Philadelphia Reflections

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

Related Topics

Some ruminations about health financing, written while we wait for the Supreme Court to announce its decision on King v.Burwell.

Multi-year Issues

Lifetime Health Savings Accounts (L-HSA) would differ from ordinary HSA in two major ways, and the first is obvious from the name. In addition to meeting each medical cost as it comes along, or at most managing each year's health costs, the lifetime Health Savings Account would try to project whole lifetimes of medical costs, and make much greater use of compound income on long-term invested reserves. The difference in net income would be quite considerable. The concept seeks new ways to finance the whole bundle more efficiently; one of which is that health expenses are increasingly crowded toward the end of life, preceded by many years of good health, which can build up unused reserves and earn rising rates of income on them. Since the expanded proposal requires major legislation to make it work, it must be presented here in concept form only, for Congress to think about and possibly modify extensively.

This proposal does not claim to be ready for immediate implementation. It is presented to promote the necessary legal (and attitudinal) changes first needed to implement its value. And frankly, a change this large in 18% of GDP is really best phased in gradually, starting with those who feel adventurous. Experience has already shown HSA is most popular with the age group 30-40. By the time the timidest among us have joined up, the transition will have become more predictable. As a first step, let's add another early proposal for the present Congress to consider:

Proposal 17b. Tax-exempt Hospitals Should be Required to Accept the DRG method of payment for inpatients from any Insurer, although the age-adjusted rates should be negotiable based on a percentage surcharge to Medicare rates. The DRG should be gradually restructured, using a reduced SNOMED code instead of enlarged ICDA code, and designed to be used as a Google-like search engine on hospital computers rather than numerical look-ups, except for very common hospital diagnoses.
Overfunding and Pooling. Lifetime Health Savings Accounts, besides being multi-year rather than annual, are unique in a second way : they overfund their goal at first, counting on mid-course corrections to whittle down toward the somewhat secondary goal of precision -- amounting to, "spending your last dime, on the last day of your life". To avoid surprising people with a funding shortfall after they retire, we encourage deliberate over-estimates, to be cut down later and eventually added to retirement income. For the same reason, it is important to have attractive ways for subscribers to spend such surpluses, to blunt suspicions the surpluses might be confiscated if allowed to grow. An acknowledged goal of ending with more money than you need runs somewhat against public instincts and is only feasible if surpluses can be balanced with pleasing alternatives.

Saving for yourself within individual accounts is more tolerable than saving for impersonal groups within pooled insurance categories, but nevertheless must constantly defend itself against the administrative urge to pool. Pooling should only be permitted as a patient option, which creates an incentive to pay higher dividends for it. It also eases the concern about confiscation by making it more cumbersome; very likely, inflation is a realistically greater concern. The menace of rising health cost at the end of life induces more tolerance of pooling in older people, whereas small early contributions compound more visibly if pooling is delayed. Young people must be taught the instinctive belief it gets cheaper if you don't spend it. The overall design of Lifetime HSAs is to save more than seems needed, but provide generous alternative spending options, particularly the advantage of pooling later in life. Because it may be difficult to distinguish whether underfunded accounts were caused by bad luck or improvidence, the ability to "buy in" to a series of single-premium steps should, both create penalties for tardy payment, as well as incentive rewards for pooling them. This point should become clear after a few examples.

Proposal 17c: Where two groups, by age or other distinguishing features, can be identified and matched, as permanently in revenue/expense deficit, or surplus, internal borrowing at reduced rates may be permitted between the two groups to the extent they consistently match. Borrowing for other purposes (such as transition costs) shall be by issuing special purpose bonds. These bonds may also be used to make multi-year intra-family gifts, such as grandparents for grandchildren, or children for elderly parents.

Smoothing Out the Curve. There is a considerable difference between individual bad luck with health, and mismatches between average costs of different age groups. Let's explain. An individual can have a bad auto accident and run up big bills; as much as possible, the age group should smooth out health costs by pooling within the age cohort to pay the bill. On the other hand, compound investment income follows one curve, while illnesses predominate in bulges on a different curve. It isn't bad luck that concentrates obstetrical and child care costs into a certain age range, it is biology. No amount of pooling within the age cohort can smooth out such a systemic cost bulge, so the reproductive age group will have to borrow money (collectively) from the non-reproductive ones.

With a little thought, it can be seen that subsidies between age groups are actually more nearly fair, than subsidies based on marital status or gender preference, or even employers, who tend to hire different age groups in different industries, and can accordingly game their health costs in various ways. On the other hand, if interest-free borrowing between age cohorts is permitted, there must be some agency or special court to safeguard that particular feature from being gamed. All of these complexities are vexing because they introduce bureaucracy where none existed; it is simply a consequence of using individual ownership of accounts to attract deposits which nevertheless must occasionally be pooled, later. Because these borrowings are mainly intended to smooth out awkward features of the plan, every effort should be made to avoid charging interest on these loans or bonds. However, if gaming of the system is part of the result, heightened interest may have to be charged. Using bonds to borrow between age groups is probably cheaper than constant bookkeeping, and more reassuring about the political risk.

Proposal 17d: A reasonably small number of escrowed accounts within a funded account may be established for such purposes as may be necessary, particularly for transition and catastrophe funding. Where escrowed accounts are established, both parties to an agreement must sign, for the designation to be enforceable. (2606)

Escrowed Subaccounts. Buying Out of Medicare.Both Obamacare and Health Savings Accounts are presently expected to terminate when Medicare begins, at roughly age 65. Nevertheless, we are on the subject of lifetime coverage, where we have a rough calculation of the cost ($350,000) and the Medicare data is the most accurate set, against which to make validity comparisons. We want to start with $350,000 at the expected date of death, spend some of it in roughly 20 installments, and see how much is left for the earlier years of an average life. Then, we repeat the process in layers down to age 21 and hope the remainder comes out close to zero. There are several things missing from this, most notably how to get the money out of the fund, but let's start with this much, in isolation for the Medicare age bracket, age 66-85, more or less. For simplicity, we are going to assume a single-premium payment at age 66, which both life expectancy and inflation in the future will increase in a predictable manner, while changes in health and health care eventually reduce healthcare costs, not increase them. Not everyone would agree to the last assumption, but this is not the place to argue.

We know:

(a) The average cost of Medicare per year ($10,900)

(b) How many years the beneficiaries on average are in the age group (18).

(c) Therefore, we know how much of the $350,000 to set aside for Medicare ($196,200),

(d) And know how much a single premium at age 65 would have to be, in order to cover it. ($196,000 apiece)

(e) We thus know how much all the working-age groups (combined as age 21 to 66, 60% of the population) must accumulate, in advance for their own health care costs, when they reach Medicare age($196,000 apiece).

(f) And by subtraction therefore how much is left for personal healthcare within ages 21 to 66 ($128,800).

(g) We can be pretty certain average Medicare costs will exceed those of anyone younger, setting a maximum cost for any age.

Shifts in the age composition of the population will produce very large changes in total national costs in any given year, but should by themselves not change the average lifetime costs. What they may do soon is increase the proportion of the population on Medicare, thereby paradoxically making both Obamacare and Health Savings Accounts relatively less expensive. Obamacare can predict its future costs with information already available. But the Health Savings Account must still adjust its net future costs for whatever income is produced by investments. We don't yet know is how much each working person must contribute each year, because we haven't, up to this point, yet offered an assumption about the interest rate they must produce. Let's construct a table of the outcome of what seems like reasonable income results. There are four relevant outcomes to consider at each level: the high, the low, and the average. Plus, a comparison with what Obamacare would cost. But there are two Medicare costs: the cost from age 21 to 66, and the cost from 67-85, advancing slowly toward a future life expectancy of perhaps 91-93. These two estimations are necessary for displaying the relative costs of Medicare and also Obamacare.

Someone is sure to notice the apportionment for children is based on income rather than expenses. The formula can be adjusted to make that true for any age bracket, and a political decision must be made about where to levy assessments if income is inadequate; we made it, here. We have repeatedly emphasized that if investment income does not match the revenue requirement, at least it supplies more money that would be there, without it. Somewhat to our surprise, it comes pretty close, and we have exhausted our ability to supply more. Any further shortfalls must be addressed by more conventional methods of cost-cutting, borrowing, or increased saving. In particular, attention is directed to the yearly deposit of $3300 from age 25-65, which is what the framers of the HSA enabling act set as a limit, perhaps arbitrarily.

And finally but reluctantly, the figures include provision for phasing out Medicare, which everyone treats as a political third rail, untouchable. But gradually as I worked through this analysis, I came to the conclusion that uproar about medical costs is never likely to come to an end until the Medicare deficit is somehow addressed. I believe we cannot keep increasing the proportion of the population on Medicare, paying for it with fifty-cent dollars, and pretending the problem does not exist. So it certainly is possible to balance these books by continuing our present approach to Medicare for a while. But it would be a sad opportunity, lost. In summary, we have concocted a guess of the outer limit of what the American public is willing to afford for lifetime health coverage ($3350 per person per year, from age 21 to 66), and added an estimate of compound income of 6% from passive investing, to derive an estimate of how much we can afford. From that, we subtract the cost of privatizing Medicare if our politicians find the courage for that ($196,000) and thus derive an estimate of how much is available for health care of the rest of the population ($128,000).

No doubt, data from the Obamacare program will soon reveal some information, but it will derive from a non-representative sample group. It would seem dangerous at this point to use data from that source to make a judgment about the soundness of this one. Because of the longer time spans available for compound income, at 6% it would actually cost more out-of-pocket to finance the $128,000 than the $196,000; it would actually be financially better to do it. The non-investment cost would only be small, for an expense which otherwise seems almost insurmountably crowding out everything else in the national budget.

What remains is to see if it is possible to finance the average healthcare of younger people with a budget of $128,000. All told, we could foresee a limit of a dollar a day for life, but that would assume using all of the revenue-enhancements together. If for no other reason, a slow transition would make this unrealistic.

Originally published: Saturday, June 20, 2015; most-recently modified: Friday, May 24, 2019