Highlights
Editor’s Note: This is the second of three posts in a roundtable on men and women at work that the Institute for Family Studies is hosting this week.*
The web is abuzz with discussion of a “manifesto” informally circulated at Google by James Damore, who, until yesterday, worked there as an engineer. Damore’s memo criticized Google’s diversity programs as “authoritarian,” labeling them “arbitrary social engineering” projects that ignore the biological differences between men and women, which he believes explain many of the gender disparities Google is trying to change—the dearth of women in tech and higher management and the gender pay gap. He suggested that Google should focus more on intellectual diversity, showing more respect for views like his.
Respecting intellectual diversity is one thing. Respecting intellectual blindness is another. Damore’s memo missed one huge thing: Abundant and rigorous scientific studies—by sociologists, psychologists, and economists—have demonstrated that gender and race biases adversely affect women and people of color in the workplace. In this post, I’ll tell you about a few studies that I believe provide convincing evidence of gender or race discrimination at work.
Sociologists Devah Pager, Bruce Western, and Bart Bonikowski investigated hiring for low-wage jobs. They used what is called an audit study, considered the “gold standard” for assessing hiring discrimination. They gave white, black, and Latino men equivalent résumés and sent them out to apply for hundreds of entry-level jobs. They found that black applicants were only half as likely as equally qualified whites to receive a callback or job offer. In fact, black and Latino applicants with no criminal record fared no better than white applicants who reported that they were just released from prison. This is strong evidence of discrimination in hiring.
To reduce the possibility of bias, more and more symphony orchestras have adopted the practice of “blind” auditions, where musicians perform behind a screen so they are not seen by the selection committee. Economists Claudia Goldin and Cecilia Rouse used data on many auditions in multiple cities across many years as this change had occurred some places and not others. They found that using the screen helped women. In fact, the same woman was more likely to be picked in an audition behind a screen than one where she could be seen. The results suggest that, without the screen, there was bias in favor of men.
We can disagree over the best way to reduce bias, but refusing to face its existence is not the answer.
Once hired, minorities and women still face bias. MIT Business Professor Emilio Castilla used Human Resources records from a large company to study whether racial or gender bias occurred in the assignment of raises among people already employed. He had access to managers’ numerical performance evaluations of professional workers, as well as the percent the workers’ salaries were raised in each of multiple years. On average, women and minorities got raises of a lower percent—even when their supervisors had given them identical performance evaluations! If Castilla hadn’t had access to the performance evaluations, we might wonder if the smaller average bonuses given to women and minorities simply reflected that they performed less well. But the striking finding was that bias somehow crept in between the numerical performance evaluations and the decisions about raises, which are often made by a different person.
A study I did with Asaf Levanon and Paul Allison focused on how much one job pays compared to another, using national data on hundreds of occupations. We wanted to know if employers view a job as worth less pay when most of the workers in it are women. The tough thing about studying this is that if a mostly-female job pays less than a mostly-male job, we can’t be sure whether the difference is because employers think women’s work deserves less or because the male job is just more demanding. We reasoned that if the same job was paid less when the percentages of female workers in the occupation increased, this would be strong evidence that employers are responding to the gender of those doing the work rather than the demands of the job. In a statistical analysis of data ranging from 1950 to 2000, we found that when the percent of women in an occupation goes up, the pay relative to other jobs goes down. We concluded that employers start seeing a job as worth less when women enter in large numbers. This contributes to women earning less than men in the economy at large. But this kind of sex discrimination against women can hurt men, too. That’s because, when the pay goes down in a job because it is “feminizing,” the pay of both the men and the women in the job goes down.
Of course, I am not claiming that all gender and race disparities are caused by bias in the workplace—there are many other factors at play. But this body of research convinces me that companies are doing the right thing when they implement programs to safeguard against gender and race bias. Damore decries diversity initiatives as authoritarian social engineering. But if companies do nothing to reduce bias, that amounts to social engineering, too—a type of social engineering that perpetuates the second-class status of minorities and women. We can disagree over the best way to reduce bias, but refusing to face its existence is not the answer.
Paula England is a Silver Professor of Sociology at New York University. The author of Comparable Worth: Theories and Evidence, Professor England is also past president of the American Sociological Association.
*The views and opinions expressed in this article are those of the author and do not necessarily reflect the official policy or views of the Institute for Family Studies.
