Women have made huge strides in the workforce in a remarkably short span of time. Many of us forget that in 1970, not so long ago, men still earned 99% of dentistry degrees, over 95% of law degrees and MBAs, and over 90% of medical degrees awarded in the U.S.
It suddenly made more sense for women to invest years in earning those degrees after the birth control pill was approved for married women in 1960 and for single women circa 1970. Pregnancy was now no longer a constant risk to degree completion, so women could delay marriage and childrearing, and invest in their careers.
In the almost fifty years since, the proportion of women with a college degree has more than tripled, their labor force participation rates have soared, they have entered previously male-dominated occupations in droves, and the earnings gap between men and women has narrowed markedly. Of course, substantial gender gaps in employment and earnings still remain. Here are the main reasons, and some evidence.
- Childbearing, differences in gender roles and the gender division of labor
For whatever reason (biology, preferences, culture, and patriarchy are all candidates), women do far more unpaid childcare and housework than men. According to a recent OECD report, mothers in paid employment spend more time on childcare even than non-working fathers. Even in dual-physician and dual-lawyer “power couples,” women spend an hour more daily on housework and childcare.
When women shoulder more of the household responsibilities, that reduces the time that they spend on paid labor. It also effectively supports their partners’ ability to spend more time on paid labor or to accept longer commutes and more frequent business travel for higher salaries. So women doing more household work widens the pay gap in both directions.
While women spend more time cooking and cleaning, studies find that the bulk of the earnings gap can be traced to the arrival of children. One study finds that the birth of a first child creates a long-term gender gap in earnings of around 20%. Vox.com made recent scientific research on the topic more accessible with this arresting set of graphics showing a stark divergence in earnings for women, but not for men, upon a child’s birth.
- Women’s workforce interruptions and shorter hours
Because women tend to take greater responsibility for raising children, they also tend to experience more workforce interruptions and to seek jobs with shorter, and more flexible hours. But many jobs are still structured around working long hours on fixed, non-flexible schedules. In many occupations, earning promotions or climbing the advancement ladder requires a kind of frenetic and uninterrupted hustle during one’s 20s and 30s—the prime childbearing years when women are most likely to experience labor force interruptions. One study finds that these interruptions explain around 19% of the wage gap.
The preeminent expert on women’s labor force participation, Harvard University’s Claudia Goldin, finds that workforce interruptions and working hours are the key factors explaining the persistent earnings gap. She writes: “The gender gap in pay would be considerably reduced and might vanish altogether if firms did not have an incentive to disproportionately reward individuals who labored long hours and worked particular hours. Such change has taken off in various sectors, such as technology, science, and health, but is less apparent in the corporate, financial, and legal worlds.”
In this tight labor market, employers are increasingly offering flexible hours and work days as one way to attract candidates. In just the past year, the number of job postings on ZipRecruiter offering flexible schedules has risen 110%. That should present an exciting opportunity for women who took a break from the labor market to raise children, and are now looking to ease back in.
- Gender differences in occupations and industries
Perhaps anticipating the need for more workforce interruptions and more flexible hours, women are less likely than men to work in high-wage industries or high-wage firms. Conversely, there is considerable evidence that women work in industries with better non-pay characteristics, like better hours, work-life balance, and safety standards.
Although women now outnumber men in grad school by 137 to 100, and earn a majority of doctoral degrees in 7 of 11 fields tracked by the Council of Graduate Schools, including Biology, Health and Medical Sciences, they are still under-represented in some of the highest-paid fields: Business, Engineering, Math and Computer Science, and Physical and Earth Sciences.
Why are there still so few women in STEM (science, technology, engineering, and mathematics)? Many explanations have been proposed, including differences in preferences, a lack of female role models in STEM, the lasting power of stereotypes, and discrimination. Much of the difference in STEM training emerges before college, in high school. Culture may play a role. One study finds that immigrant girls with parents from more gender-equal countries perform better at math than girls with parents from less gender-equal countries.
But there’s a puzzling gender-equality paradox in STEM: According to a recent study published in the journal Psychological Science, the more gender equality in a country, the fewer women in STEM. For example, in Finland, Norway, Sweden, Switzerland, and the Netherlands, countries with more gender equality, there is a higher gender gap in STEM. One possible implication is that when women are truly free to choose, they choose what they like—and, on average, they like other fields more than STEM.
It is true that no women made Forbes’s most recent list of the ten highest-paid athletes—but it is also true that no men made Forbes’s list of the ten highest-paid fashion models. Even rank-and-file female models earn a lot more than male models. That probably has something to do with the fact that women spend much more time and money shopping for clothes and watching fashion-related content, whereas men spend more time watching sports. Perhaps, then, some of the wage gap is merely a reflection of differences in preferences and priorities.
- Psychological attributes or noncognitive skills
This somewhat controversial suggestion leads to a fourth set of explanations for the gender gap, that emphasize differences in the psychological attributes of men and women. There is a large and growing body of evidence, for example, that men are more competitive than women and perform better in competitive environments. Differences in willingness to compete partly explain why women choose to specialize in mathematics less frequently.
Women are also found to be significantly more risk averse, with the result that they choose less risky asset investments, and settle for lower returns. There isn’t just an earnings gap, but also an investing gap. Aversion to risk and competition may also cause some women to shy away from joining startups or working as entrepreneurs. Substantially fewer women than men are self-employed and just 17% of startups have a female founder–although that may be because of differences in access to venture capital dollars. (Just 2% of VC dollars went to female founders in 2017.)
The differences go on. There is evidence that women more frequently volunteer to perform low promotability tasks that nobody really wants to do, like taking notes in meetings or sitting on boards. The highly publicized recent study of the 7% Uber pay gap in favor of male Uber drivers highlighted further behavioral differences. Twenty percent of the pay gap is because men tend to drive more profitable routes and neighborhoods. And 50% is due to men driving faster or more efficiently, i.e. covering a greater distance per driver-hour.
One possible reason is that men may simply be more driven by money. Perhaps they devote more effort to figuring out ways to game the system—even if these require taking on greater risk (e.g. by driving too fast). Remember, women have substantially higher life expectancy than men (81.3 versus 76.5 years)—partly because so many more men are lost to accidents (6.8% of total deaths among men versus 4.0% of total deaths among women).
In our own ZipRecruiter marketplace, we find that women apply to fewer jobs and ask for less money. The percentage of job seekers who apply after clicking on a job posting is higher for men (36%) than for women (33%). Among those who apply, the self-reported desired salaries of men are $11,103.65 higher than those of women. In the Northeast, the gap is even larger, with men reporting an average desired salary of $80k per year, vs only $61k for women. In a recent ZipRecruiter survey of 50,000 job seekers, 46.5% of men said they had turned down prior job offers, compared with only 38.7% of women. When they accepted their last job, 38.5% of men say they negotiated for higher pay rather than accepting the opening salary offer, compared with only 30.4% of women.
Differences in job search behaviors could have real earnings effects down the road. A study conducted in 2000 found that for most women, their first job out of college was also their only job offer. Applying more widely and getting more offers leads to a better selection of first jobs, which leads to higher pay and job satisfaction.
A final reason for the gap may be lingering discrimination against women—even among the most educated, scientific, and self-professedly rational professionals. There is plenty of evidence in the literature. Several studies find that women economists are less likely to be promoted to tenure than men, and while productivity and the effect of children partly explain the gap, a significant portion of the promotion gap is not explained by observable characteristics. Women also appear to get less credit for co-authored research than men.
One might assume that discrimination would become less of a problem as more women rise into leadership roles. Surely female supervisors would be more inclined to hire and retain female managers? Some studies find evidence that this is the case, but other studies suggest it isn’t always so. There is a very large body of evidence, however, that in fields where they are underrepresented, women are more likely to persist in a field if they have a female instructor or mentor.
Men in fields like education, where they are underrepresented, similarly become more likely to pursue the field when they have a male instructor. For whatever reason, same-gender instructors and mentors appear to positively influence interest and achievement in a subject. Perhaps this is due to discrimination by opposite-gender supervisors, but perhaps it is simply due to a greater affinity, understanding, and identification between same-gender trainees and supervisors.
Some suggestions for narrowing the gender earnings gap
Both employers and workers can take steps to shrink the gender earnings gap. The most important change employers can make is to evaluate whether rigid working schedules are really necessary and to adopt more flexible scheduling systems where possible. Technology is making it ever more possible for many jobs to be performed anywhere, any time, and for organizations to use shift-scheduling systems that put more weight on employee needs and preferences, without compromising the performance of the mission. Given the strong positive impact of same-gender supervisors, another step employers can take is to ensure that there is gender-diversity within each department.
Women can improve their chances of finding high-paying jobs with desirable working conditions by applying to a larger number of positions, gathering more information about typical salary levels, and negotiating for competitive pay, rather than settling for the first offer. Information is power. Women who find that having female mentors and role models increases their interest in, and attachment to their subject should by all means seek such female mentors, if not within their organizations then through social media or good old fashioned networking.
Women should also be encouraged to learn more about investing and invest more of their earnings. Finally, women should not meekly tolerate being given all the office “housework” by default, but rather propose, say, a roster in which low promotability tasks, such as taking meeting minutes, are rotated fairly across all team members. Oh, yes, and the same goes for the actual housework too.