Designing Ideal Health Insurance
The modern era has inherited two models of health insurance: the fee-for-service model and the HMO model. Both models create perverse incentives for patients and their doctors.
As I wrote in my recent book, Priceless: Curing the Healthcare Crisis, virtually all recent variations on these two models are attempts to ameliorate and control those perverse incentives—usually by introducing features that have a new set of perverse incentives. It is probably no exaggeration to say that the evolution of health insurance is one of cascading perversions, with each new wave of design trying to overcome the bad outcomes of the previous designs.
Under the fee-for-service model, insurance is designed to pay a separate fee for each service rendered, with patients responsible for some portion of the fee— in the form of a deductible, coinsurance or co-payment amount. Under the HMO model, providers receive a fixed fee, irrespective of the amount of service rendered.
When healthcare is perceived as free (the HMO model), patients will have an incentive to consume it until its value at the margin approaches zero. Since the cost of care is well above zero, this implies that unconstrained patients will consume healthcare resources very wastefully. The deductibles and co-insurance that are features of a typical fee-for-service plan are only a small improvement on these distorted incentives. If patients pay 20 percent of the bill, for example, their incentive is to consume care until its value at the margin is worth only 20 cents on the dollar.
On the provider side, the fee-for-service model encourages overprovision— since more service results in higher income for the doctor, hospital or other supplier of care. The HMO model, by contrast, encourages underprovision, since any portion of the fixed fee that is not spent on medical care is available to the providers as take home pay or some other form of compensation.
Readers may wonder why either model was ever found appealing to anyone in the past. The short answer is that both models are the product of the technocratic approach to healthcare I discussed earlier in my book. Both, in other words, ignore economic incentives.
Both models, for example, implicitly assume that (1) the amount of sickness is limited and largely outside the control of the insured, (2) methods of treating illness are limited and well defined, and (3) because of patient ignorance and asymmetry of information, treatment decisions will always be filtered by physicians, who will make decisions based on their own knowledge and experience or clinical practice guidelines. In this way, both models implicitly assume—one way or another—that economic incentives can be ignored.
Although the HMO model is often viewed as the more contemporary, it is actually less compatible with the changes the medical marketplace is undergoing. The traditional HMO model is fundamentally based on patient ignorance. The basic idea is a simple one: make healthcare free at the point of consumption and control costs by having physicians ration care, eliminating options that are judged “unnecessary” or at least not “cost effective.”
But this model works only as long as patients are willing to accept their doctor’s opinion. And that only works as long as patients are unaware of other (possibly more expensive) options.
However, an explosion of technological innovation and the rapid diffusion of knowledge about the potential of medical science to diagnose and treat disease have rendered these assumptions obsolete.
We could spend our entire gross domestic product on healthcare in useful ways. In fact, we could probably spend the entire GDP on diagnostic tests alone—without ever treating a real disease. The new reality is that patients are becoming as informed as their doctors—not about how to practice medicine, but about how the practice of medicine can benefit them. Combine the potential of modern medicine to benefit patients with a general awareness of these benefits and zero out-of-pocket payments, and the HMO model is simply courting disaster. The fee-for-service model is only a slight improvement.
Some believe that managed care and practice guidelines can solve these problems. Imagine grocery insurance that allows you to buy all the groceries you need; but as you stroll down the supermarket aisle, you are confronted with a team of bureaucrats, prepared to argue over your every purchase. Would anyone want to buy such a policy? Traditional health insurance isn’t designed to work much better.
Accordingly, I propose a new approach. It combines an old concept, casualty insurance, with two relatively new concepts: universal Health Savings Accounts (to control demand) and a proliferation of centers of excellence or “focused factories” (to control supply). I will be posting more on this later. I believe this is the approach that would naturally emerge if we relied on markets, rather than regulators, to solve our problems.
[Cross-posted at Psychology Today]