The gender pay gap does not exist (updated)

David Green blogs, at the Telegraph.co.uk, on the controversial topic that The gender pay gap does not exist. Green writes in response to Harriet Harman's, the UK Minister for Women and Equality, claims that women in the UK earn on average 22.6% less per hour than men and assumes that this difference is the result of discrimination against women by men. Green claims that the UK Government's own figures do not support such a conclusion. Green writes

Three official surveys are used by the Office for National Statistics (ONS): a survey of employers called ASHE (the annual survey of hours and earnings), a survey of households (the Labour Force Survey) and the panel dataset of the New Earnings Survey, which provides information from 1975 to 2006. [...]

But as the ONS said in its monthly Economic & Labour Market Review, published last year just before the Government committed itself to forcing the Equality Bill through, the gender pay gap 'does not necessarily indicate differences in rates of pay for comparable jobs'.

He goes on to say

According to ASHE, in 2007 a gender pay gap does not open up until women reach about 30 years of age. From ages 18-29 there is hardly any difference and, according to the Labour Force Survey (LFS), women aged 22-29 are paid on average slightly more per hour than men. As the ONS concludes, having children is the decisive factor, not being a woman. Historical data confirm this conclusion. Based on the New Earnings Survey panel data, in 1975 there was a pay gap from the age of 18 onwards, but in 2006 no such gap existed until age 34. Why? In 1975 women tended to have children in their 20s and by 2006 it was more common to have them in their 30s. As the average age of child-rearing increased so too did the age at which the pay gap kicked in.

Then there are some other disparities, which are very hard to explain as the result of discrimination. Overall the hourly rate of pay is higher for men, but not when they work from 10-20 hours per week or from 20-30 hours per week, when on average women earn slightly more. If discrimination is the explanation, why would (male) employers discriminate against women who work more than 30 hours a week but not those who work fewer hours? Moreover, women who work part-time earn more than men in companies with 500 or more employees but less than men in companies with fewer then 25 employees.

Green argues that what the data tends to suggest is that the vital difference is not between men and women but between women with dependent children and those without, whether they be male or female. Green notes,

The hourly rate of pay for women who are neither married nor cohabiting is slightly higher than for men in the same situation. For men and women who are married or cohabiting the hourly pay gap is 14.5% and the gap widens with the number of children. Women with one dependent child earn on average 12.3% less than men and with four or more dependent children 35.5% less.

Thus the UK government's emphasis on the gender pay gap of 22.6% looks more like an an abuse of official statistics, than a certain fact, and to infer that the difference in the average hourly rate is the result of discrimination is an abuse of logic. But given how often we find governments abusing both statistics and logic, this is no surprise.

Update: Thanks to Jim Donovan in the comments I was alerted to this piece from Tim Harford

Katz and Goldin, with Marianne Bertrand of Chicago’s Booth School of Business, have now produced a new study, examining the experience of Booth’s MBA alumni – a high-flying group from whose ranks one would expect future CEOs to emerge. The outstanding feature of this research is the very detailed data available on this group: their pre-MBA experience, the courses they took and the grades they earned, their career progression afterwards, and the timing of their families. Women did achieve worse grades, and avoided hardcore classes in finance: but the differences were tiny. Far more important was what happened when children came along. If you look only at promotions and earnings, childless women are all but indistinguishable from men. The moment children arrive on the scene, a big gap opens up.

“The penalty for career interruptions is huge,” Bertrand told me in a recent interview. New mothers are derailed from the fast track in investment banking or consulting, and their potential earnings fall by about 40 per cent. The gap is aggravated by the fact that many of these women are married to men so rich that they decide to drop out of the labour force altogether. (Emphasis added.)