by Donald Tomaskovic-Devey , Karen Brummond, Skylar Davidson and Stephanie Tsamasiros
Noting that women make 77 cents on the male dollar, President Obama signed an executive order instructing the Department of Labor to collect pay data from federal contractors. The familiar criticisms of this statistic are that women and men are often in different jobs and sometimes choose different work schedules because of family responsibilities, so simple gender comparisons overstate the discrimination case. Collecting data would let us ask a much more useful question, which employers have high levels of gender disparity and which have none?
For many people the notion of a singular gender pay gap makes little sense. After all we know that gender and gender bias are not constants, but vary across contexts, always intersect in practice with other bases of distinction. Most of us can look around and see men and women working in the same jobs in the same workplaces. Women are bosses, CEOs, lawyers, and occasionally engineers. Men are nurses, elementary school teachers, and occasionally secretaries. Women sometimes have higher paid jobs than men. In many jobs, especially jobs that require a college degree, women and men work side by side.
Do these observations imply that there is now no male advantage in employment? Clearly not, but it does suggest that we should expect a great deal of variation across workplaces in the degree and sources of gender disparity. It is easy to criticize President Obama for using a naïve statistic, but this reflects in part our refusal to collect workplace level data. The President’s executive order that the Department of Labor begin to collect such data has the potential to make visible the real pay gaps that exist out there, including identifying workplaces with small or no gender pay gaps at all.
The sex typing of jobs is one well known source of gender pay differentials and varies widely across workplaces. Using data collected by the US Equal Employment Opportunity Commission we can observe the degree to which men and women are hired into different occupations in private sector workplaces. In 2012, the latest year for which data is available the index of dissimilarity averaged .32 on a scale of 0 to 1, where 1 means no women or men work together in the same occupation in any workplace [These data understate true segregation levels since they are at the major occupation, not job level.]
Notably, there is a wide distribution across workplaces. In some workplaces gender job sorting by this metric is close to zero, in others it is total. Not all employers separate men and women equally. Importantly, there are some workplaces that completely segregate their male and female employees.
While we cannot yet observe pay gaps in workplaces, we can estimate the within industry gender pay gap using data that the Bureau of the Census already collects. When we statistically adjust pay differences to account for gender differences in education, labor market experience, and local pay rates, we find an average gender pay gap of 19%.
But again more interesting is the dramatic range of gender pay gaps across industries. The figure shows that the gender pay gap varies from nearly zero to 35%. This strongly suggests that when the Department of Labor starts collecting data from federal contractors they will discover that some workplaces have no gender pay gap at all, many will have small gaps, but some will be outrageously high.
Today the largest gender pay gaps are in knitting mills (37%), information services (34%), water transportation (34%), ambulatory health care services (33%), and financial services (33%). Industries with small male pay advantages include nursing homes (1.6%), the U.S. Postal Service (3.1%), and religious and civic organizations (4.9%).
When we lack quality data to show the absence or presence of significant gender (or race or sexual orientation, or age or disability) wage gaps, responsible employers are at risk for unjust accusations of wage discrimination. Irresponsible employers can hide behind a veil of ignorance. What we really need is good data collection, not only on federal contractors, but on all workplaces, private and public alike. We should monitor equal opportunity in employment in the same way we do insider trading or environmental pollution, by documenting their actual practices.
If job seekers had these data they would know which firms to avoid. If consumers knew which firms were most biased against women or minorities they might decide to shop elsewhere. Communities contemplating subsidies to lure employers, might be able to decide if this is really the kind of employer they want in their community. And of course, lawyers might know which companies to sue over discrimination and which to leave alone.
Donald Tomaskovic-Devey , Karen Brummond, Skylar Davidson and Stephanie Tsamasiros are developing systemic report cards to rank firms, industries, and communities in terms of their gender and racial employment disparities. With Kevin Stainback, Tomaskovic-Devey is the author of Documenting Desegregation: Racial and Gender Segregation in Private Sector Employment since the Civil Rights Act (Russell Sage, 2012).