It's inaccurate to claim the man has made no correct statements, or no empirical statements. And it's naive and foolish to consider him stupid, as many people do. I don't know about "empirically correct", but his discussion of the gender pay gap is certainly empirically grounded.
-
-
Yes, and if their choices are limited then we can not include them as statistical controls. Imagine controlling for occupation in the 1950s. You would find no wage gap. It's a meaningless regression.
-
The fact that you’d have to “go back to the 1950’s” rather makes the point. You couldn’t validly “correct for profession” when comparing pay by gender in the ‘50s because entire fields were closed to women. But that’s no longer the case.
-
Yes, but I'm making a point about regressions. You can't claim an absence of discrimination on the basis of a regression that would show the same thing under circumstances where there is discrimination.
-
It’s very doable. You can validly compare the rate of pay of, say, respiratory therapists in Houston, right? And nurse anesthetists in Cleveland? And junior accountants in Miami? And we can then see whether, there is a gender gap, after correcting for hours, seniority, etc.
-
John, you can't do that without invalidating your analysis. Including those occupational controls incorporates a selection bias into your regression.
-
Ha! There it is. The classic untestable hypothesis! “You can’t disprove my claim of a persistent, discrimination-based gender gap because discrimination is so pervasive that the effect can’t be measured.” It’s brilliant.
-
No, you can test it directly with experiments.
-
An experiment? Like in a lab? Silly me. I thought the way to determine whether there’s a gender gap in the pay of nurse anesthetists in Cleveland is to look at the pay of nurse anesthetists in Cleveland.
- 8 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.