4 ms·
Two words: disparate impact. I'd be very concerned that they're somehow charging more to a protected class. Yeah, pricing is a bit of a black box, but that isn
by ThrustVectoring 9y ago
Two words: disparate impact.
I'd be very concerned that they're somehow charging more to a protected class. Yeah, pricing is a bit of a black box, but that isn't a defense against "blacks end up getting charged $1.2 for every $1 whites get charged for equivalent rides". There's enough protected classes that you can basically do p-fishing. "Uber price discriminates against black people" not statistically significant? Run the numbers again with Asian instead, and see if it comes up. Or women. Or religion (rides to church service more expensive? Whoops.)
- benmarten 9y agolol! any proof?
- LoSboccacc 9y agoI think parent means to search it in reverse: split the pricing by class enough and eventually you'll find something that correlates. That's a quest for a modern ambulance chaser to be sure, but I can see some merit in it.
- ThrustVectoring 9y agoYeah, it'd be something that could possibly be dug up. Many location-based systems for price discrimination can run into similar issues - see https://en.wikipedia.org/wiki/Redlining https://en.wikipedia.org/wiki/Redlining Note that being purely profit-driven through your pricing scheme isn't a good defense. There could easily be a "taxis are racist and won't go into certain areas, therefore Uber has less competition serving black neighborhoods, therefore Uber can increase their prices there."