clustering standard errors

[kamble vikrant]

Hi, This question may not be related to STATA but econometrics in general.
Still, I am asking to seek help here because I want to incorporate this information into the STATA command.

My dependent variable is the State level but the independent variable is the district level. Districts are part of states. In this case, should I cluster the errors district level or state level?

Should the STATA command be :
reghdfe Y X, absorb (stated year) cluster(stateid)
Or
reghdfe Y X, absorb(districtid year) cluster(districtid)


Thank you.

Regards,
Vikrant

[Bruce Weaver]

I guess reghdfe is a user-contributed package. I don't know anything about it. But your post reminded me of these papers, which talk about how to handle things in a multilevel model when the outcome variable is at level 2 rather than the usual level 1. Perhaps these papers will give you some useful ideas.

Croon, M. A., & van Veldhoven, M. J. (2007). Predicting group-level outcome variables from variables measured at the individual level: a latent variable multilevel model. Psychological methods, 12(1), 45.

Foster-Johnson, L., & Kromrey, J. D. (2018). Predicting group-level outcome variables: An empirical comparison of analysis strategies. Behavior Research Methods, 50, 2461-2479.