Saturday, March 28, 2009

I Cannot Has Fail? Also: Risk-Averse Megan McArdle


Something I always thought would signal the economic end of the world has come to pass: We are on the verge of seeing the abandonment of the very idea of insurance.

I always thought one of the most alarming things about the Clintons was their belief that insurance companies should stop charging people higher premiums if they’re in higher risk categories (such as possessing pre-existing conditions). No one ever wants to pay a cent more for anything, of course, but the whole point of insurance is that each party in the transaction is making a bet based on risk assessment, and without that — and the wonderful, socially-beneficial incentive effects it creates (such as discouraging being a smoker) — insurance really does not exist at all. Welfare, maybe. Disaster relief. But not insurance.

Well, pressured by the government, a group of insurance companies, in typical self-destructive cowardly corporate fashion, is now offering to preempt further government regulation by “voluntarily” ceasing to charge riskier, sicker people more. Of course, that just means having to charge everyone more to compensate, which you’d think would cost the insurance companies some business. Ah, but they’ll only stop the risk-rating if government goes forward with Obama’s plan to mandate health “insurance” for all, the government presumably subsidizing it for those who can’t afford it. In other words, the insurance companies have decided to become a form of welfare (and thus an arm of the state) instead of a form of risk-rating hedging, as long as they’re guaranteed to come out of it with more customers.

Pathetic — and to me, particularly sad, since I’ve long said that insurance actuarial tables are the single most rational thing about the human race: real probabilities instead of wishful thinking, paranoia, hunches, prejudices, or egalitarian pretenses. (And Mom is a former insurance office manager, I admit.) This news about not charging anyone more, then, is a bit like the stock market announcing that by legal fiat (and through government subsidy) there will no longer be losses — great news for about five minutes, and then reality crushes civilization like a deranged bug. (Not to be confused with a gold bug, which is probably what the survivors will have been.)


Not everyone sees the tidy connection between actuarial thinking, risk-taking, markets, and the moral responsibility to pay for your own messes that I do, of course (and if you assume I do see a tidy chain of connections there, it makes a lot more sense of my combination of robotic rationalism and conservative moralism, for anyone who still finds that confusing). Megan McArdle is lazily thought of as more-or-less-libertarian but has, I think, a much stranger and more complex underlying philosophy of risk than the one underlying this humble blog post.

You may have noticed she was an early proponent of a banks bailout (though not quite of the sort that came to pass), is keener to take measures against global warming than most libertarians, favored mandatory health insurance purchases before it was cool, thinks the industrial revolution never would have happened without the principle of limited liability, likes sumptuary laws believe it or not, yesterday even suggested that capital moving around the globe might be too destabilizing (I mean, heck, having an economy is a risk, too), and in general is more sympathetic than most libertarians to macroeconomics-influenced (and business school-inculcated) notions about there being optimal levels of investment and risk-taking for society as a whole that may not jibe with individuals’ own preferences.

We need the risk-takers even when their risk-taking is irrational from an individual cost-benefit perspective, I think she’d argue, a position that could even logically lead to “subsidizing entrepreneurship” in some cases, absent an accompanying fear of government bungling. (This is not so unlike Adam Smith’s view in Theory of Moral Sentiments, by the way, that entrepreneurs are often crazy or emotionally dysfunctional people but ones for whom the rest of us should be very, very grateful. Really, he said that, I swear.)

Add to all this the fact that Megan admits to being a bit neurotic and hypochondriac (I’m not badmouthing, just summarizing), and you get a picture of someone with almost a “precautionary principle” approach to economics and other big systemic issues, at least as compared to most of us rootin’-tootin’, let the chips fall where they may, laissez-faire libertarians, if I’m understanding it all.

But without going as far as my girlfriend Helen Rittelmeyer, who sometimes suspects we’d be better off if we simply stopped using statistics (a radical view you can discuss with her after you hire her to be your best assistant editor), I’m more and more inclined to think, in anarchist fashion, that any talk at all about what the social system as whole “should” do is a formula for disaster or at best disappointment (I for one wasn’t surprised the real bailout was a stupid one — were you?). The optimal savings rate — or the optimal level of health insurance — is probably whatever the hell happens when society becomes libertarian enough to stop talking about “what everyone else should do” in some broad macro sense, because that always ends badly. Doesn’t logically have to end badly. Just always does.


Todd said...

I’m an actuary myself, and agree wholeheartedly with your assessments both regarding insurance and Megan McArdle.

The information contained in actuarially sound insurance rates can often be a bitter pill to swallow, but it’s almost always good medicine. Some people say you can learn a lot about others by looking through their trash. I say you can learn a lot about yourself by reviewing your bill from GEICO.

Megan McArdle is one of my favorite bloggers. She’s a good writer with strong opinions, but her knowledge of economics and support for individual freedom have always struck me as weak. I was very surprised by her endorsement of the bailouts, and my opinion of her has yet to recover fully. It seems like there is precious little space in her worldview for recognition of uncertainty without correspondent fear.

Tim said...

Re: no-loss stock market – there are such things as “stable-

value funds,”
but essentially yields are given up to pay


(Saw “stable-value funds” in an article about AIG)

Tim said...

AIG’s role in insuring “stable-value funds” was one factor

in Bernanke’s reasoning concerning bailing it out, so your

stock market analogy isn’t far off for that category of

investors (truth not far from fiction)

jenny said...

you’re missing the pooled risk component of insurance. the bet isn’t just between the individual and the insurer, but between the insurer and the entire category of insureds. this shifts the incentives somewhat.

Todd Seavey said...

No, that’s my _entire_ point. The whole exercise becomes meaningless — or rather becomes welfare — if it ceases to be based on individualized risk calculations. Force everyone to behave as if their risks and resulting premiums are equal, and it’s a bit like saying, “Some of you should train for months for this upcoming race — really give it your all, people, no slacking — while others should continue to be morbidly obese and watch TV all day, and fear not, because you’ll all be judged as though you ran a fifty-minute mile, in keeping with basic justice.”

Insurance companies have every right to decide how fine-grained the risk ratings should be to make it worth their while to _be_ an insurance company, just as a race run merely to separate the fast half from the slow half might not bother with photo finishes, but saying “Insurance is about everyone together at the same time hedging for collective risk” is as absurd as saying “All the runners are in the race together, so no information is lost if they’re treated as though they ran the same speed.”

Maximum insanity. Let’s give all kids straight A’s while we’re at it and claim we’ve done no damage to education. This is how everything humanity has ever worked for ends, all things reduced to their lowest and most mindless level. If your best and brightest can’t intuit the problem, we’re done for. You’ll notice I don’t speak of a “future” as much I used to.

Diana said...

Slightly tangential to the topic above, since very many of the most prevalent diseases that people die from (and insurance companies spend the most money on) are preventable is there a cost-benefit analysis between insurance companies charging everyone more or micromanaging the people on their plans. For instance I’ve heard of insurance companies asking their clients to call a number from the gym every time they go in. If they go in 120 days or more a year they get reduced premiums.

In this particular case I’m against everyone shouldering more burden when so much insurance money is spent on people who are irresponsible with their own health.

How far away are we from insurance companies putting a pedometer on you or having you photograph everything you eat and drink for a month so you can get lower premiums?

Todd Seavey said...

I think the less regulated the economy, the more likely we’d be to see a drift away from health insurance plans provided through employers (or governments) and that individualized plans would tend to be tailored to customers’ desires: If you wanted to pay slightly less by proving to the insurance company that you were diligently maintaining your health, you could do so, but surely there’d be plans for people less inclined to put up with such intrusions.

One thing McCain had going for him (though not enough — and I didn’t vote for him, and I’m not confident we’d be any better off if he’d won) was the desire to enact health reforms specifically aimed at separating health insurance from employers, which would likely make plans more numerous and fluid _and_ have the potentially even more beneficial side effect of making people more willing to hop around between jobs, less beholden to a given employer and less frightened of losing their coverage if they made a change.

In general, though, I can’t blame insurance companies for wanting you to strive to be less risky — and, as with making people check off smoker or non-smoker, they might even be nudging you toward greater happiness in the process.

I also think insurance companies are the logical de facto arbiters of the “legitimacy” of gay marriage, but that’s a whole different, complicated story.

Brandon Berg said...


While, I don’t like this, it does address a very real problem with medical insurance, namely that you can develop a health problem through no fault of your own, and while the insurance company will cover it, they’ll then jack up your premium to reflect the risk of recurrence. People with congenital health problems have the same problem.

Ideally there would be a way to insure against this sort of thing (i.e., for an additional charge your insurance will not only covers treatment, but also pay out a lump sum to cover the increased premiums). But for whatever reason (probably the fact that very few people actually have to buy insurance individually), this product does not, to the best of my knowledge, exist.

Without a market solution, this is the next best thing. However, I would want the insurance companies to be able to continue to discriminate on the basis of lifestyle factors.

jenny said...

todd – you can’t base insurance on “individualized” risk calculations – the entire mathematical model underpinning the concept is based on population aggregates. attempting to tailor to the individual rather than to broader risk profiles is impossible – and if anyone’s trying to sell insurance based on individualized assessments, it’s worse than impossible – it’s fraudulent.

Sean Dougherty said...

The best thing about differentiated insurance is it puts a price on misbehavior and you can decide whether or not to misbehave. Want to smoke? Eat too much? Fine, but it will cost you. I’ve been known to exceed the speed limit on occasion, and my insurance premiums reflected that for a while. It was worth it.

We’re already seeing the collective risk pool concept applied to such personal decisions and I don’t think anyone likes it.

Sammler said...

Brandon Berg: It is not hard to buy term life policies which have a maximum rate increase (in exchange for a higher initial rate). This is an established example of the “insure my insurance premiums” policy you describe. I am not aware of such policies for medical, rather than life, insurance — but I have never looked.

Sammler said...

Jenny: the point is that insurance companies can deduce risk factors over a large population — the morbidly obese, for example, tend to have a higher risk of early death — and then factor those into the price of insurance for an individual. Thus the individual is penalized for having things in common with identifiable high-risk populations. This is a reasonably blunt instrument, but Mr. Seavey and I both think it is a legitimate role of insurers (who should enjoy a strong presumption in favor of being able to measure the risk they are taking).

I’m not sure what you are saying is “fraudulent”, but I hope it’s not this practice.