[Joao Santos Silva]

Dear Guest,

If these are not included, there you will not have any fixed effects. Notice, however, that the fixed effects should not be included explicitly as regressors, but absorbed with the a() option (please check the help file). The results will be the same, but estimation is much faster and results will be easier to interpret. Note also that some of your variables are likely to drop when you used fixed effects in this way; that is because they become redundant when you have fixed effects.

Best wishes,

Joao

[Guest]

Thank you very much.

[Guest]

Dear Guest,

If these are not included, there you will not have any fixed effects. Notice, however, that the fixed effects should not be included explicitly as regressors, but absorbed with the a() option (please check the help file). The results will be the same, but estimation is much faster and results will be easier to interpret. Note also that some of your variables are likely to drop when you used fixed effects in this way; that is because they become redundant when you have fixed effects.

Best wishes,

Joao

Joao Santos Silva Professor, how can I fixed those redundant problem? Two of my variables Time and Importer fixed effects were omitted when I tried ppmlhdfe with a().

[Joao Santos Silva]

Dear Guest,

If you include the variables in a() they should not be included also as regressors, and if they are they will be omitted because otherwise they would enter twice.

Best wishes,

Joao

[Guest]

Joao Santos Silva I am sorry for making you confused. My project is the impact of importing countries' (20 countries) non-tariff measures on Myanmar's Agricultural Exports between 2010-2022 (Year). In which, agricultural production variable is incorported with time lag1. Regarding that when I run ppmlhdfe a(Year Importer), the distance variable and production variables become omitted. Could you please guide me how can overcome it?

[Joao Santos Silva]

Dear Guest,

Thanks for clarifying. If I understand correctly, Myanmar is the only exporter in your sample. If that is the case, Myanmar's production will be captured by the year dummy, and distance by the importer dummy, and that is why they have to drop.

Best wishes,

Joao

[Guest]

Joao Santos Silva a follow up question. So, should it be (1) ppmlhdfe Export LnGDP LnDistance LnPOP LnProduction LnTariff NTM (#no a() command) (2) ppmlhdfe Export LnGDP LnDistance LnPOP LnProduction LnTariff NTM a(Year Importer) (#production and distance are omitted) (3) ppmlhdfe Export LnGDP LnPOP LnTariff NTM a(Year Importer) (#dropping distance and production variable) or else.....? Thank you.

[Joao Santos Silva]

Dear Guest,

2) and 3) should give you exactly the same results. Whether this is better than 1) it is up to you to decide :-)

Best wishes,

Joao

[Guest]

Dear Guest,

2) and 3) should give you exactly the same results. Whether this is better than 1) it is up to you to decide :-)

Best wishes,

Joao

Professor Joao Santos Silva Q.1 is it possible to use the (1) as it does not include any fixed effects? Q.2 Since it is only one exporter, can I just neglect importer fixed effects and only use Year fixed effects? Thanks again.

[Joao Santos Silva]

Dear Guest,

I am sorry, but only you can decide on that.

Best wishes,

Joao

[Guest]

Thank you Professor.