Hello everybody!
I need your help today.
I try to estimate anticipation on trade at announce of brexit. So, I used this gravity model and estimate for two periods : 1995-2015 and 2016-2021.
My model is : ppmlhdfe x l_pib_i l_pib_j l_pop_i l_pop_j l_dist comlang_off colony contig eu eu2 uk_eu eu_uk exp_uk* imp_uk* if ......, a(exp_id imp_id t) cluster(pair_id).
Here, exp_uk* are dummies for each country pair between uk exporter and european importer country, reverse to imp_uk.
So i have two questions:
- First, how can i interprete the evolution of dummies coefficients? For exemple, if the coefficient of uk_eu was 0.3 before annoucement then 0.1 after, can i say trade between uk and european union have decreased of : 100*(exp(0.1)-1)-100*(exp(0.3)-1) ?
- Second, i want to do joint hypothesis test between these two regression to see if coefficients are significantly different between two regression. But, when i try to do, stata say me that standards errors need to not be robust. Or, i can't run ppml with not standard error robust. How can i do?
Thank's for your help.
Romain Andre
I need your help today.
I try to estimate anticipation on trade at announce of brexit. So, I used this gravity model and estimate for two periods : 1995-2015 and 2016-2021.
My model is : ppmlhdfe x l_pib_i l_pib_j l_pop_i l_pop_j l_dist comlang_off colony contig eu eu2 uk_eu eu_uk exp_uk* imp_uk* if ......, a(exp_id imp_id t) cluster(pair_id).
Here, exp_uk* are dummies for each country pair between uk exporter and european importer country, reverse to imp_uk.
So i have two questions:
- First, how can i interprete the evolution of dummies coefficients? For exemple, if the coefficient of uk_eu was 0.3 before annoucement then 0.1 after, can i say trade between uk and european union have decreased of : 100*(exp(0.1)-1)-100*(exp(0.3)-1) ?
- Second, i want to do joint hypothesis test between these two regression to see if coefficients are significantly different between two regression. But, when i try to do, stata say me that standards errors need to not be robust. Or, i can't run ppml with not standard error robust. How can i do?
Thank's for your help.
Romain Andre
