I'm supposed to take seriously a man who tells people what to do based on what lobsters do, and gets even that wrong? What are these empirically correct statements which he's supposed to have made? Honestly curious.
-
-
Helen, that's simply not true. Please read the statistics textbooks I linked earlier, or the Blau and Kahn review, which says the same thing:pic.twitter.com/hjEfpmLpF6
-
It's not true that the wage gap is only a problem if it isn't chosen freely?
-
You are confusing moral and statistical issues. The wage gap isn't a problem if it's chosen freely, but the regression model you are advocating for does not demonstrate it is chosen freely
-
Research-active economist here. The models used to analyze the wage gap cannot tell whether sex differences in occupation are due to women being shut out of some professions or due to women's preferences. Other evidence is needed to distinguish between those hypotheses.1/
-
IMHO, the Scandinavian data on occupational choice, with which Helen is familiar, provide strong evidence that the preference hypothesis explains most sex differences in occupation. But, as is typical in economics, we can't completely rule out the other hypothesis. 2/
End of conversation
New conversation -
-
-
People suggest that women work more or men work less to get rid of the pay gap but the pay gap itself is only a problem if men or women are limited in their choices..https://www.theguardian.com/world/2018/may/10/help-men-work-less-to-close-gender-pay-gap-says-thinktank …
-
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.
- 9 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.