Entitled to Leisure?
Tuesday, September 13, 2011
The growing length of retirement for men in part reflects a decline in the number of years spent working. Is this a good thing?
While Senator Harry Reid might demur, most experts believe raising the retirement age for Social Security is a good idea to address its long-term fiscal imbalance. Indeed, both Social Security and Medicare are fiscally unsustainable in part because life expectancy has increased substantially since these programs began. Consequently, seniors spend an ever-increasing number of years in retirement, during which taxpayers finance a large fraction of their incomes and medical expenses. What is little-discussed is that the growing length of retirement for men in part reflects a decline in the number of years spent working.
Since 1900, male life expectancy at age 20 has risen by 14 years, yet working-life expectancy is currently lower than it was when Theodore Roosevelt was first elected president. While life expectancy rose rather steadily throughout the 20th century, working-life expectancy for men generally declined slowly starting in 1950 (although it has increased slightly since 1990).
A baby born in 1900 had a life expectancy of only 47 years. A baby born in 2007 has a life expectancy of 77.9 years. The health sector cannot take all the credit for this 30-year increase: public health measures such as improved sanitation and clean drinking water certainly played a role. But we also know that rapid growth in the healthcare sector in the United States made at least some contribution to these remarkable gains in years of life. For example, the 70 percent decline since 1960 in mortality for low birth-weight infants born in the United States was almost entirely the result of improved medical care.
At the start of the last century, a 20-year-old man could expect to live an additional 42 years, during which he could expect to work 38 years (figure 10.6a). The average period of retirement was thus relatively short. By 2004, life expectancy for a typical 20-year-old man had climbed to 56 years, yet his working-life expectancy was still 38 years! With a longer life expectancy and no change in working-life expectancy, the expected duration of retirement rose to 18 years, a considerable increase over the four years of a century earlier. Another way to look at this is to consider that in 1900, a man surviving to age 20 could expect to work 90 percent of his remaining life; by 2004, that share was less than 65 percent.
Women have also experienced gains in life expectancy, but their working-life expectancies have taken a very different course from their male counterparts. Female life expectancy has risen even more than for men over the same period—from 44 to 61 years for a 20-year-old woman (figure 10.6b). Working-life numbers for women also rose more rapidly, as women’s participation in the labor force has increased. In 1940, the average woman at age 20 could expect to be actively working in paid employment for only 12 years—less than 25 percent of her remaining years of life. This was 28 years fewer than the comparable number for men. By 2004, this male-female difference had decreased to only five years. It remains to be seen whether this differential is eradicated entirely in the decades ahead.
By 2004, this male-female difference in working-life expectancy had decreased to only five years.
These metrics are far from perfect. First, working-life expectancies actually only measure whether someone is in the labor force, meaning they are either employed or unemployed and looking for a job; moreover, employment includes full-time and part-time work, which would include those working only part of the year due to the seasonal nature of their job (e.g., fishing, agriculture). Thus, depending on one’s education, level of skill, and type of job, a non-trivial fraction of one’s working life might be spent not working.
Second, these are averages that mask large differences within population sub-groups. Careful research has shown that working-life expectancies generally rise with level of education, but are higher for whites compared to blacks even after controlling for education, and, not surprisingly, are much shorter for those with disabilities compared to those without.
Finally, the statistic only focuses on paid work: the figures are not meant to imply that unpaid work such as child-rearing or informal caregiving to family members or friends does not have any value. But social insurance systems such as Social Security and Medicare are built on payroll taxes paid by workers. Thus, their fiscal integrity will unavoidably hinge on what fraction of time the population is working rather than receiving benefits.
Today, men still have a life expectancy that is nearly five years less than women, yet they devote five years more than women to paid employment. Women spend far more time in paid employment than they did a century ago, but even now such work accounts for only approximately half (53 percent) of their adult lives. If these trends continue, increased longevity for men will be associated with increased leisure time. In contrast, increased longevity for women will result in a larger share of their lives being devoted to paid work.
Working-life expectancies generally rise with level of education, but are higher for whites compared to blacks even after controlling for education.
So long as a sizable fraction of retirement spending is tax-financed, two fundamental questions underlie debates over “entitlements” such as Social Security and Medicare. First, what fraction of retirement spending should be bankrolled by taxpayers in the first place? Should Social Security be structured as a “comprehensive package of protection” (as some of its original proponents had hoped), or simply to “pay benefits that provide a minimum degree of social security” (as the program was originally designed)? Likewise, should Medicare be structured to provide comprehensive medical benefits to virtually all retirees regardless of income, or should it instead be limited to catastrophic protection and/or health coverage only for those who could not otherwise afford their care? Second, for those who do obtain taxpayer-financed support, how long should retirement be?
That men age 20 now devote only two-thirds rather than 90 percent of their remaining lifetimes to paid work will be regarded by many readers as a sign of progress. But would we be better off if such men “caught up” to women by devoting only half of their remaining lifetime to paid work? Or would it be better if instead women caught up to men by devoting two-thirds of their remaining lifetimes to paid work? From the standpoint of financing entitlements, the answer is clear. But whether that answer maximizes social welfare will be a topic of heated discussion in the months and years ahead.
Christopher J. Conover is a research scholar at Duke University’s Center for Health Policy and Inequalities Research and an adjunct scholar at the American Enterprise Institute. The charts shown are from his new book American Health Economy Illustrated, to be released in January 2012 by AEI Press. See PowerPoint versions of Figure 10.6a and Figure 10.6b and Excel spreadsheet on trends in life expectancy and worklife expectancy for data, sources, and methods.
FURTHER READING: Conover also writes “Feeling Poorer? Healthcare Bears Some Blame,” “Government Share of Healthcare Is Far Bigger Than Advertised,” and “Health Is the Health of the State.”
Image by Rob Green | Bergman Group