Having completely failed to anticipate the global economic meltdown, academic macro-economists have run headlong into "intellectual insolvency", according to Justin Wolfers at Freakonomics.
For far too long, they've relied on elegant theory instead of empirically grounded data. He is, however, hopeful that economists will be able to exit a bankruptcy of their own.
As a whole, the economics profession has become more empiricallygrounded. New large datasets offer the prospect of truly understanding
individual behavior in a way that paying lip service to “micro-founded
models” doesn’t. Many are engaged in the tricky business of writing
more psychologically grounded models that are closely tied to real
human behavior. Computational advances allow us now to take differences
in people, and how they respond, far more seriously. The absence of
available data doesn’t necessarily require more complex theory; we are
also learning how better to measure the objects we model.