Skip to content

Retirement Coverage of Women and Minorities: Analysis from SIPP 1998 Data


The female labor force participation rate has been rising gradually for the past several decades; the annual growth in women's labor force participation has exceeded that of men – 2.3 percent vs. 1.2 percent in the 1980s, and 1.4 percent vs. 0.9 percent in the 1990s. Nevertheless, women have different work history patterns from men. Interrupting their careers to take care of children often reduces women's opportunities to acquire retirement coverage. A study using the 1988 Current Population Survey found that women were less likely to be offered pension benefits than men due to occupational segregation and the fact that women work in lower-paying jobs. Also, marital disruptions can jeopardize access to employment-based benefits derived from a spouse (Currie, 1995).

This paper discusses current pension and overall retirement coverage of women (during their working career) by their marital status, race, and age using 1998 data from the Topical Module (Wave 7) of the Survey of Income and Program Participation (SIPP), a longitudinal survey conducted by the U.S. Census Bureau.

Figure 1 shows the marital status of working women age 16 and older who had a job or owned a business in 1998. "Married, spouse present" constitutes the largest share of this group (55 percent). The remainder is divided into never married (25 percent), divorced (13 percent), separated and widowed (3 percent each), and "married, spouse absent" (1 percent).

Women's Retirement Coverage

Gender and racial differences in retirement coverage exist regardless of how retirement coverage is defined. Men and white workers in general are more likely to have pension coverage at their current job as well as during a career than are women and nonwhites, respectively. The gender difference in retirement coverage narrows if IRAs are included, but the race difference widens (see Table 1.)

Coverage by Race. Pension coverage among all women at their current job does not vary noticeably by race. However, when marital status is considered, nonwhite female workers who are "married, spouse absent," widowed, separated, and never married have higher rates of pension coverage than their white counterparts by statistically significant percentages ranging from 3 to 8 percentage points (see Figure 2).

(Click here to see Figure 2: Percent of All Working Women Age 16 and Older Who Had Pension Coverage at Current Job/Business by Marital Status and Race in 1998.)

The reason for higher coverage among nonwhites is that a larger proportion of nonwhite than white female workers are engaged in gender-specific jobs, such as nursing, teaching, and office secretarial work and stay for a long period of time, thus acquiring pension coverage. A large proportion of nonwhite female workers are also found in government jobs and unionized industries, where they acquire pension coverage (Mitchell et al., 1999).

When the retirement coverage during an entire working career (including pensions from current job, current second job, previous jobs, and IRAs) is analyzed, nonwhite female workers have a rate of pension coverage that is 8 percentage points lower than that of white female workers (see Figure 3). Nonwhite female workers who are "married, spouse present," divorced, or widowed have a rate of pension coverage over a career that is 8 percentage points lower than the coverage of their white counterparts. Only nonwhite separated female workers do slightly better (51 percent vs. 48 percent) than white female workers?the difference is small but statistically significant.There is no race difference in retirement plan coverage among women who are never married or "married spouse absent."

(Click here to see Figure 3: Percent of All Working Women Age 16 and Older Who Had Any Retirement Coverage (including IRA) during Career by Marital Status and Race in 1998.)

IRAs. One reason why white female workers have higher retirement coverage during their career than their nonwhite counterparts is because of their higher rates of coverage under IRAs. As shown in Figure 4, white female workers have higher IRA coverage overall and in every marital status than their nonwhite counterparts (21 percent vs. 8 percent). IRA ownership is higher among white than among nonwhite widows (32 percent vs. 9 percent), and higher among white than nonwhite women who are "married, spouse present" (25 percent vs. 12 percent). Among those women who are separated, divorced, and "married, spouse absent," IRA ownership is 8 to 10 percentage points higher among whites than among nonwhites. Also, never-married white women have higher rates of IRA coverage (9 percent vs. 5 percent) than nonwhite women (Figure 4).

(Click here to see Figure 4: Percent of All Working Women Age 16 and Older Who Had IRAs by Marital Status and Race in 1998.)

Coverage by Age. Among women, racial differences in retirement coverage also exist by age groups, but their magnitude varies. As shown in Figure 5, the racial difference is 15 percentage points among younger boomers (64 percent vs. 49 percent), 5 percentage points among older boomers (71 percent vs. 66 percent), and 9 percentage points among preboomers (71 percent vs. 62 percent). The white-nonwhite difference is 13 percentage points among retired women (57 percent vs. 44 percent) but only 2 percentage points among younger female workers.

(Click here to see Figure 5: Percent of All Working Women Age 16 and Older Who Had Retirement Coverage from Any Source by Age and Race in 1998.)

Types of Plans. Women also differ by race and age group in the type of pension coverage at their current job (see Table 2). Overall, nonwhite female workers are much more likely to have defined benefit (DB)-only coverage than are white female workers (21 percent vs. 18 percent) and less likely to have defined contribution (DC)-only coverage (12 percent vs. 16 percent). These patterns may occur because of the jobs nonwhite female workers hold (such as government and unionized jobs), which offer DB plans more frequently than DC plans. The DB-only coverage is 4 to 5 percentage points higher among nonwhite female older boomers, preboomers, and retired workers than their white counterparts. By contrast, white female workers have significantly higher rates of DC-only coverage in all age groups. Among younger boomers, white female workers have almost 7 percentage points higher DC-only coverage than their nonwhite counterparts. The white-nonwhite difference is almost 3 percentage points among younger workers, 4 percentage points among preboomers, and 5 percentage among retired women.

(Click here to see Table 2: Retirement Coverage of Women by Marital Status and Age in 1998.)

Among nonwhite women, there is a 2.2 percentage point difference between DB-only and DC-only coverage among younger workers. This difference increases to 15 percentage points among older boomers, almost 8 percentage points among younger boomers, 14 percentage points among preboomers, and 10 percentage points among retired women.

Among white women, however, there is only a 1.2 percentage point difference in DB-only and DC-only coverage among younger boomers, a 1 percentage point difference (which is statistically insignificant) among retired women, and 7.6 and 6.4 percentage point differences among older boomers and preboomers, respectively. Among those having dual coverage (both DB and DC plan), there are no statistically significant differences in coverage between whites and nonwhites.


Women's pension coverage, particularly among minorities, is subject to many labor market distortions. Occupational segregation and lifetime work experience reduce future retirement incomes of women. Anticipated retirement incomes of nonwhite men and women, according to Mitchell et al. (1999), do not differ markedly, but both have low levels of wealth to draw on in retirement. The SIPP 1998 survey further confirms the low coverage of nonwhite women in all age groups.

New policies are needed to promote retirement savings, especially among vulnerable population groups such as women and minorities, because coverage tends to decline as workers reach their early 50s.


  1. Currie, Janet (1995). "Gender Gaps in Benefits Coverage," in David Lewin, Daniel Mitchell and Mahmood Zaidi (eds.), The Human Resource Management Handbook: JAI Press.
  2. Kreider, Rose, and Jason M. Fields (2001). "Number, Timing and Duration of Marriages and Divorces: Fall 1996. Current Population Reports. Washington DC: U.S. Census Bureau.
  3. Olivia, Mitchell, Phillip Levine, and John Phillips (1999). "The Impact of Pay Inequality, Occupational Segregation and Lifetime Work Experience on the Retirement Income of Women and Minorities." AARP Issue Paper # 9910.


  1. Employed Civilians by Occupation, Sex, Race and Hispanic Origin, Table No. 588, Labor Force, Employment and Earnings, U.S. Census Bureau, Statistical Abstract of the United States: 2002.
  2. It is also true that the cumulative divorce rate from the first marriage is higher among nonwhites than among whites, nonwhite Hispanics, Hispanics, and Asians (Kreider et al., 1996).

Written by Satyendra Verma, AARP Public Policy Institute
October 2003
©2003 AARP
All rights are reserved and content may be reproduced, downloaded, disseminated, or transferred, for single use, or by nonprofit organizations for educational purposes, if correct attribution is made to AARP.
Public Policy Institute, AARP, 601 E Street, NW, Washington, DC 20049