From an old David Brooks article:
"It turns out that it's hard to change the destinies of nations and individuals just by pulling economic levers."
For the most part, economists—and macroeconomists in particular—used to think of the economy as a machine. They looked at the various inputs of the system (interest rates, capital levels, investment expenditures, etc.), related them with the outputs (chiefly GDP), and derived neat mathematical formulas to express these relationships. From there it seemed obvious that if you set public policy consistent with the theory, you'd have a prosperous society: if all the cogs were whirring along harmoniously, then a nation had no choice but to get richer.
We now know, of course, that reality is much more complicated. Prosperity, it seems, is not just about what the central bank does; fiscal and monetary policy do not fully contain the "destiny of nations." Instead, as we're slowly realizing—and as Mr. Brooks points out—it's the "murky" socio-anthropological stuff—like social norms, historical trajectories, culture, and institutions—that really counts.
In the 21st century, then, I think we're seeing the discipline move into increasingly murky waters. The old "economist-as-engineer" paradigm is giving way to "economist-as-anthropologist/sociologist"—which, I think, can only be a good thing. Although the latter approach is messier, it's the only natural next step; it's the only honest thing to do.