Thursday, April 10, 2008

A pretty good economist

I have been reading webbed articles by Austan Goolsbee, widely described as Barack Obama's economic advisor. Most of them are pretty good; he's obviously a real economist in the Chicago style, someone who sees economics as a powerful and exciting tool for explaining the world. And he generally favors markets, incentives, and the like.

On the other hand ... . Take a look at his Slate piece on the American health care system. It takes the form of a critique of Michael Moore's proposals but includes Goolsbee's own views of what is wrong and what should be done about it. He writes:

Economists call this "adverse selection" and when there is too much adverse selection—when the health of the people in the uninsured pool is extremely different from the average person in the country—the market may fail completely. Insurance companies may just deny people coverage entirely.

This is a problem at the core of our health care woes. Moore finds scores of examples—people with tumors, heart problems, lost limbs and digits, you name it. And in each case the insurance company finds a way to deny paying for people's illness even though the people actually have health insurance. He also shows people who simply cannot get insurance because they have pre-existing conditions, are too heavy, are too light, and on and on.

Without any rules against cream-skimming, the insurance companies have every incentive to keep dumping the sick people—often retroactively, after they become sick.

This confuses several different issues. One is the failure to enforce insurance contracts, with the result that the insurer who has lost his bet fails to pay off. That may be a serious problem but it has nothing to do with adverse selection or cream skimming.

That case aside, the argument is simply wrong. Insurance companies free to set the price of what they sell have no incentive to avoid insuring people who are bad risks. They can make money insuring good risks at good risk prices and bad risks at bad risk prices.

Adverse selection, as Goolsbee surely knows, requires asymmetric information—a situation where one party to a transaction has information the other does not. If the customer knows more about his health than the insurance company then the decision to buy insurance will be taken as a signal that the purchaser is a worse than average risk, insurance companies will price accordingly, and people who know they are good risks but cannot prove it will be unwilling to buy good risk insurance at a bad risk price. That is the standard problem of adverse selection in insurance and it is the precise opposite of cream skimming. The bad risks end up insured—at a bad risk price—and the good risks uninsured. In Ackerlof's famous sketch of the problem, set in the used car market, lemons sell, cream puffs don't.

All of this assumes that insurers are free to set their prices. Suppose instead that they are required to charge the same price to everyone, or at least restricted in ways that prevent them from charging bad risks the true cost of insuring them. In that situation it will indeed be in the interest of the insurance companies to try to avoid insuring bad risks—to skim the cream off the top. But the problem there is produced not by the market but by price control. The solution is to eliminate the restriction.

What does Goolsbee propose?

Addressing cream-skimming is at the heart of every responsible program for U.S. health-care reform .... These plans take aim at "pooling," for example, by allowing insurance companies to insure an entire state or region as a whole in exchange for serving everyone in that pool—no dropping, no denials, no shenanigans.

For the requirement that the insurance company serve everyone in the pool to have any teeth, it must include restrictions on the prices insurance companies can charge to those they serve. So Goolsbee's solution to a problem created by price control is—unless I badly misread him—price control.

There is, of course, another problem in the background—but one that has nothing to do with adverse selection. Someone with bad health will, on a free market, end up paying more for health care, directly or through insurance, than someone with good health. Many people, quite possibly including Goolsbee, see that as a bad thing that we should do something about. But it is not a problem that insurance can be expected to solve. Once the dice—for bad health or anything else—have been rolled, it is too late to bet on them.

Unless I am missing something, the analysis in that particular piece is simply bad economics, including the misuse of a technical term that the author surely understands. Nonetheless, my reading of Goolsbee's work leaves me more, not less, favorably inclined to Obama. His economic advisor may get some things wrong, but overall he is an economist and one inclined to favor the market.

Rather like Alfred Kahn, another Democratic economist, to whom we owe airline deregulation.

[Readers interested in a more detailed explanation of adverse selection may want to go to the relevant chapter of my webbed Law's Order and search for "adverse selection."]


Anonymous said...

> All of this assumes that insurers are free to set their prices. Suppose instead that they are required to charge the same price to everyone, or at least restricted in ways that prevent them from charging bad risks the true cost of insuring them.

I don’t know what the system in the US looks like (I am from Germany). But in which way are the private insurance companies restricted to inspect their potential customers and to reject them or only accept them for a higher price?

Is there a short summarization of the US health system somewhere on the web that someone could recommend?

Adolfo David said...

Psss, its not perfect (an Austrian economist), but its good for a democratic candidate. I agree.

jimbino said...

The major problem with insurance is that, even if all the risks were known to all parties, only persons incapable of acting in their own self-interest would be insured.

Health and car insurance pay out only about 50% ("claims ratio") of their premiums in benefits, and the injured insured gets less than that 50% because some of the "benefit" goes to fraud, waste and abuse, not to true compensation for injury.

Title insurance is the worst, since the claims-ratio is around 2%, but anyone who freely subscribes to any insurance is a fool, unless he has important information that he doesn't share with the insurer.

If Hillary or Barack wins, call me Amish.

Anonymous said...

I was quite encouraged by the news reported a month ago that Hillary Clinton (or Barack Obama?) had approached the Canadian government to reassure them they were just playing politics by being opposed to free trade.

Good to know that they are simply lying to the public. Makes me think their economic policies would not be significantly worse than the current administration.

Unknown said...

If people differ in their degree of risk-aversion, then you cannot easily infer whether someone asking for insurance is a "bad" risk. An agent who is risk-neutral or nearly so, will never insure at rates that differ significantly from the expected cost. So some good-risk customers insure are simply acting on their risk-aversion (or loss-aversion, perhaps) and some bad-risks don't insure because they are nearly risk-neutral.

If the insurance company can't perfectly screen ex-ante the very bad risks (multi-dimensional), some of those may pool under terms targeted at a different group, then it's no surprise that the insurer attempts some ex-post screening (perhaps some contract clause existed in advance to allow for it).

There are two somewhat distinct events: trying to deny coverage to a bad draw from a good-risk distribution vs. trying to deny coverage to an average draw from a very-bad-risk distribution (who are ex-ante loss-leaders under their pooling contract). From a moral point of view, these may be similar but may strategically be different for the insurer.

David Friedman said...

Pablo writes:

"was quite encouraged by the news reported a month ago that Hillary Clinton (or Barack Obama?) had approached the Canadian government to reassure them they were just playing politics by being opposed to free trade."

That was, in fact, Goolsbee. He was reported to have told some Canadian official not to take Obama's anti-trade comments too seriously. There was a minor flap over it.

The risks of telling the truth.

David Friedman said...

Libertarian is concerned that Goolsbee isn't an Austrian economist. He is, however, a Chicago economist, as am I, so I don't regard that as a problem.

Jonathan said...

It is encouraging that Obama has an adviser you can approve of, albeit with some reservations. Perhaps he'd make a tolerable president and not an Obamination.

But politicians listen to advisers only when they choose. Your own father was regarded by some as the eminence grise behind Ronald Reagan and Margaret Thatcher, but I'm sure he regretted how little of his advice they took.

Adolfo David said...

Lol David F., I know this is not a problem for you. It is just a joke from an Austrian-Rothbardian like me. I usually dont write here, so I wanted to point a curious story: Chelsea Clinton's spokesman admit to have been asked her 10 times about gold standard issue. Lol, only ronpaulmania can explain this.

Apart of jokes, I belive I prefer Obama over John McMilitarist.

Libertarian greetings from Europe.

Donald Pretari said...

If pooling is mandatory, in any sense, it would seem to require some people to pay more money than they would based on their risk. In other words, it a subsidy. The question for some politicians seems to be how to sell this subsidy to the U.S. Taxpayer. If you call it a tax, some people will rightly see it as a subsidy. If you call it pooling, some people will not see it as a subsidy but as a general aspect of insurance and not a subsidy. Am I missing something?

David Friedman said...

Azaria asks about how pooling differs from a subsidy. The answer is that with pooling the subsidy is being paid by the low risk customers. If joining the pool isn't mandatory—and Obama, to his credit, seems uncomfortable with the idea of forcing people to buy insurance—the low risk customers have the option of dropping out of the pool, creating a body of uninsured people—precisely the problem that candidates have been complaining about.

Beastin said...
This comment has been removed by the author.
Beastin said...

I agree that a number of issues are being conflated in these health insurance debates, but I think the core problem being discussed when they trot out the unfortunates is just this: Many people wind up with a serious medical condition and no medical insurance. (Often this is their own fault.) At that point, for perfectly sensible market reasons, there is no way for them to purchase medical insurance at a price that they can afford to pay. That leaves two basic options: treat them anyway or let them go without care. In the extreme these options become pay for their care or let them die.

We do a bit of both in America, but mostly we treat people anyway (once they become sufficiently sick). This is not a very cost-effective way of doing things, and it effectively of penalizes people who buy insurance.

A few years ago I maxed out my graduate-student health insurance. The cost of my initial hospitalization was far greater than my lifetime insurance cap and was nearly an order of magnitude greater than my yearly income. Worse, continuing treatment would cost more each year than I made (before taxes). Naturally, no insurance company would come near me.

The state of New Mexico picked up the tab. That allowed me to continue living, finish graduate school, and get a job with good health insurance. I'm very grateful. I am also an example of the problem with our current system. Ultimately, I was forgiven for having inadequate health insurance.

So that's the problem. What do you think should be done? Treat them anyway or let them die? Remember though, if you choose the first answer you've got to pay for it somehow.

I primarily see universal health care as a way of treating everyone while distributing the costs more fairly.

Anonymous said...

My reasons to be against Obama for president:

Obama ...

* Favors "card check" to allow massive unionization of companies throughout the U.S. (And once this sets in, it could take many years to undo the damage.)

* Favors squelching foreign trade. His advisor sings a different tune. In such a case it makes sense to take the word of the boss, not the subordinate (who can be dismissed at any time or whose influence can be eclipsed by a different advisor).

* Wants to raise taxes on capital gains and dividends to promote "Social Justice". Never vote for anyone who uses the term "Social Justice" in earnest. This type of tax increase also tends to be counter-productive (produces less revenue).

* Shows no indication of any grasp of economics. He will be vulnerable to being conned by his cleverest advisor. (This is what happened to FDR.)

* Is perfectly willing to rob Group A (the "rich") to give the money to Group B (his constituents). If you are in Group A, or you object to this situation on a moral basis, Obama is not for you.

Anonymous said...

Those individuals that are not well insured and consequently get in trouble have made themselves dependent on charity.

I do not think anybody would like to let them just die but the notion that mandatory charity is therefore justified makes me even more uncomfortable.

Health care is so crucial to society that I'd prefer it if government stayed out of it.

Anonymous said...

There is perhaps another possibility - a loan. Health care costs are paid and afterwards the individual must make monthly payments to pay off his debt.

David Friedman said...

Beastin asks the perfectly reasonable question of what I think should be done. I'm afraid the answer is that I don't know enough about what things are wrong with the present system to be willing to offer suggestions about how to fix it—which is part of the reason that my post was on what is wrong with Goolsbee's arguments, not on what should be done to deal with the problem.

In principle, individuals should be responsible for arranging to pay for their own medical services and depend on voluntary charity if they don't. But there are fairly obvious reasons why, at least in a rich society, it's hard to hold to that line. The problem would be less serious if government didn't do a variety of things to hold up the cost of medical services, ranging from restrictions on the development of new drugs by the FDA to restrictions on the supply of physicians via medical licensing. But even in an entirely free market, the problem Beastin raises would still exist.

You can find my views on the relevant economic arguments at:

but that piece doesn't deal with the problem of the tension between the logic of the argument and our gut level unwillingness to let people die when we could keep them alive.

David Friedman said...

Anonymous argues that what Obama says is a more reliable indicator of what Obama will do than what Goolsbee says is. I'm not sure I agree.

In order to get the Democratic nomination, Obama has to take an anti-trade position, so all his taking it tells us is that he wants the nomination and is, like most politicians, willing to say what will get him elected. Once he is nominated, or elected, his incentives are different.

It's true that Goolsbee has no power to make Obama do things. But the fact that Obama chose Goolsbee is some evidence both that Obama has some understanding of economics and that he has some sympathy for a pro-market approach.

Anonymous points out that Obama is willing to rob A and give the money to B. So are both of the other major party candidates.

Beastin said...

Anonymous correctly points out that, in a free-market system, a loan would have been a workable solution to my health-care dilemma. That is because my education level at the time of my illness indicated that, should I survive, I was likely to be able to make payments in the future. In that respect I am not a particularly good example of the hard choice to be made. More commonly, persons with severe chronic diseases that cannot get health insurance are terrible credit risks. They basically have no prospect of being able to pay back the loan. In such a situation, free-market loans will not save them.

Regarding candidates, I like Obama, and I was generally disheartened to read about his positions on free trade. I was also disheartened to hear his economic advisor suggesting that he was just playing to the crowd with his protectionist talk. I would really like to believe that it is possible for a reasonable, competent, intelligent individual to get elected president without lying to the public.

David, as usual, your article was a very interesting and enjoyable read. However, I am, I confess, among those disreputable individuals who argue for government interventions based on, not an ideal world, but a law written just so. It is difficult (and I'm not convinced warranted) to declare a proposed solution unachievable because of generic distrust of government. I like to see specific examples of how government will botch a particular plan.

The following off-topic comments were spurred by your article:

For the most part you seem to treat political support as continuous, but voting is a very discrete system. I think a lot of preferences don't get expressed because they are overridden by preferences with larger weights (applied to a small set of choices). A big part of politics seems to consist of convincing people that some signature issue trumps their other concerns.

Doesn't your discussion of the inefficiencies of government suggest that laws passed which benefit very large groups at the expense of small groups are likely to be Marshal efficient?

You have argued before that variations in intelligence are largely genetic. There is, however, a moderately good correlation between intelligence and income. (I suspect that a real lack of intelligence correlates very strongly with low income.) This would imply that the ability to make a sizable income is partly inherited, making income less a matter of personal utility. Doesn't that weaken your argument against generic subsidies for the poor?

Anonymous said...

I think that the root of the problem is much deeper. Under the current system, the people who are supposed to keep us healthy (doctors and pharmaceuticals) get money when we are sick. I know about high ethical standards and all, but this incentive is quite powerful.

The arrangement that I would prefer is one where I pay a constant amount to my doctor to help me stay healthy, and s/he pays all the costs that are incurred by me getting sick. The doctor is also in a much better position to get insurance for unexpected medical expenses than I am.

Anonymous said...

You can find my views on the relevant economic arguments at:

but that piece doesn't deal with the problem of the tension between the logic of the argument and our gut level unwillingness to let people die when we could keep them alive.

I think that's an important part of the problem, to be sure. My take is that human life has finite value, in that we aren't prepared to spend an infinite amount of money on keeping people alive; but most people don't like confronting that fact personally, either by saying "I can't afford the treatment for condition X" or by saying "this person has a condition that's too expensive to treat, so we should let them die" (consider, for example, what would be the political reaction to doing away with all subsidies for dialysis).

In countries where the government pays for health care, the government ends up rationing care by delays and bureaucratic approval rules, and even, apparently, in some cases saying that you can't purchase faster or more extensive health care if you're prepared to spend your own money on it. (Not an entirely irrational position; if someone with assets can afford expensive treatment X, there will be pressure on government health services to provide X to everyone.) In the United States, the job has gotten shoved onto HMOs. Effectively, I think, they're scapegoats, doing a necessary but brutally unpopular job of rationing and getting resented for it.

But if you came out and said, "If you have a condition that costs more to treat than you and your family can pay, do your estate planning and try for a comfortable death," the position would be wildly unpopular. Obama's occasional incautious statements wouldn't come close.

Anonymous said...

Adverse selection doesn't require asymmetric information. If one party is unable to act on the information, it still happens. (E.g. if an insurance company has to accept all comers at a specified price, the fact that it knows some of them are bad risks doesn't help it.)

jimbino, the reason title insurance payouts are such a small percentage is that most of the cost is for the title search (legal fees). If other insurance really paid out only 50%, why aren't insurance companies more profitable?

Utility isn't linear in money, so it can be quite correct for someone to buy insurance with a low payout.

Automobile insurance is mandatory, with insurance companies competing for the low-risk customers and an assigned pool at government-specified pricing for the high-risk.

jimbino said...


If you think auto insurance pays out more than 50% of the premium, give me the citations please. Ditto for 2% and Title Insurance. (The fact that attorney's fees might take a cut is irrelevant to the person who has the option to go "bare.")

Furthermore, you should know that auto insurance is NOT mandatory in Wisconsin and New Hampshire and, if you are capable of reading the statutes, you will find that it is not really mandatory in California or Texas (nor in many other states) either!

Anonymous said...

William H. Stoddard:

That is something that has amazed me as well.

How come dying on a waiting list is regarded as so much better than dying with an empty wallet? Is that purely a product of socialist environment, or is there more going on? I really dont know, but its interesting.

Anonymous said...

"How come dying on a waiting list is regarded as so much better than dying with an empty wallet?"

That is indeed interesting! Does it have anything to do with a difference between what we prefer for ourselves and what we want others to have? Do the questions

"Would you rather die on a waiting list or because of an inability to pay?"
"Would you prefer a system where people die on a waiting list or one where an equal number die because of an inability to pay?"

have the same or different answers?

Anonymous said...

That is indeed interesting! Does it have anything to do with a difference between what we prefer for ourselves and what we want others to have?

I dont know. Even assuming that a market system would not produce more cures (generous, i know), that would still leave half the population with a better shot at getting their treatment under a market system. Yet by far most prefer 'competing' on a waiting list, even if they are far above average in income.

Im torn between plain irrationality or indoctrination teaching to favor equality over absolute outcomes.

Anonymous said...

The arrangement that I would prefer is one where I pay a constant amount to my doctor to help me stay healthy, and s/he pays all the costs that are incurred by me getting sick.

Isn't this close to how Kaiser Permanente operates?

Anonymous said...

I see only one problem with "allowing insurers to set their own prices in accordance with individual customers' risk": it means that anybody who is both high-risk and poor can't afford insurance at all.

Does this mean the person doesn't get health care? No, because our society frowns on turning people away from emergency rooms. In practice, it means that this person will get all of his/her health care at public expense, through emergency rooms and ambulances, which are among the least cost-effective parts of our health care system.

In other words, David's free-market answer makes perfect sense if health care is treated as a commodity that people can either buy or not. It doesn't make sense if health care is treated as (at least in part) a human right.

See this post (and I'd really appreciate constructive criticism from the folks over here).

Furious Ennui said...

A "good" economist is a bit like saying someone is an "accurate soothsayer", surely?