Help with my Panel Data Analysis

[Akinyemi Afolabi]

Dear All,

I need help with my panel data analysis.
Topic: The Impact of Trade Integration on Poverty Reduction

Data properties: N: 15 countries, T: 10 years

The analysis I did
1. Hausman Test: Recommended Fixed Effect
2. There was no auto-correlation but there is Heterogeneity.
3. To resolve this, I used FGLS to finally analyse the data.

Question: Is FGLS suitable, and is it able to solve the endogeneity problem? If not kindly assist with possible suggestions.

Thank you

[Andrew Musau]

I would keep to the regular fixed effects estimator and wild bootstrap confidence intervals due to the small number of clusters. See

help wildbootstrap

[Akinyemi Afolabi]

I would keep to the regular fixed effects estimator and wild bootstrap confidence intervals due to the small number of clusters. See

help wildbootstrap

I have tried checking the wildbootstrap, I am using stata 15 and I have not been able to do that. Kindly help.

[Andrew Musau]

There is a community-contributed command boottest that does the same thing.

ssc install boottest
help boottest