11 ms·

> So minority candidates are given an advantage in getting their foot in the door, but still have to prove themselves qualified for the job by doing well in int

> So minority candidates are given an advantage in getting their foot in the door, but still have to prove themselves qualified for the job by doing well in interviews. Discrimination in selecting who to interview is absolutely a form of discrimination. Imagine I tell my recruiters to exclusively interview white Catholics, and I respond "well, those white Catholics still had to pass the skill-based interview. Had we interviewed any non-whites or non-catholics, the interview would be unbiased towards them" Is that a non-discriminatory hiring process? The fact that non-catholics and non-whites weren't even given a chance to interview is rendered irrelevant by the fact that the White Catholics that were still had to pass a skill-based interview?

tptacek 4y ago
None of this engages with the comment it replies to. One way to see that is to simply remove the quoted sentence, and read the comment to see if it would remain coherent at the top of the thread.
- Manuel_D 4y ago
  It absolutely engages with the parent comment. The parent comment is trying to argue that discrimination in selecting which candidates to interview isn't a form of bias. This is incorrect as I explain in my comment.
  - chipotle_coyote 4y ago
    The apparent Microsoft policy is that every interview pool “must include at least one external African-American, Black, Hispanic, or Latina/o and one external female candidate”. Your counterexample is “every interview pool has to be entirely white.” I was originally going to write that you’re playing the “I’m technically correct” trick here, but I don’t think your argument actually rises to the level of technical correctness. Setting aside the debate over whether the former policy is desirable, it is clearly not the same as the latter policy. If Microsoft had said “everyone interviewed cannot be a white male,” or even “most people interviewed cannot be white males,” then you could more credibly try to make the case you’re aiming for. But they simply didn’t.
    - Manuel_D 4y ago
      How big is said interview pool? With a pool of four people, which isn't uncommon in my experience, that means that 50% of the interviews are locked behind racial and gender requirements. My example just made it bluntly obvious that discrimination in the interview stage is still discrimination, there's no "technically correct trick here". Mandating that X% of your interviews be of a particular race or gender is discrimination, no matter the value of X. Setting X to 100 just makes it very clear.
      - chipotle_coyote 4y ago
        > Mandating that X% of your interviews “2 of the candidates you interview must meet these gender/racial requirements” is a mandate, but it’s not a percentage mandate. I know that’s a little pedantic, but I think it’s important pedantry, because you keep using percentages: > With a pool of four people, which isn’t uncommon in my experience, that means 50% of the interviews are locked behind racial and gender requirements. That’s only true if you are limited to just four people, which obviously you are not. You may feel it’s an unfair burden on the hiring manager to expand the candidate pool if necessary to meet the racial and gender requirements, but that’s not an argument about discrimination. > Setting X to 100 just makes it very clear. I sincerely believe setting X to 100 makes it a different argument. :) “All of your candidates must be X” is manifestly not the same as “some of your candidates must be X”. (The former may require you to leave out candidates you think are qualified, the latter does not, for a start, which strikes me as an extremely important distinction in this context.)
        Manuel_D 4y ago
        > The former may require you to leave out candidates you think are qualified, the latter does not, for a start, which strikes me as an extremely important distinction in this context.)? Incorrect. If you mandate that 20% of candidates be Y, but only 10% of candidates in the applicant pool are Y and 90% are X then on average you need to exclude 50% of qualified non-X candidates. If I have a pool of 90 X and 10 Y candidates and I have a quota of 80,20 then even if I include all 10 Y candidates I can only include 40 of the 90 X candidates. Sure, if I said that 100% have to be Y then all of the X candidates would be excluded. But even lower quota values still result in the out-group being limited. > “2 of the candidates you interview must meet these gender/racial requirements” is a mandate, but it’s not a percentage mandate. I know that’s a little pedantic, but I think it’s important pedantry, because you keep using percentages: Since there's a finite number of candidates it's still ultimately a percentage. The percentage is variable based on the total number of candidates, but it's still a percentage in the end. Quotas and caps are two sides of the same coin. Instituting a minimum representation of one group, is fundamentally the same thing as capping the representation of those who don't belong to said group.
- steve76 4y ago
  [dead]
- missingrib 4y ago
  It directly responds to the comment even without the quoted sentence.
sascha_sl 4y ago
- Manuel_D 4y ago
  > You addressed the wrong argument. The intent is to give groups that are disadvantaged in the hiring process a chance. The intent is equity. Your example distorts that by shifting the balance in favor of an already advantaged group. Then anonymize the resumes so that recruiters can't tell which candidates are men or women, or which is white, Asian, Black, etc. You don't eliminate discrimination by setting caps on how many interviewees can belong to each race. It certainly could be the case that whites are advantaged (curious why you focus on whites despite Asians being far more overrepresented, by the way). Put the proverbial veil between the candidate and the hiring manager, and we'll find out the truth. One of my previous workplaces rejected proposals to anonymize our interview process, on the grounds that it would inhibit our diversity initiatives. Interviewing.io did an experiment relative to gender with anonymized phone interviews, and the result were the opposite of the traditional narrative [1]. 1. https://blog.interviewing.io/we-built-voice-modulation-to-mask-gender-in-technical-interviews-heres-what-happened/ https://blog.interviewing.io/we-built-voice-modulation-to-ma...
  - ealexhudson 4y ago
    Blinding an interviewer by changing pitch/modulating the voice does little to erase the actual disadvantage. Few are arguing "people that sound like women are discriminated against", they're saying that the systems are set up in such a way that there is a bias against women that encompasses the evaluation of their experience, their work profiles, the topics they're interested in, the projects they've completed previously, etc. It's a systemic disadvantage, and that experiment isn't getting at the issue being claimed. As just one example, women still do the majority of child rearing, especially babies. People who want children make that choice, but men typically take a few weeks out of their career whereas women take months or more. That's a systemic disadvantage women suffer.
    - Manuel_D 4y ago
rrradical 4y ago
> Discrimination in selecting who to interview is absolutely a form of discrimination. Exactly. That's what the parent comment is saying. But they are thinking about the entire funnel, not just the end of it. By the time a slate of candidates reaches a company's hiring process, there has already been an immense selection bias against minority candidates. Two people growing up in different places (not different cities, but different neighborhoods within the same city) have lived in completely different worlds. Their schools are different; their health care is different; their safety is different; their opportunities are different; the people they know are different. And much of the time there's a stark racial difference in the makeup of those places. Historically this was very much intentional; but even if it were no longer intentional, the effects won't dissipate for a long time. So when you get a slate of candidates that all happen to be white, it's not just a random coincidence. Imagine if a slate of candidates were all black. That would seem kind of odd, right? Now obviously the best thing would be to fix all the other environmental factors that led to an all-white candidate slate. But that's not going to happen any time soon. So a good thing to do is apply some pressure on the funnel to elevate candidates that just barely miss out. In other words, candidates that are strong, but, say, don't know anyone that works at microsoft (no surprise there... two worlds) or perhaps don't think they're good enough. The article points to a rising black employee population has some kind of evidence of injustice, but, if the company works harder to find qualified black candidates then obviously the percentage would rise. Unless we think that skin-color is a predictor of performance (ugh, I hope no one actually does) then improving a hiring process would result in an employee population that more closely matches the demographics of the population at large.
- knewter 4y ago
  A person that truly believe this would find themselves advocating for an increase in the interview pool of West Virginian / Appalachian white kids. I have seen exactly zero people suggest that. That's how I know it's made up.
- tomp 4y ago
  This is a nice way to excuse your racism. A lot of people use this perverted logic. But at the end of the day, it's still evil racism. Different "worlds" (neighbourhoods, schools, health care) doesn't happen because of skin color, it happens because of wealth/poverty. So if you apply a racist filter on top of the (implicit) wealth filter, you're just being racist against poor Asian & white people.
tshaddox 4y ago
It's essentially impossible to have a job opening without making choices that will directly affect which people see that job opening. If you go to some university's job fair, you're selecting for the students at that university. If you had chosen to go to a different university's job fair instead, the students who were exposed to your job posting would likely have different qualities. There is no "default behavior that doesn't constitute making a choice," thus the only thing we can debate is the virtues of which particular choice was made, not whether a choice was made.

tptacek 4y ago

None of this engages with the comment it replies to. One way to see that is to simply remove the quoted sentence, and read the comment to see if it would remain coherent at the top of the thread.

Manuel_D 4y ago
It absolutely engages with the parent comment. The parent comment is trying to argue that discrimination in selecting which candidates to interview isn't a form of bias. This is incorrect as I explain in my comment.
- chipotle_coyote 4y ago
  The apparent Microsoft policy is that every interview pool “must include at least one external African-American, Black, Hispanic, or Latina/o and one external female candidate”. Your counterexample is “every interview pool has to be entirely white.” I was originally going to write that you’re playing the “I’m technically correct” trick here, but I don’t think your argument actually rises to the level of technical correctness. Setting aside the debate over whether the former policy is desirable, it is clearly not the same as the latter policy. If Microsoft had said “everyone interviewed cannot be a white male,” or even “most people interviewed cannot be white males,” then you could more credibly try to make the case you’re aiming for. But they simply didn’t.
  - Manuel_D 4y ago
    How big is said interview pool? With a pool of four people, which isn't uncommon in my experience, that means that 50% of the interviews are locked behind racial and gender requirements. My example just made it bluntly obvious that discrimination in the interview stage is still discrimination, there's no "technically correct trick here". Mandating that X% of your interviews be of a particular race or gender is discrimination, no matter the value of X. Setting X to 100 just makes it very clear.
    - chipotle_coyote 4y ago
      > Mandating that X% of your interviews “2 of the candidates you interview must meet these gender/racial requirements” is a mandate, but it’s not a percentage mandate. I know that’s a little pedantic, but I think it’s important pedantry, because you keep using percentages: > With a pool of four people, which isn’t uncommon in my experience, that means 50% of the interviews are locked behind racial and gender requirements. That’s only true if you are limited to just four people, which obviously you are not. You may feel it’s an unfair burden on the hiring manager to expand the candidate pool if necessary to meet the racial and gender requirements, but that’s not an argument about discrimination. > Setting X to 100 just makes it very clear. I sincerely believe setting X to 100 makes it a different argument. :) “All of your candidates must be X” is manifestly not the same as “some of your candidates must be X”. (The former may require you to leave out candidates you think are qualified, the latter does not, for a start, which strikes me as an extremely important distinction in this context.)
      - Manuel_D 4y ago
        > The former may require you to leave out candidates you think are qualified, the latter does not, for a start, which strikes me as an extremely important distinction in this context.)? Incorrect. If you mandate that 20% of candidates be Y, but only 10% of candidates in the applicant pool are Y and 90% are X then on average you need to exclude 50% of qualified non-X candidates. If I have a pool of 90 X and 10 Y candidates and I have a quota of 80,20 then even if I include all 10 Y candidates I can only include 40 of the 90 X candidates. Sure, if I said that 100% have to be Y then all of the X candidates would be excluded. But even lower quota values still result in the out-group being limited. > “2 of the candidates you interview must meet these gender/racial requirements” is a mandate, but it’s not a percentage mandate. I know that’s a little pedantic, but I think it’s important pedantry, because you keep using percentages: Since there's a finite number of candidates it's still ultimately a percentage. The percentage is variable based on the total number of candidates, but it's still a percentage in the end. Quotas and caps are two sides of the same coin. Instituting a minimum representation of one group, is fundamentally the same thing as capping the representation of those who don't belong to said group.
steve76 4y ago
[dead]
missingrib 4y ago
It directly responds to the comment even without the quoted sentence.

sascha_sl 4y ago

Manuel_D 4y ago
> You addressed the wrong argument. The intent is to give groups that are disadvantaged in the hiring process a chance. The intent is equity. Your example distorts that by shifting the balance in favor of an already advantaged group. Then anonymize the resumes so that recruiters can't tell which candidates are men or women, or which is white, Asian, Black, etc. You don't eliminate discrimination by setting caps on how many interviewees can belong to each race. It certainly could be the case that whites are advantaged (curious why you focus on whites despite Asians being far more overrepresented, by the way). Put the proverbial veil between the candidate and the hiring manager, and we'll find out the truth. One of my previous workplaces rejected proposals to anonymize our interview process, on the grounds that it would inhibit our diversity initiatives. Interviewing.io did an experiment relative to gender with anonymized phone interviews, and the result were the opposite of the traditional narrative [1]. 1. https://blog.interviewing.io/we-built-voice-modulation-to-mask-gender-in-technical-interviews-heres-what-happened/ https://blog.interviewing.io/we-built-voice-modulation-to-ma...
- ealexhudson 4y ago
  Blinding an interviewer by changing pitch/modulating the voice does little to erase the actual disadvantage. Few are arguing "people that sound like women are discriminated against", they're saying that the systems are set up in such a way that there is a bias against women that encompasses the evaluation of their experience, their work profiles, the topics they're interested in, the projects they've completed previously, etc. It's a systemic disadvantage, and that experiment isn't getting at the issue being claimed. As just one example, women still do the majority of child rearing, especially babies. People who want children make that choice, but men typically take a few weeks out of their career whereas women take months or more. That's a systemic disadvantage women suffer.
  - Manuel_D 4y ago

rrradical 4y ago

> Discrimination in selecting who to interview is absolutely a form of discrimination. Exactly. That's what the parent comment is saying. But they are thinking about the entire funnel, not just the end of it. By the time a slate of candidates reaches a company's hiring process, there has already been an immense selection bias against minority candidates. Two people growing up in different places (not different cities, but different neighborhoods within the same city) have lived in completely different worlds. Their schools are different; their health care is different; their safety is different; their opportunities are different; the people they know are different. And much of the time there's a stark racial difference in the makeup of those places. Historically this was very much intentional; but even if it were no longer intentional, the effects won't dissipate for a long time. So when you get a slate of candidates that all happen to be white, it's not just a random coincidence. Imagine if a slate of candidates were all black. That would seem kind of odd, right? Now obviously the best thing would be to fix all the other environmental factors that led to an all-white candidate slate. But that's not going to happen any time soon. So a good thing to do is apply some pressure on the funnel to elevate candidates that just barely miss out. In other words, candidates that are strong, but, say, don't know anyone that works at microsoft (no surprise there... two worlds) or perhaps don't think they're good enough. The article points to a rising black employee population has some kind of evidence of injustice, but, if the company works harder to find qualified black candidates then obviously the percentage would rise. Unless we think that skin-color is a predictor of performance (ugh, I hope no one actually does) then improving a hiring process would result in an employee population that more closely matches the demographics of the population at large.

knewter 4y ago
A person that truly believe this would find themselves advocating for an increase in the interview pool of West Virginian / Appalachian white kids. I have seen exactly zero people suggest that. That's how I know it's made up.
tomp 4y ago
This is a nice way to excuse your racism. A lot of people use this perverted logic. But at the end of the day, it's still evil racism. Different "worlds" (neighbourhoods, schools, health care) doesn't happen because of skin color, it happens because of wealth/poverty. So if you apply a racist filter on top of the (implicit) wealth filter, you're just being racist against poor Asian & white people.

tshaddox 4y ago

It's essentially impossible to have a job opening without making choices that will directly affect which people see that job opening. If you go to some university's job fair, you're selecting for the students at that university. If you had chosen to go to a different university's job fair instead, the students who were exposed to your job posting would likely have different qualities. There is no "default behavior that doesn't constitute making a choice," thus the only thing we can debate is the virtues of which particular choice was made, not whether a choice was made.