Justifying Health Insurance
Recent discussions about revising or replacing the Affordable Care Act (ACA) raise philosophical questions about the rationale for having a health insurance system. Among these philosophical questions are the extent to which such insurance should be compulsory, and, relatedly, the extent to which the cost of compulsory insurance should depend on risk and ability to pay. As lawmakers continue to debate the path forward for health policy, it is helpful to review the economic and moral justifications for health insurance.
From a utilitarian, or “welfare economics,” perspective, the main purpose of insurance is redistribution. Insurance redistributes money collected from a broad group to those who suffer some misfortune that can be mitigated with money, such as a treatable illness. Those who suffer such misfortunes find greater utility from the money than those who pay premiums but have no misfortune, so this redistribution increases total utility.
An increase in total utility also justifies other forms of redistribution. Progressive taxes—which scale up payment rates as income rises—can increase total utility by taking more from the rich, who lose less utility from a given sacrifice of money than the poor do. Redistribution of income must be limited, however, because it can reduce incentives for work and innovation. If we ignore these incentives, the optimal outcome would be for everyone’s after-tax income to be the same. But income also functions to reward and encourage useful work. Progressive taxes operate as a compromise between these two concerns. But other forms of redistribution, including health insurance, do not have anywhere near the same harmful effect on incentives, as Allison Hoffman has recently discussed.
Consider two extremes, what I will call “market insurance” and “social insurance.” At one extreme, market insurance provides a fixed level of coverage at a price determined by each individual’s level of risk. From an individual’s perspective, for example, the potential loss of $10,000 with an annual probability of 10 percent is worse than the certainty of paying an annual premium of $1,200. The insurance company can make an average of $200 gross profit by selling policies to cover this loss, and the individual still benefits. The greater the risk, the higher the premium. Some insurance markets come close to working this way, such as life insurance.
Market-based health insurance based on risk introduces a source of inequality because it forces those with higher risk to pay more in premiums for the same coverage. Those who must pay high premiums for high-risk insurance are penalized for situations mostly beyond their control, that is, situations not subject to the incentive effects of insurance that might otherwise exist. It is as if people were entered into a lottery that determined who would be at risk, and the losers would have to pay more. We could reduce the resulting inequality by taking from those with low risk and giving to those with higher risk, because the latter, being poorer as a result of the higher premiums, can make better use of extra money in ways that increase their utility.
At the other extreme, “social insurance” provides the same coverage, with the cost covered by progressive taxation (or income-dependent payments, as is partially implemented for Medicare). Because the rich pay more, they may over-pay for the services that social insurance provides, and they would opt out if they could get the same insurance through a market system, as would those at low risk. Thus social insurance functions only with mandatory participation. Such compulsory enrollment increases total utility by making insurance coverage available to those who need it, when they can make good use of the money it provides. Compulsion is a standard function of governments: Different national governments require their citizens to pay taxes, to get some vaccines, to show up for jury duty, to serve in the military, to vote, to respond to census takers, to report some crimes, and, more generally, to follow the law.
Between the two extremes are regulations that limit the information that can be used to determine a person’s risk. If no information is used, then everyone pays the same, and some of those at low risk will go without insurance, if they are allowed to do so, leading to higher prices for those who remain, starting a vicious circle.
In sum, policymakers face a major tradeoff between compulsion, which is required for social insurance, and the utility lost from uncorrected inequality in market-based insurance.
The ACA is a compromise between the two extremes, but closer to the social insurance system. It uses weak compulsion in the form of penalties that increase over time. Yet the ACA simultaneously allows some price differences according to risk—specifically, age. Young people probably still pay more than they get back, on the average, but less so than they would if all ages paid the same. The combination of this small overpayment with weak penalties is supposed to encourage most young people sign up. The age difference is also somewhat justified by distributional considerations: Without it, those who die young would be subsidizing those who live to old age, thus reducing their lifetime utility more than the increase in utility that older people get from a lower cost.
Consider some other possible compromises, closer to the market system. For example, as a supplement the market, the government could subsidize pools of high-risk individuals. A problem with this idea, though, is that it seems to require a cutoff for “high risk,” which thus puts those just on one side of the cut-off line at a disadvantage compared to those just on the other side. Second, because the high-risk category is a minority, the operators of health insurance pools have an incentive to reduce benefits to this minority.
A second compromise would be to try to compensate for the inequality created by a market system through other means, such as more progressive taxation. This is inefficient. Progressive taxation will either prove insufficient to undo the induced inequality or it will redistribute too much to everyone else, thus reducing incentives. For a simple example, suppose that people categorized as “high risk” had to pay an additional $10,000 per year for health insurance. These individuals would thus see their effective pre-tax annual income reduced from $50,000 to $40,000. To fully remedy this effect, policymakers would have to tax the difference between $40,000 and $50,000 at 100 percent, so that people in the lower-risk pool would take home the same level of after-tax income regardless of their pre-tax income in this range. Such policy kills any monetary incentive to earn within this range.
A third compromise would involve separating health insurance into parts: some social—hence compulsory—insurance would remain, while other pieces would be market based, hence optional. The approach reflected in this third compromise is already in effect for most groups covered by the ACA. A basic compulsory package plus optional purchases of additional coverage is technically equivalent to a larger package with optional opt-outs.
It would be difficult to allow people to choose which part of health insurance to opt out of. Emergency services could not reasonably be denied to those who show up at emergency rooms having opted out of insurance for that. Other services have social value beyond their value to the individual patient, such as vaccinations, birth control, or assistive services that allow disabled people to work.
Problems also arise if we allow people to opt out of highly beneficial services on the basis of very low risk for particular services, such as allowing men or infertile women to opt out of coverage for childbirth. Such opt-outs would create the very sort of risk-based inequality that social insurance is designed to avoid.
Should people be allowed to opt out of part of a policy because of some moral objection, such as objecting to coverage for birth control or abortion? This is a difficult question deserving of more attention than I provide here, but it is worth noting that such opt-outs are routinely available in many areas, however odd the “moral objection” seems to be. Some people feel very strongly about these matters, and they are willing to jump through several hoops in order to avoid vaccination, military service, or, indeed, health insurance that covers contraception. In the case of vaccination, states in the United States differ in the opt-out procedure. Although it appears that some states are too lenient, thus allowing vaccination rates to drop below the level required for broader population immunity, in general it seems possible to permit some opt-outs that do not create serious damage, if the procedure for certifying them is sufficiently rigorous.
Considering all the alternatives, it seems to me that the most justifiable policy would make basic insurance (as defined by a formal procedure such as cost-effectiveness analysis) compulsory, but it would also permit add-ons and opt-outs based on strongly held moral reasons. Costs should be covered either from general tax revenues or from individual payments on a sliding scale. Any less coercive policy will not increase utility because more lenient policies would make high-risk people worse off without compensating improvement for anyone else.