Do economists simply disregard social norm in their analysis?
Contrary to common belief, economists have tried to incorporate social norms in at least three ways. First, in collaboration with psychologists, they tried to model an agent's maximizing utility behavior with additional constraint of psychological bias. Second, they also tried to take into account other's utility function into the representative agent's utility function; known as endogenous preference approach.
This first two approaches still hold up a single representative agent behavior as micro-foundation of a social phenomena.
Akerlof and Kranton, in their latest book, Identity Economics, 2010, offer a different perspective. They use more than one representative agent and respective utility function (they called it identities) in looking at a certain social pattern.
I have read this interesting and readable book for the general public, but haven't had time to peruse the more technical academic journal articles behind this book --those on the bibliography. With that in mind, this is my knee-jerk reaction: how do we decide the number of agents' types and utility functions we need to put into account in analyzing, say, gender and labor market in Indonesia? Two, three, four? What is the general rule for determining that number?
Nevertheless, the book shows that economics, after all, has never been an isolated subject full with stubborn students. It did not stop with Milton Friedman's book on Price Theory.