4 ms·
Purely hypothetically, if there are string indications that parental socio-economic status influence the chances of their children. And that socio-economic stat
by randomswede 5y ago
Purely hypothetically, if there are string indications that parental socio-economic status influence the chances of their children. And that socio-economic status is somewhat (to strongly) correlated with previous race-influenced policies, is a generation or two of compensating for that "inflicting racial injustice"? Is a failure to compensate for it not explicitly "continuing racial injustice"?
I mean, I literally have no answer. Except, possibly, "yes" and "yes", and "the former is on the whole a lesser evil than the latter"?
- rebelos 5y ago> Purely hypothetically, if there are string indications that parental socio-economic status influence the chances of their children. And that socio-economic status is somewhat (to strongly) correlated with previous race-influenced policies, is a generation or two of compensating for that "inflicting racial injustice"? Is a failure to compensate for it not explicitly "continuing racial injustice"? This makes two assumptions that I would challenge: that affirmative action as it exists at present is benefitting the most socioeconomically vulnerable underrepresented minorities and that the problem you're describing cannot be tackled with race neutral wealth redistribution and other need-aware policymaking. Also, if you want to go down the racial grievance accounting rabbithole, then the Irish, Japanese, Jews, etc would like a word as well. You're be hard-pressed to find a race that doesn't have some such claim.
- randomswede 5y agoI am specifically NOT trying to go down "redress historical grievances", only "try to correct for currently existing imbalances". There's probably also something to be said for having policies trying to get some geographic rebalancing in place as well. But, what I am trying to say is that redressing recent-past wrongs MAY require slight over-compensation. To make anything other than sweeping generalisations, I would need numbers. Oh so many numbers.