Russell Korobkin, Relative Value Health Insurance: The Behavioral Law and Economics Solution to the Health Care Cost Crisis, Mich L. Rev. (forthcoming 2013), available at SSRN.
Nearly all health insurance contracts currently sold in the U.S. cover all medically necessary, non-experimental services, subject to only specifically listed exclusions. As a result, the coverage provided is what those in the benefits industry would refer to as “rich” coverage. If the treatment is non-experimental and is expected to have a positive clinical benefit, no matter how small, it is covered regardless of cost. This rich coverage leads to some predictable problems. Because individuals typically have little incentive to decline treatment that might benefit them, utilization is high and costs rise accordingly. This, in turn, makes health insurance more expensive for all purchasers. Our health system has tried to remedy this issue by adopting managed care structures to create incentives for providers to limit utilization of a treatment where it has only marginal benefits. And, more recently, consumer-driven health care has been developed to create incentives for patients themselves to reduce utilization of marginally beneficial treatment.
Russell Korobkin’s new article seeks to address this well-known problem through a novel use of comparative effectiveness data to create health insurance contracts that only cover services that provide a given level of cost effectiveness. He refers to this type of insurance as “relative value health insurance.” The basic idea is to start with an index of treatments based on cost-effectiveness, with a proposed scale of 1 for highly cost-effective treatments to 10 for treatments with low cost-effectiveness. Health insurance contracts could then be sold based on the level of cost-effectiveness they will cover. For example, insurers might offer a policy that covered all treatments with a rating of 3 or above for $X, while charging significantly more for a policy that covers all treatments with a rating of 7 or above. Korobkin’s basic argument is that relative value health insurance would greatly simplify an individual’s tradeoffs between medical care and competing goods and services.
The idea of offering a health insurance contract that provides a lower level of coverage for less money is not new. But while there has been theoretical interest in these “pay less to get less” contracts, no one has yet come up with a practical way for insurance companies to structure such contracts, particularly to structure them in such a way that consumers could meaningfully shop for and compare policies. Professor Korobkin’s proposal seeks not only to create a reasonable way to buy “less” health insurance, it also relies on behavioral insights to create a choice architecture that will allow consumers to make meaningful choices, or at least better informed choices, than are possible under current contract structures.
I admit to being skeptical initially of this basic proposal. In particular, I wondered whether any insurance purchaser would actually want to buy insurance that provided perhaps less-than-complete coverage based on cost-effectiveness. What if you develop a life-threatening illness and the treatment your physician tells you is necessary to give you the greatest chance of survival happens to score low on the cost-effectiveness scale? Korobkin’s reply to this concern, which I found helpful, was to make an analogy to car purchasing decisions:
“Few people purchase the safest possible car, completely ignoring the tradeoffs this would entail, as they would do if they were truly to adopt a non-compensatory decision strategy that refused to trade off health and safety against other product attributes. In this sense, the relative value health insurance purchase decision would probably look much the same to most consumers. The promise of infinite and unlimited medical care would be nice to have, just as the safest care that technology can produce would be nice to have. The reality, however, that resources are scarce and dollars spent on medical care cannot be spent on other things likely would encourage boundedly-rational decisionmakers to employ a consciously compensatory decision making approach, leading to more efficient resource allocation decisions.”
There are clearly big hurdles to implementing a system of relative value health insurance, most of which Korobkin acknowledges. First and foremost, we would need to have the data necessary to construct a meaningful index of cost-effectiveness. While the Affordable Care Act funds the Patient-Centered Outcomes Research Institute (PCORI) in order to generate better data to support evidence-based medicine, we may be decades away from having sufficient data to make a relative value index possible.1 And once we have the necessary data on outcomes, we’d need to agree on the methodology for converting the data into an easy to understand cost-effectiveness index. Assuming we could get the index up and running, there are also significant adverse selection problems with relative value health insurance that would need to be thought through and addressed.
To be clear, the article does not present a ready-made perfect solution to growing health care costs that can be implemented tomorrow. Indeed, there are many details that would need to be thought through and ironed out. But the reason this article is important is that it gets the ball rolling on needed discussions of how we can allow consumers meaningful choice in health insurance without simply suggesting that costs and burdens should be shifted to consumers. And it does so using thoughtful insights from behavioral law and economics, and by proposing a system that will take advantage of the clinical outcomes data to which we should (relatively soon) have access.
- The ACA in fact prohibits PCORI from calculating cost-effectiveness measures, although presumably private insurers could provide such calculations relying on PCORI’s underlying clinical effectiveness data. For more information on PCORI and the limitations on cost-effectiveness calculations, see Elizabeth Weeks Leonard, Death Panels and the Rhetoric of Rationing, 13 Nev. L.J. 872 (2013), available at SSRN. [↩]