Wednesday, August 15, 2018

The Gross Gender Earnings Gap Revisited

I have written about the gross and net earnings gaps* many, many times.  It's the latter which we want to use when looking for possible sex- or race-based (or both) discrimination in the labor market.

That's because the net earnings gap gets as close as observational data can to the ideal where we would compare two individuals, identical in all other characteristics except the one or the ones we study (say, sex, race, age etc.).

It's not exactly the same as that theoretical ideal, but it's loads better than the gross earnings gap which does not take into account differences in education, experience, and other things which affect how much people earn, on average.  In some forms it doesn't even take into account differences in hours, days or weeks worked.

But the net earnings gap does that.  More about it in this post and here.   And that's why the feminist left should not use the gross gap in earnings as the measure of labor market discrimination.

But the anti-feminist right does something worse when it argues that studies have conclusively shown that there is no gender gap in earnings after women's own free-and-voluntary choices are taken into account.   Good studies have shown no such thing, if they sample a proper cross-section of workers.

The studies the right decides to focus on are only about very young workers, just beginning their working lives or careers. Given that the Equal Pay Act of 1963 makes it illegal to pay men and women different wages for exactly the same work, and given that Title VII of the Civil Rights Act makes discrimination in hiring illegal, it's hard** for employers to discriminate against any particular group among brand-new workers.

It's the passage of time*** which allows any possible discriminatory motives to be satisfied.  For example, men and women can be promoted at different rates, irrespective of their productivity and other characteristics, or on-the-job training can be offered to one group but not to the other group etc.

This post discusses one study that has been very popular among the anti-feminists.  You probably notice by now that because it fails to standardize for education it's comparing apples to oranges, even if the results were only applied to the group of young, single, metropolitan workers.  But they have been applied much more widely than that.


* The gross gap between, say, women and men would be calculated by dividing the earnings of an average woman by the earnings of an average man, preferably both expressed in earnings per hour, week and so on.

The net gap corrects the gross gap for hours worked, education, experience and other relevant factors which affect how much someone works and are mostly viewed as non-discriminatory.  The net gap is what still remains after those calculations are carried out.

**  Not impossible, given that earnings information is usually kept hidden in the US, but harder than it would be without those laws. As an aside, the UK right spreads the same propaganda about the wage gap not existing after women's "voluntary" choices are taken into account.

***   The right-wing argument is that the earnings differences accruing over time are due to women's choices to focus more on their children and less on their careers.  But note that we cannot conclude anything of the sort from the data, because discriminatory effects also accrue over time.