Philadelphia Reflections

The musings of a physician who has served the community for over six decades

Related Topics

Healthcare Reform: Looking Ahead (2)
The way to make certain you have enough -- is to have too much.

How Do You Withdraw Money From Lifetime Health Insurance?

Four ways should be mentioned: Debit cards for outpatient care, Diagnosis pre-payment for hospital care, Transfers from escrow, and Gifts for specified purposes. The comments which follow apply to regular, old, single-year HSAs. The multi-year variety has more similarity to insurance than to retail banking and probably would favor the "cash balance" approach used to withdraw money from whole-life insurance. In the long run, that would probably lead to lower costs, but actual retail experience does produce a different culture.

Special Debit Cards, from the Health Savings Account, for Outpatient care and Insurance Deductibles. Bank debit cards are cheaper than Credit cards, because unpaid credit card payments are a loan, whereas the money is already in the bank for a debit card. It could be argued credit cards are a little safer than debit cards, because "possession is nine points of the law". Sometimes pressure has to be applied to banks or they won't accept debit cards with small balances. Somehow, the banks must be made to see that you start with a small account and only later build up to a big one. So it's probably fair, for them to insist on some proof you will remain with them. The easiest way to handle this issue is to make the first deposit of $3300, the maximum you are allowed to deposit in one year. Even better would be a family account with a $6000 deductible, which probably gets to the $10,000 threshold in less than two years. That's difficult for little children and poor people, however, so some way ought to be devised to have family accounts for children. At the moment, you just have to shop around, that's all. Unfortunately, the tendency of banks to merge into bigger entities headquartered in another city leads to powerlessness at the local level.

After negotiating that hurdle, you should pay your medical outpatient bills with the debit card, although we advise paying out of some other account when you can, so the balance can more quickly build up to a level where the bank allows more latitude. Remember this: the only practical difference between a Health Savings Account and an ordinary IRA, is that medical expenses are tax-exempted when paid with money proven to come from an HSA. Both debit and credit cards are tax-sheltered for deposits, and both (in normal economic times) internally generate income, un-taxed. If you can scrape together $6000, you are completely covered from Obamacare deductibles, and since co-payment plans are to be avoided, an HSA with Catastrophic Bronze plan is your present best bet. If you have a bronze plan, you probably get some money back if you file a claim form, but those rules are still in flux at this writing. The expense of filing and collecting claims forms is one of the reasons the Bronze plan is more expensive, but that's their rule. The bronze plan is thus easier to get, but harder and more expensive to use, and carries a political risk of changing rules with political motives. Another curiosity is that big banks tend to be more customer-friendly than small ones, although that may well be temporary. The tendency of traditional HSAs would be to act like banks: checking accounts with reinsurance in the background for emergencies. The multi-year approach would probably behave like insurance with occasional withdrawal privileges, very likely treating cash withdrawals as a nuisance which increases costs. Their experience is with "cash balances" which are somewhat smaller than true balances, and a preference for big-ticket hospital payments.

There are some other important things to say about outpatient vs. inpatient care, but first, it seems best to describe how inpatient care is envisioned to work in this system, before returning to the tension between one-year and multi-year approaches. Increasing ease of use might create the problem of making it a little too easy to spend money foolishly.

Payment by Diagnosis Bundles, for Inpatient care. In 1983 a law was included as a largely unnoticed section of the annual Budget Reconciliation Act, which nevertheless later proved to have a huge effect on the hospital financing arrangement. The proposal was to stop paying for Medicare inpatients on the basis of a bill for itemized services, but rather to pay a lump sum based on each patient's elaborated diagnosis. The argument was accepted that most cases of a given diagnosis were pretty much the same, so small variations soon average out. Such a casual approach to the complexity was justified by arguing any patient sick enough to be in a hospital bed, was too overwhelmed by his frightening situation to dispute what was done to him. Market mechanisms, in short, were futile is situations with such imbalances of power. Consequently, why waste money on accounting systems to arrive at prices which were actually arbitrary.

This overly simple argument prevailed in a Congress desperate about relentless cost increases. Misgivings that the hospital accounting system was a large part of its administrative information system, were brushed aside. To the extent such objections were valid, they could be addressed later. In retrospect, it can be seen the administrative and medical parts of a hospital act largely independently of each other, communicating through prices as a sort of abbreviated language. The administrative mission of bottom-line efficiency thus became even more insulated from those who saw patient satisfaction as far more important. In fact, the unresisted expedient emerged, for prices of the DRG ( diagnosis "related" groupings) to migrate toward a 2% profit margin on the bottom line, no matter how delicate the medical issues happened to be. You might suppose anyone could see a 2% profit margin was unsustainable during a 2% inflation, but normal hospital behavior is to seek uncomplaining workarounds.

The hospitals might have rebelled, or might have collapsed. Instead, they just accepted 2% for inpatients as additional administrative nonsense and set about adjusting the cost-accounting to aim for 15% profit margin on the Emergency Room, and 30% profit on outpatient services. Cost shifting of established cost accounting was difficult to achieve at first, so Emergency rooms were enlarged, and much-expanded outpatient facilities were built, requiring hospitals to purchase physician practices to keep them filled. The entire healthcare system was put under strain, and hardball became the game of the day. New lifesaving drugs were priced at $1000 per pill, less expensive institutions were merged out of existence, the office practice of medicine was in turmoil, and a year in business school could make someone a millionaire if he could appear calm in the midst of such confusion.

I tell this story to explain why, with great reluctance, I advise the management of Health Savings Accounts to base their inpatient payment system on some variation of Diagnosis Related Groups. It's a terrible system, designed for other purposes and adopted for hospital billing by Congressmen. It does protect the paying agency from being fleeced, once it gets past negotiated rebalancing of a reduced list of prices, aggregating toward a politically dictated bottom line. It chases everyone else out of attempting to understand it, with the consequence that a handful of people have brought hospitals dangerously close to quick destruction by a sudden change in the rules. Whatever it may call itself, it is a rationing system. And rationing invariably leads to shortages.

Resolving Tension Between The Two Payment Systems. Evidently, some shrewd thinking by some smart people have brought them to the ruthless conclusion that a two-class system of medical care is preferable to the way we were otherwise going. Rich people will have their way if their own health is at stake, and poor people will have their way if they exercise their votes. Both of these conclusions were correct, but they lead to Medieval monks retreating into monasteries. The cure for cancer and a few brain diseases might make monasteries unnecessary, and so would a drastic reduction in health care costs. Huge research budgets and major regimentation are big-government approaches, of willingness to accept some loss of freedom to achieve equality of outcome.

But we can't completely depend on either choice, so the remaining choice is to undermine a lot of recent culture change, by devolving back to leadership on the local level of small states and big cities. This is a small-government approach, willing to accept wider inequalities in order to find the freedom to act. Mostly using the licensing power, the competition will reappear if retirement villages and nursing homes are licensed to be hospitals. If not, nurses and pharmacists can be licensed as doctors. Some of this could become pretty brutal, and all of it leads to patchy results. But of its ability to restrain prices temporarily, there can be little doubt.

Escrow Subaccounts within HSA Accounts. Whether anything can restrain reckless spending of "found" money, is quite a different matter, however. It may be that supply and demand will balance, even if it takes generations. There is some satisfaction to be gained from watching reckless teenagers become penny-pinching millennials, but dismal reminders of improvidence will also be found in ninety-year-old millionaires marrying teen-aged blondes, further reinforced by watching the blondes run off with stable-boys. The net conclusion is that if certain portions of a Health Savings Account must be set aside for mandatory later expenses, then the money should be set aside within partitions, like an escrow account. Even that will have limits to its effectiveness, as I have noticed when trust-fund babies in my practice worked around the restraints their grandfather's lawyer took care to put in place.

Specified Gifts to be Encouraged. Only limited restraints on spending the client's own money can ever be justified, but certain types of gifts can still be better justified than others. One of them would be the special $6000 escrow fund for deductibles and caps on out-of-pocket spending. Particularly in the early transitional years, the fund's solvency may be threatened by leads and lags, where these escrow funds could save the day. Therefore, if someone accumulates large surpluses in his account by the fortuitous conjunction of events, he should be encouraged to consider donating a $6000 escrow to one of his grandchildren or other impecunious relatives. Quite often, a prudent gift to a grandchild can lighten the burdens of his parents or other members of the family. If they wish, any number of $6000 transfers to the escrow funds of others should be encouraged.

Spending Health Savings Accounts. Spending Less. In earlier sections of this book, we have proposed everyone have an HSA, whether existing health insurance is continued or not. It's a way to have tax-exempt savings, and a particularly good vehicle for extending the Henry Kaiser tax exemption to everyone, -- if only Congress would permit spending for health insurance premiums out of the Accounts. To spend money out of an account we advise a cleaned-up DRG payment for hospital inpatients, and a simple plastic debit card for everything else. Credit cards cost twice as much like debit cards, and only banks can issue credit cards. Actual experience has shown that HSA cost 30% less than payment through conventional health insurance, primarily because they do not include "service benefits" and restore the patient to a position of negotiating individual item prices, or be fleeced if he doesn't. Not everybody enjoys haggling over prices, but 30% is just too much of a penalty to ignore.

No Medicare, no Medicare Premiums. We assume no one wants to pay medical expenses twice, and will, therefore, want to drop Medicare if investment income is captured in lifetime Health Savings Accounts. Such a change of attitude might take twenty or more years, however. The major sources of revenue for Medicare at the present time fall into three categories: half are drawn from general tax revenues, a quarter come from a 6% payroll deduction among working-age people, and another quarter are premiums from retirees on Medicare. All three payments should disappear in time, but the 50% subsidy may actually block it. Therefore, the benefit available for dropping Medicare would differ in type and amount, related to the age of the individual. Eliminating the payroll deduction for a working-age person would still find him paying income taxes in part for the costs of the poor, as it would for retirees with sufficient income.

Retirees might pay no further Medicare premiums. Illnesses of the elderly make up 85% of Medicare cost, but at present only contribute a quarter of Medicare revenue. They first contribute payroll taxes without receiving benefits, and then later in life pay premiums while they get benefits, to a total contribution of 50% toward their own costs. But the prosperous ones still contribute to the sick poor through graduated income taxes. There might be some quirks of unfairness in this approach, but its rough outline can be seen from the size of aggregate contributions. At any one time during a transition, working-age and retirees would both benefit from about the same reduction of money, but the original working age people would eventually skip payments for twice as long. Invisibly, the government subsidy of 50% of Medicare costs would also disappear as beneficiaries dropped out, so the government gets its share of a windfall, in proportion to its former contributions to it. One would hope they would pay down the foreign debt with the windfall, but it is their choice. This whole system -- of one quarter, one quarter, and a half -- roughly approximates the present sources of Medicare funding and can be adjusted if inequity is discovered. For example, people over 85 might well cost more than they contribute. For the Medicare recipients as a group, however, it seems like an equitable exchange. This brings up the subject of intra- and extra-group borrowing.

Escrow and Non-escrow. When the books balance for a whole age group, the managers of a common fund shift things around without difficulty. However, the HSA concept is that each account is individually owned, so either a part of it is shifted to a common fund, or else frozen in the individual account (escrowed) until needed. It is unnecessary to go into detail about the various alternatives available, except to say that some funds must be escrowed for long-term use and other funds are available in the current year. Quite often it will be found that cash is flowing in for deposits, sufficient to take care of most of this need for shifting, but without experience in the funds' flow, it would be wise to have a contingency fund. For example, the over-85 group will need to keep most of its funds liquid for current expenses, while the group 65-75 might need to keep a larger amount frozen in their accounts for the use of the over-85s. In the early transition days, this sort of thing might be frequent.

The Poor. Since Obamacare, Medicaid and every other proposal for the poor involves subsidy, so does this one. But the investment account increasingly pays a larger share, so the cost of the subsidy is considerably reduced. HSA seemingly makes it somewhat cheaper to pay for the poor.

Why Should I Do It? Because it will save large amounts of money for both individuals and the government, without affecting or rationing health care at all. To the retiree in particular, who gets the same care but stops paying premiums for it. In a sense, gradual adoption of this idea actually welcomes initial reluctance by many people hanging back, to see how the first-adopters make out. Medicare is well-run, and therefore most people do not realize how much it is subsidized; even so, everyone likes a dollar for fifty cents, so there will be overt public resistance. When this confusion is overcome, there will still be the suspicion that government will somehow absorb most of the profit, so the government must be careful of its image, particularly at first. Much depends on allowing individuals to drop Medicare if they wish, rather than eliminating the choice, or even poisoning it with benefits reduction. Medicare now serves two distinct functions: to pay the bills and to protect the consumer from overcharging by providers. Providers must also exercise prudent restraint. To address this question is not entirely hypothetical, in view of the merciless application of hospital cost-shifting between inpatients and outpatients, occasioned in turn by DRG underpayment by diagnosis, for inpatients. A citizens watchdog commission is also prudent. The owners of Health Savings Accounts might be given a certain amount of power to elect representatives and negotiate as a group what seem to be excessive charges.

We answer this particular problem in somewhat more detail by proposing a complete substitution of the ICDA coding system by SNODO coding, within greatly revised Diagnosis Related Groupings,(if that is understandable, so far) followed by linkage of the helpless inpatient's diagnosis code, to the same or similar ones for market-exposed outpatients. (Whew!) All of which is to say that DRG has been a very effective rationing tool, but it must not persist unless it becomes generally proportional to market prices. We have had entirely enough talk of ten-dollar aspirin tablets and $900 toilet seats; we need to understand how such prices are arrived at. In the long run, however, medical providers are highly influenced by peer pressure, so again, mechanisms to achieve price transparency are what to insist on. These ideas are expanded in other sections of the book. An underlying theme is those market mechanisms will work best if something like the Professional Standards Review Organization (PSRO) is revived by self-interest among providers. Self-governance by peers should be both its theme and its reality, ultimately enforced by fear of a revival of recent government adventures into price control. Those who resist joining must be free to take their chances on prices. Under such circumstances, it would be best to have multiple competing PSROs, for those dissatisfied with one, to transfer their allegiance to another. And an appeal system, to appeal against local feuds through recourse to distant judges.

Deliberate Overfunding. Many temporary problems could be imagined, immediately simplified by collecting more money than is needed. Allowing the managers some slack eliminates the need for special insurance for epidemics, special insurance for floods and natural disasters, and the like. Listing all the potential problems would scare the wits out of everybody, but many potential problems will never arise, except the need to dispose of the extra funds. For that reason, it is important to have a legitimate alternative use for excess funds as an inducement to permit them. That might be payments for custodial care or just plain living expenses for retirement. But it must not be a surprise, or it will be wasted. Since we are about to discuss doing essentially the same thing for everybody under 65, too, any surplus from those other programs can be used to fund deficits in Medicare. But Medicare is the end of the line, so its surpluses at death have accumulated over a lifetime, not just during the retiree health program.

That outline may not be more accurate, but it displays its assumptions better. Michigan Blue Cross has calculated we calculate lifetime costs and Obamacare costs by starting with lifetime average health costs of $325,000 and subtracting Medicare. Although Medicare is reported by CMS to have average costs of $11,000 a year, for which we prefer to assume a Health Savings Account "present value" cost of $80,000 on the 65th birthday (at a 6.5% interest rate). At the same 6.5% rate, a $3300 annual deposit from age 25 to 65 (the earning years) would total $132,000 of deposits. The striking fact is, however, that Medicare alone could be pre-paid by an escrow of $150 to $350 a year, from age 26 to 65, providing it can generate 8% compounded investment income. The entire staggering cost of Medicare would hardly add any expense, within a lifecare financing system. Preliminary goals for a hypothetical average person are: To accumulate $57, 000 in the Medicare escrow fund by the age of 65, to pay off the 25-year health costs of 2.0 children per couple as a gift to them, and to pay his own relatively modest average healthcare costs from 25-45, somewhat higher costs 45-65. The Medicare goal of $57,000 is what is estimated to be what is required for a single-deposit investment fund (paid on the 65th birthday) to pay the health costs for an average person aged 65-93,(a guessed-at future average longevity), with an estimated compound investment income continuing at 8%, also guessed. Inflation is ignored, assuming revenue and expenses will inflate at the same rate. Our average consumer will have to set aside $150-350 per year from age 25 to 65, and earn 8% compounded, to do it. Different contributions at different interest rates will produce different results. We defend 8% in a later chapter.

Those who disagree with the underlying assumptions should feel free to substitute their own assumptions. The interest rate of 8% is deliberately high, in order to make room for disagreements which are higher. The upper limit of life insurance ($132,000) is set to match the HSA contribution limits of 3300 times 40, becoming hypothetically the upper bound of revenue which can ever be anticipated, and from which $150-350 is escrowed for Medicare replacement. Anticipating two children per couple and full employment from 25 to 65, this revenue effectively covers one full lifetime, from cradle to grave. Childhood illnesses and elderly disabilities notwithstanding, this is all the revenue we allow ourselves in this particular example. Quite frankly, $3000 per year for age 26-65 is the weakest part of the estimation, because it is most dependent on the general state of the economy, the number of indigent immigrations we permit, and the competition of other worthy goals for the same resources.

Let us assume that an average person can start contributing to an H.S.A. at the age of 25, even though perhaps a quarter of the population at that age are burdened with college debts, etc. and cannot. We are well aware of the Pew Foundation poll that many of those under 30 are still living with their parents, and many others have college debts. The present ceiling of $3300 annual contribution is otherwise taken as the upper boundary of what is possible for the sake of example, and theoretical deficits have to be made up from whatever surplus is created by such maximums. To plunge ahead with the example, our average person sets aside $3300, starting at age 25 toward lifetime health costs. To simplify the example, he does so whether he can afford it or not, and what he can't supply himself is provided by a subsidy or a loan. Since present law prohibits spending from the H.S.A. for health insurance premiums (this should be reconsidered by Congress, by the way), an estimated premium of $300 for his own Catastrophic insurance is taken from the set-aside, and the remainder is placed in the H.S.A., paying an estimated 8% tax-free. Within this, he eventually needs to set aside a Dependent Escrow premium (remember, this example covers lifetime expenses, even though everyone has Medicare), which for twenty years (until age 45) is zero for Medicare and available for medical gifts to children. After that, it is exclusively used for Medicare, as explained in later sections.

Health Savings Accounts are tax-exempt, and they can earn tax-free investment income. Except it isn't all it could be. Professor Ibbotson of Yale, the acknowledged expert in the long term results of investment classes, has regularly published data going back nearly a century. In spite of military and economic disasters of the worst sort, investment classes have remained remarkably steady throughout the past century and presumably will maintain the same relationships for some time to come. John Bogle of Philadelphia has translated that into index funds of investment classes, with negligible administrative costs. (Caution: Many index funds are sold with very high trading costs, typically in hidden charges when money is withdrawn. Be careful of your counterparty, particularly if he specifies the index fund, because he may limit it to one who gives kickbacks to him.) With this warning, there is a reasonably good chance of getting gross returns approaching 10% for investments in index funds of well-known American stocks, even though the typical HSA at present is yielding less. This investment income can grow to the point where it constitutes a fairly large part of the health revenue.


Instead of starting at birth and ending at death, this book reverses the process for financial reasons. For social and political purposes however, that may not be where further expanding the program can make the most difference. Let me explain. During the first two years of life, it seems likely excellent care would do the most enduring good. The same can be said of the last two years of life because they contain the highest proportion of mortal illness. But after the first two years, there are many decades before healthcare makes the same difference. The same is true of terminal care in reverse; it's preceded by decades of golf, bridge, and television. If we must concentrate expenditures, these four, bookend, years of a lifetime are where to do it most effectively.

There is also a big transition problem in alternative proposals, since voters will be of different ages, and the system must work without gaps. It will take decades to prove any of them have much effect. Concentrate in these four years, however, and changes will be both prompt and wide-spread, a politician's dream. Everybody has already been born, and for a long time to come, everybody will have a piece of his life behind him that he does not want to pay for. The time has passed when Lyndon Johnson could solve the transition problem by simply giving a gift of many years free coverage to most of the new entrants to his system. So, although it will probably spook a number of old folks just to hear the discussion, let's begin with Last Year of Life Coverage, where the data is most accurate. Two years may be a little safer. Next, for political reasons, we would jump to First Two Years of Life coverage. If it is planned to have anything permanent, these are the two minimum goals you would start with. In our wildest dreams, after we have cured just about everything, these are the two features which would remain. Both of those apply to 100% of Americans, and in one sense would be basic coverage. Other end-games are possible, like universal health insurance, or universal good health, or universally top-notch quality care for everybody. But only the year of birth and the year of death are universal and finite. Only these two would be essential to any other scheme of healthcare reform, and therefore teach us the most. If we had to retrench, these two would be the last to disappear. If any health insurance should be universal, these four years have the strongest medical arguments. Unfortunately, right now, they seem to have the least chance of political success. Therefore, it is likely that they will be voluntary and self-pay if they are adopted at all.

Footnote:That isn't quite the case however. Since third party (insurance) payers were placed in the middle of the transaction, and after electronic computers arrived, piles of individual payment data made analysis irresistible. That approach was repeatedly discredited when everyone with a computer found out that increasing the volume of useless data never improves its lack of relevance. The watchword of the 1960s became GIGO, garbage in, garbage out. Expanding the dataset with large volumes of medical data is nevertheless a dream lingering on, eventually running up against a new stone wall. It makes no economic sense to shift the clerical data-entry burden to a physician, the most expensive employee in the system. Although the Affordable Care Act mandates something close to that, it is safely predicted we will restrain the impulse when the cost is fully appreciated. Meanwhile, the utility of just applying more reasoning to aggregate data opened up the vista of a reversed health insurance system. In a sense, this book is a product of that line of thinking; more pieces of data contribute very little, but a new concept changes everything. Unfortunately, although a radical idea can be developed in six months, it may take decades to prove it had the predicted effect.

Originally published: Tuesday, November 18, 2014; most-recently modified: Sunday, July 21, 2019