New Evidence on Gender Differences in Promotions and Pay
Controlling for all variables, ... men's promotion rates were ... 2.2 percentage points higher than women's, ... large relative to the overall mean promotion rate of 9 percent. However, in marked contrast, ...there were essentially no gender differences in overall wage growth at the establishment.
In a market economy, wages and promotion are linked because promotions often come with a wage increase. One important question is whether equally qualified men and women with similar characteristics have an equal probability of advancing their careers through promotion. Others are whether there are gender differences in the wage increases that accompany promotions, and whether any observed gender difference in promotion produces gender differences in overall wage growth at the establishment. However, determining whether equally qualified men and women have been treated similarly can be difficult. If, on average, men and women differ in their (unobserved) preferences for schooling, the amount of time they spend on housework, their attachment to the labor force, or their working environment, then comparisons of male and female promotion or wage rates may uncover differences in both. Those differences may appear to be the result of prejudice but may in fact be the result of individual differences in tastes an d productivity.
In New Evidence on Gender Differences in Promotion Rates: An Empirical Analysis of a Sample of New Hires (NBER Working Paper No. 12321) co-authors Francine Blau and Jed DeVaro use data from the Multi-City Study of Urban Inequality, a large cross-sectional telephone survey of 3510 employers in Atlanta, Boston, Detroit, and Los Angeles, conducted between 1992 and 1995. The survey asked about an establishment's most recently hired worker: was he or she promoted by the date of the survey, or expected to be promoted in the next five years; the worker's wage at the time of the survey, the worker's starting wage; and the wage the worker could expect to receive if promoted. The survey is also unique in that it allows the authors to control for the worker's job-specific performance rating, as well as nonprofit status, industry, establishment size, franchise status, and unionization. It is quite unusual to observe all these characteristics for such a broad sample of workers.
The authors analyze the extent to which reported productivity and observable characteristics account for any observed gender differences in promotions and wage growth. Without controlling for any gender differences in characteristics, the authors find that 10.6 percent of men had received promotions versus 7.6 percent of women, a statistically significant gender gap in promotion rates favoring men. There were only slight, statistically insignificant, differences in wage growth and expected wage growth between men and women, though.
The authors then control for differences in the observed characteristics of men and women, including the fact that they were not distributed equally, either in occupational categories or across industries. For example, men were more likely to be scientists, engineers, doctors, and lawyers; more likely to work in craft, construction, and transportation jobs, and as production workers and laborers; and more of the men worked in the for-profit sector. Women were more likely to be teachers, librarians, and counselors, and to work in administrative support occupations.
The additional controls slightly reduce the gender difference in promotion rates but, controlling for all variables, including worker performance ratings, men's promotion rates were still 2.2 percentage points higher than women's. This is large relative to the overall mean promotion rate of 9 percent and the raw gender difference in promotion rates of 3 percentage points. However, in marked contrast, the authors find that after controlling for measured characteristics, promotions and expected promotions continued to yield comparable wage increases for both men and women. And, there were essentially no gender differences in overall wage growth at the establishment, with or without promotion.
-- Linda Gorman