Transcript Presentation of SG`s report

Global Internet Connectivity:
A critical success factor in the
information economy
Dr Tim Kelly,
International
Telecommunication Union,
Telenor Carrier Venue,
2001
The views expressed in this paper are those of the author and may not necessarily reflect the opinions of the
ITU or its membership. Tim Kelly can be contacted at [email protected]
.
Agenda
 Connectivity and reach




Internet and the digital divide
Some international comparisons
Some common myths
“The Internet latecomer” problem
 Connectivity and richness
 Broadband deployment
 Barriers to broadband deployment
 Connectivity and economic opportunity
Number of Internet host computers,
in millions, and annual growth in %
104.0
135%
100%
72.4
107%
93%
95%
81%
79%
70%
14.4
0.1
0.2
0.4
Jan89
Jan- Jan90
91
1.3
2.7
5.8
Jan- Jan92
93
Jan94
Jan- Jan95
96
0.7
43.2 67%
29.7 46%
21.8 36%
44%
Jan- Jan97
98
Jan01
Source: ITU, adapted from Internet Software Consortium <www.isc.org>.
Jan- Jan99
00
Digital divide = Infrastructure divide
User distribution, by income group, Jan 2000
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
300
million
490
million
912
million
6 billion
15 %
High income
Upper-mid income
58 %
Lower-mid income
69 %
82 %
Low income
Internet
users
Mobile
users
Telephone
lines
Source: ITU World Telecommunication Indicators Database.
Population
Barriers to connectivity
 Regulatory barriers
 46 countries permit only one ISP
 A significant minority of countries restrict access to
international IP bandwidth to a single IAP
 A majority of countries place limitations on use of
international IP bandwidth for voice over IP
 Economic barriers
 High costs for int’l leased lines in some markets,
esp. those without infrastructure competition
 For countries with only low levels of IP demand, unit
bandwidth costs are higher than for countries with
higher levels of demand (economies of scale)
 Many countries are not served by international
cables (e.g., landlocked countries, small islands)
More connectivity = more ISPs =
more connectivity
ISPs and Leased Line Internet connections in countries with more than 1 million inhabitants
ISPs per 1 milion inhabitants .
100.00
R2 = 0.7851
10.00
0.01
0.10
1.00
1.00
10.00
100.00
0.10
0.01
Leased lined connections to the Internet per 1 million inhabitants
Source: OECD, adapted from Netcraft.
1000.00
Market liberalisation and
investment brings down prices
Band-X: Bandwidth Price Index (2Mbit/s). Oct ’98 =100
120
London Sydney
London Hong Kong
Composite World .
Los Angeles Tokyo
New York London
Los Angeles Hong Kong
London Amsterdam
London Frankfurt
London Milan
London Madrid
London Brussels
Composite European
New York Frankfurt
London Paris
100
80
60
40
20
0
Oct-98
Source: Band-X.
Dec-98
Dec-99
Dec-00
Connectivity: Myths and misses
 The total bandwidth available in the United
States is increasing three-fold every year.
Therefore there will soon be universal
connectivity.
 Some 264 economies have Internet access.
Therefore the whole world is connected.
 Bandwidth across the Atlantic will soon
amount to more than 1 Mbit/s for every person
in North America and Europe. Therefore usage
will soon be too cheap to meter
Connectivity: Myths and misses
 The total bandwidth available
in
the
United
Connectivity is about
States is increasing three-fold every year.
distribution, not
Therefore there will soon be universal
about total quantity
connectivity.
WRONG!
WRONG!
The
majorityaccess.
of
 Some 264 economies have
Internet
have less
Therefore the whole worldcountries
is connected.
connectivity than a
single ADSL link
 Bandwidth across the Atlantic
will
soon
The vast majority of fibre
amount to more than 1 Mbit/s for every person
optic cable is left dark
in North America and Europe. Therefore usage
and is made available
will soon be too cheap to meter
WRONG!
only when price is right
Inter-regional Internet backbone
357 Mbit/s
19’716 Mbit/s
AsiaPacific
USA &
Canada
2’638
Mbit/s
Latin
America &
Caribbean
56’241 Mbit/s
468
Mbit/s
171
Mbit/s
Europe
Arab
States,
Africa
127 Mbit/s
Source: TeleGeography Inc., Global Backbone Database. Data valid for Sept. 2000.
Inequalities in Inter-continental
Internet bandwidth
 Singapore
 Estimated 950’000 Internet users (Dec 1999)
 Around 800 Mbit/s international capacity (June 2000)
 Around 20 kbit/s per user (assuming 1 hour per day)
 Egypt
 Estimated 200’000 users (Dec 1999)
 Around 26 Mbit/s international capacity (May 2000)
 Around 3 kbit/s per user (assuming 1 hour per day)
 Uganda
 Estimated 25’000 users (Dec 1999)
 Around 2 Mbit/s international capacity (Feb 2000)
 Around 1.9 kbit/s per user (assuming 1 hour per day)
Telecom connectivity and Internet
connectivity: What’s the difference?
Telecom world
 Platform is jointlyprovided circuits (costsharing)
 Operators typically
guarantee end-to-end
connectivity
 Financial arrangements
based on negotiated
settlement rates
 Cash flows from core to
periphery of network
Internet world
 Platform is leased circuits
(full-circuit paid by
connecting operator)
 Internet exchange points
typically guarantee only
efficient routing of traffic
 Peering between Tier 1
IXPs; traffic exchange
payments from Tier 2 IXPs
 Cash flows from periphery
to core of network
The “Internet latecomer” problem
 In the beginning, the Internet was:
 North American
 Not-for-profit
 Used primarily for email & file transfer
 By the mid-1990s, the Internet had:
 Spread to the OECD countries
 Begun “privatisation” of the backbone
 Become primarily a channel for the Web
 By 2005, the Internet may be:
 Available throughout the world
 A fully commercial undertaking
 Used primarily for multimedia streaming
USA &
Canada
AsiaPacific
Europe
Latin
America &
Caribbean
Arab
States,
Africa
Two LDC “success stories”
 Nepal
 16-fold increase in IP connectivity in 8 months
following liberalisation of VSAT market in 1999
 Lowest IP access prices in South Asia
 BUT, opening up VSAT market has lead to a drastic
fall in incoming telephone traffic and settlements
 Uganda
 Rapid increase in network growth following
introduction of second national operator, MTN
Uganda
 First country in Africa where mobilephone
outnumbered fixed-line telephones
 BUT, entire national capacity is less than 2 Mbit/s
Source: Internet Diffusion Case Studies at: http://www.itu.int/ti/casestudies
Broadband in Europe, installed
lines, per 100 inhabs, Dec. 2000
UK
Spain
DSL
Italy
Cable modems
Germany
Fibre
Portugal
France
Norway
Switzerland
Finland
Iceland
Denmark
Sweden
Belgium
Netherlands
Austria
0
Source: OECD, ITU.
0.5
1
1.5
2
Broadband outside Europe, installed
lines, per 100 inhabs, Dec. 2000
New Zealand
DSL
Australia
Cable modems
Japan
USA
Canada
Korea (Rep.)
0
Source: OECD, ITU.
2
4
6
8
10
Barriers to broadband
deployment
 Technical barriers
 Technical constraints on DSL technologies and cable
modems on older cable TV networks.
 Regulatory barriers
 Where infrastructure competition is not permitted,
roll-out had been slower and prices higher.
 Where there is cross-ownership between cable TV
operators and incumbent telcos, roll-out is slower.
 Economic barriers
 High costs for network roll-out = high prices
 Financial crisis is slowing down deployment
 Perceived lack of “killer applications”
Connectivity and economic
opportunity
 Connectivity in the 21st Century will play an equivalent
role to trade routes in earlier eras
 Trade in information goods will outweigh trade in goods
 Connectivity differences between economies are much
greater than would be predicted by differences in wealth
or market structure:
 There is an International IP connectivity divide between
developed and developing nations
 Disparities in broadband access between developed countries
 International digital divide initiatives have focused on
using ICTs to extend access to healthcare, education,
governance etc
 BUT, without connectivity, demand side initiatives are largely
irrelevant
ITU Internet Diffusion
Case Studies
 2000
 6 studies covering Bolivia,
Egypt, Hungary, Nepal,
Singapore, Uganda
 2001
 Regional focus on ASEAN
 Purpose:
“To understand the factors
that may promote or retard
the development of the
Internet in different
countries”
www.itu.int/ti/casestudies