Followers are opinion leaders: The role of people in the flow of political communication on and beyond social networking sites

Abstract

One of the defining characteristics of the increasingly essential social networking sites is the network structure of the new medium. Opinion leaders are essential for the flow of communication in networks and consequently should be crucial for the flow of communication on social networking sites. Hence, to assess the political role and impact of social networking sites in electoral politics, we must find out more about the followers of parties and politicians. The aim of this article is therefore to investigate to what extent opinion leaders are more likely than others to follow parties and politicians on Facebook and whether opinion leaders are especially active in their online and offline networks, using a unique national representative web survey with more than 5700 respondents. The findings show that followers of parties and politicians on Facebook to a great extent are opinion leaders and that these opinion leaders are especially active in online and offline networks. This increases the significance of social networking sites and increases the role of opinion leaders in the flow of political communication in the society.

Keywords

Opinion leaders political communication political parties social media social networking sites two-step flow

Once, media research was characterized by the image of the atomized audience. This implied that every member of the audience received mass media content in isolation and that the messages had immediate effects on the individual.¹ However, while studying communication patterns in US election campaigns in the early 1940s, Lazarsfeld and his colleagues discovered that media messages were further mediated by what the scholars called opinion leaders (OLs) who diffused and interpreted messages to their personal networks. Hence, the role of individuals and group interaction became essential in scholars’ perceptions and models of communication flows (e.g. Katz and Lazarsfeld, 1955). Recent developments in society and media technology have made Bennett and Manheim (2006) argue for a return to a one-step flow of communication. The two-step flow of communication hypothesis implies a first step of mediated messages issued to a mass audience by the mass media. In the second step, small group interactions powered by OLs interpret and contextualize this message (Katz and Lazarsfeld, 1955; Lazarsfeld et al., 1944). According to Bennett and Manheim (2006), societal changes have isolated increasing numbers of individuals from the groups that traditionally provided these vital cues (cf. Putnam, 2000), and audience selection and microtargeting have replaced peer group interaction (see also Bennett and Iyengar, 2008).² Although new media technology arguably mostly contributes to increasing fragmentation and isolation (e.g. Blumler and Kavanagh, 1999; Chaffee and Metzger, 2001), innate qualities with aspects of the new technology might counteract the technological push towards isolation. More specifically, the network structure – one of the defining characteristics – of the increasingly essential social networking sites (SNSs) arguably connects people to a greater extent than isolate them. Platforms such as Facebook constitute digital network architecture for people to connect and create overlapping social networks. The flow of communication in networks of individuals depends on these individuals passing on information. Hence, OLs, who in the original conceptualization by Lazarsfeld and colleagues were emphasized as the nodes that connect the mass medium and the interpersonal networks, are most likely equally essential in the flow of communication on SNSs. Now, OLs are ingrained as nodes in the very structure that defines the medium.

Against this background, in this article, we address the role of people in the flow of political communication on SNSs. Although an increasing number of individuals include social media and SNSs such as Facebook in their daily media and communication diets, people seem less keen to use social media for political use, at least in traditional electoral politics. In this article, we emphasize the network aspect of the new medium and suggest that SNSs are more important to the flow of political communication than the relatively low number of political followers suggests. Individuals are now connected in digital networks, and in these networks, individuals will be exposed to a wide range of messages they do not actively seek out. What they are exposed to depends to a great extent on the interests and behaviour of the ones they are connected to, for example, their friends on Facebook. Hence, rather than omit individuals from models and our perceptions of political communication, we must treat individuals and especially OLs as essential for the flow of communication on today’s most innovative and increasingly essential media platform. More specifically, in this article, the first empirical aim is to investigate whether OLs are more likely to follow parties and politicians on Facebook, and whether this implies that the group which follows parties and politicians mostly consists of OLs.³ The second aim is to scrutinize the individuals who follow politicians and parties and study whether they are more active and more likely to pass on information than others.

We investigate the research questions by applying the classic self-designated measurement of OLs in a unique national representative web survey about social media use conducted in 2011 in Norway, consisting of 5737 respondents. Although relatively few follow politicians and parties on Facebook, the large dataset offers the possibility of studying these followers much more closely than is possible with an ordinary representative survey. Norway is an interesting case as the use of social media, and especially Facebook, is widespread: 73% of the population used Facebook weekly in 2011.⁴

The article proceeds as follows. In the first section, we substantiate the research questions first through a discussion of the notion of the two-step flow of communication and the role of OLs and then through a discussion of the nature of SNSs and the importance of OLs for the flow of communication in these networks. In the next section, we discuss the Norwegian case. The empirical analysis is divided into two parts. In the first part, we investigate to what extent followers of politicians and parties on Facebook are OLs, and in the second part, we study the size of the OLs’ networks and activity in online and offline networks. In the last section, we discuss the results and suggest avenues for future research.

Internet and politics

The equalization versus normalization theses put forward by Margolis and Resnick (2000) have arguably been the dominant theoretical framework in studies of online political communication and web campaigning. While the former thesis argues that the Internet equalizes competition between political actors, the latter claims that web campaigning will be politics as usual (see also Gibson et al., 2003). To some extent, this framework is also found in studies of the role of social media in campaigns (e.g. Strandberg, 2013). In this perspective, the impact of the new medium, to what extent it matters for the end result, has naturally received a lot of attention. For instance, although political actors seem to embrace social media and SNSs (e.g. Jackson and Lilleker, 2011; Karlsen, 2011a; Strandberg, 2013), voters seem much less inclined to use them for political purposes. A case in point is the use of social media in the 2011 Finish campaign which was very low (Strandberg, 2013).

In this article, we emphasize the network quality of the new medium and suggest that SNSs play a more important role for party politics and political communication than the relatively low number of political followers suggests. As mentioned, on SNSs individuals will be exposed to messages they do not actively and intentionally seek to find. In this manner, the so-called web 2.0 differs fundamentally from web 1.0. The digital architecture which defined web 1.0 entailed that people had to actively seek out information from, for example, party websites or online newspapers. On SNS, different messages will be diffused by members in the network and appear in the individual’s ‘feed’. This means that political actors have both primary and secondary audiences on SNS like Facebook and Twitter (cf. Vaccari and Valeriani, 2013), and consequently, active individuals who contribute to the diffusion of messages in networks, OLs, are of particular importance.

The two-step flow and OLs

The idea of a two-step flow of communication originates from Lazarsfeld and colleagues’ work on American election campaigns in the 1940s reported in the seminal The People’s Choice (Lazarsfeld et al., 1944). When the first study was carried out in Erie County in 1940, the campaign was expected to be decisive for the outcome of the election. However, the research group discovered that the media had little influence on voters.⁵ Indeed, most people had decided well before the campaign who to vote for. Moreover, the people who changed their voting intention during the campaign listed other people – not the media – as the reason for the change. Consequently, a new idea of media effects emerged: the idea of a two-step flow of communication. Basically, ideas flow from the media to OLs and from them to the less active sections of the population (Katz, 1957; Katz and Lazarsfeld, 1955; Lazarsfeld et al., 1944).

The two-step flow of communication model has been criticized for being a too-simple description of communication flows in society (e.g. Robinson, 1976; Weiman, 1994). For instance, communication not only flows from the mass media to OLs and from them to ordinary citizens: communication also flows from citizens to OLs as well as between OLs (e.g. Robinson, 1976; Troldahl and Van Dam, 1966). Surely, communication flows are more complex than the two-step model implies. However, the essential aspect in this context is the basic principle in the model that certain people can be considered OLs and are likely to pass on information and influence more passive individuals.

The most important aspect of the two-step flow perspective was that the prior traditional idea of mass media effects had to make way for people as an intermediary variable, and OLs are pivotal in this model. Before The People’s Choice, scholars had generally assumed that opinions were formed by the elite of the community and then passed down from one social stratum to the next (Katz and Lazarsfeld, 1955). Now scholars found that OLs were distributed in all occupational groups and at every socioeconomic level (Katz and Lazarsfeld, 1955). Moreover, OLs were experts in one field, not several. For example, political OLs were not the same individuals as OLs for fashion or cultural matters. OLs were also greater consumers of media than others – seeking information from several sources.

In the years following the publication of The People’s Choice and Personal Influence, the OL concept was applied in a wide variety of studies, often related to diffusion processes.⁶ However, during the last few decades, to our knowledge, the concept has been used only sporadically in political communication research. The exceptions include Norris and Curtis (2007), who addressed the role of the Internet in a two-step flow of political communication. They argued that political websites often ‘preach to the converted’ rather than expand the pool of engaged citizens. The influence of the Internet in a campaign therefore, Norris and Curtis argued, is mostly indirect through OLs. Due to the defining network characteristic of SNSs, the idea that individuals who use political websites are OLs should be even more relevant for political use of these media platforms. Vaccari and Valeriani (2013) also study to what extent Italian politicians reach a primary audience with their tweets and secondary audience through retweets. Although most who follow politicians are hardly active, they suggest that the potential for indirect communication lies in the ‘vital middle’ of the Twitter population who are more active than average. In the next part, we further discuss how the network structure of SNSs makes the concept of OLs arguably more relevant than ever.

OLs, networks and SNSs

A network is a pattern of interconnections among a set of things (Easley and Kleinberg, 2010), a set of interconnected nodes (Castells, 2009). Social networks are social ties between friends, colleagues, acquaintances and so on. SNSs are digital infrastructures that provide social ties between individuals. Hence, SNSs are defined by their qualities as a medium and innate network structure. The special appeal of online SNSs such as Facebook is therefore that they allow users to network with each other. Users construct a public or semipublic profile, make a list of other users to share a connection with, and view and traverse their list of connections and those made by others within the system (Boyd and Ellison, 2007).⁷ An SNS profile typically consists of three elements: a description of the individual, his or her friends (connections) list and communication exchange between profile owners (e.g. through the Facebook wall; Ancu and Cozma, 2009).

The original formulation of OLs emphasized them as the nodes that connected the medium and the interpersonal networks. This is visualized in the model shown in Figure 1(a). Regarding SNSs, OLs and other individuals are ingrained as the nodes that constitute the structure that defines the medium. This is visualized in the model shown in Figure 1(b).

The basic logic is the same in both models: OLs will diffuse and interpret messages to more passive individuals. However, in the mass media situation (Figure 1(a)), the passive individuals will be exposed to some of the same messages as the OLs, but the OLs will interpret and contextualize these messages. On SNSs, the diffusion of messages depends on active individuals such as OLs.

There are two main ways for information to diffuse/spread on Facebook. First, someone can ‘like’ or comment on a post. Then, the post is (at least most of the time) visible for everybody who is ‘friends’ with that individual. The second main way is how several individuals can share a post, for example, by linking to an external website, such as a blog, newspaper or a party website. Consequently, although Facebook is a semipublic SNS, in the sense that users can choose to let only their ‘friends’ see what they do or post, diffusion through ‘likes’, ‘commenting’ and ‘sharing posts’ makes the potential audience of any message much vaster than the size of the initial network.⁸ Hence, if a politician posts something on Facebook, the number of individuals exposed to this post depends on the activity of his or her followers. For example, politician (P) has four followers (see Figure 1). However, if the three OLs who are linked to her share her posts, she reaches out to an additional 14 people. Moreover, these OLs are likely to be connected to other OLs. If these OLs comment, like or share the post, the potential audience for the message is enormous and can lead to network effects like information cascades (Easley and Kleinberg, 2010; Velasques, 2012). For example, some of the success of the Obama campaign in 2008 has been credited to some active individuals online passing on the information (Cogburn and Espinoza-Vasquez, 2011).⁹ Moreover, Messing and Westwood (2014) show that social endorsement on Facebook increases the likelihood for people to select the endorsed content, emphasizing the role of individuals.

Research questions and expectations

The main aim of this article is to investigate the role of people in the flow of political communication on SNSs. Based on the theoretical discussion and the literature review above, we identify four research questions that we investigate empirically.

The classic studies of two-step flow have shown that OLs are active and seek to retrieve information to a greater extent than other parts of the population. We assume that this also applies for SNSs and ask, (1) Are OLs more likely to follow parties and politicians on Facebook, and if so, does the group that follow parties and politicians mainly consist of OLs?

If followers are OLs (based on self-designated opinion leadership measurements), the next step is to investigate whether they also behave like and have the characteristics we should expect from OLs. OLs’ media use is characterized by their instrumental search for relevant information. Hence, we ask, (2) Are OLs’ motives for following parties and politicians more related to (political) activity and information gathering than others? The classic studies on two-step information flows have also shown that OLs are centrally placed in networks. If OLs are to be essential in the flow of political communication on social media, they have to be centrally placed and active in these online networks. Hence, we ask, (3) Do the OLs who follow politicians and parties have larger online networks, and are they more active and more likely to pass on information in these online networks than others? Most, if not all, of the individuals (the nodes) in SNS networks have other offline (e.g. family, close friends, acquaintance, colleagues) networks. Hence, OLs on Facebook can also be expected to influence other people in these offline networks. If so, politicians will reach individuals who are likely to diffuse messages in online networks, but also influence others in offline settings. Thus, we ask, (4) To what extent are OLs who follow parties and politicians on Facebook active in their offline social networks?

The Norwegian case

Norway is an interesting case to investigate the research problem as Internet diffusion in society, and the use of social media, is relatively high. According to TNS Gallup, in 2011 the share of the total population who used Facebook weekly was 73%.¹⁰ Social media and SNSs have also been embraced by political parties. During the 2009 national election campaign, more than 43% of all candidates who ran for election used Facebook in their campaign communication (Karlsen, 2011a). However, in 2009, just 8% of the electorate reported that they had visited a politician’s blog or profile on a SNS (Karlsen, 2011b).

The Norwegian political system is characterized by a parliamentary government, a stable multiparty system and well-organized membership parties. Campaigning in Norway is typically party-centred, centralized and nationwide, and candidates campaign as part of the greater party campaign organization (Karlsen and Skogerbø, 2013). However, much media focus is devoted to party leaders and other top politicians, and in this sense, Norwegian politics is also personalized. Hallin and Mancini (2004) identified the Norwegian media system as a typical example of the democratic corporatist model, a model characterized by the historical coexistence of commercial media and media tied to organized social and political groups and by the relatively active but legally limited role of the state (Hallin and Mancini, 2004).

Data and operationalization

The data comprise a – to our knowledge – unique web survey about social media use conducted in 2011 in Norway. The survey was conducted on a representative sample of the Norwegian population and a representative sample of Norwegian social media users. The survey was conducted in March and April 2011. The total number of respondents for the population sample was 1587,¹¹ whereas the total number of respondents for the social media sample was 4151. These two samples are combined in the analyses when the followers on Facebook were compared to the population as a whole. Special weights were designed and used in these analyses. The relatively large sample for the social media users secured a sufficient number of respondents who use SNS to follow politicians and parties. At the same time, the data are representative, for the population as a whole and for the social media population.

To measure interaction with politicians and parties on Facebook, we used the following question: What groups are you a member of, and what sites do you follow on Facebook? Included in the alternatives are top politicians, local politicians and political parties. The answer categories are several, one and none. In Table 1, the share of the population who follows a party or a politician is described.

Table 1.

Descriptive statistics.^a

	Age (years)
	15–24	25–44	45+	All
Top politicians	16	10	8	10
Local politicians	6	5	6	6
Parties	16	10	8	10
At least one	22	15	12	15
N	823	2407	2507	5373

The share in the population that follows a top politician, a local politician and a political party on Facebook.

The motivations for following politicians and parties on Facebook were measured with a follow-up question: To what extent do you agree with the following statement? I follow politicians/parties on Facebook to … nine alternatives were listed (see Figure 3), and the categories were strongly agree, agree somewhat, neither agree nor disagree, disagree somewhat and strongly disagree.

The OL concept was measured by using the two original questions from Lazarsfeld et al.’s Erie County study: Have you recently tried to convince anyone of your political ideas? Has anyone recently asked you for advice on political questions? Similar to the Erie County study, we were interested in political opinion leadership, and thus, these two questions fit our purpose.¹² These questions are so-called self-designated opinion leadership measurements (cf. Hamilton, 1971), and by using them it is not possible to show that OLs influence other more passive segments of the population (cf. Katz, 1957). Hence, the second aim of the article was to investigate the behaviour and activity of OLs more closely.

To measure the size of the online networks, two questions regarding the number of friends on Facebook and followers on Twitter were used. To measure activity in networks, we used questions about online and offline behaviours. First, two questions were related to Facebook activity: How often do you link to news about current affairs and politics? How often do you discuss links about current affairs and politics? Second, we asked questions about discussing politics: How often do you discuss politics? The categories were daily, a couple of times a week, more seldom and never. In addition to a general question, we also asked about discussions with four groups: family members, friends, colleagues and fellow students, and members of the same organization as the respondent. Third, activity was also measured with a battery of questions regarding traditional political participation (see Table 4 for the nine activities).

Empirical analysis

The empirical analysis is divided into two main parts. In the first part, we investigate to what extent OLs are more likely to follow parties and politicians than other segments of the population and the OLs’ motives. In the second part, we study the size of OLs’ networks and their activity in networks, first online and then offline.

Followers and OLs

In our sample, about 26% tried to convince someone of their political ideas, and 18% had been asked about advice on political matters. If we combine these questions, about 33% of the population are OLs.¹³ There was no difference between the population as a whole and Facebook users, reflecting the great diffusion of Facebook use in Norway. The first research question asked whether OLs are more likely to follow politicians and parties than other individuals. The data indicate that this indeed is the case. Although only 15% of the population followed a politician or a party, 40% of the OLs did. The young and the politically interested were also more likely to follow a politician or a party on Facebook than others, while gender had no effect. However, the multivariate analysis reported in Table 2 shows that the effect of opinion leadership holds even when age and political interest are controlled for.

Table 2.

The effect of opinion leadership on following politicians and parties on Facebook.^a

	1		2
	b	SE	b	SE
Opinion leader	1.332**	.070	.874**	.076
Gender			−.126	.073
Age			−.007**	.002
Political interest			.957**	.064
Constant	−1.66**		−2.72**	.185
Nagelkerke’s R²	.11		.19

SE: standard error.

Only Facebook users are included in the analysis.

Dependent variable (follow politician or party = 1; else = 0).

Entries are logistic regression coefficients, standard error and Nagelkerke’s R² (N = 4685).

p < .01 and *p < .05.

OLs are following politicians and parties on Facebook to a greater extent, and the result of this is that the individuals who follow politicians and parties on Facebook constitute a special group that consists mostly of OLs. Figure 2 shows the share of OLs who follow top politicians, parties and local politicians, compared to the rest of the Facebook population.

Three main points need to be emphasized. First, individuals who followed parties and politicians were mainly OLs.¹⁴ Around 70% of the individuals who followed several parties or politicians were OLs, while just 30% of the individuals who did not follow parties or politicians were OLs. Second, individuals who followed several politicians were more likely to be OLs than individuals who followed just one top politician. The differences are not great, but they are consistent for all three groups. The difference in percentage points is around 15 points for top politicians (14) and parties (18) and 9 points for local politicians. Third, the results are similar for top politicians, local politicians and parties. However, OLs were even more overrepresented in the groups that followed local politicians and parties. This suggests that although some non-OLs are attracted to the celebrity of top politicians, OLs have other motives. As Figure 3 shows, OLs were more instrumental in their motives for following politicians and parties on Facebook. The figure shows the share of OLs compared to others who agreed (strongly and somewhat) that they followed politicians and parties for nine different motives.

Motives which involved activity and information gathering tended to be more important for OLs. The difference between OLs and others was greatest for motives related to getting something out of following the politician or party, such as ‘to discuss with the politician/party’, ‘to meet other supporters of the politician/party’, ‘an easy way to get information’ and to ‘find out what others are saying’. OLs were less likely to say that motives such as ‘killing time’ and ‘a coincidence’ were important.

Online network size and activity

In the previous part, we saw that individuals who follow parties and politicians online to a greater extent are OLs, and they have more instrumental motives for following parties and politicians than others. This indicates that these individuals are more active and aid in spreading political messages in networks. In this section, we investigate this proposition in detail. Are the OLs who follow parties and politicians characterized by the distinctive hallmarks we should expect from OLs? More specifically, we investigate whether they have greater online networks and whether they are more active in online and offline social networks.

In the following analysis, the OLs who followed politicians and parties are compared to three other groups. First, OLs are compared to others who followed parties and politicians. Second, they are compared to OLs who did not follow politicians and parties. Third, these OLs are compared to the rest of the Facebook users or alternatively to the rest of the population when this is relevant.

First, we investigate the size of OLs’ online networks. The average user on Facebook had 239 friends, and the average user on Twitter had 94 followers. Figure 4 shows the average number of followers on Facebook and Twitter for the four groups.

OLs who follow parties and politicians reach out to a much larger network than others. In 2011, on average, these OLs had 302 friends on Facebook and 166 followers on Twitter. Others who followed politicians and parties had 274 individuals in their network on Facebook and 152 followers on Twitter. OLs who did not follow politicians had, on average, a network on Facebook of 220 individuals and 76 followers on Twitter. The rest of the population had, on average, 206 friends on Facebook and 56 followers on Twitter. Consequently, they are more centrally placed in networks in the sense that these OLs had larger networks than all the other groups.

OLs who followed parties and politicians were clearly more active than others on Facebook when it comes to behaviour that contributes to diffusing political messages (Figure 5). Of the OLs who follow parties and politicians, 32% commented and discussed links to news about politics and current affairs often, compared to only 8% of the OLs who were not followers and 2% of other Facebook users. In addition, 23% of OLs who followed parties and politicians linked to news about politics and current affairs often or very often, compared to only 6% of other OLs and 2% of other Facebook users. The fact that the OLs (who did not follow a party) were less active than the non-OLs who followed a party suggests that these OLs were not active users of Facebook and might be more active in offline networks and activity.

Offline network activity

Thus far, we have found that individuals who follow parties and politicians on Facebook are OLs. These OLs have larger networks than others and are more active in these online networks when it comes to political behaviour. They comment and link more to news about politics and current affairs than others. This behaviour contributes to diffusing political messages in the network. This supports the expectation that active individuals (OLs) contribute to the flow of political communication on SNSs. Now, we turn to offline activity and the general question of how often individuals discuss politics. Table 3 shows the percentage share that discussed politics daily and a couple of times a week in general, and with family members, friends, colleagues or fellow students, and members of the same organization as the respondent.

Table 3.

Opinion leaders who follow parties/politicians discuss more politics.

	Discuss	Daily	A couple of times a week	N
In general
Opinion leader and follower of party/politician	96	52	44	468
Follow party/politician	72	17	55	356
Opinion leader	89	38	51	1403
Others	58	12	46	3380
With family
Opinion leader and follower of party/politician	73	25	48	468
Follow party/politician	52	11	41	356
Opinion leader	69	28	41	1403
Others	48	11	37	3380
With friends
Opinion leader and follower of party/politician	79	16	63	468
Follow party/politician	49	4	45	356
Opinion leader	64	10	54	1403
Others	35	2	33	3380
With colleagues/fellow students
Opinion leader and follower of party/politician	74	28	46	468
Follow party/politician	55	9	46	356
Opinion leader	60	14	46	1403
Others	42	5	37	3380
With members of an organization
Opinion leader and follower of party/politician	36	9	27	468
Follow party/politician	9	1	8	356
Opinion leader	16	3	13	1403
Others	3	0	3	3380

The proportion who discussed politics daily and a couple of times a week in general, with family members, friends, colleagues, or fellow students and members of the same organization.

(Smallest) N = opinion leader and follower of party and politician (468), follow party and politician (356), opinion leader (1403) and others (3380).

More than half of the OL followers discussed politics daily and another 44% weekly. Hence, almost all the OL followers discussed politics at least weekly. They also discussed politics more frequently with family members, friends, colleagues and fellow students, but the difference is particularly big for discussing with members of an organization. OLs who did not follow parties and politicians were relatively more active here than in activities related to Facebook. This suggests that these OLs are active in social networks, but that a type of digital divide exists between these OLs and the OLs who followed politicians and parties on Facebook.

Although the level of discussion is high in Norway, the OLs who followed politician and parties appeared to be particularly active OLs, offline as well. This notion is further supported by the findings reported in Table 4. It shows the share of the four groups that participated in nine different types of political actions.

Table 4.

Opinion leader followers participate in a broad range of political activities.

	Opinion leader and follower	Follow party/politician	Opinion leader	Others
Petition	61	49	44	29
Boycotting	43	31	39	19
Supported an action	40	29	40	16
Newspaper piece	35	9	16	3
Public discussion	35	7	17	3
Raised an issue in a political party	33	6	17	2
Contacted a politician	30	8	17	4
Demonstration	22	7	11	4
A written complaint	11	3	10	3
Not participated in any	8	31	22	53
N	468	356	1403	3380

The proportion of four groups that participated in the action.

(Smallest) N = opinion leader and follower of party and politician (468), follow party/politician (356), opinion leader (1.403), others (3380).

The OLs who followed parties or politicians had the highest level of participation for all nine actions surveyed. Only 8% did not participate in any action, while more than half of the rest of the population (group 4) had not participated in any. The difference between OL followers and OLs who do not follow parties on Facebook is very small when it comes to boycotting and supporting an action and is greater when it comes to participating in public discussions and writing newspaper pieces. Hence, these OLs are particularly active when it comes to activities related to political communication.

Discussion and conclusion

SNSs are characterized by their qualities as both media and networks, and active individuals (OLs) are essential for diffusing messages in networks. Hence, in this article, we investigated to what extent followers of politicians and parties on Facebook are OLs, as the significance of SNSs in electoral politics depends on who these individuals are.

The findings show that OLs are much more likely than others to follow parties and politicians on Facebook. The group that followed politicians and parties therefore mostly consisted of OLs: about 70% of the ones who followed several politicians and parties were OLs, compared to about 30% in the rest of the population. The pattern was similar for followers of top politicians, local politicians and political parties. OLs also had more instrumental motives than non-OLs to follow politicians and parties. For OLs, motives such as an ‘easy way to get information’, ‘meeting other supporters’ and ‘wanting to discuss with the politician’ were more important.

One of the defining characteristics of OLs is that they are active in networks, pass on information and influence others. Hence, in addition to using the self-designated OL question to identify OLs and their motives, we investigated the activity of these OLs who followed politicians and parties. The results show that they had larger online networks than others; they had more friends on Facebook and more followers on Twitter; they were more active when it came to diffusing political messages in these networks as they commented and discussed, as well as linked to, content about politics and current affairs to a much greater extent than other groups.

OL activity was not limited to online networks. Although around half the population discussed politics weekly, the OL followers discussed to an even greater extent: almost all of them discussed politics at least weekly. Moreover, they were more active than the rest of the population regarding discussing with friends, family members, colleagues or fellow students. OL followers particularly stood out regarding discussing with other members of organizations, suggesting that many of these individuals are active in networks beyond friends, family and colleagues. The OL followers also participated in all nine political activities we studied to a much greater extent than the rest of the population.

The results indicate that OLs are essential for the flow of information on SNSs and for the flow from SNSs to other social networks. Consequently, Facebook is an important platform for political parties and political candidates to reach OLs who will contribute to diffuse political messages in online and offline networks. Hence, applying a two-step flow perspective on political communication on SNS has proved necessary to understand the impact of SNS on electoral politics and political communication. However, this study, although it serves to show that followers of parties and politicians are OLs who contribute to spread information in both digital and personal networks, has limitations. In this article, we suggest a secondary audience, and the role of OLs is to diffuse political messages to secondary audiences. The extent and manner in which such messages are received on Facebook and other social media should be a topic for future research. Moreover, future research should also investigate whether OLs merely contribute to diffusing messages or whether SNSs also are a medium for OLs to contextualize and interpret messages for more passive individuals. The results in this article indicate that contextualization and interpretation are part of the process: OLs frequently discuss links to news about politics and current affairs. By doing this, they probably contribute to interpret and contextualize news for people in their network.

Vaccari and Valeriani (2013) show that most followers of Italian politicians on Twitter are hardly active and do not retweet messages to a great extent. This study of Norwegian Facebook users suggests that followers on Facebook are more active. This could be related to differences between the countries, but also to differences between Facebook’s and Twitter’s affordances. The threshold for liking or commenting something on Facebook is probably lower than the threshold for retweeting or commenting on tweets on Twitter. Future research, both qualitative and quantitative, can shed more light on the role of OLs in the flow of political communication on Facebook compared to other SNSs.

OLs’ activity on SNSs increases the significance of the medium in the flow of political communication in society. Simultaneously, SNSs arguably increase the significance of OLs. Some scholars argue for increasing isolation and a one-step flow of communication in today’s societies. Although Norwegian data are used in this article, and there is less evidence of increasing isolation in Norwegian society than in the United States,¹⁵ the network capabilities of Facebook and the role of active individuals in diffusing messages on these networks should apply in any society. Hence, developments in media technology have increased the network capabilities of some individuals, while others arguably become more isolated. The result might very well be an increasing network gap in society.

Footnotes

Funding

This research received funding from the Norwegian Research Council.

Notes

References

Ancu

Cozma

(2009) MySpace politics: Uses and gratification of befriending candidates. Journal of Broadcasting & Electronic Media 53(4): 567–583.

Asp

(1986) Mäktiga Massmedier [Powerful Massmedia]. Stockholm: Akademilittertur.

Bennett

Iyengar

(2008) A new era of minimal effects? Journal of Communication 58: 707–731.

Bennett

Manheim

(2006) The one-step flow of communication. Annals of the American Academy of Political and Social Science 608: 213–223.

Blumer

(1946) The mass, the public, and public opinion. In: Lee

(ed.) New Outline of the Principles of Sociology. New York: Barnes & Noble, pp. 185–193.

Blumler

Kavanagh

(1999) The third age of political communication: Influences and features. Political Communication 16: 209–230.

Boyd

Ellison

(2007) Social network sites: Definition, history, and scholarship. Journal of Computer Mediated Communication 13(1): 210–230.

Castells

(2009) Communication Power. Oxford: Oxford University Press.

Chaffee

Metzger

(2001) The end of mass communication? Mass Communication & Society 4: 365–379.

10.

Cogburn

Espinoza-Vasquez

(2011) From networked nominee to networked nation: Examining the impact of Web 2.0 and social media on political participation and civic engagement in the 2008 Obama campaign. Journal of Political Marketing 10: 189–213.

11.

Easley

Kleinberg

(2010) Networks, Crowds, and Markets: Reasoning about a Highly Connected World. Cambridge: Cambridge University Press.

12.

Friedson

(1951) Communication research and the concept of the mass. American Sociological Review 18: 313–317.

13.

Gibson

Margolis

Resnick

. (2003) Election campaigning on the WWW in the USA and UK. Party Politics 9: 47–75.

14.

Granovetter

(1973) The strength of weak ties. American Journal of Sociology 78(6): 1360–1380.

15.

Hallin

Mancini

(2004) Comparing Media Systems. Three Models of Media and Politics. Cambridge: Cambridge University Press.

16.

Hamilton

(1971) Dimensions of self-designated opinion leadership and their correlates. Public Opinion Quarterly 35(2): 266–274.

17.

Holbert

Kelly Garret

Gleason

(2010) A new era of minimal effects? A response to Bennett and Iyengar. Journal of Communication 60: 15–34.

18.

Iyengar

Van den Bulte

Valente

(2010) Opinion leadership and social contagion in new product diffusion. Marketing Science 30(2): 195–212.

19.

Jackson

Lilleker

(2011) Microblogging, constituency service and impression management: UK MPs and the use of Twitter. Journal of Legislative Studies 17: 86–105.

20.

Karlsen

(2011a) A platform for individualized campaigning? Social media and parliamentary candidates in the party-centered Norwegian campaign. Policy & Internet 3(4): article 4.

21.

Karlsen

(2011b) Velgernes valgkamp [The Voters’ Election Campaign]. In: Aardal

(ed.) Det politiske landskap. En studie av stortingsvalget i 2009. Oslo: Cappelen Damm, pp. 41–63.

22.

Karlsen

Skogerbø

(2013) Candidate campaigning in parliamentary systems. Individualized vs. localized campaigning. Party Politics. Epub ahead of print 7 June. DOI: 10.1177/1354068813487103.

23.

Katz

(1957) The two-step flow of communication: An up-to-date report on an hypothesis. Public Opinion Quarterly 21: 61–78.

24.

Katz

Lazarsfeld

(1955) Personal Influence: The Part Played by People in the Flow of Mass Communication. New York: Free Press.

25.

Lazarsfeld

Berelson

Gaudet

(1944) The People’s Choice. How the Voter Makes up His Mind in a Presidential Campaign. New York: Duell, Sloan and Pearce.

26.

Margolis

Resnick

(2000) Politics as Usual: The Cyberspace ‘Revolution’. Thousand Oaks, CA: SAGE.

27.

Messing

Westwood

(2014) Selective exposure in the age of social media: Endorsement trump partisan source affiliation when selecting news online. Communication Research 41: 1042–1063.

28.

Noelle-Neumann

(1984) The Spiral of Silence. Chicago, IL: University of Chicago Press.

29.

Norris

Curtis

(2007) Getting the message out: A two-step model of the role of the Internet in campaign communication flows during the 2005 British general election. Journal of Information Technology & Politics 4: 3–13.

30.

Putnam

(2000) Bowling alone: The Collapse and Revival of American Community. New York: Simon & Schuster.

31.

Robinson

(1976) Interpersonal influence in election campaigns: Two step-flow hypotheses. Public Opinion Quarterly 40: 304–319.

32.

Rogers

(1995) Diffusion of Innovations, 5th edn. New York: Free Press.

33.

Shah

Scheufele

(2006) Explicating opinion leadership: Nonpolitical dispositions, information consumption and civic participation. Political Communication 23: 1–22.

34.

Strandberg

(2013) A social media revolution or just a case of history repeating itself? The use of social media in the 2011 Finnish parliamentary elections. New Media & Society 15: 1329–1347.

35.

Troldahl

Van Dam

(1966) Face-to-face communication about major topics in the news. Public Opinion Quarterly 29: 626–634.

36.

Vaccari

Valeriani

(2013) Follow the leader! Direct and indirect flows of political communication during the 2013 general election campaign. New Media & Society. Epub ahead of print 11 November. DOI: 10.1177/1461444813511038.

37.

Velasques

(2012) Social media and online political discussion: The effect of cues and informational cascades on participation in online communities. New Media & Society 14: 1286–1303.

38.

Webster

(2005) Beneath the veneer of fragmentation: Television audience polarization in a multichannel world. Journal of Communication 55: 366–382.

39.

Weiman

(1991) The influentials: Back to the concept of opinion leaders? Public Opinion Quarterly 55: 267–279.

40.

Weiman

(1994) The influentials. New York: State University of New York Press.

41.

Wirth

(1948) Consensus and mass communication. American Sociological Review 13(1): 1–15.

42.

Wollebæk

(2011) Norge og Nordens sosiale kapital i europeisk kontekst [Social capital in Norway and the Nordic countries in a European context]. In: Wollebæk

Segaard

(eds) Sosial kapital i Norge [Social Capital in Noway]. Oslo: Cappelen Damm, pp. 51–77.