The academic employment situation has been described in terms of rough global comparisons-black-white or male-female. Finer breakdowns are necessary in order for us to determine the effects of many variables which differ between the groups whose economic conditions are being compared. Some of these intergroup differences have already been mentioned-educational differences and differences in publications, for example-but there are others as well. If discrimination is to mean unequal treatment of equal individuals, then comparisons must be made between individuals who are similar with respect to the variables which generally determine employment, pay, and promotion. Only insofar as we succeed in specifying all these variables can we confidently refer to the remaining economic differences as "discrimination." One of the perverse aspects of this residual method of measuring discrimination is that the more determining variables that are overlooked or ignored, the more discrimination there seems to be. Since no study can specify all relevant variables, the residual pay differences between minority and female academics, on the one hand, and white males, on the other, must be understood as the upper limit of an estimate of discriminatory differences.
          For both sets of comparisons, the data sources are the American Council on Education (ACE) and the National Academy of Sciences (NAS). The AGE data are based on a sample of 60,028 academicians surveyed in 1969 and a sample of 50,034 academicians surveyed in 1972. The NAS data are from (1) a National Science Foundation survey conducted in 1973, based on a stratified sample of 59,086 doctorates in the social and natural sciences and engineering1 and (2) a longitudinal compilation by NAS of biennial surveys of the same target population by the National Science Foundation during 1960-70.

Minorities. Existing studies of black faculty members show many ways in which their job-related characteristics differ from those of faculty members in general. All these differences tend to have a negative impact on employment, pay, and promotion for academics in general:

(1) A smaller proportion of black faculty than of white faculty holds a doctoral degree.2
(2) The distribution of black doctoral fields of specialization is biased towards the lower-paying fields-particularly education (roughly one-third of all black doctorates) and the social sciences (one-fourth)-with very few (about 10 percent) of the doctorates in the natural sciences.3
(3) The bulk of black faculty is located in the South4-a lower-paying region for academics in general,5 as well as for others.
(4) Blacks complete their Ph.D.s at a later age than whites6-a reflection of both financial and educational disadvantages-and academics in general who complete their Ph.D.s at a later age tend to be less "productive" in research publications.
(5) Black academics, both at black colleges and at white institutions, publish much less than white academics.7  Among the factors associated with this are much higher teaching loads and late completion of the Ph.D.
(6) Black academics are less mobile than white academics-and less mobile academics tend to earn lower salaries. Forty percent of the black professors in the Mommsen study had not moved at all, despite an average of three or four job offers per year,8 and the median pay increase which they considered necessary to make them move was a $6,134 per year raise.9 By contrast, among faculty in general, "the academic career is marked by high mobility,"10 and "professors expect to switch schools several times, at least, during their careers."11
(7) Women constitute a higher proportion (20 percent) of black doctorates than of doctorates in general (13 percent)12-and women earn less than men among both blacks and whites.
          With all these downward biases, it is worth noting once again that the academic salaries of white doctorates averaged only $62 per year more than those of black doctorates in 1970. On a field-by-field basis, black doctorates were generally earning more than white doctorates in the same area of specialization and receiving more job offers per year13-all this before the affirmative action program under Revised Order No. 4 in 1971. In other words, the effect of the straightforward antidiscrimination laws of the 1960s and of the general drive toward racial integration had created a premium for qualified black academics, even before HEW's goals and timetables. Moreover, the improvements that have occurred since then need not be due to HEW pressures but may be thought -of as a continuation of trends already evident before affirmative action programs began.
          The data from the American Council on Education permit a standardization for degree level, degree quality, field of specialization, and number of articles published, so that the salaries of blacks, whites, and Orientals who are comparable in these respects may be compared. Table 1 omits field of specialization to give a general view of race and salary in the academic world as a whole. Degree rankings in the table are based on surveys conducted by the AGE to determine the relative rankings of Ph.D.-granting departments in twenty-nine disciplines, as ranked by members of those respective disciplines. (I have collapsed the two departmental rankings, "distinguished" and "strong," into one category in order to maintain a large sample size.) Articles published were selected as a proxy for publication in general, avoiding the problem of trying to convert books, monographs, conference papers, and articles into some equivalent.
          My results for 1973 (Table 1) are generally not very different from those of Professor Mommsen for 1970: white faculty earned slightly more than black faculty in general ($16,677 versus $16,037). But when degree level and degree quality are held constant, blacks earned more than whites with doctorates of whatever ranking, while whites had an edge of less than $100 per year among academics without a doctorate. The overall salary advantage of whites over blacks-$640 per year-is a result of a different distribution of the races among degree levels and degree qualities, as well as a different distribution among publication categories. For example, 11 percent of the white faculty members in the AGE samples had Ph.D.s from departments ranked either "distinguished" or "strong" by their respective professions, while only 4 percent of the black faculty came from such departments. Only 18 percent of the black academics in this sample had a doctorate at all, compared to 38 percent of the white academics. Thirty-one percent of the white faculty had published five or more articles while only 12 percent of the black faculty had done so. Blacks who had published at all had higher salaries than whites with the same number of publications.
          Orientals present a somewhat different picture. Only those Orientals with "distinguished" and "strong" Ph.D.s received slightly higher salaries than their white counterparts ($18,235 versus $17,991), and even this difference was not uniform across publications categories. Among the lower-ranked doctorates, both whites and blacks earned more than Orientals, and among those with less than a doctorate, considerably more. The overall salary average of Orientals was only slightly below that of blacks, but solely because Orientals were far more concentrated in the higher degree levels and higher degree qualities. Less than half of the Oriental faculty members lacked the Ph.D. and more than 40 percent of all Oriental faculty had published five or more articles. In short, just as group differentials do not imply discrimination, so an absence of such differentials does not imply an absence of discrimination. Orientals receive less than either blacks or whites with the same qualifications, and only the fact that the Orientals have generally better qualifications than either of the other two groups conceals this.
          When field-by-field comparisons are made, very similar patterns emerge. In the social sciences, blacks have higher salaries than whites or Orientals, and especially so among holders of Ph.D.s from "distinguished" and "strong" departments (Table 2). In the natural sciences (Table 3) and the humanities (Table 4), whites lead, with blacks second in the humanities and Orientals second in the natural sciences. A comparison of overall sample size from one table to another reveals very different distributions of these racial groups among academic fields: 37 percent of all black faculty members were in the social sciences, 23 percent were in the humanities, and only 16 percent were in the natural sciences. By contrast, 44 percent of the Orientals were in the natural sciences, 28 percent in the social sciences, and only 16 percent in the humanities. Whites were distributed more or less midway between blacks and Orientals: 30 percent in the social sciences, 24 percent in the humanities, and 25 percent in the natural sciences. Again, the net effect of these distributions is to exaggerate the overall salary differences between blacks and whites and to understate salary differences between Orientals and whites.
          The National Academy of Sciences data confirm some of these patterns and reveal some new ones. NAS data for full-time doctoral scientists and engineers (academic and nonacademic) show blacks earning slightly more than whites, with Orientals last-and a spread of only $1,500 per year over all three groups (Table 5). Publications data are not available for this survey but age was tabulated as a proxy for experience. Degree quality was again available, and again Orientals with given credentials quality had lower salaries than either blacks or whites in the same categories. In all three groups, salary rises with age, but the relative positions of blacks and whites are reversed in the oldest and youngest age brackets. Young black doctorates-under thirty-five-earn more than their white counterparts in either degree quality category, but older blacks-over fifty-earn less than their white counterparts in either degree quality category. These results hold up when the sample is broken down into natural sciences and social sciences. It is also consistent with a larger study by Professor Finis Welch of UCLA which showed a much higher rate of return to education for younger blacks than for older blacks-both absolutely and relative to their white counterparts." Two important factors are involved here: (1) the older blacks were educated in an era when their public school education was inferior not only by various quality measures but also in sheer quantity (black schools had fewer days than white schools in their respective school years)," and this poorer preparation could not help affecting later capability, and (2) the level of job discrimination was also greater when the older blacks began their careers, and this too could not help affecting the later course of those careers, making it difficult for these blacks to exploit new opportunities as readily as the younger blacks just beginning their careers. A further implication of all this is that global comparisons of blacks and whites capture many existing effects of past discrimination, while an age-cohort breakdown of the same data permits a better look at the current results of current policies and the trends to expect in the future.
          In summary, the salary differentials among these three racial or ethnic groups are small, both in the academic world and among holders of the doctorate in the social or natural sciences (academic and nonacademic). With such variables as credentials, publications, and experience held constant, blacks equalled or surpassed whites in 1973-but they also equalled or surpassed whites with fields held constant in 1970. Without these variables held constant, the overall black-white differential was $62 per year in 1970 and $640 in 1973. Given that these are different samples, it is perhaps best to say that there were negligible overall differences among black and white academics in both years-that affirmative action has achieved nothing discernible in this regard. But if an arithmetic conclusion is insisted upon, then it must be said that there has been a negative effect of affirmative action as far as black-white differences are concerned.

Women. The classic study Academic Women, by Jessie Bernard described women as "overrepresented in college teaching." This was based on the fact that women were only 10 percent of the Ph.D.s but constituted more than 20 percent of college and university faculties." This was written in 1964-before affirmative action. Unlike HEW's crude "underutilization" measures, this study (by an academic woman) considered not only the number of women with the usual degree requirements but also "the large number of educated women-30.6 percent of those with five years or more of college-who are not in the labor force."17 Withdrawal from the labor force is only one of many career characteristics which have a negative effect on the employment, pay, and promotion of academic women. Some others are:

(1) Female academics hold a doctorate less frequently than male academics-20 percent as against 40 percent in 1972-73.18
(2) Female academics publish only about half as many articles and books per person as do male academics,19 and females are especially underrepresented among frequent publishers.20
(3) Academic women are educated disproportionately in lower-paying fields of specialization, such as the humanities,21 and they prefer teaching over research more so than academic men, not only in attitude surveys,22 but also in their allocation of time23 and in the kinds of institutions at which they work24-which are the low-paying teaching institutions more so than the top research universities with high salaries. 
(4) Academic women more frequently subordinate their careers to their spouses' careers, or to the general well-being of their families, than do academic men. This takes many forms, including quitting jobs they like because their husbands take jobs elsewhere,25 interrupting their careers for domestic reasons,26 withdrawing from the labor force (25 percent of women Ph.D.s)27 doing a disproportionate share of household and social chores compared to their husbands in the same occupations28 and a general attitude reported by women themselves of putting their homes and families ahead of their careers much more often than do male academics.29 All this goes to the heart of the question of the actual source of sex differentiation-whether it is the home or the work place, and therefore whether "equal treatment" as required by the Constitution and envisioned in the Civil Rights Act would eliminate or ensure unequal results by sex.
          None of these factors disproves the existence of sex discrimination; but they do mean that attempts to measure sex discrimination must be unusually careful in specifying the relevant variables which must be equal before remaining inequalities can be considered "discrimination." Unfortunately, such care is not evident in HEW pronouncements or in much of the literature supporting affirmative action. Even the comprehensive studies by Helen S., Astin and Alan E. Bayer make the fatal mistake of holding marital status constant in comparing male-female career differences.30 But marriage has opposite effects on the careers of male and female academics, advancing the man professionally and retarding the woman's progress. Not only do the men and women themselves say so,31 but the Astin-Bayer data (and other data) also show  it.32 Therefore to treat as "discrimination" all residual differences for men and women of the "same" characteristics-including marriage-is completely invalid and misleading.
          Marriage is a dominant-and negative-influence on academic women's careers. A study of academics who had received their Ph.D.s many years earlier showed that 69 percent of the total-mostly men-had achieved the rank of full professor, as had 76 percent of the single women but only 56 percent of the married women.33 In short, many of the statistical differences between the broad categories "men" and "women" are to a large extent simply differences between married women and all other persons. It is an open question how much of the residual disadvantages of single academic women is based upon employer fears of their becoming married academic women and acquiring the problems of that status. One indication of the difficulty of successfully combining academic careers with the demands of being a wife and mother is that academic women are married much less frequently than either academic men or women Ph.D.s in nonacademic fields34 are divorced more frequently,35 and have fewer children than other female Ph.D.s.36
          Much of the literature on women in the labor market denies that "all" women, "most" women, or the "typical" woman represent special problems of attrition, absenteeism and other characteristics reflecting the special demands of home on women. For example, the "typical woman economist" has not given up her job to move because of her husband's move, but 30 percent of the women economists do, while only 5 percent of male economists accommodate their wives in this way." Similarly, while most female Ph.D.s in economics have not interrupted their careers, 24 percent had interrupted their careers prior to receiving the degree (compared to 2 percent of the men) and "another 20 percent" afterwards (compared to I percent of the men).38 These are clearly substantial percentages of women and several-fold differentials between men and women.
          The literature on women workers in general makes much of the fact that most women "work to support themselves or others," not just for incidental money.39 However, this does not alter the facts (1) that women's labor force participation rates are substantially lower than men's40 and (2) that married women's labor force participation declines as their husbands' incomes rise.41 This is also true of academic women.42
          In considering global male-female differences in career results, the question is not whether "most" women have certain negative career characteristics but whether a significant percentage do and whether that percentage is substantially different from that of men. Moreover, it is not merely the individual negative characteristics that matter but their cumulative effects on male-female differentials in employment, pay, and promotion. Nor can these differences in career characteristics be dismissed as subjective employer perceptions or aversions.43 They represent in many cases choices made outside the work place which negatively affect women's career prospects. As one woman researcher in this area has observed: "One way of insuring that the academic husband's status will be higher than his academic wife's is to allow the husband's job opportunities to determine where the family lives."44   But regardless of the wisdom or justice of such a situation, it is not employer discrimination, even though it may lead to statistical male-female differences between persons of equal ability.
          One of the fertile sources of confusion in this area is the thoughtless extension of the "minority" paradigm to women. It makes sense to compare blacks and whites of the same educational levels because education has the same positive effect on black incomes and white incomes, though not necessarily to the same extent. It does not make sense to compare men and women of the same marital status because marital status has opposite effects on the careers of men and women. Minorities have 'serious problems of cultural disadvantages, so that faculty members from such groups tend to have lower socioeconomic status and lower mental test scores than their white counterparts,45  and black colleges and universities have never been comparable to the best white colleges and universities,"' whereas female academics come from higher socioeconomic levels than male academics," female Ph.D.s have higher IQs than male Ph.D.s in field after field,"' and the best women's colleges have had status and student SAT levels comparable to those of the best male or coeducational institutions. Women have been part of the cultural, informational, and social network for generations, while blacks and even Jews have been largely excluded until the past generation. While minorities have been slowly rising in professional, technical, and other high-level positions over the past 100 years, women have declined in many such areas over the same period, even in colleges institutionally operated by women,' so that employer discrimination can hardly explain either the trend or the current level of "utilization" of women. Marriage and childbearing trends over time are highly correlated with trends of women's participation in high-level occupations, as well as being correlated with intra-group differences among women at a given time. In short, women are not another "minority," either statistically or culturally.
          When male-female comparisons are broken down by marital status and other variables reflecting women's domestic responsibilities, some remarkable results appear. Although women in the economy as a whole earn less than half as much annually as men,50 with this ratio declining from 1949 to 1969,51 the sex differentials narrow to the vanishing point-and in some cases are even reversed-when successive corrections are made for marital status, full-time as against part-time employment, and continuous years of work. For example, in 1971 women's median annual earnings were only 40 percent of those of men, but when the comparison was restricted to year-around, full-time workers, the figure rose to 60 percent, and when the comparison was between single women and single men in the same age brackets (thirty to forty-four) with continuous work experience, "single women who had worked every year since leaving school earned slightly more than single men."52 These are government data for the economy as a whole.
          The severe negative effect of marriage on the careers of women is not a peculiarity of the academic world. Other nation- wide data on sex differences show single women's incomes ranging from 93 percent of single men's income at ages twenty-five to thirty-four to 106 percent in ages fifty-five to sixty-four53-that is, after the danger of marriage and children are substantially past. For women already married, the percentages are both lower and decline with age-ranging from 55 percent of married men's incomes at ages twenty-five to thirty-seven to only 34 percent at ages fifty-five to sixty-four.54 Apparently early damage to a woman's career is not completely recouped-at least not relative to men who have been moving up occupationally as they age while their wives' careers were interrupted by domestic responsibilities. In the early years of career development, single women's labor force participation rates are rising sharply, while those of married women are declining sharply." Again, the data suggest that what are called "sex differences" are largely differences between married women and all others, and that the origin of these differences is in the division of responsibilities in the family rather than employer discrimination in the work place. The increasing proportion of married women in the work force over time " has been a major factor in the decline of the earnings of women relative to men.
          Academic women show similar patterns. For example, the institutional employment of married women is "determined to a large extent" by the location of their husbands' jobs,57 and this contributes to a lower institutional level for academic females than for male Ph.D.s. Academic women apparently find it harder than other women of similar education to combine marriage and a career. One study of "biological scientists receiving their degrees during the same time period" found only 32 percent of such academic women married compared to 50 percent of their non-academic counterparts, even though initially virtually identical percentages were married before receiving their Ph.D.58 A more general survey of women holding doctorates found only 45 percent to be married and living with their husbands.59 Although there were more married than single women among women doctorates in general,60 in the academic world there were more single than married female doctorates.61 Moreover, female academics had divorce rates several times higher than male academics.62 Another study of college teachers found 83 percent of the men but only 46 percent of the women to be married.63 Women in other high- level, high-pressure jobs requiring continuous full-time work show similarly low proportions married.64
          Childbearing is also negatively associated with career prospects. Among Radcliffe Ph.D.s, those working full time had the fewest children, those working part time next, those working intermittently next, and those not working at all had the most children.65 Various surveys show that "female Ph.D.s who are married are twice as likely to be childless as women in the same age group in the general population" and even when they do have children, to have fewer of them.66 The husband's prospects also have a negative effect on women doctorates' careers: a woman married to a "highly educated man with a substantial income was less likely to work" or, if she did, was more likely to take a part-time job.67 This parallels a negative correlation between married women's labor force participation and their husbands' incomes in the general economy.68
          In research output, "the woman doctorate who is married and has children was less likely than the single woman doctorate or a childless married woman doctorate to have many scientific and scholarly articles to her credit."69 I it is not surprising that the married woman doctorate "tended to make a lower salary than the single woman, even if she was working full time."70  Unfortunately, studies of academic women have not simultaneously controlled for marital status, full-time continuous employment, publications, and degree level and quality.
          The National Academy of Sciences data permit comparisons of the salaries of male and female doctorates who worked full time both in 1960 and in 1973 and who responded to all the biennial surveys of the National Science Foundation from 1960 through 1973 (see Table 6). This gives an approximation of full-time continuous employment, but does not show whether the respondent was employed full time in each of the years during which a survey was made or whether the respondent worked at all in the non-survey years. These data show female salaries at 83 percent of male salaries in 1970 (before affirmative action) and 84 percent in 1973 (after affirmative action)-a smaller proportion than in other data which controlled for other variables such as publications and degree quality. It also indicates no discernible effect of affirmative action programs.
A 1968 study of full-time academic doctorates found women's salaries ranging from 89 percent to 99 percent of men's salaries in the same field, with similar length of employment, and in broadly similar institutions (colleges versus universities).71 These higher percentages-as compared with the results in Table 6-suggest that the distribution of women by institutional type and ranking and by years of employment explains a significant part of the male-female salary differences among academics. Moreover, since women academics with Ph.D.s in this 1968 study earned 92.2 percent of the income of men academics with Ph.D.s (even without controlling for publications), these figures indicate how small the sex differential was for even roughly similar individuals before affirmative action.
          Even more revealing patterns appear in our tabulations of ACE data by marital status (Table 7). In 1969, academic women who never married earned slightly more than academic men who never married. This was true at top-rated institutions and at other institutions, for academics with publications and for academics without publications. The male salary advantage exists solely among married academics and among those who used to be married ("other" includes widowed, divorced, et cetera). The male advantage is greatest among those married and with dependent children. Being married with children is obviously the greatest inhibitor of a woman's career prospects and the greatest incentive to a man's. The salaries of women who never married were 104 percent of the salaries of their male counterparts at the top-rated institutions and 101 percent at other institutions. For women who were married but had no dependent children, the percentages fell to 88 and 84 percent, respectively. For married women with dependent children, the percentages fell still further, to 69 and 70 percent. For women and men without publications and in nonranked-essentially nonresearch-institutions, the "never married" women earned 145 percent of the "never married" men's incomes-confirming a general impression that women prefer teaching institutions, and therefore a higher proportion of top-quality women than of top-quality men end up at such places by choice. It also suggests that employers are not unwilling to recognize such quality differentials with salary differentials in favor of women.
          In the literature on sex differentials and in the pronouncements of governmental agencies administering affirmative action programs, sinister and even conspiratorial theories have been advanced to explain very ordinary and readily understandable social phenomena: (1) academic individuals who are neither aiding nor aided by a spouse make very similar incomes, whether they are male or female, (2) academic individuals who are aided by a spouse (married males) make more than unaided individuals, and (3) academic individuals who aid a spouse (married females) make less for themselves than do the other categories of people. The social mores which lead women to sacrifice their careers for their husbands' careers may be questioned (as should the high personal price exacted from academic career women, as reflected in their lower marriage and higher divorce rates). But social mores are not the same as employer discrimination. The fact that single academic women earn slightly higher salaries than single academic men suggests that employer discrimination by sex is not responsible for male-female income differences among academics. Moreover, even as regards social mores, it must be noted that academic women report themselves satisfied with their lives a higher percentage of the time than do academic men72-a phenomenon which some explain by saying that women do not put all their emotional eggs in one basket as often as men,73  and which others explain by treating high research creativity as a somewhat pathological and compensatory activity of the personally unfulfilled.74 The point here is that the evidence is not all one way, nor the logic overwhelming, even as regards apparently inequitable social mores. On the basic policy issue of employer discrimination, such evidence as there is lends no support to this as an explanation of male-female career differences, and the slight but persistent advantage of single females over single males undermines the pervasive preconception that employers favor men when other things are equal.

  1 National Academy of Sciences, Doctoral Scientists and Engineers in the United States: 1973 Profile (Washington, D. C.: National Academy of Sciences, 1974), p. 30.
2 Lester, Antibias Regulation, p. 50.
3 Mommsen, "Black Doctorates," pp. 103-104.
4 Mommsen, "Black Ph.D.s," p. 258.
5 David G. Brown, The Market for College Teachers (Chapel Hill: University of North Carolina Press, 1965), p. 83.
6 Lester, Antibias Regulation, p. 49.
7 Thompson, Private Black Colleges at the Crossroads, p. 155.
8 Mommsen, "Black Ph.D.s," pp. 258-259.
9 Ibid., p. 262.
10 Caplow and McGee, Academic Marketplace, p. 41.
11 Brown, The Mobile Professors, p. 26.
12 Lester, Antibias Regulation, p. 48.
13 Mommsen, "Black Ph.D.s," pp. 262, 259.
14 Finis Welch, "Black-White Differences in Returns to Schooling," American Economic Review, vol. 63, no. 5 (December 1973), pp. 893-907.
15 Ibid., p. 894.
16 Bernard, Academic Women, p. 52.
17 Lester, Antibias Regulation, p. 42.
18 Alan E. Bayer, Teaching Faculty in Academe: 1972-73 (Washington, D. C.: American Council on Education, 1973), p. 15.
19 Frank Clemente, "Early Career Determinants of Research Productivity," American Journal of Sociology, vol. 79, no. 2 (September 1973), p. 414; Lester, Antibias Regulation, p. 47. See also Brown, Mobile Professors, p. 78, and Bernard, Academic Women, p. 148.
20 Brown, Mobile Professors, pp. 76-78. See also Lester, Antibias Regulation, p. 42.
21 Bernard, Academic Women, p. 180; Brown, Mobile Professors, p. 81; Helen S. Astin and Alan E. Bayer, "Sex Discrimination in Academe," Educational Record, Spring 1972, p. 103; Helen S. Astin, The Woman Doctorate in America (Now York: Russell Sage Foundation, 1970), pp. 20-21.
22 Bernard, Academic Women, pp. 151-152.
23 Astin, The Woman Doctorate, p. 73; Lester, Antibias Regulation, p. 42.
24 Brown, Mobile Professors, pp. 79-80.
25 Barbara B. Reagan, "Two Supply Curves for Economists? Implications of Mobility and Career Attachment of Women," American Economic Review, vol. 65, no. 2 (May 1975), pp. 102,103. See also Astin, The Woman Doctorate, P. 102.
26 Reagan, "Two Supply Curves," p. 104.
27 Brown, Mobile Professors, p. 78.
28 Bernard, Academic Women, p. 221; Lester, Antibias Regulation, p. 39.
29 Reagan, "Two Supply Curves," p. 103. See also Bernard, Academic Women, pp. 151-152,181-182; Astin, The Woman Doctorate, pp. 91-92.
30 "The regression weights of the predictor variables that emerged in the analysis of the men's sample were applied to the data for the women's sample to assess the predicted outcome when the criteria for men were used.... To award women the same salary as men of similar rank, background, achievements and work settings ... would require a compensatory average raise of more than $1,000. . . ." Astin and Bayer, "Sex Discrimination in Academe," P. 115.
31 Bernard, Academic Women, p. 217. 
32 Astin and Bayer, "Sex Discrimination in Academe," p. 111; Lester, Antibias Regulation, pp. 36-37.
33 Helen S. Astin, "Career Profiles of Women Doctorates," Academic Women on the Move, ed. Alice S. Rossi and Ann Calderwood (New York: Russell Sage Foundation, 1973), p. 153.
34 Lester, Antibias Regulation, p. 41.
35 Bernard, Academic Women, pp. 113, 206.
36 Ibid., p. 216.
37 Reagan, "Two Supply Curves," pp. 101-103.
38 Ibid., p. 104.
39 U.S. Department of Labor, Underutilization of Women Workers (Washington, D. C.: U.S. Government Printing Office, n.d.), p. 1.
40 William G. Bowen and T. Aldrich Finegan, The Economics of Labor Force Participation (Princeton: Princeton University Press, 1969), pp. 41, 101, 243. 
41 Ibid., p. 132.
42 Astin, The Woman Doctorate, p. 60.
43 Jerolyn R. Lyle and Jane L. Ross, Women in Industry (Lexington, Mass.: D. C. Heath and Co., 1973), p. 13; Reagan, "Two Supply Curves," p. 104.
44 Quoted in Reagan, "Two Supply Curves," p. 102.
45 Horace Mann Bond, A Study of the Factors Involved in the Identification and Encouragement of Unusual Academic Talent among Underprivileged Populations (U.S. Department of Health, Education, and Welfare, Project no. 5-0859, Contract no. SAE 8028, January 1967), p. 117.
46 Sowell, Black Education, pp. 255-259; Jencks and Riesman, "The American Negro College."
47 Bernard, Academic Women, pp.,xx, 77-78; Alan E. Bayer, College and University Faculty: A Statistical Description (Washington, D. C.: American Council on Education, June 1970), p. 12; Astin, The Woman Doctorate, pp. 23, 25. 
48 Bernard, Academic Women, p. 84.
49 Ibid., pp. 39-44.
50 James Gwartney and Richard Stroup, "Measurement of Employment Discrimination According to Sex," Southern Economic journal, vol. 39, no. 4 (April 1973), pp. 575-576; and "The Economic Role of Women," in the Economic Report of the President, 1973 (Washington, D. C.: U.S. Government Printing Office, 1973), p. 103.
51 Gwartney and Stroup, "Employment Discrimination," p. 583.
52 "The Economic Role of Women," p. 105.
53 Gwartney and Stroup, "Employment Discrimination," p. 582.
54 Ibid.
55 Kreps, Sex in the Marketplace, p. 32.
56 Ibid., pp. 4, 19.
57 Bernard, Academic Women, p. 88.
58 Ibid., p. 113.
59 Astin, The Woman Doctorate, p. 27.
60 Ibid.
61 Ibid., p. 71.
62 Bernard, Academic Women, p. 216.
63 Ibid., p. 313.
64 Ibid., pp. 313-314.
65 Ibid., p. 241.
66 Lester, Antibias Regulation, p. 38.
67 Astin, The Woman Doctorate, p. 60.
68 Bowen and Finegan, Labor Force Participation, p. 132. 
69 Astin, The Woman Doctorate, p. 82.
70 Ibid., p. 90.
71 Bayer and Astin, "Sex Differences in Academic Rank and Salary among Science Doctorates in Teaching," Journal of Human Resources, vol. 3, no. 2 (Spring 1968), p. 196. "Science" in the title includes the social sciences.
72 Bernard, Academic Women, p. 182.
73 Ibid., p. 152.
74 Ibid., p. 156.