[CAO DUC SON]

Dear professor,
1. I have 5.8% missing data of GDP variable (source WB)
Should I manipulate the missing GDP by adding from the other source, but it may take much time
Or, if I just let it as the omitted variable is ok?

2. When I ran the PPML with importers, exporters fixed the model indicated some fixed effect are omitted because of collinearity=> is it ok

thank you so much!

[CAO DUC SON]

after 1 mins I have the result but I am confusing about the collinearity. #16. is it ok?
Thank you!

[Joao Santos Silva]

Dear CAO DUC SON,

On the GDP question, of course it would be better to get the missing data, but it may be OK just to drop the observations with missing data.

Dropping some fixed effects is normal but, depending on how you specified your model, may make the results more difficult to interpret. I suggest you use ppmlhdfe rather than ppml; the results should essentially be the same, but are easier to interpret.

Best wishes,

Joao

[CAO DUC SON]

Dear professor @Joao Santos Silva Thank you so much!