Arnold Kling  

Real Health Insurance

PRINT
Becker and Posner vs. Medicare... Euro-America...

Extending the remarks I made earlier, in my latest essay I look at a recent study of medical-related bankruptcies and conclude that cost-reimbursement health insurance is ineffective and should be replaced by event-based insurance.


Event-based insurance means receiving a lump-sum payment at the time of diagnosis. Thus, even if the treatment goes on for years, the consumer has money to pay for it. The consumer does not have to worry about losing coverage with a pre-existing condition, because the consumer already has received benefits for that condition.

It would be reasonable to include a disability component to event-based insurance. If an ailment is going to cause a lapse in work, then the insurance payment can include a lump sum based on the typical amount of work time lost with that ailment.


UPDATE: For more on medical care issues, including economics-related issues, see Grand Rounds.

For Discussion. What problems would event-based insurance solve, and what problems would it create?


Comments and Sharing





TRACKBACKS (3 to date)
TrackBack URL: http://econlog.econlib.org/mt/mt-tb.cgi/200
The author at Different River in a related article titled Event-based Health Insurance writes:
    One of the problems with health insurance is that it isn't really insurance. Normally, the idea behind insurance is you trade a small probability of a large loss (say, a 0.0001% chance that your house burns down) for a certain (100% probability) smal... [Tracked on February 24, 2005 9:30 PM]
COMMENTS (7 to date)
Boonton writes:

Benefit(s):

1. A lump sum would motivate the person to seek the most economical treatment possible.

Problems:

1. It would create the incentive to misdiagnose illnesses, especially ones that lack clear objective criteria such as ADD.

2. The benefit could be subverted because I doubt society will really let people who lost their lump sum money on an unproductive treatment simply die if other treatments were promising.

3. Complications would limit its ability to become more than a niche player in insurance. I can see people looking for a lump sum policy against cancer and diabetes and also for an organ transplant. Beyond that do you really see people pricing insurance policies with varying payouts for the thousands of possible medical events that could happen?

B. Tental writes:

Let's say I get $1 million for a heart condition. I can go to India get the surgery at half cost then return to the US. By the way, Indians account for 20-30% of the physicians in the US (the percentage is much higher for cardiologists and anesthesiologists.) I can get the necessary surgery and get a life insurance policy of 500K at the same time, when can I sign up? This is currently happening in Europe for "non-essential" surgeries.

--BT

Randy writes:

I like the idea of events based medical insurance, but I doubt very much that it would be cheap enough to sell to the target audience, lower middle class working people. Truth is, there isn't a plan that would be cheap enough. Health insurance is simply not a priority in these people lives - until it is - at which point the solution is bankruptcy. When you live on the edge, you don't plan for the future. You live for today and hope for the best.

Lawrance George Lux writes:

It has all been said: a) misdiagnosis for fun and profit; and b) how to get Medical Underwriters to give away a quarter-million dollars at a shot. They would do the later item, if and only if, the premiums were set as 50% per year with a $40 Deductible. lgl

El Presidente writes:

Randy hit the nail on the head. The incentive for the poorest to "misunderestimate" (discount) the future dissuades them from playing in the first place. It also places them in a compromised bargaining position. I know from my own experience with lump sum settlements from insurance claims. Setting a number, a standard value for a type of injury, serves as a point from which insurers can chisel. When its you against them and its all or nothing who's gonna win the negotiation? Plus, how would you spend money to get a second opinion or basic care while your payment is delayed and reduced if you're poor and in dire need of help? Who sets the value of your injury? The only upside I can see is with work comp or disability where insurers currently over-reserve on claims to leverage the employer against an employee who wants to pursue a lengthy treatment, even under doctor's orders. They would no longer be able to over-value an injury that they consistently under-compensated. Audit trail.

Different River writes:

This is sort of a long response, and I've posted it on my blog as well. Feel free to post comments any comments that reference this to my blog as well.

In principle, this is a great idea. It would make health insurance more like regular insurance, and -- more important -- it would also rationalize the market for health care by giving patients incentives to consider the monetary costs, as well as the benefits (and side-effect costs) of various treatment options and various health care providers. (The reason for this is that if they get cured for less than the lump-sum payment, they could keep the change -- and if they couldn't get cured for that amount, they'd have to pay the difference or do without.)

However, one problem with event-based (better, "diagnosis-based") insurance is that it doesn't interact well with the technology of how medicine is actually practiced. In particular, event-based health insurance would make payments after a diagnosis is determined -- but a huge percentage of medical costs these days is in diagnostic tests and images. (MRIs and CAT scans are expensive!). For example, in 2002, 17.7% of Medicare spending was for diagnostic images and tests (see page 78 of the document in this PDF file), and that's not including the substantial share of physician visits devoted to diagnosis -- not to mention things like biopsies and exploratory surgeries (which, since they involve surgery, count as "procedures" not "tests"). I would not be surprised if it turned out that more than a third, maybe more than half, of total health care spending is for diagnosis rather than treatment.

Furthermore, medicine is often practiced in a "sequential" manner, in which treatments lead to diagnoses, as well as the other way around. For example, suppose a patient comes in with chronic breathing trouble. The doctor could immediately order a CAT scan of the lungs and an MRI, looking for everything from lymphoma to pulmonary fibrosis to lung cancer. All of these diagnoses are possible, but the highest probability event is that it's asthma. So, instead of ordering all these tests, the doctor writes a prescription for a couple of asthma drugs, and tells the patient to try them. If it works, the problem is solved and the diagnosis is simultaneously confirmed. The doctor may not even order the definitive test for asthma. If the asthma drugs don't work, then the doctor goes on to try other asthma drugs and/or test for asthma, then test for other diseases, most likely in order of their probability (unless some tests are very cheap, in which case they'll be done anyway).

Because some tests are very expensive, and some treatments are relatively cheap, this could actually be the optimal approach to treatment given both the costs of treatments and tests, and the probabilities of various diseases. If the probability (given known symptoms) of a disease is high, and the treatment for that disease is low in cost (both in terms of money and side-effects), and the tests for alternative, low-probability diseases are expensive, then it can be optimal to "treat and see if it works" rather than "test for everything before treating for anything."

However, event-based health insurance is dependent on the latter approach, since no payment is given without a definitive diagnosis. As a result, event-based health insurance could actually increase the cost of health care for some classes of symptoms, and could delay treatment for a majority of patients -- since tests take time and by definition most patients have the common diseases.

In other words, part of the reason health insurance is not "real" insurance is due to historical and regulatory issues, but part of the reason is due to the inherent nature of health care.

I should add, however, that the case for event-based (or diagnosis-based) insurance is not completely lost. It is possible that an optimal health insruance system could involve a diagnosis-based component for certain well-defined diseases and conditions. For example, I have actually received in the mail offers for something called "cancer insurance" that pays a set amount for any diagnosis of cancer. They sell it as a way to pay for deductibles, co-payments, and items not covered by ordinary health insurance, but in principle there's no reason why an insurance scheme couldn't have a complex system of different payments for different types of cancers, perhaps will additional payments if cancers proceed beyond specificed stages. However, not all diseases are as well-defined as all that, and you still need a way to cover the diagnostic tests.

Another problem I'd like to see dealt with is insurance for chronic conditions. For example, a diagnosis of Type I ("juvenile") diabetes is basically a guarantee of a lifetime of large health expenses. It would be nice to have insurance against this. There are two problems, however.

First, diabetes (and many other chronic conditions) is usually diagnosed in children and adolescents, but most of the payments are made when those same people are adults. So by the time you get to buy insurance, you already have the event that should trigger payment. Insurance would have to be purchased by parents for their children, but that's not likely to happen as long as parents have the reasonable expectation that their children will be able to obtain, as adults, health "insurance" that will pay these expenses.

Second, people diagnosed with chronic conditions by definition live a long time after they are diagnosed. And, medical technology is developing quickly and all indications are it will continue to do so. It's not all that difficult to calculate the cost of treating diabetes for a lifetime -- if you get to assume that technology will remain at today's level for that entire lifetime. But it won't. Improvements in medical technology can either increase or decrease costs (I have written on this in my "real life" and should post about it some time). A patient insuring against diabetes would want to have enough to pay for (say) an artificial pancreas in case one is invented in his/her lifetime, even if it's more expensive than a (remaining) lifetime of test strips and insulin shots. An insurance company has no good way to estimate that cost -- and furthermore, the insurance company would want to be "protected" against the chance that test strips and insulin injections become much cheaper over the patient's lifetime.

By the way, diabetes is just an example -- the same sort of concerns would hold true for numerous other chronic conditions, including genetic conditions.

Who is to say how much a particular condition or injury will cost to treat? What about complications? In an ideal world this may work but the variables are too great. Most people given a lump sum of money for a condition that needs to be treated for a number of years will spend the money foulishly within a few months. What do you do with the guy on the heart monitor whos wife drives up in the new Benz (she just wrecked) and and says she has no money?
Our President although I generally disagree with him, did a good thing in creating Health Savings Accounts (My HSA Blog). I beleive this is a better way of creating what it seems that you are looking for, a consumer driven healthcare industry. 'Do I need all those X-rays doc?' 'Do I need to go see a doctor for this bump on my knee?' Its alot easier to spend the insurance companies money. (They got so much, right?) 'Why not, it only costs me $10!' And the doctor has to make you feel like he's doing something for you so he gives you a prescription for a pill thats the same as two Advil.

Comments for this entry have been closed
Return to top