FINAL SEAT PROJECTION

Is Our Projection Better Than Your Guess?

[OTTAWA – October 13, 2008] – Throughout this campaign, EKOS has been offering seat projections based on the results of our EKOS tracking poll. We have repeatedly said when we offered them that seat projections are inherently fraught with more difficulties than surveying public opinion because small shifts in relative party support can make big differences in seat distribution.

Nonetheless, as we have repeatedly said, pollsters, journalists and commentators often speak of one party “nearing a majority” or another “falling out of contention”, all based at least implicitly on a projection of what a certain level of popular support would translate into in terms of seats.

In that spirit, we offer our final projection of the campaign. Using a formal mathematical model, we have taken into consideration the patterns of support in even the smallest provinces, as well as sub-regions within the larger provinces.

We have also “tweaked” the numbers to take account of what we see as the relatively high level of commitment of Conservative and Liberal voters, and the concentration of support for the NDP and Greens among socio-economic groups that traditionally are less likely to vote.

So this is better than an educated guess, but something less than a prediction. It is a carefully reasoned conjecture.

Click here to download PDF: election-08-final-seat-projection-oct13