Which part of DEI do you hate?

You're viewing part of a thread.

Show Context

325 comments

"Badly implemented colorblind policies didn't stop racism in this one country, so let's have explicitly racist policies."
If they are still racist, they are not colorblind. Make stronger colorblind policies and enforce them. Color aware policies don't do anything either if they are only on paper.
Besides, you ignore the point of my criticism. Color aware policies don't prevent inequity, they shift it elsewhere. They keep some places and aspects of life racist while having other be reverse racist. On individual level, the inequity increases, but people pat themselves on the back because when you only look at it based on color, it averages out. It is like saying we should increase the pay of Billionaires to increase average wages. The statistic looks better, but it did not help most people.
- I think we may be operating on different suppositions, so addressing that rather than wasting time clarifying details about France's choice to never record demographic stats for things would be best.
  Do you think systemic racism exists and is a large problem in the USA or France?
  
  Do you think systemic racism exists and is a large problem in the USA or France?
  Yes.
  
  OK, we agree on that.
  To what extent do you think that implicit or unconscious bias cause visible minority groups to need to have to work harder and be more exceptional to get a position, role, or responsiblity, or a n on-category specified grant, assistance, or similar?
  
  I think it would considerably vary from place to place, even workplace to workplace. In some (rare) places not at all, in some places considerably. I would be entirely guessing if I was to say what the average was.
  
  Replying to this one because newer. Have read and taken the other reply of yours into account too.
  I agree that we're off on a vibes and feels thing here because we don't have the data, and obviously it will vary between workplaces and individuals (even if to put systemic issues as individual choice/responsibility just covers for those systemic issues).
  We do have data from France showing that their entirely colourblind governance has not helped, despite targeting on socio economic or geographic bounds.
  When surely, if colourblind policies would do better at undoing systemic racism, wouldn't France have had better outcomes from them?
  
  Look, I don't know what exactly France did, maybe colorblind measures are not very effective. Maybe France picked stupid ones and implemented them badly. Let's not pretend there is only one way to do colorblind hiring.
  But my counter question is this. You say it did not help in France. How do you measure that? If one black person has it much easier while another was not helped at all, is that success? That is what I have issue with. Color-aware policies are extremely likely to just fake the statistics about groups, while if you actually compare random person to random person, it is just as (if not more) unfair as before. I believe it does not create real equity, it just fools statistics.
  You should not measure inequity between arbitrary groups. You should measure inequity between individuals to get a reliable metric.
  
  First, maybe this will help fill in as a starter on the French situation.
  Secondly, I do agree that targets and statistics inevitably distort and pervert any goals. So it will tend towards failure, but that's government. It never really works, and since I assume we're talking about the system we're in rather than a new one I don't think it's a deal-breaker.
  Thirdly, and most pertinently: due to systemic racism/prejudices there is a barrier to various arbitrary socially constructed groups that other arbitrary socially constructed groups do not need to deal with.
  By ignoring that there is a barrier to some in the form of systemic prejudice you don't actually help those more discriminated against groups. You just help the arbitrary groups that are less discriminated against. Maybe you have less inequality overall because the discriminated against group is a minority, but I don't think either of us think that that makes it "better".
  This is in fact where France has gotten to in its starting to analyse it's own colourblindness.
  
  First, maybe this will help fill in as a starter on the French situation.
  So they picked extremely stupid ones, got it.
  Secondly, I do agree that targets and statistics inevitably distort and pervert any goals. So it will tend towards failure, but that's government. It never really works, and since I assume we're talking about the system we're in rather than a new one I don't think it's a deal-breaker.
  Depends on how much they get "perverted and distorted". It absolutely is a deal breaker if it makes things worse than before.
  Thirdly, and most pertinently: due to systemic racism/prejudices there is a barrier to various arbitrary socially constructed groups that other arbitrary socially constructed groups do not need to deal with.
  By ignoring that there is a barrier to some in the form of systemic prejudice you don't actually help those more discriminated against groups. You just help the arbitrary groups that are less discriminated against. Maybe you have less inequality overall because the discriminated against group is a minority, but I don't think either of us think that that makes it "better".
  I don't think we understand each other. I am not saying we should do nothing. We should try to create policies that enforce color blind hiring, rewarding, etc. E.g. have people evaluate work before knowing whose work it is where possible. I am not saying there can't be any color/gender-aware policies anywhere. I am certainly not saying we should stop collecting statistics and put our heads into the sand. But we shouldn't hire/promote/reward people based on their race/gender in either direction.
  How would such a policy even work? You measure by how much is each minority disadvantaged on average and give them advantage by that amount via whatever mechanism? So the individuals that were already treated fairly now have an advantage even compared to the majority, those that were disadvantaged most are still disadvantaged, but a bit less and some random people from the majority are disadvantaged, because hiring is a zero sum game. You arguably did not make the system any more fair. The only good part is that it probably reduces by how much the most disadvantaged people are disadvantaged by.
  More importantly, you do nothing to fix the impression people have, that minorities are doing less/worse work, yet show everyone they are treated preferentially. This will cause people from the majority to wonder with every failure, whether it is because of the unfair advantage minorities are given. You can't even try to disprove it, because it is true in some cases. Rare cases perhaps, but very few people would care.
  Then act surprised when this creates conscious racists and the majority tells you to fuck off and elects a candidate that cancels DEI initiatives entirely. See the issue?
  In a democracy, you will never be able to enact policies that fix subconscious racism without fixing peoples perceptions. You will get voted out. That's why the policies have to be color-blind, even if they are less effective (take longer to work).
  And if we are lucky and do the policies well, we may even fix plenty of other biases unrelated to race and gender and eventually have much better results than color-aware.
  PS: If you know how to say color-aware and color-blind in a way that includes gender and other minorities, can you let me know? I think you understand what I mean but it still bothers me I am using the wrong word.
  
  Status aware... I don't like that. Protected groups... Even worse. Minority groups... Feels odd applying that to women, and various intersections...
  Statistically Disadvantaged/discriminated identities?
  Ehh, it is hard. I think that's why govs haven't managed to do a good job with naming it.
  On the main point, I agree that there is often a perception/"PR" problem for these policies.
  But then, in the UK where the policy was just "when deciding between two equally qualified candidates, choose the under represented one" still got done in the right wing media as "law mandating hiring on unqualified individuals", so I don't think that adjusting would do a huge amount of work.
  I think the contention is that I think that colouring hiring policies have been shown to not work, because it's very hard to implement in practice. At least collecting identity data would stave off the level of head-in-the-sand France reached.
  If the hiring process has an interview stage, how to make it identity-blind?
  How to deal with the perception of people, especially women, in a management position?
  I do agree that the main thing is hitting the underlying perception issues, but how to do that without creating a world where they're visibly untrue is trickier. But if it was an easy problem there'd probably be less division on how to tackle it.
  
  On the main point, I agree that there is often a perception/"PR" problem for these policies.
  I think calling it perception/PR problem is misleading, because it implies it is just misunderstood. If that was the case, it would just be a question of how to inform people. The issue is informed people dislike these policies as well, because they genuinely are unfair towards individuals and people are rarely willing to to be treated unfairly for the good of the society. People are also extra sensitive to governments and other people in positions of power treating people unfairly, arguably for a good reason.
  But then, in the UK where the policy was just "when deciding between two equally qualified candidates, choose the under represented one" still got done in the right wing media as "law mandating hiring on unqualified individuals", so I don't think that adjusting would do a huge amount of work.
  Yeah, obviously political parties fear mongering about the policies to get votes is a very big issue on it's own. But even if the criticism is way over exaggerated, can you rationally defend even that policy? Two simple points:
  There are no two equal candidates in practice, you can always add more tie-breaker criteria, like their expected salary, volunteering activities, ... So despite how it is worded, in reality, it just gives the decision to hiring managers, since they can just decide to only check basic qualifications and call all candidates equal, if they want to preferentially hire the diverse ones or they can keep adding criteria until the candidates are not equal if they dislike the policy.
  Winning tie-breakers is a significant advantage. Casinos never lose, and in most games it is just because they win in case of ties. Meaning it can still be significantly unfair.
  And again, I don't see how you could defend it other than insisting it makes things more fair overall even if it is unfair to individuals.
  If the hiring process has an interview stage, how to make it identity-blind? How to deal with the perception of people, especially women, in a management position?
  Yeah, it is a really difficult issue that probably does not have a single answer that can be applied everywhere. There probably have to be individualised solutions for various scenarios. There even may be situations where it can't be fixed at all until peoples perceptions improve and biases erode. Hopefully, showing people their biases are incorrect in different situations will be enough to do that. I really think normalizing diversity through means people perceive as fair could do that.
  I do agree that the main thing is hitting the underlying perception issues, but how to do that without creating a world where they're visibly untrue is trickier. But if it was an easy problem there'd probably be less division on how to tackle it.
  Yeah, unfortunately, it is one of the most difficult issues our society faces. :(
  I feel like instead of trying to implement one solution right now, we maybe should try to encourage workplaces to experiment with various policies and collect data back. Try to find working solutions by iterating and continuously improving policies like we do in engineering. Hopefully, it can help find decent enough solutions to chip at peoples biases.
- Make stronger colorblind policies and enforce them
  Any suggestion of such policies and ways to enforce them?
  
  I have general ideas in what direction to look at, but I don't see how I could post them without this inevitably devolving into off-topic discussion of hiring policies.

325 comments