The Legend of Log(NAICS)
This Harvard Professor Doesn't Understand Basic Math. Or, even worse, he does.
This story is perhaps the single most legendary piece of lore in the academic economics/finance profession, but nobody has ever written about it, and few people beyond a small circle of economics and finance professors are aware of it.
Yes, it is a funny inside joke, but this is not merely a story about one absurd regression variable. This episode has become shorthand for corruption and exemplifies much of what is wrong with academic prestige, data misconduct, and the incentives at play in the finance job market: who is scrutinized, who is protected, and how serious errors can disappear if your advisor has a Nobel Prize.
What makes it so striking is the gap between the error and the status of the scholar involved. At best, using log(NAICS) was an elementary econometric mistake. At worst, it’s outright fraud. Either way, the person at the center of this scandal now teaches econometrics at Harvard.
