Pandering, protesting, engaging. Norwegian party leaders on Facebook during the 2013 ‘Short campaign'

Abstract

While a comparably large amount of research has looked into the uses of Twitter at the hands of politicians, relatively little work has been done how Facebook is being used in this regard. The current paper expands our insights into the uses of social media by politicians by presenting a study of party leaders' uses of Facebook during the 2013 Norwegian election campaign. The paper focuses on two overarching areas: gauging the different themes brought up by the party leaders in their posts and the types of feedback (understood here as likes, comments and shares) that these activities appear to result in. Results indicate that the types of content least provided by the politicians – acknowledging the support of others or criticizing the actions by political peers or media actors – emerge as the most popular in this regard. Results further show that the most common type of feedback is likes – a finding that suggesting that a reassessment of the viral qualities of Facebook for purposes like these is necessary.

Keywords:

• politics
• social media
• web 2.0
• Facebook
• Norway
• e-Government

Introduction

Accompanied by hyped concepts such as ‘the third age of political communication' (Blumler & Kavanagh, 1999) or ‘new politics' (Calenda & Meijer, 2009), the Internet is often believed to have played a large part in the development of political campaigning and communication as we know and understand these processes today. Indeed, while websites hosted by parties as well as individual candidates have become virtually mandatory in a series of different political contexts and systems (Druckman, Kifer, & Parkin, 2007; Karlsen, 2009; Strandberg, 2013), recent years have seen similar demands for politicians to make use of so-called social media in attempts to further their success rate at the ballots.

While all media forms essentially can be understood as social to some extent (Klinger & Svensson, 2014), the characteristics of the online services popularly understood as social media are that they potentially open up for interaction between different users, between sender and receiver (Bechmann & Lomborg, 2012). These types of interactions were, of course, a possibility also in the earlier phases of the Web, where some technical skill was nonetheless required to curate one's own online presence. Still, the networking qualities and supposed ease of use often associated with services such as Twitter or Facebook are often thought of as having revolutionized online political engagement practices – sometimes reminiscent of the buzzwords discussed previously. Focusing on the latter of these two services, the present study gauges the degrees to which Norwegian party leaders make use of their Facebook presences, as well as the levels of engagement – understood on the platform at hand as likes, shares and comments – that result from these uses. As such, while the focus of the work performed lays primarily on the activities undertaken by political actors, the analytical design also incorporates the outcomes that these activities yield.

Norway has about 5.1 million citizens, and high levels of the citizenry make use of the Internet on a daily basis (Vaage, 2012). The political systems can be characterized as moderate pluralistic, featuring seven main parties. The paper at hand is based on data gathered during a one-month period leading up to the 2013 parliamentary election – arguably a period of heightened political activity. At the time of the election, the Social Democrats under Jens Stoltenberg were the heads of minority government, challenged primarily by the two other major parties – the conservative Høyre (under the leadership of Erna Solberg) and the right-wing populist Progress Party (spearheaded by Siv Jensen). The current focus thus provides an interesting contrast to findings from other contexts, as Norway can be understood as an advanced nation when it comes to Information- and Communication Technologies (ICTs) use. Given that a 2012 survey found 60.5% of Norwegians to be on Facebook (Enjolras, Steen-Johnsen, & Wollebaek, 2013), this becomes especially true with regard to the service under scrutiny.

As such, the paper at hand provides insights into politicians' uses of social media in a context other than the United States or UK (as suggested by Hermans & Vergeer, 2012). Moreover, it does so by moving beyond a focus on adoption of social media by politicians (Strandberg, 2013), instead looking at their actual uses of the service at hand. Finally, while the number of studies on Twitter in this context is rising (Jungherr, 2014), comparably few research efforts look into the political uses of the arguably more popular Facebook platform (Vergeer & Hermans, 2013). Thus, this study makes a clear contribution to the broader field of online political communication by providing insights into the uses of Facebook at the hands of leading political actors.

Politicians providing content online – a proposed typology

As alluded to in the introduction, hopes regarding the political potential of the Internet in the mid-1990s and later, the advent of social media by the mid-2000s, were often held high (Blumler & Kavanagh, 1999). For citizens, both waves of technological advancement were thought by the most optimistic of pundits and scholars as baring the potential to function as a ‘magic elixir [ … ] to reinvigorate the masses to participate in the process of government' (Stromer-Galley, 2000, p. 128). For politicians, similar claims were aired by proponents of what has been labelled as the innovation hypothesis, largely suggesting that especially smaller parties would be able to balance their structural disadvantage towards larger competitors by making use of online services to communicate and engage with potential supporters. While early studies in particular have given some merit to such claims of innovation or equalization (Larsson & Svensson, 2014), later empirical work has largely favoured the contrasting normalization hypothesis, suggesting ‘politics-as-usual' (Margolis & Resnick, 2000) as the favoured practice of political actors when going online. This latter hypothesis suggests that rather than providing a shift with regard to political campaigning, the Internet led instead to an enhancement of existing behavioural patterns as exhibited by politicians, generally providing information rather than opportunities for participation (Larsson, 2013; Vaccari, 2008a, 2008b; Ward & Vedel, 2006). While there are exceptions, such an approach would largely suggest these uses to function as ‘vehicles for downward dissemination of information rather than recruitment of users’ opinions and the promotion of participation' (Gibson, 2004, p. 96).

Building on these hypotheses, this section outlines previous research performed on the uses of the Internet in general and social media in particular at the hands of politicians. As plenty of scholarship has been devoted to the study of Twitter in this regard (Jungherr, 2014), much of this literature review will be based on results from such research – as well as on findings from studies focused on ‘website feature analysis' (Vergeer, Hermans, & Cunha, 2012). While the differing characteristics of each distinctive platform hinder parallel comparisons of the more technical features available, our focus here is on the types of content provided by the politicians through their respective Facebook Pages. As such, the current study builds on previous, feature-based efforts, attempting to move beyond the practice of ‘simply documenting usage patterns' (Lomborg, 2012, p. 229) in these online political contexts. Instead, the focus here is placed more clearly on what is actually being said in the postings of politicians (as suggested by Bruns & Stieglitz, 2012; Larsson & Moe, 2012). As such, this study draws on the comparably early effort by Blumler and Gurevitch (1995), where the authors suggest that ‘technological innovations – such as [the] television and now the Internet – can offer political actors new opportunities for projecting their messages' (p. 204). Our current efforts, then, are focused on these uses at the hands of politicians on Facebook – and the way that these activities appeared to resonate with those following the politicians on the service at hand.

With the focus of the paper at hand in mind, previous research suggests six broad, yet distinct, themes when it comes to the contents posted by politicians online. The themes identified – stating Acknowledgements, providing Campaign Reports, Informing followers about prioritized policy issues, uttering Critique towards competitors or other actors, utilizing the Facebook platform for attempts at citizen Mobilization, and providing glimpses into one's Personal sphere – will serve as the basis for the coding rationale employed in the study and are described in the following.

Acknowledgements

One of the characteristics of the Internet in general and perhaps of social media services in particular is the speed with which issues, ongoing events and specific themes can gain traction throughout a wider network of followers or friends – tendencies that are often described as the potential to ‘go viral’ (Klinger, 2013). Undoubtedly, the potential for such fast-paced spread of information is of interest for politicians up for election. Research has found a clear linkage between conjuring emotions and virality (Berger & Milkman, 2010) – a result that appears to hold especially true when it comes to the sharing of positive sentiments (Wu, Hofman, Mason, & Watts, 2011). In the context of political actors online, such positive news are regularly associated with giving thanks and acknowledgement to supporters and co-workers, often at the end of election campaigns (Bronstein, 2013; Graham, Broersma, Hazelhoff, & van 't Haar, 2013). Indeed, as the ‘victory tweet’ sent by Barack Obama following the 2012 presidential campaign is supposedly one of the most shared and redistributed in all of social media (Griggs & Mendoza, 2012), we might expect Norwegian party leaders to approach Facebook in a similar fashion.

Campaign reports

Politicians up for election spend plenty of time on the campaign trail. This category of posts, then, make up for the glimpses of ‘road life’ that are sent – often in terms of ‘information from party conventions' (Klinger, 2013, p. 724) or ‘references to campaign events such as rallies, speeches or debates' (Conway, Kenski, & Wang, 2013, p. 1600). By providing information regarding their whereabouts and current campaign-related undertakings, politicians supposedly provide an easy way for their supporters to keep a check on current electioneering developments. The always-on character of the Internet combined with the relative ease with which a candidate (or a staffer, for that matter) could potentially employ a plethora of mobile phones and other similar devices for these purposes would seem to lend themselves especially well towards such quick campaign reports from speeches, debates, meetings and the likes.

Informing

Arguably one of the established modes of political campaigning, practices of informing (Foot & Schneider, 2006) largely serves as a ‘Burkean, Top-Down communication strategy' (Jackson & Lilleker, 2009, p. 246) in providing policy statements and discussion of specific issues of political relevance. While too much information provision can be seen as thwarting opportunities of ‘strategic ambiguity' (Stromer-Galley, 2000, p. 125) – practices of non-disclosure regarding certain issues that allow politicians to ‘becloud their policies in a fog of ambiguity' (Downs, 1957, p. 136) in order to appeal to voters of varying political persuasions, such ‘downward flows' (Lilleker & Jackson, 2010, p. 80) of information have been commonplace in much earlier research (Hermans & Vergeer, 2012; Larsson, 2011a; Strandberg, 2009). Indeed, while the bulk of studies has been performed in a supposedly mostly one-directional ‘Web 1.0’ context, suggestions have been made that the ‘philosophical perspective which underpins Web 2.0' (Lilleker et al., 2011, p. 198) should diminish such approaches, sometimes described as ‘electronic brochures' (Kamarck, 1999) or ‘virtual billboards' (Sadow & James, 1999). Given the plus ça change nature of findings with regard to these practices, we should perhaps expect more of the apparent established practice of issue informing than popular debate would sometimes have us believe.

Critique

Sometimes also referred to as negative campaigning and mainly involving various forms of criticisms of campaign opponents (Glassman, Straus, & Shogan, 2010), our current purposes seek to broaden this definition to also include critical remarks in general. Specifically, the category not only encompasses critical utterances directed towards political competitors, but also includes those statements made by politicians that critique other societal actors, such as media outlets. Indeed, as negative news appear to play a substantial part in the mediated coverage of politics (Schweitzer, 2012), we might expect politicians to react towards being portrayed in a certain way, or towards having their issue statements described incorrectly – as they themselves see it. As negative information is more easily acknowledged and memorized by recipients, regardless of their political preferences (Lau, 1982), politicians might find it useful to let their scorn show as this might lead the recipients of such messages to hold the competition in doubt. Be that as it may – the party-centred Norwegian political system in general, and its few centrifugal forces in particular (Sartori, 1990), could be expected to function as a check on such overtly negative or critical Facebook Page postings.

Mobilization

One of the often-aired suggestions of the Internet as related to political campaigning is its supposed potential for increasing political involvement among citizens (Chadwick, 2003; Norris, 2002). Such a potential among the citizenry, then, is mirrored in the proposal that politicians, with the help of online technologies, would be able to communicate more directly with potential voters, potentially mobilizing them for campaign purposes. As such, the category of mobilization relates to the interactive potential of the Internet, mapping out the degree to which the politicians under scrutiny make use of their respective Facebook Pages to generate feedback from the electorate, or to encourage the recipients to partake in campaign meetings, ask questions regarding specific policies or other issues, join the campaign by canvassing or discussing politics with their friends, vote and the likes (Conway et al., 2013; Foot & Schneider, 2006; Graham et al., 2013; Jackson, 2011; Lilleker et al., 2011). Such forms of activation of the citizenry through mobilization efforts of the politicians could indeed take place both on- and offline; with regard to the former, pleads from a party leader to share a specific post on Facebook are to be expected; for the latter, the need to gather sizeable crowds at physical meetings should not be underestimated when it comes to building campaign stamina and self-confidence. Given the ‘hype’ surrounding these types of employment, we might expect the Norwegian context to feature a certain level of mobilization efforts.

Personal

While the concept of personalization with regard to politics makes up a rich body of research on its own (Enli & Skogerbø, 2013; Kruikemeier, van Noort, Vliegenthart, & de Vreese, 2013), the current employment of the term personal seeks to assess the degree to which politicians make use of their Facebook Page presences to ‘present themselves in an individual way' (Hermans & Vergeer, 2012, p. 72). A choice at the hands of the politicians to gear their online presences and postings towards such ‘backstage’ topics along the lines of ‘personal affairs, feelings' (Herkman, 2012, p. 376), ‘private information' (Vergeer et al., 2012, p. 136) or more generally ‘everyday activities of a non-political nature' (Jackson & Lilleker, 2011) might be a wise move from a pure campaigning perspective, as this signals a levelling or democratization of the relationship between citizens and those competing to govern them (van Zoonen, 2005) – something which in turn might equally diminish the supposed psychological distance between these two groups (Vergeer, Hermans, & Sams, 2011). A less narrow perspective, taking the role and influences of campaigning within the broader confounds of democratic society into account would most likely not approve of what could perhaps slightly controversially be described as personalized pandering (Habermas, 1989; Sennett, 1992). As such, levels of use among the politicians under scrutiny is difficult to assess beforehand – on the one hand, they might chose to appeal to voters by providing extensive features regarding their personal life; on the other, alleged professional concerns might come into play and secure the priority of more traditionally politically oriented content.

In summary, then, the typology presented earlier provides us with a six-part rationale, informed by previous, similar scholarship, with which to approach Facebook activities of politicians. While a heuristic like the one featured here might not provide specific detail regarding each individual post, the argument made here is that it does indeed allow for an overarching mapping of the types of content being made available on these Pages.

Method

Facebook started out as a platform primarily geared towards individuals, who could use host Profile pages. Since 2008, uses also involve the possibility for organizations to create so-called Fan Pages (or Pages for short), on which to entertain a community of fans for specific products, causes, etc. Such uses of Pages have spread into politics, with politicians adopting such online presences in a variety of contexts (Gulati & Williams, 2013; Larsson & Kalsnes, 2014).

Our focus on Facebook Pages can be motivated by two considerations in particular. First, for ethical reasons, while we can assume that politicians might choose to employ their personal Profile pages in a professional manner, our emphasis was placed on the professionally geared Pages instead. Indeed, while politicians are public figures and should therefore expect a higher amount of scrutiny, they have the right to privacy with regard to their online activities. As such, our choice to exclude such more privately themed Profile pages is limiting, but was nevertheless deemed reasonable. Second, the focus on Pages also allows us to say something about the degree to which the uses of Facebook at the hands of the political actors under scrutiny could be said to be professionalized (Vaccari & Nielsen, 2012).

Data were collected using the Netvizz service (Rieder, 2013). In brief, the service collects Facebook Page posts, as well as the corresponding metadata (timestamp, likes, shares, etc.). While metadata of the type collected and examined here are often the product of the actions (i.e. clicks) undertaken by other Facebook users, the internal mechanisms of the platform and the previous choices made by account holders undoubtedly play a part in making certain content more or less visible (Bucher, 2012). As the goal here was to gather the ‘raw numbers’ rather than obtaining a visual, end-user view of the Pages under scrutiny, the adopted approach appeared to be reasonable. Nevertheless, in order to provide as neutral a starting point for archiving as possible, a new Facebook Profile was created to be used exclusively for the purposes of data collection – in essence, extracting data from them by means of Netvizz.

The focus on metadata from each individual post allows for some variety in the modes of analysis. While previous studies have emphasized studying the number of likes received by individual posts (Bronstein, 2013; Vaccari & Nielsen, 2012) or the Pages that feature them (Jaeger, Paquette, & Simmons, 2010; Kalnes, 2009), the study at hand adopted a wider approach, also gauging the number of comments and shares. While likes can certainly be seen as measurement of popularity, the two latter metrics arguably serve as better indicators with regard to quantity of discussion among followers as well as the potential to go viral. As such, by comparing these metrics side by side, we are able to discern specific differences in online engagement – if any – that Facebook activity at the hands of politicians appears to have.

In order to avoid ‘fake’ social media profiles, the approach adopted here was inspired by Klinger (2013), selecting only Pages that were officially sanctioned by the political parties and/or politicians. Utilizing Netvizz, data were downloaded focusing on the ‘short campaign' – the intensive final stretch of the election campaign season, leading up to election day on 9 September 2013. Specifically, data were archived from 1 August to 10 September, in order to catch some of the election-related aftermath. In total, 548 posts were collected.

Content analysis was applied in order to determine the main theme of each post. Specifically, the six-part rationale discussed earlier was utilized for these purposes. Following similar, previous scholarship, the quality of coding was assessed by testing for intra-coder reliability, recoding a randomly selected 20% sample of the full 548 post set (Klinger, 2013; Schweitzer, 2012; Small, 2012). Given the nominal nature of the data, and following the recommendations of Lombard, Snyder-Duch, and Bracken (2002), Krippendorf's α was employed by means of the macro available for IBM SPSS (Statistical Package for the Social Sciences) (Hayes & Krippendorff, 2007). Reliability was gauged at .76, which is considered a satisfactory result given the conservative nature of the α coefficient (Lombard et al., 2002, p. 596).

Results

First, with regard to the themes of the posts made by the party leaders, the results from the content analysis is presented in Table 1.

Table 1. Main themes of Facebook Page posts by party leaders (% (N)).

	Acknowledgements	Campaign Reports	Informing	Critique	Mobilization	Personal	Total
Grande	3.7 (1)	44.4 (12)	33.3 (9)	0 (0)	11.1 (3)	7.4 (2)	100 (27)
Hareide	1.7 (1)	41.4 (24)	15.5 (9)	1.7 (1)	13.8 (8)	25.9 (15)	100 (58)
Jensen	2.1 (1)	29.2 (14)	27.1 (13)	16.7 (8)	20.8 (10)	4.2 (2)	100 (48)
Lysbakken	3.1 (2)	23.4 (15)	48.4 (31)	0 (0)	18.8 (12)	6.3 (4)	100 (64)
Marcussen	3.3 (1)	50.0 (15)	6.7 (2)	6.7 (2)	23.3 (7)	10.0 (3)	100 (30)
Navarsete	2.6 (2)	22.1 (17)	35.1 (27)	6.5 (5)	3.9 (3)	29.9 (23)	100 (77)
Nissen	4.2 (1)	33.3 (8)	25.0 (6)	0 (0)	29.2 (7)	8.3 (2)	100 (24)
Solberg	3.2 (3)	34.4 (32)	23.7 (22)	3.2 (3)	15.1 (14)	20.4 (19)	100 (93)
Stoltenberg	2.4 (3)	58.3 (74)	12.6 (16)	0.8 (1)	11.8 (15)	14.2 (18)	100 (127)
Total	2.7 (15)	38.5 (211)	24.6 (135)	3.6 (20)	14.4 (79)	16.1 (88)	100 (548)

While the percentages presented in Table 1 suggest some similarities between the party leaders with regard to their individual employments of the Page functionality, certain themes appear to be more fervently employed by some of the politicians than others. Cramer's V coefficient was employed to test the strength of these supposed associations. Following guidelines suggested by Hair (2010), the relation between party leader and post theme was found to be highly significant and of moderate strength (p < .001, Cramer's V = .218). As such, while the associations discernible in Table 1 might not be of the strongest variety, this still provides some confidence in the results presented.

First, the theme of Acknowledgements stands out as the least employed in the posts made (2.7%, N = 15 of all Page posts). Perhaps not very surprising, a closer look at the data reveals that the bulk of these posts were made towards the end of the campaign itself. Typically, these posts feature the party leaders offering thanks to their supporters. As the percentages for this category employed by all party leaders must be considered rather small, it is difficult to make any certain claims with regard to these results. Percentage-wise, Nissen, one of the two leaders of the environmental party (the other being Marcussen), appears as being more ardent in this way; however, this is due to the rather limited number of posts made by this particular user. Focusing instead on actual numbers, we see that while all politicians post at least one message characterized by acknowledgement, Solberg and Stoltenberg, the leaders of the two largest parties, provide three. The fact that the leaders of the two major parties are most active in this regard could be seen as a result pertaining to party size – they have more people to thank. A tentative, yet simplistic, conclusion, this points to the need for future research to look further into the online behaviour of politicians.

Second, Campaign Reports emerge as the largest theme occupying the Pages operated by the politicians (38.5%, N = 211 of all Page posts). This would indicate, then, that these types of ‘reports from the road' lend themselves particularly well to the Facebook platform as operated by the party leaders. As for the most zealous employer of such reporting, the results suggest that Prime Minister Jens Stoltenberg drove a campaign particularly geared towards providing updates from his many campaign stops (58.3%, N = 74 of Stoltenberg's posts). As with the theme of Acknowledgements, this could be interpreted as a consequence of party size, with the leaders of larger parties having more resources at hand to go on an extensive campaign trail, thereby hosting more events from which to report. While her efforts in this regard are largely overshadowed by those undertaken by Stoltenberg, Solberg emerges as the ‘runner-up’ with regard to these types of reports (34.4%, N = 32 of Solberg's posts), a result that seems to corroborate this suggested size effect.

Third, the theme of Informing emerges as the second largest one in our sample (24.6%, N = 135). The leader with regard to these types of posts was Audun Lysbakken, leader of Socialist Left Party (48.4%, N = 31 of Lysbakken's posts). As the leader of a rather small, yet incumbent party at the time of the election campaign, the comparably frequent use of Facebook to present political priorities could be seen as an attempt to differentiate the party from the more dominant Social Democrats, with whom Lysbakken's party was governing at the time of the election. The fact that Navarsete, who was in a similar position before the election, was the runner-up in the category at hand (35.1%, N = 27 of Navarsete's posts) appears to further strengthen this suggestion.

While the other themes are touched upon to some extent by the eight party leaders, our fourth theme, Critique, is employed only by five of them (3.6%, N = 20 of the total sample of posts). As such, the prediction made earlier regarding the relative scarcity of this type of content appears valid. Of the five politicians who chose to utilize Facebook to provide criticisms in this way, Siv Jensen can be pointed to as an interesting case in particular. Percentage-wise, Jensen takes a clear lead among her peers (16.7%, N = 8 of her posts are mainly characterized by criticism) – a result that can perhaps stem from the somewhat marginalized role of the Progress Party in Norwegian society. While Jensen and her party are firmly established and even gained governmental representation following the election studied here, they can also be understood as marginalized in terms of proposing a stigmatized ideology – right-wing populism (Larsson, 2014). Indeed, a large portion of Jensen's critiquing posts deal with how she herself and the Progress Party in general are wilfully misunderstood by established media actors, criticizing what the politician herself sees as biased reporting.

Fifth, the use of Facebook Pages for Mobilization purposes is dominant in 14.4% (N = 79) of the total number of posts. The lead users in this regard – at least based on the percentages reported – are Marcussen (23.3%, N = 7 of Marcussen's posts) and Nissen (29.2%, N = 7 of Nissen's posts), the two spokespersons for the Environmental party. While percentages might be misguiding in this case – seeing how the actual numbers are compared to many of the other party leaders – the relative degree to which these representatives of a minor party make use of Facebook in attempts to activate potential voters is nevertheless interesting. Additionally, as environmental or green parties were often found to make more progressive use of the online during the initial popularization of the Internet for political purposes (Gibson, 2004), the findings presented here concurs with latter scholarship which suggests that such progressive uses at the hands of these types of parties are found also in the current context (Vergeer et al., 2011).

The sixth and final category points to the use of Personal content, effectively providing short glimpses into the private spheres of the party leaders. As this theme emerged as the third most employed (16.1%, N = 88), this suggests that Norwegian party leaders have indeed started to embrace this mode of communicating with their constituents. Again remembering the problems associated with discussing percentages when the actual numbers are rather low, we can conclude that Hareide (25.9%, N = 15 of Hareide's posts) and Navarsete (29.9%, N = 23 of Navarsete's posts) chose to employ relatively more personal detail in their Facebook Page activities. While other politicians might have posted more frequently with this type of thematic content, the fact that both Hareide and Navarsete chose to devote up to one-third of their Page presence to such personalized matters stands out – perhaps especially since their respective parties must be said to enjoy rather limited popularity. As with Campaign Reports, it is difficult to provide any firm assessment of the role of party size in this regard, but the results do indeed provide some interesting tendencies that would at least tentatively suggest such an outcome.

While details regarding the degree to which the party leaders referred to certain themes in their Facebook Page posts certainly provide useful insights into the use patterns and online priorities of different types of high-level political actors, our current efforts to study the results of these activities – here understood as audience engagement – require different analytics. With this in mind, Figure 1 gauges the median amounts of likes, comments and shares per posts made by each party leader.

Figure 1. Median likes, comments and shares per post made by each party leader. Error bars indicate 95% confidence intervals.

The error bars visible in Figure 1 indicate the confidence intervals for the reported medians, suggesting considerable variations regarding these metrics – especially with regard to the median amount of likes per post. With these distributions in mind, we can nevertheless conclude that especially in terms of likes, the leaders of the three largest parties – Stoltenberg (Md = 1587), Solberg (Md = 970) and Jensen (Md = 3878) – emerge as the most popular. As such, the suggestion made by Vaccari in his study of political websites during the 2007 French presidential elections that larger parties ‘usually have stronger ICT infrastructures due to their superior resources' (2008b, p. 6) appears to hold true also in a social media context (Gibson & McAllister, 2014). Jensen in particular stands out – especially when it comes to the statistics of comments and shares. Here, the leader of the right-wing populist Progress Party appears to have enjoyed comparably larger amounts of success in terms of raising discussion and leveraging the viral aspects offered by the sharing functionality.

Figure 2 presents the median amounts of likes, comments and shares for posts classified according to different themes.

Figure 2. Median likes, comments and shares per post classified according to theme of post.

As several of the error bars visible in Figure 2 appear comparably larger than those reported in Figure 1, the spread around the reported medians must be taken into account – especially when it comes to feedback in the form of likes, which again comes up as most common across all themes. Likes, then, appear as especially common in relation to posts featuring Acknowledgements (Md = 2494) – giving thanks to supporters, especially at the very end of the campaign – as well as in relation to posts involving Critique (Md = 1881). As such, while showing gratitude towards or criticizing others appears to result in ample amounts of feedback, levels of engagement yielded as a result of other thematic uses must be regarded as minimal in comparison.

Discussion

The results presented earlier found two categories of posts to be most commonly used. Campaign Reports – insights or glimpses from the campaign trail – accounted for 38.5% of the posts made (N = 211), while the second most popular category saw the politicians Informing about political stances or issues to be prioritized should they do well at the ballots (24.6%, N = 135). While the methodological design employed here allows for more overarching insights as to the different types of posts made, it cannot provide any in-depth understandings regarding the stance taken in each individual posting. With this in mind, the results nonetheless suggest that these two themes, together accounting for well over half of the posts made, indicate that more traditional modes of communicating with the electorate were favoured by the party leaders, suggesting a tendency towards the aforementioned normalization hypothesis. As such, while the theme of Mobilization – posts that more directly urge readers to take political action in some way – account for 14.4% (N = 79) of the posts made, the majority of posts were found to be of a disseminating nature – a result that could be interpreted in the light of the cautiousness often displayed by political actors when adopting novel media outlets. Such cautiousness, then, has been reported from early work (Jackson & Lilleker, 2004; Stromer-Galley, 2000) as well as by later efforts (Lilleker & Malagón, 2010; Vergeer & Hermans, 2013), further strengthening the view of these ongoing developments as primarily characterized by ‘politics as usual' (Margolis & Resnick, 2000).

Figures 1 and 2 provided insights regarding the feedback received – understood here as the median number of likes, comments and shares per post – per politician and per theme. While several of the medians presented in the figures featured rather extended confidence intervals, we can nonetheless point a few interesting details. First, while the results presented in Table 1 suggest that she was not the most adamant user of Facebook, Siv Jensen certainly receives comparably large amounts of feedback on the posts she made during the studied period. As shown in Figure 1, Jensen emerges as the most successful in terms of raising interest on Facebook – not only in terms of likes, but also especially with regard to the reported median shares and comments per post. While Jensen's Progress Party is a major actor in the Norwegian political landscape, the party and Jensen herself are often described as controversial due to their right-wing populist political agenda which often challenges the popular consensus regarding issues such as immigration. Such controversial underpinnings, then, would seem to have a positive effect on the levels of attention gained. Accounting for varying degrees of ideological differences between controversial parties, the result that a comparably provocative party leader would fare well online in this regard is supported in findings from Sweden (Larsson, 2014) as well as from the UK (Lilleker & Jackson, 2010), indicating that contentious political actors might find it fruitful to redress lack of attention in mainstream media by using online technologies, such as Facebook.

Second, with regard to the feedback yielded by posts featuring different themes, Figure 2 suggested ample amounts of likes and to some degree also comments and shares for posts featuring Acknowledgements (Md = 2494) and Critique (Md = 1881). Curiously, while these topics appear to raise interests on Facebook, the results reported in Table 1 suggest that these themes are the least favoured by politicians. As Acknowledgements was found to be the main theme of 2.7% of posts (N = 15) and as Critique was featured in 3.6% (N = 20), the combined results presented here suggest a clear distinction between what topics the party leaders chose to write about in their postings on the one hand, and what topics seemed to activate the Facebook audience on the other. As such, it would appear that criticizing your opponents as well other societal actors is a suitable tactic – at least in terms of gaining interest and promoting online activity. With the increased focus of post-election analytics in a professionalized political environment in mind (Strömbäck, 2007), we might ask ourselves if such a move towards the stylistics of negative campaigning is to be expected in future elections. From the perspective of raising attention, the results presented in this paper suggest that such an approach might be deemed feasible by politicians and political consultants alike. From a more normative democratic perspective, however, the implications of these results would most likely be seen in a different light (Lau, Sigelman, & Rovner, 2007).

Both figures indicated the dominance of likes as the apparent preferred form of feedback, followed by comments and shares. While Facebook popularity is typically measured as Engagement (Rieder, 2013) – essentially a combination of the three measurements – this lumping together of what essentially must be seen as apples and oranges provides a metric that to a certain extent conceals the various types of feedback received. While the comparably simple act of clicking might be more feasible to undertake for the average Page visitor, the true value in the Facebook platform lies in the potential for the content to ‘go viral’. As this potential is unlocked primarily through the content being commented upon or shared from user to user, the results presented here suggest that a more careful assessment of this potential is suitable. Indeed, while certain posts, baring the characteristics described earlier, certainly reach high amounts of shares and comments, these figures are dwarfed in relation to the number of likes received. As such, what is sometimes discussed in terms of lower or less demanding forms of engagements appear as most common (Chung & Yoo, 2008; Larsson, 2011b) – suggesting a clear presence of ‘clicktivists' (Karpf, 2010), as well as the need to critically appraise the function of the Internet in parliamentary-political settings.

In closing, while this study has provided important insights into the uses of Facebook at the hands of high-end politicians, it has limitations that need to be duly mentioned. First, while the fact that the Norwegian party leaders provided a total of 548 posts during the month leading up to election could be seen as an interesting result in itself, one could also argue that the claims made here are based on limited data. While the time period studied is suitable, future research might find it fruitful to adopt longitudinal designs – perhaps also covering the activities of other, less senior, politicians.

Second, the methodological design employed here allowed us to gauge – to some extent – the degree to which various forms of feedback vary on Facebook. Future research should consider to take the responsiveness of the politicians into account (Vaccari, 2014). While attempts at dialogue from politicians might be expected to be few and far between (Sweetser & Lariscy, 2008), the effect that such an activity might have on the levels of likes, comments and shares should be assessed more clearly.

Third, interested researchers might find it rewarding to take the audiovisual qualities of social media more clearly into account. As pictures and videos can be said to be an integral part of Facebook – and indeed of political campaigning in itself (Lilleker et al., 2011; Vaccari, 2013) – the inclusion of such content could be expected to carry some weight in determining the popularity of specific posts.

Notes on contributor

Anders Olof Larsson (PhD, Uppsala University, 2012) is a postdoctoral fellow working in the Social media and Agenda-Setting in election campaigns project at the Department of Media and Communication, University of Oslo. During his PhD work, which was funded by The Swedish Research School of Management and Information Technology, Larsson took part in the 2011 Oxford Internet Institute Summer Doctoral Programme. His PhD thesis – ‘Doing Things in Relation to Machines – Studies on Online Interactivity' (pdf) – was awarded the 2012 Börje Langefors award (for best Swedish thesis within the field of Informatics) and the 2012 FSMK Doctoral Dissertation award (for best Swedish thesis within the field of media and communication studies). Anders Olof Larsson's research interests include the use of online interactivity and social media by societal institutions and their audiences, and online political communication and methodology, especially quantitative methods. [email: [email protected]]