Neil Irwin, a few weeks ago:
The scholars mentioned by name are Alvin Roth, Raj Chetty, and Matthew Gentzkow, all truly great scholars. Actually Roth has done a fair amount of “theoretical modeling” in his career, and by all accounts his work has illuminated our understanding of “how things work in the real world”. Pure empirics has no monopoly on real-world insight; the "theory vs real world" heuristic so popular among econ outsiders (and some insiders) often does not mean what they think it means.
Am I correct in putting Chetty and Gentzkow primarily into the context of the “credibility revolution”? This consensus applied micro approach, which in my view has dramatically elevated the value and usefulness of the profession, typically produces results that are local to the data used. Often it's reasonable to assume that the "real world" is approximately linear locally, which is why this research agenda is so useful and successful. But I wrote about some of its limitations here, with respect to the minimum wage; the short version is that the usefulness of such results declines as the policies motivated by them get further from the specific dataset with which the results were derived. The only way around this is to make assumptions about the linearity of the “real world”, a nontrivial theoretical proposition.
This topic came up a few months ago when Noah Smith had a nice post quoting Angrist and Pischke:
This is fine with all the right caveats. But if we take the next step of applying the results to some real-world problem, such as policy, we start to get very model dependent in a hurry; again, we’re dependent on the linear functional form of the model (among other assumptions). In some cases those assumptions are probably fine; in others, maybe not. But you can’t escape theory. Pretending it’s not there doesn’t make it go away.
Beware of those peddling theory-free economic levers, for they are hiding something.
UPDATE 9/28/15: A few people have read this post as a critique of Chetty and Gentzkow, or of applied micro economists generally. It is not.
![]()
The specialties of the new recruits [to Stanford econ] vary, but they are all examples of how the momentum in economics has shifted away from theoretical modeling and toward “empirical microeconomics,” the analysis of how things work in the real world.
The scholars mentioned by name are Alvin Roth, Raj Chetty, and Matthew Gentzkow, all truly great scholars. Actually Roth has done a fair amount of “theoretical modeling” in his career, and by all accounts his work has illuminated our understanding of “how things work in the real world”. Pure empirics has no monopoly on real-world insight; the "theory vs real world" heuristic so popular among econ outsiders (and some insiders) often does not mean what they think it means.
Am I correct in putting Chetty and Gentzkow primarily into the context of the “credibility revolution”? This consensus applied micro approach, which in my view has dramatically elevated the value and usefulness of the profession, typically produces results that are local to the data used. Often it's reasonable to assume that the "real world" is approximately linear locally, which is why this research agenda is so useful and successful. But I wrote about some of its limitations here, with respect to the minimum wage; the short version is that the usefulness of such results declines as the policies motivated by them get further from the specific dataset with which the results were derived. The only way around this is to make assumptions about the linearity of the “real world”, a nontrivial theoretical proposition.
This topic came up a few months ago when Noah Smith had a nice post quoting Angrist and Pischke:
A principle that guides our discussion is that the estimators in common use almost always have a simple interpretation that is not heavily model-dependent.
This is fine with all the right caveats. But if we take the next step of applying the results to some real-world problem, such as policy, we start to get very model dependent in a hurry; again, we’re dependent on the linear functional form of the model (among other assumptions). In some cases those assumptions are probably fine; in others, maybe not. But you can’t escape theory. Pretending it’s not there doesn’t make it go away.
Beware of those peddling theory-free economic levers, for they are hiding something.
UPDATE 9/28/15: A few people have read this post as a critique of Chetty and Gentzkow, or of applied micro economists generally. It is not.