In another paper on the US magazine industry, Depken (2004) uses the same data as
Depken and Wilson (2004) to show that both reader income and, in contrast to Koschat
and Putsis (2000a,b)
, age have positive effects on advertising rates.
While these studies use magazine data and readership characteristics data at the
magazine level,
Chandra (2009) uses zip-code level on circulation and readership char-
acteristics for US newspapers. He finds that newspapers that operate in a more compet-
itive environment charge lower cover prices but higher ad rates compared to similar
newspapers that face less competition. To explain his results, Chandra subsequently
shows that newspapers in more competitive markets are better able to segment readers
according to their location and demographics, thereby catering to a more homogeneous
“targeted” audience that is appreciated by advertisers.
Given that magazines are highly segmented across reader characteristics (one can
think of fashion magazines where it is sometimes hard to distinguish between advertise-
ments and content), and therefore almost arbitrarily targetable, it can be argued that
magazine readers may be more likely to appreciate advertisements than newspaper
readers since newspapers are targeted to a lesser extent and instead cater to a geograph-
ically segmented audience. At the same time, it may be the case that advertisements are
more easily skipped in newspapers, precisely because it can be difficult to distinguish ads
from content in magazines. However, studies that deal systematically with such differ-
ences do not currently exist.
9.6. ANTITRUST ISSUES IN NEWSPAPERS AND MAGAZINES
Issues related to antitrust and market power in print media deserve a special discussion. As
this section will show, numerous authors have noted that the antitrust economics of
media, and of two-sided markets in general, have unexpected or counter-intuitive
features.
Evans (2002) points out that, in two-sided industries, an analysis of market
definition and market power that focuses on a single side will be misleading. As
Rysman (2009) notes, two-sided markets generally exhibit network effects and are there-
fore liable to tip toward a single dominant platform, which makes these markets of
interest to competition authorities.
In addition, the common feature exhibited by media, such as newspapers, of setting
price below marginal cost on one side of the market can lead to surprising policy pre-
scriptions with regard to mergers and market concentration. As with other two-sided
industries, mergers in these markets can theoretically raise prices, for both sets of con-
sumers (
Evans and Schmalensee, 2012). This section provides an overview of research
concerning market power and mergers in print media. We direct the reader to
X-FKS on media mergers for a more general treatment. We will also discuss two
antitrust-related topics that are of special interest in the newspaper industry: joint oper-
ating agreements and vertical price restrictions.
427Newspapers and Magazines
The first issue we discuss is estimating market power in the newspaper industry.
Argentesi and Filistrucchi (2007), briefly discussed in Section 9.3, estimate a structural
model of newspaper demand by both sets of newspaper consumers. Their goal is to exam-
ine whether the observed pattern of prices in the Italian newspaper industry is consistent
with competitive behavior rather than with coordinated behavior, and they analyze this
issue separately for each side of the market while taking into account the two-sided nature
of the newspaper industry. As they point out, a naı
¨
ve examination of price elasticities on
one side of the market does not necessarily imply anything about the degree of market
power that firms enjoy. The authors specify a nested logit model of demand on the sub-
scriber side and a simple logit model on the advertiser side, for the four main national
newspapers in Italy. Two points are worth noting: first, they assume that readers are
neutral toward advertising and, second, they assume single-homing on both sides of
the market. While this assumption may be reasonable for readers, it is a definite simpli-
fication on the advertiser side, but is driven by data limitations. Finally, the authors specify
the supply side by modeling newspaper publishers as setting both prices simultaneously.
Argentesi and Filistrucchi estimate the implied markups that publishers would set
under four different scenarios, which correspond to competition or collusion on each
side of the market. They then compare these implied markups with estimates of the actual
markups that publishers set, based on data on newspapers’ revenues and costs. They con-
clude that the data are most likely to be consistent with competition on the advertising
side but collusion on the subscription side.
9.6.1 Mergers
Ownership consolidation and mergers are a particularly important topic in newspaper
markets. As with any industry, consolidation leads to concerns about higher prices,
and this is especially the case in a market such as the US newspaper industry, which
already tends toward local monopolies. But the newspaper industry also raises concerns
about the diversity of opinion, and as a result this matter is particularly controversial.
Anderson and McLaren (2012) state: “The controversy is both political and economic:
even if a media merger increases profit, it affects how well informed is the public and
hence political outcomes. This means that traditional IO merger analysis is inadequate
for media mergers, and until recently policy debates have been dominated by non-
economists.”
Nevertheless, in recent years there have been a number of studies that examine the
issue of newspaper mergers, both from the traditional Industrial Organization perspective
of prices, and from the issue of diversity of opinion.
George (2007) studies the effect of
ownership consolidation on the variety of topics covered by US daily newspapers. As
multi-product firms internalize business-stealing externalities, she points out that mergers
can lead owners to eliminate duplicative products and change the content of others.
428 Handbook of Media Economics
She measures the variety of topics covered by newspapers using Burelle’s Media
Directory, which provides data on the titles of newspaper staff. She examines the period
from 1993 to 2001, which saw a large number of newspaper acquisitions. Her results
show that a reduction in the number of newspaper owners in a market leads to an increase
in the degree of separation among the existing newspapers. Moreover, the aggregate
number of topics covered per market increases with ownership consolidation. Thus,
there is support for the notion that consolidation may actually benefit consumers by
increasing the variety of topics covered by daily newspapers. George also finds that
the increased ownership concentration did not reduce newspaper readership.
Chandra and Collard-Wexler (2009) also examine the issue of ownership consolida-
tion in newspapers. Their study focuses on the price effects of mergers among Canadian
newspapers, in contrast to the focus on content in
George (2007). They first develop a
Hotelling model of newspaper competition for readers and advertisers which shows that
joint ownership of newspapers has no clear effect on prices for either subscribers or adver-
tisers. A key feature of their model is that advertisers value not just the number of readers
at a given newspaper, but also their characteristics. Given heterogeneity in reader char-
acteristics, it is possible that in a duopoly equilibrium some readers provide a negative
value to the newspaper publishers. These readers are the least desirable from the point
of view of advertisers, yet continue to enjoy the per-reader subsidy that newspapers
implicitly provide by setting price below marginal cost on the subscription side. Thus,
duopoly newspaper firms may end up setting higher prices in equilibrium, in order to
try to screen out these undesirable readers. However, under joint ownership of these
newspapers, prices will fall because the monopolist will internalize the effect of high
prices on both newspapers, in an analog of the traditional Hotelling model where joint
ownership raises prices since the marginal consumer provides positive value to firms.
They also show that advertising prices will move in the same direction as subscription
prices, i.e., the effect on advertising prices is ambiguous as well.
Chandra and Collard-Wexler then empirically examine the price effects of ownership
consolidation, relying on a series of newspaper mergers in Canada in the late 1990s, when
about 75% of Canada’s daily newspapers changed hands. They find that ownership con-
solidation had no discernible effect on either circulation or advertising prices.
Fan (2013) develops a structural model of the newspaper industry to analyze the wel-
fare consequences of newspaper mergers. Her paper accounts not just for post-merger
price changes, but also for newspapers adjusting their product characteristics. In addition,
she generalizes the model of demand for newspapers by allowing households to purchase
at most two daily newspapers, in contrast to most previous work, which assumed
single-homing on the subscription side. She uses county-level circulation data on US
newspapers between 1997 and 2005.
Fan uses the structural estimates to perform counterfactual simulations. In particular,
she examines a proposed merger in the Minneapolis market that was blocked by the
429Newspapers and Magazines
Department of Justice. She shows that an analysis of reader surplus that only focuses on
price effects, and ignores changes to newspaper quality, understates the loss in consumer
welfare. Both newspapers will raise prices post-merger, but Fan’s analysis of endogenous
product characteristics shows that they will also reduce product quality, which then fur-
ther reduces circulation and reader welfare. Advertiser welfare also falls by more when
product characteristics are endogenized. Not surprisingly, the surplus captured by news-
paper publishers is higher when they are permitted to adjust newspaper quality. In addi-
tion, Fan also simulates the effects of newspaper mergers in all US markets with two or
three daily newspapers, obtaining results similar to the specific case of the Minneapolis
market.
Filistrucchi et al. (2012) examine a hypothetical merger in the Dutch newspaper
industry. They point out that in a number of recent newspaper mergers, competition
authorities in various European jurisdictions have analyzed either a single side of the mar-
ket, or each of the two sides separately, instead of incorporating the feedback between the
two sides. On the subscription side, Filistrucchi et al. model consumer demand as a dif-
ferentiated products discrete-choice problem. On the advertising side, similar to
Rysman
(2004)
and other prior studies, the authors assume that the decision to advertise in any
given newspaper is separable from advertising decisions at other publications. They
model the quantity of advertising demanded at each newspaper as a function of the adver-
tising price per reader, acknowledging that this variable is endogenous. As an instrument,
they use the number of content pages in the newspaper, reasoning that content affects
total subscriptions, and hence the advertising price per reader, but should not otherwise
affect advertising demand.
They use a method laid out in a compa nion paper, and re cover estimates of news-
paper publishers’ marginal costs. Somewhat surprisingly, their results suggest that
newspaper publishers make positive margins on the r eadership side, and in fact higher
margins than on the advertising side. Their results suggest that readers attach a positive
value to newspape r a dvertising. Their m ain contribution comes from the analysis of a
hypothetical merger in the Dutch newspapermarket.Theirresultssuggestthatsuch
a merger would n ot directly af fect advertising prices. However, it would raise sub-
scription prices, and the resulting loss of subscribers would reduce advertising demand
and would also raise advertising prices per reader; nevert heless, the estimated effects
are small.
9.6.2 Newspaper Joint Operating Agreements
A unique feature of newspaper markets in the United States is the JOA. As discussed in
Section 9.1, this is a consequence of the Newspaper Preservation Act of 1970, which
endeavored to preserve the diversity of newspaper voices.
Romeo et al. (2003) explain
the rationale behind the JOA:
430 Handbook of Media Economics
“Under the protected JOA arrangement, two previously competing papers maintain separate news
gathering, news reporting, and other editorial functions while combining their advertising and
circulation functions: a single entity sells subscriptions to both papers and sells advertising in both
papers.”
Gentzkow et al. (2012) point out that newspaper JOAs are one of three instruments that
policymakers have employed to increase ideological diversity, the other two being limits
on joint ownership, and explicit subsidies.
Thus, JOAs are intended to permit a diversity of opinion and news in markets that
would otherwise only be able to sustain a single newspaper, and publishers are given a
special exemption from antitrust laws to allow them to combine their advertising and
circulation operations.
What are the welfare effects of this policy? Current research on this issue is divided.
Gentzkow et al. (2012) argue that allowing newspapers to form JOAs leads to a rise in
both economic surplus as well as diversity. They show that allowing newspapers to col-
lude on circulation prices alone leads to inefficient outcomes because the rise in news-
paper profits does not offset the loss of surplus to consumers and advertisers, and also
reduces the share of households who read diverse papers. By contrast, allowing papers
to collude on advertising prices increases both economic welfare and diversity. This is
because, in this situation, publishers slash circulation prices in order to increase readership
and thereby profit in the advertising market in which they now have substantially greater
market power. Even though publishers now have an incentive to differentiate from com-
petitors, the effect is weak, and many more households end up reading diverse papers. In a
JOA, where publishers coordinate prices on both sides of the market, the advertising
effect dominates.
Antonielli and Filistrucchi (2012), by contrast, question the rationale for JOAs. Sim-
ilar to
Gabszewicz et al. (2001, 2002), they allow publishers to first choose their political
position and then advertising and circulation prices. They analyze two forms of newspa-
per collusion: In the first, newspapers are allowed to jointly set prices on both sides of the
market and also cooperate on their editorial lines. In the second, they can cooperate on
prices but not on political position, which is exactly the situation with JOAs. The authors
find that editorial lines converge much more in the latter situation, using reasoning sim-
ilar to
George (2007). Antonielli and Filistrucchi conclude, therefore, that the logic of
JOAs is self-defeating. A possible explanation of their result is that they model newspaper
readers as single-homers. By contrast,
Gentzkow et al. (2012) argue that the multi-
homing of readers is a critical component in their finding that newspaper JOAs raise over-
all welfare.
Romeo et al. (2003) focus exclusively on the advertising side of the market in their
examination of the economic consequences of JOAs. They point out that, since JOAs are
intended to allow once competing newspapers to combine their advertising operations,
the effect should normally be to raise advertising prices to monopoly levels. However,
431Newspapers and Magazines
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.