Thread by @lewbel

As a student at MIT, I heard the expression, "The Iron Law of Econometrics," referring to OLS coefficients being biased toward zero, due to measurement error attenuation bias and other data issues. #EconTwitter #Econometrics
1/4

But who coined the term, "iron law of econometrics?"
Hausman (2001 JEP) says, "At MIT I have called this the 'Iron Law of Econometrics' - the magnitude of the estimate is usually smaller than expected."
This would seem to answer the question...
2/4

However, In a blog post, Mankiw attributes the phrase to Franklin Fisher. But the Fisher and Monz (1991) textbook says that the iron law of econometrics is, "data are never available in the form in which one wishes to use them."
3/4

And even earlier, a 1989 issue of The Economist says "the iron law of econometrics is that there is no such thing as an unambiguous result."

So my now question is: Who actually coined the term "iron law of econometrics," and what does it really mean?

4/4