Early in 2000, following the publication in 1999 by MIT of a landmark self-study that had found definitive evidence of bias against its senior women faculty, MIT convened a meeting of presidents and provosts of nine leading universities, including Harvard, to discuss these issues and plan further steps. At the time, the group agreed to meet again a year later (i.e. in early spring 2001) to compare results, but no meeting actually took place until 2002. The Committee for the Equality of Women at Harvard (CEWH) had strongly urged President Rudenstine to undertake and publish a self-study similar to MIT's but he had replied that Harvard was "different" and did not need such a study. As of this date, Harvard has not published such a report, nor have any findings of the last (spring 2002) meeting been announced. CEWH continues to believe that such a self-study is badly needed at Harvard. The urgency of such an effort is greatly enhanced by the recent publicity concerning the diminishing rate of appointing senior women in the last few years and the resulting disclosure of discontent among the University's senior women faculty.

Before outlining the details of our concerns we wish to state in the strongest possible terms the need for candor and transparency in reporting the outcome of such a study. Harvard's continuing secrecy concerning many issues of women faculty serves only to arouse mistrust and the suspicion that there is much to hide. It is known that at least two studies were conducted in the last several years but their results have been inaccessible. We are well aware that some aspects of the issues involved need to be confidential, but we emphasize that comparable institutions seem to feel less need for secrecy. Duke University has published a very comprehensive self-study report and has taken action to implement its recommendations. Both Princeton and Yale have published similar but less far-reaching reports.

Harvard's secrecy is all the more puzzling in an institution established for the discovery and dissemination of knowledge. Among other problems generated by secrecy is the failure to make use of the potential input from the larger university community and to generate the sense of common problems and shared ownership that a broad spectrum of participants could contribute.

The only data freely available on gender issues are the annual figures, comprising numbers and proportions of women faculty (among much other information) at ladder and senior levels, which Harvard reports in its Affirmative Action (AA) Plan, as required by law.The report also contains a few summary figures on students, breaking down degree candidates, but not enrollments, by gender as well as race. The most serious problem is that faculty data are reported only by academic division (humanities, social sciences, and natural sciences) and thus do not permit analyses by department. Departmental figures are, however, essential to an understanding of past successes and failures in recruitment and retention because aggregate numbers for whole divisions may, and probably do, conceal large and persistent imbalances in individual units. Such imbalances may or may not be evidence of bias. It is obvious from faculty rosters that some departments are seriously at variance with the expected distributions.

The numbers we use are taken from Harvard's published AA reports except where otherwise noted.The following issues within FAS are of most concern to CEWH:

FACULTY

SENIOR FACULTY

The proportion of women full professors increased rapidly during the late 1990s, after a very slow start during the previous two decades, and we acknowledge and appreciate the successful efforts Harvard made during this period.The last few years, since 2001, however, have seen a dramatic annual decline in tenure offers made to women, and it is this decline that some twenty-six senior faculty women are currently protesting. Although there seems to be some dispute about the precise figures (perhaps Harvard should issue an official count), recent press reports indicate that tenure offers to women amounted to 37 percent of all such offers in 2000-2001 but have declined each year since then to a low of 11.1 percent in 2003-2004. It is obvious that if this trend continues, the gains of the 90s will erode rapidly.

Much room for improvement remains in any case. Harvard ranks just above the middle but well below the top with respect to proportions of women faculty among the group of ten comparable universities. The numbers of tenured women are particularly low in the natural sciences, where they have increased from a total of 8 in the entire division in 1988 to only 14 in 2004, out of a total of 157 positions in 1988 and 161 in 2004. Such proportions are woefully small compared to the supply of women doctorates, whose proportion of science degrees has been more than one-fifth in the physical sciences and well over forty percent in biological sciences for a number of years. In the next section we explore what might be reasonable expectations for tenure appointments for women based on their presence in the labor pool and their qualifications.

AVAILABILITY AND UTILIZATION

By federal regulation of AA, representation of women, known as "utilization", is measured against their "availability", defined (for faculty) as the proportion of an "appropriate pool" of candidates with the necessary degrees, experience, and distinction to fill a particular position. There is obviously much room for interpretation of such a standard; a commonly accepted proxy for it is the proportion of Ph.D.s earned by women in a particular field at a relevant time. The reason this works is that extensive research over the last three decades, much of it by the National Academy of Sciences, has demonstrated that the distribution of women candidates with respect to "quality" resembles that of men. As always, the devil is in the details: it is obvious that the definition of a particular position can easily be tailored to eliminate a priori various candidates who may be unacceptable to a department for reasons unrelated to their competence. We say this not to imply that such discrimination is necessarily occurring but to point out one of the major pitfalls between the articulation of an effective AA policy and its execution.

The "relevant time" mentioned above refers to the number of years it takes for a candidate to attain the desired qualifications. While not immutable, in general a scholar needs about ten to fifteen years to become qualified for a senior appointment but that period may be extended when a "star" is being recruited. For ladder faculty, the normal period is anywhere from zero to about five years, depending on the need for postdoctoral training, which varies widely across fields. In short, most candidates for senior rank in 2000, say, would have earned their doctorates around 1985 to 1990, and those for ladder positions about 1995 to 2000.

A comparison of Harvard's published availability and utilization data over time, as shown in Table 1, demonstrates the importance of looking critically at such figures. The availability of women candidates for senior positions is generally listed as less than half of what the figures collected annually by the National Academy of Sciences actually show. Further, Harvard's availability numbers have remained unchanged each year since at least 1996 for each of the three major divisions within FAS, despite the considerable actual growth of the candidate pool. Such an outcome is, of course, extremely unlikely, and suggests that no one overseeing affirmative action at the University has lately taken a serious look at the reporting procedures, at the very least. The situation casts serious doubt on other aspects of the process: how carefully are departments' claims regarding their recruitment practices actually being examined?

SALARIES

Salary equity is a standard test for equality of treatment, working condiditons, and hence opportunity. In the group of ten comparable institutions, Harvard ranks close to the bottom on this measure, with senior women being paid only 92 percent of men's salaries, compared to, e.g. Princeton and Stanford at 97 percent. (These figures come from the data published annually by the American Association of University Professors. Harvard does not include them in their AA reports.) Some salary inequality can be explained based on an individual's length of service. More recently hired people (e.g. women) will make less than longer-term employees. In public universities, salary information is published, with no deleterious effects, but private institutions do not usually make it available. But it would be possible in many cases to preserve confidentiality by publishing at least average salaries by department, disaggregated by sex and length of service.

Note that we do not argue for any kind of quota or fixed fraction of appointments for women. We do maintain, however, that the proportion of women on the faculty needs to bear some reasonable relation to the numbers of women who have the necessary degrees, experience, and eminence.

This last point raises the issue of what Harvard should be doing to produce its share of fully qualified senior women, an activity that the regulations require under the heading of "good faith efforts", especially at institutions whose proportion of women is below expectations. Part of such an effort must be the development of a cadre of ladder faculty, either for promotion at Harvard (now that the old policy of not promoting from within is being relaxed) or for appointments elsewhere. Some of these concerns are addressed in the next section.

The recent publicity concerning the declines in tenure offers and hires of women, extensively covered in the daily press, in Science , and in The Chronicle of Higher Education , yields somewhat confusing figures, which vary from one account to another. Without trying to sort these out, in Table 2 we give the numbers supplied to CEWH by Dean William Kirby in a recent letter.

Ladder Faculty

Harvard concedes that for ladder faculty "utilization continues to lag behind availability... in the Humanities and in the Natural Sciences". (2004 AA Report, p.38) In the natural sciences, the reported utilization slipped from 20.3 percent in 1997 to 9.4 percent in 2001 and has since then crept back to 17.5 percent. Given that much hiring in this division has been in the life sciences, where women earned 45 percent of Ph.D.s in 1998 and more since then, the inference has to be that no very great effort was made to hire women. The availability of women for natural sciences is given as 27.5 percent, almost certainly an undercount since even in physical sciences women are earning about a quarter of doctorates. Even the smallest field for women, engineering, awards 16 percent of doctorates to women so that Harvard's figures seem distinctly open to question.

In the last several years our concern that Harvard's number of women at ladder ranks is inadequate has increased greatly. Between 1998 and 2004 the total number of women rose by only 4, from 58 to 62, although the total faculty at this rank rose from 172 to 195. Women's share of ladder appointments thus grew by less than 2 percent in six years, at a period when women constituted about 43 percent of the total new doctorate pool, and over 49 percent of US citizens in that group. Oddly enough, women in the humanities - an area in which women have earned about half of all doctorates since 1990 and fully one-third since 1920 - number a mere 23 in 2004, up from 20 in 1988, out of a total of 64 positions. They are slightly better represented in the social science fields, where they hold 33 out of 74 positions, up from 29 in 1988, but much less so in natural sciences, holding 11 of 63 ladder positions, an increase of only 3 in sixteen years.

These figures must be viewed in the context of an overall declining number of ladder positions since 1988, from 220 to 195 in 2004. This decline is steepest in social sciences and next largest in humanities; natural science totals have remained unchanged. There are several possible reasons for such changes and of course we do not know what reasons actually exist in this case because there has been no offical comment on the matter. Since all the decreases are in the numbers of male faculty, it is certainly possible that more men than women have been promoted to tenure. That is the kind of information that a self-study should reveal.

Salary differences also exist in this group, with Harvard paying women 93 percent of men's salaries in 2002; this is a decrease since 2000, when the figure was 97 percent, putting Harvard at the top of its group of peer institutions then but barely in the middle of the pack in 2002. Salaries at this rank should be identical for men and women, since both are at the beginning of their careers; however, differences by field are common. It is likely that if Harvard had an adequate number of women in natural sciences, where salaries are often higher, this difference would be much reduced or eliminated.

Working Conditions

To our knowledge, working conditions for women faculty have not been examined as they were in the MIT report, which revealed considerable inequities in space allocation, collegial integration, recognition, and rewards. But these factors are vitally important in promoting a climate in which all members of a group can be productive scholars. Anecdotal accounts from some women faculty members suggest that Harvard wins no prizes in this area.

STDENTS

This year Harvard admitted its first class ever with equal numbers of men and women - more than two decades after women became the majority of college graduates nationwide; they are currently about 57 percent of college graduates, but only 47 percent at Harvard. The AA report contains very little information about either undergraduate or graduate students beyond overall numbers. Data we obtained in the early 1990s showed that in some fields the participation of women as graduate students was well below national norms. Is that still the case? The National Science Foundation at that time found that while Harvard ranked ninth nationally as a baccalaureate source of men who earned doctorates in natural sciences, it ranked only seventeenth for women - a discrepancy that may arise from from differential admissions or differential treatment. Have these data changed, or been investigated? In 2002 a group of undergraduates published the Women's Guide to Harvard , which includes data on women concentrators by field. They report, for example, that in the two years from 1998-1999 and 2000-2001 women in mathematics dropped from 22 to 14 percent of concentrators; that is a very long way from national figures, where women are over 47 percent of baccalaureate earners and which has been among the most sex-neutral majors for decades, popular wisdom to the contrary not withstanding.

Such differences raise many questions. In about 1969-1972, when most formerly all-male institutions began admitting women, they did so primarily to balance the books; it was assumed that women would enroll heavily in humanities courses where universities were overstaffed with tenured faculty, compensating for the men who were flocking into occupation-related fields. It is entirely possible that Harvard's admissions were skewed in favor of women expressing intentions to pursue humanities concentrations. Such practices, even after they end, cast long shadows. Faculty, advisers, and students themselves tend to follow precedents. Has Harvard biased its admissions against women in mathematics?

How much do such differences matter to students' educational experience? Much evidence shows that they do. Detailed studies find that women faculty are more aware of and responsive to the needs of women students than are men. The ratio of male faculty to male students is about 1:4, but for women faculty and students it is 1:11. Does this yield an equal education for women, or are they getting less than they pay for?

CONCLSION

In the absence of enough information to really analyze such numbers as the ones we have presented here, we can only say that they raise doubts about Harvard's "good faith" efforts to integrate women fully into the university community. The figures given year after year for availability are patently inaccurate and misleading and the University needs to offer an explanation for the fact that they remain unchanged over a decade of phenomenal growth in the pools of women candidates. The Harvard community is entitled to know in which departments there are problems and what specific efforts are made to remedy them. Reporting for entire divisions is inadequate and possibly misleading. In the meantime, the administration should reflect on the fact that an annual increase of 1 to 2 percent in the proportion of women faculty is a rate they would find unacceptable for endowment growth.

Similar considerations apply to students, with the added emphasis that the University must insure that women students are indeed getting the same excellent education and guidance as men.

Members of the administration have recently argued that the failure to hire more women rests with individual departments at Harvard. We note that this is almost universally true and Harvard is no exception here. But the significance of the numbers we have cited lies precisely in the fact that they reflect attitudes and conditions which the administration has tolerated. Dozens of careful studies have confirmed that administrative leadership, however exercised, is an essential component of successful diversity policies action, and Harvard should provide no less.

Involving the wider University community in a thorough self-study will help to provide not only the necessary broad coverage but will serve to engage more of its members in a productive fashion, enhancing their sense of ownership and participation and very likely their willingness to contribute funds as well as effort. Above all, this must be a candid and public undertaking, and we reiterate our urgent request that it be undertaken with all deliberate speed.

Lilli S. Hornig, Ph.D. Harvard '50
The Committee for the Equality of Women at Harvard.

URL: http://world.std.com/~cewh/
November 2004