Nova Scotia aftermath -- Did the pandemic measurably affect turnout by party? Do you think the gap had more to do with real behavioural change or polling error? Is there a polling error trend developing in favour of Conservatives?
Overall, it looks like turnout went up slightly from 2017, so it doesn't seem like the pandemic had much of an impact. Did it impact parties differently? That's impossible to know for certain, since we can't really separate out people who stayed home from people who switched parties.
Considering that the polls turned out to be pretty close to the result, I'd suspect the error was more to turnout effects (i.e. older people vs. younger people, rural vs. urban) and undecideds than the polls just being done poorly.
As to whether the Conservatives could benefit from a polling error, that's really tough to say. The Nova ScotiaPCs beat their polls, but the B.C. Liberals (the right-of-centre option there) didn't in 2020. The Saskatchewan Party beat their polls last year, as did the New Brunswick and N&L PCs. So, you'd be right to say it is more likely than not, but it isn't a given. In the last two federal campaigns, both the Conservatives AND the Liberals beat their polls by a bit.
The low turnout numbers were predictable but still distressing I think, continuing a steep downward trend in voting NS since the 80s/90s. I wonder if there's a larger phenomenon of decreasing voter turnout across Canada at all levels over the past few decades or if it is level/region/province specific. Do pollsters ever ask people if they intend to vote/spoil or abstain, and if so why or why not? I'm sure there is work on why the decreasing turnout numbers in general, but it seems something major has shifted in some parts of Canada, like in NS, in terms of how they view the (non)importance of elections. This isn't to say people are less political, as there are many ways to be so besides voting, but there does seems to be at least parts of Canada a growing disengagement from the electoral process.
Isn't the correct answer to why the polling was wrong that people have agency and can change their preferences at any time. Admitting that should mean you loss clients. Then the next step is to look at ways to get more frank and reliable insights for voters.
Of course they can change their preferences at any time, but people usually don't. People usually go through with what they tell pollsters they will do (or already have done, in the case of early voters). It's the small number of people who don't, and the undecideds, that make the difference. The polls were dead-on at 21% for the NDP and were only 1-3 points off for the Liberals and PCs. That's not by chance!
Nova Scotia aftermath -- Did the pandemic measurably affect turnout by party? Do you think the gap had more to do with real behavioural change or polling error? Is there a polling error trend developing in favour of Conservatives?
Overall, it looks like turnout went up slightly from 2017, so it doesn't seem like the pandemic had much of an impact. Did it impact parties differently? That's impossible to know for certain, since we can't really separate out people who stayed home from people who switched parties.
Considering that the polls turned out to be pretty close to the result, I'd suspect the error was more to turnout effects (i.e. older people vs. younger people, rural vs. urban) and undecideds than the polls just being done poorly.
As to whether the Conservatives could benefit from a polling error, that's really tough to say. The Nova ScotiaPCs beat their polls, but the B.C. Liberals (the right-of-centre option there) didn't in 2020. The Saskatchewan Party beat their polls last year, as did the New Brunswick and N&L PCs. So, you'd be right to say it is more likely than not, but it isn't a given. In the last two federal campaigns, both the Conservatives AND the Liberals beat their polls by a bit.
The low turnout numbers were predictable but still distressing I think, continuing a steep downward trend in voting NS since the 80s/90s. I wonder if there's a larger phenomenon of decreasing voter turnout across Canada at all levels over the past few decades or if it is level/region/province specific. Do pollsters ever ask people if they intend to vote/spoil or abstain, and if so why or why not? I'm sure there is work on why the decreasing turnout numbers in general, but it seems something major has shifted in some parts of Canada, like in NS, in terms of how they view the (non)importance of elections. This isn't to say people are less political, as there are many ways to be so besides voting, but there does seems to be at least parts of Canada a growing disengagement from the electoral process.
Isn't the correct answer to why the polling was wrong that people have agency and can change their preferences at any time. Admitting that should mean you loss clients. Then the next step is to look at ways to get more frank and reliable insights for voters.
Of course they can change their preferences at any time, but people usually don't. People usually go through with what they tell pollsters they will do (or already have done, in the case of early voters). It's the small number of people who don't, and the undecideds, that make the difference. The polls were dead-on at 21% for the NDP and were only 1-3 points off for the Liberals and PCs. That's not by chance!