This is a quick post to note a minor modification to the model. The relevant methodology post has been updated. I have also revised the first two projections (here and here) in order to maintain comparability going forward.
Addendum: I have also updated the adjustment for Beauce based on this Mainstreet poll. (Minor changes to Quebec regional adjustments were also made.) This update will be in effect as of the next projection, but it wouldn't change the current projection anyway.