Moral Hazard in Health Insurance (Kenneth J. Arrow Lecture Series) 1st Edition

The backbone of the insurance industry is actuarial science. It is impossible to know whether any given car will be stolen or any given individual will get lung cancer. However, with a sufficiently large population, it is quite possible to estimate the statistical risk of such occurrences – which the insurance industry calls a hazard – to a near certainty. This allows the insurer to establish premium levels sufficient to make a profit from pooling these risks.

Moral hazard is the difference between the actuarial risks in the uninsured pool and the actual claims history when those in the pool obtain insurance coverage. Experience shows that actual claims tend exceed an insurer’s actuarial assumptions for two reasons. First, there is typically an adverse selection in those seeking insurance. People are forward looking, so when they see themselves as especially likely to need insurance, they tend to be the first to seek it. Second, those with insurance coverage have less incentive to use preventive care, and this tends to result in higher loss experience than if they had not had insurance.

For the most part, there is nothing immoral about moral hazard. Let’s face it, if you have a little workshop in your garage, it’s probably more important to have fire insurance. Similarly, if your wife discovers that breast cancer runs in her family, she’ll want to make sure she and her daughters have good health coverage. To the insurance company, these factors may represent an adverse selection, but from the standpoint of the insurance seeker it’s just common sense.

As for changed behavior when one obtains insurance, the whole purpose of insurance is to make an unacceptable risk more acceptable. It’s pretty obviously immoral to leave your keys in the car in the hopes that someone will steal it, but it’s hardly immoral to park near your favorite little out-of-the-way restaurant in a poor neighborhood, even though you have a sense that the neighborhood is not as safe as the gated community where you live.

Whenever moral hazard is at issue, the concept of asymmetrical information is never far behind. As described above, the insurance company only knows how its actuarial assumptions apply to the group as a whole. In contrast, the individuals buying insurance know a lot more about why they initially sought insurance and how much care they’re exercising to avoid claims, and it’s this asymmetry of information that creates the moral hazard.

In the case of health insurance, the information gap between doctor and patient is even greater than between the patient and the insurer. When a car is damaged or stolen, insurance tends to shift the economic burden to the insurance company, but quantifying the loss occurs pretty much on a level playing field. However, in the case of a knee problem or an inflamed gall bladder, neither the insurance company nor the insured is in the driver’s seat. Both depend on the doctor to determine the appropriate treatment, but doctors have their own economic incentives. This adds a further level of moral hazard to the economics of health insurance, hence the importance of a book on the subject.

It’s difficult to know how to describe this particular book. It’s billed as part of a lecture series, but calling it a lecture is something of a misnomer. For one thing, the supposed “lecture” comprises only a quarter of the page count. The rest is foreword, introduction, commentaries, panel answers, and even a 40-page paper from 1963 all loosely strung together. Who are these contributors? It seems they’re all big names in health care economics:

Joseph E. Stiglitz is a former chief economist of the World Bank, and his 2002 book, “Globalization and its Discontents,” is an excellent read. He is perhaps best known for his work on how information asymmetry prevents markets from being efficient, including of course, health insurance.

Joseph P. Newhouse led the famous Rand Health Insurance Experiment, a study involving thousands of people, randomly given various levels of health coverage from 1974 to 1982. The lessons from that study are the foundation of Finkelstein’s lecture, which itself is a valuable contribution because his own book, "Free for All? Lessons from the RAND Health Insurance Experiment" (1993), despite the catchy title, is comprehensible only to those with a PhD in econometrics.

Kenneth J. Arrow was apparently the first to apply the concept of moral hazard to the medical-care industry way back in 1963. His 40-page paper, “Uncertainty and the Welfare Economics of Medical Care,” is apparently the seminal work in the field and is included in this book.

Jonathan Gruber, until recently a lesser light, is credited as a behind-the-scenes architect of the 2010 Affordable Care Act. He only became famous in 2014, when videos surfaced showing him crowing about how they were able to get Obamacare passed due to the stupidity of the American voter. His star has been descending ever since, and I find it mildly surprising that he was included as a commenter for this lecture.

In summary, I give this a mixed review. One Amazon reviewer aptly called it a 5-star lecture with a 3-star translation into a book. It reminds me of the aphorism that a camel is a horse designed by committee. Despite its weaknesses, however, this book has value. Health care is one of the key challenges facing America. These economists have identified certain fundamental forces that shape how our legislative solutions are likely to fare in the real world, and their approach is scientific rather than political. It’s the best work I’ve found on the subject.

The stars shine on. We wish moral hazard did not exist, and maybe it is not as bad as Mark Pauly says it is. But the Oregon data say yes, it's there in health and merits our continuing attention.

A 5 star lecture, with a 3-star effort at translating a lecture into a book.
Accessible to a general audience but still interesting to those on the research frontier.

Timely Volume. Highly recommended.