![Sample our Social Sciences journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110822/TOC-Subject-Sample-2021-Social-Sciences.jpg"})
Abstract
This study uses the AHEAD survey to examine the effects of increasing the Social Security retirement age in the United States by modeling the labor force participation decisions of men and women over the age of 65. Separate probit analyses by gender with interaction terms indicate that race/ethnicity is only marginally important in these decisions after controlling for key factors such as health, physical and mental disabilities, education, and nonlabor income. However, detailed examination reveals highly significant differences by race and ethnicity in all of these critical factors, as well as in the distribution of previous occupations. Although elderly blacks and Hispanics have greater financial need than whites, they have significantly less ability to continue working for pay. Thus, seemingly race- and ethnicity-neutral policies such as increasing the full Social Security retirement age may have disproportionate negative effects on elderly members of minority groups in the US.
ACKNOWLEDGMENTS
The author would like to thank Don Bellante, Marianne Ferber, Marjorie Honig, Gabriel Picone, and Jennifer Ward-Batts for helpful comments on earlier versions. The research reported here was supported by a grant from the University of South Florida Institute on Aging.
Notes
JEL Codes: J140, J150, J180
Estimates are from Bureau of the Census projections (Middle Series). There will, however, be a greater ratio of working adults to dependent children than has been true in the past, due to the declining birthrate. Because the labor force participation of women continues to rise, a larger proportion of the labor force will be comprised of women.
Whether this is actually the case is a matter of controversy. Dean Baker and Mark Weisbrot (Citation1999) point out that those claiming that the Social Security system is in serious trouble base much of their argument on the increasing costs of Medicare. However, Medicare's looming funding difficulties arise from the inflation of healthcare costs for everyone, costs which they claim need to be dealt with by healthcare reform in the private sector. A comprehensive discussion of the funding issues is beyond the scope of this paper.
In 2003, the limit was $87,000.
The FRA began increasing in 2000; it is scheduled to reach 67 years of age by 2022.
Earnings are indexed according to increases in the average wage between the year in which they were received and the year of retirement.
The earnings limit was $11,520 in 2003.
The median retirement age in the US is now 62.
If an individual was married or living with a partner, the partner was also interviewed, regardless of age. Although some spouses as young as 39 were interviewed, for the present study only those over 65 years old were included.
The six “activities of daily living” with which an individual may have difficulty or need help are walking, dressing, bathing, eating, getting into or out of bed, and using the toilet.
Strictly speaking, these would be employment rather than labor force participation rates. However, since in the US unemployment rates among those 70 and over are only about half as large as those of the population in general, these numbers are a fairly close approximation. For example, in 2003, the labor force participation rate for those 70 and over was 8.8 percent. The employment rate, calculated by omitting those reported as unemployed, was 8.5 percent.
Information is provided separately for white and black Hispanics. For purposes of the analysis in this paper, black Hispanics were included with non-Hispanic blacks. Whites, on the other hand, include only non-Hispanic whites, and “Hispanics” refers only to white Hispanics.
The oldest employed woman, who reported working 56 hours per week during a typical week, was 89 years old at the time of the interview; the oldest man, who reported spending an average of 40 hours per week at work, was 90.
Studies have been conducted about returns to Social Security contributions that take into consideration these differences. For examples of fairly recent studies, see Michael D. Hurd and John B. Shoven (Citation1985), Nancy Wolff (Citation1987), and James E. Duggan, Robert Gillingham, and John S. Greenlees (Citation1993), and Charles W. Meyer and Nancy Wolff (Citation1993). However, these types of studies do not address the issue of the abilities or the desires of these individuals to continue working late in life.
Source: National Projections Program, Population Division, US Census Bureau, Washington, DC 20233.
These figures are somewhat larger than those reported by Margaret E. Weigers and Susan K. Drilea (Citation1996), whose source was the 1996 Medical Expenditure Panel Survey Household Component. They found that among those 65 and over, 25.3 percent of whites reported being in fair or poor health, 40.4 percent of blacks, and somewhat fewer Hispanics, 36.9 percent. However, since the ages of most of those in our sample are 70 and over, it is not surprising that we find larger percentages.
In cases where the respondent was unable to answer the questions posed by the interviewer due to illness or incapacity, a proxy answered the questions. Those with no apparent cognitive impairment and those with only a small amount were lumped together.
Weigers and Drilea (Citation1996) found, among those 65 and over, a greater percentage of blacks than whites suffered from cognitive limitations, and still more Hispanics suffered from cognitive limitations.
In reporting their research, Weigers and Drilea (Citation1996) did not separate out those over 65 needing help with ADLs from those needing help with IADLs (instrumental activities of daily living) such as shopping and paying bills, but found that almost twice as many blacks as whites needed such help. Hispanics fell between the two groups.
In 2002, over twice as many blacks and Hispanics fell below the poverty level, 22 percent and 21 percent, respectively, compared to 9 percent of whites. These lower average incomes generally mean less access to adequate medical care. Moreover, researchers have found that African Americans in the US tend to suffer more from high blood pressure and other stress-related diseases than do members of other demographic groups (Arline T. Geronimus Citation2000).
The well-known occupational segregation by gender dictates the necessity of examining the distributions separately.
Due to rounding these percentages differ slightly from the sum of the figures in and .
Cordelia Reimers (Citation1985) found that Hispanic wives may actually be more likely to participate in the labor force than white wives after controlling for language, etc. However, since not all of the variables she used are available for this study, a simple dummy variable was used here.
However, we have little historical knowledge about trends in labor force participation rates of Hispanics in the US since these statistics were not reported separately until 1994.
In this model, any labor income of a spouse is counted as nonlabor income for the respondent.
Results from the same model run using only those with some employment history showed only small differences in the size or significance of the coefficients.
This is not to deny that the same types of financial incentives influence women, but rather to suggest that the widely accepted notion that women are primarily responsible for household work has tended to constrain their work behavior.
Preliminary versions of these regressions showed there to be no greater tendency to work among those with high-school diplomas than among high-school dropouts.
The model could not be estimated with all interaction terms since there were no employed black men with college degrees in our sample. Neither were there any employed Hispanic respondents with graduate degrees.
This age group, the youngest targeted by the AHEAD study, was chosen rather than an older group because these are the ones most likely to be affected by increases in the full retirement age for Social Security. (Recall that those between the ages of 66 and 69 were only included in our sample if they were spouses of the primary respondents.)
