I want to combine Canadian census data with provincial and municipal voting data. I have voting data, which is sub-divided into polls, which map nicely to Forward Sortation Codes (FSAs)
I can find Canadian census data by FSAs for 2016, but not for 2021. Should I stitch together Dissemination Areas (DA)? Does a nice mapping exist?
Please boost to me connect with the right person. Thank you!
@benevolentprof The StatCan CFSAs (Census FSAs) are stitched together from DAs. I have not seen a file with the mapping DA->CFSA, but it *should* be fairly straight-forward to produce that by spatial match.
Another option would be to keep it at the polling district level and try to match that up with census data.
And I am guessing you would then run an ecological inference model for the analysis.
@vb_jens Thanks for the reply!
I like the second option. I'm going to map DA->Polls, as the municipal polls don't map cleanly to FSAs. It's also more extensible for federal elections.
Not familiar with ecological inference. I'll look it up. My plan was to start with some regression and feature engineering and see how far that gets me.
Susan Sim, Ph.D.
Jens von Bergmann