Hello,

I am trying to understand how binscatter plots work and I have a somewhat general question in case of panel data. For example, let the variables be a, b, and c, for the period from 1990-2020. I would like to do a binscatter of the variable a against the average value of variable b by municipality for each decade, controlling for variable c, something like this:

by municipality decade, sort: gen b_m=mean(b)
binsreg a b_m c, nbins(15) if decade==1990

Given that I want on x axis to have the average value of variable b by municipality, does it make sense to run a simple binsreg like this "binsreg a b_m c, nbins(15) if decade==1990" or do I also need to compute the average values of both of b and c variables before doing binsreg?

Thank you very much for your help.