[jscott2]

The metric 'percent-of-race-to-be-flagged' is more a measure how how homogenous the names of people of that race are and inadequate for this analysis.

If they actually wanted to show prejudice they should've normalized for that factor, then shown that certain people were still targeted at a higher rate.

That a strategy works better or worse for certain races may be racial, but is only racist if it was designed to work that way.