EKOS VERSUS THE ACTUAL RESULTS
“The Liberal Party of Canada is poised to return to power with at least a clear minority and possibly a majority.”
-EKOS’ final prediction, October 18th, 2015
[Ottawa – October 20, 2015] After an exciting 78-day campaign, the 42nd Canadian general election has come to a close. We at EKOS believe we did a very good job in charting the direction of what was a rollercoaster campaign. In the end, we correctly predicted that the Liberals would return to power (and we noted the possibility of a majority), although we acknowledge that our final estimate of the Liberal support fell slightly outside the margin of error. We were, however, spot-on in terms of the final Conservative, NDP, and Bloc numbers, although we overestimated Green Party support.
Areas of uncertainty
In our final release, we identified five areas of uncertainty. Now that the final election results are in, we believe that we are able to address the issue of how these potential sources of error played into the final results.
1. Will the Liberals trend further upward?
As noted by a number of pollsters, there was a last-minute surge in Liberal support. In our final release, we clearly noted that the Liberals might trend upward even further. It appears that this was indeed the case, particularly in Quebec and British Columbia.
Our final poll was based on a three-day roll-up (Friday-Sunday), but if we were to drop the Friday results and restrict our analysis to the final two days of the campaign, we see a large uptick in Liberal support. Indeed, on Sunday, we had the Liberals at 42 points. However, we did not report on these smaller roll-ups for two reasons:
- To meet Sunday’s reporting deadlines, we had to shut down calling at 4:00 pm EST which, given Canada’s time zone differences, gave us a calling window of just two hours in some regions. We simply had too few cases in Western Canada to offer a robust dataset and we felt that it would have been irresponsible to use these figures as our final prediction.
- We feel that it is inappropriate to report on multiple predictions and then, after the election, root around in the bin for the most accurate set.
2. Will cellphone-only households actually show up?
As always, we are going back and conducting a post-election survey to look at the demographic differences in turnout but, based on Canada’s much higher than usual turnout (~68 per cent), it is clear that cellphone-only households did indeed show up to the polling stations.
3. Is the late senior conversion to the Liberals genuine?
Again, we will be better positioned to answer this question once our post-election survey is complete, but we would speculate that the senior conversion to the Liberals was indeed genuine.
4. Why are Liberals higher with a live interviewer than with the robot?
We do not have a clear answer to this question at this time, but we will be reporting back on this soon.
5. How will Quebec and British Columbia conclude?
In our final release, we warned that Quebec and British Columbia were highly confused and unclear races. While we had excellent regional representation in this election, particularly in Ontario and Atlantic Canada, we underestimated Liberal support in British Columbia and Quebec and this is ultimately what threw off our final prediction.
In closing, we would like to thank the 124,425 Canadians who generously responded to our surveys since January. We would also like to acknowledge sincere appreciation of our partners at iPolitics who did a great job in covering this election. We would also like to thank those who have provided advice, comments, and feedback throughout the election, particularly in the world of social media. Finally, we congratulate all those who participated in the political process and recognize the passion and effort that they bring to Canadian democracy. Justin Trudeau and the Liberal Party are to be congratulated on having won a clear and decisive mandate to lead the country.
Click here for PDF version of this article: Full Report (October 20, 2015)