Below you will find the federal polling averages for the month of September. The averages combine four federal polls (Léger, Mainstreet, and Forum x 2) and one Quebec poll (CROP), altogether surveying 10,481 Canadians.
Compared to the August 2016 averages, the Liberals were up 1.8 points, the Conservatives were unchanged, the New Democrats were down two points, and the Greens were down 0.5 points.
Monthly tracking chart
The tracking chart below shows the monthly polling averages stretching back to January 2009. Elections and campaigns as well as the arrival of new federal leaders are also included.
You can click or tap on the chart above to magnify it.
The chart below shows how many seats each of the parties would have won in an election held in this month. This seat projection uses the current first-past-the-post system. For full methodology, see here.
The tracking chart below shows the maximum and minimum seat ranges (which are wider than the likely ranges above) projected for each party since the 2015 federal election.
Seat projections with alternate electoral systems
The chart below shows potential seat outcomes using alternative electoral systems.
In addition to first-past-the-post (FPTP), the chart shows estimations for proportional representation (PR) and alternative voting (AV).
For PR, each province retains the number of seats they currently have. The number of seats each party receives is rounded up or down according to the vote share received in each province, and any leftover seats are awarded to the party that finished in first place in the region.
A very simple calculation is done for AV. Because the Liberals and New Democrats tend to be each other's second choice, they are awarded any seat where they are projected to be in first place (along with the Greens). Any seat that the Conservatives or Bloc Québécois leads with 45 per cent or more is awarded to that party. Any seat where the Conservatives or Bloc Québécois is in first place but with less than 45 per cent is given to the Liberals, the NDP, or the Greens, depending on which of these parties was in second place.
Though a crude method, past experience with more sophisticated methods have yielded virtually identical results in the current political landscape.
These projections also assumes no change of behaviour by the parties based on the system in place, no change in the behaviour of voters, and no other parties on the ballot. All of these assumptions are likely to be greatly tested in any change to the electoral system.