Public(s) and Public Opinion with David Coletto

Season: 1 Episode: 5 [download this episode (MP3, 20.2MB)]

David Coletto, pollster and CEO of Abacus Data, chats with Elizabeth about what public opinion polling is, how publics are identified, and why what "the public" thinks is politically useful. They also talk about how social media (and all the data we leave behind as we use them) are changing how we can get to know different publics. David takes a quiz to differentiate between networked publics and calculated publics (spoiler: he passes!).

Additional Resources:

• In the academic journal article: Social media analysis and public opinion: The 2010 UK general election, Anstead and O'Loughlin do a great job of explaining how ideas about public opinion and how it is measured have changed over time.

• Kennedy and Moss also explore the idea of publics in their academic journal article: Known or knowing publics? Social media data mining and the question of public agency. They do a great job of connecting their work to the ideas of networked and calculated publics.

• Dubois (yep, your host), Gruzd and Jacobson consider how social media data can be used to infer public opinion in this Policy Options article: Journalists should be careful when using social media posts to report what the public thinks, and should tell readers how the data was collected.

---

Episode Transcript: Public(s) and Public Opinion with David Coletto

Read the transcript below or download a copy in the language of your choice:

English (PDF)

French (PDF)

Elizabeth 0:04

Welcome to Wonks and War Rooms, where political communication theory meets on the ground strategy. I'm your host Elizabeth Dubois, and today we’re talking about the idea of public opinion, publics, networked publics, calculated publics... all of the publics! My guest today is David Coletto and I'm going to get him to introduce himself.

David 0:20

My name’s David. Pleasure to be here. For the last ten years I’ve been running and started a polling firm based in Ottawa called Abacus Data, and we do market research, political communications research, public affairs research.

I think the best connection between what you’re all studying and what we do on a day-to-day basis is, you know, we do quite a bit of work trying to understand how to communicate with Canadians, how to communicate with different audiences. And we do that by first understanding what they think; more importantly, what they perceive the world to be like; and then through a lot of the research (whether qualitative or through quantitative surveys), try to gauge what messages, what framing, what arguments are most effective at getting people to where my clients want them to be. So, it’s a mix of all of that and so, almost everything we do is about communications broadly, and a lot of the public affairs work whether we’re working for an association, for corporation, for a non-profit, for a charity—we don’t do political work but if we did, the same applies, and that is: How you persuade people and motivate them to do the things you want to do?

Elizabeth 1:37

Perfect. Thank you. I'm really looking forward to our chat.

I'm going to go out on a limb and say you've probably heard of the idea of public opinion before. Have you heard of this distinction that in academia we make between the idea of public and publics?

David 1:52

Yeah, they're—absolutely, and I think, you know, public opinion, the definition of that has been contested and debated for years. It's a relatively new concept in the history of, I guess, social sciences. But even in the last 100 years, it's evolved in terms of how we define it and what it is. But I think, you know, first you have this sense that public opinion is, you know, today, we traditionally, I think, or we typically define it as what the average person or what a population—so let's say Canadians—think about something. But that's evolved; it used to be what the majority of people think. That came out of thinking about opinion leaders, and groupthinks, and how differences across groups perceive issues.

But, I think, public opinion typically refers to what we in our business call the "general public," right? So those aged 18+ who live in the country or live in a jurisdiction, and our goal is to measure and try to represent the views of everybody through the survey research that we do.

Elizabeth 3:04

Right. And is that only survey research that you do to gather that information? Or do you ever use other kinds of data?

David 3:10

I think survey research is the primary way, or at least it's the way that most people would expect to measure public opinion. If you're going to advocate to a political leader or a journalist and say "Here's what the public thinks", their starting point will be "Well show me the survey", right?

But we have other ways of measuring public opinion. Historically, or traditionally, people would look to the media as a voice of the public. So editorials and the content that, particularly newspapers would produce, would be perhaps a measure of public opinion. Now today we've got social media. We have the ability to not not just ask people questions through our traditional surveys, but hear or see what people are saying, passively. And so social media has become a way that we now measure public opinion. Is it the best measure? It's contested. But, it is a measure of public opinion.

And then there are other things. Oftentimes you'll hear an MP say, "Canadians think this" or "Canadians think that" and he or she might be concluding that based on a handful of conversations he or she has in their constituency or by phone calls they receive at their office. So I think everybody has a different conception of how to interpret or determine what public opinion is.

But I think, you know, I'm in the business of doing surveys. I will say: perhaps the most consistent way is by doing a representative survey, but it's not the only way that you can get a good feel about what a public is thinking about something.

Elizabeth 4:52

Yeah, that tracks really well with the kind of academic perspective from pol comm literature that we've seen.

You know, the way you described public opinion from representative surveys, it's, you know, it goes back to roughly the 1930s with Gallup being like, "Hey, I have a better way of predicting who's gonna win the next US election. I can guess who the president's gonna be because I've talked to this sample of Americans—not all of them, but a selection of them that roughly mirror the larger population." And I think you're right to say that, when we say "the general public" or "public opinion", we often are thinking of those kinds of representative surveys. And we have this base in statistics that tells us why that should make sense.

But today, I want to focus more on that idea of "the public". So, let's take Canadians as an example: that's one version of what "the public" could be. But we could also do provincially focused samples, or even down to a particular community. You can still have representation to a population that is not the entire Canadian population...

David 6:05

Right.

Elizabeth 6:05

... And so there, we get into that idea of there possibly being multiple publics. And so, I wonder, when you're working with clients to try and establish what kind of survey you're going to run and what the population is going to be, do you guys have conversations about which public they care about? And, what kind of language do you use when you're talking to them about establishing that population?

David 6:28

Absolutely. So let me give you some examples.

One that is perhaps the least relevant to your class—but probably the easiest to understand—is if you are a cat food maker, the most relevant public for you is owners of cats. They're the ones going into the pet shops, and/or going on Amazon and buying cat food. So, if you're going to want to understand the ad audience that matters to you, well, then "cat owners" is particularly important.

But from a more political comms perspective, or public affairs perspective, one of the things we often do when we talk about political research is the difference between those who can vote and those who can't. Eligible voters, from a public affairs lens, are more valuable than those who can't [vote, because] they have a direct say in the country. There's lots of people who live in Canada who can't vote and are impacted by the decisions that political leaders make, but they actually have no say—or direct say—and who gets to govern them. So that would be one way of slicing it up.

But then we can also take it one step further and say: "Well, I'm only interested in likely voters". So even if you've got 20+ million Canadians eligible to vote, we know that in the last few federal elections, for example, about 65 to 70% are going to turn out. So, you distill that even [further] down. And that means that public is probably a little bit older, a little bit more wealthy. And, so we're talking about a different group of people. So the more that you distill down to a smaller group, who you're talking about will inevitably change.

One of the other subgroups, or publics, that we often talk about in public affairs is "engaged Canadians" or "informed Canadians". Those—and it's about 30% of the country—who actively consume large amounts of media, perhaps are most likely to participate in political life beyond just voting. So they may be members of parties, they may be donors, they may actively tweet or share on Facebook their political views. And that public is particularly important from an advocacy or communications perspective, in that not only are they consuming more information, but they are influencers themselves. And so if you understand the influencers, if you understand where they're leading everybody else (so they think X), then maybe in a few months the rest of the country will follow them. So it's a particularly important public that we often look to because they are signals to where perhaps the rest of the country might be going.

Elizabeth 9:11

Yeah, that's super interesting. And, you've used the term "opinion leader" a little bit- I've actually got another episode about opinion leaders, so the students are going to be introduced to that idea. We talked about opinion leaders in the "on the ground, getting out votes" context with a campaigner. But what you're describing strikes me as being connected to the idea of early adopters. So, looking to the people who, [when] new tech comes out, they want it first. Or [if] there's a new idea percolating in the public, they're the ones who are going to inform themselves first. And we see them in a bunch of different areas, which is super interesting.

But it brings to mind a question for me of: What's the difference between the idea of a public and just an audience or a segment of the population? Is there a meaningful difference between how we conceptualize "public" and just segmenting out the group?

And so, maybe I can give you a bit of background on how pol comm literature has talked about publics. They've looked at the idea of publics as being sometimes connected through community, connected through shared experiences, [or] connected through shared goals or aspirations. And the more we get into those sorts of cultural definitions of public, the harder it is to do nice, clear, systematic assessment (which obviously is necessary for survey research). I was wondering what you think about that idea of "public" needing some sort of social connection to be considered a public?

David 10:45

Yeah, I'm increasingly a believer that that matters to the outcome. And when I say, "to the outcome," I mean: "to what that public does, or how they move or act, collectively." So if I, as a researcher, take a sample—say I interview 1000 Canadians—and I decide I want to split them based on their views towards climate change, or government spending, or some demographic profile. That's useful to some extent for me, in terms of describing who they are, and why they matter. But, I think, to your conception of public, it will matter more if those groups I create (and I'm creating them based on my own decisions) actually have an identity—they identify with that group, they see common cause. And when you have that, that is a public that I think is much more meaningful because it's going to be first a durable kind of structure that then allows us to think about that group as having certain attributes. And I think that is, in my mind—and we're talking somewhat conceptually here—but that's the basis for us to understand how to structure it.

But that conception of public as being something that's first felt by an individual within that group, or is led by a leader or set of leaders, gets us to a point where they matter more. And the reason it I think matters is you hear all the time, political leaders, those who have power, referencing publics, right? Referencing the views of these groups. And that's where, if [that public is] well defined, if they have common connections, they are more likely to be influential and to be listened to. And so I think that's the value in it.

But I do think that there is a distinction between a segmentation or just defining people by their characteristics or their views or their behaviours (which is what we normally do in market research in terms of segmentation). We never use the term "publics" to describe the subgroups. Because I think there's something deeper in that public view.

I'll give you an example. When I was doing my study in grad school, I was fascinated by the identity that Québecers and Scots might have to their sub-national or national state, and the role that that played in their views towards sovereignty or independence. And so when I think about Scotland, for example, there's a question that researchers often ask: "Do you see yourself as more Scot than British, or equally British and Scottish, or more British than Scottish?" We use that question to divide publics—for example, in Scotland, you do see a distinction and that's partly built around identity. So we can speak to those who identify more Scottish as a subgroup (or a public of that nation)—that, I think, affects their behaviour, and there's a common identity. And we now see a strong independence moving in in Scotland and a referendum a few years ago—that, I think, was built on [this idea of identity and sub-national publics]. And so, when the current First Minister of Scotland [Nicola Sturgeon]—she herself supports separation, or the National Party—when she speaks of the public in Scotland, she, I think, is speaking about [Scottish nationalists] as opposed to the broader population because they have alignment and common cause.

Elizabeth 14:23

Yeah, that's a really, really super clear example. Thank you for that. It really illustrates the difference between the idea of a public and segmentation. Like segmentation: we can choose whatever characteristics we want, and we are going to get information out of it by cutting up the larger population in those ways. But that doesn't necessarily align with those cultural constructs.

David 14:49

Right.

Elizabeth 14:50

Super interesting.

Okay, I want to switch gears just a little bit because earlier you mentioned that social media data among other kinds of data can start to be used, to start to, I guess, infer public opinion to a certain extent. And one of the things that in academia has started to come out are these ideas of networked publics and calculated publics. Have you heard of either of those terms before?

David 15:16

I've heard of networked but not calculated. Tell me more.

Elizabeth 15:19

Okay, great.

So the idea of network publics is these are publics that basically are formed of users of a particular social media platform or other digital tool. And so the idea is, if you are on Facebook, you become part of that Facebook public because you are contributing to conversations there, and you're having experiences there, and you're building relationships and fostering them, continuing them. The idea is not to say that the experiences you have on Facebook are distinct from the rest of the world—like, just because you're part of a networked public on Facebook doesn't mean you're no longer part of the rest of the publics. But it has this specific focus on the social media experience because social media tools are developed in a way that afford us particular kinds of interactions and so that could create a new kind of public for us. And then we think of that as being useful because we can think about going to Facebook groups and pulling out information about, like, "Oh, yeah, you think that more funding should go to Canada Parks because you really are a park enthusiast."

The flip side is the idea of calculated publics, where we're looking at the ways that social media platforms design algorithms that essentially create publics for us. And so we have things like, "Customers like you bought this book" when you have bought something on Amazon. And so they're envisioning a public, they're envisioning what you and "your people" are like based on people's purchasing habits. Or, you know, Twitter having "This is trending in Canada"—like they've come up with some definition of "the Canadian public" in order to tell us what's trending there. And so, those are the calculated publics.

David 17:20

Like Instagram knows that I loved road cycling because on my general feed, it's like 80% of cycling stuff, right? Like, and I've never told it [that] I do [like road cycling]. But it knows based on what I'm looking at, and who I follow, that that's what I'm interested in. So it's connected me to that community in a way that I didn't choose to join it. So is that the basic [idea]—that the algorithm has forced me there?

Elizabeth 17:41

Yeah. And it's like, they have envisioned a public that you are part of without you self identifying as being part of that public.

David 17:52

Which is in some way, if you expand it to what [we talked about in] our conversation earlier, it's no different than me as a researcher deciding who fits into my nice, neat box of a segmentation.

But in this way, they're using an algorithm and data—big data—to determine that, right?

Elizabeth 18:07

Yeah, exactly, exactly. And so then there's, you know—based on the conversation we had before where we distinguished usefully between the more cultural-based ideas of what a public is and the power that having those social relationships as part of our conception offers, versus the segmentation approach that isn't necessarily informed by those same social structures—you could question whether or not these calculated publics are in fact publics in the useful sense of the term, right?

David 18:40

Well, that makes total sense. And I think of—and I don't use Reddit, actively or hardly at all, but I was listening to a podcast by the founders of Reddit and what their vision was: to create a platform that created communities—purposely created communities. And so when I think about Reddit [I think about it] as built around networked publics that are meaningfully connected, that participate and are actively doing it.

You know, if I am going to use that platform as a way to measure that public's opinion, or the views of that [public], I have to be aware of what being a member of that public means and what's involved. Versus a [calculated] public [which] is probably a lot more artificial and it's not, you know, you're not connected to it. That's an interesting way of seeing it. And I think you see parallels outside of social media. I mean, social media is really just a microcosm of life, maybe it exaggerates certain parts of it, but I think you see the same kinds of distinctions in just our general life and how society and people are connected and network with each other.

Elizabeth 19:46

Yeah, I think you're right.

And that idea of networked public—I think the example of Reddit is really useful because people choose to go and join a subreddit. They choose to participate in it, and they engage in it, and the roles of moderators on Reddit really highlight the social structures that create that public or inform that public.

Whereas these calculated publics, you know, it's kind of like, well, we can sell you the idea of "You have access to this public because we have all of this big data and algorithms, and you don't know how they work. So we can convince you that they're magic, right?" And so, then there's one argument that it's a marketing technique: it's a strategy for segmenting your population. You're not doing it by people like yourself making decisions about how it should be segmented, but you are trusting an algorithm that's been designed by people to do that. And ultimately, there's this sort of weird tension between describing what you've actually done—which is chunking up the population based on their behaviours online—and the idea that, well, it's social media, so you should get that social information, too.

David 21:00

Right. I wonder if the effectiveness that we've seen some examples, if you've seen— I mean, the HBO movie Brexit is not the best description of the Cambridge Analytica [scandal], but it's an effectively entertaining way of showing what happened there—where they use these calculated publics to in a way manipulate them into voting in a certain way. And so, is it more effective or less effective when you're trying to mobilize these publics?

Elizabeth 21:35

Mm hmm.

David 21:36

When, if you're a member of a calculated public, you don't know you are, right? You don't know that you've been assigned a score of likelihood to vote Brexit or not, or vote Conservative or not, or vote Trump, or Liberal, or whatever. Then it becomes a question about ethics and how do you actually engage these publics? Whereas if I'm a member—I know I'm a Canadian, or a member of the electorate—that probably creates an ability for me to engage with that within that public, as opposed to one in which I'm a member of but I don't know I am, and I've been artificially placed there.

Now we're going maybe above and talking about ethics and philosophy a little bit... way outside my—but those—that's an interesting—I always try to think about, like, "What motivates people?" And, "Why do we get them to do what they do?" And not knowing that you are being targeted and grouped based on your past behaviour, [and] that we'll be able to predict your future behaviour is both scary, but also likely very powerful.

Elizabeth 22:41

Yeah, I think the ethics around that are really interesting. I think it is a topic for a different episode, but it makes me think about why we care about the public at all, right? And why public opinion has become part of our democratic systems.

The basic argument is public opinion helps people see what other people think and care about and then [people can] further develop their own opinions. It helps [to] kind of get a sense of whether or not their political systems are reflective of their needs and wants and desires. It helps political elites, as we tend to call them—so like politicians and government and journalists—it helps them figure out what to focus on. And so, we think about public opinion as playing this really crucial role in that information process and policy development process and communication process. And if people don't know that they're part of the process, can it still be that useful in our democratic systems? Like, if we were to only be relying on calculated publics, would we lose something?

David 23:53

Great question. I think we would, I just don't know what. I think it's so new—it's such a new concept. And I think you lose that sense of community, you lose the motivation, the ability to feel connected to it, but it just feels so artificial, right? It feels... I think it sounds like that world none of us want to live in.

And I think it also adds to the point of why public opinion is so powerful a concept: it is often used to justify decisions. And as you said, we live in a democratic society in which every person has a vote, and therefore there's a natural connection—a natural relationship—between the concept of public opinion in that democratic system because we care what each individual thinks of things. But, that's a relatively new version of public opinion; in the past, leaders of social groups—elite opinion—was actually the indicator of public opinion. Anyways, it's a really interesting way of thinking about our community—our collective community—and how we then both self-identify but also use that concept to justify things that we do.

Elizabeth 25:10

Yeah, definitely. Alright, so my last question for you is just a quick little quiz. And so I'm wondering if you can explain to me the difference between networked publics and calculated publics.

David 25:25

Oh, okay. So a networked public—and we're talking [about] this through the lens of social media, is how I'm conceptualizing it—a network public would be one in which I actively participate on say Reddit, or Facebook or Twitter, and may through that get connected with a group or a broader group. And that's defined by my participation and my activity through it.

A calculated network would be one that is created through some what we would call [an] algorithm—some artificial structure that I don't even know I'm a member of—and I'm not actively participating. So it's inferring that I'm a member of it through my other behaviours or things I've done, as opposed to me knowing I'm a part of it by actively participating in it.

Elizabeth 26:11

Yeah, that's pretty great. That's very great! The only little thing: you said "calculated network" is "calculated public", but I understood what you meant.

David 26:20

Right. Because we were talking publics and networks. Yes, yes.

Elizabeth 26:23

I know, all of the terms interweaving with each other.

David 26:26

All right, I'll take, I'll take an 8 out of 10 there. That's okay.

Elizabeth 26:29

At least 8.5

David 26:30

Okay!

Elizabeth 26:35

All right, that was our episode on the idea of public opinion, network publics, calculated publics, multiple publics.

We talked a lot about how public opinion is often thought of as kind of the aggregate of all of, of populations, opinions or ideas on a particular thing. But then we dove into the idea of having multiple publics depending on who your population is, who you actually care about, who you conceive of as being part of a community.

We talked about the cultural aspects and social dynamics of what constitutes a public and contrasted that to simple segmentation, which often happens in survey research.

Then we jumped into the idea of network publics versus calculated publics. Network publics being ones that are communities of users of a given social media platform. Whereas calculated publics are the publics kind of conceived of by the platforms themselves—the algorithmically driven creators of "people like you" and "friends of friends" on Facebook, and those people who are contributing to whatever's trending in your region on Twitter.

If you'd like to know more about any of these concepts or theories, go ahead and check the show notes or polcommtech.ca for more resources.