Marginal Revolution: Adverse selection is NOT the problem

May
S	M	T	W	Th	F	St
				1	2	3
4	5	6	7	8	9	10
11	12	13	14	15	16	17
18	19	20	21	22	23	24
25	26	27	28	29	30	31

« Risk vs. uncertainty | Main | Tyler Cowen begs for hate mail »

Adverse selection is NOT the problem

The adverse selection story is a wonderful example of McCloskey's argument that great rhetoric persuades even when it shouldn't. The market for lemons is simple enough for your friends to understand but profound enough for them to be impressed at your learning, so it's a hard story not to tell!

The facts of the matter, however, are that adverse selection is not an important part of the market for automobiles (trucks), or of auto, life insurance or health insurance (on the latter see below).

One reason adverse selection may not be that important in practice is because buyers and sellers use testing and certification to remove the most important information asymmetries. You can buy a decent used car, for example just get it inspected or certified. Only if such adjustments are illegal, or in some other way not allowed, will adverse selection become important.

Second, the asymmetry may run in favor of the sellers. Do I really know more about my own life expectancy than an insurance firm that has access to sophisticated actuarial models? And, assuming that I do have extra information is it all that important? After all "the race is not to the swift, nor the battle to the strong, neither yet bread to the wise... but time and chance happen to them all." Or, more prosaically, the signal is near irrelevant when the noise to signal ratio is high.

Third, propitious selection can be more important than adverse selection. What sort of person buys a lot of life insurance? Is it people who expect to die soon? Or is it the sort of person who is so worried about not leaving their family in trouble that not only do they buy life insurance they also buckle their safety belt and eat healthy? The price of life insurance falls the more you buy so evidently insurance companies believe it is the latter.

Everyone talks about adverse selection in the market for health insurance but in fact non-group policies in these markets are not relatively expensive and not hard to get. The national average annual premium for reasonably generous coverage for a single person is just $2,268.

Sure, that's a lot of money but the point is that it's not a lot relative to what an employed person and their employer would pay for similar coverage in the group market. There is no evidence for an adverse-selection death spiral in the market for health insurance. That's not surprising because non-group health insurance is medically underwitten (i.e. medical inspections just like car inspections). Most people are accepted a few are not. Only in states that require insurance companies to accept all or most buyers are rates high relative to the group market (rates in New Jersey, an outlier, are almost three times as high as the national average.)

There are problems in the health insurance market, including a lack of long term insurance, job lock and the inequity of affordability, but adverse selection is not one of them.

Thanks to Bryan Caplan, Robin Hanson, Tyler Cowen, Tim Harford, and Ray Lehmann for discussion.

Addendum: Comments are open.

Posted by Alex Tabarrok on December 13, 2005 at 07:16 AM in Economics | Permalink

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/t/trackback/3576/3840127

Listed below are links to weblogs that reference Adverse selection is NOT the problem:

» Adverse news for adverse selection from Asymmetrical Information
Alex Tabarrok has a marvellous post up on adverse selection. As he puts it, the central insight of adverse selection ". . . is simple enough for your friends to understand but profound enough for them to be impressed at your learning. so it's a hard st... [Read More]

Tracked on Dec 13, 2005 12:51:44 PM

Tracked on Dec 17, 2005 9:57:54 AM

» Texas Plans To Really Screw You, Hardstyle. from The One-Handed Economist
I'm not much for local politics, or for updating my blog regularly, but a headline in today's Express-News made me angry. Superpowers notwithstanding, I shall blog. Texas is considering, as part of a tax-reform package, a measure that would charge... [Read More]

Tracked on Apr 7, 2006 3:44:38 PM

» Adverse selection is so the problem from Whimsley
Alex Tabarrok at Marginal Revolution claims that Adverse selection is NOT the problem. This is not surprising given his enthusiasm for Markets Everywhere. Most of the comments address the empirical aspects of one particular case, which is the insurance... [Read More]

Tracked on Apr 23, 2006 6:31:38 PM

Tracked on Apr 23, 2006 9:10:19 PM

Comments

"There is no evidence for an adverse-selection death spiral in the market for health insurance. That's not surprising because non-group health insurance is medically underwitten (i.e. medical inspections just like car inspections). Most people are accepted a few are not."

I'm not really sure what you're trying to show. It is correct that "adverse selection death spirals" are not a problem in the health insurance marketplace because, as you note, insurance companies carefully avoid insuring sick people to avoid going broke by paying their health care bills. But I don't think that when people complain about the problem of "adverse selection" in the health insurance marketplace they are concerned with the welfare of the insurance companies; rather, they are worried about the sick people who can't get coverage. I may be missing your point here so I'd appreciate any clarification you can offer.

Posted by: alkali at Dec 13, 2005 9:53:34 AM

Non-group health insurance for a single person may be relatively affordable in most states, but family coverage tends to be very expensive.

Posted by: Peter at Dec 13, 2005 10:02:49 AM

To answer alkali's question I am mostly speaking to economists who use the adverse selection argument to suggest that there is in a technical sense a "market failure" in health insurance markets. It's not a market failure, however, if bad cars don't sell for much. Nor is it a market failure if people who are born with medical problems have to pay a lot for health insurance. These are issues of distribution not efficiency. Issues of distribution are important, but we need to correctly understand the problem we are dealing with to have an appropriate solution.

Peter, the national average premium for family coverage is $4,424 - again not expensive relative to what you would pay in the group market.

Posted by: Alex Tabarrok at Dec 13, 2005 10:16:40 AM

You fail to understand the difference between individual and group coverage. The HIPAA regulations regarding preexisting conditions, the COBRA requirements, the lack of medical underwriting, the renewal clauses. These are incredibly valuable options and comparable coverage is not available in the individual market. Thus, individuals, who are deemed basically healthy (medical underwriting via tests) are required to pay (on average) the same as groups (no medical underwriting and valuable protections). Your own data thus points to a vast disparity in the two markets.

Posted by: elliottg at Dec 13, 2005 10:35:03 AM

My gut reaction is your on crack. I have never heard of anybody at anytime ever say that individual coverage is close to the cost of group coverage. I've looked into individual coverage during a period of self-employment, the cost was factors more than the $50 I pay now (my company is cheap, so I assume they don't pay more than 50, probably 0).

At any rate, if there was a problem, it wouldn't be markets, which I think is your point. Health insurance is probably the most regulated industry in the economy.

Posted by: cb at Dec 13, 2005 11:03:50 AM

Argh!! I just clicked through for your source and found it was the AHIP. I should have done that before I wasted bits responding at any level; serves me right. If you use BS sources then your arguments will sound like BS - a variation of the garbage in, garbage out maxim.

Posted by: elliottg at Dec 13, 2005 11:09:09 AM

[Everyone talks about adverse selection in the market for health insurance but in fact non-group policies in these markets are not relatively expensive and not hard to get. The national average annual premium for reasonably generous coverage for a single person is just $2,268.]

I have a real problem with people using statements about pricing as evidence one way or the other on adverse selection. It's meaningless. If you can price for a risk, it's not a case of asymmetric information. What we need to hear about is the excess or co-pay; how does this differ between pooled and individual policies? Also that average is a somewhat misleading number because it's the average over underwritten policies in cases when the denials rate was as high as 20%. If you cut out 20% of cases, of course the average is going to be lower; we would need to know the cost of providing insurance to the cases that were not underwritten to get a meainingful measure because not insuring a medical risk doesn't make it go away.

And Alex's "Second" above is a really bad argument about adverse selection. The overall signal/noise ratio may be low, but it will be higher in some cases than others and the high cases are the ones that create the problem.

Posted by: dsquared at Dec 13, 2005 11:25:01 AM

elliottg, can you explain why the AHIP data is garbage? And even if there are prices differences between group and individual premiums, couldn't this reflect the fact that the government subsidizes many kinds of group premiums?

Posted by: Javier at Dec 13, 2005 11:27:54 AM

Alex's point is a good one, but easy to misunderstand.

The main problem with the "health insurance" system is one of distribution,
not insurance. We want to be able to efficiently redistribute from healthy
people to sick people. Large group employer policies do this to some degree,
but in general the free market is not designed to provide redistribution.

Imagine we had technology to perfectly predict everyones future health. Then
there would be no hidden information, no risk, and no need for classic "insurance."
But society would still want a way to give resources to the sick people.

I wish Alex would tell us his opinion on why New Jersey has so much higher rates than states without community rating.

Posted by: ed at Dec 13, 2005 11:51:21 AM

I don't understand why everyone is so hostile to Alex's very good points. In insurance markets, I find that when people complain about adverse selection or "uninsurable risks", they are really demanding free insurance. Just because a premium is costly does not mean the market has failed. Plenty of allegedly uninsurable risks (such as flooding, hurricanes, sickness, etc.) can in fact be priced efficiently by the market. Consumers just balk in a childlike fashion when they are confronted with the true cost of their lifestyle decisions and/or desires. Most allegations of market failure in insurance are in fact demands for a subsidy. As Alex points out, this is a question relating to distributional fairness, not market failure. Too many people believe "affordable health insurance" means a right to unlimited free health care. Simply pretending that something shouldn't be costly doesn't mean that it won't be in the real world.

Posted by: Will at Dec 13, 2005 12:19:09 PM

I wish Alex would tell us his opinion on why New Jersey has so much higher rates than states without community rating.

Alex, the above quote points something out, albeit indirectly, adverse selection can be a problem when laws prohibit the market from working.

Take "community rating," for example. Depending on how fully implemented the system is, community rating forces insurers to sell to all people for the same price, regardless of health condition and other factors. Surely adverse selection is a problem in this case. People have an enormous incentive to buy more health insurance as their health worsens, and the insurance companies cannot price people differently. Since the insurance companies price insurance the same for everyone, and people can buy insurance at any time, it's always worth it for sick people to buy insurance, but not worth it for healthy people to subsidize the sick people. If you're healthy, it makes sense to wait until you get sick.

Adverse selection definitely happens, but it happens most strongly in situations where various consumer protection laws prevent sellers from acting on publically available information but buyers obviously can. There are situations where regulated markets are worse than both freer markets or socialized ones.

Posted by: John Thacker at Dec 13, 2005 1:05:33 PM

AHIP is simply a trade lobbying group. I have yet to read a report of theirs that does not support their lobbying positions. That would be fine except that most of the reports that I have read clear through have logical errors or facts that directly contradict other, more disinterested (in a financial sense) research. I admit that I did not read this particular report clear through since it's conclusions seemed intuitively wrong to me. Had it been someone other than AHIP, I would have read it to challenge my preconcieved notion, but experience with previous AHIP literature made me unwilling to go any further.

To Alex and Will, I fail to see how the "correct" pricing of insurance which leaves many unable to afford it and at risk for worse health outcomes (at a high cost to society)as anything but a market failure. You might argue that my view essentially dictates a universal health scheme, that's correct. I don't understand how wanting to stay healthy is a "lifestyle" choice nor do I believe that this desire is associated at all with "unlimited free health". Even for those which the market works (I have had medical insurance 100% of my life) I have never seen a plan that entitles people to unlimited care.

Finally, the evidence of the individual market and the group market charging similar rates (assuming AHIP did good research here) when the group plans have important, costly, demonstrably valuable options means that the individual market (which draws from a healthier pool of people once you consider medical underwriting) is suffering distortion.

If you wanted to show a lack of market failure then you could easily do a study for COBRA insured. Here the problem of adverse selection is obvious, weel-known, and priced into the group policy premium. This means that healthy individuals who want health insurance should be able to easily find cheaper healthcare in the individual market compared to the COBRA rate offered. The studies I've seen show that people who forego COBRA choose no health insurance. The number of people who voluntarily give up COBRA in favor on an individual policy is negligible. If there was no market failure then it should be obvious that healthy individuals would fare better on their own.

Posted by: elliottg at Dec 13, 2005 1:11:14 PM

Alex is right in diagnosing a common fallacy: the problem of adverse selection (i.e. the efficiency loss) in health insurance is that low-risk individuals get priced out of the market, not that high risks have to pay too much. Similarly, the problem in used cars is that people with good used cars hold on to them, rather than sell them, because they can't get a good price.

However, it is not useful simply to point to insurance rates for individuals (vs. group rates) as evidence that there is no adverse selection problem. Far more telling is the fact that so many individuals (young and in relatively good health) don't buy it, at the supposedly bargain price of $2,268. I don't know much about their utility functions (and I suspect Alex doesn't either), so my inclination is to regard this pattern of market preference as significant. The more plausible conclusion is that the group rate is too high for many as well, but individuals who get health insurance through their employer either have no choice but to participate, don't see the true cost of their policies, can't opt out with full reimbursement (esp. regarding tax treatment of the benefit), etc.

With this in mind, consider the following argument: The fact that any solution to the problem of the uninsured seems to involve massive cross-subsidization (as one finds in group plans) is itself an argument in favor of social insurance. At least then the cross-subsidization has a principled and consistent basis, rather than being ad hoc and morally arbitrary.

Finally, to get back to used cars for a second -- there have been a number of major initiatives in the past few decades, by manufacturers, dealers, and in consumer protection law, aimed precisely at correcting adverse selection problems in this market. The practice of offering warranties on used cars, which many dealers now do, is an example of a relatively recent innovation, aimed at correcting these problems. Note also that the price of used cars has steadily risen in the past decade as well (at least where I live). So one must be cautious in saying adverse selection is "not the problem" in this market. It's not the problem anymore, but only because the market has been tweaked and regulated in various ways to fix the problem. It would become the problem again very quickly if these kludges were removed. Furthermore, because the problems involve information asymmetries, things have the potential to change very quickly. The most recent problem I had with my car involved a software glitch. It was solved by bringing it into the dealer for a firmware upgrade. My suspicion is that these sorts of problems would probably go undetected in a traditional car inspection.

Posted by: Joseph Heath at Dec 13, 2005 1:20:49 PM

The practice of offering warranties on used cars, which many dealers now do, is an example of a relatively recent innovation, aimed at correcting these problems.

Yes, and it definitely can work. Note, however, that if "consumer protection" laws force used car dealers to offer the same warranties on all used cars for the same price, adverse selection can still occur.

Posted by: John Thacker at Dec 13, 2005 1:23:45 PM

Isn't the fact that insurance companies (and credit institutions etc...) have put in place tools to reduce information asymetry a proof of the existence of the adverse selection?

These tools allow insurers or creditors the ability to better gage a person's risk, and thus better price their offer. Its a cost the insurers have to bear because there is not much one can do to convincingly signal one's true risk level. In addition, the incentives in this respect work against the insurer: the "inseree" will try to present the best profile possible.

Posted by: Philippe at Dec 13, 2005 3:32:23 PM

A comment to Ed above. Insurance in states with community rating (e.g., New Jersey and New York) costs more than states without it (e.g., California and Tennessee) because insurers are unable to price the risks they underwrite appropriately. NJ and NY also have guaranteed issue, which means they are forced to accept all applicants and can't exclude unacceptable risks.

As for your comment about redistribution of resources to sick people, with the exception of charity, "we" don't do this, the state does--by using coercion.

Posted by: Bill Stepp at Dec 13, 2005 3:59:53 PM

"I don't understand how wanting to stay healthy is a "lifestyle" choice..."

Because hardly anyone does even the basic things that are generally considered to be the components of a healthy lifestyle. Such things include regular exercise, maintaining a healthy weight, eating mostly healthy food, and avoiding constant stress. A study I read at

http://www.healthfinder.gov/news/newsstory.asp?docID=525339

estimated only 3% of the US population practices such healthy habits routinely.

Wanting to stay healthy and actually doing what it takes are not the same thing.

Posted by: Unknown at Dec 13, 2005 4:48:11 PM

Here's a data point, for what it is worth.

I am self-employed and have an HSA with a 2,500 deductible and a one million dollar lifetime cap on coverage. This costs me $85 a month or $1020 a year - well below the $2268 average cited. I'm healthy and 37 and I live in Iowa. If I were employed I'd want the same thing instead of the ridiculous $300-400 per month my past employers have negotiated (of which I paid $20-$80 of that) with a group plan provider that provided 100% coverage except for little co-pays. I'd rather have the extra $200-300 in my pocket for other things.

Posted by: Unknown at Dec 13, 2005 4:52:54 PM

I hope you stay well unknown, but that, to me, sounds like pretty lousy coverage at a relatively high price and that's without knowing your copays and drug coverage. Also, Iowa health costs are less than the national average so you benefit from that as well.

Posted by: elliottg at Dec 13, 2005 5:18:51 PM

It's major medical coverage elliotq - there are no copays. I pay for everything below the annual deductible. It makes perfect sense for me. $1020 a year (which is tax deductible) vs. $4800 a year that I and my prior employers paid for my co-pay only coverage. Which coverage is lousy?! I pay for all medicine out of pocket. Since I only consume allergy medicine a few times a year, I don't spend more than about $200 on pharms which I buy with tax-free money from my HSA.

Since it is an HSA ( so I can put $2500 a year into a tax sheltered account and use it for any medical costs. If I don't spend the money on medical costs, I can use it (but taxed) for anything else including letting it grow just like it was in an IRA. So I have a strong financial incentive to stay as healthy as I can.

Posted by: Unknown at Dec 13, 2005 5:29:26 PM

Remember, just because you gamble and win, doesn't mean it was a good bet.

Posted by: ellliottg at Dec 13, 2005 5:32:16 PM

What ARE you talking about?! Do you understand the benefits of the HSA? Is your standard assumption that everyone is going to incur catastropic medical costs year in and year out? Most people do not and that's what makes the super-care insurance that is standard at most companies overkill.

Posted by: Unknown at Dec 13, 2005 5:34:11 PM

Do you also understand that the $4800 a year (to continue my case) for coverage (individual coverage btw) that employers pay comes out of the employee's compensation? The marginal difference between the cost of an HSA and the cost of standard big company insurance is wasted money that could be put to more productive use elsewhere.

Posted by: Unknown at Dec 13, 2005 5:36:38 PM

The absolute first thing I did after being diagnosed with lymphatic cancer in 1996 was drive to my place of employment, go to the Human Resources department, and increase my employer-provided life insurance to the maximum allowed.

Posted by: Steve Sailer at Dec 13, 2005 8:15:00 PM

To me Alex's argument fails. The low figure cited for individual insurance is for a LIMITED TIME, usually one year. After which you policy is reevaluated and your premiums skyrocket. This doesn't happen in a large group policy where there is still always a certain percentage of healthy people to share the costs.

Posted by: David Y. at Dec 13, 2005 9:28:17 PM

I can't get at the empirical stuff (being outside the institutional firewall) -- the information it contains is too costly. Speaking of which, let me move away from the insurance industry aspect for a moment. For me, the key line in the original post is this:

You can buy a decent used car, for example just get it inspected or certified. Only if such adjustments are illegal, or in some other way not allowed, will adverse selection become important.

But the whole thing about asymmetric information is surely that information is not non-existent, but it is costly. No one thinks that the lemons scenario is more than a first-order approximation, and signalling and screening make second-order corrections, but the problem may still remain and be important. Alex Tabarrok is very close to saying that costs can be neglected as long as adjustments are "illegal, or in some other way not allowed". This claim that transaction costs are unimportant seems to be one of the dividing lines between market enthusiasts (AT) and market sceptics (me). I would say that asymmetric information becomes a fixed cost, leading to economies of scale in the insurance industry (which at least in Canada is an oligopoly with fewer members by the year).

I've posted a bit more at my blog (click name to get there).

Posted by: Tom Slee at Dec 13, 2005 9:39:33 PM

Just a quick question. What is the definition of 'no adverse selection' used here? Is it that the market leads to the complete information allocation and prices? How does the evidence cited support that?

I thought we could have an equilibrium in a model with advsere selection where everyone is insured, but the allocation is not optimal: a pooling (or even partially pooling) equilibrium. How does the evidence cited rule that out?

Posted by: anon at Dec 13, 2005 9:40:56 PM

"Or, more prosaically, the signal is near irrelevant when the signal to noise ratio is high."

What you intended to say is understood, but what you actually said is backwards. Signal-to-noise ratio is exactly that: the ratio, (signal)/(noise). Thus a *high* s/n ratio indicates that the signal is very strong relative to the noise, and a low ratio means that the signal is very weak relative to the noise.

Posted by: anony-mouse at Dec 13, 2005 10:46:05 PM

Here's my comment on your LIFE insurance statement. You say, in part, "not only do they buy life insurance they also buckle their safety belt and eat healthy? The price of life insurance falls the more you buy so evidently insurance companies believe it is the latter."

Since life insurance payout, unlike that for health insurance, is an all-or-nothing deal, aren't the life insurance risk managers looking mainly at the likelihood of your dying within a set period (during which they can invest your premiums), rather than just the dollar payout of your policy? If that's true, then isn't the declining price of additional insurance more a case of declining marginal cost (for the company) for each $X-thousand of coverage rather than a link between the total dollar value of your policy and your risk of dying during that period?

Posted by: Dottore at Dec 14, 2005 6:26:34 AM

How is affordability inequitable? It's the same for everyone.

Posted by: Robert Speirs at Dec 14, 2005 9:25:18 AM

anony-mouse is correct, I meant to say the signal is near irrelevant when the noise to signal ratio is high. For the record, I have corrected in the original post.

Posted by: Alex Tabarrok at Dec 14, 2005 10:29:11 AM

Market advocates like varied premiums, and tie them in their minds to "lifestyle choices." While I think such choices are a factor, I think many of the most expensive diseases are the outcome of a genetic/environmental crapshoot. Maybe you have a tendency toward cancer X when exposed to environment Y. Is it a "market sucess" if you are charged for that?

I gotta say, maybe it is if you value the market more than the guy with cancer. Personally, I think we're all in this together ... I'm not going to say "screw you, with your congenital heart valve malfunction ... pay for it you bastard!"

(if you don't like the language, imagine Cartman saying it. probably gets the point across better that way anyhow.)

Posted by: odograph at Dec 14, 2005 11:19:45 AM

So give him some of your money, odograph.

Posted by: joshg at Dec 14, 2005 11:50:31 AM

I actually did drop a bit on Doctors Without Borders, back in the bubble-day. You could find my real name in their annual report, if you are industrious.

Just curious, do you also give the medically needy some of your money, or are you a little more ... Darwinist in your attitude?

Posted by: odograph at Dec 14, 2005 12:10:05 PM

Alex,

I think one bit of evidence you cited against adverse selection is actually evidence that there is adverse selection:

"Only in states that require insurance companies to accept all or most buyers are rates high relative to the group market (rates in New Jersey, an outlier, are almost three times as high as the national average.)"

So in states that require insurance companies to sell to everyone at the same price, prices are higher. This would be true only if, in those markets, people buying individual health insurance were sicker than average. Unless you believe that states with those laws have populations three times as sick as the national average, this can come about only through healthy people dropping out of the market -- that is, only through adverse selection.

In addition, Devon M. Herrick at NCPA found that "[A]lmost one-third of the uninsured now live in households with annual incomes above $50,000 and one in five live in households earning more than $75,000 annually."

If individual health insurance is as affordable as you say, we can only conclude that these people believe they are healthy enough that health insurance is not worth the price. Assuming they are not extraordinarily risk-seeking (opposite of risk-averse), this means their lack of insurance is evidence of adverse selection.

The fact that there is no "adverse-selection death spiral" (outside of New Jersey and other states that essentially implement adverse selection by law) does not mean that there is no adverse selection -- it just means there is an equilibrium in which some healthy risk-averse (but not sufficiently so) people find health insurance to be too actuarially unfair for them to buy it.

Posted by: Robert A. Book at Dec 14, 2005 12:28:41 PM

Shouldn't a comparison of group and individual policies take into account taxes?

Since employer health coverage is treated much more favorably than individual policies can't we conclude that individual coverage, on an after-tax basis, is much more costly than employer-based coverage?

Posted by: Bernard Yomtov at Dec 14, 2005 12:46:51 PM

In my personal case Benard, the individual coverage - after factoring in taxes - is substantially less expensive. This is in part because the premium is so much lower (, in part because my premium is tax deductible (I'm self-employed), and in part because I put tax-free dollars into an HSA account.

Posted by: Unknown at Dec 14, 2005 1:07:06 PM

Odograph -

A study here:

http://www.healthpromotionjournal.com/Merchant2/merchant.mvc?Screen=PROD&Store_Code=AHJP&Product_Code=JV15I145

stated that:

The relationship between 11 modifiable risk factors (stress, current or former smoking, sedentary lifestyle, obesity, nutrition, depression, high blood pressure, high cholesterol, alcohol use, and blood glucose) and group level medical care costs were examined for 46,026 employees working in six companies. Overall, these risk factors accounted for approximately 25% of total medical care costs for these companies. Almost 8% of costs were attributed to high stress, and just over 8% to current plus former smoking."

I suspect that your belief that many of the most expensive medical conditions are due to an environmental/genetic crapshoot might need reconsidering. Human choice plays a significant role and I suspect studies that compare life expectancies and other health indicators between the US and other countries will show that lifestyle is a significant - if not the #1 - factor. I suspect that the US population is fairly risk-prone compared to other populations that live longer.

Posted by: Unknown at Dec 14, 2005 1:20:13 PM

Can you save me a little time? Doesn't 25% leave the other 75%? On the surface you are telling me that the "11 modifiable risk factors" cause a minority "of total medical care costs for these companies."

Posted by: odograph at Dec 14, 2005 1:24:12 PM

ohh, time and money .. $10 report.

Posted by: odograph at Dec 14, 2005 1:25:29 PM

Sorry to tripple post, but note that some of these 11 "modifiable" factors also have been shown to be at least in part "heritable." Twin studies.

Posted by: odograph at Dec 14, 2005 1:33:23 PM

I think your points are valid. I didn't read the actual study either, but it is suggestive that choice plays a role. There may be addditional factors that the study did not examine which also play a role.

I also read a study that claimed only 3% of the US population actually makes healthy lifestyle choices (eating plenty of fruit and veges, exercise, low stress, maintaining reasonable weight) on a regular basis. That makes me more suspicious that lifestyle is a significant factor in outcomes. I haven't been able to find anything that attempts to correlate lifestyle choices and health outcomes between nations, but I think it could be telling. How do we explain why some nations have significantly higher life expectancies than the US? Japan's LE is over 4 years more than the US. How much is genetics, how much is choice?

Posted by: Unknown at Dec 14, 2005 1:51:30 PM

Unknown,

First, I think you don't have the full picture and are full of arrogance about your good health. I start with the assumption that the insurance companies only underwrite when they can make a profit. Without risk aversion, there is no market for insurance (although there is still a reason to provide social insurance). Secondly, I'll observe that $400 per month for a 37 year-old male with no dependents in a group plan is outrageous. I'm not saying you're wrong, I'm saying the HR manager who is paying that in Iowa needs to be fired or investigated for receiving kickbacks. Finally, your personal situation has nothing to do with the policy debate we are engaging in. A 37 year old single male who is not so risk-averse that he is willing to tolerate a 1 million lifetime cap and limited renewal options if the problem spans several years is not the target of any reasonable policy person. I suspect that you are most excited about the tax advantages of your HSA and would consider, if legally allowed, to dispense with insurance entirely if you could retain the HSA.

Posted by: elliottg at Dec 14, 2005 1:59:35 PM

How much is genetics, how much is choice? ... just to be glib, it will take another 20 years of research to determine, and another 50 years after that for society to accept.

Posted by: odograph at Dec 14, 2005 2:04:01 PM

Elliottq-

First, I think you don't have the full picture

And I’ve never claimed to have it. My first post about my personal situation was prefaced with “for what it is worth” and I think the other comments were pretty well moderated, until you tossed in the non sequitir about betting.

and are full of arrogance about your good health.

Hardly. I’m grateful for it and I know genetics plays a role, but in a family with incidences of high blood pressure and cancer, I also know that my diligence in taking care of myself plays a role as well. I do exercise, eat well, and avoid stress to the best of my ability.

Secondly, I'll observe that $400 per month for a 37 year-old male with no dependents in a group plan is outrageous.

But common. These were standard in every company I’ve worked for since 1993 – and they were in CA, MN, and IA. The reason I think it occurs is because of competition for employees. Maybe it’s different in other industries, but in mine – software – these sorts of plans are completely normal and expected by most.

Finally, your personal situation has nothing to do with the policy debate we are engaging in.

My info was offered to contrast the line of thinking that individual insurance is as expensive or more than group insurance. Clearly that is not universally true. I am hardly the only case.

I suspect that you are most excited about the tax advantages of your HSA and would consider, if legally allowed, to dispense with insurance entirely if you could retain the HSA.

You mean have an IRA or 401k? You are wrong. I hardly want to be accountable for tens of thousands or more in medical bills if something catastrophic happened out of my control. If we didn’t have insurance when my wife became sick and needed 100k in care we would be in trouble. But that’s what insurance is supposed to be for – massive and uncommon medical costs. What I don’t want to do, on the other hand, is subsidize the bad habits of others to the extent possible.

Posted by: Unknown at Dec 14, 2005 2:26:49 PM

My sincere apologies unknown. I thought you were single.

Posted by: elliottg at Dec 14, 2005 2:47:49 PM

Odograph: "Maybe you have a tendency toward cancer X when exposed to environment Y. Is it a "market sucess" if you are charged for that?"

I won't say that it's a good thing, but yes, it is a market success. If you and the insurance company both know that you'll cost significantly more than average, your payments would also be significantly higher than average; otherwise it's just the other people subsidizing you. We as a society may decide that we want to subsidize the unusually sick, but I don't think that group insurance plans are the right way to do that.

Posted by: jadagul at Dec 14, 2005 3:21:15 PM

"How do we explain why some nations have significantly higher life expectancies than the US?"

IIRC, in many other developed countries infants who die within a short period after birth (24 or 48 hours, maybe more in some countries) are classified as stillbirths and do not enter into the life expectancy numbers. Therefore they don't pull down the overall life expectancy figures.
If you look at the years of life remaining for a person at a particular age, for example 25, I believe the United States doesn't look quite so bad. This is not a complete explanation of our lower life expectancy, but it's a start.

Posted by: Peter at Dec 14, 2005 4:07:08 PM

Haven't we already decided, and it's just that our method (mandates upon insurance companies) is less than perfect?

I'm in California where there are both regulations to encourage insurance companies to cover everyone, and a Major Risk Medical Insurance Program to pick up people refused by private carriers. We also, in an attempt to reduce the ability of insurance companies to play the genetics angle, and restrict their ability to genetic screening.

It is all in all, neither fish nor fowl, it is an attempt to hammer the private system into a socialized medicine work-alike. It becomes an adversarial system between insurance and oversight. I actually am a fair free marketeer, but think I might dislike such intensively "managed" markets more than I would hate a straight public service to replace them.

Posted by: odograph at Dec 14, 2005 4:28:26 PM

Okay, odograph, good point. We as a society have decided to subsidize the unusually sick; I still don't think group insurance plans are a good way to do that. Among other things, they lead to exactly the problem you describe: we've managed to craft a system with many of the flaws of a socialized universal single-payer healthcare system without reaping the benefits. I would incline toward scrapping some of the more socialized elements of our system and moving to a more market-like system, but even I get the impression that a well-constructed universal system could potentially be better than the disaster we have now.

Posted by: jadagul at Dec 14, 2005 8:23:27 PM

Wow, four grand sounds pretty good to me. I will pay more than $11,000 for health insurance next year for my family, and that's a small group rate HMO with a $25 copay. That rate is up 6.5% from this year. At least everything else is cheap in New York and the taxes are low.

I'd be just slightly more worried about adverse selection in a free market in health insurance if we actually had a free market in health insurance. For example, it would be great if that nice insurance company in Iowa could legally offer me a policy.

Am I just blinded by frustration or is it folly to be discussing an industry so intensely regulated by states, distorted by the historic third-party purchase by employers, and skewed by historically lower pre-tax costs for C corporations, as if it's some textbook example of market failure?

Posted by: david at Dec 15, 2005 4:24:58 AM

Elliotq writes: "To Alex and Will, I fail to see how the "correct" pricing of insurance which leaves many unable to afford it and at risk for worse health outcomes (at a high cost to society)as anything but a market failure. You might argue that my view essentially dictates a universal health scheme, that's correct."

I think the point being made in the original article has to do with a quite technical sense of the term "market failure". (BTW, I'm also a strong believer in universal health schemes, but that's not the central topic here.) Briefly, "market failure", when used as a technical term, means something quite different from failure of markets to provide a socially equitible outcome that would satisfy your or I.

Here's a little cartoon example of "market failure" (in the technical sense) in health insurance. Consider a population of 100 people, with 90 people healthy people, and 10 unhealthy people with hereditary cancer. Suppose the healthy people spend on average $500 per year (some less, and some a lot more, because some do get ill, but they don't know about that ahead of time). Suppose, they all want insurance, but very few want to pay more than $2000 per year for it, and the number willing to buy it drops as the cost increases. The unhealthy people have costs that average $10,000 per year, and they know about this in advance, so they all want to buy insurance.

In a correctly functioning market (in the technical sense) for full individual insurance based on individual risk, where both the insurers and the purchasers know all the risks, the healthy people would pay annual premiums of $500 + administrative costs. The unhealthy individuals would pay annual premiums of $10,000 + administrative costs. (Personally, I wouldn't call this socially equitable, but in a technical sense, it is a correctly functioning market that accurately prices individual risks).

The possibility of "market failure" arises when the insurers do not know who is healthy and who is not, and when individuals have the choice of whether they buy insurance. So, insurers must offer insurance at the same cost to everyone.

Suppose in year one they offer insurance at $1000 per person, and suppose every one buys it. (We can ignore administrative costs because they don't change the argument.) Suppose that, at the end of each year, insurers raise premiums to what would have allowed them to break even in the year that just ended.

So, in year 1 the insurers collect $100,000 in premiums. However, they pay out 10 x $10,000 + 90 x $500 = $145,000. Oops! Have to raise premiums, let's try to cover costs, so raise to $1450 per person.

In year 2, some healthy people decline the insurance because of the price increase. Suppose 70 healthy people buy it, and still all 10 unhealthy people buy it. So, the insurers collect 80 x $1,450 = $116,000 in premiums. They pay out 10 x $10,000 + 70 x $500 = $135,000. Oops again! Have to raise premiums again. To cover costs, raise to $1,687.50 (= $135,000/80).

Now in year 3, even more healthy people decline the insurance because of the price increase. Suppose now, 50 healthy people buy it, and still all 10 unhealthy people buy it. So, the insurers collect 60 x $1,687.50 = $101,250 in premiums. They pay out 10 x $10,000 + 50 x $500 = $125,000. Oops again! Have to raise premiums again. To cover costs, raise to $2,083.33 (= $125,000/60).

In year 4, yet again more healthy people decline the insurance because of third big price increase. Suppose now, 20 healthy people buy it, and still all 10 unhealthy people buy it. So, the insurers collect 30 x $2,083.33 = $62,500 in premiums. They pay out 10 x $10,000 + 20 x $500 = $110,000. Oops again! Have to raise premiums again, let's try to cover costs, so raise to $3,666.66 (= $110,000/30).

Now in year 5, no healthy people buy the insurance because it's just too expensive...

The point here is that this example is a death spiral -- the premiums increase until the only people who will buy insurance are people who know they are unhealthy. The people who are healthy, but who would like insurance because there is a small chance they will face large costs, find themselves unable to buy insurance at anything close to a reasonable price. This is the "market failure". This "market failure" has two necessary conditions: (1) "information assymetry" (purchasers know more than insurers) and (2) purchasers can choose whether or not to buy insurance.

Note that the various other issues that some people bring into the discussion, such as "lifestyle choices", are peripheral to the possibility of "market failure" in health insurance. I'm not saying that these other issues are not important, or that they're completely irrelevant to "market failure", just that we can discuss the existence of "market failure" without invoking other issues that sometimes bring more heat than light into the discussion.

Also, it is possible that "correctly priced" insurance (in the economic sense) is out of reach of many people (it might be very expensive because of known health risks). While I would argue that this is a socially unjust outcome, it is not a "market failure".

There are two simple ways to prevent this market failure: (1) remove the major elements of choice from the purchasers of insurance (universal insurance); (2) remove the information assymetries so that insurance companies know as much as purchasers of insurance (the US system). Under (2), people with known health problems will pay much higher premiums, and this is considered a correctly functioning market (while one can still debate whether it is desirable or not). The question Alex was addressing is how many information assymetries exist in the US market, and whether they are sufficient to induce noticable market failure.

Many people do claim that "market failure" is widespread in the health insurance market in the US. The existence of this type of market failure is a strong economic argument (quite distinct from a social justice argument) for "fixing" the system in one way or another. Alex Tabbarok, in his initial post was making the controversial claim that "market failure" is not nearly as widespread as some claim. This is a different argument to "universal insurance is more socially just" vs "private choice in insurance is the only system consistent with American values."

Posted by: Tony Plate at Dec 15, 2005 1:47:58 PM

That's a neat trick Tony. Set the initial conditions such that the insurance company can never be profitable. Maybe there are some other reasons health insurance is so expensive besides adverse selection:

1. No deductible. All expenses after a miniscule co-pay are covered. In essence your health plan is part insurance and part prepaid expenses, usually prepaid by someone else. Why not run to the doctor for every little sniffle? Talk about subsidizing other's lifestyles. HSA's may help eventually but, at least in my state, they are not yet a competitive product.

2. Zero competition. One way to maximize profits is to offer insurance to only the healthy; the other way is to spread the risk across many people as possible. To some that means a universal system. Why not try a national market in health insurance first? Why average costs over county-sized parcels (community rating) when there's a whole wide world of conusmers out there? Let insurance providers compete nationally.

3. Values skewed by tax policy and third-party purchasing. Buy a non-group plan and not only do you pay a premium 30% higher than even a group of two, you pay with after-tax money. No wonder it's corporations buying all the insurance. Employers get value out of each health insurance dollar that individual buyers don't see. Health insurance has value as alternative untaxed compensation to attract and retain employees. Get employers out of the business of providing and paying for insurance, and let everyone (or no one) pay with pre-tax money.

Posted by: david at Dec 15, 2005 3:21:12 PM

"Second, the asymmetry may run in favor of the sellers. Do I really know more about my own life expectancy than an insurance firm that has access to sophisticated actuarial models?"

Supported by December 18th's Dilbert - see Supported by December 18th's Dilbert (at http://www.dilbert.com/comics/dilbert/archive/dilbert-20051218.html).

Posted by: Tom Slee at Dec 18, 2005 8:54:50 AM

I was of the oppinion that the primary cause of "market failure" in the case of provision of medical care was simply non-price factors of demand. Wouldn't that be more important that the existance or non-existance of information asymetry? To put it another way:

Those who argue that its a distributional issue are correct, because even if the insurance providers did have perfect information in reguards to an individuals health, it would still be priced inefficiently in accordance to actual demand. (E.g., I want people I've never met to be vaccinated against communicable diseases even if they can't afford the cost of the shot.)

However information asymetry does exist, but the analysis that certification and independant verification reduces the effect of a "market of lemons" is also correct. The primary basis of the successful utilization of these information correcting processes is the reputation and objectivity of the 3rd party however. If the system of trust breaks down between buyer, seller, and certifier, then a market of lemons prevails. Likewise, if for some reason the data-set involved in accurately forecasting information is simply too variable, the reliability of data-correction through 3rd party certification is increadibly suspect. The health of any given person is a good example of that. And in fact the asymetry is likely to favor neither the buyer nor seller, as its likely to simply put the market well under equilibrium. A particular individual may have knowledge his insurance company does not have, but the insurance company is likely to experience significant losses if it didn't attempt to account for lack of particular knowledge. Ergo, you may also have knowledge of your own particular health which may be hard or impossible to prove to an insurer, that would qualify you for a much lower rate. However the insurer has to assume you may have knowledge that they don't which would put you in a higher category.

Thus over-all demand is diminished due to pricing having to be somewhere above what a person might be on a completely "objective" scale of health, while at the same time those who do fit a perfectly objectively healthy archetype are overcharged.

Posted by: UberIcarus at Dec 20, 2005 5:29:06 AM

Adverse selection in health care is primarily, although not entirely, a social equity issue. Is it "fair" for an otherwise healthy-living individual who just happens to get cancer to go bankrupt when it just as easily could have been someone else? That's a social preference issue with no positive answer.

The far greater efficiency problem in health care finance is moral hazard. People will overconsume a good, relative to the social optimum, when the marginal cost is zero. This covers both benefit-less consumption (seeing a doctor for a sniffle), as well as lifestyle tradeoffs (buying health through medical care instead of producing it through behavior).

I've long advocated a dual financing scheme: MSAs (Medical Savings Accounts) for routine care and catastophic coverage for major stochastic illnesses. The major hangup for effectively priced catastrophic care is identifying the line between behavioral and random factors in major illnesses. You want price differentials (the two pack a day smoker pays more than the non-smoker, ceterus paribus) to distribute costs based on behavioral contrasts, but maintain true risk pooling for factors beyond members control, including genetic predispositions. I've always maintained that, at some point, technological advances (primarily in information technology) will negate market failures (see peakloak pricing on freeways).

For the sake of brevity, I'll pass on the subject of outcome prioritization and rationing. See the Oregon Health Plan and the Prioritized List.

And to address a subtheme here, I'm using market failure to denote any market outcome that is suboptimal in the Social Welfare sense. Yes, I'm assuming a unique solution, or at least a finite set of unique solutions from which the benign dictator can randomly select.

Posted by: denadn at Jan 24, 2006 4:19:21 PM

Paris

Hover above Paris as you dine on a three-course dinner in the bustling first-floor restaurant of the Eiffel Tower. Then board a boat and cruise under the Seine's illuminated bridges.

http://geocities.com/bblessme1st2/classic_blue.html

Posted by: icci harambre at Apr 15, 2006 4:47:39 PM

lesbisch pissend ^^^ nympho kont ^^^ ebony gif ^^^ bionde ubriache nel bagno ^^^ jambes ciel ^^^ non sens poupee irve ^^^ sodomie quarante six ^^^ poupee trente six ^^^ desiderio asiatiche amore ^^^ stravagante amatoriali strip ^^^ oerfaren tonaring tillverkningen alska ^^^ amator alska i bostaden ^^^ infermiera anale fotti in casa ^^^ retiring lesbiche sesso ^^^ mye handjager ^^^ kjoligere tullete stump ^^^ sentimental kino ^^^ nydelig sjuskete ^^^ exeretikos mora ^^^ i omorfoteri efivos praxi ^^^ papa action ^^^ papa fetiche ^^^