The evolution of vote intentions during the 2015 Canadian election campaign

By André Blais and Jean-Michel Lavoie
Université de Montréal

Campaign dynamics

Support for the Conservatives remains remarkably stable during the whole campaign, hovering between 28% and 32%. The Conservatives reached their bottom, at 27%, in late August, possibly because of the Duffy trial. The drop in Conservative support is observed only in Ontario, where Conservative vote intentions decrease from about 37% before the appearance of Wright (August 12) to about 27% by the end of the month. By mid-September the Conservatives were back to about 33% nationally, suggesting that the impact of the Duffy trial had gradually evaporated.
For the rest of the campaign the Conservatives remain at slightly above 30%, with little movement. There is some shift in Quebec, however, as their support increases in the second half of September. Note that the Conservatives also made some gain in British Columbia during the campaign.
Things are quite different for the Liberals. They start at about 27% and end at 38%, an increase of about 10 points. There is a small increase in Liberal support in late August, which is most apparent in Ontario, from about 27% to about 30%, possibly related to the Duffy trial. Liberal support stays at about 30% for the whole month of September. Liberal vote intentions more or less constantly and progressively rise throughout October.
NDP support, for its part, stays around 32% from the start of the campaign to mid-September. This is followed by a very gradual and constant decline throughout the last five weeks of the campaign.

The niqab issue

The big question is whether the NDP decline was the result of the niqab issue. If so, we should observe that the NDP drop was strongest in Quebec since this is where the issue was most strongly debated. That seems to be the case: support for the NDP in Quebec went from about 43% in mid-September to about 25% by election day.
In order to conclude that the niqab issue was the crucial factor in the NDP decline, we would also have to observe that the decline started earlier in Quebec. That seems to be the case as well. The NDP descent in Quebec starts around September 18, which is the day the Conservatives announced their appeal; the party’s descent in Ontario begins only in early October. The evidence suggests that the niqab had a big influence in Quebec, producing a drop of about 10 points for the NDP in late September. Logically, that loss should have been to the benefit of the Conservatives and the Bloc, which were taking an anti-niqab stance. Is it the case? The party that makes the biggest gain in Quebec during that period is the Bloc (from 14% to 20%).

The Liberal surge in October

The last question is: How can we explain the NDP’s continuous decline throughout October, to the benefit of the Liberals? The fact that the movement is very gradual suggests that this is not due to any specific crucial event.
One interpretation is that many voters, especially in Ontario, abandoned the NDP for the Liberals as it became obvious that only the latter could defeat the Conservatives. We cannot directly test that interpretation with these data. We observe, however, that the Liberal gains in Ontario during the month of October come at the expense of the Conservatives as well as of the NDP, which suggests that it was simply the image of the Liberal party and of Justin Trudeau that was constantly improving in the last three weeks of the campaign.

Conclusion

Two major conclusions can be drawn from this analysis. First, specific campaign events do not seem to account for most of the campaign dynamics. The Liberal gains and the NDP losses appear to be mostly gradual, as if Canadians came to like the Liberals a little bit more and the NDP a little bit less as the campaign progressed. The Duffy trial had a temporary effect in Ontario but that is all. There is no trace of any significant shift after any of the leaders’ debates or after the Liberals’ announcement, on August 27, that they would run a deficit if elected, or when the story broke about the drowned Syrian boy (September 2). The only exception is the niqab affair, which had a big impact in Quebec, to the detriment of the NDP.
The second conclusion is that while the Liberals gained ground and the NDP declined during the campaign in each of the three provinces, there were substantial differences in campaign dynamics across the provinces. The shifts were strongest and most gradual in Ontario. In Quebec, the Conservatives and the Bloc also made some modest gains when the NDP support began to tumble. And in BC, the Liberal surge was particularly modest and late.
NB: We thank Claire Durand for sharing the poll data. We thank her and Laura Stephenson for comments on a previous draft.

Figures 1 to 13 present the evolution of vote intentions during the 2015 Canadian election campaign, starting August 2, until October 18, the day before the election.
Daily estimates correspond to the mean of all the polls that were in the field that day.

figure 1

graph1

figure 2

graph2

figure 3

graph3

figure 4

graph4

figure 5

graph5

figure 6

graph6

figure 7

graph7

figure 8

grpah8

figure 9

graph9

figure 10

graph10

figure 11

graph11

figure 12

graph12

figure 13

graph13

This entry was posted in Blog. Bookmark the permalink. Both comments and trackbacks are currently closed.