presentation - Columbia Institute for Tele

Self-Image: “You can’t tell a
dog on the Internet”
Internet Concentration
and What it Tells Us
About the Problems of
the Information
Economy
Eli M. Noam
Professor of Finance and Economics
Columbia University Graduate School of
Business
Director
April 2004
Columbia Institute for Tele-Information
©2003 Eli M. Noam
1
Questions:
• Has the Internet sector
become more concentrated?
• More concentrated than other
information and media
industries?
• Causes and implications for
the information sector
generally
©2003 Eli M. Noam
3
Industries of the Internet
Sector: Basic Instrumentalities
• Backbones
• ISPs
• Portals
• Broadband Providers
• Internetworking Equipment
• Browsers
• Search Engines
• Modems
• Media Player Software
• IP Telephony
©2003 Eli M. Noam
©2003 Eli M. Noam
2
• No past empirical study to answer
the question of concentration
• My study looks at 95 US information
sector industries
– backbones, ISPs, long-distance telecom,
broadcast TV, micro-computers, ISPs,
mobile telecom, PBXs etc.
• Over the past 20 years
• Thousands of company reports and other
sources over a 20 year period
©2003 Eli M. Noam
4
Not Included
• Telecom conduits
• Applications
• Content
• Computer Hardware and OS
5
©2003 Eli M. Noam
6
1
Concentration index #1:
• Herfindahl-Hirschman Index (HHI)
f
HHI = ∑ S
i =1
2
i
90
Percent of Market
2000
70
C4
60
1500
HHI
50
40
1000
30
20
WC 4
sec tor
52
=∑
j =1
m
M
j
4
∑S
i =1
ij
j = a sub-industry (e.g mobile handsets) within an industry (e.g. telecom
equipment)
mj = total revenue of a sub-industry
M = total revenue for the information sector
i = firm in a sub-industry
©2003 Eli M. Noam
2500
80
i = firm in a sub-industry
Si = market share of firm given in a given sub-industry
7
Backbones
Where: SI = firm's market share of a given sub-industry
Where:
HHI < 1,000 Unconcentrated Market
1,000 < HHI, Moderately Concentrated Market
1,800 < HHI, Highly Concentrated Market
100
C4 Index
4
C4 = ∑ S
i
i
A weighted aggregate C4 for
the entire information sector
then is:
Antitrust enforcement guidelines
classify markets
©2003 Eli M. Noam
Concentration Index #2:
8
Backbones 2002
WorldCom
AT&T
Genuity
C4
HHI
US Total Revenues ($ mil)
41
13
11
74.0
2174
10,500
500
10
0
0
1984
1984
1988
1992
C4
1996
2001
2001
HHI
9
©2003 Eli M. Noam
ISPs
100
©2003 Eli M. Noam
7000
90
6000
C4
Percent of Market
80
5000
70
60
HHI
4000
50
3000
40
30
2000
20
1000
10
0
1984
1984
1988
1992
C4
1996
0
2001
2001
10
ISPs 2002
AOL Time Warner
MSN
United Online
Earthlink
Prodigy (SBC)
AT&T WorldNet
Other
C4
HHI
Total Subscribers
44.2
11.5
8.1
6.9
5.2
2.0
22.1
70.7
2226
69.4
HHI
©2003 Eli M. Noam
11
©2003 Eli M. Noam
12
2
Portals
100
90
Percent of Market
6000
C4
80
5000
70
60
4000
HHI
50
3000
40
30
2000
20
1000
10
0
1984
1984
1988
1992
C4
Portals (% users 2002)
7000
1996
AOL Time Warner
Yahoo
MSN* (Microsoft)
Other
US Total Revenues (in millions)
C4
HHI
29.5
22.6
21.7
28.9
2908
85.6
2036
0
2001
2001
HHI
13
©2003 Eli M. Noam
Media Player Software 2002
49.8
27.6
13.5
6.2
2.9
153
97.1
3462
Browser Software
100
9000
80
7000
60
6000
50
5000
40
4000
30
3000
20
2000
10
1000
1
1995
2
1996
3
1997
4
1999
C4
15
5
2000
0
6
2001
HHI
16
©2003 Eli M. Noam
Browser Software 2001
Internetworking Equipment
3000
100
13.0
86.0
300.0
100
7565
90
Percent of Market
Netscape (AOL TimeWarner)
Microsoft
Total Revenue ($ mil)
C2
HHI
8000
HHI
70
0
©2003 Eli M. Noam
10000
C4
90
Percent of Market
RealNetworks
Microsoft Windows Media
Apple Quick Time
AOL Winamp
Other
Revenue ($ mil)
C4
HHI
14
©2003 Eli M. Noam
C4
80
2500
70
2000
60
50
1500
HHI
40
1000
30
20
500
10
0
1984
1984
1988
1992
C4
©2003 Eli M. Noam
17
1996
0
2001
2001
HHI
©2003 Eli M. Noam
18
3
Internetworking Equip 2001
Cisco
Avaya
Marconi
Nortel
Alcatel
IBM (USA)
Siemens
Industry Revenue US ($mil)
C4
HHI
48
20
8
7
5
5
2
14,840
83.0
2871
19
©2003 Eli M. Noam
Aggregate
Concentration
(weighted by size)
21
©2003 Eli M. Noam
3500
3000
2500
2000
1500
1000
500
0
1988
1992
1996
2001/2
Internet Industries 1984 - 2001/2
©2003 Eli M. Noam
DSL
SBC
Verizon
Bell South
Qwest (US West)
Broadwing (Cincinnati Bell)
12.8
11.5
6.0
4.3
0.6
Cable
AOL Time Warner
Comcast Corporation
Cox Communications
Charter Communications
Cablevision Systems
Aldelphia Communications
18.4
23.6
8.5
5.8
4.9
3.6
Total Revenues (mil)
C4
HHI
4,762
66.3
1389
20
©2003 Eli M. Noam
Weighted Aggregate HHI
n
WAHHI = ∑
j =1
mj
∑m
f
∑S
j i =1
2
ij
Where:
• j = an industry
• mj = total revenue of an industry
• Si = each firm’s market share of an industry
• n = number of industries
• f = number of firms in an industry
©2003 Eli M. Noam
22
Comparison #1: Internet
vs Other Media
Internet Industries 1984 - 2002
1984
Broadband Providers 2002
23
• Print
• Film
• Broadcasting
• Internet
• Broadband Internet
©2003 Eli M. Noam
24
4
Media Concentration Trends: Old
Media to Internet 1984 - 2002
6,000
Print Sector
Media Concentration Trends: Old
Media to Internet 1984 - 2002
6,000
5,000
5,000
4,000
4,000
3,000
3,000
2,000
2,000
1,000
1,000
-
-
1984
1988
1992
1996
1984
2001/2
Print Sector
1988
1992
Film Sector
25
©2003 Eli M. Noam
Media Concentration Trends: Old
Media to Internet 1984 - 2002
1996
2001/2
Print Sector
26
©2003 Eli M. Noam
Media Concentration Trends: Old
Media to Internet 1984 - 2002
6,000
6000
5,000
5000
4,000
4000
3,000
3000
2,000
2000
1,000
1000
0
-
1984
1988
Broadcast Sector
1992
1996
Film Sector
1984
2001/2
1988
New media Sector
Print Sector
27
©2003 Eli M. Noam
Media Concentration Trends: Old
Media to Internet 1984 - 2002
1996
Film Sector
2001/2
Print Sector
28
©2003 Eli M. Noam
Media Concentration Trends: Old
Media to Internet 1984 - 2002
6000
6000
5000
5000
4000
4000
3000
3000
2000
2000
1000
1000
0
1992
Broadcast Sector
0
1984
Internet Sector
Film Sector
1988
1992
New media Sector
1996
2001/2
Print Sector
©2003 Eli M. Noam
1984
1988
1992
Internet Sector
Broadcast Sector
Print Sector
Broadcast Sector
29
1996
2001/2
New media Sector
Film Sector
Broadband Providers, Local
©2003 Eli M. Noam
30
5
Findings
Reasons?
• The younger the medium
Higher investment needs,
greater economies of scale,
greater network effects,
greater risk from
competition
–the more concentrated
–The stronger concentration trend
after 1996
–
31
©2003 Eli M. Noam
Comparison #2:
Internet vs. Other Info
Sectors
Internet concentration likely to
increase in near term
–
–
Economies of scale
Profitability potential of
oligopoly
• Internet
• Telecom
• Mass Media
• IT
33
©2003 Eli M. Noam
34
©2003 Eli M. Noam
Concentration of the Four Major
Segments of the Information Sector
Concentration of the Four Major
Segments of the Information Sector
4000
HHI
4000
32
©2003 Eli M. Noam
HHI
M a s s M e d ia
3500
3500
3000
3000
2500
2500
2000
2000
1500
1500
1000
1000
500
500
0
0
1983 1984
1
1
1983 1984
1988
1992
1996
2001
©2003 Eli M. Noam
1988
1992
Mass Media
Mass Media
35
1996
2001
Information Technology
©2003 Eli M. Noam
36
6
Concentration of the Four Major
Segments of the Information Sector
4000
HHI
HHI
4000
3500
3500
3000
3000
2500
2500
2000
2000
1500
1500
1000
1000
500
500
0
1
1983 1984
1988
Mass Media
1992
Information Technology
1996
2001
Telecommunications
37
©2003 Eli M. Noam
Concentration of Total Information Sector
3000
HHI
In f or ma t oi n Se c t o r In c ul d in g IT
0
1983 1984
1
Mass Media
1988
1992
Information Technology
1996
2001
Telecommunications
©2003 Eli M. Noam
Internet
38
Findings
• Mass Media concentration, though
receiving much attention, and while
growing, is the lowest of the 4 infoindustry sub-sectors
• Internet (and telecom) are the most
concentrated sub-sectors
2500
2000
1500
1000
500
0
Concentration of the Four Major
Segments of the Information Sector
1983 1984
1
1988
1992
1996
–Only ones above DoJ threshold of
“highly concentrated”
2001
Information Sector Including IT
©2003 Eli M. Noam
39
3. Internet (and Mass Media) have
highest growth in concentration
4. After 1996, telecom & Internet
concentration up sharply
– Mass Media steady up
©2003 Eli M. Noam
41
©2003 Eli M. Noam
40
Trends of Vertical
Integration in the
Internet
©2003 Eli M. Noam
42
7
Vertical Measure I:
The Participation Index (PI)
shows extent of top firm
n
Even if a firm does not
dominate any specific
market, its presence in
several markets might,
in combination, become
powerful
1
1
PIn =  ∑ PI i =  ∑ 0,1
n
43
©2003 Eli M. Noam
 n  j =1
i
Where:
• i = firm (top n firms in terms of information
revenues)
• j = sub-industries
• n = number of sub-markets in which a firm
participates
44
©2003 Eli M. Noam
Participation Index Top 10 Internet Companies in
Total Information Sector 1984 - 2001/2
Vertical Measure II:
70
The Sector Share Index (SSI)
60
SSI =∑ SSI = ∑ s m
M
Where:
50
j
40
sector
30
i
firmi
j
j
• sj = Firm i’s Share in Market j, in percent
• mj = Market revenues of sub-industry j
• M = Revenues of Total Sector
20
10
-
1984
1988
1992
1996
2001/2
Participation Index
©2003 Eli M. Noam
45
Percentage of Top 10 Companies in Internet
Industries' Revenue of Total Information Industry
50.0
Revenue 1984 - 2001/2
45.0
Mind the Scale
©2003 Eli M. Noam
46
Percentage of Top 10 Companies in Internet
Industries' Revenue of Total Internet Industry
Revenue 1984 - 2001/2
100
90
40.0
80
35.0
70
30.0
60
25.0
50
20.0
40
15.0
30
10.0
20
5.0
10
-
1984
1988
1992
1996
2001/2
Percentage of Top 10 Companies in Internet Industries' Revenue of Total Information
Industry Revenue 1984 - 2001/2
©2003 Eli M. Noam
47
0
1984
1988
1992
1996
2001/2
Percentage of Top 10 Companies in Internet Industries' Revenue of Total Internet
Industries Revenue 1984 - 2001/2
©2003 Eli M. Noam
48
8
Share of Total revenue of Top 10 Companies in
Internet Companies, of total information sector 1984
100.0%
- 2001/2
90.0%
• Internet sector increasingly
part of
80.0%
70.0%
60.0%
50.0%
–Telecom industry
–Cable TV
–Mass Media conglomerates
40.0%
30.0%
20.0%
10.0%
0.0%
1984
1988
1992
1996
2001/2
Percentage of Top 10 Companies in Internet Industries' Revenue of Total Information
Industry Revenue 1984 - 2001/2
49
©2003 Eli M. Noam
Vertical Measure III:
1200
The Company Power Index (CPI)
50
©2003 Eli M. Noam
Company Power Index of Top 10 Internet
Companies 1984 - 2001/2
1000
2
CPI
= ∑ CPI = ∑
s m
j
j
800
M
Where:
• sj = firm’s share in market j
• mj = total revenue of sub-market j
• j = sub-industries, ranging from 1 to 52
(consisting of the 52 sub-industries)
•M = Revenues of total information sector
industry
i
firm
©2003 Eli M. Noam
j
600
400
200
0
1984
51
Company Power Index Top 10 Internet Companies Internet Activities Only, within Internet Industry
2500
Revenue 1984 - 2001/2
1988
1992
1996
2001/2
Company Power Index of Top 10 Companies in Internet Industries, including PCs, Microprocessors
(MPUs) & Operating System Software
©2003 Eli M. Noam
60
52
Company Power Index Top 10 Companies in
Internet Industry 1984 - 2001/2 (Internet only)
50
2000
40
1500
30
1000
20
500
10
0
0
1984
1988
1992
1996
2001/2
Company Power Index Top 10 Internet Companies - Internet Activities Only, within
Internet Industry Revenue 1984 - 2001/2
©2003 Eli M. Noam
53
1984
1988
1992
1996
2001/2
Company Power Index Top 10 Internet Companies - Internet Activities Only 1984 2001/2
©2003 Eli M. Noam
54
9
800
700
Company Power Index of Top 10 Internet
Companies in Internet Industry 1984 - 2001/2
(including activities in other industries).
Related Industries
600
500
400
300
200
100
0
1984
1988
1992
1996
2001/2
Company Power Index Top 10 Internet Companies - Total Activities 1984 - 2001/2
55
©2003 Eli M. Noam
56
©2003 Eli M. Noam
Future:
Similar trends:
• Broadband: even greater trend to
concentration
–Cost high
–Already prices high in US duopoly
DSL-Cable
• Applications
–eBay
–Amazon
• Major software
–Microsoft
• Key hardware
–Intel
57
©2003 Eli M. Noam
Operating System Software
100
10000
90
Percent of Market
8000
70
7000
60
6000
5000
HHI
40
4000
30
3000
20
2000
10
1000
0
0
1984
1984
Microcomputers
3000
9000
C4
80
50
58
©2003 Eli M. Noam
1988
1992
C4
1996
2001
2001
2000
1500
1000
500
0
1984
HHI
©2003 Eli M. Noam
2500
59
1988
1992
Microcomputers
©2003 Eli M. Noam
1996
2001/2
60
10
Applications: Online Book
Retailing
100
10000
90
9000
C4
Percent of Market
80
8000
70
7000
60
6000
50
5000
40
4000
HHI
30
3000
20
2000
10
1000
0
1984
1984
1988
1992
C4
1996
2001
Likely Implications
• Slowing of innovation & upgrade
• Higher consumer prices, higher
profits
• Restrictiveness on content and
applications (cable TV model) and
content access pricing
• Global strength and presence
0
2001
HHI
©2003 Eli M. Noam
61
• Greater vulnerability to disruption
• More proprietary standards and
protocols?
• Internet governance replaced by
cartel
• Cross-subsidies and vertical
extension of market power
• Gov regulation to deal with such
problems of market power
©2003 Eli M. Noam
63
©2003 Eli M. Noam
62
Internet Past
©2003 Eli M. Noam
64
Internet Future?
Why
Concentration?
http://www.gocollect.com/product_display/products/product_page.asp?id=28126&refcode=affiliate-929
©2003 Eli M. Noam
65
©2003 Eli M. Noam
66
11
Response to structural
instability of many
information industries
©2003 Eli M. Noam
67
Information Economy will be
• Volatile
• Cyclical
• Unstable
• Needed: Macro-economics of the
information society
©2003 Eli M. Noam
69
Price Problems
•
•
•
•
•
•
Music industry
Online publishers
Phone calls
Cell phone
Web advertising
TV and radio
• Software
• Academic
articles
• Newspaper
• Digital photos
• Semiconductors
©2003 Eli M. Noam
71
•
•
•
•
•
•
•
•
•
Information Sector Crisis
Dot com bubble
Telecom crash
Music bust
E-publishing stagnation
PC drop
Wireless saturation
Semi-conductor slump
Newspaper recession
R&D crisis
©2003 Eli M. Noam
68
Common Revenue Problem of
Information Industries
• Information has become cheaper
for many a decade.
• It is now becoming difficult to
charge anything for it.
©2003 Eli M. Noam
70
Two Basic Explanations for
Crises
• “Perfect Storm” scenario
• “Fundamental instability”
©2003 Eli M. Noam
72
12
“Fundamental
Instability”
The entire information
sector is subject to a
gigantic market failure in
slow motion.
©2003 Eli M. Noam
73
Basic Structural Reasons
1. High fixed costs and low
marginal costs
2. Network effects
3. More competitive
4. Commodification
5. Inelastic demand
©2003 Eli M. Noam
75
Exceptions: Market Power
• Cable TV
• Sports rights
©2003 Eli M. Noam
74
Consequence: a secular
trend of price deflation in
information products and
services
©2003 Eli M. Noam
76
Reaction by providers
• Price is marginal cost, which is
close to zero
• Most likely below average cost,
which means that it does not
cover total cost
©2003 Eli M. Noam
77
• Price discriminate
• Innovate
• Hedge
• Cut cost
• Industry consolidation
©2003 Eli M. Noam
78
13
Same economic factors
lead again to a new cycle
of investment overproduction and a new
price collapse
©2003 Eli M. Noam
79
The information
industries are becoming
interdependent under
digital convergence
©2003 Eli M. Noam
81
Cyclical instabilities
common to other industries
with similar characteristics
of high fixed cost and low
marginal costs
–Airlines
©2003 Eli M. Noam
80
Price deflation in the
information sector will
drag down the rest of the
economy, too, through a
multiplier effect.
©2003 Eli M. Noam
82
Business Responses
• As societies become
information economies, they
also become more volatile
economies.
©2003 Eli M. Noam
83
• Try to consolidate and cartelize
to create pricing power
• Price discriminate
• Technological innovation
• Outsource and offshore
• Protection of property rights to
differentiate product
©2003 Eli M. Noam
84
14
The role of government
©2003 Eli M. Noam
85
Macroeconomic instruments
of governments are based on
the experience of the
industrial economy
©2003 Eli M. Noam
87
Another policy approach: Assure a
diversification of the economy
• Reconsider emphasis towards information
sector
• Encourage industries outside of the
information sector
–Often low-tech industries
• Back to the basics, the industrial sector
and the basic industries
• “Info-industrial” sector
©2003 Eli M. Noam
89
Finland, Nokia
• 35% of all exports
• 12% of GDP
• Plus indirect contributions
©2003 Eli M. Noam
86
• Keynsian demand generation?
• Industrial support policy?
• Monetary policy?
• Competition policy?
©2003 Eli M. Noam
88
Alternative: the Civil Society
Sector?
• Volunteerism has structural
problems too, the classic
“tragedy of the commons”
problems
©2003 Eli M. Noam
90
15
Conclusion
• In the information economy
–1. Volunteerist activities are subject to
the “tragedy of the commons”
–2. Private information markets are
likely to frequently fail in “the tragedy
of the digital market”
–3. Governments do not have many tools
to deal with these failures
©2003 Eli M. Noam
91
Economics, the “dismal
science”
©2003 Eli M. Noam
92
Next Questions:
• What are the implications for
–Innovation, business strategy, stability
of the industry, public policy?
• Discussed in my next books
– “The Dark Side of the Internet”
–“Media Concentration and Ownership
in America”
END OF
PRESENTATION
Thank You.
©2003 Eli M. Noam
93
©2003 Eli M. Noam
94
16