QnAs with Paula England

Fifty-five years after the passage of the Civil Rights Act in the United States preventing discrimination based on sex, gender equality in the workplace remains unattained. After rapid gains in the 1980s and 1990s, progress has slowed considerably and, by some measures, stalled, according to Paula England, chair of the Sociology Department at New York University. In her Inaugural Article (1), England, elected to the National Academy of Sciences in 2018, examines a panoply of trends and indicators of gender equality in the workforce. England discussed her work with PNAS.

Paula England. Image courtesy of New York University Abu Dhabi.

PNAS: You majored in sociology and psychology at Whitman College, Washington before getting your PhD in sociology at the University of Chicago. What piqued your interest in science?

England: As a kid, when you hear the word “science,” you think of the physical sciences. So I didn’t even know that the social sciences are sciences, but of course, they are. I got interested in sociology in high school, mainly out of an interest in doing work that would help disadvantaged communities. I envisioned myself as a social worker working on problems of poverty.

But when I got to college and studied sociology, I began to think that the social problems that I was interested in trying to solve could be addressed by doing good scientific research on what causes things like discrimination and inequalities by class background, race, and gender. It was with that vision that I applied to doctoral programs in sociology. I got interested in gender, I think, because A) I was a woman and B) I was a feminist. To me, feminism is about figuring out why we have gender inequality and doing something about it.

PNAS: What was the landscape for gender studies at the time you were entering the field of sociology in the early 1970s?

England: At no time in either my undergraduate or graduate education did I have any class that had gender or women in the title. Few courses ever mentioned gender. I was taking courses about things like explaining racial group differences in pay. I realized that you could apply the same methods and related concepts to answering questions about gender inequalities, and I set out to do it although almost no one was studying these things. But it was clear to me that, while biology has something to do with gender differences, many facets of the differences in men’s and women’s outcomes could be explained by basic sociological principles. I was fortunate that I had a dissertation committee that supported me, although none of them were telling me that these were important questions.

PNAS: A large portion of your early work focused on comparable worth, which you call “a complicated notion of discrimination.”

England: It’s complicated because when most people think of employer discrimination against women, they think either of A) men and women with the same seniority and performing equally well in the same job but getting paid differently, or B) equally qualified men and women applying for the same job and employers being less willing to hire the women. Those are important types of discrimination, but I discovered something more subtle and nearly invisible going on as well: People in predominantly female occupations are paid less than those in predominantly male jobs, even when the jobs, while different, are comparable on what we’d expect them to pay based on things like how much education they require, their physical demands, whether they are unionized, and so forth.

My early research convinced me that employers were responding, perhaps unconsciously, to the sex composition of a job when they set its pay band. I arrived at that conclusion because predominantly female jobs pay less to both the men and women in them than you would expect based on the amount of education they require or a lot of other measurable factors about the jobs and the people in them (2, 3).

There are lots of factors that affect why one job pays more than another, and you can statistically control for those things if you can measure them in big representative datasets. But no matter how many statistical controls I put in, I still kept finding that men or women faced a pay penalty for being in a female occupation. This was true for predominantly female jobs at every educational level.

PNAS: What was the most convincing analysis you did to look for a pay penalty for predominantly female occupations?

England: In response to some of my early analyses, critics suggested an explanation other than employer devaluation of occupations because they were filled with women. They would say things like, “Well, maybe the less ambitious of both men and women go into the female occupations, and you don’t have a measure of that, so can’t control for it.” So I thought, “Okay, what could I do that would address that possibility?”

I started using panel data where you have data on the same person over multiple years; you see year by year if they’ve changed occupation, pay, and so forth. With these kind of data you can get a little closer to seeing what is causing what with a “person fixed-effects model.” Because lots of people change jobs at some point in their life, you can use those changes to see if there’s a tendency for the very same person, whether they’re male or female, to make more when they’ve moved to a male occupation and make less when they’ve moved to a female occupation (4). My group did this analysis, and it, too, showed a pay penalty for being in a female-dominated occupation, but this analysis gave us more confidence that the low pay wasn’t a function of a less ambitious people selecting into female occupations.

The next step—in a paper I published in 2009 (3)—examined whether occupations that feminized more (i.e., increased their percent female more) between 1950 and 2000 saw a decline in their relative pay. This is exactly what we found.

PNAS: In your Inaugural Article (1) you revisit an analysis of multiple measures of gender equality in the workplace that you published almost a decade ago. What did you find?

England: This article takes a descriptive look at multiple indicators of gender inequality (1). It assesses trends from 1970 to the present. The earlier paper had looked at fewer indicators and had cut off before 2010 (5). But in this paper (1), I get the same conclusion: That gender inequality fell a lot in the 1970s through some time in the 1990s, but progress has slowed or stalled since then.

As an example, the percent of women who are employed for pay increased dramatically after 1970, but has changed little since 2000. The segregation of occupations—how much men and women are concentrated in different jobs—has decreased continuously, but the rate of change has really slowed. The pay gap decreased the most in the 1980s. It is still closing, but much slower now. If we look at how segregated the fields of study in which people get bachelor’s or doctoral degrees are, there was substantial desegregation, but that progress has been stalled for at least 20 years. It is not an overly optimistic picture. One way to think about it is that the low-hanging fruit has been picked, and the institutional and cultural forces that perpetuate gender inequality are not easy to change. But many of us will not give up, either on studying the subject or trying to change gender inequality.