Competing Narratives

E-Day minus 36

This week the polls seem to have different and conflicting stories to tell. The Liberals are on the move. The Conservatives are sinking fast. Wait, the Conservatives are recovering. The NDP is sliding. Hold on a second, the NDP is still in the lead.

Here’s the weekly summary: the Grenier CBC Poll Tracker weighted average at the end of the week, with the numbers from the very beginning of the campaign in brackets, show NDP at 32% (33.2), CPC 29.7% (30.9) and LPC 29.4%(25.9). The NDP and the Liberals are virtually unchanged from last week, with a difference of .1% each. The Conservatives are up 2.0% points. If one simply averages the polls that reported during the last week without weighting them, the results are the same as Grenier’s weighted average at week’s end, NDP 32, LPC and CPC 30 each.

Grenier applies a weighting to each poll that he incorporates into his averages -- favouring Nanos, for example, and discounting Forum. His weighting is based on an assessment of their past performances and their polling methods. He also adjusts his averages with the publication of each new poll. This means that his totals are constantly shifting, creating headache-producing variations in the narrative. Better to check in once a week.

The swings in the narratives are largely the result of the fact that individual polls swing back and forth a bit, and sometime quite a bit. No single poll is precise. Studies of past results shows that no polling service nails it every time. Each polling company takes a turn getting it more or less right. And the media often further distorts the narrative by how they report and headline each individual poll.

The lesson? Don’t be concerned with a single poll result up or down, particularly the provincial or regional subsets. And don’t get preoccupied with Nanos' “daily tracking poll”. There is a case that other polls are more accurate most of the time than Nanos.

My own version of the narrative is this:

1. The Conservative core is holding. Harper still has an almost impossible task getting back to a majority. And time is beginning to run out.

2. The NDP and the Liberals are still locked in a tight battle for winning the most seats with the NDP correctly projected as ahead on that measurement. Some of the Liberal strengths especially in Ontario and to some extent in the Lower Mainland of B.C. are being felt. No one is breaking away yet. And they probably won't.

3. In the crucial “battleground” of Ontario there were wide variants in individual national polls. The average was CPC and LPC 35, NDP 28. That seems more or less accurate to me today. On those numbers in Ontario, as noted, the Conservatives will fall far short of a majority across Canada, and the NDP will win more seats in the House of Commons than the Liberals.

Let’s take a look at the "break away" theory. The idea is that in a close race suddenly a consensus seems to develop among the electorate, a click, a sort of group decision, and off everyone goes towards one party or another. 

My picture of this theory comes from the year I spent with my family in East Africa in 1972-73. While camping on the Serengeti, we watched a herd of gazelle stampeding across the African plain led by a single wildebeest. Only a few gazelle at the front of the herd would have known where they were running and why. Probably the wildebeest didn’t know either. He or she was likely trying to get away from the gazelle. But surge they did.

In 2011 Quebec voters collectively surged towards Jack Layton. Trudeaumania in 1968 looked like that. Maybe Rachel Notley benefited from a similar surge. In 1990 Ontario showed that in the last few weeks of an election campaign it could unpredictably break for the NDP. In this election as the theory goes, the anybody-but-Harper voters will as a group decide to pick one of the anti-Harper or “progressive” parties and move as a block in that direction. It is suggested that Ontario is poised to go one way or the other, likely for the new Trudeau. Generally the proponents of this theory are Liberals who live in the hope that voters will surge toward their candidate, sweeping them back into power and into their "rightful" role as the Government of Canada.

Could happen. Probably won’t.

If it occurs at all it would only happen in Ontario and possibly parts of the Lower Mainland. The CPC’s hegemony in parts of the Interior of B.C., Alberta and the Prairies, the NDP’s apparent lock on coastal B.C., parts of the Lower Mainland and Francophone Quebec and the Liberal hold on most of Atlantic Canada all seem impervious to that kind of sweep.

Ontario has a deeply embedded three-party tradition going back to 1945 and arguably 1919. During that time there has seldom been the kind of sweep any of the parties would need to win a majority other than in the unusual circumstances of the 1990s when the federal Liberals had the field to themselves. Now the party that hopes to sweep Ontario faces two other long-standing parties with significant strengths who will not cede the field easily.

The Conservatives in Ontario are quiet in this election buffeted by events and a sense of impending doom. They have, however, a strong core in half the ridings in Ontario and will not be pushed out easily. Their Ford-like core will hold, as it did for the embattled Tim Hudak, the Progressive Conservative leader in last year’s provincial election. The NDP is showing much more energy across the province this year than in the 2014 provincial campaign when it got 24% of the vote and 21 seats. Arguably, the NDP might yet catch fire in Ontario, building on its leads in Quebec and B.C. as it surges towards Ottawa.

Today, the tight three-way race for total votes and the projected seat profile shields from view that voting loyalties to party and leader are probably solidifying. At the finish line, the race may well look more like it does today than the one led by the Serengeti wildebeest.