There’s a peculiar pattern to religious beliefs worldwide and across history. Most of the religions likely practiced throughout history don’t include an all-powerful, morally concerned god. Yet, most religious believers on earth today believe in exactly this sort of Big Moralizing God. What explains this pattern?
I think one of the best explanations out there right now to explain this pattern goes something like this: People’s supernatural beliefs can have tangible effects on their behavior. People who believe that an all-powerful god is monitoring their behavior, cares about that behavior, and can reward or punish that behavior, might make different choices than do people who believe in a god without this skill set. For example, if people feel like a morally concerned god is monitoring their behavior, they might be more cooperative with others, or resist temptation in other ways. Beyond individuals, societies full of people who check their behavior because they believe in a Big Moralizing God might be more successful than societies full of people who do not. Over time, those groups that were best able to overcome cooperative dilemmas (and other sorts of dilemmas that disrupt the smooth functioning of society) would have a competitive advantage. The result? The most successful religions in the world tend to be those with beliefs and practices that interact to produce stable sets of norms that promote group cohesion and competitiveness.
This example focused primarily on cooperation, but we can imagine all sorts of norms that would be more easily stabilized by certain religious beliefs that would also have interesting long-term consequences for the prosperity of societies. For example, many religions spend a lot of time dealing with mating and marriage systems. Where norms related to mating and marriage yield different reproduction rates across groups with various belief systems, there is again a case for potential competitive advantages. So, the basic idea is that the specific contents of religious systems—including the kind of god a group believes in—can have profound effects on the success of that group. This can produce dynamics where cultural group selection is possible. (For those who are uncomfortable talking about “cultural” or “group” selection, let alone both, feel free to use “differential success of societies based on how norms, beliefs, and practices common with those groups affect the choices of individuals” instead. I’ll stick with “cultural group selection” because it’s shorter.)