4 ms·

> Aren't these the real reasons why people are disadvantaged? No, or at least, not exclusively. There's tons of social science research in which, for example,

> Aren't these the real reasons why people are disadvantaged? No, or at least, not exclusively. There's tons of social science research in which, for example, two identical fabricated resumes were submitted for a battery of jobs, differing only by, say, a headshot of a black vs. white person, or a stereotypically-African-American name vs. a stereotypically white (or not-obviously-race-signaling) one. In these studies there are huge differences in response rates and callbacks, even when everything else about the resumes (educational attainment, experience, etc.). is identical. It's true that many black people suffer structural disadvantages because of systemic lack of access to resources, etc. But they also suffer disadvantages because they're black.

luckylion 6y ago
> No, or at least, not exclusively. But primarily, right? Bias plays a role, so everything else being equal, a black applicant will receive fewer callbacks than a white or asian one, but (and that's a big but) the primary issue is that usually not all other things are equal. The average black person comes from a different socioeconomic background than the average white or asian person, so they'll usually not have the same level of education, but less so because of being discriminated against because of their race, but because of their class. More single-parent households, lower household income, their parents can't simply afford to hire tutors and put them through college etc etc. That's what makes these things somewhat weird. Yeah, sure, a black guy from a rich family that attended MIT will have a harder time finding funding than his white or asian friends from the same background ... but he'll have a much easier time than some white kid from a broken home. I believe two things to be true: class is the primary factor, not race. And it's more convenient to focus on race because it's much easier and feels less threatening to the (usually upper middle class and beyond) proponents. "We (as white people) need to share more" puts them in a group with (for the US) 60% of the population and is pretty abstract. Great for virtue signalling, terrible for concrete actions. "We (as rich people) need to share more" puts them in a much, much smaller group and suggests they could actually do something besides yelling at people to check their privilege: share some of the money they inherit. Great for doing something, terrible for virtue signalling because "well, put your money where your mouth is" is so obvious.
- apendleton 6y ago
  But there are programs that focus on access to resources based around class, and to the extent that there are factors besides class that matter (like race), exclusively-class-oriented programs fail to fully level the playing field. That this program exists to help combat bias doesn't stop someone else from starting a different program aimed at attacking a different angle with different eligibility criteria. The more the merrier.
- baddox 6y ago
  > Yeah, sure, a black guy from a rich family that attended MIT will have a harder time finding funding than his white or asian friends from the same background ... but he'll have a much easier time than some white kid from a broken home. Even if your claims are supported by evidence, why is it a bad thing to want to correct that disparity? The fact that there also exist other disparities in the world doesn’t seem like a good reason to oppose correcting one particular disparity.
  - luckylion 6y ago
    > Even if your claims are supported by evidence, why is it a bad thing to want to correct that disparity? It's not "bad" as in "omg don't do that", it's just focusing on a less significant problem instead of the large problem, which it conveniently shifts focus from. I don't think that it's intentional, but I'm absolutely sure that addressing the lacking social mobility in the US would benefit minorities much more than racism, because they are much, much more affected by it.
donw 6y ago
I have a question, just because I don't know if there is data on this. I've seen a bunch of studies on things like "stereotypically-african-american-name" vs "stereotypically-white-name", which are good evidence as to why a good hiring process should remove personally-identifying details from resumes. Not just the name of the applicant, either -- school names and locations can also introduce unfair biases. Company names are unavoidable, I think: there are only so many companies that fit the description of "Social Network, Global, >10k employees, uses lots of Javascript" or "Privatized Intelligence Service, Global, >100k employees, uses lots of Go". Buuuuuuut... is there research on what happens when you expand the number of dimensions? E.g., add in a mix of broad cultural backgrounds (Pan-European, Latin, Arab, Chinese, African, etc.), test different raraties ("Smith" is common, "Chaucer" is rare, both are Anglo) Does "Victoria Henderson" garner a statistically different response to "Vitalija Herrera" (Anglo vs Lithuanian)? What about "Nyala Aregawi" vs "Nadya Ahane" (Mixed Russian-Japanese)? What happens when we mix in popular fictional names? Solid bet there are boys out there named "Thor" and "Bilbo" by parents that are fans of Star Wars. No idea what actual outcomes might be, the data could yield interesting results on hiring biases.
- baddox 6y ago
  I don’t know if there is such evidence, and while it could be interesting to know just out of curiosity, it doesn’t strike me as particularly valuable to test the impact of adding additional irrelevant information to job applications.
  - donw 6y ago
    Fair enough. Kind of a data geek, hence the ask.
- theredbox 6y ago
  What even is a stereotypically-white-name? Anglo name?

luckylion 6y ago

> No, or at least, not exclusively. But primarily, right? Bias plays a role, so everything else being equal, a black applicant will receive fewer callbacks than a white or asian one, but (and that's a big but) the primary issue is that usually not all other things are equal. The average black person comes from a different socioeconomic background than the average white or asian person, so they'll usually not have the same level of education, but less so because of being discriminated against because of their race, but because of their class. More single-parent households, lower household income, their parents can't simply afford to hire tutors and put them through college etc etc. That's what makes these things somewhat weird. Yeah, sure, a black guy from a rich family that attended MIT will have a harder time finding funding than his white or asian friends from the same background ... but he'll have a much easier time than some white kid from a broken home. I believe two things to be true: class is the primary factor, not race. And it's more convenient to focus on race because it's much easier and feels less threatening to the (usually upper middle class and beyond) proponents. "We (as white people) need to share more" puts them in a group with (for the US) 60% of the population and is pretty abstract. Great for virtue signalling, terrible for concrete actions. "We (as rich people) need to share more" puts them in a much, much smaller group and suggests they could actually do something besides yelling at people to check their privilege: share some of the money they inherit. Great for doing something, terrible for virtue signalling because "well, put your money where your mouth is" is so obvious.

apendleton 6y ago
But there are programs that focus on access to resources based around class, and to the extent that there are factors besides class that matter (like race), exclusively-class-oriented programs fail to fully level the playing field. That this program exists to help combat bias doesn't stop someone else from starting a different program aimed at attacking a different angle with different eligibility criteria. The more the merrier.
baddox 6y ago
> Yeah, sure, a black guy from a rich family that attended MIT will have a harder time finding funding than his white or asian friends from the same background ... but he'll have a much easier time than some white kid from a broken home. Even if your claims are supported by evidence, why is it a bad thing to want to correct that disparity? The fact that there also exist other disparities in the world doesn’t seem like a good reason to oppose correcting one particular disparity.
- luckylion 6y ago
  > Even if your claims are supported by evidence, why is it a bad thing to want to correct that disparity? It's not "bad" as in "omg don't do that", it's just focusing on a less significant problem instead of the large problem, which it conveniently shifts focus from. I don't think that it's intentional, but I'm absolutely sure that addressing the lacking social mobility in the US would benefit minorities much more than racism, because they are much, much more affected by it.

donw 6y ago

I have a question, just because I don't know if there is data on this. I've seen a bunch of studies on things like "stereotypically-african-american-name" vs "stereotypically-white-name", which are good evidence as to why a good hiring process should remove personally-identifying details from resumes. Not just the name of the applicant, either -- school names and locations can also introduce unfair biases. Company names are unavoidable, I think: there are only so many companies that fit the description of "Social Network, Global, >10k employees, uses lots of Javascript" or "Privatized Intelligence Service, Global, >100k employees, uses lots of Go". Buuuuuuut... is there research on what happens when you expand the number of dimensions? E.g., add in a mix of broad cultural backgrounds (Pan-European, Latin, Arab, Chinese, African, etc.), test different raraties ("Smith" is common, "Chaucer" is rare, both are Anglo) Does "Victoria Henderson" garner a statistically different response to "Vitalija Herrera" (Anglo vs Lithuanian)? What about "Nyala Aregawi" vs "Nadya Ahane" (Mixed Russian-Japanese)? What happens when we mix in popular fictional names? Solid bet there are boys out there named "Thor" and "Bilbo" by parents that are fans of Star Wars. No idea what actual outcomes might be, the data could yield interesting results on hiring biases.

baddox 6y ago
I don’t know if there is such evidence, and while it could be interesting to know just out of curiosity, it doesn’t strike me as particularly valuable to test the impact of adding additional irrelevant information to job applications.
- donw 6y ago
  Fair enough. Kind of a data geek, hence the ask.
theredbox 6y ago
What even is a stereotypically-white-name? Anglo name?