Smith-Doerr and colleagues in AJS on Gender Pay Gaps in Federal Agencies

Laurel Smith-Doerr, past ISSR Director
While government employment is commonly believed to be controlled by neutral, formal pay structures, new research from a team of researchers led by University of Massachusetts Amherst sociologist Laurel Smith-Doerr has found that in a number of science-based federal agencies, gaps and differential implementation in current standardization schemes create gendered outcomes. The research, published in the American Journal of Sociology, indicates that the resulting pay gaps between men and women at these agencies can also be associated with the cultural gender frames of the agencies’ field of research.
The researchers analyzed federal employment data from 1994-2008 at the organizational level, focusing on seven U.S. science agencies that span physical, biological and engineering disciplines, as well as interdisciplinary research: the National Oceanic and Atmospheric Administration (NOAA), the National Institutes of Health (NIH), the Department of Agriculture (USDA), the National Science Foundation (NSF), the Centers for Disease Control and Prevention (CDC), the Environmental Protection Agency (EPA), and the Department of Energy (DOE). They received de-identified data from the Office of Personnel Management in 2012 as the result of a Freedom of Information Act request and to their knowledge have been the only recipients of the population level data.

In the agencies based on physical sciences and engineering – the sciences culturally framed as more masculine – the researchers found that more of the pay gap can be attributed to inequalities within jobs, or “within-job discrimination,” so that men are paid more than women in the exact same jobs at the same agency locations, even women at the same pay grade and with the same work experience.

In the agencies based on more gender-neutral sciences, such as life sciences and interdisciplinary agencies, they found that more of the pay gap can be attributed to differences in individual characteristics, so that men and women of different educational and racial backgrounds are hired into different jobs in the agency in a way that reinforces gender hierarchies, but does not produce much within-job discrimination.

“It is very easy for us in the U.S. to think in terms of individuals and to think about people making decisions, but actually it is really important to look at how organizations make decisions, and a lot of times that’s less visible to us,” Smith-Doerr says regarding the importance of an organizational analysis of pay. “Therefore, you need to have this kind of population-level data and statistical analyses in order to be able to see these trends, and that’s what we have here.”

Joining Smith-Doerr, a professor of sociology at UMass Amherst, in the study were: Sharla Alegria, assistant professor of sociology at the University of Toronto; Kaye Husbands Fealing, chair and professor at the Georgia Institute of Technology School of Public Policy; Debra Fitzpatrick, co-director of the University of Minnesota Center on Women, Gender and Public Policy; and Donald Tomaskovic-Devey, professor of sociology and director of the Center for Employment Equity at UMass Amherst. The researchers present their findings in “Gender Pay Gaps in U.S. Federal Science Agencies: An Organizational Approach,” published today in the American Journal of Sociology.

“When the findings of this study are shared with members of the scientific community, there is typically surprise that measurable disparities in earnings are persistent between women and men even at Federal agencies that use the General Schedule,” says Husbands Fealing. “We have a way to go toward reducing pay disparities in STEM fields. Our study shines a bright light on the persistence of earnings disparities between women and men, even when there is an appearance that the Federal system is designed to avoid these pay gaps.”

In addition to the differences between the gender-neutral and masculine science agencies in what explains gender pay gaps, the researchers found a striking range of variation at individual agencies. Differences in how agencies sort workers by their individual characteristics, for example, account for approximately 84% of the gender pay gap at the NSF, compared to about 47% at NOAA. Meanwhile, occupational segregation and pay grade variations account for about 33% of the gender pay gap at the USDA, compared to less than 9% at the NSF.

“The federal government is made up of hundreds of smaller agencies, and while they share some things, they each have local decision-making processes and workplace cultures,” says Alegria. “Most studies of workplace inequality examine only a few workers from lots of different workplaces or all the workers in just one workplace. These studies can show us that there are important differences in how inequality happens in different places, but without the kind of large-scale employer-employee matched data that we have it's very difficult to say what those important differences are or how they take shape across different workplaces. As far as we know our data are unique for allowing comparisons for how inequality operates across so many American workplaces.”

“Our research shows the importance and value of the type of systemic, periodic collection and analysis of employer level pay data that will be proceeding within the private sector under new EEOC reporting beginning September 30,” Fitzpatrick says. “With the new data, employers and the EEOC can get a better handle on organization level pay gaps and potential causes. Our research provides a model for how to do just that.”

The complete report, “Gender Pay Gaps in U.S. Federal Science Agencies: An Organizational Approach,” is available online at the AJS website.

Contact: Laurel Smith-Doerr, Professor of Sociology (UMass Amherst).