Eric Crampton Has an Answer

I really should have written this post earlier, but just when it was about time to write it on Wednesday evening, I suddenly felt an irresistible urge to vomit like it was becoming illegal tomorrow. At first I thought I'd eaten something wrong, but the next day my doctor opined that I had the same flu that everybody in this city has this month. Based on my colleagues experiences, that should have lasted for about a week, but about seven hours ago I woke up and felt much improved. So either my doctor opined incorrectly or my immune system - or The Army of Steel as I like to call it - decided that if Olympique Lyonnais can do the seemingly impossible, so can it.

Anyway -

Arguably one of the more excellent features of blogs is that it allows you to have conversations with people around the world, all of which come with their own brands of perspectives, opinions and expertise. I recently decided to exploit that feature by asking Eric Crampton, an economist in New Zealand whose blog is Offsetting Behaviour, a question that had been sitting at the back of my mind for quite a while. The context is differences in pay between men and women; here's my comment (including typo):

I have read, I think, a grand total of two academic papers on this subject, but here's somehting I have been wondering for a while; maybe you as an economist can point out where my reasoning is wrong.

Given that many countries mandate that an employer keeps paying a portion of an employee's salary when she goes on leave due to pregnancy, shouldn't we expect a male-female wage gap for people below 40 even after controlling for all of the other stuff (because the expected utility of a female employee is less, all other things equal)? And if we don't find such a gap, wouldn't that suggest discrimination against men?

What about countries in which the government mandates no such thing?

The context, as I didn't mention, is that I may have read only 2 academic papers on the subject, but a shitload of articles in the press, and the rather obvious-seeming point that women can get pregnant, to the detriment of the employer, never gets mentioned - so maybe there was an error in my reasoning?

Eric Crampton very kindly and quickly anwered with what feels like a full-blown lecture. For one thing, it turns out my reasoning was not incorrect - but do read the link for lots, lots more.

To be sure, paying women less because they might get pregnant is discrimination - a form of what social scientists call statistical discrimination - but it's the kind of discrimination that has nothing to do with a dislike for the group in question.