Death of Community Rating Has Been Greatly Exaggerated

The idea that everyone in a given geographical area should pay the same for health coverage has come under assault in recent years.

The potential demise of community rating concerns people who think the health care insurance market is moving toward adverse risk selection. They say community rating is the bedrock of our employer-based health care system. Their concerns focus on four market factors: an aging population, an apparent growth in self-selected consumer-driven health care plans, the increasing number of low income and unemployed people who can’t afford individual insurance, and changes in information technology.

Although each of these is certainly having an effect on the market, we are not yet seeing the end of community rating, if it is ever to occur.

Adverse risk selection occurs when an organization that bears financial risk, a health plan or provider group, attempts to avoid enrolling or caring for sick patients who have conditions that result in high and continuing costs.

“Both insurers and providers are concerned that undertaking quality improvement and publicizing quality outcomes will attract patients more likely to have conditions that make them high users of care and high-cost,” say the authors of the Institute of Medicine’s 2002 study of health care quality, Crossing the Quality Chasm.

“Crude” mechanisms

“Much discussion is evolving about an increasing need to segment risk,” says Humphrey Taylor, chairman of the Harris Poll at Harris Interactive in New York, “especially in light of all this interest in consumer-driven health care. Taylor has studied health insurance trends in a series of surveys over the last several years. Some people believe that cost and other problems can be avoided through risk adjustment, with insurers covering older and higher-risk members receiving higher premiums from the employer than those with younger, lower-risk members. But risk-adjustment mechanisms are crude at best. It will be many years before we know whether or not we can adjust risk effectively for the employed population.”

One issue leading to risk segmentation discussions is whether there will be too few healthy people to support the care of sicker Americans in coming years. No doubt the market is changing: An American turns 50 every 90 seconds. In 10 years, that will be nearly 30 percent of the population, and old people use a lot of health care resources. So as the population ages, questions begin to arise about whether the healthy young have the ability — or the will — to subsidize the health care of older generations. That willingness is, of course, at the core of what health insurance is all about.

“In health insurance, the healthy pay for the care needed by the sick,” says Taylor. “But the young and healthy assume, rightly, that their chances of needing expensive medical care are much lower. So if people can choose plans based on their health assumptions, the young and healthy will no longer subsidize the cost of care for the old and sick.”

Consumer-directed health care (CDHC) is defined by most experts today as high-deductible health plans accompanied by consumer contributions to tax-free health savings accounts (HSA). One reason the idea has been receiving a lot of attention is that advocates say it will help control the increase in health care costs by shifting more costs to consumers and giving them some incentive to just say no to tests, procedures, and drugs.

This creates concern for Taylor and others, who say that traditional managed care plans could be left with risk pools containing only the sickest beneficiaries. If that happens — and there is still doubt about whether consumer-driven health care will ever be a major market force — then community rating could become impossible, dying of its own weight.

Appropriate use

And some people believe that CDHC could result in worsening health care inflation, in part because consumers are ill-prepared to make the treatment decisions necessary to make appropriate use of consumer-driven health plans. “What the CDHC concept is missing is any strategy to promote quality health care, and there’s plenty of evidence that improving the quality of care for the 10 percent of chronically ill people who account for more than two-thirds of expenditures is the best way to control costs,” says Karen Davis, president of the Commonwealth Fund. “Also, these plans may be hard to understand for many consumers and their Internet-based tools may not be very easy to use.”

Davis also worries whether the plans exclude many people who need health insurance the most, such as the working uninsured poor who cannot afford the high deductibles and who won’t benefit from the tax breaks that come from investing in HSAs, and she worries whether CDHC will draw healthy people from traditional health plans, thus driving up premium costs.

She has reason for concern, according to a recent study titled “Risk Segmentation Related to the Offering of a Consumer-Directed Health Plan” by researchers at Kaiser Permanente’s Institute for Health Policy. It was published in an August 2004 supplement to the journal Health Services Research. Using a case study approach, the authors examined whether risk segmentation occurred when the employees in a large employer group were offered two consumer-directed plans alongside more traditional, comprehensive PPO and HMO plans.

When looking only at the age, gender, and other demographic characteristics of enrollees, researchers found that the consumer-directed plans did not appear to attract a group that was healthier than the members of other plans. But when they looked at proxies for health status — including prior use of health services, health spending, and pharmacy utilization — they found that workers who enrolled in the consumer-directed plans appeared healthier than their counterparts who remained in the more comprehensive, higher premium types of plans.

“Without risk-adjustment formulas, such significant segmentation of the risk pool could result in a death spiral for more comprehensive types of plans,” says Laura Tollen, MPH, lead author and senior policy consultant at Kaiser. “In a large sense, that’s what consumer-driven care is about. It’s being sold to consumers as ‘buy only what you need.’ In health care, no one seems willing to talk about how what one purchaser does affects what other purchasers do.”

In a seminal article titled “How and Why the Health Insurance System Will Collapse” in the November/December 2002 issue of Health Affairs, Taylor points to two examples of health insurance markets that began to unravel because of adverse selection in the 1980s and 1990s: the coverage of mental health care in the Federal Employees Health Benefits Program (FEHBP) and the New York and New Jersey individual insurance markets.

But, says Taylor, “The most relevant example may be the Australian system, in which young Australians did not opt in to subsidize health care for the old. In these cases, governments intervened to prevent the systems’ collapse. If the employer health insurance market begins to unravel because of adverse selection and defined contribution, presumably the government would intervene to prevent its collapse.

“In health insurance, an educated consumer with a choice of a broad range of health plans would be our worst nightmare, leading to adverse selection and the collapse of employer-sponsored health insurance,” Taylor concludes.

A significant paradox in all this is that risk segmentation does not appear to be a particularly effective means of cost control, and community rating apparently is not a primary factor in health care inflation. A study published in January 1999 by the National Bureau of Economic Research in Cambridge, Mass., concluded that community rating does not lead to what University of California-Irvine economist Thomas Buchmueller calls an “adverse selection death spiral.”

“Community rating continues to be a source of considerable controversy,” says Buchmueller in a NBER paper titled “Did Community Rating Induce an Adverse Selection Death Spiral? Evidence from New York, Pennsylvania and Connecticut.” “Some industry groups and economists have argued that, contrary to the intended effect of insurance reform, community rating might actually reduce insurance coverage.”

The result, people feared, would be an “adverse selection death spiral” in which, in an unregulated market, premiums will be risk-rated, says Buchmueller. “Since community rating requires insurance companies to charge the same premium to all consumers, the rate charged to younger individuals may rise in response to the higher costs imposed by older persons with the same coverage. If enough younger consumers react to this price increase by dropping coverage, expected claims for the covered group will rise, which will necessitate a further increase in premiums,” says Buchmueller.

Tale of 3 states

But that’s not what happened. In the early 1990s, the three states diverged sharply relative to health reform: New York enacted comprehensive small group and individual market reforms that included pure community rating in both markets; Pennsylvania enacted no reform; and Connecticut enacted moderate reforms.

What Buchmueller found flew in the face of conventional wisdom, he says. “Specifically, the percentage of individuals in small firms covered by insurance did not fall in New York relative to Pennsylvania or Connecticut. New York did experience a dramatic shift away from indemnity insurance toward HMOs. While this shift took place during a period of nationwide increases in the managed care enrollment, the increase in HMO penetration in New York’s small group and individual markets was significantly greater than in either Pennsylvania or Connecticut…. Contrary to the worst-case scenarios depicted by the insurance industry and other critics of reform, there is no evidence that the reforms led to a significant increase in the number of uninsured individuals.”

A study published in the Journal of Risk and Insurance and titled “Why Are Managed Care Plans Less Expensive: Risk Selection, Utilization, or Reimbursement?” also has some interesting results. Researchers at the University of Pennsylvania observed the degree of risk selection among people who switched between HMO and indemnity plans.

Using a large database of nonelderly enrollees, they found that people who switched from a non-HMO to an HMO plan used 11 percent fewer medical services in the period prior to switching than people who remained in a non-HMO plan. This relatively low use persisted once they enrolled in an HMO. In addition, people who switched from an HMO to a non-HMO plan used 18 percent more medical services in the period prior to switching than those who remained in an HMO plan. They concluded that “there may be little need for employers to risk-adjust insurance premiums or otherwise restrict employee choice of plan types.”

Individual health market

An area where adverse risk selection is increasingly prevalent is in the individual health market, a market where there are few state demands for community rated products. And an increasing number of low-wage earners (who generally shift their health care burden to taxpayers) are finding health insurance virtually impossible to afford. But political and practical realities seem to dictate that community rating will remain a vital part of employer-sponsored health care. For example, virtually all states require that HMOs offer a community rated product.

Also, driven in part by the demands of an aging generation of baby boomers, an increasing number of HMOs are offering disease management products that address chronic care. That implies at least that they are adapting their services to meet the evolving needs of their beneficiaries, rather than designing their products to eliminate the chronically ill. “If a health plan has a good program for identifying and managing those with diabetes, it is likely to enroll more diabetic patients,” say the IOM authors.

It appears that for now, at least, adverse risk selection by managed care plans that offer employer-sponsored health care policies — which covered 60 percent of the population last year — is not occurring at a significant rate.

Another 27 percent of the population is covered by government health insurance programs, which generally can’t practice adverse risk selection. The exception is Medicare Advantage, under which HMOs are allowed some degree of risk selection based on health status. That program has about 4.5 million enrollees, about 12 percent of Medicare beneficiaries.

What will happen tomorrow remains anyone’s guess. “It is likely that insurers will game the system, finding new ways to attract and recruit good risks, so that risk adjustment slows but does not prevent the death spiral of adverse selection,” says Taylor.

WHO defines risk

The World Health Organization described the concept of risk in health care in the World Health Report 2002, defining it as “a probability of an adverse health outcome, or a factor that raises this probability.” Other risk-related definitions are:

  • Prevalence of risk — the proportion of the population exposed to a particular risk. For example, the prevalence of smoking might be 25 percent in a particular population.
  • Relative risk — the likelihood of an adverse health outcome in people exposed to a particular risk, compared with people who are not exposed. For example, if people who smoke for a certain time are, on average, 15 times more likely to develop lung cancer than those who do not smoke, their relative risk is 15.
  • Population attributable risk — the proportion of disease in a population that results from a particular risk to health.
  • Attributable burden — the proportion of current disease or injury burden that results from past exposure.
  • Avoidable burden — the proportion of future disease or injury burden that is avoidable if current and future exposure levels are reduced to those specified by some alternative, or counterfactual, distribution.
MANAGED CARE November 2004. ©MediMedia USA

Our most popular topics on