View metadata, citation and similar papers at core.ac.

uk brought to you by CORE

provided by ScholarSpace at University of Hawai'i at Manoa
Proceedings of the 50th Hawaii International Conference on System Sciences | 2017

The Political Economy of New Media Revisited:

Platformisation, Mediatisation, and the Politics of Algorithms
Elizabeth Van Couvering
Karlstad University
[email protected]

Abstract technologies and the increasing prominence of the

This paper defines media platforms in terms of the “platform” as the new model of media which crosses
theory of traditional two-sided media markets, then categories between different businesses like Facebook
goes on to develop the theory to include content and Google and media forms as different as radio and
providers as a third side of the market (the video gaming. In tandem with studies on platform
“platformisation of media”), the widespread economics and the meaning of platforms has come a
introduction of sellable meta-information about the renewed interest in algorithms and software studies as
platform network (the “mediatisation of platforms”), a way to approach modern digital content
and the importance of social networking technologies infrastructures. This paper analyses the structure of
to the media platform. Issues of concern raised by this online media platforms in order to answer the
include privacy, intellectual property, and equity. following questions:
Genres of resistance to unwanted visibility and
invisibility, such as spoofing, spamming, fingering and • What can an economic view of platforms
silencing are also noted. contribute to our understanding of platform
media? And, conversely, how does an analysis of
platform media contribute to our general
1. Background understanding of other platform businesses?

It is increasing difficult to distinguish any actual It could be argued that instead of platform media
analytical category that separates “retailers” such as the term algorithmic media [7-9] might be used.
Amazon eBay and Apple from “search engines” like However, although an empirical focus on algorithms is
Google or Bing or “social media” such as Facebook warranted, it is not (or not entirely) the presence of
and Twitter or even networked services such as Uber digitized logic that gives these markets their character.
or AirBnB. Nearly all of these companies include Platforms existed prior to digitized algorithms – good
features for searching content; for saving profiles; for examples are credit cards and newspapers. Yet,
recommending, rating, and commenting on content; for algorithms make the functions of the platform much
contributing content; for marking other profiles or more efficient and available. A secondary question
content as favoured; and for locating and grouping that this paper will seek to answer, therefore, is which
audiences and targeting them with content. Obviously algorithms are central to platform media.
the scale, scope, and emphasis of these features is in As platforms develop and algorithms increase in
each case is different, but it is not a stretch to argue both power and complexity, what is at stake is
that they all represent instances of a similar meaningful public oversight over these developments
phenomenon. They are, by and large, distributors who and their implications for society. Concerns about
do not create content but who acquire and redistribute privacy, content ownership, and cultural diversity are
it, and who sell access to and information about the often handled on a case-by-case basis, but the analysis
people who read, write, play, buy and connect using in this paper seems to show systemic features that need
their services, while at the same time attracting those to be carefully examined by academics and regulators.
people by giving them tools to find their way in a sea First I outline what I mean by a media platform. I
of content, products, and people. then describe media platforms from both a technical
Work has been done on the political economy of and an economic perspective, using the concept of a
new media [1] and on some genres within it [2-6]. multi-sided market in order to understand the relations
Nevertheless, chief among the changes to the wider between different interest groups. I focus on the
Web in the last decade has been the rise of social centrality of social networking technologies as

1
URI: http://hdl.handle.net/10125/41374
ISBN: 978-0-9981331-0-2
CC-BY-NC-ND 1812
essential means of control of the platform and its area in which exchanges are coordinated and in which
dimensions, and discuss the centrality of surveillance the presence of the platform gives additional benefits
and of algorithms of measurement, ranking, and sorting the parties in the exchange [12]. A market with such a
for platform businesses. I illustrate these mechanisms platform is also called a two-sided or sometimes multi-
with examples from current media. sided market (or sometimes, multi-sided business or
multi-sided platform) [13].
2. Platform Economics and Media In these type of markets, the platform creates value
by bringing disparate groups together (for example,
Platforms readers and advertisers) and allowing them to interact
in such a way that both parties are better off
As literature on multi-sided and two-sided markets (advertising messages reach readers, and readers better
has developed within the field of economics, often understand the world and are entertained). Rochet and
using media businesses as examples, media scholars Tirole (who won the Nobel prize in 2014 for his work
have also been attempting to grapple with what the on understanding and regulating these markets) give
“platformisation” of media might mean for issues of several media examples to illustrate their definition of
content diversity, privacy, freedom of expression, and two-sided markets in their key 2003 paper: video
the relation between free and paid labour. Mansell games (with consumers and software developers as the
argues strongly that new empirical evidence is needed two sides), streaming media (consumers and servers),
because “[t]he structural features of ‘platformised’ browsers (users and web servers), portals (“eyeballs”
markets are continuously changing in ways that are not (quotation marks in the original paper) and
captured in the assumptions in economists’ theoretical advertisers), newspapers (readers and advertisers) and
models” [10]. TV networks (viewers and advertisers) [12].
The “platform” is an increasingly common type of According to Rochet and Tirole, the platform sets
online organizational form. Online platforms include the price structure and price level, and a key strategic
search engines like Google, retailers like Amazon and or competitive goal is to ensure that both sides of the
eBay, content providers like Wikipedia, and social market sign up to the platform. In order to ensure this,
networks like Facebook, Twitter, Instagram, as well as the platform may – and often does – structure prices so
some parts of more traditional software and hardware that one side of the market operates at a loss while the
manufacturers like Microsoft (Bing) and Apple (iTunes other side generates all the profit of the platform. To
and App Store). Online platforms are central to what take an example from modern newspapers, the profit of
might be called the content or media infrastructure of a free newspaper (like Metro or The Evening Standard
the Web – that is to say, the infrastructure through in the UK) is what it charges its advertisers, less what
which digital content and media, both privately and it costs to “pay” its readers in news and entertainment.
industrially created, are distributed throughout the In a free newspaper the readers are entirely subsidized
Web. Companies based on the platform model, such as by the advertisers, and one side subsidizing another
Google and Facebook, rather than the giant content- side is a common structure for two-sided markets.
creating companies (such as TimeWarner or Disney),
have emerged as the distinctive new media industrial
form of the digital era.
The word platform is ambiguous and multimodal.
Gillespie [11] highlights the discursive aspects and
rhetorical functions of the platform, citing its
computational, architectural, figurative and political
meanings. Implicit in all these definitions is the
platform as a “raised level surface,” which is “designed
to facilitate some activity that will subsequently take
place,” [11]. The definition of platform, according to
Gillespie, implies a place in which all are equal, visible
and assisted in fulfilling their aims. As Gillespie
describes, platform-based companies use this definition
Figure 1. Two-sided (traditional) media market
strategically to deflect criticism and to elide differences
(newspaper example).
between different constituencies.
One aspect not taken up by Gillespie is the
Why do advertisers subsidise readers? In fact, they
increasing use of the word platform within the
only do so in publications whose readership is includes
economics literature. As with the other meanings of
enough of the advertiser’s customers to make it worth
platform, within economics the platform is a central
their while. In a multi-sided market, demand for the

1813
platform (and relatedly, the price it can charge) is, in to the platformisation of the media (see also Manovich,
large part, determined by how many participants are who discusses this separation as a key property of
present on each side of the market. The more readers media software [16]). The addition of content
of the right type, the more attractive the paper is to providers as a third side of the market in turn implies
advertisers. In economic language, when the benefit of that indirect network effects in terms of strengthening
a network expands with its size, it is called a network demand from consumers, and therefore advertisers,
effect. When the benefits of the network size reach will apply to platforms that supply a great amount or
non-participants in the network – so the size of the variety of content (as suggested by the success of
readership benefits another group, the advertisers, it is content aggregators such as NetFlix, iTunes and
an indirect network effect. In traditional media Google).
markets, generally it is only the advertiser side that
increases demand due to indirect network effects [13].
Platforms exercise power in this example through
setting prices on both sides –in the example of Metro, a
price of zero to readers – which serves to attract and
retain enough customers on one side (the reader side)
for network effects to kick in and attract the other side
(the advertisers).
Multi-sided markets with platforms exist in many
businesses. For example, the literature contains
references to credit and debit cards [12-14],
manufacturing [15], and services like real estate [12].
Yet, very many examples in the literature relate to
classic media markets such as newspapers and
television, newer media such as video games and
streaming media, and internet media such as search and
social media, and other types of computer software and
operating systems. In other words, media markets
have been seen as platforms or multi-sided businesses Figure 2. Three-sided new media market
since economists began to develop theories about these (search engine example).
markets.
The separation of content from distribution is not a
2.1 The Platformisation of Media Businesses new trend in media, (one can think, for example, of the
ever-increasing trend for freelance journalists within
the field of journalism), but the development of
If traditional media types such as newspapers,
computerized content management systems has
radio, and television are already included in the
enabled this separation in such a way that it represents
definition of a multi-sided or platform market, what
a qualitative and not just a quantitative shift in media
then might it mean for media to be “platformised,” as
business.
Mansell [10] suggests? There are (at least) three
dynamics at play in this platformisation process.
The first dynamic is the continuing separation of 2.2 The Mediatisation of Platform Businesses
media content from media distribution methods (in the
1990s this was referred to as convergence) so that The second dynamic might perhaps be termed the
content provision becomes a third side of the market. mediatisation of the platform. This is the development
In our previous example from Rochet & Tirole’s of meta-information about the platform as a key part of
original descriptions, “newspapers” were the platform the platform business, for re-sale (for example, to
and “readers” and “advertisers” were the sides of the advertisers) as an add-on or additional product. In this
market. Today’s media markets are perhaps better way, non-media platform companies begin to resemble
understood if we instead took the analogy that paper media. For existing media businesses, the collection of
was the platform and the news organization or information about their users, content providers, and,
freelance writers contributed news content. It’s easy to to a lesser extent, advertisers, is extended in the
see this structure with Google, which has advertisers “mediatized” model.
and searchers looking for content provided by web site An example of a platform business which is in the
developers, or with iTunes, where listeners find content process of mediatisation is Uber, a platform-based taxi
provided by record labels and individual artists. The business with passengers and drivers as the two sides
abstraction of content from delivery method is central of the market. It collects and analyses information

1814
about its network of passengers and drivers and their development by Kane et al [20], but with more weight
behavior, and has begun to use this information in its given to content-related features since content is a key
business, sometimes for PR purposes and sometimes side of any media market, and the relationship between
for revenue [17]. In the diagram below, the platform content and profile is deep and complex.
graphic is changed to a cylinder form, which Although they are commonly used on the
conventionally denotes a database, to indicate the “consumer” side of media markets, social technologies
presence of this meta-information. may be incorporated, in addition, into the content
provider or developer sides of the market. It is also
possible to incorporate these tools into the advertiser
side as well, though this is much less common.
In social network-enabled platforms, a different
network effect, called a direct network effect, is
created: the number of users on the platform directly
benefits the user, since the more people are connected,
the more they can interact with their friends. This
direct network effect stimulates demand on the
consumer side in addition to the indirect network
effects on advertising and content demand. Thus,
social technologies can increase market share
(consumers would like to be on the network where
most of their friends are) or stabilize a side of the
market that might otherwise be vulnerable to
competitive pressures (consumers will tend not to leave
that network if their friends do not also leave). There
are some vulnerabilities, of course, for example the
Figure 3. Multi-sided media market with generational effect that leads young people to seek
mediatised platform. different media from their elders.
Once again, the focus on amassing information
about users is not a new trend within media (consider
the importance of Nielsen audience metrics within
television), or even within business in general, where
“relationship management” theories are built on
customer databases, but within modern platforms these
trends are made much more prominent and intense
using digital technologies.

2.2 Social Networking Technologies and

Platform Media

Evident in practice, but not much mentioned by in

the literature on platform economics is the frequent Figure 4. Mediatised platform with a social
introduction of social networking technologies to one network enabled among users.
or more sides of the market. Social networking
technologies contribute to the development of the size Social technologies serve other functions, as well,
of the platform, the separation of content from for the mediatised platform business in creating the
distribution, and the amassing of information all at the meta-information about the platform, and in
same time. strengthening demand from both users and advertisers.
By social networking technologies I mean On the profile side, they describe (and to some sense,
technologies which allow users firstly to create and create) the network topologically through “friend”
manage an identity and community, that is, to set up a links and structured profile information. On the
profile and to find, view, link to, and traverse other content side, they incorporate at least part of the third
profiles; and, secondly, to create and manage content: side of the platformised media market (content
to post and edit content, link to content, share content provision) by generating content themselves. On
with others, and rate or rank content. This definition Instagram, for example, users provide the photographic
partly follows Ellison & boyd [18; 19] and its content which other users browse; on Etsy, business

1815
owners provide discussions and tips for other business human intervention, institutions, and practices are key
owners. In addition, these technologies provide to the continuing function of these algorithms [8; 27].
information about quality of content on the platform However, the algorithms provide a useful focus for
(through rating and ranking functions such as likes on examining, evaluating and comparing media platform
Facebook or ratings on TripAdvisor or businesses.
recommendations on Amazon) and the structure of the The delivery of media content to the user can be
content (eg through playlists on YouTube or links handled by a browsing interface (for example, a
between patterns and yarns on the knitting network channel or electronic program guide), a searching
Ravelry). The use of social technologies in media interface (the query box), or a simple display interface
platforms is compelling for many reasons, although (featured programming). Most media platforms
their implementation and configuration varies. provide all three types of delivery. Users of the
network generally experience these information reveals
as guidance or navigational help.
3. Platforms and algorithms Algorithms for finding and displaying content
therefore include simple functions such as listing, but
It is evident that the mediatised and social-network- usually the list is classified or sorted, and, most often,
enabled platform that is the modern media business ranked. We might think here of Google’s search
must have a considerable technical infrastructure. Not engine results, but another interesting example of this
only must physical datacenters store and transmit the is Snapchat, which has a time-limited display
content and information produced by the different sides mechanism. Increasingly content may be visualized
of the market, but complex computer algorithms must (this is, presented in a non-traditional format, such as a
manage and process that data. The centrality of map or chart or other information graphic), and it can
algorithms to modern media businesses (and other also be personalized; that is, the part of the network
mediatised platform businesses, as in the earlier Uber that is revealed to you could be quite different than that
example), has been noted in the literature in the past revealed to your neighbor (for example, your feed on
few years, prompting discussion of an “algorithmic Facebook is different to anyone else’s).
turn” in media studies, and potentially of “algorithmic As discussed above, mediatised platforms also,
media” as the new paradigm [11; 21-24]. There have importantly, store and process information about the
been discussions of algorithmic cultures [25] and network itself, and these are the algorithms that
algorithmic epistemologies. constitute the second part of the algorithmic
Gillespie calls for an investigation of “public ecosystem. Specifically they store and process
relevance algorithms”. Napoli suggests, rightly, that information about users, about content, and about the
algorithms are probably best viewed as sociotechnical relationships between users and content. (Rarely, if
assemblages with a distinctly institutional character. ever, do they reveal to outsiders information about
This paper aims to contribute to the discussion, not by advertisers or relations between advertisers and users
suggesting how to investigate algorithms, but by or advertisers and content.) This requires algorithms
outlining which types of algorithms deserve the focus that trace and track user behavior. Initially these
of media scholars, based on the previous discussion of operated on page visits and clicks, traces of which
the economic features of platform businesses. were built into the very first Web servers in the form of
http:// requests and referral records. But increasingly
high value information comes in the form of links of
4. Types of media platform algorithms different types, both user-to-user (eg “follow” links on
Twitter) and content-to-content (e.g., hyperlinks, the
The core economic function of the platform is to first-exploited source of network information). Some
facilitate exchanges between the different sides of the links come with information about the type of link. For
market, as we have seen. In the media market, the example, on Flickr you may identify “friends” but also
central tripartite “exchange” has generally been the “family.” User-to-content relationships are established
delivery of content to the user and at the same time not just by page visits and clicks on a mediatised
user attention to the content provider and the advertiser platform, but also by user-evaluation mechanisms such
[26]. as likes, tags, recommendations, ratings and rankings.
The algorithms related to this exchange are one part Similarities in these user-to-content relationships are
of the algorithmic ecosystem of a media platform. In then often used to identify implicit user-to-user
this context, algorithmic ecosystem simply refers to a relationships or content-to-content relationships as the
complex and interdependent set of algorithms. This is topology of the network is more and more described.
not to say that these algorithms are unchanging or
independent from human intervention. On the contrary,

1816
 The meta-information collected by the mediatised Making other network members or network content
platform is then made available for searching and more visible may also be desirable to some. One may
display to the advertiser or other market participant. wish to indicate disagreement with the views or actions
These two classes of algorithms – search and of network members and fingering people or content
display, and tracking and tracing – are linked by a third with tags (for example disagreeing their posts or
set which matches, groups, and sorts content and posting a private exchange to a public space.) One
people dynamically. may wish to have better access to content by
downloading it for use outside of the network.
5. Platform media algorithms and politics Strategies for excluding or blocking unwanted
network members or content are also found. from
While the use of social networking technologies certain discussions, etc.
and mediatised platforms makes economic sense in .
many cases, the implications of these platforms are
troubling. At least three key areas of public concern 6. Conclusions
have been identified called into question with this new In this paper I have laid out the economic
economic structure. The first is privacy; the second is discussion about multi-sided platform markets, and
intellectual property; and the third is equity or fairness considered their implications for media platforms as
in representation. When viewed a certain way, each of we know them today. I have argued that media is
these issues of concern is about control of information becoming more “platformised” as more sides are
– how much should be shared, and with whom, and in introduced to the market, notably in the form of the
what way? These issues are central to the politics of abstraction of content production from content
media platforms because access to information and distribution. And, I have argued that platforms of
content on the network and about the network is the different types have become more “mediatised” as they
heart of the platform media business. increasingly store a layer of meta-information about
Clearly algorithms do not exist in a social and users and content for resale primarily to advertisers.
political void, and in fact struggles for the management Finally, I have argued for the importance of social
of the network away from its algorithmic base are networking technologies to the economic performance
common. Without access directly to the algorithms of of media platforms. Competitive logics within the
revelation, network members turn to different tactics to media platform sphere focus heavily on ensuring a
protect or promote themselves and others, materially critical mass of users that will attract advertisers, which
affecting, in some cases, the operation of the platform. in turn means either securing attractive (and ideally
Not everyone wants themselves or their actions to exclusive content), either from traditional content
be made visible. Criminals, for example, do not want providers or alternative from content generated within
to reveal their crimes; but neither do adulterers want to a social network. Secondly, the focus is on building up
show their relationships, nor employees their off-hours information about that network of users (and content
hobbies. Companies and institutions don’t want their providers, potentially) to sell on to others, primarily
internal discussions known. Dissenters may not wish advertisers.
to showcase unpopular political views. The right to The central algorithms of the platform media
privacy of both persons and corporate persons is ecosystem are firstly algorithms of search and display
enshrined in the UN laws. Unwanted visibility may be and secondly algorithms of tracking and tracing. A
combatted by a variety of techniques (other than third, meta-class of algorithms of matching and sorting
simply refusing to use the media platform): cloaking, links these first two classes. While these algorithms
or hiding your information; spoofing, or pretending to may be opaque and complex as a group, each smaller
be someone you are not. element of the class is potentially more available for
Conversely, some would like to be more visible study by social scientists.
than the algorithms of revelation show them. This The platform business creates revenue by
category could include aspiring celebrities; ideologues controlling access to content stored within the platform
who believe their views should be more widely spread; and information about the platform. Social issues of
or companies with something to sell; or content privacy, intellectual property, and equity are all
providers in the margins who want to move into the directly and immediately implicated in the
mainstream. Unwanted invisibility may be combatted algorithmically-made decisions about who can and
by spamming, or giving false information; or who cannot have access to the platform’s information
boostering, paying other network members for their and media content. Users, advertisers, and content
links, recommendations, ratings, tags, comments and providers all have adopted strategies to increase and
similar. decrease their own and other’s visibility and
invisibility within platform frameworks, and these

1817
issues will continue to be at the centre of platform
media design, management, regulation and governance
in the future.

7. Reference List

Handbook on International Antitrust Economics, R. Blair

[1] R. Mansell “Political economy, power and new and D. Sokol, eds, Oxford University Press, 2015.
media”, New Media and Society, 2004, pp. 74-83.
[14] D. Auer and N. Petit “Two-Sided Markets and
[2] M. Bauwens “The political economy of peer the Challenge of Turning Economic Theory into Antitrust
production”, CTheory, 2015, pp. 12-1/2005. Policy”, The Antitrust Bulletin, 2015, pp. 426-461.
DOI=10.1177/0003603X15607155.
[3] N. S. Cohen “The valorization of surveillance:
Towards a political economy of Facebook”, Democratic [15] A. Gawer “Bridging differing perspectives on
Communiqué, 2008, pp. 5-22. technological platforms: Toward an integrative
framework”, Research Policy, 2014, pp. 1239-1249.
[4] C. Fuchs “The political economy of privacy on DOI=10.1016/j.respol.2014.03.006.
Facebook”, Television & New Media, SAGE Publications,
2012, pp. 139-159. [16] Manovich, L., Software takes command,
Bloomsbury Academic, New York, NY, 2013.
[5] E. Van Couvering “New media? The political
economy of Internet search engines”, Citeseer, 2004, pp. [17] J. Cassano “How Uber profits even while its
25-30. drivers aren’t earning money”, Vice.com, 2016.

[6] T. Wall “The political economy of Internet [18] N. B. Ellison “Social network sites: Definition,
music radio”, Radio Journal: International Studies in history, and scholarship”, Journal of Computer Mediated
Broadcast & …, 2004. Communication, Wiley Online Library, 2007, pp. 210-230.
DOI=10.1111/j.1083-6101.2007.00393.x.
[7] F. R. McKelvey “Algorithmic media need
algorithmic methods: Why publics matter”, Canadian [19] N. B. Ellison and D. Boyd “Sociality through
Journal of Communication, 2014. social network sites”, The Oxford handbook of internet
studies, Oxford University Press Oxford, 2013, pp. 151-
[8] P. M. Napoli “Automated media: An 172.
institutional theory perspective on algorithmic media
production and consumption”, Communication Theory, [20] G. C. Kane, M. Alavi, G. J. Labianca, and S. P.
2014, pp. 340-360. DOI=doi:10.1111/comt.12039. Borgatti “What’s different about social media networks? A
framework and research agenda.”, MIS Quarterly, MIS
[9] P. M. Napoli “Social media and the public Quarterly, 2014, pp. 275-304.
interest: Governance of news platforms in the realm of
individual and algorithmic gatekeepers”, [21] S. Barocas, S. Hood, and M. Ziewitz “Governing
Telecommunications Policy, 2015. algorithms: A provocation piece”, Available at SSRN
2245322, 2013.
[10] R. Mansell “Platforms of power”, Intermedia,
International Institute of Communications, 2015, pp. 20-24. [22] R. Kitchin “Thinking critically about and
researching algorithms”, Information, Communication &
[11] T. Gillespie “The politics of ‘platforms’”, New Society, 2016, pp. 1-16.
Media & Society, 2010, pp. 347-364. DOI=10.1080/1369118X.2016.1154087.
DOI=10.1177/1461444809342738.
[23] M. Mahnke and Uprichard, E., “Algorithming
[12] J. C. Rochet and J. Tirole “Platform competition the Algorithm”, in Society of the Query Reader:
in two sided markets”, Journal of the European Economic Reflections on Web Search, R. König and M. Rasch, eds,
Association, Wiley Online Library, 2003, pp. 990-1029. 2014.
DOI=10.1162/154247603322493212/abstract.
[24] N. Seaver “Knowing algorithms”, Media in
[13] D. S. Evans and Schmalensee, R., “The Antitrust Transition, 2013, pp. 1-12.
Analysis of Multi-Sided Platform Businesses”, in Oxford

1818
[25] T. Striphas “Algorithmic culture”, European
Journal of Cultural Studies, 2015, pp. 395-412.
DOI=10.1177/1367549415577392.

[26] D. W. Smythe “Communications: blindspot of

western Marxism”, CTheory, 1977, pp. 1-27.

[27] E. Van Couvering “Is relevance relevant?

Market, science, and war: Discourses of search engine
quality”, Journal of Computer Mediated Communication,
Wiley Online Library, 2007, pp. 866-887.
DOI=10.1111/j.1083-6101.2007.00354.x.

8
1819