Members of the subcommittee on gender and salary equity have found evidence of a gender discrepancy in the salaries of Williams professors.
Specifically, a complex statistical model has revealed that veteran male professors – those who have worked at the College for more than 15 years – are making significantly more than would be expected.
Associate Professor of Biology Heather Williams said the College formed the subcommittee of the faculty compensation committee last year to investigate the possibility of a gender discrepancy among the academic departments similar to a discrepancy discovered in the athletic department.
Professor Williams said last year the administration learned that coaches of men’s sports teams earn more than coaches of women’s sports teams, and thus decided to investigate whether or not there were similar inequities in the professors’s salaries.
Williams, who is one of the four members of the gender and salary equity subcommittee, said the committee was formed despite the initial reluctance of the administration to supply faculty salary data.
“Eventually, [the administration] decided that it would be in everybody’s best interest to [supply faculty data],” Williams said.
Dean of the Faculty David L. Smith said originally the administration was concerned about releasing the salaries, but trusted the faculty and the subcommittee members.
“The vast majority of the faculty is committed to confidentiality,” Smith said. “Because the subcommittee is elected it’s appropriate for it to look at the data. We have no problem trusting the committee.”
Associate Professor of Mathematics Richard De Veaux presented the committee’s findings and recommendations at a faculty meeting on September 16.
De Veaux explained the process the committee used to analyze faculty salaries. He said the first task of the committee was to create a model of the faculty salaries that could be used to predict individual salaries. Such variables as department, year of Ph.D., and whether or not the faculty member came to Williams with tenure were factored into the model.
Analysts then calculated the discrepancy between predicted salaries and actual salaries. De Veaux explained that a positive discrepancy, or residual, implies that a faculty member is earning above the predicted salary after all the relevant factors have been taken into account. Similarly, a negative residual implies a faculty member is earning less than the predicted salary.
The first finding of the subcommittee revealed that there are a group of associate professors who are paid between $4,000 and $6,000 less than other professors of the same rank. De Veaux explained that in the 1980s the administration raised salaries in an effort to adjust Williams salaries to match those at similar colleges. The associate professors were not included in the raises, and a corrective measure put in place a few years ago failed to remedy the situation.
“The problem is [the administration] miscounted our benefits a few years ago,” De Veaux said. “They gave quite large increases to all the faculty, but they gave even larger ones to the full professors. If that compounds over time, the increase difference rises.”
A large proportion of this group of professors was female.
But members of the subcommittee said the high percentage of female professors in this group was coincidental, and should not be interpreted as discrimination.
“(It) in no way should be construed as a case of gender discrimination,” Professor Williams said. “Because that group was substantially depressed it comes out looking as if there’s a gender problem.”
The committee recommended that the administration correct this discrepancy and institute any necessary retroactive changes.
The committee also discovered a difference in the average residuals (the differences between the actual and predicted salaries) of male and female professors. Professor Williams explained that when the average residual of all male salaries was compared to the average residual of all female salaries, the male average residual was approximately $1000 higher.
Professor Williams said the committee then restricted the number of professors considered, comparing only the mean residuals of male and female professors who have worked at Williams College for fewer than 15 years. When only these professors were considered, the difference in the average residual dropped to $200.
“It’s clear that at the bottom end the men and women are fairly evenly mixed,” Williams said. “The difference comes in at the top.” She added that the 12 professors with the highest residuals were all male, and of the top 20 only one was a woman.
De Veaux said further research needs to be done before the results can be substantiated. Specifically, he said, other factors should be included in the model used to predict the difference between predicted and actual salaries.
“It is important to point out that no notion of merit was included in the model,” he said. “We didn’t know how to measure it. We don’t know as a faculty how much that should count, but there was no attempt to put that in the model.”
Professor Williams noted that the first finding related to associate professors is easier to explain than the second finding of gender discrepancy.
“We don’t really know how to interpret [this discrepancy],” she said. “I think that is what people are wrestling with now. How do we interpret that and how do we look at that question.”
De Veaux added: “We want the administration to look at it and tell us how this happened that all the top 12 professors were male.”
At tomorrow’s faculty meeting, Provost of the College Stuart Crampton will speak on the committee’s findings.
Williams added that the committee has recommended that an ad hoc committee be created to investigate further the why the gender discrepancy in residuals.
Dean Smith said the College will look continue to investigate the issue.
“We are very pleased with the good work of the subcommittee,” he said. “This report is an excellent report. They have recommended a further study into one of their findings and we are currently in the process of looking into it. We think we have a good cooperative process and that it will work out well.”
Williams explained that the purpose of further investigation would not to point fingers at anyone in particular.
“We’d like to have a sense of what reasons have been considered to be appropriate to raise someone’s salary above what the model predicts,” she said. Professor Williams cited such outside factors as job offers, outstanding work, and heavy committee work as possible explanations for the high residuals of those 12 professors.
“What I, personally, am hoping is going to come out of this is some sense of demystification of how salaries work at Williams,” she said. “We all want a place where people don’t worry about their salaries.”
Professor of English Suzanne Graver, who will serve on the new ad hoc committee, said the committee plans to meet for the first time in the upcoming week. She said the committee will meet for one month and then report its findings to the administration.
Graver said she suspects that a high departure rate among “distinguished” female faculty members may have contributed to the discrepancy.
“Over the years there have been some very distinguished female faculty members who have been picked off to go elsewhere after the College tried very hard to keep them,” she said. “If they had stayed, they may very well have been in tha
t group of high salaries. That might be one factor, but there may very well be other factors.”
“The ad hoc committee is going to be able to illuminate to some extent why those people are in that group of high salaries,” she added. “However, it won’t be able to explain why the rest of the people are not in that group.”
Both De Veaux and Williams emphasized that it is important to keep in mind that the gender discrepancy in the salaries of Williams professors is relatively small.
“I think that there are some issues that are somewhat troubling, but I don’t think there are the kind of gross systematic gender problems in salaries that some people were expressing,” Williams said. “There were, however, major problems in the athletics department.”
“One thing that needs to be said is that Williams is incredibly consistent in salary compared to most places,” De Veaux added. “We’re talking about relatively few dollars. It’s really important to put this all into perspective.”