Rational Health Insurance
One of the most innovative thinkers in the field of health insurance is University of Chicago Professor John Cochrane. I’ve been meaning to write about him for some time and am spurred to do so now by a new policy report from the Cato Institute.
Here is the back story. After hail damages the roof of your house, your homeowner’s insurance is supposed to pay for the repairs. But there is no requirement that you continue paying premiums while your roof is being repaired. And if you switch to a new insurer, the new insurer doesn’t pay for damages incurred while the previous insurance was in force. These same principles apply to auto collision insurance and every other form of casualty insurance.
Yet health insurance is different. If you get sick, your insurer won’t keep paying medical bills unless you keep paying premiums. If you are unable to switch plans (because your new pre-existing condition causes you to be rejected or face exorbitant premiums), you are stuck in a continuing relationship with your existing insurer — regardless of the quality of service or the premiums charged. If you are able to switch plans, the new insurer has to start paying your medical bills, even though the illness (and all the premiums paid up to that point) occurred while you were on some other plan. This is why the new insurer doesn’t really want you and has no incentive to treat you well after you arrive.
To make matters worse, healthy people always have an incentive to leave a plan after some of its members get sick. The reason: the new plan formed by healthy people can charge much lower premiums. Meanwhile, premiums in the original plan (which now has only sick people) must rise to ever higher levels to keep paying the medical bills.
In a very real sense, health insurance isn’t insurance at all. It’s the artificial product of unwise tax and regulatory policies. But fret not. There is an ingenious solution to all this.
httpv://www.youtube.com/watch?v=okd3hLlvvLw
“You may say I’m a dreamer, but I’m not the only one.”
Instead of paying the premium for all-purpose insurance, Cochrane proposes two premiums for two different kinds of insurance. The first premium is for, say, a year’s worth of health insurance for a healthy person. The second premium covers the risk of changes in health status that potentially increase premiums in future years. Suppose that during year one you are diagnosed with a costly-to-treat medical condition. If you shop for a new health plan in year two, your premium will be higher than the rate healthy people pay, as a result. But your health status insurance (which you purchased in year one) will pay the extra premium cost.
Voila. We have in one simple step converted a completely dysfunctional market into a real market that solves real problems — for everyone.
Insurance for Pre-existing Conditions. Instead of trying to force insurers to ignore them, the Cochrane approach allows everyone to insure against them. In the Cochrane world, you don’t have to worry about financial consequences of developing a pre-existing condition. Your health status insurance will pay those costs.
Premiums for Pre-existing Conditions. In Cochrane’s world, premiums for pre-existing conditions would be determined in the marketplace. This means that the cost of a year’s worth of insurance for diabetics, asthmatics, cancer patients, heart patients, etc., would be transparent and competitively priced. By contrast, regulation in many insurance markets today tries to force insurers to ignore both the conditions and the cost of insuring them.
A Market for the Care of Pre-existing Conditions. Once there is transparency and competition in the market for insuring for pre-existing conditions, a natural extension is a competitive market for efficient, high-quality care for those conditions. Providers who find ways to lower the cost of care will allow insurers to be able to lower the cost of insuring that care.
The Cost of this Proposal. Some might suppose that adding health status insurance to routine health insurance would be very expensive. But since the only purpose of insurance is to pay medical bills and since the medical bills would be basically the same as under the current system, there would likely be little, if any, increase in total insurance premiums. In fact, to the degree that this proposal leads to more efficient chronic care, total premium payments may actually decrease.
Choice of Health Plans. Although, in general, I think long-term relationships with health plans are better than short-term ones, there is no reason in principle why people could not choose a different health plan every year under this proposal. The reason: Individuals (through premium payments made by them or on their behalf) would always pay an amount equal to the full expected cost they bring to any health plan they enter. No insurer would ever be forced to take an enrollee it did not want and no insured would ever be stuck in a plan he/she did not want to be in.
The Casualty Insurance Model. Cochrane’s proposal is a clever way of introducing the casualty insurance model into the world of health insurance. Alert readers will remember that my own proposal for ideal health insurance argues for incorporating other features of the casualty model as well. The proposal is also consistent with incentive-compatible health insurance, proposed by Brad Herring and Mark Pauly.
Ths is the most sensible thing I have seen on health insurance in a long time. Maybe ever.
John, do you know that John Lennon’s song is a paean to socialism, aethism, and one world ism. Other than that, its a perfect pairing.
Vicki, communist song, but capitalist solution to the health insurance market. That’s a reasonable trade.
Interesting idea, so do you envision two premiums for each person. Say currently the charge for insurance is 1200 per month for a family of four, what do you envision the cost for the two insurance plans would be for that same family? Would an insurance company be allowed to deny someone health status insurance due to family history? What about heath savings accounts? Those are a great way for people to save money and take responsibility for their health, yet they have been slow to catch on because most people can’t understand a world without copays – even if at the end of the day they are paying thousands in copays that are capped better with the HDHP plans. If they don’t get that, trust me they won’t get the health status insurance. All they want is insurance that pays for everything that they don’t have to pay for or think about. Trust me, I make my living talking to employees about their healthcare.
Partial response to trosen: the whole idea is to create a world in which insurers are free to price risk accurately. So of course insurers would be free to deny coverage to someone, or in the new workd envisioned here, they would more likely charge a higher premium for higher risk. But the reason for health status insurance is to pay the extra premium.
And by pricing risk accurately, I would also assuming all government mandated coverage would be out which would drive down the cost for insurance considerably. For example, in my state of IL, insurers are required to cover infertility treatments, including invitro. You can imagine what this does to the rates of a woman in childbearing years, or a group that has a large number of such women. Mandated hospital stays and unnecessary treatments that have to be covered no matter what, have outpriced insurance premiums. If insurance companies can accurately price, the basic insurance would be relatively cheap – however, I think the health status insurance could get costly. And again, Americans won’t understand it – they barely understand how HMO and PPO’s work. So that’s a huge roadblock, it’s the one that has made HDHP’s, which again transfers a large amount of the risk to the employee, along with HSA’s have been hard to catch on mostly because people won’t try to understand how it works or sit down and do a little leg work to see if it works for them – even though I present all the tools for them to do so. By and large they are lazy and have a huge sense of entitlement that is going to drive the politicians who can’t seem to do what is right, only what is popular, to pass stupid socialization or one payer or nationalized health care. And when that happens, I’m moving to an island!
There was a Cato forum a week ago where Dr Cochran spoke. Also on the agenda was Dr. Bradley Herring from Emory.
Dr. Herring has collaborated for many years with Dr. Mark Pauly of the Wharton School on “Incentive -compatible, Guaranteed Renewable Health Insurance”.
In the discussion at the forum it became clear to all concerned that Cochrane’s ideas are nearly completely equivalent to such ICGRHI. The major difference is that with Cochrane’s idea it is easy to switch insurance companies.
Incentive-compatible means that a policy can be written such that it is not appealing for a healthy person to leave the insurance company pool which first wrote the guaranteed-renewable policy in order to get a cheaper policy at a different company; hence leaving behind the sick individuals whose premiums hence rise.
The basic idea is that the premiums are higher than needed for the first year so that they may be lower than needed in the last year.
Herring and Pauly have investigated the insurance market in non-community-rated, non-guaranteed-issue states and shown that insurance companies actually are almost all spontaneously selling such policies to almost all customers.
A good place to start reading about these policies, very technical, however, would be “Incentive-compatible guaranteed rewable health insurance premiums”, Bradley Herring and Mark Pauly, Journal of Health Economics 25 (2006) 395-417. Available at http://www.sciencedirect.com.
These private policies are those which are at the core of my approach to universal health care which can be seen at http://www.plan.bipartisanhealhplan.com .
At the Cato meeting I asked if the policies could be extended from birth to death; the paper above considers only from age 15 to 65.
The discussion at the meeting suggested that it could, premiums from age 0 to 15 would build up credit for 65 to death; but detailed analysis would be necessary to be perfectly certain. In my approach I have the government give individuals $2000/year, vouchers for health care only, from birth to age 21; and parents are required to buy a policy for their children, possibly with that money.
Good ideas? Generally, yes. New and original to this Cochrane fellow? No way. I, and other health actuaries, have been proposing and debating the need to move health insurance from the current “services rendered” model to a “claim occurrence” model for decades. It is the difference between a health policy today that covers charges for medical services rendered/incurred during a defined calendar period and one that covers the resultant medical services that extend from the point of first occurrence of a condition/event, at least for some period of time into the future. The latter is more typical of the way a disability policy works – it provides payments for a given future period of time IF the status first begins during the policy’s coverage period, and only depends on the continuation of the conditional status (and the contractual period of time for the payments), NOT whether the underlying policy of coverage is renewed or not. Actually, if you go back to the 50s and 60s, you will find a number of medical expense policies that provided what was called “per cause” coverage as opposed to “per covered service” coverage. This is kind of along the same lines as what is being proposed. No pre ex conditions were covered because the cause/condition/event had to happen WHILE the coverage was in force, but then payment for services related to that cause/condition/event was made for 2, 5, or even 10 years. Sometimes there was no time limit, but a dollar limit. Where did these policies go? Well, of course, the lawyers and doctors teamed up to make sure that ANY service rendered once a person was “in benefit” got attributed to the particular condition or cause that triggered the benefit status, insurers had a difficult time pricing the products and couldn’t re-rate if blocks of business deteriorated, and it killed the things off. The only related remnant you find today are the “defined benefit” policies exemplified by Critical Illness products. Definitions are pretty important with a policy like this, and sorting out what care is unrelated, and what care is related, so you can sort out who/what policy is financially responsible for what service is where the proverbial rubber meets the road with this kind of approach. Having said that, you also have the model of workers compensation, which is a “per cause/event” type of coverage, and the expenses relating to it are, theoretically, the responsibility of the risk taking entity that had the coverage at the time of the event. As to the other issues, such a system is merely a different way to carve up actuarial risk, it doesn’t change the fundamental nature, nor ultimate size, of the risk picture being carved up. So, issues relating to guaranteed insurability, availability, and affordability will still exist, you don’t eliminate them by merely re-arranging the risk structure. To reiterate, though, if people would have listened to actual practitioners of health care financing (the actuaries) instead of academics and people working for think-tanks, the problems of health care reform could have been dealt with years ago. Of course, that would have required killing all the politicians first.
John Goodman is a major thought leader in health care. He provokes many to think more broadly and creatively about the health care delivery. I respect his vision and his understand of the issues. However, in this item he suggests developing two types of health insurance. One for the times that people are well and one for times when people are not. Perhaps I am missing a major element, but what causes people to buy both elements? It is like buying insurance for preventive care and then buying an additional policy for the rest of their care.
I am not so sure this is ingenious. It may only fragment the health care market further. Perhaps a thoughtful reader can set me straight.
Sorry, the URL left out a “t” :
http://www.plan.bipartisanhealthplan.com
A question for trosen (or others) to consider: could the health status insurer be the federal government? Is this the one piece of the health care financing puzzle that should be public mandate? There would be no need to “sell” the underinformed public about the need to purchase a new type of insurance. The premiums could be a flat rate for all people with health insurance. This still does not address the uninsured, I realize, but could it be a first step toward a rational healthcare financing approach?
The existence today, in the free market, of “Incentive-Compatible, Guaranteed Renewable Health Insurance”; as discussed by Herring and Pauly (2006)J Health Econ, 25, 3, 395-417 suggests that it would not be necessary for the government to offer health status insurance, since such insurance is substantially “embedded” in GR insurance.
On the other hand, it would be desirable for the government to strongly regulate such insurance to ensure that companies do not over-promise and leave customers without insurance after paying for GR insurance for many years.
John, someone else may have mentioned this, but would it not be just as effective to allow insurers to offer an “option” at the time of the purchase of the health insurance policy which “guaranteed” that person’s insurabiity for “x” number of years (possibly even committing to no more than “x” increase each year)?
John –
This is the same issue that has plagued product liability insurance for years – long tail liabilities and which insurer pays for them.
John:
Not having given this much thought, I see two problems.
First, doesn’t this just push the adverse selection problem logically upstream one step? In other words, won’t pre-existing condition insurers prefer people who are less likely to have them?
Second, competitive pricing is nice – I’m all for it – but it requires relatively thick markets. This means the framework Cochrane envisions might suffer from an inherent catch-22. To work well it must finely discriminate between various kinds of pre-existing conditions, but the more finely this discrimination is made, the thinner is the market and, hence, the less parametric in nature is the market price. As such, all the benefits of competitive pricing are weaker. Additionally, the more finely we discriminate between pre-existing conditions, the smaller will be insurance pools.
The first problem is related to the second. The more we move away from sharply tailored policies to avoid the problems I just mentioned, the more this just looks like standard insurance, with all the adverse selection problems that normally accompany it.
Again, and I can’t stress this enough, I have not given this much thought. I haven’t even read the original message carefully. So in your response please be generous, and kind.
Fantastic summary. Great stuff.
John:
Thanks for the info. I always read what you send and forward it onto others.
The question I have for you and others is that everyone always talks, as you did below, about car insurance, home insurance and life insurance. All of those forms of insurance are rated by the size of your car, driving records, size of your house, location, etc. and for life insurance all kinds of questions are asked such as do you smoke, high blood pressure, cholesterol, etc. Why isn’t health insurance rated like other forms of insurance? I strongly believe that if you have high cholesterol, high blood pressure, are of child bearing age, are overweight, have diabetes, etc. and there was a point system for determination of premiums, if people had to pay more there would be an incentive for those individuals to do something to improve their health. This wouldn’t solve all of our challenges, but we don’t give people any incentive to lead health lifestyles when there is no differentiation in premiums. Before computers, this would have been impossible to keep track of, but with computers, while it wouldn’t be easy at the start, it’s doable, and I firmly believe that incentivizing individuals would go a long way to having healthier people who used health insurance less.
I have trouble believing that most of the risk factors used in establishing risk would be those that individuals would have much direct control over. How many people do you know who have successfully lost a large amount of weight and kept it off permanently? And how many skinny people, who might be inclined to credit their condition on their healthy lifestyle, suddenly gain 30 pounds on reaching a certain age? You say hypertension should result in higher rates? If you take medication to control it, won’t being on the medication also be grounds for higher rates? Same with cholesterol.
If you want to punish unhealthy behavior, wouldn’t it make more sense to punish the behaviors directly, rather than years later based on a presentation that may or may not be caused by “unhealthy” behavior? And which behavior do you punish– salt consumption, for example? Half the time we don’t even understand the relationship between behavior and health outcome. I suspect that many of the factors used in underwriting are simply correlations between measurable data and average costs. We see the same kinds of correlations all the time medical journal articles. And then we see them contradicted six months later. And the whole time nobody has ever proved a causal relationship either way.
I’m sure that underwriting does a fair job of predicting costs given a large number of people with similar measurable characteristics, but I’m also sure that it fails a fair amount of the time. Some people with high cholesterol don’t get heart disease, and some who pay preferred rates fall over dead at 60. Aren’t these exceptions being billed unfairly or given a free ride?
I’m not saying that risk rating is completely useless, but be realistic. How much of it is worth the overhead?
Interesting, yes. Pragmatic? I’m not so sure, but that’s probably irrelevant. The elephant in the room is the likelihood of Congress passing legislation that would prohibit the pre-existing condition exclusion. If that little ditty passes as part of a comprehensive health reform agenda, it would render this idea obsolete. What would the private insurance industry do then? Exactly what they are doing now: Lobbying for an individual mandate.
This is an apples and oranges comparison. Hail damages the roof of your house at a single point in time. An automobile accident also represents a single point incident. A comparable health incident would be a time-limited event caused by, say, food poisoning. But the vast majority of expensive health care coverage applies to long-term issues – diabetes, hypertension, CAD, COPD, major depression, etc. One cannot expect insurance purchased for a specific time period to cover a life-long illness forever.
Car insurance and homeowners insurance, in fact, do not cover “illness” at all. If my 10 year old car starts to fall apart, requiring one expensive repair after another, State Farm isn’t going to cover that. Nor will they cover shingles falling off my house due to age. Imagine how high car and home insurance premiums would be if we tried to purchase products that did try to make our cars and homes last forever.
I don’t mind Goodman’s approach to rebuilding the health insurance marketplace through a new system, but it makes little sense to build the argument through an inappropriate comparison. And truth be told, I’d much rather we move to a system in which personal responsibility and preventive treatment take the lead.
[Reply to Stuart]
Interesting reply.
Either way, we need to insert the patient back in the financial equation of health care.
The concept of free care, or unlimited care for a $20 copay does not work.
A model of catastrophic coverage for things such as hospitalizations, medications, expensive tests, and certain illness based would be best.
Patients would pay for their care up to a certain income based amount, until they reach a cap where insurance kicks in.
Preventive care and personal responsibility could be used to lower the cap or lower the costs.
John,
Cochrane is correct. There are TWO possible types of insurable events in a given plan year. You insure against event risk (i.e. against a medical event that will be treated in the next year) and against reclassification risk (the event that you are re-classified into a higher risk class that will impact future years). As Kenneth Arrow pointed out decades ago (and everyone should know already) reclassification risk is a separately insurable event.
You may recall that my book co-authored with Jim Henderson which you reviewed, Health Care for Us All, explained this more than a year or two ago. I have attached a short selection that talks about insuring for both types of risk that may interest you.
Some disconnected thoughts:
1) I like David Rose’s second point about progressively finer discrimination in underwriting leading to very thin markets. Also, it seems to me that the resulting complexity would make it increasingly more difficult and expensive to make rational decisions. So you end up with insurers trying to balance the competitive advantage of accurate underwriting against the cost of that accuracy.
2) It’s tempting to think of risk assessment as simple science or mathematics, but health-related risk seems fundamentally different from, say, quantum uncertainty or even auto insurance. The latter involve random or pseudorandom events, while in health insurance “uncertainty” usually means ignorance, as in incomplete science. Another difference is that you can’t make the errors go away by averaging them together, since individuals who receive false low ratings don’t compensate for those with false high ratings (even though the errors cancel out from the corporate point of view). Again, I’m not saying there’s no role for underwriting, but it seems too messy to treat as a science. At minimum, underwriting is necessary for one insurer to stay competitive with others, and to police for deliberate free riders.
“Stuart” claims that Cochrane’s model is inappropriate because cars or homes are not the same as bodies. Too true: The best thing that can happen to you in a car accident is that the insurer buys you a new car. And that’s the worst thing that can happen from the insurer’s perspective. Obviously, for health care, the liability is more open-ended and less predictable.
But this is exactly why Cochrane’s proposal is so beneficial: A year is the time it takes the earth to revolve around the sun, but it has nothing to do with health status. Health-insurance polices are written for one year because of regulation and it suits employers, who manage most health insurance.
As Cochrane points out, the market is starting to offer products along the lines of his model: He points out United Health Group’s product. However, these products would have arisen much faster without the prejudice against families buying their own policies. I also suspect that HIPAA and COBRA have crowded out this innovation, too.
Health Care Plan for the Obama Administration
1. Government single payer.
2. Administration of plan to be contracted out by private insurers at a negotiated fee.
3. Universal coverage – hearing aids, glasses, alternative and holistic medicine, preventive medicine, prescription drugs, vitamins, supplements, etc.
4. Abolish FDA and set up a government run laboratory to test conventional and unconventional medical products and procedures for safety and efficiency.
5. Government to insure all doctors against malpractice. However, any doctor who losses his case in court more than once in any 10-year period would have his license revoked or not permitted to participate in the plan.
6. Break the monopoly of the medical profession by allowing competing methodologies and medicines to be available to the public.
7. Prohibit selling of body parts.
8. No patents on biological organisms.
9. All patents on new medical technology to remain in public domain with fair royalties paid by government for valid patents for a reasonable period of time.
10. Plan paid through Medicare tax increase by employer/employee with no prior limit on earnings plus a ½% health tax on all financial transactions.
11. Copayments required to control usage based on sliding scale keyed to gross adjusted income.
12. Vitamins and supplements approved by government health laboratory as to purity, safety and efficacy.
13. Set up a National Health Academy to test new approaches to health in all categories and train public service doctors for manning U.S. public hospitals.
14. Set up standards for healthy food, nutrition and exercise easily understood by the public. Clean up contaminated food supply by truth in labeling. Upgrade municipal water supplies
15. Provide a tax rebate incentive to all persons remaining healthy and using Medicare below a set minimum.
16. Fees paid to doctors to be negotiated with government every two years to account for inflation and latest technology.
17. Free medical scholarship to worthy, qualified students who agree to serve for five years in public health services and/or rural areas devoid of adequate medical personnel and facilities.
18. No public advertising of medicines or procedures on TV and radio.
19. Have states bid on pilot projects funded by the Federal government to come up with the most innovative health care delivery systems.
20. All hospitals and nursing homes have to be operated by private, not-for-profit entities on the same basis as public health hospitals.
Great piece in this month’s issue. Thanks for the continued leadership.
I am going to try to understand the John Cochrane piece your wrote, and I downloaded your chapter 24.
Under the banner “There is nothing new under the sun” comes Dattilo (and by extension, Racer) with a 90-10 or 80-20 idea. And the reason I send this to you is I bet in you billions of brain cells, a few recall similar ideas from the past.
Simple enough idea. I don’t think the numbers work. But the goal is get coverage for sick people. They need the insurance. Not healthy people (who only need insurance if they get sick – but you already know that).
Everybody pays the same premium. At the end of the year, each carrier tallies it up. Makes two piles. The cost of care for the 90%, and the cost of care for the 10% highest. The amount that exceeds the cost of providing care for the 90% is billed to a pool, and the insurance company gets reimbursed. The pool is paid for by a “sick tax” just as Minnesota does now (Health Care Access Fund).
Insurance remains affordable because everyone pays the same.
I am hedging on this, because Greg really likes the idea. But I worry that the cost of that top 10 is so high that it renders the idea impossible.
You may recall Don Larson writing a book circa 1994 on a 90-10 idea, in which everyone would pay up to 10% of their annual household income for health care after which the feds would pay the rest of the bill. He was a big fan of yours, by the way. Don died probably 10 years ago now.
[Response to Dave R.]
Basically, yours is an idea for government reinsurance and it is a completely separate idea. It could be combined with Cochran’s proposal or not combined. Personally, I don’t see any reason to have government involved here – unless there is a plague or a terrorist bombing that overwhelms the system.
John
Response to David Rose and Bart:
The underwriting market for sick people might be very thin if the only players are insurers. But why only consider insurers (and here I may well be departing from the thinking of Cochrane, Pauly, and Herring)? Why not let doctors and groups of doctors bid for the opportunity to take care of chronic patients? There might be one fixed price, say, for the next 12 months or a fixed plus variable (payments from an HSA) price. What I am envisioning is literally a market for the care of sick people in which all manner of entrepreneurs can profit by figuring out how to provide high quality care for a lower price.
Thanks to John Goodman and to all the many commenters.
As I’ve thought about it lately, there is another equivalent way to put the idea that may make it more attractive or understandable. Current individual (GR/IC if you want to be techical) insurance combines two things: medical insurance for one year, and the right to buy medical insurance in the future at a preset rate. My idea comes down to separating those two functions, and then having insurance companies occasionally post collateral on the second. As some commenters have noted there is no extra cost relative to individual insurance, we’re just separating these two things that come bundled in an individual insurance policy.
Separating the right to buy from the actual insurance is useful in a lot of ways. For example, someone who bought individual insurance and then got a job where an employer provides it could still keep the right to future insurance alive while he or she gets current insurance from an employer. Similarly, if you’re already in an employer plan, you could just buy the right part, knowing you’ll be covered later. People in current financial difficulty could by rights to more expensive coverage in the future, while suffering through a bare bones plan now.
The “health status account” can be thought of as collateral. The company selling you the right to future insurance has a long term debt. Why not make them post some collateral on that long term debt so that if they go out of business or otherwise stop performing, you have the resources to go elswhere.
I hope this helps, thanks for the comments
John Cochrane
[…] Standard insurance would cover the health needs of people during the insurance period, while health status insurance would pay future premium increases people face if they have a change in health status and then try […]
[…] Standard insurance would cover the health needs of people during the insurance period, while health status insurance would pay future premium increases people face if they have a change in health status and then try […]
[…] Standard insurance would cover the health needs of people during the insurance period, while health status insurance would pay future premium increases people face if they have a change in health status and then try […]
Thanks so much for your intelligent arcitle;this is the words that keeps me on track through these day. I have been looking around for your site after being referred to them from a colleague and was thrilled when I found it after searching for long time. Being a demanding blogger, I’m cheerful to see others taking initivative and contributing to the community. Just wanted to comment to show my appreciation for your post as it is very appleaing, and many writers do not get acknowledgment they deserve. I am sure I’ll drop by again and will send some of my friends.