The study says nothing about the underlying reasons for that discrimination though. The point of discounting those two roles in my post was to test my personal hypothesis of the underlying cause based on the data at hand - not to trivialise the findings of the authors.