Mr. Trudeau and Mr. O'Toole are attempting to accomplish two different tasks. Mr. Trudeau wants to win a majority government while Mr. O'Toole wants to replace the Trudeau government with a government lead by him and if he could do that with a majority of his own he would love that.
It should come as no surprise that Mr. Trudeau has the easier task. That would seem to be counterintuitive looking at the polls for the past week but if you look past the top-line numbers it does seem to be true at this point in the campaign.
If you look at all of the polls, without focusing on just one, you will note some interesting results.
In BC the polling averages indicate that if an election were held today the Liberals could gain up to 8 seats. For the sake of this exercise let's split the difference and call it 4 seats.
On the Prairies it is 7 additional seats so let's call it 3.
In Quebec the Liberals could win up to 20 additional seats so let's call it 10.
In the Eastern Provinces they could win 4 additional seats so let's call it 2.
So when the election was called the Liberals had 155 seats. Adding the number of possible additional seats to that gives you: 155+4+3+10+2=174 seats. A slim majority.
Looking at Ontario the polls indicate that the Liberals could win anywhere from 60 to 80 seats. Again, let's split the difference which means, 70 seats, 7 seats off of their current total in that province.
So when we take the seat gains above and apply the Ontario losses we get: 174-7=167, a strong minority government. If the Liberals just beat the mean in a couple of regions they find themselves back over the majority government mark and beating the mean significantly puts them into comfortable majority territory.
Let's contrast that with the task Mr. O'Toole has ahead of him.
The polls indicate that if an election were held today the Conservatives would lose between 6-12 seats in the West. As above we will split the difference and call it 9 seats. As well, it would appear he could lose up to 4 seats in the Eastern Provinces. We will call it 2.
When the election was called the Conservatives had 121 seats. So looking at their losses that gives the Conservatives: 121-11=110 seats. Looking at where they might gain the current polls are indicating that they may gain up to 5 seats in Quebec, let's call it 3. As well, they may gain from 5 to 15 seats in Ontario which we will call 8 additional seats.
So taking their totals with all gains and losses accounted for we have: 110+11=121 seats, essentially back to where they started.
As an aside, it is interesting that the only part of the country where the Liberals are down is Ontario. In all other provinces and regions in the country the Liberals are either at their 2019 levels or higher while the Conservatives are mostly at their 2019 levels or lower. I would be curious to know why Ontario is the outlier here. Are they a harbinger of things to come in the rest of the country? Or will they eventually come back into line with the rest of the country in the last weeks of the campaign?
So just looking at these numbers it is obvious that after two weeks of the campaign, dominated by bad public polls for the Liberals, the Liberals are still closer to achieving their objective than the Conservatives. Currently the Liberals can afford to lose some seats in Ontario as long as they can pick up seats in the other regions and provinces and still obtain their majority. At the very least they, if the election were held today, they would retain government. The Conservatives on the other hand would need to virtually sweep Ontario and Quebec, while limiting his losses elsewhere, to win government.
Could the Conservatives change the math by September 20? Of course they could. Is it probable that they will change it? I would say no at this point.
3 comments:
I know you're a lot more bullish on the Liberals' chances than I am, but what's your take on the "summer cottage theory" as an explainer for why the Liberals' vote is softening in Ontario?
https://scrimshawunscripted.substack.com/p/canada-reset-ontario-polls-and-can
Basically the idea that, since most of the polls that have come in which were bad for the Liberals were IVR robopolls conducted by telephone, they may not be picking up on a segment of the population who are away for the end of summer (and might not be as likely to vote Conservative).
I mean, I personally find it to be a stretch and an example of "copium," but there are at least some questions to be had as to why, in contrast to the IVR polling, the live telephone rolling polls (Nanos) show a relative tie, while the online panels that have been sparsely conducted so far have still shown relative strength for the Liberals. And more so as to why the polling of individual ridings done by Mainstreet show a rosier picture for the Liberals than their national daily trackers. None of the discrepancies make any sense.
The only certainty right now is volatility, but I still want to believe that what the Liberals are saying is true, that things may pick up for them after Labour Day with the release of their platform and the debates.
I will say, however, that the Liberals have done a poor job communicating their accomplishments and policy proposals thus far, and that seems to be a weakness of theirs. The CPC and NDP are winning the air war, which at least lends credence to the notion that the Liberals are not doing as well as they had hoped. The Liberals have plenty of announcements but the media is ignoring them. They have no ads.
LOL, Mr. Scrimshaw and I came to the same conclusion independently.
I believe he may be oversimplifying it a bit. Certainly, polling in the last two weeks of the summer would probably kick the hell out of a pollster's response rates, possibly introducing bias but there are many other factors that could be involved.
Statistical issues such as the quality of the frame, sample size and design, questionnaire design, collection strategy, edit and imputation strategies and potential bias introduced in the data analysis stage could account for the wildly different estimates from the pollsters. Survey work is as much art as science that, to be done properly, takes time and meticulous attention to detail. That can be the exact opposite of how political polls are done.
Also let's remember that media organizations using a specific pollster are probably not paying the pollsters or at best are only paying them on a cost recovery basis. As for pollsters that do not have a media sponsor they are paying for the polls out of their own funds, which means they would be doing it on the cheap. In either case the pollsters are not spending too much to conduct and publish their polls and as the old saying goes "you get what you pay for".
I have other suspicions but I am not going to right them down.
I saw another comment just now from Evan Scrimshaw. It calls to mind something that I had wondered as well: why are we only getting daily IVR tracker polls from two firms this campaign? Entered into the seat aggregators, they make the Liberals look like they're in the doldrums after just two weeks, and the media picks up on a particular bandwagon narrative that Trudeau is toast. It almost feels like the media wants to push a particular self-fulfilling prophecy... 😢
Post a Comment