Social preferences research (behavioral economics in general, I'm sure) is so plagued by identification issues. By that I mean, there are many possibilities for what it is that motivates people to do things, and it's really hard to tell them apart. And in behavioral economics, for the first time, we really are trying to figure out how people think, rather than just how they act as though they think.
It's still as-if economics, as meant by Friedman when he used that phrase, in the sense that people don't actually maximize complicated utility functions. But whereas it was ok to assume people are perfectly rational perfectly selfish actors when studying marketplaces (because they act as if they are, in those contexts), it's of course not ok to assume that when what you're trying to figure out is what makes people act different ways in different situations.
And all the sudden there is an explosion of possibilities to explore, and my job is hard.