Philadelphia Reflections

Multi-Year Catastrophic Health Insurance

Since we propose a monitoring agency for health insurance, we might as well add a few suggestions for its formative years. Insurance companies with experience in the field should, of course, be represented, and although an actuary might represent a company, actuaries as a profession should be invited to the governing and advisory committees. My own professional experience suggests the President of a large and influential institution will often nominate himself. Then, after a meeting or two, he sends his chief deputy. But scattered through the field are apt to be occasional revered experts, working for obscure institutions with little political clout. These are the ones you want on the brainstorming committees, and they are more likely to be nominated by the professional organization than by the employing ones. For example, let's imagine multi-year catastrophic insurance has been proposed for multi-year pricing.

Catastrophic health insurance is an indemnity plan, and its tradition is to be a one-year renewable one. That is, it's term insurance when we are starting to propose the whole-life model might be cheaper and have some other advantages. What are the pros and cons of multi-year Catastrophic coverage? It's probably cheaper because of lower marketing costs but lacks the flexibility of changing its premium frequently, in rapidly changing marketplaces. And it provides the opportunity to drop a troublesome customer, which is to say it shifts power somewhat in the direction of the insurer. At least in the life insurance market, most of the profit derives from customers who drop their policies, an unfortunate incentive arrangement. A particular hazard is to attract a disproportionate number of clients who have hidden information about impending health problems, leading to "adverse selection" of clients. That leads to requiring physical examinations, which raise costs.

Although some arbitrary time period, let's say five years, may be arbitrarily selected, everybody would recognize that accurate statistics might suggest a better time interval, let's say three or seven years. The management of the company would naturally prefer a shorter period in which to make mistakes, but it might be just as satisfactory (and therefore cheaper) to pick a longer period. Such decisions are often made by a board member with a booming voice, but it would be better to base them on statistics derived from pilot studies. For example, in seven years the client is seven years older at the end of it, and more likely to have serious illnesses intervene.

That may or may not be decisive. With the one-year term, the client is likely to apply for the insurance "on the way to the hospital", that is after he suspects he is sick, but before the insurer can prove it. That leads to waiting periods of variable length, again decided by the booming voice unless you can produce data that he is wrong, and often not even then. Statistics are the timid actuary's friend when he is in conflict with an aggressive executive who came up through sales. The best statistics are derived from pilot studies aimed at the particular question you are asking.

A committee might have better suggestions, but my guess is there are two curves: a slowly increasing one, with age. And a sharply decreasing one with the duration of time since application. My guess is that cheaters will decline in a year, hypochondriacs in three years, and cancer victims in five. The incidence of cancer is age-related, possibly the others are, too. There may be some other factors at work, which will surface in the first study, and have to be examined in a second one. But eventually, it should be possible to prove just what added risk appears at what age, and how much that is attenuated by adverse selection.

Knowing the added risk, it should be possible to set age-related risk adjustments, modified by different surcharges for the first year, the second, etc. And then, it should be possible to judge the best number of years to cover, since the adverse selection should only appear in the first round, not much in the renewals. And finally, it should be possible to see whether lifetime catastrophic rates might be a commercially viable possibility. Once that is established, it becomes like the fable of Columbus and the egg; everyone can do it.

Co-insurance. This seems like a good moment to introduce the subject of co-insurance. Co-insurance is the main reason for supplemental insurance to cover it, resulting in two insurance policies to pay for one illness. One presumes it was the reason Obamacare does not include it. Although it masqueraded as a utilization control, giving the customer some skin in the game, a 20% co-pay feature had very little effect on utilization. It was customary because a 20% co-pay makes the premium 20% smaller, a 30% co-pay would make the premium 30% smaller, etc. It was a handy tool for midnight labor negotiations, but otherwise, it was a burden.

But in the example we are following, it is very handy to have a fixed relationship between the surcharge and the added risk. Therefore, a demonstrated added risk of, let's say, 14.7% in the first year after initial purchase, would be adequately covered by a 14.7% surcharge on the base rate for the age and gender group. In the second year it would be less, and in subsequent years, still less. The insurance design is simple, once you can define the risk. Almost any other risk would be subject to the same rules: tell me the risk, and I'll tell you the premium.

Originally published: Thursday, August 13, 2015; most-recently modified: Friday, May 24, 2019