The top figure illustrates the manner in which money flows from society to the providers of health care. That flow now goes through a myriad of independent and uncoordinated pipes of varying size. Most of these pipes are so small, relative to the overall money flow, that persons controlling their valves cannot exert much influence over the market for health care. Should they seek to constrain prices, they can easily be threatened by providers with loss of access to health for the insured covered by that money pipe. It is a system designed to maximize the transfer of Gross National product from the rest of society to the providers of care.

Indeed, it is a system so constructed that it effectively shields most Americans from knowing, at any time of the year, precisely how much their families spend on health care. The money flow from households to pay for uninsured services, coinsurance for partially covered services, contributions to insurance coverage, and so on is so complex that it takes a special effort to tabulate it all for one year.

It can be argued that this is precisely as it should be, that most other markets in the economy also are characterized by a myriad of money pipes to the suppliers. The counter-argument that has prevailed in

B

B

Most other countries have therefore all but relieved the individual patients from the role of a cost-controlling consumer. Instead of conceiving of the healthcare market as a set of economic transactions between individual patients and providers, they have transferred the economic facet of these transactions to a middle plane where prices are negotiated between associations of third-party payers and association of physicians. In other words, they have concentrated the money flow to providers into one (or at most a few) large pipes whose valves are operated through negotiation, as is illustrated in the bottom panel of Figure 4.

In the Canadian health system, for example, there really is but one money pipe to providers per province. That pipe originates in the provincial government, which administers health insurance in that country. In West Germany, on the other hand, there are over a thousand pipes going from the sickness funds or private insurance carriers to the providers. The money flow through these pipes, however, is coordinated at the level of the state (Land) into allpayer systems which offers the payer a degree of market power similar to that enjoyed by truly singlesource payers. For that reason, it is not inappropriate to think of West Germany's health system as effectively a single-pipe system as well.

Every German household knows what it spends per year for comprehensive health service, for that spending can be read off a simple barometer: the percentage of gross-compensation paid to the sickness fund, or the insurance premium paid to a private carrier. This simple index is very carefully read by the citizenry, by employers and by the government. Figure 5, taken directly from the trade journal of the Local Sickness Funds, illustrates this barometer graphically. The previously cited Health-Care Reform Act of 1988 was the government's reaction to the sharp increase in the contribution rate after 1984. The chief objective of that Act has been to stabilize that contribution rate.

First, the systems enable those who directly and ultimately pay for health care to procure health services at lower monetary transfers to provider per unit of real health service than is paid under the looser American system. Second, these systems enable those who control the valve of the single pipe literally to determine that physical capacity of the health system. They can make favorable turns of the valve contingent upon regional health planning, as is the case in all of these nations.

In short, then, the single-pipe approach probably more so than any other factor enables these countries to allocate to health care a much smaller slice of the Gross National Product than is being allocated in the United States. If Americans are unwilling to countenance that approach and prefer to continue with their myriad-pipe system, they are likely always to pay more for health care, per unit of service and overall, than they otherwise would.

But these single-pipe systems also have drawbacks that should be acknowledged.

Clearly, a single-pipe system is vulnerable to possible mistakes made by the few who wrangle at the single valve. Such a system may allocate less to health care than the citizenry actually would like to see allocated, if it had its choice, and it may also provide less variety in health care than the citizenry might want.

Furthermore, such a system makes it far more risky for entrepreneurs to venture their funds in search of new medical technology, because those at the valve. may capriciously refuse to pay for that technology.

Finally, from the viewpoint of providers single-pipe systems are clearly undesirable. They manifestly tend to reduce the providers' income. There is no reason why normal, income-seeking providers of health care should favor a system that serves to shrink their income.

11.5

Avoiding Adverse-Risk Selection

The term "adverse-risk selection" has varied meanings in discussions on health policy, and it is often misused. For purposes of this discussion, adverse-risk selection refers to situations in which the distribution of actuarial risks among different insurance pools is not perfectly matched by the distribution of premium income needed to cover those risks.

Suppose, for example, that the likelihood of particular levels of health expenditures for an individual

could be accurately ascertained by both the individual and the insurer issuing that individual a health insurance policy. Suppose next that it were possible to group people into distinct cohorts so that all people within a cohort are exactly alike in their likelihood of incurring particular levels of health expenditures. There would thus be cohorts of relatively healthy people and cohorts of relatively sickly people. For a large enough cohort, one could then quite accurately predict the total health expenditure that would be incurred by that cohort for a given future period. On dividing that total by the number of persons in the cohort, one would obtain an average predicted expenditure level. In the jargon of insurance actuaries, and of economists, an insurance premium set equal to this average predicted level would be said to be an "actuarially fair" premium." The premium would be considered "fair" in the sense that it would not force cohorts of relatively healthy persons to subsidize with their insurance premiums cohorts of relatively sickly persons, a cross-subsidy actuaries define as "unfair." 8

An insurance industry that always charged perfectly actuarially fair premiums could be said to be free of adverse-risk selection, because a particular insurance pool's premium income would always fully reflect and cover its own mix of risk.

Two factors of the real world intrude upon the actuary's idyllic world.

First, the individual's likelihood of future health expenditures usually cannot be ascertained with such accuracy. And even if the individual could ascertain it, he or she would surely not reveal it to a prospective insurance carrier if that revelation would drive up the actuarially fair premium. From the insured's perspective, the selective withholding of facts has always been considered fair game in this context. In this asymmetry of information-the fact that the buyer of insurance knows so much more about his or her health status than the prospective insurer ever will know-lies a major source of adverse risk selection.

Adverse risk selection, of course, can originate also on the insurer's side when insurers have some discretion in composing the risk-mix of their members in response to externally determined insurance premiums. It might occur, for example, when a government announced that it will pay a voucher of $ X year per member in some cohort-e.g., the Medicare cohort-and insurance carriers then seek to enroll at that predetermined premium only individuals whose

"Of course, to this "actuarially fair" premium there would be added an allowance for administrative costs.

In fact, only last year the American Council of Life Insurers declared such a cross subsidy as unfair in a nationwide advertising campaign.

actuarially fair premium is below $ X. From the insurer's perspective, such judicious risk picking is probably considered fair game as well.

Quite aside from the problem of adverse-risk selection, however, most of the world considers the actuary's conception of "fairness" ethically repugnant. This is the second intrusion of the real world on the actuary's idyllic model.

The dominant notion throughout Europe and in Canada is that health insurance should not only help smooth the individual's own outlays for health care over time, but that it should also force chronically healthy people to bear part of the health-care costs incurred by the chronically sick, and that the insurance mechanism is the ideal vehicle to effect this redistribution. This notion is fundamental to these countries' idea of community and nationhood. As noted, they call it the Principle of Social Solidarity.

West Germany's Statutory Health Insurance system is a perfect expression of that principle. The system openly uses health-insurance premiums to redistribute income from healthy and high-income households to low income households and to the sick. To cope with the adverse-risk selection such a premium structure might otherwise engender, the system has traditionally limited the choice individuals and their families have among different sickness funds.

Traditionally, the sickness fund for compulsorily insured West Germans has been dictated by location and/or employment. Blue-collar workers among these typically have had a choice, if any, among only one or a few RVO Funds, although white-collar workers had a wider choice among RVO Funds and Substitute Funds. Even the 25 percent or so of the population who are not compulsory members of the Statutory system, but may join that system voluntarily, have much less effective freedom of choice than is typically available in the United States. All Statutory health insurance funds, for example, must offer the same catalogue of prescribed benefits, and they cannot tailor their premiums to the individual's actuarial risk.

Although private insurers do have greater freedom in this respect, their policies also are much more uniform in basic parameters and much more restrictive than is typical in the free-wheeling United States health-insurance market. And, as already noted, their premium structure prescribed by statute-builds in a

• Economists and actuaries typically counter that notion with the argument that the proper redistribution is not from the healthy to the sick, but from the wealthy (healthy or sick) to the poor sick. Such distributions, they argue, had be effected through taxes and transfers, and not through prices (insurance premiums) in the market place.

non-transferable old-age reserve that effectively locks an insured into a particular private carrier for life.

Threats to the Principle of Solidarity

Not surprisingly, this approach to eliminate adverse-risk selection and, at the same time, to effect a redistribution of income through the Statutory system's premium structure has led to considerable strain within the system. Because the individual sickness fund is empowered to levy whatever payroll-tax is necessary to cover the risk-mix of its own members, these taxes vary widely among the funds. Table 3 illustrates this phenomenon with the most recent data. It is impossible to defend these highly different payroll-tax rates with any appeal to social equity, and one must wonder why these differentials have not led to a more open revolt among the insured and their employers alike. But that revolt is starting.

Round One in that revolt is the increasing tendency among large business firms to take their employees out of high-premium sickness funds and to fold them instead into newly established company-based sickness funds (Betriebskrankenkassen) whose actuarial healthcare costs may be lower. As Table 3 shows, such a switch might yield these companies substantial savings, at least in the short run, while their fund members are still young. Naturally, this move is vigorously opposed by the other funds in the system, particularly by the Local Sickness Funds that cover a demonstrably higher average morbidity.

The Local Sickness Funds have entered the fray from another corner. They would like to see legislation mandating inter-fund financial transfers to compensate for differences in the risk borne by individual

funds. Such transfers are already being made in some states within particular types of sickness funds (for example, within the company-based funds), but not yet across types of funds.

Although the Local Sickness Funds do have a legitimate point, their plea is opposed by other funds in the system on the ground that, pushed to its logical conclusion, the policy would convert the Statutory Health Insurance system into one single, national fund, a Universalkasse with one level of payroll tax and one benefit package for all members. Such a result, although perfectly equitable on its face, would make the Statutory system resemble more and more a fully government-financed system, such as Canada's.

Yet another attack on the present system comes from West Germany's neoclassical economists who, like their colleagues on this side of the Atlantic, regard as the only fair and efficient health-insurance system one priced strictly on actuarial principles and one offering every citizen the widest conceivable choice among competing funds. Neoclassical economists believe that governmental paternalism should not override individual myopia and that the individual should be made to suffer the consequences of his or her myopia. Neoclassical economists also believe that health insurance premiums are not a proper vehicle for the redistribution of income-that if society wishes to assist poor and sickly citizens, it should do so with taxes and cash transfers (or, at most, vouchers).

The West German government plans to arbitrate the emerging fight over West Germany's health insurance system with legislation scheduled for 1992. Considerable public debate will precede that legislation. If the philosophy of the Local Sickness Funds were to prevail, the West German health-insurance system would move substantially toward a government-financed system on the Canadian model. On the other hand, if the neoclassical school of thought were to prevail, the system would abandon the Principle of Solidarity altogether and become more like the American health system. The status quo is unlikely to be tenable over the long run.

West Germans, however, would be unlikely to move toward the American health-insurance model which is decried, throughout Europe, as not only wasteful, but also grossly inequitable. A move towards the Canadian model probably would be more palatable to West Germans unless, as seems more likely, the system can evolve towards a more workable middle ground. One such compromise might be a