Earlier I alluded to the cost risk allocation problem in health care. Costs are high. Who should bear the risk, individuals, insurers, physicians, the government? I suggested that the risk should be allocated differently for different populations. The elderly and ill should, perhaps, bear less risk, the young and healthy more, and so forth. ACOs are an attempt to shift risk to provider organizations. CDHPs are based on shifting more to individuals. Both models can do some good, but for different populations. That’s why I never claim that either is the solution or totally wrong. It’s more complicated than that.
Risk can be allocated differently by diagnosis or treatment too. This dimension of risk allocation is compatible with ACOs and CDHPs, or any number of other designs. The concept is value-based insurance design, something I’ve written about before. It’s a way to take the full conclusions of the RAND health insurance experiment and other, related work seriously and to fine-tune risk allocation accordingly.
From that work, we know that some people–those with chronic illnesses or the elderly–really do suffer worse outcomes when they have to pay for more of their care, while others–healthy, working-age individuals–generally do not. To the extent that a specific type of care would improve outcomes (relative to other types of care or to no care) and/or reduce future costs, that should be reflected in cost sharing. One should pay less for care that is more beneficial.
But if an individual pays less, some other entity pays more. The cost risk is shifted, but to whom? The answer is, “It depends.” Insurers can bear some risk. So can the government. What’s best can be settled via a competitive pricing system. I’ll have a lot more to say about that next week, so forgive me for just leaving it hanging there.
Rather than repeat a fuller description of value-based insurance design and some of the research relating to it, I refer interested readers to a this prior post.
I think it is worth emphasizing the key underlying idea here. The health care cost problem is a risk allocation problem. If risk is allocated optimally, costs will come down (relative to trend) with minimal reduction in (or better an increase in) quality. To recognize that simple fact is already to admit that there is an optimum, we’re likely not at it, and that it can vary by population and even health condition. So, it is complex in detail and not amenable to a one-size-fits-all insurance design.
If past experience is any guide, some will find something to disagree with in what I’ve written above. That baffles me. Does anyone want to spend more for worse outcomes? (It seems so.) Does anyone not think financial incentives and price signals matter in health care as in all other industries? (Apparently.) That’s all I’m talking about, harnessing them to obtain far better results. Welcome to health economics.
