Oh, I dunno. I don’t think it’s a particularly sensible thing to try to evaluate a president on after four years. These things change slowly, and their effects propagate slowly. We could maybe measure, e.g., whether the Trump administration itself is biased against women when hiring, but gender equality in society at large? There’s not that much Trump can do about it, and any effects he might have will take a long time to show up and be very difficult to disentangle from other causes.
But the sort of thing you could do is: take a selection of jobs; look at men and women in those jobs and try to pair up men and women who are comparable in various plausbly-relevant respects (e.g., similar IQ, similar number of years’ experience, same hours worked, etc.), and compare their pay. The matching process should minimize the impact of any actual differences in ability (note: this might not work well for very “extreme” cases, so e.g. I wouldn’t recommend it for assessing putative biases in theoretical physics, but it should be OK elsewhere). It’s important to control for full-time versus part-time because it’s known that women are much more likely to work part time. (That could itself be the result of discrimination of some kind, but best to leave that aside for now.)
A small difference wouldn’t necessarily indicate anything bad; there are all sorts of non-prejudicial mechanisms that could make men’s and women’s pay not come out exactly equal. But big differences would be cause for concern, and if the differences are not small then movement towards equality would probably indicate reduction in bias.
That addresses just one aspect of gender equality, of course. Some others would be really hard to measure. (E.g., women sometimes complain that they aren’t taken as seriously as men; e.g., they say something and get ignored and then a man in the same meeting repeats the exact same thing they said and everyone listens. Maybe this is a real effect, but I wouldn’t want to try to measure it accurately—but if it is real, it could be a big deal.)
Although I don’t think ArisC had inequalities going “the other way” in mind, I think some of them would be good to measure and to try (with the usual caveats about small differences) to get more equal. For instance, men don’t live as long as women; it would be great to make male lifespans more like female ones. But most likely this is mostly biological and improving it would be a medical, not a social, problem. For another instance, James alluded to the gap in college attendance: more boys than girls leave the educational system early. That might be the result of some sort of social messed-up-ness (e.g., it might come from overcorrecting for discrimination against girls), and if so it would be good to fix it.
Oh, I dunno. I don’t think it’s a particularly sensible thing to try to evaluate a president on after four years. These things change slowly, and their effects propagate slowly. We could maybe measure, e.g., whether the Trump administration itself is biased against women when hiring, but gender equality in society at large? There’s not that much Trump can do about it, and any effects he might have will take a long time to show up and be very difficult to disentangle from other causes.
But the sort of thing you could do is: take a selection of jobs; look at men and women in those jobs and try to pair up men and women who are comparable in various plausbly-relevant respects (e.g., similar IQ, similar number of years’ experience, same hours worked, etc.), and compare their pay. The matching process should minimize the impact of any actual differences in ability (note: this might not work well for very “extreme” cases, so e.g. I wouldn’t recommend it for assessing putative biases in theoretical physics, but it should be OK elsewhere). It’s important to control for full-time versus part-time because it’s known that women are much more likely to work part time. (That could itself be the result of discrimination of some kind, but best to leave that aside for now.)
A small difference wouldn’t necessarily indicate anything bad; there are all sorts of non-prejudicial mechanisms that could make men’s and women’s pay not come out exactly equal. But big differences would be cause for concern, and if the differences are not small then movement towards equality would probably indicate reduction in bias.
That addresses just one aspect of gender equality, of course. Some others would be really hard to measure. (E.g., women sometimes complain that they aren’t taken as seriously as men; e.g., they say something and get ignored and then a man in the same meeting repeats the exact same thing they said and everyone listens. Maybe this is a real effect, but I wouldn’t want to try to measure it accurately—but if it is real, it could be a big deal.)
Although I don’t think ArisC had inequalities going “the other way” in mind, I think some of them would be good to measure and to try (with the usual caveats about small differences) to get more equal. For instance, men don’t live as long as women; it would be great to make male lifespans more like female ones. But most likely this is mostly biological and improving it would be a medical, not a social, problem. For another instance, James alluded to the gap in college attendance: more boys than girls leave the educational system early. That might be the result of some sort of social messed-up-ness (e.g., it might come from overcorrecting for discrimination against girls), and if so it would be good to fix it.