حق بیمه تجدید پذیر تضمین تشویقی سازگار

Theoretical models of guaranteed renewable insurance display front-loaded premium schedules. Such schedules both cover lifetime total claims of low-risk and high-risk individuals and provide an incentive for those who remain low-risk to continue to purchase the policy. Questions have been raised of whether actual individual insurance markets in the US approximate the behavior predicted by these models, both because young consumers may not be able to “afford” front-loading and because insurers may behave strategically in ways that erode the value of protection against risk reclassification. In this paper, the optimal competitive age-based premium schedule for a benchmark guaranteed renewable health insurance policy is estimated using medical expenditure data. Several factors are shown to reduce the amount of front-loading necessary. Indeed, the resulting optimal premium path increases with age. Actual premium paths exhibited by purchasers of individual insurance are close to the optimal renewable schedule we estimate. Finally, consumer utility associated with the feature is examined.

Many illness-related events do not begin and end within the single-year time frame that is typical for private health insurance policies. A person initially in good health who develops a chronic illness may expect to have above-average expenses in subsequent years. If the annual insurance premium is set proportional to expected expense in each year, someone who contracts a multi-year condition would face a substantial and unexpected jump in premiums—something public policy finds undesirable and something which a risk-averse person would prefer to avoid. A potential solution to this problem is for the insurance policy purchased when the individual is still in good health to contain a guaranteed renewability (GR) provision which stipulates that no insured's future premium for the given policy will increase more than any other insured's premium increases. Thus, people who unexpectedly become high-risk will pay the same premium as those who remain low-risk. While the Health Insurance Portability and Accountability Act (HIPAA) of 1996 mandates GR for almost all individual insurance in the United States,1 and while that feature in theory solves the “reclassification risk” problem, there are a few potential problems with successfully implementing GR. In a world of competitive unsubsidized insurance firms, the insurer offering GR faces a problem. If some individuals who are low-risk in period one will become high-risk in period two, a second-period premium incorporating these higher expected expenses may drive away from the plan those who remain low-risk. Low-risk individuals will be attracted to other firms promising to charge them no more than their own expected expenses. In this sense, the promise to “community rate” insurance in the future for a population of initial purchasers is not sustainable in competitive insurance markets. There is, however, a solution in theory to this problem, as suggested by Cochrane (1995) and Pauly et al. (1995). The premium the insurer should charge the low-risk individuals in period one should be higher than the expected expense for that population in period one. The “extra” premium is used to cover the subsequent above-average expenses of those who became high-risk between period one and period two. It would then be possible to charge a premium equal to the low-risk individuals’ expected expense in period two. At this premium, both those who became high-risk and those who remain low-risk would remain with the insurer.2 Moreover, risk-averse individuals will prefer paying the higher premium in period one to paying a lower premium in period one and facing the uncertainty of above-average premiums in period two. The expected lifetime premiums are the same under both scenarios, but “front-loading” the first-period premium produces an “incentive-compatible” schedule of premiums preferable to facing single-period risk-rated insurance.3 However, there is considerable disagreement on whether this feature works in practice as it does in theory. One practical problem is that the “front loading” of the GR premium schedule (in the sense that the premiums in the initial time periods are higher than current-period expected expenses) may be rather large in a multi-period model. Frick (1998) has suggested that these high premiums may pose a problem for young families that are credit constrained, and so they will not seek insurance with effective GR. Fig. 1 illustrates what the multi-period incentive-compatible premium schedule defined by Pauly et al. (1995) would look like for an initially low-risk population facing a constant per-period probability of becoming (and remaining) high-risk; we assume here that high-risk expense is four times the magnitude of low-risk expense and that half of the population has become high-risk by the final period. The incentive-compatible premium schedule PGR under this set of simple assumptions would be high at young ages (in order to cover future claims for those who become high-risk) and would decrease with age (as the number of years requiring pre-funding high-risk claims decreases).4 Thus, the additional premium for a 25-year-old to cover 40 or more years of above-average spending might be quite high.

This paper provides evidence that guaranteed renewability provisions appear to be effective in providing protection against reclassification risks in individual health insurance markets. We have begun by estimating the breakeven GR premium schedule for a benchmark plan using medical claims data. We found that the amount of front-loading necessary to effectively fund GR insurance is mitigated by several factors: low-risk expected expenses increase with age, the likelihood of becoming high-risk increases with age, and high-risk people either recover or die. As a result, the incentive-compatible GR premium schedule increases with age. We then compared this hypothetical premium schedule to data for actual premiums in the individual health insurance market and found them to be rather consistent. Health insurers should be and evidently are familiar with the need for some frontloading of premiums. But what is less clear is whether they are aware of the desires of the youngest low-risks who need to be attracted to the pool. For the youngest of persons with relatively high discount rates, there may some instances of non-purchase due to what is perceived as excessively-high front-loading. Providing subsidies such as tax credits for lower-income families could help alleviate this potential problem. Regardless, it does seem that existing premium schedules come reasonably close to the optimal incentive-compatible patterns of premiums we estimate.