9
Race, Class, and Occupational Mobility
Black and White Women in Service Work in the United States
Marilyn Power and Sam Rosenberg
The economic experience of women of different race-ethnicities is affected by the complex interaction of race, class, and gender. An analysis of these experiences must acknowledge the “interlocking, interactive nature of these systems” (Glenn 1992, 1), and a growing body of literature has begun the difficult process of understanding this dynamic in the United States as it is played out in many venues.1 These studies demonstrate that class and racial-ethnic differences are reflected in different life experiences, different world views, and, important for the current study, different labor force experiences. Additionally, for all women, labor force experience cannot be understood without an understanding of unpaid labor in the household as well; but both the content and the impact of this unpaid labor may be affected by class and race-ethnicity (e.g., by the presence or absence of extended family; by the ability or inability to afford household help or commodity substitutes for family labor).
This study examines and compares the occupational mobility of black and white women who worked in service occupations in their late teens and twenties. Rather than focus on a narrow set of variables, the study replaces conventional hypothesis-testing with an exploratory, storytelling approach, which more effectively illuminates the complexity of the interaction of gender, race-ethnicity, and class in the lives of women. This descriptive methodology enables us to examine how being a service worker when young contributes to a different “life story” for women of different race and class.
Table 9.1
Percentage of Black and White Women Employed in Service Occupations, 1960 and 1994
White |
Black |
|||
1960 |
1994 |
1960 |
1994 |
|
Private household workers |
6.5 |
1.3 |
37.4 |
2.0 |
Other service workers |
13.8 |
15.4 |
22.7 |
23.7 |
Total |
20.3 |
16.6* |
60.1 |
25.7 |
Source: Zalokar 1990, 45; Employment and Earnings, January 1995, 204.
* Discrepancy due to rounding.
Historically there has been considerable occupational segregation by race among women workers in the United States, with black women confined to the lowest level of manual occupations, especially domestic work, farm work, and unskilled factory jobs. However, black women achieved a dramatic improvement in their occupational distribution in the postwar period, largely through a movement out of service and into professional and clerical occupations (Albelda 1986; King 1992; Malveaux 1985b; Zalokar 1990). In this period, the proportion of black women employed in service occupations dropped dramatically, while white women experienced a much smaller decline (see Table 9.1).
Among black women who work full-time year-round, improved occupational distribution was reflected in a rapid increase in their median earnings relative to white women’s, increasing from 74.6 percent of white women’s earnings in 1967 to 96.3 percent in 1975. However, the ratio has declined somewhat since the mid-1970s, and has fluctuated in the past decade between 90 and 92 percent of white women’s median earnings (U.S. Bureau of the Census 1992).
The convergence of median earnings between black and white women during the 1960s and 1970s has led some authors to conclude that occupational segregation by race is no longer an issue for women workers (cf. Fosu 1988). Yet, as the data indicate, black women remain considerably more likely to be service workers than white women. In addition, cross-sectional data on occupational distribution by race may understate differences in mobility patterns for black and white service workers. Because many service occupations are relatively easy to enter, they serve for some young women as a temporary means of earning money while preparing for other, more lucrative jobs. In other cases, women may be stuck in service employment over the long term. Longitudinal data allow us to investigate the extent to which black women are more likely than white women to get stuck in low-wage service employment.
The Current Study: Descriptive Characteristics
The focus of the current study is occupational mobility of black and white women working in service occupations. Service occupations have historically been a major source of employment for black women and continue to employ a significant, though declining, proportion of both black and white women. They are for the most part low-paid jobs with no clear promotional path (the exceptions by and large are the overwhelmingly male protective services). Their relatively easy entry and frequently flexible hours attract a diverse group of women.
Our sample consists of women surveyed by the National Longitudinal Survey (NLS) of Young Women who reported working in a service occupation in 1972, and who also reported an occupation in 1988. The women were 14 to 24 years of age in 1968. Our study picks them up in 1972 when they were 18 to 28, in order to examine a sample with greater labor market experience.2 These women are a fascinating cohort to study because they were beginning their work lives just as the civil rights laws and affirmative action guidelines were beginning to be enforced (however inadequately). At the same time, labor force participation for women was becoming a long-term fact of life, and the civil rights and women’s movements were challenging established racial and gender stereotypes. The sixteen years covered in this study take the cohort through the years when they would be likely to be completing education, forming families, and establishing labor force strategies.3
The women in our sample were employed at least one week in both 1972 and 1988, and worked in service occupations in 1972 (as defined by the 1960 BLS occupational codes). In total there were 261 white women and 135 black women in the sample. Mobility was measured in two ways.4 First, we looked at the broad (one-digit) occupational category of 1988 occupations. Although there are differences in pay and prospects among service occupations, most female-dominated service jobs are very low paid with little room for advancement. Occupational mobility, in many cases, involves leaving the service category altogether. Second, we divided occupations into ranks based on 1970 census data on mean earnings in three-digit occupations in 1969 of the experienced female labor force who worked fifty to fifty-two weeks.5 Mean earnings were used in developing our ranking system rather than the earnings of the members of the sample because mobility is better measured by potential earnings in an occupation than by the earnings of an individual at a moment in time (e.g., the earnings of a new law school graduate may not reflect the relative stature of her job). Occupations were divided by earnings into fifteen ranks as follows:
Rank |
Earnings Range |
1. |
$0-$1,999 |
2. |
$2,000-2,999 |
3. |
$3,000-3,999 |
4. |
$4,000-4,999 |
5. |
$5,000-5,999 |
6. |
$6,000-6,999 |
7. |
$7,000-7,999 |
8. |
$8,000-8,999 |
9. |
$9,000-9,999 |
10. |
$10,000-10,999 |
11. |
$11,000-11,999 |
12. |
$12,000-12,999 |
13. |
$13,000-13,999 |
14. |
$14,000-14,999 |
15. |
$15,000-15,999 |
|
The direction of occupational mobility is defined as follows: Upward mobility: Rank 1988 > Rank 1972 Downward mobility: Rank 1988 < Rank 1972 No change: Rank 1988 = Rank 1972 |
|
NLS Service Workers and Occupational Mobility
1. Mobility by Occupation
Of all the women in the NLS cohort who were interviewed in 1972, the black women were considerably more likely than the white to be service workers: 28 percent of the black women who reported occupations, and 17.4 percent of the white women, were in service work. For both groups of women the most common occupation was clerical work, which employed 42.4 percent of the white women and 31.5 percent of the black women.
White women were considerably more successful at leaving service work than were black women. By 1988, among our sample of 1972 service workers, 74 percent of the white women had left the service category, mostly for jobs in the professional/technical or clerical occupations. By contrast, only a minority, 48 percent, of black women had left service occupations; and those who had were as likely to have moved into operative occupations as into clerical or professional/technical occupations. These observations suggest that service work may be more likely to serve as a temporary occupation for young white women who are preparing themselves for better jobs, while for black women service is more likely to become a long-term job category. Further, for those black women who did leave service work, a higher percentage than the white women moved into blue-collar occupations. It is important to note, however, that few women of either race entered highly paid predominately male occupations; among the white women who entered professional/technical occupations, half became noncollege teachers or nurses (as did nearly half of the much smaller number of black professional/technical workers). And for women of both races who left service work, the most represented occupational category was clerical work, itself relatively low paid and heavily female-dominated (see Table 9.2).6
Table 9.2
1988 Occupations of White and Black 1972 Service Workers (percentages)
Occupation |
White (n = 261) |
Black (n = 135) |
Professional, technical, and kindred |
23.0 |
11.1 |
Managers, officials, and proprietors |
11.5 |
3.7 |
Clerical and kindred |
24.5 |
15.6 |
Sales workers |
3.8 |
1.5 |
Craftsmen, foremen, and kindred |
1.1 |
0.7 |
Operatives and kindred |
8.4 |
11.8 |
Private household workers |
1.5 |
5.9 |
Other service workers |
24.5 |
46.7 |
Laborers, including farm |
1.5 |
2.9 |
Total |
99.8* |
99.9* |
* Varies from 100 percent due to rounding.
2. Occupational Mobility by Rank
In addition to being more able to leave service occupations, white women started out in better paying service positions. Both the low pay and the low range of pay within service occupations are illustrated by our ranking system: traditionally female service occupations ranged in rank from 1 to 4 (representing mean earnings in 1969 of 0–$4,999). Only the male-dominated protective services rose above the rank of 4; three of the white women, and no black women, had entered these occupations, scoring ranks of 5. Nevertheless, even within this narrow range the white women were clustered in the higher-ranking occupations in 1972, with 31 percent at rank 4 (compared to 11 percent of the black women). The mean rank for white service workers was 2.75, significantly higher than the black women’s mean of 2.36 (see Table 9.3).
Table 9.3
Occupational Ranks in 1972 and 1988 of 1972 Service Workers (percentages)
White (n = 261) |
Black (n = 135) |
|||
Rank |
1972 |
1988 |
1972 |
1988 |
1 |
16.9 |
1.5 |
28.1 |
5.9 |
2 |
24.9 |
5.4 |
18.5 |
3.7 |
3 |
25.7 |
16.5 |
42.2 |
36.3 |
4 |
31.4 |
23.5 |
11.1 |
23.7 |
5 |
1.1 |
19.5 |
14.8 |
|
6 |
17.2 |
8.9 |
||
7 |
11.5 |
6.7 |
||
8 |
3.1 |
|||
9 |
||||
10 |
||||
11 |
||||
12 |
||||
13 |
0.4 |
|||
14 |
||||
15 |
1.1 |
|||
100 |
100 |
99.9* |
100 |
|
Mean rank# |
2.75 |
4.85 |
2.36 |
3.91 |
Standard deviation |
1.11 |
1.98 |
1.01 |
1.47 |
Notes:
* Varies from 100 percent due to rounding.
# Difference in mean rank by race significant at the .01 level in both 1972 and 1988, two-tailed test.
Over the period in question both black and white female service workers moved up the occupational hierarchy. However, the extent of the upward mobility was considerably greater for the white women. By 1988, the gap in rank between the black and white women had increased and remained statistically significant. The mean rank for black women was now 3.91, while that for whites was 4.85. The highest rank attained by black women was 7, and 70 percent continued to work at occupations ranked 4 or below. And while 95 percent of the white women remained at or below rank 7, only 47.1 percent were working in jobs of rank 4 or less. In addition, eight white women achieved rank 8, one rank 13 (a lawyer), and three rank 15 (doctors) (see Table 9.3).
The more favorable labor market prospects facing white women can be seen further by controlling for initial occupational position. The likelihood of upward mobility is related to the job held at the beginning of the period. Individuals located near the bottom of an occupational hierarchy are, virtually by definition, more likely to move up since there is little room to fall further.
Thus, service workers are divided into two groups, those with 1972 ranks of 1 or 2 and those with a rank of 3 or above. For both black and white women, the likelihood of upward mobility is greater and the likelihood of downward mobility is smaller the lower the initial rank. For example, 89 percent of whites with a job in 1972 of either rank 1 or 2, but only 57.2 percent of those with a job in 1972 of rank 3 or above, were upwardly mobile.
However, regardless of the rank of initial job, white women fared better. For both high- and low-ranked groups, the white women began and ended at significantly higher ranks, with the gap widening over the period. For example, the racial gap in average rank rose from 0.2 in 1972 for service workers in ranks 1 and 2 to 1.29 in 1988 (see Table 9.4).
Furthermore, holding a low-paying service position in 1972 did not seem to negatively influence the jobs held by white women in 1988. There was no significant difference in 1988 rank between those in the higher- and lower-ranked jobs in 1972. The same was not true for black women. Rather, for black 1972 service workers, being in an occupation ranked 1 or 2 was a significant disadvantage. Women in the higher-ranked occupations achieved a 1988 rank of 4.21, compared to 3.57 for those in the lower-ranked groups.7
Thus, service jobs, even low-paid ones, may be more of a bridge to better positions for young white women and more of a trap for young black women. And as we will discuss below, most of the lower-ranked white women were babysitters and waitresses, and for many of them these occupations served as temporary jobs while they prepared themselves for better employment. Most of the lower-ranked black women were domestics not elsewhere counted and did not seem to have access to the same degree of upward mobility.
3. Methodology
Prior to investigating the role of specific occupations in influencing mobility, we analyze three factors—family size and family structure, education, and social class background—which may also affect women’s job prospects. Our analysis is more exploratory than confirmatory.8 We create a series of tables documenting racial differences in family size, educational attainment, and social class. We develop a “story” by interweaving these racial differences with the racial differences in occupational mobility. We complement this “story” with a discussion of two occupations—domestic work and waitressing—and the differential opportunities they provide black and white women for occupational advancement. We do not test specific hypotheses since we believe that existing theory is incomplete, not being sufficiently complex to account for race and class differences among women. Rather we feel that our exploratory analysis may help to develop hypotheses for confirmatory testing in the future.
Furthermore, we do not use the more conventional technique of regression analysis. First, we believe that family size and family structure, educational attainment, and social class background are intertwined, not amenable to being analyzed via a conventional linear regression. More complex modeling might be feasible, but given the existing state of theory, it too would be little more than an exploratory analysis. Second, the detailed analysis of domestic work and waitressing could not be integrated into the regression analysis. Third, in some categories, there are just a few individuals. These small cell sizes would make it very difficult to run meaningful regressions.
4. Work/Family Differences
Care of children is an important aspect of women’s unpaid labor and can create work/family conflicts. Women’s labor force strategies, and their ability to participate in wage work, are affected by their responsibilities for children. The presence of children may affect women differently by race and by class, partly for cultural reasons, but importantly because of the presence or lack of other economic resources. That is, it is possible that women with more access to resources may respond to child-care demands by reducing their labor force attachment, while women with fewer resources may increase their wage work, both in hours and in efforts to advance. On the other hand, their limited resources, combined with the pressures of work and family, are likely to seriously impede their upward mobility. Black women have a long history of combining work and family, unlike white women; and they also are likely to have more limited access to economic resources other than their own wage-earning ability.9
In our sample, the black women were more likely to face work and family conflicts than the white women, since they had more children and had them at a younger age (see Table 9.5).
Table 9.5
Number of Children per Family by Race, 1972 and 1988 (percentages)
1972 |
1988 |
|||
Number of children |
White |
Black |
White |
Black |
0 |
60.2 |
35.6 |
17.0 |
10.4 |
1 |
21.1 |
24.4 |
17.4 |
15.6 |
2 |
13.4 |
12.6 |
33.6 |
15.6 |
3 |
4.6 |
18.5 |
21.2 |
23.0 |
4 |
0.8 |
3.7 |
5.8 |
17.0 |
5 |
— |
3.7 |
3.5 |
8.9 |
6 |
— |
0.7 |
0.8 |
4.4 |
7 |
— |
0.7 |
0.8 |
3.0 |
8 |
0.7 |
|||
9 |
0.7 |
|||
10 |
0.7 |
|||
Mean |
0.65 |
1.48* |
1.99 |
2.96* |
* Difference significant at the .01 level, two-tailed test.
Although there was no significant difference in rank between women with and without children in 1972, the presence of children in 1972 was associated with lower mobility for both groups. By 1988, the white women who had children in 1972 had achieved a rank of 4.5, compared to 5.1 for those who had not had children in 1972. Black women who were parents in 1972 achieved a mean rank of 3.7 in 1988, compared to 4.2 for women who had not been parents.10
The black women in our sample were apparently much more likely to be single parents: only 56.3 percent of the black mothers, compared to 83.7 percent of the white mothers, reported being married in 1972. Single parenting can make it very difficult for young women to gain the training and experience that could help them advance in the labor force; indeed, the small group of white women who were single mothers in 1972 were at a significantly lower rank than other white women in both 1972 and 1988. However, the considerably larger group of black 1972 single mothers were not lower in rank than other black service workers in 1972, and only slightly, albeit significantly, lower in rank in 1988.11 This difference may reflect greater cultural familiarity with combining work and single parenting among black women, and perhaps more acceptance and assistance from other family members; but it also reflects the poor mobility of the black women as a whole, whether or not they were single mothers. In addition, as we have seen, most of the black women were mothers, whether single or married, and the presence of children was associated with significantly lower mobility for both white and black women.
Table 9.6
Educational Attainment by Race, 1972 and 1988 (percentages)
1972 |
1988 |
|||
White |
Black |
White |
Black |
|
Mean educational attainment (years) |
12.0 |
10.8* |
13.0 |
11.9* |
Percentage with less than 12 years |
21.5 |
53.6* |
7.3 |
24.4* |
Percentage attending school in 1972 |
23.7 |
15.6** |
||
* Difference significant at the .01 level, two-tailed tests. |
||||
** Difference significant at the .05 level, two-tailed tests. |
||||
5. Education
Many service occupations have low educational requirements; few require more than a high school diploma, and for many even this is not necessary. We might expect, therefore, that both the white and black women in our sample would have relatively low levels of education. On the other hand, service work can be used as temporary employment by women preparing for more lucrative occupations (e.g., college students working as waitresses or babysitters during the year or in the summer); this group could be working in low-ranked service occupations despite relatively high levels of education. In addition, they may be more likely to be attending school at the same time that they are working as service workers. And they are likely to be more successful at leaving service work and achieving a higher occupational rank by 1988.
The educational variables show large and striking differences between the black and white women, with black women substantially lower in educational attainment at both the beginning and the end of the period. The black women were also significantly less likely to be attending school at the time of the 1972 interview (see Table 9.6).
Table 9.7
Rank in 1972 and 1988 by School Attendance (percentages)
Attending school |
Not attending school |
|
White |
||
Rank in 1972 |
2.53 |
2.82*** |
Rank in 1988 |
6.19 |
4.43* |
Black |
||
Rank in 1972 |
2.86 |
2.27** |
Rank in 1988 |
4.71 |
3.76** |
* Difference significant at the .01 level.
** Difference significant at the .05 level.
*** Difference significant at the .10 level; all two-tailed tests.
White women were apparently more likely to be using the flexibility of service work to support further education, a strategy that seems to have paid off extremely well in terms of mobility. In fact, white women in school in 1972 held service jobs of significantly lower rank than white women not in school. The opposite was the case for black women. Black women also benefited from school attendance, but their upward mobility was considerably lower (see Table 9.7). This lower mobility may reflect in part the lower starting level of education for the black women: 38.1 percent of the black women attending school in 1972 were in grade 12 or lower, compared to only 3.2 percent of the white women.12
6. Class Differences
These observations about education may suggest underlying class differences between the black and white women (and, as we will discuss below, within each racial group). There are considerable disagreements among theorists about how to measure class position (and further disagreements about how to measure class position for women—cf. Power 1984). For the purposes of this project, our interest is in the differing access to economic resources. Because the women were still young in 1972, we will measure these resources by the resources of their families of origin, as measured by their fathers’ occupations at the time of their first interviews in 1968. Mothers’ occupations are obviously also relevant,13 and the discussion will refer to them as well, but almost half the white women and more than half of the black women did not report an occupation for their mothers.
Table 9.8
Father’s Occupation in 1968 for Women in Service Occupations in 1972 (percentages)
Occupation |
White |
Black |
5.0 |
17.0 |
|
Professional, technical, and kindred |
12.3 |
2.21 |
Managers, officials, and proprietors |
8.8 |
0 |
Clerical and kindred |
6.9 |
1.5 |
Sales workers |
3.4 |
0.7 |
Craftsmen, foremen, and kindred |
20.7 |
6.7 |
Operatives and kindred |
14.6 |
19.3 |
Private household workers |
1.1 |
7.4 |
Other service workers |
3.8 |
9.6 |
Farmers and farm managers |
14.6 |
14.1 |
Farm laborers |
1.5 |
5.2 |
Laborers |
6.5 |
16.3 |
Armed forces |
0.8 |
0 |
Total |
100.0 |
100.0 |
There are striking occupational differences between the black and white fathers. Considerably more black than white respondents reported no occupation for their fathers, either because no father was present or because fathers were not working. Black fathers for whom occupations were reported were most likely to be factory operatives, service workers (including a substantial proportion in domestic work), laborers, and farmers and farm managers. Only three black fathers were reported to be in professional or technical occupations, and none were in managerial occupations. White fathers were likely to work in professional/technical or managerial occupations, in the crafts as operatives, or as farmers or farm managers (see Table 9.8).
Only 41.5 percent of the black women reported occupations for their mothers; but among mothers who were reported to be working, 82.1 percent were in service occupations, including 32.1 percent in domestic work. Among white women, 54 percent reported mothers’ occupations; of these, 31.2 percent were in service occupations (including 7.8 percent in domestic service), 20.6 percent were clerical workers, and 15.6 percent were in professional/technical occupations.
This information about parents’ occupations makes it clear that the black women came from families with fewer economic resources. Black parents were clustered in lower-paid occupational categories at a time in their daughters’ lives (ages 14 to 24) when money to pay for college or further training would be particularly important. This, of course, is not a surprising finding, given a long history of discrimination in access to education and occupational segregation by race. But it does help explain why a substantial number of the white respondents were able to use service work as a temporary stopping place, while the majority of the black respondents became mired down.
Domestics and Waitresses
We have seen that for a substantial proportion of the white 1972 service workers, service work was a temporary occupation on the way to better-paid employment. This employment pattern was available to many fewer of the black women in our sample. The difference in ability to use service work as a stepping stone may begin with access to a different set of service occupations; the white and black women in our sample differed markedly in the service occupations they held. One-fourth of the white 1972 service workers were waitresses, with hospital attendants a distant second at 14.2 percent, followed by hairdressers and cosmetologists and private babysitters. One-fourth of the black women were hospital attendants (the only occupation with substantial representation from both races), closely followed by domestic workers not elsewhere classified; no other occupation employed as many as 10 percent of the black women (see Table 9.9).
Both babysitting and waitressing were low-ranked occupations: babysitting had a rank of 1, while waitressing had a rank of 2. But both occupations have the flexibility in hours that permits them to be combined with schooling or training to achieve upward mobility. This flexibility seemed to be a more important determinant of mobility for white women than the rank of the occupation: while black women with 1972 ranks of 1 or 2 were in significantly lower-ranked occupations in 1988 than black women who had started in higher-ranked service occupations, there was no difference in 1988 rank for white women from low-versus high-ranked service occupations.
1. Domestic Work
A number of authors have observed that domestic work may serve as a “bridging” occupation for white women, a transitional job on the way to better-paid work, while women of color are more likely to become ghettoized and unable to leave the occupation (cf. Glenn 1981; Rollins 1985; Romero 1988; 1992). Our data seem to confirm these observations. While the majority of the black 1972 domestic workers in our sample were able to move out of housework, few escaped from low-ranked service occupations. The white 1972 domestic workers had a much more positive experience.
Table 9.9
Top Four Service Occupations by Race, 1972
White |
Black |
||
Occupation |
Percent |
Occupation |
Percent |
Waitresses |
23.7 |
Hospital attendants |
23.0 |
Hospital attendants |
14.2 |
Private household, not elsewhere classified |
21.5 |
Hairdressers and cosmetologists |
12.6 |
Chambermaids |
8.1 |
Private babysitters |
11.9 |
Waitresses |
8.1 |
Total |
62.4 |
Total |
60.7 |
1972 |
1988 |
|||
White |
Black |
White |
Black |
|
Attending school in 1972 |
29.5 |
7.9* |
||
Educational attainment (years) |
12.3 |
9.9* |
13.5 |
10.2* |
Percentage with children in 1972 |
27.3 |
71.0* |
||
Mean rank in 1988 |
5.14 |
3.24* |
||
*Difference significant at the .01 level, two-tailed test. |
||||
For both groups of women, recent arrivals in the labor force in 1972, private household work was a frequent occupation: 16.9 percent of the white women and 28.1 percent of the black women worked in private households. However, as we have seen, most of the white women worked as babysitters, while most of the black women were in the category “private household, not elsewhere classified,” which includes all domestic workers except babysitters, housekeepers, and laundresses (two white women in our sample were housekeepers in 1972; there were no laundresses).
The white women in domestic work clearly fit a different pattern from the black women. In fact, they were among the most highly educated and upwardly mobile of the white women in our sample, while the black domestics had educational and mobility averages considerably below the mean for the black sample as a whole. The white women were also far less likely than the black women to be balancing work and family responsibilities (see Table 9.10).
Only 20.4 percent of the white domestics were still service workers in 1988 (and only one of those, or 2.3 percent of the sample, was still in domestic work), and 27.3 percent had become professional/technical workers. By contrast, only one of the black women (2.6 percent of the sample) had entered a professional/technical occupation, while 60.5 percent had remained in service work, including 15.8 percent still working as domestics.
2. Waitresses
Waitressing illustrates some of the differences in the experience of black and white women in service work, and also illuminates some of the differences within the sample of white service workers. The most striking observation is that this category, by far the most common occupation for white service workers in our sample, employing one-quarter of the white women, employed only eleven black women (8.1 percent of our sample). That black women were so much less likely to work as waitresses may be at least in part the result of a familiar form of racial discrimination, in which people of color are excluded from jobs involving direct contact with customers.14 This exclusion is significant because waitressing may offer the flexibility of hours that enables it to be combined with education or training to achieve occupational mobility. It can also combine with child-rearing for the same reasons, a possibility that apparently attracted many of the white women in our sample. In our discussion, we will compare white and black waitresses on some variables, but it is important to note that the small number of black waitresses makes such comparisons problematic.
The white waitresses as a group did not vary significantly from the rest of the white 1972 service workers in the percentage in school in 1972, presence of children in 1972, educational degree, or rank attained in 1988; they had a slightly lower level of education in 1972, at 11.6 years compared to 12.2 for the rest of the white service workers.15 These averages, however, conceal important distinctions among the white women. When we divide the sixty-two white waitresses in our sample by whether or not they had children in 1972, striking differences appear, suggesting the existence of class differences among white waitresses that may be both determined by and reflected in the presence of children. The thirty-three white waitresses without children began and ended our survey period with higher levels of education than the twenty-nine women with children, moving from a mean of 12.5 years in 1972 to an impressive 14.3 years in 1988, compared to 10.8 and 11.3 respectively for the women with children.16 Half of those without children, and none of the women with children, were attending school in 1972; and by 1988 those without children in 1972 had achieved a considerably higher occupational rank, at 5.1 compared to 4.1 for the women with children.17 These findings confirm casual observation: waitressing can be a temporary occupation for women who are continuing their education and training; and it can also be a dead end for less educated women with fewer economic resources. The apparent importance of children in this division is perhaps not surprising: the presence of children makes it difficult to further an education and work; in addition, women striving for upward mobility may postpone childbearing.
The small number of black waitresses were quite similar to the white waitresses; they diverged from the total black sample in their lower likelihood of having children (45.5 percent did), and higher level of education in 1988 (12.5 years). While their mean 1988 rank, at 4.12, was lower than white waitresses, it was higher than the black sample as a whole. The clear divergence from the white waitresses comes in the likelihood of combining the work with schooling: only one of the black waitresses was attending school at the time of the 1972 interview. There were none of the dramatic differences observed for the white waitresses between the five black women with and the six without children (perhaps not surprising given the tiny sample).
By 1988 only 24.2 percent of the white waitresses were still service workers, almost half of them still waitresses (a relatively high degree of occupational stability compared to other service occupations). The remainder were most likely to have moved into clerical work (25.8 percent) or professional/technical occupations (19.3 percent). Among the eleven black waitresses, five (45.4 percent) remained in service work, none of them as waitresses; four (36.4 percent) had become clerical workers, one (9.1 percent) had entered the professional/technical category, and one had become a laborer.
Conclusion
In summary, we found that the black women in our sample were considerably more likely to get stuck in service occupations than the white women, or to leave them for other low-ranked occupational categories. White women were often able to use service work, even the lowest-ranked occupations, as a temporary stopping place while preparing themselves for better-paying jobs (mostly in traditionally female clerical and professional occupations). For both white and black women, responsibility for children led to lower mobility; however, considerably fewer of the white women had children when the study began. Finally, the white women were more likely to come from families in which parents worked in higher-paid occupational categories, giving them potentially more access to the resources they needed for upward occupational mobility. For most of the black women, and some of the white women in our sample, service work was a dead end. For most of the white women, it was not.
The 396 women in our sample started out in the same occupational category. They were the same age and were living through the same political/economic era. Yet the outcomes they experienced in terms of economic mobility were clearly affected by race and by class background, complexly intertwined with parenthood and marital status. From a methodological point of view, these findings bring into question the meaningfulness of cross-sectional analysis of occupational distributions. From a theoretical point of view, they confirm the argument that the effects of race and gender oppression are not simply additive—the outcome is a vastly different life experience—and they emphasize the importance of considering class background as part of this analysis as well. The findings also suggest the importance of considering the presence of children, not unidirectionally as an impediment to economic mobility, but as part of the complex picture of women’s differing economic prospects. Politically, the findings emphasize the continuing and substantial differences in access to economic mobility by race, differences that must be addressed through public policy.
Notes
This chapter is reprinted from Marilyn Power and Sam Rosenberg, “Race, Class and Occupational Mobility: Black and White Women in Service Work in the United States,” Feminist Economics 1, no. 3 (fall 1995): 40–59, with permission from Routledge.
1. Cf. Zavella 1991; Zinn and Dill 1994; Andersen and Collins 1992. Specifically with regard to women working in service occupations, see Palmer 1989; Sacks 1988; Rollins 1985; Glenn 1981, 1992; Romero 1988; Cobble 1991.
2. The black women in our sample were one year older, on average, than the white women. Mean age in 1968 was 17.8 years for white women and 18.7 years for black women (difference significant at the 1 percent level, two-tailed test).
3. Black women were oversampled proportionally to the population, to allow for a significant sample size: of the 5,159 young women in the original sample, 1,459, or 28.3 percent, were black. Neither Hispanics nor any other racial-ethnic group than blacks were identified separately; 62 women in the total sample were identified as “other,” and are not included in our study. The National Longitudinal Survey is a valuable, if flawed, tool to use to track the economic experience of black and white women. Some flaws (and frustrations) of this study come from holes in the data: the women were not interviewed every year, and retrospective questions were not consistently asked about the noninterview years. In addition, although the retention rate for the sample is very impressive, a quarter of our sample of service workers was missed for at least one interview year. As a result, while we can be sure what each of the women in our sample was doing in 1972 and 1988, our information about the intervening years is necessarily incomplete.
4. This study focuses on occupational mobility, which is defined as movement among occupations with differing earnings potentials. This is a more narrow focus than that taken by studies of social mobility, which may include cultural and social characteristics (e.g., neighborhood of residence), as well as measures of the total array of resources (e.g., spouse’s income) in a determination of mobility. Nor does this study address intergenerational mobility (cf. Tickamyer and Blee 1990), or the subjective experience of mobility, which may bring stress as well as rewards to women of color who find themselves separated from their communities of origin (cf. Higginbotham and Weber 1992).
5. Since the NLS uses the 1960 census occupational codes, it was first necessary to recode the occupations to the 1970 classification scheme. This was done using the system provided in John Priebe (1972).
6. Having escaped service work for clerical occupations, further, black women may again experience lower mobility than white women (cf. Power and Rosenberg 1993).
7. Difference significant at the 1 percent level, two-tailed test.
8. For a discussion of the differences between exploratory and confirmatory data analysis, see Erikson and Nosanchuk (1992) and Tukey (1977).
9. Cf. Geschwender and Carroll-Seguin (1990). Because respondents were not sampled every year, and because one-quarter of the sample missed at least one interview year, we cannot compare the amount of time the black and white women in our sample spent out of the labor force.
10. Differences for both black and white women significant at the 10 percent level, two-tailed test.
11. There were only seventeen white single parents in 1972; they held a rank of 2.3 in 1972 and 4.1 in 1988, compared to 2.8 in 1972 and 4.9 in 1988 for all other white women. Differences in both years were significant at the 10 percent level, two-tailed test. There were thirty-eight black single parents in 1972; their rank of 2.2 in 1972 was not significantly different from the rank of 2.4 achieved by all other black women. In 1988 black single parents held a rank of 3.6, compared to 4.0 for all other black women; this difference was significant at the 10 percent level, two-tailed test. This finding is consistent with Elaine McCrate’s argument that young black women confined to the secondary labor market may have little incentive to postpone childbearing, since their chances for upward mobility are slight in any case (McCrate 1990).
12. A detailed analysis of relative returns to education is beyond the scope of this paper. However, we found that even the minority of black women who had achieved twelve or more years of education in 1972 experienced less upward mobility than their white counterparts, achieving a 1988 mean rank of 4.36, compared to 5.04 for white women with twelve or more years of education in 1972 (black/white difference significant at the .01 level).
13. Tickamyer and Blee (1990) find mother’s occupation has a more significant effect than father’s occupation in determining a daughter’s occupational mobility.
14. According to data compiled by Dorothy Sue Cobble, waiting tables became a female-dominated occupation in the twentieth century, as white women displaced white and black men; black women were either relegated to the least desirable, lowest paid waitressing work or excluded entirely (Cobble 1991, 27–29). In 1993 the occupational category waiters and waitresses was 80 percent female, but only 4.6 percent black (Employment and Earnings January 1994, 207).
15. Difference significant at the 10 percent level, two-tailed test.
16. Differences significant at the 1 percent level, two-tailed test.
17. Difference in school attendance significant at the 1 percent level, difference in 1988 rank significant at the 5 percent level, two-tailed tests.