Why some economies succeed with
broadband
Sam Paltridge
OECD*
http://www.oecd.org/sti/telecom
ITU Promoting Broadband Workshop
9th April 2003
*The views expressed are those of the author and may not reflect those of the OECD or its Member countries.
1
Changes in the proportion of Communication in
disposable households incomes
*Communication includes Telecommunication equipment and services and Postal services. Note: Hungary, Norway, Slovak Republic, Switzerland and
Turkey are not included in this index.
Source: OECD, SNA database
160
Index: 1990=100
Communications*
150
Health
140
Education
130
Water, electricity and gas
120
Recreation and Culture
110
Transport
100
Restaurants and hotels
90
80
Alcohol, tobacco and
narcotics
Households equipment
70
Clothing and footwear
Food
60
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2
Why is broadband important to
telecommunication carriers?
Telecommunications revenue in the United States
Trends in Revenue reportedby UnitedStates Telecommunication Providers
Local (1992=100)
1200
Other
400
Long Distance (1992=100)
350
Wireless (1992=100)
300
Wireless
1000
2002 data for
Other not yet
available
Other including Internet Access
(1992=100)
600
Long Distance
Local
250
Billion $
800
Internet Access included
under Other. 2002 data not
yet available.
200
150
400
100
200
50
0
0
1992
1993
1994
1995
1996
1997
Source: FCC
1998
1999
2000
2001
2002
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
Source: FCC
3
2002
Why is broadband important to the rest of us?
!
!
!
!
Broadband provides first “always-on, affordable, high-speed” access for
residential users, some public services, small business, and the possibility
for employees of larger firms to tele-work in new and innovative ways.
Role of ICTs in development is an ongoing debate but the economic and
social benefits are still evident amid the current slow-down in the telecom
services sector (which by the way is still growing).
Broadband access is the next step in a series of ICT developments but
penetration is low (5 subscribers per 100 inhabitants across OECD and 0.2
per 100 outside OECD).
Broadband digital divide:
– Some OECD countries have barely started while one country is pondering whether it has
reached a penetration ceiling.
– DSL availability ranges from “not offered” through to 98% population coverage.
– Some sell broadband at ISDN speeds while others are dramatically increasing baseline
offers and extending the reach of fixed broadband via wireless LANs.
!
At this stage, it is easier to answer why broadband succeeds in some
economies than why some economies succeed with broadband!
4
What is the potential market for
broadband access?
!
!
!
!
!
!
Alcatel puts the value of the global DSL services market, in 2002, at US$14 billion.
Adding other broadband access would lift that number well above $22 billion. Small in
terms of overall telecoms market but at very early stage of broadband development.
How much can market grow? One approach is to look at how many Internet subscribers
there are and calculate how many will adopt broadband. In OECD area somewhere
between 18% to 25% of all fixed network Internet subscribers already have broadband
access (albeit with huge variations across countries).
Anecdotal evidence suggests most users migrate from dial-up to broadband. On the other
hand, Telekom Austria claims that 40% of its broadband subscribers previously had no
Internet access!
Broadband access via Wireless LANs adds another dimension. The increasing
incorporation of wireless enabled chips in less inexpensive communication devices and
development of prepaid cards for W-LANs mean that any predictions of market size may
be as accurate as those for mobile telephony.
Value added services? Broadband impact on growth and distribution of revenue in the
information industries?
How much capacity is enough to meet demands of users on the fixed network? Is there
demand for wireless broadband via LANs? Japan and Korea may provide the first
indications.
5
DSL Broadband Divides
1. Japan – excludes Fibre at 100 Mbps & Korea excludes VDSL at 20 Mbps.
2. Iceland and USA highest capacity aimed at business users (e.g. Verizon: 7.1 Mbps = $204) .
Luxembourg
Italy
Australia
Netherlands
Switzerland
Mexico
Finland
Turkey
Slovak Rep.
Denmark
Iceland
Czech Rep.*
Hungary
New Zealand*
Portugal
Poland
Ireland
France*
UK
Sweden
Spain
Austria
Norway
Germany
United States
Canada*
Belgium
Korea
Japan
0
2
*Excluding offers below 256 kbps
4
6
8
10
12
14
Mbps
6
Question for carriers is what users want from broadband: Will users migrate
from 8-12 Mbps to 100 Mbps? Japan may provide first demand side clues.
7000000
DSL Subscribers in Japan
5000000
Cable Modems Subscribers in
Japan
Fibre to the Home
4000000
Yahoo BB (IP Telephony)
3000000
2000000
1000000
0
Ja
n0
M 0
ar
M 00
ay
-0
0
Ju
l-0
Se 0
p0
No 0
v0
Ja 0
n0
M 1
ar
M 01
ay
-0
1
Ju
l-0
Se 1
p0
No 1
v0
Ja 1
n0
M 2
ar
M 02
ay
-0
2
Ju
l-0
Se 2
p0
No 2
v0
Ja 2
n03
Number of Subcribers .
6000000
7
Wireless LAN 802.11b services in Korea
140000
KT WLAN Subscribers
120000
100000
80000
Hanaro WLAN Subscribers
KT WLAN ARPU =
US$10.24 per Month
(January 2002 )
60000
40000
20000
Flat Monthly Rate = US$20
Nespot Lite = US$8 for 5
hours (1.5 cent per additional
minute)
0
Mar-02 Apr-02 May-02
Jun-02
Jul-02
Aug-02
Sep-02
Oct-02 Nov-02 Dec-02
Jan-03
Feb-03
8
NT
T
120
Monthly
300 minutes over 5 days
100
One Day
Te
lia
140
TBT
M
ob
ile
(U
SA
)
AT
W
&T
ay
po
rt
Si
(U
pa
SA
nd
)
Su
rf
(U
SA
)
Ae
rv
P
CC
ik
(N
W
Sw
eth
er
iss
lan
co
ds
m
)
(m
eg
ab
ea
Te
m
lst
)
ra
Sk
yn
W
et
iF
ix
(F
M
ran
etr
on
ce
et
)
(A
us
t ri
a)
Ca
na
Bo
da
)
ing
De
o(
fau
US
lt
Ci
A)
ty
(S
we
de
n)
Sp
otn
ik
(
KT
US$
Wireless-LANs: 802.11b pricing in selected
countries: Carriers are testing the market and
experimenting with pricing.
160
80
60
40
20
0
9
The good news is that broadband (or high speed access) is growing quickly
with more than 62 million subscribers in the world by end 2002. This
represented a 70% increase over 2001 and 4th quarter 2002 was highest
quarterly growth yet.
70
60
Broadband outside OECD Area (from 2001
onwards)
Other Platforms in OECD Area
Subscribers, Million
Cable Modems in OECD Area
50
DSL in OECD Area
40
30
20
10
0
1999
2000
2001
2002
10
Ko Kore
ng ,
a
Ch
in a
Ch
C
ine anad
se
Ta a
ip
I ce ei
la
Be n d
lg i
De um
nm
a
S rk
Ne wede
th
n
Un erlan
i ted d s
Sw State
itze s
rla
Au nd
s tri
a
Jap
a
n
Fin
Sin land
gap
No ore
r
Ge way
Ma
r
cao many
,C
hin
a
Sp
ain
Fra
nce
Isr
ae
Un
i ted Po rtu l
g
Kin al
gdo
Est m
o
Au nia
s tra
lia
Ne
Ita
w Z ly
Lu eala
x em nd
b ou
rg
Ch
i
Hu le
n ga
ry
Cz
ech Ch i
n
a
Re
p ub
l
Ire ic
lan
Me d
xic
Po o
lan
Slo
Tu d
rk
vak
R e ey
pu b
Gr lic
eec
e
Ho
ng
Broadband Subscribers per 100 Inhabitants, 2002
Broadband penetration is very uneven throughout the
world so we naturally look to leading countries for
what works
25.00
20.00
15.00
10.00
5.00
0.00
11
2.5
Competition over DSL
as per ITU case study
Anyone like to share? How to
get 6 Euro access!
Competition: Cable versus DSL
2.0
Competition over incumbent local loops - $21 for 8 Mbps!
1.5
1.0
DSL versus cable = 3 Mbps and 98% DSL coverage
Dancing on the ceiling?
More than half Korean
households have broadband
so new subscriber growth
has slowed while existing
subscribers are upgrading
to higher speeds.
0.5
0.0
Ic
el a
F nd
Sw inla
itz nd
er
lan
Ja d
p
Be an
lg
ium
Ne Fran
the ce
rla
No nds
De rwa
nm y
a
Au rk
st
Sw ri a
ed
Ca en
n
Po ada
rtu
Un
ga
l
i te
K
or
d
Ki ea
ng
do
m
Un
i te Spa
i
d
St n
Ge ates
rm
an
y
Ita
A ly
Ne ustr
w ali
Ze a
a
Hu land
ng
ar
Cz
ec Irel y
h R an
Lu epu d
xe bl
mb ic
ou
Po rg
la
M nd
ex
i
Tu co
rk
e
Sl
ov Gr y
ak
e
Re ece
pu
bli
c
New subscribers per 100 inhabitants, 4thQ 2002
Who is growing the fastest: Broadband
Growth Rates in OECD countries
12
40.0
SMAT Vs can change t heir business
plan and receive t he signals from T DC
or ST OFA.T his decision must ,
however, be t aken at t he headend of
t he SMAT Vs and not at individual basis
35.0
30.0
Red - Incumbent telecommunication
carrier owns part of cable industry.
Blue - Independent cable companies
Aust ralia and P ort ugal
have cable companies
compet ing down t he same
st reet in some regions.
25.0
20.0
NZ broadband pricing
plans are not popular met ered downloads.
DSL compet it ion
very recent so
used capit al
elsewhere.
15.0
10.0
5.0
Po
ra
lia
K
U
us
t
rtu
ga
l
Sw
ed
en
N
or
w
ay
Fi
n
N
ew land
Ze
al
an
d
Fr
an
ce
H
un
ga
ry
Tu
rk
ey
Ire
la
nd
G
er
L u ma
x e ny
m
bo
ur
g
N
A
us
et tri a
he
r la
U
n
ni
ted d s
St
at
es
Ja
Sw p a
itz n
er
la
n
Be d
lg
iu
m
D
en
m
ar
k
Sp
ai
n
ad
a
A
Ca
n
or
e
a
0.0
K
Cable Modem subscribers as % of homes passed by cable
If one company owns both available platforms there
is a lower take-up of cable modems and lower
overall market growth. This impacts on a third of
OECD countries.
13
Telecommunication carriers have a low
broadband take up rate on their cable networks.
30
25
%
20
Cable ModemSubscribers to
homes passed (Independent
Companies)
Cable ModemSubscribers to
homes passed (Telcos)
15
10
5
0
14
What doesn’t work
!
!
!
!
Monopolies
Telecommunication Carriers owning cable networks
Truck Rolls (Need Self Install)
No evidence that low speed ‘always-on” offers (e.g. 128 kbps, 144 kbps) are
popular compared to broadband but data are hard to find:
– In Denmark where a 144kbps cable service is on offer only 13% of cable users and
less than 5% of overall market use that option.
– “If poorly understood, consumers may be seriously disappointed when they realise
that although they had been promised high-speed Internet access, in practice, it is not
much better than dial-up.” ART-Telecom “Internet, a review of the French
market”, March 2003
!
Unattractive Pricing
– Low take-up rates in Australia and New Zealand suggest many potential users do not
like low download caps (e.g. 500 Mbytes) and metered pricing as they limit use of
streaming media. The contradiction is recognised in differentiation of pricing, by
incumbents, between their own content/services and those of others.
– Very high prices, of course, are constraining growth in a number of OECD countries.
15
What doesn’t work (continued)
!
Subsidies to suppliers:
– Korea’s success was not due to government funding. Low cost loans initiated
during Asian financial crisis but operators, such as Hanaro, soon found they could
get less expensive capital elsewhere as crisis eased. Korean government did
stimulate the backbone market by financing capacity and then being repaid by the
utilisation of government agencies. But backbone markets are not the barriers to
broadband in developed countries and a number of governments have funded
backbones in various ways and not experienced success in broadband access.
Subsidies are no substitute for competition and a lack of access competition usually
means investment is deployed ahead of demand. Liberalisation is the best way to
stimulate backbone construction.
– But “middle mile” backhaul problem clearly does exist in some rural areas. Demand
aggregation by public sector users is one option as long as it is applied in a way that
does not distort competition. Any government funding should be for users rather
than suppliers to address problem.
– Universal service policies for broadband, in terms of last mile access, are
premature. What would be provided? Low cost solutions are being found by users
and new entrants. For example, increasing use of wide area wireless LANs is
addressing the ‘last mile’ problem for small rural towns. An increasing number of
WISPs (Wireless ISPs) are serving rural areas. (Antennas on locations such as
wheat silos are being found to have extensive coverage areas and the technology is
rapidly evolving).
16
What does work?
!
Policies aimed at promoting competition
– Facilities competition is the best option (e.g. Korea, Canada)
– Unbundling and line sharing are tools to open the market to competition and can accelerate
growth (e.g. Denmark, Iceland, Japan)
– Cable divestiture by incumbent telecommunication carriers have proven its worth in
countries such as the Netherlands and Switzerland.
– Ensuring spectrum is available for innovative solutions.
– Independent regulator
!
Actual Broadband
– The baseline offers in Belgium, Korea and Japan all target much higher broadband
performance levels including superior upstream capabilities.
!
Pricing
– Metered pricing can be an option but need to have a reasonably priced flat rate options for
those users that prefer certainty and want to use broadband for streaming media.
– High growth rates are invariably linked to operators reaching a price point and structure
attractive to users.
!
Benchmarking
– Introduce regular reporting on the availability and take-up of broadband services.
!
All of above
!
– Competitive entry with innovative technology, services, pricing and regulatory safeguards.
Much else but that is usually not in the hands of telecommunication policy makers and
telecommunication carriers.
17