3 ms·
> “institutional racism" deals with is if one race is disproportionately affected by a policy over others, which is more often than not caused by socioeconomic
by rajin444 4y ago
> “institutional racism" deals with is if one race is disproportionately affected by a policy over others, which is more often than not caused by socioeconomic factors that weren't considered during the creation of the policy
How do you prove these things are caused by socioeconomic factors? At a minimum we have nowhere near the genetic understanding to control for those aspects. At a higher level, human society, culture, and the systems created by it are infinitely complex. Grouping people on 1 axis then saying the reason for this grouping is “socioeconomic” seems…optimistic at best.
There are so many reasons one racial grouping may be disparately impacted by something. We should focus on fixing the downsides of that impact for all people, not asking why X arbitrary category we came up with doesn’t match Y (because the answer to question is far beyond our technical ability).
And our metric of success cannot be “x group more affected” - if we make that our metric, we’ll find a way to hit that target, correct or not.
- soco 4y agoThis sounds very nihilistic, or maybe libertarian I can't tell the difference. My point is, we must do something and as engineers we work based on data. Now you can argue successfully that data is not perfect, but if the alternative is (institutional) paralysis, then what? A state cannot afford to be that zen regarding the well-being of its citizens.
- choko 4y agoThis sounds authoritarian, or maybe progressive. I can't tell the difference. My point is that data without context is almost worthless when crafting policy.
- rajin444 4y agoI agree we must do something! It's much more reliable to talk with individuals and assess instances of racism, discrimination, etc. on a case base case basis (taking into account other evidence). We don't have to be paralyzed.