Changes in work force, not pay, narrowing the gender wage gap

"Though decades of economic research suggest men and women are equalizing in the labor market, the notion that today's working women are being paid more and treated better than ever before is simply wrong," said Rubinstein, assistant professor of economics. "The growing equality between genders reflects the entry of the most able women to the workforce rather than better pay. While there may be more women holding high-power positions today, they are still being paid as their counterparts were three decades ago."

After years of a fairly constant gender wage gap in the United States, women's wages grew from the late 1970s to the mid 1990s, and the gap seemed to narrow. At that same time, wages became much less equal within gender groups. Although previous economic observers have called these simultaneous growths "curiously coincidental," Rubinstein and Mulligan connect these two phenomena and show that growing wage inequality within gender groups was actually a catalyst for bringing "highly able" women into the labor market.

Rubinstein explained that in the 1970s, the labor market had an increased demand for "skilled workers." Because most of the "skilled men" were already in the workforce, the demand increasingly pulled in a pool of smart, skilled, and "highly able" women — those who were previously choosing to be at home. As a result, the United States saw an increase in how much the average working woman earned. The authors show that this wage growth for women might not have happened if the workforce composition had been held constant.

The authors suggest that growing inequality within gender, through its effect on women's selection into the labor force, their labor force attachment, and their human capital investment, is a major reason why the wages of the female workforce have grown relative to men's — thus giving the impression that they are being treated more fairly than they were 30 years ago. Using data from the Current Population Survey and IQ data taken from the National Longitudinal Survey, the authors used three different empirical approaches to measure the existence and importance of these effects.

Source: Brown University