Keywords

1 Introduction

The media landscape for political communication has never been as complicated as it is today. Fake news, bots, polarization, right- (and left-) wing activism, echo chambers and a plethora of new online news sources with an agenda has been highly debated in recent years [1,2,3]. After the election of Barack Obama in 2008 and the Arab spring in 2009–2010, scholars and media experts were highly optimistic about the democratic potential of the Internet and social media, talking about a new dawn for enlightened debate and freedom of speech. This optimism is now slowly turning to a more pessimistic, or perhaps more balanced, view of the relationship between social media and democracy following the election of US president Donald Trump, the Brexit referendum and other cases [3].

While a lot of this is centred around the American context, there is also research on the Scandinavian countries, for example a study of the right-wing Sweden democrats, [4], or more general studies of campaigning and participation in social media, i.e. [5, 6]. A common variable in many of these examples seems to be populism, where certain prominent individuals use social media to circumvent traditional media channels in order to present a popular and direct message to the people [4].

The media channels themselves might contribute, but this trend is resting on real societal challenges. A special issue of the journal American Ethnologist examined the Brexit referendum and Trump campaigns, both of which have been framed as social media campaigns. The articles point to several underlying explanations such as increased inequality in society, leading to increased nationalism, concerns about immigration and a sudden rise in anti-globalization sentiment [7]. Data from the European Social SurveyFootnote 1 shows a general lack of trust in traditional media, political parties and political institutions which could further help explain the rise of populist politicians using social media to reach out to disgruntled citizens. There are changes going on in the public sphere, and public opinion is at the same time both divided and polarized, but also empowered, and more research is needed to understand these changes [8].

In Norway, the power and democracy project was already in 2003 concerned about a decline in representative democracy, with voters moving between parties depending on single issues and media attention [9]. While Norwegians in general have somewhat higher trust in both media and political institutions [10], Norwegian politicians are signalling that they want more citizen dialogue and user-involvement in the political process [11], and they are increasingly attempting to achieve this through social networking systems (SNS’) and other digital communication channels in order to reach out and communicate directly with voters [12].

This paper responds to Mindus’ [8] call for more research on the current changes to the public sphere by examining communication genres in the 2017 Norwegian parliamentary election. Following the same research design as studies of the 2009 and 2013 elections, we seek to answer the following research questions:

  • RQ1: Which genres were used during the 2017 election campaign?

  • RQ2: Given the events of recent years, have Norwegian political communication changed compared to previous campaigns?

The rest of the paper is structured as follows: Sect. 2 provides an overview of related research, specifically on the topics of democracy, the public sphere and genre theory, which informs the findings and discussion sections. Section 3 presents the research approach of the study, and Sects. 4 and 5 present the findings and conclusions with some possible directions for future research.

2 Related Research

2.1 Theoretical Lens: Democracy, Participation and the Public Sphere

Democracy can be conceptualized in a number of ways [13]. There are several models of democracy in literature defining everything from direct democracies to the parliamentary/representative democracy we find in most western countries today [14]. This paper applies the traditional representative model, where our role as citizens is to vote in elections and participate in societal debate as members of an informed public, in line with the writings of Habermas [15] and Dewey [16].

Habermas’ concept of the public sphere as the “domain in social life in which such a thing as public opinion can be formed” [15] (p. 261) has often been used as theoretical lens for studies of online democracy, as it can be understood as a mediating layer between politicians and citizens where “the interaction between citizens, civil society, and the state, communicating through the public sphere, that ensures that the balance between stability and social change is maintained” [17]. However, to use the concept in today’s fragmented media landscape, we need to discuss not one, but several overlapping public spheresFootnote 2. Trenz and Eder [18] presents four different archetypes of public sphere; discourse, political protest, political campaigning, or consensus. We can talk about a mainstream public sphere in mainstream media and politics, but with SNS’, alternative and marginalised groups can have a voice [19], creating their own “counter” public spheres [20]. An effective protest-based or counter public sphere should follow three conditions: (1) The intention of protest should be to address issues relevant to the democratic community. (2) Protest should provide an alternative to, or new information for, the mainstream discourse. (3) It should not promote discourses incompatible with the public sphere principle of inclusion, or “aim to force the alteration of a decision” [21]. In SNS’, the lines might be even more blurred, due to “trench warfare” dynamics where confirming and conflicting arguments both tend to reinforce existing attitudes [22]. Further, those with a strong interest in politics tend to seek out and engage with a variety of political news sources [23]. The findings and discussion will illustrate how the 2017 election can be interpreted as both campaign and counter public sphere at the same time.

2.2 Analytical Lens: Genre Theory

A genre can be defined as “a conventional category of discourse based in large-scale typification of rhetorical action” [24]. Genre theory can be applied to classify communication practices, and has been applied to several eParticipation studies for classification and understanding [25,26,27,28,29], as well as for modelling purposes [30]. Genre theory provides us with a lens for detailed understanding of political communication, beyond the observation of technological functionality [31]. Genres are recognized by having similar form and content, where form refers to physical and linguistic features, and content to themes and topics of the genre [32]. As digital media has become more common, functionality of the medium delivering the genre has been added as a third construct [33]. Genres can be defined by examining form, functionality and content, by using the 5w1h-method [34, 35]:

Where tells us where the communication takes. Why explains the purpose of the genre. When refers to the time where communication takes place. Who defines the actors involved in communication, the sender and receiver of the genre. What is the content of the genre and How describes the technical needs for delivery of the genre.

The genres used by a given community can be seen as a genre system [35] and this system can reveal a “rich and varied array of communicative practices” shaped by community members in response to norms, events, time pressure and media capabilities [31]. Genres are useful for studying communication in SNS’, as the introduction of new media over time often leads to new communication practices which genre theory allows us to map and analyze [27]. Genre theory, including the technological functionality of the medium the genre is enacted within, allows us to better understand the interplay between the social and the technical [33].

3 Research Approach

The objective of this paper is to examine how the genre system used by Norwegian political parties has evolved since the last election and to discuss this considering current trends in political communication as presented in the introduction. The study was conducted using a qualitative, interpretive approach.

Data Collection:

Data for this study has been collected over three periods: The elections of 2009, 2013 and 2017. Data for the 2009 study was collected through semi-structured interviews with representatives from the seven political parties that were represented in the parliament before the election (Socialist Left, Labor, Center Party, Liberals, Christian people’s party, Conservatives and the Progress Party). In 2013 and 2017, follow-up interviews were done electronically to confirm findings from 2009. Further, SNS content (posts, comments and interactions from the pages of the political parties) during the main campaign period in June to election day in September, has been archived and analyzed using Nvivo and Tableau software. In addition, statistics from Likealyzer.com, the European Social Survey, and the polling company TNS Gallup has been used to examine trust in media and politics.

Data Analysis:

The combination of interviews and SNS content made it possible to compare what informants say with what we can observe happening. This is used to map the genre system in SNS political communication. For this study, only Facebook data has been analyzed, since Facebook is by far the most used channel in Norwegian politics. The genre systems have been analyzed using the 5W1H method presented in Sect. 2.2. Of the around 6000 posts collected, a selection has been coded until saturation (no new genres emerging from further study). When no new genres were identified, the remainders of the posts were quickly scanned to see which genre category they matched. Due to space limitations, the findings are presented using the “form/function/content” constructs [31, 32].

4 Findings

4.1 Summary of Findings, 2009 and 2013 Elections

The interviews made in 2009 revealed that the political parties agreed on three objectives for political communication in SNS’: Dialogue with citizens, contributions from citizens, and involvement in party activities. When asked if these objectives remained the same, the parties agreed in 2013. In 2017, they still agreed that these were the overall objectives, but several respondents pointed out that they have evolved and developed a more fine-grained set of strategies, objectives and goals for different channels. In terms of channel use, blogs were popular in 2009, almost gone in 2013. Facebook emerged as the most important channel, and there were some experiments with Instagram. One of the parties said SNS communication had been moved from communications to marketing. The objectives are presented in Table 1.

Table 1. Political party objectives for SNS participation
Full size table

In 2009, seven genres were identified:

  • Policy comments are comments from citizens on party policy. These come in many forms: Wall or discussion posts on Facebook, in Twitter messages and blog comments.

  • Calls for action mainly originate with the party but are often distributed through citizens supporting the party making the call. This genre incorporates calls for volunteers, competitions and calls for action in specific cases. Several parties have created Facebook groups for specific parts of their policy. Calls are presented in video, with links to the video posted to Facebook and Twitter.

  • The Q&A genre is perhaps the genre that citizens are least satisfied with. Many questions on Facebook walls remain unanswered, or are answered unsatisfactorily. Some citizens ask why politicians bother having a presence in SNS when they do not engage in conversations with citizens.

  • Appeals to the party are similar to policy comments. The difference is that where policy comments reflect directly on the party’s political program, appeals are more specific, asking what the party intends to do with this or that matter. There is some frustration among citizens when these are not answered.

  • Greeting is an interesting genre. At his birthday, Prime Minister Jens Stoltenberg received hundreds of greetings wishing him a happy birthday. In 2013, we saw the sme related to birthdays and other personal occasions. This genre, while not directly political, could be seen as narrowing the gap between politician and citizen, creating a sense of personal attachment.

  • Personal accounts are mainly found in blogs, as response to politicians asking for the stories of individual citizens. The most interesting example is where the minister of health asks for people’s stories as input to a major health reform. In 2013, this genre was still present, but had mostly evolved so that personal accounts were incorporated into other genres such as debate and policy comments.

  • Video responses from citizens are rare, but some examples exist. These are typically posted as responses to competitions where parties ask citizens to contribute. There are also responses between parties, where video is used in a similar manner to newspaper debates, and responses between politicians belonging to the same party. This genre disappeared in 2013.

In 2013 five new genres emerged:

  • Debate was not present in 2009, perhaps because overall activity was low. In 2013 there was much more activity in SNS’, and this led to several rational debates on several policy issues.

  • Support and non-support Citizens showed their support (“steady course. Four new years of labour”) or lack of support (“about time someone else takes the wheel”) for the party.

  • Disgruntlement is like non-support, but different in form. Here we found Sarcastic comments about the party, unpleasant comments about the party and its politicians.

  • Link as genre simply consists of links to news articles and other sources. This is often accompanied by a short statement (“Do something about this, please!”) or question (“Why is this allowed/not allowed”?). Linking to content to support a position shows the richness of digital communication, and the easy by which relevant information can made available to people.

In summary, the 2009 election showed an emerging genre system for SNS campaigning, but there were many voters who were unhappy with a lack of response from the parties. Responses to party calls for input on specific issues received a lot more comments than other politician-initiated genres, indicating that citizens want to be heard and feel that their input is used for something if they are to participate. 2013 introduced several new genres, indicating that SNS’ were moving towards a richer genre system for communication between citizens and parties. The parties had listened and were much quicker to reply in 2013. They also asked for input on a wide range of policy issues and received hundreds of replies. The main challenge in 2013 was that the form (language use) of genres varied greatly. Citizens communicated with a language ranging from highly informal, with lots of typing errors, exclamation marks and capital letters to the formal language more common in political communication, and the border between a post or comment coded as “debate” and one coded as “disgruntlement” was sometimes quite blurry.

4.2 Genre System 2017 Election

The 2017 election saw some interesting variations from the previous two campaigns. While 2013 gave the impression that the political parties were moving towards a form of “politics 2.0”, focusing on interaction and feedback from voters [36], this is less visible in 2017, with most parties being more focused on getting the message out to the public. The political parties are heavy users of the following genres:

  • We want to is the most commonly used genre from all the parties. The content is directly related to the party program, with statements such as “we want to [do something] because [of some reason]”.

  • We have is only used by the current governing parties. In this genre, the ruling parties present their accomplishments from the last parliamentary session. Sometimes accompanied by the phrase “you know what you have, do you dare vote for something untested”. Video and images are frequently used.

  • Non-support is frequently used by most parties. In this genre the party attacks the policy and policy consequences of other parties. Political parties have always done this, but the tone is harder than in previous elections. Making fun of the other parties has become a lot more common, as exemplified by the Conservative’s image of sun lotion with the text “don’t be red this summer, vote Conservative”

  • Slogan is related to we want to, but in place of concrete policy issues and references to the party program the slogan is more idealistic in nature and is not supported by arguments as to why the statement is true: “We are the best party for young people!” or “Vote for us if you want change”

  • Personal accounts come in two forms: One is promoting popular politicians in the party, the other is “interviews” with typical voters from large voter groups.

  • Contribute is where parties ask voters to participate. This can be in the form of Q&A sessions or, more commonly, by asking voters to register for updates, become members of the party or act to support the party.

  • Society & Context involves parties posting links and updates about current affairs they somehow believe reflects on the values and ideology of the party. For example, the greens post quite a lot about global warming and the conservatives wish people happy pride or post content about the importance of reading.

  • Experiments is a genre where parties try out different formats of communication, using podcasts or live streaming, giving someone a GoPro to document a day in their lives and similar. Not all parties try this, and the genre is not frequently used. However, this is a sign that there is still some experimentation going on in SNS’.

The citizens commenting and posting use the following genres:

  • Non-support and Support are popular both in comments and posts created by users, even more in 2017 than in 2013.

  • Greetings is a popular way of showing support, as it was in 2013. Popular politicians celebrating their birthday or other major life event get a lot of congratulations also in 2017.

  • Disgruntlement is another genre that emerged in 2013 and is sadly a growing genre. There is a lot of sarcasm and outright hostility towards most of the parties. In fact, most comments and user posts fall into categories arguing for or against the party. This can be interpreted as a sign that polarization is occurring also in Norwegian politics.

  • Debate and policy comments are present, but very little compared to the three genres above. There are a few examples of users attempting to start a debate based on evidence, facts and arguments, but most often these posts are taken over by non-supportive or disgruntled comments.

Genres Have Common form and Functionality

It is quite clear that the parties are done experimenting and are relying more on data and statistics in 2017. The genres all have similar form and functionality: Posts are short, most are within around 200 words or less. There is a video, link or image attached to almost all of them and each post focuses on one simple idea or issue from the party program. Looking at the timeline of posts, the parties have more or less the same frequency of posting (2–3 posts every few days, growing to 8–10 posts closer to election day).

Engagement and Effects of Genres

Putting a face on policy seems to become more and more important, as posts with the name of popular politicians are frequent in the top 10 posts receiving engagement from voters. Other genres creating engagement include we want to, we have, and non-support. However, the clearest observation is that popular politicians create a lot of engagement, both supportive and non-supportive in form. While several parties make some attempts at two-way communication with citizens, for example by creating monthly Q & A sessions, asking for input on specific policy issues etc., none of these ranks high in the list of posts receiving a lot of engagement. Likealyzer is an online service that analyses Facebook pages. Analysing the Norwegian political parties using this tool shows that the parties could improve when it comes to debate. With a response rate varying between 20 and 44%, and little interaction with other pages, the Likealyzer statistics strengthens the impression that the political parties view SNS more as a one-way campaign tool than a channel for interaction and debate. Facebook in 2017 is more about marketing the party’s program than about dialogue, and the voters play along, acting like supporters in a game of football.

Themes and topics receiving engagement vary between parties, depending on the issues they have chosen to put high on the agenda. However, some themes create a lot of engagement across party lines. As we have seen in other countries, themes of inequality, social dumping and people being left behind are common, but blame is placed differently depending on people’s political beliefs. Immigration is one issue that has really contributed to polarization, especially following the rise of asylum seekers from Syria in 2015. There is also a lot of criticism of globalisation and the EU. A lot of people seem to blame increasing inequality on these factors. Others, especially the far-left opposition, instead blame the ruling government’s policies. As the data collection tool anonymises the author of posts and comments, it is unclear if this is caused by a few very active users, or if this is a larger trend.

4.3 Growth in SNS Use – How Effective Is Facebook for Reaching Out?

There is little doubt that Social networks are increasingly important as a communication channel for political parties. Most of the parties have seen a massive growth from 200–6000 followers in 2009, up to 16–160.000 followers in 2017. The media use survey from TNS Gallup confirms this, showing that TV, newspapers and Social networks are equally important when citizens seek information related to politics. Two elections ago, TV and newspapers scored a lot higher than any digital medium. Figure 1 shows the growth in Facebook followers from 2009–2017, and the number of votes the parties received in the three elections.

Fig. 1.
figure 1

Follower growth and votes received 2009–2017

Full size image

There seems to be little if any correlation between the number of votes received and the increase in followers. The progress party is by far the most popular Facebook party, but also the party with the largest drop in votes from 2009–2013. Labour has lost more than 100.000 votes in the same period, while gaining an equal number of followers. The Center party doubled their vote from 2013 to 2017, but only have 25.000 Facebook followers. On the other hand, the green partyFootnote 3 was elected to parliament in 2013 and cited social networks as an important factor. However, with 60.000 followers they are the fourth largest party on Facebook, but still has less than 4% of the total vote. Finally, the numbers from both 2013 and 2017 seem to confirm an increased focus on person over party. The leaders of the three largest parties (Labour, Conservatives, Progress party) all have more followers than their respective parties. Prime minister Erna Solberg has almost twice the number of followers of the Conservative party she belongs to.

5 Discussion – Implications for the Public Sphere

The populist rhetoric from other countries and campaigns seems to have influenced communication in the Norwegian election of 2017. Political parties are less eager in seeking two-way communication, and even when they do, voters seem more concerned with showing their support or non-support of the parties. The topics and issues discussed, such as immigration, globalization and inequality have been on the agenda before, but the tone of the conversation is more aggressive than it has been, and polarization seems to be a factor in Norwegian political communication (at least on Facebook). This study supports the findings of Dubois [23], showing that SNS does not equal echo chambers, as a lot of the comments given to all the political parties are negative.

Examining the election campaign in SNS’ using democracy models [14] and the public sphere [15, 18], SNS seem to be falling in line with traditional representative democracy, as the attempts at two-way communication from the past elections is less visible today. SNS have become yet another channel where parties seek to convince voters to vote for them. As for the other part of representative democracy, a public engaging in reasoned debate, there is little evidence of that in the genre system of the Norwegian political parties, despite a few efforts at reasoned debate. Instead, we see a form of hybrid public sphere. A mix of what Trenz and Eder [18] would call a consensus-based and protest-based public sphere. This supports other research stating that echo chambers are not as much of a problem as previously thought [23], but does little to alleviate the fears that politics is becoming increasingly polarised.

6 Conclusion

A genre system for political campaigning emerged in 2009, and in 2013 this had matured significantly, with more genres, more users and experiments with two-way communication. In 2017, however, SNS’ are mostly used as a one-way communication tool, with participation mostly limited to cheering or opposing the statements from the parties, confirming that Norway has not been immune to the issues and events happening in other countries. This paper points to several possibilities for future research: Data-driven methods [37] could be applied for a more detailed analysis of each and every comment, post and engagement, in order to quantify the findings of this content analysis. An in-depth (network) analysis of the people being active and commenting would help understand if the activity in SNS’ is just a few people being very active, or if this is a general societal trend, especially if this was linked to other analyses of people’s opinions and actions outside of social networks. This type of research should also examine the broader social world, to identify any differences between SNS’ and real-world conversations. Finally, broader studies of sentiment towards issues such as immigration and globalisation should be carried out, in order to better understand the underlying motivations and processes driving these issues to the forefront.