Should your employer offer you a single health plan type or is it better to be offered a menu of health insurance options that vary by generosity level. At first glance, the answer seems that more choice is better; different individuals may have different risk preferences; some may prefer higher premiums and more generous coverage while others prefer lower premiums with higher cost sharing. In fact, the Affordable Care Act Health Insurance Exchange plans offer different generosity levels (e.g., Gold, Silver Bronze) to allow for these differences in preferences. Further, other markets benefit from more choice. However, one may worry that allowing such choices could be problematic due to adverse selection; individuals who are sicker would be more likely to choose the more generous plans. In the UK, all Britons receive coverage the National Health Service.
So how would one determine whether more choice of insurance generosity or less is better? A paper by Marone and Sabety (2022) cite papers by Arrow 1965, Pauly 1968 and Zeckhauser 1970 and state that optimal insurance levels are set when coverage sets the marginal benefit of risk protection equal to the marginal cost of utilization induced by insurance (i.e., moral hazard). However, the key issue identified by Marone and Sabety is that “…consumers with higher willingness to pay are not necessarily the consumers with a higher efficient coverage level.” In fact, the authors claim to show that:
…the key condition determining whether the optimal menu features vertical choice is whether consumers with higher willingness to pay have a higher efficient level of coverage
Here the term “vertical choice” means offering more than one level of health plan generosity. One could offer multiple health plans allowing for different service levels but there is only vertical choice if different levels of cost sharing (i.e., deductibles, copayments, coinsurance) are allowed.
The econometric approach used to test their theory is as follows:
We estimate the model using data from the population of public school employees in Oregon [run by the Oregon Educators Benefit Board]. The data contain health insurance plan menus, plan choices, and the subsequent health-care utilization of nearly 45,000 households over the period 2008 to 2013. Crucially for identification, we observe plausibly exogenous variation in the plan menus offered to employees. The variation is driven by the fact that 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, providing isolated variation along our focal dimension.
By having school districts pick from a closed set of options at the state level, it allows for easier comparison. The authors model household utility following Cardon and Hendel (2001) and Einav et al. (2013). The authors findings are as follows:
- There is a high value of insurance. All households have high value of coverage, which may range from a 10,000 deductible with full coverage after (least generous) to 100% first dollar coverage. Insuring everyone at the highest “gold” coverage leads social welfare that is close to optimal.
- In some cases, a single contract is preferred, particularly when the contract is fairly generous. When contracts’ maximum out-of-pocket cost must be spaced along $2,500 intervals, the authors find that it is optimal to offer a single contract set at a relatively generous (i.e., Gold) level.
- In other cases, multiple contracts are (slightly) preferred. When contract’s maximum out-of-pocket cost can be spaced along $250 unit intervals, the authors claim that it is optimal to offer 4 contracts, however, the benefits of doing so are fairly modest.
These results are driven by the fact that (i) there is a high value of generous insurance and (ii) sick people have the highest willingness to pay. Fact (i) implies that a single contract would be fairly generous; fact (ii) implies that if one allows for deviations from this relatively generous plan, this will largely shift cost for sick patients but have little insurance benefit. For instance, if there was $5,000 out of pocket max but I know that I would have $20,000 of cost next year due to my chronic condition, changing the out-of-pocket max will just change my cost, but it won’t really insure against any risk.
Note that the author’s approach does have a number of limitations. First, they assume that increased cost sharing is good in once sense as it decreases moral hazard. For some products (e.g., pharmaceuticals) cost sharing may decrease medication adherence which could result in higher medical cost. This adherence level impact on cost is not taken into account. Second, the paper assumes that social planner is able to pick the optimal single plan. However, would would happen if the social planner was not very skilled and picked a very poor plan. In that case, vertical choice may be highly beneficial since it may include the optimal plan. One could of course say that these additional plans also would be suboptimal, but by offering people choice there has to be–by definition–a chance that some plans are better than others. An extension of this paper with an imperfect social planner would be interesting. The authors do mention that “If health-care providers charge supracompetitive prices, or if there are positive externalities of health-care utilization, it may well be the case that using insurance to induce additional utilization is desirable” and that liquidity constraints are not addressed in this model.