Future of Democracy Series: The State of Political Polarization in Canada (FON1-V21)

Transcript: Future of Democracy Series: The State of Political Polarization in Canada

[The CSPS logo appears on screen.]

[Video shows animated close-up view of a spiderweb. Dramatic music plays in the background. As the video progresses, the web is shown to cover all continents on a map of the Earth.

Text on screen appears: Canada School of Public Service; École de la function publique du Canada; And University of Toronto's Munk School of Global Affairs and Public Policy; Et l'École Munk des affaires internationals et des politiques publiques à l'Université de Toronto; Present Future of Democracy Series; Présente la série L'avenir de la démocratie; The State of Political Polarization in Canada; La polarization politique au Canada.]

[Aiesha Zafa appears on screen. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: Hello everyone, and welcome. My name is Aiesha Zafa, a faculty member at the Canada School of Public Service, and I'm pleased to moderate today's session in our series on The Future of Democracy. I'm speaking to you from Burlington, Ontario, land which is part of the treaty lands in the territory of the Mississaugas of the Credit. These lands are rich in the history and modern traditions of many first nations in the Métis, and they span from Lake Ontario to the Niagara Escarpment. I would also like to acknowledge the land upon which I was born and raised, where I met many indigenous people, and experienced their traditions growing up. However, it was not until I joined the public service that I truly understood the significance of these lands, the history of Canada and the indigenous people, and about reconciliation. The land upon which I was born was a traditional territory of the Bisset, known as Thompson, Manitoba. And while I was born in Thompson and I currently live in Burlington, Ontario, you may be on different land or territory, and I really encourage you to learn more about the traditional lands of the indigenous people of Canada.

For the next 90 minutes, we're going to be focusing on the state of political polarization in Canada, an incredibly timely and important topic, which I'm sure many of us have thought about over the past few years. This is the second event in our new series on the Future of Democracy, following the first discussion on how governments can meet future challenges in this space. Before we dive in, I want to make sure that as always, you have the best experience you can joining us today. So I have a couple of tips for you. Today's event will be in English, simultaneous interpretation as well as the CART. So that's the real time captioning is available if you need it, and if you want to follow in the language of your choice. To access these features, just click on one of the icons directly from the webcast interface, or you can look at the email reminder that was sent to you by the school, and the instructions are in there. To optimize your viewing experience, we also recommend that you disconnect from your VPN if you're currently connected to one, or use a personal device so you can watch the session. If you're experiencing technical issues: a great old IT trick, just turn it off and relaunch the webcast link that's provided to you, it usually does the trick. During the event as well, you may submit your questions at any time by pressing the raised hand icon, which is on the top right hand corner of your screen. We have planned time for a question and answer period, and it's going to be a lot of really interesting information, I'm sure you've got some questions, so make sure you submit them as you think of them and I will do my best to get to them for you.

So let's jump into this because there is so much that we can talk about. We are going to start off today with a presentation by Eric Merkley, Assistant Professor of Political Science in the Department of Political Science at the University of Toronto. He is going to share with us his research which broadly focuses on the relationships between elite behaviour, news media, and public opinion, particularly as they relate to issues of expert or scientific consensus. So think about all of the stuff that's been going on in the first couple of years of the pandemic. He's also currently working on a book, which I'm really looking forward to reading when it comes out because it's going to be on political polarization in Canada. Something that I know many of you have probably talked about with your friends or colleagues or at the dinner table, and looking at the data and researching this topic, which really hasn't been done before. We're all really, really looking forward to hearing more about that. So without further ado, welcome Eric and thank you for joining us today.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: Good morning. My name is Eric Merkley and I'm an Assistant Professor of Political Science at the University of Toronto. It's a pleasure to be invited here to talk to you all about the state of political polarization in Canada, which I think is a super important topic and one that's quite timely because there is a nagging sense that things have changed over the last decade or so.

[Video shows cut-outs of newspaper headlines: "Welcome to the Americanization of Canadian politics."; Things fall apart in the United States – and Canada takes a hard look in the mirror"; "Donald Trump unleashed dark politics in America. Could it happen in Canada?"; "Whatever happened to Canadian's famous pursuit of balance?"; "Canadians' appetite for climate action growing, but be aware of polarization"]

Eric Merkley: Some folks have raised alarms about growing polarization. For example, Akaash Maharaj, CEO of the Mosaic Institute notes that "The moderate middle has largely disappeared. Increasingly political rhetoric is used to incite rage against opponents and fear of electing another party." And Susan Delacourt of the Toronto Star writes that "People are absolutely convinced the right, and everybody else, is absolutely wrong. The middle is the scariest place to be in Canadian politics." Plenty of other pundits have alluded to similar themes and I'm not just cherry-picking helpful anecdotes, I downloaded articles focusing on polarization in the big three newspapers.

[Video shows a graph, the X-axis shows number of articles on polarization, the y-axis shows the years from 1980 to 2018.]

[TEXT ON SCREEN: The Big 3 Newspapers: National Post; Globe and mail; Toronto Star]

Eric Merkley: This graph here on the screen shows us that there has been a sharp rise in this topic over the last decade or so. And a common thread throughout all these accounts is a belief that Canadians might be polarizing, just like Americans.  So for the rest of this talk, I want to unpack the narrative here. Are Canadians polarized? Are we polarizing over time? And if so, what are the consequences of this phenomenon, and what can we do about it? I'll present eight findings related to these questions for the remainder of this talk. The first finding that I want to draw attention to is that people who support political parties increasingly do dislike parties and their supporters that are their ideological opposites. That is, for example, Conservatives increasingly dislike the NDP and Liberal party, and vice versa for NDP and Liberal supporters. This is what scholars call Affective Polarization.

[Video shows a graph, as described by narrator.]

Eric Merkley: And we can see this in Canada by using 'feeling thermometers". This is where we ask survey respondents how warm or cold they feel towards political parties on zero to a hundred scales. These can be found in the Canadian Election Study. I'm going to be using this a lot over the course of this presentation because it's the best academic study out there that allows us to observe changes in Canadian public opinion over time. This graph on the screen will plot the average feeling thermometer score at zero to a hundred scale for one's own party that they support, and their ideologically opposing party. So for instance, for Liberal and NDP supporters, that would be the Conservative party, and vice versa. Across elections, which are on the x axis, and as you can see here, we have seen a steady decline in people's feelings towards opposing parties, along with increasing warmth towards the party that they support. And it's important to note that this has been a very long-term process. It's not confined to the Harper and Trudeau years, nor is it confined to the social media era. This has been going on for decades.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: One potential cause of this growing inter-party hostility, it leads me to finding number two. It turns out that NDP and Liberal party supporters on the one hand, and Conservatives on the other, increasingly disagree with one another about the proper role of government in the economy and in social life. That is, they're becoming ideologically dissimilar. And this is what we tend to call partisan polarization. You can also evaluate this phenomenon using the Canadian election study.

[Video shows a graph, as described by narrator.]

Eric Merkley: Here I use a question that asks our survey respondents where they would place themselves on a zero to 10 scale from left to right. Here, I calculate a statistic that tells us the dissimilarity of responses to this question between two pairs of parties, say the Liberals and the Conservatives. A value of one means the two parties are perfectly dissimilar from one another, and a value of zero means they overlap perfectly. And we should expect to see an upward trend in this statistic if there was growing polarization between any two parties. And this is exactly what we see. The ideological differences between the Liberals and Conservatives have increased over a hundred percent since 1997. The Conservatives and NDP have also become more dissimilar, but they were already pretty dissimilar to begin with. And at the same time, the Liberals and NDP have converged over time in their ideological beliefs, signified that downward trend we see on the graph.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: But, and this is a big but, there are three important caveats to this polarization narrative. First, Canadians aren't actually becoming more extreme in their beliefs. Now wait, I just said that ideological beliefs of Conservatives in the one hand, and Liberals in the NDP on the other, are diverging.

Isn't that a sign Canadians are moving to the extremes? It's important to note that that's not actually necessarily the case. There are a lot of Canadians that don't identify with political parties. And people could change their partisan loyalties to better match with their preexisting ideological views, rather than changing those views themselves to become more extreme.

[Video shows a graph, as described by narrator.]

Eric Merkley: So I use zero to ten ideological self-placement scales, that measure that I just used before, from the Canadian Election Study. Here I calculate something called a bimodality co-efficient. This will tell us whether people are starting to cluster at the extremes of that zero to ten scale. 0.55 on this scale is the threshold for bimodality, and it runs from zero to one. And what we see is that there's very little movement over time. If you squint really hard perhaps you can see on the graph some initial depolarization that's been reversed since 2008. But the low numbers also suggest a very unimodal distribution. And what I mean by that is that Canadians are clustering at the center point on that zero to ten scale. They're moderates. There's no evidence of the growth of the extreme left or right.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: The second biggest caveat to the polarization narrative relates to one of the most notable characteristics of US polarization. That is that Republican and Democratic voters have become increasingly socially distinct. Evangelical high income whites clustered in the Republican party; non-religious low income, racialized minorities have gravitated towards the Democratic party. These differences aren't just about policy or ideology, it's about something even more fundamental than that. And this is a process known as Social Polarization. So, what's the story here in Canada?

[Video shows a list of correlations, as described by narrator.]

Eric Merkley: Here I used data collected with a 2019 digital democracy project. What I asked respondents to do is to identify different social groups that they belong to and how important these groups are to their sense of self. Such as: their region of residence; their class; their race; et cetera. And I examined how correlated those social identities are with the intensity of a person's support for a given political party. And what we see is that, on the whole, correlations between social identities and partisan identities are, on average, quite weak with the exception of region of residence. And they're much weaker than what we see in the United States. So that's good news.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: Then finally, the last caveat I want to draw attention to is the reality that we actually think we are far more polarized than is actually the case. We tend to adhere to exaggerated stereotypes of political parties. We think that Conservatives for example are much more wealthy evangelical and right wing than they really are. We think NDP and Liberal supporters are more left wing racialized and cosmopolitan than they really are.

[Video shows a graph, as described by narrator.]

Eric Merkley: In the Digital Democracy Project surveys, I asked respondents to give their best guess of the share of Conservative, NDP and Liberal supporters that are members of certain social groups. These groups can be stereotypical. So for instance, evangelicals are associated with the Conservative party. These groups can also be counter stereotypical. Visible minorities for example, aren't typically associated with the Conservatives.

The graph I'm going to show you here is going to tell us the net error in people's evaluations. And what it shows effectively is that higher values indicate that people overestimate the share of partisans from a given stereotypical group, compared to their estimate of the same counter stereotypical group. So just as an example, how much do people exaggerate the prevalence of evangelical Conservatives compared to their estimate of NDP and Liberal evangelicals? And we see considerable evidence of bias across all groups. People exaggerate the share of party supporters from a stereotypical group by about 10 percentage points on average. These are very large differences, though not nearly as stark as what we've observed in the United States. So to recap: yes, there is some polarization in Canada. Partisans increasingly dislike and disagree with one another, but we're not becoming more extreme, and we're not as polarized by race, religion, or class as Americans are. And finally, and also problematically, we think we are more polarized than we actually are, and these misperceptions actually matter in their own right.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: So this discussion brings me to the second part of this talk. Well, why does this matter? If we are polarizing, why should we care? US scholarship has identified a number of different social, cognitive, and democratic consequences of polarization. I'm going to spend just a little bit of time evaluating the scope of a few of these problems in Canada, but there is much more work to be done on all of these issues in our country. Most of the research so far is done in the US.

One concern that's been raised about polarization is that it may lead to social alienation. And this is implied by Affective Polarization. If you really dislike your political opponents, you're just not going to be likely to engage with them in your day-to-day lives. This could undermine social trust and social cohesion, which is the foundation for a healthy market economy and for democratic life. So it would be a big problem.

[Video shows a graph, as described by narrator.]

Eric Merkley: In the Digital Democracy Project surveys, I measured people's willingness to have opposing party supporters as friends, neighbours, or in-laws. From these, I construct an index of social alienation, where a hundred is the maximum level of alienation, so another zero to a hundred scale. And we can see that alienation towards opposing party supporters. It does rival feelings people have towards Muslims. But on this index, it's still relatively low, about 27 on a zero to a hundred scale. So much closer to the zero, where there's no social alienation than to a hundred. And these results are more muted than what we see in the US. So for instance, 24% of Canadians would be upset to some degree with an opposing party supporter as an in-law, compared to over 40% of Americans.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: And this has implications for the relationships we pursue.

Another thing that US scholars have found is a striking trend towards political homogeneity in romantic relationships. And this is problematic because it could reproduce polarization through generations when children are politically socialized in increasingly homogenous families.

[Video shows data from the experiment described by narrator.]

Eric Merkley: I conducted an experiment where I gave respondents a series of dating profiles of hypothetical individuals. And in these profiles attributes like political views and religion, they were randomized. So it allows me to say which of these factors matter in shaping people's decisions. They were asked their willingness to respond to messages, to message, to date, or to form friendships with the people featured in these profiles. So did politics matter? Turns out it does, and it matters at the same level as matching on religion, or matching on education, which is striking. But there is some good news here. Getting an opposing party supporter on a profile on average didn't matter.

People gravitated towards people that shared their partisanship, they weren't necessarily repelled by those that were opposed to them. And further analysis showed it's really only a small segment of people that are intensely hostile to opposing parties where it did matter. So that's generally good news. Polarization may be shaping people's willingness to form social relationships with political opponents, but it's not nearly as extreme here as it is in the United States.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: There are also cognitive consequences of polarization. First and foremost, what's known as Selective Exposure. You've probably heard this before, people tend to choose information or sources of information like the news that support one's own politics and their world views. And even worse, people might cocoon themselves in echo chambers that shield them from uncomfortable information. As the importance of those identities increase with polarization, so too does the importance of selective exposure.

[Video shows data from the experiment described by narrator.]

Eric Merkley: I can shed light on this with an experiment in the Digital Democracy Project. In this case, I provided people with four pairs of hypothetical news story profiles. They're asked to pick a story that they'd like to read. In these profiles I randomized the source so that some received a major national news outlet like the CBC or CTV. Some got a local news source, and others got partisan news, either left wing or right wing in nature, like say Rebel Media. I also randomized the source of the headlines. Some were right leaning headlines; some were left leaning headlines.

[Video shows a series of graphs, as described by narrator.]

Eric Merkley: This graph here on the screen will show us the share of people selecting a story at different levels of partisanship along the X axis. If people select headlines aligned with their partisanship, we should see a downward sloping line in red for left leaning headlines, and an upward sloping line in blue for right leaning headlines. And that's what we see.

So among strong supporters of the Liberals or NDP, 57% of respondents are expected to select the left leaning headline, compared to 46% of respondents for strong Conservative partisans. But it's worth noting though that these effects are quite modest in size. And there's even better news from this experiment. There's no selective engagement with partisan sources of news whatsoever. This graph here will plot the share of people selecting national news, a national news story starting on the left, then local news, and left or right slanted news. Liberal and NDP partisans will be in red, and Conservatives will be in blue. And higher values mean people are more likely to select that source. And what we see is that stories from major national news outlets are more likely to be selected than partisan news sites by both groups of strong partisans. And in fact, people hardly make a distinction between sources of news slanted in their favour or against. And so that is for instance, Conservatives are just as unlikely to select a right-wing sourc, compared to a left-wing source compared to major national news.

So, this is a bit of a paradox. Why don't our respondents show some love for partisan news? We see this going on in the US, why is that not the case here? That's likely a mix of two factors. First, there's less awareness of these sources. People don't know necessarily what Rebel News is, for instance, and there's lower levels of trust in partisan news relative to mainstream news. There's evidence of both of these things in the survey data.

But the big kind of consequence of this, and I really want to draw your attention to, is that as a result, the size of partisan media audiences in Canada is extremely small. In 2019, we tracked the online search behaviour of a sample of Canadians over the course of two weeks during the election. And in that sample of about 750 people, only seven people visited Rebel News at any point over two weeks. Two people visited the Post Millennial, no one visited rebel.ca, compared to 33% who visited CBC Radio Canada. The overwhelming majority of Canadians consume no partisan news at all. And this result holds even when we include American sources like Fox News.

[00:21:45  Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: So in short, people do prefer political information that backs their political views. This should not surprise us, but the effects are small. And so far, Canada's media environment really hasn't caught up to this demand. And we might worry about echo chambers existing, but they functionally do not exist at any level in Canada, despite growing polarization. And finally without a doubt, the biggest concern raised about polarization is it's toxicity for democracy itself. If citizens increasingly disagree with and dislike their political opponents, states of political conflict skyrocket. Politics becomes more than just about debates over inane tax policy, for instance, but about institutions and rules and norms, and people might be less willing to concede political defeat in that sort of context. We see this dynamic spiraling towards political violence in the United States, but luckily there's no evidence so far that this is a problem in Canada.

[Video shows a series of graphs, as described by narrator.]

Eric Merkley:  I personally haven't done work in this area yet, so I want to give a shout out to the Environics Institute for monitoring these issues in Canada and the US. This graph on the screen shows the percentage of Canadians, in that left panel, that are satisfied with democracy across different levels of ideology. And unlike in the US, we see no growing gulf of democratic satisfaction that varies across election results. This next graph shows the same for the percentage of people who have a lot of trust in elections. And again, unlike the US, there are minimal differences between the left and the right. They're broadly satisfied with democracy and trust in what our elections produce.

In short, Canadians aren't necessarily willing to socially wall themselves off from their political opponents, even amidst polarization. While we might prefer politically aligned news content that backs up our politics, they haven't meaningfully turned away from mainstream news and the remains cross partisan consensus on matters relating to democracy. This is all pretty good news. We're polarizing, but maybe the consequences that a lot of people fear just aren't there.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: But that's not to say this will always be true. The US spiraled towards increasingly destructive polarization in a remarkably short span of time. So is there anything that we can do about it? Unfortunately, I'm more of an identifying problems than identifying solutions kind of academic at this stage in my career. And I worry that there's no magic wand that can be waved to depolarize Canadians. And I think that's true because the responsibility mostly rests with our political parties.

Partisan and affective polarization are likely consequences of our major parties becoming more ideological over time. We see that the Liberals have moved to the left, and the Conservatives to the right, and this has left the center vacant as measured by the content of party platforms by scholars. So ideologically charged rhetoric and policy don't go unnoticed by Canadians. It has consequences, it polarizes them.

And it's also possible that social media is further sharpening political conflict, though it isn't a cause of polarization per se. It can do this by making it easier for people to hear perspectives that reinforce their own beliefs, but probably even more importantly, it exposes people to very inflammatory content from the other side. So we do need greater transparency in recommendation algorithms across social media platforms, more access for scholars to study these things. And changes are needed to reduce the virality of provocative content which travels at light speed through social media.

And most of this isn't driven by bots, even though we like to point them out as the boogeyman. Social media isn't just problematic because of its circulation of misinformation. It has polarization consequences as well, even when factual information is disseminated. Ultimately though, I'm pretty skeptical of the prospects of depolarization. It's a tall order. I do think it might be more practical to focus on reducing challenges associated with polarization.

For one, as I talked about earlier, people do think we're more polarized than we really are. The news media is almost certainly the principle source of these misperceptions. Journalists and editors have commercial incentive to hype up political conflict. So we really do need more reflection by journalists on how they can responsibly cover politics. We need to think carefully about how we talk about polarization because just by talking about it, we might increase perceptions that it's a problem.

We also do a really poor job at preparing children, citizens to be responsible participants of democracy. This could mean building media literacy skills to build up resilience to misinformation on online spaces, to bolster trust and mainstream news by educating people about what exactly journalists do. We could teach children how to engage in healthy and constructive political dialogue. We do have some work in political science, for instance, that shows that respectful cross partisan dialogue does depolarize. So normalizing those sorts of interactions I think is extremely important.

We can also make people aware of their own political blind spots, their tendency to prefer information that backs their own beliefs to engage in motivated reasoning. That is the selective acceptance or rejection of information based on its convenience for one's politics. Making people aware of these cognitive biases is also very important.

And finally, we need to adapt communication strategies for a polarized era. I think across a lot of different domains, it's really important to build areas of cross partisan consensus. I think of climate change, for instance. And this means we need to identify those who are resistant to persuasion on a certain issue, and find ways to bridge that divide. And this can be simple, we don't have to overthink it, it can be as simple as building relationships with trusted messengers in targeted communities. We can make arguments in ways that are compatible with people's values. In other words, we need to meet people where they are, and respectively engage with their beliefs and try to persuade them from that perspective.

Polarization also isn't necessarily a bad thing, and I want to leave us off with this thought. Politics should be a contest of ideas and values where reasonable people can disagree. But we can manage the potential bad consequences of that polarization. We can do that better to make the process healthier, and more constructive. So, thank you very much, and I look forward to the discussion.

[Aiesha Zafa appears on screen. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: Wow, Eric! I don't think I was expecting some of that, honestly. There's a lot to think about there. The last couple of years the conversations that we've all had, again at our dinner table, with our friends, doesn't really jive with what you just said. So I want to dive into that a little bit more.

But first before we get into the discussion, I want to remind our participants that you can ask questions. So use the chat function, submit your questions, and I will be able to ask them to our speakers. One of whom I'd like to introduce now is Scott Matthews. So, Scott Matthews, also joining us on our panel today, is the Associate Professor in the Department of Political Science at Memorial University of Newfoundland. Scott brings a wealth of knowledge in the study of elections, voting and public opinion in Canada and the United States, and looks at things like the psychology of political learning, the role of uncertainty and support for costly public investments and how that affects us. So really, really interesting stuff to dive into there as well. Welcome, Scott.

Scott Matthews: Thanks, Aisha.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: So maybe I'll go to you first, because I know I have a ton of questions, and I have a lot of things that I want to dig into Eric's findings. But with what you just saw, can you give us your thoughts on Eric's findings, how it relates to some of the research that you're doing, and any other points to start us off with, policy makers or just regular Canadian citizens should actually consider?

Scott Matthews: Yes, absolutely! I'll resist the urge to just praise everything and agree with everything that Eric said, but that's actually a really hard thing for me not to do because it's basically true.

[Scott Matthews appears on screen. A sign language interpreter appears in a smaller window on the side.]

Scott Matthews: But the one little bit of praise is, a couple of years ago just before the pandemic, it was my last trip. I was invited out to BC to give a talk on polarization. And in constructing that talk, I really had to rely almost entirely on evidence from the US. And the kinds of dimensions Eric's talking about, I tried to speak to a lot of those, but there really was no Canadian evidence, and now we have it. So my main reaction to what Eric has done, and is in the process of doing, is that this is just really critical for understanding Canada, especially because there are some real big differences here.

Other reactions: I wanted to build on Eric's last comment about polarization not being such a bad thing. I guess it's sort of an academic tick to want to disagree with everything you hear, or want to disagree with every bit of conventional wisdom. But, not long ago, the conventional wisdom in Canada and the US about parties was that they were too similar. And when I'm saying not long ago, I'm thinking of the 1980s, maybe a bit before the eighties in the case of the US, but the Democrats and Republicans were sometimes criticized as being just Tweedledum and Tweedledee, essentially interchangeable and Americans didn't have a real choice. In fact, there was a really famous report among political scientists that was written about the lack of a responsible and ideological party system, in the fifties I think, because the parties were too similar, they weren't polarized enough.

We had the same kind of arguments in Canada, and much more recent, as late as the 1980s. Janine Brodie and Jane Jensen, and a couple of influential scholars of Canadian politics went as far as to suggest that there was a sort of bourgeois conspiracy joining the Liberals and Conservatives in essentially closing debate on all kinds of economic issues in order to preserve capitalism. We don't need to go that far to see that the parties actually were quite a lot more similar not that long ago, and not everyone liked that. So it's just to reinforce the idea that polarization as difference is not an inherently bad thing, and we might complain about a lack of difference.

The last thing I wanted to mention to start us off, that I think is really important and what Eric said, is that this process of polarization in Canada is a long-term phenomenon. And, in a way, even more long term than in the US. It was really in the 1990s that the party system blew up. That's the metaphor we all use, with the Conservative party essentially collapsing and giving birth to these two alternatives: The Reform party, later on the Alliance; and the Block Quebecois. By and large, Canadians were pleased by that development. Suddenly they found parties that were closer to them in ideological terms. That's reflected in Eric's data, where you see not just people finding their least liked party more disagreeable, the party they like best, they like better than they used to because, maybe you're a Conservative and suddenly you have a truly Conservative party like the Reform party.

The flip side of that is that difference between the parties also drove people to participate in politics. I've done some work on that in the past, on voter turnout that shows that among other things, and there's a few pieces to the story, but one of the things is that people became less indifferent about the alternatives. When you have two really different parties competing, it makes it more worth your time to get out and participate, to learn, to vote. That's one of the upshots of polarization. So those are some of my starting comments. But generally, I think what Eric's presenting really important, and important lessons for all of us.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: Yes, I think it definitely surprises a lot of people out there. I'm going to make that assumption. You know, we have over 1500 participants registered today for this event, and I'm sure people are going through their Twitter feed right now, or their Instagram, and they're like, "Look, here is clear polarization. It's true, I've lost friends over this in the past couple of years."

So, Eric, to go into a bit of your methodology, how do we know that when people are responding to the surveys or whether you're asking questions or monitoring their behaviour, how do we know that they're actually telling the truth? And are there reasons why there may be some kind of hidden agenda in the responses that you're getting? Or maybe walk us through a little bit of your methodology.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: Yes, so this is just basic survey methodology, so using the Canadian election studies. So that's a study that's run every Canadian election. So there's a campaign period survey, a post-election survey and a mail-back survey through most of these. Though there have been some methodological changes recently moving towards online survey sampling. Typically, the concern that most academics have is on questions that have social desirability issues. That is that there is a norm in society that you're supposed to believe "X", and it makes people more likely to say that they believe "X", or that they behaved in a certain way. So the classic example of this is when we ask people whether they turned out to vote, a lot more people say they voted than actually voted in surveys.

So that's an example of the social desirability issue. In the Covid context, there's some literature that shows that people are more likely to say that they engaged in precautionary behaviours. So that's another example of social desirability. Here, we're just asking people to place their ideology on a zero to ten scale, or in my other work, just their basic attitudes, nothing particularly inflammatory, just basic questions about the role of government in social life and economic life. And we typically don't see issues with social desirability on those sorts of questions. So, certainly, there's always issues with survey respondents not paying particularly careful attention to their responses. Not everybody is fully motivated when they're responding to a survey. But on average those responses tend to wash out with effective and proper survey design. And that's why relying on an academic study done by the best discipline when crafting these sorts of questions, I think is really important. So, I don't suspect those issues arise with anything that I've shown today.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: Yes. You mentioned ideology, and I think that was one of the things that really, not necessarily shocked me but struck me was when you mentioned that the ideological differences between the Liberals and the Conservatives have increased over a hundred percent since 1997. And Conservatives, NDP, I think you said were already kind of on the opposite ends of the spectrum, but over a hundred percent! And, if it's the ideology that's actually causing that polarization, can you delve into that a little bit more?

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: Yes, absolutely. So, the argument that I make in my work is that people increasingly disagreeing with one another on basic questions related to the role of government is a driver of who they like and dislike. So it generates that sort of affective response. And that's a conservative estimate, that hundred percent. It depends on the indicator that I use. So in my work, I have that self-placement scale that I show. That's the most conservative estimate. But I also use people's policy attitudes, I separate those policy attitudes to economic views and social views. So attitudes towards say, same sex marriage and abortion, those sorts of questions. And there's even more action on social questions. So, on social questions, the dissimilarity between Liberals and Conservatives has increased 270% over the time of the Canadian election study. So, these are big numbers, and there's also increasing disagreement between the NDP and Conservatives as well on those sorts of questions too. So these are big differences, and they are likely drivers of the tendency to increasingly dislike your political opponents.

But that doesn't necessarily mean Canadians are becoming more extreme, and there's all the other caveats that I talked about. Disagreeing with one another on the role of government is not necessarily a bad thing on its own. And you can manage the consequences for affective polarization in a number of ways. So, nothing that I showed is necessarily bad in an important sense. And I agree with everything that Scott said about how there are normative benefits to polarization. Our institutions, unlike the United States I would argue, are better structured to manage polarization when it occurs, for a variety of reasons. And so, nothing I've shown concerns me in a serious way.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: And I think that's a really important distinction when you're talking about polarization. Not always a bad thing, but it's the extreme, becoming more extreme in our beliefs. And that's where I think generally when we're just talking at the dinner table, we mix those two things up, or we use them interchangeably. So our beliefs are not becoming more extreme. And what's really important for public servants, and this is why I love the research and the data, and we need data to be able to make good policies and to present options to government for the different things that are happening. But when we're basing those policies and those recommendations and options on assumptions that we are more extreme in our beliefs and that there is this huge political polarization, we might not be putting the right policies and recommendations forward.

So, this data is really humbling for I think, all of us to sit back and say, OK, we've had these assumptions for a few years based on everything that's out there, but now there's actually been research and data done.

So, maybe one more question before I go to Scott, and for you, Eric and Scott, feel free to comment on it as well, if you like, but as federal public servants, as policy makers, so if Canadians are not becoming more extreme in their beliefs, there is political polarization, what can we learn from other countries where it is increasing? Where there is that increase in beliefs, the extreme in beliefs and where the polarization is intensifying? And maybe not so much the US, because I do want to get into the US in a moment, but are there other countries that we can learn from in terms of policy making?

Eric Merkley: Yes. I think the lesson here is mostly for politicians. Politicians and political parties set the ability to construct coalitions of people that are going to support them on election day. They do that by taking policy stances to appeal to certain groups of citizens. They take rhetorical stances, even if it's not necessarily reflected in policy, that appeal to certain groups of citizens. They have choices about who they court, and who they don't. And in Canada over time, there's increasingly been this tendency towards the minimum winning coalition. You get your base out and a little bit more just to get the majority government.

So that is kind of the starting point. I think countries where you see it being the most toxic are where politicians are crafting concerns around racial religious identity that it inhibits their ability to be able to reach out to broader groups in society.

And so, in the United States, you have the Republicans and Democrats. Because of the rhetorical signals sent, especially by the Republican party, that they are polarizing on these social dimensions: race; income; religion, in ways that it's not just about policy, it's much more fundamental than that. And it raises the stakes of political conflicts. It makes it much more toxic.

And that all flows down to political elites and the choices that they made. The Republican party had the choice to court racial Conservatives in the Nixon administration. Under Reagan they had the choice to court social Conservatives. And now under Trump, ethnic nationalists and white supremacists to the extreme, and they've had the choice of who to court.

And so that is the lesson there, and that's where we really should be concerned, is if political parties start making those choices in that sort of fashion. And I don't think that's where we are here, and so avoiding that in the future I think is critical.

Aiesha Zafa: All right. So kind of on the same vein, Scott, and feel free to comment on that one, but even taking it one step further. So, from everything that we've been listening to so far, the situation is very different in Canada than in other parts of the world, definitely not like it is in the United States, or some of the Latin American countries. So we found to be that it seems like we're more a policy difference than the polarity. Canadians are really thinking more about how much should government be involved in our day-to-day lives and our social lives than other things.

So, considering that then, what kind of strategies can policy makers adopt so that we can actually engage Canadians, and stakeholders, and policy decisions where there are those strong differences in how much government should be involved, and then kind of move away from that assumption that we have these extreme beliefs, but really focus on what Canadians are really thinking?

So how can we engage Canadians more? And Scott, after you we'll go straight to Eric for comments on that too.

[Scott Matthews appears on screen. A sign language interpreter appears in a smaller window on the side.]

Scott Matthews: Well, I have a few thoughts around this. One is that, and I think this compliments what Eric presented, facts still matter. There's this increasingly current idea, and it's part of the false polarization narrative that I think the media and academics, to some degree, are responsible for disseminating. But there's this idea that basically people choose their own facts now. They're so absorbed by, so intoxicated with, that's the metaphor that has been used by colleagues of Eric's and mine in political science, they're so intoxicated by partisanship that they're totally impervious to information. And that one set of arguments has to be prepared for one group, and a different set for another group.

It's certainly the case that partisanship and other social identities and strong attitudes of all kinds can interfere with people's ability to make rational sense of the information they're exposed to. But that's not all that's going on. There's plenty of information, plenty of evidence I should say, that people respond to information, particularly on the kinds of matters that are not so easily absorbed into partisan debate. Things that are kind of ambiguous where there's diversity of moral standards, where understanding is low, those are things where partisanship really bites the most, makes the biggest difference.

Some of my own work has been about how information about the economy can be a source of convergence in political attitudes, at least when economic circumstances are relatively extreme. In work that I've done with Mark Pickup at Simon Fraser University on Canadian elections, we find not entirely consistent, but pretty strongly suggestive evidence that when the economy is going particularly well or particularly badly, an increase in reporting of that economic situation in the media tends to bring partisans together around a common view of economic circumstances. That's significant, because most of the time partisans disagree quite a lot about how the economy is going. It's one of the big things governments are expected to do. And so, out partisans, those who are partisans of the party not in government, have a strong motivation to see things as going worse than those in government, and vice versa. In circumstances when economic conditions are more clear, less ambiguous, it seems that more information can bring partisan groups together.

And just to bring in the US example, there's a similar study prior to ours, with a lot more data, that shows this over a really long period of time in American attitudes. So the economy is perennially important to political choice, to decision making in elections, and it seems to be one area where more information really does make a difference. And there's lots of findings like that really.

So, I suppose this is a kind of pedantic point, but it's an important one. It's not all or nothing. It's not all partisanship, and people polarized into divergent social camps, or total rationality. The truth is actually in the middle. And I think that for government, for civil servants, information still matters and so continuing to do the kinds of things that civil servants do, supporting the political executive but also at the departmental level and in the more service capacity, I still think high quality research expertise, the things that public service has, still count.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: Yes. Just to build on that point, Scott, I totally agree, and there's a lot of, even in the US context, there's evidence that things like fact checking are still effective that, if you try to correct Republican on the facts, they'll move their attitudes towards factual information. Even in the context of really high toxic polarization, facts still matter.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: In the Covid-19 pandemic, the vast majority of Canadians trust experts. The vast majority of Conservatives trust experts. The vast majority of Conservatives vaccinated, double vaccinated, have high opinions of experts, are persuaded by experts. And so, we tend to lose sight of those things. So, facts still matter, expertise still matters. Conservatives are not opposed to any of that to a strong degree in Canada. We tend to exaggerate those differences because in part, of what we see going on in the United States.

There's also we are all probably very politically engaged. There's a lot of people out there that don't live for the cut and thrust of politics. So, there's a lot of people that are not on Twitter all the time being exposed to all this content, and so there's a whole other world out there that probably doesn't reflect our own social circles. These people aren't that polarized. They might not strongly identify with the party. That's a lot of Canadians right there that you also have to bring on board for policy and the like. And starting off with an assumption that everyone's polarized, no one can be convinced, it's not reflective of the reality on a number of different levels.

But that being said, there is more disagreement out there in Canadian politics than there used to be.

So if you could find ways of meeting people where they are, and trying to persuade them, there is a difference between facts and values. Some questions have normative foundations where people are going to disagree and in those cases, just saying "Well, the experts say this" is probably not going to work the way you think it is. So, some sensitivity is called for when normative value based questions are at play, and then having respectful dialogue and all that. I think that's important for political parties, for civil servants, for educators. Across the board, it's something we need to do better at.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: Okay. So let's maybe take that angle then. If we're talking about political leadership, and they do play a crucial role in deescalating some of the divides, are there examples, and I'll post this to both of you, are there examples in other regions or other jurisdictions of initiatives that have actually been effective in fostering cross partisan collaboration? Because, from what I'm hearing, if a lot of this and a lot of what we believe or a lot of what we see in the assumptions that we're making, are based on the political parties and the divide between them, then what examples are out there to close that divide a little bit?

Eric Merkley: Is this for Scott or me?

Aiesha Zafa: Either one of you? You both look like you have something to say, so maybe we'll go to Scott first. [Laugh].

[Scott Matthews appears on screen. A sign language interpreter appears in a smaller window on the side.]

Scott Matthews: So, in thinking about this question, I admit that I am what I am. I'm a psychologically oriented political scientist, and I think about mass politics, not so much about political leaders, I'm concerned about how citizens respond to what political leaders are putting out there. And so, what I have to say about something like that is really about how citizens might respond to different approaches.

One of the most encouraging findings in the last number of years that I've seen is from a study that looked at what difference it makes if people are told that there is some overlap between political leaders in different parties or on different sides of an issue. There's one study in particular, and Eric will know the one I'm talking about as I start to describe it, that compares essentially how susceptible people are to persuasive arguments on a political issue under three different conditions: When those arguments are not attributed to any partisan sources; when those arguments are attributed to partisan sources; and when those arguments are attributed to those same partisan sources, but people are told that, for example, most Republicans are on this side of the issue, but there are a few on the other side. The amazing thing about this research, and they replicated it with a few different issues, is that people were as susceptible to the arguments in the in-between condition - there's a few people on the other side - as they were when there was no partisan information at all presented.

And that's a practical kind of lesson but it links into a larger idea psychologically, which is that one of the things that challenges people to change their minds and to see the perspective of the other side, is that essentially politics is presented in an all or nothing, black or white, absolutist kind of way, as opposed to a space where reasonable people disagree, there's room for discussion, there's room for disagreement, you're not a bad person. Some people on my side actually happen to agree with the folks on the other side. These kinds of interventions are significant because people want to, among other things, they want to feel that their attitudes are justified, that they make sense, that they're compatible with the belief that they're a good person. And when you create an environment that makes politics very binary and when political claims are advanced in a highly categorical way, I think it's quite toxic for people's ability to be open and say, "Maybe I'm wrong about that", or, "I don't agree with you, but I don't dislike you. I don't think you're an evil person." And so, I think that's a lesson for political leaders, and for all Canadians, that the quality of the way we talk about political conflict it's not just something that sounds nice, it's something that is demonstrably important to the psychology of the way people can be persuaded and can come together in politics and society on a whole host of issues.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: Yes, and there are certain attributes of the Canadian institutions that heighten these sorts of polarization dynamics that elites engage in. So you have winner take all, single member plurality elections, where the stakes of conflict increase. And so some scholars would say, well some forms of proportional representation might be able to mute some of these dynamics. That it's not all or nothing, that you have to be responsible coalition partners with other parties, and so it tones down the rhetoric and tones down the temperature. So you could imagine, not all forms of PR, but some forms of PR might help with that. Obviously, you would want some guardrails to prevent really extreme parties from coming in. But that being aside, and then we have a Question Period, the whole notion of the parliamentary structure is oppositional. And so it kind of breeds that sort of conflict, putting the cameras on in parliamentary committees, it allows for grandstanding. So, there's all these things that have kind of really supercharged things in Ottawa, and provincial legislatures.

I'm not sure what can be done to soften those sorts of tendencies? Institutional reform that encourages collaboration, I don't know what that looks like really. But there are inherent features of the Canadian political system that definitely sharpen things.

Aiesha Zafa: Yes.

Scott Matthews: That's actually, It just popped, Oh, sorry.

Aiesha Zafa: No, go ahead. Go ahead

[Scott Matthews appears on screen. A sign language interpreter appears in a smaller window on the side.]

Scott Matthews: I was just going to sort of reinforce what Eric said, these are inherent features and they have been actually getting a little bit worse over the last 20 years or so, or aggravated by the frequency of elections. You know, the idea of the permanent campaign that we're always in. The parties are always in election mode for obvious reasons. It does heighten the desire to grandstand or makes it harder to compromise. To see the other side makes it more dangerous politically. So, some of this stuff, in a time of minority governments maybe that's part of the story. I'm not at all suggesting that's the whole story, but I think that's part of the story too.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: I think those are fascinating observations and one of the questions we got in, I'm not getting to the audience questions right now, but another plug, if you do have questions, please put them in the chat and we will ask our speakers soon.

But one of the questions that you just answered was, "Why is there such a disconnect between politicians and the population?" And I think what you both just mentioned, some of it is institutionalized, like Question Period. Yes, absolutely. It's very different than some of the other jurisdictions that we see: always in election mode, the minority government, "Here we are going into elections again." So, what are the big and perhaps polarizing issues that we need to focus on as Canadians, or as the elections go through. So, something that I don't think I've really thought about in those terms.

So when we are, as public servants, I guess ourselves, or maybe not just public servants, but there's a lot of immigration still happening in Canada. We have new Canadians who come in, I don't know what their views might be about the state of polarization in Canada before they get here, but what should we actually consider when we're educating the new generation of Canadians, whether immigrants or our kids who are being raised in Canada on this topic. What should we be focusing on when it comes to polarization so that they don't have the same assumptions we do, and then maybe dig a little deeper into what's actually going on.

Eric Merkley: I think the biggest thing, this is probably more for children than for new immigrants, but I remember my Civics class back in grade 10. A half-credit course, the teacher put his shoes on the table, gave us a reading and that was it. That was Civics class, basically.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: We need to do a lot better job at that. And, I think there's a tendency to "Well, okay, if you want to avoid polarization, you need to just avoid conflict." But that's not really healthy or helpful. People just suppress what they think, they're resentful for it. We have to allow for, and educate citizens into having respectful conversations that highlight disagreement, and allow for some form of accommodation in that regard.

We have research that suggests that cross partisan dialogue can reduce affective polarization, make people like the other side a little bit more. Part of the reason for that is it probably reduces that false polarization I talked about, that we tend to exaggerate beliefs about the other side. There's still more work being done on this. So there's all sorts of caveats based on the design, the research design, the topics that are discussed and all that. But there's definitely some potential there of educating citizens, children, about how to have those conversations with one another at an early stage that I think can be very helpful.

And also, just recognizing their own cognitive biases. Everybody thinks the other side is biased when we all are, because of our own values. And so instilling these sorts of notions early on I think is really important. It doesn't start with children, we all could use a helpful refresher on some of those issues too, but I think the earlier, the better.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: But how do you do that in the age of social media where, and I've said this before, as much as I like to think that I'm very well read, and I read a whole news article and I read a variety of different sources, the reality is I get a lot of information pushed to me. And we know how the algorithms work and everything else. So, your busy day-to-day life, you're having these conversations. First thing in the morning, I turn on my Google and I say, "Hey, Google, tell me the news." And it's three minutes of headlines across the world, and that's what I'm going to take with me for the rest of my day. So, the points you're making Eric, are great. We need to start having those conversations. I like the cross partisan dialogue, et cetera. But the reality is, it's hard for me, and part of my job is actually going out and finding the news and what's happening in the world. It's hard for me to do that. So how does the average Canadian start to think about things differently?

[Scott Matthews appears on screen. A sign language interpreter appears in a smaller window on the side.]

Scott Matthews: [Laugh] I'll jump in to say that yes, I think it's a tricky, tricky problem to think about how we change our media system. Not least because we have a free, kind of commercial open society sort of media system. And not a whole lot of this is in the hands of government, which is just to say it's not in our hands collectively to do much about it as Canadians.

So, I like the direction of Eric's suggestions about civic education. All I would add is I think the idea that civic education is the solution to our democratic ills doesn't have a great reputation in political science but I think that's because we usually think of it as potentially a place where people can learn about policy issues and learn how to make good vote choices. It's never going to happen.

But civic education as a place to inculcate values, but values of civil discourse, values of democracy, like openness to the other side. The importance of making arguments in ways that others can at least in principle accept. Respecting others. So those kind of process values for democratic politics. I'm not saying there's nothing else civic education should do, but that sounds to me like a really important one, a kind of way to go.

And then I think that would be complimentary with media literacy. I know that there are multiple NGOs, one in particular totally escapes my name that's working on this in particular, that going into schools and teaching young people how to evaluate the credibility of the information they encounter. Lateral reading and so on, these kinds of strategies, I think that should be part of a civic education curriculum.

Probably less institutional detail, more basic democratic principles and strategies that can be applied no matter what comes in the future.

So, I think that's one thing I'd say. That's not so much for new [Canadians]. Like Eric, I guess that's for young people, not so much for immigrants. My own view is that a lot of people emigrating to Canada are coming from places that are demonstrably, vastly more polarized than Canada. And so I think we should respect [laugh] that they may have some things to teach us about what is going right here, and also perhaps things that should be changed. But anyway, that's a slightly flippant remark, but when I think of the kinds of students I encounter, most of the immigrants I encounter nowadays are students, people studying here. I'm kind of struck by just how good we have it. I know it's kind of trite, but that's my reaction.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: That's actually a really, really thought-provoking reaction though, Scott, because typically, as policy makers or as public servants, when we're going out and we're doing our stakeholder engagement, we're often thinking about who are the groups that we need to talk to for a certain topic?

And I don't know, first of all, full disclosure, I don't know that we've actually gone out and done anything with regards to polarization but immediately thinking of the immigrant population or recent immigrants to Canada, and what their views are on Canadian politics and political polarization wouldn't have come to my mind, but how fascinating would their views be given that they have experiences in other countries that are potentially vastly different than ours. So, certainly something for all of us to think about as public servants on the earlier question of who do we engage and how do we engage the public when it comes to this? It's not just people who necessarily do have political party alliances and feel strongly about how government should be and how much they should be involved in our lives. It's people who have experiences elsewhere that can also help us. Eric, did you want to jump in there? I might have cut you off.

Eric Merkley: Yes, I had one quick point just to build on Scott's point of a media literacy. Typically those sorts of interventions are used mostly to combat misinformation. But in a broader sense, media literacy skills could also. There's one stream of research that suggests that all inculcation strategies work at combating misinformation. What those are, is basically educating people about the strategies that people that disseminate fake news, like what watch out for, these are the strategies they use.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: This is kind of what we should look for, and then it inculcates them. So, they may see a piece of fake news and they're, "Oh, I know how to spot that." So, you can imagine an analog for social media. Young children are going to be on social media, and they're going to be bombarded with very inflammatory content, stuff that goes viral, stuff that heightens that political conflict. So, being clear early on that this is a very distorted picture of reality. Journalists can use this lesson too when they write stories based on a handful of tweets that they happen to come across. But getting to children early with this is important. Research has shown that younger people do have higher levels of media literacy, so they might be able to more effectively, wade through social media space without necessarily being overtly influenced by the inflammatory content that they see. So I'm hopeful for the future and I think things are moving in the right direction, but certainly more can be done on that. It's not just about misinformation and fake news, it's also about all of the toxicity that's out there. It is not reflective of the vast majority of the public.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: So that's an interesting point and I won't go too deep into it because I know we want to get to questions from the audience but you just mentioned younger people might be able to navigate social media better. I think another one of our assumptions is that they're the ones who get sucked into all of the negativity and the fake news that's out there.

So again, to all of my public servants there's over 1500 watching us right now, we've got a new generation of public servants coming in and I think using that expertise that these new public servants have on just navigating social media and where to go for news sources and the spotting fake news, I mean, how fantastic would that be for us to be able to then know what's actually happening out there as we're making our policy decisions and our recommendations? So, another great point, Eric.

Okay. So, I could continue to ask questions all morning and/or afternoon depending on where you are. I am going to go to the audience, because we do have a number of questions that have been coming in. So I've grouped a couple of them together. The first ones are around, again, your methodology, Eric, so people curious to delve in a little bit deeper. And the first question about your methodology is whether or not your study includes news or information that was shared on Facebook or other social media platforms. And if so, how that would actually influence people's views when they're responding.

Eric Merkley: So, I'm not sure which part of the blizzard of findings that I presented that refers to? Can you repeat the question?

Aiesha Zafa: Does the study include news or information that's been shared on Facebook or other social media platforms, and how that would influence people's views?

Eric Merkley: So some of the data I draw on is from the Digital Democracy Project and part of that project also did a social media scrape of Facebook, Twitter, a whole bunch of different platforms over the course of that election.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: And so if there's one little bit of a difference in the findings is that there does tend to be a little bit more echo chamber-y sort of behaviour in terms of what people pay attention to on Twitter. So who they follow, who they retweet. All of that tends to be more homogenous than people's overall selection of the news. So, I guess that's where social media data is useful, that is one caveat.

Still not an overwhelming tendency. People are, I think for reasons of virality and all that, they are exposed to voices on the other side on social media, not the voices that'll be the most constructive but they are nonetheless exposed to them. But that is one key difference I think.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: And can you speak about the representativeness or the diversity of participants of the data sets and responses, and have you noticed differences among those participants, or those responses related to racial backgrounds or other kind of identifying factors?

Eric Merkley: Unfortunately one of the key limitations of the Canadian Election Study. So it's a great survey. It is representative of the Canadian population generally.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: But some of the questions that I have to rely on are from the mail back survey, just because of how researchers structured the survey early on. So once you get to a sample of 800 respondents or so, if you want to pick up the share of visible minorities that are in that, you're really slicing and dicing the sample pretty fine there. And so unfortunately with the data that currently exists and for the questions that I use it's not really possible to look at patterns in such fine grain detail. It would be very interesting to see. And that's true especially going back to 1993, the percentage of respondents that are ethnic and racial minorities has gone up over time. But starting at the beginning of this series, it is quite small. So unfortunately it's not possible to slice and dice that fine to evaluate that.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: The CES now though is moving towards big online samples so there's a tradeoff there that I'm happy to speak to but these sorts of concerns moving forward will be less of a problem. But unfortunately then if you're interested in a long term trend, you have to wait a whole bunch of election cycles to be able to do that.

Aiesha Zafa: [Affirmative] It's definitely interesting though I think you mentioned that we think Conservatives are wealthier or they're right wing, et cetera, and we think that NDP supporters are left wing and racialized, and maybe that's not the case. So it would be fascinating to kind of correlate the two or put them together.

Eric Merkley: Yes, it is certainly true. These tendencies are modest, but they still exist. So we do have exaggerated perceptions on those dimensions. And a crucial point that I didn't fully get into with the presentation is that those perceptions are also an important correlate of how intensely people dislike the other side. And that's accounting for policy differences, ideological differences, all those sorts of things. So those perceptions might matter in their own right at intensifying affective polarization.

Aiesha Zafa: Interesting, Interesting. So we do have a few questions in about the United States so maybe we can pivot back to that. And Scott I'll start with you. So, what do you think about the state of polarization in the US and how that actually affects the perception of political conflict in Canada, but particularly situations like the fundamental human rights changes down south? And linked to that because they're combined or they're related questions is, how much of what we think is polarization here in Canada simply a cultural bleed off from the amount of US content we consume? You grow up watching these US shows, I lived down in the States for two years, I was like, "Whoa, culture shock!", but how much of it is from the content?

Scott Matthews: I think that's a really sharp question slash observation. I think Canadians, especially English Canadians, are really immersed in American culture, and increasingly that's American politics too.

[Scott Matthews appears on screen. A sign language interpreter appears in a smaller window on the side.]

Scott Matthews: I have been really struck over the last few years, I'll use [laugh] the typical thing that academics do or in my field, which is my mom suddenly knows all of these prominent figures in American politics. She knows the senate majority, or minority leader who she really dislikes. And she got really invested in his senate race in 2020. I mean 10 years ago, I'm positive she wouldn't have known. And she's not totally uninterested politically and she's a very intelligent person, but American politics just didn't interest her. But now she's disturbingly obsessed with it.

And I think a lot of Canadians are like that. The conflict of American politics attracts interest for its entertainment value if nothing else, but also for what it might be telling us about the future of our most important trading partner, social partner. In any case there's certainly have been some efforts to try and document these effects bleeding over, particularly through the Trump era. And Eric may know a bit more about that than I do.

One thing that I have been trying to speak to in some of my recent work with another fellow at Toronto, Randy Besco, Eric Merkley's colleague, is about the racialization of Canadian's views of the political parties. How racial attitudes are increasingly coming to inform evaluations of the parties, presumably because of perceptions of the parties. And this is something that has really intensified over the past seven, eight years. This dynamic of the parties being evaluated in terms of racial attitudes. And I think that that story's impossible to tell without acknowledging that the era of the Trump presidency took the already quite racialized American political scene and dialed it up in a really remarkable way with the kinds of explicit racialized rhetoric we saw in that era.

But I also think this has resulted in the reframing of all kinds of political conflicts. And it's goes beyond Trump but the movement that arose following the killing of George Floyd, a global movement for racial justice, I really think helped to advance new, and I personally would say important ways of thinking about how law and order issues are racialized. I think for Americans that wasn't news. That's an issue that's been racialized for a long time but I think it's being increasingly racialized in Canada and that concepts like systemic racism is now on every politician's lips and many Canadians understand that term now as well. That's a new development and that is really a reflection I think, of the American political scene and our proximity to it.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: Eric, any comments you want to make on that?

Eric Merkley: Yes, I certainly think that there is something there about US influence in Canada and it, to some degree, starts with journalists that love the narratives that Canada equals the US on anything. So anything that happens down there, it's "Well, what's the story here? Are we the same?" And it just pervades everything. So many interviews that I get from journalists have some sort of framing like that. And so those sort of story frames kind of spill over and affect people's views as well.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: But there are a lot of people, a lot of Canadians that do consume American media. And so, I have some work that shows that those that use social media and consume a US news, they're more likely to harbor misperceptions related to Covid-19. So there is some spill over there. We see that in actual observed social media data. So those that follow more US accounts are more likely to share fake news on social media spaces.

So, there's certainly some things there but it's tricky to really identify a US driven effect on a lot of these questions. So people that watch and read a lot of American news, they're more polarized than people that aren't. Well, they also watch and read a lot of more domestic news and so it's tricky to get the causal effect of that American media influence on anything. It remains to be seen how big the effect is, where it applies. I think there's something there but it's hard to know.

But one thing to keep in mind though, to the extent that we would expect that in Quebec with a distinct media system that is less dependent on the US while there may be things aren't happening in Quebec, but actually everything that I've shown, it applies in Quebec too. And so we see that that distinction between the Liberals and the NDP on the one side, Conservatives on the other, same affective polarization, same partisan sorting, partisan polarization, is happening in Quebec as well. Same magnitudes, change, though they start from a lower baseline.

So, these long trend trends towards polarization, I don't think it's coming from the United States.

There might be something recently that's aggravating certain things but a lot of the long-term trends that I'm talking about, I don't think it comes from the US.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa

That's really fascinating what you said about Quebec there. Because we do have a couple of questions again about, "Are these just trends?" And like everything else, Scott, you mentioned systemic racism and Canadians finally understanding what that means and how it applies in the Canadian context. So the questions were about, "Are we just five or ten years behind the US on this trend, and it's just going to get there?" What I'm hearing is that it's tricky to really figure out why, if we do get there, if it's because of US media and media sources.

Eric Merkley: Yes, I think this is a really good point that you raise. Are we ten years behind? And I think it's important to note that there is one, notwithstanding the media environment and political discourse and all of that, there are very important institutional differences between Canada and the United States that make me kind of skeptical that that's the case.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: In the US, you have institutions structured to basically benefit white rural Americans and protect their power in the American political system. And so the Republican party has been able to avoid discipline by public opinion by catering to those groups and exploiting the electoral college, their advantage in the Senate, gerrymandering, all that. We don't have these things in Canada. And so the Conservative party is much more disciplined by electoral politics than they are. They have to win in swing seats in the 905. They have to win in swing seats in Vancouver and politicians want to win elections more than anything. It's kind of the overriding theme of electoral politics. And so incentives matter. And so unless those incentives change, I don't see it happening.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: Interesting. What about the impacts of a potential economic recession, and what would those impacts have on the future of democracy, the trust in public institutions, and also polarization in Canada? And maybe Scott, I'll throw that to you first.

Scott Matthews: At the risk of rehearsing what I said earlier, I think recessions are the kinds of events in the social world that bring people together.

[Scott Matthews appears on screen. A sign language interpreter appears in a smaller window on the side.]

Scott Matthews: Not in a good way, I suppose, but when a recession, particularly like around 2008 for instance, which was a really profound economic disruption, when these kinds of events happen, partisans tend all to recognize and agree that it's happening. We have lots of evidence that that was the case. Well, I have some evidence for Canada, but in the US and in the UK there's plenty of evidence of this. In the 2008 presidential election when the Lehman Brothers collapsed in the final phase of the campaign when politics was at its height, most polarized, the Lehman Brothers collapse led to essentially a convergence about three quarters of the difference between Republicans and Democrats on the state of the economy disappeared in the space of weeks. So from the respective of polarization, and if polarization is a bad thing, recessions are a good thing. I know that's a perverse thing to sort of say but I think it's true. Now of course, that's kind of the immediate effect of a recession.

In the long run, if recessions introduce a kind of politics of retrenchment, then I think in the medium term looks not so good. And I think political conflicts may become intensified under those circumstances. Of course, the recession in the US was followed by the last six years of the Obama presidency, which were very divisive and set up very well the even more divisive Trump presidency. That's one thing I want to say about American politics is the argument about polarization has been going on since the late eighties, the early two thousands, there was already this huge literature on polarization and fierce debates going on. Is it polarizing? Is it not in the US? I remember reading a book in about 2008 that talked about George Bush being the most polarizing president in history and it couldn't possibly get worse, then President Obama became [laugh] president and the same story and then Trump, good lord! So anyway, it's all to say that this is a really long run sort of phenomenon. I kind of lost track of what point I was making there.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: I think one thing that Scott raised that I think is crucial there, the recession might have de-polarizing effects. And we can see that with the Covid-19 pandemic, too. You know, economic collapse.

[Eric Merkley appears on screen. A sign language interpreter appears in a smaller window on the side.]

Eric Merkley: But government responded with a lot of measures to support individuals that had very massive cross partisan consensus. I have some work that shows that people took a very collectivist tone early on in the pandemic through their support for government redistribution.

What's crucial is the aftermath of the recession. Some people recover faster than others. There's important inequities there. And so after the financial crisis, after the fiscal supports ended for the Covid-19 pandemic and things started return to normal, some returned to normal faster than others, and some people were left holding the bag. And that's where the tensions really emerged. And I think with housing costs and all that, there is an intergenerational tension now in Canadian politics that maybe hasn't been this acute before. And so that might be something to watch. So while recessions might not polarize, their aftermath might play an important, might play some role.

[Aiesha Zafa, Eric Merkley, and Scott Matthews appear in video chat panels. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: Really interesting. I know we're pretty much at time but I have to ask one more question because I think it kind of wraps it up for all of us who are watching, most of whom are public servants. And so maybe if I can ask both Scott and Eric, your views just in a sentence or two.

What role do you think the public service, so that's us, right? We're separate from government, we're the public servants. What role do you think public servants actually have in managing the national perception of the state of polarization, and extreme beliefs, et cetera? And Scott?

Scott Matthews: Sure, maybe I'll go back to the point that both Eric and I were sort of making earlier. There's something about the quality of the political discourse that we have in this country that I think public servants can be important in shaping, especially when they're allowed to communicate directly with Canadians which hasn't always been the case. But when they're allowed to do that, and of course at all times they're producing information that shapes how we understand our country. And I think remaining committed to the kind of dispassionate values, recognizing the existence of a plurality of values and perspectives in this country, we're a deeply diverse country in a number of ways. And I think that as long as the public service is reflecting that through its recruitment, certainly, and through the way it interacts with Canadian society then I think it's probably doing most of its job as far as polarization is concerned.

Aiesha Zafa: Anything you want to add, Eric?

Eric Merkley: Yes, just that people listen when civil servants talk especially at the higher levels. And when we saw through the pandemic, civil servants taking a very active role communicating to the public through the media and people listened, and they cared what they had to say and their messages are more likely to be seen as credible than coming from a political party, and the like.

So there is a communication role for civil servants, and that's likely to be even more important if they're able to tap into unifying themes with their messaging, like Canadians as a whole, reminding people of their common citizenship. Those sorts of things that bring people together. I think there was a lot of good in what the civil service did through the pandemic, not just the federal level, provincial levels probably even more important for some of the policies that were made. But that all matters and just because Canadians are polarizing doesn't mean that suddenly those things don't matter.

Aiesha Zafa: Yes. Well, we are at time. I know we could talk about this for days because the research, the data that you've shown us, your perspectives, very fascinating. Definitely a lot for us to think about.

And I'll say the same thing I said at the beginning, which is we as public servants absolutely need this data. We need to understand it better. We need to know what is actually happening out there with Canadians so that we can make the right policies and provide the right recommendations to government and make sure that we're using the right assumptions and not the wrong ones.

So, thank you so much Eric and Scott, it was great talking to you. I want to thank all of our participants across the country who've joined us today.

[Aiesha Zafa appears on screen. A sign language interpreter appears in a smaller window on the side.]

Aiesha Zafa: I also invite you to participate and register for our next session in the Future of Democracy.

It's the third event, on National Identity and Challenges of the Democratic Cooperation.

So, that's going to be on October 20th. You can certainly register through the Canada School of Public Service website. You'll learn about the ways in which government policies can foster a uniquely democratic sense of national belonging and in turn, strengthen the trust of citizens in the government and of each other, which is another important topic, that topic of public trust that we've talked about over the last few months as well.

So thank you once again for allowing us to be a part of your day and for all of our participants, I challenge you to share a bit of what you've learned today with a colleague or somebody in your life so that we can continue this important conversation. Thank you and goodbye.

[The video chat transitions to  the CSPS logo.]

[The Government of Canada logo appears and fades to black.]