> > Furthermore, no one's arguing because studies show women to be less effective negotiators than men that we should give up.
> I'm sure a lot of people would argue that, but since neither I nor you nor Damore seem to argue that, I agree with the connotation.
I only mean that Damore's argument is (roughly, mind you) "studies show the gender gap is likely due to biological differences so we should give up", and if we're comparing Google's pro-diversity hiring initiatives to Stretch, it's important to note that when Google noted the research on women and negotiating, their response wasn't "oh it's biological differences, we should give up". I don't know if Stretch is effective, but at least it's a proactive, supportive response rooted in research.
> There seems to be a lot of science on the benefits of pair programming... I don't know about part time work and making work less stressful, but they don't seem like universally bad ideas either.
I don't think they're universally bad ideas, but Google's gender gap is something like 70-30. There's no research to support the notion that pair programming, part time work and low stress jobs can address a 40 point spread like that, but there is research that pro-diversity hiring and support policies do, so I think it's actively harmful to advocate for replacing the latter with the former.
> I'm definitely arguing for fairness of outcome here. If you observe that some people are worse negotiators than others, then to achieve fairness of outcome, you have to offer them help. ... Helping only women is better than nothing, but it is not optimal, because you are adjusting the wrong variable.
Sure that makes sense, but the goal isn't to get every employee's negotiating skill to a certain level, it's to narrow the gender pay gap. In that context, it makes sense to work only with women.
> If someone is already taking their status into account, sure, you need to take that into account to counteract their biases. But that's a kludge and hard to balance correctly, if you can instead remove the influence of that person altogether, you should do that.
It is really hard to quantify, definitely. But these issues aren't limited to "that person"; we're all, every single one of us, subject to unconscious bias when it comes to race, gender identification, sexual orientation, and other ascribed statuses because of the culture and society we grew up in. Therefore we all need to adjust, and pro-diversity policies and affirmative action policies help us do that.
> The article is about racial bias and not the gender gap, the alternative is not completely race-blind, and it doesn't show any skyrocketing. In fact, the effect is much weaker than I'd have expected. The situation for Hispanics looks more like noise. Maybe there aren't many affirmative action programs for Hispanics even in states that allow them?
Sorry "skyrocketing" was a poor characterization (it was laaaaaaate :) Here's what 538 says about Black enrollment:
"...only two research universities in states with affirmative action bans have at least the same proportion of black students as the state’s college-age population, and one of those, Florida A&M University, is a historically black college or university (HBCU). ...only one school, Florida International University, has at least the same proportion of Hispanic students as the state’s college-age population.
...
Researchers looked at the effect race had on admissions and saw a 23 percentage point drop in the chance of admission for minority students in states with bans, compared with a 1 percentage point drop in other states, relative to nonminority students."
That's rough, no matter how you look at it.
> Personally, I think that affirmative action in college admissions shouldn't be based on race either. As I understand it, most racial differences in the distribution of applicants are due to economic reasons.
538 addresses this too:
"Opponents of affirmative action argue that aiming for diversity in areas other than race, such as socioeconomic class, can ensure sufficiently diverse student bodies. The most common race-neutral policy used as an alternative to affirmative action is a plan that the University of Texas already uses, in which a percentage of graduates from every high school get automatic admission. These policies have been shown to increase racial and ethnic diversity on campus, but research[1] on whether they’re as effective[2] as more explicit race-based affirmative action policies has been mixed[3], and critics say that it doesn’t make sense to use a proxy when so many colleges continue to struggle with racial diversity."
There are similarities between the experience of lower income Americans and Americans of color, but not all Americans of color are lower income, and policies that focus on evening out the income divide overlook the disadvantages people of color face because of their race.
> I agree that biases shouldn't be ignored, but I don't like it when the countermeasures assume that disadvantages only happen across a few categorizations. There are all kinds of reasons some people have worse outcomes than others, and to only pay attention to them when they coincide with membership in one of your favorite protected groups, is a kind of bias in itself.
Agreed, but at the same time, I don't think we need to show up at every discussion about gender issues and remind everyone that men also face challenges. We can advocate for policies that help straight cis White men who may be disadvantaged for other reasons without derailing discussions about race and gender issues.
> I'm sure a lot of people would argue that, but since neither I nor you nor Damore seem to argue that, I agree with the connotation.
I only mean that Damore's argument is (roughly, mind you) "studies show the gender gap is likely due to biological differences so we should give up", and if we're comparing Google's pro-diversity hiring initiatives to Stretch, it's important to note that when Google noted the research on women and negotiating, their response wasn't "oh it's biological differences, we should give up". I don't know if Stretch is effective, but at least it's a proactive, supportive response rooted in research.
> There seems to be a lot of science on the benefits of pair programming... I don't know about part time work and making work less stressful, but they don't seem like universally bad ideas either.
I don't think they're universally bad ideas, but Google's gender gap is something like 70-30. There's no research to support the notion that pair programming, part time work and low stress jobs can address a 40 point spread like that, but there is research that pro-diversity hiring and support policies do, so I think it's actively harmful to advocate for replacing the latter with the former.
> I'm definitely arguing for fairness of outcome here. If you observe that some people are worse negotiators than others, then to achieve fairness of outcome, you have to offer them help. ... Helping only women is better than nothing, but it is not optimal, because you are adjusting the wrong variable.
Sure that makes sense, but the goal isn't to get every employee's negotiating skill to a certain level, it's to narrow the gender pay gap. In that context, it makes sense to work only with women.
> If someone is already taking their status into account, sure, you need to take that into account to counteract their biases. But that's a kludge and hard to balance correctly, if you can instead remove the influence of that person altogether, you should do that.
It is really hard to quantify, definitely. But these issues aren't limited to "that person"; we're all, every single one of us, subject to unconscious bias when it comes to race, gender identification, sexual orientation, and other ascribed statuses because of the culture and society we grew up in. Therefore we all need to adjust, and pro-diversity policies and affirmative action policies help us do that.
> The article is about racial bias and not the gender gap, the alternative is not completely race-blind, and it doesn't show any skyrocketing. In fact, the effect is much weaker than I'd have expected. The situation for Hispanics looks more like noise. Maybe there aren't many affirmative action programs for Hispanics even in states that allow them?
Sorry "skyrocketing" was a poor characterization (it was laaaaaaate :) Here's what 538 says about Black enrollment:
"...only two research universities in states with affirmative action bans have at least the same proportion of black students as the state’s college-age population, and one of those, Florida A&M University, is a historically black college or university (HBCU). ...only one school, Florida International University, has at least the same proportion of Hispanic students as the state’s college-age population.
...
Researchers looked at the effect race had on admissions and saw a 23 percentage point drop in the chance of admission for minority students in states with bans, compared with a 1 percentage point drop in other states, relative to nonminority students."
That's rough, no matter how you look at it.
> Personally, I think that affirmative action in college admissions shouldn't be based on race either. As I understand it, most racial differences in the distribution of applicants are due to economic reasons.
538 addresses this too:
"Opponents of affirmative action argue that aiming for diversity in areas other than race, such as socioeconomic class, can ensure sufficiently diverse student bodies. The most common race-neutral policy used as an alternative to affirmative action is a plan that the University of Texas already uses, in which a percentage of graduates from every high school get automatic admission. These policies have been shown to increase racial and ethnic diversity on campus, but research[1] on whether they’re as effective[2] as more explicit race-based affirmative action policies has been mixed[3], and critics say that it doesn’t make sense to use a proxy when so many colleges continue to struggle with racial diversity."
There are similarities between the experience of lower income Americans and Americans of color, but not all Americans of color are lower income, and policies that focus on evening out the income divide overlook the disadvantages people of color face because of their race.
> I agree that biases shouldn't be ignored, but I don't like it when the countermeasures assume that disadvantages only happen across a few categorizations. There are all kinds of reasons some people have worse outcomes than others, and to only pay attention to them when they coincide with membership in one of your favorite protected groups, is a kind of bias in itself.
Agreed, but at the same time, I don't think we need to show up at every discussion about gender issues and remind everyone that men also face challenges. We can advocate for policies that help straight cis White men who may be disadvantaged for other reasons without derailing discussions about race and gender issues.
[1]: http://www.mitpressjournals.org/doi/abs/10.1162/003465304312...
[2]: http://onlinelibrary.wiley.com/doi/10.1002/pam.21800/abstrac...
[3]: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2137126