pull down to refresh

I think people would love to, but haven't been able to. Race is a statistically significant predictor, even after controlling for things like income, education, etc, in many sociological studies. It probably proxies for cultures and values that are difficult to observe through other means.
The problem is that if your metrics discriminate, any policies made as a result of the metric will discriminate too.
Does it matter what someone's race is when you measure non-marriage as a "Percent of Americans" as the title implies?
reply