Earlier quoted context omitted.
My beliefs aside [0], you do bring up a good point. I can't find the HN thread, but there was a good TEDx presentation on it as well: Given two curves on the same graph, you are not guaranteed to be able to min/max them simultaneously. For example: say you are a mortgage broker at a bank and you have to give out mortgages to people in your community. Obviously, the people in your community are diverse. There are men,…
If the bank does not take race or gender into its decision-making process, it is not subject to race/gender discrimination. Period. If the outcome produced is that people of a certain race or gender receive more or less loans, that does not mean that there is anything wrong with that system. The banks isolated system is not at all discriminatory, it just receives different inputs that happen to differ, on average, be…
No matter what you do, you'll be discriminatory (to a very high likelihood). Even if you don't look at the break-downs of your data, you will be discriminating against some group or another. It is incredibly unlikely that you can max/min two curves at once, let alone for all the classes, races, sexes, castes, genders, etc. that exist.
I agree that root causes should be addressed first and is the best use of time and resources. However, that's not how journalists see it. They see a red-meat story and they go with it.
Federal investigators, journalists, community organizers, etc don't really care what metrics the banks are using/not-using, they care about the results. And I'm saying that there will very nearly always be discrimination if you use nearly any metric. And that's not a bad thing. Being in a Catch-22 is part of life. Trying to get out of one is the thing that matters.