I am in Cologne Germany at GESIS teaching a five day workshop on diff-in-diff. On the first day, roughly four or five people requested that I cover continuous diff-in-diff. As I hadn’t planned to, I’ve been trying to get the last part of the Callaway, Goodman-Bacon and Sant’Anna paper prepped. I’m basically having a hard time with my attention span lately, and keep getting distracted when I get to their TWFE decompositions and with the new nonparametric estimator. But something I wanted just mention today had to do with what I think is true, and that’s that Jerzy Neyman’s original notation for potential outcomes was actually not based on binary treatments. If I’m not mistaken, his 1923 Polish masters thesis was about continuous treatments.
Today I’ll try to be really brief as this really is just a passing thought as I prep. Neyman’s masters thesis in 1923 (which was only translated into English in 1990 I think), from what I have learned, makes at least two contributions. One is…
Keep reading with a 7-day free trial
Subscribe to Scott's Substack to keep reading this post and get 7 days of free access to the full post archives.