NL 15 pp.

15–32 Intellect Limited 2017

Northern Lights
Volume 15
© 2017 Intellect Ltd Article. English language. doi: 10.1386/nl.15.15_1

Mattias Ekman
Örebro University

Andreas Widholm
Stockholm University

Political communication in
an age of visual connectivity:
Exploring Instagram
practices among Swedish
politicians

Abstract Keywords
This article explores the specific features of Instagram as a platform for visual politi- Instagram
cal communication. Drawing on theories of mediatization and celebrity politics, it celebrity politics
analyses how various forms of symbolic connectivity are expressed and performed by social media
sixteen leading politicians in Sweden, and moreover how their social media use relates political
to news media. The study leans on a content analysis (n=800) and results show that communication
journalism still holds a strong symbolic value, even when politicians are in charge journalism
of the political discourse. In addition, it reveals how the platform logic of Instagram Sweden
contributes to the formation of digital lifestyle politics, where symbolic connec-
tions between politicians and a variety of actors are staged through new mediatized
relations. Visual political communication does inherit a democratic and interactive
potential. However, according to the analysed data, most politicians avoid public
interaction. Instead, they are preoccupied with the branding of their public persona.

15
Mattias Ekman | Andreas Widholm

Introduction
Current transformations in the media industry towards digital production
and distribution have had severe consequences for media markets across
the globe, sparking discussions of a forthcoming ‘death of journalism’ (e.g.
McChesney 2013: 172ff). In Sweden, two of the largest newspaper compa-
nies are in such a bad condition that bankruptcy, devastating rationalization
of the work force and structural conversion from paid quality newspapers to
free dailies have been suggested as probable solutions for the future. In a criti-
cal comment on this crisis, cultural journalist and editor Björn Wiman (2016)
wrote in Sweden’s most influential newspaper, Dagens Nyheter, that Alice Bah
Kuhnke, minister of culture and democracy and responsible for the current
government’s media policy, ‘should log out from Instagram and take a firm
grip on [Swedish] media policy’. Wiman’s comment is an indication not only
of a newspaper industry in crisis, but also of a structural shift in which new
actors challenge journalism’s centrality in the democratic discourse. For exam-
ple, politicians are no longer dependent on the traditional interaction with
news media, but can instead circumvent them and communicate through
their own social media accounts. The image-sharing application Instagram
has grown fast over the past years and has, up to late 2016, 600 million users,
yet it is still an emerging platform when it comes to political communication.
In this article, we explore what we previously have defined as a perform-
ative turn in online political communication, arguing that politicians’ self-
imagery on social media platforms is part of a broader process of political
‘celebritization’ where ‘performed’ connectivity plays a substantial role
(Ekman and Widholm 2014). Visual political communication on social media
platforms merges the personal and the professional realm of politicians’ lives
through specific modes of addressing the audience (cf. van Zoonen et al. 2010).
These new expressions of celebrity politics also emanate from broader societal
changes related to the current era of mass-self communication (Castells 2013).
Drawing on van Dijck’s (2012: 5) argument, that the dominance of commer-
cial social media platforms has transformed online communication from a
‘participatory culture’ into a ‘culture of connectivity’, we argue that this also
has implications for contemporary political communication. Since politicians
are using the same commercial social media platforms (e.g. Facebook, Twitter,
Instagram, etc.) as everybody else in their daily communication with citizens,
political communication practices are partly structured by the socio-technical
affordances of these particular platforms. Moreover, since contemporary social
media communication is centred on the ‘connectedness’ with other users and
publics, connectivity also comes to the forefront in political communication.
A considerable amount of scholarly work has been devoted to textual
political communication, particularly on Twitter, whereas visual communica-
tion practices are still highly understudied. This article draws on the conten-
tion that visual political communication needs to be assessed both from the
communalities of contemporary social media communication, and from the
particular affordances of different platforms. The aim of this study is therefore
to explore the platform-specific features that Instagram has brought to politi-
cal communication in Sweden, looking specifically at: the interrelationship
between self-imagery and symbolic connectivity; how political communica-
tion on Instagram relates to news media; and differences in strategies mainly
between political parties and individual politicians. The article is structured
as follows. We start with the theoretical framework that tries to interlace two

16   Northern Lights

 Political communication in an age of visual connectivity

strands of research that are central to contemporary political communication:

mediatization (with a particular focus on the spheres of politics and the news
media) and celebrity politics (particularly its connective aspects). The subse-
quent section is devoted to methodology, which is followed by a content analy-
sis of a representative sample (n=800) of images derived from the Instagram
accounts of sixteen politicians in Sweden, two from each party represented in
the parliament and who have a prominent position on the platform. The arti-
cle ends with a concluding discussion including suggestions for future studies
of visual political communication.

Politics–media relations
A growing trend in political communication research is to relate changes
in information activities of political actors to the increasingly central role of
mediatization in contemporary societies. As an analytical concept, mediati-
zation has been a term that underlines important shifts in the relationship
between the spheres of politics and the media (and especially political actors’
relationship with news journalism). A common view has been to understand
media and politics as constituted by different logics, and that media logic
gradually has become increasingly dominant in relation to what has been
described as political logic or sometimes more specifically party logic (cf.
Strömbäck 2008, 2011). When media logic reigns, political actors need to
adapt their strategies so that they fit the modus operandi of the mass media,
notably the narrative conventions, professional expectations and techno-
logical and distributional features that have characterized large news media
institutions (Esser 2013; Asp 2014). According to Kunelius and Reunanen,
this view largely stems from ‘the imaginaries of the mass media era’, where
strong journalism institutions functioned as ‘the main proxy for public opin-
ion, both for the politicians and the public itself’ (2016: 2). Contemporary
political communication takes place instead in a new media environment
of radical structural changes, characterized by an apparent crisis of journal-
ism (Zelizer 2015; Blumler 2010) including worrying forecasts of an inevi-
table death of the printed press (Brock 2013; Siles and Boczkowski 2012).
In essence, this crisis has its roots in the changing conditions for produc-
tion and dissemination of information that came in conjunction with the
development of the Internet, challenging the dominance of traditional news
media institutions in favour of not just new media actors, but actors from
practically all sectors of society, politicians included. While what we might
call ‘forced’ adaption to the logic of the news media can be taken as the
foremost consequence of the initial phase of mediatization, we now witness
the advent of a new phase where this adaption becomes harder and harder
to make a case for (cf. Kunelius and Reunanen 2016; Ekman and Widholm
2015). Today, citizens, organizations, corporations as well as politicians take
advantage of the new communicative possibilities enabled by the fragmented
landscape of social media. It is no longer a given that media logic should
be thought of as something pertaining mainly to the professional culture of
the news media, since journalism has received intensive competition from
alternative information producers, which may be governed by completely
different ideals and working practices. Hence, journalism’s traditional role
as gatekeeper is in steady decline (Bro and Wallberg 2015) and its central
role in the democratic discourse is moreover challenged due to a growing
dependence on external distribution platforms for the news (cf. Lasorsa et al.

www.intellectbooks.com   17
Mattias Ekman | Andreas Widholm

2012). Against such a background, it has been argued that even journalism
has been ‘mediatized’ in the sense that its production processes, distribu-
tion techniques and economic models have been redesigned to cope with
new ‘social’, ‘networked’ or ‘connective’ media logics (van Dijck and Poell
2013; Klinger and Svensson 2014; Kammer 2013). This is particularly evident
in the increasingly desperate struggle to obtain and increase user traffic to
online news media outlets – resulting in what have popularly been defined
as the ‘click economy’ of news.
News journalism is still a central source of information that people use
when informing themselves about political issues (Strömbäck 2011), but
political actors are no longer completely dependent on the news media
in order to reach out, as many of them now have become influential and
powerful media producers themselves (Chadwick 2013; Papacharissi 2015).
Donald Trump’s 2016 presidential election campaign in the United States
is a powerful example of how politics is battled not only within or through
journalism, but against it; a situation that had been unthinkable without
social media as a central communication tool for Trump’s campaign admin-
istration. Swedish and Scandinavian politics is less personalized compared
to the United States, but social media play an important role in political
parties’ as well as individual politicians’ public communication (cf. Klinger
and Svensson 2014; Larsson 2015). Interaction between journalists and poli-
ticians does not only take the form of published interviews in newspapers
or in newscasts on television. On the contrary, this dialogue has been an
increasingly open process online, as journalists and politicians interact in
online spaces, forming a new type of mediatized interdependency (Ekman
and Widholm 2015). Journalists and politicians are also undergoing simi-
lar transformations with regard to personalization, as the private dimension
of their personalities often intertwines with the professional dimension in
their public persona. Social media centre public attention to the personal
aspects of the politician and reciprocally advance personalization in the
communicative features of the individual politician. Thus, the underlying
practices of politicians’ social media communication need to be addressed
with a particular focus on how political communication and popular and
celebrity culture merge in social media platforms. Instagram, which is the
object of study here, accentuates this development in very specific ways.
As a platform, Instagram constitutes a digital space of ‘selfies’ (along many
other image types) where people’s self-branding snapshots of their daily
lives are strongly interlaced with senses of immediacy, mobility and inti-
macy. Differently put, Instagram accentuates a constant delivery of self-
centred visual imagery, offering symbolic access to other people’s personal
and professional lives (Schroeder 2016). Although Instagram allows for text
as well as images and video, the platform leans on what we might call an
‘image first’ logic since it – in contrast to other social media platforms –
presupposes that the users communicate visually while writing text is an
option. Since the platform exhorts a strong form of individualism, Instagram
users have also proved to be highly performative in their self-display, staging
their personalities for immediate public consumption in the continuous race
for likes and comments. Hence, Instagram provides a culturally and techno-
logically specific context for communication (Hjorth and Hendry 2015) and
the central question here is not whether the platform steers political commu-
nication towards personalization, but how and to what extent personaliza-
tion occurs among different parties and actors.

18   Northern Lights

 Political communication in an age of visual connectivity

Porous and ‘celebritized’ politics online

As stated in the foregoing section, the digital transformation of the media land-
scape has radically changed the relationship between the spheres of politics
and the news media as well as the techniques and modes of political commu-
nication. Already during the early stage of this development, Blumler and
Gurevitch called for a fundamental rethinking of the study of political commu-
nication because of the ‘complex, volatile and chaotic’ aspects surrounding
the new transitional media environment (2005: 105). They specifically high-
lighted the ‘embryonic’ modes of political communication on the Internet,
but also what they coined ‘porous politics’ as important themes for future
research. The concept of porous politics relates to the entry of popular culture
into politics in the form of television talk shows and online political entertain-
ment, etc., where politics ‘break out of the shells of respect, deference and
distance from people’s daily lives in which had formerly been enclosed’ (2005:
113). Central to porous politics is that the raw materials for political discus-
sions become fragmented and politicians increasingly judged with regard
to personal aspects including, among other factors, social life, sexuality and
family relations. Blumler and Gurevitch (2005) distinguished between ‘infinite
politics’ where practically everything can be included in political discourse,
not least entertainment, and a more traditional form of ‘bounded’ politics
which takes place in and through conventional state institutions and the news
media. Without doubt, porous or infinite politics are core elements of politi-
cal communication online, even when individual politicians are in charge of
constructing their own public persona. Online spaces have been increasingly
central for political performativity and identity making, and politicians lean on
similar celebrity management strategies as those found among entertainment
actors in the global media industry (Ekman and Widholm 2017; Marwick and
Boyd 2011). This reflects that the mediatization of politics has been intensified
with regards to personalization and media adaption, but also that this adap-
tion to an increasing extent is steered towards platforms and formats beyond
the traditional news media.
The fact that popular culture has been an integrative part of political
communication has recently been problematized in fruitful ways in terms of
theories of celebrity politics (Street 2010; van Zoonen 2005; Wheeler 2013).
Digital celebrity politics can be seen as a specific trajectory of mediatiza-
tion; it underlines how politicians lean on the styles and aesthetic conven-
tions of being well known, through performed connectivity online (Ekman
and Widholm 2014). When politicians enter the sphere of celebrity, they are
‘coming to terms with the media age and consumer culture, attempting to
personalize or brand their leadership, and are constantly adapting their politi-
cal communication strategy to communicate through evolving media’ and
thereby they are ‘celebritisizing’ their personal image (Marsh et al. 2010:
325). Moreover, as van Zoonen (2005) argues, celebrity politics is tainted by
particular gendered norms – that vary depending on political and societal
contexts. On the one hand, female politicians traditionally have had ‘family
life’ as a frame of political capital, and in tandem with celebrity culture their
other options have been ‘predicated on appearance, sexuality and – most
recently – continuous experimentation with styles of femininity for the perfor-
mance of a convincing female political persona’ (2005: 95). On the other
hand, research on international female political leaders shows that ‘female
heads of state and female politicians in general, more extremely than before,

www.intellectbooks.com   19
Mattias Ekman | Andreas Widholm

1. Information on are “others” to dominant images of femininity while remaining “others” in

followers for Aron
Modig was fetched
the political sphere, due to their minority position’ (van Zoonen 2006: 297).
from Instagram on This implies that, to be accepted, successful female politicians avoid associa-
19 January 2017 due to tion with hegemonic ideas of femininity (i.e. ‘the hyperfemininity of current
technical problems in
the original data set. celebrity culture and post-feminism, with fashion, sexuality, glamour and
Modig’s numbers were consumption as core’) (van Zoonen 2006: 297). An interesting question is thus
probably slightly lower how female politicians navigate the ruling gendered norm on platforms such
in June 2016 when the
Instagram posts were as Instagram (adhering to the aforementioned ‘hyperfeminity’). Considering
collected. the relatively advanced gender equality in the context of Sweden, the ques-
tion is if the visual communication of female (and male) politicians mirrors a
more diverse gendered landscape of political identities, or if it adheres to the
gendered duality expressed in previous research on female politicians.

Methodology
The analysis is based on a content analysis (Neuendorf 2002) of Instagram
postings from sixteen Swedish politicians active on a national level. To assess a
broad range of political actors, yet at the same time politicians that are prolific
on a national level, we selected two politicians each from the eight parties
represented in the Swedish parliament. Four of the politicians are ministers of
the current government and the remaining twelve are members of parliament.
Three criteria were applied in the selection of politicians: (1) we only selected
open accounts accessible to anyone with Internet access; (2) we selected one
woman and one man from each party; and (3) we selected the politicians with
most followers (in their respective party) on Instagram. The final criterion was
related to both social media impact and network centrality. Politicians with a
large number of followers are more likely to reach out to both the public and
to other media actors. Politicians with many followers are also more likely

Name Party Political pos. Posts Followers

Rosanna Dinamarca Left Party MP 638 7337

Jens Holm Left Party MP 426 552
Ylva Johansson Social Democratic Party MG 920 2364
Gabriel Wikström Social Democratic Party MG 154 3673
Alice Bah Kuhnke Green Party MG 419 14300
Gustav Fridolin Green Party MG 380 9076
Birgitta Ohlsson Liberals MP 3484 5666
Fredrik Malm Liberals MP 531 1406
Annie Lööf Center Party MP 277 11300
Emil Källström Center Party MP 272 1317
Ebba Busch Thor Christian Democrats MP 369 9000
Aron Modig Christian Democrats MP 830 930
Anna Kinberg Batra Moderate Party MP 462 10600
Johan Forsell Moderate Party MP 611 1077
Paula Bieler Sweden Democrats MP 418 794
Linus Bylund Sweden Democrats MP 1995 862

Table 1: Politicians, number of posts and followers (160628).1

20   Northern Lights

 Political communication in an age of visual connectivity

located in the centre of (political) social media networks, i.e. they attract more
user traffic, comments and likes compared to politicians with less followers.
Since we are interested in the broader characteristics of the politicians’
communication rather than that of a specific time period, the analysis draws
on a systematic random sample of all images posted on Instagram, mean-
ing that the time frame differs between the politicians depending on for how
long and how often they use the platform. The sampling process began with a
random start between 1 and x. First a skip interval was established, calculated
from the known population of images/posts in each Instagram account. The
sample size was set to 50 units from each account regardless of the size of
the population in each account. Thus, when calculating the skip interval, we
used the following model: N/n=N/50=y. So, if the total amount of posts in an
Instagram-account is 400, then 400(N)/50(n)=8, so here we would select every
eighth unit, beginning randomly within the first eight images (Neuendorf
2002: 85). The systematic random sampling ensures a data sample ranging
from the latest uploads to the start-up of the accounts. The final sample (n)
is made up of 800 Instagram posts (image and text) of a total of 12,186 (N)
posted images on the sixteen accounts. Our data set thereby contains 6.6 per
cent of the total population. The data sample was collected and stored through
screenshots of the image and the textual features (including comments).
The coding unit consists of each uploaded image/video including the
textual description made by the account holder. We coded the amount of
comments for each unit and the presence/absence of textual interaction (i.e. if
the politicians responded to comments). Moreover, we coded for the amount
of likes for each post and the number of persons present in the image. Since
the socio-technical affordances of social media platforms such as Instagram
(tend to) merge the professional and the private life of users (cf. van Dijck
2012; Papacharissi 2015), we coded for the presence of political message (if the
unit contains a political message it could be seen a part of a political communi-
cation strategy), and what political–spatial level the political message belongs
to (distinguishing between a local, national and international political level).
Moreover, we coded for the presence of every-day life imagery/postings (unre-
lated to politics and the professional realm of the politician). We also coded
for different types of performances and practices related to the professional
(and semi-professional) aspects of political life, distinguishing between every-
day professional footage (e.g. imagery of meetings and other day-to-day duties/
tasks), political performances (superseding the every-day duties of politicians),
media appearance (both meta and message level), and attendance at celebrity
events and public demonstrations, respectively. We included negative comments
directed to political opponents (i.e. negative campaigning), social media activ-
ity and comments on media content (not relating to the politician in question).
Finally, we coded for the use of Instagram posting as a digital poster (publish-
ing more traditional party or political poster imagery/messages in their feeds).
As we have argued elsewhere (Ekman and Widholm 2014, 2017), politi-
cians’ use of selfies could be understood as a form of ‘performed connectiv-
ity’, meaning that politicians’ imagery on social media platforms is strategically
selected to showcase various forms of interrelationships with other people
(such as other politicians, celebrities, media professionals, ordinary people,
etc.). In order to capture actor connectivity, we analyse if the image and/or
textual feature relate to other individuals/groups (manifested in the imagery or
the textual message). We coded for the presence of intra-political connectivity
(distinguishing between connectivity with party colleagues, allied politicians,

www.intellectbooks.com   21
Mattias Ekman | Andreas Widholm

political opponents and international politicians). We coded for media connectivity

(the presence of journalists or media professionals), and celebrity connectivity
(distinguishing between cultural, sport and royalty actors). Moreover, we coded
for connectivity with businesses actors, authorities/civil servants and civil soci-
ety representatives. And finally, we coded for connectivity with family members
and common people. All variables measuring ‘practice’ and ‘connectivity’ were
categorical, meaning for example that one unit could express connectivity with
more than one specific type of actor. We also looked for inter-media references,
coding if posts included references to journalism. The socio-technical aspects of
the communicated self (cf. Papacharissi 2015) was captured through the coding
of self-portrayal (if the politician is present in the imagery), composition tech-
niques (e.g., the use of close-ups or full body shots), the visual perspective (the
angle of the camera), and the specific attitude of the post (if the visual/textual
message is positive, neutral negative in character). Results of the content analy-
sis will form the basis of a typology of visual political performances.

A brief note on the Swedish political context

Before we go to results, it is necessary to briefly describe the power- and party
constellation in Swedish politics after the latest national election. The Swedish
political landscape is divided into two major blocs. A centre-right constel-
lation of four parties including the Centre Party, the Liberals, the Moderate
Party and the Christian Democrats has held the power between 2006 and
2014. After the 2014 election, the Social Democratic Party and the Green
Party formed a minority government. The far-right populist party the Sweden
Democrats constitutes a third political force in Sweden. The party managed
to get seats in the Swedish parliament for the first time in 2010, drawing on
anti-immigration policies and highly conservative ideals similar to the popu-
list right wing tendencies that have been consolidated in other European
countries. In the 2014 election, they doubled their votes and became the third
largest party in Sweden. One question is thus to what extent the political
landscape is reflected in the way individual politicians use Instagram.

Results
In this section, we present the results of the content analysis, and it has been
structured into two main sections. First, we discuss connective features, distin-
guishing between six actor fields to which politicians – in varying degrees –
are connected. Second, we show how political communication on Instagram
relates to traditional media and the mediated ‘self’.

Types and levels of connectivity

The socio-technical affordances of Instagram facilitate a variation of poten-
tial actor connectivity. The constant flow of images published in the politi-
cians’ Instagram platforms, discloses numerous self-produced and mass-self
communicated images mirroring the classical political ‘photo-op’. For exam-
ple, by displaying affiliation to a certain celebrity or international politician,
blending with common people at various locations, or showcasing internal
unity and solidarity within a party or a coalition, politicians express differ-
ent forms of connectivity. In the various flows of images, we notice a broad
repertoire of actor connectivity. The practices of politicians’ social media
communication reflect the production of ‘closeness’ or ‘togetherness’, which

22   Northern Lights

 Political communication in an age of visual connectivity

2. All the displayed

correlations are
significant at the 0.001
level.

Figure 1: Level of connectivity for political parties (per cent, n=800).

embodies the ideological structure (and infrastructure) of a visual social media

platform such as Instagram. In the following we start with an analysis of the
intensity of displayed connectivity and relate this to both party belonging and
individual politicians. Second, we assess different forms of actor connectivity
displayed in the sixteen Instagram accounts.
All the represented parties and all the politicians display various forms
of connectivity. There are some significant differences between the amounts
(intensity) of connectivity displayed in the various accounts. When the rela-
tion between political party and connectivity is assessed, there are also some
interesting differences (see Figure 1). However, the inter-party differences
are most accurately explained on the level of individual politicians. There is a
stronger (yet still moderate) correlation (Cramer’s V=0.486)2 between connec-
tivity and individual politicians, compared to the correlation between politi-
cal party and connectivity (Cramer’s V=0.400). We have clustered the various
levels of actor-connectivity displayed in the sixteen politicians Instagram

Hyper connectivity High connectivity Moderate connectivity Low connectivity

96 % Bah Kuhnke 72 % Bush Thor 54 % Holm (Left Party) 30 % Modig

  (Green Party)   (Christian Dem.) 54 % Malm (Liberals)   (Cristian Dem.)
94 % Fridolin 72 % Ohlsson 48 % Dinamarca 18 % Bieler
  (Green Party)  (Liberals)   (Left Party)   (Sweden Dem.)
94 % Wikström 66 % Kinberg Batra 46 % Bylund
  (Social Dem.)   (Moderate Party)   (Sweden Dem.)
90 % Lööf 44 % Källström
  (Center Party)   (Center Party)
82 % Johansson 42 % Forsell (Moderate
  (Social Dem.) Party)

Table 2: Level of connectivity for individual politicians (per cent, n=800).

www.intellectbooks.com   23
Mattias Ekman | Andreas Widholm

accounts in four categories, demonstrating the amount of connectivity with

other actors. We label these: The hyper-connected politician (politicians with
more than 80% connectivity), the high-connected politician (60–79% connec-
tivity), the moderately connected politician (40–59% connectivity) and the politi-
cian with a low level of connectivity (less than 40%).
The level of connectivity is partially related to the position of the indi-
vidual politician, four out of five of the hyper connected politicians are members
of the current Swedish government, and the fifth is a party leader. The other
two party leaders in the sample (Bush Thor and Kinberg Batra) are positioned
in the category of high connectivity. When considering gender and the level
of connectivity, the female politicians reveal slightly more connectivity (eleven
percentage points) than the male politicians as a total. The most common
form of connectivity is with party colleagues (which counts for 25 per cent of
all images displaying connectivity). When using clustered categories of actor
connectivity, the results show that political connectivity is visible in almost half
of the images exhibiting contact with actors. The number of images display-
ing other forms of actor distribution is fairly evenly distributed (see Figure 2).
The types of actor connectivity displayed in the images differ between politi-
cians and tend to reflect not only individual politicians’ professional and private
lives, but also their party affiliation. There is a span ranging from insignificant
(Cramer’s V=0.100) to moderate correlation (Cramer’s V=0.400) between party
affiliation and various types of actor connectivity, where most forms of connectiv-
ity display low or low/moderate correlation related to party affiliation. While the
most common form of connectivity is with party colleagues, there is one exception:
the Green Party. Twenty per cent of their connective images display connectiv-
ity with media professionals/journalists. In relation to intra-party connectivity,
the two MGs (Members of Government) of the Social Democratic Party display
the highest figures. Nearly every third connective image (32%) contains party
colleagues, indicating a different communicative strategy compared to other
parties (whose figures range from nine to eighteen intra-party connectivity). The
most notable party differences are detected in connectivity with common people.
The three centre-left parties have quite a high level of connectivity with common
people (16, 20 and 20%), whereas the four centre-right parties and the populist

Figure 2: Types of actor connectivity (per cent of images displaying connectivity, n=501).

24   Northern Lights

 Political communication in an age of visual connectivity

far-right party display very low connectivity with common people (ranging from
0 to 5%). Noticeable is the zero connectivity with common people within the two
accounts belonging to MPs of the Sweden Democrats.
There are some noticeable differences between parties (and between party
blocks) in relation to different types of actor connectivity. The Green Party and
the Left Party display relatively high connectivity with cultural actors (16%
and 10%, respectively), and the MGs from the Green Party and the Social
Democratic Party have high connectivity with civil society actors (22 and 12%,
respectively). The two MPs of the Christian Democrats, the Liberals and the
Sweden Democrats display relatively high family connectivity (16, 13 and
10%, respectively). These inter-party differences could partially be explained
by assessing party political strategies, but also by looking at the professional
level. One of the Green Party MGs is the minister of culture, and meeting
celebrities from the cultural field is part of the professional work. Showcasing
a high family connectivity could be an assumed strategic choice for politicians
in a conservative and ‘family value’ oriented party.
When assessing connectivity with other platform users, we also meas-
ured the level of interactivity (responses to comments) for each politician. The
result reveals significant individual differences. The level of interactivity ranges
from only 2% for three politicians (Wikström, Johansson and Bylund) to 86%
for the minister for culture and democracy, Alice Bah Kuhnke. This reflects a
possible change in Bah Kuhnke’s strategy as her Instagram bio explicitly has
stated that questions should be directed to the ministry of culture rather than
to herself (Ekman and Widholm 2017). The second most interactive politician
is Lööf (46%). Bah Kuhnke and Lööf are also the two most followed politi-
cians in the sample, so engaging in user-interactivity seems to be a way of
obtaining and sustaining user traffic.

Politics, media and performances of professional and

private selves
As noted in the introduction to this article, results from previous stud-
ies of political communication in Sweden (e.g. Ekman and Widholm 2015,
2017) have shown that the politics–media relationship is articulated in new
ways online, but also that there are differences between platforms when it
comes to type of published content. Instagram is a unique platform based on
its visual focus, but it is also special when it comes to the type of content that is
published there. About one third, 35%, of all posts analysed contained politi-
cal messages in some form (see Table 3). Differently put, a clear majority of the
communication deals with other dimensions of politics than the strictly ideo-
logical. The politicians of the Green Party stand out, however, being the only
politicians in the sample that place actual political issues at the very centre of
their communication, yet often in combination with a highly personalized and
self-centred style reflected in the exceptionally high figures with regard to self-
imagery (83% in total). On the opposite side, we find the Sweden Democrats,
whose communication is almost empty when it comes to political issues (only
7% contained political messages). More generally, our results disclose that
news media references seem to hold a strong symbolic value. As displayed in
Table 3, one fifth of all postings contained references to journalism. In addi-
tion, as much as 36% of the ‘political’ postings, (e.g. those containing a politi-
cal message) referred to activities taking place in the news media. Examples of
media references are images or screenshots of opinion articles written by the

www.intellectbooks.com   25
Mattias Ekman | Andreas Widholm

Reference to Political Everyday Everyday

news media message professional private

Left party 9 42 30 48
Social Dem. Party 12 29 45 24
Green Party 30 59 41 6
Liberals 20 36 31 37
Center Party 30 37 36 26
Moderate Party 29 42 39 20
Christian Democrats 19 26 37 40
Sweden Democrats 7 9 31 61
All parties 20 35 36 33

Table 3: Presence of four information features (per cent within each party) (n=800).

politicians themselves as well as video clips from television news interviews.

In this way, communicated messages are often re-mediated and translated
from one media type to another, and references are often used as a tool for
the provision of criticism against political opponents. Neither journalists nor
journalism in general seem to be the object of such criticism (less than 1% of
the entire material expressed a negative stance towards journalists or specific
articles). Media references appear in digital flows across the entire political
spectrum, but there are also noticeable variations between different parties.
Political parties located more in the centre of the left-right spectrum
(Green Party, Liberals, Centre Party) seem to rely on news media to a greater
extent than those located at the ends of the same spectrum (especially the
Left Party and the Sweden Democrats). The Sweden Democrats’ low degree
of news media connectivity can be explained by their broader political orien-
tation. For the Sweden Democrats, journalists belong to the political ‘estab-
lishment’ in Sweden, and it is precisely that establishment that they seek to
defeat in their rhetoric (cf. Oja and Mral 2013). The party is still controversial
in Sweden due to its anti-immigrant ideology and right wing extremist past,
and they receive far more negative than positive coverage in the media. In that
sense, they have less to gain by referring to images of themselves in the news
media compared to other parties represented in the parliament. The Social
Democrats are somewhat of an anomaly here, since they are the largest politi-
cal party in Sweden, leading the current government, and by that they receive
constant media coverage that they easily could draw on and benefit from in
their visual communication. Despite being on opposite sides of the political
spectrum, The Sweden Democrats and the Left Party share another central
feature, namely that they provide a glimpse into their private lives more often
than politicians of other parties (61 and 48 per cent, respectively). This is nota-
bly higher than the share for all parties (33 per cent).
Instagram is a platform centred on visual self-communication and visual
performativity of the ‘self’ is at the heart of its social media logic (Ekman and
Widholm 2017). The patterns of distribution across the data reveal that there
are indications of both gender and party differences when considering the use
of self-images (see Figure 3).
There is a moderate correlation between party affiliation and the use of self-
ies (Cramer’s V=0.365), but a weak correlation considering gender (Cramer’s

26   Northern Lights

 Political communication in an age of visual connectivity

Figure 3: Gender differences in self-portrayal (per cent, n=800).

V=0.143). Thus, the general gendered pattern across political parties shows
that female politicians in six parties are posting self-images to a greater extent
compared with their male colleagues (59 per cent of images in the female
politicians’ accounts are self-images compared with 45 per cent of the male
politician’s images). Selfie-images generate more likes and comments from
other users/publics, compared with non-selfie images. The arithmetic mean
of likes for selfie-images across the whole data is A=175, compared to a mean
of A=81 likes for non-selfie images. So, by publishing selfies the politician can
possibly attract more attention to his/her account, thus increasing the central-
ity in the social network of users.
The most followed account in the sample belongs to the minister for culture
and democracy Bah Kuhnke, who also posts selfies to a considerably higher
proportion than any other politician in the sample (92% of Bah Kuhnke’s
images are selfies). There are also notable gender differences concerning the
production of visual intimacy (i.e. closeness) on the politicians’ accounts. As
Berger (1991) argues, various photographic compositions are also coded with
meaning. For example, images of face close-ups, or half-body shots, connote
intimacy and personal relationship, whereas full-body shots and long shots
connotes social relationship and public distance, respectively (van Zoonen
1994: 76). There is an obvious risk of over-emphasizing the specific use of
image-composition, but the overall pattern of selected compositions in rela-
tion to gender can shed some light on gendered visual practices among politi-
cians. When assessing the relation between gender and use of selfie close-ups
and half-body shots (head and torso) we can see that the female politicians
publish 165 selfie images of that category compared with 96 images for the
eight male politicians. This is partly related to the fact that the female politi-
cians publish more selfies in total, compared to the male politicians. But when
assessing the use of close-ups and half-body images within the category of
selfie-images, there is still a minor, yet notable, difference between female
and male politicians (66% of all female selfies are close-ups and half-body
shots compared to 56% of all the male selfies). So, the female politicians do
not only publish more selfies, but their selfies are ‘intimate’, to a higher extent,
compared to the male counterparts. This indicates that traditional gendered

www.intellectbooks.com   27
Mattias Ekman | Andreas Widholm

photographic conventions, reproduced on social media platforms, are visible

to some extent among the politicians studied (cf. van Zoonen 2005).

Concluding discussion
Commercial social media platforms available to anyone with Internet access
and a smart phone have clearly broadened the communicative repertoires of
elected representatives (Enli and Skogerbø 2013). This article has explored new
modes of political communication enabled by the platform-specific features of
Instagram. More specifically we have analysed how connectivity in a variety of
forms is expressed and performed by politicians in Sweden with the largest
number of followers. In a broad sense, connectivity can be seen as a strategy
that creates symbolic ties between individual politicians and various spheres
of society, from the micro-political level of the family to the macro-political
level of global international relations. The analysis shows that type of imagery
and composition differ between politicians and that differences can be traced
to party affiliation, individual strategies and to some extent gender. Political
connectivity is by far the most common type, while other identified categories
such as celebrity, media, family and civil society are more equally represented.
Despite the fact that social media have enabled politicians to communicate
more independently from news media, many of the actors included in this
study still rely on news media, not only as an arena but as a symbolic centre
for the political discourse. Twenty per cent of all studied images included
references to news media, while 36 per cent of the images that contained a
political message did the same. However, if the crisis of journalism deepens, it
is not a given that this centrality will remain.
The mixture of professional and private imagery is an apparent feature
that we have identified across the whole sample, but there are also varia-
tions between the politicians, especially concerning the private realm. The
Sweden Democrats stand out as they combine low or moderate connectiv-
ity and few media references with a depoliticized and highly privatized form
of communication that sets the focus on the everyday life. This should be
seen in light of the party’s anti-establishment rhetoric, where low connectiv-
ity can function as a way of underlining a ‘folksy’ and ‘ordinary’ ideology. It
is also reminiscent of the ‘porous’ features of online political communica-
tion as the private life becomes increasingly important on behalf of tradi-
tional ‘bounded’ politics (cf. Blumler and Gurevitch 2005). The opposite
of this strategy can be seen in the Green Party’s communication which is
highly political, highly professional and strongly connected to news media.
Other parties are located in the middle of these extremes. Despite the differ-
ences, however, the incorporation of self-produced imagery into the daily
communication strategies of politicians is strongly related to an ongoing
popularization of the political discourse. The ability to produce and publish
images opens new modes of contact between politicians and members of
the public. Many images express a successful life and key to these perfor-
mances is the construction of networked ties between politicians, follow-
ers and other prominent public actors who actively redistribute images to
the wider audience (see also Ekman and Widholm 2017). Simultaneously,
politicians appear as hard working, emphasizing the duties for which they
were elected and therefore the visual communication produces a form of
imaginary accountability. Politicians’ visual political communication certainly
inherits a democratic potential, for example by reducing the communicative

28   Northern Lights

 Political communication in an age of visual connectivity

distance between politicians and citizens. However, the dominant mode of

communication on Instagram centres on the visual branding of individual
politicians rather than on interaction with the public (although there are a
few exceptions in the sample). This is particularly evident in the more profes-
sionally managed accounts, notably the ones belonging to party leaders and
to the minister of culture and democracy. Visual communication differs in
this respect from the more interactive practices commonly associated with
Twitter and to some extent Facebook.
Politicians’ Instagram use can be understood as a way of producing visual
flows of professional and private practices in which the authenticity of every-
day political life takes centre stage. Through the dissemination of daily snap
shots politicians are able to construct strategically self-managed ‘galleries’ in
which work practices and glimpses of their private life becomes visible. This
is a typical feature of celebrity culture, where actors ‘must constantly shift
between performing their stage persona, concealing or revealing personal
information, and creating intimacy and authentic self-presentation’ with their
followers (Marwick and Boyd 2011: 151). In relation to politics, this form of
‘celebritization’ contributes to a depoliticization of public discourses, primar-
ily by shifting focus from ideological questions to lifestyle politics. Even when
political issues are manifested, they come to the fore with an emphasis on the
persona or on the achievements of the particular individual. Depoliticization
should not be seen as a general characteristic of politicians’ communication
through social media. Rather it should be seen as an adaption to the socio-
technical affordances of Instagram as a communicative platform where style,
looks and visual performativity are central features. Images do involve limita-
tions when it comes to communication of complex political issues. In addition,
Instagram does not allow for links to external platforms in individual posts,
which makes it even more complicated to provide ‘thick’ political messages.
It should be emphasized though that visual communication is, of course, not
inherently non-political, and our data clearly shows that there are large differ-
ences between the actors as to whether they use Instagram for dissemination
of clear political messages or for more personal and intimate branding of their
digital identities. However, politicians with a less self-centred style have lower
network centrality and less followers, which indicates that they need to adapt
to Instagram’s aesthetic conventions in order to be successful.
The visual practices of politicians self-communication is still a highly
under-researched field of enquiry. This calls for further analyses, not least
comparative ones, focusing on how visually ‘performed’ politics relate to
different political systems and media cultures. Moreover, the overall question
of whether social media and politics inherit and advance democratic prac-
tices, or if they foremost mirror a post-democratic development, calls for more
interdisciplinary and multi-methodological empirical approaches dealing with
the overall mediatization of politics.

References
Asp, K. (2014), ‘News media logic in a new institutional perspective’,
Journalism Studies, 15:3, pp. 256–70.
Berger, A. A. (1991), Media Analysis Techniques, rev. ed., Newbury Park and
London: Sage Publications.
Blumler, J. G. (2010), ‘The two-legged crisis of journalism’, Journalism Practice,
4:3, pp. 243–45.

www.intellectbooks.com   29
Mattias Ekman | Andreas Widholm

Blumler, J. G. and Gurevitch, M. (2005), ‘Rethinking the study of political

communication’, in J. Curran and M. Gurevitch (eds), Mass Media and
Society, New York: Oxford University Press, pp. 104–21.
Bro, P. and Wallberg, F. (2015), ‘Gatekeeping in a digital era: Principles, practices
and technological platforms’, Journalism Practice, 9:1, pp. 92–105.
Brock, G. (2013), Out of Print: Newspapers, Journalism and the Business of News
in the Digital Age, Philadelphia and London: Kogan Page Publishers.
Castells, M. (2013), Communication Power, Oxford: Oxford University Press.
Chadwick, A. (2013), The Hybrid Media System: Politics and Power, Oxford:
Oxford University Press.
Davis, A. (2010), Political Communication and Social Theory, London and New
York: Routledge.
Dijck, J. van (2012), The Culture of Connectivity: A Critical History of Social
Media, Oxford: Oxford University Press.
Dijck, J. van and Poell, T. (2013), ‘Understanding social media logic’, Media
and Communication, 1:1, pp. 2–14.
Ekman, M. and Widholm, A. (2014), ‘Twitter and the celebritization of
politics’, Celebrity Studies, 5:4, pp. 518–20.
—— (2015), ‘Politicians as media producers: Current trajectories in the
relation between journalists and politicians in the age of social media’,
Journalism Practice, 9:1, pp. 78–91.
—— (2017), ‘Performative intimacies and political celebritization’, in A.
Kuntsman (ed.), Selfie Citizenship, Cham: Springer, pp. 65–74.
Enli, G. S., and Skogerbø, E. (2013), ‘Personalized campaigns in party-centred
politics: Twitter and Facebook as arenas for political communication’,
Information, Communication & Society, 16:5, pp. 757–74.
Esser, F. (2013), ‘Mediatization as a challenge: Media logic versus political
logic’, in H. Kriesi, F. Esser, J. Matthes, D. Bochsler, M. Bühlmann and
S. Lavenex (eds), Democracy in the Age of Globalization and Mediatization,
Basingstoke: Palgrave Macmillan, pp. 155–76.
Hjorth, L. and Hendry, N. (2015), ‘A snapshot of social media: Camera phone
practices’, Social Media+ Society, 1:1, pp.1–3.
Holmes, S. and Redmond, S. (2006), Framing Celebrity: New Directions in
Celebrity Culture, London and New York: Routledge.
Kammer, A. (2013), ‘The mediatization of journalism’, MedieKultur: Journal of
Media and Communication Research, 29:54, pp. 141–48.
Klinger, U. and Svensson, J. (2014), ‘The emergence of network media logic
in political communication: A theoretical approach’, New Media & Society,
17:8, pp. 1241–57.
Kunelius, R. and Reunanen, E. (2016), ‘Changing power of journalism: The
two phases of mediatization’, Communication Theory, 26:4, pp. 369–88.
Larsson, A. O. (2015), ‘Pandering, protesting, engaging. Norwegian party
leaders on Facebook during the 2013 “short campaign”’, Information,
Communication and Society, 18:4, pp. 459–73.
Lasorsa, D. L., Lewis, S. C. and Holton, A. E. (2012), ‘Normalizing Twitter:
Journalism practice in an emerging communication space’, Journalism
Studies, 13:1, pp. 19–36.
Marsh, D., ‘t Hart, P. and Tindall, K. (2010), ‘Celebrity politics: The politics of
the late modernity?’, Political Studies Review, 8:3, pp. 322–40.
Marwick, A. and Boyd, D. (2011), ‘To see and be seen: Celebrity practice on
Twitter’, Convergence: The International Journal of Research into New Media
Technologies, 17:2, pp. 139–58.

30   Northern Lights

 Political communication in an age of visual connectivity

McChesney, R. W. (2013), Digital Disconnect: How Capitalism is Turning the

Internet Against Democracy, London and New York: The New Press.
Neuendorf, K. A. (2002), The Content Analysis Guidebook, Thousand Oaks, CA:
Sage Publications.
Oja, S. and Mral, B. (2013), ‘The Sweden Democrats came in from the cold:
How the debate about allowing the SD into media arenas shifted between
2002 and 2010’, in R. Wodak (ed.), Right-Wing Populism in Europe: Politics
and Discourse, London: Bloomsbury, pp. 277–92.
Papacharissi, Z. (2015), Affective Publics: Sentiment, Technology and Politics,
Oxford: Oxford University Press.
Russell, A. (2011), Networked: A Contemporary History of News in Transition,
Cambridge: Polity.
Schroeder, R. (2016), ‘Online entertainment: The globalization of on-screen
sociability: Social media and tethered togetherness’, International Journal of
Communication, 10:2016, pp. 5626–43.
Siles, I. and Boczkowski, P. J. (2012), ‘Making sense of the newspaper crisis:
A critical assessment of existing research and an agenda for future work’,
New Media and Society, 14:8, pp. 1375–94.
Street, J. (2010), Mass Media, Politics and Democracy, Basingstoke: Palgrave
Macmillan.
Strömbäck, J. (2008), ‘Four phases of mediatization: An analysis of the
mediatization of politics’, The International Journal of Press/Politics, 13:3,
pp. 228–46.
—— (2011), ‘Mediatization and perceptions of the media’s political influence’,
Journalism Studies, 12:4, pp. 423–39.
Wheeler, M. (2013), Celebrity Politics, London: Polity Press.
Wiman, B. (2016), ‘Dags för Alice Bah Kuhnke att logga ut från Instagram
och ta tag i mediepolitiken’, Dagens Nyheter, 20 August, http://www.dn.se/
kultur-noje/bjorn-wiman-dags-for-alice-bah-kuhnke-att-logga-ut-fran-
instagram-och-ta-tag-i-mediepolitiken/. Accessed 1 September 2016.
Zelizer, B. (2015), ‘Terms of choice: Uncertainty, journalism, and crisis’, Journal
of Communication, 65:5, pp. 888–908.
Zoonen, L. van (1994), Feminist Media Studies, London and Thousand Oaks:
Sage Publications.
—— (2005), Entertaining the Citizen: When Politics and Popular Culture Converge,
Lanham, MD: Rowman and Littlefield.
—— (2006), ‘The personal, the political and the popular a woman’s guide to
celebrity politics’, European Journal of Cultural Studies, 9:3, pp. 287–301.
Zoonen, L. van, Vis, F. and Mihelj, S. (2010), ‘Performing citizenship on
YouTube: Activism, satire and online debate around the anti-Islam video
Fitna’, Critical Discourse Studies, 7:4, pp. 249–62.

Suggested citation
Ekman, M. and Widholm, A. (2017), ‘Political communication in an age of
visual connectivity: Exploring Instagram practices among Swedish politi-
cians’, Northern Lights, 15, pp. 15–32, doi: 10.1386/nl.15.15_1

Contributor details
Mattias Ekman is senior lecturer at Örebro University, Sweden. He is currently
heading a research project on ‘Interactive racism’ that analyses how racism is
discursively produced, shaped and distributed in dialogic interaction between

www.intellectbooks.com   31
Mattias Ekman | Andreas Widholm

online publics, media actors and political actors. Ekman is also studying politi-
cians’ strategic use of social media, focusing on the changing communicative
dynamics between journalism and politics. Ekman has published in interna-
tional research journals such as Journalism Practice, Ethnic and Racial Studies,
Celebrity Studies and Mediekultur.
Contact: School of Humanities, Education and Social Sciences, Örebro
University, SE-701 82 Örebro, Sweden.
E-mail: [email protected]

Andreas Widholm is associate professor of journalism at Stockholm

University, Sweden. His research addresses the relationship between media,
politics and culture with a particular focus on the changing role of journalism
in the new digital landscape of social media. His research also explores global
media forms and constructions of political and cultural identities. Widholm
has published in international media research journals such as Journalism
Practice, Digital Journalism, Convergence and Celebrity Studies.
Contact: Department of Media Studies, Stockholm University, 115 93
Stockholm, Sweden.
E-mail: [email protected]

Mattias Ekman and Andreas Widholm have asserted their right under the
Copyright, Designs and Patents Act, 1988, to be identified as the authors of
this work in the format that was submitted to Intellect Ltd.

32   Northern Lights