Good economics starts with troubling facts, makes some guesses based on what we already know about human behavior and theories elsewhere shown to work, uses data to test those guesses, refines (or radically alters) its line of attack based on the new set of facts, and eventually, with some luck, gets to a solution.
