Stanford Ed School prof, Myra Strober, once related a nice paradigm for relating psychology, sociology, and economics. When trying to predict what a person will do, the psychologist wants to know about his past, the sociologist wants to know about his present societal influences, and the economist wants to know how the person's decision will affect his future. Likely the truth depends on all three.
(Yes, that is a gross caricature, be prepared, there will be many more, but caricatures are often useful... Disclaimer, the following is pure speculation on my part, I invite you to tell me if you disagree.)
So I had a conversation with an Illinois psychologist Sharon Shavitt, that helped another big difference between economics and psychology click into place. My classes in the soul searching political science department (where every class began with two weeks of justifying itself as a science) helped. Psychology concerns itself with laws, while economics concerns itself with theorems.
A law is an observed regularity of behavior -- e.g. the law of gravity (pre-Einstein) -- without any attempt to explain why. Psychologists are primarily concerned with finding regularities of behavior, and expressing them in as universal terms as possible. The reasons for the behavior don't matter so much. The economist always wants to explain things. Start with basic assumptions/axioms, and derive the reason behind why these regularities of behaviors exist. At first glance, this just sounds like better science (esp in a Popperian sense). Though in order to make these derivations tractable, we wind up basing our derivations on axioms we know not to be true, in a sollipcistic (perhaps the first time i've used that word non-ironically) attempt to have our explanations stand up on a vaporous foundation.
Which helps me understand why my co-authors in our fashion paper think the finding that the meaning of fashion is socially constructed (in econo-speak, fashion has signal value only as a result of Nash equilibrium between all players). Curious that it is actually economics that would find itself more comfortable with this fundamentally po-mo (post modern) idea. The psychologist focuses on universals of behavior, ignoring from whence it came.
It also explains why psychologists don't seem bothered by the fact that attribution theory is really more law, than theory. A large literature exists describing how attributions operate in a vast variety of circumstances, but there is little attempt to explain why. So my clumsy attempt to explain it, is thus met with indifference, disinterest, maybe scorn.
I guess social psychologists take behavior as a primitive. Economists aren't much better. We tend to take preferences as a primitive. Evolutionary psychology and evolutionary economics take both fields further, but neither is well developed.