Risk Selection

Risk selection occurs when health plans differ systematically in the expected costs or utilization of their enrollees. Economists use the term somewhat differently from the way actuaries use it. Economists emphasize the broad concept that if competing health plans differ in any observable dimension from each other, then plans will tend to attract enrollees that attach greater importance to this dimension. If people who value these plan features differ from average in their health care utilization and costs, then risk selection causes plans to differ systematically in their expected costs and health care utilization. For example, a health plan that offers a prescription drug benefit (or access to a specific teaching hospital, or expanded mental health treatment) tends to attract people for whom this benefit is relatively important. Because people for whom these benefits are important tend to be more expensive than average in using that feature, if any plan offers greater coverage in these dimensions then risk selection will occur: High-cost enrollees end up in high-coverage health plans, and low-cost enrollees end up in low-coverage health plans. Plans receiving the high-cost enrollees experience “adverse selection,” whereas plans enrolling low-cost enrollees experience “favorable selection,” which collectively are called “biased selection.” Biased selection is a particular problem when health plan premium payments are capitated rather than experience rated. Biased selection is perceived to be a serious problem in the U.S. Medicare program (Greenwald, Levy, & Ingber, 2000).

Actuaries use the term risk selection to describe what happens when underwriters introduce eligibility criteria to restrict the kinds of people who enroll in an insurance plan. For example, in the individual health insurance market a health plan may deny coverage to individuals with certain pre-existing conditions (such as cancer, AIDS, or diabetes) or require a medical examination as a precondition for offering insurance. Another form of explicit risk selection is selective disenrollment, whereby unprofitable enrollees are dropped or encouraged to leave a health plan. These forms of explicit risk selection are often illegal or prohibited by employers and the government. The U.S. Medicare program, for example, prohibits explicit risk selection. Implicit risk selection mechanisms, through benefit design, for example, may also cause biased selection. Health plans with lower deductibles and copayments or higher coverage limits tend to attract higher-cost enrollees. Rothschild and Stiglitz (1976) provide the classic description of this process. Recent literature has highlighted that individual choice of health insurance may be influenced in subtle ways by such factors as level of coverage for high-cost chronic conditions, numbers of specialists, degree of access to specialty hospitals or primary care physicians, or even waiting times. Health plans that offer only certain types of providers or that utilize observable forms of care management may also lead to risk selection. Risk adjustment (that is, making payments that vary with assessed health risk) is the most commonly recommended strategy for addressing risk selection among capitated providers.