Dear Joao Santos Silva Tom Zylkin ,
Currently, I'm doing an investigation about the effects of Free Trade Agreements on Peruvian trade (imports and exports) in the period 2000-2018. For that purpose, I'm using the gravity model as a benchmark and the PPML as the econometric technique. However, some problems raised with my model specification. In particular, I have problems with the variable LnDistance since it is the same for each pair of countries every year (distance between countries does not change): I have a singular matrix. So, I have some questions.
1) Is correct to use the PPML technique with a dynamic panel? Is there a stata comand for that?
2) If I include fixed effects for each exporter-importer pair, can I overcome the problem of the singular matrix (i.e. the fixed effect absorb the variable LnDistance)? Is it correct?
Currently, I'm doing an investigation about the effects of Free Trade Agreements on Peruvian trade (imports and exports) in the period 2000-2018. For that purpose, I'm using the gravity model as a benchmark and the PPML as the econometric technique. However, some problems raised with my model specification. In particular, I have problems with the variable LnDistance since it is the same for each pair of countries every year (distance between countries does not change): I have a singular matrix. So, I have some questions.
1) Is correct to use the PPML technique with a dynamic panel? Is there a stata comand for that?
2) If I include fixed effects for each exporter-importer pair, can I overcome the problem of the singular matrix (i.e. the fixed effect absorb the variable LnDistance)? Is it correct?
Comment