Philadelphia Reflections

Lifetime Healthcare, Using Health Savings Accounts (1)

Can it be done? What would it cost? Since no one can predict future healthcare costs, no one knows how to pay for them. Conclusions like that over-state the difficulty. It's fairly easy to predict minimum available revenue, and fairly good cost extrapolations already exist. Payment feasibility can amount to comparing "no more than" with "no less than". With luck, we can judge the relative probability of success among payment proposals. Since that still sounds like a scam, what follows is step-by-step. We apologize to actuaries and mathematicians, who may find simplified explanations tiresome.

Medical Costs. We use someone else's estimate of present costs and the foreseeable rate of growth. Blue Cross of Michigan, confirmed by federal agencies, estimates the average lifetime cost in America today roughly approximates $350,000 for a male, and about 10% more for females, using the year 2000 dollars. An 83-year overall lifespan necessarily includes history and prediction. The gender difference rests mainly on women's somewhat longer life expectancy, as well as the statistical convention of attributing all obstetrical costs to the mother; it's likely to be a stable ratio. A million dollars for a family of three is pretty daunting. Because there have been so many changes in medical care in the past century, likely to be repeated in the next century as well, $350,000 must be considered a soft number. Most extrapolation errors will arise from it, so an analysis reduces to this question: How close could we come to cover a $350,000 cost, without distorting medical care? To simplify the explanation, the cost goal is reckoned in the year 2000 dollars, so inflation and other adjustments are taken from revenue, to make them match. Inflation has remained steady at 3% in the past century, so 3% per year is accepted for the future in this analysis. Medical inflation is somewhat different; greater than 3% in the past, recently diminished. Only revenue projections are discussed, with costs taken as a given. It isn't perfect, but it can serve as a base for mid-course corrections, providing a margin for error is adequate.

Revenue, per average person. It is easier to estimate future bulk numbers for the whole nation than to predict an individual's future cost. Therefore, our approach is to take national costs, divide them by the population, and concentrate final analysis into a hypothetical "average" person. Since the goal is to compare average costs with average revenue, it is important to avoid using revenue as a basis for costs. The temptation is great, however, because of the well-known tendency of costs to rise to the level of available funding. To the extent possible, such behavioral adjustments are reserved for "dynamic scoring."

Other Data Points. Longevity has risen by 30 years in the past century but is now relatively stable at age 83. Where practical, we have extended calculations to 93, which seems a reasonable guess for where longevity might go in the coming century. How much an average person could or would be likely to accept as a personal expenditure is hard to say, just as it is hard to say how much cost the country would be willing to pay for subsidies to the poor. There is a temptation to use present costs as a surrogate because there is so much uproar about them, but that approach has too much circularity to its logic. After all, the public is complaining. So, whenever practical, we have presented a family of curves which permit the reader to choose between alternatives. The final data point is the maximum achievable interest rate, a matter of enough complexity to require special discussion, which follows after voicing an opinion about the medical future underlying this subject.

The medical side of it. There is fair reason to believe most or all late-developing diseases might originate in the dozen or so complete genes in the mitochondria of cells. These genes are only inherited through the mother and probably originated in the plant kingdom. So the conquest of our currently most expensive diseases -- diabetes, cancer, Alzheimer's disease, Parkinson's disease, and arteriosclerosis -- during the next century -- is not a totally unreasonable prediction. Furthermore, new cure discoveries, while generally expensive at first, eventually become cheap. Mix it all together, and while the costs of the next century may at times be towering, it seems entirely conceivable healthcare payments could become self-sustaining without financial intervention, a century from now. If we generate the means to get to that point, curiously we should give some credit to financiers, like Warren Buffett and John Bogle. If that sounds confusing, read on.

What passive investment income for a Health Savings Account is generally achievable? Essentially, this proposal advocates saving in advance instead of paying after the fact (often called "Pay as you go.") That translates into being paid interest instead of being charged interest. For this, we steer the reader toward investing his savings as much as possible in an HSA (Health Savings Account), rather than an IRA (Individual Retirement Account), or a 401(k) plan, the employer-based equivalent to IRA. There's nothing the matter with these other tax shelters; they are just not as good as an HSA. The only qualified American savings plan to contain a tax shelter on both deposits and withdrawals is the HSA, and even its tax shelter on withdrawals is limited to approved medical expenditures. The Canadian savings plans do have this dual advantage, but in every other qualified American plan, it is necessary to reduce either the deposit or the withdrawal by its estimated taxes at different ages. All graphical representations of IRAs and 401(k)s likewise require a mental adjustment for taxes. In amounts, this large, taxes make a vital difference. After-tax savings vehicles are necessarily less generous, and are not discussed.

So far, so good. Probably the greatest reluctance this proposal will encounter will come from an almost absolute need to invest the HSA in common stocks, which many people sincerely feel is a form of gambling. But to reflect on history for a moment, it took over a century for Americans to overcome their resistance to banks, which are now found on practically every street corner. Pioneer families were obedient to Shakespeare's, "neither a lender nor a borrower is," which indeed remains pretty good advice in most circumstances. But banks were also the foundation of the Industrial Revolution, and their process could also be called a form of gambling. Modern index funds are a far cry from 19th Century mining and railroad stocks. Their risks, while not totally eliminated have been tamed, so the modern economy really has no savings vehicle quite as safe for those who must live in the real world. At this particular moment in time, almost no stock is as risky as bonds, and in Europe cash in the mattress can lose 50% of its value in a month, responding to central-bank changes in currency rates. True, approximately every thirty years stocks fall almost as much, but modern investment has ways of coping with the "black swan" risk, by somewhat sacrificing some of the investment return. Every company eventually comes to an end in about a century, and the only real safety comes from wide diversification of risks substituting for agility in jumping among them. The current total-market index fund model allows for investment in the whole economy at once, counting on remorseless market pressure to purge the index of failing companies, while constantly adding new ones who are succeeding. Holding several thousand successful stocks at once is the new definition of safety. Meanwhile, the new definition of success is moderate but relentless growth, from low costs and low taxes.

John C. Bogle of Philadelphia probably did not invent the notion you can't beat the index (he means the stock market averages like Dow Jones, Standard, and Poor, Russell, etc.), but he certainly evangelized the idea. Let's explain. When you finally overcome the idea of getting rich by out-performing the stock market, the idea reverses itself. The entire stock market is a proxy for the whole economy, and although some people do get rich faster than the stock market grows, hardly anybody gets appreciably richer than the index in the stock market without using leverage , and leverage is only for gamblers who can disguise the nature of their leverage.

Professor Roger Ibbotson of Yale has compiled extensive data for the previous century and demonstrates how relentlessly the American equity stock market has grown quite linearly, varying by asset class but largely disregarding stock market crashes, or numerous wars large and small. While small stocks have grown at a rate of 12.7% per year over the past century, safer Blue chip stocks have consistently grown at about 11%. With big computers, we can see investors in stocks have only received a return of 8%, which sometimes implies the financial industry is absorbing 3% for its expenses and profits. That may not be a fair comment since a considerable portion of the 11% must be invested in lower-yielding bonds to protect against periodic black swan disasters like 1929 and 2008; this point is expanded later. Vanguard, Bogle's fund, reduces overhead cost by matching his portfolio to the index and letting it run indefinitely, a process known as passive investing, which at least minimizes taxes and expenses. Perhaps, over time, ways can be found to widen the investor's lifetime return to more than 8%, but for the time being one must be satisfied with 8%, and 11% remains the ultimate goal for the far future. Warren Buffett does better than that by buying whole insurance companies and leveraging with their cash float; that's not exactly possible for other people. To rephrase the whole business, a total-market index fund offers the 8% current safe limit to passive investing, within a bumpy unsafe 11% world. Furthermore, the 8% contains steady 3% inflation, so investors better not count on more than 5% spendable return. A disappointingly low five percent, relatively safe, after-tax and after-inflation, return. What will that achieve toward paying an average lifetime cost of $350,000? Remember, this is compounded , which has a magic of its own.

Table xxx plots how $400 will grow in response to compounding, starting at birth and ending at 83 to 93 years, at 5% to 12% compound investment return. We've already described why 83, 93, and 5% were chosen, but why $400? It's a personal guess, shown at the bottom of a family of curves which go up to 12%, the current maximum. It represents the amount I guess would be privately regarded as within almost everyone's reach, and if lost wouldn't financially cripple them forever. It would admittedly have to come as a government subsidy for handicapped people who could never support themselves. And since it would be at birth, it would have to seem bearable to young parents. Also included in this family of curves, are 12% (the limit of growth in the stock market over the past century) and other 1% levels down to 5%, so the effect of taxes, overhead, inflation, and bond protection against stock crashes, can be judged. Note in particular how the curves widen around age 60, exposing the new opportunity created by the 30-year increase in longevity during the past century. The consequence of every improvement in the investment return is multiplied appreciably after you reach that bend in the curve. In my personal opinion, this growth is both staggeringly large, but disappointingly inadequate to pay for all future health care with enough margin to justify committing the whole country to risk it. It will pay for a big chunk of it, however.

Another central point of the graph, however, is that a lifetime of investing a relatively small amount -- at reasonably achievable interest rate -- has apparently come within our grasp. In my view, however, we have to do better than this. We have to tweak this basic idea enough to generate more than $400 at birth. Although prosperous people could use it to make a large reduction in their lifetime medical costs, there is not enough room for error, to permit the nation to risk so much for millions of people with only a marginal income. In the following sections, we apply a variety of other variations to convert an attractive idea into a widely useful one.

************* Compound interest always surprises people with its power, and in this example, 5% just about makes the goal. There's not much room for error or contingencies. All of the known factors are conservatively estimated, and it passes the test. What isn't covered is the unknown factor, the atom wars, a stock market collapse, an invasion from Argentina. To be on the safe side, we had better not count on this approach to pay for all of the health care. Just a big chunk, like 25%, does seem feasible. In the immediately following section, we examine the first "technical" problem. The first year of life is just as expensive as the last year of life, and you can't dip into savings.

Originally published: Thursday, March 05, 2015; most-recently modified: Sunday, July 21, 2019