Bryan Caplan  


Expressive Recycling... Unchecked and Unbalanced
Oddly, even from intellectuals, you rarely hear what is one of the strongest arguments for the bill, namely that personal genome sequencing might mean -- how many years from now? -- that many more people have pre-existing conditions than we currently are aware of.
What I find "odd," is that Tyler doesn't mention how silly this "strong" argument is.  Even Krugman admits that banning pre-existing conditions, by itself, is a recipe for disaster.  And the more pre-existing conditions we know about, the worse the disaster would be!

COMMENTS (10 to date)
david writes:

I think the intuition here is that 'in the limit' insurance pricing approaches the eventual cost of treatment, which would rather defeat the point of insurance. You would be born with your medical bill from birth (at which point we can start invoking Rawlsian veil arguments to justify universal risk spreading, etc.).

At present we wouldn't want this because we want to discourage risky or unhealthy activities - moral hazard and such. But this risk also declines 'in the limit' of improved genetic detection.

RL writes:

If genetic sequencing ever manages to answer health questions with utter detail, we'll know in advance when we'll get sick, with what, and in such context, health insurance won't be needed. Savings will be needed, to pay for known anticipated bills.

But if genetic sequencing merely tells us more details about pre-existing conditions, properly priced health insurance will still play a role.

Take Marfan's, an inherited genetic disease of the connective tissues which places one at increased risk of aortic dissections, which if not treated electively, before they occur, or emergently when they occur, will kill you.

Even if you place all people with Marfan's in a separate group with a known pre-existing condition, it is still the case that there is a spectrum of disease severity. Some dissect in their 30s, others in their 60s. There is still moral hazard: once insured there is marginally less incentive to not lift heavy objects or rigidly keep on one's blood pressure medication--heavy lifting and high blood pressure both increasing the risk of dissection. There is still adverse selection, so those with a family history of tears at 30 sign up before those with a family history of tears at 60.

That is to say, proper risk adjustment for prior conditions doesn't solve all the problems associated with the economics of health insurance, but it does allow more realistic pricing, and to the extent genetic sequencing improves--distinguishing genetically, for example, those Marfanoids more likely to tear early, allowing subsets of pre-existing conditions to be segregated out--the more we move toward savings and away from insurance in handling healthcare costs.

ed writes:

This is a useful thought experiment. What it shows is that what liberals want from a health "insurance" system is not "insurance" at all in the classic micro-econ sense, but rather *redistribution* from healthy people to sick people. We shouldn't be surprised that free markets can't do redistribution. (And this is completely separate from concerns about moral hazard or information asymmetry/adverse selection.)

Or looked at another way, we want to insure against risk as seen behind the Rawlsian veil of uncertainty.

Personally I believe this is a fine goal, so I am sympathetic to some forms of redistribution towards the sick, despite my generally libertarianish leanings.

Les Cargill writes:

We call it 'insurance', but what it really
is is just finance. But even in the usual
insurance markets, if I lose a building to fire,
and you don't, there's a transfer to me. You
gain because if you have a fire, you get the same.
The end invariant - we both still have a building -
is what's being paid for, not the intermediate
transfers to get there.

Health insurance plus medical care equals Max von
Sydow "playing chess with Death" in "The Seventh
Seal." The logic past that point is the logic of
black humor, of Creepy Suzie and Fight Club.

Arthur B. writes:

[Comment removed for supplying false email address. Email the to request restoring this comment. A valid email address is required to post comments on EconLog.--Econlib Ed.]

Guy Thomas writes:

You may be wrong that banning pre-existing conditions is a "disaster." And Krugman may be wrong too.

Neither of you you have done the math to be able to say that. You haven't done the math or simulations to clearly distinguish "disaster" from "slightly messy but very workable compromise." Economists of all ideological persuasions are like this, because it's so easy to bandy around lazy economists' memes: adverse selection, death spiral, yada yada yada. It requires little intellectual effort; and because the arguments are cute, easy, and selfish (yay! libertarian!), it makes you look superficially smart. But quantitatively, you may very well be wrong.

There is lots of recent actuarial literature on genetics and insurance, which basically says that it isn't much of a problem. And that 'adverse' selection may not be adverse after all.

For example try here

"Demand elasticity, adverse selection and loss coverage: when can community rating work?"

or here

"Multifactorial genetic disorders and adverse selection: Epidemiology meets economics"

[Tinyurls replaced with their targets. N.B. Using tinyurls or other character-saving url redirects on EconLog will cause your comments to be moderated. We are not that space-constrained that you cannot enter the full url; and we prefer our readers to be able see where they are going to be sent before they follow a link.--Econlib Ed.]

David C writes:

Where did Tyler Cowen write that a ban on pre-existing conditions should be done by itself? He's just adding one more consideration to be included with all the others. The basic point is that people don't want their rates to skyrocket just because their condition changes. If they lose insurance right when they need it because their higher rates make it impossible for them to afford it, then what's the point of having insurance in the first place? The greater the extremes in health and, therefore, in costs, then the more likely it is a person will lose their insurance when they need it.

But I was 90% certain you already understood all of this, so I have no idea what this post is getting at.

mulp writes:

The argument made at the health care summit was that the solution to preexisting conditions is increased government subsidies to high risk pools that will cover those with preexisting conditions.

If the government subsidies are increased sufficiently as required to ensure that health insurance is available all those who need it because they have a preexisting condition, the need for insurance will be eliminated, except for the coverage it provides for your normal health care spending. But if insurance tracks spending for routine care, then one should not buy health insurance until it is needed, at which point one buys from the government subsidized risk pool that makes your health care affordable.

In other words, the Republican solution was to start over and move to a single payer solution, based on government funded high risk pools which cover all preexisting conditions.

Steve Sailer writes:

Michael Kinsley made Tyler's argument a decade ago. I haven't, however, seen much medical progress toward making it come true in the ensuing years. I think we are a long way off scientifically from having much of a clue about genes and major diseases -- mostly because "your genes didn't evolve to kill you."

Dan Weber writes:

I brought this up on a previous thread, but I didn't have the citation necessary:

This seems to say that insurance companies cannot use genetic tests to decline coverage or jack up rates.

Comments for this entry have been closed
Return to top