1

Mexico’s Media Concentration and Ownership in Global Comparison

Eli Noam Professor of Finance and Economics Garrett Professor of Public Policy and Business Responsibility Director, Columbia Institute for Tele-Information Columbia University Business School April 2014 Speaker’s Draft – Preliminary A. Introduction

In the information age, the diversity, openness, competitiveness, pluralism, and innovativeness of the means of communication are central to societies and economies. Therefore, the market structures of media industries is a central question. The media business is big business. Content media account for a combined $676 billion, about 1% of the world's GDP. Platform media accounted for $1.6 trillion, 2.3% of world GDP. Add media devices and media retailing, and the total rises to $3.6 trillion, or 5.9% of world GDP. Of discretionary spending, (i.e., excluding food, housing, medical, education, transportation to work), the media spending (including telecom connectivity), accounts for almost 20%. It is even higher in terms of discretionary time – beyond the time spent for work, sleep, and meals. In the US, average annual media consumption per person, as listed by the U.S. Census Bureau, has been measured to be an astonishing 3,545 hours per year for 2005-2009, not including time for e-mail and telephone calls, or 9.7 per day. This number implies that media consumption-- including background music, multi-tasking, and multi-media, but not including email and phone time, -- would occupy or overlap with 60% of all non-sleep time, including work time. Thus, the control over such large components of wallet and time are important just as a business matter, even beyond the role of media for politics, culture, and society. Our study identifies as one of the major issues in global media concentration the concentration of media in the emerging countries. Mexico is at the front line of this issue. But for most media intellectuals around the world arguing in favor of media pluralism, the focus of attention is on advanced world media moguls such as Rupert Murdoch, Silvio Berlusconi, or the Hollywood film majors. These are indeed problems. But usually overlooked are the dominant media companies in the emerging and developing countries where most of the world’s people live. My contribution to you today will be to provide some global numbers, which will show how Mexico performs relative to the rest of the world.

2

It is, of course, useful to look at the data for a single country. But it is also useful to look at the rest of the world. Because if we can identify common trends we can distinguish between fundamental drivers of change and country-specific situations. It’s hard to deal with fundamental forces. But for country-specific variations policy intervention may make a difference. Our work covers 30 countries. 80% of world by population, and 80% by GDP. The study involved 70 researchers around the world, covering 30 countries, 13 media industries1, over a period of 10-25 years. They covered thousands of companies. What we at Columbia University added was the common methodology and the overarching data analysis. B. Who are the largest owners of media in the world?

The world’s largest media owners are listed in Table 1. Carlos Slim is the single largest individual owner, trailing only large governments (China, Japan, and Germany) and several large institutional owners (Several major mutual and hedge funds managed by State Street, Vanguard, etc.). Table 1: Top Media Owners Worldwide (as of Sept. 2013)

Owner or Asset Manager Value of Media Holdings ($ billions)

Government of China 317.2

Government of Japan 67.2

State Street (US) 64.8

Vanguard (US) 63.8

Fidelity (US) 46.5

Capital Group (US) 35.2

Government of Germany 29.9

Carlos Slim (Telmex, America Movil, Carso, Mexico) 29.2

Larry Page (Google, US) 26.7

Government of France 26.4

T. Rowe Price Assoc. (US) 26.1

1 We investigated 13 Media Industries: Newspapers; Magazines; Books; Radio; Broadcast TV; Video channels; Wireline telecom; Mobile telecom; Internet service providers; Film; Search engines; Online newspapers.

3

Sergey Brin (Google, US) 26.0

Government of Russia 25.4

BlackRock (US) 24.3

Cox family (Chambers, Kennedy, Parry-Sheden, Anthony) (Cox Communications (US) 24.0

Michael Bloomberg (Bloomberg LP, US) 24.0

David Thompson family (Thompson Reuters, Canada) 20.3

Marinho family (Globo, Brazil) 20.0

Dodge & Cox (US) 20.0

Government of India 19.8

Mark Zuckerberg (Facebook, US) 19.0

Massachusetts Finance (US) 18.7

Brian Roberts family (Comcast, US) 18.5

JP Morgan Chase (US) 17.9

Newhouse family (Advance Publications, US) 17.1

Government of Norway 16.2

Wellington Management (US) 14.0

Government of the UK 13.9

Janus Group (US) 13.8

Goldman Sachs (US) 12.3

Government of Taiwan 12.1

Sawiris Family (Orascom, Egypt) 12.0

Murdoch family (News Corp./21st Century Fox, US) 11.6

Government of South Africa 11.6

4

C. Who is the largest media organization in the world?

a. By Revenues

Graph 1: Top 10 Content Media Organizations by Revenue

b. News Attention Time

A second way is to rank media companies is by news attention time. Table 2: TOP NEWS MEDIA COMPANIES BY ATTENTION SHARE OF WORLD POPULATION, 2004/05 AND 2009-2012 (> 0.1% SHARE, OR #1 IN THEIR COUNTRY)

Entity/Company Attention Share (2009-2012) - World

Country of Origin

GOV'T OF CHINA 18.96% China

CCTV 7.28% China

SHANGHAI MEDIA GROUP 2.33% China

33076.000

25872.000

22770.000

15,537

12,80411,702

9,7378,148 7,933 6,990

0.000

5000.000

10000.000

15000.000

20000.000

25000.000

30000.000

35000.000

5

HUNAN MEDIA GROUP 1.24% China

XINHUA 1.12% China

PEOPLE'S DAILY 0.73% China

CHINA NATIONAL RADIO 0.24% China

TRENDS MEDIA GROUP 0.17% China

PRASAR BHARATI (PUBLIC) 11.82% India

GOV'T OF RUSSIA 1.13% Russia

GAZPROM MEDIA 0.54% Russia

CHANNEL ONE 0.33% Russia

VGTRK 0.27% Russia

GOV'T OF EGYPT 1.12% Egypt

ERTU 0.87% Egypt

AL AHRAM 0.09% Egypt

AL AKHBAR 0.09% Egypt

AL GOMHOURIA 0.06% Egypt

NILE TV NETWORK 0.01% Egypt

GLOBO GROUP 1.09% Brazil

MURDOCH GROUP 0.95% US

21ST CENTURY FOX 0.76% US

NEWS CORP. 0.19% US

BCCL 0.93% India

GRUPO TELEVISA 0.77% Mexico

ZEE ENTERTAINMENT 0.68% India

DISNEY 0.62% US

COMCAST 0.56% US

REDSTONE 0.49% US

CBS 0.32% US

VIACOM 0.17% US

FININVEST 0.40% Italy

DOGAN GROUP 0.38% Turkey

RAI (public) 0.35% Italy

TV AZTECA 0.34% Mexico

CTC MEDIA 0.33% Russia

NHK (public) 0.33% Japan

6

BERTELSMANN 0.32% Germany

The government of China, through its several media organizations, accesses a truly vast share of global news attention. Almost 19% of global news attention. Even if we unbundle China’s news organization, CCTV alone would still command 7.28% of the global news attention and be the second largest news media company in the world. The explanations for these high shares are simple: a huge population (1.3 billion) and state control over most news media outside several online portals and print magazines. The largest private media firm, by attention time, is Globo in Brazil, with 1.09% for the entire world. Brazil, too, has a large population (192 million) and its media is dominated by Globo. Rupert Murdoch’s two companies combined are the second largest privately owned news providers, by attention, and the largest US-headquartered news firm (0.95% for the entire globe). Mexico’s Televisa is #8 worldwide, and #4 worldwide among private companies. It is larger in news attention than any US company except for Murdoch. It is larger than any European media firm. c. Market Power

A 3rd Approach is to ask: who is the largest media company by market power? Who is, in an antitrust sense, the most powerful company in the world? It is Google. By revenues, several companies are well ahead of it. But it holds high market shares in its industry in many. And its overall index has been growing enormously. The runners up are the major media firms News Corp, Time Warner, Disney, Vivendi, Bertelsmann. And the major European telecom companies Telephonica, Orange, Vodafone.

7

Graph 2: Companies with Highest Power Index

0

1000

2000

3000

4000

5000

6000

2009/8 WeightedGlobal Power Index

2005/4 WeightedGlobal Power Index

8

Graph 3: Countries’ Pooled Overall Sector C4 – All Media

D. Global Concentration Levels

What are the levels of media concentration? The world average is a very high 3,241 and up from 3,134 in 2004/05. The arithmetic average, which reduces the weight of large (by GDP or population) countries such as the US, China, and India, is 3,904. In 2013, the five countries with the highest average media industry HHIs were China (9,840), South Africa (6,242), Turkey (5,990), Egypt (5,544), and Mexico (5,042). On average, the pooled C4 (i.e. the top 4 companies in the overall combined market (13 industries) for the world is 56.6% (up from 52.3% in 2004), which means that on average, four companies control over one half of each country’s 13 national media industries, combined. This is an astonishingly high percentage and it is based on the large size of platform media. The share of the top firm in each country’s national media market, as measured by the pooled C1 ratio, is a huge 42.7% for platforms, which is 25 points higher than for content media, and 20 points higher than for news media, as measured by revenue. Without China, the average pooled C1 is 29.9%.

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0

Chin

a…Ch

ina…

Egyp

tSo

uth

Afric

aSo

uth

Kore

aTu

rkey

Pola

ndM

exic

oBr

azil

Belg

ium

Taiw

anIs

rael

Cana

daAr

gent

ina

Wor

ld A

vgIta

lyRu

ssia

Japa

nAu

stra

liaFi

nlan

dSw

itzer

land

Net

herla

nds

Spai

nIn

dia

Fran

ceIre

land

Port

ugal

Swed

enCh

ileG

erm

any

UK US

Pooled C4 - All Media

2004/5

2009 orMost Recent

9

E. Are the world’s media becoming more concentrated?

Has this concentration been declining or rising still further? For content media, average national HHI industry concentration rose by an annual rate of 3.7% per year, a fairly strong rate. That increase was mostly due to search engines. If we exclude the Internet sector, content concentration rose by a still strong 2% per year. The top companies increased their collective market shares, with C1 and C4 rising by 2.9% per year, and on average in an industry, by 2.3% for the top firm, and by 1.2% for the top four. The number of voices, on average, rose 1.25% per year. However, for platform media, average weighted industry HHI concentration declined by -0.3% per year worldwide. (On the basis of arithmetic averaging, it fell more strongly, by -2.4%.) The dominance of the top firm (pooled C1) in a platform industry declined by 0.3% per year, while the share of the top four (pooled C4) rose by 0.9%. This suggests a shift from near-monopoly to oligopoly. Average weighted national news media HHI industry concentration declined by -0.2% per year, based on attention time. In conclusion: average concentration for content media rose by an annual 3.1%, while it declined by an annual -0.3% for platforms. The picture is more positive in an arithmetic averaging, i.e., for the average country. HHI concentration for platform media declined by -2.4%. For content media they rose by 0.3%. On average (arithmetic), the top single platform company in a country lost 1 point of market share each year, while the combined share of the top 4 firms stayed flat. For content media, the top firm gained 0.3 points per year, the top 4 firms gained 0.2, and the HHI rose by about 9 points. For Mexico, the share of the top 4 firms (C4 index) declined by 1.3% for content media, and stayed stable for platforms. The C4 was 86% for content, and 98% for platforms.

10

Graph 4: Media Industry HHI Concentrations and their Trends

F. What are media industries with particularly high– and low –concentrations? What are the explanations?

Our finding is that average industry concentration is a function of capital intensity.

11

Graph 5: Media Concentration and Capital Intensity

The more capital intense an industry, the higher fixed costs, lower variable cost, and hence the higher economies of scale are. This, in turn, favors large firms and thus industry concentration. The consequence is that if media become more capital-intensive in the future, their industry concentration will rise. The search engine industry is the most concentrated (7,648). It also experiences the highest increase in concentration (5.6%) during the observed period. In comparison, the worldwide average HHI of the wireline industry, characterized by former or current public telecom monopolies, is the second most concentrated, but has been decreasing by 3% per annum. At the lower end of concentration are the print media industries. Surprisingly, the film industry has the second lowest average HHI of all media industries, due to the absence of a single dominating firm. It is an oligopoly of 6 firms, each with about 10-15%. That industry, however, is unusual insofar as the same six firms dominate in almost all countries where they are free to operate. Worldwide, of the 13 industries, the top 6 firms hold 76.6%. The Film industry’s Worldwide HHI is second only to Search Engines. So what are some implications for trends in media concentration? We have established that media concentration is affected by the capital intensity of the medium. When we look at media trends, we see a move towards online-based media. These media are capital intensive. They have high fixed costs, low marginal costs, and they are more global and national rather than local and regional of traditional media. This then suggests that the trend of media worldwide will be one of greater concentration.

Wireless

Wireline

Search Engines

Multichannel

RadioVideo ChannelsBroadcast TV

Newspapers

ISP

BooksMagazinesFilm

0100020003000400050006000700080009000

10000

0 0.5 1 1.5 2 2.5 3

Indu

stry

HH

I

Ratio of Avg Total Assets to Avg Total Revenue

HHILinear (HHI)

12

G. Does the transition to internet-based media reduce concentration?

The ISP market is highly concentrated, worldwide, with the arithmetic average HHI standing at 3,566 (down from 3,871 in 2004/05) and ranked sixth out of all industries for concentration. The search engines industry is the most concentrated media industry of all. Its average HHI concentration was an extraordinarily high 7,648 in 2013, up from 6,864 in 2004. Internet media, after an early stage of a dynamically competitive market structure, often become highly concentrated. Various market segments have their dominant players— Amazon, EBay, Microsoft, Google, Facebook, Twitter, YouTube, Apple’s App Store, and others. The Internet sector was believed to be wide open and competitive and would open things up for other industries, but it exhibits strong concentration trends. This observation about Google raises an important question. Do the new internet media make a difference, in the way its enthusiasts fervently believe? Not really. Internet media are actually more concentrated than the old legacy media. The problems are fundamental economics of scale and network effects. There are dominant players in their respective niches—Amazon, Ebay, Microsoft, Google, Facebook, Twitter, YouTube, Apple i-store, and on. All have overwhelming market shares, and not just in America, but worldwide. These are the industries that were believed to be wide open and

Old and New Media - Average Concentration by Region

0

1000

2000

3000

4000

5000

6000

World EU North America Asia-Pacific Latin America BRICS Middle East

Traditional Media 20th Century Media Internet Media

13

competitive. But they exhibit strong concentration trends. The underlying economic factors are easy to describe. High fixed cost, low marginal cost on the supply side, and high network effects on the demand side. That creates very high advantages to scale. The underlying economics on the supply side are, high fixed cost and low marginal cost; and on the demand side, strong network effects. Because these factors will remain, this trend is likely to continue, especially if the pace of disruptive innovation in the sector slows down a bit. Online news media is less concentrated than newspapers, though not in all countries. In the US, the incumbency advantage of the major multimedia producers means this online news are almost five times as concentrated as the daily newspapers industry. In fairness, the analysis deals with “commodity news” rather than with specialized online information sources. But such “long tail” information does not negate the fact that most online news attention is focused on a few mass-audience outlets. One must therefore conclude that the internet will not overcome the problems of media concentration. To the contrary, it will accelerate it. H. What countries have particularly high media concentrations?

Media Concentration in the US is among the lowest in the World. In Mexico, it is among the highest in the world. In 2013, the five countries with the highest average HHI industry HHIs were China (9,840), South Africa (6,242), Turkey (5,990), Egypt (5,544), and Mexico (5,042). The lowest concentrations measured were for the United States (1,565), India (2,690), Canada (2,690), France (2,872), and Sweden (2,815). Looking at content media only, China has the highest average concentration at 8,317. Also high are Russia (4,132), South Africa (4,151), Ireland (4,047), and Mexico (3,816). The lowest average content industry concentrations are in the US (1,042), Spain (1,665), Taiwan (1,834), and Japan (1,850). For platform media, the list of the highest concentration countries is similar: China (10,000), Turkey (6,439), South Africa (6,570), Egypt (5,882), and Mexico (5,332). The lower concentrations are in US (1,817), India (2,526), the Netherlands (2,840) and Finland (2,929). Not a single country’s platform concentration is below the antitrust standard of “highly concentrated”.

14

Graph 7: Weighted Average HHI – Content Media

0

1000

2000

3000

4000

5000

6000Ch

ina

Egyp

tRu

ssia

Sout

h Af

rica

Irela

ndM

exic

oPo

rtug

alSo

uth

Kore

aBe

lgiu

mCh

ileIn

dia

Turk

eyIta

lyAu

stra

liaN

ethe

rland

sW

ORL

D AV

GIs

rael

Finl

and

Switz

erla

ndBr

azil

Fra

nce UK

Pola

ndG

erm

any

Cana

daSw

eden

Arge

ntin

aTa

iwan

Japa

nSp

ain US

2004/5

2009 or MostRecent

15

Graph 8: Pooled HHI - Platform Media

0

2000

4000

6000

8000

10000

12000Ch

ina

Mex

ico

Belg

ium

Switz

erla

ndJa

pan

Spai

nSo

uth

Kore

aEg

ypt

Cana

daTu

rkey

Sout

h Af

rica

Italy

Fran

ceW

ORL

D AV

GFi

nlan

dCh

ileIs

rael

Swed

enAr

gent

ina

Ger

man

yPo

rtug

alIre

land

Braz

ilN

ethe

rland

sPo

land

Aust

ralia

Russ

iaIn

dia US

UK

Taiw

an

2004/5

2009 or MostRecent

16

Graph 9: COUNTRIES’ UNWEIGHTED AVERAGE C1 – All Media

0.00%

10.00%

20.00%

30.00%

40.00%

50.00%

60.00%

70.00%

80.00%

90.00%

100.00%

Chin

a… E

gypt

Sou

th A

fric

aTu

rkey

Mex

ico

Irela

ndKo

rea

Switz

erla

ndCh

ilePo

land

Belg

ium

Wor

ld A

vgFi

nlan

dU

KAu

stra

liaPo

rtug

alAr

gent

ina

Italy

Swed

enIn

dia

Chin

a…N

ethe

rland

sRu

ssia

Isra

elFr

ance

Spai

nBr

azil

Japa

nTa

iwan

Ger

man

yCa

nada U

S

Unweighted Avg C1 (All Media)

2004/5

2009 orMostRecent

17

Graph 10: Power Index Concentration by Country

I. What are factors for high national media concentration?

The analysis trend to capture country-specific characteristics that affect concentration in that industry. A number of variables were hypothesized and tested:

• Population size • Geographical size • Income • Education level • Per Capita Spending • Regulatory quality (for regulated media industries) • Years as a democracy since 1900.

The findings are: Industry concentrations in countries affected by variables of population, wealth, and governance.

0

1000

2000

3000

4000

5000

6000

7000

8000So

uth…

Turk

eyCh

ina

Egyp

tG

erm

any

Mex

ico

Taiw

anRu

ssia

Isra

elSw

itzer

land

Aust

ralia

Spai

n

WO

RLD…

Port

ugal

Chile

Irela

ndJa

pan

Pola

nd UK

Belg

ium

Indi

aBr

azil

Arge

ntin

aSo

uth…

Cana

daIta

lyFi

nlan

dN

ethe

rlan…

Swed

enFr

ance U

S

2004/05

2009 or MostRecent

18

The size of population was a fairly decent-sized negative factor for concentration in the newspaper and multi-channel TV industries. Geographic size of a country was a moderately sized factor for newspapers (a higher concentration) and ISPs (a lower concentration). Per capita income in a country was a negative factor for the concentration in mobile and wireless industries and a positive one for multi-channel platforms. Educational levels were a factor for newspapers and for multichannel (both associated with lower concentration). The “quality of regulation” and a variety of other “good government” metrics showed no statistically significant correlation for any of the industries. J. Analyzing the Divergence in Concentration

The graphs that follow depict the divergence of a country’s concentration in a media sector (e.g., magazines, newspapers, etc.) from concentration that would be expected based on the various socio-demographic and economic variables. They are ranked in order of such divergence. The results for Mexico show no divergence for magazines, but strong divergences for newspapers, wireless, wireline, and ISPs. Graph 11: Divergence of Actual National Media Concentration over Predicted Concentration - Magazines

19

Graph 12: Divergence of Actual National Media Concentration over Predicted Concentration – Newspapers

20

Graph 13: Divergence of Actual National Media Concentration over Predicted Concentration - Wireless

21

Graph 14: Divergence of Actual National Media Concentration over Predicted Concentration – Wireline

In wireless, China shows a very high divergence in terms of actual vs. expected concentration. Also high are Mexico, Turkey, and Switzerland. India is at the other extreme. For wireline, it is Japan that is at the high end, together with Turkey, Australia, South Africa, Spain, and Mexico. At the low end are the US, the Netherlands, Finland, and India.

22

Graph 15: Divergence of Actual National Media Concentration over Predicted Concentration - ISPs

23

Graph 16: Divergence of Actual National Media Concentration over Predicted Concentration – TV Broadcasting

K. The Developing Countries and BRICS

Even where media companies are private, they might not be independent. In many emerging markets, the top individual news media owners have an amazingly high share in news attention, often well above 40%. These companies have often achieved their strong mindshare by close relations with a government in power that awarded preferential licenses. (This is not limited to developing world: in France, President Mitterand in the 1980s awarded the country’s exclusive pay-TV license to Canal Plus, a company headed by his former chief-of-staff. In the US, the FCC under President Eisenhower, awarded no TV license to any newspapers company that had editorially endorsed his Democratic rival.) Once they achieve dominance, the media companies are hard to dislodge by subsequent governments or competitors. Indeed, given their influence over public opinion, they may receive

24

additional benefits. In some cases, their principals become political players themselves, as in the case of Italy, Indonesia, or Thailand. The implicit quid-pro-quo of economic favors and media support eradicates the concept of media as “speaking truth to power”. It would be comforting to believe that the role of governments in the media world declines and with it such leverage. But that is not so. On a large range of issues – telecom infrastructure, content production support and protectionism, intellectual property rights, merger approvals, access and interconnection rules and pricing, standards, etc—the role of governments may well be increasing, and with it the potential for favoritism. A great deal of attention has been given to the media power of such Murdoch, Redstone, or Berlusconi. It does not, condone such individual media power in rich countries if one also flags the problem of media in developing and emerging countries. In a good number of these countries the market shares and mindshares of domestic media firms are, within their societies, even higher, often by a substantial margin. In Mexico, three companies dominate their respective market niches, Carlos Slim in telecom (85% in wireline and 71% in wireless), the Azcarraga family in TV (68.3%), and O.E.M. in newspapers (59.4%). The critique of media power cannot be asymmetric. Developing and emerging countries cannot get a free pass on anti-pluralism. Indeed, they need active media even more than well-developed democracies. L. Are there different market characteristics for media in the countries of the North versus those of the South?

The poorer the country, the higher the dominance of the top firm. Media concentration in news is associated with lesser economic development, not with more. !#$1=7.24−0.4018 !#% (C1 is the market share of the top firm, and Y is per capita income with a R2 of 0.5801 and a t-value for the coefficient, of -6.22. There are several factors that contribute to this: Poorer countries are economically smaller, and hence sustain fewer media firms. Poorer countries tend to be characterized by powerful governments whose TV and press dominate the media. Cause and effect in this relation are intertwined. Crony capitalism. In some instances, a strong governmental role in media is rationalized as a defense to offset powerful private media firms. This was the case, for example, in Venezuela, as it had been in France under President Charles de Gaulle. The reverse happens, too, when a strong private media firm is seen as an offset against powerful State media. An example is Italy, where Fininvest has a major share and emerged as an offset to the State RAI. The organizations with the greatest news attention are not necessarily those with the highest revenues. The media organizations of China (CCTV, SMG, Xinhua, etc, whether separate or

25

integrated), of India (Prasar Bharati, BCCL, Zee, Sun), Russia (Gazprom Media, CTC), Egypt, Brazil (Globo), Mexico (Televisa), and South Africa (SABC, Naspers) have a vastly greater hold on people’s attention than on revenues. These are the media organizations of the emerging world – the BRICS media. In contrast, the leading media organizations of the developed world are a combination of traditional media conglomerates and successful startups. Their hold on attention, large as it is, is dwarfed by the BRICS media. But when it comes to revenues, it is the other way around (see Table 37.9). Here, the Murdoch Group, Google, Comcast, Disney, Bertelsmann, and Time Warner outpace China. Vivendi and Redstone are nearly as large too. Only one other BRICS media organization is among the top 30 media groups (Globo, #17). These media companies are headquartered in OECD countries and among these countries the economically strongest account for 24 of the 30 largest media content firms by revenue: US (12), Japan (5), France (4), and Germany (3). Thus, the emerging media system of the world is one of BRICS vs. Top-OECD, of mindshare vs. marketshare. Concentration in news media is associated with less economic development, not with more. In many emerging markets, individual news media owners have an amazingly high share in news attention: the Marinho family (Globo, Brazil, 35.3%), Emilio Azcárraga (Televisa, Mexico, 44.4%), the Dogan family (DMG, Turkey, 31.6%), Alexander Rodnyansky (CTC Media, Russia, 15.6%), and the de Noble family (Grupo Clarín, Argentina, (30%). (It is also very high in Italy, where Silvio Berlusconi’s firms hold 37.7% of news attention, and in Sweden, where the Bonnier family holds 21.1%). M. Voice Pluralism

There is another way to look at things– to count the ‘voices’. Which we define as media companies with more than 1% of a media market.2

2 Where the same company cross-owns media outlets in several media industries, we still count it only as one ‘net voice’, not as several.

26

Graph 17: Countries’ Net N (Number of Voices), Companies with >1% of Market Share in its Industry, 2009 or Most Recent.

In 2010, each country has on average 42 independent content media voices (up from 40 in 2004) active in its national media market, and possessing at least 1% in one of that country’s media industries. (Several of these voices may be owned by foreign firms.) The US has a high number of voices, behind only Russia. (Russia’s high number is partly due to its huge geography, and mostly due to its multiple language nationalities). So one might conclude that the US is doing fine. However, if one looks at the number of media voices per capita, the picture changes dramatically.

27

Graph 18: Countries’ Per Capita N (Number of Voices Weighted by Population), Companies with >1% of Market Share in its Industry, 2009 or Most Recent

Now, the US is really far down the line. Relative to its population, the number of its media voices is small. And who is high up? Ireland, Israel, Switzerland, and Finland, with that combination of active politics, entrepreneurship, and small population. On the other extreme are countries that are large as well as poor, and there the number of voices per capita is the smallest– China, India, Egypt, Mexico. A regression analysis of the data finds that the number of Net Voices is ≈ 40 pop&mil*+.+-.. The number of voices (holding the voice definition constant in terms of size) rises, for each doubling of population, by about 7.5 voices.

28

N. Which companies dominate the attention for news?

The combined share of firms that are top in news attention is enormous everywhere. The C10 in news attention in 8 countries is above 90%. It is between 80-90% in ten countries. It is nowhere below 50%. Thus, in most of the world, the top 10 firms command over 80% of news attention. The share of the top firm in terms of news attention is, on average, 30%. The government of China, in the aggregate it has 19.8% of the news attention of the entire globe. CCTV alone would still command 7.3% of the world’s news attention and be the second largest news media company in the world. The largest private media firm, by news attention time, is Globo in Brazil, with 1.09% for the entire world (1.71% for our 30 countries). India’s BCCL is the third-largest privately owned media firm in the world, with 0.93% for the entire world of global news attention share in 2012. Rupert Murdoch’s two companies combined are the second largest privately owned news providers, holding a global attention share of 0.95% of the entire world. It is the largest US-headquartered news firm. Most other large firms dominate large, single-country markets. Televisa of Mexico is the fifth largest private news firm in world news attention (0.77% for the world). In most of the world, the top 10 firms command over 80% of news attention. And nowhere do they command less than 50%. In many countries, the public service broadcasters are strong at the number 1 spot: India (Prasar Bharat, 65.4%) Ireland (RTE, 50.9%) Finland (YLE, 47.5%) Portugal (RTP, 33.9%) South Africa (SABC, 40.5%) Private media firms are at the #1 spot in: Mexico (Televisa, 44.4%) Brazil (Globo, 35.3%) Turkey (Dogan, 31.6%) Argentina (Grupo Clarín, 30.0%) Spain (PRISA, 20.0%) Israel (Channel 2, 19.8%) Australia (Seven Network, 16.9%)

29

O. Where is there a strong role of public (state) ownership in media concentration?

Public (i.e. state and state-corporate) ownership has been decreasing, for all media, from 18.4% in 2004 to 15.2% in 2012. Public ownership for content fell from 17.8% to 17.5%. For platforms, it fell from 18.2% to 14.4%. It is highest in countries where public incumbents, usually in platform industries, control the national infrastructure. In radio, even after liberalization in the 1980s, the national market shares of public service radio broadcasters remained quite high (55.2% for the BBC in the UK, 36.5% for the ARD regional stations in Germany, 49% for RAI in Italy). This is remarkable insofar as these shares are not based on a formal monopoly. The countries with the lowest percentage of public owned market share are Belgium, Argentina, Brazil, Mexico, and the US, all of which had less than 1%. (In Mexico this encompasses several public-owned radio stations). The World Bank study mentioned earlier (Schleifer et al., 2001/Djankov, 2003) reviewed 97 countries and identified a state ownership, on average, of 29% of major newspapers, 60% of TV stations, and 72% of the top radio stations. Twenty-one percent of the countries had a government monopoly on newspapers, and 44% on TV stations. In Africa, governments control 61% of the top five newspapers (by circulation) and 84% of TV audiences, with 71% of countries having a state monopoly. In the Middle East and North Africa, in 2001 all countries had a state monopoly over TV broadcasting, and held 90% of newspapers by circulation. (The exception was Israel.)

30

Graph 19: % PUBLIC OWNERSHIP IN COUNTRY MARKETS – ALL MEDIA

P. What countries have a high foreign ownership of its media?

Foreign ownership of media has been increasing in most countries, rising from a world average of 21.9% in 2005 to 26% in 2013 for all media industries. The growth has been driven primarily by internationalization in telecom, especially in mobile, and in the non-broadcast audiovisual media, enabled by lowering of laws favoring domestic ownership. The countries with the highest percentage of foreign-ownership of media are Ireland (53.6%), Chile (65%), Argentina (66.8%), Brazil (60.2%), the UK (50%), and Turkey (49.9%).

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

90.0%

100.0%

Chin

aEg

ypt

Taiw

anSw

itzer

land

Pola

ndSo

uth…

Russ

iaIn

dia

Ger

man

yJa

pan

Wor

ld A

vgSw

eden

Turk

eyFi

nlan

dFr

ance

Port

ugal

Irela

ndSo

uth…

Italy UK

Net

herla

n…Au

stra

liaSp

ain

Cana

daCh

ileIs

rael

Belg

ium US

Braz

ilAr

gent

ina

Mex

ico

Public Ownership (All Media)

2004/5

2009 or MostRecent

31

Graph 20: % FOREIGN OWNERSHIP IN COUNTRY MARKETS – ALL MEDIA

Q. What countries are high exporters of media?

Table 3: COUNTRY SHARES OF WORLD MEDIA EXPORTS, 2013 Country Country's Share of

World Media Exports

Country’s Share of World Media Market

US 23.9% 35.4%

UK 14.8% 5.7%

Spain 9.5% 3.9%

France 9.2% 4.5%

Mexico 8.0% 1.1%

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

Arge

ntin

aCh

ileBr

azil

Irela

nd UK

Turk

eyEg

ypt

Port

ugal

Italy

Net

herla

n…Be

lgiu

mAu

stra

liaSw

itzer

land

Spai

nW

orld

Avg

Ger

man

yFi

nlan

dSw

eden

Sout

h…In

dia

Fran

ceRu

ssia US

Taiw

anPo

land

Cana

daM

exic

oIs

rael

Japa

nSo

uth…

Chin

a

Foreign Ownership (All Media)

2004/5

2009 orMost Recent

32

Germany 7.5% 6.7%

Japan 6.4% 8.8%

South Korea 3.5% 3.9%

Italy 3.2% 3.0%

Russia 2.7% 2.1%

Sweden 2.3% 0.9%

South Africa 1.8% 0.9% The US is highest, at 23.9% in world share of non-domestic media activities. However, The US world export share is much smaller than its share in the overall world media market, which is 35.4%. That means that the domestic role of US media is much higher than the world average, and the export activity, while large in absolute terms, is lower than world average. Export “over-achievers” are the UK (14.8% vs. 5.7%) Spain, France, Mexico (8.0% vs. 1.1%, mostly due to Carlos Slim’s America Movil), Sweden, and South Africa. Table 4: TOP MEDIA COMPANIES BY EXPORT PERCENTAGE, 2009 OR MOST RECENT (> 10% OF TOTAL REVENUE FROM EXPORTS)

Company Total Revenue (mil $)

Non-domestic revenues (Export) (mil $) Export%

Vodafone (UK)a 73,766 65,681 89.0 Pearson (UK) 9,043 7,841 86.7 Google (US) 50,175 39,829 79.4 America Móvil (Mexico) 59,300 45,731 77.1 Liberty (US) 12,197 8,775 72.0 MTN (South Africa) 11,667 8,213 70.4 Naspers (South Africa) 2,719 1,914 70.4 Softbank (Japan) 50,804 34,497 67.9 Murdoch Group (US) 64,424 41,792 64.8

- 21st Century Fox 30,718 21,168 68.9 - News Corp. 33,706 20,561 61.0

Bertelsmann (Germany) 16,065 10,259 63.9 For the top 50 media companies, the average share of exports (defined as non-domestic revenues) as part of their overall turnover was 38.6% in 2013. The company with the greatest

33

share of revenue outside of its home country was Vodafone (89% of revenue outside of the United Kingdom). Pearson (UK, 86.7%), Google (US, 79.4%), America Móvil (Mexico, 77.1%), and Liberty (US, 72%). R. Is there a convergence in market structure among the various media industries?

We find a convergence in the concentration of media industries relative to each other. The content media sector’s concentration trend is converging with that of ICT. High-concentration industries have mostly been trending downwards, while low-concentration industries have risen (the exception are search engines.) The standard deviation – the measure of the industries’ divergence from the average -- in the concentration of the 4 media industry sectors dropped from 1,512 in 2000 to 1,026 in 2013. Ratio of Avg Total Assets to Avg Total Revenue The equilibrium of media industries is likely to reach an oligopoly structure. The fundamental economic reasons are: High fixed costs Low marginal cost In competition, this leads to over-investment, price deflation, failure, consolidation to achieve price stability. S. Who are the media owners, and what do they own?

There are 3 major kinds of owners – individual, institutional, and state/public. There are major differences in the ownership of content media and platform media. Many of the platform companies, even after the corporatizations and privatizations of the 1980s, have major ownership stakes by their governments. This includes NTT in Japan (33%), Deutsche Telekom in Germany (32%), Orange in France (27%), China Telecom (78%), Svyazinvest in Russia (53%) and Telkom in South Africa (50.7%). These governmental ownership stakes are particularly high given the fragmentation of the private stock holdings in platform companies among numerous small investors. Few of the top 20 platform companies have major individual owners. The main exception is America Movil in Mexico and other Latin American stakes. In Mexico Carlos Slim is holding 52%. Several other large platform firms have a high individual ownership stakes but originated as a media or internet company with a presence in platforms. Most of these companies have a dual stock structure.

34

US institutional investors have good-sized ownership stakes in many of the top 20 platform companies around the world, though rarely over 5%. Content firms tend to be privately owned. Most of the top content companies have major individual owners. The World Bank study mentioned (Shleifer et al, 2001) of major media enterprises in 97 countries finds that only 4% of them were widely held by shareholders. 2% were owned by employees. In contrast, families control 57% of newspapers and 34% of TV stations. What are the reasons for such high levels of individual or family ownership? Economists, going back to Demsetz (1989) concluded that the non-financial “amenity potential” of controlling media outlets, such as fame, influence, and favorable policy are high and therefore create incentives to acquire control. Control is often assured by foregoing a shareholder democracy. Great discrepancies exist for the equity percentage and voting percentage of many major content companies: Google; News Corp/21st Century Fox; Comcast; Lagardère; Bertelsmann; Liberty; Televisa; Bouygues; CBS; Viacom; etc. In each case, a founder or his heirs protect their control through complex stock structures. On the individual level there are 56 media billionaires as persons or families. Two-thirds substantially created their media properties themselves (37) rather than inherit established firms (19). The “new media billionaires” were mostly active in internet (12), mobile telecom (11), and information services (2). A significant share of the 56 media billionaires are US citizens (22); 4 each are from Japan and France; 3 from India; 2 each from Germany, Canada, Turkey, and China; 5 from Latin America; and 2 from Africa. Overall, there are 17 media billionaires in developing countries and the BRICS. The top 10 individual owners hold, in aggregate, $225 billion worth of media companies. The top 20 individual owners hold $313 billion. And the top 50 such owners had $422 billion. The share of the top 10 individual owners in the world’s media companies is then about 4.8% of the top 20 owners about 6.7%, and of the top 50 owners, 9.0%. A lower multiple, such as 1, would double these shares. Generally, the stake of institutional investors’ is larger than those of individuals. In 2013, the investment company State Street Corp. had $65 billion invested in major media companies, mostly in behalf of the clients of its investment funds. Rupert Murdoch, in comparison, had “only” $11.6 billion. Janus Capital, with over $11 billion, had more money tied up in media than Berlusconi, Malone, Redstone, and Lagardère combined. The popular belief that technological convergence in the information sector, plus worldwide mergers, have resulted in a small group of media moguls is not an accurate one. Such individual owners do exist, of course, and the 56 media billionaires attest to that. Each country seems to have several such large-scale private media owners. But for the developed world, larger in financial terms and growing faster is ownership by fairly anonymous funds owning, on behalf of their fund investors, narrow slices of a very big pie.

35

We identify at least 30 institutional owners with more than $1 billion in media assets. The total media assets of the top 10 institutional owners add up to $332.5 billion; the top 20 have $423.4 billion. And the top 30 have $449 billion. Of overall media value, estimated as $4.7 trillion before including state-owned media), the top 10 institutional owners hold 6.1%, and the top 30 hold 9.6%. Of the headquarters of the top 30 asset management companies (in terms of media holdings), 73% are based in the US (22 companies). In conclusion, the overall ownership of the media sector is as follows: The Top 30 individual owners: 7.9-15.8% The Top 30 institutional owners: 9.6-19.2% The 30 government holdings: 13.4% These 90 owners, then, account for about 30-50% of all media assets. Institutional investors will likely grow their ownership share, while state ownerships will likely continue to decline in the developed world. Individual non-entrepreneurial ownership will also decline, though it is being kept around through various forms of economically inefficient multiple share classes. T. What are the priority problems in Media Concentration?

In this presentation, we have striven to create a fact base as the basis for public policy. Adding policy recommendations would only detract from the analysis. That said, we identify several key problems. 1. The Problem of Rising Content Media Concentration

Globally media concentration (in terms of HHI) grew by 3.1% per year for a weighted average and 0.3% per year on the basis of an arithmetic averaging (which reduces the weight of the US, China, India). In the US, content concentration, though low in international comparison, (HHI=1042) was growing at a rate of 4.2% per year, and by 2.5% without the internet content media, much faster than the world average. The rise in content media concentration was only partly offset by the decline in the platform industry concentration. That sub-sector is still very highly concentrated.

36

2. The China Problem

The share by the government of China in the world’s media ownership and especially in news attention, through its domestic dominance, is truly extraordinarily high.

Even countries that are market oriented, such as Japan and Germany, have substantial state ownership in the media sector. But those are used to extend pluralism through transmission over content-neutral telecom, and in providing, over public service TV, content that is often not produced by the commercial sector. As China becomes an economic superpower, such centralized control might either decline as part of more general liberalization, or remain as a major question mark for the world’s overall media diversity. 3. The Google Problem

The media organizations with the highest revenues are not necessarily the ones with the highest news attention or the highest market power globally. Google, with just $3 billion in 2004/05, became the company with the second highest content revenues in 2009. Google’s growth rate will make it shortly the world’s largest content company. It has the highest Power Index of any company, with a score of 288. US and EU antitrust actions deal primarily with the manifestations of market power in vertical markets, rather than with the market power itself. Even aside from its talent, the company’s sheer market power and entry barriers in a high fixed-cost, low marginal-cost industry create problems that are hard to overcome conceptually or practically. 4. The Developing Countries and BRICS Problem

Supporting a media protectionism that serves as a shield for major domestic companies under the guise of media autonomy does no favor to these societies. Of course, a delicate balance must be found to domestically grown content and distribution, but it is not found in the domestic dominance of politically well-connected companies. The incentives to national media empires in poorer countries is still rising. For now, cross-ownership is higher in well-developed media markets. This suggests that the presence of media firms in multiple media industries rises with economic and media development. In consequence, as countries develop economically and their media grow in technological and business complexity, one should expect further trends to cross-ownership, without necessarily the counter-forces of markets and regulation.

37

5. The Internet Problem

The internet has been the great hope for media pluralism. It is supposed to open established powerful media and create pluralism around the world, impervious to efforts by governments to control it. This optimistic scenario is unlikely to happen. First, because governments can control its use, as China has demonstrated. More fundamentally, the underlying economics of the internet media is one of great economies of scale, and of high network effects. Our findings, more generally, show that differences in the concentration levels of different media industries are correlated with the capital intensity of that media industry. The more capital intense an industry, the higher fixed costs, lower variable cost, and hence the higher economies of scale are. This, in turn, favors large firms and thus industry concentration. The consequence is that if media become more capital-intensive in the future, their industry concentration will rise. The internet will play a positive role in niche markets of the ‘long tail’, where entry becomes easier and scale is low, and competition from large players is not a major factor. But off the long tail – at the center of media activities, with general news, mass entertainment, infrastructure, and central nodes such as clouds—the opposite is the case. The internet raises capital intensity of key segments of media. Companies such as Netflix, Apple iStore, Amazon.com, Google, and its YouTube, have global market shares that far exceed those of conventional media. Successful internet companies with high market share in some of their operations are also able to extend them into other media activities, as Apple or Google have demonstrated. This goes beyond online. The example of Jeff Bezos’ (founder of Amazon.com) buying the Washington Post might be part of a broader pattern. Internet and concentrations are becoming high enough and large enough in key activities that they have been pulling up the average concentration of content industries. Thus, the internet, far from being the solution to media power, will prove to be a force of concentration. 6. The Globalization Problem

In an increasingly global media system, the number of voices per capita will decline, and the aggregate number of voices will decline. E.g., instead of two countries having each 30 voices, a joint media market might have 45 voices. On the positive side, users will have more options in that global media system than they had in a national market. But on average, the national production of media voices will decline as media markets become more global. The number of voices per capita will also decline. And the content, in order to be successful in the larger world market, will be less domestic in orientation.

38

7. Institutional Ownership

Public attention has centered on highly visible moguls. But the reality of media ownership is that of institutional investors that hold small to medium-sized pieces of many media companies. Institutional investors will likely grow their ownership share. There are several potential issues associated with such an ownership system: too much control; not enough control; and the absence of localism. The first issue is that an institutional owner or a small group of such owners would affect media behavior. It is true that they do not normally exercise a direct role in management in the way that personal owners do, but in an indirect way, through their setting priorities on short-term stock performance, and through their buy-and-sell decisions, they set behavioral parameters for the actual managers. The second potential problem is one of giving managers too much of a free hand, without the vision and civic responsibilities of proprietors or their heirs. But this faith in the nobles oblige romanticizes such essentially feudal arrangement. The third and most real type of problem is that of absentee ownership—a lack of sensitivity and concern by institutional owners for distant localities. 8. The Problem of Low Impact Regulation It is generally undesirable to give governments an undue role in the structure or behavior of media, since such powers are inevitably abused or corrupted. To that principled objection is added one of practicality. Media regulation was easier accomplished before digital convergence and global information flows. But now, through tools such as licenses, ownership ceilings, cross-ownership rules, etc. these regulatory tools are becoming ineffectual. How would they deal with the market power of Google? With quality issues of Skype? The internet replaces a tight national oligopoly with a less tight but international one. There would be more diversity but less national content? How would one deal with that? In the end, the most effective tool for government to assure pluralism, outside of curbing clear abuses, is to help generate alternative media, and to protect their access and interconnection. It is important for academics, public-policy analysts, NGOs, companies, and governments to think creatively about new approaches to these issues, balancing innovation, investment, and the public interest in the emerging environment. Outlook

39

The issue of media concentration has a long past and will have an even longer future. It is one of those fundamental issues of distribution of power and wealth that every generation needs to resolve. Now, that task has reached us. The data presented here, I hope, will help Mexico in that process.

Thank you! noam@columbia.edu