How Long You'll Live: A Guide to Guessing

May you live a long life
Full of gladness and health,
With a pocket full of gold
As the least of your wealth.
--Irish blessing

May you live until 120.
--Jewish blessing

Live long and prosper.
--Vulcan salute

In many cultures, to live a long, presumably happy, life is the very definition of success. But in the realm of retirement planning, living a long life can be a mixed blessing.

As life expectancies in the U.S. continue to increase as a result of medical advances, so, too, do the chances that retirees who haven't saved enough will outlive their money. Many seniors who assumed they were financially set for life have learned the hard way that a retirement plan based on a life expectancy of, say, 85 may require serious cutbacks to avoid running out of money as they approach and surpass that age. At the same time, most retirees would prefer not to have to pinch pennies throughout their golden years just because it's possible they might live much longer than life-expectancy statistics would suggest. In some ways, estimating how much one will have saved by the time he stops working is the easier part of the retirement-planning equation; it's figuring out how long that money will have to last that's the real challenge.

Of course, it's impossible to know years or decades in advance just how long each of us will live, and most people would just as soon avoid the subject entirely. But estimating how long each of us will live is just as important as other key retirement-planning variables such as savings rate, rate of return on your investments, and withdrawal rate. In attempting to answer this question, many people will look no further than government statistics about average life expectancies for men and women at various ages, but as we'll see in a moment, those can be of limited value.

Seeking Out a Fuller Picture of Life Expectancy
One of the first places many people turn to estimate their longevity is the Life Expectancy Calculator found on the Social Security Administration's website. The calculator estimates life expectancy based on the user's gender and birthday, and the results show the number of additional years the user can expect to live on average based on current age, at age 62 (when first eligible to receive benefits), at full retirement age, and at 70 (the latest age to start benefits).

The tool is good as a starting point, but the problem, as with most life-expectancy estimates, is that it is too general to be of much use. Life expectancy is typically forecast as the SSA calculator expresses it: as an average for a given population. For example, current life expectancy for 65-year-old males is 19.2 years, or to age 84.2, according to the SSA calculator. That means that 50% of 65-year-old males can expect to live longer than that and 50% can expect to die sooner than that. This average figure may be useful for the SSA, which has to project the program's future costs while helping seniors estimate their benefits, but it tells us nothing about a 65-year-old male's odds of reaching 90 or 100. The average acts as a blunt instrument when what we really need is a more comprehensive picture of life expectancy--one that is more effective for retirement-planning purposes.  

Fortunately, the SSA does far more actuarial work than what is represented by its online calculator. In fact, its actuaries also produce tables that estimate a broad range of life expectancies for people born in various years. Using one of these tables, we have created the following chart, which shows how long those who have turned or will turn 65 this year are expected to live by the percentage reaching various ages, expressed as "age at death."

  

By showing the full range of projected life expectancies for today's 65-year-olds, the chart illustrates that many men and women in this group are expected to live well beyond the average. For example, 29% of 65-year-old males and 39% of 65-year-old females are expected to live to 90 while about 3% of males and 5% of females from this age group are expected to make it to 100.

Looking at figures for younger age cohorts suggests even larger percentages of people living beyond these thresholds. For example, among current 45-year-olds, 30% of males and 41% of females are expected to live until at least 90 while nearly 4% and 7%, respectively, are expected to live until at least 100. For current 25-year-olds the trend toward longer life expectancy continues, with 34% of males and 45% of females projected to live to at least 90 and 5% and 9%, respectively, expected to make it at least to the century mark.

With around one third or more of each of these age groups expected to live until at least age 90, using an average life expectancy that reaches only into the mid- to late 80s for each age group would leave a substantial portion of new retirees without enough to live on in their later years, to say nothing of the increasing number who will reach 100.

For individuals who want to be as certain as possible of not outliving their money, planning a retirement that lasts well beyond 100 may be a worthy, if unrealistic, goal. The longer the money needs to last, the less you can afford to withdraw each year along the way. Trying to preserve your nest egg on the remote chance that you'll live longer than almost anyone else of your generation involves a trade off in terms of quality of life. You may find that planning based on a life expectancy projection that puts you in the 80th percentile for your age and gender--in other words, with just a 20% chance of outliving your money--strikes a better balance between guarding against longevity and not depriving yourself of the lifestyle you envisioned for your golden years.

Of course, the SSA estimates are not the only ones the federal government produces. The Centers for Disease Control and Prevention publishes its own life-expectancy tables, with figures that usually are in the same ballpark as the SSA's. (The CDC also breaks down life expectancy by race and ethnicity. Its figures, found at the end of Table 18 in this document, generally show Hispanics as having life expectancies that are slightly longer than whites, with blacks' life expectancies below that of whites.) All life-expectancy projections should be seen as just that: projections. Any number of factors could alter the actual outcome for the broader population or for individual members. For retirement-planning purposes, the only life expectancy that really matters is your own, along with that of your spouse or partner. It's here that more personal information comes into play when guessing at one's own longevity, including the risk factors that could determine where you'll end up along the life-expectancy probability curve.  ​

Key Risk Factors
To help identify some of these factors we consulted a pair of experts at analyzing life expectancies--Maureen Leydon, vice president of individual underwriting for
 MetLife (MET), and Joe Tomlinson, an actuary and certified financial planner based in Greenville, Maine. Both caution that they are not in the business of estimating individuals' life expectancies but instead look at broad trends for specific populations--for example, how an individual's health profile compares with that of other life insurance customers. However, learning how these pros look at longevity risk factors among broad populations can help us identify risk factors in our own lives.

Here's how Leydon and Tomlinson responded when asked about the role each of the following factors plays in life expectancy.

Health profile: Not surprisingly, Leydon says an individual's health profile is among the most important considerations in the underwriting process. "In general the biggest risk factors that we consider also correlate with what the largest causes of death are [in the United States]," says Leydon. She mentions cancer and cardiovascular, respiratory, and blood diseases as among the top risk factors but adds that factors such as family longevity, height and weight, and whether the person uses tobacco products also come into play.

From there she says she would ask questions about the individual's personal health history, asking whether he has been diagnosed with any kind of chronic health problem, whether he visits a doctor on a regular basis, and whether he takes any medication and why.

Tomlinson mentions many of the same factors, along with whether the individual is a heavy drinker and whether the individual gets regular exercise.

Occupation: Leydon says that after a person's health profile, the next obvious factor to look at is his line of work. In particular, Leydon says an underwriter would take note of whether the person has a dangerous occupation or is in the military. Along with the occupation itself is whether the person receives health benefits from work because a lack of access to health care is another risk factor.

Income: Although Leydon says that income in and of itself is not seen as a risk factor in evaluating someone for life insurance, both she and Tomlinson point to it as a factor that can have an impact on an individual's health. "There's indication that overall life expectancy has been improving, and there's some evidence that it's increasing more for the more upscale," says Tomlinson. A 2008 brief by the Congressional Budget Office, titled "Growing Disparities in Life Expectancy," reported that in 1980 life expectancy at birth was 2.8 years greater for those at the top end of the socioeconomic spectrum than for those at the bottom, but by 2000 that gap had risen to 4.5 years. The life-expectancy gap between 65-year-olds in the top and bottom groups had expanded from 0.3 years to 1.6 years in that time span, says the report, which points to several possible explanations for this growing disparity, including gaps between the groups in the rate of smoking-related diseases, obesity, healthy lifestyles, and access to health care.

Time spent abroad: Whether living overseas for work or just visiting, spending significant time in a country with limited access to health care or one that has exposure to diseases such as drug-resistant tuberculosis can increase someone's risk of premature death, Leydon says.

Other factors: Risky hobbies such as skydiving would also be a red flag in the underwriting process, Leydon says. Other behaviors that increase mortality risk include alcohol or drug abuse or a driving record that includes multiple speeding tickets, accidents, or DUIs. Spending time in prison can also have a negative effect on life expectancy, she adds.

An Educated Guess at Best
None of these risk factors guarantees a life expectancy shorter than the average. Indeed, how long you live could be determined by just one of the factors mentioned above, a combination of factors, or none at all.

Tomlinson, for one, thinks too many people run the risk of outliving their money. "I think people tend to underestimate how long they might live," he says. "They look at life expectancy from birth. They may look at Social Security tables from age 65. If they are in good health and take good care of themselves, they may well live longer than the Social Security tables suggest."

Tomlinson sees the trend toward rising life expectancies continuing. "If mortality continues to improve, then the kinds of ages we see people dying at today are going to be different than the ages we see them dying at in 30 years," he says.

He cautions that tomorrow's longer life spans require today's retirement savers to think not just about having enough money to support a longer retirement but also enough to pay for additional costs, such as long-term care.

One last point to remember is that life expectancy is not a fixed measure, and it's wise to revisit it as you get older. An individual who experiences serious health problems early in retirement would likely make different planning decisions than someone who remains reasonably healthy well into his 70s or 80s. Folks in the latter group might want to revisit their plans and assume a somewhat higher life expectancy than they did in their mid-60s.

Also be aware that, statistically speaking, life expectancy rises as we age. That's because life expectancy measured at birth factors in those who will die young, which pulls the average age of death down for the group. Average life expectancy measured at later ages won't include those earlier deaths and will thus be higher. For example, for men who are currently 65 years old, life expectancy at birth was 80 whereas today it's closer to 85.

The point is not to fixate on one number when calculating life expectancy for retirement-planning purposes. Build some flexibility into your plan so that if things change as you go along, you'll be able to adjust accordingly.