A New York Times article looks at the findings of several recent studies on this puzzle:
A new study from researchers at Cornell University found that the difference between the occupations and industries in which men and women work has recently become the single largest cause of the gender pay gap, accounting for more than half of it. In fact, another study shows, when women enter fields in greater numbers, pay declines — for the very same jobs that more men were doing before.
Consider the discrepancies in jobs requiring similar education and responsibility, or similar skills, but divided by gender. The median earnings of information technology managers (mostly men) are 27 percent higher than human resources managers (mostly women), according to Bureau of Labor Statistics data. At the other end of the wage spectrum, janitors (usually men) earn 22 percent more than maids and housecleaners (usually women).
Bolds are mine.
Occupational segregation (1) by sex has long been known to be one important reason why women, on average, earn less than men, on average: Because many more women than men work in low-pay industries and occupations. Almost all predominantly female industries and occupations pay less than almost all predominantly male industries, and job titles traditionally associated with women's jobs (say, secretary) pay less than job titles traditionally associated with men's jobs (say, administrative vice president for office work) (2).
Now tease apart that giant knot! How much of this is because of direct sex discrimination? How much because of indirect sex discrimination: The steering of girls into certain fields, the pressures put on boys and girls to have jobs compatible with traditional gender roles at home? So that boys expect to have to earn more than half of the family's income one day, while girls expect to have to do the bulk of child-care one day?
Or, as the gender essentialists argue, do men just happen to like and thrive in jobs which pay well, whereas women just happen to like and thrive in jobs which do not? Are men better at asking raises and competing? Would the same strategies even work for women?
Individuals might steer themselves, too, while not even thinking about it that much. Young heterosexual men with traditional values or expectations might expect to have to earn enough to provide the bulk of income support for a family one day, with some "junior partner" type financial support from the wife who will do the bulk of childcare and cleaning etc. Young heterosexual women with traditional values or expectations might expect to need extra flexibility one day, to do the hands-on caring for those future children, and that could lead them into picking flexible careers, even if they are not that well paid (3).
I can make dozens of those kinds of mini-theories, in various directions, but they are not statistical evidence.
One promising way to start teasing the knot apart is by asking what happens when an occupation turns from predominantly male to predominantly female or vice versa.
Why? Consider this: Say that an occupation is currently predominantly male and pays pretty well. Then lots of women enter it. They can't have entered it because it's one of those flexible-but-low-pay jobs which fit well with parenting: It isn't one at that point.
But it can become one, at least a less well-paid one, and it looks like the cause is the influx of more female workers (4). The New York Times article quotes the authors of one study on what happens when women start entering an occupation in larger numbers:
A striking example is to be found in the field of recreation — working in parks or leading camps — which went from predominantly male to female from 1950 to 2000. Median hourly wages in this field declined 57 percentage points, accounting for the change in the value of the dollar, according to Professor Levanon. The job of ticket agent also went from mainly male to female during this period, and wages dropped 43 percentage points.It would be fascinating if someone created a very detailed study of the changes in just one occupation over time. Maybe such a study exists, but I haven't come across one. I'd like to know the exact point in the tipping process at which average earnings start to drop, I'd like to know the exact point when men start avoiding the field that is turning female-dominated, I'd like to know everything about technological change in the field during the process.
The same thing happened when women in large numbers became designers (wages fell 34 percentage points), housekeepers (wages fell 21 percentage points) and biologists (wages fell 18 percentage points). The reverse was true when a job attracted more men. Computer programming, for instance, used to be a relatively menial role done by women. But when male programmers began to outnumber female ones, the job began paying more and gained prestige.
I'd also love to understand why some occupations don't seem to result in sex segregation, and how those differ from the ones that are prone to tipping (moving from male-dominated to female-dominated or vice versa).
I hope this isn't quite the mess it looks to me. To return to the title of the post: It was a sneaky one, because I forced you into accepting a false either-or set of choices. Both could of course be true at the same time. But if the New York Times summary of the studies is correct, we should seriously address why jobs become lower-paid when more women enter them (5).
(1) Occupational segregation by sex refers to the simple fact that some industries or occupations have mainly male workers, some industries or occupations have mainly female workers. It says nothing about the reasons why the industries got that way, it says nothing about how voluntary or involuntary those gendered choices are and it says nothing about the reasons why female-dominated jobs pay less. All that is a separate question.
(2) I made up that second job title, but you get the point. Which is that it's possible to pay someone less by just changing labels.
(3) The less-well-paid female dominated jobs are not necessarily more flexible or more compatible with major responsibilities for child-care; an important point which is often ignored by those who argue that women "choose" lower paying occupations for their greater flexibility. Hospital nursing is not very flexible and neither are female dominated factory or sales jobs. Teaching is the one exception to that rule. It's also worth noting that studies suggest men are paid more, on average, even in female-dominated fields and women paid less, on average, even in male-dominated fields.
(4) Or reverse the same example for an occupation that rapidly goes from at least mixed if not predominantly female to male. The initial reason for the larger numbers of men entering the field cannot be higher earnings in that occupation in general, if the underlying theory is to hold.
(5) The quotes around the verb to choose are explained in this post.