Philadelphia Reflections

Details of Lifetime Health Savings Accounts (L-HSA)

If we propose to adopt the whole-life model for Health Savings Accounts, then why don't we just add it as a new product for the companies who are already in the whole-life business? It's a good question, and most of the answer is I don't happen to own an insurance company. Somebody has to invest a pile of money to own one. You almost never hear of corporate pirates attempting a take-over, and many insurance companies make their profits on subscribers who drop their policies, although that's mostly term insurance. Come to think of it, these are mostly 19th Century organizations who sort of had the good luck to encounter windfall profits when subscribers lived longer than was necessary to break even, and then even kept living on some more. It isn't exactly the background of people who start new businesses with new ideas. Nevertheless, they do sell their products to young people, invest the premiums for many years, and eventually pay their bills to old folks, on time and cheerfully. And there would seem to be plenty of incentive. Aggregate retirement income fifty years from now will probably be many times as large as the present face value of insurance, and probably include a larger proportion of the population. They already have actuaries on their payroll who could do the math, and who yearn for the day a new product would give them a shot at being CEO. Like me, they have already had a look at the C-suite offices, and like me, compare them favorably with the Temple of Karmac.

Who will run L-HSA, once it is legal?

However, I don't know any of them personally, so the assumption must be made that L-HSA will grow out of companies that package C-HSA at present and sort of have a term insurance point of view. Elements of it have been around for almost two centuries, but even the life insurance industry might become dubious to hear whole-life coverage of health insurance presented as an investment. As the level of income taxation rises, tax-free internal transfers assume new value, but resistance to higher taxes will grow. Extending life expectancy gives compound interest much longer to grow, it thus transforms what it can do. In addition, if you desire intergenerational cost shifting for health costs, you probably must incorporate some form of insurance as a pooled transfer vehicle. This combination also enables funds to shift within the account to a later time in life. It's a more attractive individual incentive for savings, than threateningly proposing your generation must support mine.

As a final feature, Catastrophic high-deductible is here added, providing stop-loss protection. Call it re-insurance if you prefer. It's single-purpose coverage, based on the idea that the higher the deductible, the lower the premium. So it follows that the longer you are a customer, the more catastrophic insurance you can afford. Cost saving runs through all multi-year ideas, but lifetime coverage is a cost-saving whopper, because of the way Aristotle discovered compound interest turns up at the far end. (By the way, that's why I suspect we have rules against perpetuities of inheritance.) It transforms Health Savings Accounts into a transfer vehicle for funds, from one end of life to the other, and must add debit-card health insurance for current expenses. Forward from the surplus of the present. And backward from the compound interest of the future. The last-year-of life could be chosen as an example because the last year comes to 100% of us, and is usually the most expensive year in healthcare, not greatly different from the face value of life insurance. But needs differ, and a ton of money sounds pretty good at any age. A Health Savings Account can also be used as a substitute for day to day health insurance. Another synonym might be Whole-life Health Insurance, although multi-year health insurance is probably more precise. The idea behind presenting this concept piecemeal is to provide flexibility for both overfunding and underfunding, since the time periods for coverage can be so long (and the transitions so variable) that both eventualities would occur simultaneously to different individuals.

The simple idea is to generate compound investment income -- not presently being collected -- on currently unconsumed health insurance premiums. And eventually, to apply the profit to reducing the same individual's future premiums. Even I was then startled, to realize how much money it could save. It's a scaled-up version of what whole-life life insurance does for death benefits. Since lessened premiums generate greater investment income, the math is complicated even when the theory is simple, but every whole-life insurer has experience with smoothing it out. For example, if someone had deposited $20 in an HSA total market Index fund ninety years ago, it would now be worth $10,000, roughly the average present healthcare cost of the last year of life. Neither HSAs nor Index funds existed ninety years ago, and of course, we cannot predict medical costs ninety years from now. This is only an example of the power of the concept, which we can be pretty certain would save a great deal of money, but skirts the guarantees about just how much.

There's one other advantage to using HSAs within the whole-life insurance model. It has always bothered me that life insurance tends to gravitate toward bond investments, matching fixed-income revenue with fixed-outgo expenses. But insurance companies largely support the bond market, which is many times as large as the stock market. In effect, their situation encourages them to increase the amount of leverage in the economic system, thereby increasing its volatility, and its tendency to experience black swans.

Furthermore, the insurance industry has accumulated a great many special tax preferences, based on the notion its social value is a good one. Placing life insurance in competition with non-insurance providers of the same services would justify extending the tax preferences to the others as well. The resulting competition would invigorate what has become a pretty stolid plodding citizen, with somewhat unique power over state legislatures. State legislatures, in turn, would benefit from increased competitive points of view among their lobbyists.

People would be expected to join at different ages, so the ones who join at birth in a given year have accumulated funds which would be matched by late-comers. In our example, if a person waited until age twenty (and most people would wait at least that long), he would need to deposit $78 -- not $20 -- to reach $10,000 at age 90. It's still within the means of almost anyone, but the train is pulling out of the station. Participation is voluntary, but no one saves any money by delaying and learns a bitter lesson when he tries. Notice, however, no one pays extra for a pre-existing condition, either; it costs more to wait, but it does not cost more to get sick while you wait. If the government wants to pay a subsidy to someone, let the government do it. But nothing about the whole-life retirement system compels increased premiums for bad health or justifies lower premiums for good health.

Whole-life health insurance takes advantage of the quirk that the biggest medical costs arise as people get older, and similarly, health insurance premiums are collected early in life when there is considerably less spending on health. The essence of this system is to reform the "pay as you go" flaw present in almost all health insurance. Like most Ponzi schemes, the new joiners do not pay for themselves, they pay for the costs of still-earlier subscribers, a system that will only work if the population grows steadily and/or prices rise. When the baby boomers bulge a generation, they bankrupt the system, but only when they themselves start to collect. Everybody knows that. What is less generally known is that "pay as you go" systems fail to collect interest on idle premium money; the HSA system does that, and it turns out to be a huge saving unless the Industrial Revolution stops. Medicare and similar systems don't collect interest during the many-year time gap between earlier premiums and later rendered service; potential compound interest is therefore lost because payroll deductions are used for other purposes. "Pay as you go" is only half of a cycle; adding a Health Savings Account converts it into a full cycle like whole life insurance, and furthermore returns the savings to the individual, rather than using them for insurance company purposes. Whole-life life insurance is more than a century old, but health insurance somehow got started without half of it, the half which could lower the premiums. Nobody stole those savings, they just weren't part of the gift.

All this creates an incentive to overfund the Health Savings Account. The surplus which remains after death is a contingency fund, probably useful for estate taxes or other purposes; but on the other hand, the uncertainty of estate taxes creates an incentive not to overfund by much. Most people would watch this pretty carefully, and soon recognize the most advantageous approach of all would be to pay a lump sum at the beginning, at birth if possible. Before someone roars in outrage about the uninsured, let me say this would work for poor people with a subsidy, and it begins to look as though the Affordable Care Act won't work unless it is subsidized. In that case, a downward adjustment doesn't reduce premiums, it reduces the subsidy.

Investment It seems best to confine the investments of a nation-wide scheme to index funds of a weighted average of the stocks of all U.S. companies above a certain size, and thus offering to pool for those who are (rightly) afraid of investing. This will disappoint the brokerage industry and the financial advisors, but it certainly is diversified, fluctuates with the United States economy, and has low management costs. In a sense, the individual gets a share in nation-wide whole-life health insurance which substitutes long-run equities for conventional fixed income securities. It removes the temptation to speculate on what is certain to occur, but on dates which are uncertain. Treasury bonds might be added to the mix, but almost anything else is too politically vulnerable to political temptations. Even so, it will have downs as well as ups, and therefore participation must be voluntary to protect the index manager from political uproar when stocks go down, as from time to time they certainly will.

One danger seems almost certainly predictable. This book has chosen 6.5 percent assumed return, mostly because it happens to make examples easy to calculate. The actually required return is probably closer to 4% plus inflation. Supposing for example that 7 % is the right number, there is little doubt a steady investment return is only achieved on an average of constant volatility, sometimes returning 20% in some years, and sometimes declining as much or more in other years. Judging from past experience, there will be a temptation for some people to make withdrawals in years of bull markets, which could reduce average returns to 3 or 4 percent in bear market years, and fall short of the 7% average at the moment it is needed. In addition, the officers of Medicare are likely to be tempted to pay Medicare more than a 7% average in windfall years, leaving the running annual average to decline below 7%, just as the trust officers of pension funds once deluded themselves by temporary runs of bull markets. Ultimately this issue reduces itself to a question whether a temporary surplus is really temporary, and if not, whether the subscribers should benefit, or the insurance company. After that is decided, extending or contracting the accordion would get consideration. It seems much better to negotiate these philosophical questions of equity in advance, and establish firm rules before sharp temporary fluctuations are upon us.

Ensuring the Uninsured. Because universal coverage has great appeal, I have gone through the exercise of calculating whether the impoverished uninsured might be included by using subsidy money to provide a lump sum advance premium on their behalf. It would work, in the sense, it would be less costly, but I do not recommend beginning by including it. Reliable government sources have calculated that even after full implementation, the Affordable Care Act will leave 31 million people uninsured. That is, there are 11 million undocumented aliens, 7 million people in jail, and about 8 million people so mentally retarded or impaired, that it is unrealistic ever to expect them to be self-supporting. In my opinion, it is better to design four or five targeted special programs for these people and keep their vicissitudes out of conventional insurance. Better, that is, than to include them in any universal scheme which the mind of man can devise. But to repeat, the mathematics are adequate to justify the opinion that it would save money to include them in this plan with a front-end subsidy of about five thousand dollars, adjusted backward for fund growth since birth. I refuse to quibble about investment size since no one can be certain what either investments or medical science will do in the future. It seems much better to make annual recalculations for inflation and medical discoveries, and then make adjustments through an accordion approach for coverage. There seems to be no need to make precise predictions since any benefit at all is an improvement over relying on taxpayer subsidies, which now run 50% for Medicare itself. This plan will help somewhat, no matter what the future brings, and as far as I can see, it would make the presently unmanageable financial difficulties, more manageable.

George Ross Fisher, M.D.

Originally published: Thursday, August 13, 2015; most-recently modified: Thursday, May 16, 2019