Hi all,
I'm trying to instrument a fixed effect model with variable Z, I'm using ivregdhfe from SSC
The code I'm using is:
The error I get is:
I try to run the command without cluster, and it executed succesfully so I suppose the problem is with cluster(). My occupation variable takes 400+ values, my state variable takes 47 values. How should I cluster at occupation-state level then?
Thank you!
I'm trying to instrument a fixed effect model with variable Z, I'm using ivregdhfe from SSC
The code I'm using is:
Code:
ivreghdfe Y (X= Z), absorb(i.state#i.occupation i.state#i.year i.occupation#i.year) vce(cluster i.occupation#i.state) first
Code:
(MWFE estimator converged in 29 iterations) factor-variable and time-series operators not allowed
Thank you!
Comment