Telling stories about women and work: Judicial interpretations of sex segregation in the workplace in title VII cases raising the lack of interest argument

Harvard Law Review

Volumn 103, Issue 8, 1990, Pages 1750-1839

  Schultz, Vicki ^a  

^a UNIVERSITY OF WISCONSIN   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 11944260248     PISSN: 0017811X     EISSN: None     Source Type: Journal    
DOI: 10.2307/1341317     Document Type: Article

References (444)

1. Although the degree of sex segregation declined modestly during the 1970'S, work remains highly segregated by sex. Throughout the 1980'S, for example, roughly 60% of all men and women workers would have been required to switch to occupations atypical for their sex to achieve sex integrated occupations. See, e.g., J. JACOBS, REVOLVING DOORS: SEX SEGREGATION AND WOMEN'S CAREERS 20, 28-29 (1989); Beller, Trends in Occupational Segregation by Sex and Race 1960-1981, in SEX SEGREGATION IN THE WORKPLACE: TRENDS, EXPLANATIONS, REMEDIES 11 (B. Reskin ed. 1984) [hereinafter SEX SEGREGATION IN THE WORKPLACE]. As recently as 1985, over two-thirds of working women were employed in occupations in which at least 70% of the workers were female. See Jacobs, Long-Term Trends in Occupational Segregation by Sex, 95 AM. J. Soc. 160, 160 (1989). These estimates of occupational segregation understate the degree of sex segregation, because even workers employed in apparently sex-neutral occupations often work in industries, firms, departments, and jobs that are highly segregated by sex. See, e.g., Bielby & Baron, A Woman's Place Is with Other Women: Sex Segregation Within Organizations, in SEX SEGREGATION IN THE WORKPLACE, supra, at 27, 35 (finding that in a random sample of 393 California firms, 90% of the workers were in job titles to which only one sex was assigned); Gutek & Morasch, Sex-Ratios, Sex-Role Spillover, and Sexual Harassment of Women at Work, J. Soc. ISSUES, Winter 1982, at 55, 61-62 (finding that in a representative sample of 1232 Los Angeles workers, 42% of the women in male-dominated occupations were nonetheless in female-dominated jobs). For general discussions of sex segregation at work, see B. BERGMANN, THE ECONOMIC EMERGENCE OF WOMEN (1986); P. ENGLAND & G. FARKAS, HOUSEHOLDS, EMPLOYMENT, AND GENDER: A SOCIAL, ECONOMIC, AND DEMOGRAPHIC VIEW 121-96 (1986); J. JACOBS, supra; WOMEN'S WORK, MEN'S WORK: SEX SEGREGATION ON THE JOB (B. Reskin & H. Hartmann eds. 1986) [hereinafter WOMEN'S WORK, MEN'S WORK].
2. Although the degree of sex segregation declined modestly during the 1970'S, work remains highly segregated by sex. Throughout the 1980'S, for example, roughly 60% of all men and women workers would have been required to switch to occupations atypical for their sex to achieve sex integrated occupations. See, e.g., J. JACOBS, REVOLVING DOORS: SEX SEGREGATION AND WOMEN'S CAREERS 20, 28-29 (1989); Beller, Trends in Occupational Segregation by Sex and Race 1960-1981, in SEX SEGREGATION IN THE WORKPLACE: TRENDS, EXPLANATIONS, REMEDIES 11 (B. Reskin ed. 1984) [hereinafter SEX SEGREGATION IN THE WORKPLACE]. As recently as 1985, over two-thirds of working women were employed in occupations in which at least 70% of the workers were female. See Jacobs, Long-Term Trends in Occupational Segregation by Sex, 95 AM. J. Soc. 160, 160 (1989). These estimates of occupational segregation understate the degree of sex segregation, because even workers employed in apparently sex-neutral occupations often work in industries, firms, departments, and jobs that are highly segregated by sex. See, e.g., Bielby & Baron, A Woman's Place Is with Other Women: Sex Segregation Within Organizations, in SEX SEGREGATION IN THE WORKPLACE, supra, at 27, 35 (finding that in a random sample of 393 California firms, 90% of the workers were in job titles to which only one sex was assigned); Gutek & Morasch, Sex-Ratios, Sex-Role Spillover, and Sexual Harassment of Women at Work, J. Soc. ISSUES, Winter 1982, at 55, 61-62 (finding that in a representative sample of 1232 Los Angeles workers, 42% of the women in male-dominated occupations were nonetheless in female-dominated jobs). For general discussions of sex segregation at work, see B. BERGMANN, THE ECONOMIC EMERGENCE OF WOMEN (1986); P. ENGLAND & G. FARKAS, HOUSEHOLDS, EMPLOYMENT, AND GENDER: A SOCIAL, ECONOMIC, AND DEMOGRAPHIC VIEW 121-96 (1986); J. JACOBS, supra; WOMEN'S WORK, MEN'S WORK: SEX SEGREGATION ON THE JOB (B. Reskin & H. Hartmann eds. 1986) [hereinafter WOMEN'S WORK, MEN'S WORK].
3. Although the degree of sex segregation declined modestly during the 1970'S, work remains highly segregated by sex. Throughout the 1980'S, for example, roughly 60% of all men and women workers would have been required to switch to occupations atypical for their sex to achieve sex integrated occupations. See, e.g., J. JACOBS, REVOLVING DOORS: SEX SEGREGATION AND WOMEN'S CAREERS 20, 28-29 (1989); Beller, Trends in Occupational Segregation by Sex and Race 1960-1981, in SEX SEGREGATION IN THE WORKPLACE: TRENDS, EXPLANATIONS, REMEDIES 11 (B. Reskin ed. 1984) [hereinafter SEX SEGREGATION IN THE WORKPLACE]. As recently as 1985, over two-thirds of working women were employed in occupations in which at least 70% of the workers were female. See Jacobs, Long-Term Trends in Occupational Segregation by Sex, 95 AM. J. Soc. 160, 160 (1989). These estimates of occupational segregation understate the degree of sex segregation, because even workers employed in apparently sex-neutral occupations often work in industries, firms, departments, and jobs that are highly segregated by sex. See, e.g., Bielby & Baron, A Woman's Place Is with Other Women: Sex Segregation Within Organizations, in SEX SEGREGATION IN THE WORKPLACE, supra, at 27, 35 (finding that in a random sample of 393 California firms, 90% of the workers were in job titles to which only one sex was assigned); Gutek & Morasch, Sex-Ratios, Sex-Role Spillover, and Sexual Harassment of Women at Work, J. Soc. ISSUES, Winter 1982, at 55, 61-62 (finding that in a representative sample of 1232 Los Angeles workers, 42% of the women in male-dominated occupations were nonetheless in female-dominated jobs). For general discussions of sex segregation at work, see B. BERGMANN, THE ECONOMIC EMERGENCE OF WOMEN (1986); P. ENGLAND & G. FARKAS, HOUSEHOLDS, EMPLOYMENT, AND GENDER: A SOCIAL, ECONOMIC, AND DEMOGRAPHIC VIEW 121-96 (1986); J. JACOBS, supra; WOMEN'S WORK, MEN'S WORK: SEX SEGREGATION ON THE JOB (B. Reskin & H. Hartmann eds. 1986) [hereinafter WOMEN'S WORK, MEN'S WORK].
4. Although the degree of sex segregation declined modestly during the 1970'S, work remains highly segregated by sex. Throughout the 1980'S, for example, roughly 60% of all men and women workers would have been required to switch to occupations atypical for their sex to achieve sex integrated occupations. See, e.g., J. JACOBS, REVOLVING DOORS: SEX SEGREGATION AND WOMEN'S CAREERS 20, 28-29 (1989); Beller, Trends in Occupational Segregation by Sex and Race 1960-1981, in SEX SEGREGATION IN THE WORKPLACE: TRENDS, EXPLANATIONS, REMEDIES 11 (B. Reskin ed. 1984) [hereinafter SEX SEGREGATION IN THE WORKPLACE]. As recently as 1985, over two-thirds of working women were employed in occupations in which at least 70% of the workers were female. See Jacobs, Long-Term Trends in Occupational Segregation by Sex, 95 AM. J. Soc. 160, 160 (1989). These estimates of occupational segregation understate the degree of sex segregation, because even workers employed in apparently sex-neutral occupations often work in industries, firms, departments, and jobs that are highly segregated by sex. See, e.g., Bielby & Baron, A Woman's Place Is with Other Women: Sex Segregation Within Organizations, in SEX SEGREGATION IN THE WORKPLACE, supra, at 27, 35 (finding that in a random sample of 393 California firms, 90% of the workers were in job titles to which only one sex was assigned); Gutek & Morasch, Sex-Ratios, Sex-Role Spillover, and Sexual Harassment of Women at Work, J. Soc. ISSUES, Winter 1982, at 55, 61-62 (finding that in a representative sample of 1232 Los Angeles workers, 42% of the women in male-dominated occupations were nonetheless in female-dominated jobs). For general discussions of sex segregation at work, see B. BERGMANN, THE ECONOMIC EMERGENCE OF WOMEN (1986); P. ENGLAND & G. FARKAS, HOUSEHOLDS, EMPLOYMENT, AND GENDER: A SOCIAL, ECONOMIC, AND DEMOGRAPHIC VIEW 121-96 (1986); J. JACOBS, supra; WOMEN'S WORK, MEN'S WORK: SEX SEGREGATION ON THE JOB (B. Reskin & H. Hartmann eds. 1986) [hereinafter WOMEN'S WORK, MEN'S WORK].
5. Although the degree of sex segregation declined modestly during the 1970'S, work remains highly segregated by sex. Throughout the 1980'S, for example, roughly 60% of all men and women workers would have been required to switch to occupations atypical for their sex to achieve sex integrated occupations. See, e.g., J. JACOBS, REVOLVING DOORS: SEX SEGREGATION AND WOMEN'S CAREERS 20, 28-29 (1989); Beller, Trends in Occupational Segregation by Sex and Race 1960-1981, in SEX SEGREGATION IN THE WORKPLACE: TRENDS, EXPLANATIONS, REMEDIES 11 (B. Reskin ed. 1984) [hereinafter SEX SEGREGATION IN THE WORKPLACE]. As recently as 1985, over two-thirds of working women were employed in occupations in which at least 70% of the workers were female. See Jacobs, Long-Term Trends in Occupational Segregation by Sex, 95 AM. J. Soc. 160, 160 (1989). These estimates of occupational segregation understate the degree of sex segregation, because even workers employed in apparently sex-neutral occupations often work in industries, firms, departments, and jobs that are highly segregated by sex. See, e.g., Bielby & Baron, A Woman's Place Is with Other Women: Sex Segregation Within Organizations, in SEX SEGREGATION IN THE WORKPLACE, supra, at 27, 35 (finding that in a random sample of 393 California firms, 90% of the workers were in job titles to which only one sex was assigned); Gutek & Morasch, Sex-Ratios, Sex-Role Spillover, and Sexual Harassment of Women at Work, J. Soc. ISSUES, Winter 1982, at 55, 61-62 (finding that in a representative sample of 1232 Los Angeles workers, 42% of the women in male-dominated occupations were nonetheless in female-dominated jobs). For general discussions of sex segregation at work, see B. BERGMANN, THE ECONOMIC EMERGENCE OF WOMEN (1986); P. ENGLAND & G. FARKAS, HOUSEHOLDS, EMPLOYMENT, AND GENDER: A SOCIAL, ECONOMIC, AND DEMOGRAPHIC VIEW 121-96 (1986); J. JACOBS, supra; WOMEN'S WORK, MEN'S WORK: SEX SEGREGATION ON THE JOB (B. Reskin & H. Hartmann eds. 1986) [hereinafter WOMEN'S WORK, MEN'S WORK].
6. Although the degree of sex segregation declined modestly during the 1970'S, work remains highly segregated by sex. Throughout the 1980'S, for example, roughly 60% of all men and women workers would have been required to switch to occupations atypical for their sex to achieve sex integrated occupations. See, e.g., J. JACOBS, REVOLVING DOORS: SEX SEGREGATION AND WOMEN'S CAREERS 20, 28-29 (1989); Beller, Trends in Occupational Segregation by Sex and Race 1960-1981, in SEX SEGREGATION IN THE WORKPLACE: TRENDS, EXPLANATIONS, REMEDIES 11 (B. Reskin ed. 1984) [hereinafter SEX SEGREGATION IN THE WORKPLACE]. As recently as 1985, over two-thirds of working women were employed in occupations in which at least 70% of the workers were female. See Jacobs, Long-Term Trends in Occupational Segregation by Sex, 95 AM. J. Soc. 160, 160 (1989). These estimates of occupational segregation understate the degree of sex segregation, because even workers employed in apparently sex-neutral occupations often work in industries, firms, departments, and jobs that are highly segregated by sex. See, e.g., Bielby & Baron, A Woman's Place Is with Other Women: Sex Segregation Within Organizations, in SEX SEGREGATION IN THE WORKPLACE, supra, at 27, 35 (finding that in a random sample of 393 California firms, 90% of the workers were in job titles to which only one sex was assigned); Gutek & Morasch, Sex-Ratios, Sex-Role Spillover, and Sexual Harassment of Women at Work, J. Soc. ISSUES, Winter 1982, at 55, 61-62 (finding that in a representative sample of 1232 Los Angeles workers, 42% of the women in male-dominated occupations were nonetheless in female-dominated jobs). For general discussions of sex segregation at work, see B. BERGMANN, THE ECONOMIC EMERGENCE OF WOMEN (1986); P. ENGLAND & G. FARKAS, HOUSEHOLDS, EMPLOYMENT, AND GENDER: A SOCIAL, ECONOMIC, AND DEMOGRAPHIC VIEW 121-96 (1986); J. JACOBS, supra; WOMEN'S WORK, MEN'S WORK: SEX SEGREGATION ON THE JOB (B. Reskin & H. Hartmann eds. 1986) [hereinafter WOMEN'S WORK, MEN'S WORK].
7. Although the degree of sex segregation declined modestly during the 1970'S, work remains highly segregated by sex. Throughout the 1980'S, for example, roughly 60% of all men and women workers would have been required to switch to occupations atypical for their sex to achieve sex integrated occupations. See, e.g., J. JACOBS, REVOLVING DOORS: SEX SEGREGATION AND WOMEN'S CAREERS 20, 28-29 (1989); Beller, Trends in Occupational Segregation by Sex and Race 1960-1981, in SEX SEGREGATION IN THE WORKPLACE: TRENDS, EXPLANATIONS, REMEDIES 11 (B. Reskin ed. 1984) [hereinafter SEX SEGREGATION IN THE WORKPLACE]. As recently as 1985, over two-thirds of working women were employed in occupations in which at least 70% of the workers were female. See Jacobs, Long-Term Trends in Occupational Segregation by Sex, 95 AM. J. Soc. 160, 160 (1989). These estimates of occupational segregation understate the degree of sex segregation, because even workers employed in apparently sex-neutral occupations often work in industries, firms, departments, and jobs that are highly segregated by sex. See, e.g., Bielby & Baron, A Woman's Place Is with Other Women: Sex Segregation Within Organizations, in SEX SEGREGATION IN THE WORKPLACE, supra, at 27, 35 (finding that in a random sample of 393 California firms, 90% of the workers were in job titles to which only one sex was assigned); Gutek & Morasch, Sex-Ratios, Sex-Role Spillover, and Sexual Harassment of Women at Work, J. Soc. ISSUES, Winter 1982, at 55, 61-62 (finding that in a representative sample of 1232 Los Angeles workers, 42% of the women in male-dominated occupations were nonetheless in female-dominated jobs). For general discussions of sex segregation at work, see B. BERGMANN, THE ECONOMIC EMERGENCE OF WOMEN (1986); P. ENGLAND & G. FARKAS, HOUSEHOLDS, EMPLOYMENT, AND GENDER: A SOCIAL, ECONOMIC, AND DEMOGRAPHIC VIEW 121-96 (1986); J. JACOBS, supra; WOMEN'S WORK, MEN'S WORK: SEX SEGREGATION ON THE JOB (B. Reskin & H. Hartmann eds. 1986) [hereinafter WOMEN'S WORK, MEN'S WORK].
8. Although the degree of sex segregation declined modestly during the 1970'S, work remains highly segregated by sex. Throughout the 1980'S, for example, roughly 60% of all men and women workers would have been required to switch to occupations atypical for their sex to achieve sex integrated occupations. See, e.g., J. JACOBS, REVOLVING DOORS: SEX SEGREGATION AND WOMEN'S CAREERS 20, 28-29 (1989); Beller, Trends in Occupational Segregation by Sex and Race 1960-1981, in SEX SEGREGATION IN THE WORKPLACE: TRENDS, EXPLANATIONS, REMEDIES 11 (B. Reskin ed. 1984) [hereinafter SEX SEGREGATION IN THE WORKPLACE]. As recently as 1985, over two-thirds of working women were employed in occupations in which at least 70% of the workers were female. See Jacobs, Long-Term Trends in Occupational Segregation by Sex, 95 AM. J. Soc. 160, 160 (1989). These estimates of occupational segregation understate the degree of sex segregation, because even workers employed in apparently sex-neutral occupations often work in industries, firms, departments, and jobs that are highly segregated by sex. See, e.g., Bielby & Baron, A Woman's Place Is with Other Women: Sex Segregation Within Organizations, in SEX SEGREGATION IN THE WORKPLACE, supra, at 27, 35 (finding that in a random sample of 393 California firms, 90% of the workers were in job titles to which only one sex was assigned); Gutek & Morasch, Sex-Ratios, Sex-Role Spillover, and Sexual Harassment of Women at Work, J. Soc. ISSUES, Winter 1982, at 55, 61-62 (finding that in a representative sample of 1232 Los Angeles workers, 42% of the women in male-dominated occupations were nonetheless in female-dominated jobs). For general discussions of sex segregation at work, see B. BERGMANN, THE ECONOMIC EMERGENCE OF WOMEN (1986); P. ENGLAND & G. FARKAS, HOUSEHOLDS, EMPLOYMENT, AND GENDER: A SOCIAL, ECONOMIC, AND DEMOGRAPHIC VIEW 121-96 (1986); J. JACOBS, supra; WOMEN'S WORK, MEN'S WORK: SEX SEGREGATION ON THE JOB (B. Reskin & H. Hartmann eds. 1986) [hereinafter WOMEN'S WORK, MEN'S WORK].
9. The most serious problem associated with segregation is its effect on women's earning power. A substantial portion of the male-female wage disparity is attributable to women's concentration in lower-paying, female-dominated occupations and jobs. See, e.g., BUREAU OF THE CENSUS, U.S. DEP'T OF COMMERCE, CURRENT POPULATION REPORT, HOUSEHOLD ECONOMIC STUDIES, SERIES P-70, No. 10, MALE-FEMALE DIFFERENCES IN WORK EXPERIENCE, OCCUPATION, AND EARNINGS: 1984, at 9-10 (1987); WOMEN, WORK AND WAGES: EQUAL PAY FOR JOBS OF EQUAL VALUE 33-38 (D. Treiman & H. Hartmann eds. 1981). Entry-level female jobs are often on short mobility ladders that offer little or no opportunity for advancement, see infra note 317, and the jobs women do tend to have less prestige than the jobs men do. Even when women work in male-dominated jobs, segregation creates a context in which they are perceived to have significantly less prestige than their male counterparts. See, e.g., Powell & Jacobs, Gender Differences in the Evaluation of Prestige, 25 Soc. Q. 173 (1984). For a discussion of these and other negative consequences of sex segregation for working women, see WOMEN'S WORK, MEN'S WORK, supra note 1, at 9-17.
10. The most serious problem associated with segregation is its effect on women's earning power. A substantial portion of the male-female wage disparity is attributable to women's concentration in lower-paying, female-dominated occupations and jobs. See, e.g., BUREAU OF THE CENSUS, U.S. DEP'T OF COMMERCE, CURRENT POPULATION REPORT, HOUSEHOLD ECONOMIC STUDIES, SERIES P-70, No. 10, MALE-FEMALE DIFFERENCES IN WORK EXPERIENCE, OCCUPATION, AND EARNINGS: 1984, at 9-10 (1987); WOMEN, WORK AND WAGES: EQUAL PAY FOR JOBS OF EQUAL VALUE 33-38 (D. Treiman & H. Hartmann eds. 1981). Entry-level female jobs are often on short mobility ladders that offer little or no opportunity for advancement, see infra note 317, and the jobs women do tend to have less prestige than the jobs men do. Even when women work in male-dominated jobs, segregation creates a context in which they are perceived to have significantly less prestige than their male counterparts. See, e.g., Powell & Jacobs, Gender Differences in the Evaluation of Prestige, 25 Soc. Q. 173 (1984). For a discussion of these and other negative consequences of sex segregation for working women, see WOMEN'S WORK, MEN'S WORK, supra note 1, at 9-17.
11. The most serious problem associated with segregation is its effect on women's earning power. A substantial portion of the male-female wage disparity is attributable to women's concentration in lower-paying, female-dominated occupations and jobs. See, e.g., BUREAU OF THE CENSUS, U.S. DEP'T OF COMMERCE, CURRENT POPULATION REPORT, HOUSEHOLD ECONOMIC STUDIES, SERIES P-70, No. 10, MALE-FEMALE DIFFERENCES IN WORK EXPERIENCE, OCCUPATION, AND EARNINGS: 1984, at 9-10 (1987); WOMEN, WORK AND WAGES: EQUAL PAY FOR JOBS OF EQUAL VALUE 33-38 (D. Treiman & H. Hartmann eds. 1981). Entry-level female jobs are often on short mobility ladders that offer little or no opportunity for advancement, see infra note 317, and the jobs women do tend to have less prestige than the jobs men do. Even when women work in male-dominated jobs, segregation creates a context in which they are perceived to have significantly less prestige than their male counterparts. See, e.g., Powell & Jacobs, Gender Differences in the Evaluation of Prestige, 25 Soc. Q. 173 (1984). For a discussion of these and other negative consequences of sex segregation for working women, see WOMEN'S WORK, MEN'S WORK, supra note 1, at 9-17.
12. The most serious problem associated with segregation is its effect on women's earning power. A substantial portion of the male-female wage disparity is attributable to women's concentration in lower-paying, female-dominated occupations and jobs. See, e.g., BUREAU OF THE CENSUS, U.S. DEP'T OF COMMERCE, CURRENT POPULATION REPORT, HOUSEHOLD ECONOMIC STUDIES, SERIES P-70, No. 10, MALE-FEMALE DIFFERENCES IN WORK EXPERIENCE, OCCUPATION, AND EARNINGS: 1984, at 9-10 (1987); WOMEN, WORK AND WAGES: EQUAL PAY FOR JOBS OF EQUAL VALUE 33-38 (D. Treiman & H. Hartmann eds. 1981). Entry-level female jobs are often on short mobility ladders that offer little or no opportunity for advancement, see infra note 317, and the jobs women do tend to have less prestige than the jobs men do. Even when women work in male-dominated jobs, segregation creates a context in which they are perceived to have significantly less prestige than their male counterparts. See, e.g., Powell & Jacobs, Gender Differences in the Evaluation of Prestige, 25 Soc. Q. 173 (1984). For a discussion of these and other negative consequences of sex segregation for working women, see WOMEN'S WORK, MEN'S WORK, supra note 1, at 9-17.
13. 628 F. Supp. 1264 (N.D. Ill. 1986), aff'd, 839 F.2d 302 (7th Cir. 1988).
14. Title VII is the major federal statute prohibiting discrimination in employment. It prohibits employers with 15 or more employees from discriminating on the basis of race, color, religion, sex, or national origin. See 42 U.S.C. §§ 2000e to 2000e-17 (1982).
15. Between 1973 and 1980 the median hourly wages for first-year commission salesworkers were about twice as high as those for all noncommission salesworkers. See Plaintiff's Pretrial Brief - Commission Sales Issues at 27, EEOC v. Sears, Roebuck & Co., 628 F. Supp. 1264 (N.D. Ill. 1986) (No. 79-C-4373).
16. Between 1973 and 1980, for example, women constituted 61% of all full-time sales applicants at Sears, but only 27% of the newly hired full-time commission salesworkers. In contrast, women made up approximately 75% of Sears' noncommission salesforce. See Brief for the Equal Employment Opportunity Comm'n as Appellant at 7, EEOC v. Sears, Roebuck & Co., 839 F.2d 302 (7th Cir. 1988) (Nos. 86-1519 and 86-1621) [hereinafter EEOC Brief].
17. Sears conceded that it had established no objective qualifications for the commission sales job. See Sears, 628 F. Supp. at 1290. Nonetheless, the EEOC performed two types of multiple regression analyses that controlled for any differences between the male and female sales applicants on various characteristics that may have influenced their selection into commission sales, including age, education, job applied for, job type experience, product line experience, and expanded commission sales experience. See id. at 1296-98; EEOC Brief, supra note 6, at 20-26.
18. See Sears, 628 F. Supp. at 1305. The EEOC could not reconstruct a precise pool of applicants who specifically preferred commission sales, because Sears' application form did not provide separate boxes for commission and noncommission sales jobs. See EEOC Brief, supra note 6, at 7-8. The EEOC's regression analyses did control for any preferences the applicants had written in by hand, see id. at 22-23, 128-29, even though fewer than half of those hired for commission sales had written in a preference for that job, see Sears, 628 F. Supp. at 1296 & n.21; EEOC Brief, supra note 6, at 27-28.
19. Sears, 628 F. Supp. at 1324.
20. See id. at 1324-25. The judge reached this conclusion despite evidence that women who worked at Sears needed to maximize their incomes. For example, a 1981 survey of married women employed in noncommission sales jobs showed that 28% had unemployed husbands, 35% had husbands who earned below the $20,260 national median income for men, and an additional 19% had husbands who earned below $25,000 per year. See Written Testimony of Eileen Appelbaum at 18-19 & n.27, EEOC v. Sears, Roebuck & Co., 628 F. Supp. 1264 (N.D. Ill. 1986) (No. 79-C-4373).
21. According to the judge, women shunned the "big ticket," "hard" lines of merchandise, such as home improvements, hardware, and men's clothing, which were more likely to be sold on commission at Sears; they felt more comfortable with the "small ticket," "soft" lines, such as jewelry, cosmetics, and women's clothing, which were sold on a noncommission basis. See Sears, 628 F. Supp. at 1306. In addition, women "disliked the perceived dog-eat-dog competition" and "financial risk" of commission sales, preferring the "security" and "more enjoyable and friendly" nature of noncommission sales. See id. at 1307.
22. The news media reported developments in the case from its inception and speculated about its effect on the future of title VII enforcement. See, e.g., Hunter, U.S. Files Five Suits Charging Sears with Job Bias, N.Y. Times, Oct. 23, 1979, at A1, col. 4; Williams, Despite Class-Action Doubts, EEOC Presses Sears Bias Case, Wash. Post, July 9, 1985, at A1, col. 1; Pear, Changes Weighed in Federal Rules on Discrimination, N.Y. Times, Dec. 3, 1984, at A1, col. 6; Greenhouse, Federal Judge Rules for Sears in Sex Bias Case, N. Y. Times, Feb. 4, 1986, at A21, col. 2; Lewin, Statistics Have Become Suspect in Sex Discrimination Cases, N.Y. Times, Feb. 9, 1986, § 4, at E8. Two prominent women's historians testified on opposite sides of the issue of women's "interest" in nontraditional work, and both sought to vindicate their positions publicly after the suit was resolved. Compare Kessler-Harris, Equal Employment Opportunity Commission v. Sears, Roebuck and Company: A Personal Account, 1986 RADICAL HIST. REV. 57 (defending her testimony for the EEOC) with Rosenberg, What Harms Women in the Workplace, N.Y. Times, Feb. 27, 1986, at A23, col. 1 (defending her testimony for Sears). In addition, numerous scholars have analyzed the Sears case, discussing its significance for the contemporary debate about gender difference in our society. See Finley, Choice and Freedom: Elusive Issues in the Search for Gender Justice, 96 YALE L.J. 914, 937-40 (1987); Milkman, Women's History and the Sears Case, 12 FEMINIST STUD. 375 (1986); Scott, Deconstructing Equality-Versus-Difference: Or, the Uses of Poststructuralist Theory for Feminism, 14 FEMINIST STUD. 33, 38-47 (1988); Williams, Deconstructing Gender, 87 MICH. L. REV. 797, 814 (1989).
23. The news media reported developments in the case from its inception and speculated about its effect on the future of title VII enforcement. See, e.g., Hunter, U.S. Files Five Suits Charging Sears with Job Bias, N.Y. Times, Oct. 23, 1979, at A1, col. 4; Williams, Despite Class-Action Doubts, EEOC Presses Sears Bias Case, Wash. Post, July 9, 1985, at A1, col. 1; Pear, Changes Weighed in Federal Rules on Discrimination, N.Y. Times, Dec. 3, 1984, at A1, col. 6; Greenhouse, Federal Judge Rules for Sears in Sex Bias Case, N. Y. Times, Feb. 4, 1986, at A21, col. 2; Lewin, Statistics Have Become Suspect in Sex Discrimination Cases, N.Y. Times, Feb. 9, 1986, § 4, at E8. Two prominent women's historians testified on opposite sides of the issue of women's "interest" in nontraditional work, and both sought to vindicate their positions publicly after the suit was resolved. Compare Kessler-Harris, Equal Employment Opportunity Commission v. Sears, Roebuck and Company: A Personal Account, 1986 RADICAL HIST. REV. 57 (defending her testimony for the EEOC) with Rosenberg, What Harms Women in the Workplace, N.Y. Times, Feb. 27, 1986, at A23, col. 1 (defending her testimony for Sears). In addition, numerous scholars have analyzed the Sears case, discussing its significance for the contemporary debate about gender difference in our society. See Finley, Choice and Freedom: Elusive Issues in the Search for Gender Justice, 96 YALE L.J. 914, 937-40 (1987); Milkman, Women's History and the Sears Case, 12 FEMINIST STUD. 375 (1986); Scott, Deconstructing Equality-Versus-Difference: Or, the Uses of Poststructuralist Theory for Feminism, 14 FEMINIST STUD. 33, 38-47 (1988); Williams, Deconstructing Gender, 87 MICH. L. REV. 797, 814 (1989).
24. The news media reported developments in the case from its inception and speculated about its effect on the future of title VII enforcement. See, e.g., Hunter, U.S. Files Five Suits Charging Sears with Job Bias, N.Y. Times, Oct. 23, 1979, at A1, col. 4; Williams, Despite Class-Action Doubts, EEOC Presses Sears Bias Case, Wash. Post, July 9, 1985, at A1, col. 1; Pear, Changes Weighed in Federal Rules on Discrimination, N.Y. Times, Dec. 3, 1984, at A1, col. 6; Greenhouse, Federal Judge Rules for Sears in Sex Bias Case, N. Y. Times, Feb. 4, 1986, at A21, col. 2; Lewin, Statistics Have Become Suspect in Sex Discrimination Cases, N.Y. Times, Feb. 9, 1986, § 4, at E8. Two prominent women's historians testified on opposite sides of the issue of women's "interest" in nontraditional work, and both sought to vindicate their positions publicly after the suit was resolved. Compare Kessler-Harris, Equal Employment Opportunity Commission v. Sears, Roebuck and Company: A Personal Account, 1986 RADICAL HIST. REV. 57 (defending her testimony for the EEOC) with Rosenberg, What Harms Women in the Workplace, N.Y. Times, Feb. 27, 1986, at A23, col. 1 (defending her testimony for Sears). In addition, numerous scholars have analyzed the Sears case, discussing its significance for the contemporary debate about gender difference in our society. See Finley, Choice and Freedom: Elusive Issues in the Search for Gender Justice, 96 YALE L.J. 914, 937-40 (1987); Milkman, Women's History and the Sears Case, 12 FEMINIST STUD. 375 (1986); Scott, Deconstructing Equality-Versus-Difference: Or, the Uses of Poststructuralist Theory for Feminism, 14 FEMINIST STUD. 33, 38-47 (1988); Williams, Deconstructing Gender, 87 MICH. L. REV. 797, 814 (1989).
25. The news media reported developments in the case from its inception and speculated about its effect on the future of title VII enforcement. See, e.g., Hunter, U.S. Files Five Suits Charging Sears with Job Bias, N.Y. Times, Oct. 23, 1979, at A1, col. 4; Williams, Despite Class-Action Doubts, EEOC Presses Sears Bias Case, Wash. Post, July 9, 1985, at A1, col. 1; Pear, Changes Weighed in Federal Rules on Discrimination, N.Y. Times, Dec. 3, 1984, at A1, col. 6; Greenhouse, Federal Judge Rules for Sears in Sex Bias Case, N. Y. Times, Feb. 4, 1986, at A21, col. 2; Lewin, Statistics Have Become Suspect in Sex Discrimination Cases, N.Y. Times, Feb. 9, 1986, § 4, at E8. Two prominent women's historians testified on opposite sides of the issue of women's "interest" in nontraditional work, and both sought to vindicate their positions publicly after the suit was resolved. Compare Kessler-Harris, Equal Employment Opportunity Commission v. Sears, Roebuck and Company: A Personal Account, 1986 RADICAL HIST. REV. 57 (defending her testimony for the EEOC) with Rosenberg, What Harms Women in the Workplace, N.Y. Times, Feb. 27, 1986, at A23, col. 1 (defending her testimony for Sears). In addition, numerous scholars have analyzed the Sears case, discussing its significance for the contemporary debate about gender difference in our society. See Finley, Choice and Freedom: Elusive Issues in the Search for Gender Justice, 96 YALE L.J. 914, 937-40 (1987); Milkman, Women's History and the Sears Case, 12 FEMINIST STUD. 375 (1986); Scott, Deconstructing Equality-Versus-Difference: Or, the Uses of Poststructuralist Theory for Feminism, 14 FEMINIST STUD. 33, 38-47 (1988); Williams, Deconstructing Gender, 87 MICH. L. REV. 797, 814 (1989).
26. The news media reported developments in the case from its inception and speculated about its effect on the future of title VII enforcement. See, e.g., Hunter, U.S. Files Five Suits Charging Sears with Job Bias, N.Y. Times, Oct. 23, 1979, at A1, col. 4; Williams, Despite Class-Action Doubts, EEOC Presses Sears Bias Case, Wash. Post, July 9, 1985, at A1, col. 1; Pear, Changes Weighed in Federal Rules on Discrimination, N.Y. Times, Dec. 3, 1984, at A1, col. 6; Greenhouse, Federal Judge Rules for Sears in Sex Bias Case, N. Y. Times, Feb. 4, 1986, at A21, col. 2; Lewin, Statistics Have Become Suspect in Sex Discrimination Cases, N.Y. Times, Feb. 9, 1986, § 4, at E8. Two prominent women's historians testified on opposite sides of the issue of women's "interest" in nontraditional work, and both sought to vindicate their positions publicly after the suit was resolved. Compare Kessler-Harris, Equal Employment Opportunity Commission v. Sears, Roebuck and Company: A Personal Account, 1986 RADICAL HIST. REV. 57 (defending her testimony for the EEOC) with Rosenberg, What Harms Women in the Workplace, N.Y. Times, Feb. 27, 1986, at A23, col. 1 (defending her testimony for Sears). In addition, numerous scholars have analyzed the Sears case, discussing its significance for the contemporary debate about gender difference in our society. See Finley, Choice and Freedom: Elusive Issues in the Search for Gender Justice, 96 YALE L.J. 914, 937-40 (1987); Milkman, Women's History and the Sears Case, 12 FEMINIST STUD. 375 (1986); Scott, Deconstructing Equality-Versus-Difference: Or, the Uses of Poststructuralist Theory for Feminism, 14 FEMINIST STUD. 33, 38-47 (1988); Williams, Deconstructing Gender, 87 MICH. L. REV. 797, 814 (1989).
27. The news media reported developments in the case from its inception and speculated about its effect on the future of title VII enforcement. See, e.g., Hunter, U.S. Files Five Suits Charging Sears with Job Bias, N.Y. Times, Oct. 23, 1979, at A1, col. 4; Williams, Despite Class-Action Doubts, EEOC Presses Sears Bias Case, Wash. Post, July 9, 1985, at A1, col. 1; Pear, Changes Weighed in Federal Rules on Discrimination, N.Y. Times, Dec. 3, 1984, at A1, col. 6; Greenhouse, Federal Judge Rules for Sears in Sex Bias Case, N. Y. Times, Feb. 4, 1986, at A21, col. 2; Lewin, Statistics Have Become Suspect in Sex Discrimination Cases, N.Y. Times, Feb. 9, 1986, § 4, at E8. Two prominent women's historians testified on opposite sides of the issue of women's "interest" in nontraditional work, and both sought to vindicate their positions publicly after the suit was resolved. Compare Kessler-Harris, Equal Employment Opportunity Commission v. Sears, Roebuck and Company: A Personal Account, 1986 RADICAL HIST. REV. 57 (defending her testimony for the EEOC) with Rosenberg, What Harms Women in the Workplace, N.Y. Times, Feb. 27, 1986, at A23, col. 1 (defending her testimony for Sears). In addition, numerous scholars have analyzed the Sears case, discussing its significance for the contemporary debate about gender difference in our society. See Finley, Choice and Freedom: Elusive Issues in the Search for Gender Justice, 96 YALE L.J. 914, 937-40 (1987); Milkman, Women's History and the Sears Case, 12 FEMINIST STUD. 375 (1986); Scott, Deconstructing Equality-Versus-Difference: Or, the Uses of Poststructuralist Theory for Feminism, 14 FEMINIST STUD. 33, 38-47 (1988); Williams, Deconstructing Gender, 87 MICH. L. REV. 797, 814 (1989).
28. The news media reported developments in the case from its inception and speculated about its effect on the future of title VII enforcement. See, e.g., Hunter, U.S. Files Five Suits Charging Sears with Job Bias, N.Y. Times, Oct. 23, 1979, at A1, col. 4; Williams, Despite Class-Action Doubts, EEOC Presses Sears Bias Case, Wash. Post, July 9, 1985, at A1, col. 1; Pear, Changes Weighed in Federal Rules on Discrimination, N.Y. Times, Dec. 3, 1984, at A1, col. 6; Greenhouse, Federal Judge Rules for Sears in Sex Bias Case, N. Y. Times, Feb. 4, 1986, at A21, col. 2; Lewin, Statistics Have Become Suspect in Sex Discrimination Cases, N.Y. Times, Feb. 9, 1986, § 4, at E8. Two prominent women's historians testified on opposite sides of the issue of women's "interest" in nontraditional work, and both sought to vindicate their positions publicly after the suit was resolved. Compare Kessler-Harris, Equal Employment Opportunity Commission v. Sears, Roebuck and Company: A Personal Account, 1986 RADICAL HIST. REV. 57 (defending her testimony for the EEOC) with Rosenberg, What Harms Women in the Workplace, N.Y. Times, Feb. 27, 1986, at A23, col. 1 (defending her testimony for Sears). In addition, numerous scholars have analyzed the Sears case, discussing its significance for the contemporary debate about gender difference in our society. See Finley, Choice and Freedom: Elusive Issues in the Search for Gender Justice, 96 YALE L.J. 914, 937-40 (1987); Milkman, Women's History and the Sears Case, 12 FEMINIST STUD. 375 (1986); Scott, Deconstructing Equality-Versus-Difference: Or, the Uses of Poststructuralist Theory for Feminism, 14 FEMINIST STUD. 33, 38-47 (1988); Williams, Deconstructing Gender, 87 MICH. L. REV. 797, 814 (1989).
29. The news media reported developments in the case from its inception and speculated about its effect on the future of title VII enforcement. See, e.g., Hunter, U.S. Files Five Suits Charging Sears with Job Bias, N.Y. Times, Oct. 23, 1979, at A1, col. 4; Williams, Despite Class-Action Doubts, EEOC Presses Sears Bias Case, Wash. Post, July 9, 1985, at A1, col. 1; Pear, Changes Weighed in Federal Rules on Discrimination, N.Y. Times, Dec. 3, 1984, at A1, col. 6; Greenhouse, Federal Judge Rules for Sears in Sex Bias Case, N. Y. Times, Feb. 4, 1986, at A21, col. 2; Lewin, Statistics Have Become Suspect in Sex Discrimination Cases, N.Y. Times, Feb. 9, 1986, § 4, at E8. Two prominent women's historians testified on opposite sides of the issue of women's "interest" in nontraditional work, and both sought to vindicate their positions publicly after the suit was resolved. Compare Kessler-Harris, Equal Employment Opportunity Commission v. Sears, Roebuck and Company: A Personal Account, 1986 RADICAL HIST. REV. 57 (defending her testimony for the EEOC) with Rosenberg, What Harms Women in the Workplace, N.Y. Times, Feb. 27, 1986, at A23, col. 1 (defending her testimony for Sears). In addition, numerous scholars have analyzed the Sears case, discussing its significance for the contemporary debate about gender difference in our society. See Finley, Choice and Freedom: Elusive Issues in the Search for Gender Justice, 96 YALE L.J. 914, 937-40 (1987); Milkman, Women's History and the Sears Case, 12 FEMINIST STUD. 375 (1986); Scott, Deconstructing Equality-Versus-Difference: Or, the Uses of Poststructuralist Theory for Feminism, 14 FEMINIST STUD. 33, 38-47 (1988); Williams, Deconstructing Gender, 87 MICH. L. REV. 797, 814 (1989).
30. The news media reported developments in the case from its inception and speculated about its effect on the future of title VII enforcement. See, e.g., Hunter, U.S. Files Five Suits Charging Sears with Job Bias, N.Y. Times, Oct. 23, 1979, at A1, col. 4; Williams, Despite Class-Action Doubts, EEOC Presses Sears Bias Case, Wash. Post, July 9, 1985, at A1, col. 1; Pear, Changes Weighed in Federal Rules on Discrimination, N.Y. Times, Dec. 3, 1984, at A1, col. 6; Greenhouse, Federal Judge Rules for Sears in Sex Bias Case, N. Y. Times, Feb. 4, 1986, at A21, col. 2; Lewin, Statistics Have Become Suspect in Sex Discrimination Cases, N.Y. Times, Feb. 9, 1986, § 4, at E8. Two prominent women's historians testified on opposite sides of the issue of women's "interest" in nontraditional work, and both sought to vindicate their positions publicly after the suit was resolved. Compare Kessler-Harris, Equal Employment Opportunity Commission v. Sears, Roebuck and Company: A Personal Account, 1986 RADICAL HIST. REV. 57 (defending her testimony for the EEOC) with Rosenberg, What Harms Women in the Workplace, N.Y. Times, Feb. 27, 1986, at A23, col. 1 (defending her testimony for Sears). In addition, numerous scholars have analyzed the Sears case, discussing its significance for the contemporary debate about gender difference in our society. See Finley, Choice and Freedom: Elusive Issues in the Search for Gender Justice, 96 YALE L.J. 914, 937-40 (1987); Milkman, Women's History and the Sears Case, 12 FEMINIST STUD. 375 (1986); Scott, Deconstructing Equality-Versus-Difference: Or, the Uses of Poststructuralist Theory for Feminism, 14 FEMINIST STUD. 33, 38-47 (1988); Williams, Deconstructing Gender, 87 MICH. L. REV. 797, 814 (1989).
31. The news media reported developments in the case from its inception and speculated about its effect on the future of title VII enforcement. See, e.g., Hunter, U.S. Files Five Suits Charging Sears with Job Bias, N.Y. Times, Oct. 23, 1979, at A1, col. 4; Williams, Despite Class-Action Doubts, EEOC Presses Sears Bias Case, Wash. Post, July 9, 1985, at A1, col. 1; Pear, Changes Weighed in Federal Rules on Discrimination, N.Y. Times, Dec. 3, 1984, at A1, col. 6; Greenhouse, Federal Judge Rules for Sears in Sex Bias Case, N. Y. Times, Feb. 4, 1986, at A21, col. 2; Lewin, Statistics Have Become Suspect in Sex Discrimination Cases, N.Y. Times, Feb. 9, 1986, § 4, at E8. Two prominent women's historians testified on opposite sides of the issue of women's "interest" in nontraditional work, and both sought to vindicate their positions publicly after the suit was resolved. Compare Kessler-Harris, Equal Employment Opportunity Commission v. Sears, Roebuck and Company: A Personal Account, 1986 RADICAL HIST. REV. 57 (defending her testimony for the EEOC) with Rosenberg, What Harms Women in the Workplace, N.Y. Times, Feb. 27, 1986, at A23, col. 1 (defending her testimony for Sears). In addition, numerous scholars have analyzed the Sears case, discussing its significance for the contemporary debate about gender difference in our society. See Finley, Choice and Freedom: Elusive Issues in the Search for Gender Justice, 96 YALE L.J. 914, 937-40 (1987); Milkman, Women's History and the Sears Case, 12 FEMINIST STUD. 375 (1986); Scott, Deconstructing Equality-Versus-Difference: Or, the Uses of Poststructuralist Theory for Feminism, 14 FEMINIST STUD. 33, 38-47 (1988); Williams, Deconstructing Gender, 87 MICH. L. REV. 797, 814 (1989).
32. The news media reported developments in the case from its inception and speculated about its effect on the future of title VII enforcement. See, e.g., Hunter, U.S. Files Five Suits Charging Sears with Job Bias, N.Y. Times, Oct. 23, 1979, at A1, col. 4; Williams, Despite Class-Action Doubts, EEOC Presses Sears Bias Case, Wash. Post, July 9, 1985, at A1, col. 1; Pear, Changes Weighed in Federal Rules on Discrimination, N.Y. Times, Dec. 3, 1984, at A1, col. 6; Greenhouse, Federal Judge Rules for Sears in Sex Bias Case, N. Y. Times, Feb. 4, 1986, at A21, col. 2; Lewin, Statistics Have Become Suspect in Sex Discrimination Cases, N.Y. Times, Feb. 9, 1986, § 4, at E8. Two prominent women's historians testified on opposite sides of the issue of women's "interest" in nontraditional work, and both sought to vindicate their positions publicly after the suit was resolved. Compare Kessler-Harris, Equal Employment Opportunity Commission v. Sears, Roebuck and Company: A Personal Account, 1986 RADICAL HIST. REV. 57 (defending her testimony for the EEOC) with Rosenberg, What Harms Women in the Workplace, N.Y. Times, Feb. 27, 1986, at A23, col. 1 (defending her testimony for Sears). In addition, numerous scholars have analyzed the Sears case, discussing its significance for the contemporary debate about gender difference in our society. See Finley, Choice and Freedom: Elusive Issues in the Search for Gender Justice, 96 YALE L.J. 914, 937-40 (1987); Milkman, Women's History and the Sears Case, 12 FEMINIST STUD. 375 (1986); Scott, Deconstructing Equality-Versus-Difference: Or, the Uses of Poststructuralist Theory for Feminism, 14 FEMINIST STUD. 33, 38-47 (1988); Williams, Deconstructing Gender, 87 MICH. L. REV. 797, 814 (1989).
33. See, e.g., Lewin, supra note 12; Pear, supra note 12. One legal scholar stated this assumption explicitly, arguing that the Sears court's acceptance of the lack of interest argument represented "a dramatic reversal of existing Title VII law," Williams, Deconstructing Gender, supra note 12, at 819, which succeeded for the first time in "enshrining gender stereotypes at the core of Title VII," id. at 814.
34. Employers also resort to this argument to justify sex segregation outside the litigation context. As one researcher summarized managers' views: "The conclusion is: 'It's not our fault' . . . . [E]veryone places the blame, far away from the workplace. A unanimous chorus repeats: it's the parents' fault, it's the teachers' fault, it's the fault of the career advisers. And, of course, fundamentally, it's women's fault: 'they are their own worst enemies.'" C. COCKBURN, MACHINERY OF DOMINANCE: WOMEN, MEN AND TECHNICAL KNOW-HOW 165 (1985). For similar descriptions of how managers have rationalized segregation, see V. BEECHEY & T. PERKINS, A MATTER OF HOURS: WOMEN, PART-TIME WORK AND THE LABOUR MARKET 102-19 (1987); and L. HOWE, PINK COLLAR WORKERS: INSIDE THE WORLD OF WOMEN'S WORK 90-91 (1977).
35. Employers also resort to this argument to justify sex segregation outside the litigation context. As one researcher summarized managers' views: "The conclusion is: 'It's not our fault' . . . . [E]veryone places the blame, far away from the workplace. A unanimous chorus repeats: it's the parents' fault, it's the teachers' fault, it's the fault of the career advisers. And, of course, fundamentally, it's women's fault: 'they are their own worst enemies.'" C. COCKBURN, MACHINERY OF DOMINANCE: WOMEN, MEN AND TECHNICAL KNOW-HOW 165 (1985). For similar descriptions of how managers have rationalized segregation, see V. BEECHEY & T. PERKINS, A MATTER OF HOURS: WOMEN, PART-TIME WORK AND THE LABOUR MARKET 102-19 (1987); and L. HOWE, PINK COLLAR WORKERS: INSIDE THE WORLD OF WOMEN'S WORK 90-91 (1977).
36. Employers also resort to this argument to justify sex segregation outside the litigation context. As one researcher summarized managers' views: "The conclusion is: 'It's not our fault' . . . . [E]veryone places the blame, far away from the workplace. A unanimous chorus repeats: it's the parents' fault, it's the teachers' fault, it's the fault of the career advisers. And, of course, fundamentally, it's women's fault: 'they are their own worst enemies.'" C. COCKBURN, MACHINERY OF DOMINANCE: WOMEN, MEN AND TECHNICAL KNOW-HOW 165 (1985). For similar descriptions of how managers have rationalized segregation, see V. BEECHEY & T. PERKINS, A MATTER OF HOURS: WOMEN, PART-TIME WORK AND THE LABOUR MARKET 102-19 (1987); and L. HOWE, PINK COLLAR WORKERS: INSIDE THE WORLD OF WOMEN'S WORK 90-91 (1977).
37. By "nontraditional" jobs, I refer to the jobs in which women workers are significantly underrepresented in the particular establishments they are suing. I also refer to such jobs as "male-dominated," "traditionally male," or "male" jobs, throughout this Article. By contrast, I use the terms "traditional," "traditionally female," "female-dominated," or "female" jobs to refer to the jobs in which women workers are overrepresented in the particular establishments they are suing. I use these terms only as convenient labels; I do not mean to suggest that there is any historical consistency, cross-cultural agreement, or even uniformity across or within firms about which jobs are appropriate for men and women.
38. See infra notes 80-81 and accompanying text.
39. See infra notes 74 and 82 and accompanying text. Faced with the intersecting disadvantage of racial and sexual discrimination, women of color have often been excluded from work deemed appropriate for white women. For accounts of labor market discrimination against African-American, Hispanic, and Chinese-American women, see, for example, J. JONES, LABOR OF LOVE, LABOR OF SORROW: BLACK WOMEN, WORK, AND THE FAMILY FROM SLAVERY TO THE PRESENT (1985); P. ZAVELLA, WOMEN'S WORK AND CHICANO FAMILIES: CANNERY WORKERS OF THE SANTA CLARA VALLEY (1987); Glenn, Racial Ethnic Women's Labor: The Intersection of Race, Gender, and Class Oppression, in HIDDEN ASPECTS OF WOMEN'S WORK 46 (C. Bose, R. Feldberg & N. Sokoloff eds. 1987); and Malveaux & Wallace, Minority Women in the Workplace, in WORKING WOMEN: PAST, PRESENT, FUTURE 265 (K. Koziara, M. Moskow & L. Tanner eds. 1987).
40. See infra notes 74 and 82 and accompanying text. Faced with the intersecting disadvantage of racial and sexual discrimination, women of color have often been excluded from work deemed appropriate for white women. For accounts of labor market discrimination against African-American, Hispanic, and Chinese-American women, see, for example, J. JONES, LABOR OF LOVE, LABOR OF SORROW: BLACK WOMEN, WORK, AND THE FAMILY FROM SLAVERY TO THE PRESENT (1985); P. ZAVELLA, WOMEN'S WORK AND CHICANO FAMILIES: CANNERY WORKERS OF THE SANTA CLARA VALLEY (1987); Glenn, Racial Ethnic Women's Labor: The Intersection of Race, Gender, and Class Oppression, in HIDDEN ASPECTS OF WOMEN'S WORK 46 (C. Bose, R. Feldberg & N. Sokoloff eds. 1987); and Malveaux & Wallace, Minority Women in the Workplace, in WORKING WOMEN: PAST, PRESENT, FUTURE 265 (K. Koziara, M. Moskow & L. Tanner eds. 1987).
41. See infra notes 74 and 82 and accompanying text. Faced with the intersecting disadvantage of racial and sexual discrimination, women of color have often been excluded from work deemed appropriate for white women. For accounts of labor market discrimination against African-American, Hispanic, and Chinese-American women, see, for example, J. JONES, LABOR OF LOVE, LABOR OF SORROW: BLACK WOMEN, WORK, AND THE FAMILY FROM SLAVERY TO THE PRESENT (1985); P. ZAVELLA, WOMEN'S WORK AND CHICANO FAMILIES: CANNERY WORKERS OF THE SANTA CLARA VALLEY (1987); Glenn, Racial Ethnic Women's Labor: The Intersection of Race, Gender, and Class Oppression, in HIDDEN ASPECTS OF WOMEN'S WORK 46 (C. Bose, R. Feldberg & N. Sokoloff eds. 1987); and Malveaux & Wallace, Minority Women in the Workplace, in WORKING WOMEN: PAST, PRESENT, FUTURE 265 (K. Koziara, M. Moskow & L. Tanner eds. 1987).
42. See infra notes 74 and 82 and accompanying text. Faced with the intersecting disadvantage of racial and sexual discrimination, women of color have often been excluded from work deemed appropriate for white women. For accounts of labor market discrimination against African-American, Hispanic, and Chinese-American women, see, for example, J. JONES, LABOR OF LOVE, LABOR OF SORROW: BLACK WOMEN, WORK, AND THE FAMILY FROM SLAVERY TO THE PRESENT (1985); P. ZAVELLA, WOMEN'S WORK AND CHICANO FAMILIES: CANNERY WORKERS OF THE SANTA CLARA VALLEY (1987); Glenn, Racial Ethnic Women's Labor: The Intersection of Race, Gender, and Class Oppression, in HIDDEN ASPECTS OF WOMEN'S WORK 46 (C. Bose, R. Feldberg & N. Sokoloff eds. 1987); and Malveaux & Wallace, Minority Women in the Workplace, in WORKING WOMEN: PAST, PRESENT, FUTURE 265 (K. Koziara, M. Moskow & L. Tanner eds. 1987).
43. Without intending to deny or ignore the unique discrimination experienced by women of color, this Article analyzes the courts' treatment of sex discrimination claims brought by working women of various racial and ethnic groups in an effort to obtain jobs traditionally done by men. Since the enactment of title VII, women of color have made more progress entering occupations traditionally held by white women than they (or white women) have made in entering occupations traditionally held by men. Researchers typically estimate the extent of occupational segregation with a construct called the index of dissimilarity, which measures the proportion of workers who would have to switch to occupations atypical for their sex or race in order for occupations to be integrated. While the index of race segregation between minority women and white women workers declined substantially from 46.8% in 1960 to 17.2% in 1981, the index of sex segregation between minority women and minority men declined much more modestly during the same period from 54.0 to 47.9. See Albelda, Occupational Segregation by Race and Gender, 1958-81, 39 INDUS. & LAB. REL. REV. 404, 405-06 (1986). Similarly, while the index of race segregation between minority men and white men declined from 38.4 in 1960 to 23.8 in 1981, the index of sex segregation between white women and white men declined more modestly from 59.6 to 53.0. See id. The greatest difference in occupational distribution between any two groups was, of course, between minority women and white men. The index of segregation began at 67.2 in 1960 and declined to only 58.1 in 1981. See id. For other studies reporting similar findings, see J. JONES, cited above in note 17, at 301-03; WOMEN'S WORK, MEN'S WORK, cited above in note 1, at 19; and Beller, cited above in note 1, at 20.
44. Historically, African-American women, immigrant women, and low-income native-born white women have engaged in wage work in large numbers, despite being relegated to undesirable jobs. See, e.g., J. JONES, supra note 17; S. KENNEDY, IF ALL WE DID WAS To WEEP AT HOME: A HISTORY OF WHITE WORKING-CLASS WOMEN IN AMERICA (1979); A. KESSLER-HARRIS, OUT TO WORK: A HISTORY OF WAGE-EARNING WOMEN IN THE UNITED STATES (1982); Milkman, A Statistical Portrait, in 2 N. HEWITT, WOMEN, FAMILIES, AND COMMUNITIES: READINGS IN AMERICAN HISTORY 249 (1990). Estimates of women's historical labor force participation are known to be too low because of the U.S. Census Bureau's practices of undercounting work done by women. See, e.g., Bose, Devaluing Women's Work: The Undercount of Women's Employment in 1900 and 1980, in HIDDEN ASPECTS OF WOMEN'S WORK, supra note 17, at 95.
45. Historically, African-American women, immigrant women, and low-income native-born white women have engaged in wage work in large numbers, despite being relegated to undesirable jobs. See, e.g., J. JONES, supra note 17; S. KENNEDY, IF ALL WE DID WAS To WEEP AT HOME: A HISTORY OF WHITE WORKING-CLASS WOMEN IN AMERICA (1979); A. KESSLER-HARRIS, OUT TO WORK: A HISTORY OF WAGE-EARNING WOMEN IN THE UNITED STATES (1982); Milkman, A Statistical Portrait, in 2 N. HEWITT, WOMEN, FAMILIES, AND COMMUNITIES: READINGS IN AMERICAN HISTORY 249 (1990). Estimates of women's historical labor force participation are known to be too low because of the U.S. Census Bureau's practices of undercounting work done by women. See, e.g., Bose, Devaluing Women's Work: The Undercount of Women's Employment in 1900 and 1980, in HIDDEN ASPECTS OF WOMEN'S WORK, supra note 17, at 95.
46. Historically, African-American women, immigrant women, and low-income native-born white women have engaged in wage work in large numbers, despite being relegated to undesirable jobs. See, e.g., J. JONES, supra note 17; S. KENNEDY, IF ALL WE DID WAS To WEEP AT HOME: A HISTORY OF WHITE WORKING-CLASS WOMEN IN AMERICA (1979); A. KESSLER-HARRIS, OUT TO WORK: A HISTORY OF WAGE-EARNING WOMEN IN THE UNITED STATES (1982); Milkman, A Statistical Portrait, in 2 N. HEWITT, WOMEN, FAMILIES, AND COMMUNITIES: READINGS IN AMERICAN HISTORY 249 (1990). Estimates of women's historical labor force participation are known to be too low because of the U.S. Census Bureau's practices of undercounting work done by women. See, e.g., Bose, Devaluing Women's Work: The Undercount of Women's Employment in 1900 and 1980, in HIDDEN ASPECTS OF WOMEN'S WORK, supra note 17, at 95.
47. Historically, African-American women, immigrant women, and low-income native-born white women have engaged in wage work in large numbers, despite being relegated to undesirable jobs. See, e.g., J. JONES, supra note 17; S. KENNEDY, IF ALL WE DID WAS To WEEP AT HOME: A HISTORY OF WHITE WORKING-CLASS WOMEN IN AMERICA (1979); A. KESSLER-HARRIS, OUT TO WORK: A HISTORY OF WAGE-EARNING WOMEN IN THE UNITED STATES (1982); Milkman, A Statistical Portrait, in 2 N. HEWITT, WOMEN, FAMILIES, AND COMMUNITIES: READINGS IN AMERICAN HISTORY 249 (1990). Estimates of women's historical labor force participation are known to be too low because of the U.S. Census Bureau's practices of undercounting work done by women. See, e.g., Bose, Devaluing Women's Work: The Undercount of Women's Employment in 1900 and 1980, in HIDDEN ASPECTS OF WOMEN'S WORK, supra note 17, at 95.
48. Historically, African-American women, immigrant women, and low-income native-born white women have engaged in wage work in large numbers, despite being relegated to undesirable jobs. See, e.g., J. JONES, supra note 17; S. KENNEDY, IF ALL WE DID WAS To WEEP AT HOME: A HISTORY OF WHITE WORKING-CLASS WOMEN IN AMERICA (1979); A. KESSLER-HARRIS, OUT TO WORK: A HISTORY OF WAGE-EARNING WOMEN IN THE UNITED STATES (1982); Milkman, A Statistical Portrait, in 2 N. HEWITT, WOMEN, FAMILIES, AND COMMUNITIES: READINGS IN AMERICAN HISTORY 249 (1990). Estimates of women's historical labor force participation are known to be too low because of the U.S. Census Bureau's practices of undercounting work done by women. See, e.g., Bose, Devaluing Women's Work: The Undercount of Women's Employment in 1900 and 1980, in HIDDEN ASPECTS OF WOMEN'S WORK, supra note 17, at 95.
49. See, e.g., Working Women's 1980 Platform, reprinted in P. FONER, WOMEN AND THE AMERICAN LABOR MOVEMENT: FROM THE FIRST TRADE UNIONS TO THE PRESENT 497 (1980) (demanding the elimination of job segregation by sex as one of the highest priorities for working women).
50. I use the term "workworld" to refer to all aspects of wage work as a sphere of activity or interest. People's experiences in the workworld include both the experience of searching for work in particular labor markets and the experience of working in jobs in particular workplaces. In stating that courts assume that women form their work aspirations in private realms that are "prior to" the workworld, I mean to suggest a definition that captures a dimension of both temporal and metaphysical priority. The temporal dimension is reflected in the view that women form stable work aspirations through early sex-role socialization that is completed long before they ever begin working or searching for work. This in turn implies an almost metaphysical dimension: women's job preferences are "fixed" in advance, so that they constitute predetermined inputs to the labor market. Throughout this Article, I refer to the realms in which courts imagine women's job preferences to be formed as "pre-work," "pre-workworld," or "pre-labor market" realms. I refer to the view that women form stable job preferences in such realms as the "pre-labor market" explanation for sex segregation. See infra Part V.
51. Throughout this Article, I use the term "gender" as a noun to refer not to the biological sex of a human being, but rather to the complex of social and cultural differences the dominant society attributes to biological men and women. See 6 OXFORD ENGLISH DICTIONARY 428 (2d ed. 1989) (recognizing the modern feminist distinction between sex and gender by defining gender as the "social and cultural, as opposed to the biological, distinctions between the sexes"); cf. 15 id. 107-10 (defining sex). Similarly, I use gender as a verb to refer to the processes of attributing social and cultural differences to biological men and women and, by extension, to the activities commonly associated with them. A major theme of this Article is that the workworld is a central site where people's social identities as men and women are created and contested. Thus, when I state that gender is constructed at work or that work processes gender people and jobs, I mean that employers structure labor markets, work, and workplaces in ways that assign and actually create differences between biological men and women as well as the work they do.
52. As my use of the metaphor of interpretation suggests, I believe that what courts say about sex segregation influences more broadly how people not involved in the immediate legal contest understand that reality. For earlier work in the interpretive tradition, see, for example, Delgado, Storytelling for Oppositionists and Others: A Plea for Narrative, 87 MICH. L. REV. 2411 (1989), which examines the use of storytelling in the struggle for racial justice; Fineman, Dominant Discourse, Professional Language, and Legal Change in Child Custody Decisionmaking, 101 HARV. L. REV. 727 (1988), which explores how social workers helped shape legal discourse in child custody cases so as to disempower custodial mothers; and Minow, Interpreting Rights: An Essay for Robert Cover, 96 YALE L.J. 1860 (1987), which argues for an interpretive approach to law that defends rights as tools for expressing communal aspirations.
53. As my use of the metaphor of interpretation suggests, I believe that what courts say about sex segregation influences more broadly how people not involved in the immediate legal contest understand that reality. For earlier work in the interpretive tradition, see, for example, Delgado, Storytelling for Oppositionists and Others: A Plea for Narrative, 87 MICH. L. REV. 2411 (1989), which examines the use of storytelling in the struggle for racial justice; Fineman, Dominant Discourse, Professional Language, and Legal Change in Child Custody Decisionmaking, 101 HARV. L. REV. 727 (1988), which explores how social workers helped shape legal discourse in child custody cases so as to disempower custodial mothers; and Minow, Interpreting Rights: An Essay for Robert Cover, 96 YALE L.J. 1860 (1987), which argues for an interpretive approach to law that defends rights as tools for expressing communal aspirations.
54. As my use of the metaphor of interpretation suggests, I believe that what courts say about sex segregation influences more broadly how people not involved in the immediate legal contest understand that reality. For earlier work in the interpretive tradition, see, for example, Delgado, Storytelling for Oppositionists and Others: A Plea for Narrative, 87 MICH. L. REV. 2411 (1989), which examines the use of storytelling in the struggle for racial justice; Fineman, Dominant Discourse, Professional Language, and Legal Change in Child Custody Decisionmaking, 101 HARV. L. REV. 727 (1988), which explores how social workers helped shape legal discourse in child custody cases so as to disempower custodial mothers; and Minow, Interpreting Rights: An Essay for Robert Cover, 96 YALE L.J. 1860 (1987), which argues for an interpretive approach to law that defends rights as tools for expressing communal aspirations.
55. See infra note 71.
56. There are many documented cases of dramatic increases in female participation in non-traditional jobs in response to court orders and court-supervised consent decrees. See, e.g., COUNCIL ON ECONOMIC PRIORITIES, WOMEN AND MINORITIES IN BANKING: SHORTCHANGE/ UPDATE 68 (1976) (recording a 166% increase in female participation in managerial, professional, technical and salesworker jobs, and an even greater increase in female participation in blue-collar jobs, at a major bank between 1971 and 1975); K. DEAUX & J. ULLMAN, WOMEN OF STEEL: FEMALE BLUE-COLLAR WORKERS IN THE BASIC STEEL INDUSTRY 85 (1983) (recording a 170% increase in female participation in production, maintenance, and craft positions in two steel mills between 1976 and 1979); Appendix D, in EQUAL EMPLOYMENT OPPORTUNITY AND THE AT&T CASE 343 (P. Wallace ed. 1976) (recording a 119% increase in female participation in craft jobs and a 46% increase in female participation in managerial jobs between 1973 and the end of 1974).
57. There are many documented cases of dramatic increases in female participation in non-traditional jobs in response to court orders and court-supervised consent decrees. See, e.g., COUNCIL ON ECONOMIC PRIORITIES, WOMEN AND MINORITIES IN BANKING: SHORTCHANGE/ UPDATE 68 (1976) (recording a 166% increase in female participation in managerial, professional, technical and salesworker jobs, and an even greater increase in female participation in blue-collar jobs, at a major bank between 1971 and 1975); K. DEAUX & J. ULLMAN, WOMEN OF STEEL: FEMALE BLUE-COLLAR WORKERS IN THE BASIC STEEL INDUSTRY 85 (1983) (recording a 170% increase in female participation in production, maintenance, and craft positions in two steel mills between 1976 and 1979); Appendix D, in EQUAL EMPLOYMENT OPPORTUNITY AND THE AT&T CASE 343 (P. Wallace ed. 1976) (recording a 119% increase in female participation in craft jobs and a 46% increase in female participation in managerial jobs between 1973 and the end of 1974).
58. There are many documented cases of dramatic increases in female participation in non-traditional jobs in response to court orders and court-supervised consent decrees. See, e.g., COUNCIL ON ECONOMIC PRIORITIES, WOMEN AND MINORITIES IN BANKING: SHORTCHANGE/ UPDATE 68 (1976) (recording a 166% increase in female participation in managerial, professional, technical and salesworker jobs, and an even greater increase in female participation in blue-collar jobs, at a major bank between 1971 and 1975); K. DEAUX & J. ULLMAN, WOMEN OF STEEL: FEMALE BLUE-COLLAR WORKERS IN THE BASIC STEEL INDUSTRY 85 (1983) (recording a 170% increase in female participation in production, maintenance, and craft positions in two steel mills between 1976 and 1979); Appendix D, in EQUAL EMPLOYMENT OPPORTUNITY AND THE AT&T CASE 343 (P. Wallace ed. 1976) (recording a 119% increase in female participation in craft jobs and a 46% increase in female participation in managerial jobs between 1973 and the end of 1974).
59. In classwide title VII cases, a plaintiff makes a prima facie case of discrimination by demonstrating a gross disparity between the number of protected class members hired by the employer and the number of qualified protected class members who would have been available for hire in the absence of discrimination. See, e.g., Hazelwood School Dist. v. United States, 433 U.S. 299, 307-08 (1977); International Bhd. of Teamsters v. United States, 431 U.S. 324, 339 n.20 (1977). In its most dramatic form, statistical proof reveals that the employer has never hired any women or minorities or, more commonly, has never hired them for the higher-paying jobs. See Teamsters, 431 U.S. at 342 n.23 (referring to the "inexorable zero" as powerful proof of intentional discrimination).
60. Teamsters, 431 U.S. at 340 n.20 (quoting United States v. Ironworkers Local 86, 443 F.2d 544, 551 (9th Cir.), cert, denied, 404 U.S. 984 (1971), and citing other early courts of appeals decisions).
61. The earliest case I found that addresses the lack of interest argument is Cypress v. Newport News General & Nonsectarian Hospital Association, 375 F.2d 648, 653 (4th Cir. 1967). For a fuller discussion of this case, see pp. 1772-73 below.
62. 431 U.S. 324 (1977).
63. 30433 U.S. 321 (1977).
64. 433 U.S. 299 (1977).
65. Although Teamsters and Hazelwood are race discrimination cases, they provide precedent for how to analyze the lack of interest argument in sex discrimination cases as well. Title VII's anti-discrimination provisions make no distinction between race and sex discrimination. See 42 U.S.C. § 2000e-2(a)(1)-(2) (1982). The Supreme Court has never suggested that race and sex discrimination are to be analyzed differently under title VII. To the contrary, the Court's title VII decisions have consistently used the same approaches to analyze race and sex discrimination, citing precedent from the two contexts interchangeably. See, e.g., Johnson v. Transportation Agency, 480 U.S. 616, 631 (1987) (citing United Steelworkers of America v. Weber, 443 U.S. 193, 197 (1979), to uphold the validity under title VII of an affirmative action plan for women); Meritor Sav. Bank v. Vinson, 477 U.S. 57, 66-67 (1986) (citing title VII decisions recognizing a cause of action for racial harassment to support recognizing a cause of action for sexual harassment creating a hostile work environment). Indeed, in Dothard, a sex discrimination case, the Court supported its analysis of the lack of interest issue with a reference to its earlier analysis of the same issue in Teamsters. See infra p. 1762.
66. Classwide disparate treatment suits typically are tried in two stages. The first stage addresses whether the employer has engaged in a regular practice of discriminating against protected class members as a group. See Teamsters, 431 U.S. at 336 & n.16. The second stage allocates relief to individual victims of discrimination if liability is established. See id. at 361-ïó. Classwide disparate treatment cases include both private class actions initiated under Federal Rule of Civil Procedure 23 and pattern or practice cases brought by the Department of Justice or the EEOC under § 707(a) of title VII, 42 U.S.C. § 2000e-6(a) (1982).
67. See Teamsters, 431 U.S. at 342-43.
68. Id. at 365.
69. Id. at 369.
70. See id. at 371-72. The Court held that a minority non-applicant who proves that he or she would have applied but for discrimination stands in the same position as one who did apply. At that point, the employer bears the burden of proving that he would not have hired the person anyway, because of a legitimate nondiscriminatory reason such as the lack of a vacancy, even in the absence of the already proven pattern of discrimination. See id. at 357-62, 367-68.
71. The lack of interest argument involves different considerations when asserted during the liability, as opposed to the remedy, stage of a classwide discrimination suit. In raising the lack of interest argument during the remedial phase, the employer is claiming that some individual class member would not have been interested in some nontraditional job, even if the employer had always welcomed protected class members into that job. This claim rests on an assertion about the work preference of a particular claimant, as an individual, that is not necessarily related to that person's race or sex. See id. at 369-70 (acknowledging that even some whites, who had never been subjected to discrimination, had failed to express interest in the line-driver job). But in raising the lack of interest argument during the liability phase, the employer is claiming that minorities or women are, as a group, systematically less interested than whites or men in the nontraditional work at issue. This claim necessarily rests on an assertion that there is something distinctive about minorities' or women's work preferences that is directly attributable to their race or sex.
72. Dothard v. Rawlinson, 433 U.S. 321 (1977).
73. In Johnson v. Transportation Agency, 480 U.S. 616 (1987), the Supreme Court upheld the validity under title VII of an affirmative action program for women, holding that the plan was justified by the "'manifest imbalance' that reflected underrepresentation of women in 'traditionally segregated job categories'" in Santa Clara County's workforce. Id. at 631 (citing United Steelworkers of America v. Weber, 443 U.S. 193, 197 (1979)). Although the majority did not mention the issue of women's interest in the county's nontraditional jobs, Justice Scalia argued in dissent that the absence of women from road maintenance work was attributable not to historical discrimination, but rather to women's own lack of interest in such work. See id. at 668 (Scalia, J., dissenting). Because Johnson involved the validity of a voluntary affirmative action plan rather than the county's liability for sex discrimination against women, I do not analyze the case here. Johnson did not influence the lower court decisions in my study: only three cases were decided after Johnson, and none of them cited Johnson in analyzing the lack of interest issue. See EEOC v. General Tel. Co., 885 F.2d 575 (9th Cir. 1989); EEOC v. Sears, Roebuck & Co., 839 F.2d 302 (7th Cir. 1988); Palmer v. Shultz, 815 F.2d 84 (D.C. Cir. 1987).
74. In a disparate impact case, a plaintiff need not prove that the employer acted with discriminatory purpose or intent. To establish a prima facie case of discrimination, the plaintiff need only show that apparently neutral selection criteria operated to exclude protected class members at a disproportionate rate. See, e.g., Griggs v. Duke Power Co., 401 U.S. 424, 429-30 (1971). The employer bears the burden of justifying the challenged criteria. The Supreme Court recently held that the employer's burden is merely one of "producing evidence of a business justification for his employment practice[s]." Wards Cove Packing Co. v. Atonio, 109 S. Ct. 2115, 2126 (1989). Before Wards Cove, lower courts had held that the employer's burden is the heavier one of persuading the court of the business necessity of the challenged practices. See id. at 2130 & n.14 (Stevens, J., dissenting) (citing earlier courts of appeals' decisions).
75. Plaintiffs' statistics showed that only 58.9% of all women in the United States between the ages of 18 and 79 could meet Alabama's height and weight requirements, whereas 99.8% of men could do so. Only 12.9% of the state's prison guards were women. See Dothard, 433 U.S. at 329-30 & n.12.
76. Id. at 348 (White, J., dissenting).
77. If the disparate impact model is to have any meaning, plaintiffs must be able to establish the disproportionate impact of the requirements they are challenging with reference to some pool of potential workers besides the applicant pool. Otherwise, employers could always argue that only women or minorities who satisfy the requirements were interested enough in the job to apply; by definition, the requirements would not operate to exclude those who already meet them.
78. Dothard, 433 U.S. at 330 (citing International Brotherhood of Teamsters v. United States, 431 U.S. 324, 365-67 (1977)).
79. See id. at 331; see also id. at 337-39 (Rehnquist, J., concurring) (emphasizing that the state had made virtually no effort to challenge the plaintiffs' prima facie statistical showing in the trial court).
80. See 433 U.S. at 331 (stating that the district court was entitled to find that the height and weight requirements had a disparate impact where the defendant introduced no evidence to the contrary); see also id. at 338 (Rehnquist, J., concurring) ("It is for the District Court . . . to determine whether [the] statistics appear sufficiently probative of the ultimate fact in issue - whether a given job qualification requirement has a disparate impact on some group protected by Title VII.").
81. Hazelwood School Dist. v. United States, 433 U.S. 299 (1977).
82. See id. at 302-03 & nn.2 & 4, 304 n.7.
83. Eighty percent of Hazelwood's teachers had come from the city of St. Louis or the surrounding county, and one-third from the city alone. See id. at 315 n.2 (Stevens, J., dissenting).
84. In 1970, 15.4% of all teachers in the city of St. Louis and the surrounding county (which included Hazelwood) were black. Yet in 1972 and 1973, blacks constituted less than 2% of Hazelwood's teaching staff, see 433 U.S. at 303-05, and only 3.7% of its newly hired teachers, see id. at 310.
85. In ruling against the government, the district court had relied in part on the fact that the racial composition of Hazelwood's teaching staff differed only slightly from the racial composition of its students. The Supreme Court affirmed the Eighth Circuit's holding that this comparison was erroneous as a matter of law, because the proper basis of comparison was the racial composition of the qualified school teachers in the relevant labor market. See id. at 304-05, 308-09.
86. Hazelwood also contended that the court of appeals had erred in failing to consider whether the district had discriminated in hiring after title VII became applicable to public employers in 1972. See Equal Employment Opportunity Act of 1972, Pub. L. No. 92-261, 86 Stat. 103 (amending 42 U.S.C. § 2000e(a) (1982)). The Supreme Court agreed that because figures showing the racial composition of Hazelwood's teaching staff built in the effect of hiring decisions made before 1972, the most appropriate figures to review were those showing the racial composition of teachers the district had hired since 1972. See Hazelwood, 433 U.S. at 308-10.
87. See id. at 312 (directing the trial court to consider the extent to which black teachers employed by the city would prefer working for other school districts such as Hazelwood); id. at 317 (Stevens, J., dissenting) (arguing that the city's 50% goal provided no basis for excluding it from the relevant labor market because there was no proof that black teachers employed by the city preferred working there instead of for Hazelwood).
88. Blacks constituted only 5.7% of the teachers employed in the county, excluding the city. See 433 U.S. at 303.
89. Id. at 309.
90. See id. at 310-13.
91. See id. at 308 n.13, 310, 313 n.21.
92. See id. at 312 ("Only the trial court is in a position to make the appropriate determination after further findings."); see also id. at 313 (Brennan, J., concurring) ("[T]oday's opinion revolves around the relative factfinding roles of district courts and courts of appeals.").
93. See 433 U.S. at 309-10 (referring to the possibility that Hazelwood might be able to rebut the "prima facie statistical proof"); see also id. at 314 (Brennan, J., concurring) (interpreting the majority opinion as holding that the government's statistical proof sufficed to make a prima facie case); id. at 347-48 (White, J., concurring) (same).
94. See id. at 314 (Brennan, J., concurring) (emphasizing that on remand it would be up to Hazelwood to "come forward with more focused and specific applicant-flow data in the hope of answering the Government's prima facie case"); id. at 347-48 (White, J., concurring) (same).
95. See Dothard v. Rawlinson, 433 U.S. 321, 330 (1977); International Bhd. of Teamsters v. United States, 431 U.S. 324, 365 (1977).
96. See Hazelwood, 433 U.S. at 312 (enumerating five factors that the trial court should consider).
97. In individual disparate treatment cases, where the inference of discriminatory purpose to be drawn from the plaintiff's prima facie case is often relatively weak, the Supreme Court has held that the employer's rebuttal burden is merely that of producing evidence of (or "articulating") a legitimate nondiscriminatory reason to explain the plaintiff's unfavorable treatment. The ultimate burden of persuading the court that it was the plaintiff's race or sex, rather than the employer's stated reason, that motivated the employer remains on the plaintiff. See Texas Dep't of Community Affairs v. Burdine, 450 U.S. 248, 252-56 (1981); Segar v. Smith, 738 F.2d 1249, 1268 (D.C. Cir. 1984), cert, denied, 471 U.S. 1115 (1985).
98. See, e.g., EEOC v. Sears, Roebuck & Co., 839 F.2d 302, 309 (7th Cir. 1988); Piva v. Xerox Corp., 654 F.2d 591, 594 (9th Cir. 1981).
99. In practical terms, this means that the plaintiff must offer convincing proof of women's interest in order to meet her overall burden of persuasion. See, e.g., Sears, 839 F.2d at 309, 334-38.
100. See, e.g., Griffin v. Carlin, 755 F.zd 1516, 1526-28 (11th Cir. 1985); Segar, 738 F.2d at 1267-70; see also Bazemore v. Friday, 478 U.S. 385, 397-404 (1986) (per curiam) (Brennan, J., concurring in part) (holding that the employer may not meet its burden of rebutting a plaintiff 's prima facie case through speculative assertions that factors omitted from the plaintiff 's statistical analyses might explain the apparent disparities on a basis other than race).
101. See, e.g., EEOC v. General Tel. Co., 885 F.2d 575, 581-82 (9th Cir. 1989); Palmer v. Shultz, 815 F.2d 84, 101 (D.C. Cir. 1987).
102. See Anderson v. City of Bessemer, 470 U.S. 564, 573 (1985); Pullman-Standard v. Swint, 456 U.S. 273, 287-88 (1982).
103. See, e.g., Sears, 839 F.2d at 310 (applying this reasoning to conclude that the trial court's finding with respect to the interest issue is subject to reversal only for clear error).
104. I searched the following two sources: (1) the Fair Employment Practice Cumulative Digests, published by the Bureau of National Affairs, Volumes 1-46, covering the period from 1965 to 1987 and (2) the West digests, including the Eighth Decennial Digest, Ninth Decennial Digest, the General Digest, Sixth Series, Volumes 1-53, and the General Digest, Seventh Series, Volumes 1-2, together covering the period from 1966 through 1986. Although I searched systematically for all cases in which the lack of interest argument was raised, I may not have found every such case. That would raise no methodological problem, however, so long as my search strategy did not produce a biased selection of cases. There is no reason to suspect that my search yielded any such bias. I do not claim that the decisions included in my study are representative of any larger sample of title VII cases. Published judicial decisions are probably not a random sample of all judicial decisions. See, e.g., S. Olson, Studying Federal District Courts Through Published Cases (June 8-11, 1989) (unpublished manuscript of a paper presented at the annual meeting of the Law and Society Association) (suggesting that civil rights cases are overrepresented among published decisions of federal district courts). In addition, lawsuits tried to judgment are probably not a random sample of all lawsuits that are filed. See, e.g., Priest & Klein, The Selection of Disputes for Litigation, 13 J. LEGAL STUD, 1 (1984). A more complete picture of "the law" would review the terms of pre-trial settlement agreements, since the overwhelming majority of lawsuits are settled before they ever reach trial. See Galanter, Reading the Landscape of Disputes: What We Know and Don't Know (And Think We Know) About Our Allegedly Contentious and Litigious Society, 31 UCLA L. REV. 4, 26, 32 (1983). Furthermore, even filed lawsuits are a small portion of legally actionable grievances that might have been brought to the judicial system. See Felstiner, Abel & Sarat, The Emergence and Transformation of Disputes: Naming, Blaming, Claiming . . ., 15 LAW & SOC. REV. 631, 636 (1980-1981); Trubek, Sarat, Felstiner, Kritzer & Grossman, The Costs of Ordinary Litigation, 31 UCLA L. REV. 72, 85-87 (1983). There is evidence that people who experience employment discrimination grievances are less likely to take steps to vindicate their rights, including filing lawsuits, than are people who experience other types of grievances. See B. CURRAN, THE LEGAL NEEDS OF THE PUBLIC 141-42, 146, 262 (1977); see also Miller & Sarat, Grievances, Claims, and Disputes: Assessing the Adversary Culture, 15 LAW & Soc. REV. 525, 537, 540-42, 544-45, 563-64 (1980-1981) (reporting the same finding for people who experience discrimination generally).
105. I searched the following two sources: (1) the Fair Employment Practice Cumulative Digests, published by the Bureau of National Affairs, Volumes 1-46, covering the period from 1965 to 1987 and (2) the West digests, including the Eighth Decennial Digest, Ninth Decennial Digest, the General Digest, Sixth Series, Volumes 1-53, and the General Digest, Seventh Series, Volumes 1-2, together covering the period from 1966 through 1986. Although I searched systematically for all cases in which the lack of interest argument was raised, I may not have found every such case. That would raise no methodological problem, however, so long as my search strategy did not produce a biased selection of cases. There is no reason to suspect that my search yielded any such bias. I do not claim that the decisions included in my study are representative of any larger sample of title VII cases. Published judicial decisions are probably not a random sample of all judicial decisions. See, e.g., S. Olson, Studying Federal District Courts Through Published Cases (June 8-11, 1989) (unpublished manuscript of a paper presented at the annual meeting of the Law and Society Association) (suggesting that civil rights cases are overrepresented among published decisions of federal district courts). In addition, lawsuits tried to judgment are probably not a random sample of all lawsuits that are filed. See, e.g., Priest & Klein, The Selection of Disputes for Litigation, 13 J. LEGAL STUD, 1 (1984). A more complete picture of "the law" would review the terms of pre-trial settlement agreements, since the overwhelming majority of lawsuits are settled before they ever reach trial. See Galanter, Reading the Landscape of Disputes: What We Know and Don't Know (And Think We Know) About Our Allegedly Contentious and Litigious Society, 31 UCLA L. REV. 4, 26, 32 (1983). Furthermore, even filed lawsuits are a small portion of legally actionable grievances that might have been brought to the judicial system. See Felstiner, Abel & Sarat, The Emergence and Transformation of Disputes: Naming, Blaming, Claiming . . ., 15 LAW & SOC. REV. 631, 636 (1980-1981); Trubek, Sarat, Felstiner, Kritzer & Grossman, The Costs of Ordinary Litigation, 31 UCLA L. REV. 72, 85-87 (1983). There is evidence that people who experience employment discrimination grievances are less likely to take steps to vindicate their rights, including filing lawsuits, than are people who experience other types of grievances. See B. CURRAN, THE LEGAL NEEDS OF THE PUBLIC 141-42, 146, 262 (1977); see also Miller & Sarat, Grievances, Claims, and Disputes: Assessing the Adversary Culture, 15 LAW & Soc. REV. 525, 537, 540-42, 544-45, 563-64 (1980-1981) (reporting the same finding for people who experience discrimination generally).
106. I searched the following two sources: (1) the Fair Employment Practice Cumulative Digests, published by the Bureau of National Affairs, Volumes 1-46, covering the period from 1965 to 1987 and (2) the West digests, including the Eighth Decennial Digest, Ninth Decennial Digest, the General Digest, Sixth Series, Volumes 1-53, and the General Digest, Seventh Series, Volumes 1-2, together covering the period from 1966 through 1986. Although I searched systematically for all cases in which the lack of interest argument was raised, I may not have found every such case. That would raise no methodological problem, however, so long as my search strategy did not produce a biased selection of cases. There is no reason to suspect that my search yielded any such bias. I do not claim that the decisions included in my study are representative of any larger sample of title VII cases. Published judicial decisions are probably not a random sample of all judicial decisions. See, e.g., S. Olson, Studying Federal District Courts Through Published Cases (June 8-11, 1989) (unpublished manuscript of a paper presented at the annual meeting of the Law and Society Association) (suggesting that civil rights cases are overrepresented among published decisions of federal district courts). In addition, lawsuits tried to judgment are probably not a random sample of all lawsuits that are filed. See, e.g., Priest & Klein, The Selection of Disputes for Litigation, 13 J. LEGAL STUD, 1 (1984). A more complete picture of "the law" would review the terms of pre-trial settlement agreements, since the overwhelming majority of lawsuits are settled before they ever reach trial. See Galanter, Reading the Landscape of Disputes: What We Know and Don't Know (And Think We Know) About Our Allegedly Contentious and Litigious Society, 31 UCLA L. REV. 4, 26, 32 (1983). Furthermore, even filed lawsuits are a small portion of legally actionable grievances that might have been brought to the judicial system. See Felstiner, Abel & Sarat, The Emergence and Transformation of Disputes: Naming, Blaming, Claiming . . ., 15 LAW & SOC. REV. 631, 636 (1980-1981); Trubek, Sarat, Felstiner, Kritzer & Grossman, The Costs of Ordinary Litigation, 31 UCLA L. REV. 72, 85-87 (1983). There is evidence that people who experience employment discrimination grievances are less likely to take steps to vindicate their rights, including filing lawsuits, than are people who experience other types of grievances. See B. CURRAN, THE LEGAL NEEDS OF THE PUBLIC 141-42, 146, 262 (1977); see also Miller & Sarat, Grievances, Claims, and Disputes: Assessing the Adversary Culture, 15 LAW & Soc. REV. 525, 537, 540-42, 544-45, 563-64 (1980-1981) (reporting the same finding for people who experience discrimination generally).
107. I searched the following two sources: (1) the Fair Employment Practice Cumulative Digests, published by the Bureau of National Affairs, Volumes 1-46, covering the period from 1965 to 1987 and (2) the West digests, including the Eighth Decennial Digest, Ninth Decennial Digest, the General Digest, Sixth Series, Volumes 1-53, and the General Digest, Seventh Series, Volumes 1-2, together covering the period from 1966 through 1986. Although I searched systematically for all cases in which the lack of interest argument was raised, I may not have found every such case. That would raise no methodological problem, however, so long as my search strategy did not produce a biased selection of cases. There is no reason to suspect that my search yielded any such bias. I do not claim that the decisions included in my study are representative of any larger sample of title VII cases. Published judicial decisions are probably not a random sample of all judicial decisions. See, e.g., S. Olson, Studying Federal District Courts Through Published Cases (June 8-11, 1989) (unpublished manuscript of a paper presented at the annual meeting of the Law and Society Association) (suggesting that civil rights cases are overrepresented among published decisions of federal district courts). In addition, lawsuits tried to judgment are probably not a random sample of all lawsuits that are filed. See, e.g., Priest & Klein, The Selection of Disputes for Litigation, 13 J. LEGAL STUD, 1 (1984). A more complete picture of "the law" would review the terms of pre-trial settlement agreements, since the overwhelming majority of lawsuits are settled before they ever reach trial. See Galanter, Reading the Landscape of Disputes: What We Know and Don't Know (And Think We Know) About Our Allegedly Contentious and Litigious Society, 31 UCLA L. REV. 4, 26, 32 (1983). Furthermore, even filed lawsuits are a small portion of legally actionable grievances that might have been brought to the judicial system. See Felstiner, Abel & Sarat, The Emergence and Transformation of Disputes: Naming, Blaming, Claiming . . ., 15 LAW & SOC. REV. 631, 636 (1980-1981); Trubek, Sarat, Felstiner, Kritzer & Grossman, The Costs of Ordinary Litigation, 31 UCLA L. REV. 72, 85-87 (1983). There is evidence that people who experience employment discrimination grievances are less likely to take steps to vindicate their rights, including filing lawsuits, than are people who experience other types of grievances. See B. CURRAN, THE LEGAL NEEDS OF THE PUBLIC 141-42, 146, 262 (1977); see also Miller & Sarat, Grievances, Claims, and Disputes: Assessing the Adversary Culture, 15 LAW & Soc. REV. 525, 537, 540-42, 544-45, 563-64 (1980-1981) (reporting the same finding for people who experience discrimination generally).
108. I searched the following two sources: (1) the Fair Employment Practice Cumulative Digests, published by the Bureau of National Affairs, Volumes 1-46, covering the period from 1965 to 1987 and (2) the West digests, including the Eighth Decennial Digest, Ninth Decennial Digest, the General Digest, Sixth Series, Volumes 1-53, and the General Digest, Seventh Series, Volumes 1-2, together covering the period from 1966 through 1986. Although I searched systematically for all cases in which the lack of interest argument was raised, I may not have found every such case. That would raise no methodological problem, however, so long as my search strategy did not produce a biased selection of cases. There is no reason to suspect that my search yielded any such bias. I do not claim that the decisions included in my study are representative of any larger sample of title VII cases. Published judicial decisions are probably not a random sample of all judicial decisions. See, e.g., S. Olson, Studying Federal District Courts Through Published Cases (June 8-11, 1989) (unpublished manuscript of a paper presented at the annual meeting of the Law and Society Association) (suggesting that civil rights cases are overrepresented among published decisions of federal district courts). In addition, lawsuits tried to judgment are probably not a random sample of all lawsuits that are filed. See, e.g., Priest & Klein, The Selection of Disputes for Litigation, 13 J. LEGAL STUD, 1 (1984). A more complete picture of "the law" would review the terms of pre-trial settlement agreements, since the overwhelming majority of lawsuits are settled before they ever reach trial. See Galanter, Reading the Landscape of Disputes: What We Know and Don't Know (And Think We Know) About Our Allegedly Contentious and Litigious Society, 31 UCLA L. REV. 4, 26, 32 (1983). Furthermore, even filed lawsuits are a small portion of legally actionable grievances that might have been brought to the judicial system. See Felstiner, Abel & Sarat, The Emergence and Transformation of Disputes: Naming, Blaming, Claiming . . ., 15 LAW & SOC. REV. 631, 636 (1980-1981); Trubek, Sarat, Felstiner, Kritzer & Grossman, The Costs of Ordinary Litigation, 31 UCLA L. REV. 72, 85-87 (1983). There is evidence that people who experience employment discrimination grievances are less likely to take steps to vindicate their rights, including filing lawsuits, than are people who experience other types of grievances. See B. CURRAN, THE LEGAL NEEDS OF THE PUBLIC 141-42, 146, 262 (1977); see also Miller & Sarat, Grievances, Claims, and Disputes: Assessing the Adversary Culture, 15 LAW & Soc. REV. 525, 537, 540-42, 544-45, 563-64 (1980-1981) (reporting the same finding for people who experience discrimination generally).
109. I searched the following two sources: (1) the Fair Employment Practice Cumulative Digests, published by the Bureau of National Affairs, Volumes 1-46, covering the period from 1965 to 1987 and (2) the West digests, including the Eighth Decennial Digest, Ninth Decennial Digest, the General Digest, Sixth Series, Volumes 1-53, and the General Digest, Seventh Series, Volumes 1-2, together covering the period from 1966 through 1986. Although I searched systematically for all cases in which the lack of interest argument was raised, I may not have found every such case. That would raise no methodological problem, however, so long as my search strategy did not produce a biased selection of cases. There is no reason to suspect that my search yielded any such bias. I do not claim that the decisions included in my study are representative of any larger sample of title VII cases. Published judicial decisions are probably not a random sample of all judicial decisions. See, e.g., S. Olson, Studying Federal District Courts Through Published Cases (June 8-11, 1989) (unpublished manuscript of a paper presented at the annual meeting of the Law and Society Association) (suggesting that civil rights cases are overrepresented among published decisions of federal district courts). In addition, lawsuits tried to judgment are probably not a random sample of all lawsuits that are filed. See, e.g., Priest & Klein, The Selection of Disputes for Litigation, 13 J. LEGAL STUD, 1 (1984). A more complete picture of "the law" would review the terms of pre-trial settlement agreements, since the overwhelming majority of lawsuits are settled before they ever reach trial. See Galanter, Reading the Landscape of Disputes: What We Know and Don't Know (And Think We Know) About Our Allegedly Contentious and Litigious Society, 31 UCLA L. REV. 4, 26, 32 (1983). Furthermore, even filed lawsuits are a small portion of legally actionable grievances that might have been brought to the judicial system. See Felstiner, Abel & Sarat, The Emergence and Transformation of Disputes: Naming, Blaming, Claiming . . ., 15 LAW & SOC. REV. 631, 636 (1980-1981); Trubek, Sarat, Felstiner, Kritzer & Grossman, The Costs of Ordinary Litigation, 31 UCLA L. REV. 72, 85-87 (1983). There is evidence that people who experience employment discrimination grievances are less likely to take steps to vindicate their rights, including filing lawsuits, than are people who experience other types of grievances. See B. CURRAN, THE LEGAL NEEDS OF THE PUBLIC 141-42, 146, 262 (1977); see also Miller & Sarat, Grievances, Claims, and Disputes: Assessing the Adversary Culture, 15 LAW & Soc. REV. 525, 537, 540-42, 544-45, 563-64 (1980-1981) (reporting the same finding for people who experience discrimination generally).
110. See, e.g., Dothard v. Rawlinson, 433 U.S. 321, 330 (1977); supra pp. 1761-63 (discussing Dothard).
111. A single employment discrimination lawsuit may of course raise a host of different claims, ranging from discrimination in hiring, promotion, transfer, or discharge to claims of retaliation or harassment. In such cases, the employer may assert the lack of interest argument to defend against one or more of these claims (for example, hiring and promotion), but not all of them (for example, discharge and retaliation). In cases in which the court analyzed the lack of interest argument separately as a defense to more than one claim of discrimination, I have included each claim as a separate "case" for purposes of analysis. The primary justification for doing so is that the evidence presented on the interest issue is sometimes very different for separate claims; it would have been impossible to capture the full range of evidence that influenced the outcomes without including each claim as a separate unit. Furthermore, it seemed methodologically indefensible to exclude arbitrarily one claim or the other. In any event, the outcomes are identical regardless of whether all claims or only one claim is included for each case. When all claims are included, plaintiffs prevail in 57.4% (31 out of 54) of the claims, and when only one claim is included, plaintiffs prevail in 57.4% (27 out of 47) of the claims.
112. Of the 54 sex discrimination claims analyzed in this study, in 44 (81.5%) the plaintiffs alleged only sex discrimination in connection with their underlying claim. In five (9.3%) more, plaintiffs alleged both race and sex discrimination, and the employer tried to defend the sex claim by asserting that women lack interest, but chose not to defend the race claim by arguing that minorities do. In the remaining five (9.3%) cases, plaintiffs alleged both race and sex discrimination, and the employer tried to defend the sex claim by asserting that women lack interest and tried to defend the race claim by asserting that minorities do. For these last five cases, only the claims for which employers asserted women's lack of interest are designated sex claims and included in this study. This analysis shows that women of color face a potential double risk with respect to the lack of interest argument. As women, they may face the argument that they lack interest because of their gender; as racial minorities, they may face the argument that they lack interest because of their race.
113. See Logan v. General Fireproofing Co., 6 Fair Empl. Prac. Cas. (BNA) 140 (W.D.N.C. 1972). The dearth of cases before 1972 is consistent with other studies. See Mills, On the Use of Equal Employment Laws, 24 PAC. Soc. REV. 196, 201 (1981) (finding that between 1967 and 1972, district courts in five circuits decided only 20 cases alleging any type of sex discrimination in employment); cf. Stidham, Carp & Rowland, Women's Rights Before the Federal District Courts, 1971-1977, 11 AM. POL. Q. 205, 208 (1983) (limiting a study of federal district court cases involving sex discrimination of all types to the period between 1971 and 1977, because "[p]rior to 1971 there were too few women's rights cases to code as a separate case category").
114. See Logan v. General Fireproofing Co., 6 Fair Empl. Prac. Cas. (BNA) 140 (W.D.N.C. 1972). The dearth of cases before 1972 is consistent with other studies. See Mills, On the Use of Equal Employment Laws, 24 PAC. Soc. REV. 196, 201 (1981) (finding that between 1967 and 1972, district courts in five circuits decided only 20 cases alleging any type of sex discrimination in employment); cf. Stidham, Carp & Rowland, Women's Rights Before the Federal District Courts, 1971-1977, 11 AM. POL. Q. 205, 208 (1983) (limiting a study of federal district court cases involving sex discrimination of all types to the period between 1971 and 1977, because "[p]rior to 1971 there were too few women's rights cases to code as a separate case category").
115. See EEOC v. Gen. Tel. Co., 885 F.2d 575 (9th Cir. 1989).
116. Of the 54 claims in the study, only eight (14.8%) involved individual claims of discrimination. Employers typically assert the lack of interest argument in individual cases for the same reason they do in classwide discrimination cases: to discredit the statistical evidence. In individual cases, the plaintiff often introduces statistical evidence to persuade the court that her treatment was only one instance of a larger pattern of discrimination by the employer. See, e.g., Davis v. Richmond, F. & P. R.R., 593 F. Supp. 271 (E.D. Va. 1984), aff'd in part and rev'd in part on other grounds, 803 F.2d 1322 (4th Cir. 1986).
117. Thirty-nine of the 54 claims (72.2%) alleged disparate treatment, while 15 (27.8%) alleged disparate impact.
118. Fifteen claims (27.8%) involved government participation by the Department of Justice or the EEOC, and 39 (72.2%) involved private plaintiffs only.
119. Thirty-nine claims (72.2%) involved blue-collar jobs, while only 15 (27.8%) involved white-collar jobs. Of the blue-collar claims, over half (22, or 56.4%) involved jobs in factories. The remaining blue-collar claims (17, or 43.6%) involved laborer jobs with state or local governmental employers, jobs in the baking or transportation industries, or jobs in law enforcement. My study includes a higher proportion of blue-collar claims than other studies of sex-based employment discrimination cases. See, e.g., Burstein, Attacking Sex Discrimination in the Labor Market: A Study in Law and Politics, 67 Soc. FORCES 641, 648 (1989) (finding that only 12.5% of all federal appellate decisions involved plaintiffs in blue-collar and operative occupations); Mills, supra note 75, at 203 (finding that only 39.8% of a sample of federal trial court decisions involved blue-collar jobs). This finding may suggest that claims challenging sex segregation are more likely to involve blue-collar work than other types of claims, or that employers are more likely to assert the lack of interest defense when blue-collar work is involved.
120. Two of the 15 white-collar claims (13.3%) involved university faculty jobs, another two (13.3%) involved foreign service positions, and one more (6.7%) involved public school administration. The majority of the white-collar claims involved far less prestigious positions as sales agents (six out of 15, or 40%) or low-level supervisors (four out of 15, or 26.7%).
121. I attempted to ascertain from each opinion the race and sex of the individual or group of individuals who initiated the lawsuit (that is, the named plaintiffs in a private class action or any persons who filed charges of discrimination in cases brought by the government). Six of the 54 sex claims (11,1%) appear to have been initiated by minority women, and an equal number (11.1%) by white women. Another claim (1.85%) was initiated by minority women and men, and one more (1.85%) by minority and white women. For the vast majority of the sex claims (40 or 74.1%), however, the opinions fail to disclose the race or ethnicity of any individual women who are identified as having initiated the lawsuit.
122. Only a few other scholars have performed content analyses of employment discrimination cases. None of them has analyzed the relationship between the outcomes and the evidence in the cases, as I do in this study. See Burstein, supra note 80; Eisenberg, Litigation Models and Trial Outcomes in Civil Rights and Prisoner Cases, 77 GEO. L.J. 1567 (1989);
123. Only a few other scholars have performed content analyses of employment discrimination cases. None of them has analyzed the relationship between the outcomes and the evidence in the cases, as I do in this study. See Burstein, supra note 80; Eisenberg, Litigation Models and Trial Outcomes in Civil Rights and Prisoner Cases, 77 GEO. L.J. 1567 (1989);
124. Mills, supra note 75;
125. P. Burstein, Legal Mobilization as a Social Movement Tactic: The Struggle for Equal Employment Opportunity (1988) (unpublished manuscript of a paper presented at the annual meeting of the American Political Science Association) ;
126. P. Burstein & S. Pitchford, Social-Scientific and Legal Challenges to Educational Credential and Testing Requirements in Employment (June 8-11, 1989) (unpublished manuscript of a paper presented at the annual meeting of the Law and Society Association).
127. Throughout this Article, I use the term "conservative" to refer to courts who have accepted the lack of interest argument and the term "liberal" to refer to courts who have rejected that argument.
128. In a forthcoming piece, I analyze how courts have treated the lack of interest argument in race discrimination cases over time. Preliminary results suggest that after a twelve-year period in which courts almost universally rejected employers' attempts to attribute racial segregation to minorities' lack of interest in more remunerative jobs, judges have since 1978 become increasingly more willing to accept this explanation. See V. Schultz & S. Petterson, Privatizing Work Preferences: A Study of Outcomes and Evidentiary Approaches in Title VII Cases Challenging Job Segregation (unpublished manuscript on file with the author). The Supreme Court's recent decision in Wards Cove Packing Co. v. Atonio, 109 S. Ct. 2115 (1989), may provide enhanced legitimacy for the notion that minorities' underrepresentation in higher-paying work reflects their own cultural preferences rather than employer discrimination. See id. at 2122-23 (holding that plaintiffs failed to make a prima facie case of discrimination by showing gross disparities between the proportion of minorities employed in lower-paying and higher-paying jobs, even where the skills required for the two sets of jobs were the same, because most minorities did not seek the higher-paying jobs and plaintiffs failed to prove that the employer's practices discouraged them from doing so); cf. Crenshaw, Race, Reform, and Retrenchment: Transformation and Legitimation in Antidiscrimination Law, 101 HARV. L. REV. 1331, 1379 (1988) (arguing that the new conservative rationalization for black economic disadvantage is "cultural inferiority").
129. In a forthcoming piece, I analyze how courts have treated the lack of interest argument in race discrimination cases over time. Preliminary results suggest that after a twelve-year period in which courts almost universally rejected employers' attempts to attribute racial segregation to minorities' lack of interest in more remunerative jobs, judges have since 1978 become increasingly more willing to accept this explanation. See V. Schultz & S. Petterson, Privatizing Work Preferences: A Study of Outcomes and Evidentiary Approaches in Title VII Cases Challenging Job Segregation (unpublished manuscript on file with the author). The Supreme Court's recent decision in Wards Cove Packing Co. v. Atonio, 109 S. Ct. 2115 (1989), may provide enhanced legitimacy for the notion that minorities' underrepresentation in higher-paying work reflects their own cultural preferences rather than employer discrimination. See id. at 2122-23 (holding that plaintiffs failed to make a prima facie case of discrimination by showing gross disparities between the proportion of minorities employed in lower-paying and higher-paying jobs, even where the skills required for the two sets of jobs were the same, because most minorities did not seek the higher-paying jobs and plaintiffs failed to prove that the employer's practices discouraged them from doing so); cf. Crenshaw, Race, Reform, and Retrenchment: Transformation and Legitimation in Antidiscrimination Law, 101 HARV. L. REV. 1331, 1379 (1988) (arguing that the new conservative rationalization for black economic disadvantage is "cultural inferiority").
130. Specifically, I do not intend to suggest that since 1965, plaintiffs have experienced greater success in cases challenging race segregation than in cases challenging sex segregation. There are defenses other than the lack of interest argument available to employers to defend segregation in their workforces. An employer might contend, for example, that the statistical disparity is attributable to minorities' or women's relative "lack of qualifications." See Wards Cove, 109 S. Ct. at 2122; EEOC v. Sears, Roebuck & Co., 839 F.2d 302, 338-39 (7th Cir. 1988); Segar v. Smith, 738 F.2d 1249, 1274-75 (D.C. Cir. 1984), cert. denied, 471 U.S. 1115 (1985). Although my data do not permit me to test this proposition, employers may have used the lack of qualifications argument more often to justify racial segregation and the lack of interest argument more often to justify sex segregation.
131. This vision seems to be receding from the Supreme Court's view. A number of the Court's recent decisions have cut back on broader approaches to dismantling race- and sex-based employment discrimination developed by the lower courts. See, e.g., Patterson v. McLean Credit Union, 109 S. Ct. 2363 (1989); Wards Cove, 109 S. Ct. 2115; Price Waterhouse v. Hopkins, 109 S. Ct. 1775 (1989); City of Richmond v. J.A. Croson Co., 109 S. Ct. 706 (1989); see also Supreme Court cases cited infra note 318. Nonetheless, the lower courts retain the authority to determine the validity of the lack of interest argument, and the Court's recent decisions do not prevent them from taking an approach similar to the one taken in early race discrimination cases. Indeed, Bazemore v. Friday, 478 U.S. 385 (1986) (per curiam), prohibits the trial courts from rejecting a plaintiff's statistical analyses based solely on the employer's speculative assertion that factors omitted from those analyses might explain the disparity on a neutral basis. See id. at 397-404 (Brennan, J., concurring in part).
132. In all but two of the cases, the courts rejected the lack of interest argument outright. See Carter v. Gallagher, 452 F.2d 315 (8th Cir. 1971), cert. denied, 406 U.S. 950 (1972); Boudreaux v. Baton Rouge Marine Contracting Co., 437 F.2d 1011 (5th Cir. 1971); Parham v. Southwestern Bell Tel. Co., 433 F.2d 421 (8th Cir. 1970); Jones v. Lee Way Motor Freight, Inc., 431 F.2d 245 (10th Cir. 1970), cert. denied, 401 U.S. 954 (1971); United States v. Sheet Metal Workers Int'l Ass'n, Local Union No. 36, 416 F.2d 123 (8th Cir. 1969); United States v. Hayes Int'l Corp., 415 F.2d 1038 (5th Cir. 1969); Cypress v. Newport News Gen. & Nonsectarian Hosp. Ass'n, 375 F.2d 648 (4th Cir. 1967); United States v. Central Motor Lines, 338 F. Supp. 532 (W.D.N.C. 1971); United States v. Local No. 86, Int'l Ass'n of Bridgeworkers, 315 F. Supp. 1202 (W.D. Wash. 1970), aff'd, 443 F.2d 544 (gth Cir.), cert. denied, 404 U.S. 984 (1971); United States v. Plumbers Local 73, 314 F. Supp. 160 (S.D. Ind. 1969). In the two remaining cases, plaintiffs won at least a partial victory. See United States v. H.K. Porter Co., 491 F.2d 1105 (5th Cir. 1974) (vacating the district court's decision accepting the lack of interest explanation when at the court of appeals' prompting, the parties entered a consent decree that substantially revamped the company's hiring practices); Castro v. Beecher, 334 F. Supp. 930, 945 (D. Mass. 1971) (attributing minorities' low rate of application in part to their own lack of interest, but nonetheless ordering the defendant to submit a comprehensive recruiting plan designed to convince minorities that they were "not only welcome but eagerly sought"), aff'd in part and rev'd in part on other grounds, 459 F.2d 725 (1st Cir. 1972).
133. 375 F.2d 648 (4th Cir. 1967).
134. Id. at 653.
135. See id. at 654-55.
136. See id. at 653 nn.7 & 8.
137. See Hill, Myth-Making as Labor History: Herbert Gutman and the United Mine Workers of America, 2 INT'L J. POL., CULTURE & SOC'Y 132, 136 n.* (1988).
138. Of the 12 early race discrimination cases, at least six involved no proof that the employer had ever operated an overt system of racial discrimination. See Carter v. Gallagher, 452 F.2d 315, 323 (8th Cir. 1971), cert. denied, 406 U.S. 950 (1972); Parham v. Southwestern Bell Tel. Co., 433 F.2d 421, 426-27 (8th Cir. 1970); Jones v. Leeway Motor Freight, 431 F.2d 245, 247 (10th Cir. 1970), cert. denied, 401 U.S. 954 (1971); Cypress, 375 F.2d at 653-55; Castro v. Beecher, 334 F. Supp. 930 (D. Mass. 1971), aff'd in part and rev'd in part on other grounds, 459 F.2d 725 (1st Cir. 1972); United States v. Plumbers Local 73, 314 F. Supp. 160, 161-63 (S.D. Ind. 1969).
139. See, e.g., United States v. Sheet Metal Workers Int'l Ass'n, Local Union No. 36, 416 F.2d 123, 139 (8th Cir. 1969); United States v. Central Motor Lines, 338 F. Supp. 532, 551, 559-60 (W.D.N.C. 1971); United States v. Local 86, Int'l Ass'n of Ironworkers, 315 F. Supp. 1202, 1236 (W.D. Wash. 1970), aff'd sub nom. United States v. Ironworkers Local 86, 443 F.2d 544 (9th Cir.), cert. denied, 404 U.S. 984 (1971).
140. Early courts were virtually unanimous in concluding that word-of-mouth recruiting systems were discriminatory because they built upon patterns of racial segregation in non-work realms to perpetuate segregation in the workplace. See, e.g., Parham, 433 F.2d at 427; Sheet Metal Workers, 416 F.2d at 139; Central Motor Lines, 338 F. Supp. at 551, 558, 561-62; Ironworkers, 315 F. Supp. at 1225-26, 1237; Plumbers, 314 F. Supp. at 163-64.
141. See, e.g., Parham, 433 F.2d at 425, 429 (holding that although the telephone company had made progress in hiring blacks, the company was still liable and subject to continuing judicial supervision); Central Motor Lines, 338 F. Supp. at 549-50, 565-66 (holding that although the trucking company had made extensive efforts to recruit black clericals and had recently hired black applicants at a higher rate than whites, the company was still liable and subject to numerical goals and timetables).
142. Sheet Metal Workers, 416 F.2d at 139; accord Jones, 431 F.2d at 248; United States v. Hayes Int'l Corp., 415 F.2d 1038, 1044 (5th Cir. 1969); Ironworkers, 315 F. Supp. at 1235; Plumbers, 314 F. Supp. at 163-64.
143. See, e.g., Jones, 431 F.2d at 247 ("True, no specific instances of discrimination have been shown. However, because of the historically all-white make-up of the Company's [job] category, it may well be that negroes simply did not bother to apply."); Sheet Metal Workers, 416 F.2d at 127 (same); see also Carter v Gallagher, 452 F.2d 315 (8th Cir. 1971) (holding that presence of individual victims of discrimination is not required to justify numerical relief), cert. denied, 406 U.S. 950 (1972).
144. Sheet Metal Workers, 416 F.2d at 139.
145. The Fifth Circuit's decision in Morrow v. Crisler, 491 F.2d 1053 (5th Cir. 1974) (en banc), illustrates this perspective. The court held that the relief granted by the district court was inadequate to eliminate the effects of the Mississippi Highway Patrol's long-standing history of racial discrimination. The Patrol attempted to excuse its record by arguing that it had hired every qualified black person who had applied to be a patrol officer. Rejecting this explanation, the court stated that "[i]f this be true, it is apparent that either the qualifications are discriminatory in effect, or the State has not conducted a sufficient recruitment campaign to enlist blacks who meet those requirements." Id. at 1055-56. The court added: [We] are not sanguine enough to be of the view that benign recruitment programs can purge in two years a reputation which discriminatory practices of approximately 30 years have entrenched in the minds of blacks in Mississippi. . . . Imaginative initiative is needed from all present members of the Highway Patrol. Id. at 1056-57.
146. For three eloquent descriptions of how law can express and expand the aspirations of disempowered groups, see Hartog, The Constitution of Aspiration and "The Rights That Belong to Us All," in THE CONSTITUTION AND AMERICAN LIFE 353 (D. Thelen ed. 1988); Schneider, The Dialectic of Rights and Politics: Perspectives From the Women's Movement, 61 N.Y.U. L. REV. 589 (1986); and Williams, Alchemical Notes: Reconstructing Ideals from Deconstructed Rights, 22 HARV. C.R.-C.L. L. REV. 401 (1987).
147. For three eloquent descriptions of how law can express and expand the aspirations of disempowered groups, see Hartog, The Constitution of Aspiration and "The Rights That Belong to Us All," in THE CONSTITUTION AND AMERICAN LIFE 353 (D. Thelen ed. 1988); Schneider, The Dialectic of Rights and Politics: Perspectives From the Women's Movement, 61 N.Y.U. L. REV. 589 (1986); and Williams, Alchemical Notes: Reconstructing Ideals from Deconstructed Rights, 22 HARV. C.R.-C.L. L. REV. 401 (1987).
148. For three eloquent descriptions of how law can express and expand the aspirations of disempowered groups, see Hartog, The Constitution of Aspiration and "The Rights That Belong to Us All," in THE CONSTITUTION AND AMERICAN LIFE 353 (D. Thelen ed. 1988); Schneider, The Dialectic of Rights and Politics: Perspectives From the Women's Movement, 61 N.Y.U. L. REV. 589 (1986); and Williams, Alchemical Notes: Reconstructing Ideals from Deconstructed Rights, 22 HARV. C.R.-C.L. L. REV. 401 (1987).
149. This early race discrimination law may, however, have taken sex segregation as a given. Most of the cases involved jobs held by white men, and the courts probably did not envision women of color as victims among those who had been discriminatorily denied these jobs. In one case the Second Circuit vacated a 30% hiring goal because it was based on the entire population of minorities in the local area, and thus included minority women. Noting casually that "women have never sought to become steamfitters," the court held: "Absent racial discrimination, . . . the non-white members of the Union would have been drawn from the male workforce over 18 years of age in the Union's jurisdiction." Rios v. Enterprise Ass'n Steamfitters Local 638, 501 F.2d 622, 632 (2d Cir. 1974) (emphasis added). In other cases, minority women are mentioned only in connection with clerical jobs. See, e.g., United States v. Central Motor Lines, 338 F. Supp. 532, 548-52 (W.D.N.C. 1971). A burgeoning new literature discusses other examples of the legal "erasure" of women of color. See, e.g., Austin, Sapphire Bound!, 1989 WIS. L. REV. 539; Crenshaw, Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of Antidiscrimination Doctrine, Feminist Theory, and Antiracist Politics, 1989 U. CHI. LEGAL F. 139; Scales-Trent, Black Women and the Constitution: Finding Our Place: Asserting Our Rights, 24 HARV. C.R.-C.L. L. REV. 9 (1989).
150. This early race discrimination law may, however, have taken sex segregation as a given. Most of the cases involved jobs held by white men, and the courts probably did not envision women of color as victims among those who had been discriminatorily denied these jobs. In one case the Second Circuit vacated a 30% hiring goal because it was based on the entire population of minorities in the local area, and thus included minority women. Noting casually that "women have never sought to become steamfitters," the court held: "Absent racial discrimination, . . . the non-white members of the Union would have been drawn from the male workforce over 18 years of age in the Union's jurisdiction." Rios v. Enterprise Ass'n Steamfitters Local 638, 501 F.2d 622, 632 (2d Cir. 1974) (emphasis added). In other cases, minority women are mentioned only in connection with clerical jobs. See, e.g., United States v. Central Motor Lines, 338 F. Supp. 532, 548-52 (W.D.N.C. 1971). A burgeoning new literature discusses other examples of the legal "erasure" of women of color. See, e.g., Austin, Sapphire Bound!, 1989 WIS. L. REV. 539; Crenshaw, Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of Antidiscrimination Doctrine, Feminist Theory, and Antiracist Politics, 1989 U. CHI. LEGAL F. 139; Scales-Trent, Black Women and the Constitution: Finding Our Place: Asserting Our Rights, 24 HARV. C.R.-C.L. L. REV. 9 (1989).
151. This early race discrimination law may, however, have taken sex segregation as a given. Most of the cases involved jobs held by white men, and the courts probably did not envision women of color as victims among those who had been discriminatorily denied these jobs. In one case the Second Circuit vacated a 30% hiring goal because it was based on the entire population of minorities in the local area, and thus included minority women. Noting casually that "women have never sought to become steamfitters," the court held: "Absent racial discrimination, . . . the non-white members of the Union would have been drawn from the male workforce over 18 years of age in the Union's jurisdiction." Rios v. Enterprise Ass'n Steamfitters Local 638, 501 F.2d 622, 632 (2d Cir. 1974) (emphasis added). In other cases, minority women are mentioned only in connection with clerical jobs. See, e.g., United States v. Central Motor Lines, 338 F. Supp. 532, 548-52 (W.D.N.C. 1971). A burgeoning new literature discusses other examples of the legal "erasure" of women of color. See, e.g., Austin, Sapphire Bound!, 1989 WIS. L. REV. 539; Crenshaw, Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of Antidiscrimination Doctrine, Feminist Theory, and Antiracist Politics, 1989 U. CHI. LEGAL F. 139; Scales-Trent, Black Women and the Constitution: Finding Our Place: Asserting Our Rights, 24 HARV. C.R.-C.L. L. REV. 9 (1989).
152. Throughout this section I report the results from tests of statistical significance. Because my search strategy was designed to locate every published decision that addressed the lack of interest argument, the collection of claims in my study approximates a universe rather than a sample. Strictly speaking, it is unnecessary to test for statistical significance of the relationship between variables in a universe, as opposed to a sample, of cases. I report p-values, however, to give the reader some sense of the relative importance of the variables under consideration. The p-values may be particularly useful in connection with the analyses which show the relationship between case outcomes and certain types of evidence. A p-value of .05 or less means that the probability that the outcomes would have differed as much as they did across cases where that evidence was present and absent, by chance alone, is one in 20 or less. Similarly, a p-value of .10 or less shows that the same probability is one in 10 or less. The reported p-values were calculated using chi-square tests, unless the numbers were too small. Where any of the expected cell values in the contingency table was less than five, a Yates correction was used. Where the total size of the sample was less than 20, a Fisher's Exact Test was used. For a very readable description of these tests, see W. MENDENHALL, L. OTT & R. LARSON, STATISTICS: A TOOL FOR THE SOCIAL SCIENCES 321-36 (1974).
153. The early period begins in 1972, when the first sex discrimination case raising the lack of interest argument was decided, and continues through 1977, when the Supreme Court decided Teamsters, Dothard, and Hazelwood. The modern period spans from 1978 through 1989, when the last sex discrimination case in my data set was decided.
154. The outcomes vary only slightly in individual as opposed to classwide claims of discrimination. Plaintiffs prevailed on the lack of interest argument in 50% (four out of eight) of the individual claims and in 58.7% (27 out of 46) of the classwide claims (p = 1.0).
155. p = 1.0.
156. Some theoretical work in the sociology of law predicts that disadvantaged groups will tend to lose legal confrontations with powerful groups having greater resources and better organizational capability. The classic article is Marc Galanter's Why the 'Haves' Come Out Ahead: Speculations on the Limits of Legal Change, 9 LAW & SOC'Y REV. 95 (1974). Other models of litigation predict that under certain conditions, plaintiffs will win approximately 50% of all cases tried to judgment, regardless of whether the legal standard is favorable or unfavorable to them. See, e.g., Priest & Klein, supra note 71, at 4-5. This work suggests caution in interpreting plaintiffs' success rates in tried cases as evidence of the favor or disfavor with which they are treated by the legal system, because these success rates do not reveal how the parties fared in cases settled before trial. See id. For an empirical study examining the extent to which these models explain success rates in employment discrimination litigation, see Eisenberg, cited above in note 83.
157. In the only comprehensive study of employment discrimination cases that reports separate results for sex discrimination cases, Burstein found that plaintiffs won at least a partial victory in 58% percent of the 672 cases decided by federal appellate courts and published in the Fair Employment Practices Reporter between 1963 and 1985. See Burstein, supra note 80, at 655-57; see also Mills, supra note 75, at 206 (finding that plaintiffs won 32% of all sex-based employment discrimination cases decided by federal district courts in five circuits between 1967 and 1975); cf. Eisenberg, supra note 83, at 1578 (finding that plaintiffs won 22% of employment discrimination cases decided by the federal district courts between 1978 and 1985, but reporting no separate results for sex discrimination cases). For a more detailed comparison of success rates in race and sex discrimination cases raising the lack of interest argument in various historical periods, see V. Schultz & S. Petterson, cited above in note 85, at 38-52.
158. Those who spoke in favor of the amendment adding the prohibition against sex discrimination to the original 1964 Act focused primarily on the injustice of sex segregation in the labor market. See, e.g., 110 CONG. REC. 2579-80, 2580-81 (1964) (remarks of Reps. Griffiths and St, George). Moreover, when Congress amended title VII in 1972, both the House and the Senate made clear that they considered sex segregation to be the primary evil that the statute was designed to address. See H.R. REP. No. 92-238, 92d Cong., ist Sess. 4-5, reprinted in 1972 U.S. CODE CONG. & ADMIN. NEWS 2137, 2140 [hereinafter HOUSE REPORT] ("[W]omen are placed in the less challenging, the less responsible and the less remunerative positions on the basis of their sex alone. Such blatantly disparate treatment is particularly objectionable in view of the fact that Title VII has specifically prohibited sex discrimination since its enactment in 1964."); S. REP. No. 415, 92nd Cong., 1st Sess. 7 (1971) [hereinafter SENATE REPORT] (including similar statements).
159. 111 Even those who take the position that title VII should not be interpreted to require pay equity argue that the way to eliminate the male-female wage disparity is to enforce title VII and the Equal Pay Act of 1963, 29 U.S.C. § 206(d) (1982), vigorously to ensure that women are integrated into nontraditional jobs and that they are paid the same as men in those jobs. See, e.g., American Nurses' Ass'n v. Illinois, 783 F.2d 716, 730 (7th Cir. 1986); Abram, Concurring Statement of Morris B. Abram, in U.S. COMM'N ON CIVIL RIGHTS, COMPARABLE WORTH: AN ANALYSIS AND RECOMMENDATIONS 76, 77 (1985) [hereinafter COMPARABLE WORTH]; Pendleton, Concurring Statement of Clarence Pendleton, Jr., in COMPARABLE WORTH, supra, at 73, 75.
160. 111 Even those who take the position that title VII should not be interpreted to require pay equity argue that the way to eliminate the male-female wage disparity is to enforce title VII and the Equal Pay Act of 1963, 29 U.S.C. § 206(d) (1982), vigorously to ensure that women are integrated into nontraditional jobs and that they are paid the same as men in those jobs. See, e.g., American Nurses' Ass'n v. Illinois, 783 F.2d 716, 730 (7th Cir. 1986); Abram, Concurring Statement of Morris B. Abram, in U.S. COMM'N ON CIVIL RIGHTS, COMPARABLE WORTH: AN ANALYSIS AND RECOMMENDATIONS 76, 77 (1985) [hereinafter COMPARABLE WORTH]; Pendleton, Concurring Statement of Clarence Pendleton, Jr., in COMPARABLE WORTH, supra, at 73, 75.
161. 111 Even those who take the position that title VII should not be interpreted to require pay equity argue that the way to eliminate the male-female wage disparity is to enforce title VII and the Equal Pay Act of 1963, 29 U.S.C. § 206(d) (1982), vigorously to ensure that women are integrated into nontraditional jobs and that they are paid the same as men in those jobs. See, e.g., American Nurses' Ass'n v. Illinois, 783 F.2d 716, 730 (7th Cir. 1986); Abram, Concurring Statement of Morris B. Abram, in U.S. COMM'N ON CIVIL RIGHTS, COMPARABLE WORTH: AN ANALYSIS AND RECOMMENDATIONS 76, 77 (1985) [hereinafter COMPARABLE WORTH]; Pendleton, Concurring Statement of Clarence Pendleton, Jr., in COMPARABLE WORTH, supra, at 73, 75.
162. Employers sometimes try to substantiate the lack of interest argument with data showing that women have underapplied for the work relative to their representation in the relevant labor market. But as Dothard, Teamsters, and the early race discrimination cases teach, applicant data do not "prove" anything about women's relative interest in the work. One needs to know why women were less likely than men to apply.
163. The only other defense that draws on these sorts of broad generalizations is the bona fide occupational qualification (BFOQ) defense permitted under § 703(e) of title VII, 42 U.S.C. § 2000e-2(e) (1982). With the BFOQ defense, the employer seeks to justify an overt policy of excluding all women by proving that "all or substantially all women would be unable to perform safely and efficiently the duties of the job involved," Weeks v. Southern Bell Tel. & Tel. Co., 408 F.2d 228, 235 (5th Cir. 1969), or that "the essence of the business operation would be undermined by not hiring members of one sex exclusively." Diaz v. Pan Am. World Airways, 442 F.2d 385, 388 (5th Cir. 1971) (emphasis in original), cert. denied, 404 U.S. 950 (1971); accord Dothard v. Rawlinson, 433 U.S. 321, 333 (1977) (citing both the Weeks and Diaz formulations approvingly). Although there are no studies showing how often employers succeed in establishing the BFOQ defense, the perception is that the courts have accepted it only infrequently. See, e.g., B. SCHLEI & P. GROSSMAN, EMPLOYMENT DISCRIMINATION LAW 341, 353 (2d ed. 1983). Courts have held that the BFOQ defense is to be interpreted narrowly precisely because it draws upon the sort of broad generalizations about women that title VII was intended to prohibit. See, e.g., Dothard, 433 U.S. at 333-34; Diaz, 442 F.2d at 387; Weeks, 408 F.2d at 235.
164. See, e.g., Piva v. Xerox Corp., 654 F.2d 591, 595 (gth Cir. 1981); Hill v. Western Elec. Co., 12 Fair Empl. Prac. Cas. (BNA) 1175, 1179-80 (E.D. Va. 1976), rev'd on other grounds, 596 F.2d 99 (4th Cir.), cert. denied, 444 U.S. 929 (1979).
165. See, e.g., EEOC v. Sears, Roebuck & Co., 628 F. Supp. 1264, 1305 (N.D. Ill. 1986), aff'd, 839 F.2d 302 (7th Cir. 1988); EEOC v. Mead Foods, Inc., 477 F. Supp. 1, 3 (W.D. Okla. 1977); Logan v. General Fireproofing Co., 6 Fair Empl. Prac. Cas. (BNA) 140, 144 (W.D.N.C. 1972).
166. p = 1.0.
167. Indeed, in Sears, defense lawyer Charles Morgan went so far as to argue in his summation that it is ridiculous to suggest that sex discrimination even exists. "Strange, isn't it," said Morgan, "that we live in a world where there is supposed to be a monopoly of white men who somehow get up every morning trying to find a way to discriminate against their wives, their daughters, their mothers, their sisters." Closing Argument, Trial Transcript at 19,064, EEOC v. Sears, Roebuck & Co., 628 F. Supp. 1264 (N.D. Ill. 1986) (No. 79-C-4373). Of course, if sex discrimination does not exist, it follows that women have chosen their lot in life, including their "place" at work. That many judges have condoned such arguments suggests that they may share with employers and defense attorneys the dominant cultural image of women as marginal workers. As a counter-vision, one can imagine a world in which judges would consider employers' willingness to argue that women "choose" to segregate themselves into the lowest-paying, most menial jobs as itself evidence of a sexually discriminatory attitude. Cf. Arnold v. Ray, 21 Fair Empl. Prac. Cas. (BNA) 793, 795 (N.D. Ohio 1979) (finding that the "most telling piece of evidence" of intentional discrimination was the mayor's own testimony that blacks "don't want to be policem[e]n").
168. The "neither party wins" category includes four cases in which the court of appeals remanded the underlying claim to the trial court for further consideration so that neither party achieved a clear victory, as well as two cases in which the interest issue was raised during the remedial phase of the litigation so that the discrimination claim was no longer at issue. In the two remedy cases, the employers urged women's lack of interest in an attempt to persuade the courts that there was no need to impose numerical relief. See United States v. Virginia, 22 Fair Empl. Prac. Cas. (BNA) 936, 939 (E.D. Va. 1978); Jordan v. Wright, 417 F. Supp. 42, 45 (M.D. Ala. 1976).
169. p = .0000.
170. This is only one possible explanation for the observed pattern. In some cases, it may be that there simply were no plausible alternative explanations for women's underrepresentation; pre-trial discovery may have shown, for example, that the women were as well qualified as the men. In other cases, employers may have asserted defenses other than women's lack of interest, but the same evidence and attitudes that led courts to reject the interest argument led them to reject the alternative explanations as well. See, e.g., Kilgo v. Bowman Transp. Co., 570 F. Supp. 1509, 1517, 1526-28 (N.D. Ga. 1983) (rejecting the lack of qualifications defense on the ground that the company's experience requirement was a pretext for discrimination intended to discourage females from applying, and finding that this and other evidence of discouragement also invalidated the lack of interest defense), aff'd, 789 F. 2d 859 (11th Cir. 1986).
171. p = .07.
172. The relationship between evidence of past discrimination and outcome observed in Table 3 remains even after controlling for the presence or absence of anecdotal evidence of discrimination. In cases in which anecdotal evidence was introduced, plaintiffs won only 61.9% (13 out of 21) of the cases in which they presented evidence of past discrimination, compared to 81,8% (nine out of 11) of the cases in which they did not. Similarly, in cases in which no anecdotal evidence was introduced, plaintiffs won only 28.6% (four out of 14) of the cases in which they presented evidence of past discrimination, compared to 62.5% (five out of eight) of the cases in which they did not. The relationship between evidence of past discrimination and outcome in Table 3 also remains after controlling for the presence or absence of evidence of special efforts. In cases in which evidence of special efforts was introduced, plaintiffs won only 28.6% (four out of 14) of the cases in which they presented evidence of past discrimination, compared to 71.4% (five out of seven) of the cases in which they did not. Similarly, in cases in which no evidence of special efforts was introduced, plaintiffs won only 61.9% (13 out of 21) of the cases in which they presented evidence of past discrimination, compared to 75.0% (nine out of 12) of the cases in which they did not. These analyses do not control for possible interaction among all three types of evidence. A separate logistic regression analysis showed, however, that even after controlling simultaneously for anecdotal evidence and evidence of special efforts, evidence of past discrimination still decreased plaintiffs' likelihood of winning. The T-ratio for the evidence of past discrimination variable was -1.837, with a corresponding p-value of .07.
173. Of the 11 claims for which plaintiffs presented direct or anecdotal evidence of past discrimination, the courts found in all 11 (100%) that the employer had engaged in past discrimination. But of the 24 claims for which plaintiffs relied on statistical evidence alone, courts found in only 11 (45.8%) that the employer had engaged in past discrimination (p = .0069). Statistical evidence of past discrimination is no different from statistical evidence of current discrimination, except that it covers the time preceding the liability period. Typically, the plaintiff will show that, until a certain date, the employer hired no women at all (or only a few women) for nontraditional jobs.
174. More precisely, all 11 (100%) of the courts who inferred past employer discrimination from statistical evidence alone rejected the interest argument, and 12 out of 13 (92.3%) of those who refused to infer past discrimination from statistical evidence accepted that argument (p = .0000).
175. I use the term "conservative" here to refer generally to courts that have accepted the lack of interest argument. I also use it more specifically here to refer to courts that refused to infer past employer discrimination from statistical evidence. Although not all courts that accepted the interest argument were presented with statistical evidence of past discrimination, part of what leads courts to attribute present patterns of segregation to women's choice is a vision of an uninterrupted history in which women have always been free to choose whatever work they wanted.
176. 466 F. Supp. 1 (W.D. Okla. 1977).
177. Id. at 3-4.
178. Id. at 2-3.
179. Id. at 3 (emphasis added).
180. 17 Fair Empl. Prac. Cas. (BNA) 954 (D.S.C. 1978), aff'd on other grounds and remanded, 662 F.2d 256 (4th Cir. 1981).
181. See id. at 959.
182. The company had never hired a single woman for its hardwood division and had employed women in only five of nine job categories in its cabinet division. See id. at 958.
183. Id. at 959. The court's refusal to find sex discrimination is particularly disturbing because the EEOC had alleged not only that the company was segregating its jobs by race and sex, but also that it was discriminating specifically against black women. See id. at 955. The court's rulings prohibited the company from denying black men access to the higher-graded jobs held by white men, but they left the company free to exclude black (and other) women from the same jobs.
184. I am using the term "liberal" here to refer to courts who have rejected the lack of interest argument, as I do throughout the Article. As shown earlier, though, there is a significant relationship between judges' willingness to reject the lack of interest explanation and their willingness to infer past employer discrimination from statistical evidence alone. See supra note 124 and accompanying text.
185. Only one court has explicitly invoked the principle that an employer has an affirmative duty to correct for past discrimination by attracting women to the jobs from which they formerly had been excluded. See Chrapliwy v. Uniroyal, Inc., 458 F. Supp. 252, 262 (N.D. Ind. 1977). In a few additional cases, courts did not invoke this affirmative duty, but nonetheless seemed to place women's current interest in nontraditional jobs in the context of the employer's historical discrimination. See Harless v. Duck, 619 F.2d 611, 618 (6th Cir.) ("[The police department's] policies had been engraved in stone since the 1920's. Every woman on the police force and every woman interested in becoming a member of [the force] knew her restricted role."), cert. denied, 449 U.S. 872 (1980); Thompson v. Boyle, 499 F. Supp. 1147, 1150, 1161-62 (D.D.C. 1979) (citing the history of sex discrimination in the bookbinder industry to explain women's diminished interest in the craft), aff'd sub nom. Thompson v. Sawyer, 678 F.2d 257 (D.C. Cir. 1982)
186. 711 F.2d 647 (5th Cir. 1983), cert. denied, 466 U.S. 927 (1984).
187. See id. at 662.
188. See id. at 653, 660-61. The company tried to excuse its sexually discriminatory advertising by emphasizing that a female personnel officer had placed the ads and that she had not intended to discriminate against women. When asked at trial why she placed the ads in the male-only section of the newspaper, the personnel manager replied: "'Well, that was because managers were always males. So, I put it in the male column.'" Id. at 660.
189. Between 1973 and 1977, women were only 9.2% of those the company hired as managers. See id. at 652. By the Fifth Circuit's own account, "[t]he results of [plaintiff's statistical] tests . . . consistently showed the probability of such disparate hiring occurring by chance to be less than one in 10,000." Id.
190. Indeed, the court emphasized that the company had hired zero women during the early 1965-1972 period, and observed that hiring even two or three women would have gone a long way toward showing the company's "willingness to consider women as equals in firm management." Id. at 662.
191. 828 F.2d 1260 (8th Cir. 1987), cert. denied, 485 U.S. 1021 (1988).
192. See Catlett v. Missouri Highway & Transp. Comm'n, 589 F. Supp. 929, 933-34 (W.D. Mo. 1983), aff'd in part and rev'd in part on other grounds, 828 F.2d 1260 (8th Cir. 1987), cert. denied, 485 U.S. 1021 (1988).
193. See Catlett, 828 F.2d at 1266.
194. Evidence in the record directly tied the low number of female applicants to the state's historical hiring practices. With each additional woman hired, the state experienced an increase in applications from women. In 1975, before the state had hired a single woman as a road maintenance worker, only nine women applied for the job. In 1976, when the state hired its first woman, the number of women applicants rose to 16. In 1977, when the State hired two more women, the number of women applicants rose to 55. In 1978, when the State hired five more women, the number of women applicants rose to 120. See Catlett, 589 F. Supp. at 934.
195. See Catlett, 828 F.2d at 1268-69.
196. The district court had explicitly rejected this position: "While defendants argue that their voluntary efforts in changing their hiring practices are enough to preclude additional affirmative relief, the Court finds the long history of discrimination and the comparatively short history of attempts to end such discrimination warrant further measures . . . ." Catlett, 589 F. Supp. at 1020.
197. Indeed, one cannot always distinguish clearly between the two approaches in judicial decisions. There is slippage between the view that women aspire to "women's work" because they are women and the view that women aspire to "women's work" because society has socialized them to be women. The allegedly more historical liberal approach converges with the ahistorical, conservative approach insofar as both approaches attribute the historical change in women's work aspirations to "societal" forces that are so vague and ill-identified as to constitute essentialist explanations for women's experience and work consciousness. This slippage is analogous to that observed in some cultural feminist work. See Schultz, Room to Maneuver (f)or a Room of One's Own? Practice Theory and Feminist Practice, 14 LAW & Soc, INQUIRY 123, 128 (1989); Scott, Gender: A Useful Category of Historical Analysis, 91 AM. HIST. REV. 1053, 1065 (1986).
198. Indeed, one cannot always distinguish clearly between the two approaches in judicial decisions. There is slippage between the view that women aspire to "women's work" because they are women and the view that women aspire to "women's work" because society has socialized them to be women. The allegedly more historical liberal approach converges with the ahistorical, conservative approach insofar as both approaches attribute the historical change in women's work aspirations to "societal" forces that are so vague and ill-identified as to constitute essentialist explanations for women's experience and work consciousness. This slippage is analogous to that observed in some cultural feminist work. See Schultz, Room to Maneuver (f)or a Room of One's Own? Practice Theory and Feminist Practice, 14 LAW & Soc, INQUIRY 123, 128 (1989); Scott, Gender: A Useful Category of Historical Analysis, 91 AM. HIST. REV. 1053, 1065 (1986).
199. For examples of recent work documenting the history of labor market discrimination against women, see P. FONER, cited above in note 20; M. GREENWALD, WOMEN, WAR, AND WORK: THE IMPACT OF WORLD WAR I ON WOMEN WORKERS IN THE UNITED STATES (1980); J. JONES, cited above in note 17; S. KENNEDY, cited above in note 19; A. KESSLER-HARRIS, cited above in note 19; and R. MILKMAN, GENDER AT WORK: THE DYNAMICS OF JOB SEGREGATION BY SEX DURING WORLD WAR II (1987).
200. For examples of recent work documenting the history of labor market discrimination against women, see P. FONER, cited above in note 20; M. GREENWALD, WOMEN, WAR, AND WORK: THE IMPACT OF WORLD WAR I ON WOMEN WORKERS IN THE UNITED STATES (1980); J. JONES, cited above in note 17; S. KENNEDY, cited above in note 19; A. KESSLER-HARRIS, cited above in note 19; and R. MILKMAN, GENDER AT WORK: THE DYNAMICS OF JOB SEGREGATION BY SEX DURING WORLD WAR II (1987).
201. For examples of recent work documenting the history of labor market discrimination against women, see P. FONER, cited above in note 20; M. GREENWALD, WOMEN, WAR, AND WORK: THE IMPACT OF WORLD WAR I ON WOMEN WORKERS IN THE UNITED STATES (1980); J. JONES, cited above in note 17; S. KENNEDY, cited above in note 19; A. KESSLER-HARRIS, cited above in note 19; and R. MILKMAN, GENDER AT WORK: THE DYNAMICS OF JOB SEGREGATION BY SEX DURING WORLD WAR II (1987).
202. For examples of recent work documenting the history of labor market discrimination against women, see P. FONER, cited above in note 20; M. GREENWALD, WOMEN, WAR, AND WORK: THE IMPACT OF WORLD WAR I ON WOMEN WORKERS IN THE UNITED STATES (1980); J. JONES, cited above in note 17; S. KENNEDY, cited above in note 19; A. KESSLER-HARRIS, cited above in note 19; and R. MILKMAN, GENDER AT WORK: THE DYNAMICS OF JOB SEGREGATION BY SEX DURING WORLD WAR II (1987).
203. For examples of recent work documenting the history of labor market discrimination against women, see P. FONER, cited above in note 20; M. GREENWALD, WOMEN, WAR, AND WORK: THE IMPACT OF WORLD WAR I ON WOMEN WORKERS IN THE UNITED STATES (1980); J. JONES, cited above in note 17; S. KENNEDY, cited above in note 19; A. KESSLER-HARRIS, cited above in note 19; and R. MILKMAN, GENDER AT WORK: THE DYNAMICS OF JOB SEGREGATION BY SEX DURING WORLD WAR II (1987).
204. I borrow this phrase from Richard Delgado, who borrowed it from Milner Ball. See Delgado, On Taking Back Our Civil Rights Promises: When Equality Doesn't Compute, 1989 WIS. L. REV. 579, 580 & n.6 (citing Ball, Stories of Origins and Constitutional Possibilities, 87 MICH. L. REV. 2280 (1989)). Here the "story of origins" refers to the constitutive myths and ideals we ascribe both to title VII's ban on sex discrimination, and to ourselves as a nation and as a people in having enacted the legislation.
205. I borrow this phrase from Richard Delgado, who borrowed it from Milner Ball. See Delgado, On Taking Back Our Civil Rights Promises: When Equality Doesn't Compute, 1989 WIS. L. REV. 579, 580 & n.6 (citing Ball, Stories of Origins and Constitutional Possibilities, 87 MICH. L. REV. 2280 (1989)). Here the "story of origins" refers to the constitutive myths and ideals we ascribe both to title VII's ban on sex discrimination, and to ourselves as a nation and as a people in having enacted the legislation.
206. See, e.g., C. BIRD, BORN FEMALE (1968); A. SMITH, EMPLOYMENT DISCRIMINATION LAW 327 (1978); Developments in the Law - Employment Discrimination and Title VII of the Civil Rights Act of 1964, 84 HARV. L. REV. 1109, 1167 (1971). At least one scholar has challenged this view. See Gold, A Tale of Two Amendments: The Reasons Congress Added Sex to Title VII and Their Implication for the Issue of Comparable Worth, 19 DUQ. L. REV. 453, 457-69 (1981) (arguing that even if members of Congress who introduced the amendment adding sex did so for the purpose of trying to defeat the bill, the majority of Congress did not have that motivation when they approved the amendment).
207. See, e.g., C. BIRD, BORN FEMALE (1968); A. SMITH, EMPLOYMENT DISCRIMINATION LAW 327 (1978); Developments in the Law - Employment Discrimination and Title VII of the Civil Rights Act of 1964, 84 HARV. L. REV. 1109, 1167 (1971). At least one scholar has challenged this view. See Gold, A Tale of Two Amendments: The Reasons Congress Added Sex to Title VII and Their Implication for the Issue of Comparable Worth, 19 DUQ. L. REV. 453, 457-69 (1981) (arguing that even if members of Congress who introduced the amendment adding sex did so for the purpose of trying to defeat the bill, the majority of Congress did not have that motivation when they approved the amendment).
208. See, e.g., C. BIRD, BORN FEMALE (1968); A. SMITH, EMPLOYMENT DISCRIMINATION LAW 327 (1978); Developments in the Law - Employment Discrimination and Title VII of the Civil Rights Act of 1964, 84 HARV. L. REV. 1109, 1167 (1971). At least one scholar has challenged this view. See Gold, A Tale of Two Amendments: The Reasons Congress Added Sex to Title VII and Their Implication for the Issue of Comparable Worth, 19 DUQ. L. REV. 453, 457-69 (1981) (arguing that even if members of Congress who introduced the amendment adding sex did so for the purpose of trying to defeat the bill, the majority of Congress did not have that motivation when they approved the amendment).
209. See, e.g., C. BIRD, BORN FEMALE (1968); A. SMITH, EMPLOYMENT DISCRIMINATION LAW 327 (1978); Developments in the Law - Employment Discrimination and Title VII of the Civil Rights Act of 1964, 84 HARV. L. REV. 1109, 1167 (1971). At least one scholar has challenged this view. See Gold, A Tale of Two Amendments: The Reasons Congress Added Sex to Title VII and Their Implication for the Issue of Comparable Worth, 19 DUQ. L. REV. 453, 457-69 (1981) (arguing that even if members of Congress who introduced the amendment adding sex did so for the purpose of trying to defeat the bill, the majority of Congress did not have that motivation when they approved the amendment).
210. For sources showing that Congress considered sex segregation to be a primary evil that title VII was intended to address, see note no above. For sources showing that Congress intended the courts to treat sex discrimination with the same seriousness as discrimination on other grounds, see SENATE REPORT, cited above in note no, at 7 ("While some have looked at . . . women's rights as a frivolous divertissement, this Committee believes that discrimination against women is no less serious than other prohibited forms of discrimination, and that it is to be accorded the same degree of concern given to any type of similarly unlawful conduct."); and HOUSE REPORT, cited above in note no, at 5, which contains similar statements. See also Barnes v. Costle, 561 F.2d 983, 987 (D.C. Cir. 1977) (concluding that "Congress was deeply concerned about employment discrimination based on gender, and intended to combat it as vigorously as any other type of forbidden discrimination").
211. By refusing to infer from statistical evidence that employers even engaged in past discrimination, many courts have eliminated the factual foundation for invoking the futility doctrine and the accompanying affirmative duty in sex discrimination cases. In some cases, judges have accompanied their refusal to find past discrimination with explicit announcements that title VII does not require employers to make any efforts to recruit women. See, e.g., Gilbert v. East Bay Mun. Dist., 19 Fair Empl. Prac. Cas. (BNA) 304, 307 (N.D. Cal. 1979) ("Defendant was under no obligation to affirmatively solicit female applicants for these supervisory positions."); EEOC v. Korn Indus., 17 Fair Empl. Prac. Cas. (BNA) 954, 959 (D.S.C. 1978) ("There is no obligation on the defendant to waste its time and money trying to find women of extraordinary size, strength, and stamina."), aff'd on other grounds and remanded, 662 F.2d 256 (4th Cir. 1981). Some courts have gone even farther and held that even if the employer discriminated against women in the recent past, title VII still does not require the employer to correct for that discrimination by taking affirmative steps to attract women to formerly segregated jobs. See, e.g., Ste. Marie v. Eastern R.R. Ass'n, 650 F.2d 395, 403-04 (2d Cir. 1981); Lewis v. Tobacco Workers' Int'l Union, 577 F.2d 1135, 1146-47 (4th Cir. 1978), cert. denied, 439 U.S. 1989 (1979).
212. p = .08. A logistic regression analysis showed that, even after controlling simultaneously for evidence of past discrimination and anecdotal evidence, evidence of special efforts increased employers' likelihood of prevailing. (The T-ratio for the evidence of special efforts variable was -1.524, with a corresponding p-value of .13.)
213. Of the 21 claims of special efforts, in only seven (33.3%) had the employers promulgated written affirmative action plans that appeared to include goals and timetables for women. See, e.g., EEOC v. Sears, Roebuck & Co., 628 F. Supp. 1264, 1292-94 (N.D. Ill. 1986), aff'd, 839 F.2d 302 (7th Cir. 1988); EEOC v. Cook Paint & Varnish Co., 24 Fair Empl. Prac. Cas. (BNA) 51, 52 (W.D. Mo. 1980); United States v. County of Fairfax, 19 Fair Empl. Prac. Cas. (BNA) 753, 758, 760 (E.D. Va. 1979), vacated on other grounds, 629 F.2d 932 (4th Cir. 1980), cert. denied, 449 U.S. 1979 (1981).
214. 601 F. Supp. 1377 (E.D. Ark. 1985).
215. See id. at 1378-79.
216. See id. at 1379.
217. Id. at 1385 (quoting the company's affirmative action plan).
218. See id.
219. For another case in which a court was able to conclude easily that the employer's affirmative action plan was woefully lacking in substance, see Capaci v. Katz & Besthoff, Inc., 711 F.2d 647 (5th Cir. 1983), cert, denied, 466 U.S. 927 (1984). Capaci held that the company's alleged affirmative action program did little to dispel the inference of sex discrimination, where the company's president testified that "[t]he sex aspect of the discrimination was passed as a part of the Civil Rights Act as an afterthought" and offered few specifics other than a written statement of good faith intentions to treat women on a nondiscriminatory basis. Id. at 657-58.
220. In a few cases, employers presented data showing that women had applied or bid for the jobs at issue at disproportionately lower rates than men. See, e.g., Durant v. Owens-Illinois Glass Co., 517 F. Supp. 710, 716 (E.D. La. 1980), aff'd, 656 F.2d 89 (5th Cir. 1981); Davis v. City of Dallas, 483 F. Supp. 54, 61 (N.D. Tex. 1979). Applicant data merely beg the question, however, for they do not prove whether women underapplied because of lack of interest or because the employer's discrimination discouraged them.
221. See, e.g., Ste. Marie v. Eastern R.R. Ass'n, 650 F.2d 395, 403 (2d Cir. 1981); EEOC v. Mead Foods, Inc., 466 F. Supp. 1, 3-4 (W.D. Okla. 1977).
222. There are a variety of networks that employers might use to target their recruiting efforts specifically toward women. All over the country, there are community-based programs designed to help women enter nontraditional employment. See Law, "Girls Can't Be Plumbers" - Affirmative Action for Women in Construction: Beyond Goals and Quotas, 24 HARV. C.R.-C.L. L. REV. 45, 53-55 & n.33 (1989) (describing such organizations). Employers might also contact other working women's organizations, broader women's rights groups, traditionally female educational institutions, or a variety of other organizations that serve women, including welfare rights groups. See id. at 53 (reporting that of the 1500 women who contacted Non-Traditional Employment for Women in New York in an effort to find nontraditional work, most received public assistance). Employers might also advertise in newsletters distributed by such organizations, or in magazines, newspapers, or radio shows designed to appeal specifically to a female audience. All these sorts of targeted outreach efforts find precedent in the race discrimination context. See, e.g., Movement for Opportunity v. Detroit Diesel, 18 Fair Empl. Prac. Cas. (BNA) 557, 571-72 (S.D. Ind. 1978), aff'd sub nom. Movement for Opportunity & Equality v. General Motors Corp., 622 F.2d 1235 (7th Cir. 1980); United States v. Central Motor Lines, 338 F. Supp. 532, 565 (W.D.N.C. 1971).
223. 483 F. Supp. 54 (N.D. Tex. 1979).
224. See id. at 59.
225. See id. at 61.
226. See id. at 59.
227. Id. at 61.
228. See id.
229. Id. (emphasis added).
230. Anecdotal evidence shows that the employer discriminated against live "victims." For this reason, it is sometimes referred to as evidence of "individual instances of discrimination," or "victim testimony." A woman who failed to apply for a job because the employer's sex discrimination persuaded her that it would be futile to do so is referred to in legal doctrine as a futility victim. This term should not mislead us into seeing such women as passive. By failing to submit herself to an employment process she perceives as discriminatory, the victim is actively seeking to avoid the experience of victimization. See Bumiller; Victims in the Shadow of the Law: A Critique of the Model of Legal Protection, 12 SIGNS 421 (1987).
231. Courts were approximately three times more likely to reject the lack of interest argument in cases in which plaintiffs presented anecdotal evidence than in cases in which they presented no such evidence (p = .04).
232. The p-value for Table 6 is .09. Furthermore, a logistic regression analysis showed that anecdotal evidence greatly increased plaintiffs' likelihood of prevailing, even after controlling simultaneously for evidence of past discrimination and evidence of special efforts. (The T-ratio for the anecdotal evidence variable was 1.958, with a corresponding p-value of .05.)
233. A few liberal courts have been willing to reject the lack of interest argument even in the absence of anecdotal evidence, but most of these were "easy" cases in which the evidence showed that the employer was operating an overt system of sex segregation during the period when the employer claimed women lacked interest in the work. See, e.g., Harless v. Duck, 619 F.2d 611, 614, 618 (6th Cir.), cert, denied, 449 U.S. 872 (1980); Mitchell v. Mid-Continent Spring Co., 583 F.2d 275, 280 (6th Cir.) (finding sex discrimination in a plant whose superintendent testified at trial that he assigned women only to lower paying positions because "'that's the type of work that's open for them to do'"), modified, 587 F.2d 841 (6th Cir. 1978), cert. denied, 441 U.S. 922 (1979); Chrapliwy v. Uniroyal, Inc., 458 F. Supp. 252, 262, 264 (N.D. Ind. 1977). At the other end of the spectrum, a few conservative courts have accepted the lack of interest explanation even though plaintiffs presented anecdotal evidence of discrimination or discouragement. See, e.g., Peltier v. City of Fargo, 396 F. Supp. 710, 722 (D.N.D. 1975) (rejecting three women's testimony that they failed to apply for police officer jobs because they believed the city would not hire women, as mere "feelings . . . subjectively dependent on rumors that could not be substantiated"), rev'd on other grounds, 533 F.2d 374 (8th Cir. 1976).
234. 628 F. Supp. 1264 (N.D. Ill. 1986), aff'd, 839 F.2d 302 (7th Cir. 1988).
235. Id. at 1324-25.
236. See id.
237. See, e.g., Piva v. Xerox Corp., 654 F.2d 591, 595 (9th Cir. 1981); EEOC v. H.S. Camp & Sons, Inc., 542 F. Supp. 411, 446-47 (M.D. Fla. 1982); Gilbert v. East Bay Mun. Dist., 19 Fair Empl. Prac. Cas. (BNA) 304, 307 (N.D. Cal. 1979); EEOC v. Korn Indus., 17 Fair Empl. Prac. Cas. (BNA) 954, 959 (D.S.C. 1978), aff'd on other grounds and remanded, 662 F.2d 256 (4th Cir. 1981).
238. 480 F. Supp. 1015 (W.D. Va. 1979).
239. Id. at 1027-28.
240. See, e.g., Mitchell v. Mid-Continent Spring Co., 583 F.2d 275, 278-79 (6th Cir.), modified, 587 F.2d 841 (6th Cir. 1978), cert. denied, 441 U.S. 922 (1979); Kilgo v. Bowman Transp. Inc., 570 F. Supp. 1509, 1515 (N.D. Ga. 1983), aff'd, 789 F.2d 859 (11th Cir. 1986); Ostapowicz v. Johnson Bronze Co., 369 F. Supp. 522, 537, 538 (W.D. Pa. 1973) (acknowledging that the employer had produced "a large amount of testimony that certain women were happy in the plant and thought there was no discrimination" but relying on the fact that a number of women had not opted out of the class to reject the lack of interest argument), aff'd in part and vacated in part on other grounds, 541 F.2d 394 (3d Cir. 1976), cert. denied, 429 U.S. 1041 (1977).
241. To illustrate, suppose the plaintiffs' evidence shows that, of 100 people hired by the employer, only five (5%) were women. By comparison, of the 1000 workers that plaintiffs contend were eligible for hire in the local area, 400 (40%) were women and 600 (60%) were men. Now suppose the employer proves that of the 400 women workers plaintiffs contend were eligible for hire, 200 (50%) lacked interest in the employer's work. The employer might contend that women's lack of interest explains their underrepresentation. But if 50% (300) of the men in the original pool of 1000 lacked interest in the work, too, the number of eligible male workers would decline from 600 to 300. This would leave a final pool of 500 eligible and interested workers, 200 (40%) of whom are women and 300 (60%) of whom are men. Thus, the proportions of men and women in this final pool are identical to those in the pool originally proposed by plaintiffs. Obviously the lack of interest argument cannot explain the fact that the employer's hirees included only five (5%) women, where the pool of eligible and interested workers included 200 (40%) women. To show that the employer hired women in proportion to their representation among eligible and interested workers, the employer must show that of whatever number of eligible workers in the original pool of 1000 were interested in the work, only 5% were women and 95% were men. This would mean that of the original pool, women were 19 times less interested than men in the work.
242. This is the point the EEOC's lawyers were trying to make in Sears, when they argued that "'where 47,000 hires and promotions were at issue . . . it would have been impossible to present enough individual demonstrations of discrimination to meaningfully reflect on the statistics.'" EEOC v. Sears, Roebuck & Co., 839 F.2d 302, 311 (7th Cir. 1988). The Seventh Circuit rejected this argument, remarking that "[e]ven a few examples would have helped bring 'cold numbers convincingly to life.'" Id. at 311-12 (citation omitted). The court failed to explain exactly what these examples would have proved.
243. Each of these stories is, of course, an "ideal type" drawn from a reading of the cases in the aggregate. The ideal type does not necessarily apply in every case.
244. I have borrowed heavily from Martha Fineman's insightful description of the role of narrative in legal decisionmaking about child custody. See Fineman, supra note 23, at 753-58.
245. Courts have been almost equally likely to attribute segregation to women's "choice" in blue-collar and white-collar cases. They rejected the lack of interest argument in 59.0% (23 out of 39) of the claims involving blue-collar jobs, and in 53.3% (eight out of 15) of the claim involving white collar jobs.
246. EEOC v. Mead Foods, Inc., 466 F. Supp. 1, 3 (W.D. Okla. 1977).
247. See id. at 4.
248. EEOC v. Service Container Corp., 19 Fair Empl. Prac. Cas. (BNA) 1614, 1616 (W.D. Okla. 1976). The judge in this case seemed to believe that women who work in relatively unskilled jobs in factories do not live in "very poor" sections of the city, evidence of a class bias that colors a number of these opinions.
249. Mazus v. Department of Transp., 489 F. Supp. 376, 388 (M.D. Pa. 1979), aff'd, 629 F.2d 870 (3d Cir. 1980), cert. denied, 449 U.S. 1126 (1981).
250. Ste. Marie v. Eastern R.R. Ass'n, 650 F.2d 395, 403 (2d Cir. 1981).
251. For a description of the bona fide occupational qualification defense, see note 113 above.
252. In EEOC v. H.S. Camp & Sons, Inc., 542 F. Supp. 411 (M.D. Fla. 1982), for example, a meat processing plant explicitly barred women from a number of departments, on the ground that the jobs there were too physically demanding for women. The court accepted this reasoning, even though the company produced no evidence other than the owner's subjective opinion that women were incapable of doing these jobs. See id. at 429. Ironically, the company did hire women for other lower-paying departments in which the jobs appeared to require equal physical strength. Women predominated in the meat packing department, for example, where they were required to carry boxes of meat weighing from 80 to 90 pounds. In the all-male receiving department, by contrast, the men did not rely on brute strength but used hydraulic jacks to unload heavy boxes from delivery trucks. In addition, the company had hired no women for at least three other departments that, according to the court's own description, required no particular physical prowess. See id. at 422.
253. Logan v. General Fireproofing Co., 6 Fair Empl. Prac. Cas. (BNA) 140, 144 (W.D.N.C. 1972).
254. Id.
255. EEOC v. Mead Foods, Inc., 466 F. Supp. 1, 3 (W.D. Okla. 1977).
256. I found only one blue-collar case in which an employer appealed to women's family roles to explain their alleged preferences for traditionally female work. See Parker v. Siemens-Allis, 601 F. Supp. 1377, 1385 (E.D. Ark. 1985) (noting the company's assertion in its affirmative action plan that women prefer electrical wiring to higher-paid machine-shop work because the former is "clean . . . [and] routine work, which once learned, gives the female the opportunity to plan the family budget, menu and other responsibilities directly related to family ties").
257. 198 See, e.g., Kraszewski v. State Farm Ins. Co., 38 Fair Empl. Prac. Cas. (BNA) 197, 222 (N.D. Cal. 1985); EEOC v. Akron Nat'1 Bank & Trust Co., 497 F. Supp. 733, 748 (N.D. Ohio 1980). It is not clear why employers use different explanations for sex segregation in the blue-collar and white-collar contexts. Perhaps employers have realized that it would be implausible to try to attribute sex segregation to women's domestic roles in blue-collar settings, in which women with family responsibilities have long labored in jobs that demand as much of their time as the higher-paying jobs done by men. See generally L. WEINER, FROM WORKING GIRL TO WORKING MOTHER 86-87 (1985) (describing the relegation of married women to lower-paid blue-collar work before 1940). In numerous cases in this study, women were assigned to lower-paying female jobs in factories, even though those jobs were apparently on the same shifts as the higher-paying, male-dominated jobs. See, e.g., Parker, 601 F. Supp. at 1385; Chrapliwy v. Uniroyal, Inc., 458 F. Supp. 252, 266 (N.D. Ind. 1977); Ostapowicz v. Johnson Bronze Co., 369 F. Supp. 522, 527 (W.D. Pa. 1973), aff'd in part and vacated in part on other grounds, 541 F.2d 394 (3d Cir. 1976), cert. denied, 429 U.S. 1041 (1977); see also Mitchell v. Mid-Continent Spring Co., 583 F.2d 275, 279 (4th Cir. 1978), cert. denied, 441 U.S. 922 (1979) (noting that a plant permitted male-only machine set-up employees to transfer between the day and night shifts, but refused to permit female-only machine operators to do so, because "'there was a shortage of men. However, females were easier to hire.'"). Conversely, employers may have realized that they could not plausibly defend women's absence from white-collar work with images of physical difference, because white-collar work is light, clean work of the type associated with femininity in the blue-collar context. Thus, in white-collar cases, employers have had to resort to imputed social and psychological characteristics to ground their conceptions of femininity and masculinity.
258. 528 F. Supp. 433 (E.D. Ark. 1981), aff'd on other grounds, 692 F.2d 529 (8th Cir. 1982).
259. Id. at 437.
260. EEOC v. Sears, Roebuck & Co., 628 F. Supp. 1264 (N.D. Ill. 1986), aff'd, 839 F.2d 302 (7th Cir. 1988).
261. Id. at 1308.
262. Rosalind Rosenberg was the historian who testified for Sears. The gist of her testimony is captured in the following excerpt: Women's role in American society and in the American family unit has fostered the development of 'feminine' values that have been internalized by women themselves . . . . Throughout American history women have been trained from earliest childhood to develop the humane and nurturing values expected of the American mother. Women's participation in the labor force is affected by the values they have internalized. For example: Women tend to be more relationship-centered and men tend to be more work-centered. Women tend to be more interested than men in the cooperative, social aspects of the work situation. These differences in female and male self-perception present difficulties for women in traditionally masculine occupations. Offer of Proof Concerning the Testimony of Dr. Rosalind Rosenberg, EEOC v. Sears, Roebuck & Co., 628 F. Supp. 1264 (N.D. Ill. 1986) (No. 79-C-4373) [hereinafter Testimony of Rosalind Rosenberg].
263. Sears, 628 F. Supp. at 1290.
264. Id. at 1300.
265. Id. at 1300 n.29.
266. The fact that Sears' characterization of the commission sales job varied dramatically from the way the job was denned in an earlier era shows that there is nothing necessary or inevitable about the way Sears characterized it. Susan Porter Benson has shown that from 1890 to 1940, when department store managers were eager to attract women to retail sales jobs (including commission sales) in the newly expanding service sector, managers defined the essence of "good selling" in terms of stereotypically feminine traits rather than the masculine traits emphasized by Sears. See S. BENSON, COUNTER CULTURES: SALESWOMEN, MANAGERS, AND CUSTOMERS IN AMERICAN DEPARTMENT STORES, 1890-1940, at 130-31 (1986).
267. See, e.g., EEOC v. H.S. Camp & Sons, Inc., 542 F. Supp. 411, 446 (M.D. Fla. 1982); EEOC v. Mead Foods, Inc., 466 F. Supp. 1, 4 (W.D. Okla. 1977).
268. EEOC v. Cook Paint & Varnish Co., 24 Fair Empl. Prac. Cas. (BNA) 51, 56 (W.D. Mo. 1980).
269. Ostapowicz v. Johnson Bronze Co., 369 F. Supp. 522, 537 (W.D. Pa. 1973), aff'd in part and vacated in part on other grounds, 541 F.2d 394 (3d Cir. 1976), cert. denied, 429 U.S. 1041 (1977).
270. See C. MACKINNON, SEXUAL HARASSMENT OF WORKING WOMEN 122-23 (1978).
271. The Supreme Court has adopted this form of anti-stereotyping reasoning in both the title VII and fourteenth amendment contexts. See, e.g., City of Los Angeles Dep't of Water & Power v. Manhart, 435 U.S. 702, 707-08 (1978) (holding that under title VII "[e]ven a true generalization about the class is an insufficient reason for disqualifying an individual to whom the generalization does not apply"); Frontiero v. Richardson, 411 U.S. 677 (1973) (adopting a similar rationale in the fourteenth amendment context).
272. Ostapowicz, 369 F. Supp. at 537.
273. See, e.g., Mitchell v. Mid-Continent Spring Co., 583 F.2d 275, 281-83 (6th Cir.), modified, 587 F.2d 841 (6th Cir. 1978), cert. denied, 441 U.S. 922 (1979); EEOC v. Cook Paint & Varnish Co., 24 Fair Empl. Prac. Cas. (BNA) 28, 51, 56 (W.D. Mo. 1980); EEOC v. Rath Packing Co., 40 Fair Empl. Prac. Cas. (BNA) 559, 566 (S.D. Iowa 1979), aff'd in part and rev'd in part on other grounds, 787 F.2d 318 (8th Cir. 1986); Chrapliwy v. Uniroyal, Inc., 458 F. Supp. 252, 262-63 (N.D. Ind. 1977).
274. See, e.g., Kohne v. Imco Container Co., 480 F. Supp. 1015, 1027-28 (W.D. Va. 1979); Ostapowicz, 369 F. Supp. at 537-38.
275. Rath, 40 Fair Empl. Prac. Cas. (BNA) at 566; accord Mitchell, 583 F.2d at 281; Chrapliwy, 458 F. Supp. at 278.
276. See supra pp. 1797-98.
277. See, e.g., Palmer v. Shultz, 815 F.2d 84, 110 (D.C. Cir. 1987); Kohne, 480 F. Supp. at 1027-28 & n.6; Rath, 40 Fair Empl. Prac. Cas. (BNA) at 565-66; Ostapowicz, 369 F. Supp. at 537-38. Employers tend to present testimony from women who were hired for traditionally female jobs, rather than from women who were rejected from employment altogether, as examples of women who prefer traditionally female work. But as Judge Thornberry has observed, the fact that women already working in traditionally female jobs have grown accustomed to them says little about whether women who were denied employment altogether "might well have taken [nontraditional jobs], if not precluded from doing so by a discriminatory hiring policy." Durant v. Owens-Illinois Glass Co., 656 F.2d 89, 91 (5th Cir. 1981) (Thornberry, J., dissenting).
278. As one court who accepted the lack of interest argument stated: "To be sure there are some females who would be interested in this type of physical [road maintenance] work but a reliable percentage has not yet been developed." Mazus v. Department of Transp., 489 F. Supp. 376, 388 (M.D. Pa. 1979), aff'd, 629 F.2d 870 (3d Cir. 1980), cert. denied, 449 U.S. 1126 (1981); see also EEOC v. Sears, Roebuck & Co., 628 F. Supp. 1264, 1314 (N.D. Ill. 1986) (dismissing the EEOC's historical examples of women who have responded to nontraditional job opportunities as isolated instances involving only "small groups of unusual women" rather than "the majority of women"), aff'd, 839 F.2d 302 (7th Cir. 1988).
279. See EEOC Brief, supra note 6, at 127, 141. Among the characteristics the EEOC controlled for in its statistical analyses were age, education, job applied for, job type experience, product line experience, and expanded commission sales experience. See supra note 7.
280. EEOC Brief, supra note 6, at 38.
281. See EEOC v. Sears, Roebuck & Co., 839 F.2d 302, 360-66 (7th Cir. 1988) (Cudahy, J., concurring in part and dissenting in part).
282. See id. at 361-62.
283. See id. at 362.
284. Id. (emphasis added).
285. There is now an extensive feminist literature taking the position that women's and men's divergent life experiences lead them to develop different perspectives, attitudes, and values. Carol Gilligan's book is perhaps the most prominent example of work in this tradition. See C. GILLIGAN, IN A DIFFERENT VOICE (1982). The popularity of Gilligan's work attests to the fact that people find the cultural "difference" thesis intuitively plausible. For critiques of Gilligan's work, see Auerbach, Blum, Smith & Williams, On Gilligan's In A Different Voice, 11 FEMINIST STUD. 149 (1985); Scott, cited above in note 147, at 1065; and Women and Morality, 50 Soc. RES. 487 (1983).
286. There is now an extensive feminist literature taking the position that women's and men's divergent life experiences lead them to develop different perspectives, attitudes, and values. Carol Gilligan's book is perhaps the most prominent example of work in this tradition. See C. GILLIGAN, IN A DIFFERENT VOICE (1982). The popularity of Gilligan's work attests to the fact that people find the cultural "difference" thesis intuitively plausible. For critiques of Gilligan's work, see Auerbach, Blum, Smith & Williams, On Gilligan's In A Different Voice, 11 FEMINIST STUD. 149 (1985); Scott, cited above in note 147, at 1065; and Women and Morality, 50 Soc. RES. 487 (1983).
287. There is now an extensive feminist literature taking the position that women's and men's divergent life experiences lead them to develop different perspectives, attitudes, and values. Carol Gilligan's book is perhaps the most prominent example of work in this tradition. See C. GILLIGAN, IN A DIFFERENT VOICE (1982). The popularity of Gilligan's work attests to the fact that people find the cultural "difference" thesis intuitively plausible. For critiques of Gilligan's work, see Auerbach, Blum, Smith & Williams, On Gilligan's In A Different Voice, 11 FEMINIST STUD. 149 (1985); Scott, cited above in note 147, at 1065; and Women and Morality, 50 Soc. RES. 487 (1983).
288. There is now an extensive feminist literature taking the position that women's and men's divergent life experiences lead them to develop different perspectives, attitudes, and values. Carol Gilligan's book is perhaps the most prominent example of work in this tradition. See C. GILLIGAN, IN A DIFFERENT VOICE (1982). The popularity of Gilligan's work attests to the fact that people find the cultural "difference" thesis intuitively plausible. For critiques of Gilligan's work, see Auerbach, Blum, Smith & Williams, On Gilligan's In A Different Voice, 11 FEMINIST STUD. 149 (1985); Scott, cited above in note 147, at 1065; and Women and Morality, 50 Soc. RES. 487 (1983).
289. Testimony of Rosalind Rosenberg, supra note 204, at 766; see also Davis v. City of Dallas, 483 F. Supp. 54, 61 (N.D. Tex. 1979) (attributing women's failure to apply for police work to "job preferences . . . born of attitudes conditioned by societal sexist values").
290. See EEOC v. Sears, Roebuck, & Co., 628 F. Supp. 1264, 1302-08 (N.D. Ill. 1986), aff'd, 839 F.2d 302 (7th Cir. 1988).
291. See, e.g., Palmer v. Shultz, 815 F.2d 84, 110 (D.C. Cir. 1987); EEOC v. Akron Nat'l Bank & Trust Co., 497 F. Supp. 733, 748 (N.D. Ohio 1980). But see Kraszewski v. State Farm Gen. Ins. Co., 38 Fair Empl. Prac. Cas. (BNA) 197, 221 (N.D. Cal. 1985) (acknowledging that women were interested in careers as insurance sales agents "despite difficult family situations").
292. Studies have universally found that women do far more child care and other domestic work than men and that married men increase their share of housework very little in response to increases in their wives' paid employment. See, e.g., B. BERGMANN, supra note 1, at 261-69; S. BERK, THE GENDER FACTORY: THE APPORTIONMENT OF WORK IN AMERICAN HOUSEHOLDS (1985); M. GEERKEN & W. GOVE, AT HOME AND AT WORK: THE FAMILY'S ALLOCATION OF LABOR (1983). See generally P. ENGLAND & G. FARKAS, supra note 1, at 94-99 (1986) (summarizing these studies and the prevailing explanation for men's low participation in house-work).
293. Studies have universally found that women do far more child care and other domestic work than men and that married men increase their share of housework very little in response to increases in their wives' paid employment. See, e.g., B. BERGMANN, supra note 1, at 261-69; S. BERK, THE GENDER FACTORY: THE APPORTIONMENT OF WORK IN AMERICAN HOUSEHOLDS (1985); M. GEERKEN & W. GOVE, AT HOME AND AT WORK: THE FAMILY'S ALLOCATION OF LABOR (1983). See generally P. ENGLAND & G. FARKAS, supra note 1, at 94-99 (1986) (summarizing these studies and the prevailing explanation for men's low participation in house-work).
294. Studies have universally found that women do far more child care and other domestic work than men and that married men increase their share of housework very little in response to increases in their wives' paid employment. See, e.g., B. BERGMANN, supra note 1, at 261-69; S. BERK, THE GENDER FACTORY: THE APPORTIONMENT OF WORK IN AMERICAN HOUSEHOLDS (1985); M. GEERKEN & W. GOVE, AT HOME AND AT WORK: THE FAMILY'S ALLOCATION OF LABOR (1983). See generally P. ENGLAND & G. FARKAS, supra note 1, at 94-99 (1986) (summarizing these studies and the prevailing explanation for men's low participation in house-work).
295. Studies have universally found that women do far more child care and other domestic work than men and that married men increase their share of housework very little in response to increases in their wives' paid employment. See, e.g., B. BERGMANN, supra note 1, at 261-69; S. BERK, THE GENDER FACTORY: THE APPORTIONMENT OF WORK IN AMERICAN HOUSEHOLDS (1985); M. GEERKEN & W. GOVE, AT HOME AND AT WORK: THE FAMILY'S ALLOCATION OF LABOR (1983). See generally P. ENGLAND & G. FARKAS, supra note 1, at 94-99 (1986) (summarizing these studies and the prevailing explanation for men's low participation in house-work).
296. This process works in the opposite direction, too, when employers create low-paying jobs and then recruit in a way that is designed to attract women. After title VII took effect a number of employers in the cases in this study transfered some of the responsibilities of an exclusively male job to a new, much lower-paid job, and then proceeded to construct the new job as a
297. Some multinational corporations justify hiring women for the least skilled jobs in ideological terms, citing women's "natural patience" and "manual dexterity"; they justify paying women low wages by claiming the women do not need to work and will quit when they get married anyway. See A. FUENTES & B. EHRENREICH, WOMEN IN THE GLOBAL FACTORY 11-15 (1983).
298. A classic study by Bem and Bem illustrates this point. Whereas only 5% of the women surveyed expressed interest in nontraditional telephone "lineman" and "frameman" jobs when the ad described those jobs in sex-biased language, 25% expressed interest when the language was sex-neutral, and fully 45% expressed interest when the ad was written to appeal to women. Bem & Bem, Does Sex-Biased Job Advertising 'Aid and Abet' Sex Discrimination?, 3 J. APPLIED Soc. PSYCHOLOGY 6 (1973). Even apparently sex-neutral language can communicate to women that they are not who the employer had in mind for the job. Cockburn describes the following ad for an electronics engineer: "Enthusiasm, along with creativity, drive and a clear understanding of your personal contribution are needed in a business where technological limits are constantly being tested and new frontiers broken and explored." C. COCKBURN, supra note 14, at 181. As Cockburn states: On the face of it, this is not discriminatory wording. But women know how women are usually defined - not with words like 'drive,' 'limits,' 'test.' [A] woman is likely to read [this] as addressed not to women but to men. To many women it will be more of a warning than an invitation. Id.
299. It was not until 1969, however, that the EEOC guidelines on sex discrimination prohibited employers from placing neutrally worded advertisements under "Male" or "Female" newspaper headings. Campare 29 C.F.R. § 1604.4(b) (1966) (permitting advertising placements under "Male" and "Female" headings to convey "that some occupations are considered more attractive to persons of one sex than the other") with 29 C.F.R. § 1604.5 (1989) (prohibiting such placements unless sex is a bona fide occupational qualification for the job).
300. See Wetzel, 372 F. Supp. at 1150-51, 1154 (recognizing that the company's use of separate, heavily gendered recruiting brochures was designed to communicate to women that the only position appropriate for them was the lower-paying claims representative job, but not the higher-paying claims adjuster job); see also Kraszewski v. State Farm Ins. Co., 38 Fair Empl. Prac. Cas. (BNA) 197, 230-31 (N.D. Cal. 1985) (using a similar analysis).
301. See, e.g., Chrapliwy v. Uniroyal, Inc., 458 F. Supp. 252, 266-67, 284 (N.D. Ind. 1977); Hill v. Western Elec. Co., 12 Fair Empl. Prac. Cas. (BNA) 1175, 1179 (E.D. Va. 1976), aff'd in part and rev'd in part on other grounds, 596 F.2d 99 (4th Cir.), cert. denied, 444 U.S. 929 (1979); Peltier v. City of Fargo, 396 F. Supp. 710, 727 (D.N.D. 1975), aff'd in part and rev'd in part on other grounds, 533 F.2d 374 (8th Cir. 1976). Peltier illustrates the pitfalls for plaintiffs caused by the courts' failure to develop an analysis of the deeper gender dynamics of sexually discriminatory advertising. The plaintiffs were three women who had been hired in 1973 for the police department's newly created, all-female car marker position. They claimed that they had been discriminatorily denied the opportunity to become regular police patrol officers. They had never applied to be patrol officers, but they testified that they believed it would have been futile to do so in light of the fact that the department had never hired any women for the job. Id. at 713-14. The police department had advertised the job as one for males only until the spring of 1973, when it stopped advertising altogether and began to rely on word-of-mouth recruiting. Even in the face of this evidence, however, the court held that the fact that the plaintiffs' failure to apply for the patrol officer job defeated their claims. The court recognized that the sex-segregated advertising was discriminatory, but it concluded that the advertising could not have discouraged the plaintiffs from applying, because the last male-only ad had been placed shortly before the plaintiffs were hired as car markers. See id. at 723.
302. As one court stated, "[n]o notices of vacancies were ever posted, nor had the Union ever recommended any of its women members for these positions [as meat-cutters]. Consequently plaintiffs were never informed of the vacancies for which they could apply." Babrocky v. Jewel Food Co., 773 F.2d 857, 867 (7th Cir. 1985). Of the 54 sex discrimination claims in my study, in 17 claims the employer used word-of-mouth recruiting. Of these 17 claims, in only eight did the court rely even in part on the employer's use of word-of-mouth recruiting to reject the lack of interest argument. See, e.g., Wheeler v. City of Columbus, 686 F.2d 1144, 1148, 1152 (5th Cir. 1982); Davis v. Richmond, F. & P. R.R., 593 F. Supp. 271, 278 & n.1, 279 (E.D. Va. 1984), aff'd in part and rev'd in part on other grounds, 803 F.2d 1322 (4th Cir. 1986). Kraszewski is the only one of these cases in which the court portrayed the use of word-of-mouth recruiting as an active form of constructing a male-oriented occupational culture rather than a merely passive form of failing to provide information about job opportunities to already interested women. Kraszewski, 38 Fair Empl. Prac. Cas. (BNA) at 228-29 (citing from the company's own employment manual to show how word-of mouth recruiting inevitably produces candidates who look just like the men who are recruiting them).
303. An extensive literature documents that male workers are more likely to share job information with other men than with women. For this reason, men are more likely to secure their jobs through personal contacts, while women are more likely to use formal job-search methods. For a summary of this literature, see Roos & Reskin, Institutional Factors Contributing to Sex Segregation in the Workplace, in SEX SEGREGATION IN THE WORKPLACE, cited above in note 1 at 235, 241-42, 245-46.
304. 654 F.2d 388 (5th Cir. 1981), vacated on other grounds, 459 U.S. 809 (1982).
305. See id. at 399.
306. See id. at 396 n.8, 399.
307. Id. at 400.
308. This, of course, is not true. The New York Times reported the findings of a recent study which is only "the latest in a steady stream of research over the last two decades showing that while they may be sitting side by side, male and female [college students] have substantially different educational experiences." Fiske, Lessons, N.Y. Times, Apr. 11, 1990, at B8, col. 1. According to Catherine G. Krupnick, the Harvard Graduate School of Education researcher who conducted the study: "'College catalogues should carry warnings: The value you receive will depend on your sex.'" Id. Like earlier researchers, Krupnick's study found that "faculty members consistently take male students and their contributions more seriously than females and their ideas." Id.
309. Other courts also have refused to characterize word-of-mouth recruiting as sexually discriminatory. In EEOC v. Service Container Corp., 19 Fair Empl. Prac. Cas. (BNA) 1614 (W.D. Okla. 1976), for example, the plaintiff contended that the plant's reliance on its almost-exclusively male workforce to recruit new shop workers discriminated against women. The court disposed of this contention in a sentence, noting that the plaintiff herself "came to work as a referral." See. id. at 1616. But the plaintiff had applied because she had learned specifically that the plant was replacing another female worker, not because she had heard that the plant was hiring workers generally. See id.
310. EEOC v. Mead Foods, Inc., 466 F. Supp. 1, 3 (W.D. Okla. 1977).
311. See Marini & Brinton, Sex Typing in Occupational Socialization, in SEX SEGREGATION AND THE WORKPLACE, supra note 1, at 192 (describing studies that show differences in young people's early occupational aspirations and discussing the mechanisms through which their aspirations come to be gendered).
312. Researchers who explain segregation as a function of pre-labor market socialization have done little to explore this link. Typically, they point to evidence showing that the level of sex segregation in young people's occupational aspirations corresponds to the level of sex segregation in the labor force. See, e.g., id. at 203-04 (concluding that the "overall similarity between the set typing of occupational aspirations and attainments indicates that influences prior to labor market entry play an important role in the determination of occupational outcomes for individuals"). However, they fail to demonstrate any direct connection between the sex-type of early aspirations and the sex-type of later occupational outcomes.
313. See J. JACOBS, supra note 1.
314. Jacobs examined data from the National Longitudinal Survey of Young Women (NLS Voung Women), a survey of a representative sample of more than 5000 women between the ages of 14 to 24 in 1968. Each year between 1968 and 1980, the women were asked to name the occupation in which they were employed that year, as well as the occupation to which they aspired at age 35. See id. at 10. Jacobs' analyses treat the sex-type of occupations held or aspired to as a continuous variable, with each occupation coded from 0 to 100 percent female. See id. at 94.
315. For women who changed their aspirations between 1970 and 1980, less than 1% of the variance in the sex-type of the occupations to which they aspired in 1980 was explained by the sex-type of their 1970 aspirations. See id. at 94-96.
316. See id. at 96-97.
317. See id. at 97-98.
318. Jacobs also examined data from the NLS Survey of Mature Women, a representative sample of over 5000 women between the ages of 30 and 44 in 1967. See id. at 11. A majority of these women (58%) changed occupations between 1967 and 1977. See id. at 141 (derived from table 7.1). For those who changed occupations, there was no correlation between the sex-type of the occupations in which they were employed in 1967 and 1977. See id. at 140-42. Indeed, 61% of the occupation-changers changed the sex-type of their occupations. See id. at 141 (derived from table 7.1).
319. See id. (table 7.1). Jacob's findings are consistent with those of other studies of women's mobility patterns. See Corcoran, Duncan & Ponza, Work Experience, Job Segregation, and Wages, in SEX SEGREGATION IN THE WORKPLACE, supra note 1, at 171, 177-78 (finding that 34% of the women in the Panel Study of Income Dynamics survey changed the sex-type of their occupations between 1975 and 1979); Rosenfeld, Job Changing and Occupational Sex Segregation: Sex and Race Comparisons, in SEX SEGREGATION IN THE WORKPLACE, supra note 1, at, 56, 63 (finding a low correlation between the sex-type of original and destination occupationas for a sample of women who changed employers between 1972 and 1973).
320. See id. (table 7.1). Jacob's findings are consistent with those of other studies of women's mobility patterns. See Corcoran, Duncan & Ponza, Work Experience, Job Segregation, and Wages, in SEX SEGREGATION IN THE WORKPLACE, supra note 1, at 171, 177-78 (finding that 34% of the women in the Panel Study of Income Dynamics survey changed the sex-type of their occupations between 1975 and 1979); Rosenfeld, Job Changing and Occupational Sex Segregation: Sex and Race Comparisons, in SEX SEGREGATION IN THE WORKPLACE, supra note 1, at, 56, 63 (finding a low correlation between the sex-type of original and destination occupationas for a sample of women who changed employers between 1972 and 1973).
321. see J. JACOBS, supra note 1, at 103, Jacobs found that 49% of the NLS Young Women aspired to, and 44% worked in, a male-dominated occupation in some survey year between 1968 and 1980. See id.
322. See id. at 148-49. Jacobs also found that other independent variables - including number and ages of children, weeks employed, and hours worked per week - did not dramatically alter women's occupational mobility patterns. See id. at 149-50. Jacobs' findings are consistent with Rosenfeld's, who found that, for both black and white women, the likelihood of changing the sex-type of their occupations was independent of marital status and whether they had interrupted their careers to care for children. See Rosenfeld, supra note 254, at 72-76. Ironically, Rosenfeld found that "[t]he only effect of family responsibility . . . [was] for white men," who were less likely to move from a male-dominated to a female-dominated occupation if they were married. Id. at 74 (emphasis in original).
323. See, e.g., Beller, Occupational Segregation by Sex: Determinants and Changes, 17 J. HUM. RESOURCES 371, 383 (1982); England, The Failure of Human Capital Theory To Explain Occupational Sex Segregation, 17 J. HUM. RESOURCES 358, 367 (1982).
324. See, e.g., Beller, Occupational Segregation by Sex: Determinants and Changes, 17 J. HUM. RESOURCES 371, 383 (1982); England, The Failure of Human Capital Theory To Explain Occupational Sex Segregation, 17 J. HUM. RESOURCES 358, 367 (1982).
325. See, e.g., Corcoran, Duncan & Ponza, supra note 254, at 188; England, supra note 257, at 368.
326. See Beller, supra note 257, at 384-85; Daymont & Statham, Occupational Atypicality: Changes, Causes and Consequences, in 5 DUAL CAREERS 107 (L. Shaw ed. 1981). Indeed, Beller found that contrary to conventional predictions, the probability that a woman was nontraditionally employed actually increased slightly with her number of children. Moreover, Beller found that, even if women had been identical to men in terms of a number of personal characteristics associated with family responsibilities (for example, marital status, number of children, number of weeks worked, part-time versus full-time status, and whether the reason for working part-time was "home specialization"), the probability that a woman would have worked in a nontraditional occupation in 1974 would have increased by only 1.1%. See Beller, supra note 257, at 384-85.
327. See Beller, supra note 257, at 384-85; Daymont & Statham, Occupational Atypicality: Changes, Causes and Consequences, in 5 DUAL CAREERS 107 (L. Shaw ed. 1981). Indeed, Beller found that contrary to conventional predictions, the probability that a woman was nontraditionally employed actually increased slightly with her number of children. Moreover, Beller found that, even if women had been identical to men in terms of a number of personal characteristics associated with family responsibilities (for example, marital status, number of children, number of weeks worked, part-time versus full-time status, and whether the reason for working part-time was "home specialization"), the probability that a woman would have worked in a nontraditional occupation in 1974 would have increased by only 1.1%. See Beller, supra note 257, at 384-85.
328. See J. JACOBS, supra note 1, at 190-91 ("[M]arital and family responsibilities simply are not powerful factors in producing mobility from male-dominated into female-dominated occupations."); Rosenfeld, supra note 254, at 77 ("For neither white nor black women was there much support for the idea that extent of family responsibilities influences the chance to move from or to a sex-typical occupation.").
329. See supra note 231.
330. The fact that women with primary family responsibilities are about as likely as women without such responsibilities to be found in, or to move to, nontraditional occupations may reflect a number of underlying phenomena. First, many nontraditional jobs probably do not pose any greater barriers to family life than do traditionally female jobs. See supra note 198. Indeed, portraying the jobs men do as inherently more demanding than the jobs women do is part of the ideological framework that stigmatizes women as marginal workers and justifies keeping them out of the higher-paying "men's jobs." This is part of the insight of the comparable worth movement. See Clauss, Comparable Worth - The Theory, Its Legal Foundation, and the Feasibility of Implementation, 20 U. MICH. J.L. REF. 7 (1986). Second, to the extent that certain nontraditional jobs may make demands that are difficult for primary family caretakers, many such women are willing to undertake those demands despite the back-breaking double burden. Particularly for working-class women, the higher wages for nontraditional jobs enable them to give their children greater opportunities, such as a college education. See, e.g., M. WALSHOK, BLUE-COLLAR WOMEN: PIONEERS ON THE MALE FRONTIER 252 (1981). As Walshok explained: Most of the women with children indicated they would like more time with their children, but none defined their need for employment as a hardship on children because "what could I do for my kids and husband just sitting around the house all day!" This is not to suggest the women had no problems combining work and family. They clearly did. . . . In fact, their descriptions of a typical workday showed that most were fully occupied twelve to fourteen hours a day.
331. The fact that women with primary family responsibilities are about as likely as women without such responsibilities to be found in, or to move to, nontraditional occupations may reflect a number of underlying phenomena. First, many nontraditional jobs probably do not pose any greater barriers to family life than do traditionally female jobs. See supra note 198. Indeed, portraying the jobs men do as inherently more demanding than the jobs women do is part of the ideological framework that stigmatizes women as marginal workers and justifies keeping them out of the higher-paying "men's jobs." This is part of the insight of the comparable worth movement. See Clauss, Comparable Worth - The Theory, Its Legal Foundation, and the Feasibility of Implementation, 20 U. MICH. J.L. REF. 7 (1986). Second, to the extent that certain nontraditional jobs may make demands that are difficult for primary family caretakers, many such women are willing to undertake those demands despite the back-breaking double burden. Particularly for working-class women, the higher wages for nontraditional jobs enable them to give their children greater opportunities, such as a college education. See, e.g., M. WALSHOK, BLUE-COLLAR WOMEN: PIONEERS ON THE MALE FRONTIER 252 (1981). As Walshok explained: Most of the women with children indicated they would like more time with their children, but none defined their need for employment as a hardship on children because "what could I do for my kids and husband just sitting around the house all day!" This is not to suggest the women had no problems combining work and family. They clearly did. . . . In fact, their descriptions of a typical workday showed that most were fully occupied twelve to fourteen hours a day.
332. See K. GERSON, HARD CHOICES: How WOMEN DECIDE ABOUT WORK, CAREER, AND MOTHERHOOD 53-55 (1985). Gerson did a life-history analysis of a group of women between At ages of 27 and 37, with varying socioeconomic backgrounds and work histories. When asked about parental expectations while growing up, only 32% of the women responded that their parents had stressed the importance of marriage and family above all else. An almost equal proportion (27%) reported that their parents had placed greater value on education, work, and economic self-sufficiency than on marriage and parenting. Another 17% could not recall their parents expressing any strong expectations. The remaining 25% stated that they had received mixed messages: the importance of education, work and economic independence had been stressed, while marriage, motherhood, and homemaking had been assumed. See id. at 54.
333. See id. at 198.
334. Gerson confirms this theme of change and development over time. Of the women she studied, 55% stated that they had emerged from childhood with "domestic" orientations, in which they expected their primary roles to be marriage, mothering, and homemaking. The other 45% reported early "nondomestic" orientations, in which they looked with indifference or disdain on domesticity and expected to devote themselves primarily to wage work instead. See id. at 59-65. Both groups of women deviated substantially from their original orientations as they moved into adulthood. Of those who reported early domestic orientations, 67% changed their orientations to nondomestic ones as adults, and of those who reported nondomestic origins, 63% became domestically oriented as adults. See id. at 67. Furthermore, neither parental expectations nor maternal role models predicted whether the women took domestic or nondomestic paths. Of those whose parents had stressed marriage and family, only 47% were domestically oriented as adults; of those whose parents had stressed education, work, and self-sufficiency, only 62% held primarily nondomestic orientations as adults. See id. at 55.
335. See C. WILLIAMS, GENDER DIFFERENCES AT WORK: WOMEN AND MEN IN NONTRADITIONAL OCCUPATIONS 78-79 (1989).
336. Few women joined the Marines out of a desire to defy traditional sex roles. They joined the Marines for pragmatic reasons, most notably a desire for financial security and career advancement. They almost universally planned to combine childrearing with their military careers; many enlisted to support their children. See id. at 72-74.
337. Id. at 75.
338. Id. at 6. Many of the women also refused to accept the cultural construction of the military as inherently "masculine." Some of them insisted, for example, that the Marine's basic training in discipline and deference was no different from Catholic schoolgirls' education. See id. at 141.
339. Id. at 79. If the women did not view nontraditional work as inconsistent with their womanhood, most were also disdainful of women who tried to "use" their womanhood to obtain special favors. True "femininity," for them, lay in a sense of dignity and self-worth that was incompatible with "using the fact that you're feminine . . . . to get out of things." See id. at 77.
340. See M. WALSHOK, supra note 262. Two more recent books include detailed interviews with women in nontraditional blue-collar occupations. See M. MARTIN, HARD-HATTED WOMEN: STORIES OF STRUGGLE AND SUCCESS IN THE TRADES (1988); J. SCHROEDEL, ALONE IN A CROWD: WOMEN IN THE TRADES TELL THEIR STORIES (1985).
341. See M. WALSHOK, supra note 262. Two more recent books include detailed interviews with women in nontraditional blue-collar occupations. See M. MARTIN, HARD-HATTED WOMEN: STORIES OF STRUGGLE AND SUCCESS IN THE TRADES (1988); J. SCHROEDEL, ALONE IN A CROWD: WOMEN IN THE TRADES TELL THEIR STORIES (1985).
342. See M. WALSHOK, supra note 262. Two more recent books include detailed interviews with women in nontraditional blue-collar occupations. See M. MARTIN, HARD-HATTED WOMEN: STORIES OF STRUGGLE AND SUCCESS IN THE TRADES (1988); J. SCHROEDEL, ALONE IN A CROWD: WOMEN IN THE TRADES TELL THEIR STORIES (1985).
343. M. WALSHOK, supra note 262, at 115-16.
344. See K. DEAUX & J. ULLMAN, supra note 25, at 74 (reporting that of the 103 women in nontraditional jobs in the steel industry, the vast majority had previously been employed in traditionally female jobs); M. MARTIN, supra note 271 (revealing that of the 18 women who discussed their employment history, 14 had previously worked in female-dominated jobs); J. SCHROEDEL, supra note 271 (revealing that of the 25 women in nontraditional blue-collar Jobs, 20 had previously held female-dominated jobs); U.S. DEP'T OF LABOR, WOMEN IN TRADITIONALLY MALE JOBS: THE EXPERIENCE OF TEN PUBLIC UTILITY COMPANIES 35-37 (1978) (revealing that of women in nontraditional jobs in utility companies, most had begun in clerical jobs).
345. See M. WALSHOK, supra note 262, at 121.
346. Id. at 120. All of the women, including college graduates, reported that their "employment opportunities [had been] limited primarily to clerical jobs, waitressing, housecleaning, and factory work." Id. at 121.
347. See id. at 138. As one female mechanic put it: "I never made a conscious decision to become a mechanic, y'know, for that to be my skill or my trade. I don't remember that happening. It is now. And 1 think this is pretty much my life's work." Id. at 156.
348. As one female aircraft production control trainee explained: "I didn't know anything about aircraft other than to see them fly. . . . I didn't know what to expect, but I knew that I would be trained." Id. at 136. Another female stationary engineer did not even know what her own father had done for a living until after she had decided on her own to enter his trade. See M. MARTIN, supra note 271 at 39. While her story may seem amusing, it is also telling. It reveals that many fathers do not pass on knowledge of their trades to daughters the way they have passed it on to their sons and nephews. Thus, informal social networks for sharing occupational knowledge and skill remain gendered, even if some courts refuse to acknowledge this truth. See, e.g., Wilkins v. University of Houston, 654 F.2d 388, 399-400 (5th Cir. 1981), vacated on other grounds, 459 U.S. 809 (1982); supra pp. 1813-14 (discussing Wilkins).
349. M. WALSHOK, supra note 262, at 132-33 (footnote omitted).
350. See id. at 132. Walshok suggests that the assumption that people form their work aspirations before they enter the workworld is class-biased. Since women from upper-middle-class families do not necessarily take paid employment for granted, they have the luxury of clarifying their work values and interests before they enter the labor market. Working-class women, however, must begin working early in life in whatever jobs they can get. Thus, "employment may be the first step and the primary context in which [their] values and interests become solidified. . . . [They] might discover a whole world of unanticipated interests and abilities on the job which then become the impetus for training or education at a later phase life." Id. at 272. The assumption that career development proceeds in a linear sequence with stable values and interests leading people to enter certain lines of work seems to capture the experience of professionals better than that of working-class people. Jacobs suggests, however, that this linear model may not accurately describe the experience of many middle-class women, either. In his case studies of women in medicine and law, he found that substantial numbers of women entered these professions mid-career, after beginning in female-dominated occupations. Thus, contrary to the pre-labor market perspective, women physicians and lawyers were not necessarily raised with atypical sex-role expectations. Many contested their early socialization and entered male-dominated professions when the opportunity became available. See J. JACOBS, supra note 1,at 155-64.
351. See J. JACOBS, supra note 1, at 48.
352. As Rosabeth Moss Kanter has written: [T]o a very large degree, organizations make their workers into who they are. Adults change to fit the system. . . . [O]rganizations often act as though it is possible to predict people's job futures from the characteristics they bring with them [to] a recruiting interview. What really happens is that predictions get made on the basis of stereotypes and current notions of who fits where in the present system; people are then "set up" in positions which make the predictions come true. R KANTER, MEN AND WOMEN OF THE CORPORATION 263 (1977).
353. See id. at 9.
354. See P. ENGLAND & G. FARKAS, supra note 1, at 140. "[R]esearch [in this tradition] implies that individuals develop the psychological styles required to survive in the structural Position they hold. . . . [B]ehavioral differences between groups are a product of the jobs they have been allowed to enter, rather than being exogenous to actual work experience." Id. at 138.
355. See R. KANTER, supra note 281, at 251-52.
356. Id. at 252.
357. See J. JACOBS, supra note 1, at 141. Overall, 70.4% of the NLS mature women who were employed in female-dominated occupations in 1967 remained in female occupations in 1977, but only 46.9% of the women who were employed in male-dominated occupations remained in male occupations a decade later. See id. at 141-42. This pattern reflects a disproportionate - and alarmingly high - rate of female attrition from male-dominated occupations. Of the women who left male-dominated occupations between 1967 and 1977, only 19.5% moved to other male-dominated occupations, while 43.2% switched to female-dominated occupations and 37.4% to more sexually integrated ones. But of those who left female-dominated occupations, 45.1% moved to other female occupations and 35.3% to more sexually integrated ones, while only 19.6% switched to male-dominated occupations. See id. at 141.
358. Because the absolute number of women who began in female-dominated occupations was much greater than the number who began in male-dominated occupations, the relatively low proportion of women who shifted from female to male occupations between 1967 and 1977 caused a net increase in the overall share of women in male occupations. In other words, in absolute numbers, more women entered male-dominated occupations than left them, and, thus, the overall level of sex segregation declined slightly during the 1970's. Had there been less attrition of women from male occupations, however, the level of segregation would have declined much more dramatically. Jacobs found that this same "revolving door" pattern continued into the 1980's: In recent years, for every 100 women in male-dominated occupations who were employed in two consecutive years, 90 remained in a male-dominated occupation, while 10 left for either a sex-neutral or female-dominated occupation. At the same time, 11 entered a male-dominated occupation from one of these other occupation groups. Thus, the revolving door sends out 10 for every 11 it lets in. Id. at 4.
359. C. COCKBURN, supra note 14, at 167. A number of other British and Australian researchers have done theoretical and empirical work from this perspective, examining how gender both structures and is constructed within the labor process itself. See, e.g., V. BEECHEY & T. PERKINS, supra note 14; A. GAME & R. PRINGLE, GENDER AT WORK (1983); Scott, Industrialization, Gender Segregation, and Stratification Theory, in GENDER AND STRATIFICATION 154 (R. Crompton & M. Mann eds. 1986).
360. C. COCKBURN, supra note 14, at 167. A number of other British and Australian researchers have done theoretical and empirical work from this perspective, examining how gender both structures and is constructed within the labor process itself. See, e.g., V. BEECHEY & T. PERKINS, supra note 14; A. GAME & R. PRINGLE, GENDER AT WORK (1983); Scott, Industrialization, Gender Segregation, and Stratification Theory, in GENDER AND STRATIFICATION 154 (R. Crompton & M. Mann eds. 1986).
361. C. COCKBURN, supra note 14, at 167. A number of other British and Australian researchers have done theoretical and empirical work from this perspective, examining how gender both structures and is constructed within the labor process itself. See, e.g., V. BEECHEY & T. PERKINS, supra note 14; A. GAME & R. PRINGLE, GENDER AT WORK (1983); Scott, Industrialization, Gender Segregation, and Stratification Theory, in GENDER AND STRATIFICATION 154 (R. Crompton & M. Mann eds. 1986).
362. C. COCKBURN, supra note 14, at 165.
363. By focusing on these two phenomena, I do not mean to suggest that they are the only features of the workplace that contribute to women's disempowerment. See, e.g., Roos & Reskin, supra note 238, at 235-60 (describing numerous institutional mechanisms that limit women's ability to enter nontraditional jobs); supra pp. 1811-13 (discussing how employers' recruiting strategies discourage women from aspiring to nontraditional work).
364. See, e.g., R. BLAUNER, ALIENATION AND FREEDOM (1964); E. CHINOY, AUTOMOBILE WORKERS AND THE AMERICAN DREAM (1955); T. PURCELL, BLUE COLLAR MAN: PATTERNS OF DUAL ALLEGIANCE IN INDUSTRY (1960); Dubin, Industrial Workers' Worlds: A Study of the "Central Life Interests" of Industrial Workers, 3 Soc. PROBS. 131 (1956); Guest, Work Careers and Aspirations of Automobile Workers, 19 AM. Soc. REV. 155 (1954); see also R. KANTER, supra note 281, at 140, 143, 147-48 (summarizing these and other similar studies).
365. See, e.g., R. BLAUNER, ALIENATION AND FREEDOM (1964); E. CHINOY, AUTOMOBILE WORKERS AND THE AMERICAN DREAM (1955); T. PURCELL, BLUE COLLAR MAN: PATTERNS OF DUAL ALLEGIANCE IN INDUSTRY (1960); Dubin, Industrial Workers' Worlds: A Study of the "Central Life Interests" of Industrial Workers, 3 Soc. PROBS. 131 (1956); Guest, Work Careers and Aspirations of Automobile Workers, 19 AM. Soc. REV. 155 (1954); see also R. KANTER, supra note 281, at 140, 143, 147-48 (summarizing these and other similar studies).
366. See, e.g., R. BLAUNER, ALIENATION AND FREEDOM (1964); E. CHINOY, AUTOMOBILE WORKERS AND THE AMERICAN DREAM (1955); T. PURCELL, BLUE COLLAR MAN: PATTERNS OF DUAL ALLEGIANCE IN INDUSTRY (1960); Dubin, Industrial Workers' Worlds: A Study of the "Central Life Interests" of Industrial Workers, 3 Soc. PROBS. 131 (1956); Guest, Work Careers and Aspirations of Automobile Workers, 19 AM. Soc. REV. 155 (1954); see also R. KANTER, supra note 281, at 140, 143, 147-48 (summarizing these and other similar studies).
367. See, e.g., R. BLAUNER, ALIENATION AND FREEDOM (1964); E. CHINOY, AUTOMOBILE WORKERS AND THE AMERICAN DREAM (1955); T. PURCELL, BLUE COLLAR MAN: PATTERNS OF DUAL ALLEGIANCE IN INDUSTRY (1960); Dubin, Industrial Workers' Worlds: A Study of the "Central Life Interests" of Industrial Workers, 3 Soc. PROBS. 131 (1956); Guest, Work Careers and Aspirations of Automobile Workers, 19 AM. Soc. REV. 155 (1954); see also R. KANTER, supra note 281, at 140, 143, 147-48 (summarizing these and other similar studies).
368. See, e.g., R. BLAUNER, ALIENATION AND FREEDOM (1964); E. CHINOY, AUTOMOBILE WORKERS AND THE AMERICAN DREAM (1955); T. PURCELL, BLUE COLLAR MAN: PATTERNS OF DUAL ALLEGIANCE IN INDUSTRY (1960); Dubin, Industrial Workers' Worlds: A Study of the "Central Life Interests" of Industrial Workers, 3 Soc. PROBS. 131 (1956); Guest, Work Careers and Aspirations of Automobile Workers, 19 AM. Soc. REV. 155 (1954); see also R. KANTER, supra note 281, at 140, 143, 147-48 (summarizing these and other similar studies).
369. See, e.g., R. BLAUNER, ALIENATION AND FREEDOM (1964); E. CHINOY, AUTOMOBILE WORKERS AND THE AMERICAN DREAM (1955); T. PURCELL, BLUE COLLAR MAN: PATTERNS OF DUAL ALLEGIANCE IN INDUSTRY (1960); Dubin, Industrial Workers' Worlds: A Study of the "Central Life Interests" of Industrial Workers, 3 Soc. PROBS. 131 (1956); Guest, Work Careers and Aspirations of Automobile Workers, 19 AM. Soc. REV. 155 (1954); see also R. KANTER, supra note 281, at 140, 143, 147-48 (summarizing these and other similar studies).
370. See E. CHINOY, supra note 201, at 114-15, 130, 132-133; Dubin, supra note 291, at 131, 135.
371. See E. CHINOY, supra note 291, at 74, 82-83, 85-86, 118-20; Guest, supra note 291, at 158-59, 162-63; Mayer & Goldstein, Manual Workers as Small Businessmen, in BLUE COLLAR WORLD, 537, 539-43 (A. Shostak & W. Gomberg eds. 1964).
372. See E. CHINOY, supra note 291, at 74, 82-83, 85-86, 118-20; Guest, supra note 291, at 158-59, 162-63; Mayer & Goldstein, Manual Workers as Small Businessmen, in BLUE COLLAR WORLD, 537, 539-43 (A. Shostak & W. Gomberg eds. 1964).
373. See E. CHINOY, supra note 291, at 74, 82-83, 85-86, 118-20; Guest, supra note 291, at 158-59, 162-63; Mayer & Goldstein, Manual Workers as Small Businessmen, in BLUE COLLAR WORLD, 537, 539-43 (A. Shostak & W. Gomberg eds. 1964).
374. See, e.g., Pennings, Work-Value Systems of White-Collar Workers, 15 ADMIN. Sci. Q. 397, 401-02 (1970); Tichy, An Analysis of Clique Formation and Structure in Organizations, 18 ADMIN. Sci. Q. 194, 195-96, 204-05 (1973). See generally R. MERTON, SOCIAL THEORY AND SOCIAL STRUCTURE 319-22, 347 (1968) (noting that in high-mobility systems, individuals compare themselves with those higher in rank in a process of "anticipatory socialization," but that in low-mobility systems, individuals compare themselves with their peers and develop group solidarity).
375. See, e.g., Pennings, Work-Value Systems of White-Collar Workers, 15 ADMIN. Sci. Q. 397, 401-02 (1970); Tichy, An Analysis of Clique Formation and Structure in Organizations, 18 ADMIN. Sci. Q. 194, 195-96, 204-05 (1973). See generally R. MERTON, SOCIAL THEORY AND SOCIAL STRUCTURE 319-22, 347 (1968) (noting that in high-mobility systems, individuals compare themselves with those higher in rank in a process of "anticipatory socialization," but that in low-mobility systems, individuals compare themselves with their peers and develop group solidarity).
376. See, e.g., Pennings, Work-Value Systems of White-Collar Workers, 15 ADMIN. Sci. Q. 397, 401-02 (1970); Tichy, An Analysis of Clique Formation and Structure in Organizations, 18 ADMIN. Sci. Q. 194, 195-96, 204-05 (1973). See generally R. MERTON, SOCIAL THEORY AND SOCIAL STRUCTURE 319-22, 347 (1968) (noting that in high-mobility systems, individuals compare themselves with those higher in rank in a process of "anticipatory socialization," but that in low-mobility systems, individuals compare themselves with their peers and develop group solidarity).
377. See E. CHINOY, supra note 291, at 47-50, 59-60, 62, 111-12; T. PURCELL, supra note 291, at 125-26, 269; Guest, supra note 291, at 157-59.
378. For a critique of earlier sociological studies analyzing men's relationship to employment on a "job model" while analyzing women's on a "gender model," see Feldberg & Glenn, Male and Female: Job Versus Gender Models in the Sociology of Work, 26 Soc. PROBLEMS 524 (1979).
379. See, e.g., Bielby & Baron, supra note 1, at 27, 35; Gutek & Morasch, supra note 1, at 61-62.
380. See Hartmann, Internal Labor Markets and Gender: A Case Study of Promotion, in GENDER IN THE WORKPLACE 59, 59-66 (C. Brown & J. Pechman eds. 1987) (reviewing studies documenting the existence of separate internal career ladders for men and women); Roos & Reskin, supra note 238, at 248-51 (same).
381. R. KANTER, supra note 281, at 159 (emphasis in original); see also Laws, Psychological Dimensions of Labor Force Participation of Women, in EQUAL EMPLOYMENT OPPORTUNITY AND THE AT&T CASE, supra note 25, at 125, 141 ("When women lower their occupational aspirations, this may reflect a realistic assessment of their chances for success.").
382. See R. KANTER, supra note 281, at 69-91.
383. See id. at 91-99.
384. There are several ways in which employers have structured traditionally female jobs so as to require or encourage women to display behaviors that are commonly viewed as preexisting attributes of womanhood. See, e.g., V. BEECHEY & T. PERKINS, supra note 14, at 45-76, 77-101 (showing how employers have built gender into the way they structure hours, achieving flexibility in male jobs by adding overtime to full-time jobs, but doing so in female jobs by constructing them as part-time); S. COHN, THE PROCESS OF OCCUPATIONAL SEX-TYPING: THE FEMINIZATION OF CLERICAL LABOR IN GREAT BRITAIN 91-115 (1985) (showing how employers have forced women to quit work when they marry as a way of lowering labor costs in certain female jobs and have legitimated such "marriage bars" by describing women as uncommitted to wage work); B. GUTEK, SEX AND THE WORKPLACE 134-36, 142-46 (1985) (showing how employers have built sexual attractiveness into the very definition of what is required to do traditionally female jobs).
385. There are several ways in which employers have structured traditionally female jobs so as to require or encourage women to display behaviors that are commonly viewed as preexisting attributes of womanhood. See, e.g., V. BEECHEY & T. PERKINS, supra note 14, at 45-76, 77-101 (showing how employers have built gender into the way they structure hours, achieving flexibility in male jobs by adding overtime to full-time jobs, but doing so in female jobs by constructing them as part-time); S. COHN, THE PROCESS OF OCCUPATIONAL SEX-TYPING: THE FEMINIZATION OF CLERICAL LABOR IN GREAT BRITAIN 91-115 (1985) (showing how employers have forced women to quit work when they marry as a way of lowering labor costs in certain female jobs and have legitimated such "marriage bars" by describing women as uncommitted to wage work); B. GUTEK, SEX AND THE WORKPLACE 134-36, 142-46 (1985) (showing how employers have built sexual attractiveness into the very definition of what is required to do traditionally female jobs).
386. Kanter found that the mean score on a measure of motivation to be promoted was lower tor nonmanagerial women workers (mostly secretaries) than for nonmanagerial men. However, the men's objective prospects for promotion were also better. Indeed, the women consistently rated their promotional prospects as more desirable than likely. They believed they had no chance of escaping low-level, female-dominated jobs. See R. KANTER, supra note 281, at 140-42.
387. See id. at 149-52.
388. See id. at 151.
389. See id. at 159.
390. Many women understood that they had obtained their jobs only as a result of legal pressures. See, e.g., M. MARTIN, supra note 271, at 150 ("The company was pushing affirmative action, because it had a class-action suit brought against it by a group of women in the mines in 1973 I was hired four years after the suit was filed, but even then, there were only a few Women working for the company."); id. at 71 ("The process of entering the San Francisco Police Department . . . started for me in 1973. That's when community groups got together and filed a suit to open up the job to women and minorities . . . .").
391. As one former secretary explained: "I didn't start thinking about non-traditional work until I heard the carpenters were looking for women. . . . But as soon as the possibility was mentioned, my imagination went with it." J. SCHROEDEL, supra note 271, at 35.
392. One woman described the transformative power of seeing women doing nontraditional work as follows: When I came out here I fell in with some women who worked in the trades and they had some potlucks for women in the building trades and I went there and I saw all these women and I was real excited - I thought, "Oh, yeah, that's who I am, I'm like those women over there." M. WALSHOK, supra note 262, at 137-38; see also id. at 163-64 (describing a similar transformation).
393. As one woman who became a sailor explained: [I]t wasn't until I moved to Seattle when I was surrounded by organizations and groups that seemed encouraging of this - just seeing flyers about workshops on women in nontraditional trades, having Mechanica available where you could learn the details about steps in joining a union. That's when it became a real possibility. J. SCHROEDEL, supra note 271, at 77; see also M. WALSHOK, supra note 262, at 167-68 (discussing the importance of community-based programs in inspiring and helping women enter nontraditional trades); Law, supra note 163, at 45-46, 53-55, 72-76 (same).
394. M. WALSHOK, supra note 262, at 140.
395. Id. at 142.
396. See id. at 147-48.
397. Id. at 145. Other studies have reported that women in blue-collar trades value the challenge, freedom, and intrinsic rewards of the job, just as their male co-workers do. See, e.g., K. DEAUX & J. ULLMAN, supra note 25, at 131-33. The women speak movingly of the exhilaration that comes with challenge and freedom on the job. See, e.g., M. MARTIN, supra note 271, at 167-68. One female firefighter stated: For nine days, I was part of the biggest [fire] incident I ever expect to see. . . . I've never worked as hard as I did on some of those hot afternoons, pulling those lines around in the mud and rocks. . . . Events that demand everything you can give leave you with an unconquerable feeling of exuberance that lasts well beyond fatigue. Given the choice, there was no place in the world I would rather have been. Id.; see also J. SCHROEDEL, supra note 271, at 14 ("That is the greatest feeling in the world - the music, the sun, and wheeling along the freeway [in my truck]. . . . I like being on the road where you haven't got somebody looking over your shoulder, bitching all the time.").
398. See M. WALSHOK, supra note 262, at xix-xx.
399. For suggestions for how to change work structures in ways that will both reduce sex segregation and improve the quality of worklife for all workers, see C. COCKBURN, cited above in note 14, at 242-44; and R. KANTER, cited above in note 281, at 267-84.
400. In several cases, plaintiffs have alleged that women in entry-level female jobs were systematically denied promotion or transfer into higher-paid male jobs on separate career ladders. See, e.g., Kyriazi v. Western Elec. Co., 461 F. Supp. 894 (D.N.J. 1978), aff'd on other grounds, 647 F.2d 388 (3d Cir. 1981); Wetzel v. Liberty Mut. Ins. Co., 372 F. Supp. 1146 (W.D. Pa. 1974), aff'd, 511 F.2d 199 (3d Cir. 1975), vacated and remanded on other grounds, 424 U.S. 737 (1976); see also B. BERGMANN, supra note 1, at 106-10 (discussing these two cases). In other cases, women employed in what we think of as nontraditional positions (such as management) have alleged that they were disproportionately denied promotion into the upper echelons of were given discriminatory work assignments that decreased their chances for promotion later. See, e.g., Palmer v. Shultz, 815 F.2d 84 (D.C. Cir. 1987); Leisner v. New York Tel. Co., 358 F. Supp. 359 (S.D.N.Y. 1973).
401. Although traditional goals and timetables probably provide the best incentive for employers to change practices that lead to segregation, some lower courts who have held employers liable for sex segregation have declined to impose such numerical relief. See, e.g., Catlett v. Missouri Highway & Transp. Comm'n, 828 F.2d 1260, 1268-69 (8th Cir. 1987), cert. denied, 485 U.S. 1021 (1988); Jordan v. Wright, 417 F. Supp. 42, 45 (M.D. Ala. 1976). Recent Supreme Court decisions permit the trial courts to use flexible long-term goals as remedies for proven patterns of discrimination, but establish relatively high standards that may make it even easier for trial courts who are so inclined to refuse to grant such relief. See United States v. Paradise, 480 U.S. 149, 167 (1987) (holding numerical relief permissible under the fourteenth amendment when the employer has engaged in "pervasive, systematic, and obstinate" conduct); Local 28, Sheet Metal Workers' Int'l Ass'n v. EEOC, 478 U.S. 421, 445 (1986) (holding numerical relief Permissible under title VII when the employer has engaged in "persistent or egregious discrimination or when necessary to dissipate the lingering effects of pervasive discrimination").
402. See, e.g., G. NIELSEN, FROM SKY GIRL TO FLIGHT ATTENDANT: WOMEN AND THE MAKING OF A UNION 81-103 (1982) (describing how early title VII decisions helped redefine the flight attendant job in less patronizing, more professional terms).
403. In contrast to my definition of harassment, the legal system focuses on conduct that is explicitly "sexual" in nature. EEOC guidelines, for example, define harassment in these terms: Unwelcome sexual advances, requests for sexual favors, and other verbal or physical conduct of a sexual nature . . . when (1) submission to such conduct is made explicitly or implicitly a term or condition of an individual's employment, (2) submission to or rejection of such conduct by an individual is used as the basis for employment decisions affecting such individual, or (3) such conduct has the purpose or effect of unreasonably interfering with an individual's work performance or creating an intimidating, hostile, or offensive work environment. EEOC Guidelines on Discrimination Because of Sex, 29 C.F.R. § 1604.11(a) (1989) (emphasis added). Drawing on the guidelines, legal doctrine recognizes two different types of sexual harassment: (1) "quid pro quo" harassment, in which women workers are asked to grant sexual favors at the risk of forfeiting some employment benefit; and (2) "hostile environment harassment," in which conduct by supervisors or co-workers is "sufficiently severe or pervasive, to alter the conditions of [the victim's] employment and create an abusive work environment." Meritor Savings Bank v. Vinson, 477 U.S. 57, 67 (1986) (quoting Henson v. Dundee, 682 F.2d 897, 904 (11th Cir. 1982)); see also C. MACKINNON, supra note 212, at 32-47 (describing these forms of harassment).
404. For women in male-dominated jobs, harassment is less likely to take the form of supervisors' demands for sexual favors and more likely to take the form of sexual taunts and other actions by co-workers that are part of a larger pattern of hostility intended to drive the women away. Foremen and supervisors usually tolerate or cooperate in the harassment. See Crull, Searching for the Causes of Sexual Harassment: An Examination of Two Prototypes, in HIDDEN ASPECTS OF WOMEN'S WORK, supra note 17, at 225, 228-30; Pollack, Sexual Harassment: Women's Experience vs. Legal Definitions, 13 HARV. WOMEN'S L.J. 35, 37, 50 n.51 (1990).
405. For women in male-dominated jobs, harassment is less likely to take the form of supervisors' demands for sexual favors and more likely to take the form of sexual taunts and other actions by co-workers that are part of a larger pattern of hostility intended to drive the women away. Foremen and supervisors usually tolerate or cooperate in the harassment. See Crull, Searching for the Causes of Sexual Harassment: An Examination of Two Prototypes, in HIDDEN ASPECTS OF WOMEN'S WORK, supra note 17, at 225, 228-30; Pollack, Sexual Harassment: Women's Experience vs. Legal Definitions, 13 HARV. WOMEN'S L.J. 35, 37, 50 n.51 (1990).
406. See M. MARTIN, supra note 271, at 10. This broader form of harassment is so much a part of the "normal" environment of traditionally male-dominated trades that some researchers do not even attempt to measure it. Mary Walshok observed, for example, that it is "normal" for men in blue-collar trades to question the sincerity of the woman's interest and commitment to a man's job, to wonder about whether or not the woman was going to get married and take off or get pregnant, to question whether the woman had technical or mechanical competence or the physical strength and agility to do the job, and to resent women because they perceived them as taking away a job from one of their own. M. WALSHOK, supra note 262, at 211. This led Walshok to define "negative work environments" as those that go beyond this "normal" treatment to involve "actual acts of hostility or sabotage, withholding of opportunities for information and training, persistent sexual innuendos, and open harassment." Id. at 211-12. Even using this narrow definition, Walshok found that approximately half of the women had negative relationships with their supervisors and co-workers during their first year on the job. See id. at 188, 221.
407. Id. at 221-22.
408. The literature documenting the effects of skewed sex ratios on work groups makes this clear. See, e.g., B. GUTEK, supra note 302, at 129-52; R. KANTER, supra note 281, at 206-42.
409. This literature is now far too extensive to cite. Some of the earliest surveys documenting the existence of sexual harassment are described in C. MACKINNON, cited above in note 212, at 26-30. More recent studies include B. GUTEK, cited above in note 302; U.S. MERIT SYSTEM PROTECTION BD., SEXUAL HARASSMENT IN THE FEDERAL WORKPLACE: IS IT A PROBLEM? (1981); and U.S. MERIT SYSTEMS PROTECTION BD., SEXUAL HARASSMENT IN THE FEDERAL GOVERNMENT: AN UPDATE (1988).
410. This literature is now far too extensive to cite. Some of the earliest surveys documenting the existence of sexual harassment are described in C. MACKINNON, cited above in note 212, at 26-30. More recent studies include B. GUTEK, cited above in note 302; U.S. MERIT SYSTEM PROTECTION BD., SEXUAL HARASSMENT IN THE FEDERAL WORKPLACE: IS IT A PROBLEM? (1981); and U.S. MERIT SYSTEMS PROTECTION BD., SEXUAL HARASSMENT IN THE FEDERAL GOVERNMENT: AN UPDATE (1988).
411. See, e.g., Gutek & Morasch, supra note 1, at 67-68 (finding that women in maledominated occupations and jobs were more likely to report harassment and to have experienced negative consequences from it than women in other work settings); Martin, Sexual Harassment: The Link Joining Gender Stratification, Sexuality, and Women's Economic Status, in WOMEN: A FEMINIST PERSPECTIVE 57, 61 (J. Freeman 4th ed. 1989) (citing studies showing that the greater the proportion of men in a work group, the more likely women were to be harassed).
412. See, e.g., O'Farrell & Harlan, 29 Soc. PROBS. 252, 259 (1982) (finding that half the women in white-collar, female-dominated occupations who considered moving into blue-collar, male-dominated occupations expected that they would be subjected to harassment if they did so).
413. See B. GUTEK, supra note 302, at 119 ("By making insulting comments and touching women sexually, some men may try to 'make life miserable' for women in the [nontraditional] jobs, encouraging them to leave. The relatively high turnover rate among women in [these jobs] suggests that this is a successful strategy to force women out."); Gutek & Morasch, supra note 1, at 68 (finding that 20% of women in nontraditional work quit a job at some point because of sexual harassment, while only 9% of the larger sample did so). This research exposes a methodological problem in most surveys attempting to measure the extent of harassment, particularly among women in male-dominated occupations. Most of the surveys are not based on longitudinal data and thus do not include women who left because of harassment. For this reason, they probably underestimate the prevalence of harassment. See J. JACOBS, supra note 1, at 153.
414. See, e.g., M. MARTIN, supra note 271, at 11; Eisenberg, Women Hard Hats Speak Out, NATION, Sept. 18, 1989, at 272-73; Pollack, supra note 321, at 37-38.
415. See, e.g., M. MARTIN, supra note 271, at 11; Eisenberg, Women Hard Hats Speak Out, NATION, Sept. 18, 1989, at 272-73; Pollack, supra note 321, at 37-38.
416. See, e.g., M. MARTIN, supra note 271, at 11; Eisenberg, Women Hard Hats Speak Out, NATION, Sept. 18, 1989, at 272-73; Pollack, supra note 321, at 37-38.
417. A recent newspaper article noted, for example, that a growing number of women engineers have become so discouraged by their discriminatory treatment that they are leaving engineering to pursue alternative careers. See Arundel, Stagflation for Female Engineers, N.Y. Times, Oct. 1, 1989, at F32. Other studies have documented the disproportionate attrition of women attorneys from law firms, see, e.g., Menkel-Meadow, Exploring a Research Agenda on the Feminization of the Legal Profession: Theories of Gender and Social Change, 14 LAW & Soc. INQUIRY 289, 307 (1989); Weisenhaus, Still a Long Way To Go for Women, Minorities, Nat'l L.J., Feb. 8, 1988, at 48, and there is evidence that women leave in part because of discrimination, see, e.g., Liefland, Career Patterns of Male and Female Lawyers, 35 BUFFALO L. REV. 601, 609-11 (1986); Quade, Myth v. Ms.: Why Women Leave the Law, 13 BARRISTER 28 (1986).
418. A recent newspaper article noted, for example, that a growing number of women engineers have become so discouraged by their discriminatory treatment that they are leaving engineering to pursue alternative careers. See Arundel, Stagflation for Female Engineers, N.Y. Times, Oct. 1, 1989, at F32. Other studies have documented the disproportionate attrition of women attorneys from law firms, see, e.g., Menkel-Meadow, Exploring a Research Agenda on the Feminization of the Legal Profession: Theories of Gender and Social Change, 14 LAW & Soc. INQUIRY 289, 307 (1989); Weisenhaus, Still a Long Way To Go for Women, Minorities, Nat'l L.J., Feb. 8, 1988, at 48, and there is evidence that women leave in part because of discrimination, see, e.g., Liefland, Career Patterns of Male and Female Lawyers, 35 BUFFALO L. REV. 601, 609-11 (1986); Quade, Myth v. Ms.: Why Women Leave the Law, 13 BARRISTER 28 (1986).
419. A recent newspaper article noted, for example, that a growing number of women engineers have become so discouraged by their discriminatory treatment that they are leaving engineering to pursue alternative careers. See Arundel, Stagflation for Female Engineers, N.Y. Times, Oct. 1, 1989, at F32. Other studies have documented the disproportionate attrition of women attorneys from law firms, see, e.g., Menkel-Meadow, Exploring a Research Agenda on the Feminization of the Legal Profession: Theories of Gender and Social Change, 14 LAW & Soc. INQUIRY 289, 307 (1989); Weisenhaus, Still a Long Way To Go for Women, Minorities, Nat'l L.J., Feb. 8, 1988, at 48, and there is evidence that women leave in part because of discrimination, see, e.g., Liefland, Career Patterns of Male and Female Lawyers, 35 BUFFALO L. REV. 601, 609-11 (1986); Quade, Myth v. Ms.: Why Women Leave the Law, 13 BARRISTER 28 (1986).
420. A recent newspaper article noted, for example, that a growing number of women engineers have become so discouraged by their discriminatory treatment that they are leaving engineering to pursue alternative careers. See Arundel, Stagflation for Female Engineers, N.Y. Times, Oct. 1, 1989, at F32. Other studies have documented the disproportionate attrition of women attorneys from law firms, see, e.g., Menkel-Meadow, Exploring a Research Agenda on the Feminization of the Legal Profession: Theories of Gender and Social Change, 14 LAW & Soc. INQUIRY 289, 307 (1989); Weisenhaus, Still a Long Way To Go for Women, Minorities, Nat'l L.J., Feb. 8, 1988, at 48, and there is evidence that women leave in part because of discrimination, see, e.g., Liefland, Career
421. A recent newspaper article noted, for example, that a growing number of women engineers have become so discouraged by their discriminatory treatment that they are leaving engineering to pursue alternative careers. See Arundel, Stagflation for Female Engineers, N.Y. Times, Oct. 1, 1989, at F32. Other studies have documented the disproportionate attrition of women attorneys from law firms, see, e.g., Menkel-Meadow, Exploring a Research Agenda on the Feminization of the Legal Profession: Theories of Gender and Social Change, 14 LAW & Soc. INQUIRY 289, 307 (1989); Weisenhaus, Still a Long Way To Go for Women, Minorities, Nat'l L.J., Feb. 8, 1988, at 48, and there is evidence that women leave in part because of discrimination, see, e.g., Liefland, Career Patterns of Male and Female Lawyers, 35 BUFFALO L. REV. 601, 609-11 (1986); Quade, Myth v. Ms.: Why Women Leave the Law, 13 BARRISTER 28 (1986).
422. In Walshok's study, when the women were asked about the negative aspects of their job during the first year, the most frequently voiced criticism (expressed by 68% of the women) was that they felt they were being trained poorly. See M. WALSHOK, supra note 262, at 188.
423. Stories like electrician Sue Eisenberg's are still far too common: For some men, getting rid of the invaders was a personal mission. Ron, one of my first foremen, constantly warned me of the ways I might get killed in this dangerous trade: be electrocuted, have my head severed from my body, be boiled alive by steam. Without giving any instruction on how to do it safely, he told me one day to open up a 200-foot-long snake . . . . A snake is a thin piece of steel, used by electricians to pull wires through pipes. It comes tightly coiled, bound with wire ties, and if not opened carefully, will spring apart with great force. "I had a Chinese kid open one up," Ron told the crew, laughing. "He got it caught up his nose and wound up in the hospital. Quit right after that." I haven't opened up a snake since without remembering how I sweated through it that first time, while my co-workers hid. Eisenberg, supra note 329, at 272. For other stories of how women have been subjected to acts by foremen or co-workers that threatened them with, or caused them, physical harm, see, for example, M. MARTIN, cited above in note 271, at 33-34 ("[The men] didn't want the women to replace them, so they pulled stunts. Someone cut the chain holding up a big motor mount I was welding. It fell down on me and burned my arm to the bone."); id. at 257 ("I had to start checking all the parts on my machine because Dick would loosen stuff on it, which could Kill you."); and J. SCHROEDEL, cited above in note 271, at 256-57 ("I went in the women's room, and I cried, because a man pushed me under a machine. . . . The men admit they think it's a man's job and a woman has no right out there.").
424. For example, one of Atlanta's most prestigious corporate law firms, King & Spalding, planned to hold a "wet T-shirt" contest featuring its female summer associates, even while the firm faced a sex discrimination lawsuit in the Supreme Court. See Burleigh & Goldberg, Breaking the Silence: Sexual Harassment in Law Firms, A.B.A. J., Aug. 1989, at 46. (The lawsuit was Hishon V. King & Spalding, 467 U.S. 69 (1984).) After complaints, the firm decided to hold a swimsuit competition instead. One of the firm's partners later told the Wall Street Journal that the "winner" of the competition had been offered a job upon graduation, remarking: "She has the body we'd like to see more of." Burleigh & Goldberg, supra, at 46.
425. In one recent case, for example, the lone female resident in a general surgery program was forced to endure sexual advances and touching; sexually explicit drawings of her body and other pornography in public meeting rooms; her supervisors' refusal to talk to her, permit her to operate, or assign her work tasks; discriminatory standards for evaluating her performance; sabotage of her work, including falsification of medical records to make it appear as though she and another female resident had made an error, see Lipsett v. University of Puerto Rico, 864 F.2d 881, 886-94 (1st Cir. 1988), and "a constant verbal attack, one which challenged their capacity as women to be surgeons, and questioned the legitimacy of their being in the Program at all," id. at 905.
426. See, e.g., Price Waterhouse v. Hopkins, 109 S. Ct. 1775 (1989). Ann Hopkins brought suit against Price Waterhouse, perhaps the nation's most prestigious public accounting firm, claiming that she was discriminatorily denied partnership. Among other outstanding achievements, she had helped secure a multimillion dollar contract with the Department of State, an accomplishment that none of the other candidates for partnership that year had matched. See id. at 1782. But, when it came time to consider her for partnership, her colleagues evaluated her by criteria by which no man would be judged: One partner described her as "macho" . . . ; another suggested that she "overcompensated for being a woman" . . . ; a third advised her to take "a course at charm school" . . . . Several partners criticized her use of profanity. . . . [Another advised her to] "walk more femininely, talk more femininely, dress more femininely, wear make-up, have her hair styled, and wear jewelry." Id. (citations omitted).
427. Some researchers emphasize that men are motivated by economic incentives, because they feel that their job security and high wages are threatened by the presence of women. See, e.g., R. MILKMAN, supra note 148, at 7-8, 158. Other writers have stressed patriarchal motives, arguing that traditional sex roles "spill over" into the workplace. See, e.g., B. GUTEK, supra note 302, at 149-51.
428. See V. BEECHEY & T. PERKINS, supra note 14, at 102-19; C. WILLIAMS, supra note 266, at 88-130. These studies show how the cultural constructions of the same job vary depending upon whether men or women do it. For example, one machine tools company that employed only men as crane operators explained that women did not want to drive cranes because "that was hot, heavy, dirty work and women didn't do that sort of work." V. BEECHEY & T. PERKINS, supra note 14, at 105. But another such company that employed women defined the job in feminine terms, and suggested that women were actually better crane operators because they had a "sensitive touch" learned through knitting. See id. at 106.
429. J. SCHROEDEL, supra note 271, at 20-21; see also C. WILLIAMS, supra note 266, at 61 (noting that the Marine Corps segregates basic training because "[r]egardless of whether training standards are compromised in fact, the sight of women mastering the feats of basic training makes it appear that the training is not rigorous enough").
430. One of the tactics men use to make nontraditional women workers feel deviant is "lesbian-baiting." See M. MARTIN, supra note 271, at 14. Lesbian-baiting is only one extreme example, however, of behavior intended to divide and alienate women. See Pollack, supra note 321, at 78 & n.178.
431. Kanter discusses how male workers subject token women to "loyalty tests," in which the women are required to affirm their loyalty to the dominant culture by turning against other women. Pressuring women to tolerate sexist jokes and comments are examples. See R. KANTER, supra note 281, at 227-29. Racist jokes and comments are often used in the same manner, The invitation to participate in racist "humor" is one of the ways white women are often "welcomed into the club." M. MARTIN, supra note 271, at 13. The pervasive racism of many nontraditional blue-collar environments makes them particularly difficult for women of color, who often encounter a virulent combination of sexual and racial stereotyping. For some of their stories, see id., at 71-80, 118-21, 150-55, 187-92; and J. SCHROEDEL, cited above in note 271, at 99-139. One of the most shocking stories is police officer Rose Melendez's. One night her partner drove her to a secluded area and stopped the patrol car. When Melendez asked why they were stopping, he said he wanted to shoot rats. He then opened his shirt, pulled out a handgun, and pointed it directly at her, saying; "I just want to see how fast you women cops can run." M. MARTIN, supra note 271, at 74.
432. See C. COCKBURN, supra note 14.
433. Id. at 172.
434. Id. at 185.
435. Id.
436. Id. at 186.
437. Id. at 101 (emphasis in original).
438. Id. at 188.
439. Id. at 176.
440. The contradictions within the men's ideological justifications are exposed as the men's careers unfold. Early in their careers, when they do hands-on machine work, male engineers defend the masculinity of their jobs in terms of a hard/soft dichotomy that defines "hard," physical work as masculine and "soft," intellectual work as feminine. In middle age, however, many of these same engineers must move on to managerial desk jobs that they once denigrated as unmanly. Id. at 195. They then adopt an intellectual/non-intellectual dichotomy that, ironically, associates masculinity with the intellectual and femininity with the physical. See id. at 196-97.
441. It remains unclear, for example, whether hostile but not sexually explicit behavior of the type so often encountered by nontraditional women workers even falls within the scope of sexual harassment prohibited by title VII. Compare McKinney v. Dole, 765 F.zd 1129, 1138 (D.C. Cir. 1985) (holding that acts that are hostile but not sexually explicit may constitute sexual harassment prohibited by title VII) with Rabidue v. Osceola Ref. Co., 805 F.2d 611, 619 (6th Cir. 1986), cert, denied, 481 U.S. 1041 (1987) (implicitly refusing to recognize hostile conduct as harassment when it is not explicitly sexual in nature). For discussions of this issue, see Austin, Employer Abuse, Worker Resistance, and the Tort of Intentional Infliction of Emotional Distress, 41 STAN. L. REV. 1, 12-13 (1988); and Pollack, cited above in note 321, at 75 & n.164.
442. It remains unclear, for example, whether hostile but not sexually explicit behavior of the type so often encountered by nontraditional women workers even falls within the scope of sexual harassment prohibited by title VII. Compare McKinney v. Dole, 765 F.zd 1129, 1138 (D.C. Cir. 1985) (holding that acts that are hostile but not sexually explicit may constitute sexual harassment prohibited by title VII) with Rabidue v. Osceola Ref. Co., 805 F.2d 611, 619 (6th Cir. 1986), cert, denied, 481 U.S. 1041 (1987) (implicitly refusing to recognize hostile conduct as harassment when it is not explicitly sexual in nature). For discussions of this issue, see Austin, Employer Abuse, Worker Resistance, and the Tort of Intentional Infliction of Emotional Distress, 41 STAN. L. REV. 1, 12-13 (1988); and Pollack, cited above in note 321, at 75 & n.164.
443. See, e.g., Collins & Blodgett, Sexual Harassment . . . Some See It . . . Some Won't, 59 HARV. Bus. REV. 76, 90 (1981) (noting that a majority of managers responding to a survey said that women employees should be able to handle on their own whatever sexual harassment comes their way).
444. The clearest expression of this attitude appears in the Sixth Circuit's opinion in Rabidue v. Osceola Refining Co., 805 F.2d 611 (6th Cir. 1986), cert. denied, 481 U.S. 1041 (1987). The plaintiff was the only woman with managerial responsibilities over male employees in a refining company. Despite the fact that the work environment was extremely hostile and degrading to women, the court held that it was not so offensive as to have interfered with the work performance, or to have affected seriously the "psychological well-being of a reasonable person." Id. at 620. To support this holding, the majority quoted favorably from the following passage from the district court's opinion: [I]t cannot seriously be disputed that in some work environments, humor and language are rough hewn and vulgar. Sexual jokes, sexual conversations and girlie magazines may abound. Title VII was not meant to - or can - change this. . . . Title VII is the federal court mainstay in the struggle for equal employment opportunity for the female workers of America. But . . . Title VII was [not] designed to bring about a magical transformation in the social mores of American workers. Id. at 620-21 (quoting Osceola v. Rabidue, 584 F. Supp. 419, 430 (E.D. Mich. 1984)).