When politicians go native: The consequences of political native advertising for citizens’
trust in news
Manuscript accepted for publication in Journalism.
Authors: Magnus Hoem Iversen and Erik Knudsen
Affiliation: Department of Information science and media studies, University of Bergen,
Fosswinckelsgate 6, 7802 Bergen, Norway.
Corresponding author: Erik Knudsen, Department of Information science and media
studies, University of Bergen, Fosswinckelsgate 6, 7802 Bergen, Norway. Email:
When politicians go native: The consequences of political native advertising for citizens’
trust in news
Declining revenues from offline and online ads has led publishers to pursue new avenues,
such as native advertising: camouflaging ads as news. Critics of native advertising claim that
this form of advertising blurs the boundaries between editorial and commercial content, and
can reduce the audiences’ trust in editorial content. However, little research has assessed the
possible effects of native ads on audiences' trust in news. With an experimental design
embedded in an online survey (N=733) representative of the Norwegian population, this study
explores the consequences of political native advertising for citizens’ trust in political news.
The present paper discusses how political native advertising poses a challenge to the boundary
between journalism and advertising as well as the boundary between journalism and powerful
elites. Our study examines 1) how prominently native advertisements should be labelled in
order for readers to recognise them as advertising content, and 2) whether exposure to such
ads reduces readers' trust in political news. Our most important finding shows that, when
explicitly labelled, native advertising by political parties can reduce people’s trust in political
Native advertising, boundaries of journalism, journalistic ideology, trust, political news,
The journalistic profession is reconfiguring its boundaries in regard to what are considered
acceptable advertising practices. Advertising is crucial to news media companies, since they
are also businesses (Albarran, 2010; Picard, (2002) 2011) and are thus dependent on a stable
income. Independent journalism cannot exist as a service without a working business model
to support the often costly, and not always very profitable, ‘good quality journalism’ that is
key to the monitoring of the laws, practices and institutions of modern democracies
(Strömbäck, 2005). The current business model of journalism is facing a threat to two of its
main income sources: audiences’ lack of will to pay for news media and declining income
from advertisers, leading to plummeting revenues (Kaye and Quinn, 2010: 5–6). Thus, the
need for innovation and solutions to the problems are pressing. Concerning the problems of
advertising revenue, the growing phenomenon of ad-avoidance among news consumers paired
with technology that allows for so-called ad-blocking makes the problem even more pertinent.
As a response, news organizations are increasingly moving away from traditional banner ads
and are spending more resources on so-called native advertising.
Native advertising mimics editorial content in both form (Coddington, 2015: 14) and
positioning within the editorial space (Carlson, 2014: 2). As publications such as The New
York Times and The Guardian not only feature these types of advertisements in their online
papers but actively integrate their own departments dedicated to the creation and proliferation
of the practice into their own institutions, the phenomenon appears to be here to stay.
The advent of native advertising, both from commercial and political sponsors, has
sparked heavy debate between professional practitioners of both journalism and advertising.
Native advertising, in addition to being controversial, offers a glimmer of hope for newsrooms
in a milieu in which revenue from advertising has looked dreary for quite some time.
However, advertising that seeks to blend in with the editorial content could also pose a
challenge to readers’ trust in independent journalism. Native advertising could blur the
boundary separating commercial and editorial spaces and interests within individual news
publications. More importantly, it could blur the lines between these interests for the audience
of journalism. This is problematic, because it may lead to a decline in credibility for ordinary
editorial content. If audiences come to view all content with the same scepticism and disdain
they subject regular advertising to, journalism is running the risk of diminishing its most
valuable asset: credibility. Native ads as a solution to revenue loss could undermine one of the
very foundations for making revenue at all.
As this article will demonstrate, corporations are not alone in employing native
advertising. Politicians and political parties are also increasingly advertising through this
format. The boundaries between journalism, advertising and political actors are changing. As
we will argue, native advertising is a new and rapidly changing phenomenon. Because of this,
research on its consequences is remarkably insufficient – especially regarding its potential
effects on citizens’ trust in news (for notable exceptions investigating neighboring aspects, see
Austin and Newman, 2015; Cramer, 2015; Cole and Greer, 2013; Rodgers, 2007; Tutaj and
van Reijmersdal, 2012; Wojdynski and Evans, 2015). Empirically, this paper seeks to remedy
this by investigating the effects of political native advertising on readers’ trust in political
news. Theoretically, our main contribution is to understand native advertising by political
actors as a challenge to two of the established boundaries of journalism: the separation of
editorial content from commercial content and the separation of journalism from the influence
and interests of powerful institutions such as political parties. We adopt the view of
journalism as ever reliant on continuous boundary work. In the words of Singer: ‘The survival
of journalism as an occupation depends on its credibility, which is gained through the
collective behaviour of its practitioners’ (2015: 22). Journalistic boundary work attempts to
define and regulate this collective behaviour. Understanding the profession of journalism as
being constantly constructed and boundary lines being continually drawn between ‘in groups’
and ‘out groups’ as well as ‘good practice’ and ‘bad practice’ (Carlson, 2015: 2), this paper
will examine boundaries defined by both political and economic tensions in the daily
workings of journalism.
In the following, we first discuss the consequences of native ads for the relationship
between journalism and advertising, understood as the strict separation of the two forms based
on journalistic ideology and professional norms and standards.
Second, we discuss the consequences of political native ads for the relationship
between journalism and politicians and powerful elites, understood through the professional
norm of the press as the fourth estate and journalism’s social contract.
Third, as political native advertising challenges the two abovementioned boundaries at
the same time, we employ an experimental design embedded in a survey (N=733) to examine
the effects of political native advertising. Empirically, our study makes a novel contribution
by showing that citizens, at least in the Norwegian context, are able to distinguish between
editorial content and political native advertising and that explicitly labelled native advertising
by political parties can reduce people’s trust in political news. Based on this, we discuss the
potential implications of political native advertising for journalistic credibility.
Native ads and the boundary between commercial and editorial content
The boundary between commercial and editorial space in a newspaper points to a central
tension in journalism, as the field is torn between market values and democratic values
(Altheide and Snow, 1991). On the one hand, journalism is dependent on the audiences
trusting that the journalists are autonomous from their owners in editorial choices in order to
be perceived as a credible source of information (Schudson, 2011). On the other hand,
journalism is dependent on funding, as producing journalism is costly, and the publications
have owners desiring profits. To cope with this conflict of loyalty, the needs of citizens on the
one hand, and the needs of owners on the other, publishers have traditionally attempted to
keep business decisions separate from editorial decisions (Carlson, 2014) – known as the
separation between ‘church and state’, or ‘the wall’ between news and business (Coddington,
2015). When it comes to advertising, this division has manifested itself as a trade-off in which
commercial actors are granted access to audiences but left with no power over news content
(Carlson, 2014: 3).
However, as Coddington (2015: 69) points out, the strict business-news-wall was
never as impenetrable as the metaphor suggests. There has been a tendency in news
organizations to accommodate the interests of advertisers rather than prioritizing certain
democratic norms (Baker, 1994: 44). ‘The wall’ has served a symbolic function, as something
that has helped shape a journalistic identity, and as a marker for what has been a professional
norm, serving as a so called boundary device (Singer, 2015: 22). Such norms are often
codified. For instance, a sentence in The Society of Professional Journalists’ Code of Ethics
reads, ‘Distinguish news from advertising, and shun hybrids that blur the lines between the
two (...)’ (Society of Professional Journalists, 2014).
These boundaries are being contested and reconfigured by native advertising. Two
examples provide arguments for why this could be the case. The first example is an article in
the New York Times, complete with a by-line and well-crafted illustrations as well as video
and audio segments, exploring the trials and tribulations of women in American prisons
(Deziel, 2014). Through subtle visual cues, one is informed about the fact that it is a ‘paid
post’ produced by T Brand Studio and sponsored by Netflix – to promote a new season of the
series ‘Orange Is the New Black’. T Brand Studio is part of the New York Times’ advertising
department. At the very bottom of the webpage there is a small disclaimer: ‘The news and
editorial staffs of The New York Times had no role in this post’s preparation’. The article is
an advertisement resembling ‘normal’ online news content.
Second, in 2013, The Atlantic published an article titled ‘David Miscavige Leads
Scientology to Milestone Year’. This story was an advertisement resembling a ‘normal’ news
story, with a subtle clue to its actual genre at the top of the page reading ‘sponsored content’.
That the sponsor was the Church of Scientology was revealed at the bottom of the article. The
ad spurred much debate and controversy and was eventually retracted by The Atlantic (for a
detailed account, see Carlson, 2014).
The form of boundary work taking place here can be understood as an expansion,
shifting the borders of what is considered journalism (Carlson, 2015: 10) in terms of both
participants and practices. In terms of participants, non-traditional professionals are being let
into newsrooms. A clear example is the new profession called ‘content creator’, a hybrid
between journalist and advertiser. These are increasingly given space and resources within the
newsrooms themselves. Furthermore, an expansion of practices is taking place – as native
advertising is employed, newsrooms are implicitly stating that it is acceptable. There is
resistance to native advertising both outside and inside newsrooms. Consequently, there are
attempts at the expulsion and protection of autonomy (Carlson, 2015: 10). An example of
such attempts are revisions to the Society of Professional Journalists’ Code of Ethics in
America, specifying the need for a strict divide between editorial and commercial content.
Even though advertisers are being let into newsrooms, the boundary could exist within
individual newsrooms. Following a critical piece on the Norwegian newspaper Aftenposten
and their newly established hybrid advertising department within their newsroom (Flatø,
2015), an editor of the publication maintained that sponsored content has nothing to do with
journalism: ‘Employees of Aftenposten Brand Studios are not journalists. They are
advertisers’ (Borud, 2015). However, audiences of journalism do not necessarily see or know
about these boundaries – as they only experience the end product.
As native advertising seeks to ‘camouflage church as state’ (Ferrer, 2016), it raises
questions concerning the importance of the boundaries between editorial content and
commercial interest. Some (Carlson, 2014; Coddington, 2015; Ferrer, 2016) perceive native
advertising to be a challenge – that the wall could potentially become a curtain (Coddington,
2015), with negative consequences for credibility, the legitimacy of the profession and the
institution of journalism in society.
If people both liked and trusted regular advertising as much as editorial content, there
would be no reason for native advertising to exist. Readers are more motivated to read and
pay more attention to editorial content than commercial messages (Cameron, 1994). A central
function for native advertising is to be perceived as editorial content by the reader by
mimicking the shape, form and function of the mentioned traditional content made by
journalists. A form of trickery is essential to a native ad functioning as intended (Carlson,
2014). Readers have to believe that they are actually consuming journalistic content. The
native ad is a commercial message attempting to piggyback on the trustworthiness that the
traditional journalistic message carries: ‘The goal is not merely to generate greater exposure
than display advertising, but “to appropriate the format of the surrounding publication and
harness its credibility to strengthen the authority and persuasiveness of the advertising”’
(Wasserman, 2013 in Carlson, 2014). What the communicating text is actually perceived as is
The literature thus far indicates that readers are somewhat confused when faced with
native advertising. Previous research shows that readers are less sceptical of native ads
compared to banner ads (Tutaj and van Reijmersdal, 2012) in that they have a harder time
recognizing sponsored content than traditional banner ads. Others have suggested that this is
due to the fact that users fail to realise that they are reading an advertisement (Howe and
Teufel, 2014: 79), which studies by Kim et al. (2001) and van Reijmersdal (2009)
substantiate. Wojdynski and Evans (2015) found that readers struggle with recognizing native
ads as advertising. They mention that one explanation for this could be that readers were
asked whether they encountered advertising when reading a story – not explicitly prompted to
answer whether the story they read was advertising.
The contested novelty of native advertising
The issues raised by native advertising are not novel. The novelty of the format itself is
contested. Some claim that this is an old practice, referring to examples such as the Michelin
Guide or the even clearer parallel of advertorials, defined as “editorial-like advertising”
(Cameron and Ju-Pak, 2000: 65) in print newspapers and magazines. This format has received
a wide range of scholarly attention (i.e. Cameron et al., 1996; Cameron and Ju-Pak, 2000;
Cooper and Nownes, 2004; Eckman and Lindlof, 2003; van Reijmersdal et al., 2005).
Cameron and Curtin found that advertorials could have adverse effects on readers’ ability to
remember message content yet called for more follow-up studies to assess the potential effects
of loss of credibility (Cameron and Curtin, 1995: 185). Addressing concerns on advertorials,
Cameron and Ju-Pak claimed that “newspapers should safeguard editorial credibility by
making advertorials distinct from editorial content” (Cameron and Ju-Pak, 2000: 65) due to
the possible risk of tainting the credibility of editorial content.
Investigating labelling, Kim et al. (2001) found indications of advertorials fooling
‘(…) readers into greater involvement with the advertisement message’ and that participants
failed to recall the labelling (Kim et al., 2001: 4). Another previous example of blurred lines
between politics and news through advertising are newsads – political television ads
mimicking news formats (Jamieson, 1992: 148–150).
Native advertising can be seen as the latest iteration in a long line of advertising
formats attempting to borrow credibility from news. Indeed, it is possible to view native
advertisements as ‘digital advertorials’. Our point in this paper is not necessarily to enforce a
strict divide between the two genres. However, we argue that several key characteristics
warrant the ‘native’ term. Native advertising goes beyond advertorials by adapting to ‘the
look-and-feel, the visual design, the usability and the ergonomics of the publisher’s website’
(Matteo and Dal Zotto, 2015: 177). This characteristic emphasizes the technical novelty that
online features bring. More important from a journalistic point of view is the shift in practice
regarding the editorial/commercial boundary. Newsrooms are now actively integrating, even
welcoming, this form of advertising to a greater degree than before. It is native not only in the
technical sense – but in its integration in newsrooms as well. The practice of advertising
mimicking news has moved from a necessary evil to ‘openly commerical journalistic
ventures’ (Ferrer, 2016: 9) and is a sign of boundaries that are being reconfigured as a result
of severe advertising revenue loss. Part of the novelty of native advertising is the changed
situation of journalism.
Native ads and the boundary between journalists and politicians
The boundary between journalists and politicians is often manifested as part of a strong desire
to be both independent from and critical towards institutions of power in society (Schudson
and Anderson, 2010). Similar to the business-news-wall, this boundary has been a fluid and
changing phenomenon and is a professional ideal that has emerged over time. As journalism
exchanged its former passive roles of covering politics with investigative journalism
(Hjarvard, 2013: 49), distinguishing itself from other institutions as an independent actor, or
institution (Cook, 2006: 160), a general critical stance against political actors emerged
(Lengauer et al., 2011: 3). In this setting, journalism positions itself in the role of the fourth
estate, often called the public watchdog – monitoring how politicians and the state exercise
power on behalf of the people (Strömbäck, 2005: 332). Thus, the normative ideal of
journalism states that it should be critical and vigilant in the face of various powerful actors in
society and towards the three estates in particular. A central ideal in the practice of journalism
can be considered as a social contract, promising to hold these powerful actors to account on
behalf of the people (Sjøvaag, 2010).
In Norway, the construction of this professional boundary has been particularly
important since the 1970s, as it was during this period that individual newspapers started
breaking free from their historically tight affiliations with political parties. The Norwegian
newspapers have travelled from party press to a more politically independent press.
Considering this, one could argue that maintaining the boundary between political parties and
journalism in terms of autonomy is particularly salient in the Norwegian context to maintain
readers’ trust. In other countries, such as the US – political parties and press outlets parted
ways at an earlier point in time. However, the increased partisanship of journalism both on
television and online (e.g. Stroud, 2011) gives the discussion new relevance. The potential
problems of mixing editorial and political interests are still pertinent.
Norway, like the rest of the Nordic countries, is a ‘consensus democracy’ with an
egalitarian population, both socially and economically, and high levels of trust in both private
and governmental institutions (Syvertsen et al., 2014: 4–8). This trust extends in part to the
media as an institution (Aalberg and Curran, 2012: 196). Furthermore, Norway has a
particularly high newspaper readership, both in print and online (Syvertsen et al., 2014: 33).
A political native advertisement has a political party or organization as the source
sponsor. Although not as prolific as native ads originating from commercial brands, there are
several examples of political native advertising. BuzzFeed has cooperated with Obama for
America to create native ads consisting of campaign videos on pages that are similar to other
BuzzFeed content (Ellis, 2012). The American case is not unique. The Norwegian Labour
party employed a native advertisement as a part of their local election campaign of 2015.
Within the regional online newspaper Nordlys, the Labour party’s candidate was presented
with text, video and images, all brandishing the design and look of the newspaper within the
paper’s own webpages (Nordlys.no, 2015). The labelling was originally a grey text box next
to the newspaper logo stating ‘sponsored content’. There was no indication of who the
sponsor actually was until the article attracted controversy due to this very lack of labelling.
Subsequently, the article was explicitly labelled: ‘The content on this page is paid for by the
Considering journalism’s normative understanding as a fourth estate, this form of
advertising presents a new development. This could be considered a colonization from
political, not merely commercial, interests. This could potentially, as in the Atlantic-
Scientology example (see Carlson, 2014), pose a threat to journalistic autonomy and the
watchdog role. The shifting boundaries of professional journalism, both in terms of
commercial and political native advertising thus leads to a challenge on two fronts. The
emergence of political native advertising is a clear contrast to the practice of political
adwatching: journalists openly upholding boundaries by putting forward critiques of
advertising content, even on their own pages and in their own programs. Adwatching is a
clear example of journalism attempting to maintain boundaries and separate itself from
political advertising, yet the practice has mostly died out (Jamieson, 2000: 122).
As commercial and political interests put pressure on the boundary devices
maintaining journalistic autonomy, what is essentially at risk is a loss of credibility – that
people will not trust regular news content. If citizens become unsure of where the loyalty of
publications ultimately lies, or if it is unclear whether political actors enforce some sort of
influence over news content, people could experience reduced trust in political editorial
content. For instance, Wired’s David Dobbs (2013 in Carlson, 2014) stated that sponsored
content ‘borrow – no, steal – some of the credibility’ of journalists. Thus, the blurred
boundaries between journalism and advertising can leach credibility from regular editorial
content (Carlson, 2014: 8) and negatively impact audiences’ trust in political news. This could
have wider ramifications. Trust is an essential component for news media (Kohring and
Matthes, 2007) and journalism’s key democratic functions (Strömbäck, 2005). Through the
informational function of the press, citizens and governments alike are informed about the
effects of laws, regulations and procedures on the people and society – and are thereby
enabled to take action on the basis of well-informed reasoning and information (Strömbäck,
2005: 332). When political advertising mimics editorial content in form and placement, it
begs the question of whether it can harm the credibility of news media and journalism.
Previous research and hypotheses
Native advertising is part of a rapid development. Despite previous research on similar
formats such as advertorials, there is a profound lack of research on the actual effects of
native advertising (and similar formats of content marketing) on citizens’ trust in political
news. Some notable exceptions are Cole and Greer (2013), as they investigated how so-called
brand journalism, a type of content marketing that ‘allows businesses to target customers with
useful, tailored editorial content while promoting their brand, values, and products’ (Cole and
Greer, 2013: 1), influenced the credibility of the advertiser and the message. In a similar vein,
Boerman et al. (2012), Ashley and Tuten (2015) and van Reijmersdal (2011) examined
various effects of content marketing and product placements in terms of the ads’ ability to
persuade readers. Cramer (2015) further examined the effects of native advertisements’
quality and content-relevance on readers’ perception of content quality.
As part of the Digital News Report by the Reuters Institute (Newman et al., 2015),
Austin and Newman (2015) employed an online survey to ask US and UK citizens about
attitudes towards native advertising and found that 33% of the respondents in the UK, and
43% in the US have felt disappointed or deceived after they read an article they later found
out was sponsored (Austin and Newman, 2015). They also found that the majority of the
respondents in both the UK and the US believed that ‘sponsored content has neither a
positive nor negative impact on either the brand in question or the news organization that
carries the content’ (Austin and Newman, 2015). Tutaj and Van Reijmersdal (2012) compared
native ads to banner ads and studied the effects on a news site’s credibility, thereby providing
an important starting point for further research on effects on journalistic credibility. However,
considering that the native ad attempts to pass for normal editorial content, it is essential to
compare audiences’ levels of trust between native ads and editorial content. To the best of our
knowledge, empirical research has yet to comparatively examine this in an experimental
setting. In this paper, we empirically outline the consequences for audiences’ trust in political
news by introducing a political party as an advertiser in the native format. Through a unique
experimental design embedded in an online representative survey, we test how clearly such
native advertisements should be labelled in order for readers to recognize that it is an ad and
whether exposure to it reduces readers’ trust in regular political news.
The abovementioned debate, paired with our observation of the increasingly changing
boundaries on two fronts and the apparent lack of knowledge in the field, leads us to
formulate some hypotheses. We are interested in the question of whether people understand
the difference between a native advertisement and editorial content: do people ‘spot’ the often
quite discrete labelling that should reveal to the reader what kind of content they are being
exposed to? This question leads us to the following (H1 and H2 also serve as manipulation
checks to see whether our manipulations perform as intended):
H1: People are more likely to perceive a story as an advertisement if it states (either
explicitly or less explicitly) that it is an ad compared to a story featuring only editorial
H2: People are more likely to perceive a story as an advertisement if it explicitly states
that it is an ad as well as explicitly stating the source of the message compared to a
less explicit labelling of ‘sponsored content’.
We assume that blurred lines between commercial and editorial spaces in journalistic
platforms combined with the introduction of political actors as sponsors could undermine the
audiences’ trust in political news (as argued above). Given a scenario in which after exposure
to a single political native ad one is faced with the thought of a news organization being paid
to print favourable political stories attempting to pass as independent journalism, one could
expect people to experience a drop in trust towards political news in general. We base these
assumptions on theories of priming effects: short time changes in standards for evaluations
(Iyengar and Kinder, 1987: 63). Priming is a memory-based model of information processing.
It assumes that people form attitudes based on the considerations that are most salient and
accessible when they make decisions (Hastie and Park, 1986). If people evaluate and express
an attitude towards political news in general directly after being exposed to a political native
ad, priming theory dictates that they will incorporate the most salient and readily available
information upon evaluating. This should include the impression that a newspaper is
attempting to trick them into believing that an ad is independent journalism. We consequently
assume the following:
H3: People subjected to a native advertisement from a political party are less inclined
than people subjected to editorial content to trust political news in general.
Experiments embedded in online surveys are increasingly employed to assess the effects of
different media messages. Such experimental settings reach a larger audience than the
traditional lab-experiment. To test our hypotheses, we embedded an experimental design
within a larger representative online national survey conducted by the Norwegian Citizen
Panel in March of 2015. The panel’s respondents were gathered through the postal
recruitment of 25,000 individuals over 18 years. The individuals were randomly selected for
recruitment from Norway’s National Registry: a list of all individuals who either are or have
been a resident in Norway, maintained by the official Tax Administration. The recruitment
rate was 20%. The current study is a part of the fourth wave of the Norwegian Citizen Panel
(N=6297, panel response rate= 69 %). A randomly selected subsample of 743 respondents
were presented with our experimental design. Ten of these respondents did not answer one or
more of the questions in the present study, leaving a total of 733 respondents that participated
in and completed the study. Gender, age and education biases in the response rate were low
(Skjervheim and Høgestøl, 2015). Demographically, 58% of our respondents had higher
education (bachelor level and above), 49% were female and the median age was between 46
and 55 years.
Experimental design and procedure
As displayed in Figure 1, the experimental setup consisted of exposing the participants to two
news articles that were supposedly published on the online version of Aftenposten.no –
considered to be a trusted news outlet in Norway: 67% of the respondents reported ‘high’ or
‘very high’ confidence in Aftenposten in 2014 (Ivarsflaten et al., 2015). First, the participants
read a story about a political party surging in the polls. Second, they read a story about the
Norwegian state budget. We measured two key dependent variables: to what degree the
participants regarded article 1 as editorial or commercial content and to what degree the
participants trusted the type of news in article 2.
To isolate the effects of exposure to native advertising and degree of how prominently
a native ad is labelled, we randomly assigned the respondents to one out of three conditions.
We performed group comparisons and found no significant differences between the
experimental conditions with regard to gender, age, education, political interest, general trust
in the media, attitude towards the Labour Party leader and attitude towards the Labour Party.
We manipulated the introduction of the first news story with a 1 X 3 (no labelling, less
explicit labelling or explicit labelling of ‘sponsored content’) between-subjects factorial
design. The control group, or ‘no labelling’ condition, featured the introduction of a story
about a political party leader (the party leader of the Labour Party, Jonas Gahr Støre) being
favoured as the next prime minister in an opinion poll – thus being regular editorial content.
In the ‘explicit labelling’ condition, the article featured the same story as the control condition
but with the following label above the story: ‘[t]his article is a commercial paid for and
written by the political party the Labour Party’. In the ‘less explicit’ condition, the article was
labelled with ‘sponsored content’.
Not wanting to risk deceiving participants, we explicitly stated that the two articles were
hypothetical. However, the articles were designed to resemble actual and realistic news stories
about political polls and the consequences of policy priorities (state budget), respectively. We
exposed individuals to one of these manipulations with the following introduction and
question: ‘This example shows the introduction to a hypothetical article which could appear in
online Norwegian papers. Based on the information provided in this article, do you consider
this to be a typical example of a news article (written by a journalist) or a typical example of
an advertisement (written by an advertiser)?’ This was measured with the following five-point
scale: 5 (This is a typical news article), 4 (Looks more like a news article), 3 (This is just as
much a news article as it is an advertisement), 2 (Looks more like an advertisement) to 1
(This is a typical advertisement). This measure was used to assess whether the manipulations
were successful and whether the degree of labelling matters (for a similar approach, see Kim
et al., 2001: 272).
Next, we exposed all the respondents to the second news story, without any
manipulations (thus being editorial content), with the headline reading ‘This is how the new
state budget will affect your life’. We immediately measured their trust in such political news
in general with the following introduction and question: ‘This is also an example which shows
the introduction to a hypothetical article which could appear in online Norwegian papers.
How much confidence do you have in the information provided in articles of this type?’. This
was measured on a scale from 1 (No confidence at all) to 5 (Very high confidence).
We initially analysed the differences as to the degree participants perceived article 1 as an ad
or as editorial content. Figure 2 displays the mean differences between the three groups in the
experiment, with 95% confidence intervals. Figure 2 provides the answers to the question
regarding whether people perceived the article as an advertisement or editorial content –
showing that the mean answers in the editorial condition and those in the two native
advertisement conditions (explicit labelling and less explicit labelling) are pushed in the
expected, opposing directions. This not only indicates that people are more likely to perceive
a story as an advertisement if it is labelled as an ad (H1), but it also points to the fact that the
manipulations were successful. This is due to the results showing that the control conditions
are more likely to consider the stimulus material as editorial content, and the treatment groups
are more likely to consider the stimulus material as an ad. These results are robust if we
include sociodemographic variables as controls. This leads us to confirm H1.
Turning to H2, we expected that people are more likely to perceive a story as an
advertisement if it is explicitly labelled rather than less explicitly labelled. Figure 2 further
shows that although the differences between the treatment groups point in the expected
direction, the differences between the two native advertising groups are insignificant (t(507) =
1.71, p = .088). This leads us to reject H2, assuming that people are not more likely to
perceive a story as an advertisement if it explicitly states the ad sponsor rather than a more
Table 1. Effects of exposure to native advertising on trust in political news in general
Editorial content (no ad)
Attitude towards the Labour Party (the advertiser)
General trust in the media
Note: Coefficients are unstandardized beta coefficients with standard errors in parentheses. Dependent variable
is general trust in political news. Attitude towards the Labour Party is coded on a scale from 1(extremely dislike
the Labour Party) to 7(extremely like the Labour Party). General trust in the media is coded on a scale from
1(very high mistrust) to 7(very high trust). N=601 in model 2 due to some respondents not answering questions
about general trust in the media, education, and attitude towards the Labour Party.
* p<.05, ** p<.01, *** p<.001
Our third hypothesis addressed the possible effects of native advertising on citizens’
trust in political news. In general, our respondents assign some trust in political news content
(M= 2.85, SD=.029) on a scale from 1 (no confidence at all) to 5 (very high confidence).
Moreover, the trust in political news among the editorial versions of the articles was slightly
higher (M=2.92, SD=.81).
H3 holds that people subjected to native advertising (either explicitly labelled or less
explicitly) will be primed to assert less trust in political news content in general than those
subjected to editorial content. To investigate this, we regressed trust in political news in
general on dummy variables for each of the experimental conditions, using the ‘no labelling’
(editorial content) conditions as the reference category. Because the respondents were
randomly assigned to specific conditions, all other characteristics of the respondents are held
constant. Thus, all the differences in trust across the conditions in the two different articles
can be attributed to the differences in the labelling of sponsored content.
Model 1 in Table 1 displays the results of this OLS regression. The results in the first
model in Table 1 shows that native ads indeed yield a statistically significant yet small effect
(η2 = .011) on people’s trust in political news, given that the ad is explicitly sponsored by a
political party. These results were replicated and remained statistically significant when we
used ANOVA or ordered logistic regression instead of OLS regression. For ease of
interpretation, we concentrated on the latter.
The two models in Table 1 show that the effect of being subjected to native
advertising rather than editorial content points in the expected direction: exposure to an
explicitly labelled native ad reduces people’s trust in political news. As displayed in Model 2
in Table 1, these effects are robust if we include sociodemographic variables, attitudes
towards the Labour Party (the sponsor of the ad) and general trust in the media as controls.
Thus, H3 is supported. The effects of the less explicit version also point in the expected
direction of reduced trust but do not reach statistical significance compared to the editorial
condition (p=.45) or the explicit condition (p=.11).
This study is, to our knowledge, among the first to provide insights into the consequences of
native advertising on audiences’ trust in political news. Our participants clearly distinguished
between editorial content and native advertising (either explicitly labelled or non-explicitly).
We further predicted that participants exposed to native advertising (either explicitly labelled
or non-explicitly) would be primed to report less trust in the type of news displayed in article
2 compared to those exposed to the editorial version. We did indeed find clear, but small,
indications of such effects when the sponsor was explicitly mentioned in the native ad.
Returning to our first finding, a previous study suggests that people are more likely to
recognize traditional banner ads as commercial messages rather than native advertisements
(Howe and Teufel, 2014). In our case, people seem to recognize native advertisements as
commercial messages, at least when explicitly prompted to choose between the two
The fact that our respondents were able to discern between editorial and commercial
content has potential consequences for the boundary between commercial and editorial
content. This is an indication that our respondents were attentive news readers, aware of
labels identifying type of content, even when subtle. Our experimental design has a possible
limitation given that the abovementioned is only valid in a situation in which respondents are
actively prompted to evaluate and choose between the two conditions of ‘advertisement’ and
‘journalistic content’. Wojdynski and Evans (2015: 10) asked respondents more generally if
they had encountered any advertising while reading the stimulus. This stands in contrast to
our approach, pointing to the possible limitation that respondents become primed to evaluate
the condition of the stimulus material in a more active and attentive way than could otherwise
have been the case.
However, this finding can speak to the debate over the business-news-wall. If the
labelling of native advertisements performs as intended, the format could be less problematic
than initially thought, granted that it is labelled properly. In the wake of the Atlantic-Church
of Scientology controversy, there were a lot of assumptions of betrayal on behalf of the
audience. Concerns were voiced as to whether people were being tricked by the publication.
These assumptions arose uncritically (Carlson, 2014: 7). The results of the present paper can
speak to this observation. Our findings do not indicate that readers necessarily are being
duped. Understood in the terms of boundaries, clearly visible and explicit labelling could be
enough to maintain the division between editorial and commercial spaces in the eyes of the
audience. That said, native advertising could still pose a threat to journalism’s credibility.
This leads us to our most important finding: we identify a slight, yet statistically
significant reduction of trust in political news among respondents having been exposed to an
explicitly labelled native advertisement with an explicit partisan source. The result suggests
that political native advertising can reduce people’s trust in political news. This has
implications for the boundary between journalists and politicians and could trigger some
pessimism towards the future of audience trust in political journalism. Native advertising is
increasingly gaining acceptance, and political parties are already starting to experiment with
such new avenues of political marketing.
Given that we find that a single exposure to a political native advertisement reduces
trust in political news in general, there is cause for concern. Granted that the reduction in trust
is small, there is no cause for immediate alarm, but the very direction in which the trust is
headed is problematic. This is further accentuated because participants were prompted to
generalize from hypothetical stimuli to a real, generalized genre of news. One could assume
these effects to be even greater with a more specific dependent variable and in real-life
situations. Further research should deploy authentic stimuli to examine the effects on trust to
the same news story and the specific publication. As our experiment measures only a single
drop in trust based on a single exposure, longitudinal studies are needed. The effects of
political native advertising on trust could prove to be greater on a long-term basis after
repeated exposures and a gradual reduction of credibility over time.
Through the practice of political native advertising, audiences could gain the
impression that the professional journalistic ideology calling for a well-kept boundary
between journalists and those in power is merely a façade, or even impotent, as it can easily
be bypassed through payment. If similar results can be observed in different contexts, a bleak
picture is painted for the future of journalistic autonomy and the watchdog function of
journalism as a whole. The practice of political native advertising could be harmful, because it
diminishes one of journalism’s most important business assets: that people actually trust the
publication, the individual journalists, or journalism as a profession. If journalism ultimately
ends up with the same credibility and level of trust as advertising as a result of one imitating
the other, then the added revenue of native advertising could prove an overly costly venture.
There would not be much of a product left to sell without any credibility.
In sum, the abovementioned results can be interpreted both as reassuring and
unsettling. Shifting the boundary between editorial and commercial content could prove less
problematic as long as other boundary devices such as clear and understandable labelling is
put in place. On the other hand, political native ads labelled in a manner that people
understand seem to reduce trust in journalism. Given that journalism as an institution and a
profession is reliant on this trust to uphold its informational function as well as its watchdog-
role – newsrooms would do well to reconsider, or at least tread carefully around, the practice
of political native advertising until we have more knowledge of the phenomenon. One could
do worse than to seriously consider Cameron and Ju-Pak’s strong recommendation regarding
the parallel format of advertorial: to keep it distinct from editorial content to safeguard
credibility (Cameron and Ju-Pak, 2000).
Suggestions for future research
To fully grasp the consequences of native advertising, further inquiries are necessary.
Amongst others, we need more studies of how audiences respond to this form of
communication from the perspective of journalism research. Our study can serve as a starting
point in this regard. Even though our findings suggest that the political dimension of native
advertising should be examined more closely, other types, such as native brand advertising,
should also be the subject of further research. It is pertinent to reveal what differences exist
between political and commercial native advertising in terms of loss of journalism’s
credibility. A relevant question in this regard is whether one can expect the same effects of
trust reduction outside a political context, in the realm of ‘soft news’. Blurred lines between
travel journalism, consumer journalism or fashion journalism and their promotional
counterparts should be examined. Future research should also take into account that everyday
news consumption is not as neat and orderly as it is when simulated in an experimental
setting. How audiences experience native ads outside this setting could be a different story
and should be studied accordingly.
Aalberg T and Curran J (2012) Conclusion. In: Aalberg T and Curran J (eds) How Media
Inform Democracy: A Comparative Approach. New York: Routledge, 189–201.
Albarran AB (2010) The Media Economy. New York and London: Routledge.
Altheide DL and Snow RP (1991) Media Worlds in the Postjournalism Era. New Jersey:
Ashley C and Tuten T (2015) Creative strategies in social media marketing: An exploratory
study of branded social content and consumer engagement. Psychology & Marketing
Austin S and Newman N (2015) Attitudes to sponsored and branded content (native
advertising). University of Oxford: Reuters Institute Digital News Report 2015.
Baker EC (1994) Advertising and a Democratic Press, Princeton: Princeton University Press.
Boerman SC, van Reijmersdal EA and Neijens PC (2012) Sponsorship disclosure: Effects of
duration on persuasion knowledge and brand responses. Journal of Communication
Borud H (2015) Aftenposten lurer ingen – Flemmen lurer seg selv [Aftenposten does not fool
anyone – Flemmen is fooling himself]. Journalisten.no, 21 October. Available at:
Cameron GT (1994) Does publicity outperform advertising? An experimental test of the third-
party endorsement. Journal of Public Relations Research 6: 185–207.
Cameron GT and Curtin PA (1995) Tracing sources of information pollution: A survey and
experimental test of print media’s labeling policy for feature advertising. Journalism
& Mass Communication Quarterly 72: 178–189.
Cameron GT and Ju-Pak K-H (2000) Information pollution? Labeling and format of
advertorials. Newspaper Research Journal 21: 65–76.
Cameron GT, Ju-Pak K-H and Kim B-H (1996) Advertorials in magazines: Current use and
compliance with industry guidelines. Journalism & Mass Communication Quarterly
Carlson M (2014) When news sites go native: Redefining the advertising–editorial divide in
response to native advertising. Journalism: DOI: 10.1177/1464884914545441
Carlson M (2015) Introduction: The many boundaries of journalism. In: Carlson M and Lewis
SC (eds) Boundaries of Journalism: Professionalism, Practices and Participation.
New York: Routledge, 1–18.
Coddington M (2015) The wall becomes a curtain: Revisiting journalism’s news-buisness
boundary. In: Carlson M and Lewis SC (eds) Boundaries of Journalism. New York:
Cole JT and Greer JD (2013) Audience response to brand journalism: The effect of frame,
source, and involvement. Journalism & Mass Communication Quarterly. DOI:
Cook TE (2006) The news media as a political institution: Looking backward and looking
forward. Political Communication 23: 159–171.
Cooper CA and Nownes AJ (2004) Money well spent? An experimental investigation of the
effects of advertorials on citizen opinion. American Politics Research 32: 546–569.
Cramer H (2015) Effects of ad quality & content-relevance on perceived content quality. In:
proceedings of the 33rd annual ACM conference on human factors in computing
systems. Seoul, Republic of Korea: ACM, 2231–2234.
Deziel M (2014) Women inmates: Why the male model doesn’t work. Available at:
Dobbs D (2013) The Atlantic, Scientology and the theft of credibility. Wired, 16. January.
Available at: http://www.wired.com/2013/01/the-atlantic-scientology-and-stolen-
Eckman A and Lindlof T (2003) Negotiating the gray lines: An ethnographic case study of
organizational conflict between advertorials and news. Journalism Studies 4: 65–77.
Ellis J (2012) BuzzFeed adapts its branded content approach to political advertising, and
Obama’s in. Available at: http://www.niemanlab.org/2012/10/buzzfeed-adapts-its-
Ferrer R (2016) Camoflaging church as state: An exploratory study of journalism’s native
advertising. Journalism Studies. DOI: 10.1080/1461670x.2016.1165138
Flatø E (2015) Da Aftenposten startet reklamebyrå [When Aftenposten started an advertising
agency]. Morgenbladet, 16 October. Available at:
Hastie R and Park B (1986) The relationship between memory and judgment depends on
whether the judgment task is memory-based or on-line. Psychological Review 93:
Hjarvard S (2013) The Mediatization of Culture and Society. New York: Routledge.
Howe P and Teufel B (2014) Native advertising and digital natives: The effects of age and
advertisement format on news website credibility judgments. #ISOJ Journal 4(1): 78-
Ivarsflaten E, Tvinnereim E and Aars J (2015) Norsk Medborgerpanel [Norwegian Citizen
Panel] Round 4: Data file. The University of Bergen.
Iyengar S and Kinder DR (1987) News That Matters: Television and American Opinion.
Chicago: University of Chicago Press.
Jamieson KH (1992) Dirty Politics: Deception, Distraction and Democracy. Oxford: Oxford
Jamieson KH (2000) Everything You Think You Know About Politics... And Why You’re
Wrong. New York: Basic Books.
Kaye J and Quinn S (2010) Funding Journalism in the Digital Age. Business Models,
Strategies, Issues and Trends. New York: Peter Lang Publishing.
Kim B-H, Pasadeos Y and Barban A (2001) On the deceptive effectiveness of labeled and
unlabeled advertorial formats. Mass Communication and Society 4: 265–281.
Kohring M and Matthes J (2007) Trust in news media development and validation of a
multidimensional scale. Communication Research 34: 231–252.
Lengauer G, Esser F and Berganza R (2011) Negativity in political news: A review of
concepts, operationalizations and key findings. Journalism: DOI:
Matteo S and Dal Zotto C (2015) Native advertising, or how to stretch editorial to sponsored
content within a transmedia branding era. In: Siegert G, Förster K, Chan-Olmsted SM
and Ots, M (eds) Handbook of Media Branding. Springer International Publishing,
Newman N, Levy DA and Nielsen RK (2015) Reuters Institute Digital News Report 2015.
Reuters Institute for the Study of Journalism.
Nordlys.no (2015) Det handler om å lede [It’s About Leadership]. Available at:
Picard RG ((2002) 2011) The Economics and Financing of Media Companies. New York:
Fordham University Press.
Rodgers S (2007) Effects of sponsorship congruity on e-sponsors and e-newspapers.
Journalism & Mass Communication Quarterly 84: 24–39.
Schudson M (2011) The Sociology of News. New York: W.W. Norton.
Schudson M and Anderson C (2010) Objectivity, professionalism, and truth seeking in
journalism. In: Wahl-Jorgensen K and Hanitzsch T (eds) The Handbook of Journalism
Studies. New York: Routledge, 88–101.
Singer JB (2015) Out of bounds: Professional norms as boundary markers. In: Carlson M and
Lewis SC (eds) Boundaries of Journalism: Professionalism, Practices and
Participation. New York: Routledge, 21–36.
Sjøvaag H (2010) The reciprocity of journalism’s social contract: The political-philosophical
foundations of journalistic ideology. Journalism Studies 11: 874–888.
Skjervheim Ø and Høgestøl A (2015) Norwegian Citizen Panel 2015, Fourth Wave
Methodology Report. Ideas2Evidence. Available at:
Society of Professional Journalists (2014) Code of Ethics. Available at:
Stroud NJ (2011) Niche News: The Politics of News Choice. New York: Oxford University
Strömbäck J (2005) In search of a standard: Four models of democracy and their normative
implications for journalism. Journalism Studies 6: 331–345.
Syvertsen T, Enli G, Mjøs OJ and Moe, H(2014) The Media Welfare State: Nordic Media in
the Digital Era. Ann Arbor: The University of Michigan Press.
Tutaj K and van Reijmersdal EA (2012) Effects of online advertising format and persuasion
knowledge on audience reactions. Journal of Marketing Communications 18: 5–18.
van Reijmersdal EA (2009) Brand placement prominence: Good for memory! Bad for
attitudes? Journal of Advertising Research 49: 151–153.
van Reijmersdal EA (2011) Mixing advertising and editorial content in radio programmes:
Appreciation and recall of brand placements versus commercials. International
Journal of Advertising 30: 425–446.
van Reijmersdal EA, Neijens PC and Smit E (2005) Readers’ reactions to mixtures of
advertising and editorial content in magazines. Journal of Current Issues & Research
in Advertising 27: 39–53.
Wasserman E (2013) Advertising goes native, and deception runs free. The Huffington Post,
30 January. Available at: http://www.huffingtonpost.com/edward-wasserman/native-
Wojdynski BW and Evans NJ (2015) Going native: Effects of disclosure position and
language on the recognition and evaluation of online native advertising. Journal of
Advertising 45(2): 1–12.