technical assistance in bridging the “digital divide” - a cost benefit analysis for broadband connectivity in europe - pwc_final_report

Prepared by:

Final Report
Technical assistance in bridging the “digital divide”:
A Cost benefit Analysis for Broadband connectivity in Europe

6 October 2004
With:



EC/ESA Digital Divide CBA – Final Report 6 October 2004

Contents
1 Introduction 4
1.1 Background and Objectives 4
1.2 Issues 4
1.3 Our Approach 5
2 Existing level of broadband connectivity across Europe 7
2.1 Methodology 7
2.2 Terrestrial broadband availability 8
2.3 Terrestrial broadband availability forecasts: 2004 - 2013 11
2.4 Satellite broadband availability 14
2.5 Broadband take up: 2004 – 2013 16
2.6 Potential broadband demand remaining unserved 22
2.7 Availability and take up of terrestrial broadband - conclusion 25
3 Combined system and user costs of broadband connection 27

3.1 Costing considerations 27
3.2 Capital costs 28
3.3 Operational costs 36
3.4 Typical cost breakdown by technology 36
3.5 Average cost per user 39
3.6 Total costs by Country Group 40
3.7 Costs of providing service to Objective 1 regions 41
3.8 Summary of technology mix 42
4 Optimal technical solution for broadband connectivity 44
4.1 ADSL and Derivatives 45
4.2 Satellite 53
4.3 Hybrid Satellite/BFWA/WiFi 60
4.4 Affordability and User Willingness to Pay 63
4.5 Conclusions on Optimal Mix of Technologies 68
5 Benefits of the provision of broadband connectivity across Europe 71
5.1 Introduction 71
5.2 Economic concepts 71
5.3 Implementing the methodology 78
5.4 Direct benefits to subscribers 84
5.5 Benefits to public sector suppliers of services (Step 3) 97
5.6 Regional analysis of benefits 118
5.7 Conclusions: Overall value of broadband benefits 120
6 Conclusions on cost benefits for broadband connectivity 122
6.1 Assumptions for the cost benefit calculation 122
6.2 Results of the cost benefit calculation 127
6.3 Conclusions 135
7 Framework for the preparation of a business plan 137
7.1 Background and Objectives 137
7.2 Issues 138
7.3 Our Approach 138

Appendix 141
(1)



EC/ESA Digital Divide CBA – Final Report 6 October 2004


Glossary

ADSL Asymmetric digital subscriber line
ADSL2 and
ADSL2+
Enhanced ADSL – greater bandwidth on
short-mid length lines
ARPU Average Revenue Per User
B2C Business to Customer
BFWA Broadband Fixed Wireless Access
CAPM Capital Asset Pricing Model
CATV Cable Television
CBA Cost Benefit Analysis
EC European Commission
EMRP Equity Market Risk Premium
ESA European Space Agency
EU European Union
FTTB/U Fibre To The Building/User
FTTC Fibre To The Cabinet/Curb
GCSE General Certificate of Secondary
Education
HFC Hybrid Fibre Co-ax

ICT

Information and Communications
Technologies
LAN Local Area Network
NPV Net Present Value
PLC Powerline Communications
PPP Public Private Partnership
PwC PricewaterhouseCoopers LLP
SME Small & Medium-Size Enterprises
SoHo Small Office Home Office
VAT Value Added Tax
VoIP Voice Over Internet Protocol
VSDL Very high speed digital subscriber line
WACC Weighted Average Cost of Capital
WiFi WLAN used in the wide area
WLAN Wireless Local Area Network

(2)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

ESA STUDY MANAGER

Pierluigi Mancini Ph.D.
Applications Strategy Manager
Directorate of EU and Industrial Programmes
Email:



This document has been prepared on the instructions of the European Space Agency (“ESA”) and with only
ESA's interests in mind. PricewaterhouseCoopers LLP, its members, partners, employees and agents neither

owe nor accept any duty or responsibility to other parties (“you”), whether in contract or in tort (including
without limitation, negligence and breach of statutory duty) or howsoever otherwise arising, and shall not be
liable in respect of any loss, damage or expense of whatsoever nature which is caused by any use you may
choose to make of this document, or which is otherwise consequent upon the provision of the document to
you. Without conferring any greater rights than you would otherwise have at law, we accept that this
disclaimer does not exclude or indemnify us against any liability we may have for death or personal injury
arising from our negligence or for the consequences of our own fraud. This report represents the Phase I
analysis on behalf of the ESA of the state of the Digital Divide in Europe,, on a technology neutral basis. The
report will provide inputs for a second phase to develop a business case to address the Divide, where further
analysis and/or refinement of the findings will be carried out as necessary to gain greater specificity.



(3)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

1 Introduction
1.1 Background and Objectives
In line with the strategic vision adopted by the European Union, the European society
of tomorrow will be a 'knowledge-based society' in which access to knowledge is for
all. Information outreach must therefore be considered as a major objective for
Europe. This means extending access to electronic communication services to
everyone, notably in those regions of our continent, which up until now have been
disadvantaged or neglected (rural and mountain areas, islands, far-flung outlying
regions). The enlargement of the Union adds a further dimension to the digital-divide
issue.
The European Space Agency and the European Commission have decided to give

joint consideration to the use of space-based technologies to help bridge 'the digital
divide'. In its White Paper on Space Policy, the Commission has placed this issue -
satellites contributing to bridge the digital divide - at the forefront of Europe's
strategic priorities in the run-up to 2007.
With this as a background, PricewaterhouseCoopers LLP, together with Ovum and
Frontier Economics their subcontractors, have been invited by ESA, to prepare a
detailed, technology-neutral Cost Benefit Analysis on broadband connectivity for an
expanded Europe.
The study’s objectives are:
• Through both primary and secondary research, identify and quantify, where
possible, the benefits of providing broadband connectivity across an extended
Europe;
• Prepare an objective analysis of the optimal technical solutions for broadband
connectivity, region by region;
• Provide an independent perspective on the Cost Benefit ratio for technology-
neutral provision of broadband for Europe over a period of at least 10 years; and
• Clarify the potential role of satellite-based solutions for broadband connectivity
and thus bridging the digital divide in an extended Europe.
1.2 Issues
Some of the issues the Study considered include:
• The extent to which Europe (and not the USA) will derive economic benefit from
any EU-initiated project to tackle the digital divide;

(4)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

• Quantification of user benefits – from societal and environmental, to economic

and industrial;
• Understand what benefits are actually incremental, resulting directly from new
user broadband connectivity; and
• How best to identify the optimal technical solution for the provision of
broadband service to new users across an extended Europe.
1.3 Our Approach
Our approach to the study is summarised below.
• We reviewed the existing level of broadband connectivity across Europe
• We analysed the combined system and user costs of a broadband connection by
each technical solution
• We identified the optimal technical solution for provision of future broadband
connectivity across different regions of Europe
• We quantified the benefits of the provision of broadband connectivity across
Europe
• We calculated the net benefits of broadband connectivity across an extended
Europe
• We set out a framework for the next steps following this study, including the
preparation of a business plan based on our findings
Throughout our study we have analysed two different scenarios of broadband
availability and take-up - a base case and extended take-up case. For clarity, the base
case and extended case assumptions were as shown in the tables below:
Table 1: Broadband connectivity and forecast take-up in Europe
A
vailabilit
y
Base Case Extended Case
A
vailabilit
y
Base Case Extended Case

Country Group 1
Urban 98% 54% 56% 99% 76% 82%
Rural 92% 36% 39% 95% 54% 64%
Country Group 2
Urban 97% 37% 41% 97% 74% 81%
Rural 62% 19% 20% 78% 39% 42%
Country Group 3
Urban 96% 45% 50% 97% 75% 81%
Rural 61% 12% 13% 75% 28% 30%
Country Group 4
Urban 94% 13% 14% 95% 43% 47%
Rural 15% 1% 2% 34% 8% 10%
2008 2013



(5)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

PwC prepared a number of Technical Notes in the course of this study which were
individually presented and discussed in detail with a large number of representatives
from the European Space Agency and the European Commission. The content of these
Technical Notes, as well as the comments received from the stakeholders within both
ESA and the EC are reflected in this final report.

(6)

EC/ESA Digital Divide CBA – Final Report 6 October 2004

2 Existing level of broadband connectivity across Europe
2.1 Methodology
The current and estimated future broadband connectivity in Europe has been
determined based on research of the current levels of broadband availability and take
up, historic trends of broadband availability and take up, reference to other published
material and modelling of estimated future availability and demand. The methodology
used in the study to determine the estimated availability and take up of broadband
services is briefly described in this Section of the report, and described in more detail
in the appendix to this report. The costs associated with the deployment of services are
also described in the appendix.
Each of the 25 EU countries has different characteristics, e.g. internet usage, GDP,
price of services, market competitiveness, which cause the provision and use of
broadband to vary from country to country. To assist the modelling, the 25 countries
of the EU were split into 4 country groups, in order that one or two countries per
group could be considered in detail and the results from this detailed analysis then
scaled to the other countries in the group. The factors considered most important when
the groups were determined were: the expected levels of adoption of broadband within
the forecast period of 2004-2013; GDP; the split of urban and rural population
1
and
the current percentage of population in EU Objective 1 areas; the current penetration
of telephone lines and the take-up of broadband services. The methodology by which
the country groups are determined is described in the appendix to this report. This
analysis was underpinned by preliminary regional analysis within two countries, Italy
and the UK. If the modelling were extended to cover the non-EU ESA member
countries, Switzerland and Norway would come within Country Group 1.

With high current and expected future levels of broadband availability and take-up,
medium to low levels of rural population and high disposable income, the countries in
Group 1 are: Austria; Belgium; Denmark; France; Luxembourg; Malta; Netherlands;
Sweden; United Kingdom. The take up and development of broadband services in the
UK and France have been used as a basis for modelling Group 1 countries.
Countries in Group 2 have a higher percentage of population in EU Objective 1
2
areas
(10-35%), lower average current levels of broadband adoption, and have high


1
The cost modelling used same data for urban and rural population as used for the benefits analysis.
The data was sourced from the European Social Survey, and is shown in Figure 2.2.
2
A region whose development is lagging behind others in the EU is given Objective 1 status to
promote its development and structural adjustment. Objective 1 areas may be urban as well as rural.
/> defines Objective 1 regions for the period 2000-
2006.

(7)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

disposable income. The countries in Group 2 are: Finland; Germany; Ireland and Italy,
with Italy being used studied in more detail for the project.
Those countries in Group 3, with high levels of the population living EU Objective 1
regions, medium levels of rural population and reasonable levels of current broadband

adoption are: Cyprus; Estonia; Portugal; Slovenia and Spain.
Group 4 is mainly comprised of the 2004 EU member countries of Central and Eastern
Europe, together with Greece, which has very low levels of broadband take-up. These
countries, which have medium-high levels of rural population and medium levels of
GDP and disposable income, are: Czech Republic; Greece; Hungary; Latvia;
Lithuania; Poland and Slovakia. The take up and development of broadband services
in Poland been used as a basis for modelling Group 4 countries.
Brief country profiles, for UK, France, Italy, Slovenia, Poland and Greece, together
with four regional case studies (two in Italy and two in the UK), which highlight the
issues facing different regions of Europe and which provide a qualitative analysis of
the issues faced during the development of broadband and infrastructure across the
digital divide, are provided in Appendix C to this report.
2.2 Terrestrial broadband availability
The scope of the problem of broadband provision in Europe can be seen in Figure 2.1,
showing the variations in population density across Europe. This shows the population
density at the NUTS 3
3
level. There are large rural regions of thinly populated space in
Sweden, Finland, Ireland, Greece, central France and central Spain. Low population
density is also found in north-eastern Germany and in most of the Baltic States.
Within these areas of low population, it must be recognised that there are many
distribution profiles. This uneven population distribution is a challenge for service
providers.
The availability of broadband in the EU is dominated by ADSL and cable modem
services and this is expected to continue. Cable modem services, built on top of cable
TV infrastructure are primarily provided in urban areas and their provision is very
varied across the EU. They are rarely cost effective to deploy in isolated urban areas
and rural areas. DSL services are provided in urban areas and increasingly in more
rural locations. There are currently low levels of availability of fibre to the user /
building (FTTU/B), broadband fixed wireless access (BFWA) and WiFi (wireless

3
Nomenclature of Territorial Units for Statistics (NUTS) are regional areas defined to allow a
common statistical classification to enable the collection, compilation and dissemination of
harmonised regional statistics in the Community. NUTS 3 regions have a minimum population of
150,000 and a maximum population of 800.000.


(8)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

LAN technology used in the wide area). However, where FTTB is available, it has a
dramatic effect on the speed and range of services that can to be offered, as can be
seen in Italy and Sweden. In the latter years of the forecast period, when demand for
services at above 10Mbps is growing particularly among business users, it is
anticipated that deployment of fibre will have started in many countries, but few will
have significant FTTB networks, which will be limited to urban areas. However, fibre
deployment to street cabinets (FTTC) is expected to be more widely used, taking
advantage of high capacity DSL services (e.g. ADSL2+ and VDSL).
Figure 2-1 Europe's population density for NUTS3 levels areas

Source: IIASA European Rural Development Project

(9)

EC/ESA Digital Divide CBA – Final Report 6 October 2004

Figure 2-2 Urban population in EU countries

Urban
Population
Urban
Population
Urban
Population
Austria
56%
Greece
68%
Poland
56%
Belgium
46%
Hungary
60%
Portugal
64%
Cyprus
45%
Ireland
54%
Slovakia
45%
Czech Republic

68%
Italy
64%
Slovenia
45%
Denmark
70%
Latvia
45%
Spain
56%
Estonia
45%
Lithuania
45%
Sweden
69%
Finland
57%
Luxembourg
49%
UK
74%
France
74%
Malta
45%
Germany
74%
Netherlands

53%
Source: European Social Survey



The availability of these DSL and cable modem services cannot be considered
cumulative. The country groupings, and the range of broadband availability to (as %
of population) in each country group are given in Figure 2.3 and the availability of
broadband in 2004 is shown in Figure 2.4.
There are currently low levels of availability of fibre to the user / building (FTTU/B),
broadband fixed wireless access (BFWA) and WiFi (wireless LAN technology used in
the wide area). However, where FTTB is available, it has a dramatic effect on the
speed and range of services that can to be offered, as can be seen in areas of Italy and
Sweden.

Figure 2-3 Broadband availability by Country Group in 2004
Countries in Group Broadband availability –2004
Group 1 UK, France, Belgium, Denmark,
Luxembourg, Netherlands,
Austria, Sweden and Malta
78% (Sweden) – 98% (Denmark)
Group 2 Italy, Germany, Finland and
Ireland
62% (Ireland) – 85% (Germany and
Finland)
Group 3 Spain, Portugal, Estonia,
Cyprus and Slovenia
50% (Estonia, estimate) – 92% (Spain)
Group 4 Czech Republic, Hungary,
Poland, Slovakia, Greece,

Latvia and Lithuania
Very low (Greece) – 85% (Lithuania)
Source: Ovum


(10)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

Figure 2-4 Broadband availability by country - 2003
0.0%
20.0%
40.0%
60.0%
80.0%
100.0%
Belgium
Denmark
Spain
Luxembourg
UK
Finland
Germany
Lithuania
Netherlands
France
Austria
Italy

Latvia
Sweden
Poland
Hungary
Ireland
Portugal
Czech
Estonia*
Malta*
Slovenia*
Cyprus*
Slovakia*
Greece*
Cable modem DSL * Ovum Estimate

Source: OECD
4
(DSL – YE2003; Cable 2003) and Ovum
The availability of broadband in all countries and especially the 10 EU accession
countries demonstrates that it is not difficult to make rapid progress in providing
broadband access to large proportions of the population. The challenge lies in
reaching beyond the urban and more tightly grouped rural regions. This is being
achieved through a combination of commercial WiFi, BFWA and satellite services,
public sector financial support and other initiatives such as regional aggregation and
awareness campaigns.
2.3 Terrestrial broadband availability forecasts: 2004 - 2013
To develop the broadband take up model, we have made assumptions about the basis
of availability by access speed in 2004 and assumptions about the way in which
terrestrial broadband service availability will increase in each proxy country, by
access speed and type of user. The forecasts for each proxy country are derived from

Ovum’s detailed broadband forecast methodology. The Ovum broadband team works
closely with network operators, service providers, industry bodies and other players in
the market. It undertakes analysis of what has gone before – the size and rate of
market development, its regulatory environment and value drivers (e.g. content,
services). The model forecast takes into account these factors and also the
development of economic and demographic factors.



4
DSTI/ICCP/TISP(2003)7/FINAL: The Development of Broadband Access in Rural and Remote
Areas, 10-May-2004

(11)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

The broadband availability forecasts also assume that broadband services will be
made available where it is commercially viable for service providers to do so, which is
related to national prosperity. The availability of cable TV networks stimulates the
broadband market, as cable service providers have sunk their initial costs for cable to
the home (in more densely populated areas) and can easily offer attractive service
bundles, without the greater costs that DSL service providers have to bear.
The drivers for increasing availability and take-up of broadband will be similar in all
country groups. Initially consumers will require those services available with narrow-
band internet access, shown in Figure 2.5 at the lower service levels. As consumers
become more familiar with the service and new products become available, so usage
and bandwidth demand will increase.

Figure 2-5 Bandwidth and service level requirements of selected consumer applications

Source: Point Topic
Figures 2.6 – 2.9 show the forecast availability of broadband services for residential
and SoHo users. It is assumed in all cases that SMEs and large enterprises will have
higher broadband availability as they are a desirable customer sector and business
areas are amongst the first to be served. The charts showing service availability for all
broadband user groups are provided in Appendix A to this report. The rate of growth
and the limit of take up will vary between country groups. The major factors are
discussed below:
• Group 1 countries have very mature telecoms markets in both urban and rural
regions. The regulatory environment has adapted to create a strong competitive
market which is assisted by high levels of GDP and annual disposable income.
Those areas which do not have terrestrial broadband services available are few,

(12)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

and the incumbent is often leading the activity to reach nearly 100% availability,
enabling the remaining smaller rural exchanges for DSL. The assumed
availability by access speed for residential and SoHo users in urban and rural
areas is shown in Figure 2.6.

Figure 2-6 Broadband availability assumption – Group 1 countries
Residential and SoHo - Urban Availabity
-
10,000

20,000
30,000
40,000
50,000
60,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
(,000 )
Urban <512kbps Urban 512kbps-10Mbps Urban >10Mbps
Residential and SoHo - Rural Availability
-
2,000
4,000
6,000
8,000
10,000
12,000
14,000
16,000
18,000
20,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
(,000 )
Rural <512kbps Rural 512kbps-10Mbps Rural >10Mbps

Source: Ovum
• Group 2 countries are similar to Group 1 countries. They have mature,
competitive markets, and high GDP levels and high annual disposable income,
but have large areas which are less attractive to serve, owing to lower GDP,
terrain or general level of development, leading to overall lower levels of
broadband adoption. The assumed availability by access speed for residential and

SoHo users in urban and rural areas is shown in Figure 2.7
• Group 3 countries have a high proportion of their population in Objective 1
designated areas. There are medium levels of annual disposable income per
household and the availability of broadband is varied. The countries have good
broadband availability in urban areas, but face a challenge in rural areas. The
assumed availability by access speed for residential and SoHo users in urban and
rural areas is shown in Figure 2.8
• Group 4 includes most accession countries which have medium levels of rural
population and medium levels of disposable household income. The broadband
market in these countries is just starting to develop but is restricted by lower
levels of technology and telephone line adoption. Operators are primarily
concentrating their broadband offerings in urban areas. The assumed availability
by access speed for residential and SoHo users in urban and rural areas is shown
in Figure 2.9.
Terrestrial broadband availability is complemented by the availability of satellite
services, which are discussed in Section 2.4.

(13)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

Figure 2-7 Broadband availability assumption – Group 2 countries
Residential and SoHo - Urban Availabity
-
5,000
10,000
15,000
20,000

25,000
30,000
35,000
40,000
45,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
(,000 )
Urban <512kbps Urban 512kbps-10Mbps Urban >10Mbps
Residential and SoHo - Rural Availability
-
2,000
4,000
6,000
8,000
10,000
12,000
14,000
16,000
18,000
20,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
(,000 )
Rural <512kbps Rural 512kbps-10Mbps Rural >10Mbps

Source: Ovum
Figure 2-8 Broadband availability assumption – Group 3 countries
Residential and SoHo - Urban Availabity
-
1,000
2,000

3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
(,000 )
Urban <512kbps Urban 512kbps-10Mbps Urban >10Mbps
Residential and SoHo - Rural Availability
-
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
(,000 )
Rural <512kbps Rural 512kbps-10Mbps Rural >10Mbps

Source: Ovum
Figure 2-9 Broadband availability assumption – Group 4 countries
Residential and SoHo - Urban Availabity

-
2,000
4,000
6,000
8,000
10,000
12,000
14,000
16,000
18,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
(,000 )
Urban <512kbps Urban 512kbps-10Mbps Urban >10Mbps
Residential and SoHo - Rural Availability
-
1,000
2,000
3,000
4,000
5,000
6,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
(,000 )
Rural <512kbps Rural 512kbps-10Mbps Rural >10Mbps

Source: Ovum

2.4 Satellite broadband availability
In theory broadband satellite services have 100% availability outside local blackspots
such as urban canyons (where there is no line of sight to the satellite). Europe is

(14)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

served by several different satellite operators whose spacecraft provide excellent
coverage of the entire EU area. This notably includes the members of the European
Satellite Operators’ Association: SES Global, Eutelsat, New Skies, EuropeStar,
EurasiaSat, Hellassat, Hispasat, Inmarsat, NSAB, Telenor and Telespazio. Most of
these operators have substantial capacity available at Ku-band and to a much lesser
extent at C-band. In addition, the major operators are incrementally introducing new
capacity at Ka-band, specifically intended to serve broadband markets. Finally, the
mobile satellite operator Inmarsat is launching new L-band satellites this year, one of
which is intended to support broadband mobile services across Europe from 2005.
Yet the real availability of broadband satellite services depends upon many other
things besides in-orbit capacity. Experience has shown that the takeup of such services
requires all of the following to be simultaneously in place:
• national and Europe-wide licensing of both spectrum and user terminals (which
must be straightforward and inexpensive)
• effective distribution channels (for both user terminals and integrated services
packages)
• proven high QoS levels
• customer support services (with rapid turnout)
• affordable usage charges
• affordable user terminal prices.
At present few of the above elements are in place to a satisfactory degree in any EU
country. On the other hand, 2004 is the first year in which most of these issues are
starting to be resolved. Their current status and future prospects are reviewed in detail

later in this report, with particular consideration given to:
• the status and capabilities of distribution channels, including customer support
• the importance of customer faith in good QoS
• the competitiveness of satellite usage charges compared to terrestrial services,
and as a subset of this the impact of the price of user terminals upon demand,
including setup and licensing costs and the effect of subsidies and grants where
these exist.


(15)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

2.5 Broadband take up: 2004 – 2013
2.5.1 Market development
The forecasts for take up are based on the market continuing to develop with healthy
competition and product development, taking into account:
• broadband services being made available where it is commercially viable for
service providers to do so
• lower service prices, driven by competition, efficiencies of scale and network
operator cost reductions
• lower differentials between narrow-band and broadband prices, making it
increasingly attractive for consumers to switch to broadband
• increasingly attractive service packages and bundled services (TV, video-on-
demand, Internet access, voice and content)
• continuation of some national and regional initiatives to stimulate availability
and take-up.
The forecasts do not take into account any new initiatives to make significant

intervention into the market.
2.5.2 Established broadband take up in 2004
At the end of 2003 there were 22.7 million ADSL and cable modem broadband users
in the 25 countries that now make up the EU. Research carried out in Q1 2004
5
shows
that in the second half of 2003:
• overall broadband growth in the major EU countries ranged from 15% in Austria
to over 57% in Italy. France, Hungary, Italy, Portugal and the UK all showed
more that 44% growth. However, it should be noted that these growth figures
may be from a very low base
• DSL growth ranged from 16% to 75%, with strongest growth in France,
Hungary, Italy and the UK
• cable modem growth ranged from 0% to 43%, with strongest growth in Finland
and Spain.
• These growth figures will also be affected by churn as consumers switch
suppliers, particularly where service packages and competitive pricing make


5
Source: Point Topic

(16)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

change an attractive option. Figure 2.10 shows the take-up of cable modem and
DSL, relative to the number of telephone lines in use in each country

Figure 2-10 Cable and DSL broadband take up per telephone line in use – Q4 2003
0%
5%
10%
15%
20%
25%
30%
BE NL DK AU FI SE ES EE PT FR MT UK DE IT SI LU HU LT LV IE CY PL CZ SK GR

Source: Ovum

There is still plenty of scope for increasing broadband take up in Europe, and this is
evident from Figure 2.11, which shows that the highest take-up in the EU is in
Belgium, with 12 broadband lines per 100 population. This compares with 24
broadband lines per 100 population in South Korea, one of the world’s highest users
of broadband services. If the number of residential broadband lines per household is
considered, then we see that in Belgium 27% (Q1 2004), of households have a
broadband connection which compares with a figure of 62% (Q1 2004), in South
Korea where the number of broadband lines grew by 2.2% in the quarter from Q4
2003 to Q1 2004, which is a slowing down of the growth rate experienced in earlier
recent periods.
Figure 2-11 Broadband take up (cable and DSL total) per head of population
0%
2%
4%
6%
8%
10%
12%

14%
BE DK NL SE FI AU FR DE UK ES MT PT EE IT LU SI HU CY IE LT LV PL CZ SK GR

Source: Ovum

(17)



EC/ESA Digital Divide CBA – Final Report 6 October 2004


Can South Korea be used as a good model for estimating the growth of broadband in
Europe? The main reasons for the great penetration of broadband in Korea are:
a) the Korean government has an initiative for carriers to offer universal access
to broadband Internet with speeds up to 1 Mbit/sec by 2005 to 84% of South
Korean households. The government has subsidised equipment purchases and
consistently pursued deregulation and liberal market principles
b) there have been sustained high levels of investment in technical education to
help stimulate infrastructure build and an environment suitable for rapid
adoption of the new communication services
c) Korea's closely-knit and homogeneous culture has led to rapid take up of
broadband Internet services. More than 93% of homes are within 4 km from
end offices allowing high speed DSL access. Around 94% of Internet
customers in Korea use broadband services.
d) the online gaming industry in Korea is well developed and very popular.
e) by 2005, the government plans to build high-speed broadband networks using
fibre optic cables offering speeds up to 20 Mbit/s. It also plans to introduce
FTTH in 2005. The telecommunications companies are being encouraged to
invest a total of 18.4 trillion KRW (EUR13 billion) to build a countrywide

ultra-speed information network by providing them with loans and incentives.
f) in April 2002, the Korean government announced a series of loans totalling 80
billion KRW (EUR 56 million) to Korean ISPs to increase the number of
homes passed by broadband from 55% of Korean homes to 70% by the end of
2002. The Korean government plans to spend over EUR 8.3 billion to deliver
VDSL or fibre to over 80% of the Korean population by the end of 2005.
Whilst some of these factors do not have direct parallels in Europe, they indicate the
level of broadband adoption that can be achieved when attention and investment is
stimulated. The Korean initiatives are wide ranging and they are far more than being a
subsidy which reduces the cost of broadband for end-users. Further analysis would be
needed to determine the relative importance of each of the Korean initiatives and the
extent to which demographic and cultural differences between Korea and Europe (e.g.
multi-tenant buildings, the strong emphasis on education and the enthusiasm for
online gaming) would effect the take up of broadband in Europe if similar initiatives
were undertaken.

(18)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

2.5.3 Take up by country group: 2004 – 2013
Two sets of data have been developed for modelling the costs of broadband
deployment over the period 2004 - 2013:
• a Base Case. This is the growth in broadband take up that can reasonably be
expected to develop over the period 2004 – 2013. The forecasts are different for
each country group and they assume that:
− the regulatory environment will continue to develop and support the roll-
out of broadband

− service prices will continue to drop – driven by competition and
manufacturing efficiencies achieved through high volumes of product
shipments
− governments and agencies will encourage the adoption of broadband for
the benefits that it brings
− applications, content and service bundles stimulate the market
• an Extended Case. In each country group a modest increase is assumed over and
above the base case. This case assumes that the terrestrial availability remains
similar to the base case. The increased take up might be due a more stimulating
market environment or intervention to attract new users.
2.5.4 Base case for take up
The Base Case for broadband take up is derived from Ovum’s broadband forecast
methodology. This model has been developed through many years of experience and
is refined at each iteration. The Ovum broadband modelling methodology is
developed from historic demand information, with country scenarios for
telecommunications infrastructure and services derived from published statistics, desk
research and expert knowledge. This is complemented and reviewed through
correspondence and interviews with country sources such as regulators, operators and
equipment vendors. The model incorporates the main types of infrastructure deployed
and service segments, breaking each category down into consumer and business, voice
and data, and narrowband and broadband segments, as relevant.
The Base Case model developed for this study provides forecasts for take up by user
group (Consumer/SoHo, SME and Large Enterprise), access speed and urban / rural
split. The forecasts for residential / SoHo users in each country group are provided in
Figure 2.12 to Figure 2.15 and summarised with a comparison with the Extended Case
in Figure 2.16. Detailed charts of the broadband Base Case are provided in Appendix
A. The charts show:

(19)

EC/ESA Digital Divide CBA – Final Report 6 October 2004

• higher take up in urban areas than in rural areas. This reflects the greater service
availability in urban areas and also the higher levels of competition causing
service providers to offer attractive service bundles
• an expectation that businesses will be much more attracted to higher bandwidth
services for desk-based applications, but will still find lower bandwidth services
sufficient for many specific applications, e.g. point of sale terminal connections
• the profile for Country Groups 3 and 4 show a significant lag behind Country
Group 1 and Country Group 2 countries.
Where take up grows to over 100% in the SME and large enterprise sector, this
reflects the fact that one enterprise may have more than one broadband connection.
Figure 2-12 Broadband take up in residential/ SoHo sector in Country Group 1
Residential and SoHo - Urban Take-up
-
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
(,000 )
Urban <512kbps Urban 512kbps-10Mbps Urban >10Mbps
Residential and SoHo - Rural Take-up

-
2,000
4,000
6,000
8,000
10,000
12,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
(,000 )
Rural <512kbps Rural 512kbps-10Mbps Rural >10Mbps

Source: Ovum

Figure 2-13 Broadband take up in residential/ SoHo sector in Country Group 2
Residential and SoHo - Urban Take-up
-
5,000
10,000
15,000
20,000
25,000
30,000
35,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
(,000 )
Urban <512kbps Urban 512kbps-10Mbps Urban >10Mbps
Residential and SoHo - Rural Take-up
-
1,000
2,000

3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
(,000 )
Rural <512kbps Rural 512kbps-10Mbps Rural >10Mbps

Source: Ovum


(20)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

Figure 2-14 Broadband take up in residential/ SoHo sector in Country Group 3
Residential and SoHo - Urban Take-up
-
1,000
2,000
3,000
4,000
5,000
6,000

7,000
8,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
(,000 )
Urban <512kbps Urban 512kbps-10Mbps Urban >10Mbps
Residential and SoHo - Rural Take-up
-
500
1,000
1,500
2,000
2,500
3,000
3,500
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
(,000 )
Rural <512kbps Rural 512kbps-10Mbps Rural >10Mbps

Source: Ovum

Figure 2-15 take up in residential/ SoHo sector in Country Group 4
Residential and SoHo - Urban Take-up
-
1,000
2,000
3,000
4,000
5,000
6,000
7,000

8,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
(,000 )
Urban <512kbps Urban 512kbps-10Mbps Urban >10Mbps
Residential and SoHo - Rural Take-up
-
200
400
600
800
1,000
1,200
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
(,000 )
Rural <512kbps Rural 512kbps-10Mbps Rural >10Mbps

Source: Ovum

2.5.5 Extended Case for take up
The model develops costs for each country for high levels of broadband take-up, at
different access speeds. The levels of high take-up are based on assumptions that the
market will be stimulated in some way. This might be caused by:
• the availability of market-stimulating content or applications which have not
been considered in the base level scenarios
• regulatory intervention, which for example:
− drives down the cost of unbundled and wholesale DSL services
− establishes broadband universal service obligations

(21)

EC/ESA Digital Divide CBA – Final Report 6 October 2004

• service provider market and pricing strategies which cause stronger take up than
in the base case
• market intervention (beyond the initiatives that are already envisaged in the base
case) to stimulate demand, for example:
− community schemes to bring user groups together
− company schemes to bring employees onto broadband, enabling flexible
working
− subsidies from municipalities or regional development agencies
− utilities or supermarkets being more active in broadband service resale as
part of loyalty schemes.

Figure 2.16 shows the levels of take up in 2008 and 2013 for the residential user
segment and allows comparison with the base case. Detailed charts of the Extended
Case take up profiles, broken down by user segment and access speed, are provided in
Appendix A.
Figure 2-16 Residential / SoHo segment take up
Base Case Extended Case Base Case Extended Case
Country Group 1
Urban 54% 56% 76% 82%
Rural 36% 39% 54% 64%
Country Group 2
Urban
37% 41% 74% 81%
Rural
19% 20% 39% 42%
Country Group 3

Urban
45% 50% 75% 81%
Rural
12% 13% 28% 30%
Country Group 4
Urban
13% 14% 43% 47%
Rural
1% 2% 8% 10%
2008 2013
Take Up

Source: Ovum

2.6 Potential broadband demand remaining unserved
Despite considerable efforts on the part of service providers to deploy services, this
study estimates that by 2013, terrestrial broadband services will remain unavailable to
some 4.7m would-be broadband users in the EU. About 1.3m of these will be business
users, predominantly micro-enterprises.

(22)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

The situation varies considerably between different country group. Rural areas will be
less well served during the period to 2013 and the demand remaining unserved in
these areas is of most interest. As shown in Figure 2.15, in 2004 there is estimated to
be an unserved market of 6.2m households and businesses in Group 1 countries

without access to terrestrial broadband services. If the number of would-be users in
this group is assumed to be in line with take-up where services are available, then the
estimated unmet demand among Group 1 countries is 380,000 would-be users in 2004,
rising to 490,000 in 2013.
Figure 2.18, 2.19 and 2.20 show the Base Case for broadband availability, take-up,
unserved users and unmet demand for Country Groups 2, 3 and 4.
As can be seen from the charts, the levels of unmet demand is much lower in Country
Group 1 than in the other country groups. This reflects the very high availability of
broadband services achieved in Country Group 1, whose member countries are all
wealthy and have highly developed telecommunications infrastructures. In contrast,
Country Groups 2 and 3 have large rural areas than incorporate Objective 1 regions,
often with low population densities in those areas. The level of unmet demand peaks
in the middle of the study period, reducing progressively thereafter as terrestrial
rollouts extend into at least the outer peripheries of these areas.
We estimate that by 2013, the level of unmet demand in Country Group 2 will be 2.2
million and over 0.5 million in Country Group 3. Country Group 4, which includes
the larger accession countries and also Greece has a high level of unmet demand. In
total there are estimated to be 1.5 million would-be users in the unserved areas of
these countries, reflecting the large proportions of their total land mass that remains
outside of terrestrial broadband coverage even in 2013.
The Extended Take-up Case, which represents a greater level of broadband take-up,
might be expected to show a lower level of unmet demand. Yet in fact the opposite is
true. This is because availability of broadband services is unchanged between the two
cases, with only the level of take-up varying. The Extended Take-up Case represents a
situation in which more customers take-up broadband in those areas where it is
available. It is assumed that the potential demand for broadband will be the same
whether it is available or not. Consequently, the Extended Take-up Case produces
both a higher level of take-up in areas with broadband availability and a higher level
of unmet demand where it is not available. The analysis indicates a potential of about
7m users, including 2m business users, with a requirement for broadband services that

cannot be met by the anticipated market-driven rollout of terrestrial networks.


(23)



EC/ESA Digital Divide CBA – Final Report 6 October 2004

Figure 2-17 Unserved demand – Country Group 1, rural areas
Group 1 Countries - rural availability and take-up
-
5,000,000
10,000,000
15,000,000
20,000,000
25,000,000
30,000,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Total Rural Consumers (inc bus.) Rural Availability Rural Take-up
Group 1 Countries - rural unserved market and
rural unmet demand
-
5,000,000
10,000,000
15,000,000
20,000,000
25,000,000
30,000,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

Total Rural Consumers (inc bus.) Unserved market Unmet demand

Source: Ovum

Figure 2-18 Unserved demand – Country Group 2, rural regions
Group 2 Countries - rural availability and take-up
-
5,000,000
10,000,000
15,000,000
20,000,000
25,000,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Total Rural Consumers (inc bus.) Rural Availability Rural Take-up
Group 2 Countries - rural unserved market and
rural unmet demand
-
5,000,000
10,000,000
15,000,000
20,000,000
25,000,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Total Rural Consumers (inc bus.) Unserved market Unmet demand

Source: Ovum

Figure 2-19 Unserved demand – Country Group 3, rural regions
Group 3 Countries - rural availability and take-up
-

2,000,000
4,000,000
6,000,000
8,000,000
10,000,000
12,000,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Total Rural Consumers (inc bus.) Rural Availability Rural Take-up
Group 3 Countries - rural unserved market and
rural unmet demand
-
2,000,000
4,000,000
6,000,000
8,000,000
10,000,000
12,000,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Total Rural Consumers (inc bus.) Unserved market Unmet demand

Source: Ovum



(24)