Lots of people around the world aren’t prepared for retirement, as my Bloomberg News colleague Suzanne Woolley ably explained recently. There’s one obvious solution to that problem: Postpone your retirement.

A recent academic study called The Power of Working Longer, cited in the New York Times, finds that working just three to six months longer can raise your retirement income as much as increasing your savings by 1 percent every year for the last 30 years of your career.

Here’s the beauty part: Working longer is a real option, because old age isn’t as old as it used to be. (With exceptions, of course: Not everyone can work at age 65 or 70 or 75; in the U.S., the full retirement age is 67 for people born in or after 1960.)

According to the Social Security Administration’s mortality tables, an American man who turned 65 in 2014 (the most recent data available) had no greater risk of dying in the coming year than one who turned 60 in 1990, 55 in 1957, or 50 in 1900. For women, the mortality equivalent of 65 in 2014 is 60 in 1989, 55 in 1949, or 50 in 1937. Trends are similar in Japan, the United Kingdom, and other nations.

The standard way to calculate the burden of the old on society is called the old-age dependency ratio. It’s simply the number of people 65 and over, divided by the number of those of “working age,” defined as 15 to 64.

But that ratio is misleading and overly discouraging because it doesn’t take into account the decrease in mortality rates at older ages, says Andrew Scott, who is an economist at London Business School and co-author of The 100-Year Life: Living and Working in an Age of Longevity. Old people are less likely to die—and thus, presumably healthier and more capable of continuing to work—than they were in past generations, Scott wrote in a recent article for Project Syndicate called “The Myth of the ‘Aging Society.’”

Another way to say this is that someone who is chronologically 65 is biologically only, say, 60 (or whatever—it depends on the year in the past you’re using as your base of comparison). By adjusting the dependency ratio for decreases in mortality, Scott takes some people who are 65 and older out of the top part of the ratio (presumably too old to work) and moves them to the bottom part of the ratio (working, or at least capable of working). He sent me the spreadsheet he built, which I used to create these charts.

In each country, the dependency ratio, adjusted for declining mortality, is less alarming than the raw ratio. It’s actually fallen for the U.S. and the U.K. (The mortality-adjusted ratio is choppier, apparently because the mortality numbers that are used as inputs aren’t updated every year.)

Scott acknowledges that mortality rates are only a rough approximation of someone’s ability to work in old age, since people can be alive but not healthy. A better measure to adjust by would be the morbidity rate, which is the rate of sickness or disability. But Scott says data on morbidity rates aren’t as widely available—and besides, he says, mortality rates tend to track morbidity rates fairly closely.

Raising the retirement age does raise issues of fairness. People who have the most physically taxing jobs are often those in the poorest health. So while working longer might be a breeze for tax attorneys, it could be a major challenge for dock workers and tree pruners.

Scott acknowledges the concern: “Some people are very fit and some aren’t.” But he says his mortality adjustments are at least a start. “What we really need is a proper measure of aging, and chronology is not that. Only small children and governments are interested in chronological age.”