NEW YORK (Money) -- QUESTION: Why do you advise people planning for retirement to figure on living to 90 or beyond? I hope to retire at 57, but I'm not going to plan on living beyond 85. That just seems iffy at best. -- Paul, Bridgeport, Connecticut

In the 17th century, French mathematician and philosopher Blaise Pascal posed a question about belief in the existence of God that's known as Pascal's Wager. "God is, or he is not," asserted Pascal. "But to which side shall we incline?"

Pascal went on to say that even though reason cannot provide the answer to that question, we are still better off choosing to live our life as if we believe.

After all, if it turns out God doesn't exist, the worst that happens is we miss out on some earthly pleasures we might have otherwise engaged in. But if we choose to live as if God doesn't exist and we're wrong, the consequences could be hellish, so to speak.

So what do the philosophical musings of some French dude more than 300 years ago have to do with how long you should plan on living after you retire? Well, I think the principle that Pascal applied here (which I know I've oversimplified) also applies to retirement planning.

Think of it this way. Outside of having a known life-shortening condition or being in the final stages of a terminal disease, there's no way any of us can know for sure how long we'll live.

Yes, we can look at longevity statistics and see that a 65-year-old man has roughly a 50 percent chance of living to 85 and a 65-year-old woman living to 88.

But those figures represent averages. In a large group of 65-year-old men and women, many will die before they reach 65, and many will live longer, much longer.

Indeed, a 65-year-old man has about a 25% chance of living to 92. We can refine our estimate by going to a life expectancy calculator that factors in our family history, fitness level and other information, but we're still not dealing with a very high degree of accuracy.

Given the underlying uncertainty, I think that when planning for our life in retirement - whether it's estimating the returns we'll earn or the number of years we'll be around - we ought to consider the consequences of being wrong.

As a rule, if you want your nest egg to last at least 30 years, you should limit your draw in your first year of retirement to about 4 percent of your portfolio's value. Since prices for the things you'll need will continue to rise during your retirement, you'll want to increase the dollar amount you withdraw by the inflation rate each year.

You say you're not going to plan living beyond age 85. Fine. But what if you plan withdrawals from your portfolio on that basis and you're wrong? Those last five or seven or 10 or however many years might not be so golden. They might be rather grim. Who knows, maybe even hellish.

That doesn't mean there isn't a downside to planning for a long life. Perhaps you didn't take a cruise you might otherwise have taken or you cut back in some other way. Call me overly cautious, but I'd rather run the risk of living less large in retirement and possibly leaving some money behind when I go rather than go through my assets and have to eke out my final years in penury. I mean, if my portfolio has run dry at 85 and like the Energizer Bunny I'm still going and going, what am I going to do? Start a second career and build another nest egg?

The reasonable approach, it seems to me, is to build a margin of comfort into your retirement plan. If the life expectancy stats say you've got another 20 years, then, hey, I'd want to tack on at least another five for purposes of gauging how long my money needs to last. If it turns out I depart this earth at age 85 and I still have enough money to get me to 90 or 95, that's okay by me.

I prefer that to finding myself in a situation where I'm 85 and my portfolio is running dry, but I've got miles to go before I sleep.

____________

A get-started retirement plan

Send Walter Updegrave a question

Money 65: Best funds

The last 401(k) guide you'll ever need