Wave a clove of garlic and chant “randomization”

Since hard and inconclusive thought is needed to transfer the results learned from randomized experiments into other domains, there must therefore remain uncertainty and ambiguity about the breadth of application of any findings from randomized experiments… (You and I know that truly consistent estimators are imagined, not real.)

Angrist and Pischke understand this. But their students and their students’ students may come to think that it is enough to wave a clove of garlic and chant “randomization” to solve all our problems just as an earlier cohort of econometricians have acted as if it were enough to chant “instrumental variable.”

That is Ed Leamer responding to Angrist and Pischke’s article on the causality revolution econometrics. Here is Leamer interviewed by EconTalk.

Those and other articles in the latest issue of JEP. Well worth reading. Sadly gated. Please post any ungated links in comments if you know them.

6 thoughts on “Wave a clove of garlic and chant “randomization”

  1. I liked the point made by Angrist and Pischke (at the bottom of p. 24 I believe) that some critiques of experiments seem to equate triviality with narrowness of scope. Seemingly trivial experiments can sometimes reveal potent results on behavioral issues- this is not to say that Freakonomics-type work is whats called for, just that we should not dismiss such work ex ante as being wholly useless for the “big questions”.

  2. Joke from the beginning of Michael Keane’s paper in the same issue:

    An official at Treasury asks three experts, “What’s 200 billion plus 200 billion?” The first expert, a mathematician, immediately responds, “Four hundred billion, of course.” The second, an economist, kind of grimaces and says, “Well, that depends…” But the third expert, an econometrician, doesn’t immediately answer. Instead, he gets up and quietly closes the office door. Once he’s sure no one is listening, he leans over and whispers in the official’s ear, “What do you want it to be?”

    Thought it worth sharing, for those haven’t been able to access ungated copies.