This accessible verbal summary of the state of modern macro by Narayana Kocherlakota is stunningly good on several dimensions. It provides a clear explanation of how modern academic macro works, a short, broad history of the development of macro over the last 40 years and some thoughtful reflections on modern macro and the great recession.
I particularly like the call for more attention to the nature of the shocks and to the integration of bubble behavior into macro models.
The one thing I would add is that another limitation on the usable complexity of macro models (or, indeed, of some structural models in microeconomics) is simply the capacity of the human mind to understand them. I don't think we are at that frontier yet, but at some point we will be.
I would be really enjoy reading the analogous article on climate models, which face many of the same issues.
Full disclosure: I overlapped with Narayana at Chicago, though we did not interact much due to our differing interests and being a couple of classes apart, and he married a friend from my year in the program.
Who was my favorite student this term?
9 months ago