areg or xtreg, recover s.e. of absorbed fixed effects

[desmochada]

I am wondering whether it is possible to recover the standard errors for the absorbed fixed effects when using areg or xtreg (or xtivreg2). It is straightforward to recover the *coefficient estimate* using postestimation commands, but I am also looking to recover an estimate of the standard error for each coefficient.

Since there are many fixed effects and the sample size is large, it is infeasible to do this using STATA factor variables.

Any and all help is greatly appreciated.

[Sebastian Kripfganz]

The easiest way of getting fixed effects estimates and their standard errors is using simply the regress command with dummy variables for all units. However, as you mention that your (cross-sectional) sample size is large (relative to the time dimension) those estimates will not be consistent, neither will be their standard errors (irrespective of the command you are using to recover the estimates).

[desmochada]

Thanks for the speedy response. As mentioned in the original query (somewhat obliquely), the dataset and number of fixed effects are large enough that it is infeasible to use regresss + dummy variables for all units. Which is why I am asking if there is another option.

[Sebastian Kripfganz]

What do you mean exactly by "infeasible"? Does Stata issue an error message? In any case, the main point I wanted to make is: Even if you find a command that does deliver estimates and standard errors for the fixed effects in your case, they are by no means reliable exactly because you have such a large number of fixed effects parameters.

[FernandoRios]

I dont think that is possible.
Most strategies that "deal" with fixed effect methods do dont estimate those parameters directly, and as such, they do not estimate their corresponding Standard errors.
HTH,
Fernando