In addition to general overhead items, which are fixed and therefore have to be spread over fewer enrollees, certain cost components are especially likely to be higher for small groups. The first is the commission paid to the insurance agent or broker who sells the policy. This averages 8.4 percent of claims for a group of fewer than five enrollees, and less than 1.0 percent of claims for groups of 500 or more. A second is the risk or risk and profit charge, which includes both the insurer's profits and some cushion against unexpectedly high claims cost. The risk charge averages 8.5 percent for groups of fewer than five, in sharp contrast to 3.5 percent or less for groups of 500 or more. 16 Small groups face higher risk charges in part because costs for these groups may be

less predictable than for larger ones. But smaller groups also pay a higher risk charge because of their limited bargaining power.

In addition to facing higher initial costs for coverage, small employers sometimes find that their premiums increase sharply after the first year of coverage. A group may start out with a healthy pool, since the group members with a known, immediate need for health services have been excluded. Over time, however, those left in the pool will gradually use medical care at higher rates, leading to increasing costs claimed under the policy. Accordingly, many insurers raise the rates after the first year of the policy. (Some may offer, as an alternative, to re-underwrite the

group, again separating the currently sick from the currently healthy.) Insurers may also quote artificially low rates for the first year in order to gain new business.

One other rating practice that was once rare in the small group market but now appears to be gaining currency is some form of group-specific rating, especially for years after the initial year of the policy. Experience rating, in which premiums are based on actual costs for the particular group, is still uncommon in the small group market. However, some insurers may classify small employers into broad ranges, or tiers, by claim experience. A group with unexpectedly high costs during a year may be reclassed into a new, higher-rate tier.

Finally, employers who present the best risks may respond to annual rate increases by seeking a new insurer who will offer a more affordable first-year rate, a process known as churning. As a result, while insurers are competing more vigorously for the groups that present more favorable risks, their renewal business may increasingly consist of higher-risk groups unable to find a lower price. This spiraling process, which makes insurance less and less affordable for the higher-risk groups, has accelerated in the past decade.

Defects of Public Health Care Coverage

Government provides health coverage for lowincome people through the federal/state Medicaid program. Medicaid has accomplished a great deal: its enactment undoubtedly improved access to care for the segments of the poor population it has reached. But basic flaws in the design of Medicaid prevent it from covering many of those who need help. Some states have tried to fill in with other programs of their own. But funding limitations and other problems prevent these programs from helping most of the millions of poor people without health coverage..

Medicaid Medicaid provides coverage to about 26 million people who are aged, blind, disabled, or members of families with children. Each state designs and administers its own Medicaid program within broad federal guidelines. 17 The federal government pays an average of 55 percent of the cost of the program, a share that ranges from 50 percent to 80 percent depending on a state's per capita income. Combined federal and state expenditures for the program are expected to reach $70 billion in fiscal year 1990, just over half of which (55 percent) will go for hospital, physician and other acute care services. 18, 19

Medicaid does not reach more than a fraction of the low-income population. In 1987, just 42 percent of people in poverty were covered. Even among extremely poor people-those with family incomes below 25 percent of the poverty line-nearly a quarter did not receive Medicaid or any other coverage in that year.20

There are two reasons why so many low-income people do not get Medicaid coverage. An applicant not only must meet income and asset tests to qualify, but also must be in one of the protected populations defined in Medicaid law. Medicaid was designed to cover the welfare population-families with children. receiving Aid to Families with Dependent Children (AFDC) and aged, blind, and disabled persons receiving Supplemental Security Income (SSI). Although current law defines some 40 different population groups potentially eligible for Medicaid coverage, either by federal mandate or under state option, all Medicaid beneficiaries must be aged, disabled, or members of families with children. Almost three out of four Medicaid beneficiaries are welfare recipients.

"Obtaining the benefits one is entitled to by law is not easy. It can be a time-consuming, complex, and demeaning journey. I have three documents here with me today.... One is an application for food stamps. It's 10 pages. On the second page of the application there are 38 questions. This is an application for [Supplemental Security Income] benefits. It is 15 pages long. I quit counting the questions at 100. This is an application for a $100,000 home mortgage. It's three pages long."

-Herb A. Sanderson, deputy director, Division of Aging and Adult Services, Arkansas Department of Human Services

Completely omitted from the program, even if they are literally penniless, are single people and childless couples who are under 65 and do not meet disability tests. And within the protected populations, Medicaid rules continue to use distinctions and exclusions carried over from the welfare programs. An intact, two-parent family headed by a full-time worker, for example, cannot be covered as a family. The children may be covered and the mother is eligible during the period of a pregnancy, but the father is never eligible.

Along with these complex categorical exclusions go enormous variations in the financial standards im

posed by the states. For many of the eligible groups there are no minimum federal standards. A family of three in California could get Medicaid in 1990, for example, with a monthly income of up to $934, about 106 percent of poverty. The same family could get Medicaid in Alabama in 1990 only if its income was $118 a month or less. This cutoff is just 13 percent of poverty.21 Further aggravating the disparities, many states fail to adjust their financial standards adequately to reflect inflation, making the real standards more restrictive over time. The median state AFDC payment level, which determines family Medicaid eligibility

limits in many states, dropped 30 percent in real or inflation-adjusted dollars between 1970 and 1988.22

In the last several years, the Congress has mandated coverage up to higher income levels for a few groups, beginning to separate eligibility for Medicaid from eligibility for welfare. The major focus since 1984 has been on expanding eligibility for pregnant women and young children. The most recent expansion, the Omnibus Budget Reconciliation Act of 1989 (OBRA 1989), requires states, beginning April 1990, to cover pregnant women and children under age six with

The consequences of having no insurance. can be devastating. Diana and Melvin Seeger of Grand Rapids, Michigan, know this only too well.

M MULL

MORIAL
SPITAL

Diana Seeger could not qualify for Medicaid or afford care at a hospital close to home; her baby sustained brain damage at birth.

Mr. Seeger's job as a logger did not provide insurance. And, although he earned just $9,000 a year, the couple's assets made them ineligible to qualify for medical assistance. So Mrs. Seeger couldn't afford to go to either of two hospitals within 20 minutes' drive of their home for the delivery of their third child. Instead, they chose a more affordable hospital an hour and a half away.

On their way to the hospital, Mrs. Seeger went into labor shortly after leaving home. The couple wound up at an Indian out-clinic unequipped to handle deliveries. The baby was deprived of oxygen and as a result was severely brain damaged.

After spending most of his first year in the hospital, their son David was released to medical foster care. Rehospitalized several times for surgery, David died when he was three and a half years old. "It's really tragic," Mrs. Seeger told the Commission. "Had we had health insurance or been eligible for medical assistance, we could have gone to a closer hospital, and David would be alive today."

The bitter irony for the Seegers is that David's care cost $183,000-all of which was paid by medical assistance. Both prenatal and maternity coverage would have cost $2,000-a fraction of that amount. "[A]n entire family went through months of counseling and years of agony," Mrs. Seeger testified before the Commission, and "a baby suffered for three and a half years. In a country that is rich in resources, our baby's life was taken because someone thought they could save $2,000."

family incomes up to 133 percent of poverty. It is too soon to know the full effect of the 1989 changes, but preliminary CPS data for March 1989-reflecting the earlier phases of the eligibility expansion initiativeindicate that the number of poor children covered by Medicaid may actually have dropped between 1987 and 1988.23 The fear is that the effects of expanded coverage for younger children have been outweighed by steadily more restrictive standards (again, in real dollars) for older children. In addition, not all members of the newly eligible populations are actually applying for coverage, since both stigma and administrative barriers deter participation.

Congress also helped Medicaid recipients who lose their eligibility for the program when their income increases. 24 In 1988, Congress required states to continue Medicaid coverage for a year for this group of beneficiaries. While these people are obviously helped by the extended benefits, when their Medicaid coverage ends, they often face a dilemma: working for higher income but receiving no health benefits or slipping back onto the cash assistance program in order to qualify for Medicaid.

All state Medicaid programs, by federal mandate, must cover a basic set of services, including inpatient and outpatient hospital care and physician, laboratory, and X-ray services. Most states also cover a variety of optional services, such as prescription drugs (covered in all states) and dental care. However, on mandatory as well as optional services, states are permitted to restrict the amount of the service, for example, the number of inpatient hospital days or physician office visits, that Medicaid will cover in the course of a year.

Even if the state nominally covers a service, beneficiaries may find that doctors or other providers are unwilling to accept Medicaid or have limited the numbers of Medicaid patients they will serve, largely because Medicaid's reimbursement rates are so low.

For physician services, Medicaid rates are consistently well below those paid by Medicare or by private insurers. In 1986, for example, the maximum Medicaid payment for an appendectomy averaged 61 percent of the Medicare maximum in the same state. Of 47 states reporting, 14 reimbursed physicians at a rate below 50 percent of the Medicare maximum. New York paid just 14 percent of that maximum.25 Obstetricians are especially reluctant to provide deliveries to mothers on Medicaid, claiming that the Medicaid payment rate often will not cover even the cost of the malpractice premium associated with the delivery.26

Older children may lose Medicaid coverage as the federal government expands coverage for the very young.

The effect of Medicaid reimbursement rates on access to other services is less clear. 27 There is no statistical evidence that Medicaid patients face widespread barriers to hospital care, although there are anecdotal accounts that some hospitals discourage or refuse Medicaid admissions. 28 Hospitals frequently report that states' Medicaid payment rates are less than the cost of care.29 Where this is true, hospitals' willingness to continue accepting Medicaid patients may depend on their ability to make up for Medicaid losses by increasing their private rates. It is possible that recent pressure by private payers for discounts is limiting the capacity of hospitals to engage in such cost shifting. If so, Medicaid beneficiaries, like uninsured patients, could face barriers to care.

Other Public Health Coverage Programs-Many states operate partially or wholly subsidized programs for persons who are not eligible for Medicaid coverage.30 These include programs for low-income persons who cannot qualify on categorical grounds, and programs for persons at medically high risk. In some states these programs provide benefits as comprehensive as those furnished by the Medicaid program. In other states benefits are much more limited. States may operate the programs directly, or the programs may be administered and partially or wholly funded at the county level.

According to the Intergovernmental Health Policy Project, 22 states sponsored general assistance programs in 1987, providing medical benefits (and usually cash welfare) to very low-income persons who do not meet categorical tests, such as single men. These states reportedly spent more than $1.66 billion on such programs in 1987. Twenty states operated comparable medically indigent programs for other target populations. 31

General assistance and other medically indigent programs are often called state-only programs, because they are financed without federal assistance. One 1985 survey found that state general assistance and state-only medically indigent programs covered about 1.3 million people. 32 More recent data on the total number of persons covered by these programs are not available.

State-only programs have many of the same problems as Medicaid, compounded by the lack of federal funding. Financial eligibility standards are often even more restrictive than those used for Medicaid, program benefits often more limited, and reimbursement levels often lower. For example, the Illinois general assistance program paid hospitals a flat $500 per admission in 1983, at a time when the state's Medicaid program was paying an average of $274 per day. One study found that enrollees in the general assistance program were refused care by Chicago-area hospitals almost as often as the uninsured. 33

Some states have also developed programs aimed at individuals who are not necessarily low income but who cannot purchase insurance in the private market because they present high medical risk. In some states, public funds are used to help subsidize the programs and make premiums more affordable for participants. In others, the programs are supported through contributions by health insurers, who are thus in effect pooling the cost of covering high-cost individuals. State programs for the uninsurable have thus come to be known by the general term risk pools, even though not all programs use this method of financing.

These programs tend to attract a small number of extremely costly enrollees. As of the end of 1988, risk pools were operational in 13 states, with a total enrollment of about 30,000 members. 34 The extremely limited enrollment is partly because many of the programs are inadequately publicized and partly because, even when premiums are subsidized, the rates are often beyond the reach of most people needing coverage. In some pools, for example, rates are set at 125 percent or more of the average cost of nongroup coverage in the state. Despite these high rates, most pools

EARL DOTTER

suffer large losses. 35 A pool consisting only of sick people cannot rely on the sharing of risks that might be possible if a program enrolled a representative sample of the community. Such a pool must find some external form of subsidy. The growth of state risk pools is therefore directly constrained by the willingness and ability of states (or insurers within the states) to finance them.

WHAT CARE DO THE UNINSURED GET?

More than 13 million people, 6 percent of all Americans, were unable to obtain needed medical care according to a 1986 survey estimate, because they did not have adequate financial resources. And an estimated 1 million Americans tried to obtain needed medical care but were turned away by doctors or hospitals for financial reasons. 36

Just how much less care do the uninsured receive and who pays for the care they get?