areg test for homoscedasticity and autocorrelation

[Joshua Joseph]

Hello, I am new to Stata and econometric analysis.

In my Master Thesis I run a fixed effects regression on country level panel data. I would like to analyse the influence of taxes (EATR) on FDI. Additionally I got three control variables: GDP, CPI (Corruption Perception Index), average wages. On the recommendation of my supervisor, I used "areg" to run fixed effects regression. I would like to test for homoscedasticity and autocorrelation. As far as I know, you can't use xttest3 or xtserial in an "areg" setting. Is there any alternative?

Thank you, I look forward to your contributions!

Joshua

[Carlo Lazzaro]

Joshua:
I'm surely out of topic here, but why using -areg- when you have -xtreg,fe- available in Stata?

[Joshua Joseph]

Carlo Lazzaro Thank you for your answer. I have used "xtreg, fe" before but after a consultation with my supervisor she suggested to use "areg" instead. She argued that "areg" is more suitable, because my dataset has many groups and the number of groups is not growing when I increase the sample size.

Best Regards

Joshua

[Carlo Lazzaro]

Joshua:
thanks for clarifying.
I am not aware of any community-contributed modules to do what you're after.
I'd compare the default vs the cluster robust standard errors and make a decision after that.

[Joshua Joseph]

Thank you for your advice!

Best Regards

Joshua

[George Ford]

cluster(Country_ID)

You'll end up doing it anyway. It corrects for both hetero/autocorr, so you're done.

[Joshua Joseph]

George Ford Thank you for the reply!