svyset , svy, store results

[Chiara Tasselli]

Good evening everyone,

I am working with a dataset which has individual weights. I would like to calculate the regional wage gap for descriptive statistics.
Without weights, I would have used the following command:

bysort year region : egen meanf = mean(cond(Female == 1, retribuzione_sum, .)) 
bysort year region : egen meanm = mean(cond(Female == 0, retribuzione_sum, .))
gen diffI = (meanm - meanf) / meanm

How can I achieve a similar outcome using svy? My main issue is generating a new variable that contains the wage gap or even just exporting the table whit diffI for each region and year.

Many thanks in advance for your help and time

[Clyde Schechter]

Although -egen- does not support -svy:-, it is not hard to calculate weighted means with it.

bysort year region: egen numf = total(cond(Female == 1, pwt*retribuzione_sum, .))
bysort year region: egen denf = total(cond(Female == 1, pwt, .))
gen wtd_meanf = numf/denf

and analogously for males. (Evidently, replace pwt in the above by the actual name of the pweight variable.)

[Chiara Tasselli]

Clyde Schechter Thank you very much for your response; it was extremely helpful!