I am using Stata 13 and have the most updated version of reghdfe. My read of the help file for reghdfe is that the DoF calculation is the same as that used for xtreg however I fail to get the same se's. I have tried several of the reghdfe options for adjusting the DoF to try to get the same se's as xtreg but am unable to replicate.
xtreg does not drop singleton observations and reghdfe does so I used the "keepsingle" option to see if that was resulting in different se's but was still unable to replicate xtreg results. The se's for reghdfe are smaller than xtreg so I am inclined to stay with xtreg to be more conservative. The speed of reghdfe, however, is superior so if possible I'd like to replicate the xtreg se's in reghdfe and use this user generated command because I have to run my model several times for subgroup analysis.
Hoping someone else has stumbled across this issue but my initial search on the listserve didn't address this exact issue.
xtreg does not drop singleton observations and reghdfe does so I used the "keepsingle" option to see if that was resulting in different se's but was still unable to replicate xtreg results. The se's for reghdfe are smaller than xtreg so I am inclined to stay with xtreg to be more conservative. The speed of reghdfe, however, is superior so if possible I'd like to replicate the xtreg se's in reghdfe and use this user generated command because I have to run my model several times for subgroup analysis.
Hoping someone else has stumbled across this issue but my initial search on the listserve didn't address this exact issue.
Comment