Negotiating the Political Self on Social Media Platforms An In-Depth Study of Image-Management in an Election-Campaign in a Multi-Party Democracy

: The elections 2010 were the first in Sweden where social media platforms were used to a large extent by politicians and parties in their campaigns. In this paper we follow the liberal parliamentarian Nina Larsson, who in tandem with traditional election campaigning used social media platforms with the guidance of a local communication agency, Hello Clarice. The paper is theoretically grounded in an understanding of our time as late modern, of social media use as expressive and web campaigning as to large extent revolving around image-management. The research question that will be attended to in this paper is how Nina Larsson used social media platforms in her campaign negotiate the image of herself. The methods used for empirical data-gathering are inspired by (n)ethnography, with both participant observation online and offline, interviews as well as content analyses of Nina's social media postings. Results indicate that she used social media platforms to control her political image, to amplify selected text - texts that often originated in offline/broadcast media – and to negotiate a position within the Liberal Party rather than to deliberate with potential voters.

n this paper we follow the Swedish politician Nina Larsson during the election year of 2010.Nina is a young female politician in her 30's representing Folkpartiet (the Liberal Party) voters from the Värmland constituency, midwest of Sweden.Nina was elected into the Swedish Parliament 2006 and for her re-election campaign 2010 Nina contracted the communication agency Hello Clarice for guidance and help conducting a campaign online through Facebook, Twitter, a special campaign website and two blogs, as a complement to traditional campaigning.
There is hype around social media platforms when it comes to election campaigning.Politicians are no doubt turning to social media in increasing numbers (see Anduiza 2009;Jackson and Lilleker 2009;Montero 2009;Zafiropoulos and Vrana 2009;Goldbeck et al. 2010;Grusell and Nord, 2011;Jackson and Lilleker, 2011).However, many studies have revealed that the effect of such campaigning on electoral participation and opinion formation is small (Gibson et al. 2008;Jackson and Lilleker 2011;Larsson and Moe 2011).The question that then arises is for what purposes do candidates use social media platforms?Except for strategic communicators -arguing for the potential of social media platforms to strategically target voter groups (see Zafiropoulos and Vrana 2009;Schweitzer 2010;Jackson and Lilleker, 2011) -deliberative democrats have underlined the potential of social media platforms to further communicative deliberations between politicians and the electorate (see Coleman and Blumler 2009;Fenton 2010).Having studied and participated in interactions on Nina's social media platforms, I could conclude that she neither used social media platforms primarily for deliberation -as she herself stated was her aim -nor strategically to increase visibility and attract followers -as the communication agency stated was the aim when coaching her.Rather, she used social media platforms reflexively, to express and promote herself and negotiate her political image within a party political network (Svensson 2011a;

Image-Management and Election-Campaigning on Social Media Platforms
Social media is a contested term since it implies that traditional media would not encompass social dimensions.What is often referred to when talking about social media are online communication platforms were the social seems to refer to the possibility of users to influence and interact with the content and each other in some way or another.O'Reilly (2005) claims that for a website to be defined as social the user him/herself should be able to participate and contribute to the content of the site.The user should be able to take control over his/hers information and the overall design should be interactive and user-friendly (ibid.).The more elaborate definition of social network sites (SNS) focuses on the possibility for users to articulate their social networks and make them visible to others (Ellison and boyd 2007, p. 2).SNS are defined as web-based services that allows users to construct a public, or half-public profile, tie this profile to other users, sometimes self-selected, whose contacts in turn are made available by the service (ibid.).In this article I will attend to Nina's uses of Facebook, Twitter, an interactive campaign website and two blogs.Since neither the blogs nor the interactive campaign website would fit the SNS definition by Ellison and boyd, I will use O'Reilly's more encompassing definition of social media.
The rise of social media platforms are interlinked with heightened processes of individualization.The experience of increased personal autonomy and expressions of this are among the most debated trends in our time (see Lasch 1979Lasch /1991;;Giddens 1991;Bauman 2001).In accounts of the late modern era, processes of individualization are given priority over collectively shared cultural frames of references that dominated social spaces and their organization in modernity (such as family, nation, class, party affiliation et cetera).In previous work I have discussed political participation in late modernity as increasingly characterized by expressive motivations (see Svensson 2011aSvensson , 2011bSvensson , 2012Svensson , 2013)).In light of the increasing importance of reflexivity in late modern societies, the idea of expressive rationality relies on an understanding of preference expression as part of identity negotiation (see Svensson 2011b).Giddens' (1991) understanding of self-identity as dependent on a story/ biography, a sense of ontological coherence and continuity in our everyday life, is important for this argument.Out of his reasoning it becomes rational to make our story/ biography explicit by expressing our identities.Such identity expressions, I argue, are being increasingly afforded by social media platforms (Svensson, 2011b).With the rise of digital media, processes of identity negotiation and biography maintenance, tend to become more networked in character (Castells 2001, pp. 122-125).Since we lack bodily cues in computer mediated communication, connections to others become essential in the management and negotiation of ourselves and our stories online (Miller 2008).Hence, managing our identities on social media platforms largely revolves around connecting ourselves to other nodes in the network (users, fan pages, groups et cetera).It would therefore be wrong to conceive of late modern individualization and a sense of collectiveness as mutually exclusive (see also Lasch 1979Lasch /1991)).
Here I find the concept of networked individualism particularly illuminating for understanding the practice of linking the individual to different groups online (see Castells 2001, pp. 129-133;Wellman 2001).In this way the emerging digital media landscape gets conflated with pluralization of lifestyles, tastes and subcultures that in turn work in tandem with our times manageable and negotiable individuality and life story (Donath and boyd 2004;Dahlgren 2009, p. 152).When sociocultural processes, media and patterns of communication mutually reinforce each other in this way, I believe it is appropriate to speak of a digital late modernity (see also Svensson 2011b).
If we now turn to politicians and their practices of identity negotiation in digital late modernity, the internet is increasingly used by elected representatives and discussed as tools within the field of political communication (Bimber and Davies 2003;Gulati 2004;Foot and Schneider 2006;Gibson et al. 2008;Stanyer 2008;Anduiza 2009;Costa 2009;Jackson andLilleker 2009 Montero 2009;Goldbeck 2010;Schweitzer 2010;Dimitrova et al. 2011;Grusell and Nord, 2011;Jackson and Lilleker 2011;Larsson 2011;Karlsson et al. 2012).Effective selfprojection through the media is essential for politicians building a bond with their voters.It is about projecting the right image in the media.Connected to theories of late modern individualization and expressive rationality is an understanding of social media usage as a way for politicians to negotiate an attractive image in front of their constituency.Only by being online politicians wish to convey a message -especially to younger segments of the electorate -that they are up-to-date and with their time (indicating a kind of jump-on-the-bandwagon mentality, see Jackson and Lilleker 2011, p. 88).The rise of electronic media is thus intertwined with personalization of politics.Today it is argued that politicians are scrutinized for authenticity and style by an electorate that is increasingly visually and emotionally literate (Stanyer 2008, p. 415), hence a need for politicians to manage an attractive image online.Many political communication researchers have translated Goffman's (1959) well-known work on how individuals seek manage what impression they give to others into motivations for politicians to use social media (see Gulati 2004;Stanyer 2008;Jackson and Lilleker 2011).Studies have shown how the internet presents a front region for politicians to present themselves to the electorate (Stanyer 2008), to negotiate different images (Gulati 2004) and to promote themselves (Goldbeck et al. 2010).
But how are social media platforms used for image-management?Jackson and Lilleker's (2011) study of politicians twittering in the UK found that most of it revolved around impression management through of self-promotion.As with their use of other social media platforms (see Jackson and Lilleker 2009), Twitter seems to be largely about politicians promoting themselves by negotiating an image as professional politicians and/or as individuals with a life and personal interests.Such image-management is often done through selective disclosures about their private lives (Jackson andLilleker 2009, 2011), and in this way carefully managing the information they reveal to influence the impressions the electorate construct.In the US, Goldbeck et al. ( 2010) conducted a study of congress people using Twitter and found that they primarily used the platform to disperse information, particularly links to news articles about themselves and to their blog posts.Almost 55 percent of their tweets were such informational posts, providing a fact, an opinion, a link to an article, or a position on an issue et cetera.The vast majority of these informational tweets (72 percent) contained links.The second most common tweets (27 percent) were posts on the location or activities of the congress people.These were posting about daily activities and events like meetings, dinners, and workouts.Also Stanyer (2008) shows how the web is used by politicians for self-promotion, to trumpet accomplishments and success, for emphasizing qualifications, family relations, appear empathetic and having a bond with their voters.By conducting a content analysis of 100 websites, 50 from the US and 50 from the UK, he found that politicians wanted to project a personable, approachable, knowledgeable and friendly image.They were keen to be seen as dedicated to serve their constituents and they sought to project themselves as active and working hard.But at the same time politicians wanted to portray themselves as not only dedicated to the job but also concerned about local issues.In Gultati's (2004) content analysis of the images displayed on the home pages of 344 US congress people, he could outline two distinct presentations: an insider style and an outsider style.Insiders tried to manage an impression that they were representatives who had influence, who were policy leaders, and who had access to power.In contrast outsiders left the impression that the similar to the voters -one of them -that they were representatives who have not lost touch with the people or the district.
These studies above provide us some insights into the practice of politicians imagemanagement online and can thus be used to elaborate the rather broad research question of how Nina used social media platforms to manage/ negotiate the image of her as a politician in her election campaign 2010.Given these studies, questions to be asked to the empirical material is how Nina promoted herself on her social media platforms?What parts of her private life did she disclose?How did she portray herself as a person and a politician?What and who were Nina linking to in her postings?What locations and activities was she sharing?Did she use family relations or bonds with voters in her image-management and was she trying to be seen as dedicated to the constituency and local issues?Was she managing an image of herself as an insider or outsider?Finally what does all this say about the image she wanted to portray of herself?These questions will guide the upcoming analysis of Nina Larsson's image-management on her social media platforms during the election campaign 2010 to the Swedish Parliament.But before that, first some notes on the context, the case of Nina's campaign and methods for data-gathering.

Context and Case
Sweden is a multi-party representative democracy.This is important since most studies referred to in the above section are from candidate-centered political systems.According to Stanyer (2008, p. 427) political candidates in such systems tend to emphasize local connections and family to a much greater extent than those in party-dominated systems, who are more likely to highlight party symbols and affiliations in their image-management.Stanyer based his study comparing the US and UK.Sweden is even more different from the US than the UK since it is a multi-party representative system with seven parties having more than four percent of the vote (and able to get over the four percent threshold to gain seats in the Parliament) during the time of Nina's election campaign.This multi-party system makes alliances and cross-party cooperation a common feature of Swedish politics, and has paved way for a political culture rather based on consensus and cross-party dialogue than conflict.
Sweden has a very high percentage of ICT use.In the election year of 2010 broadband penetration reached 92,5 percent of the population (World Internet Stats, www.internetworldstats. com/eu/se.htm,retrieved April 17 th 2012).68 percent of the population aged between 9-79 used the internet daily and 35 percent used social media platforms (Nordicom Internet Barometer).The use of internet and social media platforms in Värmland, Nina's constituency, reflects the nation at large.70 percent use internet daily or several times a week and 35 percent of the respondents claim to use social media daily or several times a week (compared to 46 percent who claim not to have used social media at all during 2010) i .
But do voters use social media platforms for political purposes?A recent panel study on the 2010 Swedish elections found that reading blogs about politics or current affairs were the most common social media platform used with 8 percent (Dimitrova et al. 2011).4 percent followed a politician on Facebook and only 1 percent used Twitter to follow a politician or or political party (ibid.).These numbers are further confirmed in Värmland.Only 35 percent of those that used social media several times a week had participated in an online campaign/ petition through e-mail or social media in 2010 (17 percent of the total sample), and less than 14 percent have discussed politics at all on an online debate forum during the election year 2010 (7 percent in the total sample).These numbers should be related to that 53 percent of the respondents in Värmland state that they are very or quite interested in politics in general and it should also be related to that 53 percent claimed to have discussed politics during the last month, but apparently not on social media platforms.
If we now return to politicians and their uses of social media in campaigning, there are some studies of individual candidates using social media especially from the US and the UK as referred to in the previous section.I have found no such studies in Sweden.Gustafsson (2012) made focus groups interviews with party members 2009 (among other groups) in order to study the effects of social media on political participation.Studies from the 2010 Swedish elections have either been focusing on political parties' (Grusell and Nord 2011;Karlsson et al. 2011;Larsson 2011), or the electorate's uses of social media (Dimitrova et al. 2011;Larsson and Moe 2011).
CC: Creative Commons License, 2012.Karlsson et al. (2012) made a longitudinal study of parties usage of social media platforms and found that all ten parties in the study had social media presence on Facebook, Twitter and YouTube, but that activities on these platforms were heavily concentrated to the weeks running up to the elections (see also Larsson 2011).Grusell and Nord's (2011) study of the political parties use of social media in the 2010 elections concludes that social media are generally not yet mastered by politicians and political parties, and not yet integrated in parties overall communication strategies (see also Karlsson et al. 2012).
Hence, there is a need to study in more detail how individual candidates use social media platforms in a multi-party representative democracy.And there is an increasing incentive for individual candidates in Sweden to conduct their own election-campaigns since in future elections it will be easier for the electorate to circumvent the by the party decided candidate list, by ticking their own favorite candidate.Nina Larsson is not a typical case since she did integrate social media into her campaign and she did attempt to master them with the help of professionals (as compared to the study by Grusell and Nord 2011).But even though a deviant case her conscious use of social media makes her suitable for an in-depth explorative study.Nina is an early adopter and other politicians might follow her example since social media is likely to continue playing an increasing role in future election campaigns.This study is thus explorative in its aims rather than hypothesis testing (see Gerring 2007, pp. 37-39).Nina is an example of a future generation of politicians and her Värmland web-campaign provides us insights into how social media platforms are used for image-management in a multi-party representative democracy and with professional input from a communication agency.At the same time we need to keep in mind that Nina is just one candidate and there is a need for future quantitative studies to confirm the representativeness of this case.
Which social media did Nina Larsson use?During the election campaign she was active on Twitter (http://twitter.com/NinaLarsson),Facebook (http://www.facebook.com/ninalarsson),a campaign website (www.kryssaninalarsson.nu)and on two blogs, one connected to the regional newspaper VF (Värmlands Folkblad, http://blogg.vf.se/ninalarsson) and one personal blog Ninalarsson.setied to her as politician representing Folkpartiet (the Liberal Party, http://www.Ninalarsson.se).Her uses of these platforms will be further attended to in the result section.Next I will attend to how the data was gathered and analyzed.

Notes on Methodology
The study of Nina Larsson has been inspired by (n)ethnography (ethnography being a good methodology for case study research, see Gerring 2007, pp. 68-69).The aim of netnographic research is to understand the social interaction taking place online, hence a focus on the interaction and user-generated information flows (Berg 2011, p. 120).The netnographic approach thus suits the aims of this research project since I am studying how Nina used social media platforms.Doing netnography, I followed Nina Larsson on all her five social media platforms, took field notes and screenshots when I observed something I deemed particularly interesting.I have followed Nina since 2009 but the main period of study for this article started January 2010 and continued to the election day September 19 th 2010.Screenshots have been selected every week from all five social media platforms.All postings on Twitter, the VF blog and Ninalarsson.sewere saved for the studied period and later used for different content analyses.In this sense I used her social media platforms as archives of information (see Berg 2011, p. 126), but I have also created my own archive with weekly screenshots since data and interactions on social media are instantaneous and may be changed or disappear.
For better understanding the online interactions, I participated on some of the debates on Ninalarsson.seas well as twittered and facebooked with her.My interventions with Nina followed a simple plan; when I reacted on, or felt I wanted to get clarification, information or just agreeing on something she posted, I interacted with her (wrote on the wall, retweeted a tweet, commented on, or liked a posting et cetera).My interventions most often concerned statements on education policies and infrastructure (since I work at a university and used to commute between Karlstad and Stockholm).Examples of interactions on her Facebook was giving thumbs up when she posted her train was on time, or posting comments about my train delays and asking for her ideas of improving the railway tracks in Värmland.Our twittering mostly revolved around when to meet, or asking for quick information on liberal politics, or sending condolences when I tweeted I had a cold et cetera.
Netnography works well in combination with a more traditional ethnographic method, especially when inquiring into what considerations are behind (inter)actions online.Hence online observations and interventions were complemented with continuous and open ended interviews online and offline with both Nina Larsson herself and her coaches Olle Nilsson and Gunnar Bark at Hello Clarice.Besides online chatting and e-mailing shorter questions and requests for clarifications, I conducted one more structured interview with Olle Nilsson and Gunnar Bark offline and one with Nina Larsson.Interview data have also been used from David Kvicklund and David Samuelsson's (2010) essay for bachelor degree.Furthermore, I followed Nina offline during two weeks in September 2010 before the elections.Being close to her I often took the opportunity for posing questions and discussing her social media use (in the car to different events for example).
It is difficult to say whether conducting this study influenced her behavior online.On the one hand the mere fact that she was part of this research project indicates that she conceived of her campaign as novel and of interest.And since I continuously asked her questions, participated on her social media platforms and followed her offline, she was constantly reminded of my interest in her social media uses.On the other hand it was my impression that she was too independent and too focused on the campaign to let me distract or influence her.After a while I also became such a natural part of her network, to the extent that I believe she conceived of me more as an acquaintance among others rather than as a researcher.
The research question of how Nina used social media platforms to manage and negotiate the image of her as a politician in her election campaign will be answered by a close analysis of the empirical data gathered through interviews, content analyses, online and offline ethnography.Being inspired by results of previous studies from the US and UK on politicians imagemanagement, presented previously, I am particularly going to focus on expressions of selfpromotion, what parts of her private life she disclosed, how she presented herself, both as a person and a politician, who and what she linked too, what of her locations and her activities she shared, if she managed an image of herself as an insider or outsider.By attending to these aspects I can start outlining a broader image of how she wanted to portray of herself in her election campaign and how she used social media platforms for doing this.Having accounted for the background of image-management and election-campaigning in digital late modernity, the Swedish setting of a multi-party representative democracy and the Värmland constituency in which Nina campaigned, the methods used for data-gathering and questions guiding the analysis, I will now turn to the result section.I will attend to how she used the different social media platforms for image-management one by one before discussing some general themes across the different social media platforms in the conclusions.

4.1.
Image-Management on Ninalarsson.seNina Larsson used two blogs during her election campaign, one connected to her as a liberal party member, Ninalarsson.seand one tied to the second morning newspaper in the region (with a social democratic editorial) Värmlands Folkblad (VF).Ninalarsson.sehas been up and running since 2006, and is used to communicate ideas with a more apparent political angle.The months prior to the elections Nina posted approximately four posting a week (135 postings in total in 2010 up to the election day September 19th).On the blog there are links to Nina's Twitter and Facebook account, her campaign website as well as the VF blog.Ninalarsson.se is clearly linked to the liberal party by the choice of color -blue -and the using of the party symbol, a blue cornflower.
On Ninalarsson.secomments to her entries were made visible right away (most of the time).Nina defined a good posting on here as something that not everyone agrees upon and something that is "a little bit provocative" ii .Through coaching by Hello Clarice Nina confirmed that she had become more daring in her postings.Olle Nilsson at Hello Clarice explains that they gave Nina feedback on her postings on the two blogs in forms of thumps up or thumbs down with the purpose of getting her to mediate certain kinds of emotions and to get her to "think right" about online communication.When asking Olle what this thinking right implies, he explained that they coach her to become more personal, to dare to be more provocative and direct in the communication with her constituency.This underlines a very distinct strategy in her imagemanagement, to provoke attention through disagreement and personalization.
Was she successful in attracting attention on Ninalarsson.se?A closer study of the postings indicates that in terms of attracting/ provoking more comments from readers, she had not been particularly successful.Looking at the postings 30 days before election day ( 19th August til 19 th September) she attracted on average 1,1 comment per post.However, her postings in January 2010 attracted on average 3,8 comments.This is especially due to two postings, one addressing the pricing of electricity (21 comments) and one addressing the issue of wolf hunting (24 comments).In the post on the pricing of electricity, Nina did make some personal referencing to her own electricity bill and she ended the post by somewhat provocatively calling for more nuclear power.This suggests that the strategy outlined by Hello Clarice worked.However, she was equally personal and provocative in other postings that did not attract/ provoke an equal amount of comments.The explanation rather lies in the controversial nature of the topic, nuclear energy.This is further underlined when looking closer at the post on wolf hunting where she is very careful in how she phrases her argument and almost take a middle stance between proponents and opponents.Wolf hunting is an extremely sensitive issue in the Värmland region where there actually are wolfs (compared to the capital region).This sensitiveness was apparent in the nonpersonal/ provocative tone she used in this post.Still it was the one positing attracting the most comments during 2010.Hence, it seems it was rather the issues discussed in the postings, than the tone used, that attracted comments.
Studying the postings on Ninalarsson.seshe framed herself both as a politician and as a person.She was often referring to her role as a Parliamentarian representing the Liberal Party at the same time as she uploaded videos on the blog where she interviewed people she personally found interesting, be it researcher on public transport or a blogger she liked.However, she did not disclose a lot of personal information about her everyday life and personal activities outside of her job as a politician.Reading the blog you could find out a little bit about how much she traveled, what music she liked and sports she enjoyed watching.But these revelations were often linked to a political issue she wished to discuss.Similarly there were some references to her previous working life experience as an officer in the Swedish army when posting her opinion on the Swedish military presence in Afghanistan, as well as her experiences commuting to Stockholm, when discussing railway infrastructure.Hence, she was framing herself as a politician, but with personal experiences that she drew upon when discussing issues on the blog.
On Ninalarsson.seshe was rather portraying herself as an Parliament insider than an outsider, giving glimpses into her work in corridors of the Parliament and her trips going to meetings, seminars and conferences all over Sweden.She clearly negotiated and image as a liberal parliamentarian, but rooted in the Värmland region since she sometimes highlighted local events and sometimes framed political issues locally.Thus, a follower got a local värmland flavor when reading her blog.However, postings mentioning Värmland, or being framed in a local setting, were half as common as postings framed in a national discourse.Given that she was elected to the Parliament (and not the Municipal Council) this might not be so surprising.Her blog tagline though was "Your Värmland Liberal in the Parliament" clearly addressing the Värmland constituency as their representative.In conclusion, the local region was present in the blog by her framing her self as a Värmland representative, but when studying the issues she discussed here, she was rather managing an image as a national politician, voicing liberal opinions and attacking the opposition.
If we look at her linking practices, data from the Politometer (a widget that measured in-, and out-links to and from Swedish political blogs (see www.politometerns.se)confirmed that during the months up to the elections more than 90 percent of both links out of and links in to Ninalarsson.sewere to/ from other Liberal Party members.In more than 50 percent of the postings in Ninalarsson.se(68 of 135) I could identify links to other liberal politicians or Liberal Party websites.Traditional media was also important in her linking practice.Up to 56 percent of the postings (76 of 135) referred to media texts initially broadcasted offline (newspapers, TV or radio) by either commenting on, or spreading texts initiated in traditional offline news media.Examples are linking to debate articles in morning broadsheets that she wanted to recommend -often by fellow liberals -, and to her own media appearances in radio and television.In conclusion, the image Nina managed on Ninalarsson.sewas Nina -the liberal parliamentarian.Even though a personal tone, using a lot of "I" and revealing some personal information, the Liberal party played an important role in her image-management practices.Her personal revelations were often framed CC: Creative Commons License, 2012.
in a local discourse and she described herself as a Värmland liberal, but it was clearly an image as a national politician she managed on Ninalarsson.se.

Image-Management on the VF-blog
On VF, Nina had a blog since 2008.Her aim was one posting a day, however, postings here were rarer than on Ninalarsson.se,approximately two posting per week.During 2010 she posted 20 times in January but then less and less up with only four postings the election month of September (87 postings in total January 1 st to September 19 th ).Nina decided herself what to post on the VF-blog, the newspaper only provided the domain.On the VF-website there were also other blogs from other politicians and regional celebrities to follow.Nina stopped using this blog after the elections.On October 11 th 2010 she posted for the last time.
Image-management through disclosures of more private and personal character was more apparent on the VF blog than on Ninalarsson.se.The postings here were more personal, dealing with feelings about her life and job as a politician, and here she also disclosed selected parts of her private life here, such as posting pictures of her renewing her apartment and revealing her feelings of being given the opportunity to work as a national politician together with a blurry picture of the Parliament building (Kvicklund and Samuelsson 2010).Another example is a video of her laughing in the back seat of a hot rod with local greasers in what was obviously an attempt to recruit new liberal voters.She also posted a video of her at the McDonalds drive-in ordering a Big Mac.Yet another example is a mix all the interviews with locals (posted on Ninalarsson.se)with her saying "hello" to camera in 10 different settings."Hello hello" would later become a hallmark in her election campaign used in her TV commercial among other things.According to Gunnar Bark at Hello Clarice the strategy for the VF-blog was for the reader to get to know Nina on a more personal level.The VF blog reader should more easily embrace Nina as person, through shorter postings and more pictures and videos.Studying Nina's postings on VF this strategy was apparent.Approximately half of the postings (43 out of 87) contained information primarily about her as a person -even though she often ended a posting with a link to Ninalarsson.seor an opinion piece in a newspaper.One third (32 out of 87) posting had pictures or videos that Nina herself had recorded and often acted in front of the camera interviewing someone she liked or reporting from an event such as a car race or a meeting with a local interest group.Compared to Ninalarsson.se the postings on the VF-blog were more framed in a local setting, more specifically Nina herself in a local Värmland setting, walking the streets of Karlstad (the region capital) in her Liberal campaign jacket or on her way to the local car race.In interviews Nina herself defined a good posting on the VF-blog if she succeeded with a good picture of an exciting meeting, or an entertaining story (Kvicklund and Samuelsson 2010).Hence, while Ninalarsson.sewas very much used to negotiate an image of her as a liberal parliamentarian -voicing opinions and attacking the opposition -on the VF-blog, the image she managed was as a happy person with personal interests, humor and as rooted in the Värmland region.It might not be surprising that the local angle were more prominent here since VF is a regional newspaper.
Nina herself said in an interview that the image she wanted to transmit on the VF-blog was as engaged and hardworking, both at work and at home (Kvicklund and Samuelsson 2010).This indicates that she carefully managed her personal traits on the blog for a political purpose.In another interview she said that she wanted to combat an image of politicians as "lazy and going with cabs everywhere".By showing for example pictures of herself renewing her apartment, posting comments on her lack of free time and late arrivals back from Stockholm (always with a smile though), she was managing an image of her as someone with a personality fit for politics.What other kinds of personal info and activities did she disclose?By reading the VF blog there were many references to her lunch and dinner choices, often with recommendations and pictures.When Nina baked she let her followers know about this as well as her opinions about sports-and TVpersonalities.Nina also blogged about being stuck in front of certain TV shows, she blogged when her favorite cartoon was published and she was referring several times to the car race.In contrast to the US (Stanyer 2008) the family was absent in her postings.Asking Nina about this she said it was a conscious choice to leave her partner and family out of her political career.She even hesitated to post to which town she was going to when spending Easter holidays with her parents because of a wish to keep her family out of her political life.This suggests that personalization tendencies are cultural specific.Where the family in the US is important for managing an image as fit and reliable for politics, in Sweden being hardworking, fun and having personal interests outside of politics (cartoons, care race, baking) was more important.
Concerning comments to the blog postings, these were even rarer on the VF-blog than on Ninalarsson.se.If we look at her linking practices it shows similar tendencies as on Ninalarsson.se,even though not as accentuated.Up to 35 percent of the postings referred to media texts initially broadcasted offline (newspapers, TV or radio) and in 32 percent of the postings, I could identify links to other liberal politicians or liberal party websites.Above all Nina linked to herself on the VF blog by directing traffic to ninalarsson.seand to her campaign site.As a reader you got the impression that being personal and fun was not enough, she also wanted to link to a more serious opinion piece in the end of her more fun and personal VF blogposts.Hence, being a liberal politician was important in the management of the more personal self, and as a follower of her VFblog you were constantly reminded that it was an election year.

Image-Management on Her Campaign Website
The campaign website was launched May 26 th 2010 and was active up to the election day September 19 th .Here visitors could practice in voting for Nina through an interactive application (she collected 28000 votes on her campaign website before election day), donate money, pose her questions, reading about her political agenda, watch and listen to campaign materials.The four main links were to the four profile issues/ matters of heart, she based her campaign around.These were better schools, smarter transports, a successful integration of immigrants and making it easier for companies and businesses.There were links to her Facebook profile on the front page, and further inside the site there were also links to Ninalarsson.seand her Twitter feed.The visual design was centered around the cornflower, the liberal party symbol, and one of the main headlines read "vote for the liberals and then tick my name".Hence no-one could mistake that Nina was a liberal politician first and foremost.The campaign website was completely controlled by Nina and Hello Clarice.They decided which questions should be answered to and be made visible on the website, approximately one per day.This makes it questionable whether we could even conceive of this as a social media platform at all.On the other hand this strict orchestration makes the preferred image very apparent here.
On the campaign website, Nina and Hello Clarice were clearly managing an image of Nina as professional politician.She was very politician-like in her answers to questions selected, thanking the author for posing an important question and then systematically explain how she would work for these issues if re-elected to the Parliament.Compared to the VF-blog there were no humor and not any personal disclosures at all, except for her calendar that disclosed that she took morning walks every day.Hardly any questions dealt with local issues, even though there were postings praising Nina for appearances on local school debates and alike.Most questions that were allowed on the campaign website revolved around her four core issues.There were no links apart from to Ninalarsson.seher Twitter feed and her campaign commercial.Here, she was clearly managing an image as a liberal national politician, an insider that can get things done in the Parliament.

Image-management on Twitter
Nina Larsson uses Twitter since spring 2009.The months up to the elections Nina used Twitter almost every day and several times every day, except for breaks during Easter, during some summer weeks and also during weekends when she did not tweet at all, or tweeted considerably less.This indicates that Twitter was primarily used during office hours and thus more connected to work than leisure.In total Nina tweeted 722 times from January 1 st til election day September 19 th .The month of January 2010 she tweeted 139 times but then -as with the VF blog -she tweeted less and less as she was increasingly busy with more traditional offline campaign activities the closer she got to election day.The last 30 days before the elections she only tweeted 74 times.On Twitter, Nina only had one link to her campaign website.Nina mostly tweeted updates on her whereabouts and plans such as going to conferences and seminars as well as tweeting from them.Situations, such as being delayed on the train, spotting celebrities or having an unpleasant phone call were also tweeted about.
Compared to the two blogs and to the campaign website, Twitter was used more for networking.According to Jackson and Lilleker's study (2011, p. 87) most of UK politicians tweets were one way rather than two-way, indicating politicians rather talking about themselves than engaging in dialogues.Nina no doubt talked about herself and her opinions in her tweets.Especially the months prior to the elections she used Twitter for stating political opinions and to direct traffic to Ninalarsson.seby linking to her posts in her tweets, as well as to opinion pieces suitable for her political agenda (as with the VF-blog).However, her Twitter use was different from the UK politicians in the Jackson and Lilleker study because 60 percent of her tweets were addressed to someone with the help of @tags.Reading Nina's Twitter feed it becomes apparent that networking and broadcasting are not mutually exclusive.For example, Nina could very well broadcast her attending a seminar and @ddressing local journalists and interested friends to get their attention to this seminar in the same tweet.Similarly she used hashtags # to frame her opinion tweets in a particular discursive setting, for example using a seminar hashtag to direct her opinions to other participants in this seminar.Nina used hashtags in almost 20 percent of her tweets (138 of 722) and she used 54 different hashtags in total, mostly hashtags referring to conferences, debates and seminars she attended as well as to the Liberal Party.She also used hashtags to participate in live commenting on TV shows and sport events.
Hardly any new voters were part of Nina's Twitter network.Reading her Twitter feed, the discussions rather seemed to take place among friends and party colleagues than with potential new voters.She was for example replying friends' tweets addressed to her on her plans and whereabouts, and she was also commending work and mass media appearances of fellow liberal party colleagues.The Liberal Party collective was very present in her tweets.It could clearly be observed how she was addressing particular liberals and hence tying the image of herself to some party colleagues rather than others.Her Twitter feed could very well be used to study inner party hierarchies and positionings.Hence, while Twitter was used for networking, it did not exclude broadcasting practices and strategic intra-party positioning.
On Twitter, Nina rather managed an image of herself as a person situated in a Liberal Party network and positioning herself in this network through using @tags and hashtags.This was not an outsider position she was managing.On the contrary, her twitter-feed provided followers with glimpses -not only into Nina's schedule, participations in seminars and conferences -but also into the inside of the Liberal Party.However, the political image she was managing on Twitter, was in contrast to the political image she was managing on the campaign website.Her campaign website was used as a front stage (Goffman 1959) where you as visitor you got the impression Nina addressed an imaginary audience of voters.On Twitter though, as a follower you got the impression that Nina addressed an insider network of friends and party colleagues.However, her Twitter feed was open for the public and journalists to lurk, something Nina said she was aware of.Hence, her Twitter-feed cannot be understood as a back stage (Goffman 1959) rather as something in-between.
Nina did disclose a lot of personal information on Twitter.Reading her feed, followers could find out what music she listened to, that she enjoyed eating hamburgers, especially those at McDonalds, where she also sometimes had coffee.The recurrent theme of McDonalds can be traced back to political statements Nina had made about McDonalds, its importance for young people getting their first job et cetera.Hence, having followed Nina for some time, I understood her personal disclosures of eating at McDonalds as politically motivated.Other examples of personal disclosures are her morning walks -how long she walked and who she walked with -and that she sometimes had to shovel snow to get her car out in the morning.She also mentioned which sports personalities she liked and disliked, that she enjoyed car races a lot, and she made references to her background as an officer among other things.Following Nina on Twitter you also got a quite detailed picture of her media consumption by attending to which hashtags she used as well as tweets about what she programs she watched/ listened to and what newspapers and magazines she liked to read.In one tweet she got really personal, stating that she enjoyed silence, something that was difficult for her to enjoy at home with her partner around.Here she did mention her partner even though she said in an interview that she consciously wanted to exclude him out of her social media practices.This was the only time during the observations though.In conclusion, on Twitter we find political statements mixed with personal information which sometimes are not easy to separate.This all seemed to be done in a friendly and party political atmosphere, indicating that the Liberal network was not only important for Nina professionally, but also personally.
Nina also linked on Twitter.17 percent of her tweets (123 of 722) contained links.43 percent of these were links to Ninalarsson.se(53 of 123).Only one time did she link to her campaign website and two times to the VF-blog.Apart from links to her social media platforms, there were links to traditional established broadcast media.In 26 percent of the links (32 of 123) she linked to traditional offline media, mostly to broadsheets (both national and regional) and tabloids, but also to radio and TV -most often to shows/ debates she herself was participating in.Apart from linking to herself and traditional broadcast media, there were 10 links to the liberal party or other liberal politicians' social media platforms.You could also find isolated links to web specific sites such as blogs, Spotify, Bambuser and YouTube.Hence through her linking practices on Twitter I can conclude that she was managing an image of herself as an active liberal politician, directing traffic to her blog Ninalarsson.seand to established broadcast media, most often when she or other selected liberals appeared there.

Image-management on Nina's Facebook Profile Page
On her Facebook profile page there were links to Ninalarsson.se,her Twitter account, her campaign website and her personal page on the Liberal Party website.As with Twitter, Facebook was used for personal networking but also for stating opinions and linking to opinion pieces, TVdebates, to Ninalarsson.seand to other liberal politicians' social media platforms.This underlines a mix of personal/ intra-party networking and campaigning.However it was not always evident if she was managing an image in front of her liberal party colleagues or in front potential voters.Activities on her Facebook profile page largely consisted of thumbs up and "likings" from friends and fellow party comrades.Her network was made more visible on her Facebook profile page, both image wise, but also since interactions in their entirety were made visible for followers to read, compared to on Twitter where followers only got to read Nina's answers to questions.
Following Nina on Facebook, it became apparent that the ones that posted on her wall and the ones she engaged in discussions with, were friends from before, party colleagues and fans rather than constituents, potential new voters or opponents.As I have underlined previously, political participation on social media platforms is not deliberation (see Svensson 2012), it is more often image management (see Svensson 2011a;2013).But what image did she manage on her Facebook profile page?As with Twitter, on Facebook Nina negotiated an image of herself both as a politician and a person.Studying the groups/pages she liked and displayed on her profile page, this becomes apparent.There were an array of local Liberal Party Facebook groups/ pages from different Swedish cities she had joined and were displayed her profile page as well as more general Liberal Party Facebook group/pages -celebrating the party leader and other leading liberals.But she also displayed her membership in Facebook groups/pages of a less political character, dedicated to preferences in soft drinks, music, cartoons and sports.The kind of personal information she disclosed is thus very much in line with what she disclosed on her Twitter feed.The same goes for her linking practices with links to mass media broadcasts she had been participating in, or opinion pieces by other liberals in newspapers, as well as links to Ninalarsson.se.
The image emanating from her Facebook profile page was dominated by encouraging postings from friends and acquaintances showing their support, a support that was made visible for her Facebook friends and others (since she had an open profile).As a Facebook friends you sometimes got the feeling of lurking into a semi-back stage party-political network, and sometimes the other way around since she also used her profile page as a front stage to broadcast own media appearances and that of other liberal politicians.

Conclusion
Having gone through all the different social media platforms Nina used in her election campaign 2010, what can be concluded about the image she wanted to portray of herself and how she used social media platforms for this purpose?It is apparent that she followed a strategy for her social media practices, practices that to some extent were professionalized since she was coached by the communication agency Hello Clarice.As we have seen, different social media platforms were used differently.Ninalarsson.sewas at the centre of her social media practices with almost all her postings there broadcasted in tweets, Facebook postings and on the VF blog with a link redirecting the readers/ followers to Ninalarsson.se(but not the other way around).Ninalarsson.sewas largely used to negotiate and image as a liberal parliamentarian -making political statements and providing her views on current news stories as well as pushing issues in her political field.The VF-blog was used more to manage an image of her as a fun, personal, but still a hardworking politician, deeply rooted in the Värmland region and hence apt for representing their voters.The tightly controlled campaign website was used to present Nina as a professional liberal politician, an insider parliamentarian, explaining for the voters what agenda she would pursue if re-elected to the Parliament.Twitter and Facebook was more used for personal and intra-party networking, and for directing traffic to Ninalarsson.se.
In spite of these very different uses of her different social media platforms, some common themes can be discerned.Her social media platforms were clearly used to control her political image, to amplify selected text, texts of hers and other liberals that often originated in offline/broadcast media.Her networking and social circle online also largely consisted of friends with similar political preferences and fellow liberal politicians suggesting a use of social media platforms rather to negotiate a position within the Liberal party than to deliberate with potential new voters.
When interviewing Nina she underlined several times the possibility for her to put forward "her own version in her own media channels".Social media platforms gave Nina a possibility to use other channels she has greater control over than established media channels.As Stanyer (2008, p. 418) states, these additional channels allow representatives to disclose a large amount of detailed information which journalists may not consider newsworthy.These are also channels where the politicians have editorial control and can determine the nature of that information and how long it should be available to the public (ibid.).Hence, contrary to Sey and Castells (2004, p. 366), who claim that it is more difficult for politicians to control the information flow on social media, this is precisely the reason why Nina used social media platforms in her 2010 re-election-campaign (see also Zafiropoulos and Vrana 2009).An example of this is in a posting on Ninalarsson.sewhere she commented on an investigative journalistic TV program scrutinizing the presence of Members of Parliament during voting in the chamber, in which she was portrayed as a politician being absent from many parliamentary votes (Kvicklund and Samuelsson, 2010).This underlines the usage of social media platforms for the purpose of monitoring and controlling her political persona, to control the image that was broadcasted of her in established media.
Traditional news media seemed to be at the centre of Nina's social media practices.She used her social media platforms to rather position herself in relation to traditional news media text, whose stories and angles she has no influence over, than to circumvent them all together.As mentioned before up to 56 percent of the postings on Ninalarsson.se-and 35 percent of the postings on the VF-blog -referred to media texts initially broadcasted offline.It thus seemed that established media channels to a large extent were setting the agenda for her social media practices, what she would discuss and link to.
Nina clearly used her social media platforms to amplify selected texts and news items, or trumpeting as Stanyer (2008) labels it.If Nina for example wrote a debate article in a newspaper, or appeared on a TV or radio show, this would almost automatically generate a blog posting, a Facebook posting and a tweet -all with links to the original appearance.This resonates with Goldbeck et al. ( 2010) content analysis of members of US Congress using Twitter, who found that a majority of the tweets were information such as position on an issue, opinions, facts, and a vast majority of these informational tweets contained links to other mass media texts.
Nina's interactions with regional news journalists is worth a mention here.They are part of each others Facebook-and Twitter-networks and there are numerous examples of connections between them.She sometimes addressed them, especially on Twitter with @tags when for example visiting a company or a school, to get them to come and cover the event.In 4 percent of all her tweets I could identify @tags to traditional broadcast media and journalists.Journalists also contacted Nina directly on Twitter a couple of times during the two weeks I followed her offline, to get her quickly to the radio studio to comment on an event, to ask for clarifications on liberal policies or commenting on an opponents proposals.The high circulation of broadcast news media texts and the importance of journalists in her social media networks suggest that appearances in offline media and texts originating in traditional mass media were of prime importance in the CC: Creative Commons License, 2012.negotiation of her political image.Also the liberal party played a central role in here imagemanagement as I will attend to next.
We can understand the practices of amplifying and monitoring own and others media performances in light of increased possibilities to control the own political image that social media platforms afford.Studying Nina it becomes obvious that image-management revolved around negotiating, and positioning herself as an important liberal politician.Nina clearly used her social media platforms to promote the Liberal Party, to reinforce, retweet, and "like" political messages that other liberals have been communicating both in traditional offline media, as well as on their own social media platforms.As mentioned previously, 90 percent of both links in to, and out of Ninalarsson.sewere from/to other liberal party members and in a majority of the postings on Ninalarsson.seand one third of the VF blog links to other liberal politicians or liberal party websites could be identified.Hence, a virtual patting the backs of fellow liberal party comrades seemed to take place on Nina's social media platforms in form of multiplying and commenting each other's appearances.This result resonates in Gustafsson's (2012) study where he found that party members mostly obtained a Facebook account in order to discuss politics with other party members and to receive information from the party.It is perhaps not surprising that the political party is important for candidates' image-management in a multi-party democracy.But the extent of this was somewhat surprising studying Nina's social media practices before the 2010 elections.It has been common to frame political participation on social media platforms in discourses of deliberative democracy (see Coleman and Blumler 2009), that politicians could/ should use these platforms to build bonds with voters.This was also referred to in interviews with both Hello Clarice and Nina herself when ask why they used social media platforms in her election campaign.However observations show that this was not primarily what took place on Nina's social media platforms (with the exception of the campaign website, which could be argued is not so much a social media platform given the low level of free interaction).This study clearly shows that it was rather in front of the party and fellow party colleagues she was managing her image than in front of her constituents.Some argue that social media platforms will involve more and more citizens in an ongoing debate about political issues in many areas, not only during the weeks the campaign lasts (Turiera-Puigbò, 2009, p. 18;Gerodimos and Ward, 2007, p. 119).Certainly Nina's social media practices indicate a trend towards the constant campaign.But perhaps even more, her social media practices indicates an ongoing and constant identity negotiation.In digital late modernity, it is through the use of social media platforms that our identity will be expressed, negotiated and manifested.Studying Nina's use of social media platforms for image-management in her electioncampaign the often referred to trend of a personalization of politics could clearly be observed.However in order for Nina to be the politician she wanted to be, to negotiate the political image she wanted, she needed the Liberal party and her party colleagues.She used them as resources for managing and confirming her image of her political self.These connections, these nodes in her network, seemed to be of particular importance in the management of her political image on social media platforms.
So what happened to Nina after the elections?Nina was not immediately re-elected by the Värmland voters, but eventually she got to keep her seat in the Parliament due to a so-called adjustment mandate (my translation; utjämningsmandat) that was awarded the Liberal Party in the Värmland constituency.After the elections she was also promoted to the Liberal's Party Secretary since her predecessor was appointed Minister for Integration in the new government.If this promotion is due to her networking skills on social media platforms remains to be confirmed.