectively by waiting to see who gets sick, then assessing some plans and using the proceeds to subsidize others.

   "A good way to avoid cancer patients is to have a poor oncology department."

   All such mechanisms cause plans to attach an expected premium to each enrollee.83 And, based on the argument given above, competition will tend to wipe out any arrangement under which some people are subsidizing the health care of others. New entrants into the market (or existing plan) will find it in their self-interest to offer either a lower price or more services to those who are being overcharged. At the same time, health plans will find it in their self-interest to lower the cost of care for enrollees who are being undercharged.

   Problem No. 9: To the extent that government establishes a floor on quality " through regulation or tort law " health plans would tend to provide the minimum level of quality care to the sick.

   Health economists disagree on how far health plans could go in reducing the quality of care they deliver before the government would intervene. Consider the example given by Enthoven:
   Think of drugs like Ceredase, for people with Gaucher’s disease: recommended doses cost $360,000 a year. ... A decentralized private market cannot deal effectively with such situations. If some health plans cover Ceredase and some do not, guess which will get all the Gaucher’s patients?84

   Enthoven’s solution is to have a national health board dictate whether or not health plans must provide Ceredase. However, health economist Mark Pauly argues that these kinds of mandates are inconsistent with the spirit of managed competition. He writes that:
   a sine qua non of managed competition [is] that managed-care plans be permitted to refuse to deliver beneficial technology that they judge not to be worth the price and that negligence law not be permitted to override these choices.85

   Of course, it would be impractical for a national health board to dictate every quality decision to doctors and hospital personnel and to the health plans that employ them. But to the extent that government succeeds in establishing a floor on quality, our analysis implies that competitors will compete the level of quality down to the floor.

   Problem No. 10: Despite competition along the way, a monopoly would tend to arise as smaller competitors were forced out of the market.

   While it’s not clear that managed competition would provide enough stability to allow an equilibrium to emerge, if one does emerge there is likely to be only one firm left in the market. Our prediction of monopoly is based on the fact that the health plans that are the victims of adverse selection will experience rising costs, in what some health economists have called a "death spiral of adverse selection."86 Plans that have a disproportionate number of expensive-to-treat enrollees will have to charge above-average premiums. As healthy people leave or avoid the plan, its cost per enrollee will continue to rise, leading to even higher premiums, encouraging even more healthy people to leave.

   "Ultimately there would be a monopoly - answerable only to government."

   Note that this death spiral occurs precisely because of the community rating requirement. If health plans could charge each new enrollee a premium that reflected the expected cost and risk that enrollee brought to the insurance pool, the spiral could be avoided. Under community rating, however, new enrollees must pay not only for their own insurance (or expected health care), but also for the care of sick people already in the pool. The more sick people a health plan includes, the stronger the incentives for healthy people to flee. Losers in this system, therefore, tend to get progressively worse off. And since biased risk selection will exist as long as there are at least two firms in the market, competition based on risk selection is likely to continue until only one firm is left.

   A Qualification: The Effect of Open Seasons.

   The analysis presented here assumes that patients make choices among insurers based solely on the value of medical services those patients consume. Under that assumption, market equilibrium requires that each patient receive services whose cost is exactly equal to the premium paid. This assumption would be justified to the degree that patients can easily shift back and forth among insurers as their health needs change. The Clinton plan has an annual open enrollment period and allows more frequent changes for "good cause." Presumably, a switch of plans by a seriously ill patient to obtain better medical care would constitute good cause. The Cooper-Grandy and Chafee bills also have an open season but are more restrictive about switching plans at any time. The Nickles-Stearns bill has no open enrollment period and appears to allow continuous movement among plans. However, the Federal Employees Health Benefits Program allows plan changes only once a year, and the managed competition plan about to be implemented in the Netherlands permits changes only once every two years.87 The Jackson Hole Group and other proponents of managed competition argue that open enrollment periods should be infrequent.88

   "The easier it is for patients to switch health plans, the quicker quality will deteriorate."

   To the degree that people are constrained in their choices by open seasons, they must consider the insurance value of the plan they select as well as its direct consumption value. Consider an expectant mother choosing among competing health plans. The direct services she intends to consume are the services of well-baby delivery. However, there is some chance that there will be complications of pregnancy or that her child will be born prematurely and require sophisticated medical treatment and, in that case, the woman would benefit from highly skilled medical personnel. Thus, in selecting a plan she will be purchasing real insurance as well as specific medical services.

   For such major health problems as heart disease, cancer and AIDS, however, it seems unlikely that people will be willing to pay much to insure for expensive treatment while they are healthy - if they can switch insurers at least every 12 months. The tendency will be to select a plan that is strong on preventive and diagnostic services, secure in the knowledge that in a reasonable amount of time one will be able to switch. The existence of periodic open seasons, therefore, causes us to modify somewhat the diagram in Figure IV, in recognition of an insurance component to people’s choices. Yet even with this modification we are left with the prediction that managed care will result in a radical deterioration in the quality of care sick people receive.

   To conclude this section, consider a possible objection to our analysis. Suppose everyone reads this study and becomes aware of the problem. Wouldn’t it be in their self-interest to pay higher premiums to plans that offer expensive, sophisticated medical care just to insure that type of care would be available if and when they needed it? Collectively, it would be. But under managed competition people do not make choices collectively. They make choices as individuals. And their incentive is to join low-cost plans while they are healthy and switch to high-cost plans when they get sick.89

   ---