Eons ago, when I was a student, I heard about an experiment in psych class. University students were randomly divided into teams--they all drew colored cards from a bowl in sight of each other, so everyone was aware that team membership was random. They participated in a number of games/competitions. At the end of it, they were all asked how they felt about their team members and about the members of the other teams according to a number of parameters. Most rated the members of their own team as more honest and more honorable than the members of the other team. This is a group membership that lasted an afternoon and that was established randomly. Everyone who participated was aware of those facts. They still majoritarily based their responses on group membership.

I try very hard to keep this in mind any time the issue of a person's membership in a group is brought up with respect to their actions.