Determining Life Expectancy Is Difficult
by Geoff Colvin
Thursday, June 12, 2008
provided by Fortune


The hardest question in retirement planning is the first one: How long do you expect to live? I'm afraid recent developments are making that question even harder to answer. But it's unavoidable, so what's the smart way to think about it now?

Let's start with what we know: that life expectancy in the United States is about 78. Unfortunately, this fact is useless for retirement planning. The Centers for Disease Control and Prevention announced last year that a child born in the United States in 2005 can expect to live 77.9 years. But you're not a child, and in any case the notion of estimating (to a decimal place) how long the average child born in 2005 will live is obviously crazy, with all that's happening in medical science. The figure is actually calculated by assuming that all 1-year-olds, 2-year-olds, and so on through 100-year-olds have the mortality results of people in that age group in 2005, creating what the researchers call a "hypothetical [or synthetic] cohort." Bottom line: Let's just ignore life expectancy at birth.

What about life expectancy at your age? Those figures are still constructed from the synthetic cohort, but they're arguably less squirrelly because they aren't looking so far into the future. As of 2004, the most recent year for which detailed statistics are available, a 60-year-old man could expect to live another 20.8 years; a woman, 24 years. Now we seem to be getting somewhere; a 60-year-old man should plan on living to 81. But not so fast. Mike Heller, the top actuary for TIAA-CREF, America's largest seller of annuities, points out that the 20.8 figure is just an average. If your retirement plan assumes you'll live to 81, you probably won't hit it right on the button. More likely, either you'll die earlier than you thought or you'll live longer and run out of money. As Heller notes, "Either way, you're screwed."

A better way to think about it is to consider probabilities. Our 60-year-old man has a 19% chance of hitting 90, or to put it a bit more ominously, an 81% chance of not hitting it. Should he be comfortable with a retirement plan that disposes of his last dollar at 90? Would you? By the way, while the odds are heavily against his reaching 95, if he should happen to make it, he then faces a 22% chance of getting to 100.

That is, he does if he's an average American man. But he probably isn't; no matter who you are, you're probably not perfectly average. You're richer or poorer than average, fatter or skinnier, genetically more cancer-prone or less. To address your uniqueness, you can use one of several available online longevity calculators. Answer a few dozen questions about yourself, and they tell you your life expectancy. Trouble is, the science behind them can be squishy. Mike Heller tried one and found some of the questions "inane." It asks, for example, how many new relationships you've developed in the past 12 months. Social networks do affect longevity, but there's no reliable way to connect new relationships formed when you're, say, 25 with how long you'll live. Similarly, you may be assiduous in eating your fruits and vegetables now, but your good habits may not last.

And then, to make all our estimates seem futile, there's the chance of a major break in the past century's trend of steadily increasing life expectancy. New research shows that in several U.S. counties, life expectancy actually decreased in the 1980s and '90s, the first time such a decline has ever been observed. The main culprits were smoking, obesity, and, in men, HIV/AIDS and homicide. On the other hand, some scientists think that advances in medicine will spark a radical increase in life expectancy to ages not seen since the Old Testament, if we can just hang on another 30 or 40 years for the science to get done.

Which leaves us where? Confused and uncertain, I'd say. Yet we still have to answer the first question of retirement planning. Personally, I like the story of California wine pioneer Robert Mondavi. He reportedly drank three-quarters of a bottle of wine every day, a quarter-bottle at lunch and a half-bottle at dinner. He died a few weeks ago at 94. Now that's a plan I can live with.

Copyrighted, Fortune. All rights reserved.