THE ECONOMICS OF ONLINE NEWS AGGREGATION
AND NEIGHBOURING RIGHTS FOR NEWS PUBLISHERS
DRAFT, OCTOBER 2016
PLEASE DO NOT QUOTE OR CIRCULATE
Abstract
This paper takes an economic perspective on the copyright debate in online news. The creation of a
neighbouring right for news publishers follows the logic of the civil law approach to copyright that
emphasizes the right to remuneration for all forms of digital reproduction and communication.
Empirical evidence shows that news aggregators already have a positive impact on publishers'
advertising revenue. It is difficult to see how publishers could gain more revenue from a
neighbouring right, as the Spanish and German cases show. The economics of multi-sided platforms
may provide a better explanation for this situation than the alleged abuse of a dominant market
position by news aggregators. The common law view on copyright emphasizes the purpose of
copyright as a financial incentive to ensure production of news. While several factors have driven
the decline in newspaper revenues and in the number of legacy newspapers, news aggregation has
an opposite effect. There is no evidence that declining revenues reduced the production of news
articles as many new sources of digital news production have emerged. However, news aggregation
platforms may create risks for media plurality in news consumption, more so than in the supply of
news. Article rankings in platforms are driven by ad revenues. This weakens the editorial choices
and branding of newspaper publishers. Rankings might trigger superstar effects that reduce the
variety of news consumption. These risks cannot be addressed through copyright.
1 The author is
at the Joint Research Centre of the
European Commission
Parts of this paper were presented
The views and opinions expressed in this
paper are the author's and do not necessarily reflect those of the JRC or the European Commission.
1
link to page 1 link to page 2 link to page 4 link to page 6 link to page 10 link to page 13 link to page 16 link to page 19 link to page 22
Table of Contents
Abstract ............................................................................................................................................... 1
1.
Introduction ................................................................................................................................ 2
2.
The evolution of online news distribution models ..................................................................... 4
3.
The newspaper revenue or substitution debate ........................................................................ 6
4.
The economic view on copyright .............................................................................................. 11
5.
Market failures in online news publishing ................................................................................ 14
6.
News aggregators and media plurality ..................................................................................... 17
7.
Conclusions ............................................................................................................................... 20
Bibliography ...................................................................................................................................... 22
1.
Introduction
The shift from offline printed newspapers to online digital media has led to a decline in revenues for
legacy newspapers. Several factors have contributed to that decline (Cornia et al, 2016). Readers
moved their news consumption online and, in recent years, to mobile screens. This led to a steep
decline in print edition revenues. Newspapers are struggling to compensate this through online
subscriptions, trying to find a "freemium" balance between freely accessible and paywall protected
content. Advertising spending also moved online and is more dispersed since advertisers have a
wider choice of media channels to promote their products. Online ad revenue for newspapers only
partly compensates the loss in offline ad revenue. Up to 90% of revenues in most newspapers still
come from print, even after years of decline in print advertising and circulation and almost 20 years
of investment in digital media (Cornia et al, 2016, p 7-8). Newspapers experiment with paywalls,
native advertising and product branding, diversification into merchandising and e-commerce
activities, video advertising, etc. Others are experimenting with distribution through social media
and mobile platforms though some are reluctant and fear becoming too dependent. More sources
of news have moved online, including TV and radio broadcasters, new ad-driven online newspapers,
social media websites and blogs. They all compete for the attention of news consumers and drive
down the market price of news. Unbundling adds to that competitive pressure. While offline
readers bought a bundle of articles in a print edition, online readers can chose freely between
articles from a variety of online news sources.
In this thriving online news market, news aggregation platforms have emerged that combine short
snippets of news articles from many publishers, with links to the original full article. Readers are
attracted to these because they reduce search costs for a wider variety of sources of news. Some are
ad driven but Google News, the main operator, is ad-free. Recently, full text articles, combined with
2
ads, have emerged on the Facebook social media platform. The shift to mobile news reading adds
further complications for publishers. Small mobile screens, slow loading articles and the very short
attention spans of mobile users are a challenge. The Google News and Facebook platforms have
become major gateways to a wider online audience. Publishers feel uncomfortable with that
situation.
The contrast between thriving platforms and struggling news content publishers has led some
publishers to claim a larger share of the platform revenue cake, on top of the benefits from the
additional reach and ad traffic that news aggregators provide. Publishers want remuneration for the
copyrighted content that they provide to news aggregation platforms. That explains the origins of
the debate on ancillary or neighbouring rights for newspaper publishers.
Some EU MS have taken action. Germany introduced a neighbouring right in law in 2013 that
allowed publisher to charge websites for using their text, unless that text is reduced to very small
snippets. The law creates a situation where original news content publishers need to explicitly agree
that an aggregator or search engine can put snippets of news items on a website. Despite
expressing concern about Google sites and lobbying for the law, German news publishers eventually
opted in and the market situation did not change. In Spain, the government introduced legislation in
2014 that goes a step further and provides for mandatory payment of license fees by online news
aggregators through a collection society. Publishers are not free to opt-in to aggregation of their
content. As a result, Google withdrew its News service from Spain, though it keeps displaying news
article snippets in its search pages. In other countries courts intervened. After a court decision in a
dispute between Google Search and the Francophone Belgian newspapers, the case was settled in
2012 with an agreement for enhanced collaboration between Google and the newspapers to attract
more traffic to newspaper sites and to increase advertising revenue. A similar settlement was
reached in France between Google and newspaper publishers.
.
This paper aims to complement the legal approach taken in the copyright debate with an economic
perspective. It starts from the civil law perspective of copyright as an authors' right to remuneration
3
and examines the empirical evidence on the impact of online news aggregators on the revenue of
news publishers. It finds an overall positive impact on audience reach and ad revenue for newspaper
publishers. Moving to a common law or economic perspective on copyright as an instrument to
promote innovation in society, it finds no conclusive evidence that weak copyright has contributed
to a lack of incentives for new production.
A better
explanation for the fact that news publishers were unable to monetize their legal rights in Spain and
Germany may be found in the economics of multi-sided markets and the weak market position that
publishers have in online platforms. Last but not least the paper examines the impact of news
aggregators on media plurality. Aggregators may promote plurality because they reduce transactions
costs for consumers to access a wider variety of press sources and articles. On the other hand,
search rankings in aggregation platforms may trigger superstar effects and other distortions that
affect the consumption of a variety of news sources. Again, there is insufficient evidence on this.
This is not an issue that can be addressed through copyright reforms.
This paper is structured as follows. Section 2 provides a brief introduction to the rapidly evolving
online news markets and technologies. Section 3 addresses the question whether online news
aggregators add or divert revenue from newspapers' own online outlets. Section 4 goes beyond the
remuneration debate and takes a wider economic look at copyright protection and the impact on
innovation and social welfare. Section 5 explores potential market failures in online news
aggregation markets that may justify the intervention of competition authorities and regulators.
Section 6 investigates the impact of aggregators on media plurality in news production and
consumption. Section 7 concludes.
2.
The evolution of online news distribution models
Nearly all newspapers now have an online presence, either jointly with a print edition or online only.
Not all readers come directly to the website of a newspaper. Some get there by clicking on the
search results of a general search engine, others come via news aggregators or via links embedded in
social media or other webpages. There are different types of new aggregation business models.
They show only short snippets of news articles. Some display ads next to the articles or snippets and
share the revenue (Facebook, Yahoo, MSN); others do not display ads (Google News) and have no
direct revenue to share. Google News is a stripped-down search engine that displays only newspaper
article links for a particular country, without ads. Even with ad revenue sharing commercial
agreements fall short of a full copyright license.
Publishers pay attention to sources that direct traffic to their news sites and claim a share of the
revenue of these referral sites. For example news publishers argue that although Google puts no ads
on the News aggregator pages, it draws indirect traffic and revenue to the Google Search pages by
displaying links and snippets of news article text in the search ranking. That generates ad revenue for
Google through ads placed on the Search page. It is hard though to trace a direct link between
4
Google Search ads and clicking on search results for news articles. Users either click on the ad or on
the search result; they cannot do both.
Referral traffic led to a first court case in Belgium2 (2006), and later in France, where publishers
claimed remuneration for the use of their content in Google News, despite the fact that Google
News carries no ads. Google refused to pay and de-listed the newspapers from it Search and Google
News results. That caused a drop in traffic to newspaper websites and in ad revenue. Some of these
ads where placed through Google AdWords; Google also lost revenue as a consequence. Google and
the newspapers in Belgium and France finally came to an agreement whereby Google provided some
financial means, short of a license agreement, and helped newspapers to get more traffic to their
websites. Similar cases emerged in Spain and Germany. In these countries, law makers decided to
intervene and create an explicit neighbouring right for news publishers that enables them to claim
remuneration for the (partial) reproduction of their content in news aggregation sites. In Spain, the
law made remuneration mandatory; in German, publishers were left a choice to accept or reject
reproduction without remuneration. To date German and Spanish news publishers have been
unable to monetize these rights. After an initial rejection and chilling effect on traffic to their
websites, German publishers finally agreed. Spanish publisher did not have that option and suffered
a decline in traffic.
Recently, the shift in download speed in mobile phones has enabled consumers to shift news
reading from fixed computers to mobile, especially mobile users who are active in social media.
About 40-50% of users consume news through their mobile phone (Reuters, 2016). In some
countries, social media account for 75% of all news distribution activities, especially among the
younger generations (Reuters, 2016). Facebook is the clear market leader for social media news
consumption. Its newsfeed attracts a substantial volume of traffic. End-2015, Facebook introduced a
new service: Facebook Instant Articles (FBIA). It displays the full text of a news article, with ads in
the article. The articles are loaded onto Facebook servers and into the Facebook app environment.
That puts Facebook in a position to observe user behaviour and control ad revenue. Facebook claims
that users read more Instant Articles than ordinary articles and are less likely to abandon the article.
Instant Articles allows news publishers to retain 100% of the ad revenue on their own ads and 70%
of the revenue on Facebook-produced ads. In this way, newspaper publishers passing through
Facebook receive the same ad revenue as they get on their own websites, with the added advantage
of reaching a wider audience. That explains why hundreds of newspapers have already signed up for
this distribution channel since Facebook opened it up to all newspapers in May 2016. Ad revenue
from newsfeed now represents a large share of Facebook revenue growth (Nieborg, 2016).
However, there is no guarantee that Facebook will continue to offer such generous financial terms
to news publishers in the future.
Google responded to the challenge of FBIA by introducing Accelerated Mobile Pages (AMP). This is a
technology that improves the mobile user experience with faster loading pages in their mobile
browsers. That facilitates not only newspaper reading but also many other types of content. Today
there are over 650k mobile pages that use AMP in all sectors. Slow loading mobile pages were
diverting mobile traffic from browsers to apps, including search-directed traffic in browsers, and the
ad revenue that comes along with this traffic. Google has a strong position in advertising in
2 See https://en.wikinews.org/wiki/Google convicted in case brought by Belgian press
5
browsers; Facebook has a strong position in mobile advertising. News pages loaded through FBIA
don't allow Google ads to be displayed; they display Facebook ads – or ads produced by the news
publisher. By improving mobile page load speed, consumers will be more inclined to search and
load more pages in mobile browsers – and that brings more traffic and ad revenue back to Google.
Consumers will click more on ads if they are confident that they will be loaded fast; that improves ad
click-through rates. It boosts revenue for news publishers and many other content publishers on
mobile pages. Google AMP caches the news articles on a Google server but does not control
audiences and revenue. Ad revenue accrues to the ad publisher – where Google may also have a
stake through AdWords. Last but not least, AMP pages get a rank boost in mobile search (read:
Google Search), bringing mobile searchers quicker to faster pages.
The ad revenue battle continues on other fronts too. Ad blockers become a major risk for many ad-
driven online business models, including news publishers. Some news publishers have already
introduced a "no ads, no views" policy: they block access to free online news for users who have
installed ad blockers (Cornia et al, 2016).
3.
Newspaper revenue and the substitution debate
Legacy newspaper publishers have suffered a steep fall in revenue in the transition for print editions
to online news consumption. Online ad revenue has somewhat softened the fall but has not been
able to stop it and turn it around. Newspapers observe the thriving online social media business,
powered by ad revenue and partly driven by newsfeed produced by the publishers – and they want a
(larger) share of that revenue.
Whether online news aggregators have contributed to a decline – or an increase - in revenue for
news publishers is an empirical question. In this section we review the existing empirical evidence
and conclude that it has a positive effect.
The impact of online news aggregators on the revenue of original news publishers is the result of the
net effect of two opposing forces: a quantity and a substitution effect. The substitution effect
measures to what extent aggregators displace online traffic and reduce the revenue generated by
the publishers' own website. The quantity effect measures to what extent aggregators increase
revenue by re-directing traffic towards the publisher's own website. If the substitution effect
dominates, the original news publishers lose revenue; if the quantity effect dominates they gain
revenue. The net effect is an empirical question that cannot be settled by economic theory or legal
reasoning; only data can answer that question. In the remainder of this section we summarize the
available empirical evidence to date. The main finding is that the quantity effect dominates the
substitution effect. Aggregators are complementary rather than competing services to newspapers'
original websites. On balance, they direct additional traffic to news publishers' websites and thereby
increase rather than reduce their online revenue.
News aggregators offer snippets or small extracts of the original news article only. This may
stimulate the interest of readers to click through to the full article on the original publisher's
website. That increases traffic and ad revenue on the original newspaper site (the quantity effect).
6
Aggregators may also provide a re-worked and edited version of one or several original articles, for
instance in user-aggregated and blog aggregator sites (Isbell, 2010). That shifts more consumer
value to the aggregator site and may reduce users' interest to look for the original article (Delarocas
et al, 2015). The substitution effect depends on the difference in quality between the aggregator
service and the original service offered by the publisher. The wider the quality gap between the two
services the more likely that the quantity effect prevails. Aggregators often deliberately provide
incomplete services because they cater to the long tail of news readers with less interest in full
articles and a low willingness to pay for the full content. Tapping into that long tail may give original
news publishers new audiences that they would otherwise not reach. Stopping that unlicensed
distribution channel will not bring these customers back to fully licensed distribution websites.
Aggregators and original content producers play in slightly different markets characterised by
quantity and quality discrimination between different types of news readers. A more recent
generation of mobile service news aggregators, such as Facebook's Instant Articles and Google's
Accelerated Mobile Pages, offer the full news article for reading. In return, they offer the original
newspaper publisher between 70 and 100 per cent of the ad revenue generated by the article so
that there is no revenue substitution effect at all: the publisher receives the same ad revenue as on
his own website.
Chiou & Tucker (2011/2015) explore whether aggregation of content by a single platform
encourages users to "skim" content or to investigate in depth. The study uses data related to a
contract dispute in 2010 that led a major aggregator (Google News) to remove news articles from a
major content provider (Associated Press). These articles were typically shortened versions of stories
that appeared in a select number of AP-associated newspapers. They compare users' website visits
before and after this contract dispute relative to traffic on Yahoo! News, which continued to provide
Associated Press content during this period. In theory consumers may use platforms to scan the
extracts of content without clicking through to pursue more in-depth material ("scanning" or
substitution effect). On the other hand, consumers may use platforms to explore new material more
deeply ("traffic" or quantity effect). The authors find that after Associated Press content was
removed from Google News, fewer users subsequently visited news sites after navigating to Google
News relative to users who had used Yahoo! News. Over a seven-week period 80 million monthly
visits were lost to newspaper websites that carried AP content. They find evidence that the traffic or
quantity effect is large, as aggregators may guide users to new content. There is no evidence of a
scanning or substitution effect, as overall traffic to Google News and Yahoo! News remained
relatively comparable during that time period. Websites with stronger ties to AP suffered a drop in
traffic after the dispute.
Chiou & Tucker (2015) conclude that the "fair use" exemption relied on by aggregator sites in the US
is potentially less damaging to the original copyright holder than often thought. The decision to opt-
into (or out of) an aggregation platform should depend on whether the content provider is
considered high-quality or highly unusual. Both these characteristics appear to encourage users to
use the aggregator to explore content more deeply instead of scanning content. Horizontally or
vertically differentiated content from high-quality newspapers or from very local or specialized
themes newspapers are likely to benefit from news aggregator sites while more bland middle-
ground news producers may stand to lose. These findings are in line with earlier research by Bar-
Isac et al (2012) on the impact of search cost reductions on the distribution of online consumption.
Top performing products as well as long-tail rare products are likely to gain when search costs
7
diminish while the middle ground may lose out. New aggregators reduce search costs for consumers
and seem to have a similar impact of the distribution of news consumption.
Athey & Mobius (2012) study the introduction of a “local news” feature in Google News France in
late 2009. Users could enter their zip code and on all subsequent visits they see news from local
outlets prominently featured. The authors compare the news consumption of users who enable this
feature to the usage of a set of “control” users, over a 2-week and an 8-week period. Adoption of
this Google news feature leads to 26% increase in consumption of local news sites over a 2-week
period and Google News explains a substantial portion of that increase. There is a 5% increase in
direct navigation to local outlets (bypassing Google News altogether) and a 13% increase in clicks on
local outlets from the Google news home page. Over an 8 week period the local news treatment
effect is attenuated somewhat but remains high (over 14%). Over time the incremental local news
consumption derives primarily from increased utilization of Google News. There is a 12% increase in
the number of local outlets used. However, the Google local news feature cuts into the curation role
of newspapers. Users are sent directly to the article, bypassing the profitable home page of the
news outlet. They may subsequently read other articles in the outlet through following links they see
on the same page as the original article, and thus their browsing may never take them to the outlet’s
home page. Even though these results broadly support the hypothesis that news aggregators are
complements for local news outlets, it is important to emphasize that the impact on local news
outlets is mixed overall. Some outlets gain more than others and users spread their consumption
over a larger number of outlets. George & Hogendorn (2013) come to similar conclusions for the re-
design of Google local News in the US.
Delarocas et al (2015) conduct a series of field experiments about readers' allocation of attention
between news aggregator and the original articles they links to. They examine how the length of the
text snippet, the presence of images and the number of related articles on the same story, affect a
reader’s propensity to visit the content producer’s site and read the full article. Their findings
suggest that longer aggregator snippets reduce the probability that readers will read the full articles.
However when several related snippets compete for user attention, a longer snippet and the
inclusion of an image increase the probability that an article will be chosen over its competitors.
Huanga et al. (2013) assess the relationship between aggregators and news websites in Taiwan.
Newspapers worldwide often serve as content providers for news portals, but portals outperform
most newspaper sites in audience share. Based on empirical data collected through a large-scale
survey they find that aggregators do not compete with news websites, with the exception of Yahoo!
News. Lee & Chyi (2015) use a national survey of 1,143 U.S. Internet users to study the demand for
aggregators and other media outlets in the USA. They find a non-competitive complementary
relationship between three major news aggregators and 13 major TV, print and social media news
outlets. Google News, Yahoo! News, and Huffington Post do not compete with other media outlets.
At the request of the Spanish Association of Publishers of Periodical Publications NERA Consultants
(2015) assessed the impact of the introduction of a newspapers publishers' right in the Spanish
Copyright Act that became effective on 15 December 2014. It was promoted by a small group of
publishers despite opposition from many industry players. The revision of the Spanish law was
motivated both by the observation that aggregators benefit from the publishers’ efforts without
remunerating them, and the assumption that the substitution effect dominates the quantity effect.
8
The revised Article 32.2 of the Act established a compulsory fee to be paid by online news
aggregators to publishers for linking their content within their aggregation services. Unlike the
German case, Spanish publishers cannot opt out of receiving this fee and payments are to be made
through a copyright collecting society. This was justified to prevent a repetition of the German case
where the opt-in option led to a status quo. Using ComScore web traffic data, the NERA study found
a decrease in traffic to Spanish newspapers' websites by 6% on average during the first few months
of 2015, going up to 14% for small newspaper publications. Independently, Chartbeat, that tracks 50
Spanish newspaper sites ranging from small media outlets to the largest newspaper publishers,
confirmed that these sites saw their external traffic fall by 10-15 percent. Overall traffic did not fall
immediately as the amount of internal traffic coming from other Spanish newspapers rose. This
suggests that readers are moving more between Spanish news sites than previously, rather than
coming directly from links at Google News. This could be interpreted as an indication that user
search costs increased as the Google aggregator site stopped providing links. Users had to generate
their own news aggregation instead of Google News doing it for them.
Calzada & Gil (2016) use a more sophisticated dataset from SimilarWeb to study both the Spanish
and German cases. The data do not only track the volume of web traffic but also the sources that
generate the traffic (direct access to the news pages, aggregators, search engines, social media sites,
links from other web pages, etc.). They find an 11% drop in Spanish newspaper traffic (quantity
effect) and an 8% drop in the number of newspaper pages visited (a reduction in the variety of news)
following the change in the copyright law. The latter could be interpreted as evidence that news
aggregators increase access to a plurality of media resources rather than reduce it. More
surprisingly, the authors find a drop in search and direct traffic to newspaper sites by 12% and 14%
respectively. This is inconsistent with a substitution effect. It may show that users quickly grew
tired of the additional transaction costs to do their own news aggregation via search and direct
access to newspaper sites and simply reduced the number of news sources that they consulted.
That would once more confirm the variety or media plurality effect of news aggregators.
Aggregators may actually increase the number of direct visits to news outlets by allowing consumers
to rediscover new sources of information. For the German case they find a 7% reduction in daily
visits to news outlets controlled by the Axel Springer group that initially opted out of Google News
during a 2-week period. There was also a 10% reduction in visits via search engines and 7% direct
visits.
These empirical studies cover only the impact of news aggregators on traffic and revenue for the
original content publishers – the producer surplus. They do not examine the impact on the
production of original news content – the long-term dynamic effect of (weakened) copyright
protection on the supply of innovative products - or the impact on consumer welfare through
reduced search costs, availability of a wider variety of news sources and lower prices. Jeon & Nasr
(2015) are the first to explore these wider welfare implications in a theoretical model. They examine
how the presence of news aggregators affects the quality choices of newspapers, using an economic
model to capture the "business-stealing effect" and the "readership expansion effect" of the
aggregator. They find that the presence of the aggregator leads newspapers to specialize in news
coverage and changes quality choices from strategic substitutes to complements. This shift is likely
to increase the quality of newspapers and overall social welfare, though the effect on newspapers’
profits is ambiguous. Unfortunately there is no empirical evidence yet in support of this model.
9
We can conclude from this overview that there is no empirical evidence in support of the
substitution hypothesis and thus no evidence that online aggregators have a negative impact on
original newspaper publishers' revenue. On the contrary, aggregators may actually be complements
to newspaper websites and may help consumer discover more news and boost the number of visits.
This structured empirical evidence is corroborated by more qualitative and anecdotal evidence
collected through interviews with newspaper managers (Cornia et al, 2016). They confirm that
newspapers collaborate with large social media platforms like Facebook, Snapchat, Google, Yahoo
and Apple because they bring them in touch with large online audiences that channel a substantial
volume of traffic and ad revenue to their articles. News publishers may of course want to reach
further and claim a share of indirect ad revenue related to snippets of news text that appear in
general search engines – as Belgian and French newspaper publishers tried to do. The German
competition authority (BKarA, 2015) pointed out that indirect ad revenue via search is hard to verify
(see below). Rather than claiming copyright and a share in ad remuneration, the evidence (Delarocas
et al, 2015) suggests that reducing the length of snippets – with the help of courts - may be an easier
way to get there. It will not diminish the audience reach that search engines bring to newspapers
and it will boost the expansion effect by directing more search traffic to newspapers' own websites
and ads.
The decline in revenue in legacy news publishing started long before online news aggregation. It was
caused by a shift in readership and advertising from offline print editions to online news sites. It is
difficult for news publishers to find a good balance between free and paid online content. Many are
experimenting with the positioning of paywalls on their online news sites. Chiou and Tucker (2013)
find that the introduction of a pay wall leads to a 51% decrease in online visits and thus advertising
revenue. It is not clear how much of that loss would be compensated by an increase in subscription
revenue. Only news producers with a strong reputation and/or specialisation and a high price
elasticity of demand for their news items are in a position to put up a pay wall to increase revenue.
Moreover, moving online has resulted in unbundling of newspapers: while articles were sold as a
package in print editions, consumers have more choice in making their own selection of articles
online as they can freely move between different newspapers. Subscription fees limit that
unbundling and reduce competition between newspaper sites. Still, the volume of freely available
content is huge. Inevitably, the price of online content goes down as the volume of free supply
increases.
4.
The economic view on copyright
The strong focus on the revenue or substitution effect is a result of the Continental European civil
law approach to copyright that emphasizes the rights of authors and publishers ("
droit d'auteur") to
decide on reproduction of their works and communication to the public, as well as the remuneration
they want to receive for that use (Hugenholtz & Quintais, 2016). These rights have been put under
stress in the digital age with the invention of many new forms of reproduction and communication
to the public, including through hyperlinking and other forms of re-transmission and re-mixing,
accidental and transient copying, text and data mining (Poort, 2016) and the same is happening with
online news aggregation. The civil law tradition seeks to obtain remuneration for all digital
10
reproductions that are publicly available, subject to some exceptions that are justified by transaction
costs. Lawyers and courts are struggling to bring these new technologies in line with that right or
design extensions and modifications of copyright law to bring them into the fold of that
remuneration right paradigm.
The civil law approach can be contrasted with the common law view on copyright that is more
aligned with the economic view (for instance in Landes & Posner, 1989). The economic view sees
copyright as an economic policy instrument to stimulate innovation for the benefit of society as a
whole, not only for authors or private rights holders. The civil law and common law view coincide in
the argument that, without copyright protection, authors would have no incentive to invest in the
production of artwork because everybody else could benefit from that investment by paying only
the cost of a copy. That externality creates a market failure because it would stop production. The
common and civil law views start to split ways when it comes to the economic interpretation of
copyright as an exclusive monopoly right granted to the authors to decide who can use a copyright
protected work, under which conditions and at what price. For economists, monopoly rights are by
definition inefficient because they cause deadweight losses: potential benefits that cannot be
realized in society. However, these short-term or static losses in copyright protection are
compensated by longer-term dynamic gains because it produces a steady stream of innovative
artwork. Contrary to the Continental European civil law view that emphasizes the rights of authors,
at the expense of rights of consumers or society, the common law and the economic view seek a
balance between the rights of authors and the overall welfare of society. For that reason, copyright
is limited in time and scope. Exceptions to copyright protection determine circumstances where no
payment to the author is due. In the Continental view, these exceptions should be limited to
circumstances where transaction costs exceed the benefits of licensing for the rights holders
(Charles River Associates, 2014). In the economic view, exceptions can go further in order to ensure
an appropriate balance between the static losses and dynamic gains of copyright protection and
promote wider dissemination of the benefits of innovation in society.
Digital technologies create new means to reproduce and communicate copyright protected works to
the public. The civil law approach would seek to extend authors' rights to claim remuneration for all
these new uses. For example, the creation of a neighbouring right for newspaper publishers
constitutes a de jure extension of the scope of copyright protection. The common law or economic
view on copyright would not automatically grant an extension of authors' rights to remuneration for
uses under these new technologies. Extending copyright protection to all forms of digital
reproduction may defy the purpose of copyright as an instrument to promote innovation, not
necessarily to maximize revenue of the rights holders. Economists would carefully examine the
balance between static deadweight losses and dynamic welfare gains in the copyright equation.
They would go back to the purpose of copyright and ask whether these new technologies affect that
balance and reduce the incentive to produce innovation (Bechtold, 2016). A reduction in innovation
has a negative impact on societal welfare. In the case of newspapers for example, a reduction in the
number of new articles and in the variety of views expressed in news publications would constitute a
welfare loss that requires intervention through further copyright protection. However, if no such
negative effect on the supply of innovative artwork can be observed there is no a priori reason to
extend claims.
11
Figure 1: Social versus private benefits of copyright protection
The policy gap between the civil law legal doctrine and the common law and economic view is
illustrated in Figure 1. Too little protection stops innovation; that is bad both for authors and
society. More protection boosts benefits for authors and society. The civil and common law view go
hand in hand up to the optimal degree of protection P*. Beyond P* they split ways: more protection
still benefits authors but reduces the wider benefits to society because it slows down further
innovation. Consequently, in the economic view, the crucial test for an appropriate level of
copyright protection is not the impact on revenue for authors but the longer-term dynamic impact
on the supply of innovation for society.
Finding the optimal degree of protection P*, or the point where more protection becomes
inconsistent with wider societal benefits from innovation, is an empirical question that cannot be
settled through economic theory or legal reasoning. The impact of protection on the production of
innovation or the supply of news articles is a crucial - but difficult - test for the justification of a
possible ancillary copyright for newspaper publishers. German and Spanish news publishers did not
reduce the production of news articles when Google News refused to pay a fee. They are under
financial pressure because of other factors that affect newspaper revenues and that may affect the
production of newspapers and news articles. It would be an error of attribution to hold news
aggregators responsible for a decline in news output, if any, and give publishers additional copyright
protection against aggregators in order to boost output.
Proponents of the civil law legal doctrine often argue that revenue for rights holders and the supply
of innovation are positively correlated. In the digital age that is not necessarily true. The cost of
production and distribution of media products has fallen substantially (Waldfogel, 2013). In that
case, a fall in revenue does not necessary depress production. Music is a good example: revenue
from recorded music sales declined significantly with digitization in the in early 2000s but music
12
production increased (Aguiar & Waldfogel, 2014). A similar phenomenon may have occurred in
news. The vast increase in online news sources may be an indication.
There are no statistics on the production of newspaper articles. There are statistics on the evolution
of the number of newspapers and their circulation that show that both have declined (Cornia et al,
2O16. That does not necessarily mean that the supply of news has declined as many more new
sources of news have become available online: TV broadcasters have moved online, new fully digital
newspapers have emerged and a myriad of blogs comment on current events. Moreover, news
articles are a heterogeneous product market with some degree of substitution between articles.
Articles are not entirely unique products. They can cover the same event in different ways. The
absence of statistics on news production creates a gap in our understanding of the dynamics of the
newspaper industry in the digital age. It makes it difficult to judge whether there is a need to boost
news production through additional copyright protection or other means.
Before concluding this section it is worth pointing out that copying news items is widely practiced in
the digital age by all news media, including legacy newspapers publishers, radio and TV news
channels. This is well-documented in a study of French news media by Cage et al (2015) who
investigate the speed and modalities of online news dissemination. On average, it takes two hours
for information published by an online media outlet to be published on another news site, but less
than 45 minutes in half of the cases and less than 5 minutes in 25% of the cases. At least half of
online dissemination is copy-and-paste and does not follow rules for citing and crediting.
Information is costly to produce but cheap to reproduce. Copyright law does not protect facts, only
the particular expression of facts by an author in a news article. Some re-writing, reduction to short
excerpts or an appropriate reference to the original publisher might go a long way to satisfy the law
though it may not change the information content of the article. Automated online aggregators do
not re-write articles; that creates conflict with copyright law.
5.
Market failures in online news publishing
In order to understand the potential impact of the introduction of a neighbouring right it is worth
looking into the Spanish and German experience and why both failed to change the market
situation. The European Commission (2016) argues that inequality in bargaining power between
online news aggregator platforms and newspaper publishers contributed to this failure. In economic
terms this could be translated into an abuse of dominant position by the news aggregators, or a
market failure. If such a market failure is observed, this would justify government intervention in the
market, either through competition policy instrument or regulatory interventions – such as a reform
of the copyright law. The Spanish and German competition authorities addressed the issue of
possible abuse of a dominant market position by Google Search and Google News and the need for
intervention by the competition authorities or the regulator. Both come to a negative conclusion.
The Spanish competition authority (CNMC, 2014) issued an opinion that questions the economic
foundations for the introduction of an irrevocable neighbouring right for news publishers in the
Spanish intellectual property law. First, the CNMC observes that there are technical solutions (the
introduction of a robot.txt file in a web page, or paywalls) that newspapers can apply to avoid that
13
their articles are aggregated by online news platforms. This boils down to an opt-in/opt-out
procedure, similar to the German newspapers case. The fact that newspapers do not use this means
is an indication that the perceived benefits of aggregation exceed the costs – in line with the
evidence presented in the previous section. The introduction of "irrevocable" legal means to
achieve the same purpose would then be not only superfluous but could be harmful. Second, the
CNMC notes that it is inefficient to let collecting societies fix an a-priori price for the remuneration of
newspapers and not let the market work out a price. News aggregators generate traffic and revenue
for news publishers' websites but it is very difficult to determine the revenue that news aggregators
gain from specific news articles. Hence it would be difficult to work out the basis for the calculation
of compensation. The CNMC concludes that there is no indication of a market failure in news
aggregation because news publishers do not seem to actively oppose news aggregators.
Consequently, there is no reason for regulatory intervention in that market. Moreover, the
intervention through the modification of the Spanish intellectual property law creates new market
access barriers, both for existing news aggregators and for new service providers. The CNMC
recommends dropping at least the irrevocable character of the provisions and the intermediation of
copyright management societies. The CNMC report correctly predicted the failure of the Spanish
neighbouring right to generate any additional revenue for Spanish newspaper. Aggregators have no
incentive to pay more for the news articles that the market price which happens to be zero. By
making the remuneration irrevocable, the law disrupts the status-quo that emerged in the market
and thereby introduces a new market failure.
A similar situation occurred in Germany when a revocable neighbouring right was introduced. The
German case is more complex because if involves the Google Search engine too. Part of the debate
is about the length of the news snippets that Google can extract from the article and put in Search.
After the law came into force, VG Media, the collecting society for German news publishers
associated to the Springer group, introduced a tariff of 11% on all ad revenue derived from news
search and news aggregation activities in June 2014. The tariff was subsequently reduced to around
6% in October 2014. There was also a debate about the legality of the robot.txt protocol and what to
do about the newspapers that do not include such a protocol in their webpages. Google eventually
asked all publishers to sign a written contract with a zero price license under the "opt-in" clause, and
all of them did in order to allow Google News and Google Search to continue linking to their news
articles. The news publishers then filed a complained at the German competition authority about
the alleged abuse by Google of its dominant position in the German market. The competition
authority (Bundes Kartel Ambt, 2015) ruled that there was no need for action. Since the complaint
went beyond Google News that is not ad-driven and also covered news articles snippets ranked in
Google Search the BKarA examined the question from a multi-sided online platforms perspective.
The different sides of the market (newspaper publishers as content suppliers, consumers,
advertisers) cannot be considered as separate markets; they are linked by indirect network effects.
One side (advertisers) may subsidize the other side (consumers), thereby increasing the welfare of
all sides. Although the news publishers are not remunerated for their contribution to the platform,
the BKarA still considers this to be a market, though with a zero price. It also considered that the
"opt-in" procedure is a proportional measure that leaves newspapers free to decide whether they
want to participate in the market. It concluded that Google does not abuse its market position and
that there is no reason for intervention in this market. After the competition authority rejected their
14
claim, the publishers around VG Media filed a new complaint against Google in a civil court in
January 2016, claiming that Google owes them remuneration. This case is still pending.
Prices are settled in the market at the confluence of supply and demand. Market conditions may be
such that the equilibrium price turns out to be zero, even if the producer of the copyright protected
works has a clear and enforceable claim. Whether it is a "fair remuneration" is something that
economics cannot answer. The BKarA reached the conclusion that there is no market failure and
that zero can still be a market price because there are on-going transactions at that price. By
bringing the collecting societies as intermediaries into the market process, the legislators in Spain
and Germany have tried to increase this price to a positive non-zero figure. Collecting societies will
obviously not work for zero remuneration, partly because they are private companies that have to
cover their own operational costs and would like to earn a profit margin too. As the Spanish
competition authority pointed out, the collecting societies create additional costs in the functioning
of the market. That in itself is an inefficient solution because the market can clear at a lower cost, as
it currently does with the robot.txt protocols. Regulatory intervention in the market, implemented
via a collecting society, would make the market less efficient. If newspapers withdraw their consent,
consumers and advertisers will suffer. So far newspapers in Germany opt-in and there is no loss of
social welfare. In the Spanish case however, newspapers are legally not allowed to consent at a zero
price and the entire market disappeared as a result. Consumers, advertisers and news publishers
suffer a reduction in welfare as a consequence of the law.
This still leaves the question why news publishers could not monetize their newly acquired
neighbouring rights in Spain and Germany.
Market power in multi-sided platforms may be one possible answer. Aggregators are platforms or
multi-sided markets that bring different types of users together in a single market place: news
producers, consumers and advertisers (Caillaud & Jullien, 2003; Rochet & Tirole, 2006). Typically,
platforms benefit from direct (and indirect) network effects: the more users join the platform (on
one side of the market), the more want to join (on other sides). Platforms also benefit from
economies of scope in data aggregation (Martens, 2016). They can aggregate information across a
wide variety of products and preferences across many users. That information can be plugged into
algorithms in order to reduce the cost of matching different types of users: matching consumers
with news articles, matching advertisers with consumers, etc. Bigger datasets often increase the
efficiency of matching in very heterogeneous markets. The advantage for consumers is a reduction
in search costs and lower switching costs between different suppliers of news. The advantage for
news producers is access to a much larger potential audience than offline newspapers and individual
online websites can reach (Cornia et al, 2016). News aggregation platforms enable more efficient
matching between suppliers, consumers and advertisers, more efficient than individual newspaper
websites and print editions. As such, they provide a compelling value proposition to all user groups.
The flip side of the coin is that platforms seek to internalize part of externality benefits that they can
provide to users by driving a wedge between the interests of different user groups (Parker & Van
Alstyne, 2006). They maximise their own profits by eroding some of the potential benefits that users
could achieve. This can be done for example through pricing strategies and the manipulation of
search rankings (Ursu, 2015).
15
That ranking is a form of
news curation, similar to the role played by editors in legacy newspapers. Editors format and
position articles and ads to reflect editorial preferences, newspaper branding and commercial
interests. They decide to put some articles on the front pages or with bold headings that attract
more attention while others are moved to less eye-catching formats and places. News aggregator
platforms have no editorial line or brand reputation to defend; they mix articles from many
newspapers. They only pursue their commercial interests which are linked to traffic and ad revenue.
To the extent that they personalize rankings in function of consumer profile information, their
rankings will be to some extent constrained by consumer preferences. Trending news articles will
be pushed up the rankings and attract more expensive advertising. This drives a wedge between
consumer preferences, news content suppliers' attempts at building editorial profiles and branding,
and the commercial interests of the platform. Unlike in general search engines, such as Microsoft
Bing and Google Search, there are for the time being no news aggregation platforms with an organic
ranking of news articles, based on reader preferences and independent from advertising revenue.
There are two opposing approaches to rankings (Grimmelman, 2012, 2014). The "conduit" theory
sees search engines as a passive intermediary that makes an "objective" selection of relevant articles
in response to user preferences. The "editor" theory sees search as a subjectively curated ranking
of articles with the algorithm as an active editor. The editor view implies that there is no such thing
as search neutrality because any ranking represents the editor's view of what is best. Grimmelmann
(2014) reformulates these theories in terms of freedom of speech, a relevant metaphor in the
context of newspaper publishing: the conduit view lets newspaper publishers speak and the editor
view lets the news aggregator's ranking algorithm speak. Search results are necessarily a
combination of objective conduit and subjective editing. The ideal search engine – from a consumer
point of view – would be a "trusted advisor". It should not mislead the consumer and present results
17
that match his preferences. A search engine should fulfil that fiduciary role towards the user by "not
letting its own conflicts of interest shape the results; it should not deliberately return results it
knows not to be relevant; it must not misuse the search queries that is collects and must not conceal
important results about how it generates results" (Grimmelmann, 2014). In other words, he
reformulates the concept of search neutrality from the perspective of the consumer. Finally, he
notes that a problem with search rankings may be that they are insufficiently personalised and cater
to the average consumer only. Personalized rankings require access to more personal information,
beyond the historical record of previous queries. This creates a trade-off between privacy and
search efficiency. Can bias in news rankings be excluded? Supermarkets place products in ways that
maximize supermarket revenue. Newspaper editors select articles that maximize sales. We accept
these commercial strategies in offline markets but seem to be critical when they happen in online
markets.
Several studies (Ursu, 2015 for Expedia; Chen et al, 2015 for Uber; Fradkin, 2014 for AirBnB) have
found empirical confirmation for the view that search rankings in commercial platforms drive a
wedge between the interests of suppliers and consumers, to the benefit of the platform operator.
There are no empirical studies yet on news article rankings in aggregator platforms and how they
affect the interests of the publishers and news readers. Such studies would be very useful to assess
the impact of news aggregators on media plurality.
Even if an ideal news articles ranking could be defined, there is another problem. Rankings suffer
from an inherent problem called "superstar economics" (Rosen, 1981). Rankings have a top and a
tail end. Top-ranked products are called superstars. Empirical evidence shows that consumers have
a tendency to consume top-ranked products as a way of reducing search costs. However, search
rankings are endogenous and the direction of causality is not clear (Moraga-Gonzales et al, 2013):
Does higher ranking increase consumption or do more frequently consumed items rank higher?
Rankings may create a lock-in effect: popular products become more popular.
An important question is to what extent digitization and online search engines have favoured
superstars versus long tail sales. Building on the contribution of Johnson and Myatt (2006), Bar-Isaac
et al. (2012) address the relationship between search costs and sales concentration. Search engines
and aggregators reduce search costs. They conclude that long-tail and superstar effects arise
simultaneously, at the expense of the "middle" group of products. This increases the skewedness of
the sales distribution though not necessarily the ordinal ranking of products in a popularity-based
ranking. Lower search costs may also bring in more consumers. That may increase total demand but
also upset the relatively popularity of products and thus their rankings. There is a debate among
economists whether search engines and rankings increase or decrease the diversity of products
consumed. Much depends on the design of the search algorithm. This is an empirical question to
which there is no a-priori theoretical answer.
The Faustian deal that news aggregation platforms offer to publishers is a more important source of
concern than their impact on revenue. Platforms increase audience reach and ad revenue for
publishers; they reduce transaction costs for consumers. Neither party can refuse that deal because
there are no better alternatives. The flip side of the deal is the loss of independence for publishers
(Cornia et al, 2016). This may weaken their brand name as their content gets mixed with that of
other newspapers and consumers lose awareness of the newspaper that produced the article. They
18
lose editorial control over ranking that may be driven by superstar effects ("click baiting") and the
platform's drive to maximize revenue from ads. Platforms collect a wealth of data on news readers.
That enables them to do more targeted advertising. The news publishers have no access to these
data. That weakens their own attempts at branding and marketing of their publications and further
increases dependence on the aggregator platform.
Consumers gain in transaction costs but may lose because they get less of what they prefer and
more of what generates ad revenue for the platform. It is not clear if aggregator platforms lead to
more or to less variety in consumption. Economists have researched other media industries such as
music and books to find out if a reduction in information costs results in more or less variety in
consumption (for an overview of the literature, see Duch-Brown & Martens, 2014). There is no
empirical research on this question yet in digital news.
In any case, these issues cannot be addressed through copyright or neighbouring rights. In fact, it is
not clear if & how they can be addressed at all at this stage in our understanding of online news
media markets.
7.
Conclusions
The debate between newspaper publishers and online news aggregators is primarily driven by
concerns about the decline in revenue and the longer-term financial sustainability of the newspaper
industry as consumers and advertisers move from print editions to online news consumption.
Copyright is perceived as a tool that could contribute to improving that financial situation. The
introduction of a neighbouring right for news publishers would give a more solid legal footing to the
right to remuneration for the reproduction of text snippets from their articles and communicating
them to the public in general search engines and specialised news aggregation websites that
reproduce parts of or entire news articles. That is what the German and Spanish government
attempted and what the European Commission (2016) is now proposing at EU level. The Continental
European civil law perspective on copyright does indeed emphasize the right of authors to
remuneration for their works and the neighbouring right extends this to online publishing.
Whether they will be able to effectively monetize that right is not clear, as the German and Spanish
cases show. The law creates a right but market forces have effectively valued this right at zero. In
this paper we investigated if market failures are to blame for this situation. Based on the analysis of
the Spanish and German competition authorities there is evidence that abuse of dominant market
positions by news aggregators has led to this situation. On the contrary, regulatory intervention in
the market to determine the price of a license and channel the remuneration via collecting societies
might create a new market failure. Alternative explanations may be found in the mechanics of
(news aggregator) platform economics. Platforms benefit from network effects and economies of
scope in data aggregation across many users. This results in wider reach and lower transactions costs
that individual newspaper websites cannot match. Platform services pricing will be driven by the
price elasticities of different types of users, which may put newspapers in a weak position to
leverage the value of their content. We should heed Merges' (2008) warning against "legal
centrism", the belief that laws rather than market forces determine market outcomes.
19
There is no evidence that news aggregation platforms have a negative effect on revenue of
newspaper publishers. The decline in newspaper revenue is driven by other factors such as increased
competition between ubiquitous (free) online news channels, the shift in advertising revenue from
offline newspapers to a variety of online services and the unbundling of print editions into separate
news articles online. In fact, empirical research shows that newspapers benefit from news
aggregation platforms in terms of increased traffic to newspaper websites and more advertising
revenue. This has not entirely solved the question of ad revenue in general search. However the
bulk of online news consumption is shifting to mobile news in social media platforms. Recent
business models that focus on mobile news consumption share between 70 and 100% of ad revenue
with newspaper publishers. It is hard to see how newspaper publishers could get a better deal, with
or without neighbouring rights. Aggregators will never pay more to publishers than their own ad
earnings. The flip side of these improved reach and revenue prospects is that publishers become
more dependent on platforms as online distribution outlets. It weakens their brand image and their
control over marketing and data. This may be a cause for concern in the longer run.
From a common law or more economic perspective on copyright, the key issue is not newspaper
revenue but the supply of news articles. There is no empirical evidence that despite the fall in the
number of newspapers and in revenue, the production of news articles has decreased. The decline
in news distribution costs has led to new forms of news production and online distribution that
complement and compete with legacy news publishers. From that perspective, an extension of
copyright protection, for instance by means of neighbouring rights for publishers, might have a
negative impact on social welfare in society, including on the distribution of news and on innovation
in new production and publishing models. There are currently no statistics on the evolution of news
markets that currently permit us to judge if there is a policy gap between the actual and the optimal
level of copyright protection in news markets.
Another important dimension of the policy debate on online news aggregators is their impact on
media plurality and the diversity of news production and consumption. This subject has been
somewhat side-lined by the focus on copyright extensions as a tool to boost the revenue of
newspapers. While the evidence suggests that aggregators have increased overall revenue, it
remains silent on the distribution of that revenue across articles and newspapers. Aggregators may
promote diversity because they provide an overview of news across different sources. Consumers
who go to a newspaper website would get only one source. However, aggregators' algorithms
determine the ranking of articles and newspapers. This creates superstar and long-tail effects that
are inherent to search rankings. As such they become "speech engines" (Grimmelman, 2014) that
drive the composition of the basket of news articles that consumers access. We don't know enough
about how the shift from legacy newspaper editors, who also put their editorial and commercial
20
stamp on the ranking and formatting of articles and ads in newspapers, to algorithm-based ranking
of articles affects news consumption patterns.
21
Bibliography
Aguiar, L. and Walfogel, J. (2014) "Quality Predictability and the Welfare Benefits from New
Products: Evidence from the Digitization of Recorded Music", JRC Digital Economy working
paper.
Anderson, C W (2013) "What aggregators do: Towards a networked concept of journalistic expertise
in the digital age".
Athey S, Mobius M (2012) "The Impact of News Aggregators on Internet News Consumption: The
Case of Localization",
Bar-Isaac & Guillermo Caruana & Vicente Cunat, 2012. "Search, Design, and Market Structure,"
American Economic Review, American Economic Association, vol. 102(2), pages 1140-60,
April.
Bechtold, S. (2016) "Deconstructing copyright", mimeo.
Bundes Kartel Ambt (2015) "Google/VG Media u.a.; Auseinandersetzung um das
Leistungsschutzrecht des Presseverlegers", Entscheidung B6-126/14 nach §32c GWB, 8
September 2015.
Cage, Julia, Nicolas Herve, Marie-Luce Viaud, (2015) “The Production of Information in an Online
World”, NetInstitute.org working paper nr 2015-05.
Caillaud, Bernard and Bruno Jullien (2003): Chicken & Egg: Competition among Intermediation
Service Providers, Rand Journal of Economics 34, 309-328.
Calzada, J. and Gil, R. (2016) "What do news aggregators do? Evidence from Google News in Spain
and Germany", working paper.
Charles River Associates (2014) "Assessing copyright exceptions", report for the European
Commission (DG Markt).
Chen M, Sheldon, Ml (2015) uber surge pricing Dynamic Pricing in a Labor Market: Surge Pricing and
Flexible Work on the Uber Platform, December 11, 2015
Comision National de los Mercados y la Competencia (2014) "Propuesta referente a la modification
del articulo 32.2 del Proyecto de ley que modifica el texto refundido de la ley de propiedad
intelectual", report issued by the CNMC on 16 May 2014.
Cornia, A., Sehl, A, Nielsen, R. K. (2016) "Private sector media and digital news", a report by the
Reuters Institute Digital News Project.
Danbury, Richard (2016) "Is a publishers' right a good idea?", Centre for Intellectual Property and
Information Law, Cambridge University. Mimeo.
Delarocas, Sutanto, Calin and Palme (2015) "Attention Allocation in Information-Rich Environments:
The Case of News Aggregators", Management Science, December 2015.
Duch-Brown, N. and Martens, B (2014) "Search costs, information exchange and sales concentration
in the digital music industry", JRC Digital Economy working paper 2014.
Dunaway, J. (2016) "Mobile versus computer: implications for news audiences and outlets".
Shorenstein centre on media politics and public policy, Discussion paper August 2016.
European Commission (2015a) "Digital Single Market Strategy"
European Commission (2015b) "Copyright Commmunication"
European Commission (2016) "Copyright Directive"
22
Evans, David and Richard Schmalensee (2007): The Industrial Organization of Markets with Two-
sided Platforms, Competition Policy International 3, 151-179.
Fradkin, A. (2014). Search Frictions and the Design of Online Marketplaces. Retrieved
from http://andreyfradkin.com/assets/Fradkin JMP Sep2014.pdf
George, L and Hogendorn C. (2012) "Aggregators, search and the economics of new media
institutions", Information Economics and Policy, vol 24(1), pp 40-51.
George, L and Hogendorn C. (2013) "Local news online: aggregators, geo-targeting and the market
for local news", working paper.
Grimmelmann (2012) Some Skepticism About Search Neutrality, in The Next Digital Decade: Essays
on the Future of the Internet
Grimmelman (2014) Speech Engines, 98 Minnesota Law Review 868 (2014)
Gruenberger, M (2016) "The German news publishers’ ancillary right", slides presented at the
Amsterdam conference on neighbouring rights, organised by Institute for Information
Rights, University of Amsterdam.
Hargreaves I, Guibault, L, Handke, Ch, Martens, B and Valcke, P (2014) "Text and Data Mining:
Report from the expert group", European Commission DG RTD.
Hiller, R and Jin-Hyuk Kim (2015) "Online music sales, sales displacement and internet search:
evidence from YouTube", Center for the Analysis of Property rights and Innovation (CAPRI),
Publication 13-2.
Huanga S, Mengchieh J and I Chyi (2013) "Friend or foe? Examining the relationship between news
portals and newspaper sites in Taiwan" Chinese Journal of Communication, 2013, Vol. 6, No.
1, 1–17.
Hugenholtz, B and Quintais, J.P. (2016) "Towards a universal right of remuneration: legalizing the
nono-commercial use of works", IVIR University of Amsterdam, mimeo.
Isbell, K. (2010) "The Rise of the News Aggregator: Legal Implications and Best Practices", Berkman
Center, Research Publication No. 2010-10.
Jeon D and N Nasr (2015) "News Aggregators and Competition Among Newspapers in the Internet",
mimeo, Toulouse School of Economics.
Johnson and Myatt (2006) On the Simple Economics of Advertising, Marketing, and Product Design,
American Economic Review , vol. 96, no. 3, June 2006
Landes W and R Posner (1989) "An Economic Analysis of Copyright Law", the Journal of Legal
Studies, 1989, vol. 18, issue 2, pages 325-63
Lesley Chiou and Catherine Tucker (2013) "Paywalls and the demand for news", Information
Economics and Policy, 2013, 25 (2), 61{69.
Lesley Chiou and Catherine Tucker (2011/2015) "Content aggregation by platforms: the case of the
news media", NBER Working Paper nr 21404.
Lee & Chyi (2015) "The Rise of Online News Aggregators: Consumption and Competition",
International Journal on Media Management.
Merges, Robert (2008) "IP rights and technological platforms", UC Berkeley School of Law, mimeo.
Moraga-González, José-Luis & Petrikaite, Vaiva (2013) "Search Costs, Demand-Side Economies and
the Incentives to Merge under Bertrand Competition," CEPR Discussion Papers 9374, C.E.P.R.
Discussion Papers.
NERA Economic Consulting (2015) "Impacto del Nuevo Artículo 32.2 de la Ley de Propiedad
Intelectual. Informe para la Asociación Española de Editoriales de Publicaciones Periódicas
(AEEPP)", 9 de julio de 2015.
23
Parker, Geoffrey W. and Marshall van Alstyne (2005): Two-sided Network Effects: A Theory of
Information Product Design, Management Science 51, 1494-1504.
Poort, J. (2016) "Borderline exploitation of copyrighted works", IVIR University of Amsterdam,
mimeo.
Reuters (2016) "Digital News report".
Rochet, J.-C., & Tirole, J. (2003) "Platform Competition in Two-sided Markets", Journal of the
European Economic Association, 1(4), 990-1029. doi: 10.1162/154247603322493212
Rosen, Sherwin (1981) The Economics of Superstars, The American Economic Review Vol. 71, No. 5
(Dec., 1981), pp. 845-858
Triaille, Jean-Paul, J De Meeus A De Francquen (2014) "Study on the legal framework of text and
data mining (TDM)", De Wolf and Partners, report prepared for the European Commission,
March 2014.
Ursu R, (2015) The Power of Rankings: Quantifying the Effect of Rankings on Online Consumer
Search and Purchase Decisions (presented at Searle Conference on Internet Search and
Innovation 2015.
Weaver, Alexander Barrett (2013) "Aggravated with Aggregators: Can International Copyright Law
Help Save the News Room?", Emory International Law Review, Vol. 26, No. 2, 2013.
Westlund, O. (2015) "News consumption in an age of mobile media: Patterns, people, place, and
participation", Mobile Media & Communication, Vol 3(2), Sage Publications.
Xalabarder, R (2014) "The Remunerated Statutory Limitation for News Aggregation and Search
Engines Proposed by the Spanish Government; Its compliance with international and EU
law", Universidad Obrero de Catalunya, Internet Interdisciplinary Institute, working paper
2014-04.
24