Women's Entry into Male-Dominated Work: The Possibilities of Resegregation and Job Loss
Rosabeth Moss Kanter's classic work on gender and organizational behavior (1977) pointed out many of the problems women experience on entering male-dominated jobs. Subsequent research has looked in depth at these issues, which range from overt sexual harassment (MacKinnon 1979; Schneider 1982) to the hostile attitudes of male co-workers (O'Farrell and Harlan 1982) and the exclusion of women from social networks (Reskin and Hartmann 1986, 54–55; Roos and Reskin 1984, 245–246). This evidence suggests that rather than progressing incrementally toward gender balance in the workplace, we may in many cases be seeing the emergence of new divisions of labor within occupations and new areas of job resegregation; as the composition of particular subfields changes from being largely male to largely female, the sex typing of the field begins to shift, and integration may be in name only. Instead of asking about the extent of women's progress, researchers have begun to ask such questions as, "How can the newly integrated entry-level positions be kept from resegregating and becoming new female ghettos?" (O'Farrell and Harlan 1984, 276). In short, women's recent movement into male-dominated occupations, in part the result of affirmative action efforts, does not represent the wholesale progress once hoped for.
It is also important to bear in mind, as I emphasized in Chapter 2, that the extent of this movement has so far been very modest, and sex segregation in the workplace remains largely intact. Reskin and Hartmann (1986, 28–30), for example, point out that of all job growth between 1970 and 1980, that which resulted in increased representation of women in male-dominated occupations accounted for only 6.5 percent of the total growth in female employment. Growth in secretarial jobs alone created more employment than all such nominally integrated jobs combined. Furthermore, many prominent female-dominated jobs actually became more female-
intensive during this period. Kanter and others consider this very low level of integration as itself the cause for many of the difficulties encountered by women crossing gender boundary lines. Such analysts assume that workplace interactions will cease to be problematic as more women enter, the proportion of men and women becomes more balanced, and the category of female "token" or "pioneer" becomes a historical artifact. Unfortunately, recent research into both historical and contemporary cases suggests that this view may be too simple and too optimistic.
Historically, the entry of women into previously male-monopolized fields is interrelated with the introduction of new technology, the reorganization of the labor process and restructuring of the occupation, and a subsequent decrease in status and rewards.[2] Clerical work, which changed from a male- to female-dominated occupation in the early part of this century, is the classic example of this feminization/degradation process: as demand increased and employers routinized the labor process, women entered the new deskilled jobs at substantially lower rates of pay (Braverman 1974; Davies 1982; Strom 1987). Teachers, librarians, and bank tellers have followed similar patterns (Garrison 1974; Strober and Arnold 1987; Tyack and Strober 1981). This pattern is so prevalent historically that some argue women's recent progress in the elite male professions signals nothing more than the bifurcation of these fields into female-degraded and male-protected layers (Carter and Carter 1981).
Other researchers who have examined specific contemporary cases of occupational feminization in greater depth confirm a pessimistic interpretation of women's progress. Insurance examiners and adjusters (Phipps 1986; Reskin et al. 1990), computer systems analysts and operations researchers (Donato 1986; Reskin et al. 1990), and printing typesetters and compositors (Roos 1986; Reskin et al. 1990) are occupations with a disproportionately large number of women entrants in the past decade—in fact, in the insurance and printing job categories women have largely replaced men as the predominant sex, and some computer occupations at the bottom of the hierarchy also are increasingly female-dominated. In the
[2] The causal order in this process is complex. Generally, increased demand and a changed labor process either precede or coincide with feminization, with loss in status seeming to follow (Reskin and Roos 1987; Reskin et al. 1990).
process of feminization these jobs have been degraded: automation has increased, job skills have eroded as the work has become more routine and narrow, and status and rewards have dropped (Reskin and Roos 1987; Reskin et al. 1990).
Organizational research focusing on specific firms has also suggested negative consequences from job integration. O'Farrell and Harlan (1982) found that, as a result of affirmative action efforts in a large electrical products firm, a small but significant number of women were hired for blue-collar jobs. However, these women were placed at the lowest level in the blue-collar hierarchy, with newly hired men placed above them. The women's job grouping had the lowest wage ceiling and the fewest highly skilled positions of the six blue-collar job "families." O'Farrell and Harlan (1984, 276–277) conclude that these nontraditional jobs are soon likely to become "traditional," as the women are not likely to move up and are forming a "bottleneck" at the entry level. Moreover, while the wages are better than those previously available to women, they are not likely to increase at the rate of the men's salaries.
My case studies of San Jose and Contra Costa suggest that women workers may be aware of the possibilities of resegregation or "bottlenecking." In San Jose two women suggested to me that the staff analyst classification, a bridge linking the clerical series to the personnel professional series, may have been the victim of this process. Located just below entry-level management, staff analysts are included in the union-represented classes; they handle the routine budgeting and staffing matters in different departments. The women, each in the job for several years, told me that it had changed from male-to female-dominated in the past decade, and now was thought of as a woman's job. They also claimed that the relative wages and status of the job had fallen, an issue discussed among union members.[3] For one of these women, resentment at this reputed loss in status and rewards was the major reason for joining the strike. In fact, according to the Hay job evaluation the pay for the female-dominated staff analyst I position was 31 percent below the average wage rate for its factor point score—one of the larger gaps indicated by the San Jose study (Van Beers 1981). Al-
[3] I could not get independent validation of these claims, as EEO categories used by the city are too broad to trace feminization of specific job classifications.
though this disparity is not evidence of a comparative drop in pay over the past decade, it is consistent with the women's complaints.[4]
Examples of possible resegregation can also be found in Contra Costa County. Union clericals explained to me that the Sheriff's Department had always been an all-male preserve; sheriffs dispatcher had been one of the better-paying, male-dominated jobs that union leaders had encouraged their members to try for. Current figures indicate that this effort may have been all too successful, as the job may now be an example of a new female ghetto: in 1983, 89.2 percent of dispatchers were female (Finney 1983). It remains to be seen whether a comparative drop in wages or truncated career ladder will follow. The job was included in the category of female jobs covered by comparable worth, although it still appears to be somewhat better paid than clerical jobs (Contra Costa County 1985a).[5]
Although the comparable worth movement in San Jose helped women staff analysts reverse some of the reputed fall in relative status and earnings, and in Contra Costa it may protect women dispatchers from facing future losses, there may be other situations of "fallout" from affirmative action efforts in which comparable worth is of less help. In some instances women may have moved up only to dead-end in jobs that are classified as "managerial," and therefore ineligible for union representation and comparable worth raises.[6] Further, these women may come to identify themselves as managers who do not need representation of their interests. It may
[4] This example of possible resegregation also reflects national trends: aggregate data for the personnel Held show a disproportionately large increase in women between 1970 and 1980, from 33.4 to 47 percent female (Reskin and Roos 1987, table 1). Another possible case of resegregation in San Jose may emerge in the entry-level engineering technician position, a job women have just begun to enter. By 1984 the tech series was approximately 23 percent female (extrapolated from EEO categories listed by department in City of San Jose 1984). Whether a new female ghetto develops at-the entry level will be revealed only by future research, but reported hostility of managers and rigged exams indicate a potential threat of bottlenecking. Also, whether wages will fall behind remains to be seen. At present, the entry-level technical jobs remain a better paying alternative for the small proportion of women that can be absorbed.
[5] The pattern of high female entry into this field between 1970 and 1980 reflects national trends. Dispatchers increased from 14.6 to 31.5 percent female—as in personnel, a large, disproportionate gain (Reskin and Roos 1987, table 1).
[6] In Contra Costa County managers in female-dominated job series that are deemed traditional areas of women's work were included in comparable worth ad-justments, but this seems to be the exception rather than the rule. San Jose, following the more typical pattern, did not include managerial employees in any comparable worth adjustments.
even be that some women managers have been subject to job title inflation rather than genuine upward mobility, particularly if organizations need to create the appearance of increased integration (within broad EEOC categories). Baran (1987, 52) found evidence of this in one large insurance firm: the majority of female managers, promoted largely because of affirmative action pressures, actually earned less than white male clericals. The nominal promotion of some women out of union ranks also can undermine organizing efforts to promote women's interests, an effect that may likewise serve the interests of the organization. Regardless of the extent to which they become new female ghettos, low-level managerial positions, particularly within male-dominated categories, are not likely to be covered by comparable worth strategies negotiated through collective bargaining.[7]
Another occupation in which the pattern of feminization accompanied by degradation may be developing is accounting. Nationally, the number of women in accounting has increased; between 1970 and 1980, the field went from 24.6 percent female to 38.1 percent female (Reskin and Roos 1987, table 1i). In both San Jose and Contra Costa, accounting was a male-dominated area into which some low-paid women had succeeded in gaining entry; indeed, in Contra Costa this mobility was considered the main affirmative action victory of clericals. The Contra Costa Comparable Worth Task Force estimated that the female accounting technician position—the bridge between the clerical and accounting series—was paid at a rate 40 percent below that for comparable male-dominated jobs (Contra Costa County 1985a). This provides indirect support for the suggestion that wages may have fallen behind with the feminization of low-level accounting positions. At the bridge level the position still has union representation, so comparable worth adjust-
[7] An interesting counterexample may be a small group of female eligibility work supervisors in Contra Costa County. Although considered managerial-level workers, they had unionized over the comparable worth issue and were staunch members of the Comparable Worth Coalition. The county had punished them for unionizing by revoking some managerial benefits, but their identification remained as undervalued, pink-collar workers, in greater sympathy with the female-dominated group they supervised than with other county managers.
ments can help to narrow this gap. At the next step, however, the entry level to the professional accounting series, one is brought into a traditionally male-dominated and "managerial" category; here any lagging of rewards associated with feminization will be more difficult to address.
The resegregation tendency represents a "one step forward, two steps back" pattern for women of our era in that the hard-fought entry into male-dominated work has created the potential for new female job ghettos. Yet if even in the short run such newly integrated women benefit from higher wage levels, does this not signify at least some degree of incremental progress? It would be easier to evaluate such employment shifts more positively were it not for evidence of additional adverse consequences.
Several studies have found that affirmative action efforts tend to move women into jobs that are extremely vulnerable to elimination. These job cuts may be due to the introduction of new technology or other aspects of organizational restructuring stemming from economic contraction and an increasingly competitive world market (Baran 1987; Ginzburg 1987; S. Hacker 1982; Schreiber 1979). Moreover, some better-paying male-typed jobs appear to be far more vulnerable to recessionary cuts and layoffs than traditional clerical or service jobs—jobs in which many newly integrated women might otherwise have been employed (Baran 1987; Hunt and Hunt 1987; Reskin and Hartmann 1986, 14). Therefore, in return for a short-term gain in wages, some women may have to face the threat of displacement.
Affirmative action, moreover, can be used by management to obscure the displacement of workers and loss of jobs entailed in various strategies of organizational restructuring. Both S. Hacker (1982) and Schreiber (1979), for example, found that as a result of affirmative action women were moved into skilled craft positions that new technology was soon to eliminate. The firms they studied seemed to have manipulated affirmative action implementation to prevent employee resistance in situations where layoffs were already planned. In other blue-collar industries such as auto (Foner 1980, 522) and mining (Cowan 1987), newly integrated women with the least seniority have been the first fired when firms faced contraction. Some industries that were major targets of affirmative action efforts have since faced such severe contraction that they are
now virtually in crisis. Deaux and Ullman studied women's entrance into the steel industry following the signing of affirmative action consent decrees in 1974; they found that progress made in the five years following the settlement had been "totally wiped out by recessionary effects" by the early 1980s (1983, 106).
The use of affirmative action to help rationalize internal labor markets need not, however, indicate managements conspiratorial plans to target women; perhaps affirmative action, in a time of generally increased competition, must function in a firm's plans to streamline its work force if the policy is to be incorporated at all. In white-collar industries, the fact that many firms today are poised for contraction rather than expansion means that internal labor markets only recently opened to women have been dramatically weakened (Baran 1987; Appelbaum 1987; Noyelle 1987). Many lower- and mid-level managerial positions in which women had made large gains—in such industries as insurance, banking, and financial services—are the first to be targeted in organizational restructuring and firms' attempts to become "lean and mean" (Smith 1988). Moreover, with the introduction of office automation apparently it is the middle-level white-collar positions that become redundant (in contrast to the once-predicted massive elimination of clerical jobs); the information-gathering and synthesizing functions of these mid to lower managerial levels can be taken over by cheaper skilled clericals, for whom demand is expected to continue (Appelbaum 1987; Baran 1987; Deborah Bell 1985). If, on balance, the results of affirmative action and women's entry into betterpaying male jobs in fact do fit Hacker's description of the process at AT&T (she wrote: "While some women moved up, more moved out"; 1982, 250). it may be less risky, particularly in the current economic climate, for women to remain in pink-collar work.[8]