In Forbes, Tim Worstall looks at how the World Economic Forum came up with their scary conclusions that the pay gap between men and women won’t disappear until 2095:
And that’s it: no, really, that is what they’re basing, in its entirety, their estimations of the gender pay gap upon. They asked a few people whether they thought that men and women got roughly the same pay for roughly the same sort of job or not and that’s it. This isn’t cutting edge data science to put it very kindly indeed.
For when we go off and look at the messy details of the gender pay gap we find that we’ve not really got one, not in the industrialised countries. Once we correct for the obvious things like hours at work, years in the workforce, educational background and so on we find that the mythical gender pay gap (that “women earn 77 cents to every $ men do”) simply disappears. There might be a small residual, a few percent, left in there but not enough that we can really notice. And quite apart from anything else it’s actually illegal to pay men and women different amounts for doing the same job (if on the basis that the different pay is purely as a result of their being men or women that is).
So, no, we shouldn’t be taking this report or finding seriously. And there’s more than just the fact that they’re using a survey to measure that gap. For of course the printing of this report will lead to, as the other incorrect claims about the gender pay gap do, a certain circularity of reasoning.
Ask someone: “Are men and women paid equally?” And they’ll start thinking about whoever it was that said that 77 cents line, recall that last year the WEF said that gender pay inequality was very bad indeed. So, now we come to asking them the same question for the next WEF survey and their answer will be influenced by the cacophony of voices that have been telling them how bad the gender pay disparity is. Including, obviously, last year’s WEF report that said so. It’s entirely circular and self-reinforcing.
Really, we shouldn’t be taking this stuff seriously.
