Closed ChristianZimpelmann closed 5 months ago
Great catch, thanks a lot!
Despite my comment in #412, I think individual-level is the way to go now. It is ultimately what matters in many cases and anything else would be introducing lots of complexity on the GETTSIM side.
At some point, we need a data preparation guide... There we could explain and make suggestions like trying different ways for distributing wealth measured at some group level.
Bug description
vermögen_bedürft_bg
however this grouping is endogeneous. Also, it is used forhh
-wide transfers like Wohngeld which seems incorrect ifbg
andhh
differ. Probably should bevermögen_bedürft_hh
(which we somehow need to distribute on Bedarfsgemeinschaften) or we request individual-level wealth?See also #412