Insurance Pools and Information

Reihan Salaam is a rarity it seems these days: a reasonable conservative who is both knowledgeable about the context of his opinions and willing to engage issues in an intellectually honest capacity.  I've long wanted to find more right-wing thinkers like him to defray my somewhat tendentious reliance on liberal thinkers for political and cultural analysis.  I say somewhat tendentious because, to be fair, conservatism has taken the express train to crazyville lately.

So I was delighted to see that he's blogging at National Review, with comments (although there don't appear to be many!), and that from the looks of things he's providing some quite interesting and reasonable writing.  In a great little posting on what the increasing availability of medical predictors in patients might mean for the entire enterprise of health insurance.  He quotes fellow conservative Mankiw:

If insurance companies do not use these markers, perhaps because of regulation, will the availability of these tests cause the markets for life insurance and annuities to unravel because of increased adverse selection?

Then, economist Stephen Cecchetti:

To understand the problem, think about a simple case in which there are only two kinds of people, those with high and low expected future medical-care-costs lives. Imagine that the insurance company can’t distinguish the two types, so it charges all comers the average cost across the entire population.  For the healthy people, the cost of the insurance will look very high, so they won’t buy it.  That means that the only people who will buy the insurance are the unhealthy.  Realising this, the insurance company will have to raise their price further to compensate for the fact that only the high cost people are willing to buy insurance.  This is the classic “lemons” problem that causes markets to fail and was first described by George Akerlof. 

Salan takes this predicament seriously, avoiding any knee-jerk reaction that ideological pre-disposition might suggest. 

Agree or disagree, it is an excellent and insightful discussion of the underlying issues shaping the private insurance market of the future.

I agree that this is fascinating. Insurance pools rely on a diffusion of risk. But what happens when that risk is removed? Sure, we'll all still face accidents. But I imagine most health costs that are in theory predictable, and do not involve broken limbs or infectious diseases.

Fundamentally, this goes very deep, down to our assumptions about human agency. To what extent are we in control of our lives? Any insurance pool is weighted towards people at greater risk. So what about people who make poor decisions? One model is that they could have chosen otherwise, and so the consequences should be born entirely by them. If you eat too many Cheetos and get type II diabetes, you should pay for your own insulin (this is assuming no genetic pre-disposition to overeating is found). But another model says that we - our choices - are in large part socially determined. So it is through life experience, education, etc. that one individual is more likely to be susceptible to weight gain. That there are social predictors (demographics and social capital trends) for obesity is evidence that we are not all operating with the same level of human agency. How then is risk apportioned? Should one really be considered responsible (that is, should society *not* be responsible) for the burden of this disease?

And then there is the more morally simple question of those who indeed have a genetic (not social) predisposition to disease. Should they bear the burden alone? Society already makes many sacrifices for those we consider "deserving" of our help.

But health care, unlike say, car or fire insurance, has a number of different components. Driving or owning a house is generally egalitarian where risk is concerned. To the extent that insurers are able, they make adjustments. But for most people the difference is marginal (for those of us who don't live in fire-prone environments, or have clean driving records). But what's more, the stakes aren't nearly as high. Health care is infinitely more complex and for many people a matter of life and death, whereas auto or fire insurance is considerably less so.

In other areas of life, we just assume that society should guarantee a basic level of access no matter what - the alternative would be at best onerous, and at worst horrible. For instance public education, roads, parks, etc. are all things that we basically say, "OK, we'll just do this and everyone can have it." The idea that only certain individuals would be allowed to send their kids to school, have access to 911, drive across certain bridges, etc. is almost absurd (even though yes, this was certainly the case through much of American history).
So the discussion is complicated. There are competing interests and compromises to be made. But it is a discussion that, as Salan points out, will only become more critical as society becomes increasingly capable of delivering more useful information.