Why don't applied/quasiexperimental econ papers include simulated data to show you how to think about the DGP? Eg. here's what we would see if treatment effect is real, and here's what it would look like if it's all selection. #econsky

Are there data-driven methods to tell when you have too many fixed effects? #EconSky

Are there any results for event studies with unbalanced panels? #econsky

Are there data-driven methods to tell when you have too many fixed effects? #EconSky

So you've never seen the reverse ordering like this?

Why would the context go from positive (+5) to negative (-5)? Look at the x-axis.

Is it standard to report event study or distributed-lag coefficients from +5 to -5? #EconSky

twitter.com/michael_wieb...

Should you "trust literatures, not papers"? I replicated the literature on meritocratic promotion in China, and found that the evidence is not robust.

twitter.com/michael_wieb...