Each year, most Americans face a choice of which health insurance plan to enroll in. An important aspect of this choice is often the financial level of coverage provided by the insurance plan. New research investigates whether society benefits from allowing different levels of coverage. Perhaps surprisingly, it finds that consumers are sometimes better off when the insurer offers just a single coverage level. The researchers explain the economic theory behind this result and show its relevance in a particular case, namely health insurance for public school employees in Oregon.
The research examines choices over the financial level of coverage in health insurance. In most cases, whether this type of choice is available will be decided by the entity operating the health insurance market––for example, a firm offering employer-sponsored insurance or a state running an Affordable Care Act exchange. Ostensibly, offering this type of choice is intended to make consumers better off. But does it?
A new study examines this question, asking whether offering a choice over financial coverage levels will always benefit consumers. The theoretical part of the analysis reveals that offering choice can lead to better outcomes when consumers with higher willingness to pay for insurance also have a higher socially efficient level of coverage. (Here, a consumer’s “socially efficient” level of coverage refers to the level that would be best for society as a whole.) But there are times when offering choice is counterproductive.
Although that may seem surprising, health insurance markets are unusually complex. This is partly because when consumers choose coverage levels, they reveal private information, such as how healthy they expect to be. In addition, because insurance lowers the price of healthcare, those who are insured may use more of it than they otherwise would have done. It is these features of health insurance which ultimately imply that consumers can benefit when their choices over coverage are constrained.
The new study examines, in theory and practice, when allowing choice over financial coverage levels will be socially desirable. The answer will vary across different health insurance markets, since the theoretical condition needed for choice to be desirable will not always be met. The researchers show this in a specific health insurance market, that for public school employees in Oregon. The average consumer in that market would be better off if the insurer were to offer just a single coverage level rather than a menu of options.
In U.S. health insurance markets, buyers can often choose from a menu of different financial coverage level options, which we term “vertical choice”. A notable example is the metal-tiered plans (e.g., Gold, Silver, Bronze) offered on Affordable Care Act (ACA) exchanges. In contrast, many national health insurance schemes provide only a single level of coverage. For example, the U.K.’s National Health Service automatically enrolls U.K. citizens in a uniform level of coverage. In both the U.S. and abroad, regulation plays a central role in determining the extent of vertical choice. But as yet, research by economists has provided limited guidance to regulators on this topic.
Are consumers on ACA exchanges actually made better off by having a choice between a high-deductible and a low-deductible health plan? Would U.K. citizens be better off if they could opt into accepting a deductible, in exchange for an up-front payment? The research summarized here introduces a framework for exploring such questions theoretically and empirically.
The basic case for vertical choice is a standard theoretical argument in favor of product variety. In principle, when consumers have more options to choose from, they should be able to match with products that suit them especially well, achieving social efficiency (Dixit and Stiglitz, 1977). But this fundamental benefit of product variety relies on a key condition: that consumers privately optimal choices coincide with the choices that are optimal for society as a whole.
In textbook competitive markets, where sellers’ costs are independent of consumers’ private valuations, this condition will always be met. Prices will reflect marginal costs, and privately optimal choices will coincide with socially optimal ones. Consumers with higher willingness to pay will generate higher gains from trade. But problems arise in markets with “selection”, where those deciding on a particular choice differ systematically from those making other choices, in ways that are relevant to the seller. Health insurance markets are in this category. For example, people in poorer health are more likely to choose a high level of coverage, which affects insurers’ costs.
In health insurance markets, insurers’ costs are inextricably related to consumers’ private valuations. And, asymmetry in information between insurers and consumers (or else regulation) prevents prices from reflecting marginal costs (Akerlof, 1970; Rothschild and Stiglitz, 1976). As a result, consumers with higher willingness to pay may actually generate lower gains from trade. Even if health insurance markets are competitive, regulated, and populated by informed consumers, allowing vertical choice may not lead to better outcomes for society. Since theory is ambiguous on this point, evidence is needed, as in the new study we summarize here.
To explore how insurance can achieve social efficiency, we start from long-established principles of optimal insurance. The social value of insurance is fundamentally to protect consumers from risk, or in other words, uncertainty in out-of-pocket costs. However, insurance also leads to a social cost, because lowering the out-of-pocket cost of healthcare can lead to over-utilization, an idea health economists refer to as “moral hazard.” Maximal welfare is generated when the extra benefit of increasing risk protection is just matched by the extra cost to society of greater coverage (Arrow, 1965; Pauly, 1968, 1974; Zeckhauser, 1970).
This central tradeoff between the “value of risk protection” and the “social cost of moral hazard” plays out on a consumer-by-consumer basis. For example, a high level of coverage protects a consumer against risk, but that consumer may then be more likely to use more healthcare than they in fact need – the moral hazard element. As coverage level increases, the social benefits of insuring risk further may eventually be outweighed by the adverse effect of over-utilization of healthcare.
When deciding whether to offer vertical choice, a health insurance menu designer aims to reflect these social incentives, by inducing consumers to select their efficient level of coverage. But in doing so, the designer must naturally contend with private incentives, namely that consumers with higher willingness to pay for insurance will select higher coverage. The key challenge is that consumers with higher willingness to pay need not be the consumers with a higher efficient coverage level.
Our research considers the design problem facing a health insurance market operator that can offer a menu of coverage levels, and can set premiums. This “market operator” might be an employer offering employer-sponsored insurance, a state running an ACA exchange, or a government running a national health insurance scheme. The operator’s objective is to maximize “allocative efficiency” with respect to consumers and plans. As is standard in employer-sponsored insurance and national health insurance schemes, the market operator need not break even plan by plan, nor in aggregate. If the market runs an aggregate deficit, the operator can simply “tax” consumers’ incomes in some way to make up the difference.
If consumers vary in their choices from the operator’s menu, then the operator has offered what we are calling vertical choice. The analysis reveals the key condition determining whether the optimal menu should feature vertical choice: namely, whether consumers with higher willingness to pay have a higher efficient level of coverage.
The argument starts with a model of consumer demand for health insurance. We use the model to show that willingness to pay for insurance can be separated into two parts: one that is both privately and socially relevant (the value of risk protection), and one that is only privately relevant (the expected reduction in out-of-pocket costs). This latter portion is simply a transfer of expected spending liability from the consumer to the market operator, and therefore eventually to all other consumers. Since a portion of consumers’ private valuation of insurance is a transfer, higher willingness to pay need not imply higher “social surplus”.
To sketch a simple example, consider an unwell consumer who is “risk neutral”, who derives no benefit from avoiding uncertainty in out-of-pocket costs. Allocating higher coverage to such a consumer delivers her a private benefit (because her expected out-of-pocket costs are reduced), but generates no social benefit beyond that. More of her expected healthcare spending is simply shifted to others. If, in response to higher coverage, she consumes more healthcare than is socially valuable, it would be more efficient for her to have lower coverage.
To see how this works out in practice, we use data on all the public school employees in Oregon. The data contain health insurance plan menus, plan choices, and the subsequent healthcare utilization of 45,000 households from 2008 to 2013. The empirical study takes advantage of variation in the plan menus offered to employees, which differ across employees for exogenous reasons. The plan menus are set independently by each of 187 school districts in the state, which in turn select plans from a common superset determined at the state level. In addition, we observe several coverage levels offered by the same insurer with the same provider network. This helps us to isolate variation in the dimension of interest: financial coverage level. The model allows households to differ in their health status, propensity for moral hazard, and risk aversion. We use the model to recover the overall pattern of household types that underlies and explains observed choices.
The estimates imply that all the households studied have a fairly high efficient level of coverage, ranging between a high-deductible contract (with a $10,000 deductible and full coverage thereafter) and full insurance. Contracts outside this range can be immediately ruled out from the optimal menu, as they deliver lower social surplus for every household. Within this range, we find that households with higher willingness to pay are motivated mainly by a greater expected reduction in out-of-pocket spending, rather than by a greater value of risk protection. These high willingness-to-pay households are highly likely to spend past the stop-loss point of $10,000, and therefore face little out-of-pocket cost uncertainty under any contract in the relevant range of coverage levels.
Although there are competing factors, the negative relationship between willingness to pay and “relevant risk” dominates in terms of magnitude. As a result, we find at best only a weak relationship between willingness to pay and efficient coverage level. The optimal menu for this set of consumers consists of a single contract, with a deductible of $1,250 and an out-of-pocket maximum of $10,000.
Relative to the status quo with vertical choice, the optimal menu – the single contract – would yield gains equivalent to $315 per household per year. This gain is equal to 3 percent of average total healthcare spending, and 19 percent of average out-of-pocket cost. However, the welfare gains are not shared evenly in the population. Sicker and larger households fare best under the optimal menu (the single contract), while healthier and smaller households fare best under vertical choice. These lower willingness-to-pay households benefit from the ability to select a low coverage level, thereby reducing the extent to which they cross-subsidize households with a higher willingness to pay. Our results imply that vertical choice may persist in settings such as employer-sponsored insurance because it limits redistribution across these groups.
Importantly, our empirical results may not apply to other health insurance markets. Indeed, we find that if average risk aversion in the population were doubled, the optimal menu would feature vertical choice. Once risk protection accounts for more of the variation in willingness to pay for higher coverage, the revelation of private information through choice becomes more valuable.
On the other hand, our findings likely do apply to some other settings. The negative relationship between willingness to pay and “relevant risk” is central to our results. The relationship follows from two factors: (i) differences in willingness to pay are mainly driven by consumers’ information about their upcoming health needs, and (ii) the lowest relevant level of coverage (the $10,000 high-deductible plan) is reasonably high.
The first factor implies that the highest willingness-to-pay consumers have the highest expected healthcare spending. The second factor implies that such people actually face very little out-of-pocket cost uncertainty, even in the lowest relevant level of coverage. They are so likely to spend more than $10,000 that they know, with a good deal of certainty, what their out-of-pocket costs will be. These two factors seem especially plausible in settings where contracts span only one year, and in which exposure to substantial out-of-pocket cost uncertainty is not efficient for anyone.
It is also important to keep in mind that we assume the market operator can prevent consumers from remaining uninsured, by running an aggregate deficit in the market. So long as at least one inside option insurance contract is provided for free, all consumers will participate in the market. While the practice of providing at least some level of insurance for free is analogous to the situation in a national health insurance scheme and many employer-sponsored insurance markets, it is not analogous to the current situation in the ACA exchanges, where consumers still have a genuine option to remain uninsured. Though our estimates suggest the average consumer would be better off if all consumers participated, they also point to potentially large redistributional effects, which could be politically or economically undesirable.
In sum, our research establishes that offering consumers certain kinds of choice may not be optimal in markets with selection. In the context of financial coverage level, it seems plausible that the best possible outcome is to pool consumers at a single level of coverage. If a choice is offered, the consumers who choose higher coverage may over-consume healthcare, and the consumers who choose lower coverage may under-consume health insurance. While there are many other factors to consider when designing a health insurance market, there are times when simpler is better.
This article summarizes “When Should There Be Vertical Choice in Health Insurance Markets?” by Victoria R. Marone and Adrienne Sabety, published in the American Economic Review in January 2022.
Victoria Marone is at the University of Texas at Austin. Adrienne Sabety is at Stanford University.