Need help in probit model: probability that a developing country which is a WTO member faces a dispute.

[Leire Zain]

Hello! Hope you are all having a great week!

I need help to calculate the probability that a developing country which is a WTO member faces a dispute.

Pr(Dispute I)= F( I, exports or imports, GDP, year of dispute, controls)

For that I have 4 databases:

· Data base A: Disputes = Database containing the disputes done against developing countries coded according to different criteria
· Data base B: WITS database = Import/Export data about all the countries in the world.
· Data base C: Control Variables = Variables that are constant and unchanged for each country in order to test the relative relationship of the probability of facing a dispute and the import/export share.
· Data base D: WTO members = List of all the countries all the word saying whether they are or not WTO members and since when.

I already merge them in STATA and run a probit model with the following command:
probit dispute DImport DExport trade_share POWER lPOWER lPOWER2 EMBASISABROAD NONMILITARYGOVEXP GOVQUALITY GOVERMENTSYSTEM INSTITUTIONALFACTORS SHOCKS D_* if WTOmember==1

However, I am not sure how to interpret the results and how to answer the questions. I need that to further answer to this:

Policy Space = sum (Pr (Dispute I)) * Share i

Thank you very much for your time!

Leire

[FernandoRios]

Hi Leire
your question is very clear. You say you want to estimate a conditional probability of dispute, based on a probit model.
It seems you already got the probit model estimated. So, from there, you can do something like:
margins wto_member
or
margin, dydx(wto_member)
to determine, a) the probability of dispute of a WTO member country, or b) how being a WTO member county increases or reduces your probability of dispute. (assuming there are no other problems in the model).
Perhaps your trying to answer a different question?
Fernando

[Leire Zain]

Good morning Fernando!

yes, indeed, that is exactly the question i want to answer. I used your command but stata says that "WTOmember' not found in list of covariates r(322);". Do you know how can I add them as covariates?

Thank you very much for your time!