Philadelphia Reflections

Lifetime Health Insurance: Monitoring the Data

The longer we wait to make drastic changes, the more difficult they become, more proof of benefit will be demanded. In the proposed case of switching health insurance from term insurance to whole-life, a century of insurance development contributed insights. But remember, the past fifty years have seen plenty of dissatisfaction come to the surface, only to be dashed by a (generally correct) opinion that the old system was working better than the proposed one would.

Health Cost Monitor Center. This time, let's start in advance with establishing a monitor center, and locate it near other public data centers. Turf issues are inevitable, but bowling teams can break them down. With attention to the issue, an interagency technical center could develop into a useful employment center, attracting superior people with vastly enhanced opportunities for personal advancement. If the center is large enough and sufficiently diversified, it will resist abuse by tyrants.

By waiting fifty years, we now have big data. But we could have had answers, fifty years ago.

Perfect is the Enemy of Good Enough

Conversely, in a population as large as ours, enough people of younger ages will inquiring minds; the problem is to establish the right leadership, and the right opportunity so we can still estimate in advance what difference our proposals make to costs at almost any age. At least then, the public could judge what is actually happening, instead of relying on the pronouncements of political candidates. Half of hospital cost experience already resides in Medicare data, the last years of life are well documented, and the first years of life are fairly predictable for these purposes. So we start the data with pretty good anchors and a panoramic overview of costs. With at least a stated goal of offering suggested prices for lifetime planning, but not necessarily universal ones, we should be able to cope with a voluntary system.

Footnote: An experience forty years ago makes me quite serious about this monitoring issue. While I was on another mission, I discovered Medicare and Social Security are on the same campus in Baltimore, with their computers a hundred yards apart. So I proposed to the chief statistician, that the Medicare computers already contained the date and coded diagnosis of every Medicare recipient who had, let's say, a particular operation for particular cancer. At the same time and in the same location, the Social Security computers contain the date of everybody's death; the Social Security number links the two. So, why not merge one data set into the other, and produce a running report of how long people seem to be living, on average, after receiving a particular treatment or operation -- and how it seems to be changing, over time. (Length of survival = date of death minus date of the procedure) He merely smiled at the suggestion, and I correctly surmised he had no intention of going any further with it. This time, I resolved to write a book, to see if that might have more effect.

Let's ponder about some of the uncertainties which can only result in guesses at the moment, but which in time can be more precise, and provide the material for mid-course corrections:

Considering the sums involved, We don't watch public money very carefully.

You Cannot Have Too Much Data

Transition Calculations. Let's say we have politicians with the skill to persuade the public to phase out Medicare in order to put an end to the foreign borrowing of half its costs and persuade enough beneficiaries to do so by buying themselves out of the program. It has been our extrapolation that this could be done with a single payment of $80,000 at age 66, having earned an income of 6% since age 30. The cost would compare favorably with present payroll deductions for Medicare, although of course, you can't spend it on two different things; the present premium costs of Medicare provide a roughly equal amount. That provides a price goal, for the beneficiary to make a last-minute decision whether to go ahead with the buy-out. And during the years of accumulating the wherewithal in an escrow fund, it presents a distant steady goal. In addition, we also must recognize that a large foreign debt has been obligated to pay for shortfalls of years now past.

The size of the debt is unclear, but we estimate about an additional $80,000 would pay it off. If the numbers come out wrong, the debt payment could be shrugged off, the way the Chinese Imperial government did a century ago, and the way the Greek government wants to do at present. For argument sake, let us assume that debt to be $80,000, but we are unsure how to pay it. (If I sound like Lyndon Johnson, it is because I am dealing with the same topic.) If these numbers approach any kind of accuracy, the process of buy-out could begin, voluntarily, delaying the choice until the 66th birthday. If there is significant interest in the idea, these numbers could be sharpened, but there would always remain some guesswork. Some additional protection of the gamble could be provided by adding a catastrophic insurance policy. In the long run, this whole buy-out idea would be strengthened by migrating the center of care to a nearby retirement village and using the hospital system only for tertiary care. Obviously, it is intended that the proposed monitoring center would be actively involved, and not merely act passively collecting data without designing some large-scale experiments, also obviously voluntary.

Somehow, a large and well-respected health monitoring system could perform a significant service by participating in more demonstration programs, since so often Congress must resort to trying something because it seems like a good idea. Such an approach needs some discreet professional guidance since obviously, Congress would be reluctant to hand over power, simply because some bureaucrat had gone to graduate school or present the Congressional opposition with an opportunity to get rid of them by assigning unwelcome proposals to a protracted study they did not need. Nevertheless, a few well-designed demonstration programs might be a useful way to make some decisions which are currently only based on hunches. It seems likely the kind of monitoring agency needed, would only have the desired outcome if the first Director of it was a strong and highly talented professional, preferably with a distinguished reputation for an impartial examination of the evidence.

That particular function stresses innovative ideas, starting from their beginning. A limited amount of that is badly needed, but sensitivities must be respected. The main function of a monitoring agency is to provide reliable reports to Congress that a developing program is, or is not, on schedule, or is or is not producing the intended effects at a promised cost. Congress is busy, and it is often in a hurry. It needs to be told that something unexpected is happening, or that something expected is not happening before the news media starts to get excited. And somehow it needs to be a welcome friend to the departmental auditors, who will greatly resent an outside body finding something which has been overlooked. As the saying goes, it needs to have some money in the bank before it opens its mouth.

Originally published: Monday, September 29, 2014; most-recently modified: Thursday, May 23, 2019