well as, in some cases, detailed data on the characteristics of each user. Well-known
organizations such as comScore and Nielsen provide accurate audience figures. Media
companies that have instituted paywalls often generate databases of the characteristics
and the reading habits of their online audience.
9.3. MARKET STRUCTURE IN NEWSPAPERS AND MAGAZINES
We now discuss market structure in print media markets. The majority of this section is
devoted to an examination of newspaper markets, due to the interesting economic issues
that arise in this industry and the wealth of research in this area. As discussed in
Section 9.1, empirical research on the newspaper industry has focused on the United
States, and therefore our emphasis in this section will also be on market structure in
the US newspaper industry, although we will point out important differences between
US newspapers and those in other countries.
3
We direct readers to Chapter 1 for a more
theoretical treatment of issues related to the number of firms that markets can support.
We first briefly discuss magazine markets, since the magazine industry appears so
different from newspapers in terms of market structure. This is perhaps surprising given
that, in many regards, the two industries are similar. Both the newspaper and magazine
industries have subscription prices that are subsidized, or at least supported, by advertising
revenues. Moreover, both industries are characterized by high fixed costs and low mar-
ginal costs, which are usually favorable conditions for high concentration. But measures
of concentration depend, of course, on the definition of the relevant market, an often-
fraught issue in the Industrial Organization literature.
Unlike with newspapers, magazines tend not to have local markets defined by cities or
metro areas. This is probably a consequence of the fact that magazines’ subject matter
rarely deals with specific geographic areas, but is more often a general subject, such as
news, sports, health, fashion, etc. Consequently, advertisers in magazines tend to target
readers’ demographic characteristics, rather than their geographic location. Therefore, it
appears intuitive to define the relevant market for magazines at the national level. This
choice is reflected in empirical research in the magazine industry: see, for example,
Depken and Wilson (2004) and Oster and Scott Morton (2005) for the US, Kaiser
and Wright (2006)
for Germany, and Ferrari and Verboven (2012) for Belgium.
Newspapers, by contrast, vary considerably in their news coverage, their geographic
appeal, and their target audience. In North America, newspapers have historically been
local in nature, confining their coverage to events in the surrounding city or metropolitan
3
Research specific to other countries will be reviewed in this chapter as appropriate. Examples of research
on newspaper markets in other countries include the following examples: Australia (
Merrilees, 1983),
Belgium (
van Cayseele and Vanormelingen, 2009), Canada (Chandra and Collard-Wexler, 2009), Italy
(
Argentesi and Filistrucchi, 2007), Japan (Flath, 2012), the Netherlands (Filistrucchi et al, 2012), Sweden
(
Asplund et al., 2005, 2008), and the UK (Thompson, 1989).
409
Newspapers and Magazines
area. In these cases, the natural definition of a newspaper market is at the local level as
well. In other countries, though, newspapers are often better characterized as competing
at the national level. For example,
Thompson (1989) estimates a model of circulation and
advertising prices using a sample of 34 paid Sunday and morning papers in Britain and
Ireland.
Noam (2009) points out that countries such as the UK and Japan essentially have
national, rather than local, newspaper markets.
9.3.1 The Number of Newspapers in a Market
We first pose the question: How many newspapers can a market support? Defining the
market as a city or metropolitan area, which is generally appropriate in the case of the US,
the answer is usually no more than one. The number of cities in the United States that can
support multiple daily newspapers is fast declining, and has done so for years, as we dis-
cussed in
Section 9.2. In fact, the decline in the number of multi-newspaper cities was
most pronounced during the mid-twentieth century.
Rosse (1967) states “In 1923 and
1963, respectively, 38.7% and 3.4% of US cities with daily newspapers contained more
than one. If multiple firms could exist in 1923, why not today?”
4
On the same topic,
Rosse (1980a) states: “[T]here was only one chance in nine that a paper in 1923 did
not have face-to-face competition, while the odds were more than two out of three
in 1978.”
Dertouzos and Trautman (1990) wrote that, at the time, less than 1% of US
newspapers faced competition from a newspaper published in the same city.
In fact, numerous authors in the literature have pointed to decreasing competition in
newspaper markets, at various points of time and in different settings. As long as 50 years
ago,
Reddaway (1963), writing about UK newspapers, pointed to “ … the historic
process whereby the number of towns with more than one evening paper has steadily
dwindled – and the number with more than one independent evening newspaper has
dwindled even farther.” We will return, in
Section 9.3.2, to the specific question of
how large cities must be to sustain multiple newspapers.
We attempted to analyze data on the number of newspaper firms nationally, as well as
aggregate readership, across a sample of countries. Although there is excellent data on
newspapers in North America, from the Newspaper Associations of the US and Canada,
recent and reliable data for other countries is harder to acquire, and the year of the most
recent data varies. Moreover, since auditing standards are not necessarily the same across
countries, a cross-country comparison should be interpreted with caution. With these
caveats in mind,
Table 9.1 presents data for five countries on the number of newspapers
and total circulation, both in levels and adjusted for population.
4
Using more recent data from Editor & Publisher (2012), we estimate that only about 2.5% of US cities with
daily newspapers in 2012 had more than one, and this number drops further if we do not consider jointly
owned newspapers or those operating under joint operating agreements.
410
Handbook of Media Economics
The table shows that Japan is the clear leader in terms of total readership per capita.
Japan is, in fact, a curious outlier in the newspaper industry. It has the fewest newspapers,
adjusted for population, of any developed country. However, at the same time, it has the
highest readership per capita. This is explained by the extraordinary readership concen-
tration in Japanese newspapers; the two most popular newspapers in Japan—Yomiuri
Shinbun and Asahi Shinbun—are also the top two newspapers worldwide and each
command an audience of over 8 million readers daily.
In terms of the number of newspapers, the United States is the leader among this group
of countries, both in absolute terms and adjusted for population. This reflects the distinc-
tively local nature of the US newspaper industry. In fact, the large number of newspapers
in the United States has been the case almost since the founding of the country and was
even commented upon by Alexis de Tocqueville back in 1836. In Democracy in America,
de Tocqueville comments on the “enormous number of American newspapers.” In
Vol. 1, Part 2,
Chapter 3 he states: “[T]he number of periodicals and occasional publi-
cations in the United States exceeds all belief” and “in the United States, scarcely a hamlet
lacks its newspaper.” De Tocqueville (2004, Vol. 2, Part 2, Chapter 6) provides an
explanation based on the extent of decentralization of administration in the US, and
the relative power of local governments, rather than a centralized government, compared
to other countries. He states: “This bizarre multiplication of American newspapers has
more to do with the extraordinary subdivision of administrative power than the extensive
freedom of politics or the absolute independence of the press.”
9.3.2 The Decline of Newspapers
The number of daily newspapers in the US has declined from 1878 in 1940 to 1382 in
2011.
5
As recently as 1980 there were 1745 daily newspapers, implying a decline of over
20% in the last three decades or so. The nature of the industry has changed in other ways
too. For most of the twentieth century, the industry consisted mostly of evening news-
papers. In 1980, there were four times as many newspapers published in the evening than
in the morning. By the year 2000, however, roughly equal numbers of newspapers were
published in the morning as in the evening, and today there are more than twice as many
morning papers than evening papers. Morning papers have always had higher circulation
on average, but the disparity has grown over time and today the paid daily circulation of
morning papers is more than 10 times that of evening papers. Reasons may include a rise
in the popularity of evening news shows on television, a decline in the number of factory
jobs that let out workers in the mid-afternoon, and the migration of readers from central
cities to the suburbs where home delivery is harder (
FCC, 2011; Romeo et al., 2003).
In the last several years, a number of large US cities that earlier supported two daily
newspapers have seen one of them shut down—either entirely, or become weekly papers
5
Newspaper Association of America and Editor & Publisher International Yearbook.
411
Newspapers and Magazines
or else solely digital editions. These cities include Tucson, Denver, Baltimore,
Cincinnati, Seattle, and Albuquerque.
A number of smaller US cities that could earlier support a daily newspaper have also
lost their local papers in recent years. These include the major university towns of Ann
Arbor (MI) and Madison (WI), which would generally be considered to have the
population and demographics to support a newspaper. Starting in April 2012, the New
Orleans Times-Picayune did not publish a daily edition, making New Orleans the largest
American city without a daily newspaper; however, the newspaper resumed a daily print
edition in 2013. As of 2013, however, the major—and only—daily newspapers in Port-
land (OR), Cleveland (OH), and Newark (NJ) were moving toward reduced home
delivery or daily editions on less than 4 days a week, and increasing emphasis on their
online editions.
Noam (2009) presents data to establish the population sizes required to sustain various
levels of newspaper competition. The population size required for cities to generally be
assured of sustaining a daily newspaper was 100,000 in the year 2000, while this cutoff was
around 50,000 in 1980. Cities of a million or more could generally sustain three
newspapers in 2000, but similar-sized cities in 1980 could support more than five. These
numbers bear out at the local level the same trends that we illustrated nationally in
Section 9.2: the number of newspapers in America has simply not kept pace with the
growth of population. As a result, most newspaper readers today must live in a very large
city in order to be assured of a choice between local newspapers.
Berry and Waldfogel
(2010)
, which we discuss in more detail in the next section, also provide evidence that the
number of newspapers in a market does not increase linearly with population.
Circulation declines and newspaper closings have direct effects on other metrics in the
newspaper industry. As we discussed in
Section 9.2, advertising revenue in this industry
has plunged. While all categories of advertising have declined sharply, the biggest losses
have taken place in classified advertising, an area where local newspapers used to charge
monopoly rates for listings, but now face competition from cheap or free websites such as
Craigslist and Monster.com. Another measure of how much newspapers have declined
is in their employment: the number of full-time journalists at daily newspapers fell
from 57,000 in 1989 to around 41,000 in 2010 according to the American Society for
Newsroom Editors.
6
The empirical observation that newspapers have steadily declined in their number and
circulation over a period of decades has prompted research into the causes.
Bucklin et al.
(1989)
study predation in the newspaper industry. The stated motivation for their analysis
is the declining number of US cities that can support multiple newspapers, and they dis-
cuss conditions under which firms in a duopoly or triopoly may try to force rivals out of
the market. They stress that the high fixed costs in the newspaper industry—first-copy
6
See http://asne.org/. Similar data are provided by PEW: http://stateofthemedia.org/.
412 Handbook of Media Economics
costs are between 40% and 50% of total production costs according to research cited by
them—increase the payoff for predatory action by a duopolist.
Bucklin et al. also point out that the interdependence of advertising and circulation—
in other words, the two-sided nature of the industry, as we would describe it today—
amplifies the importance of newspaper output, since a decline in circulation hurts both
sources of newspaper revenues. Importantly, this makes it easier to financially ruin a rival,
since even a small decrease in a firm’s output can make it impractical to stay in business.
They conclude by predicting that the slide toward monopoly in US central-city news-
paper markets is inevitable. Indeed, their predictions have largely been borne out in the
24-year period since their article was written.
Rosse (1980a) discusses reasons for the decline of direct newspaper competition.
Among the many reasons he considers, two in particular stand out as interesting. First,
the rise of television and, especially, that medium’s more efficient role in the advertising
market has affected newspapers, particularly those in big cities. Second, Rosse describes
the effect on newspapers of the suburbanization of America in the post-war period. Rosse
suggests that this was a further hit to big-city papers, since there was no longer sufficient
market segmentation to support multiple newspapers in the same city. Further, suburban
newspapers faced a smaller, more homogeneous audience, and thus there was little need
for multiple newspapers in these areas.
The increased pace of newspaper shutdowns in recent years is no doubt due to stiff
competition from online sources. This competition affects newspapers both due to a loss
of readers, which then directly lowers advertising revenue, but also due to a loss of clas-
sified advertising, which is traditionally an extremely profitable revenue source for print
newspapers.
Kroft and Pope (2014) document that entry by the classified website
Craigslist has directly reduced the amount of classified advertising in print newspapers,
and
Seamans and Zhu (2014) show that these effects then propagate to the other sides
of the newspaper—circulation and display advertising.
7
We discuss the effects of the
Internet on print media in more detail in
Section 9.6.
The decline of newspapers in the US has prompted policymakers to consider changes
to help the industry. The most prominent example is the Newspaper Preservation Act of
1970, which allowed the formation of JOAs. We discuss the literature on JOAs in more
detail in
Section 9.5.
9.3.3 Economies of Scale in Newspaper Markets
A number of studies have pointed to the importance of economies of scale in newspaper
markets, and suggested that these explain the high concentration in this industry.
Dertouzos and Trautman (1990) show that there are significant economies of scale in
7
See Chapter 12 for a related discussion.
413Newspapers and Magazines
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.