Hi dear researchers! I need your contribution to solve the heterogeneity problem in panel data analysis with 37 countries and 26 years. I want to examine the impact of external debt on economic growth in developing countries. Thanks
Use all of the Panel regression methods like the OLS, Fe and Re then compare between them by using special methods like Hausman Test to choose the suitable method for your data. There are 3 videos on YouTube shows how to apply the regression methods on Stat from A to Z
Mohammad Qabaja Thank you sir. I need to solve the heterogeneity problem as you know better that panel models based on the assumption of homogeneity, and ignore each cross country heterogeneity.
Swarupa Panigrahi Thank you but GMM can't considers cross country heterogeneity
Dear Muhammad,
The GMM makes use of the orthogonality condition to allow fir efficient estimation in the presence of heteroscedasticity.
In that case, you can try heteroscedasticity-consistent OLS approach.
Kindly Google for more information.
Regards
If you're sure the underlying data set is a heterogeneous panel, detectable by unit root tests, you can employ the PMG, MG or DFE heterogeneous models based on the circumstances. The PMG for example assumes long run homogeneity but short run heterogeneity and the MG assumes both short and long runs heterogeneity. I hope this helps.
Thank you Kehinde Mary Bello Adamu Braimah Abille . In my case N is greater than T. So in this case i dont know how to consider the heterogeneity
First conduct the herogeneity test (eh Pesaran and Yamagata, 2008 homogeneity) to establish whether the panel understudy is heterogeneous or homogeneous. If there is heterogeneity use Econometric models the AMG, DCCEMG, PMG etc that are robust to heterogeneity
- Badges
- Science topic