> There's no evidence migration background was a good predictor of fraud This would be a surprising statement to be true (theres technically evidence for Jesus

> There's no evidence migration background was a good predictor of fraud This would be a surprising statement to be true (theres technically evidence for Jesus being the son of god, its just the priors/hypothesis complexity are low/high. And the fact theres more evidence against the hypothesis). If you asked the makers of the law/algorithm what do you think they'd say?