Bump
De Loecker and Eckhout go on all out attack
-
non-judgemental outsider question: is IO as a field particularly infamous for this kind of behavior? or just a coincidence with these two cases?
No coincidence. Empirical IO is very clubby, run mostly by those running the NBER IO group and coeditors at top journals. There are lots of bullies from top schools, mostly HRM assistant professors. Referee reports are nasty and aggressive against LRM; extremely permissive in favor of HRM. Seniors, including coeditors at top journals, allow and promote this behavior. What happened with De Loecker-Eeckhout, and Schiraldi-Seiler are two recent examples that went mainstream. This happens often. There are worse examples dealing with harassment and discrimination. Nasty field.
-
EJMR has become a forum for ULRM to vent their frustration and for HRM to be reassured of their intellectual superiority
The Jans backed down; not quite an apology, but close enough.
The Jans did good just before Santa arrived, to get some XMas present.
Santa forgets, EJMR does not. -
The Jans have a new set of thoughts
http://www.janeeckhout.com/wp-content/uploads/Thoughts.pdf
They are now much more respectful of Traina in footnote 1. And they discuss this weighting business. It doesn't seem to matterthe weighting business matters, and they are wrong. you can weight by sales, but please don't call it the "benchmark aggregate markup". weighting by sales, you are squaring the numerator, so what you get as "aggregate" reflects markup dispersion across firms. it does not make any sense. jans still to toilet.
-
But they claim they are weighting properly.
The Jans have a new set of thoughts
http://www.janeeckhout.com/wp-content/uploads/Thoughts.pdf
They are now much more respectful of Traina in footnote 1. And they discuss this weighting business. It doesn't seem to matterthe weighting business matters, and they are wrong. you can weight by sales, but please don't call it the "benchmark aggregate markup". weighting by sales, you are squaring the numerator, so what you get as "aggregate" reflects markup dispersion across firms. it does not make any sense. jans still to toilet.
-
they (the jans) are wrong, I assume deliberately at this point.
try it yourself. assume the simplest possible economy with homogeneous labor only as input, so that the labor share is a function of the markup only. then aggregate across firms. you'll find that the aggregate labor share is a function of the cost-weighted average markup.