I have a dataset where the DV is asylum rates and the unit of analysis is directed dyad year where country A is the country of asylum and country is b is the country of origin of asylum-seekers. The DV is a rate/percentage of those from country B granted asylum. I employed a GEE regression, clustered around the dyads, and used a logit link with a correlation of exchangeable.
I ran the margins command for one of the variables, autocracy (if country A is an autocracy or not, binary variable). I used the following command:
However, I noticed the expression.
Why is the expression saying, Pr(asylum_prop != 0) ?
This happened with any of the independent variables that I used. Did I do something incorrectly?
I ran the margins command for one of the variables, autocracy (if country A is an autocracy or not, binary variable). I used the following command:
Code:
margins, dyex(autocracy)
Code:
Average marginal effects Number of obs =29880
Model VCE : Robust
Expression : Pr(asylum_prop != 0), predict()
dy/ex w.r.t. : autocracy
Delta-method
dy/ex Std. Err. z P>z [95% Conf. Interval]
autocracy -.0047661 .0011086 -4.30 0.000 -.0069388 -.0025933
This happened with any of the independent variables that I used. Did I do something incorrectly?
Comment