Issues in the Faculty Pay Equity Exercise

Kathy Bischoping/Arts-Sociology

September 1997

The Joint YUFA-Administration Committee on Pay Equity has been negotiating a methodology for distributing gender pay equity moneys for faculty and professional librarians since the spring of 1992. The settlement for professional librarians was reached in December 1995. Although the distribution of the $650 000 pay equity money for women faculty members remains outstanding, the committee has been making steady progress toward a settlement in 1996-97. In this article, I provide an overview of key points in the pay equity methodology negotiated by the committee in the past year and outline some of the methodological issues that have arisen. Please note that this overview is a report of neither the Joint Committee nor the YUFA caucus, though it has been reviewed by other members of the YUFA caucus.

1. Identifying Anomalously Low Salaries

In order to identify women faculty whose salaries are anomalously low in comparison to men's, the committee negotiated throughout the fall about which variables to use as the basis for comparison. For example, is a woman's salary low if she earns less than men with her university teaching experience, less than men her age, less than men of her rank, less than men in her unit...or any and all of the above? The committee's response to this question was not an ideal one, so much as a feasible compromise between diverging definitions of the principle of pay equity.

The initial position of the Administration had been that years since the first full-time university teaching appointment, a variable included in the faculty data base, should be the primary basis for comparison because it is related to experience. This approach does not take into account that women's career paths before they enter academe differ from those of men, or that women tend to complete graduate school later than men. YUFA's initial position was that age should be the basis of comparison between women's and men's salaries, as it is a variable unrelated to gender that does not discriminate against women who made "late starts" in their careers as full-time academics. The compromise reached by the committee was that each of these two variables should be used in separate regression models to determine which women's salaries were anomalously low.

For the third model, the committee agreed that it would be desirable to obtain an improved measure of relevant professional experience, beyond that of full-time university teaching, and to identify women whose salaries were anomalously low when this criterion was used. After lengthy discussions about how to define the attractive but elusive concept of "experience", it was decided that "experienced" begins at the time of entering graduate school. Moreover, paid experience that women acquired in the gap between completing their undergraduate degrees and entering graduate school was also to be counted, provided this experience was relevant to the current position. Because the cvs and faculty database information available to the committee did not contain all the information that was needed, a survey questionnaire was designed by the committee in October. (More on that in section two.)

In each of the three models, a variable indicating whether faculty members are in the Alternate Stream is to be used as a control. Members of this stream are typically men in the Sciences and women in Languages. While many members of the Alternate Stream are very active researchers, they are hired to focus their efforts primarily on teaching and development of curricular materials, and are evaluated for tenure and promotion accordingly. Historically, members of the Alternate Stream have been less likely than other faculty members to complete the usual three postsecondary degrees. By including both this variable and the experience measure, the settlement conveys an ambivalence about credentials and the worth of different kinds of academic work.

2. Issues in Measuring and Crediting Experience

A number of women had not completed the typical set of three post-secondary degrees -- for example, novelist Susan Swan (Humanities), who has been published in ten countries, completed her formal education with a B.A. Swan notes that in Europe, an honorary Ph.D. is typically awarded when universities hire an author whose work is clearly equivalent to the usual Ph.D. The committee's admittedly more prosaic response to this scenario was to decide that an identical period of experience should be credited identically, regardless of the degrees that had been acquired. Different versions of the questionnaire were designed to reflect faculty members' diverse educational backgrounds, before the survey was conducted by ISR in December and January.

Once the survey and strike were completed, the committee worked to resolve a number of issues that arose when the faculty data base and survey information were examined. The faculty data base was found to contain some errors. For example, the Italian "dottore" degree that is equivalent to an Honours B.A. was occasionally treated in the data base as a Ph.D., so its holders were recorded as not having an undergraduate degree. In addition, it became apparent that our questionnaire did not sufficiently emphasize that we were measuring work only in the gap between the Bachelor's degree and the onset of graduate school, meaning that several respondents had to be recontacted. The committee also recontacted women for whom we had queries about the relevance of certain work experiences to the current academic position.

Several respondents commented that the goal of quantifying relevant work in terms of its duration and intensity could be a complex or meaningless task. This seemed to apply particularly for women in the Faculty of Fine Arts whose creative work encompassed several different areas. Thus, Holly Small in the Department of Dance wrote to the committee, "It is impossible for me to think of those years in terms of % of time spent on various activities. Everything was inter-related, one thing feeding into another...I'm not talking about a 9-5 job here. It's a vocation, a life of obsession that never seems to mesh with any of the questionnaires I'm asked to fill out at York." This serious concern is one that is difficult to address in a process that is, by necessity, quantitatively oriented.

Those involved in future salary adjustment processes that require quantification of life experiences may find it helpful to know that our rate of recontacts (in what sometimes turned into in-depth interviews) and/or use of supplemental c.v. information was approximately 20%, so that a survey is best seen as a first phase of data collection rather than a final one. Sending data sheets to respondents so that they can be verified, as the committee did in August, is also recommended.

3. Variables Omitted from the Model

The salary predictors agreed on by the joint committee in fall 1996 are few -- age, years of full-time university teaching, relevant professional experience, and the variable to control for membership in the alternate stream. Of course, several other variables could have been used, as in the "Denzel model" which employs rank, teaching subject area, highest degree obtained, and the speed at which the present rank was obtained. YUFA's concern about many of these variables is that they are themselves gender-related. Thus, statistically "controlling" for these variables would amount to agreeing that existing gender differences in these variables are themselves equitable.

Two examples of this issue may be helpful. The committee discussed whether faculty and department salary profiles should be taken into account so that the exercise would seek to raise women's salaries to the level of men's within their unit. We chose instead to use men's salary profile throughout the university as the baseline. Our rationale was that units with greater proportions of women, such as the Dance and Social Work Departments and the Faculty of Education tended by some standards to be low paid on the whole. Therefore, raising the salaries of underpaid women in these units toward their own units' levels would not address the anomalously low levels of pay of these units. We recognize, however, that this decision reinforces a less frequent type of anomalous salary. Women in highly paid units, such as the Schulich School, would be unlikely to receive any pay equity adjustment because their salaries, low in comparison to their male peers', are not unusual in comparison to men in the university as a whole.

Similarly, the committee discussed whether to control statistically for the effects of having served as an administrator at the level of Dean or Vice-President on salaries. (Yes, Virginia, administrative stipends are sometimes added to salaries even after a term as an administrator has been served.) The YUFA position was to argue against this, on two grounds. First, because administrators are disproportionately male, controlling for this factor would reduce the estimate of the salary gap that would be attributed to gender. Second, controlling for this factor would tacitly acknowledge that administrative stipends were rightly added to salaries after administrators had served their terms -- a perspective we were reluctant to endorse. Once again, the limitation of omitting the variable can be identified: women who had served as administrators and who had not benefitted from it to the extent of men, would not be identified by the model.

4. Allocating Funds

As this article is going to press, women faculty members are making the final checks of their survey data and bringing forward new questions about the methodology, and the committee is addressing numerous issues in the methods for allocating funds. The deadline for agreement has been set to October 31, 1997. In this section, I sketch out a few of the more clearcut issues in ongoing negotiations in some detail, setting aside those issues that are newly emerging.

One group of decisions to be made concerns the statistical methods to be used to determine payments. For example, since a survey of men's relevant professional experience was not conducted, it will need to be imputed. The women's survey responses and the faculty data base together indicate that the probability of following a traditional path from undergraduate degree to graduate degree to full time university work vary by gender, age (and presumably economic period). An imputation method to take these factors into account is being developed by Georges Monette, statistical consultant to the Joint Committee, aka one of the "guys with vans" during the strike. An additional statistical decision is how to carry out the bottom loading of allocations that the settlement requires -- i.e., the settlement says basically that the further "off" a woman is from a male pay line, the higher a proportion of her distance below the line is to be compensated.

Another group of issues pertains to defining who is eligible to receive payments and at what rate these payments should made. For example, are full retroactive payments to go to women who, during the retroactive period, were sabbaticants, on reduced load, or left the bargaining unit to become administrators? YUFA's position is that 100% retroactive pay should apply for the time that a person was in the YUFA bargaining unit during the retroactive period. The situation of women receiving long-term disability benefits is also of particular concern to the committee, as increments to their base salaries (via pay equity or PTRs) cannot come into effect until they return to work. Last, but not least, there's the $2500 question. The settlement stipulates that women are to be eligible for compensation if their salaries are $2500 lower than one of the three male pay lines. This decision takes into account the committee's concern that the limited funds available for gender equity payment be used to address the largest anomalies that exist. Yet, as Bettina Bradbury asked at the May ratification meeting, "If women whose anomalies are under $2500 aren't to receive money, does this mean we're accepting the position that women are worth $2500 less than men?" This interpretation, potentially embarrassing to York's administration, is one that we must continue to introduce as part of a longer-range political process.