[Maxence Morlet]

Professor Wooldridge, please correct if I'm wrong, but this is my understanding of the situation here:

The paper I mentioned, Hoechle (2007), provides a good overview of this issue: you have T>N, but T/N does not tend to infinity in your case, relatively far from it. Through Monte Carlo simulations, Hoechle (2007) demonstrates that when large T asymptotic properties cannot be achieved due to a finite number of time periods (as is the case here), Driscoll-Kraay (1998) standard errors outperform Beck and Katz (1995) panel-corrected standard errors.

This post

[Michael Lee]

This post

Thank you. got it.
But I have another problem yet: month is not regularly spaced: there are contemporaneous gap(s) across all subjects in id.
What should I do?

[Michael Lee]

Thank you. got it.
But I have another problem yet: month is not regularly spaced: there are contemporaneous gap(s) across all subjects in id.
What should I do?

Anybody?

[Carlo Lazzaro]

Michael:
what does the community-contributed module -xtscc- give you back?

[Michael Lee]

Michael:
what does the community-contributed module -xtscc- give you back?

Carlo:
exactly this error:

month is not regularly spaced: there are contemporaneous gap(s) across all subjects in id.
r(101);

[Maxence Morlet]

There are risks associated with this, but you could balance your data using

xtbalance

[Michael Lee]

There are risks associated with this, but you could balance your data using

xtbalance

It is not working in this dataset (deletes all of the observations).
it seems that as I mentioned before the -xtscc- is not working with unbalanced panel data. is it right?
So I have no choice rather than -xtgls- or -xtpcse-?

[Michael Lee]

It is not working in this dataset (deletes all of the observations).
it seems that as I mentioned before the -xtscc- is not working with unbalanced panel data. is it right?
So I have no choice rather than -xtgls- or -xtpcse-?

Can anybody help me?

[Miguel Gonzalez-Loureiro]

What is the advantage of -xtscc- over -xtpcse-?
In long unbalanced panel data?

just an example: xtscc seems not to allow factor variables or time operators (lagged variables, although one can circumvent this by creating manually the lag/lead ). But the issue of factor variables can be crucial for anyone interested on that. It does not seem to be the problem stated in this question, however.
xtscc can perform pooled OLS and fe. xtpcse is an OLS estimation. So I would check the poolability of data before choosing any of the xtscc or xtpcse, and perform the usual Breusch-Pagan test. Even, if you include dummy individuals, are significant their estimates (just to rule out fixed effects)?. Maybe the #4 suggestion should then be considered. And you should investigate whether it is AR1, AR2 or what lag.

If you use xtpcse, try the pairwise option after the comma. It could help with gaps.

[Michael Lee]

Is there any paper (other than -xtpcse-'s manual) that show that -xtpcse- considers (time? and id?) fixed effects by itself (for citation issues)?
Thank you in advance for your answers.

[Maxence Morlet]

See Beck and Katz (1995)