Older Women: Pushed into Retirement in the 1970s and 1980s by the Baby Boomers?

Article excerpt

Because baby boomers crowded the labor market and competed with older women for part-time and part-year jobs, the labor force participation of older women declined slightly from 1970 to 1985; in more recent decades, women's retirement age rose as "bridge jobs" became more available

The labor force participation of older women in the United States, like that of younger women, has changed dramatically over the past 40 years, but the patterns for the two groups have differed markedly. While the participation of women ages 25-34--particularly married women--increased dramatically in the 1970s and early 1980s before beginning to level off, the participation of women ages 55-69 actually declined marginally between 1970 and 1985, and only then began a pronounced and steady increase which has not yet abated. This article looks at why these patterns have diverged so markedly. Another time of divergence was the immediate post-World War II period, when the labor force participation of older women increased while that of young women declined.

Although changes in age at retirement affect the trends in labor force participation among older workers, the concept of retirement is notoriously difficult to define. In the Current Population Survey (CPS), the only available retirement information comes from a question asking why a woman was out of work in the previous year. But she might report herself as unemployed, or simply not in the labor force, in a period in which retirement might be defined retrospectively as having begun. As a result, this paper will use a number of variables to examine the phenomenon, including not only self-reported retirement, but also annual hours worked, the propensity to be not in the labor force, and the receipt of Social Security benefits.

Literature review

Despite a voluminous literature on older men's patterns of labor force participation and retirement, there appear to be only a few reports that look specifically at older women and a few more that look at both men and women. A frequent topic discussed in this literature is the effect of Social Security earnings tests on labor force participation. This has been addressed in articles by Jonathan Gruber and Peter Orszag, Cordelia Reimers and Marjorie Honig, and Stephen Rubb. (1)

Workers have historically had their Social Security benefits reduced by current earnings. Although these workers are later compensated for this reduction through higher Social Security benefits, the reduction is usually viewed by workers as a tax on earnings and therefore is hypothesized to affect labor force participation among people ages 65 and older. The threshold above which earnings result in a reduction in Social Security benefits was removed in 2000 for those ages 65-69. In addition, legislation that was passed in 1983 caused the delayed retirement credit to increase between 1990 and 2008, allowing benefits to increase up to age 70 for every additional year benefits are delayed.

The evidence regarding women's response to these economic incentives varies. Cordelia Reimers and Marjorie Honig found that men, but not women, are highly responsive to the earnings test; their model indicates that older women's labor force participation is increased by the delayed retirement credit, but not reduced by the earnings test. (2) Stephen Rubb similarly found little earnings-test effect on women's labor supply. (3) Jonathan Gruber and Peter Orszag, however, found opposite results, with little or no significant effect of the earnings test among men, but some evidence of an effect for women. (4) And Marjorie Honig, specifically analyzing effects on married women, found them responsive to their own pension wealth and, to a lesser extent, to Social Security benefits. (5) Responsiveness to the Social Security delayed retirement credit, taken together with the increased and then eliminated earnings-test threshold, might to some extent be expected to have contributed to the patterns observed in chart 1. …