Real insurance

In the FT this week, Tim Harford writes about health insurance. He starts with the public case of a woman denied a new breast cancer treatment and the difficulties of trading-off problems in a public health care system. He then proposes that the best insurance for health concerns would be simply to give patients a check if they are diagnosed with an illness (according to some scale) and let them choose what treatment to spend it on.

Harford has a point: this would result in real insurance. To see this, notice that health insurance currently works to restore health but whether it be private or public treatment, patients have few options. For instance, a patient does not internalise the cost of the treatment at all. When a claim comes in they may as well take treatment because it is free to them. However, for all we know, a patient factoring in all of the risks (including that the treatment might not work), even if they had the money, would not choose it. If that is the case, there appears to be an inefficiency. When we are talking about expenditures in the 100s of thousands, this inefficiency is hardly trivial.

Harford's solution breaks through this by giving the patient the money instead of the treatment. At the very least, they will spend that money on the treatment. But they could do something else. They might choose not to spend it (telling us that is probably the efficient thing) or they might put the money towards a more expensive treatment that is currently denied them by their insurance provider. Again, a more efficient outcome has occured because the patient has signaled their willingness to pay for the additional cost.

While theoretically sound as a base idea, as usual practical issues abound as always happen when you give people money they might look for ways to get their hands on it (other than being truely ill). But, nonetheless, it is certainly worthy of closer examination.