I've mentioned a few times now on the blog - including the last post - my concern about the fact that their paper is essentially a pre-post test. That's fine descriptively (Christina Romer has done a lot of work like this), but it's not as good for policy analysis.
So what might be different between the pre- and post-1913 period that might justify a difference-in-differences approach? I can think of a few things:
1. The move from extensive to intensive growth. I've referred to this previously as "the closing of the frontier" in the Cafe Hayek comment section, at which point George Selgin rather unceremoniously dismissed my views. But intensive growth and extensive growth certainly shouldn't be so controversial.
2. The move from very little of the labor force working for wages to a lot of the labor force working for wages.
3. The increased financialization of the economy (this is perhaps one of the most important points).
We could also think about the gold standard - is there a case for differentiating between the gold standard Fed and the post-gold standard Fed? The paper may have addressed this point - it's been several months since I've read it.
Can anyone think of anything else?
The paper is coming out in the Journal of Macroeconomics. If anyone has the time or inclination to run a difference-in-differences, perhaps with Great Britain or some other country as the counterfactual, I think that would be a valuable contribution. Macroeconometrics can learn a lot from microeconometrics.