« Stem Cells | Main | Mac: By Intel »

Clinging To Orthodoxy

I'm somewhat saddened to see the Century Foundation hopping on the self-insurance mandate (PDF) bandwagon normally associated with the New America Foundation. The basic idea is that the government will specify a mandatory minimum level of coverage that everyone needs to buy from some insurer or other. The government will then offer individuals subsidies, as necessary, to make sure they can afford the premiums. If you want to offer additional services beyond the minimum, you're welcome to do so, and if you, as an individual, want to buy such a plan you're welcome to pony up the additional premium money out of pocket. Companies will need to offer whatever plans they offer on the basis of a single price for whoever wants to buy in. The idea is to get universal coverage, large risk-pools, and other good stuff while still maintaining the benefits of competition.

I think this doesn't work. The basic problem is that in a normal market, if you offer a better product than your rivals, you get more customers, and therefore make more money. It's not at all clear that this would hold under the sort of schemes TCF and NAF are proposing. Some customers will have above-average health care costs, others will have below average costs. Adding customers with above-average costs will lose money for the insurance company.

If I were running a company under this scheme, here's what I would do. First, look at the basic stuff I need to offer. Then, put together a restricted network of physicians to offer the required services. I would put this network together by trying to deliberately weed out doctors who have good reputations. You don't want anyone so inept he's lost his license, but you're looking for bad doctors within the realm of reason. Ideally, you also look for people whose offices are inconveniently located, dirty, etc. You're meeting all the minimum standards, but you're doing it in a super bare-bones way. Next, you add-on some nifty additional benefits. Free gym memberships. Massage therapy with a copayment. Health care schwag, basically. Organize local sports leagues for plan members or something. And invest in a marketing campaign aimed at making your plan look cool.

What you've got, basically, is a plan that will be attractive to young, healthy people and unattracted to older people or folks who have reason to worry about their health. You're within the basic minimum mandates, but people who do things like spend time asking about doctor quality aren't going to join your plan, since you've restricted the network to crappy physicians. People who don't pay much attention to this sort of thing may just get suckered in by your marketing, may carelessly choose your plan, or -- best of all -- will be attracted by the free gym memberships and so forth. Now your plan's well-positioned to make a lot of money. You're getting national-average premiums for each of your customers, but paying out sub-average amounts in health care costs, since the vast majority of your enrollees will have above-average health.

But say a second plan opens up business in town with a similar model. Now you've got competition. Faced with competition, you need to improve your business model. So you need to sign up some less-crappy doctors, right? Wrong. Improving your core health care services will be bad for business, since the additional customers you'll get will be the less-healthy sorts. You want to try and make your core health services worse, driving your less-healthy customers into the arms of your rival, while beefing up spending on tangential goodies and marketing and so forth.

The basic problem is that for a market to produce the sort of good things we associate with market competition -- lower prices, better products, etc. -- you need to have yourself a real market. Real markets have downsides, of course, in terms of distributive consequences and so forth. Try to regulate those consequences away, however, and you don't just get to keep the good aspects of your former market. Once you create an oddly-structured market, with mandates and price controls and so forth, you're going to get a market that functions oddly. An insurer who's allowed to charge what he wants to whomever he wants has an incentive to make his product attractive. Each customer will be priced in a way designed to maximize expected profits, so you want to sign up as many customers as possible. An insurer who's forced to take on certain clients under terms where he's expected to lose money has incentives to do the reverse -- try to make his product unattractive to large swathes of the population in order to avoid getting stuck with the money-losing clients.

May 24, 2005 | Permalink

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8345160fd69e200d83423726653ef

Listed below are links to weblogs that reference Clinging To Orthodoxy:

» More Confusion from The Bit Bucket
I've been beating up on Matt, so let me state for the record that this is a dead-on analysis of the downsides of mandatory private insurance. He correctly argues that if you require everyone to have insurance at a fixed... [Read More]

Tracked on May 25, 2005 2:34:29 PM

Comments

You don't need no freakin market for the health insurance.

Sure, you could use market mechanism for various components, like data processing, but the core part of it doesn't need market any more than unemployment insurance.

Posted by: abb1 | May 24, 2005 1:31:31 PM

1> Total agreement here. The basic problem is that insurance companies function best their operations are restricted to commercial functions and rich people. Then you get sustainable insurance schemes not based on cost-cutting and stock owning. Or if it is, nobody cares.

Unfortunately we have tried to stretch the concept of 'insurance company' too far, to make the pre-existing 'for rich people only' medical scheme ['I need to make a payment on my Porsche and you have a deadly disease that will cost thousands to cure'] continue to function as always. The correct cure for this problem is cheaper, better doctors and cheaper, better treatment. But there is no incentive for such a system. I doubt very much fed financing is going to change that.

2> Real markets have downsides, of course, in terms of distributive consequences and so forth. Try to regulate those consequences away, however, and you don't just get to keep the good aspects of your former market. Once you create an oddly-structured market, with mandates and price controls and so forth, you're going to get a market that functions oddly.

Now! Apply this to the stock market!

ash
['What's wrong with this picture?']

Posted by: ash | May 24, 2005 1:32:19 PM

If you want subsidization, it should be direct, not indirect; otherwise, as you point out, you get a market with perverse incentives. I think that a self-insurance mandate would work well, but eliminate the “sell to everyone at the same price” feature; that transforms the proposal into one that is a very bad deal for the young and healthy. Rather, let insurers underwrite, and make long-term deals on either underwriting level or premium level (as is done today with life insurance—I can buy 5-year term, 20-year term, or whole life—the “terms” are the length of time until I can be underwritten again); then subsidize anyone for whom the cost of insurance is too high. That way, the incentives aren’t screwed up and the system isn’t biased against the young.

Of course, for this to be even remotely fair, the system of tax deductions for insurance would need to be fixed (so that insurance isn’t tax-subsidized for those with corporate-provided insurance but not for anyone else.)

Posted by: SamChevre | May 24, 2005 1:36:12 PM

There are mechanisms called "risk adjustments" designed to address this. The basic idea is that the insurers get paid for the amount of risk they accept. If they end up with lots of sick members, they get paid more per member. If they end up with lots of sick members, they get paid less per member.

While its not a perfect system and there is some opportunity for gaming, it is already in place in some markets, such as Medicare, and can work really quite well.

To me, what matters in this discussion is whether or not we get to full coverage. The mechanism used to get there is really unimportant because the camel's nose will finally be in the tent.

Posted by: storwino | May 24, 2005 1:40:27 PM

Yes, I think the cherry-picking problem you outline is unavoidable with a system of competing insurance plans. I'd argue a much better way to introduce competition would be a single-payer system with a single risk pool but with a relatively high deductible (a deductible which would vary according to income) so that people are used to paying for routine expenses out of pocket (or out of a tax-free HSA). The idea is to protect people from medically-caused financial disaster while, at the same time, providing some level of market constraint on consumption of medical services.

Posted by: Slocum | May 24, 2005 1:44:54 PM

This, combined with corruption and out-of-control drug spending, was a part of the downfall of the TennCare experiment.

Posted by: Ted A | May 24, 2005 1:47:15 PM

Re: Rather, let insurers underwrite, and make long-term deals on either underwriting level or premium level (as is done today with life insurance—I can buy 5-year term, 20-year term, or whole life—the “terms” are the length of time until I can be underwritten again); then subsidize anyone for whom the cost of insurance is too high. That way, the incentives aren’t screwed up and the system isn’t biased against the young.

No, the way to go is with strict community rating. So what if the young and healthy pay more? That would only be unfair if the young and healthy stay young forever. They don’t. They grow old and when they do their health care costs will climb too. In fact almost everyone (unless they are “lucky” enough to die young in a traumatic accident) will incur catastrophic healthcare expenses at some point in their life.
Ultimately, no matter how you try to slice and dice it any system of healthcare financing is sustained by the healthy paying for the sick. Since we all (almost all) of us get our “fair” turn as the sick sooner or later we all benefit.

Posted by: JonF | May 24, 2005 2:00:29 PM

Matt, you and Slocum miss the point (unless you addressed this in earlier posts). Matt, the reason that single payer plans save money is because there is no competition, and the payer (the government) has the power to limit the price spent on health care. Competition increases wasteful management/administrative costs by creating a need for actuarial/accounting services to be rendered to these plans. The resources spent on those services are resources not spent on health care services for paitents. Slocum, forcing people to spend out of pocket for preventative services (office visits, etc.) will increase the amount of people who need costly services. That eliminates another benefit of single payer systems; creating a false-cost detachment from preventative care for people, encouraging people to use said preventative care. In addition, both of you ignore the risk averse nature of parental health care consumption; people tend to overspend on their childrens' health care. This will be exploited (as it is today) in any market based system.

Posted by: blogsy mcblog | May 24, 2005 2:05:57 PM

SamCherve, there is a huge difference between health insurance and life insurance. With life insurance, actuaries can determine the expected risk of an individual to an extremely close degree of accuracy. The law of large numbers works extremely well with life insurance. With health insurance, the wildly variant cost outcomes creates a much more difficult actuarial projection. Thus, there will be much more risk in the health insurance market. Just as an aside, it seems like a lot of people are tacitly endorsing price rationing as opposed to queue rationing as a method to control health care expenditures.

Posted by: blogsy mcblog | May 24, 2005 2:16:01 PM

> the reason that single payer plans save money is
> because there is no competition,

As per my comment at Kevin's place, single payer is not the same thing as single provider or no competition. I believe that in France they have single payer, national standards, private practice, and free choice. Many if not all doctors are professionals in the old-fashioned sense of the word, and given the chance they will compete against themselves and among their peers to get better locations, clients, etc. And "better" doesn't necessarily mean healthier or more affluent; by all accounts Cook County Hospital is an awful place but I knew many medical people in Chicago who itched to spend a few years there to see what their limits were.

Cranky

Posted by: Cranky Observer | May 24, 2005 2:21:58 PM

"Competition increases wasteful management/administrative costs"

Um, blogsy, I don't think that's true. If it were, why would it be limited to health care? Why not save money and get rid of this crazy competition in other industries?

There are good reasons for universal health care, but "competition is wasteful" isn't one of them.

Posted by: Steve | May 24, 2005 2:27:41 PM

> If it were, why would it be limited to health care? Why
> not save money and get rid of this crazy competition in
> other industries?

Sort of the way Ken Lay and Enron improved the efficiency of the regulated electric utility industry? Did that industry need some reorganization in the 1980s, to improve efficiency? Yes. Was unbridled competition and the destruction of the geographically compact, regulated monopoly a good thing for the consumer? I would have to say the evidence is a quite strong, NO.

Cranky

Posted by: Cranky Observer | May 24, 2005 2:30:50 PM

Since Matt has previewed this several times as his "secret plan" to make millions of dollars when the New America scheme becomes law, I suspect that the fact that he has made it public now means that either (1) he got a big raise or a big book advance and doesn't need his scheme or (2) he no longer thinks this plan will ever become law.

Seriously, though, this is a danger with this type of scheme, and hardly the only one. I think the problem with this particular scheme is that if the information about doctor-quality was so clear that a scheming company would pick out just the mediocre ones, then you could also use that information to make the market work better.

Posted by: Mark Schmitt | May 24, 2005 2:42:03 PM

Nice points Matt. And by teasing out your example we see that if the government attempts to regulate against what the company is doing (instituting range limits on doctors, surprise inspections of facilities) increases the bureacracy and costs to the government of regulation. You finally get to the point when the government is spending so much money on regulation, why not just have a government run system anyway?

Posted by: Kathleen B | May 24, 2005 2:44:04 PM

There appears to be an unargued assumption running through this post that the people who are most sensitive to the quality of medical care provided are also the higher(est?) risk people. While I can imagine mechanisms that would lead to this result, I can also see the opposite being true, that the people most sensitive to the quality of health care the healthiest. Do you have actual data on this, or just a theoretical model?

Posted by: washerdreyer | May 24, 2005 2:44:30 PM

blogsy mcblog actually the big difference between health and life insurance is that life insurance companies are (mostly) allowed to charge rates which reflect actuarial risk. This means life insurance rates can vary by a factor of 10 or more for the same coverage. People seem reluctant to allow such variations for health insurance.

Also there is less overhead for life insurance as there are fewer claims and it is a lot more clear cut what is covered and what is not covered.

Posted by: James B. Shearer | May 24, 2005 3:19:41 PM

Ah, some raise the question: is it possible to reliable attract crappy doctors? Well, one can leave it to the market forces: offer to pay them less than the others do.

I should add that Matt's scheme has one weak point: malpractice liability. As we know, the fix is in the works.

Competition in non-health markets also may lead to somewhat counter-intuitive results. There is a big deal of competition between fast food chains. Does it produce more tasty and healthier offerings? Taste is in the eye of beholder, but health values are mediocre. Competition between operating system does not need to lead to the most reliable system cornering the market. Another example, financial services. One can profitably concentrate on offering very crappy advise to the gullible. Salespeople will be specifically selected for low financial competence and scant ethical principles.

Note that the market is very efficient in finding what consumers persistently fail to understand, like reliability in operating systems, nutritional values or the quality of financial advise.

Posted by: piotr | May 24, 2005 3:56:31 PM

There are good reasons for universal health care, but "competition is wasteful" isn't one of them.

Of *course* competition is wasteful, mostly to the extent that it reduces economies of scale. Nonetheless, competition is often a good thing, because a lack of competition can be even more wasteful -- no competition means no reason to become more efficient.

Whether or not competition is *more* wasteful than a lack of competition depends on the market. In the case of health care, the much lower bureaucratic overhead of Medicare, as compared to the private health insurance industry (this article claims 2 to 3 percent for Medicare, versus 10 percent for private insurers) means that competition is indeed wasteful.

Posted by: Alex R | May 24, 2005 4:12:30 PM

I should have specified; competition to select health care consumers for health care products based upon selection of health care purchasing intermediaries is wasteful. Competition between health care providers to provide services could be wasteful, but may not be. Competition per se is always wasteful to the extent that it encourages overproduction; it is also beneficial to the extent that it encourages innovation. The key is to balance the two (duh). At any rate, if our society decides to spend X dollars on units of "health care," and "health care" can be broken down into money spent making people better (N) and money spent to determine who should be given N (administrative costs, or A), any increase in A reduces N. Insurance plans that involve ultimate consumer choice increase A and reduce N. For those of us who want to see more N, that's wasteful. A single payer system can enourage copmetition between health care providers to produce the most N for the fewest resources; there isn't necessarily a problem with that, although there could be waste. Single payer systems obviously allow the payer to have an incredible amount of leverage to force lower prices out of providers; that's where many of the savings come from.

Posted by: blogsy mcblog | May 24, 2005 5:02:55 PM

There is a big deal of competition between fast food chains. Does it produce more tasty and healthier offerings? Taste is in the eye of beholder, but health values are mediocre.

Competition most likely does produce healthier menu items, to the extent that consumers of fast food want healthier fare. I've noticed my local golden arches (not that I would actually eat there, of course) now features detailed nutritional info about its offerings in a prominent place, just in case you'd like to know how many grams of animal fat and refined carbohydrates you're about to pour down your gullet.

The same establishment also features a number of healthier, lighter choices such as fruit salads.

Do people actually eat McDonald's healthy food? Beats me. Market competition might be able to force the purveyors of fast food to offer salads, but it sure as hell can't override the evolutionary imperative to gorge oneself with fatty animal flesh and starches to counter the drought and famine just around the corner.

Posted by: P. B. Almeida | May 24, 2005 7:46:25 PM

Re: With health insurance, the wildly variant cost outcomes creates a much more difficult actuarial projection.

Given a large enough group (say several million people) I suspect that you can estimate health outcomes as well to a fairly accurate extent, absent some unforseeable catastrophe-- a pandemic, a nuke attack, something like that, which would toss the life insurance estimates in the dumper too. We do after all have valid stats on how many people will likely get such-and-such a form of cancer, how many will suffer diabetes and so forth.

Posted by: JonF | May 24, 2005 9:05:01 PM

There's a suggestion in your piece that the reason the above approach wouldn't work is because there's some sort of external force warping the market. Only, that's not accurate--the same general "adverse selection death spiral" scenario that you described above would apply even if there were no insurance mandates or laws.

I guess if there was some sort of bizarre situation in which all sick people had more money than all healthy people (in proportions relative to the cost of treatment for their disease), then a non-regulated market might work. Actually, that's kind of the idea behind risk-adjusted payment systems (although no one can figure out how the hell to do it from a technical perspective.) But as long as wealth is severely skewed and healthcare is severely skewed and the skews don't overlap, we're going to have huge problems in the health insurance market.

Anyway, my point is you don't have to add any external regulation to warp the healthcare market. Healthcare markets come pre-warped, with suicidal tendencies.

Posted by: theorajones | May 24, 2005 9:44:06 PM

No, theoraJones, the death spiral phenomenon is a direct result of regulation. If you can underwrite effectively and price based on expected cost, you can avoid the death-spiral phenomenon. Where you get killed is when you can't underwrite, or can't price based on the population you are covering.

Admittedly, health insurance is harder (actuarially) than life insurance--but life insurance is the easiest product to price. Health insurance is no harder to price than automotive insurance--in both cases, it's predict the secular trend that's hard; that's why I suggested that the guarantee might be in terms of underwriting class instead of price.

In any case--my basic argument remains: if you want subsidies, they are less market-distorting and less likely to be resented if they are direct (by paying insurance premiums over x% of income), not indirect (by forcing healthy people into pools that are a very bad deal for them).

Posted by: SamChevre | May 24, 2005 9:54:37 PM

"put together a restricted network of physicians to offer the required services. I would put this network together by trying to deliberately weed out doctors who have good reputations. You don't want anyone so inept he's lost his license, but you're looking for bad doctors within the realm of reason. Ideally, you also look for people whose offices are inconveniently located, dirty, etc. You're meeting all the minimum standards, but you're doing it in a super bare-bones way."

Plans do cherry pick patients, but not this way. If these tactics worked, you would see plans use them today, and I don't.

Posted by: Bill Gardner | May 24, 2005 11:52:48 PM

The basic problem is that in a normal market, if you offer a better product than your rivals, you get more customers, and therefore make more money. ... [In the health care market], some customers will have above-average health care costs, others will have below average costs. Adding customers with above-average costs will lose money for the insurance company.

Nail. Head. Bam. In two sentences, you have identified why the free market does not work for health insurance.

Posted by: eyelessgame | Jun 8, 2005 4:15:28 PM

The comments to this entry are closed.