Philadelphia Reflections

An Unending Capacity to Generate New Problems

There is a fair amount of seemingly unrelated detail until we reach the point of this article, where we conclude it really is possible to design and pay for lifetime health costs with the tools we already have, using individually owned lifetime policies. As a part of that, it really should be possible to substitute cost shifting between the youth and the old age of one person, rather than the present kind of cost shifting from one person or group to somebody else. People don't mind taking from one of their pockets and putting into a different one. But fierce possessiveness appears when you shift from my pocket to your pocket, and the health system is riddled with it. "Riddled with cost-shifting" seems to imply underhandedness. In fact, only the simplest businesses could survive without such flexibility. The problem with cost-shifting in medical care is there is so much it, even carried to the extreme of performing carefree, with blithe indifference about how to pay for it. Just review how accustomed we have become to cost shifting as the only possible thing to do.

From the outset, Blue plans announced their business model: patients in the private rooms supporting the care of indigent patients on the wards, up to then entirely supported by the charity. Plus a third, intermediate class of say ministers and school teachers, called semi-private, who were financed on a strict break-even basis. Summary: rich people supported poor people, and the semi-privates broke even. At first, there was just a handful of semiprivate, but after a decade or so, just about everybody was semiprivate, defined as two strangers in a room. Blue Cross had an enormous unintended effect on hospital architecture. When Medicare and Medicaid adopted the same philosophy, the semiprivate room became the standard. If the rooms were small (and cramped) the nurses didn't have to walk so far, but the main driver was the insurance reimbursement formula, which was based on square feet of floor space. A square foot of such space was used as a cost basis for non-patient space in the overhead formula. Eventually, hospital architects were receiving demands for bizarre room sizes, in order to affect the reimbursement formula. The tail was beginning to wag the dog.

t During that era, charities were payers of last resort, unless creditors were stripped by bankruptcy. Furthermore, to provide a full range of services, some services lost money, subsidized by other departments which generated a profit. Any corporate executive could tell you what came next: the profit centers start to boss the losers around. In group practices, surgeons generally still subsidize primary care ("the feeders"); state Medicaid is roughly 50% subsidized by federal Medicare, and after hospitals are paid, underpayment by Medicaid is balanced by the hospital from other sources, once again mainly from Medicare until payment by diagnosis (DRG) came along. It is when one insurance competitor is forced by internal hospital cost-shifting to subsidize its rival, that most of the outcry is heard. Employer-basing leads to different subsidies between insurances, and by a two-step process, one competitive business subsidizes its fiercest competitor. Generally, a business does not care what things cost, so long as competitors must pay the same price. In the eyes of business, trouble comes from unequal cost-shifting. Its mere suspicion is almost as bad. Working-age people subsidize the generations too young or too old to work. That is obviously what must happen indirectly and unofficially, anyway. Cost-shifting is a normal business practice, an absolutely necessary one, but the cost shifting of hospital costs is almost beyond belief. Because now, no one can tell what anything costs, and because patients who are business employees will reflect the attitude that the absolute amount doesn't matter, only that competitors must pay the same. In short, cost rises meet little resistance.

What brings the matter to a crisis is payment by diagnosis, where it doesn't matter how long a patient stays or how many tests he has, the insurance payment to the hospital is the same. Added to a determination by Medicare to cut costs, the result is that the profit margin for inpatients is around 2%. From the payment designers' point of view, it's an excellent rationing system. But it isn't, because hospital architects are directed to shift their lavishness to service areas with greater profit margins, like emergency rooms and satellite outpatient clinics. The next time you see a building crane at your local hospital, just ask them what kind of building they are putting up. Having spent a fortune twisting hospitals into one kind of shape, the reimbursement system is twisting a new shape, and rather oblivious about it. At the same time, two-bedrooms are being converted to one-bedrooms to attract a carriage trade and justify a higher price. Maybe, just maybe, the bed capacity is somewhat smaller at the end of responding to a pitiful profit margin for inpatients. Changing demographics are also a factor. Trends toward unsustainable cross subsidies grow steadily larger because the contribution of working people is certain to get proportionately smaller. Extended longevity increases the proportion of young and old dependents, boosting the costs of working people by the fact that their shrinking proportion must ultimately pay for all of it. Ultimately, all hospital revenue originates with the working segment of the population. Parents pay for their children, and payroll deductions pay for the elderly grandparents. Working people are supporting it all. Let's not overstate: disappearing infectious diseases reduced the mortality and hospitalization of working people, too. The elimination of polio and tuberculosis was a dramatic godsend but made it harder to finance a general hospital, because of the shrinking client base of employed people. The way things are going, health costs should eventually concentrate in the first and last years of life, with hardly any serious illnesses for the people in the middle years of life who ultimately pay for every bit. Hospital cost-shifting can not indefinitely support its own system because working people will have so few medical expenses it becomes impossible to hide very much within them. If you want to know why payment by diagnosis was welcomed, just reread the last three paragraphs. Unfortunately, if payment is based on diagnosis, it doesn't matter how many x-rays you have, or whether all the door handles are polished brass. We badly need a new way to charge inpatients, and just about every system has been tried. Unfortunately, it took a long time to get rid of payment by the square foot, and it will take a long time to get rid of payment by two hundred very approximate diagnosis groups, or DRG. The very least that could be done is to substitute a better diagnosis code, like SNOMed, for the private ICDA, so that payments are seen to be driven by the right diagnosis, which might tell planners something useful.

How could we have created individual policies that failed to reward customer loyalty with guaranteed renewal? Or monopoly status, to companies without guaranteed issue?

Lost Opportunity

Under the growing circumstances, it might be possible to persuade most people it was a mistake to pay current costs out of current revenues ("pay as you go"). That is, we should have created individual policies, individually owned, that included a contractual renewal right in return for customer loyalty. And while we were at it, a guaranteed issue in return for monopoly status. As it is, whole demographic groups have come to believe that others have a moral duty to subsidize their costs. But these others know that a moral right is not a contract right, and both sides know a severe economic depression is apt to sweep aside merely moral claims. Taken all together, it is probably possible to persuade most Americans that it would be best to replace one-year "term" health insurance system with a "whole life" system, chosen and owned by each individual. People would pay in their estimated costs when young, accumulate compound interest income in the meantime and thus pay less than their costs when they are elderly. The overall lifetime cost would be less, by roughly the amount of the investment income. A dozen or so television serials and books might well convince the nation. What would be hard for people to accept, however, are the huge costs and convulsive disruptions of changing existing systems during the transition from whatever we have today, to whatever this imagined system would lead to. Asked offhand, the average American would probably guess this transition would take thirty years. Most people greatly underestimate the power of compound interest, and greatly overestimate the difficulty of doing something new.

Medicare would pay terminal costs as before, but be reimbursed by the escrow fund.

Transfer Vehicle

It would be difficult, all right, but not that bad. In the first place, we have an unusual set of circumstances involving the finances of the end of life. Predicting the average stock market price eighty years from now is considerably more precise than estimating what it will be eighty days from now. Furthermore, health expenses tend to be small for children and get progressively larger until we find that absolutely everybody will die and have some medical expenses, often very large ones. The list of diseases is steadily shortening and the life expectancy is getting longer. We have already discussed how relatively easy it would be to anticipate the heavy costs of the last year of life from Medicare statistics and the timing of it from life insurance data. That's not likely to change because no matter what other costs might be lowered, everyone is going to die. While it is true that current interest rates are unusually low, even they could be estimated, since the point is not to pay terminal care costs to the penny but to reduce them by whatever income might be generated. Forward projections would be comparatively straight-forward, so it seems likely the proof of concept would emerge in a few years. The vigilance of the public about government stewardship of the escrow, and of inflation control, are less certain but reasonably secure. Just assuring transparency would make watchdogs of the public in its own behalf, particularly if the rules prevent the individual from pilfering the fund for himself. As a practical matter, Medicare would pay medical costs as before, and be reimbursed by the escrow fund during the transition period. By the time the transition is complete, many other problems will be solved, and attention can be focused on this one. By comparison, generating the funds is the easy part. Keeping the voters from giving themselves a raise will always be with us.

Malpractice costs are disproportionately concentrated in Obstetrics.

Who is doing the suing?

But what about the first years of life , whose expenses have already been spent? (The term is loosely applied to pregnancy and post-partial.) The proposal here is to do it in stages. First, get the terminal care fund established and defended, showing benefits in its first year or two, as proof of the concept. Then, start collecting a contribution to the terminal care fund for the moral debt each citizen has for his early childhood costs, and do it for twenty years. Since both first-year and last-year care costs are being paid twice, funding can soon switch from double-paying terminal care to double-paying first-year of life costs, and eventually phase out double-payment. In time, the whole system becomes fully funded but it seems likely the dumb-bell shape of lifetime medical costs would lead to a strong temptation to spend the money on non-medical costs for spurious reasons. Therefore, the retention of the present system for employee benefits would have the additional benefit of creating a watchdog for straight-forward accounting practices. Meanwhile, keep chipping away at these maternity and childhood costs. The first chip is to recognize that malpractice costs are disproportionately concentrated in this group, so the fund would greatly benefit from tort reform. Vaccine costs are also strongly influenced by liability costs. The reader may have wondered why this article has said so little about litigation costs. It is to concentrate the focus of tort reform on this revised purpose for saving the money.

The second redirection of attention would be to campaign to lower the age at which American women have their first child, greatly reducing neonatal problems, including infertility measures and congenital malformations. Absorbing the cost of having a baby ought to assist this effort, otherwise highly desirable on purely medical grounds. Unfortunately, our system of graduate education and career advancement will incentivize timing conflicts with biologic goals. Society will have to work these conflicts out in its own way, but at least we can adjust health insurance timing to be more in keeping with societal trends.

Finally, it should be said that the Health Savings Accounts are a vastly simpler way of paying for health care than using the service benefits approach, and the payment system greatly needs simplification. Using a high deductible has the potential to preserve market benchmarks for prices which are otherwise going to induce unworkable price controls, permanently. The system of "first dollar coverage" was accelerated by a wish to include as much as possible under the Henry Kaiser income tax evasion, and it will return if we neglect to correct that flaw. Experience with Health Savings Accounts has demonstrated as much as a 30% reduction in claims costs. Linking market-set outpatient costs to the same services when provided to inpatients should be an adequate price control for helpless sick people since an improved system of diagnosis-related groups should accomplish most of it. But the main advantage is to reduce these fund transfers to money without health attachments, to make unification and substitution more plausible. That is, to eliminate "service benefits" and not replace them with "diagnosis benefits" except for helpless bed patients. A return to dollar indemnity is greatly needed, although perhaps not totally.

To a considerable degree, service benefits are in conflict with indemnity benefits, in a manner resembling the conflict between debt and equity in the financial sphere. At some point, there must be a reconciliation between these two ways of paying for things, especially by keeping indemnity consistent with market prices. The best one can hope for is to shift the location of the interface between service benefits and indemnity, bringing the friction out into public view, and equalizing the power of the sponsors. Therefore, the best place to hold the debate is to treat diagnosis groups as inpatient service benefits, and outpatient costs as indemnity. With reasonable exceptions, of course. One of the main mistakes of the DRG system was to extend it to every inpatient. Inpatient psychiatry should be paid for as if it were an outpatient service, and chronic diseases such as Alzheimer's disease should also be excluded from DRG as well. Emergency room visits should be separated into two groups as well (admitted to hospital and discharged home), with reimbursement slanted to reduce the incentives for unnecessary use of the Emergency room, not the other way around as it is at present. The whole trick here is to see the double reimbursement situation as an opportunity for constantly rebalancing the two approaches, rather than allowing it to be pounced upon as a loophole.

We started by saying these issues should be chipped away, during the period when more important issues are being addressed head-on. But the list of small issues is a long one, providing ample opportunity for trade-offs in ambiguous opportunities. More than anything else, the endless capacity to develop new problems demonstrates the need for careful construction of an institution to serve as an informed and trustworthy umpire.

Originally published: Tuesday, June 11, 2013; most-recently modified: Sunday, July 21, 2019