Wages are slowly leveling out in industries where both sexes work. But occupational segregation creates a different kind of pay inequity, shutting women out of higher-paying jobs altogether.
This article was made possible because of the generous support of DAME members. We urgently need your help to keep publishing. Will you contribute just $5 a month to support our journalism?
Something happened to the U.S. gender gap in wages around 2004: It stopped shrinking. It’s no accident that the push to integrate occupations by sex also stopped during the early 2000s. After all, occupational segregation of women and men is a major cause of pay discrepancies. While some occupations are roughly integrated, many others, such as traditional blue-collar jobs and STEM jobs, are mostly male and yet others, such as nursing and secretarial work, are mostly female. And mostly male jobs pay better than mostly female jobs.
So why are women still crowding into female-dominated jobs? Do women choose to earn less, in exchange for more pleasant working conditions and greater flexibility for combining family duties with paid work, as many conservatives argue, or are women’s choices not truly voluntary, but constrained by both labor market discrimination and gender norms, as many liberals argue?
Clearly, even small shifts between occupations could increase women’s earnings, as these Bureau of Labor Statistics data show:
In 2016, the weekly earnings of janitors, who are mostly male, were 19.2 percent higher than the weekly earnings of maids and housekeepers, who are mostly female, despite the two jobs having similar responsibilities. Parking lot attendants, mostly male, earned in the same year 25.4 percent more per week than cashiers, mostly female, though the requirements in the two jobs considerably overlap. Computer and information systems managers, mostly male, earned 30 percent more per week than human resource managers, mostly female, even though the two jobs share some responsibilities and educational requirements.
We shouldn’t make too much of those examples, both because part of men’s greater earnings is caused by the longer hours men work in the labor market, on average (while women work longer hours at home), and because inside each of those jobs men also earned more than women did. Still, occupational segregation by sex matters in the overall gender earnings gap. If we could integrate the traditionally male- and female-dominated occupations, the gender gap would shrink.
That would require changes in the jobs of men and/or women. Indeed, as Francine Blau and Lawrence Kahn describe in their meta-survey of the field, the entry of many women into managerial jobs during the 1980s and 1990s caused the gender gap to shrink. But progress in occupational integration has largely stalled since the early 2000s, just as the shrinking of the gender gap itself has stalled.
If female-dominated jobs pay less, why do women stay in them? Why don’t maids and housekeepers apply to become janitors, instead? Why do many young women still elect college majors which lead to less well-paying careers than the college majors young men tend to elect?
Economists have long grappled with this question. The oldest answer begins by noting that the total compensation for a job is a package of both pay and the particular job’s general qualities. Some of these qualities are desirable (flexibility, job security, a pleasant and safe working environment), others are undesirable (being exposed to inclement weather, greater bodily risk, more stress, responsibility, and greater earnings fluctuations over time).
Jobs with many undesirable qualities must pay more to attract enough workers, while jobs with many desirable qualities may attract enough workers even if their pay is lower, all other things being equal.
Suppose that women value, say, flexibility, much more than men, on average. This could explain — if female-dominated occupations in fact are more flexible — why women end up crowded into them even if they pay less.
Flexibility is assumed to be more important for women, because cultural norms expect women to be in charge of child-rearing and other household tasks. It is women, not men, who would look for jobs where the scheduling is flexible and where re-entry into the labor market after longer absences (such as maternity leaves) is easy and not severely punished in lower pay.
Men, on the other hand, might value the pay of a job more than its other qualities, because their traditional family role is that of a breadwinner.
But do female-dominated jobs actually offer more flexibility than integrated or male-dominated jobs?
Recent evidence on this is scarce and mixed. Claudia Goldin finds that women with children are disadvantaged in high-paying jobs in law, business and medicine because those jobs require very long working hours and inflexible scheduling.
On the other hand, a recent European study finds that female-dominated jobs both pay less and offer less scheduling flexibility. Likewise, the lowest paid jobs (including such female-dominated fields as cashiers, maids and housekeepers, and restaurant servers) in the United States do not offer more flexible scheduling.
A different economic theory looks into psychological traits which might show sex differences. For example, men, on average, have been shown to be more competitive and risk-loving than women in both laboratory experiments and in some studies about sports and competition, either because of innate sex differences and/or because of the effects of gendered upbringing.
Blau and Kahn note, however, that all psychological trait differences (including risk-aversion and competitiveness) between men and women only account for a small fraction of the existing occupational segregation by sex.
Women are clearly willing to compete in college admissions and women take larger risks than men in some social situations. Women are also more competitive and no less risk-taking than men in traditional matrilineal societies. That social influences affect the observed gender differences in these traits means that they could be changed.
Some jobs can also be riskier for women than for men. Take, for example, the most common male job in the U.S. in 2016: truck driving. The package of its job qualities contains a certain risk for all drivers, given the isolated nature of the job, the necessity to drive at night and the dangers of traffic, but those qualities impose an additional risk for female truck drivers: that of sexual harassment and sexual violence. Women don’t need to be less risk-loving than men to bypass such occupations, because the actual risk they face is higher.
A third explanation for the stubbornly sticky occupational segregation by sex is more psychological: Jobs develop a gendered identity, and that identity becomes enmeshed with our views about our own masculinity and femininity. Some jobs are viewed as women’s jobs (caregiver, assistant), while other jobs are viewed as men’s jobs (traditional male-dominated blue-collar occupations):
Research in Canada has compared reactions to ads for the same jobs that used stereotypically masculine words (leader, competitive and so on) or feminine ones (such as support, interpersonal and understand). Women found the “masculine” jobs less appealing, but not because they felt they would be unable to do them. They read the words as a signal of a male-dominated workplace, where they would not belong.
Stereotypes that discourage men from female-dominated jobs are just as ingrained. Florence Nightingale, who established the principles of modern nursing in the 1890s, believed that men’s “hard and horny” hands made them unsuitable for the job, “however gentle their hearts.” Some American nursing schools started admitting men only in 1981, after a Supreme Court ruling.
If this discomfort caused by “not belonging” is strong enough, women might view well-paying but male-dominated occupations as less desirable. It could also explain why working class men left unemployed by the recent outsourcing of blue-collar jobs can find accepting the still-available service jobs so difficult. They are “ women’s work.”
Explicit labor market discrimination against women is now illegal, but it’s not a recent law. Help-wanted ads in newspapers were listed separately for men and women until the 1970s, and the famous 1970s AT&T discrimination case showed that the company had maintained sex-segregation in its workforce by, for example, giving a different list of available jobs to male and female job applicants. But gate-keeping against women can be managed through other types of discrimination, including sexual harassment at work, discouraging women from applying for certain jobs or leaving them. It’s how certain industries, including perhaps the brotopia of Silicon Valley, remain segregated for so long.
Likewise, recent statistics show that manufacturing was the third most common industry in the numbers of sexual harassment claims the EEOC received between 2005 and 2015, after two mostly female and low pay occupations (accommodation and food services, and retail trade).
Because it is mostly women who submit harassment claims and because women are only a small minority of manufacturing workers, that third-place ranking tells us that sexual harassment is a particularly serious problem in the male-dominated manufacturing industry. It could be one reason why women remain a minority among manufacturing workers.
So far we have been trying to answer the question: Why do women “choose” to enter and stay in a low-paid female-dominated occupation? But what if it’s not that women choose low-paid jobs but that jobs become low-paid after large numbers of women choose them?
Paula England and Asaf Levanon studied this question using US census data from 1950 to 2000, and found that occupations indeed start paying less after women enter (and men leave) them in large numbers, even when the workers’ other characteristics are held constant. As an example, the job of a ticket agent turned from predominantly male to predominantly female over the studied time period and its median hourly earnings fell 43 percentage points, after controlling for the effects of inflation.
Computer programming, on the other hand, has turned from a female-dominated job into a male-dominated one, and its prestige and pay have risen.
Do employers, and perhaps the society on the whole, value women’s work less, simply because it is done by women?
Which policies would best help in closing that part of the gender earnings gap which is due to occupational segregation by sex? Raising the minimum wage helps to decrease the gap, because women are over-represented in the least-paid jobs. Stronger unions in traditionally “pink-collar”occupations would level the playing field in wage negotiations.
Paid parental leave and better child-care services offer more real job flexibility for parents. Better sharing of child-care duties between parents would also raise the popularity of such programs which would no longer be seen as helping only women. And we just might want to return to the old programs which encouraged occupations to become gender-integrated. It may be a coincidence, but once those programs ceased so did the movement toward more integrated jobs.
It’s harder to change our ideas about the gendered nature of jobs themselves or about the value of work done by women. We do need a national conversation on those issues.