There is absolutely nothing new in the IFS gender pay gap paper

We all know that, on average if you count all men and women, men take home more money in pay than women. We also know that, on average, people who work sixty hours a week take home more than those who work thirty hours a week. And those in banking and law take home more than those in psychology and education.

The Institute for Fiscal Studies has a paper out today (pdf), which has been widely feted across the press, despite the fact that it simply repackages these facts with barely any extra analysis or information, though they do have some cool charts and graphs. I suspect they didn't tell the media that their data is no better than—and in my view a good deal inferior to the official Office for National Statistics numbers.

In their paper they show that there is indeed a raw gender pay gap, which is reduced substantially by various controls (like hours and having kids). They cut this up and display the data in some pretty cool ways below. But they don't mention that some pretty important facts have been left out of their narrative.

For example: do we think that all young men and women without children choose, on average, the same A-levels choices, degrees, and career paths? I'll answer the question: they don't. In the USA, 95% of the gender pay gap among new graduates is down to subject major choice—some degrees are more lucrative than others.

This, in turn, seems to be because men are more ruthlessly materialistic in major and career choice. And it reflects values men and women express in long, careful studies. Gender difference (whether socially constructed or genetically inborn, or most likely, a bit of both) predicts the world we see much better than discrimination.

When women do go into high earning, competitive fields, they do very well. In fact, one study found that women were more aggressively promoted, and earned more than men, so long as they didn't exit the workforce.

In fact, Harvard economist Claudia Goldin found (pdf) that the gender pay gap only existed in industries where there are increasing returns to hours: where 80 hours a week is worth more than double 40 hours a week, and where flexible hours are impossible for technical reasons. In these fields women were less productive because they chose to have a work-life balance that men avoided (and not just because of kids). Since they were less productive, firms paid them less.

Is this a bad thing. Well, it depends whether you think money and career success are the only important ends in life, and it depends whether you think that men and women have to have (on average) the same paths in life to have equally good paths in life. I don't think it's worse to work thirty flexible hours a week than it is to work six ten-hour days, and I think it would be dystopian to try and make everyone's careers similar.

What we want is equal work for equal pay, and according to our best analyses of the gender pay gap data, that's what we have. Today's IFS paper adds nothing to our understanding.