The gender pay gap is not a myth, it's math

But it's mostly not a discrimination story, it's a parenting story

As I’m out and about talking about boys and men, I’m often asked a question along the lines of “well, why is there still such a big gender pay gap, then?” It’s a good question. And I provide what I think it a pretty good answer, on pages 23-29 of my book Of Boys and Men. (Did I mention I’d written a book?)

If you’re more of a video kind of person, do check out my latest for Big Think, an explainer on the real causes of the gender pay gap:

But let me set out the basic argument here too.

Mind the gap instinct

When I hire a new research assistant, I ask them to read two books. The first is How to Write Short: Word Craft for Fast Times by Roy Peter Clark, an excellent guide to sharp communication in a world of blogs and tweets. The other is Factfulness: Ten Reasons We’re Wrong about the World— and Why Things Are Better Than You Think by Hans Rosling, who is something of a hero to me. Rosling, who died in 2017, was a Swedish physician who became obsessed with statistical illiteracy. In Factfulness, he describes various biases, including the “straight line instinct,” an assumption that a historical trend line will continue unaltered into the future; the “negativity instinct,” which is a tendency to think things are likely getting worse; and the “gap instinct,” which is a “basic urge to divide things into two distinct groups, with nothing but an empty gap in between.”

As Rosling puts it:

We love to dichotomize.

The gap instinct leads to two errors of perception. First, we fail to see how much overlap there is between two groups. Second, we fail to see the bigger gaps that typically exist within groups, rather than between them. The gender pay gap is a case in point. A woman at the middle of the female wage distribution (for full-time, year-round workers) earns 82% as much as a man at the middle of the male one.

When we hear about this gap, the thought that naturally gets generated is “women earn less than men.” But in fact, the distribution of women’s wages looks strikingly similar to the distribution of men’s wages, and a lot more similar today than just a few decades ago. The figure below, from the book, shows the wage distribution for men and women (working full time) in 1979 and in 2019:

As you can see, the distributions now overlap much more tightly. In fact, 40% of women now earn more than the typical man, up from just 13% in 1979. The closeness of the male and female wage distributions is of course stupendously good news on the gender equality front.

But, still, there’s a gap

So, what is causing the remaining gap? The answer to this question matters a lot, especially when it comes to potential solutions.

The basic facts are not in dispute. As I have already said, the typical (i.e., median) full-time female worker earns about 82% as much as the typical man. The real disagreement is not over whether the typical woman earns less than the typical man, but why.

There is certainly little evidence that women are paid less than men for doing the same work in the same way. Women are paid less because they do different work, or work differently, or both. But, of course, that is not the end of the story. Women may earn less because they occupy fewer senior positions, but that fact itself may be the result of institutional sexism. Similarly, it is true that women tend to be more clustered than men in lower-paying occupations and industries, which explains perhaps a third of the pay gap. But that may reflect socialized gender roles, not least in terms of family responsibilities, or a devaluation of work that is done by women, or both. In any case, while there is a pay gap between occupations, there is as big a gender pay gap within occupations.

The pay gap is a parenting gap

The one-word explanation for the pay gap is: children. Among young adults, especially if they are childless, the pay gap has essentially disappeared. “There’s remarkable evidence that earnings for men and women move in sync up until the birth of a couple’s first child,” says economist Marianne Bertrand. “This is when women lose and they never recover.” Here’s a couple of charts from Henrik Kleven and colleagues that makes the point rather dramatically:

The earnings trajectory for women who do not have children looks similar to that for men. The one for mothers does not. The more children women have, the further behind they fall in terms of both employment and earnings. But for men, if anything becoming a father increases wages and hours worked.

Some of the best proof that the gender pay gap is mostly a parenting pay gap comes from innovative studies in Sweden and Norway comparing new mothers in same-sex relationships with those in heterosexual relationships. Ylva Moberg shows that the impact on earnings for the birth mother is almost identical in both family types. Meanwhile, the nonbirth mothers in the lesbian couples show a similar earnings pattern to fathers in the heterosexual ones. Over time, the inequality seems to balance out in the lesbian couple if they have more than one child, as each takes their turn at being the birth mother. For heterosexual couples, by contrast, the gap gets wider with each child.

A study of bus and train drivers working for the Massachusetts Bay Transportation Authority (MBTA), by Valentin Bolotnyy and Natalia Emanuel, a duo of Harvard economists, provides some strong evidence here too. Women account for 30% of the drivers, and on average earn $0.89 for every dollar earned by their male peers. By focusing on men and women doing the same job for the same employer, Bolotnyy and Emanuel can tease out the various factors contributing to wage differ- ences. They conclude that:

The pay gap can be explained entirely by the fact that, while having the same choice sets in the workplace, women and men make different choices.

The men were twice as likely to work overtime (which pays extra), even at short notice. They also took fewer hours of unpaid leave, and so on. Among train drivers with children, the gaps were even wider. Fathers wanted even more overtime pay; mothers wanted more time off.

For most women, having a child is the economic equivalent of being hit by a meteorite. For most men, it barely makes a dent.

Ideology trumps the facts

As with so many issues, the gender gap gap is looked at through existing ideological lens rather than empirically. Conservatives insist that the pay gap is a “myth”, in the sense that it is largely explained by factors like occupation and parenting. Almost half of men in the U.S. (46%) now agree that the pay gap “is made up to serve a political purpose,” rather than being a “legitimate issue”.

On the other side, progressives insist that the pay gap is the result of sexist discrimination. The impression is given that the cause of the gender wage gap is that women are paid less for doing the same or similar work as men.

This is not the case, as every labor market economist knows. But strikingly, people and especially women with higher someone’s level of education are more likely to think that direct discrimination is the main cause of the gender pay gap. (To my frustration, even though I can literally see the numbers in my head, I think from Gallup, I can’t lay my hands on the survey. So for now you’ll have to trust me on this one, or even better help me find the data).

That’s clearly an ideological effect, and a useful reminder that educational level is not a reliable predictor of the accuracy of a person’s views on a particular topic. Especially if it a culture-war topic.

Progress will take more work now

None of this should be read as suggesting that there’s less to do on the gender pay gap. On the contrary: the fact that the gap is now the result of deeper factors like family and occupation, of structure rather than sexism, means that we have to work harder still to make progress.

That means reducing the sharpness of the trade off between raising a family and pursuing a career (paging Claudia Goldin), as well as increasing the share of childrearing undertaken by men.

More paid leave, anyone?