Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
I~ Concluding Remarks FEtANCINE D. BLAU Employment segregation by sex is one of the most persistent and pervasive charac- teristics of the labor market. Ike papers in this volume help us to understand the di- mensions of such segregation, its funda- mental causes, ant] its consequences for women's economic status. They also shed some light on the effectiveness of policy in- terventions. In Chapter 2, Andrea Beller gives us the heartening news that, on the basis of anal- ysis of Current Population Survey data, the tendency of men and women to be segre- gatec! by occupation declined noticeably over the 1970s, particularly among younger women. Moreover, this decline was at a rate nearly three times that of the extremely small decline that occurred during the 1960s. We must remember, however, that the mag- nitude of segregation remains high. Fur- thermore, although women increased their share of many traditionally male managerial and professional occupations during the 1970s, this was not the case for heavily male craft and operative jobs. In Chapter 3 Bielby and Baron demon- strate that estimates of the magnitude of sex segregation in employment are extremely 310 sensitive to the level of aggregation of the data. Even detailed (three-digit) census oc- cupational categories group some typically male and some typically female jobs into apparently integrated categories. To the ex- tent that men and women workers in the same occupational categories are segregated by firm, aggregation across firms will result in an underestimate of the magnitude of sex segregation in employment. Using estab- lishment-leve! data and defining segregation in terms of the employer's own job cate- gories, Bielby and Baron find a striking pat- tern of complete sex segregation by occu- pation at the establishment level for a majority of firms (including some single-sex enter- prises) and an extremely high level of seg- regation for the others. Their work raises the question of whether estimates of trends in the magnitude of occupational segrega- tion based on such microdata might differ from Beller's estimates on the basis of ag- gregate data. Less than a quarter ofthe firms in Bielby and Baron's longituclinal sample experienced any declines in the degree of occupational sex segregation during the late 1960s and early 1970s. These are the contours of occupational
CONCLUDING REMARKS 311 segregation by sex in the U. S. labor market. In assessing the significance of occupational segregation, we must know more about its causes ant! consequences. From a policy perspective, an understanding of the con- sequences is crucial for assessing how im- portant a problem it is, while an analysis of its causes helps us determine the most ef- fective tools for attacking it. The studies I reviewed in Chapter 7 sug- gest that occupational segregation does re- duce the earnings of women, although there are considerable problems in precisely es- timating the magnitude of this effect. A par- ticular problem is that overcrowding in the female sector may adversely affect women's wages in the male sector as well. This im- parts a downward bias to estimates of the wage consequences of segregation based on comparisons of women's wages in female and male occupations. It is also important to point out that the focus on earnings does not take into account the possibly negative nonpe- cuniary consequences of sex segregation in employment for women. Intuitively one feels that occupational segregation reinforces cul- tural notions that men and women differ fundamentally in capabilities, preferences, and social and economic roles. Our appraisal of the seriousness of the problem of occupational segregation may depend! in part on the ease with which women are able to move between the male and fe- male sectors. The studies in Chapters 4 and 10 do find some movement of women from predominantly female to predominantly male jobs and vice versa. In Chapter 4, Rosenfeld found that such changes in the sex labels of occupations were experienced by about one- third of women job changers over a one- year period, while in Chapter 10 Corcoran, Duncan, and Ponza found that about one- third of all employed women made such shifts over a five-year period. These findings sug- gest that the male and female sectors ought not to be viewed as watertight compart- ments, but they do not in my opinion greatly mitigate our concern over the issue of oc- cupational sex segregation. First, the mag- nitude of movement is fairly small that is, most women workers did not change the sex label of their jobs over the periods analyzed. Moreover, it is difficult to judge the mag- nitude of the probability of such moves in the abstract. We need} to know more about the desire for such moves on the part of workers and how easy it is for women rel- ative to men to move out of (generally Tower- paying) female jobs into (generally higher- paying) male jobs. It is instructive in this regard that both studies find the probability of a man moving from a predominantly fe- male to a predominantly male job to be con- siderably higher than the probability of a woman making such a move. Furthermore, women whose previous jobs were predom- inantly male were much more likely to change the sex label of their jobs than women whose previous jobs were predominantly female. Second, Bielby and Baron's findings raise the question of how much of what appear to be shifts in the sex labels of jobs as mea- sured by aggregate data actually entails changes in the sex label of the individual's job at the establishment level. The causes of occupational segregation are often classified in terms of supply- versus demand-side factors. The major supply-side explanations considered in this volume are sex-role socialization and the human capital model. In Chapter 11, Marini and Brinton describe how the socialization process influ- ences the occupational orientation of men and women as well as the role they see mar- ket work playing in their lives. It is the latter difference between men and women that is emphasized by the human capital explana- tion critiqued by Corcoran et al. According to this view, because women anticipate shorter and less continuous work lives than men, it wit! be in their economic self-inter- est to choose female occupations, which re- quire smaller human capital investments and have lower wage penalties for time spent
312 FRANCINE D. BLAU out of the labor force. Their own research and that of others summarized by Corcoran et al. does not tend to support the human capital model. In particular, women with discontinuous work careers appear to be no more likely to work at female jobs than women with more continuous work expe- rience. Nor did their findings suggest that the selection of female jobs is consistent with a strategy of maximizing lifetime income, given shorter, more disrupted work lives. On the demand side, I point out in Chap- ter 7 that employers may be motivated to exclude women from particular jobs because of the belief that they would be less stable or productive workers than men. Even if not initially correct, such views can become self-fulfi~ling prophesies if women are then given fewer incentives than men to become stable, productive workers. Boos and Reskin in Chapter 13 emphasize institutional fac- tors in their review of a variety of barriers to female employment in traditionally male jobs at four points in the job allocation pro- cess: pre-employment training, access anal assignment to jobs, mobility, and retention. The operation ofthese barriers does not rely on conscious, overt discriminatory acts on the part of employers. Rather, the everyday operation of the system works against female employment in traditionally male jobs. A clearer unclerstanding of the functions these institutional mechanisms serve is of long- run importance in devising effective strat- egies to remove these obstacles to women's advancement. While the papers in this volume do not endeavor to fully evaluate the impact offed- eral government intervention in this area, an interesting view emerges of what such an evaluation would entail. A variety of ways in which the government potentially influ- ences women's economic status are identi- fied. Some of these effects are positive and some are negative. An overall assessment of the government's impact would necessitate identifying the net effect of all its many pol- icies and programs. This point may be il- lustrated by a discussion of the impact of a few government policies. On one hand, as Boos and Reskin point out, governmental income tax and social security policy tend to discourage female labor force participa- tion. To the extent that the human capital explanation has merit, the resulting de- crease in women's work lives would increase the likelihood of their entering traditionally female jobs. Furthermore, while govern- ment training programs provide an oppor- tunity for intervention to reduce segregation by training individuals for sex-atypical jobs, Waite and Berryman's research suggests in Chapter 16 that occupational training under the Comprehensive Employment and Training Act (CETA) is sex segregated to a great extent. On the other hand, of course, the gov- ernment administers an impressive array of antidiscrimination legislation and regula- tions. There is some debate over the effec- tiveness of these activities, and it is un- doubte~y difficult to measure governmental impact in this area. But it is significant that O'Farrell and Harlan report in Chapter 15 that governmental pressure was an impor- tant factor promoting change within some firms. They also find that the employment growth of firms provides opportunities for integration and thus also facilitates change. The government's macroeconomic policies and their impact on overall business con- ditions are therefore another way in which government may have an effect on the em- ployment opportunities of women the quality of jobs obtained as well as the prob- ability of finding a job at all. While all movements toward occupational integration should be welcome, it is impor- tant to realize that the movement of women into male jobs does not always bring women significantly closer to economic parity with men. For one thing, occupational sex seg- regation may be replaced by female en- cIaves at the lower levels of male job lad- ders a process O'Farrell and Harlan term resegregation. For another, O'Farrell and
CONCLUDING REMARKS 313 Harlan note that women may gain access to male jobs just as they are becoming tech- nologicallyobsolete. Indeed, Strober argues in Chapter 8 that, in general, women are restricted to the leftovers, the jobs that men do not want. Sex segregation in employment remains a pervasive feature of the labor market and a major cause of women's lower earnings. While sex differences in socialization and the voluntary choices that women make in their selection of jobs may play a role in producing sex segregation in employment, labor market discrimination is undoubtedly also a major factor. Such discrimination is deeply entrenched: Within the workplace, a myriad of institutional mechanisms work to perpetuate segregation of work along sex lines without requiring overt, conscious acts of discrimination on the part of employers. While some government policies work to reduce sex segregation in employment, others actually help perpetuate it. Thus, change will not be easy yet such change is essen- tial if we are to move substantially closer to the goal of economic parity between women and men in the labor market.