4 ms·

This is an insanely overconfident article, given that the field in question has dozens of confounding variables that likely can't be reduced to a pat conclusion

This is an insanely overconfident article, given that the field in question has dozens of confounding variables that likely can't be reduced to a pat conclusion+prescription. I guess that makes it par for the course for Tabarrok. He reminds me of all my econ friends at unnamed Ivy League school who learned linear regressions freshman year and then started applying them to everything.

paulpauper 7y ago
agree. marginal revolution has really gone downhill in recent years
ddxxdd 7y ago
One way to easily corroborate this hypothesis is to check the acceptance rates of men and women that apply for STEM fields. This reminds me of the UC Berkeley lawsuit, where it was found that women were more likely to be accepted into almost any department, and that women were less likely to be accepted overall because they overwhelmingly applied to more competitive departments. The numbers are here: https://en.wikipedia.org/wiki/Simpson%27s_paradox#UC_Berkeley_gender_bias https://en.wikipedia.org/wiki/Simpson%27s_paradox#UC_Berkele... . The wiki article doesn't list the names of each department, but it's worth noting that every single department with a majority of men (think: STEM) is more likely to accept any given woman than any given man.
timkam 7y ago
Indeed. The "largely" in the article's headline is a blatant lie. It cannot largely explain, it probably just largely correlates, considering the samples at hand. One could write tens of similarly attention-catching articles with other variables "explaining" the issue. The article's conclusion contains a call to action: "Stop telling people to do what they are good at and instead tell them to do what pays!" Now, there is actually evidence that women in countries where there are strong socio-economic incentives to study STEM (Iran, India) are more likely to do so than women in western countries. Not that this explains the whole picture, either, but it's certainly interesting.

paulpauper 7y ago

agree. marginal revolution has really gone downhill in recent years

ddxxdd 7y ago

One way to easily corroborate this hypothesis is to check the acceptance rates of men and women that apply for STEM fields. This reminds me of the UC Berkeley lawsuit, where it was found that women were more likely to be accepted into almost any department, and that women were less likely to be accepted overall because they overwhelmingly applied to more competitive departments. The numbers are here: https://en.wikipedia.org/wiki/Simpson%27s_paradox#UC_Berkeley_gender_bias https://en.wikipedia.org/wiki/Simpson%27s_paradox#UC_Berkele... . The wiki article doesn't list the names of each department, but it's worth noting that every single department with a majority of men (think: STEM) is more likely to accept any given woman than any given man.

timkam 7y ago

Indeed. The "largely" in the article's headline is a blatant lie. It cannot largely explain, it probably just largely correlates, considering the samples at hand. One could write tens of similarly attention-catching articles with other variables "explaining" the issue. The article's conclusion contains a call to action: "Stop telling people to do what they are good at and instead tell them to do what pays!" Now, there is actually evidence that women in countries where there are strong socio-economic incentives to study STEM (Iran, India) are more likely to do so than women in western countries. Not that this explains the whole picture, either, but it's certainly interesting.