top of page
Radio show microphones

Wonks and War Rooms

RSS Feed Icon_White
Spotify Icon_White
Google Podcasts icon_White.png

The Two-Step Flow and Opinion Leaders with Nick Switalski

Nick is a long-time campaigner and former Hill staffer. He and Elizabeth talk about opinion leaders and the two-step flow hypothesis. In theory these opinion leaders use social pressure and social support to influence their friends and family but Nick argues that kind of influence just isn't that useful in most federal campaigns.

Additional Resources:


Episode Transcript: The Two-Step Flow and Opinion Leaders with Nick Switalski

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

Elizabeth: [00:00:03] Welcome to Wonks and War Rooms, where political communication theory meets on the ground strategy. I'm your host,  Elizabeth Dubois. I'm an Associate Professor at the University of Ottawa. Today we're going to be talking about the two-step flow hypothesis. And my guest is Nick Switalski. Nick, can you tell us a little bit about your background in political communication?


Nick: [00:00:22] Sure, okay. So, my name is Nick Switalski. In my everyday life, I'm a government relations professional working with a post-secondary institution. But otherwise, in my private life, I still maintain a number of campaign-related ventures. And for about ten years of my life I was on Parliament Hill in various capacities at a Member's office and a Minister's office, up until 2015. That included being the Director of Issues, Management and Parliamentary Affairs for Rona Ambrose, who went on to be the interim leader of the Conservative Party and was, at that time (when I was working with her), the Minister of Health. I managed a number of campaigns in the GTA [Greater Toronto Area]. I've worked with two leadership candidates for the federal Conservative Party; worked in different provinces and territories. And so, yeah, that's more or less my CV in a nutshell.


Elizabeth: [00:01:11] All right. So the two-step flow hypothesis [see: Two-Step Flow of Communication: An Up-to-Date Report on an Hypothesis] is a media effects theory. It basically says that instead of mass media and mainstream news being able to directly affect the general public, they do so via these opinion leaders who serve as a bit of a bridge between that news outlet and the general public. Opinion leaders care a ton about a particular issue — say [for example] it's Canadian politics. They are going to get all kinds of information from wherever they can and really inform themselves. And then they will share pertinent information with their friends and family and other everyday associates. Those people are people who probably don't care quite so much about politics, but they do care what their friends think, and they do care what their friends say is an important issue. And so that's the way that the mainstream media would have an impact on the general public.


Nick: [00:02:10] Okay.


Elizabeth: [00:02:12] Now, when we think about the way politics works, we know that it's not just the public and the media. There's a bunch of different players that come in. And so what I'm wondering is whether or not this hypothesis makes sense for you? Do you see these ideas of opinion leaders being important mediators in the work that you've done?


Nick: [00:02:33] I mean, to a certain degree. When you say opinion leader, I mean almost immediately go to the pundits. But that's not what you're talking about. You're talking about individuals in a community who consume a lot of news. Now, I mean, I'm trying to differentiate an opinion leader between everyone who watches the news. So, I'm not an opinion leader just because I watch the news, right?


Elizabeth: [00:02:51] Correct. You need to be not only somebody who watches the news or consumes other political information, you need to also be somebody who's then willing to share that information with their friends and family. We know the "Don't talk about politics at dinner, it's going to create too much of a stir" kind of mentality or the politeness approach of, "I don't want to rock the boat" — so there are people who are going to choose not to try and influence their friends and family. We're looking at opinion leaders as the ones who both have the information and are willing to share it. 


Nick: [00:03:25] Okay. So I guess this sort of concept really depends [on] the time in which we're talking about it.


Nick: [00:03:30] [In politics,] we're [often] talking in between elections - when you have governments established, opposition parties trying to get an edge, and ministers and government trying to oftentime[s] defend or push out their own message. Both [government and opposition] are tending to speak to their [political] base and speak to the individuals whose mind can be changed. And that often involves trying to negotiate directly with the media themselves because the media, in broad terms, decides the agenda. They decide what we're going to be talking about that day. I've heard one commentator refer to them as the boards in a hockey game: you can try arguing against them all you want; at the end of the day, they're going to say, "We're arguing here. What's your different angle? Is it important? Do you feel it's important? Do you feel it's not important? And how does your party feel about it? And what would they do differently if they feel that it's not a good issue to be talking about?" So oftentimes in those scenarios, we're trying to issue manage our way out of them or trying to change the channel to something we want to talk about. And the idea behind that is: we have our key messages, we have our main message, we're talking to our base, and we're trying to get accessible voters in a position where they could see themselves voting for us on election day.


Nick: [00:04:37] When it comes to an actual campaign, we're targeting a very specific subset of individuals: our base and those who we view as being accessible voters.


Nick: [00:04:46] There was a particularly gifted strategist in the Harper PMO [Prime Minister's Office] in our early- to mid- mandate named Patrick Muttart who's gone out to do bigger and better things. But he created these concepts, these individual pictures if you will, of the average conservative voter in an urban area, in a suburban area, in a rural area and, you know, average liberal voters, etcetera, as an idea of "Here's your average conservative voter. Here's John. He's a small business owner. He's concerned about this issue and that issue." So when you run into John, you should talk about this issue or that issue. You're trying to really flip the channel to issues that your party's strong in and the other party is weak in in order to get that vote.


Nick: [00:05:28] So, have we really focused on opinion leaders in this respect? I wouldn't say so - in particular when you're talking about a political narrative - because you would almost have to fashion your medium to speak directly to those individuals outside of your traditional media channels. When you're talking about social media, you've got the actual key message encapsulated within a tweet or a Facebook post so the voter can actually read it directly from the political type as opposed to the media where it's put through that medium. But all of this is scattershot.


Nick: [00:06:05] I mean, we can talk about advertisements, we can talk about geo-targeting...


Elizabeth: [00:06:09] Well let's talk for just a second about those social media posts that you're mentioning. Is there ever a thought of, "We want to use the network that these social media create"? So it's not just about getting your "I'm pretty sure they're going to vote for me anyway" secure voter to be reminded of why they're voting for you. But also to get them to try and pull in people who are like them that might not be certain yet.


Nick: [00:06:35] So I think maybe the best analog I could use for that would be a caucus endorsement, or a previous candidate or previous leader's endorsement. Because when you're talking about a political bubble, you're looking for an opinion leader not just in terms of someone who's respected in a small social circle - because that's kind of the concept you've fed me with initially is that, you know, you have one person in a group of 30 who is watching the news and then going and talking to [their] friends or family about it - but that's often not what political types, especially at the federal level, are looking for. Maybe at the municipal level (it's a bit different). But we're talking federal. We're talking about hundreds of thousands and millions of voters out there. So it's almost too granular to go to that level.


Nick: [00:07:18] But we do sometimes rely on - and this can be for better or worse, I have my views on it - individuals who we feel may carry some weight. For example, the Tories [note: “Tory” is a nickname for members of the Conservative party] are in a leadership race right now and one campaign in particular has chosen to focus very heavily on caucus endorsements. And the idea behind that is [that] if [an] MP [Member of Parliament] from riding X says you should vote for this guy or gal, then every eligible Conservative voter in that riding will do so. Can it hurt to do that? [From] what I've seen on the grounds: No. [But] I've also seen and experienced [that] they shouldn't really spend a lot of time doing it because, at the end of the day, it doesn't really translate to votes.


Nick: [00:08:02] Caucus members and previous candidates have one vote. They may have received the votes of individuals in the riding in the last election, or however many [elections] long and have supporters in the riding, but they all hold various different views themselves. In the last [Ontario] provincial PC leadership race, we saw Doug Ford go up against Christine Elliott (the two major contenders in that race), and Christine Elliott had almost all [laughs] of the PC caucus behind her. I think Doug had a small handful. I think he had two - maybe three - but it was a very, very small. And he ended up winning. I mean, at the end of the day, you are relying on these - I'll take your term - "opinion leaders" to deliver votes for you. It's almost like a cop out: "I don't want to do the hard work of making phone calls and knocking on doors and running a campaign, because that guy is going to do it for me." And I've seen many campaigns fail from that.


Nick: [00:08:54] You know, you have a ton of people who enter in your office and green campaigns saying, "I want to help you out on strategy", "I can deliver this", [and/or] "I can deliver that” [note: “green” is slang meaning new and inexperienced].  Okay, well, show me the list! Show me the people! Are they coming here? Where are they going? Like, I'll make you a poll captain, sure. But I'm not going to trust that you're going to be the one to actually run the campaign in this group that you've described to me.


Elizabeth: [00:09:13] For sure. So it's interesting because we think about opinion leaders as this really specific community based version of influence [for more information on different types of influence, see: The Multiple Facets of Influence: Identifying Political Influentials and Opinion Leaders on Twitter]. And so the idea behind opinion leaders at that community level is they're using social pressure and social support to influence their friends and family and other acquaintances. And so [opinion leaders are trying to influence] people that have shared experiences [with them]. There might be a fear of being kicked out of that community and not having that sense of belonging anymore if you do the opposite of what this opinion leader has said. But that's pretty much dependent on people's vision of their everyday life.


Elizabeth: [00:09:54] The kind of influence you were just talking about is a step up if we're [imagining] a ladder of influence. The step up [from opinion leaders] is: these are leaders in my community that I don't hang out with on a day-to-day basis but I care to some degree what they think and I take cues from them. And then we can go even further up the ladder to: celebrities or federal politicians or other people who aren't even based in your community at all, but, somehow you feel connected or you care somewhat about their views.


Elizabeth: [00:10:27] And I think that [it']s really important to to start to tease apart the different kinds of influence that show up, because I think the mechanism you use to convince your friends or family to go vote is probably different than the way you do it if you are endorsing a candidate because you yourself had gotten votes at some point. Does that track?


Nick: [00:10:48] Yeah. I think the practical utility challenge we're running into here, when I'm trying to describe the way you operate when you're actually in government or running a national campaign (Disclaimer: I've never run a national campaign - I've been involved in many, but at the local level), [is that] you run into the sheer volume of individuals you're trying to contact. And it's almost too granular to say, "Okay, well, let's contact every group leader," if you even had a list of those - because that's what it boils down to: Do you have the data and contact details and [the information on] who's in their group. You don't have that at the federal level. I don't know if you would have that at any level.


Nick: [00:11:22] But, one example that I could use that I think meets up directly with what you're talking about here is a nomination race in a riding (and this can be for any party) where your primary goal is to sell memberships [note: for more on the role of party membership sales during a nomination race, see: Membership sales being tallied as Ontario PC leadership race and Battle for the ballot: Inside the bitter nominations that divided the Ontario PCs]. In the Conservative Party that involves selling them for about 15 bucks a year now. And that [membership purchase] affords you the opportunity to be involved with the party at the local level, most particularly [to] have a vote in a candidate nomination process to choose who's going to be the candidate in that area. Or, in the case we're in right now, choose the next leader of the party itself.


Nick: [00:11:58] So, when you're selling memberships, there's a number of ways you can do that. But ridings, and even safe ridings - I live in the riding of Niagara West and we had a nomination race here in which a good friend of mine, Sam Oosterhoff, [who] is now our Member of Provincial Parliament,  had to run and win against several other challengers. The other major [challenger] was named Rick Dykstra, a former parliamentarian and, at the time, president of the PC Party [note: PC is short for the Progressive Conservative Party of Ontario]. He's been accused of a number of things afterwards, so he's disappeared off the public stage. But at that time, [Rick] was still a relatively prominent figure and had really good organization. Now, this is a very, very safe conservative riding. It's a relatively rural area, a lot of small business owners. And so the race to win this seat with these two individuals who sold what's viewed as a lot of memberships (Sam sold roughly 1200, Rick sold roughly 2000). And it came down to who can get their vote out most effectively. Now selling these memberships - because, you know, 1200 or 2000 [is] a relatively small vote share when we're talking about a provincial seat or a federal seat - but that's what it comes down to when we're talking about nominations. Because it’s a family of individuals, the Conservative membership-holding base in that riding, who gets to choose these things. So you have to [think], "Okay, who am I going to sell memberships to?" And the strategies you use for that is, "Let's start with your friends, associates, and neighbours - your “FAN” model. So, you know, walk down your street, chat with your family, say, "Hey, how's Uncle Ned doing? He's still living next door? I need him and his kids. I got to sell those memberships."


Nick: [00:13:38] And then you'll do [...] this is a strategy that I've used a number of times called hosting a coffee party. You literally just [go to] a friend of yours or an associate or a neighbour [and] say : "Hey, would you mind doing me a big favour? I appreciate that you bought a membership but, all your friends who live in the area, can you have them over and say, 'Hey, my neighbour - my friend - is running for office. He just wants a chance to meet you and [let you] hear about his case. Would you mind? Just come on by, I've got a pot of coffee on. [Would you mind] meeting them and saying hello?'" And then they show up. And then you or the candidate or whomever else is running shows up and says [to them], "Listen, I'm doing this, that and the other thing," and you give your spiel. And then afterward[s], almost all of them will buy a membership.


Nick: [00:14:17] So that is the sort of thing that you were talking about right there. But it comes into play when you're at a very granular level when you're talking about federal or provincial politics. I presume at the municipal level it has a certain degree [of applicability there] too, because we're talking the ward level.


Elizabeth: [00:14:30] Yeah, that's a really great example. It really illustrates the two-step flow hypothesis and the role of opinion leaders exceptionally well.


Elizabeth: [00:14:37] And so you're saying at the granular level it works because you're dealing with a small enough population that you can create good strategy around that personal influence mechanism. When we're looking at larger populations, bigger elections, you develop different strategies.


Elizabeth: [00:14:58] So, one of the things you've mentioned briefly in this conversation so far is the idea of what data you can get...


Nick: [00:15:08] ...Umm, yeah...


Elizabeth: [00:15:08] ... We've got data that comes from Elections Canada and [that] then on Election Day gets updated with whether or not you've actually voted - to a certain extent before [Election Day], but really the major work of that is Election Day.


Nick: [00:15:20] Yeah.


Elizabeth: [00:15:20] There's also all of the work that's been done before that to say, "Here's a mail out. Maybe she'll respond to the mail out." Presumably there's also email versions of that and other digital versions where I could sign up on the website, give new info. Cool. So you've got this chunk of information -  and we're not going [into the] Cambridge Analytica [scandal], I don't want to talk about the ethics.


Nick: [00:15:44] Nor do we!


Elizabeth: [00:15:46] [Laughs] But what I do want to do is ask you about one update to the two-step flow hypothesis that some people have been posing. And that is a version of a one-step hypothesis [see: The One-Step Flow of Communication]. And what they say is: because we can get so much data about people, we kind of eliminate the need for that personal influence. We eliminate the need to have a mediator who cares a bunch about the information, who's then going to do the work to make it relevant to the particular other people. You don't need somebody who's going to apply social pressure and social support and be like, "Oh, well, if you don't vote... Like, all of my friends vote, I don't know how you could be my friend if you're not going to go vote." You don't need that because instead parties could create things that are so targeted to each individual that they already feel like it's personal. They don't need that extra sense of community pressure to do it. Does that make any sense to you?


Nick: [00:16:48] I mean, yes and no. So, I guess I would take issue with the term you used, which is "need". Because need presupposes that, [in] the initial scenario of [the] two-step method, the need was crucial. That you had to have an opinion leader in order for political messaging or any kind of strategy to be effective; you needed opinion leaders. And I disagree with that concept. I think they exist, but I don't know if they fulfill a particular need.


Nick: [00:17:17] When you get to the granular level, like perhaps nomination contests levels (the example I used earlier, I would argue that there's weaknesses [on the effectiveness of social media] here. The need, I think, exists still for those opinion leaders and social pressures in order to sell memberships, because that's all about direct voter contact. Social media has not yet reached the level where you could sell 1000 memberships at a riding level. There's a very particular kind of person, for all parties, where you have to establish that personal contact with real people in order to get that kind of a relationship going on that relatively small scale. [In] ridings you're dealing with  [some 130,000 people], and within that very large subset, you're looking for 1000 or fewer who you might be able to sell something to or [be able to] get energized, engaged in a very specific "Help me win this nomination" [way].


Nick: [00:18:17] So social media has its weaknesses. The first which is that it doesn't really have the targeting capacities of what I would need for that sort of a thing. Like, if I am running for a nomination in Niagara West, I would be able to geotarget the riding in terms of Facebook or other social media to say, "Hey, Nick's running." But there really isn't much else that you could use on it to hit people. Maybe you could find a fan group - groups of people who say, "I subscribe to this" - but even that [has weaknesses because] Facebook will tell you "this will go out to 25 people," but I don't even think they tell you the members of the group anymore. I guess there have been privacy concerns, right? So there are severe limitations on what you can do, and it doesn't really allow you the same kind of direct voter contact that you need in order to succeed at that really ground-level politicking.


Nick: [00:19:15] So, yeah, I mean, I hear what you're saying on the opinion leader thing; I think it is very important at the granular level, [but] I'm not sure I'm with you on the higher tier.


Elizabeth: [00:19:26] Alright, that makes a lot of sense to me.


Elizabeth: [00:19:28] Do you think that the tech and our capacity to collect and analyze data could develop to the point where that could be useful? Or do you think, no, you still need those humans on the ground.


Nick: [00:19:41] It's a good question because I have actually jousted with individuals on this in the midst of a campaign. It's more of a conceptual thing...


Elizabeth: [00:19:51] ...Mhmm...


Nick: [00:19:52] ... So, in a number of nominations that I've helped with and a number of campaigns I've worked on, the social thing is very new and exciting. And it gives you a lot of data in a really cool heads up display where you can see, "Oh, look how many clicks I've got," or, "Oh, look how many impressions this, that, and the other thing [received]." It tells you a lot of numbers. But what it doesn't tell me is: John Smith who lives at 123 Fake Street is the one who hit the like button. No [it doesn't tell me that].


Nick: [00:20:20] So, you have to see this data and then you have to translate it into a very succinct set of information that you can use to get your vote out. Because you can buy all the ads you want - you can spend millions, tens of millions of dollars on ads - [but] it doesn't tell you who's looking at them. It doesn't tell you the personal information of who's looking at them. Like, I need your name and address. That's all I need in order to get you out and determine you can vote.


Nick: [00:20:48] That's the other thing too: there's no guarantee that what's looking at this tweet or this Facebook posting is even a real person, or a real voter, or lives in my writing. I need to know all of those three things in order to determine how I'm going to use that information, because otherwise it's kind of useless. So even if - I shouldn't say "even if" because I'm pretty certain I know the answer to this point - I don't think people want to get to the point where their social media interactions are able to be boiled down by political parties to where they live. I don't think people want that and political parties, I think, might even not be the worst offenders. I don't think [people] want companies knowing that. I don't think they want states knowing that. I don't think they want pretty much anyone knowing that.


Elizabeth: [00:21:34] Yeah, there's some pretty clear evidence in polling and survey research in Canada that says Canadians are not really down for that [note: for example, see Elections Canada’s Canadians’ Privacy Expectations]. And there's a whole bunch of conversations about privacy legislation and how political parties don't actually have to follow the rules companies have to follow [note: for more context, see OpenMedia’s article, Canada’s Political Party Privacy Hall of Shame]. But, at the end of the day...


Nick: [00:21:53] ... Well that's what happens to the actors that write the laws, isn't it?


Elizabeth: [00:21:56] Mmm, yeah. I wonder why!? [soft laugh]


Elizabeth: [00:22:00] I have one more question for you. Something that you hear a lot when you talk to politicians and say, "How can I support you? What can I do?" They say: "Go door knocking [to support me as an electoral candidate]. Go door knocking in your community." And the two-step flow hypothesis would suggest that the "in your community" bit is really important. Going door-to-door and seeing neighbours who might recognize you from walking your dog or dropping your kids off at school or going to the same gym. There's something about that recognition that matters from the perspective of this hypothesis. Do you think that makes sense, or really, is the value of door knocking just any human getting to the door?


Nick: [00:22:43] If it does, it's hard to qualify. How do I prove that? How do I replicate these conditions with a stranger? So that's the danger number one, and it's something that's present throughout political campaigning across the board. How do I know that this particular strategy worked in this riding versus an entirely separate one? Because you can never replicate those conditions. You'll have an entirely separate national message. You'll have an entirely separate candidate. It's very difficult to replicate the conditions, to demonstrate that this is, in fact, a thing.


Nick: [00:23:09] But what I can say, and I'll give you a particular example, is [that] candidate data that is gathered by the very specific political candidate who is stumping for votes, is viewed by campaign staff as less valuable than data that is collected by volunteers. And I'll tell you why. If I'm running for whatever - a parliamentarian - I knock on your door and I say, "Doc, I want your vote. Are you going to vote for me?" And you say, "Well, I'm undecided," or "Yes," or, "No." The reality is that you're looking me in the face, so you want to be polite. Canadians are polite. It doesn't matter where you are in the country, Canadians are generally polite. You're not going to tell me to get off your lawn. But if you are talking to a 20 year old canvasser who's like, "Oh yeah, Nick - can he have your vote?" You're like, "Get off my lawn, kid. I'm not talking to that jerk." You are much more likely to be honest about your negative views with canvassers who aren't the candidates than you are with the candidate him- or herself. Much more likely. And that's borne out time and time again. When I have some candidate out to a particular area, I'm like, "No, let's try it again with a volunteer," and always get better data.


Elizabeth: [00:24:18] That's super interesting. The socially desirable response bias - in academia, we talk about that all the time impacting how we design surveys. But to see that impacting how you decide where to send different canvassers is a nice tangible example. Thank you.


Elizabeth: [00:24:37] Alright, I want to end it off with a real quick quiz. Can you tell me what the two-step flow hypothesis is?


Nick: [00:24:43] Oh God. That's opinion leaders... Well, I mean, come on! We're talking about opinion leaders. We're talking about them getting their information from mainstream media and actually feeding that to an audience that is not as politically engaged as they are.


Elizabeth: [00:24:57] Yeah. Perfect. A+.


Nick: [00:24:59] Good! See? Look at that! I learned something. You're a good prof.


Elizabeth: [00:25:03] Thank you, Nick. This was awesome.


Nick: [00:25:04] Oh my pleasure. Anytime.


Elizabeth: [00:25:08] Well, thanks for tuning in, everybody. That was the episode on the two-step flow hypothesis with my guest, Nick Switalski, who is a veteran campaigner and former political staffer. If you'd like more information about the two-step flow hypothesis, personal influence, or the other things we talked about today, check the show notes or head over to

Wonks and War Rooms Season 1 logo


Commenting has been turned off.
bottom of page