Available Studies
- Franklin v. University of Toronto. 2001.
- UTFA. UFTA Salary Anomaly Fund Proposal. 2012: 2012-Jan-14 UTFA Anomaly Fund Proposal.pdf.
- Shillington, Richard. Gender and Salary Data Report. Tristat Resources. 2017.
- University of Toronto. Faculty Gender Pay Equity. 2019: Gender-Based Salary Equity – Division of the Vice-President & Provost (utoronto.ca).
The available documents begin with a 2002 letter regarding a legal case brought on the University of Toronto by retired female faculty and librarians. The case resulted in undisclosed settlements to retired female faculty due to gender pay inequity, but the settlement excluded librarians due to differing administrative bodies and policies overseeing the pay of librarians.
In 2012, a fund was proposed by the University of Toronto Faculty Association (UTFA) to address anomalies in pay across a number of dimensions. Gender equity was identified as one type of salary anomaly to be investigated. A committee would be created to assess anomalies and contribute funds to base salary adjustments if such anomalies were found.
Files available from 2017 show technical analysis of UFTA split into three groups: Librarians, teacher-stream, and tenure-stream instructors. The following is a summary of data, methods and results of these later documents.
The data is of 2,010 tenure stream faculty for one study, 154 librarians in another study, and 543 teaching staff in the third study. The data includes indicators of new hires and promotions as well as characteristics like rank and time since degree and time since promotion.
In each of the studies, multiple linear regression with robust standard errors were used to regress the logarithm of annual salary, however slightly different regressors were available in for each study. In the librarian study, log salary was regressed on gender, age, campus location, grouping, librarian level (similar to rank), and years in stream. For the teaching stream log salary was regressed on gender, age, campus location, division group (like faculty), rank, years in since degree, and years in rank. Finally, in the tenure stream faculty log salary was regressed on the same variables as the teaching stream faculty with the addition of an indicator of full-time status and a variable labelled CRC, and home department. However, the regression was run under multiple specifications.
In the tenure stream faculty report, a Blinder-Oaxaca decomposition was used to estimate the gender pay differential. A multi-year comparison of gender pay gap estimates was also included for the tenure stream faculty. Extensive visual analysis was also incorporated into reports, box-plots and histograms of various dimensions in the data set were included to illustrate the underlying distributions in the regression analysis. Finally, the regression analysis of 764 new hires was also included in the tenure-stream report.
For Librarians, the salaries differences between males and females are not significant but the sample size is small. There is significant evidence that salary is determined by librarian rank. Men were observed to be promoted to the highest rank of librarian, while lower level promotions were not significantly different between men and women.
For the teacher stream faculty, the results were similar to librarians. Differences in salary by gender were not statistically significant after adjusting for the other factors. However, while most of the data showed little preference for men or women in hiring, there was a clear trend of the higher ranked positions to be assigned disproportionately to males.
Finally, the analysis of tenure streamed faculty showed that the estimated gender impact on salary was between 1.2% and 2.2%, depending on the inclusion or exclusion of home department and different measures of experience. In the Blinder-Oaxaca decomposition, almost all of the gender difference in salary can be “explained” by differences in the characteristics of the faculty. The most important factors were found to be rank and division and then years at rank. The estimation of salary impact due to gender was stable over the eight years of available data, with no clear trend of increasing or decreasing.
In the analysis of new hires and promotions, a significant result for gender on starting salary was discovered. An average effect of 7.6% was found to be attributed to gender, however, the model delivering this result controlled for less of the overall variation in the data with an R-squared of 42%. Similar to other staff members, initial promotions show no gender advantage but the highest rank promotions showed a male advantage.
The administration undertook a multivariate regression model controlling for years since highest degree, rank, academic unit, and other relevant factors. The additional factors included experience prior to hire, administrative positions, and whether or not the individual had ever held a Canada Research Chair or University Professorship. The estimation was conducted separately for tenured and tenure stream faculty, and continuing stream teaching stream faculty. single methodological approach—multivariate regression analysis. The report commended a 1.3% increase to base for all tenure stream faculty.
The faculty association used a mixed quantitative and qualitative approach for its members. This approach identified a probable gender-based salary gap for women in the tenure stream of at least double the 1.3% recognized by the Administration.