Philadelphia Reflections

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

Related Topics

Pearls on a String:Further Extending Health (and Retirement) Savings Accounts
Pearls on a String: Further Extending Health (and Retirement) Savings Accounts. HSAs are the string. Retirement saving, Privatizing Medicare, and Shifting Childhood Costs-- are the Pearls. Other Pearls to follow.

Pearl #3: Medicare Supplement, Only 20 Percent of a Pearl on the String

Now that we have described Health Savings Accounts as the string linking a string of pearls, we must have a second look at one of the pearls, because in a sense it is two of them. Medicare, it may be recalled, only pays for 80% of it's patient's liability, while the other 20% is the patient responsibility. That is, Medicare has a 20% co-payment, which amounts to a 20% reduction in benefits. Most people who can afford it will purchase a secondary insurance policy from Blue Cross or a commercial insurer, to cover this 20% liability, thus restoring the 20% of benefits at their own expense. Those who cannot afford such policies will often apply to state Medicaid, to become what is known in the trade as a "dual eligible". Those who are not eligible for Medicaid will often just take a chance on their personal resources, often becoming a source of the hospital's or doctors' bad debts. It is thus a curious feature that much of a hospital's bad debts come from the lower middle class.

Co-pay has a long history and a bad reputation. Most textbooks will classify it as a form of patient participation in his costs, and a restrainer of abusive claims. But it long ago developed the practical role of adjusting premium cost to available budget during group negotiations. If you don't get sick, you will have no co-pay obligations, but if you do get sick, it extinguishes 20% of the cost. So, although Medicare co-pay secondary insurance responds to the 20% co-pay feature of Medicare, in company negotiations for group policies for younger employees, it can be 30% or 27% or some other number, because the negotiators discovered that doing so reduced the cost of the insurance by 30% or 27% or whatever. It greatly facilitated middle-of-the night negotiations for the limits of coverage, with calculations on the back of an envelope, and probably had little relationship to restraining medical overuse. Quite obviously, it created the need for secondary insurance, with a double dose of administrative costs and profits. So an expensive and largely futile feature has persisted for seventy or more years, deeply embedded in Medicare for fifty. Usually the carrier for the secondary insurance is the administrator for the 80% which the government pays, but nevertheless, two confusing reports ("explanation of benefits") for two different insurances come trickling into the patient separately, two or three months after one treatment took place. But it does save the government 20% of its cost, so it persists. You will notice this 80/20 formula makes no effort to define or assign the extra insurance company costs and profits, which are negotiated privately. If utilization is affected little by co-pays, costs are nevertheless directly escalated by using two insurances to pay for a single medical encounter. Three insurances, if you include Major Medical policies.

In the case of Medicare, there is another quirk. As mentioned earlier, about 50% of the government's share of the cost, is borrowed. Insurance companies often borrow money, too, but usually not attached to a specific policy. While there may be hidden arrangements between Medicare and its secondary carriers, on the surface it would appear the secondary carrier's 20%, actually represents 33% of Medicare's cash flow. If that's the case, nothing short of a bull in the China shop will dislodge the preposterous dual-insurance system. Furthermore, it seems likely this cash leverage is playing an important hidden role in the ACA negotiations with large group employers. Eventually, this leverage is what might threaten ACA with disruption, but that particular issue gets us off the topic of Medicare, though the issues sound similar.

Having earlier reviewed the finances of the 80% of Medicare, and found that financing it is rather precarious, let's look at the more modest goal of financing the 20% co-payment insurance with the available resources. That's a more modest goal, and a more achievable one, one which would at least remove a large source of public confusion and dissatisfaction. It might, for example, explain why it takes weeks or months for a computerized "explanation of benefits" to appear at the patient's home after he has long since forgotten the charges it matches.

* * * The main purpose of eliminating the Medicare co-pay feature is to eliminate the extra cost of a second insurance administration. Once you grasp the unlikelihood that a copayment would affect utilization if you only feel its impact after you go home from the hospital, you see the argument that it does not affect inpatient behavior. Outpatient costs might be another matter, although even that has not been demonstrated, and the alternate use of debit cards by Health Savings Accounts seems to point in the other direction (see page ). The conclusion would have to be that you have two choices: reduce the duplication of insurance companies, or reduce the duplication of policies. Just exactly which approach would save the most money requires greater access to the data and more expert analysis. Superficially, it seems likely the outpatient use is greater among younger people, resulting in a greater saving after compound interest is applied. A change of this magnitude requires more investigation than nonprofessional outsiders are likely to provide.

Nevertheless, eliminating the copay in some manner would provide a leveraged advantage. The extraction of the cost of the last four years of life would cut Medicare direct costs by 50%, and the elimination of copay would reduce it further. This is an example of the sort of leveraged cost reduction we have in mind. Migration of the center of medical care, from the hospital to the suburban retirement village would be another.

If we were commercial insurance investors dealing with a failing health insurance partner, no additional money infusions would seem sensible until Medicare stopped losing so much money. Because we are talking about a government program however, we must resort to the stance that a new program does not have to accept old debts, only new ones that it had a hand in creating. Therefore, this proposal does not include the repayment of old debts, regarding them as the government's problem to resolve. In many ways, Medicare was a noble achievement, but even the richest country in the world cannot afford to run a 50% deficit indefinitely, in an entitlement program grown so large. Undertaking to correct its mistakes does not imply assuming its debts. Furthermore looking forward, a looming retirement funding crisis, or at least equal size, threatens to replace it as the largest consequence of its heedlessness. Was this lengthening of longevity by thirty years a bad thing? Of course not. The bad thing was to let finances get into their present state before addressing them. The bad thing was to kick the can down the road, for fifty years. Because so few people seem to understand them, let's next review a quick summary of Medicare finances.

The Basic Funding Structure of Medicare. Approximately one-quarter of Medicare is paid for by its premiums, often derived from reduced Social Security payments, (a circular solution, if you regard prolonged longevity as a hidden cost of Medicare). Another quarter of Medicare is paid for by a 3% payroll withholding tax on younger, working people. (Unfortunately, this money is immediately spent, in a process quaintly known as "pay as you go"). And finally, half of Medicare expense ($260 billion annually) isn't paid for at all, it's just debt, initially laundered into general taxation and then floated away by bond issues.

Suggested Solutions:

1. Extract Income From the Float. To attack the problem we would probably need to do many complicated things, but the first step might be pretty simple. We once contemplated a transfer-entrant into this revised program be required to sign an authorization to redirect payments for Medicare cost on his behalf to his own Health Savings Account. (Incidentally, it might also include an accounting of copayments and subsidies.) From the beneficiary's point of view, nothing changes except the postal address of his payments, which becomes his Health Savings Account. If he is between the age of 25 and 65, his withholding tax is so directed; if he is already on Medicare, it is his Medicare premiums. That's a payment stream which stretches sixty years, overall. Depending on his present age, first it is one, and eventually, it is the other. That wasn't so hard, was it?

The money now starts to earn investment income, which is new money for the program, with the surplus eventually going through the Health (and Retirement) Savings Account into retirement funds. One way of looking at this rearrangement is to say the beneficiary has been given the money to pay his bills but relieved of the obligation to pay old debts. He has also been given extra latitude to invest the income and use the profit to fund his retirement. What does the government get out of it? It potentially gets an abatement to annual increases in debt, plus the hope the retirement incentive will restrain cost escalation. If you wish, you could say the principal value to the government is creating the incentive at the end of this and other programs which join the string of pearls. Meanwhile, both parties can pray that science will reduce future medical costs, not raise them. But however it turns out, this solution, unfortunately, does turn out to be too small to make a significant change in the management of Medicare's debt, so we go on to other approaches. However, it raises an interesting point for a brief digression:

Using the shorthand that Health Savings Accounts ought to produce at least a 7% return, and money at 7% doubles in ten years, a quick look at pre-financing present Medicare payments overall is a little disappointing. In the Secretary's report, Mrs. Sibelius tells us annual Medicare expenses are about $560 billion, and cash revenue aims to be half that, or $280 billion. If the cash revenue only resided in the individuals' HSAs, it might add 7% revenue or 19.6 billion per year. That sounds like a worth-while amount, and it could be approximated it would apply to each yearly age cohort for 65 years (45 years of wage withholding, followed by 20 years of Medicare premiums). Since the system has been in place for many years, it has reached a steady state, net of demographic and economic variations. So an age cohort would collect a lifetime average of 65 x 19.6, or $1,274 billion, or 1274 divided by 500 million individual recipients or $2548 per lifetime. It certainly sounds as if the maneuver would be worthwhile because the government would be no worse off, and the subscriber would have $2548 more in his HSA.

But Michigan Blue Cross has estimated the average person spends $350,000 per lifetime for health, half of which is covered by Medicare; and so 25% of that is Medicare revenue. Even by the roughest sort of estimation, this proposed re-direction of revenue would save less than 1% of the cost of Medicare, because revenue is such a small part of the cost. To put it another way, this approach might be an important funding device if indebtedness were not such a large part of Medicare's budget. We have not calculated the effect of compounding, which might theoretically reach several times its original size as stated revenue. On the other hand, neither have we recognized the annual increase in Medicare spending, which its trustees report to be 5.7% per year. Both 7% and 5.7% are fragile projections of the future, one of Medicare spending and the other of the stock market. As long as annual increases in cost are so close to investment revenue from cash revenue, any hope of substantial investment revenue is at the mercy of minor yearly volatility. and cannot be relied on. The best to be reasonably hoped for this proposal is to stop the growth of Medicare deficits. It should be done, nonetheless, but there is no great political advantage to be gained from emerging with the problem apparently unchanged.

The second weakness of this approach is a variation in proportionality between revenue and expenses among various government programs. Last year, the Social Security budget was $888 billion, while the total Medicare budget was $618. A quick glance at my secretary's pay stub reveals she has five times as much withheld for Social Security as for Medicare. The approach of investing the withholdings until the day they are spent is a good one. But if widely applied, would have drastically unexpected consequences unless other things are changed.

Additional Proposals to Supplement Medicare Income. Since Medicare is so underfunded by its revenue, the hope of extracting additional income above 7% is pretty dim. Therefore, the hope of significant cost abatement must come from three other proposals, all of which require the investment of fresh funding. That is, they are investments, not miracles:

2. The Second J-Shaped Curve, Within Medicare. All healthcare costs with the exception of premature birth, genetic disorders and the like, are migrating to older age groups. One of the main sources of disruption is the migration of costly illness from working people to people on Medicare.

But even within Medicare, costs are also migrating into later life. Half of Medicare costs are paid on behalf of the last four years of someone's life. Since Medicare extends about twenty years after retirement, half of the total Medicare cost would vanish from its annual budget as a result of placing this burden somewhere else. This might be called the Last Four Years of Life Reinsurance, a component of the First and Last Years of Life reconstruction of healthcare finance, to be described later. The consequence is partly funding forward toward death, partly funding backward toward childbirth, and so reducing the transition time. The present system, it may be recalled, always funds forward toward death, and buries childhood in "family" plans. That's the background.

But death is the end of the line; costs can't get pushed any later, although curiously, revenue just might be. Therefore, the unique features for transitioning Medicare to some other system reside not only in the universality but also the finality of the cost of terminal care. This entity has a soft lower border, but we know that half of Medicare costs are concentrated in the last four years of life, creating a simple surrogate, although not a pure one. Paying this version of terminal cost separately allows the remaining cost of Medicare to be cut in half, by spreading it over the remaining sixteen years. Transition time is also halted. Moreover, smaller pieces are considerably easier to fit into a transition scheme, so the ultimate product fits the cost curve more comfortably. By the way, the last years of life are not the same as the last years of Medicare, and can only be calculated in retrospect, after the death of the individual.. This is the reality which allows one insurance to be paid out as costs are incurred, and a second, a re-insurance, to repay the first one after the facts are in. Funding the re-insurance from birth allows compound interest to pay for most of the magic in a forward direction. And making obstetrics/pediatrics into a gift from a parent to child, allows it to fund backward. Now, we are beginning to approach the way Nature makes us pay our bills.

3. Contingency Fund. But wherever is this money, half the cost of Medicare, to come from? Terminal care is predictable the day you are born, so you might as well fund it when it is cheap. A separate fund could be imagined, but for simplicity, we lump terminal care financing into a general contingency fund. So we next propose to complete the Medicare revenue issue by adding a contingency fund, essentially substituting a subsidy of $1 at birth for a deficit sixty times as large at age 65, and depending on compound interest to make up the difference. (It may well require about $100 up front, but less if we extended its duration by two or three decades.) How fast it would actually grow in the intervening 65 years would become evident before then, and appropriate adjustments made, but the person or agency to make the decision should be specified with care. The sixty to one estimate comes from 7% doubling principal every ten years, 2,4,8, 16, 32, 64 doublings in sixty years. How much to begin with is actually the last calculation, adjusted to balance the books as experience gathers to improve the estimate, and adjustment applied to its duration.

4. Extended Contingency Fund. The contingency fund ending when Medicare begins might anyway generate a 64 to one magnification of the initial deposit. However, it could extend to 250-to-one if its boundary were the day of average death (now 84) or 1000 to one if it added 21 years to the date of death and ended where the common law now says perpetuity begins (one lifetime plus 21 years). Innovations of this sort make many people squirm, but the underfinancing of Medicare in the past leaves little opportunity for conventionality in the future. All of this magic is a function of the mathematics of compound interest; objections to it are sociological, not mathematical. With leverage of 1000 to one, it is difficult to imagine an inability to pay the front-end $100, when $100,000 is so far in excess of what actually seems needed.

Sweeping proposals of this sort, however, do tend to dump their problems at the far end, so the ultimate goal is best stated to be funding part of the individual system backward to childbirth, partly forward to death, and still having a contingency fund for safety. That is, the system as a whole may be volatile and require internal borrowing, but each individual HRSA ends up with balanced books. By implication rather than calculation, in ninety or so years, you get rid of the Medicare debt. That's approximately how long it took to create it, too. The system does not "cover" retirement, except perhaps for bare-bones Social Security. It merely closes the individual books after death as described in other sections. Last-Year Coverage is also designed to save money, and thus eventually to generate some funds for retirement, but not likely at first. First and Last Year re-insurance is intended to resemble an accordion, quickly going to the first 25 years of life and the last 4 years of life, then slowly adding other years in the middle. Somewhere along the line, it might even shrink somewhat. At the moment, it appears terminal illness usually lasts four years, a process which might be thought of as "breaking the cocoon of health". Half of Medicare expenditure occurs in the last four years of life, leaving quite a surplus when the other sixteen years of cost are redistributed. In any event, the "accordion" effect is available for use in the transitions.

This completes our proposal for refinancing Medicare. The first step is to eliminate the supplemental copayment burden by substituting pre-payment for the 20% revenue, and meanwhile letting small front-end investments grow to appreciable size for the remaining 80%. Until we see what must be done to integrate the ACA with the rest of healthcare, it's likely to use up our political capital with the public, just to make that start. Secondly and later, it reduces itself to stabilizing cost increases by first investing the float created by the J-shaped cost curve, combined with cutting forward-financing loose from the debts of the past. Even a program allowed to concentrate on its 50% forward shortfall, must employ some novel approaches to produce annually $250-300 billion in either cost cuts or new revenue. We suggest compound interest is entirely capable of achieving it mathematically by making a total stock market investment starting at birth and continuing to death, or even 21 years after death. All that is necessary to do it on paper is to invest sufficient money at first. Unfortunately, there is an invisible limit to how much the populace is willing to tolerate as an investment, even on such bargain-basement terms, which I postulate to be about $500 per newborn child. Will that suffices for a 65-year investment? Possibly. Will a hundred-year investment cover it? Almost certainly. Do we as a nation have the patience for a hundred-year investment or the degree of honesty in our agents to leave such huge amounts un-pilfered for a century? That's far less certain.

How Would This Combined Approach Make Medicare Solvent? Medicare has become so over-extended that conventional approaches are soon exhausted. Like any other proposal that might work, this one relies on approaches which are usually best avoided. First of all, it depends on such long time periods that unexpected events would be the rule, not the exception. Many Congresses of many political parties would have to understand it and leave it unharmed for a century. Secondly, such huge amounts of money are involved that tampering, embezzling and fraud are not merely possible, but inevitable. These two problems would confront any reformer. From these two obstacles emerges the third one. Individual Health Accounts would have less risk than gigantic single payers because some people will be stupid, reckless and venal. If you make up your mind in advance that you will rescue everyone who doesn't succeed, the whole system will be no better than a single gigantic reinsurer overseen by either an idiot or a crook. The opportunities for illegal gains will exceed the opportunities for honest managers. Therefore, smaller is better than bigger, simpler is better than complicated, and success is never guaranteed.

Medicare, possibly fundable, retirement income, probably not. To do the quick math in your head, it is useful to remember money at 7% doubles in 10 years. The Medicare deficit doubles every fourteen years. Since Medicare revenue is half of its expenses, its revenue invested at 7% would generate 3.5% of expenses, or just about enough to cancel out the annual rise in the deficit. Current interest rates do not achieve that, but current rates seldom do. During the eight years of the Obama administration, low-cost total market indices averaged 11% gain. Much of this never reached the average stockholder because the finance industry absorbed it, but things seem to be changing quickly. The pharmaceutical industry may possibly be over-represented in the index, but we are proposing to make the average patient become an average stockholder. Let's take the four components:

#1. The Contingency Fund. is designed to be overfunded for contingencies, so it is hard to say how much it should be. The most conservative investment period would terminate at death, but expand to whatever age is necessary, up to age 105. That means the $500 initial deposit never varies. Congress might, however, decide to vary the initial deposit to use a shorter time period. It makes no mathematical difference, but its political difference might be considerable.

#2. Delay Liquidating the HRSA at death. Although things get a little threadbare beyond this point, there is no reason to hold back borrowing for a purpose. We are at the point in the compound interest curve were holding the funds for ten years after death would multiply the original subsidy by 128 instead of 64. We are paying the Chinese much less than that for the Treasury bonds, and they would probably be greatly relieved to see a way of recovering their investment. It may not sit very well with some people, but it would surely guarantee repayment. At the moment, repayment looks rather doubtful.

#3. Investing the Pay as You Go. The problems created for others in the payment process have to be reckoned with. We propose the individuals continue to pay/go temporarily for half of the withholding tax receipts, effectively unchanged because half the cost has been transferred but the withholding tax revenue remains constant. What is essentially involved is to balance the problems of the current budget staff against the problems of passing acceptable legislation. But once more, the mathematical "sweet spot" is comparatively easy to calculate, but the political effects are more intangible. It is probably impossible for an outsider to have a firm opinion.

Additional unknowns in this equation are how much nursing home costs from state Medicaid plans would eventually emerge as Medicare deficits. It is common knowledge that although custodial costs are not allowable costs, states have found ways to make them a federal responsibility. We also understand the HRSA owner may get less than 7% income on his deposits. Although the Chinese debt would stop rising, past indebtedness remains unpaid. Current Medicare bills would have to be paid for probably another decade, and may well rise in size. Ultimately, the way to balance the books is to raise the contributions. So, privatizing Medicare might or might not make it cost less, but would greatly relieve its present costs. Funding for past debts will have to come from other sources. However, contributions from the two contingency funds could easily be increased.

#4. The Last Four Years of Life Half of Medicare costs appear in the last four years of Life. By reimbursing Medicare for the last four years from other sources, Medicare's average cost is cut in half. but the withholding tax remains the same. Therefore, we come closer to breaking even in several decades, although we probably won't quite make it.

#5. Simplicity, Simplicity. To begin with the opposite of simplicity, two quite unacceptable new ways to manage the medical payment system suggest themselves. One alternative is to consolidate the whole industry, with one corporate administrative arm assuming the payment tasks for everybody, along with the whole delivery system. That scarcely seems appropriate management for a health complex which is already too big to manage. But it seems to generate many current proposals, especially those coming from the bureaucracy itself. Another idea, based on its resemblance to whole-life insurance, proposes a giant company or government department to concentrate on health finance, doing it for everybody. It might seem suitable for an insurance company, a medical school, a computer company, or a medical society. That seems to be what these organizations would like, but it immediately creates additional complexity, because computers only work if you specify some response to every contingency in advance. In a sense, this version of "Single Payer" would be a throw-back to the days when only a big company or a big government could afford to own a computer.

Is medical finance really so complicated most people couldn't handle it by themselves? Let's remember the anguished words of the Tzar: "I don't run Russia. Ten thousand clerks run Russia." What the Tsar was saying, was the problem isn't individual complexity, the problem is the huge volume of simple problems. For example, if we proposed to butter everybody's bread, it wouldn't be hard to do, it would be hard to manage.

{top quote}
Is medical finance really so complicated most people couldn't handle it by themselves? {bottom quote}
Transfer Slips, and Monthly statements, Only. So, yes and no to computers, which what all this amounts to. Abundant cheap computers tempt us to use them for simple tasks, at the risk of making the simple task complex. (In another generation, a self-correcting code may correct this problem, at the same time it widens the opportunity for vandals.) The proposal made here instead is a confederation of otherwise free-standing organizations (The Pearls), hiring their own experts, feeding into a common channel of Health Savings Accounts owned by individual patients (The String). Individuals could hire consultants if they pleased but the decisions should be so simple the average high school graduate could cope with them.
One standardized lifetime account form, which serves as a transfer system for a single person's various balances. Sort of like a checkbook. It provides a common incentive to be frugal for future retirement, and a common way to multiply such savings.
If that won't suffice for some tasks, we are traveling down the same path as the income tax, and should re-consider such high-handedness.

There might be many networks, as long as their balances are uniformly transferable and they each link ultimately to a transferable retirement fund (The Goal) and a transferable investment fund (The Multiplier). Such networks might grow very large, but still, remain quite simple, and decisions which belong to the patient would remain within his control. The only outward purpose of such paperwork would be to transfer credits of the owner to debts of the same owner or vice versa, with the adjusted balance ultimately coming to rest in his retirement account, creating a common incentive to be medically frugal. They would maintain adequate records (which mostly no one ever reads), an information source, and a designated HSA representative, but their outward form and purpose would remain recording a transfer slip. If you want a simple system, give it to individuals who have an incentive to keep it simple. Don't give it to people who have an incentive to make it complicated.

{top quote}
If you want a simple system, give it to individuals who have an incentive to keep it simple. {bottom quote}
This particular feature has a political element. The American public now imagines it gets a bargain with Medicare, somehow getting a dollar of healthcare for fifty cents, and therefore a treasure they are unwilling to surrender. In all probability, no organization except the government could function long with such a deficit, so taking the deficit away from the government necessarily places it in the hands of someone who must balance his books. Somehow, legal protections for the patients against the debts of organizations which participate in the confederation must be established, so they can occasionally provide benefits at a loss, but only within stated limits. Called a "loss leader", the situation is a common one. Two additional savings multipliers must be added, although they will be explained shortly, along with two important investment designs. There are four large sources of new revenue within Medicare:

Investment Mechanisms.We promised to discuss two investment mechanisms which might help matters. The first is the tendency of compound interest to rise with time. We have already shown above that adding another decade to the example will have an exaggerated effect on the outcome. This is an inherent quality of compound interest which crept up on us as science has conquered early death, and should have wide application in the future. As we learn how to avoid borrowing and learn how to be successful creditors, it should become a commonplace to rearrange financing to optimize it.

The second new model is index investing. As international borrowing has vastly increased the money supply, interest rates seem to have settled at a new low. Bonds have always been a zero-sum investment, but recent trends seem to set an even lower boundary. Common stock has more risk and volatility, but John Bogle and others have shown that it is practically useless for an ordinary person to buy anything but total-market common-stock index funds since the fees charged by intermediaries wipe out any profit from active investing. We recommend a heavy emphasis on this method. Beyond that basic approach, other strategies may be considered as a way to add fractions of a percent to total returns, best avoided by people without experience, or lifetime years to recover from investment misjudgments.

In Final Summary of Privatizing Medicare. Even with considerable twisting, Medicare is so underfunded, no way can be found to self-fund it without adding two to five hundred dollars per person as a pump-primer, adjusted for being added at birth. That's a great bargain of course, but it will meet far more resistance than five hundred dollars is worth, mostly centered on the long transition time. Even with tinkering, it would require forty or fifty years at the most optimistic. In the Pearls on a String concept, the deficit might be made up by surplus generated by other programs, but it is unlikely to be able to identify such a cross-investor. The Affordable Care Act does not look as though it is going to generate a surplus, for example. In the long run, further privatization of the IPO variety seems the only way to shorten the transition time to the point where the public would accept waiting so long. Nevertheless, although complete resolution of the problem is probably presently impossible, there is no reason why partial solutions would not help.

Originally published: Thursday, June 30, 2016; most-recently modified: Thursday, May 09, 2019