Work can be tiring, all right. But how is it that people considered too old to sit at a desk are young enough to play 18 holes of golf?
These days, a lot of Americans are retiring while they still have years of good health in front of them. Given the increasing financial strain on Social Security and Medicare, some economists and politicians are contemplating putting a toe on the third rail of American politics: They're asking whether the robust good health of so many senior citizens justifies further raising the eligibility ages for the programs.
The best evidence in favor: A National Bureau of Economic Research working paper summarized here shows that most Americans are healthy enough to work longer than they actually do. The chart below shows the share of men who are working at various ages (blue bars) and, within those demographics, the share who are healthy enough to work but aren't (red bars). The baseline is men age 51-54, otherwise known as preretirement. The economists look at the health of those men and what share of them is working and compare them with men at older ages. The study, by Courtney Coile of Wellesley College, Kevin Milligan of the University of British Columbia, and David Wise of Harvard University, finds that health declines slowly with age, but work declines rapidly. The pattern is the same for women. Poor health, in other words, isn't what's pushing most people into retirement. (The authors don't take sides in the argument over raising the retirement age.)
A recent working paper concluded that years of healthy life expectancy after age 65 increased by 1.8 years between 1992 and 2008, according to David Cutler of Harvard University and three other authors. That's a big increase in a short period of time.
Senior citizens are already working more now than they did in the 1980s and 1990s, as my colleague Ben Steverman recently wrote. Alicia Munnell of Boston College's Carrol School of Management has some theories about why, among them the gradual increase in the normal retirement age, which is 66 now and rising to 67; the shift from pensions to 401(k)s, which lack incentives to retire at particular ages; and of course the improving health of older workers and the reduction in jobs requiring physical labor. As Steverman pointed out, workers age 65 and up are at a record in absolute numbers, and in percentage terms they're at the highest level since the early 1960s, before Medicare eased the financial burdens of old age.
The story doesn't end there, though. The problem with raising the normal retirement age is that it would hit the less-educated far more than the well-educated. Another chart derived from the research by Coile and company shows that men with limited education are far less likely to be working at 70-74 than college graduates of the same age (blue bars). The big red bars do show that many more of them could work given their health, but still not as much as college grads could work.
More than three-quarters of older workers with less than a high school diploma have physically difficult jobs, while less than one-quarter of older workers with advanced degrees have similarly taxing positions, according to a 2010 study by the Center for Economic & Policy Research. Less-educated workers who plan to work past 65 generally do so because they have to, not because they want to. My Bloomberg colleague Suzanne Woolley recently reported on a survey by the consulting firm Willis Towers Watson that found that employees who expected to work longer were "less healthy, more stressed, and more likely to feel stuck in their jobs than those who expect to retire earlier."
Is there a way to take advantage of the work capacity of older Americans without harming the less-educated, who are more likely to have physically difficult jobs and less likely to have the skills employers want? "It’s reasonable to be concerned about the people who may not be able to work longer," said Coile.
Unfortunately, there's no way to raise the retirement age that's problem-free. One idea is to raise the normal retirement age to, say, 70, but make it easier for older people to go on disability. Right now, qualifying for Social Security disability insurance is an arduous process that consumes the time and money of both applicants and the government. It would be hard to speed up the evaluation process without making a lot of wrong determinations.
A streamlined alternative would be to base the normal retirement age on a worker's occupation—raising it higher for deskbound jobs. The National Commission on Fiscal Responsibility & Reform proposed something along those lines in 2010 (pdf, page 50), not mentioning occupation as a criterion but "considering relevant factors such as the physical demands of labor." That could lead to a lot of gamesmanship, however, as people lobby to have their occupations declared onerous. That's what happened when Greece tried it: As of 2010, the New York Times reported, the early-retirement provision covered "radio and television presenters, who are thought to be at risk from the bacteria on their microphones, and musicians playing wind instruments, who must contend with gastric reflux as they puff and blow."
One problem with today's Social Security disability insurance is that it's all or nothing. You either can work or you can't, as far as the system is concerned. Some economists have suggested moving to something more like that used by the Department of Veterans Affairs, which rates disabilities on a severity scale of 10 percent to 100 percent, or completely disabled. A sliding scale would seem suitable for an aging workforce, since getting old is one long process of sliding downhill in terms of ability to work.
As this last chart shows, there's room for most people to work longer than they do now, if that's what the economy and the Social Security system need. The chart shows that the share of people working above the traditional retirement age, while higher than in recent decades, remains lower in percentage terms than it was in the 1940s and 1950s.
Keep in mind that life expectancy for 65-year-old men and women is longer now. It was about 14 years in 1950 vs. 19 years in 2010, according to the National Center for Health Statistics. The idea of a multi-decade period of leisure at the end of one's life is very much a modern invention, and a costly one for the economy. That's why finding ways to keep more people working longer—in an equitable way—is a priority.