In a perfect world, everyone would have health insurance and they would only use services they really need, however you figure that out. But insurance has a cost, and that means that people who need more services, i.e., are less healthy, are more likely to buy health insurance and to make claims under it. And once people have insurance, someone else is paying most of the bills, so why not use services that may not be needed. The first phenomenon is adverse selection and the second is moral hazard. Research published by the National Bureau of Economic Research examines the contribution of these insurance characteristics to the cost of insurance and the creation of inefficiencies in the insurance market. (NBER Paper) The author used the introduction of new insurance plans as an opportunity to identify how actual medical spending under a plan varies from that which would be expected if enrollment was random. The test case was one large employer who had one plan in 2005 and replaced it with three different ones of varying generosity in 2006. The actual services covered were the same under all three plans but the deductibles, copays and maximum out-of-pocket costs were different. The use of deductibles and copays is a common technique for attempting to manage moral hazard, but when an individual has a choice among plans with different levels, that may exacerbate adverse selection.
One tricky part of understanding insured consumer behavior is the difficulty of identifying what the “price” they are responding to is. One price is the premium share they opt for when they select a health plan. Once selected, if a plan has deductibles, copays and an out-of-pocket maximum, to some extent consumers may react solely to how much they would pay on a specific day they are seeking service, or they may have a longer view that recognizes that after a certain level of spending, the deductible is exhausted and they pay less. And preventive services have no cost-sharing. Good luck for any consumer trying to figure out what they will pay for a given service on a given day. Anyway, the researchers attempt to ferret out consumer behavior and to identify how much “excess” spending is due to adverse selection in a plan and how much to moral hazard.
As adverse selection theory would expect, the researchers found that older employees and those with higher medical expenditures in the previous year were most likely to select the richest plan. It follows that they would then conclude that the most generous plan design led to $3969 in additional spending compared to the least generous. If selection had been random, that difference would have been $2117, leading them to conclude that 53% of the difference in spending is due to moral hazard and the remainder to adverse selection. They find a significant price sensitivity up to the point that a deductible and out-of-pocket maximum are met; once that occurs it costs people nothing to get more care, so they get lots of it.
The researchers also conclude that if the employer only offered the least generous plan, so that everyone was enrolled in it, the premium for that plan would increase by over $1000, but total spending would have gone down quite a bit. One advantage of offering plans of differing benefit richness is that this strategy may lessen costs for relatively healthy people who choose less rich, and less expensive, plans, but it may also raise overall health spending under all the plans. One factor that was not clearly covered by the research was the premium share for each plan and what effect that might have on plan choice and subsequent expenditures. The paper tackles an important and difficult issue in health plan design and consumer behavior.