Synopsis: Ict:

INFO-Document is encrypted Fatal Error: 1: 1: Content is allowed not in prolog Exception in thread"main"org. apache. tika. exception.

Unable to extract PDF content at org. apache. tika. parser. pdf. PDF2XHTML. process (PDF2XHTML. java:

146 at org. apache. tika. parser. pdf. PDFPARSER. parse (PDFPARSER. java: 159 at org. apache. tika. parser.

Compositeparser. parse (Compositeparser. java: 244 at org. apache. tika. parser. Compositeparser. parse (Compositeparser. java: 244

at org. apache. tika. parser. Autodetectparser. parse (Autodetectparser. java: 121 at org. apache. tika. cli.

Tikacli$outputtype. process (Tikacli. java: 143 at org. apache. tika. cli. Tikacli. process (Tikacli. java: 422

at org. apache. tika. cli. Tikacli. main (Tikacli. java: 113 Caused by: java. io. IOEXCEPTION:

javax. crypto. Illegalblocksizeexception: Input length must be multiple of 16 when decrypting with padded cipher

at javax. crypto. Cipherinputstream. getmoredata (Cipherinputstream. java: 115 at javax. crypto. Cipherinputstream. read (Cipherinputstream. java:

233 at javax. crypto. Cipherinputstream. read (Cipherinputstream. java: 209 at org. apache. pdfbox. pdmodel. encryption.

Securityhandler. encryptdata (Securityhandler. java: 312 at org. apache. pdfbox. pdmodel. encryption. Securityhandler. decryptstream (Securityhandler. java:

413 at org. apache. pdfbox. pdmodel. encryption. Securityhandler. decrypt (Securityhandler. java: 386 at org. apache. pdfbox. pdmodel. encryption.

Securityhandler. decryptobject (Securityhandler. java: 361 at org. apache. pdfbox. pdmodel. encryption. Securityhandler. proceeddecryption (Securityhandler. java:

192 at org. apache. pdfbox. pdmodel. encryption. Standardsecurityhandler. decryptdocument (Standardsecurityhandler. java: 158 at org. apache. pdfbox. pdmodel.

PDDOCUMENT. openprotection (PDDOCUMENT. java: 1597 at org. apache. pdfbox. pdmodel. PDDOCUMENT. decrypt (PDDOCUMENT. java: 943

at org. apache. pdfbox. util. PDFTEXTSTRIPPER. writetext (PDFTEXTSTRIPPER. java: 337 at org. apache. tika. parser. pdf. PDF2XHTML. process (PDF2XHTML. java:

130 7 more Caused by: javax. crypto. Illegalblocksizeexception: Input length must be multiple of 16

when decrypting with padded cipher at com. sun. crypto. provider. Ciphercore. dofinal (Ciphercore. java: 750

at com. sun. crypto. provider. Ciphercore. dofinal (Ciphercore. java: 676 at com. sun. crypto. provider.

AESCIPHER. enginedofinal (AESCIPHER. java: 313 at javax. crypto. Cipher. dofinal (Cipher. java: 1970 at javax. crypto.

Cipherinputstream. getmoredata (Cipherinputstream. java: 112 19 more

Micro and Small Business in the EU What †s in it for you Table of contents

Publisher European Small Business Alliance Content ESBA Team Design Froben Ltd.,UK www. froben. co. uk

Copyright European Small Business Alliance All rights reserved Contact secretariat@esba-europe. org www. esba-europe. org

Published June 2011 2 3 1. Introduction...5 2. Small Business and the EU...7

3. Benefits from the EU...13 4. EU Funding...23 5. EU-Who do I call?..

37 6. Conclusion...45 4 5 The Internet has opened a new world to us. Any kind of information is

out there and this medium is more and more replacing printed material After years of involvement in Brussels, contributing to better legisla

-tion for small businesses, it often occurred to me that there are many positive attempts to help micro and small companies,

but very little is known to us business people This publication was created to address just these issues

a) to explain what is being done to help micro and small business by the EU Institutions

b) to list the most important websites that are relevant to micro and small businesses

website at http://ec. europa. eu/enterprise/policies/sme/files/sme definition/sme user guide en. pdf 2. 2 Why is this relevant

It has not always been evident that the European union recognised SMES as being a category of businesses, different from large enterprises.

the digital economy •Innovation Union: It consists of over 30 action points aimed at boosting research

Regarding SMES, the European commission monitors progress in the fields of research and development and innovation, resource-efficiency and employment in

as well as the links to the relevant websites and points of contact The European union has three key funding instruments to support SMES

www. access2finance. eu/en/Attachments/List of deals 23 12 2010. pdf 26 27 B). SME Guarantee Facility (SMEG The objective of the SME Guarantee Facility is to reduce the difficulties SMES face in accessing finance

initiatives of the Europe 2020 Strategy, has the goal of creating a flourishing digital economy by 2020

http://ec. europa. eu/energy/intelligent/index en. html 4. 2 7th Framework Programme (FP7 The Seventh Framework Programme (FP7) is one of the pillars of the European Research Area (ERA

which are published once a year by the European commission on FP7 website Certain topics are dedicated to SMES

and to apply for grants you can visit the website of the European Research Council

http://cordis. europa. eu/fp7/people/industry-academia en. html 2. Marie Curie Initial Training Networks (ITN) offers the opportunity to create networks for young re

http://cordis. europa. eu/fp7/people/initial-training en. html 4. 2. 4 Capacities The FP7 capacities programme intends to improve research capacities through Europe.

ftp://ftp. cordis. europa. eu/pub/fp7/docs/research smes en. pdf and Research for SME associations at a glance

ftp://ftp. cordis. europa. eu/pub/fp7/docs/research smes assoc en. pdf 2. Developing and coordinating support to SMES at national level.

website http://cordis. europa. eu/fp7/dc/index. cfm In order to provide help and advices to the participants of FP7,

states on the website of the FP7 programme. The network is based on existing national and regional

You can find a search engine of NCPS by visiting FP7 support webpage http://cordis. europa. eu/fp7/get-support en. html

Lastly, an important tool is the Enquiry Service provided by the Europe Direct Contact Centre which aims

and access to transport and telecommunications services of general economic interest http://ec. europa. eu/regional policy/thefunds/regional/index en. cfm

http://ec. europa. eu/regional policy/images/map/cooperat2007/crossborder/crossborder27 eu 07. pdf The following web link will give you information on transnational cooperation and eligible regions

http://ec. europa. eu/regional policy/images/map/cooperat2007/transnational/transnat mosaic. pdf You can find general information on accessing the funds here

http://ec. europa. eu/regional policy/thefunds/access/index en. cfm#4 For more information about regional policy in general and how to apply to Operational Programmes

website at the beginning of each year http://ec. europa. eu/transport/marcopolo/about/index en. htm 4. 4. 3 European Lifelong learning Programme

http://eacea. ec. europa. eu/llp/leonardo/leonardo da vinci en. php http://ec. europa. eu/education/lifelong learning-programme/doc1208 en. htm

http://www. erasmus-entrepreneurs. eu/page. php? pid=051 EU-Who do I call 36 37

telephone to give Mr van Rompuy a call. So who do you call It is a well-known problem to the EU that small businesses have difficulties getting

You can write to the EU SME Envoy at the following email address entr-sme-envoy@ec. europa. eu

This link provides an overview of telephone and fax numbers and postal addresses, see the list of Con

-tact points within the EU institutions, agencies and other bodies Europe Direct Tel: 00 800 67 89 10 11 (free of charge throughout the EU

http://ec. europa. eu/solvit/site/index en. htm SOLVIT is designed to solve problems encountered by both citizens and businesses in case an EU Mem

http://cordis. europa. eu/fp7/ncp en. html The National Contact Points are established by the 27 Member States as well as the associated states un

On this website you can find booklets issued by the Commission answering questions you may have

http://www. managenergy. net/smes. html The website provides you with a guide with thematic and sectorial access to locally relevant energy

information for SMES, such as key European legislation, key funding, examples of good practices and useful links

http://www. erasmus-entrepreneurs. eu/page. php? cid=05 Erasmus for Young Entrepreneurs is a grant providing promising European entrepreneurs with the skills

National contact points can be found through the web link European Documentation Centres http://europa. eu/europedirect/meet us/directory/index en. htm

Each national representation†s website provides the address of regional representations as well http://europa. eu/whoiswho/public/index. cfm?

The following website helps you to find the representations of the EU institutions in your country

+32 2 639 62 31 •Fax:++32 2 644 90 17 E-mail: secretariat@esba-europe. org

Cluster development involves identifying the Region†s core competence and putting formal structures in place to maximise the potential of that competence.

9 http://www. wheel. ie/sites/default/files/Consultation%20process%20on%20partnership%20agreement%202014%20

as a European TEN-T Core port. The ports of the Estuary are the third largest in Ireland by tonnage.

The available data indicates that while the FDI sector is of vital importance to the growth of exports

Software & Services 38 7 2 2 8 57 Industrial & Life sciences 12 2 1 4 7 26

Subdivision of data into North & South Tipperary areas not available 28 Case study: Benefi ts of participation in EU Projects

•The irish Software engineering Research Institute (LERO a global leader in software engineering research; and •The Institute for the Study of Knowledge in Society (ISKS

which supports research on political, social and cultural change, and on identity and social order, locally, regionally

-refining & Bioenergy, Data Analytics and Manufacturing Research. Two Research Centres are hosted by the University of Limerick:

c) Data pertaining to employment, turnover and exports in the Region from the Central Statistics

identifying the Region†s core competences and putting formal structures in place to maximise the potential of those competences

A cluster must be based around the core competencies of the region. This is a critical element of their successful development.

•Industry/businesses †are at the core of every cluster •Public bodies/government agencies †make policy decisions

A national clustering policy is essential to provide support and structure to cluster development. Additionally, there is a requirement for policies to ensure the creation of the type of

development and the regional benefits of clustering •To carry out the necessary prioritisation research and analysis in the Midwest Region to identify

the core regional competence •To secure public funding to engage a Cluster Facilitator to identify regional opportunities for

19 http://www. ifm. eng. cam. ac. uk/uploads/Resources/Briefings/v1n4 ifm briefing. pdf 5. 3 Open Innovation Culture

By crowdsourcing innovation, both internally and externally, GE is improving customer value and driving advancements across industries.

•To lobby nationally for policy guidance in relation to data protection, IP and competition policy to support an open innovation environment for all

and ensure its availability to future users. The establishment of Knowledge Transfer Networks (KTNS) is highly beneficial.

research or education institutes to a wide range of users promotes scientific and technological development and the future expansion of these concepts in this Region is linked closely to the

Users, NGOS Academia Researchers HEIS Figure 6. 1 Bodies/Agencies with a Role in Implementation

Users, NGOS Academia Researchers HEIS In addition to the working group of the RPG Implementation Steering Committee, there are a number of

Baseline Data: The initial sections of this Strategy form a baseline assessment of the current research

•Priority Area B-Data Analytics, Management, Security & Privacy •Priority Area C-Digital Platforms, Content & Applications

•High tech Manufacturing/Engineering •Logistics/Distribution •High Value Food & Drink •Life sciences

•ICT including Software •Logistics and Supply Chain Management •Food Sector & Agribusiness

Email: info@mwra. ie Tel: 067 33197 www. mwra. ie

International Telecommunication Union Telecommunication Development Bureau Place des Nations CH-1211 Geneva 20 Switzerland www. itu. int

ISBN 978-92-61-14661-0 SAP id 9 7 8 9 2 6 1 1 5 2 9 1 8

3 9 4 6 4 Price: 86 CHF Printed in Switzerland Geneva, 2014 Photo credits: Shutterstock

International Telecommunication Union Place des Nations CH-1211 Geneva Switzerland Original language of publication: English

International Telecommunication Union ISBN 978-92-61-15291-8 ii Foreword iii I am pleased to present to you the 2014 edition of the Measuring the Information

Its core feature is the ICT Development Index (IDI), which ranks countries†performance with regard to ICT infrastructure, use and skills.

end 2014, almost 3 billion people will be using the Internet, up from 2. 7 billion at end

While the growth in mobile-cellular subscriptions is slowing as the market reaches saturation levels, mobile broadband remains the fastest growing market segment, with

Despite this encouraging progress, there are important digital divides that need to be addressed 4. 3 billion people are still not online,

reinforcing the urban-rural digital divide. As this report finds, ICT performance is better in countries

and mobile broadband is six times more affordable in developed countries than in developing countries. Income inequalities within countries

Telecommunication Development Bureau (BDT International Telecommunication Union subscription still represents more than 5 per cent of household income for over half of the population

For these income groups, mobile broadband may be the affordable alternative An enabling telecommunication regulatory environment can significantly influence the affordability of

services. The report finds that the price of ICT services falls with better market regulation and increased

per cent and mobile-cellular prices by 5 per cent if competition and/or the regulatory framework

the lack of up-to-date data, in particular in developing countries. ITU is joining the international statistical community in looking into ways of using new and emerging data sources †such as those

associated with big data †to better provide timely and relevant evidence for policy-making. Calls for

a â€oedata revolution†are prominent in the international debates around the post-2015 development agenda, and ICTS have an important role to play in view of their capacity to produce,

huge amounts of data, as well as being a major source of big data in their own right. Big data from

mobile operators, for example, are real-time and low-cost and have one of the greatest development potentials in view of the widespread use and availability of mobile networks and services.

This report provides the reader with a comprehensive and critical overview of the role of big data from the

telecommunication sector, for use in social and economic development policy and for monitoring the future information society

I trust that the data and analysis contained in this report will be of great value to the ITU membership

including policy-makers, the ICT industry and others working towards building an inclusive global information society

Acknowledgements The 2014 edition of the Measuring the Information Society Report was prepared by the ICT Data and

Statistics Division within the Telecommunication Development Bureau of ITU. The team included Susan Teltscher (Head of Division), Esperanza Magpantay, Vanessa Gray, Ivan Vallejo, Lisa Kreuzenbeck

and Ola Amin. The following consultants to ITU provided substantive inputs: Pantelis Koutroumpis Chapter 4) and Sriganesh Lokanathan (Chapter 5). Andrã Wills, Fernando Callorda and Shazna Zuhyle

and Michael Minges to the compilation of data on international bandwidth, revenue and investment. Helpful inputs and suggestions were received

Project Support and Knowledge management Department, Telecommunication Development Bureau The report includes data from Eurostat, OECD, IMF, Informa, the UNESCO Institute for Statistics, the

United nations Population Division and the World bank, which is acknowledged greatly ITU also appreciates the cooperation of countries that have provided data included in this report

The report was edited by Anthony Pitt and Bruce Granger, ITU English Translation Section. The desktop publishing was carried out by Nathalie Delmas,

and the cover was designed by Jesus Vicente Administrative support was provided by Herawasih Yasandikusuma v Table of contents

1. 4 Revenue and investment in the telecommunication sector...13 1. 5 Use of ICTS...

2. 3 Monitoring the digital divide: Developed, developing and least connected countries...55 2. 4 Geography, population size, economic development and the IDI...

4. 2 Fixed-telephone and mobile-cellular prices...108 4. 3 Broadband prices...114 4. 4 Income inequality and broadband prices...

4. 5 The impact of competition and regulation on telecommunication prices...152 Chapter 5. The role of big data for ICT monitoring and for development...

173 5. 1 Introduction...173 5. 2 Big data sources, trends and analytics...175 5. 3 Telecommunication data and their potential for big data analytics...

181 5. 4 Big data from mobile telecommunications for development and for better monitoring...185 5. 5 Challenges and the way forward...

195 Chapter 5 Annex...207 List of references...213 Annex 1. CT Development Index (IDI) methodology...

221 Annex 2. ICT price data methodology...231 Annex 3. Statistical tables of indicators used to compute the IDI...

241 Annex 4. Statistical tables of prices used to compute the ICT Price Basket...251 viii

1. 1 Fixed-telephone subscriptions by level of development, 2005-2014 (left) and by region

1. 2 Mobile-cellular subscriptions by level of development, 2005-2014 (left) and by region

1. 4 Active mobile-broadband subscriptions by level of development, 2007-2014 (left) and by region

1. 6 Rural population covered by at least a 3g mobile network, 2009-2012.8 1. 7 Fibre and microwave routes, share of route kilometres (left)

1. 8 Total International Internet bandwidth (Gbit/s), by level of development (left) and regional share

1. 9 International Internet bandwidth (bit/s) per Internet user, by region, 2004 and 2013.10 1. 10 Percentage of households with Internet access, by level of development, 2005-2014 (left) and

1. 13 Telecommunication revenues, world and by level of development, 2007-2012, total in USD

1. 14 Annual investment by telecommunication operators, world and by level of development 2007-2012, total in USD (left) and annual growth (right...

1. 15 Individuals using the Internet, by level of development, 2005-2014 (left) and by region

1. 16 Growth in daily Google searches, 2007-2013.16 1. 17 Growth in Facebook monthly active users, 2004-2013 (millions of users...

17 1. 18 Wikipedia articles †total and English language, 2003-2013 (thousands of articles...17

1. 19 Fixed-broadband access in enterprises using the Internet, selected countries, 2005-2012.19 1. 20 E-government Development Index (EGDI), 2003-2014.20

ix 1. 21 E-government services provided by countries (transactional services, left, and e-participation

2. 2 Wireless-broadband subscriptions per 100 inhabitants, top five IDI countries, 2010-2013.47 2. 3 Wireless-broadband penetration, Bhutan, 2008-2013.50

2. 4 Proportion of households with a computer and proportion of households with Internet access, 2012-2013, Qatar...

54 2. 5 IDI by level of development...56 2. 6 IDI access sub-index by level of development...

3. 3 Mobile-cellular subscriptions per 100 inhabitants, 2012 and 2013, Africa...87 3. 4 IDI values compared with the global, regional and developing/developed-country averages

3. 5 Wireless-broadband subscriptions per 100 inhabitants, Arab States, 2012 and 2013.91 3. 6 IDI values compared with the global, regional and developing/developed-country averages

3. 7 Wireless-broadband penetration, Asia and the Pacific, 2012 and 2013.94 3. 8 IDI values compared with the global, regional and developing/developed-country averages

3. 11 Percentage of Individuals using the Internet, Europe compared to global and developed country average, 2013.100

4. 1 Fixed-telephone basket (left) and mobile-cellular basket (right), in PPP$, world and by level of

4. 2 Fixed-telephone basket (left) and mobile-cellular basket (right), as a percentage of GNI p. c

4. 11 Availability of mobile-broadband services by type of service, by level of development

4. 12 Mobile-broadband prices, in PPP$, world and by level of development, 2013.217 4. 13 Mobile-broadband prices, in USD, world and by level of development, 2013.128

4. 14 Mobile-broadband prices as a percentage of GNI p. c.,world and by level of

development, 2013.128 xi 4. 15 Mobile-broadband prices as a percentage of GNI p. c.,by region, 2013.129

4. 16 Comparison of postpaid fixed-broadband and postpaid computer-based mobile-broadband prices, in USD, by region, 2013.130

4. 17 Distribution of household disposable income (left) and household consumption (right selected countries, 2011 or latest available year...

4. 24 Variation in mobile-cellular prices(%)explained by each variable, 2013.165 List of figures

4. 1 Mobile-broadband services by type of device/plan...125 4. 2 Relationship between regulation, competition and prices...

5. 1 The five Vs of big data...176 5. 2 An overview of telecom network data...

182 5. 3 Customer profiling using telecom big data...184 xii List of boxes 1. 1 Final review of the WSIS targets:

Achievements, challenges and the way forward...26 1. 2 A decade of successful international cooperation on ICT measurement...

1. 4 What is a data revolution?..30 2. 1 ITU discussion forums on ICT statistics...

4. 2 Panel regression models for fixed-broadband and mobile-cellular prices...158 5. 1 How big data saves energy †Vestas Wind Systems improves turbine performance...

177 5. 2 How Twitter helps understand key post-2015 development concerns...179 5. 3 How mobile operators currently use data to track service uptake, business performance and

revenues...183 5. 4 Using mobile data for development...187 5. 5 How mobile network data can track population displacements †an example from the

2010 Haiti earthquake...188 5. 6 Leveraging mobile network data for transportation and urban planning in Sri lanka...

189 5. 7 Poverty mapping in CÃ'te d†Ivoire using mobile network data...190 5. 8 Using mobile-phone data to track the creditworthiness of the unbanked...

191 5. 9 Using mobile big data and mobile networks for implementing surveys...193 List of tables

1. 1 Rural population covered by a mobile-cellular signal, 2012.4 1. 2 Total Internet domain registrations by world region, 2003,2008 and 2013.18

2. 1 IDI values and changes, 2012 and 2013.41 2. 2 ICT Development Index (IDI), 2012 and 2013.42

xiii xiv 2. 3 IDI access sub-index, 2012 and 2013.43 2. 4 IDI use sub-index, 2012 and 2013.44

2. 5 IDI skills sub-index, 2012 and 2013.45 2. 6 Most dynamic countries-changes between IDI 2013 and 2012.49

2. 7 IDI by level of development, 2012-2013.55 2. 8 IDI by groups, 2012 and 2013.57

2. 9 Partial correlation analysis of IDI, population and geographic characteristics...60 2. 10 Examples of contribution of ICTS towards the Millennium Development Goals...

66 2. 11 Results of partial correlation analysis between IDI and MDG indicators...67 2. 12 Simple correlation analysis between relative change in IDI values and MDG indicators

4. 1 Fixed-telephone sub-basket, 2013.112 4. 2 Mobile-cellular sub-basket, 2013.113 4. 3 Fixed-broadband prices as a percentage of GNI p. c.,by region, 2013.116

4. 4 Fixed-broadband sub-basket, 2013.124 4. 5 Mobile-broadband prices, postpaid handset-based 500 MB, 2013.132

4. 6 Mobile-broadband prices, prepaid handset-based 500 MB, 2013.134 4. 7 Mobile-broadband prices, postpaid computer-based 1 GB, 2013.136

4. 8 Mobile-broadband prices, prepaid computer-based 1 GB, 2013.138 4. 9 Fixed-broadband prices as a percentage of household disposable income, selected

countries, 2013.144 4. 10 Fixed-broadband prices as a percentage of household consumption expenditure selected countries, 2013.145

xv 4. 11 Prepaid handset-based mobile-broadband (500 MB/month) prices as a percentage of

4. 13 Panel regression results, fixed-broadband prices and regulation...160 4. 14 Panel regression results, mobile-cellular prices and regulation...

163 4. 15 ICT Price Basket and sub-baskets, 2013.166 5. 1 Sources of big data...

175 1 Measuring the Information Society Report 2014 Chapter 1. Recent information society developments 1. 1 Introduction

global mobile-cellular market is approaching saturation levels, mobile-broadband uptake continues to grow at double-digit rates in all

The data also show a continuous increase in Internet usage, with growth in the number of

Internet users in all countries and increasing availability of online content, much of which is user-created through social media applications

and platforms (e g. Twitter, Youtube, Whatsapp With more and more applications now available through mobile platforms (mobile apps), and

the strong growth in mobile Internet uptake, an increasing number of people are joining, and

participating actively in, the information society While the information society is growing worldwide, digital divides remain â€

and are even widening †in some segments. In particular, there is a significant and persistent urban-rural digital

divide, whereby urban citizens enjoy ubiquitous mobile network coverage, affordable high-speed Internet services and the higher levels of skills

required to make effective use of online content and services, while the opposite is often the case

in rural and remote areas of many developing countries This chapter will present and discuss key indicators

and data) market segments, and considering both subscriptions and household access data. This will be followed by a presentation of the latest

trends in terms of investment and revenue in the telecom sector. Then, a number of key indicators

will be presented concerning ICT uptake by individuals, businesses and public organizations from the government and education sectors), as

social media. The final part of the chapter will discuss emerging issues related to information -society measurements, in particular in the context

WSIS+10 review, the demand for a data revolution and the role of big data for ICT monitoring

1. 2 The voice market In line with developments in recent years, fixed telephony is on the decline in all regions of the

around a dozen countries where fixed-telephone uptake has increased actually over the past year. 1

Fixed-telephone penetration decreased by about 2 per cent globally in the past year, and will drop

The decline in fixed-telephone subscriptions over the past decade was accompanied by strong growth in the mobile-cellular market

until 2010, at which point mobile-cellular growth rates dropped to single digits, and they have continued to slow down since then

In 2014, global growth in mobile penetration will be at a ten-year low of 2. 6 per cent, as the

there will be almost as many mobile-cellular subscriptions (6. 9 billion) as people On earth more than three quarters of them (5. 4 billion

does not mean that everyone has a mobile phone †since many people have more than one

subscription or SIM CARD †the total numbers and growth rates strongly point to market saturation Whether this will change in the near future, and

strongest mobile-cellular growth, and the lowest penetration rates, which will reach 69 per cent

Fixed-telephone subscriptions by level of development, 2005-2014 (left) and by region, 2014 *right

ITU World Telecommunication/ICT Indicators database 39.2 26.3 24.9 15.8 12.7 8. 7 1. 3

the fixed-telephone market is shrinking and the mobile-cellular market is tapering off. In addition mobile-cellular population coverage has reached

data reveal, however, that digital divides still exist and that some people are excluded still from access to communication networks

First, there are populations living in rural areas that are covered not by a mobile-cellular signal Table 1. 1). Even though rural population

coverage is very high, at 87 per cent globally, at end 2012 around 450 million people worldwide

or is using a mobile Chart 1. 2: Mobile-cellular subscriptions by level of development, 2005-2014 (left) and by region

ITU World Telecommunication/ICT Indicators database 162.7 124.7 109.9 108.5 96.4 89.2 69.3 0 20

For countries where data are available, the number of mobile subscriptions far exceeds the number of mobile phone users (Partnership, 2014

No regular pattern can be established, though and the difference between mobile-phone user penetration and mobile-cellular subscription

penetration ranges between 8 per cent (France and 111 per cent (Panama) across countries According to GSMA estimates, unique mobile

subscribers account for about half of mobile -cellular subscriptions, which would translate into a penetration rate of around 48 per cent globally

63 per cent in developed countries, 45 per cent in developing countries and 30 per cent in least

Third, household access to a telephone is still not the norm in many developing countries

cent of households had a telephone (up from 9 per cent ten years earlier. In addition, there were

having access to a telephone compared with 54 per cent of rural households. 3 Household

telephone penetration in Malawi stood at 36 per cent in 2011 †73 per cent in urban households

gap in household telephone access prevails in many developing countries for which data are available, but is closing with the availability of

affordable mobile-phone services in rural areas Further research and data would be necessary to determine people†s access to,

and use of, voice communications and to identify other potential barriers, such as those related to poverty, literacy

broadband Internet continues to be a priority for telecommunication service providers and governments in most countries.

This is reflected in the continuous growth in the number of mobile -and fixed-broadband subscriptions worldwide

Rural population covered by a mobile-cellular signal, 2012 Source: Partnership (2014) based on ITU data

Overall mobile-cellular population coverage %Rural population covered %Rural population covered (millions Rural population not

covered (millions Africa 88 79 498 129 Americas 99 96 171 9 Asia 92 87 2 017 309

displays the lowest growth in fixed broadband estimated at 2. 5 per cent and reaching a

ITU World Telecommunication/ICT Indicators database 27.7 16.7 14.3 9. 8 7. 7 3. 1

Chart 1. 4). Mobile broadband is growing fastest in developing countries, where growth rates over the last year are expected to be twice as

devices (smartphones) and types of plan on offer in the market Nevertheless, the divide between developed and

countries, 3g subscriptions overtook 2g subscriptions in 2010 and 3g growth is flattening In developing countries, the large majority of

subscriptions are still 2g, but 3g is growing rapidly and will overtake 2g subscriptions in a

few years. 4g4 services came onto the market only recently and 4g subscriptions still account

for only a small market share in both developed and developing countries The data on fixed-and mobile-broadband

uptake confirm what has been observed on the ground. In developed countries, fixed -broadband infrastructure and services were

countries, and before fast mobile-broadband services and smartphones entered the market This has contributed to the higher uptake of

fixed broadband in developed countries. In Chart 1. 4: Active mobile-broadband subscriptions by level of development, 2007-2014 (left) and by

region, 2014*(right Note:**Estimate Source: ITU World Telecommunication/ICT Indicators database 83.7 32.0 21.1

6. 3 0 10 20 30 40 50 60 70 80 90 Developed World Developing

Furthermore, the mobile market has benefited from a more liberal regulatory approach than the fixed market

smartphone, tablet) and SIM CARDS Looking towards the future, the growth potential for mobile broadband looks promising, as

data based on ITU and Telecom Advisory Services calculations more and more countries upgrade their mobile networks. As mentioned earlier, 2g population

coverage stands at over 90 per cent worldwide Data on 3g population coverage are less

available. According to ITU estimates, global 3g population coverage stood at around 50 per cent by end 2012,

and there were still sizeable rural -urban gaps. Rural population coverage ranged from 100 per cent in the Gulf countries of United

more and more countries are deploying 3g +technologies and services, and given the strong growth in mobile-broadband subscriptions.

At the same time, the issue of spectrum allocation will have to be addressed to ensure that the

backbone capacities and international Internet bandwidth. Indeed, without further deployment of backbone infrastructure, service providers are

New data collected by ITU on the deployment of fibre transmission capacity in countries shows

at the data also reveals major disparities across regions: Asia and the Pacific (in particular China

3g 4g %1g 2g 3g 4g %0 10 20 30 40 50 60 70 80

90 100 Developing countries Chapter 1. Recent information society developments 8 Chart 1. 6: Rural population covered by at least a 3g mobile network, 2009-2012

Source: Partnership (2014) based on ITU data Percentage of rural population covered by at least a 3g mobile

network 2012 or 2011 Percentage of rural population covered by at least a 3g mobile network 2010

Percentage of rural population covered by at least a 3g mobile network 2009 0 0 0

0 0 0 0 0 0 0 0 0 1 5 11 31 32 32

34 36 38 41 42 42 46 50 50 50 55 55 58 60 61

63 65 68 69 69 77 77 77 78 81 84 86 87 88 88

89 90 90 92 93 93 94 94 95 95 96 100 100 0 10 20 30 40 50 60 70 80 90 100

is the amount of international Internet bandwidth available in countries and regions such bandwidth being a key requirement for

delivering data-intensive applications and services through high-speed networks. While fibre transmission networks constitute an

Over the past decade, international Internet bandwidth has climbed sharply, from around 1 600 Gbit/s in 2001 to 60 400 Gbit/s in 2010 and

ITU Trends in Telecommunication Reform, 2014 of the world. Growth in international bandwidth has been strong in all regions, and the share

leads by far in terms of international Internet bandwidth, accounting for more than 50 per cent of the world†s total (2013), compared with

Europe†s leadership in international Internet bandwidth is explained by the advanced level of broadband adoption and usage

Internet. As a result, the Internet backbone network in the region is interlinked by means of several Internet exchange points (IXPS

that interconnect national networks and give them access to the global Internet. Indeed some of the world†s largest IXPS are located

in Europe and have an international reach such as for instance the German Commercial Internet Exchange (DE-CIX), the Amsterdam

Internet Exchange or the London Internet Exchange. 6 The United kingdom stands out as a prominent global hub for international

3%2 %85 %4 %6 %Africa Arab States Asia & Pacific CIS The americas 0. 4

0. 7 2. 6 1. 6 1. 3 0 0. 5 1. 0 1. 5

Total international Internet bandwidth (Gbit/s), by level of development (left) and regional share right), 2004-2013

ITU World Telecommunication/ICT Indicators database connectivity, because of the strong internal demand and also its location:

Internet bandwidth in the UK accounts for almost twice as much as Africa, Arab States and CIS combined,

available international bandwidth on Internet 0 20'000 40'000 60'000 80'000 100'000

per Internet user. This indicator has increased significantly between 2004 and 2013. There are huge differences, however, between developed

as much international bandwidth per user available in the former compared to the latter 106 000 vs 23 000 bit/s per user.

Looking at regional differences, Europe stands out by far with around 160 000 bit/s per user in 2013

compared to the global average of 52 000 bit/s per user, followed by The americas with 54 000

bit/s per user Households with Internet access Household access to the Internet is the ultimate

way of guaranteeing an inclusive information society in which all people, irrespective of age gender, employment status, etc. or possible

level of disability, can access the Internet within the privacy and proximity of their own home. A

policy aimed at universal access to broadband Internet will eventually ensure access for all households nationwide.

Household access is also mostly shared access, whereby all family members can use the same service and share the

International Internet bandwidth (bit/s) per Internet user, by region, 2004 and 2013 Source: ITU World Telecommunication/ICT Indicators database

221 420 1'213 702 4'384 11'572 8'074 19'037 21'472

43'072 53'992 -10'000 20'000 30'000 40'000 50'000 60'000

The latest ITU data show that by end 2014, almost 44 per cent of the world†s households will have

in Africa have Internet, and growth remains at a high 18.4 per cent, which is more than twice the

highest number of households with Internet Chart 1. 10: Percentage of households with Internet access, by level of development, 2005-2014 (left) and

ITU World Telecommunication/ICT Indicators database 78.0 57.4 53.0 43.6 36.0 35.9 11.1 0 10

Internet As is the case with other indicators, there is a significant urban-rural divide when it comes to

data are available, rural household access falls far below urban household access, with differences ranging from 4 per cent (meaning

that household Internet penetration in urban areas is 4 per cent higher than in rural areas

Internet than rural households (Partnership 2014). 8 Available data also show that Internet access in rural households is growing slowly

78.4 43.6 31.2 5. 0 0 10 20 30 40 50 60 70 80 %Developed

but data are not readily available for those countries. As has been illustrated earlier, network deployment is still

households from purchasing Internet services At the same time, the benefits brought by ICTS and the Internet are especially impactful in

rural areas, which often also lack access to other infrastructure and public services. Therefore connecting rural households to broadband

Internet plays a greater role in those areas. Data on public access is collected by ITU through its

household questionnaire, but only few countries report data on this indicator, let alone broken down by urban and rural populations.

Public access can be provided by commercial facilities such as privately operated Internet cafes, as well as community-type facilities, which typically

provide Internet access free of charge. Schools also constitute an important location for Internet access, especially in rural areas, although access

is limited often to students and teachers and restricted to certain hours (see section 1. 5

role in terms of providing access to the Internet they are open to the public, their branches are

the Internet, even though 31 per cent of post offices have a broadband Internet connection Chartâ 1. 11), with major differences across regions

Chart 1. 12. These numbers refer to 2012, and have increased most probably somewhat today Nevertheless, there is huge potential if all post

offices were provided with broadband Internet and offered this as a service to the public According to the Universal Postal Union (UPU

Internet access and post offices with broadband Internet access, 2012, by level of development Note: Simple averages

small towns had access to the Internet, while with 60 per cent coverage half of all rural areas

to the Internet in public libraries from 2007 to 2009.10 While the results point to improvements

in the telecommunication sector In 2012, total telecommunication revenue stagnated at around USD 1. 88 trillion,

or 2. 7 per cent of world GDP (Chart 1. 13 After the slump experienced during the

The evolution of telecommunication revenues in developed countries follows the overall pattern of their economies as a whole (in

telecommunication services. In addition to the adverse economic context, the voice market in developed countries is declining or reaching

growth in telecommunication revenues in 2012 hence mitigating the global decrease in revenues experienced in 2012.

progress of telecommunication revenues seen Chart 1. 13: Telecommunication revenues, world and by level of development, 2007-2012, total in USD

left) and annual growth (right Note: †World†includes 103 countries accounting for 96 per cent of world GDP. †Developed†includes 40 developed countries accounting for 99 per cent of total

countries†share of total telecommunication revenues increased from 26 per cent in 2007 to 32 per cent in 2012, thus approaching their

the telecommunication sector in the economic growth of the developing world. For example, in the recent revision of Nigeria†s GDP, it was found

that the telecommunication industry accounted for more than a quarter of the upgrade in GDP. 12

Despite the progress seen in several developing countries, there remains huge untapped potential in the 4 billion people not yet online in

telecommunications, which is fundamental to supporting ICT uptake and innovation. In 2012 investment grew by 4 per cent to USD 307 billion

telecommunication investment persisted in 2009 -2 per cent. The overall economic environment of restricted access to capital markets and the

telecommunication infrastructure and services has been more stable, with a smaller drop in 2008(-4 per cent) and moderate growth in the

share of global telecommunication revenues generated in developing countries 32 per cent The investment-to-revenue ratio in the

telecommunication sector stood at 17 per Chart 1. 14: Annual investment by telecommunication operators, world and by level of development

2007-2012, total in USD (left) and annual growth (right Note: †World†includes countries accounting for 91 per cent of world GDP. †Developed†includes 35 developed countries accounting for 98 per cent of total GDP

telecommunication services, USD 17 were reinvested in capital expenditure (i e. in upgrading the fixed assets needed to extend

and improve telecommunication services. The investment-to-revenue ratio was somewhat lower in developed countries (15 per cent) than

On the one hand, telecommunications is a capital-intensive industry and part of the capital investments are delivered by global

On the other hand, revenue per user in several developing countries is constrained by low income levels, which limit the margin for

Internet users ITU estimates that, by end 2014, almost 3 billion people will be using the Internet, corresponding

to a global penetration rate of 40.4 per cent (Chart 1. 15. This compares to 2. 7 billion people and 38

using the Internet, 90 per cent of whom live in the developing world. While more than three out of

Nevertheless, Internet usage is growing steadily, at 6. 6 per cent in 2014 †3. 3 per cent

countries, the number of Internet users will have doubled in five years (2009-2014), and two-thirds

of today†s Internet users live in the developing world. Growth rates are highest in LDCS (13 per

Individuals using the Internet, by level of development, 2005-2014 (left) and by region 2014*(right

ITU World Telecommunication/ICT Indicators database 78.3 40.4 32.4 8. 0 %0 10 20 30

Internet usage varies considerably across regions. In Europe, Internet usage on average is approaching saturation levels, with almost 75 per

cent penetration and low growth of 2. 3 per cent during the past year. In Africa, the region with the

lowest Internet penetration rate (19 per cent Internet usage is growing strongly at 13 per cent

and almost twice as many people will be online by the end of this year compared with only four

of the world†s Internet users live in this region The two countries combined are home to around

860â million Internet users, almost 30 per cent of the world†s total and 66 per cent of Internet users

in the Asia-Pacific region. Penetration rates in the two countries differ greatly, though, reflecting

of Internet users in China is 46 per cent, it is only 18â per cent in India

Internet usage in The americas region is relatively high: with almost 66 per cent penetration it is much higher than household Internet

access (57â per cent. This suggests that shared household access as well as access outside the home is more common in the region.

cent Internet penetration compared with 11 per cent of households with Internet access. In view of infrastructure limitations and a lack of

the Internet at locations outside the home, such as at work, school or public access facilities

Internet content and use of social media The growth in Internet users has witnessed a parallel, steep growth in the volume of Internet

content. More and more people are actively participating in the information society by creating, sharing and uploading content and

using social media and other Internet-based applications, covering a large range of topics and sectors. While measuring online content

and website use is a challenging task on account of the sheer volume of information available

an attempt to do so has been made by the Partnership on Measuring ICT for Development in its final review of achievement of the WSIS

targets, which includes an assessment of Internet content (Partnership, 2014. Some key findings featured in the report are presented below

Over the past decade, the number of websites has been growing at exponential rates and according to estimates by Netcraft, there were

185 million active sites at the beginning of 2014 Google remains the leading search engine in most countries, and accounts for around 90

per cent of the search market. 16 The number of daily Google searches reached almost 6 billion

by end 2013 (Chart 1. 16) and the total number of searches made through Google in 2013

exceeded 2 trillion Social media sites have become the most accessed websites by users in both developed

Chart 1. 16: Growth in daily Google searches 2007-2013 Source: Partnership (2014), based on http://www. statisticbrain. com/google-searches

/0 2'000 4'000 6'000 8'000 2013 2012 2011 2010 2009 2008

2007 Millions of searches 17 Measuring the Information Society Report 2014 and developing countries. Since its creation in

2004, Facebook has grown to comprise 1. 3 billion active users by end 2013, a growth of 22 per cent

over the past year (Chart 1. 17), although a single user could be operating several accounts and

therefore the numbers do not represent unique Internet users (ITU, 2011. Twitter, the leading international microblogging service, founded in

2007, has grown to comprise 646 million active registered users by end 2013 (and 115 million

active monthly users), and some 58 million tweets were posted daily in the past year. 17 The Chinese microblog service Tencent Weibo

accounts for a further 507 million subscribers out of an estimated 582 million Chinese Internet

subscribers (Partnership, 2014 More than 6 billion hours of video are being watched each month and more than 100 hours

on Youtube, the leading international video -filesharing site with services in 61 countries. As of

early 2014, Youtube boasted more than 1â billion unique visitors monthly Other top popular websites include Amazon

Wikipedia and Linkedin as well as various news and online e-market sites at the national level

see below on e-business. Wikipedia, the largest and most widely used online encyclopaedia featured more than 30 million articles by end

2013 (Chartâ 1. 18. Articles are now available in 287 languages across 30 million pages of

content. By February 2014, Wikipedia registered more than 20 billion page views per month by Internet users.

At the same time, the proportion of articles in English has decreased significantly †from 46 per cent in 2003 to 15 per cent in 2013

†while those in other languages have increased accordingly, although pages viewed are still predominantly in English and the proportion of

contributors writing in English is more than 50 per cent While these numbers illustrate the huge increase

in Internet content and usage overall, a more nuanced analysis needs to be carried out to identify digital divides.

Table 1. 2 shows that, for example, domain-name registrations are still dominated by content providers in developed

The data include both global top-level domain (gtld) and country code top-level domain (cctld) registrations

Growth in Facebook monthly active users, 2004-2013 (millions of users Source: Partnership (2014), based on http://www. theguardian. com/news

/datablog/2014/feb/04/Facebook-in-numbers-statistics, accessed 6 March 2014. Data sourced from Facebook

0 500 1'000 1'500 2013 2012 2011 2010 2009 2008 2007 2006 2005

2004 Millions of users Chart 1. 18: Wikipedia articles †total and English language, 2003-2013 (thousands of articles

Source: Wikipedia statistics at http://stats. wikimedia. org/EN/Tablesarticlestotal. htm 0 5'000 10 000

'15 000 '20 000 '25 000 '30 000 '35 000 'Total English Chapter 1. Recent information society developments

18 content generated by countries and regions and highlight the differences E-business Private-sector enterprises are early adopters of

International data on ICT access and use by enterprises are collected annually by the United Nations Conference on Trade and Development

businesses with websites was lower, accounting on average for 71 per cent and ranging from 36

are making use of social media. In 2013, around 30 per cent of European enterprises used social

mostly related to social networks (e g. Facebook followed by multimedia content sharing websites (e g. Youtube) and blogs (e g.

Twitter In the developing world, data on ICTS in enterprises are scarce and only collected by few

countries. The proportion of businesses with Internet access varies between 48 per cent in Azerbaijan and 97 per cent in Lebanon. 19 Of

these, not all have broadband access, which is essential to enable businesses to engage in, and take full advantage of, the potential of

e-business (Chart 1. 19. In addition, Internet access differs enormously according to the size and location of the enterprise †small and

Table 1. 2: Total Internet domain registrations by world region, 2003,2008 and 2013 Source: Partnership (2014.

Data supplied by Zooknic, compiled from cctld and other sources. Figures exclude fifteen cctlds which act as virtual

gtlds 2003 2008 2013 Millions%Millions%Millions %World 59.7 100.0 173.4 100.0 245.2 100.0 Developed 49.6 82.9 135.9 78.4 197.4 80.5

Developing 7. 1 11.8 34.7 20.0 45.0 18.4 Other/Unknown 3. 1 5. 2 2. 8 1. 6 2. 7 1. 1

backbone connectivity and international Internet bandwidth is still lacking in many regions of the developing world.

Not only are government entities major users of ICTS, but governments are also increasingly using the Internet to provide services to their citizens

E-government contributes to increased efficiency and greater transparency and accountability in government, reduces cost

There is little data on the use of ICTS by government organizations, and those countries that do have data are usually the more advanced

ones with high levels of connectivity in general More information is available about government services provided online, tracked by the United

The latest data show that, today, governments of all countries have established central websites and that more

than 50 per cent of countries provide links to local and/or regional government agenciesâ€

sites (UNDESA, 2014. Efforts are needed still to connect lower-tier administrations in countries When it comes to the provision of e-services

and services on government website portals increased threefold, with 70 per cent of countries providing a one-stop shop portal in 2012, as

had a government web presence, and almost all countries in Europe †and the majority of

Fixed-broadband access in enterprises using the Internet, selected countries, 2005-2012 Source: UNCTAD Information Economy Database, 2014, available at unctadstat. unctad. org

P e rc e n ta g e o f e n te rp ri

Data from United nations E-government Survey (2014 Chart 1. 21: E-government services provided by countries (transactional services, left, and e-participation

Data from United nations E-government Survey (2014 101 73 60 46 44 42 41 40

particular broadband Internet) is a basic infrastructure requirement in today†s information society. Access to high-speed Internet is

necessary to enable students to use the Internet for educational purposes, and helps enhance education administration through the electronic

exchange of forms, data and other information Internet access in schools also achieves cost efficiencies by automating manual tasks and

reducing expenses associated with the printing and distribution of textbooks. The benefits are particularly attractive for remote schools, where

Internet access provides the vehicle for online learning and access to educational content. 21 In addition to the educational benefits resulting

place where young people can use the Internet see section 1. 3 above. Therefore, they can also

The latest available data from the UNESCO Institute for Statistics (UIS) 22 show that, in

developed countries for which data are available In developing countries, school access to Internet is lower on average, although much

progress has been made in recent years. There are significant differences across countries, even within the same region and with similar income

Chart 1. 22 also shows the type of Internet access schools have, in particular the share of

Internet access (out of all schools with Internet access) is still low, suggesting that, in those

growth of mobile-broadband services, it may be expected that more and more schools will have broadband access in the near future and that the

Data on broadband in secondary schools in Bangladesh are not available. Data for Nicaragua, Philippines and Indonesia do not include upper

secondary. Data for European countries and Costa rica refer to lower secondary. Data for Guyana, Nicaragua and Indonesia refer to primary and lower

secondary. Data for Cambodia include pre-primary schools. Data for Morocco, Tunisia, Guyana, Montserrat, Dominican republic, Nicaragua, Colombia

Trinidad and tobago, Bangladesh, Philippines, Sri lanka, Azerbaijan, Bhutan, Cambodia, Kazakhstan, Malaysia, Maldives, Singapore, Belarus and the Russian Federation refer to public schools.

In Suriname, there are no private schools in upper secondary. Data for Palestine refer to West bank schools

only Source: UIS database, Partnership on Measuring ICT for Development WSIS Targets Questionnaire, 2013 P

e rc e n ta g e o f s c h o o ls

Americas 0 10 20 30 40 50 60 70 80 90 100 Internet Fixed broadband Internet

P e rc e n ta g e o f s c h o o

ls 0 10 20 30 40 50 60 70 80 90 100 Asia and Africa

tv ia N or w ay Sp ai n Au st ra lia 23 Measuring the Information Society Report 2014

partnership with the One Laptop per Child OLCP) project. Similarly, in Chile, the Enlaces initiative, which partners with the private sector

has been very effective in improving Internet access in schools, resulting in 78 per cent of

schools being connected to the Internet in 2013, compared to just 44 per cent in 2009

and the country†s more recent One Tablet per Child (OTPC) initiative has helped increase the

learner-to-computer ratio in schools Internet connectivity in schools also depends on the development of the national

telecommunication infrastructure and on whether service providers have reached out to rural and sometimes geographically difficult areas with low population density (Partnership

2014). ) As discussed earlier in this chapter, rural areas often suffer from much lower network coverage and hence ICT uptake compared with

While connecting schools to the Internet and other ICTS is essential in order to foster e-education, it is equally important to look at

some cases, computers have been introduced in schools without Internet access, which effectively Chart 1. 23: Proportion of ICT-qualified teachers versus proportion of teachers trained to teach subjects

Data for Philippines refer to primary and lower secondary. Data for Venezuela refer to primary only.

Data for Montserrat and Saint lucia refer to secondary only. Data for Palestine refer to West bank schools only.

Data for Bahrain, Belarus, Morocco and Tunisia refer to 2008. For Morocco, ICT -qualified teachers figures refer to 2008.

Data for Azerbaijan, Barbados, Jordan, Saint lucia, Singapore, Trinidad and tobago, Uruguay, Philippines and Sri Lanka refer to public schools only

Source: UIS database, Partnership Questionnaire on WSIS Indicators, 2014 Proportion of teachers trained to teach basic computer skills (or computing

%P ro p o rt io n o f te a c h e rs

t ra in e d t o t e a c h s u b

je c ts u s in g I C T %Anguilla Argentina Azerbaijan Bahrain Barbados

Available data collected by UIS at the international level shows that education systems in countries seem to put more emphasis on

computer skills or computing (i e. ICT-qualified teachers)( Chart 1. 23. In most of the countries

document also includes two targets on data and monitoring and stresses the â€oeneed to take urgent

availability of disaggregated data to ensure that no one is left behindâ€. 26 The role of ICTS as a key development enabler

collect ICT data at the national level Monitoring progress towards achievement of the WSIS outcomes has been an integral component

indicators and collecting data and statistics on ICT. Since its creation in 2004, the Partnership has

national-level data on the indicators identified to measure the WSIS targets. The results of the survey are featured in the report

along with other data sources The Final review of the WSIS targets: Achievements, challenges and the way forward report is available at:

Using different data sources and big data analytics An important element in the discussion related

to the post-2015 development agenda and the setting of measurable goals and targets has

data and statistics are collected, analysed and disseminated, in view of the large data gaps prevailing in many developing countries in basic

statistics in the areas of the economy, health education, labour, etc. all of which are crucial to

High-level Panel of Eminent Persons on the Post -2015 Development Agenda published in 2013 see Box 1. 4 on the data revolution.

Since then it has received considerable attention within the statistical community, as well as in other circles concerned with the lack of official statistics and

development data According to the discussions, a data revolution considers new data sources, in addition to existing official sources:

besides governments other stakeholders such as the private sector, civil 27 Measuring the Information Society Report 2014

Box 1. 2: A decade of successful international cooperation on ICT measurement In 2004, ICT measurement was still in its infancy, with little

telecommunication statistics which have been collected by ITU for decades. In the wake of WSIS, and with the increasing focus

calls surfaced for reliable and comparable data in order to take stock of the emerging information society, identify digital

Its core list of 57 ICT indicators, covering many aspects of the information society and economy, is used widely by countries in the course

of their national ICT data collection. The methodological work developed by the Partnership has contributed significantly to

Data availability has increased also significantly over the past decade. In particular, there are more comparable data on ICT

infrastructure, household access and Internet users. For example at the beginning of the century, only around a dozen developing countries collected data on Internet users, while today there are

almost 50 developing countries collecting this indicator through official surveys (Chart Box 1. 2). Data on household access to

the Internet or a computer are now being collected by more than 100 economies worldwide, and data on Internet use in

businesses by almost 70 countries, although not on a regular basis (Partnership UNSC 2011. Similarly, whereas no data were

available on ICT access and use in schools, they have started now to be compiled in many developing countries (see section 1. 5

At the same time, major data gaps remain, in particular in developing countries and LDCS. This concerns, notably, statistics

on ICT use by individuals, businesses, governments and other public-sector organizations, ICT-related employment data, as

well as data related to online security and cybercrime, gender and youth, and cultural and environmental aspects.

The growing information society will increasingly require more and better statistics to assess the social, economic and environmental

impacts of ICTS. The Partnership, in close collaboration with national statistical systems and the international donor

Internet user data through official surveys, by level of development Note: Chart shows countries that have collected data on the number of

Internet users through official national surveys. Data are presented in three-year intervals and include countries that have collected

data for at least one year within those intervals Source: ITU 0 2 11 24 36

49 1 3 27 37 41 40 1 5 38 61 77 89 0 20

40 60 80 100 Developing Developed World 19 97 19 98 -0 0 20 01

a set of measurable targets that would help monitor and track progress towards achieving those goals over the next five years

and foster access to and increased use of telecommunications/ICTS Target 1. 1: Worldwide, 55%of households should have access to the Internet by 2020

Target 1. 2: Worldwide, 60%of individuals should be using the Internet by 2020 Target 1. 3:

Worldwide, telecommunication/ICTS should be 40%more affordable by 2020a Goal 2. Inclusiveness †Bridge the digital divide and provide broadband for allb

Target 2. 1. A: In the developing world, 50%of households should have access to the Internet by 2020

Target 2. 1. B: In the least developed countries (LDCS), 15%of households should have access to the Internet by 2020

Target 2. 2. A: In the developing world, 50%of individuals should be using the Internet by 2020

Target 2. 2. B: In the least developed countries (LDCS), 20%of individuals should be using the Internet by 2020

Target 2. 3. A: The affordability gap between developed and developing countries should be reduced by 40%by 2020

Target 2. 3. B: Broadband services should cost no more than 5%of average monthly income in developing countries by 2020

Gender equality among Internet users should be reached by 2020 Target 2. 5. B: Enabling environments ensuring accessible telecommunications/ICTS for persons with disabilities should be established in all

countries by 2020 Goal 3. Sustainability †Manage challenges resulting from telecommunication/ICT development Target 3. 1:

Cybersecurity readiness should be improved by 40%by 2020d Target 3. 2: Volume of redundant e-waste to be reduced by 50%by 2020

Greenhouse gas emissions generated by the telecommunication/ICT sector to be decreased per device by 30%by 2020

and adapt to the changing telecommunication/ICT environment Target 4. 1: Telecommunication/ICT environment conducive to innovation

Target 4. 2: Effective partnerships of stakeholders in the telecommunication/ICT environment society and international organizations should be

involved. New data sources could include big data mostly provided by private-sector companies which could help â€oeimprove the timeliness and

completeness of data, without compromising the relevance, impartiality and methodological soundness of the statistics†(UNSC, 2014

The topic of big data is gaining momentum in the statistical community. Chief statisticians gathering at the UNSC meetings in 2013

recognized that â€oebig data constitute a source of information that cannot be ignored by official statisticians†and that â€oeofficial statisticians must

organize and take urgent action to exploit the possibilities and harness the challenges effectively†(UNSC, 2014). 30 In view of declining

surveys in a number of countries, big data could provide important sources of more timely and

contributions from Member States. c Due to data limitations, currently mobile-broadband signal coverage is considering in determining this target

d Data being compiled by the Global Cybersecurity Index (GCI Source: ITU 29 Measuring the Information Society Report 2014

potentially becoming big data sources as well At the UNSC meeting in 2014, the commission reiterated its call for the global statistical

the use of big data for official statistics. The commission requested the group to include the

of big data for official statistics at regional subregional and national levels •To address the concerns of methodology

data and legislation related to big data •To address the issue of obtaining â€oeaccess at no cost†to big data from the private sector

for official statistical purposes, as well as the issue of access to transborder data or access to data on transboundary phenomena

•To develop guidelines to classify the various types of big data sources and approaches •To develop methodological guidelines

related to big data, including guidelines for all the legal aspects •To formulate an adequate communication strategy for data providers and users on the

issue of use of big data for official statistics •To reach out to other communities especially those more experienced

in IT issues or in the use of open data platforms The UN Global Working group on Big data for

Official Statistics was launched formally in June 2014, under the auspices of the UN Statistics Division.

The mandate of the group, of which ITU is a member, includes: provide strategic vision direction and coordination of a global programme

on big data for official statistics; promote practical use of sources of big data for official statistics

provide solutions for methodological, legal and privacy issues; promote capacity building; foster communication and advocacy of the use of big

data for policy applications; and build public trust in the use of private-sector big data for official

statistics ICTS are part of the debate on the data revolution big data and, more broadly, emerging data issues

in the post-2015 development debate. First the ICT sector in itself represents a new source

of data, provided by, for example, Internet and telecommunication companies. Second, the spread and use of ICTS allow public and private

entities across all economic sectors to produce store and analyse huge amounts of data. At the

same time, however, monitoring access to and use of ICTS by people, public entities and private

enterprises will be essential in order to identify the extent to which stakeholders in the ICT sector can be used as an alternative data source.

Without ICTS, no ICT-driven data revolution will take place In view of the link between big data and ICTS, work

is under way in ITU with a view to contributing to the debate and identifying new ways and means

of exploiting the potential of big data. The focus is primarily on the telecommunication/ICT sector

as a source of big data, including players such as operators and service providers, in the fixed

mobile and Internet sectors. Delegates attending the eleventh World Telecommunication/ICT Indicators Symposium (WTIS) in Mexico city in

December 2014 recommended that ITU should further examine the challenges and opportunities of big data, in particular data coming from ICT

companies; that regulatory authorities should explore the development of guidelines on how big data could be produced, exploited and stored

and that national statistical offices, in cooperation with other relevant agencies, should look into the opportunities for big data and address

current challenges in terms of big data quality Chapter 1. Recent information society developments 30 veracity and privacy within the framework of the

fundamental principles of official statistics. 33 The big data approach taken by ITU so far focuses on the following areas and questions

Standardization: 34 •Which standards are required to facilitate interoperability and allow technology integration in the big data value chain

•Which definitions, taxonomies, secure architectures and technology roadmaps need to be developed for big data analytics and technology infrastructures

•What is the relationship between cloud computing and big data in view of security frameworks •Which techniques are needed for data

anonymization for aggregated datasets such as mobile-phone records •How is exploited big data in different industries; what are the specific

challenges faced; and how can these challenges be addressed through international standards Regulation: 35 •What are the key regulatory issues at stake

and how can and should big data be regulated •How does big data impact on the

regulation of privacy, copyright and intellectual property rights (IPR transparency and digital security issues Box 1. 4:

What is a data revolution The report of the High-level Panel of Eminent Persons on the

Post-2015 Development Agenda to the UN Secretary-general which was published in May 2013, has called for, inter alia, a data

revolution taking advantage of new technology and improved connectivity: â€oewe also call for a data revolution for sustainable

development, with a new international initiative to improve the quality of statistics and information available to people and

governments. †This has prompted a â€oerevolution†in the debates taking place in the statistical communities at both the international and national

levels on what such a data revolution could entail and how it could be implemented. While no internationally agreed concept

has thus far been defined, the following elements seem to be part of a data revolution: 32 •In view of the ubiquitous availability of communication

networks, the use of new information technologies (e g mobile technologies) should be leveraged for improving the collection and dissemination of data

•Data should be further disaggregated (by gender income, age, geography, etc. to ensure that no one is

left out; in this regard, traditional statistical processes should be made more efficient •Sustained investment in national statistical capacity

both technical and institutional, is essential and needs to receive a major push from the international donor

community •The focus should go beyond data dissemination and also include investment in the development of concepts

measurement frameworks, classifications and standards •New, nontraditional data sources should be explored and leveraged to complement existing ones and

satisfy the demand for data needs in new areas, such as big data, geospatial information and geographical

information systems •Open data policies should be envisaged to ensure accountability and promote transparency •The role of data, statistics and monitoring for policy

-making and decision-making should be increased 31 Measuring the Information Society Report 2014 •What is the link between big data

and open data (crowdsourcing, cloud computing, etc •is there need a to regulate data management and service providers

•How can market dominance in the area of big data be prevented and the rights of

the data owners protected ICT data collection and analysis •How can big data complement existing ICT statistics to better monitor

information-society developments •Which type of data from ICT companies are most useful and for which purposes

•Which new ICT indicators could be produced from big data sources •What are the key issues that need to be

addressed, and by whom, in terms of collecting and disseminating big data in telecommunications •What is the role of national statistical

offices and how can big data complement official ICT data •How can big data from telecommunications inform not only

ICT but broader development policy in real time, leading to prompt and more effective action Chapter 5 of this report addresses some of

these questions and provides suggestions and recommendations for the way forward Chapter 1. Recent information society developments

32 1 Refers to countries where fixed-telephone penetration increased by more than 1 per cent in 2014

2 See https://gsmaintelligence. com /3 http://www. censusindia. gov. in/2011census/hlo/Data sheet/India/Communication. pdf

4 4g refers to fourth-generation mobile network or service. It is a mobile-broadband standard offering both mobility and very

high bandwidth, such as long-term evolution (LTE) networks (ITU Trends 2014 5 Data collection on Europe and North america will follow in 2014

6 For a list of IXPS, see for instance http://www. datacentermap. com/ixps. html 7 For more details on international submarine fibre-optic links, see Telegeography†s Submarine cable Map 2014, available at

http://submarine-cable-map-2014. telegeography. com 8 For further discussion on progress made towards connecting rural households to the Internet,

see Partnership (2014 Chapter on Target 1 9 Universal Postal Union (forthcoming 2014. Development strategies for the postal sector:

An economic perspective 10 See http://www. ifla. org/faife/world-report 11 Source: IMF World Economic Outlook Database, April 2014

12 Source: The Economist, April 12 2014, â€oenigeria†s GDP step change†13 Telefã nica, for instance, reduced its net debt by EUR 4 819 million in 2012 after several years of sustained increases in

borrowings. Source: Telefã nica Financial Report 2012, p. 18, available at http://annualreport2012. telefonica. com/pdf/FINANCIERO 2012 ENG. pdf

14 For example, the cost of buying a mobile cell tower in Europe and in Africa may not be very different,

because only a limited group of large global equipment vendors can deliver it, and equipment is traded usually in USD.

On-site setup expenditure may however differ because of varying labour costs across countries and regions, but that is only a part of the total CAPEX of

telecommunication operators 15 For instance, the average revenue per user per month for GSM services in India was less than USD 2 in March 2012, almost

unchanged from March 2011. Source: TRAI Annual Report 2011-12, p. 2, available at: http://www. trai. gov. in/Writereaddata

/Miscelleneus/Document/201301150318386780062annual%20report%20english%202012. pdf 16 Exceptions include China, Russian Federation, Japan and Republic of korea

17 http://www. statisticbrain. com/twitter-statistics /18 Eurostat news release of 16 december 2013 and

http://epp. eurostat. ec. europa. eu/statistics explained/index. php/Social media -statistics on the use by enterprises 19 Data refer mostly to the year 2011

20 The UN E-government Development Index is a composite benchmarking indicator based on a direct assessment of the state

of national online services, telecommunication infrastructure and human capital in all countries. See http://unpan3. un. org/egovkb/global reports/index. htm

21 See ITU Connect a School, Connect a Community Toolkit of Best Practices and Policy Advice, available at

http://connectaschool. org/itu-module-list 22 See Partnership (2014), Chapter on Target 2 23 For further information, see http://www. itu. int/wsis/index. html and http://www. broadbandcommission. org

24 Information on the post-2015 development agenda process is available at http://sustainabledevelopment. un. org/index. php?

menu=1561 25 See http://unstats. un. org/unsd/broaderprogress/progress. html 26 â€oeoutcome Document †Open Working group on Sustainable Development Goalsâ€, available at

http://sustainabledevelopment. un. org/focussdgs. html 27 Available at: http://unctad. org/meetings/en/Sessionaldocuments/CSTD 2014 DRAFTRES WSIS. pdf

28 See ECOSOC Resolutions 2008/3, 2009/7, 2011/16, 2012/5 and 2013/9 29 The ITU strategic goals are under discussion

and have to be examined and approved by the 2014 ITU Plenipotentiary Conference Endnotes 33 Measuring the Information Society Report 2014

30 E/CN. 3/2014/11 31 E/2014/24 and E/CN. 3/2014/35 (UNSC Report on the 45th session, 4-7 march 2014

32 UNSC Friends of The chair (FOC) Group on Broader Measures of Progress, in its report to UNSC, highlighted a number of

elements that should be part of a data revolution 33 See final report of WTIS-13, available at:

http://www. itu. int/en/ITU-D/Statistics/Pages/events/wtis2013/default. aspx 34 For further information on the work on big data carried out by the ITU Telecommunication Standardization Bureau (TSB),

see http://www. itu. int/en/ITU-T/techwatch/Pages/big data-standards. aspx 35 A background document on big data that was prepared for GSR-14 is available at

http://www. itu. int/en/ITU-D/Conferences/GSR/Pages/gsr2014/default. aspx 35 Measuring the Information Society Report 2014

Chapter 2. The ICT Development Index IDI 2. 1 Introduction to the IDI1 The ICT Development Index (IDI) is a composite

index combining 11 indicators into one benchmark measure that serves to monitor and compare developments in information

and communication technology (ICT) across countries. The IDI was developed by ITU in 2008 and first presented in the 2009 edition of

Measuring the Information Society (ITU, 2009. It was produced in response to ITU Member States†request to develop an ICT index and publish

it regularly. This section briefly describes the main objectives, conceptual framework and methodology of the IDI

The main objectives of the IDI are to measure •the level and evolution over time of ICT

•the digital divide, i e. differences between countries in terms of their levels of ICT development •the development potential of ICTS or

-telephone subscriptions, mobile-cellular telephone subscriptions, international Internet bandwidth per Internet user households with a computer, and

households with Internet access •Use sub-index: This sub-index captures ICT intensity, and includes three ICT intensity

and usage indicators (individuals using the Internet, fixed (wired)- broadband subscriptions, and wireless-broadband subscriptions

•Skills sub-index: This sub-index captures ICT capability or skills as indispensable input indicators.

It includes three proxy 37 Measuring the Information Society Report 2014 indicators (adult literacy, gross secondary

enrolment, and gross tertiary enrolment and is given therefore less weight in the computation of the IDI compared with

the other two sub-indices. 2 The choice of indicators included in the sub -indices reflects the corresponding stage of

better data become available. For example what was considered basic infrastructure in the past †such as fixed-telephone lines †is fast

characterizing intense Internet use, and is therefore included in stage 2 (as an indicator in the use sub

•Data availability and quality. Data are required for a large number of countries as the IDI is a global index.

There is relative paucity of ICT-related data, especially on ICT usage, in the majority of developing

countries. In particular, the three indicators included in the skills sub-index should be considered as proxies until

data directly relating to ICT skills become available for more countries •The results of various statistical analyses.

data and to explore whether the different dimensions are statistically well-balanced While the basic methodology has remained the

related data availability, the indicators included in the IDI and its sub-indices are under regular

Telecommunication/ICT Indicators (EGTI) and the ITU Expert Group on ICT Household Indicators EGH)( Box 2. 1

The definitions of the following core indicators of the Partnership on Measuring ICT for Development included in the IDI were revised at

•Percentage of individuals using the Internet The suggested reference period for latest Internet usage was changed from the last

twelve months to the last three months The twelvemonth period is used still by some countries,

that Internet usage is now sufficiently frequent that the majority of users will be captured with the shorter time-frame

Chapter 2. The ICT Development Index (IDI 38 Box 2. 1: ITU discussion forums on ICT statistics

the Expert Group on Telecommunication/ICT Indicators EGTI) and the Expert Group on ICT Household Indicators (EGH

field of ICT statistics and data collection, work through online discussion forums and annual face-to-face meetings.

periodically report back to the World Telecommunication/ICT Indicators Symposium (WTIS), ITU€ s main forum on ICT statistics

•Percentage of households with a computer The definition of computer was revised to include tablet or similar handheld

computers in addition to desktop and laptop computers, so as to reflect the uptake of these devices •The definition of household access was

refined so that, in order for a household to have access to ICT equipment or services the equipment/service should generally

be available for use by all members of the household at any time, regardless of

whether it is used actually. ICT equipment may or may not be owned by the household. 5 Apart from the revisions to indicator definitions

for the indicators international Internet bandwidth per Internet user and mobile-cellular subscriptions per 100 inhabitants would be reviewed.

For international Internet bandwidth per Internet user, the methodology used in previous IDI calculations was kept,

as there is no limit to the maximum value that could be achieved by a country. The reference value employed for this

indicator is used to screen outlier values. There were three economies that were identified as outliers, namely Hong kong (China), Luxembourg

The reference value for mobile-cellular subscriptions was lowered to 120. This value was derived from examining the distribution

of countries based on their mobile-cellular subscriptions per 100 inhabitants in 2013. In order to determine the reference value, prepaid

group of countries (23 countries with a mobile -cellular penetration between 110 and 120 per

-SIM ownership is driving up mobile-cellular subscription numbers, which is an issue in prepaid and, to a lesser extent, also in postpaid

data with data based on national household surveys (demand-side indicators. An indicator such as the percentage of individuals using a

reported these data to ITU for at least one year between 2011 and 2013. It is therefore too

early to substitute the current mobile-cellular subscription data in the IDI with mobile-phone user data.

In view of the methodological difficulties in collecting harmonized data on international Internet bandwidth, a review of

the definition of the indicator is currently under discussion in EGTI The IDI was computed using the same

methodology as in the past, applying the following steps (Figure 2. 2 and Annex 1 •Preparation of the complete data set.

This step includes filling in missing values using various statistical techniques •Normalization of data.

This is necessary in order to transform the values of the IDI indicators into the same unit of

measurement. The chosen normalization method was the distance to a reference value (or goalpost. The reference values

•Rescaling of data. The data were rescaled on a scale from 0 to 10 in order to

compare the values of the indicators and the sub-indices •Weighting of indicators and sub-indices.

This chapter presents the IDI based on data from 2013 in comparison with 2012. It should be

•Country data revisions. As more accurate data become available, countries provide ITU with revised statistics for

previous years, which have been taken into consideration. This also allows ITU to identify inconsistencies and revise

percentage of individuals using the Internet changing the reference period to the last three months) and percentage of

households with a computer (updating the definition of computer to include tablet and similar handheld computers †but

excluding smartphones •Differences among countries included in the IDI. The calculation of the IDI ranking

Chapter 2. The ICT Development Index (IDI 40 ICT access Reference value %1. Fixed-telephone subscriptions per 100 inhabitants

2. Mobile-cellular telephone subscriptions per 100 inhabitants 3. International Internet bandwidth (bit/s) per Internet user

4. Percentage of households with a computer 5. Percentage of households with Internet access 60 120

787†260 *100 100 20 20 20 20 20 ICT use Reference value %6. Percentage of individuals using the Internet

7. Fixed (wired)- broadband subscriptions per 100 inhab. itants 8. Wireless-broadband subscriptions per 100 inhabitants

100 60 100 33 33 33 ICT skills Reference value %9. Adult literacy rate 10.

Secondary gross enrolment ratio 11. Tertiary gross enrolment ratio 100 100 100 33 33 33

ICT Development Index 40 40 20 Figure 2. 2: ICT Development Index: indicators, reference values and weights

added based on data availability. Overall this version of the IDI includes 166 countries/economies as compared with

Section 2. 3 analyses the global digital divide by looking at the IDI results by level of development

also displays the widest range and the lowest average value (3. 19. The minimum value is

According to data from the European union EU), 85 per cent of Danes have some level of

computer skills (compared to the EU average of 67 per cent) and 42 per cent of the population

have â€oehigh†computer skills. 7 In 2010, the digital economy accounted for more than 5. 8 per cent

with 93 per cent of households with Internet access and households with a computer by end

2013. Next-generation access, providing speeds of at least 30 Mbit/s, was available to 73 per

The Danish Internet service provider (ISP) TDC is making investments to provide access to ultra-fast speeds for over

abundant international Internet bandwidth of more than 260 000 bit/s per Internet user in 2013 Denmark tops the IDI use sub-index.

The country†s broadband market is particularly well advanced. At 107 per cent, it has one of

the highest wireless-broadband penetration rates in the world, and an equally impressive fixed-broadband penetration rate of 40 per

has a slightly higher wireless-broadband penetration †Denmark surpasses the other top five IDI countries (see Chart 2. 1). In terms of LTE

population coverage, Denmark stands way above the regional (and world) average, with 65 per cent of the population covered. 9 In January 2013

2015 users in areas with the lowest speeds will Chart 2. 1: Fixed (wired)- broadband and wireless

ITU World Telecommunication/ICT Indicators database 33 35 36 38 40 110 75 87 105

ITU World Telecommunication/ICT Indicators database have access to mobile broadband at speeds of at

least 10 Mbit/s. 10 The Republic of korea ranks second in the IDI 2013. The former leader in the IDI 2012

Korea was the first country to offer 3g services commercially in 2002, and LTE was offered first

in 2011. Two years after commercializing the first LTE network, leading operator SK TELECOM reported that it passed the 10 million LTE

subscriber mark in April 2013; this represents 37 per cent of its total mobile subscriber base

Full coverage having being achieved (by April 2012, LTE was available nationwide), the wireless -broadband market is showing signs of saturation

with little growth over the past years. From 2012 to 2013, there was only a minimal increase in

In July 2013, SK TELECOM launched the â€oeworld`s first LTE-Advanced Networkâ€, with speeds of up

to 150 Mbit/s. In 2014, the Vice-president of the European commission for the Digital Agenda

to work towards a global definition of 5g and to cooperate in 5g research. 11

In regard to fixed broadband, there is still more potential for growth, and from 2012 to 2013 more fixed-broadband than wireless-broadband

-broadband compared with 370 000 wireless -broadband subscriptions. Data also show that the Republic of korea achieves the highest

advertised fixed-broadband speeds, with all subscriptions providing at least 10 Mbit/s. This compares with 75 per cent of fixed-broadband

highest proportion of households with Internet access worldwide. A somewhat lower proportion of 81 per cent of households have a computer

International Internet bandwidth is relatively low compared to other top IDI countries, at just over

one Tbit/s in 2012. There is however a sizeable domestic demand for data driven by the high

volume of local content, and domestic Internet bandwidth was compared ten times higher with international bandwidth Third-placed Sweden records an IDI value of

8. 67 in 2013. Like the remaining EU countries in the top ten, namely United kingdom (fifth

Internet bandwidth. The United kingdom stands out as the most dynamic of the top ten IDI

The growth in wireless-broadband subscriptions is having a major impact on ICT markets, and European top performers have been at the

has a wireless-broadband penetration of 89 per cent, followed by the United kingdom (87 per

were covered by a 3g signal and 26 per cent of the population was covered by an LTE network. 12

Denmark, Sweden and Finland are the countries with the highest LTE coverage in the European

region (European commission, 2014a The European commission is partnering with the Republic of korea to work towards a definition

of 5g. Furthermore, it launched a public -private partnership on 5g (5g PPP) in late 2013

that aims to â€oedeliver solutions, architectures technologies and standards for the ubiquitous next-generation communication infrastructures

Data from the EU confirm that household access is highly correlated with regular use of the Internet, which

underlines the importance of household access. 14 All European countries included in the top ten of

households with a computer and households with Internet) of at least 88 per cent. Iceland and

The netherlands display the highest levels of households with a computer, at 97 and 95 per

cent, respectively. 15 In Luxembourg and Norway 94 per cent of households have Internet access followed by Sweden (93 per cent), Finland (89

Data from the European Commission†s Digital Agenda underline the competitiveness of the European fixed

High levels of Internet connectivity at home and the availability of mobile Internet translate into high degrees of Internet usage in the IDI€ s top

ten countries. The Nordic countries stand out with the highest percentage of Internet users globally. In Iceland, 97 per cent of the (in-scope

population is using the Internet and 95 per cent of Norwegians, Swedes and Danes are online. 19

The availability of international Internet bandwidth is critical for ICT development. All IDI top performers benefit from the abundant

availability of international Internet bandwidth Bandwidth is highest (per Internet user) in such hubs as Luxembourg, Iceland, Sweden and the

United kingdom Hong kong (China) made its entry into the top ten of the IDI 2013, up from 11th position

in 2012. The economy ranks in ninth position with an IDI value of 8. 28.

Hong kong (China) is particularly strong on the access sub-index of the IDI, in which it ranks fourth.

provision of international Internet bandwidth a policy priority in order to secure reliable and low-latency Internet connectivity (see MIS 2013

In 2013, international Internet bandwidth stood at 1. 7 million bit/s per Internet user, which is

the second highest value after Luxembourg†s Hong kong (China) has the second highest fixed-telephone penetration globally, at 63 per

cent, and relatively high levels of household ICT connectivity, at 80 per cent of households with

Internet and 82 per cent with a computer. Both fixed-broadband (31 per cent) and wireless

-broadband (95 per cent) penetration are very high in Hong kong (China Dynamic IDI countries Even though most countries do not dramatically

increase in IDI rank within a year, there are some significant and remarkable developments. Table 2. 6 lists the so-called â€oedynamic†countries, i e

Globally, the number of mobile-broadband subscriptions20 grew by 24 per cent from 2012 to 2013.

High increases in wireless-broadband subscriptions can be seen in countries that were late adopters of 3g technology.

For instance, mobile-broadband services were only commercialized in mid-2013 in Burkina faso, 21 which is among the most dynamic countries

increase in the number of wireless-broadband subscriptions from 2012 to 2013 due to a rise

increased uptake by users. In the Republic of the Congo, 3g was launched in late 2011 by Airtel

Congo, and the entry of a second operator (MTN Congo) is reflected in a significant increase in

significant amounts of international Internet bandwidth and increased international Internet bandwidth per Internet user from around 6 000

bit/s in 2012 to close to 12 000 bit/s per user in 2013. The submarine cable system spans

Africa†s west coast, from South africa to CÃ'te d†Ivoire, and connects it to Europe.

It provides international Internet connectivity, which is of particular importance for enabling an island state such as Cape verde to join the information

society. Further progress in the access sub -index is attributable to the increase in mobile

growth in the number of wireless-broadband subscriptions. From 2012 to 2013, the number of subscriptions almost doubled, reaching a

and by 2013 operator CVMOVEL had expanded 3g services to all the islands of the archipelago. 23

2008 by state-owned operator Bhutan Telecom under its B-Mobile brand), major developments took place in 2013 that helped to boost

per cent in 2013 (see Chart 2. 3). Bhutan Telecom expanded its 3g services, which had been limited

to the nation†s capital Thimphu, to 15 out of 20 districts in Bhutan. Some USD 10.9 million were

mobile network. 24 Furthermore, Bhutan Telecom launched its high-speed 4g services in Thimphu Chart 2. 3:

Wireless-broadband penetration Bhutan, 2008-2013 Source: ITU World Telecommunication/ICT Indicators database 0. 0

0. 1 0. 3 1. 0 2. 5 15.6 0 2 4 6 8 10

The launch of mobile-broadband services by the country†s only private-owned operator Tashi Cell in late 2013 has helped

access to wireless-broadband services, uptake is ascribed also to the growing popularity of social media and increased availability of

smartphones. 26 Bolivia is among the most dynamic countries on the access sub-index(+4 ranks), and also shows

good progress in the generally more dynamic use sub-index(+5 ranks. The country reported

an important increase in international Internet bandwidth. International Internet bandwidth per Internet user27 almost doubled between

2012 and 2013, climbing to 9 000 bit/s per user While this is still one of the lowest absolute

figures in The americas region (only Cuba has a lower bandwidth per Internet user), it indicates that good progress in being made in improving

connectivity in the country. Bolivia has seen an important increase in wireless-broadband penetration, from 7 per cent in 2012 to 14 per

cent in 2013 In order to increase rural connectivity, landlocked Bolivia launched its first telecommunication satellite â€oetupac Katari†in late 2013.28 ENTEL â€

Bolivia†s state-owned operator †is contracting capacity from Tupac Katari, which it will use to

connect more rural areas of the country through the establishment of 3 000 telecentres by end

Internet bandwidth (close to 82 000 bit/s per Internet user. 30 It is well-connected to its neighbouring countries in the CIS region

and to Europe through two Black sea fibre -optic cables and terrestrial links. 31 This laid the

households to the Internet †penetration increased from 27 per cent in 2012 to 35 per

of households with a computer †from 33 per cent in 2012 to close to 40 per cent in 2013

At the same time, both wireless-and fixed -broadband penetration went up significantly Wireless-broadband penetration almost doubled, to 17 per cent,

Internet stands at 96 per cent and the proportion of households with a computer at 97 per cent

in 2013 (see Chart 2. 4). A report by ICTQATAR shows that Qataris are not only almost all very

-2013, the ownership of devices such as mobile phones and laptops increased significantly within the country†s mainstream population. 32

The report also highlights discrepancies in ICT connectivity between Qataris and westerners on the one hand, and transient labourers

westerners have an Internet connection at home penetration stands at 85 per cent for the overall population.

Smartphone penetration is also much lower among transient labourers (24 per cent). ) Therefore, â€oeincreasing the penetration of

newer devices such as smartphones and tablets particularly in specific demographic segments like the transient labour population†is one of the

1. 0 Mobile-cellular subscriptions International Internet bandwidth per Internet user Households with a computer Households

with Internet Fixed-telephone subscriptions Internet usersfixed-broadband subscriptions Active mobile -subscriptions broadband Secondary enrolment

Tertiary enrolment Literacy Bhutan 2012 2013 0. 0 0. 2 0. 4 0. 6 0. 8

1. 0 Mobile-cellular subscriptions International Internet bandwidth per Internet user Households with a computer Households

with Internet Fixed-telephone subscriptions Internet usersfixed-broadband subscriptions Active mobile -subscriptions broadband Secondary enrolment

Tertiary enrolment Literacy Bolivia 2012 2013 0. 0 0. 2 0. 4 0. 6 0. 8

1. 0 Mobile-cellular subscriptions International Internet bandwidth per Internet user Households with a computer Households

with Internet Fixed-telephone subscriptions Internet usersfixed-broadband subscriptions Active mobile -subscriptions broadband Secondary enrolment

Tertiary enrolment Literacy Bosnia and herzegovina 2012 2013 0. 0 0. 2 0. 4 0. 6 0. 8

1. 0 Mobile-cellular subscriptions International Internet bandwidth per Internet user Households with a computer Households

with Internet Fixed-telephone subscriptions Internet usersfixed-broadband subscriptions Active mobile -subscriptions broadband Secondary enrolment

Tertiary enrolment Literacy Burkina faso 2012 2013 0. 0 0. 2 0. 4 0. 6 0. 8

1. 0 Mobile-cellular subscriptions International Internet bandwidth per Internet user Households with a computer Households

with Internet Fixed-telephone subscriptions Internet usersfixed-broadband subscriptions Active mobile -subscriptions broadband Secondary enrolment

Tertiary enrolment Literacy Cape verde 2012 2013 0. 0 0. 2 0. 4 0. 6 0. 8

1. 0 Mobile-cellular subscriptions International Internet bandwidth per Internet user Households with a computer Households

with Internet Fixed-telephone subscriptions Internet usersfixed-broadband subscriptions Active mobile -subscriptions broadband Secondary enrolment

Tertiary enrolment Literacy Estonia 2012 2013 53 Measuring the Information Society Report 2014 Figure 2. 3:

IDI spider charts, selected dynamic countries, 2012 and 2013 (continued 0. 0 0. 2 0. 4

0. 6 0. 8 1. 0 Mobile-cellular subscriptions International Internet bandwidth per Internet user Households with

a computer Households with Internet Fixed-telephone subscriptions Internet usersfixed-broadband subscriptions Active mobile -subscriptions

broadband Secondary enrolment Tertiary enrolment Literacy Fiji 2012 2013 0. 0 0. 2 0. 4

0. 6 0. 8 1. 0 Mobile-cellular subscriptions International Internet bandwidth per Internet user Households with

a computer Households with Internet Fixed-telephone subscriptions Internet usersfixed-broadband subscriptions Active mobile -subscriptions

broadband Secondary enrolment Tertiary enrolment Literacy Georgia 2012 2013 0. 0 0. 2 0. 4

0. 6 0. 8 1. 0 Mobile-cellular subscriptions International Internet bandwidth per Internet user Households with

a computer Households with Internet Fixed-telephone subscriptions Internet usersfixed-broadband subscriptions Active mobile -subscriptions

broadband Secondary enrolment Tertiary enrolment Literacy Gambia 2012 2013 0. 0 0. 2 0. 4

0. 6 0. 8 1. 0 Mobile-cellular subscriptions International Internet bandwidth per Internet user Households with

a computer Households with Internet Fixed-telephone subscriptions Internet usersfixed-broadband subscriptions Active mobile -subscriptions

broadband Secondary enrolment Tertiary enrolment Literacy Mali 2012 2013 0. 0 0. 2 0. 4

0. 6 0. 8 1. 0 Mobile-cellular subscriptions International Internet bandwidth per Internet user Households with

a computer Households with Internet Fixed-telephone subscriptions Internet usersfixed-broadband subscriptions Active mobile -subscriptions

broadband Secondary enrolment Tertiary enrolment Literacy Oman 2012 2013 0. 0 0. 2 0. 4

0. 6 0. 8 1. 0 Mobile-cellular subscriptions International Internet bandwidth per Internet user Households with

a computer Households with Internet Fixed-telephone subscriptions Internet usersfixed-broadband subscriptions Active mobile -subscriptions

broadband Secondary enrolment Tertiary enrolment Literacy Qatar 2012 2013 Chapter 2. The ICT Development Index (IDI

54 Looking to the future, the country released its first national broadband plan in 2013, which

prioritizes broadband infrastructure development to make services faster, more affordable and more secure. One of the core projects

of the Qatar National Broadband Network is the deployment of a fibre-optic network infrastructure. 33

Chart 2. 1: Fixed (wired)- broadband and wireless -broadband subscriptions per 100 inhabitants, top five IDI countries, 2013

Source: ITU 33 35 36 38 40 110 75 87 105 107 0 50 100 150

Sweden Iceland United kingdom Korea (Rep Denmark Per 100 inhabitants Wireless-broadband subscriptions Fixed (wired)- broadband subscriptions

Chart 2. 4: Proportion of households with a computer and proportion of households with Internet access, 2012-2013, Qatar

Source: ITU World Telecommunication/ICT Indicators database 91.5 88.1 97.2 96.4 0 10 20 30

40 50 60 70 80 90 100 Households with a computer Household with Internet access

2012 2013 P e r 1 0 0 h o u se h o ld

s Thailand is one of the most dynamic countries on the use sub-index(+34 ranks),

which led to an improvement in its overall IDI ranking from 91st in 2012 to 81st in 2013.

In particular, the country†s wireless market proved to be extremely vibrant during the period 2012-2013:

more than 7 million new mobile-cellular subscriptions and close to 28 million new wireless-broadband subscriptions

were added within one year. Penetration rates stand at 138 per cent for mobile-cellular and

52 per cent for wireless-broadband services by end 2013. This is one of the highest wireless

-broadband penetration rates in Asia and the Pacific, only surpassed by the region†s high

-income economies. The launch of 3g was much anticipated in Thailand, following the long delay

in the auctioning of 3g licences. In December 2012, licences were awarded finally to three Thai operators, 34 providing high-speed mobile

Internet connectivity to users. The rapid uptake of mobile-broadband services was spurred by heated competition among operators offering

subsidized smartphones and promotions on mobile data plans. 35 During 2013, operators further extended their wireless infrastructure and services

throughout Thailand, and are planning to provide further network updates. 36 Figure 2. 3: IDI spider charts, selected dynamic countries, 2012 and 2013 (continued

Note: These charts show normalized values of the indicators included in the IDI Source: ITU

0. 0 0. 2 0. 4 0. 6 0. 8 1. 0 Mobile-cellular subscriptions

International Internet bandwidth per Internet user Households with a computer Households with Internet Fixed-telephone subscriptions

Internet usersfixed-broadband subscriptions Active mobile -subscriptions broadband Secondary enrolment Tertiary enrolment Literacy Thailand 2012

2013 0. 0 0. 2 0. 4 0. 6 0. 8 1. 0 Mobile-cellular

subscriptions International Internet bandwidth per Internet user Households with a computer Households with Internet Fixed-telephone

subscriptions Internet usersfixed-broadband subscriptions Active mobile -subscriptions broadband Secondary enrolment Tertiary enrolment Literacy United arab emirates

2012 2013 55 Measuring the Information Society Report 2014 2. 3 Monitoring the digital divide:

Developed developing and least connected countries Tracking the global digital divide is one of the main objectives of the IDI.

The digital divide can be understood as the difference in ICT development within and between countries, regions or other

groupings. In this section, IDI performance will be analysed and compared with regard to levels of economic) development, and on the basis of IDI

groupings (based on IDI values. The digital divide is measured by looking at differences in IDI values between these different groups.

As a composite index that consolidates several ICT indicators into one single value, the IDI is an especially useful tool

for comparing differences in ICT developments Based on the 2013 and 2012 data presented in this chapter, the current (2013) global divide is

measured and differences from 2012 are identified This also serves to determine whether the digital divide has been increasing

or decreasing over the past year. Special emphasis is placed on those countries that lie at the bottom of the IDI †the so

-called â€oeleast connected countries†(LCCS The analysis of IDI values by level of development reveals a significant disparity between developed

and developing countries. Developed countries reach an average IDI value of 7. 20, while the

developing-country average is almost half that, at 3. 84. The increase in average value between 2012

and 2013 was almost the same in developing +0. 17) and developed countries(+0. 18) when

measured in absolute terms (Tableâ 2. 7). This indicates that developing countries †as a group †are not progressing enough in terms of ICT

development to close the gap. Given their lower starting point, however, the average IDI value of

developing countries increased twice as much +4. 9 per cent) compared to developed countries +2. 5 per cent) when measured in relative terms

Chart 2. 5 The range of IDI values is much higher in developing countries, a group that includes

both top IDI countries such as the Republic of Korea as well as LCCS. The measures of dispersion

and variation (Stdev, CV and range) continue to increase, indicating that disparities within the group of developing countries are rising

Furthermore, the minimum value is not only significantly lower in developing than in developed countries, but has increased also much less.

While both developed and developing countries exhibit progress in ICT development, the analysis shows that developing countries are not advancing

enough to catch up with developed countries and that within the group of developing countries disparities in ICT development continue to rise

The most important increases in the access-sub index occurred in developing countries, with an

average value increase almost three times that of developed countries. At the same time, the difference in average value (between developed

and developing countries) is lower than in the use Table 2. 7: IDI by level of development, 2012-2013

Note:**Simple averages. Stdev=Standard deviation, CV=Coefficient of variation Source: ITU IDI 2012 IDI 2013 Change in

average value 2012-2013average value*Min. Max. Range Stdev CV Average value*Min. Max. Range Stdev CV

World 4. 60 0. 93 8. 81 7. 87 2. 19 47.61 4. 77 0. 96 8. 86 7. 90

2. 22 46.44 0. 17 Developed 7. 03 4. 42 8. 78 4. 35 1. 08 15.39 7. 20 4. 72 8. 86 4. 14 1

. 03 14.24 0. 18 Developing 3. 67 0. 93 8. 81 7. 87 1. 75 47.61 3. 84 0. 96 8. 85 7. 89 1

. 80 46.93 0. 17 Chapter 2. The ICT Development Index (IDI 56 sub-index, showing that developing countries

are improving their ICT infrastructure, which is a prerequisite for intensifying ICT usage (Chart 2. 6). However, progress was slower than in the

other sub-indices of the IDI. Fixed-telephone penetration is decreasing in developing and developed countries.

Mobile-cellular subscription growth rates are slowing down, indicating that mobile-cellular penetration is nearing saturation

including in developing countries. The growth in household ICT connectivity is much higher in developing countries, where around three

-quarters of households are connected not yet to the Internet, leaving ample room for growth This also holds true for international Internet

bandwidth, which is still at very low levels in many developing countries The use sub-index is the most dynamic,

showing the biggest improvements, particularly in developing countries. However, the average value is lowest in this sub-index,

affording the biggest potential for growth. The difference in average value between developed and developing

countries is also highest in the use sub-index which underlines that significant differences exist

with regard to the intensity of ICT usage (Chart 2. 7). In many developing countries, the availability

and uptake of wireless-broadband and fixed -broadband services in particular is still relatively limited. On average, fixed-broadband penetration

reached 6 per cent in developing countries by end 2013, compared with 27 per cent in developed

countries. While many developing countries saw important increases in their use sub-index value following the introduction of 3g services in

2012/2013, there were still a few countries that had launched not services by end 2013. In 2014

close to 4. 3 billion people, most of them living in the developing world, were not using the Internet

Bringing those people online is an important task in developing countries Differences are developed smallest between

and developing countries in the skills sub -index. However, in the absence of alternative measurements, this sub-index includes three

proxy indicators that do not measure actual Chart 2. 5: IDI by level of development

Chart 2. 6: IDI access sub-index by level of development Chart 2. 7: IDI use sub-index by level of

development Note: Simple averages Source: ITU 2012 2013 ID I 4. 6 7. 0 3. 7

4. 8 7. 2 3. 8 0 1 2 3 4 5 6 7 8

9 World Developed Developing Change +4. 9 %Change +2. 5 %Change +3. 9 %2012 2013

ID I a cc e ss s u b -i n d e x 5. 3

7. 6 4. 4 5. 4 7. 7 4. 5 0 1 2 3 4

5 6 7 8 9 World Developed Developing Change +3. 6 %Change +1. 2 %Change

+2. 6 %2012 2013 ID I u se s u b -i n d e

x Change +15.4 %2. 9 5. 5 1. 9 3. 2 5. 9 2. 2

0 1 2 3 4 5 6 7 8 9 World Developed Developing Change +6. 3

%Change +10.6 %57 Measuring the Information Society Report 2014 Table 2. 8: IDI by groups, 2012 and 2013

Note:**Simple averages. Stdev=Standard deviation, CV=Coefficient of variation Source: ITU Group IDI 2012 IDI 2013

Number of coun -tries Average value*Min. Max. Range Stdev CV Average value*Min. Max.

Range Stdev CV High 42 7. 52 6. 46 8. 81 2. 35 0. 70 9. 27 7. 69 6. 70 8. 86

2. 16 0. 63 8. 22 Upper 40 5. 38 4. 50 6. 45 1. 95 0. 56 10.38 5. 63 4. 75 6. 67 1

. 91 0. 58 10.26 Medium 42 3. 69 2. 62 4. 48 1. 86 0. 54 14.61 3. 88 2. 79 4. 72 1

. 93 0. 58 14.97 Low 42 1. 83 0. 93 2. 61 1. 68 0. 44 23.77 1. 93 0. 96 2. 77 1. 81

0. 46 24.03 Total 166 4. 60 0. 93 8. 81 7. 87 2. 19 47.61 4. 77 0. 96 8. 86 7. 90

2. 22 46.44 ICT skills, but rather levels of literacy and school enrolment. Data change very little over time and

advances in skills do not show immediate effects One shortcoming of grouping countries into only two categories (developed and developing

is that these categories include countries at very different stages of ICT development. The developing-country group,

which is defined on the basis of the United nations classification, also includes ICT champions such as the Republic

of Korea, Hong kong (China) and Singapore Therefore, for the purpose of comparing levels of ICT development and analysing the digital

divide, countries are grouped also on the basis of their IDI value. To this end, four groups/quartiles

International Internet bandwidth availability is limited very, thus constraining Internet connectivity and driving up ICT prices

which in turn hampers usage of ICTS. Few households (less than 5 per cent in the majority of LCCS) are connected to the

Internet and fewer than 5 per cent of households in all LCCS have a computer.

Basic voice services are more widely available although LCCS like Eritrea (6 per cent), Myanmar (13 per cent) and

The increased availability of wireless broadband could help bring more people online in LCCS. Today, in the majority

of LCCS, few people use the Internet: an estimated less than 2 per cent the population is online in Eritrea, Myanmar, Guinea

Data not available 59 Measuring the Information Society Report 2014 the profitability of various kinds of economic

selected for their high data availability for a large number of countries Past editions of this report (see MIS 2013) have

skills and fixed telecommunication infrastructure Table 2. 9: Partial correlation analysis of IDI, population and geographic characteristics

Data on urban population, population density, land area and GNI p. c. are sourced from the World bank.

Data on population size are sourced from UNPD.**Correlation is significant at 0. 01 level

significant and persistent urban-rural digital divide On the one hand, mobile-cellular coverage for rural populations has reached very high levels

inhabitants covered by a 2g mobile-cellular signal by 2013. On the other hand, 3g mobile -cellular coverage was comparatively low for

0 1 2 3 4 5 6 7 8 9 10 0 10000 20000 30000 40000 50000 60000

pronounced when it comes to data on Internet access and use. Access to the Internet (be it

narrowband or broadband, fixed or wireless) is extremely low for rural households in developing countries, while rural households in developed

countries enjoy comparable access to their urban counterparts, albeit with slight variations in type of access and (usually) a small lag in levels of

digital divide that prevails in many developing countries. People living in rural areas, particularly in developing countries, are disadvantaged

economic means to pay for broadband Internet services, as well as the skills to make effective use of ICTS.

data are available for both sets of indicators. The following steps were performed •First, developed countries were excluded

in respect of which data pertaining to both the MDG indicators and the IDI are available for at least 16 countries. 41 The MDG

Goal 8 (fixed-telephone and mobile-cellular penetration and percentage of Internet users) were considered not,

since they are also included in the IDI •Third, a simple correlation analysis between IDI

GNI p. c. and MDG indicators was performed in order to explore the relationship between the

of multilingualism on the Internet; and ensure access to ICTS for more than half of world†s inhabitants

147 economies for which data were available for 2002 and 2013 shows that the global IDI value has doubled almost from 2. 52

would require different sets of data (including micro data) collected from official surveys Therefore, the analysis should be considered as a

first step in an attempt to quantify the relationship between ICT performance and MDG progress. The

data are available •Almost all MDG indicators that are included under MDG 1, MDG 4, MDG

os s lo ca liti es a nd r eg io ns a nd m

os o f g ir ls to b oy s in p ri m ar

os e w ho n ee d it 6 5 P ro po rti on

os is 6 9b P re va le nc e ra te s as so

os is 6 9c D ea th ra te s as so ci at ed

os is 6 10 a P ro po rti on o f t ub er

children, through the analysis of health data collected through public health applications and by serving as a platform for exchange and

basic sanitation†also displays a significant Chapter 2. The ICT Development Index (IDI 74 Box 2. 4:

of smartphones to capture essential data on the patients and monitor their treatment has accelerated progress.

malaria information system (e-MIS) uploaded on the health workers†mobile devices shows malaria volunteers where to find

patients and the status of their treatment, and helps analyse the situation and trends Source:

Furthermore, ICTS such as mobile phones and the Internet can help inform people and allow them to share information

on the use and availability of facilities. 46 Results are mixed for the other targets under MDG 7

of MDG indicators, where 2002 and 2011 data are available for both sets of indicators.

data are available for 2002 and 2011 for both the MDG indicators and the IDI

earlier section where 2011 data were analysed improvements in the level of ICT access and use between the ten-year periods have shown

Internet Goal 7 The significant positive correlation between percentage change in carbon dioxide (CO2 emissions and percentage increase in IDI

analysis will require different data sets, including micro data on ICT usage collected from official

surveys. Micro data offer analysts and researchers ample information and considerable flexibility to apply quantitative models that identify

relationships and interactions between indicators and topics covered in a survey, thereby fostering the diversity and quality of research and analysis

2 Data on the indicators included in the skills sub-index are sourced from the UNESCO Institute for Statistics (UIS.

including TV, electricity, refrigerator piped water, etc. A similar principle has been adopted for ICT equipment and services,

http://data. worldbank. org/about/country-classifications/country-and-lending-groups#High income 7 Based on 2011 data:

https://ec. europa. eu/digital-agenda/sites/digital-agenda/files/DAE%20scoreboard%202013%20-%20

3-INTERNET%20use%20and%20skills. pdf 8 http://presse. tdc. dk/pressemeddelelser/tdc-klar-til-100-mbit-s-ogsa-pa-kobberkabler-987457

9 https://ec. europa. eu/digital-agenda/sites/digital-agenda/files/DK%20%20-%20broadband%20markets. pdf

10 http://www. gsma. com/spectrum/wp-content/uploads/Digitaldividend/DDTOOLKIT/auctions-summary. html#denmark and

http://danishbusinessauthority. dk/800-mhz-auction 11 http://europa. eu/rapid/press-release ip-14-680 en. htm

12 https://ec. europa. eu/digital-agenda/sites/digital-agenda/files/DAE%20scoreboard%202013%20-%202-BROADBAND%20

MARKETS%20. pdf 13 http://europa. eu/rapid/press-release ip-14-680 en. htm and http://ec. europa. eu/research/press/2013/pdf/ppp/5g factsheet. pdf

14 https://ec. europa. eu/digital-agenda/sites/digital-agenda/files/DAE%20scoreboard%202013%20-%203-INTERNET%20

USE%20and%20skills. pdf 15 Qatar (ranked 34th) has 97 per cent of households with a computer by end 2013

15 https://ec. europa. eu/digital-agenda/en/scoreboard 17 Ofcom and the European union define superfast services as those delivering download speeds of 30 Mbit/s or more

18 https://ec. europa. eu/digital-agenda/en/pillar-4-fast-and-ultra-fast-internet-access

19 In these countries, the in-scope population for data on Internet users is aged individuals 16-74

20 Refers to the indicator active mobile-broadband subscriptions. Mobile-broadband subscriptions generally make up the

largest part of wireless-broadband subscriptions, which also includes terrestrial (fixed) wireless and satellite broadband subscriptions

21 http://www. itnewsafrica. com/2013/05/airtel-launches-first-3-75-g-service-in-burkina-faso/and

http://news. aouaga. com/documents/docs/Rapportarcep. pdf 22 http://www. thisdaylive. com/articles/a-year-after-wacs-is-faster-more-affordable-mobile-broadband-becoming-a

-reality-/156011/and http://wacscable. com/index. jsp 23 http://www. cvmovel. cv/nacional-gsm-3g-edge-e-gprs

24 http://www. telecomasia. net/content/bhutan-telecom-expand-3g-network 25 http://www. tashigroup. bt/?

/p=1058 and http://www. kuenselonline. com/tashicell-goes-3g/#./#U4rlg3ksx8e 26 http://www. kuenselonline. com/530-increase-in-mobile-broadband-users/#.

/#U4rsghksx8e 27 Purchased capacity Endnotes 81 Measuring the Information Society Report 2014 28 http://www. bbc. com/news/world-latin-america-26850393

29 http://www. entel. bo/inicio3. 0/index. php/sala-de-prensa/item/309-contrato-entel-abe and

http://www. entel. bo/inicio3. 0/index. php/sala-de-prensa/item/272-telecentro-san-juan-de-chiquitos

30 http://www. eeca-ict. eu/countries/georgia 31 http://www. submarinecablemap. com/#/landing-point/poti-georgia

as well as the household access data, excludes transient labourers, which account for a significant proportion of residents in Qatar.

According to data from ICTQATAR, â€oetransient labourers†make up 27 per cent of the overall population 33 http://qnbn. qa/qatar-vision-2030

35 http://www. telecompaper. com/news/thai-operators-reduce-prices-of-smartphone-data-plans--900198

36 http://www. telegeography. com/products/commsupdate/articles/2013/05/09/true-4g-launch-trumps-rivals-ais-claims

-800000-users-at-official-2100mhz-launch-dtac-waits-until-june/?/utm source=Commsupdate&utm campaign=40c2385114 -Commsupdate+09 may+2013&utm medium=email&utm term=0 0688983330-40c2385114-8868625

World bank, see http://data. worldbank. org/indicator/NY. GNP. PCAP. CD 39 See: Czernich, N.,Falck, O.,Kretschmer, T. and Woessmann, L. 2009), Broadband Infrastructure and Economic growth

http://www. broadbandcommission. org/net/broadband/Reports/Report 2. pdf. For a criticism see: Kenny, C. 2011

http://www. cgdev. org/files/1425798 file kenny overselling broadband final. pdf 40 http://www. unesco. org/new/fileadmin/MULTIMEDIA/HQ/CI/CI/pdf/wsis/ungis joint statement wsis 2013. pdf

41 A correlation analysis for a particular MDG indicator was performed if there is available data for at least one year between

2010 and 2012. The number of countries included in the analysis varies from 16 to 101 developing countries.

since it is the closest year to available MDG data 42 Broadband Commission. Transformative Solutions for 2015 and Beyond a Report of the Broadband Commission Task force

on Sustainable Development and http://www. broadbandcommission. org/Documents/Climate/BD-bbcomm-climate. pdf and The State of Broadband 2013 http://www. broadbandcommission. org/documents/bb-annualreport2013. pdf

43 The Pearson correlation coefficient (r is a measure of the strength and direction of association that exists between two

45 http://www. un. org/millenniumgoals/pdf/Goal 6 fs. pdf 46 http://www. multidisciplinaryjournals. com/wp-content/uploads/2014/01/THE-ROLE-OF-INFORMATION-COMMUNICATION

-pdf 47 UNCTAD (2011), Measuring the Impacts of Information and Communication Technology for Development http://unctad. org/en/docs/dtlstict2011d1 en. pdf

83 Measuring the Information Society Report 2014 Chapter 3. Regional IDI analysis This chapter, which is based on the results of

for each of the six ITU Telecommunication Development Bureau (BDT) regions (Africa Americas, Arab States, Asia and the Pacific

Europe displays by far the highest average IDI value of 7. 14. The regional IDI values of the CIS (5. 33), The americas (4. 86

increases in mobile-cellular penetration from 2012 to 2013 were registered also in Guinea from 49 per cent to 63 per cent), Mozambique

A lack of international Internet bandwidth is seriously hampering ICT development in Africa Although the region has been connected to

are lacking international Internet connectivity Chart 3. 3: Mobile-cellular subscriptions per 100 inhabitants, 2012 and 2013, Africa

ITU World Telecommunication/ICT Indicators database 0 20 40 60 80 100 120 140 160

1â 000 bit/s of international Internet bandwidth per Internet user at their disposal. Being connected to four international submarine

cable systems, Kenya has the highest amount of international Internet bandwidth, both in total and per Internet user, at 50 000 bit/s per user (see

MIS 2013. Seychelles (24 000 bit/s) and Mauritius 24 500 bit/s) also have relatively high amounts

of bandwidth per Internet user, partly because of their very small populations and hence small number of Internet users

The divide between Africa and the world becomes most visible when looking at ICT household penetration:

by end 2013, on average less than 10 per cent of households in the region had access to the Internet at home,

compared to the global average of 40 per cent and the developing-country average of 28 per cent

While 3g networks are continuing to be built and expanded across the region numerous countries saw some important

Burkina faso, 3g was launched finally in 2013 reaching a penetration of 9 per cent by end

Data from South African operators show that not only is wireless -broadband penetration reaching higher levels

consuming more data, indicating an increase in the intensity of usage. MTN reported a growth

of 63 per cent in data volumes in the first half of 2013 and Vodacom reported that on average

users were generating 75 per cent more data traffic per device than a year ago. 5 Wireless

broadband is of particular importance in the region because fixed-broadband infrastructure is lacking. The vast majority of African countries

Africa was home to 150 million Internet users by end 2013. This corresponds to around 17 per

and high levels of multi-SIM ownership (GSMA and Deloitte, 2013. Furthermore, the very high

Data from household surveys show that the actual number of people using a mobile-cellular phone is

much lower than the number of subscriptions In Tunisia, 72 per cent of individuals were using

a mobile-cellular phone, compared with a mobile-cellular penetration of 118 per cent in 2012.

of individuals using a mobile-cellular phone Fixed-telephone penetration is extremely low in the Arab States region, with a regional average

of 9 per cent in 2013. This further highlights the importance of mobile networks in the region

The Arab States region and in particular the GCC countries are well-connected to submarine Internet cables.

The United arab emirates boasts the highest amount of international Internet bandwidth per Internet user (around 52 000

bit/s per user) in the region. Furthermore, the country almost doubled its Internet bandwidth

between 2012 and 2013. Oman, too, saw a significant increase in total international Internet bandwidth, up from 17 792 Mbit/s in 2012 to 82

010 Mbit/s in 2013. In 2013, the Europe-Persia Express Gateway that connects the United Arab

Emirates and Oman to Germany via the Islamic Republic of Iran went live, increasing the region†s

international Internet connectivity. 6 Furthermore the Gulf Bridge International (GBI) system completed its â€oenorth Route†terrestrial link in

2013, which connects the Gulf region to Europe. 7 Chart 3. 4: IDI values compared with the global, regional and developing/developed-country averages

with a computer. The remaining GCC states all reach high household ICT penetration rates of

households to the Internet in 2013, penetration increasing from 39 per cent in 2012 to 46 per

households with a computer as a result of the National PC Initiative. Through this initiative eligible families (those benefiting from the

social welfare programme with at least one child enrolled in primary school, secondary school or higher education studies) in Oman

are offered one free computer per student Furthermore, Omantel provides discounted broadband Internet offers for eligible

customers. 8 Wireless-broadband penetration levels vary considerably across the region. The number of subscriptions exceeds the population in Bahrain

and upgrading their mobile -broadband networks †in Qatar LTE is available throughout the entire country9 †Algeria

and Djibouti had launched not yet mobile -broadband services in 2013 (see Chartâ 3. 5 However, there is some room for optimism

as 3g licences were awarded finally to three Algerian operators by end 2013.10 Penetration is also extremely low in Yemen (0. 3 per cent

reach Internet user penetration rates of 50 and 56 per cent, respectively. On the other hand, in the LCCS Mauritania and Djibouti, less than 10 per

ITU World Telecommunication/ICT Indicators database 2012 2013 P e r 1 0 0 i

countries have reached a mobile-cellular penetration of 100 per cent or above by end 2013.

Nepal achieved a mobile-cellular penetration rate of 71 per cent in 2013, up from

million new mobile-cellular subscriptions were added in 2013, taking the penetration rate up to 89 per cent

of international Internet bandwidth in the region, and indeed one of the highest volumes in the world, at close to 9. 5 million Mbit/s12 by

a secure and low-latency Internet connection and the telecommunication regulator has made the attraction of international submarine

cables a policy priority13 (see MIS 2013, Box 2. 10. In 2013, the Southeast asia Japan Cable

system, international Internet connectivity was boosted in these countries. Additional international Internet bandwidth is of particular

importance for sustaining ICT growth and ensuring Internet connectivity for an increasing number of users in populous countries such as

China (with an estimated 600 million Internet users) and the Philippines (with an estimated 36.5 million Internet users by end 2013.

Within the Philippines, domestic connectivity was further improved by connecting some of the Chart 3. 6:

IDI values compared with the global, regional and developing/developed-country averages Asia and the Pacific, 2013

Source: ITU ID I Developing Asia and the Ko re a R ep 0 1

2 3 4 5 6 7 8 Pacific World Developed H on g Ko ng

live in the summer of 2013.15 Regional Internet connectivity was enhanced further when the Tonga Cable, connecting Fiji and Tonga, and the

international Internet connectivity; these include in particular, the landlocked and least connected countries Afghanistan, Bhutan and Nepal, with

less than 4 000 bit/s per Internet user The regional divide in the Asia-Pacific region

of households had a computer. Data from the annual ICT household survey show that, since

2008, computers have replaced telephones as the most commonly available ICT device in Thai homes. Furthermore, the majority of households

Table 3. 5: IDI †Asia and the Pacific Note:**Simple averages Source: ITU Economy

In Thailand, where 3g was launched very late, wireless-broadband penetration went up from 11 per cent in 2012 to more than 50 per

benefit from the extension of wireless broadband to connect more people with ICTS By end 2013, fixed-broadband penetration

2014, with China Mobile entering the fixed-line market. 17 China†s broadband strategy, published

ITU World Telecommunication/ICT Indicators database 2012 2013 P e r 1 0 0 i

household penetration of 50 per cent and a 3g penetration rate of 32.5 per cent by 2015.18

Indian Internet users. Comparing the two, the proportion of the population using the Internet is much higher in China (44 per cent) than India

15 per cent. India has one of the lowest rates in the region (and globally:

proportion of Internet users. Japan (86 per cent the Republic of korea (85 per cent) and Australia and New zealand (both 83 per cent) exhibit the

The mobile markets in the CIS are predominately prepaid with typically high rates of multi-SIM ownership

Furthermore, markets are very competitive, with a relatively high number of mobile operators. In the majority of CIS countries, at least four mobile

operators are active in the market. The Russian Federation, for example, has one of the most de

-concentrated mobile markets globally, with three national operators and several regional operators competing for 143 million potential customers

Data from household surveys collected in a number of CIS countries underline that mobile -cellular penetration, measured as the number

of mobile-cellular subscriptions, can give no more than an indication of the actual number of subscribers (i e. a mobile-cellular penetration

of above 100 per cent does not mean that every inhabitant has a mobile-cellular subscriptions

For instance, whereas mobile-cellular penetration in Ukraine stood at 130 per cent by end 2012,9

of households did not have access to a mobile -cellular telephone in 2012, even though mobile -cellular penetration stood at 108 per cent

households with a computer in the region by end 2013, at 67 per cent and 70 per cent, respectively

and a computer. Kyrgyzstan and Uzbekistan display a very low ICT household penetration, with less than 10 per cent of households in the country

having access to the Internet. Total international Internet bandwidth is by far highest in the Russian Federation,

which is connected through a number of terrestrial links to both Europe and the Asia-Pacific region. 20 However, given its

Internet user the country is below most other CIS countries. International Internet bandwidth per Internet user is highest in Moldova (115 845

bit/s per user), followed by Belarus (94 797 bit/s per user) and Georgia (82 094 bit/s per user

Kyrgyzstan and Uzbekistan have limited very bandwidth, which hampers Internet connectivity and hence further development of the ICT sector

in those countries By end 2013, half of CIS countries had reached a wireless-broadband penetration of more than

45 per cent. The Russian Federation boasted the highest penetration (60 per cent), followed by Kazakhstan (56.5 per cent.

The Russian Federation was one of the first countries in the region to launch 3g services in 2007.21 Since

then, operators have expanded their mobile -broadband networks beyond the main cities Table 3. 6: IDI †CIS

Note:**Simple averages.****Until 2009, the CIS region included the above countries. Georgia exited the Commonwealth on August 18, 2009,

but is included in this report Source: ITU Economy Regional rank 2013 Global rank 2013 IDI 2013

to provide further Internet connectivity. LTE services were launched in the Russian Federation in 2012.22 The highest growth in wireless

-broadband penetration from 2012 to 2013 took place in Georgia †from 9 per cent in 2012 to 17

The slow growth in wireless-broadband penetration in Ukraine explains why the country is falling back in international comparison

-telephone penetration of 48 per cent by end 2013 Penetration is also relatively high (17 per cent) in

ITU World Telecommunication/ICT Indicators database 2012 2013 P e r 1 0 0 i

only one country (Andorra) had a mobile -cellular penetration of less than 90 per cent Most remaining countries well exceeded

most European mobile-cellular markets. Data from the Eurobarometer underlines this finding on average, 92 per cent of European union

citizens (the majority of countries in the region are members of the EU) had access to a mobile

phone in 2013 (European commission, 2014b The region benefits from an abundant supply of international Internet bandwidth.

The highest levels are reached in international hubs such as Germany, Luxembourg and the United kingdom High amounts of bandwidth per Internet user

as registered in most European countries, ensure that a large number of Internet users can go online at high speeds

Around three-quarters of European households have access to the Internet at home. The highest proportions of households connected to the

Chart 3. 10: IDI values compared with the global, regional and developing/developed-country averages

tv ia Ita ly Cr oa tia G re ec e Li th ua ni

Internet are found in Iceland (96 per cent Luxembourg (95 per cent), The netherlands (95 per

with a computer. In the majority of countries in Europe (25 out of 40), 70 per cent of households

of households have a computer. Albania ranks last in the region also in terms of household ICT

a computer and 24.5 per cent with Internet access by end 2013. Among the countries that made the

Internet from 2012 to 2013 are Italy (from 63 to 69 per cent), Czech republic (from 65 to 73 per cent

Growth in wireless-broadband penetration continued at double-digit rates from 2012 to 2013 in the majority of European countries.

operator launched its 3g services in early 2013 increasing competition in the market. 25 Operators

LTE services to customers. The top five countries in the world in terms of fixed-broadband penetration (Monaco, Switzerland, Denmark

Percentage of Individuals using the Internet, Europe compared to global and developed -country average, 2013

Data on Individuals using the Internet for Eurostat members are sourced from Eurostat. Eurostat collects data for Internet users aged 16-74 years old

Source: ITU World Telecommunication/ICT Indicators database World Developed 0 10 20 30 40 50

60 70 80 90 100 %101 Measuring the Information Society Report 2014 penetration stands at 6 per cent, all European

availability of high-speed broadband Internet access and relevant content are reflected in a higher proportion of Internet users in the region

Close to half a billion Europeans were online in 2013, which corresponds to 73 per cent of the

Internet users globally, at 96.5 per cent, followed by three other Nordic countries †Norway Sweden and Denmark †with 95 per cent of the

population using the Internet. Turkey has the lowest proportion of Internet users, at below 50 per cent.

In Romania, too, less than half of the population are online (Chart 3. 11 3. 6 The americas

2012 (see Table 3. 8). The country†s mobile-cellular Chart 3. 12: IDI values compared with the global, regional and developing/developed-country averages

os U ru gu ay St K itt s an d N ev is Co

state telecommunication-sector monopolies in the world International Internet connectivity, measured in bit/s per Internet user, is ample in the United

States and Canada, and Brazil26 also has a large amount of bandwidth. Brazil is connected within the region and across the Atlantic ocean

through a multitude of submarine cables Colombia managed to quadruple its amount of international Internet bandwidth, from around

20â 000 bit/s per Internet user in 2012 to close to 80â 000 bit/s in 2013

The americas region has a relatively high household ICT penetration. By end 2013, on Table 3. 8:

average, 55 per cent of households had Internet which is the second highest regional average

Internet: in Argentina, 54 per cent of households have Internet access, as do 53 per cent of

computer. Household Internet access remains very low in the LCC Cuba (3 per cent), as well as in Guatemala and Nicaragua, where around 9 per

partnerships (PPP), to extend and upgrade networks, in particular in rural areas, and by negotiating agreements with telecom operators

to offer discounted price plans (Galperin, 2012 Through national broadband plans, governments in The americas are recognizing the potential

Internet by end 2013, respectively. Brazil and Colombia made good progress at a somewhat higher level of household penetration, reaching

with Internet by end 2013, respectively (see Chartâ 3. 13 Wireless-broadband networks are being

LTE licences or further extended 3g coverage in 2013, spurring growth in the mobile sector

The United states has the highest wireless -broadband penetration, at 94 per cent by end Chart 3. 13:

ITU World Telecommunication/ICT Indicators database 0 10 20 30 40 50 60 70 80

It was an early adopter of LTE technology and coverage was extended massively throughout the country in 2013.

Verizon had achieved 97 per cent LTE population coverage, and the majority of all data traffic is

carried by the LTE network. 27 Very high increases were reported by Brazil, where 40 million new

wireless-broadband subscriptions were added within a year, resulting in a penetration of 52 per cent by end 2013.

LTE services were first launched in the country in early 2013.28 Antigua and Barbuda (from 23 per cent to 49 per cent

show very good progress in terms of wireless -broadband penetration from 2012 to 2013. While the majority of countries in The americas region

their wireless-broadband networks, services were still not available in Cuba, Dominica, Guyana and St vincent and the Grenadines by end 2013

display relatively high levels of more than 20 per cent penetration by end 2013. Uruguay registered the highest absolute increase from

adopters of mobile-broadband technology such as Dominica, Grenada and St vincent and the Grenadines, have fixed significantly higher

Close to 1 billion people are using the Internet in The americas region. While the highest proportion of individuals using the Internet is to

be found in the United states and Canada, more than half of the population is online in countries

http://www. itu. int/ITU-D/ict/definitions/regions/index. html 2 The standard deviation (Stdev) shows the average distance of a value to the mean.

3 http://www. cvmovel. cv/nacional-gsm-3g-edge-e-gprs 4 http://www. telegeography. com/products/commsupdate/articles/2014/04/24/mtn-reports-210-1m-subscribers

http://www. telegeography. com/products/commsupdate/articles/2012/10/22/airtel-widens-3-5g-footprint /5 http://www. gsmamobileeconomyafrica. com/Sub-Saharan%20africa me report english 2013. pdf

6 http://www. epegcable. com /7 http://www. gbiinc. com/Sitepages/News details. aspx? itmid=110

8 http://www. omantel. om/Omanweblib/Individual/Internet/pc initiative. aspx? linkid=3&menuid=420 and http://www. ita. gov. om/ITAPORTAL/Pages/Page. aspx?

10 http://www. telecompaper. com/news/algeria-awards-3g-licences-to-all-three-bidders--972965

11 Internet user data from Gulf countries are not comparable, as they refer to different populations.

Data from Bahrain and Qatar refer to the overall population, i e. including expatriate/transient workers.

Data from United arab emirates are estimated by ITU based on base data excluding the transient worker population 12 Reported in activated external capacity

13 http://www. ofca. gov. hk/en/industry focus/telecommunications/facility based/infrastructures/submarine cables/index html 14 http://submarinenetworks. com/systems/intra-asia/sjc/sjc-cable-system

15 http://www. submarinecablemap. com/#/submarine-cable/boracay-palawan-submarine-cable-system 16 http://web. nso. go. th/en/survey/data survey/560619 2012 information.

-pdf 17 http://www. digitimes. com/news/a20131227pd215. html and http://www. eurobiz. com. cn/chinas-broadband-strategy

/18 http://file. eu-chinapdsf. org/Internet/PUB/Activity4/Results%203/Broadband%20china%20introduction yu%20xiaohui. pdf

19 Belarus, Moldova, Russian Federation and Ukraine are considered as developed countries according to the UN classification

20 See ITU Interactive Transmission Map: http://www. itu. int/en/ITU-D/Technology/Pages/Interactivetransmissionmaps. aspx

21 http://www. mtsgsm. com/upload/contents/328/2008 07 25 3g strategy. pdf 22 http://www. mtsgsm. com/news/2014-01-23-61993/and http://english. corp. megafon. ru/news/20140425-1712. html

23 http://europa. eu/rapid/press-release ip-14-314 en. htm 24 http://ec. europa. eu/digital-agenda/en/digital-agenda-scoreboard

25 http://www. telegeography. com/products/commsupdate/articles/2013/03/01/the-eagle-has landed-incumbent-swoops

-into-3g-sector /26 Data reported by the country refer to 2012 27 http://www. verizonwireless. com/wcms/consumer/4g lte. html and http://www. telecompaper. com/news/verizon-wireless

-lte-reaches-500-markets--952458 28 http://www. rcrwireless. com/article/20121214/carriers/claro-first-launch-lte-services-brazil

/Endnotes 107 Measuring the Information Society Report 2014 Chapter 4. ICT prices and the role of

competition 4. 1 Introduction The price of ICT services constitutes a determining factor for ICT uptake and, as such, continues to be a

focus of attention for regulators and policy-makers Affordability remains the main barrier to Internet access at home in many developing countries.

In Brazil, for instance, 44 per cent of all households with a computer did not have Internet in 2013

because they considered it too expensive or beyond their means (CETIC. BR, 2013. In developed

countries, although not having Internet at home may be more attributable to other factors, such as lack of interest, cost still represents a barrier for

in five households without Internet cite cost as the reason, and seven out of ten of those who have

Internet state that price is the most important factor when choosing the service (European commission 2013;

national mobile phone calls because of concerns about cost (European commission, 2014 In response to the demand for global benchmarks

data following a harmonized methodology since 2008. Initially, prices were collected for fixed -telephone, mobile-cellular (voice and SMS) and

fixed-broadband services. Since 2012, the data collection has been extended to include mobile -broadband prices. These data have proved to be

useful for the international comparison of ICT prices across more than 160 countries, and for identifying those cases where prices constitute a

barrier to ICT uptake This year†s analysis of ICT prices goes beyond simply measuring affordability, and provides

apparent in the mobile-cellular market, where low prepaid prices became a key enabler for the

mass uptake of mobile-cellular services observed in the developing world in the last decade;

they also apply to other telecommunication markets. Regulation sets the framework for competition, and is thus the lever which policy

upon which telecommunication administrations may exert more direct control. They therefore merit particular attention This chapter will present

They include end-2013 data for each of the three price sets contained in the IPB (fixed-telephone

mobile-cellular and fixed-broadband services as well as the general IPB ranking combining the three sub-baskets expressed in terms of GNI per

collection of mobile-broadband prices can be found in Annex 2 4. 2 Fixed-telephone and

mobile-cellular prices Traditional voice services4 and SMS have become the most ubiquitous ICT services, overtaking

historical ICTS such as radio and television broadcasting in many countries. For instance only 9 per cent of households had a telephone

in India in 2001, compared with 32 per cent of households with a TV and 35 per cent of

109 Measuring the Information Society Report 2014 households with a radio. By 2011, however the percentages were reversed:

households had a telephone, 47 per cent a TV and 20 per cent a radio. 5

Despite the decline in fixed-telephone subscriptions over the last decade, fixed telephony remains the most widespread

telecommunication networks. Global fixed -telephone penetration stood at 16 per cent by end 2013, compared with 9 per cent fixed

wired)- broadband penetration. The coverage gaps in the fixed-telephone network have been filled by the mobile-cellular network, which

covers 93 per cent of the global population This compares with a global 3g coverage of

around 50 per cent by end 2012, highlighting that mobile-broadband services are likewise less available than mobile voice and SMS (see

Chapter 1). Subscription figures confirm this despite double-digit mobile-broadband growth rates, there are three times as many mobile

-cellular voice subscriptions as mobile-broadband subscriptions, with almost as many mobile -cellular subscriptions as people on earth

Traffic figures show that, in almost all countries the bulk of national voice traffic corresponds

to calls made from mobile networks, thus confirming the shift from fixed to mobile voice International voice traffic is also predominantly

mobile in most countries, although the number of international fixed-telephone minutes still exceeds international mobile voice minutes

in one in four countries. This situation occurs more frequently in the developed world: there is

more fixed than mobile international telephone traffic in 36 per cent of the developed countries

as against only 18 per cent of the developing countries. These differences are consistent with

the higher fixed-telephone penetration rates recorded in developed countries †almost four times higher than in developing countries

differences in mobile-cellular penetration are smaller. These findings highlight that fixed telephony is used still more than mobile

telephony in some countries for making international calls Such a ubiquitous uptake of voice services would

Chart 4. 1 shows the evolution of fixed-telephone and mobile-cellular prices in the period 2008

a prepaid low-user mobile-cellular subscription costs on average PPP$ 23.7 (or USD Chart 4. 1:

Fixed-telephone basket (left) and mobile-cellular basket (right), in PPP$, world and by level of

Based on 140 economies for which 2008-2013 data on fixed-telephone and mobile-cellular prices were available

or USD 19.5) per month for a prepaid mobile -broadband service with a 500 MB monthly data

allowance. 7 Despite the limitations of comparing such different services, the results roughly confirm that fixed-telephone and mobile-cellular

prices are the cheapest among ICT services suggesting that low prices have contributed to the widespread adoption of traditional voice and

Fixed-telephone prices have followed an almost flat evolution, with a small decrease in prices observed during the period in developing

The fixed-telephone market is the most mature segment of those included in the ITU price data collection exercise

Growth rates have been stagnating since 2008 and there have been few structural changes in the market: 78 per cent of the countries with

price data had already fully or partially liberalized their fixed-telephone market in 2008, compared

with 88 per cent in 2013. Moreover, in some cases liberalization has signalled the end of

the deregulation of retail fixed-telephone prices The cheapest fixed-telephone prices are found in

countries where there is still strong government control over the main fixed-telephone operator such as the Islamic Republic of Iran, Cuba and

cheapest prepaid mobile-cellular prices are in the Asia and the Pacific region, with Sri lanka (USD

with the lowest prepaid mobile-cellular prices in the world. These are examples of the levels

users (who can afford the service) and extend operators†customer bases, even if at lower margins for some segments. 8

user in each country, prices are presented also as a percentage of GNI p. c.,so as to provide an

insight into the affordability of fixed-telephone and mobile-cellular services from a demand-side

Fixed-telephone basket (left) and mobile-cellular basket (right), as a percentage of GNI p. c

Based on 140 economies for which 2008-2013 data on fixed-telephone and mobile-cellular prices were available

the large dispersion of fixed-telephone prices in the developing world: affordability ranges from less than 0. 2 per cent of GNI p. c. in the countries

were applied to fixed-telephone prices, there would be 35 developing countries not meeting the target in 2013, most of them from Africa.

By end 2013, a low-usage prepaid mobile-cellular service cost on average 1. 6 per cent of GNI p. c

a low-user basket in 2013), they still represented 5. 0 per cent of GNI p. c.,on account of the

Fixed-telephone sub-basket, 2013 Note:**Data correspond to the GNI per capita (Atlas method) in 2013

or latest available year adjusted with the international inflation rates.****Country not ranked because data on GNI p. c. are not available for the last five years

Source: ITU. GNI p. c. and PPP$ values are based on World bank data Rank Economy Fixed-telephone sub-basket GNI p. c

USD 2013*Rank Economy Fixed-telephone sub-basket GNI p. c USD 2013 *as%of

GNI p. c. USD PPP$ as%of GNI p. c. USD PPP$ 1 Iran (I. R.)0. 03 0. 12 0. 26 5†780 85 Romania 1. 57 11.88 16.27 9†060

2 Cuba 0. 05 0. 24-6†014 86 Equatorial guinea 1. 60 19.13-14†320

*Data correspond to the GNI per capita (Atlas method) in 2013 or latest available year adjusted with the international inflation rates.**

ranked because data on GNI p. c. are not available for the last five years

GNI p. c. and PPP$ values are based on World bank data Rank Economy Mobile-cellular sub-basket GNI p. c

uptake of mobile-cellular services in these countries, and therefore requires regulatory and policy attention

reliable Internet services. Despite the growth of mobile-broadband subscriptions, less than 3 per cent of global IP traffic corresponded to mobile

high-volume Internet applications such as file sharing (less than 1 per cent of total file -sharing traffic was transmitted through mobile

networks in 2013) and Internet video (2 per cent of total Internet video traffic was transmitted

through mobile networks in 2013. Therefore some of the potential benefits of broadband as a development enabler, such as for instance

its use in education (see Featured Insight 10 in Broadband Commission, 2013a), depend on fixed-broadband uptake in the near future.

In addition, CISCO estimates that 45 per cent of total mobile data traffic was offloaded onto fixed

networks in 2013 (CISCO, 2014), highlighting the role that fixed broadband plays in supporting the

growth in mobile-broadband networks In the period 2008-2013, the price of an entry

which 2008-2013 data on fixed-broadband prices were available Source: ITU 0 50 100

upgrade of entry-level fixed-broadband speeds in developing countries in 2013, with 1 Mbit/s

users contracting higher speeds and/or fixed broadband bundled with other services. This is in line with the findings on bundle adoption from

and data on fixed (wired)- broadband subscriptions by speed, which show that a significant share

Based on 143 economies for which 2008-2013 data on fixed-broadband prices were available Source:

Based on 165 economies for which 2013 data on fixed-broadband prices were available Source: ITU

GNI p. c. values are based on World bank data 0. 0 0. 5 1. 0 1. 5

tv ia Sl ov ak ia Bu lg ar ia Bo sn ia a nd

GNI p. c. values are based on World bank data 0 2 4 6 8 10 12

Rostelecom, Mobile Telesystems OJSC (MTS and ER-Telecom. The national fixed-broadband market in the Russian Federation is thus one of

the most de-concentrated in the world All CIS countries included in the analysis have

Mbit/s of international Internet bandwidth to share among more than 300 000 fixed (wired -broadband subscriptions in 2013.

of international Internet bandwidth is further confirmed by the fact that the entry-level plan

by Kyrgyz Telecom. These factors suggest that regulatory measures to promote competition and ease the international connectivity

4. 7). Tunisie Telecom offers regular ongoing promotions for ADSL services with some of the most advantageous prices in the region

an Internet service at speeds above 512 kbit/s The relatively low fixed (wired)- broadband

GNI p. c. values are based on World bank data for fixed-broadband uptake in the country.

Mauritel largely dominating it. 18 Moreover, international Internet bandwidth is limited very in the country: 620 Mbit/s in 2013

try to access the international Internet at the same time, they will have speed on average a below 256 kbit/s i e. narrowband

for low-income households or the promotion of public Internet access centres (based on either commercial or public schemes

GNI p. c. values are based on World bank data 0 5 10 15 20 A s

os Pe ru Ec ua do r D om in ic an R ep G

plans offered by the state-owned telecom operator ANTEL (ITU, 2013a) †Uruguay could aspire to reaching the fixed (wired)- broadband

dial-up (narrowband) Internet remains the de facto technology for Internet access by residential customers in the island. 19

GNI p. c. values are based on World bank data There are 13 countries in the Asia and the Pacific

Internet bandwidth. Indeed, the latest data on international connectivity show that this may remain an issue in Kiribati (45 Mbit/s), Marshall

Islands (45 Mbit/s), Micronesia (45 Mbit/s Samoa (135 Mbit/s), Solomon islands (216 Mbit/s), Timor-Leste (178 Mbit/s) and Vanuatu

regional Internet exchange point and sharing the cost of building a high-capacity international link from there

GNI p. c. values are based on World bank data However, a comparison with other regions shows

*Data correspond to the GNI per capita (Atlas method) in 2013 or latest available year adjusted with the international inflation rates.**

ranked because data on GNI p. c. are not available for the last five years

GNI p. c. and PPP$ values are based on World bank data Rank Economy Fixed-broadband sub-basket

telecommunication market, the only one displaying sustained double-digit growth rates since 2008 (Chapter 1). According to ITU

population are covered by a 3g network, and this figure will grow as more and more mobile -broadband networks are deployed, until

eventually 3g coverage approaches mobile -cellular coverage (93 per cent. As 3g networks become ubiquitous and therefore accessible to

most of the population, affordability will be one of the most decisive factors for actual uptake of

The dynamism of the mobile-broadband market is reflected also in prices. Unlike the fixed -broadband market, where price structures are

the main types of mobile-broadband plans for which ITU collects data on prices (Figure 4. 1

operators are adding plans for specific devices such as tablets, Mifi dongles, Blackberry, etc Some operators, like for instance Rogers in

Canada and Verizon in the United states, are allowing customers to pool the data consumed by different devices in a single subscription. 23

In addition, handset-based mobile-broadband Figure 4. 1: Mobile-broadband services by type of device/plan

Source: ITU Mobile broadband Handset-based Computer-based Prepaid Postpaid Prepaid Postpaid Chapter 4. ICT prices and the role of competition

126 plans are bundled often with mobile voice and SMS. This makes it difficult to isolate the prices

buy a package including Internet, voice and SMS than to contract only Internet Fixed-broadband and mobile-broadband

Internet prices follow different pricing structures and therefore the analysis of mobile-broadband prices cannot be based on the same parameters

used for fixed broadband. In the case of fixed Internet access, the progress from narrowband dial-up to broadband brought not only a

change in speed but also in pricing. Dial-up was priced on the basis of usage (usually billed per

whereas fixed-broadband Internet usually follows a flat-rate arrangement, whereby the customer pays a monthly fee and has

unlimited access to the Internet at a given speed with neither time nor data volume constraints

This is the common scheme in a vast majority of countries, where fixed-broadband plans are

include data volume caps, e g. USD 10 for 50 MB per month. Several operators also offer

of time and data volume limitations, e g. USD 5 for one day of use with a maximum of 50 MB

spectrum is allocated for mobile broadband for instance, part of the digital dividend) and mobile-broadband networks are upgraded to

advanced technologies (such as LTE-Advanced and Wirelessman-Advanced) that allow more efficient use of spectrum

Currently, most mobile-broadband plans are priced on the basis of the data allowance (i e. the

data volume in MB included in the plan) and not the speed. Many operators do not even advertise

the speed of the mobile-broadband service, but confine themselves to a generic mention of the

technology deployed (which provides only an indication of the speed, since the definition of †3. 5g†or †4g†may vary across operators.

cost) for mobile-broadband services based on high-speed networks. These plans are often labelled as †4gâ€

Chart 4. 11 shows that mobile-broadband plans are becoming more and more available particularly in developing countries, where

the mobile-broadband service available in the most countries is based prepaid handset, which was offered in 153 countries in 2013.

modalities of mobile-broadband services are offered In view of the dynamics of the mobile -broadband market, this section will focus on

127 Measuring the Information Society Report 2014 Chart 4. 11: Availability of mobile-broadband services by

type of service, by level of development, 2013 and 2012 Note: A mobile-broadband service is counted as available

if it was advertised on the website of the dominant operator or prices were provided to ITU through the

ICT Price Basket Questionnaire. 25 Source: ITU analysis of the 2013 prices, without comparing them with the 2012 figures.

in data allowances, bundled voice minutes and SMS, time limitations, premium speeds, etc rather than actual differences in prices for the

same mobile-broadband service The global average price for a computer-based mobile-broadband service with 1 GB monthly

data allowance was PPP$ 36.6 (or USD 24.4) for prepaid plans and PPP$ 30.0 (or USD 19.2) for

postpaid plans in 2013 (Chart 4. 12. The price difference between postpaid and prepaid plans

is also found in respect of regular mobile-cellular services, because operators will usually offer

computer-based mobile-broadband plans cost 37 per cent less than the corresponding prepaid plans in PPP terms.

prepaid and postpaid computer-based mobile -broadband plans are marked less in developing countries, suggesting that operators differentiate

-broadband service with 500 MB monthly data allowance was PPP$ 25.3 (or USD 16.9) for

compared with computer-based plans because the monthly data allowance was half as large Nevertheless, the reduction in price was not

proportional to the reduction in the data allowance, confirming that the price per GB is

lower for larger data allowances, the equivalent of a volume discount. Unlike in the case of

computer-based mobile-broadband services the prices for postpaid and prepaid handset -based mobile-broadband plans were similar

which means that operators are in most cases offering the same rates to postpaid and prepaid

smartphone customers A feature of postpaid handset-based mobile -broadband plans is that they are in some

postpaid handset-based Internet plans included free minutes and SMS in 2013. It is much less

computer-based 1gb Prepaid computer-based 1gb developing developed 160 2012 Postpaid 2013 2012 2013 2012 2013 2012

Chart 4. 12: Mobile-broadband prices, in PPP$, world and by level of development, 2013

Based on 119 economies for which data on mobile -broadband prices were available for the four types of plans

Postpaid computer -based (1gb Prepaid computer -based (1gb PPP$ Developing Developed World Chapter 4. ICT prices and the role of competition

128 common in developing countries, where fewer than one in ten countries have bundled offers as

levels of bundling in mobile-broadband plans makes it difficult to compare prices on a like-for

to streamline their mobile-broadband services and offer cheaper prices. Indeed, operating costs should be lower in developing countries and, if

This is the case, for instance, for fixed-telephone and mobile-cellular services. 26 The fact that this

These differences in mobile-broadband prices between developed and developing countries are even more apparent when looking at the

data allowance of 500 MB are about eight times more affordable in developed countries than in developing countries, on average (Chart

Computer-based services with a monthly allowance of 1 GB are about six times more affordable in developed countries, on average

price for a computer-based mobile-broadband service with 1 GB monthly data allowance corresponded to more than 20 per cent of GNI

p. c. in Africa, and around 15 per cent of GNI p. c in the case of handset-based plans with 500 MB

Based on 119 economies for which data on mobile -broadband prices were available for the four types of plans

Postpaid computer -based (1gb Prepaid computer -based (1gb USD Developing Developed World Chart 4. 14:

Mobile-broadband prices as a percentage of GNI p. c.,world and by level of development, 2013

Based on 119 economies for which data on mobile -broadband prices were available for the four types of plans

Postpaid computer -based (1gb Prepaid computer -based (1gb %GNI p. c 129 Measuring the Information Society Report 2014

in the region, particularly when compared to the lower prices in other regions. Indeed, it is

estimated that mobile-broadband penetration will reach 19 per cent in Africa by end 2014

much less than 500 MB of Internet data per month, supported by the fact that several African

plans allow only limited use of the Internet and therefore restrict the benefits that can be

For instance, Internet video cannot be consumed on the basis of such limited data allowances, and even Internet radio

would need to be limited This suggests that, if mobile broadband is to bridge the broadband gap between Africa and

the other regions, mobile-broadband services will have to become more affordable in Africa so that most applications enabled by a broadband

connection are within the means of a majority of the population Europe stands out as having the most affordable

slightly above that value in the case of computer -Chart 4. 15: Mobile-broadband prices as a percentage of

Based on 119 economies for which data on mobile -broadband prices were available for the four types of plans

Postpaid computer -based (1gb Prepaid computer -based (1gb %GNI p. c Africa Asia & Pacific

The americas Arab States CIS Europe based mobile-broadband services. The americas region also has average prices corresponding to

computer-based plans being the only ones clearly above that threshold. Country data for The americas reveal that there are a number of

countries which have high prepaid computer -based prices because minimum packages include a monthly data allowance much larger

than 1 GB (Table 4. 8). This is the case of Chile USD 73 for 14 GB), Antigua and barbuda (USD

63 for 10 GB), Haiti (USD 23 for 10 GB), El salvador USD 28 for 8 GB) and Argentina (USD 46 for

Such high monthly data allowances for prepaid mobile-broadband dongles suggest that these services target high-end customers

rather than the average user. Postpaid mobile -broadband dongles include much lower monthly data allowances in The americas (Table 3. 7

suggesting that postpaid rather than prepaid is the base offer for regular computer-based mobile-broadband customers

Average prices for computer-based mobile -broadband plans with a monthly data allowance of 1 GB suggest that mobile broadband could be

a cheaper alternative to fixed broadband in many Chapter 4. ICT prices and the role of competition

130 Chart 4. 16: Comparison of postpaid fixed-broadband and postpaid computer-based mobile-broadband

prices, in USD, by region, 2013 Note: Percentages are calculated on the basis of the total number of countries with data available in each region:

27 countries in Africa, 14 countries in the Arab States, 29 economies in Asia and the Pacific, 10 countries in the CIS, 41 countries in Europe and 27 countries in The americas

Source: ITU P e rc e n ta g e o f co u n

tr ie s Africa Cheaper mobile broadband Cheaper fixed broadband Almost no difference 0 10

20 30 40 50 60 70 <5 5-10 10-20 20-30>30 Difference in prices (USD

P e rc e n ta g e o f co u n tr ie

Cheaper mobile broadband Cheaper fixed broadband Almost no difference 0 10 20 30 40 50

Cheaper mobile broadband Cheaper fixed broadband Almost no difference 0 10 20 30 40 50

Cheaper mobile broadband Cheaper fixed broadband Almost no difference 0 10 20 30 40 50

Cheaper mobile broadband Cheaper fixed broadband Almost no difference 0 10 20 30 40 50

Cheaper mobile broadband Cheaper fixed broadband Almost no difference 0 10 20 30 40 50

fixed-broadband and postpaid computer-based mobile-broadband plans with a 1 GB monthly allowance.

This latter mobile-broadband plan is chosen because it is the best match for fixed-broadband services:

are based computer and include a monthly data allowance of at least 1 GB. Although the minimum data allowance is the same, in practice

most fixed-broadband plans allow unlimited data use (Table 4. 4), whereas most computer -based mobile-broadband plans with a minimum

monthly data allowance of 1 GB really do have a cap of 1 GB (Tables 4. 7 and 4. 8

In almost half of the African countries included in the price benchmark, mobile-broadband prices were more than USD 10 cheaper per

month than fixed-broadband prices. Taking into account the GNI p. c. levels in Africa

for a computer-based mobile-broadband plan In other African countries, mobile broadband may be a more affordable alternative to fixed

per month †but mobile-broadband prices still correspond to more than 5 per cent of GNI p. c

as many countries where mobile broadband is fixed cheaper than broadband as vice versa There are some exceptions, such as Tajikistan

region for 1 GB postpaid computer-based plans In many countries in Asia and the Pacific, there

where mobile broadband is significantly cheaper, Indonesia and Thailand are the only ones in which the 5 per cent affordability target

affordable mobile-broadband plans. In the Solomon islands, Timor-Leste and Vanuatu despite mobile broadband being more than USD

20 cheaper per month than entry-level fixed broadband, mobile-broadband prices are still high.

Pacific, such as the lack of international Internet bandwidth, also constrain mobile-broadband services There are four countries in The americas that

of cheaper mobile-broadband prices: Belize El salvador, Paraguay and Suriname. In these countries, mobile broadband is an affordable

However, the mobile-broadband market is still in its early stages, with penetration rates below

extent to which Internet users turn to mobile broadband as an affordable alternative to fixed broadband will only be seen in the coming years

*Monthly data allowance (MB as%of GNI p. c. USD PPP$ 1 Austria 0. 13 5. 31 4. 62 48'590 1'024

*Monthly data allowance (MB as%of GNI p. c. USD PPP$ 80 Maldives 2. 08 9. 7 12.6 5'600 700

*Data correspond to the GNI per capita (Atlas method) in 2013 or latest available year adjusted with the international inflation rates.**

ranked because data on GNI p. c. are not available for the last five years

GNI p. c. and PPP$ values are based on World bank data Chapter 4. ICT prices and the role of competition

*Monthly data allowance (MB as%of GNI p. c. USD PPP$ 1 Norway 0. 1 8. 34 5 102'610 500

*Data correspond to the GNI per capita (Atlas method) in 2013 or latest available year adjusted with the international inflation rates.**

ranked because data on GNI p. c. are not available for the last five years

GNI p. c. and PPP$ values are based on World bank data Rank Economy Mobile-broadband prepaid handset-based (500 MB

*Monthly data allowance (MB as%of GNI p. c. USD PPP$ 80 Antigua & Barbuda 2. 58 27.78 33.8 12'910 1'024

Mobile-broadband prices, postpaid computer-based 1 GB, 2013 Rank Economy Mobile-broadband postpaid computer-based (1 GB

GNI p. c USD 2013 *Monthly data allowance (MB as%of GNI p. c. USD PPP$

1 Austria 0. 13 5. 31 4. 62 48†590 1 2 Norway 0. 20 16.85 10.10 102†610 1

Mobile-broadband prices, postpaid computer-based 1 GB, 2013 (continued Note:**Data correspond to the GNI per capita (Atlas method) in 2013

or latest available year adjusted with the international inflation rates.****Country not ranked because data on GNI p. c. are not available for the last five years

Source: ITU. GNI p. c. and PPP$ values are based on World bank data Rank Economy Mobile-broadband

postpaid computer-based (1 GB GNI p. c USD 2013 *Monthly data allowance (MB as%of

GNI p. c. USD PPP$ 76 Montenegro 2. 89 17.49 28.93 7†260 3 77 China 2. 95 16.14 26.60 6†560 2

78 Seychelles 2. 98 31.10 44.01 12†530 1 79 Georgia 3. 03 9. 02 18.72 3†570 1

Mobile-broadband prices, prepaid computer-based 1 GB, 2013 Rank Economy Mobile-broadband prepaid computer-based (1 GB

GNI p. c USD 2013 *Monthly data allowance (MB as%of GNI p. c. USD PPP$

1 Austria 0. 13 5. 31 4. 62 48†590 1 2 Liechtenstein 0. 18 21.58-142†885 1

Mobile-broadband prices, prepaid computer-based 1 GB, 2013 (continued Note:**Data correspond to the GNI per capita (Atlas method) in 2013

or latest available year adjusted with the international inflation rates.****Country not ranked because data on GNI p. c. are not available for the last five years

Source: ITU. GNI p. c. and PPP$ values are based on World bank data Rank Economy Mobile-broadband

prepaid computer-based (1 GB GNI p. c USD 2013 *Monthly data allowance (MB as%of

GNI p. c. USD PPP$ 76 South africa 4. 82 28.90 51.17 7†190 1 77 Suriname 4. 91 37.88 64.17 9†260 5

78 Namibia 5. 09 24.75 42.88 5†840 1 79 Paraguay 5. 52 18.59 33.70 4†040 1

the region mobile broadband is more than USD 10 cheaper per month. This is a striking finding

This reflects the early launch of 3g services in Europe28 and the maturity achieved in the mobile

-broadband market, with a mobile-broadband penetration of 57 per cent by end 2013, the

highest of all regions. European countries dominate the global top ten of most affordable mobile-broadband plans, with Austria, Finland and

to making available more data on household and individual economic welfare, as well as its distribution. This section presents a refined

data on the distribution of household income or consumption expenditure are available The objective is to explore how factors such as

Data on household income, on the other hand measure only people†s economic welfare, and do not include the business sector.

Data are collected by national statistical offices by means of household income and expenditure surveys HIES) or household surveys including a module

classify income data by deciles, individuals are placed in ascending order according to the household income attributed to them, and then

availability of products, may be relevant. 34 Data on household consumption expenditure are also obtained from household surveys

international comparisons based on data on household economic welfare, because there are several methodological issues that limit

-level economic data provide information (not available from macroeconomic indicators) on the actual income and expenditure capacity

previously mentioned, these data can be used to obtain a finer-grain indication of the affordability

Data for the United states and Sweden are sourced from the OECD Database on Income Distribution

Data for South africa and Viet Nam are sourced from the World Bank†s Povcalnet and refer to 2008

4. 18 uses data on income inequality to reveal differences in the affordability of fixed-broadband

Household disposable income based on World Bank†s Povcalnet data adjusted with ITU estimates on average persons per household

inequality within countries, economic data at the household level also make it possible to determine more precisely the affordability

Household consumption based on World Bank†s Povcalnet data adjusted with ITU estimates on average persons per household

and expenditure data, the following conclusions can be drawn •In more than 85 per cent of countries

for which data are available, the richest 20 per cent of the population can afford

data are available, a basic fixed -broadband subscription still represents more than 5 per cent of household

countries for which data on household income or expenditure distribution are available, fixed-broadband plans remain

Data on household disposable income refer to 2011 or latest year available.**†Lowest 20%†refers to the price divided by the average income of the first

Household disposable income for OECD countries and the Russian Federation based on data from the OECD Database on Income Distribution

on World Bank†s Povcalnet data adjusted with ITU estimates on average persons per household Fixed-broadband prices as a%of household disposable income%households

Data on household consumption expenditure refer to 2011 or latest year available.**†Lowest 20%†refers to the price divided by the average expenditure

Household consumption expenditure based on World Bank†s Povcalnet data adjusted with ITU estimates on average persons per household

limited by lack of data on the distribution of household income/expenditure Available data suggest that basic fixed

-broadband plans are affordable for 90 per cent of the population in Jordan and Tunisia, whereas they are less affordable in

the affordability of mobile-broadband services Table 4. 11 and Table 4. 12 show the price of

with a 500 MB monthly data allowance) as a percentage of disposable household income and

four mobile-broadband services for which ITU collects price data (Chart 4. 13), and is currently the

mobile-broadband service that is available in most countries (Chart 4. 11 147 Measuring the Information Society Report 2014

which data are available. This is also the situation in Armenia, Dominican republic and Egypt In these countries, a prepaid handset-based

expenditure, suggesting that mobile-broadband affordability is an issue irrespective of income /expenditure distribution On the other hand, in developing countries

barrier to mobile-broadband adoption in these countries A comparison of fixed-broadband and prepaid

shows that mobile broadband may be the only affordable alternative for low-income households in several developing countries.

afford a mobile-broadband plan. This might be the case in countries such as Albania, Azerbaijan

subscription data show that, in developed countries, handset-based mobile-broadband subscriptions are individual, rather than shared

Data on household disposable income refer to 2011 or latest year available.**†Lowest 20%†refers to the price divided by the average income of the first

Household disposable income for OECD countries and the Russian Federation based on data from the OECD Database on Income Distribution

data adjusted with ITU estimates on average persons per household Prepaid handset-based mobile-broadband prices

Data on household consumption expenditure refer to 2011 or latest year available.**†Lowest 20%†refers to the price divided by the average expenditure

Household consumption expenditure based on World Bank†s Povcalnet data adjusted with ITU estimates on average persons per household

fewer mobile-broadband subscriptions than households in most African countries and in several developing countries in the Asia and the

developing countries, where mobile broadband may thus be the only alternative for household access In order to take into consideration both

the household has his/her own SIM CARD with a mobile-broadband plan. In this case, affordability

Data show that handset-based mobile -broadband prices (with a 500 MB monthly allowance) are affordable for a majority of the

a mobile-broadband plan would be somewhat unaffordable (i e. represent more than 5 per cent of household consumption expenditure

Data on equivalized household income and consumption expenditure per decile refer 2011 or latest available year

Equivalized household disposable income for OECD countries and the Russian Federation based on data from the OECD Database on Income

Equivalized household disposable income and consumption for other countries based on World Bank†s Povcalnet data adjusted with ITU

Considering the high mobile-cellular penetration in both countries, this suggests they are in a good position to see an increase in

Data for the Arab States are limited to four countries. In Tunisia and Jordan, individual handset-based mobile-broadband subscriptions

two countries mean that individual mobile -broadband plans would be somewhat unaffordable (i e. represent more than 5 per

need to share a mobile-broadband plan with other members of their household. In Egypt

and Sudan, the cost of a mobile-broadband plan corresponds to more than 10 per cent

major barriers for increasing mobile-broadband adoption in many African countries On the basis of the data presented, it can be

concluded that income inequality does not only have an impact on the proportion of households within a country that have access to mobile

and user preferences, but also by affordability 4. 5 The impact of competition and regulation on

telecommunication prices The impact of ICTS as development enablers depends on access to ICT services and the use

a licence to a new entrant in the mobile-cellular market. This section presents a quantitative analysis

prices for mobile-cellular (voice and SMS) and fixed-broadband services. Among all ICT services

of comprehensive data series on the prices for these two services, which makes it possible to study

data for developing countries. Quantitative studies are in several cases restricted to samples of EU and OECD countries, or else

regulatory data for up to 144 countries in the period 2008-2013. Including such a large set

and telecommunication prices, based on a worldwide representative sample, and to check to what extent the quantitative results based

on telecommunication data from EU, OECD and specific countries hold true in a global context

the level of detail of the indicators, since data availability and comparability for such a large

on telecommunication prices (e g. mandating infrastructure sharing or granting a new licence lie beyond the scope of this analysis

The fall in telecommunication prices in the last decade, and in the period analysed in this chapter (2008-2013), is linked to several

telecommunication services and the privatization of incumbent operators. In parallel, national regulators have been created to establish a

level playing field and monitor the liberalized electronic communication markets (ITU, 2013b Regulators have thus become the custodians of

competition in telecommunication services at the country level From the literature on cross-national institutional

industries like telecommunications. 43 A country†s institutional endowment determines the scope for arbitrary administrative discretion, the legal

majority of mobile-cellular and fixed-broadband markets. Regulatory and policy action can also have a direct impact on retail prices,

number of licences issued in mobile-cellular markets or by limiting foreign participation in fixed-line operators.

that affects telecommunication markets. Thus it can contribute to creating legal certainty and a level playing field, which are important

of affordable prices in telecommunication services. Chart 4. 21 shows the evolution of average entry-level fixed-broadband prices

number of subscriptions) of each Internet service provider (ISP. The result ranges from 0 (perfect

-level prices and competition in mobile-cellular markets, where the decline in prices during the

Simple averages for 140 economies with available data on fixed-broadband prices and competition for the period 2008-2013

Herfindahl-Hirschman Index (HHI) data sourced from Informa Chart 4. 22: Competition in mobile markets and mobile

-cellular prices as a percentage of GNI p. c.,2008-2013 Note: Simple averages for 140 economies with available data on mobile-cellular

prices and competition for the period 2008-2013 Source: ITU. Herfindahl-Hirschman Index (HHI) data sourced from Informa

competition and the openness of the market to private and foreign investment The scores of each cluster are combined into a

based on panel data regression. This enables us to go beyond descriptive statistics and draw some robust conclusions on the link between

telecommunication sector •Cluster 3: the regulatory regime in the different areas covered by the regulatory

analyses in this section, data from the Regulatory Tracker have been extracted for clusters 1, 2 and 3, and combined into a single value per country

econometric modelling using panel regressions for up to 144 countries based on data for the five-year period from 2008 to 2013

Panel data regression is a statistical technique which is used to assess how variations in a set

of variables over a given time period relate to Figure 4. 3: Composition of the variable measuring the regulatory environment

Panel regressions minimize problems of omitted variable bias (the omission of important variables) and multicollinearity

In addition, panel regressions have the advantage of discounting known and unknown region fixed effects. These are structural

beyond telecommunication regulation (e g. the European union acquis. 49 Such background fixed effects may be important for each region

Individual users allowed to use Voip 13. Mobile number partability implemented 12. Fixed number portability implemented

8. Co-location/site sharing mandated 7. Infrastructure sharing mandated 6. Infrastructure sharing for mobile operators allowed

1. Separate telecom/ICT regulator Regulatory authority 11. Tariff info, consumer education & complaints 10. IT

9. Internet content 8. Broadcasting content 7. Broadcasting (radio and TV transmission 6. Universal service/access

5. Spectrum monitoring and enforcement 4. Radio frequency allocation and assignment 3. Interconnection rates and price regulation

2. Licencing 1. Qos measures and monitoring Entity in charge of Regulatory mandate ++=157 Measuring the Information Society Report 2014

competition and regulation metrics, using panel regressions with fixed effects Prices for fixed-broadband and mobile voice

that characterize a telecommunication market and are often the result of the simultaneous effects of technology choices, competition

telecommunication services vary with levels of economic development Therefore, gross national income per capita (GNI p. c is included in the model

•The deployment of telecommunication networks requires large investments that operators evaluate depending on the demand for the service and the specific

GB of data allowance included per month However, entry-level fixed-broadband plans in several countries offer higher

speeds and larger data allowances. In order to measure how these enhanced features affect prices, a variable on fixed

capped data allowances are included as controls in the model for fixed-broadband prices Chapter 4. ICT prices and the role of competition

Panel regression models for fixed-broadband and mobile-cellular prices Two models are used for the regressions:

Data collected by ITU, see Annex 2 for more details on the methodology for the collection of

and represents the price by country and year for a low-user basket of mobile-cellular calls and SMS in current USD.

Data collected by ITU, see Annex 2 for more details on the methodology for the

collection of mobile-cellular prices Descriptive statistics of the dependent variables Average Standard deviation Minimum Maximum 2008 2013 2008 2013 2008 2013 2008 2013

Data sourced from Informa Herfindahl-Hirschman Index for mobile cellular (voice and SMS with Li being the number of mobile-cellular subscriptions of firm i,

and TL he total number of mobile-cellular subscriptions in the country. It is the sum of the squared market shares of each mobile-cellular service provider calculated in terms of subscriptions.

As in the case of the HHI for fixed broadband, this ranges from to 1, where k is the total number of mobile-cellular service providers

in the market. Data sourced from Informa 159 Measuring the Information Society Report 2014 Box 4. 2:

Panel regression models for fixed-broadband and mobile-cellular prices (continued Regulatory variable: The combined values of clusters 1 to 3 of the ITU ICT Regulatory Tracker.

The Regulatory Tracker is an aggregate benchmark of each country†s legal and regulatory frameworks using as a reference internationally recognized regulatory best practices

The following three clusters are used for the analysis Score of cluster 1: Regulatory authority Score of cluster 2:

Data collected by ITU, see www. itu. int/tracker for more information Fixed-broadband speed:

Speed of the entry-level fixed broadband plan in Mbit/s. Data collected by ITU

if the fixed-broadband plan includes a cap in the monthly data allowance, and 0 otherwise.

Data collected by ITU Descriptive statistics of the dependent variables Average Standard deviation Minimum Maximum 2008 2013 2008 2013 2008 2013 2008 2013

HHI mobile cellular 0. 46 0. 41 0. 16 0. 14 0. 17 0. 15 1. 00 1. 00

Descriptive statistics calculated for 124 economies that have complete data for the two models Source: ITU

The panel regression model for fixed-broadband prices has a medium explanatory power (an R-squared value of 0. 41, where 1 represents a

Panel regression results, fixed-broadband prices and regulation Variable Coefficient Statistical significance Interpretation GNI p. c. 0. 217

Fixed-broadband plans with data caps are linked to prices 31 %lower than unlimited plans Fixed-broadband speed 0. 016

telecommunication/ICT regulator that has autonomy in decision-making, enforcement power, the right to impose sanctions or penalties

data caps by fixed-broadband service providers is correlated with cheaper entry-level fixed -broadband plans, other things being equal.

ITU data collection considers a minimum of 1 GB Chapter 4. ICT prices and the role of competition

data cap of 1 GB or more, or allows unlimited traffic. In 2013, entry-level fixed-broadband

plans included a data cap in 35 per cent of the countries included in the fixed-broadband

data caps were enforced in entry-level fixed -broadband plans. This suggests that operators can offer cheaper prices in exchange for reduced

data consumption, thus indicating that capacity in fixed-broadband networks is still an issue in several countries,

additional Internet data beyond 1 GB is still non -negligible in many countries Finally, different entry-level fixed-broadband

automatic upgrades of base speeds once networks are upgraded. 54 Chart 4. 23 provides an approximation of

sector, such as operators†strategies on data caps competition in the fixed-broadband market and the ICT regulatory environment, may together be

Results for mobile cellular The results of the panel regression for mobile -cellular prices (voice and SMS) indicate that the

model constructed has a medium explanatory power (an R-squared value of 0. 41, where 1

a market with two mobile-cellular operators sharing the market equally, the entry of a new

but disruptive operators, such as mobile virtual network operators (MVNOS), may have an impact on prices Competition in mobile-cellular markets is

stronger than in fixed-broadband markets and differences in competition levels across mobile-cellular markets are on average smaller

Panel regression results, mobile-cellular prices and regulation Variable Coefficient Statistical significance Interpretation GNI p. c. 0. 147

HHI mobile cellular 0. 201 0. 082 Significant 5%level A change from a duopoly to a triopoly (with operators holding

the reduction in mobile-cellular prices that could be achieved in those developing countries with highly concentrated markets,

prices in mobile-cellular markets discussed is only valid if considered in combination with the other explanatory variables included in the

prices for mobile-cellular services: a 5 per cent increase in the percentage of the population

variations in mobile-cellular prices observed across countries in 2013. Differences in mobile -cellular prices across countries are smaller than

on the final price in the case of mobile-cellular services. This suggests that economic levels are

user voice and SMS basket) to be affordable for most of the population and, it being a mature

as a driver for lower prices in mobile-cellular markets Regulation is found to have a weak explanatory

regulation is less of an issue in mobile-cellular markets. This may be because the regulation

addition, the deployment of mobile networks tends to be less capital-intensive than the deployment of fixed-broadband networks, and

competition and lower prices in mobile-cellular services, whereas in fixed-broadband markets stronger regulatory action may be needed

Variation in mobile-cellular prices %explained by each variable, 2013 Note: Calculated taking as a reference the average of each variable and adding

difference in mobile-cellular prices that would be obtained keeping all other variables constant. The calculation does not take into consideration the

is the existence of data caps, which is correlated with lower prices. This may indicate that fixed-broadband capacity

Fixed telephone sub -basket as a%of GNI per capita, 2013 Mobile-cellular sub -basket as a%of GNI

Fixed telephone sub -basket as a%of GNI per capita, 2013 Mobile-cellular sub -basket as a%of GNI

*Data correspond to the GNI per capita (Atlas method) in 2013 or latest available year adjusted with the international inflation rates

GNI p. c. and PPP$ values are based on World bank data Chapter 4. ICT prices and the role of competition

attributable to the telecommunication sector, such as operators†strategies on data caps, competition in the fixed

-broadband market and the ICT regulatory environment, are together more of a determinant for fixed-broadband

•Mobile cellular: Differences in mobile -cellular prices across countries are smaller than the differences in fixed-broadband

differences in mobile-cellular prices observed across countries (an estimated 7 per cent), whereas differences in the

an impact in setting mobile-cellular prices, since regulation in most countries is already open enough to allow

information on the PPP methodology and data, see http://icp. worldbank. org 3 GNI takes into account all production in the domestic economy (i e.

http://data. worldbank. org/indicator/NY. GNP. PCAP. CD 4 Voice over internet services, such as Skype or Voipbuster, are excluded from the analysis in this section

because they require an Internet connection and do not have guaranteed a quality of service. They are considered under broadband services

5 Source: Census of India 2011. Mode of communication 2001-2011. Available at http://censusindia. gov. in/2011census/hlo/Data sheet/India/Communication. pdf

6 Prices for each service are calculated on the basis of a low-user basket that defines the number of calls, minutes and SMS

in the case of mobile-cellular plans) that are consumed per month. For more information on the baskets and the price

-collection methodology, see Annex 2 7 Prepaid handset-based mobile-broadband plans were only available in 122 out of the 140 countries included in the

comparison with the other telecommunication services. Therefore, the average for handset-based mobile-broadband prices

8 In the case of Sri lanka, the entry of Bharti Airtel as the fifth mobile network operator in the market led to an aggressive

and a floor rate for national mobile calls with the aim of protecting mobile operators†margins.

At present, Sri lanka continues to have some of the cheapest mobile-cellular prices in the world and the mobile market has achieved a stable financial situation.

For a more detailed analysis of the ICT sector in Sri lanka, see Galpaya (2011) and the presentation of the Telecommunications Regulatory Commission of Sri lanka on the

impact of the floor rate, available at: http://www. itu. int/ITU-D/finance/work-cost-tariffs/events/tariff-seminars/Indonesia-12

/pdf/Session4 srilanka nishantha. pdf 9 Based on 2012 and 2011 ITU data for countries accounting for 97 per cent of global fixed (wired)- broadband subscriptions, it

is estimated that 59 per cent of total fixed (wired)- broadband subscriptions were through ADSL in 2012

http://data. worldbank. org/about/country-classifications/country-and-lending-groups 11 The Communications Commission of Kenya (CCK) issued mobile virtual network operator licences to three operators in April

http://www. cck. go. ke/news/downloads/MVNO. pdf 12 Although four international operators compete in the Kenyan mobile-cellular market, the dominant mobile operator

These data suggest there is limited competition among operators 13 Advanced mobile technologies refer to standards agreed by the ITU Radiocommunication Assembly for next-generation

mobile technologies †IMT-Advanced †such as LTE-Advanced and Wirelessman-Advanced. For more details on these

standards, see http://www. itu. int/net/pressoffice/press releases/2012/02. aspx 14 Cuba is excluded from the world and developing averages of fixed-broadband prices,

16 The most visited websites in Tunisia by December 2011 were predominantly in English. Initiatives to promote Arab digital

18 Mauritel reported 7 352 fixed Internet subscriptions by end 2013,97 per cent of which trhough ADSL (source:

Telecom, http://www. iam. ma/Groupe/Institutionnel/Qui-Sommes-Nous/Filiales participations/Pages/Mauritel. aspx), out of a

/page=internet conectividad&sub=internet 20 For instance, the latest Computer literacy Survey in Sri lanka (2009) found that only 20 per cent of the household

population (aged 5-69) could use a computer on their own (Department of Census and Statistics Sri lanka, 2009.

The Ministry of Education of Sri lanka and esri Lanka have undertaken several actions to improve digital literacy (Galpaya, 2011

TEAMS€ website (http://www. teams. co. ke) and EASSY€ s website (http://www. eassy. org

/UPL389171684192165237 CP ORANGE ACE FR 191212. pdf, and for the go-live of WACS, see http://www. oafrica. com/broadband/west-africa-cable-system-wacs-technically-goes-live

23 For more information on Rogers†â€oeshare Everything†plans, see http://www. rogers. com/web/content/share-everything?

http://www. verizonwireless. com/wcms/consumer/shop/shop-data-plans/more-everything. html 24 The details of the different †4g†plans offered by Tigo can be found on the following websites

http://www. tigo. com. bo/personas/planes-y-promociones/Internet-movil-en-tu-modem http://www. tigo. com. co/4g, http://www. tigo. com. gt/personas/internet-movil/internet-movil-tigo-4g and

https://www. tigo. com. py/contenido/para-navegar-con-el-modem 25 Data for mobile-broadband services have been collected since 2012 through the ITU ICT Price Basket Questionnaire,

which is sent out annually to all ITU Member States/national statistical contacts 26 In 2013, the average price in USD for an entry-level fixed-telephone service was 49 per cent cheaper in developing countries

than in developed countries, and 14 per cent cheaper in PPP$. Likewise, a low-user mobile-cellular basket was 41 per cent

cheaper in USD, and 7 per cent cheaper in PPP$ 27 For more information on MTN Cameroon tariffs, see www. mtncameroon. net,

-broadband plans, see http://www. orange. ci/menu-mobile-3g/pass-internet-3g. html 28 The UMTS auctions took place in 2000 and 2001 in Europe (Van damme, 2002 and OECD,

2001), preceded only by Japan and the Republic of korea, where 3g licences were awarded in 2000. In most developing countries, on the other hand, 3g

licences were granted not until much later. In large emerging countries such as China and India, for instance, 3g licences

were awarded in 2009 and 2010, respectively. See for instance Xia (2011) and India†s Department of Telecommunication press

release: http://www. dot. gov. in/as/Auction%20of%20spectrum%20for3g%20&%20bwa/Auction%20results/3g auction

-Final results. pdf 29 For example, outflows of profits generated by a multinational operating in country A

and transferred back to the country of ownership of the multinational would count in country A†s GDP,

http://data. worldbank. org/about/country-classifications/country-and-lending-groups 31 Household incomes include wages, salaries, self employment incomes, capital and property income, private transfers and

http://www. oecd. org/els/soc/IDD-Tor. pdf 32 Individual income is calculated by dividing total household income by the number of persons

http://www. oecd. org/els/soc/OECD-Note-Equivalencescales. pdf 33 The first decile corresponds to the 10 per cent of the population with the lowest income.

World Bank†s website on measuring poverty: http://go. worldbank. org/W3hl5gd710 35 Differences in the equivalence scales of the source data used in this chapter are corrected roughly using ITU estimates on

the number of inhabitants per household for each country. World bank Povcalnet data on income distribution are published

per capita, and figures from the OECD Database on Income Distribution are equivalized using the square root scale

36 For OECD countries, household income is estimated by multiplying the equivalized household income by the square

According to 2008/9 survey data, mean household consumption in Angola was almost the same as in

38 Data on income distribution are aggregated per household and then equally attributed to each member of the household

In addition, data on income distribution are averaged per decile. If the price of a fixed-broadband plan represents less

In conclusion, the data presented do not determine whether 100 per cent of a population within a country can afford a fixed-broadband plan,

The World Bank†s Povcalnet data are adjusted using ITU estimates on the average number of inhabitants per household

47 Mobile termination rates are regulated in more than 120 countries. Source: ITU Tariff Policies Database 2013 (ICTEYE

http://www. itu. int/net4/ITU-D/icteye 48 For a detailed description of the ITU ICT Regulatory Tracker, see www. itu. int/tracker

market share data were available. This includes 95 economies from the developing world and 44 from the developed world

economies for which price and market share data were available for 2013. This includes 99 economies from the developing

53 The Regulatory Tracker quantifies these aspects of the regulatory framework through the indicators â€oeseparate telecom/ICT

regulatorâ€, â€oeseparate telecom/ICT regulatorâ€, â€oeenforcement powerâ€, â€oesanctions or penalties imposed by regulator†and â€oedispute

55 The mean HHI for mobile cellular is calculated as the simple average of the HHI of 140 economies for which price and market

share data were available. This includes 96 economies from the developing world and 44 from the developed world

dispersion in mobile-cellular prices is of  60 per cent around the mean 173

Chapter 5. The role of big data for ICT monitoring and for development 5. 1 Introduction

-to date and reliable data, in particular from developing countries. The information and communication technologies (ICT) sector is

of those statistics and identify new data sources In this context, the emergence of big data holds

great promise, and there is an opportunity to explore their use in order to complement the existing, but often limited, ICT data

There is no unique definition of the relatively new phenomenon known as big data. At the most basic level it is understood as being data

sets whose volume, velocity or variety is very high compared to the kinds of datasets that

have traditionally been used. The emergence of big data is linked closely to advances in ICTS. In today†s hyper-connected digital world, people

and things leave digital footprints in many different forms and through ever-increasing data flows originating from, among other things

commercial transactions, private and public records that companies and governments collect and store about their clients and citizens, user

-generated online content such as photos, videos tweets and other messages, but also traces left by the Internet of things (Iot), i e. by those

uniquely identifiable objects that can be tracked Big data have great potential to help produce new and insightful information, and there

is a growing debate on how businesses governments and citizens can maximize the benefits of big data.

Although it was the private sector that first used big data to enhance efficiency and increase revenues, the practice has

expanded to the global statistical community The United nations Statistical commission UNSC) and national statistical organizations NSOS) are looking into ways of using big data

sources to complement official statistics and better meet their objectives for providing timely and accurate evidence for policy-making. 1

value added by big data in the context of monitoring of the information society, and Chapter 5. The role of big data for ICT monitoring and for development

174 there is a need to explore its potential as a new data source. While existing data can provide

a relatively accurate picture of the spread of telecommunication networks and services there are significant data gaps

when it comes to understanding the development of the information society. Relatively little information for example, is available on the demand side

While an increasing number of countries currently collect data on the individual use of ICTS, many developing countries do not

produce such information (collected through household surveys or national population and housing censuses) on a regular basis Consequently, not enough data are available

about the types of activity that the Internet is used for, and little is known about the Internet

user in terms of age, gender, educational or income level, and so on In other areas, such as education, health or

public services, even fewer data are available to show developments over time and enable informed policy decisions.

The recently published Final WSIS Targets Review report Partnership, 2014), which attempts to assess developments in the information society

between 2003 and 2013/14, shows that little information is available to track progress over time.

It is obvious that greater efforts must be made to overcome the lack of reliable, timely

society, and that big data have the potential to help realize those efforts In addition to the data produced

and held by telecommunication operators, the broader ICT sector, which includes not just telecommunication companies but also over-the-top (OTT) service

providers such as Google, Twitter, Facebook Whatsapp, Netflix, Amazon and many others captures a wide array of behavioural data

Together, these data sources hold great promise for ICT monitoring, and this chapter will explore

the potential of today†s hyper-connected digital world to expand on existing access and

infrastructure indicators and move towards indicators on use, quality and equality of use At the same time, there is a growing debate

on the role and potential of big data when it comes to providing new insights for broader

social and economic development. Big data are already being leveraged to understand socio -economic well-being, forecast unemployment

and analyse societal ties. Big data from the ICT industry play a particularly important role because they are the only stream of big

data with global socioeconomic coverage. In particular, mobile telephone access is quasi -ubiquitous, and ITU estimates that by the end of

2014 the number of global mobile subscriptions will be approaching 7 billion. At the same time almost 3 billion people †40 per cent of the

world†s population †will be using the Internet In recent years, moreover, the strongest growth

in telecommunication access and use has been recorded in the developing economies, where ICT penetration levels have increased and where

big data hold great promise for development However, while there are a growing number of research collaborations and promising proof

the development of models to protect user privacy while at the same time allowing for the extraction of insights that can improve service

data for development, highlight advances, point to some best practices and identify challenges including in regard to the production and

sharing of big data for development The chapter will first (in Section 5. 2) describe some of the current big data trends and

definitions, highlight the technological developments that have facilitated the emergence of big data, and identify the main

sources and uses of big data, including the use of big data for development and ICT monitoring.

Section 5. 3 will examine the range and type of data that telecommunication companies, in particular mobile-cellular

operators, produce, and how those data are 175 Measuring the Information Society Report 2014 currently being used to track ICT developments

and improve their business. Section 5. 4 looks at the ways in which telecom big

data may be used to complement official ICT statistics and assist in the provision of new evidence for a host of policy domains

while Section 5. 5 discusses the challenges of leveraging big data for ICT monitoring and broader development, including in terms of

standardization and privacy. It will also make some recommendations for mainstreaming and fully exploiting telecom big data for monitoring

and for social and economic development in particular with regard to the different stakeholders involved in the area of big data

from the ICT industry 5. 2 Big data sources, trends and analytics With the origins of the term â€oebig data†being

shared between academic circles, industry and the media, the term itself is amorphous, with no single definition (Ward and Barker, 2013

At the most basic level of understanding, it usually refers to large and complex datasets Table 5. 1:

Sources of big data Sources Some examples Administrative data •Electronic medical records •Insurance records •Tax records

Commercial transactions •Bank transactions (inter-bank as well as personal •Credit card transactions •Supermarket purchases

•Online purchases Sensors and tracking devices •Road and traffic sensors •Climate sensors

•Mobile phones •Satellite/GPS devices Online activities/social media •Online search activities •Online page views

•Blogs and posts and other authored and unauthored online content and social media activities •Audio/images/videos

Source: ITU, adapted from UNSC (2013 and reflects advances in technology that make it possible to capture, store and process

increasing amounts of data from different data sources. Indeed, one of the key trends fostering

the emergence of big data is the massive â€oedatafication†and digitization, including of human activity, into digital â€oebreadcrumbs†or

â€oefootprints†In an increasingly digitized world, big data are generated in digital form from a number of

sources. They include administrative records for example, bank or electronic medical records), commercial transactions between two

example, mobile phones and GPS devices), and activities carried out by users on the Internet including searches and social media content

Table 5. 1 Big data is not just about the volume of the data. One of the earliest definitions, introduced

by the Gartner consultancy firm, describes big data characteristics such as velocity and variety, in addition to volume (Laney, 2001

â€oevelocity†refers to the speed at which data are generated, assessed and analysed, while the

Chapter 5. The role of big data for ICT monitoring and for development 176 term â€oevariety†encompasses the fact that data

can exist as different media (text, audio and video) and come in different formats (structured and unstructured.

The three-Vs definition has caught on and been expanded upon. A fourth V †veracity †was introduced to capture aspects

relating to data quality and provenance, and the uncertainty that may exist in their analysis (IBM

2013). ) A fifth V †value †is included by some to acknowledge the potentially high socio

-economic value that may be generated by big data (Jones, 2012)( Figure 5. 1 Included within the scope of big data is the

category of transaction-generated data (TGD), 2 also sometimes described as â€oedata exhaust†or â€oetrace dataâ€.

These are digital records or traces that have been generated as by-products of doing things (such as processing payments

making a phone call and so on) that leave behind bits of information. The value of this subset of big data is that it is connected directly to human

behaviour and its accuracy is generally high Most of the data captured by telecommunication companies can be classified as TGD

As is often the case with technological innovation, it is the private sector that has been

Figure 5. 1: The five Vs of big data Source: ITU at the forefront of extracting value from this

data deluge. Encouraged by promising results but also reduced budgets, the public sector is turning towards big data to improve its service

delivery and increase operational efficiency. In addition, there are uses for big data in broader development and monitoring, and there is an

increasing focus on big data†s role in producing timely (even real-time) information, as well as new insights that can be used to drive social and

economic well-being Big data uses by the private and pub -lic sectors Marketing professionals, whose constant aim

is to understand their customers, are now increasingly shifting from conventional methods such as surveys, to the extraction of customer

preferences from the analysis of big data Walmart, the world†s biggest retailer, has been one of the largest and earliest users of big data

In 2004, it discovered that the snack food known as Pop Tarts was purchased heavily by United

of data generated through large-scale datafication and digitization of information Different types and forms of

data, including large amounts of unstructured data Level of quality accuracy and uncertainty of data and data

sources VALUE Potential of big data for socio -economic development VELOCITY Speed at which data are

generated and analyzed 177 Measuring the Information Society Report 2014 a specific condition that then led Walmart to

improve its production chain †in this case, by increasing the supply of Pop Tarts to areas likely

to be affected by a disaster. Walmart has also made use of predictive analytics, which uses personal information and purchasing patterns

to extrapolate to a likely future behaviour, and to better target and address customer needs

value of big data Nor is the private sector†s use of big data techniques restricted solely to market research

Companies and whole industries (healthcare energy and utilities, transport, etc. are using such techniques to optimize supply chains and

proficient in their use of data-driven decision making have been found to have productivity levels up to 6 per cent higher than firms making

minimal to no use of data for decision-making Brynjolfsson, Hitt and Kim, 2011. Significantly industries now have the ability to conduct

How big data saves energy †Vestas Wind Systems improves turbine performance Vestas, a global energy company dedicated to wind energy, with

installations in over 70 countries, has used big data platforms to improve the modelling of wind energy production and identify

By using big data techniques based on a large set of factors and an extended set of structured and

unstructured data, Vestas was able to significantly improve customer turbine location models and optimize turbine performance

Big data have enabled the creation of a new information environment and allowed the company to manage and analyse

weather and location data in ways that were previously not possible. These new insights have led to improved decisions

ITU, based on IBM (2012 a speed that are unprecedented. Google for example, is running about a thousand experiments at any given point in time (Varian

2013a). ) Telecom network operators make extensive use of such techniques when rolling out new services, among other things for the

purpose of pricing. Telecom operators also use big data techniques to understand and control churn, optimize their management of customer

relations and manage their network quality and performance These fundamental shifts in data exploitation to generate new socioeconomic value

coupled with the simultaneous emergence of new rich data sources that can potentially be linked together

and analysed with ease have sparked also the interest of governments researchers and development agencies Encouraged by the potential of big data to

produce new insights and slimmer budgets governments (at all levels) are now looking to exploit big data and increase the application of

data analytics to a range of activities, including monitoring and improvement of tax compliance and revenues, crime detection and prediction

and improvement of public service delivery Giles, 2012; Lazer et al. 2009 To this end, governments, in addition to the

data they collect and generate themselves Chapter 5. The role of big data for ICT monitoring and for development

178 complement their official statistics by leveraging data from new sources, including crowd-sourced data generated by the public

In the United states, for example, Boston City Hall released the mobile app â€oestreet Bump†which uses a phone†s accelerometer to detect

potholes while the app user is driving around Boston and notifies City hall. 3 Some of the

richest data sources for enabling governments and development agencies to improve service delivery are actually external.

Such external data include those captured and/or collected by the private sector, as well as the digital

breadcrumbs left behind by citizens as they go about their daily lives According to a recently published White house

report, United states government agencies can make use of public and private databases and bigâ data analytics to improve public

administration, from land management to the administration of benefits. The Department of the Treasury has set up a â€oedo Not Pay†portal

which links various databases and identifies ineligible recipients to avoid wrong payments and reduce waste and fraud4 (The White house

2014 Big data for development and ICT monitoring One of the richest sources of big data is

the data captured by the use of ICTS. This broadly includes data captured directly by telecommunication operators as well

as by Internet companies and by content providers such as Google, Facebook, Twitter etc. Big data from the ICT services industry

are already helping to produce large-scale development insights of relevance to public policy. Collectively, they can provide rich

and potentially real-time insights to a host of policy domains. It should be noted that in

some countries and regions the use of big data, including big data from the ICT industry

is subject to national regulation. In the EU, for example, a number of directives require data

producers to obtain users†consent before gathering any of their personal data. 5 One of the best-known examples of leveraging

the online population†s digital breadcrumbs for development purposes is Google Flu Trends GFT). ) Following its launch in 2008,

GFT was remarkably accurate in tracking the spread of influenza in the United states, doing so more

rapidly than the Centers for Disease Control and Prevention (CDC), with a lag time of only one day

as opposed to one week. Although it has since been subject to criticism (see Section 5. 5), GFT

was held up as an outstanding example of big data in action and of the great potential of big

data for broader development and monitoring Mayer-Schã nberger and Cukier, 2013; Mcafee and Brynjolfsson, 2012.

GFT worked by monitoring health-seeking behaviour expressed through online searches, with the search terms

being correlated wherever they related to flu-like symptoms (Ginsberg et al. 2009). ) This proved to be so successful that it spawned similar efforts

focusing on the use of search-engine data to understand dengue fever outbreaks, 6 monitor prescription drug use (Simmering, Polgreen and

Polgreen, 2014), predict unemployment claims in the United states (Choi and Varian, 2009) and Germany (Askitas and Zimmermann, 2009), and

The Internet has also been a rich source of big data beyond the realm of user search terms

Online job-posting data are being used to supplement traditional labour statistics in the United States7 and other countries.

In another effort, an academic project at MIT known as the Billion Prices Project collects high-frequency price

data from hundreds of online retailers. 8 The data are used then by researchers to understand a

UN Global Pulse, a UN initiative to use big data for sustainable development and humanitarian action, has been mining Twitter data from

Indonesia (where Twitter usage is high) 9 to understand food price crises. Global Pulse was

able to identify a consistent pattern among specific food-related tweets and the daily food

analytics on the Twitter data to forecast the consumer price index several weeks in advance Byrne, 2013. As discussions on the post-2015

Pulse is also using Twitter data to understand and compare the relevance of different development topics among countries (Box 5. 2

How Twitter helps understand key post-2015 development concerns As the process of formulating the post-2015 development

Campaign are using big data and visual analytics to identify the most pressing development topics that people around the world

Users can select a country to see the number of tweets generated by its Twitter users in

regard to the highlighted topic, as compared to tweets about all the other topics. This information provides insight as to which

was phone and Internet access. By clicking on any of the data points in the chart, the application provides

Using Twitter to visualize trends in global development topics In fact, the ICT sector is itself using the Internet

as a source of big data for monitoring purposes Regulators and others are now using the

Internet to crowdsource quality of service Qos) data on broadband quality. For example the United states Federal Communications

Commission (FCC) has released mobile apps that enable consumers to check their broadband quality. The test results, which

are anonymous, are used then by FCC to understand and address coverage and quality issues in different areas. 10

Phone and Internet access Equality between men and women Chapter 5. The role of big data for ICT monitoring and for development

180 Mobile data Despite the rapid growth in Internet access 60 per cent of the world†s population is still

not using the Internet. Household Internet penetration in developing economies is expected to reach 31 per cent by the end

of 2014, as against almost 80 per cent in developed economies. In addition, as Internet penetration rates remain limited, Internet users

are not yet) representative of the population at large. For example, Internet users tend to be younger, relatively well educated, with men still

more likely to be online than women, especially in developing countries11 (ITU, 2013 Depending on the source of Internet data

results may also be biased more or less. A 2013 study into the characteristics and behaviour of Facebook users, for example, revealed

that while in many ways Facebook users have real-life behaviour and characteristics in many ways the social network fails as a

representation of society. On the one hand for example, the American Facebook user†s relationship status of â€oemarried†on Facebook

is very similar to real life (census) data on the average age when American people get

married. On the other hand, however, the average American Facebook user is much younger than the average citizen. 12 This is just

one example, but it highlights the need to take account of particular characteristics and the

limitations of producing representative results when extracting information from online users†behaviour Given the popularity of mobile-cellular

services, non-Internet-related mobile -network big data seems to have the widest socioeconomic coverage in the near term

and the greatest potential to produce relatively representative information globally particularly in developing countries. By

the end of 2014, the number of mobile -cellular subscriptions is expected to be nearing 7 billion,

and the number of mobile -cellular subscriptions per 100 inhabitants is expected to reach 90 per cent.

Mobile data are already being utilized for research and policy-making, not only in developed but also in developing economies

There are various examples of how mobile phone records have been used to identify socioeconomic patterns and migration

patterns, describe local, national and international societal ties, and forecast economic developments. 13 Data are also being

used to improve responsiveness in the event of natural disasters or disease outbreaks. Lu Bengtsson and Holme (2012) used mobile call

records to study the population displacements following Haiti†s 2010 earthquake, with a view to using such methods to improve the

effectiveness of humanitarian relief operations immediately after a disaster. Call records have also been merged with epidemiological

data to understand the spread of malaria in Kenya (Pindolia et al. 2012; Wesolowski et al 2012a), and of cholera in Haiti after the 2010

Mobile network big data have been utilized to great effect in the area of transportation helping to measure and model people†s

data from the ICT sector, and especially those available to telecommunication operators have wide applicability for informing multiple

public policy domains. Leveraging such data to complement official statistics and facilitate broader development will enable governments

as well as development agencies to better serve their citizens and beneficiaries. Less use has thus far been made of telecommunication

big data with a view to understanding its potential for producing additional information and statistics on the information society

In assessing that potential, including the potential for providing complementary 181 Measuring the Information Society Report 2014

better understand the type of data that can be made available 5. 3 Telecommunication data

and their potential for big data analytics Fixed and mobile telecommunication network operators, including Internet service providers

ISPS), are an important source of data and for the purpose of this chapter, all forms of

telecommunication big data (either volume velocity or variety) are being considered. Most telecommunication data can be considered as

TGD, 14 that is, the result of an action undertaken such as making a call, sending an SMS,

accessing the Internet or recharging a prepaid card Since the service with the widest coverage and

greatest uptake and popularity is the mobile -cellular service, data from mobile operators have the greatest potential to produce representative

results and reveal developmental insights on the population, including in developing countries and, increasingly, low-income areas

Not surprisingly, the big data for development initiatives (outlined in Section 2. 2) have mainly drawn on mobile network big data rather than

on those from fixed-telephone operators or ISPS. Figure 5. 2 illustrates some of the similarities

and differences in the type of information that mobile network operators, as opposed to fixed-telephone operators and ISPS, produce

and shows some of the additional insights, in particular in terms of the location and mobility information that mobile networks and services

generate Telecommunication data The mobile telecommunication data that operators possess can be classified into different types, depending on the nature of the information

they produce. They include traffic data, service access detail records, location and movement data, device characteristics, customer details and

tariff data. For a more detailed overview of these types of data, see Chapter 5 Annex

To collect traffic data, operators use a range of metrics to understand and manage the traffic

flowing through their networks, including the measurement of Internet data volumes, call, SMS and MMS volumes, and value-added service

VAS) volumes. Internet service providers can also use deep packet inspection (DPI), 15 which is a special process for scanning data packages

transiting the network Service access detail records, including call detail records (CDRS), are collected by operators

whenever clients use a service. They are used to manage the infrastructure and for billing purposes, and include information on the

time and duration of services used and the technology used, for example, for the mobile network (2g, 3g, etc..

These data are potentially also very useful for building a rich profile of customers, as outlined in this section

Mobile networks capture a range of movement and location variables to identify user location and movement patterns.

The degree of accuracy of this information depends on a number of factors, including the network used and device

generation, and can be classified broadly into two different types: passive and active positioning data, with the latter providing more

detailed and precise location information Since mobile user devices used to access mobile telecommunication services come with an

international mobile station equipment identity IMEI) number, operators can identify some device characteristics, including the handset

make and model and type of technology (2g 3g, LTE) employed. Mobile network operators can use the IMEI number to identify the specific

mobile handset being used by a subscriber which in turn can provide some insight as to that Chapter 5. The role of big data for ICT monitoring and for development

182 Figure 5. 2: An overview of telecom network data Source: ITU, adapted from Naef et al.

2014 Tr affi c da ta Fixed operator Mobile operator ISP Data volume Call volume

SMS/MMS volume Erlang DPI data Timestamp of use Contact network Duration of use Applicable charges

Handset type Technology utilized (2g, 3g, DSL/ADSL, etc Billing address Passive positioning data e g. cell ID

IMEI Active positioning data e g. cell triangulation, GPS MAC address Customer demographics (e g. age, gender, national ID card number

Billing address Payment history (postpaid Billing address Recharge history (prepaid Service order history Tariff sheet

Se rv ic e ac ce ss d et ai l re co rd s

Lo ca ti on da ta D ev ic e ch ar ac t -er

is ti cs Cu st om er de ta ils Ta ri ff da ta

Transaction generated data Stored warehouse data 183 Measuring the Information Society Report 2014 subscriber†s purchasing power (see below for

more details In addition, telecommunication operators hold various customer details that were captured during the customer registration process.

These can include the customer†s name, age, gender billing address and, in some cases, national

identity card number. Customer details may also include a history of the services accessed, service option preferences as well as other details (as

referred to in Chapter 5 Annex Finally, operators maintain tariff data in the form of billing records for their current and past

services, from which information on a customer†s usage patterns and preferences can be extracted The information outlined above is used at the

aggregate level to derive a range of indicators to provide operators with information on the uptake of different services and on their

How mobile operators currently use data to track service uptake, business performance and revenues Operators use their TGD to monitor the uptake and penetration

mobile-cellular and active mobile-broadband subscriptions On the basis of the detailed service-usage data collected

telecommunication operators can produce a range of detailed indicators relating to service consumption. For each customer

it is possible to determine the minutes of use (Mou), number of originating and terminating calls, SMS and MMS usage, data

upload volumes, data download volumes, level of use of different VAS, and level of use of different OTT services.

These data can be reported as averages (over time or for different categories of user), as well as at various levels of aggregation (again over time

or for different categories of user. These measures are often key performance indicators (KPIS), tracked

and used in particular by operators, but also by regulators and at the international level Finally, service consumption data are used to produce revenue

data and projections at various levels of disaggregation or aggregation. For example, the average revenue per user (ARPU) is

a KPI for operators, which identify their most important customers on the basis of the revenue they generate for the company

Similarly, revenue projections are made not just at the level of a particular service, 16 but also to identify the most important network

elements. For example, mobile operators collect indicators on the revenue being generated at the base station level, often in real

will, for instance, often associate revenue data with resource allocation to ensure that Qos at the base stations used by their

others, use aggregated revenue data to track and benchmark countries†ICT developments, monitor the evolution of the information society and

identify digital divides The telecom industry†s use of big data Telecommunication companies are actively seeking to intensify their use of big data analytics

in order to improve existing services and create new ones. For operators, big data open up opportunities for better understanding of their

customers, which in turn leads to improved sales and marketing opportunities. At the same time big data can help optimize network operations

and create new revenue streams and business lines, for example when selling data Customer profiling Telecom operators capture a range of

behavioural data about their customers Chapter 5. The role of big data for ICT monitoring and for development

184 Customer profiles include details about customers†mobility patterns, social networks and consumption preferences. Collectively, these

digital breadcrumbs enable operators to profile and segment their customers based on a variety of metrics (Figure 5. 3). Depending on the country

or region, there may be different privacy and data regulations governing the manner in which operators may keep and/or use such data.

This being the case, the extent to which behavioural profiling is used by operators may vary greatly

•Customer interests: these can be captured, or in some cases inferred, on the basis of usage levels (time spent

and/or volume) for different VAS and OTT services. DPI can also be used to categorize interests based on sites

visited (as opposed to content accessed Sophisticated clickstream analyses from DPI data17 can also generate more finely

-grained interest classifications •Socio-economic class: While customer details will often enable operators to classify their customers†socioeconomic

status, such details are not always very reliable. Big data, on the other hand can help to enhance that classification

by enabling analysis of the levels of consumption of different services including on the basis of spending (often

in relation to other services), types of device used, frequency of change of handset, and so on •Likelihood of churn:

Big data techniques can help operators understand churn better by enabling them to model the likelihood

Customer profiling using telecom big data Source: ITU CUSTOMER INTERESTS SOCIO -ECONOMIC CLASS LEVEL OF INFLUENCE OF

a large number of off-net users in a customer†s network, operators may target the subscriber and/or the off-net users

with promotions and incentives aimed at converting off-net connections into on-net users •Mobility profile:

Mobile operators accord a high priority to identifying the locations most frequented by their customers, in

Furthermore, social network insights can be used by an operator to market its services to the off-network contacts that are connected

subsidiary of SK TELECOM, uses big data to help its parent company to cut churn and generate new

revenue, and has used data mining to achieve a fourfold improvement in churn forecasting. The operator found that customers planning to quit

also seek to monetize the data they hold. The simplest way of doing this is to sell (anonymized

data to third parties. The customer insights obtained through the analysis of usage data can also help create new business lines, either

through innovation (e g. new types of VAS) or by partnering with other businesses, including credit-scoring and related financial services

One example is based the US big data startup Cignifi, 19 which obtains data from mobile operators and financial institutions to build credit

profiles and evaluate customer creditworthiness see Box 5. 8). Cross-promotions with brick-and -mortar businesses are a potentially high-growth

5. 4 Big data from mobile telecommunications for development and for better monitoring In 2013, the United nations High-level Panel of

Eminent Persons on the Post-2015 Development Agenda called for a â€oedata revolution†that draws on existing and new sources of data for the

post-2015 development agenda (United nations 2013). ) In March 2014, the forty-fifth session of UNSC, the highest decision-making body for

international statistical activities, presented a report on â€oebig data and modernization of Chapter 5. The role of big data for ICT monitoring and for development

186 statistical systemsâ€, and proposed the creation of a big data working group at the global level

UNSC, 2013). 20 Current uses of big data to complement official statistics are still exploratory but there is a growing interest in this topic, as

evidenced by the numerous initiatives being pursued by the United nations, as well as by others, including the World bank, OECD, Paris21

and NSOS There are many big data sources that can be used to monitor and assess development

results. In a world where mobile telephony is increasingly ubiquitous, it is not surprising that mobile telecommunication big data have unique

potential as a new data source, with high mobile -cellular penetration levels and the increasing

use of mobile phones, even among the poorest and most deprived, making them particularly valuable by comparison with other types of

telecommunication data. Indeed, when referring to the data revolution, the United nations High-level Panel cited the example of â€oemobile

technology and other advances to enable real -time monitoring of development results†This section will present some of the existing

and growing) evidence for the role of mobile big data in achieving development goals in various

policy areas, including disaster management and sustainable and economic development In addition to their use for development

telecommunication big data have potential as a source to enable monitoring of the information society, although they have yet to assume a critical

role in complementing the official ICT statistics that are collected and used for that purpose. As

the lead agency on global telecommunication and ICT statistics, however, ITU is exploring the potential of big data to complement its existing

and often limited, set of ICT statistics. This section presents a first attempt to help identify some of

the areas in which mobile telecommunication big data could complement existing ICT indicators to provide a more complete

comprehensive and up-to-date picture of the state of today†s information society Mobile phone big data for develop

-ment Mobile data offer a view of an individual†s behaviour in a low-cost, high-resolution, real

-time manner. Each time a user interacts with a mobile operator, many details of the interaction

are captured, creating a rich dataset relating to the consumer. Topping up airtime, making calls and sending SMSS,

downloading applications or using value-added services are all examples of interactions for which the time, location

device, user and other detailed information are captured in the operator†s system. From these interactions, information about identity

fact that these data are uniquely detailed and tractable, the information captured cannot easily be derived from other sources on such

The fact that the format of the data is relatively similar across different operators and countries creates a huge potential for

potential of mobile data for development in a number of different areas There have been a number of interesting

Big data for disaster management and syndromic surveillance21 Mobility data collected immediately after a disaster can in many cases help emergency

responders to locate affected populations and enable relief agencies to direct aid to the right

One application of such mobility data is for syndromic surveillance, especially to model the spread of vector-borne22 and

Using mobile data for development A recently published report (Cartesian, 2014) explores the potential of mobile data for development.

It points to three primary types of analysis-ex-post evaluation, real-time measurement, and future predictions and planning-in a

the development of models which protect user privacy while still allowing for the extraction of insights that can serve

mobile data Migration monitoring Text analysis economic downturn prediction Text analysis commodity fluctuation prediction Assessment

Mobile data to track food assistance delivery Geo-targeted links between Ag suppliers /purchasers Pests, bad

algorithms to anticipate prod. churn Social network targeted marketing Post-disaster refugee reunification Sentiment analysis of

research in Kenya combined passive mobile positioning data with malaria prevalence data to identify the source and spread of infections

Wesolowski et al. 2012b). ) Similar work in Haiti showed how mobile phone data was used to track the spread of cholera after the 2010

earthquake (Bengtsson et al. 2011, see Box 5. 5 The integration of mobility data from mobile

networks with geographic information frameworks, 23 supplemented with additional information, shows great potential for tracking the spread of vector-borne and other

damaged mobile network infrastructure is repaired as rapidly as possible Chapter 5. The role of big data for ICT monitoring and for development

188 Big data for better transportation planning A data-centric approach to transportation management is already a reality in many

developed economies. Transportation systems are being fed with sensor data from a multitude of sources such as loop detectors, axle counters

parking occupancy monitors, CCTV, integrated public transport card readers and GPS data derived not only from phones but also from

public transport and private vehicles (Amini Bouillet, Calabrese, Gasparini and Verscheure 2011 One advantage of mobile networks is that

even the least developed mobile network infrastructure generates passive positioning data, which, despite its limited spatial accuracy

cell ID), has great potential for transportation planning. For example, IBM researchers used CDR data from mobile operator Orange to map out

citizens†travel routes in Abidjan, the largest city in CÃ'te d†Ivoire, and show how data-driven insights

could be used to improve the planning and management of transportation services, thereby reducing congestion (Berlingerio et al.

2013). ) By simply extending one bus route and adding four new ones, overall travel time was reduced by ten

per cent. Passive mobile positioning data has also been used for transportation planning and management in Estonia (Ahas and Mark, 2005

and has provided reliable results in Sri lanka Lokanathan et al. 2014, see Box 5. 6 Box 5. 5:

How mobile network data can track population displacements †an example from the 2010 Haiti earthquake

The Figure below shows the number of people estimated to have been in Port-au-prince (Pap) on the day of the 2010 Haiti

earthquake, but outside the capital 19 days later. The circles represent the numbers of people who were displaced.

This map was produced on the basis of mobile network data to show the potential of big data in tracking population movements

Source: Bengtsson et al. 2011 Figure Box 5. 5: Tracking mobility through mobile phones Port-au-prince (Pap Number of

people displaced after earthquake 189 Measuring the Information Society Report 2014 Box 5. 6: Leveraging mobile network data for transportation and urban planning in Sri lanka

Very similar findings between the results of an official household survey assessing mobility patterns (right-hand map) with the

results of a big data analysis using mobile-phone data (left-hand map) underscore the merits of big data.

The image on the left based on mobile-phone data, depicts the relative population density in Colombo city and its surrounding regions at 1300

hours on a weekday in 2013, compared to midnight the previous day. While the yellow to red colouring shows areas in which the

density has increased relative to midnight, the blue colouring depicts areas in which the density has decreased relative to

Mobile big data (left) versus official survey data (right Both passive and active positioning data are used to

analyse traffic conditions, particularly in urban areas with higher base-station density. Active positioning data (especially GPS) produce higher precision in

location data and are therefore the most useful Operators may offer such specialized services based on passive or active location data) either

directly, or by providing data to third parties. Mobile network data are less expensive, are in real time and

are less time-consuming to produce than survey data, particularly in urban and peri-urban areas

where base-station density tends to be high In another example, the analysis of mobility

flows between two Spanish cities derived from three different data sources †mobile -phone data, geolocated Twitter messages and

the census †showed very similar results, and although the representativeness of the Twitter geolocated data was lower than the (real-time

mobile-phone and census data, the degrees of consistency between the population density profiles and mobility patterns detected by

means of the three datasets were significant Lenormand et al, 2014 Chapter 5. The role of big data for ICT monitoring and for development

190 Big data for socioeconomic analysis Data from mobile operators can provide insights in the areas of economic development and

socioeconomic status, often in near real time Big data techniques can therefore complement official statistics in the intervals between official

surveys, which are usually relatively expensive and time-consuming and therefore carried out infrequently. In many cases, insights derived

from big data sources may help to fill in the gaps rather than replace official surveys. It should

also be noted that mobile network big data are one of the few big data sources (and often the only one) in developing economies that

contain behavioural information on low-income population groups Frias-Martinez et al. 2012) developed a mathematical model to map human mobility

variables derived from mobile network data to people†s socioeconomic and income levels. The model took into account existing socioeconomic

and income-related data derived from official household surveys, and the results showed that populations with higher socioeconomic levels

are associated more strongly with larger mobility ranges than populations from lower socio -economic levels. By extending this method, the

study suggested that it was possible to create a model to estimate income levels based on data

from mobile network operators Another study, by Gutierrez, Krings and Blondel 2013), used two types of mobile network

data, namely subscriber communication data and airtime credit purchase records, to assess socioeconomic and income levels.

The authors used airtime purchase records based Box 5. 7: Poverty mapping in CÃ'te d†Ivoire using mobile network data

In CÃ'te d†Ivoire, researchers used mobile network data specifically communication patterns, but also airtime credit

purchase records) from Orange to estimate the relative income of individuals, as well as the diversity and inequality of income

levels. The research helped to understand socioeconomic segregation at a fine-grained level for CÃ'te d†Ivoire, with the

following map showing poor areas (in blue) in relation to the areas of high economic activity (yellow to red areas

Source: Gutierrez et al. 2013 Figure Box 5. 7: High-and low-income areas in CÃ'te d†Ivoire

Abidjan Road to Ghana Liberian border Roads to Mali and Burkina faso 0. 53 0. 73

on the assumption that users who make large purchases are more affluent than those who

analysis with a study of users†social networks with two users being considered as connected if they communicated with each other at least

once a month. Results showed that people tend to socialize with those who have a similar

regarding revenue data and the difficulty this poses for outside parties wishing to obtain such

data The use of mobile-operator TGD can also foster financial inclusion by facilitating the provision of

Group to Assist the Poor (CGAP) and GSM Association (GSMA) estimated that close to 2 billion

people had a mobile phone but no bank account After analysing mobile consumption variables CGAP suggested that it was possible to identify

example of how mobile big data can be used for the unbanked is Cignifi, a big data startup that uses

the mobile phone records of poor people to assess their creditworthiness when they apply for a loan

Boxâ 5. 8 Big data for understanding societal structures Social-network studies relying on self-reporting relational data typically involve both a limited

number of people and limited number of time points (usually one. As a result, social-network

analysis has generally been confined to the examination of small population groups through a small number of snapshots of interaction

patterns based on telecommunication data it has become possible to obtain insights into societal structures on a scale that was

Using mobile-phone data to track the creditworthiness of the unbanked Cignifi, a big data startup, has developed an analytic platform

to provide credit and marketing scores for consumers, based on their mobile-phone data. The Cignifi business model is founded

on the idea that â€oemobile phone usage is not random †it is highly predictive of an individual consumer†s lifestyle and riskâ€

Based on the behavioural analysis of each mobile-phone user †phone calls, text messages, data usage and, extrapolating from

these, lifestyles †the company identifies patterns and uses them to generate individual credit risk profiles.

This information could help many of the world†s unbanked to have access to insurance, credit cards and loans.

respond to changes in customer activity as the data are refreshed usually every two weeks. In addition to updating a person†s

findings from the model against historical lending data from approximately 40 000 borrowers using the mobile operator Oi†s

Chapter 5. The role of big data for ICT monitoring and for development 192 have been used to study the geographic

However, telecommunication data are also revolutionizing the study of societal structures at the micro level.

is possible to assess friendship using data from mobile network operators, and that the accuracy is compared high when with self-reported

data. Leveraging these behavioural signatures to obtain an accurate characterization of relationships in the absence of survey data could

also enable the quantification and prediction of macro and micro social-network structures that have thus far been unobservable

Big data to monitor the information society There is a case to be made for analysing data captured by telecommunication operators in

the interests of improving the current range of indicators used for monitoring the information society.

An internationally-accepted and widely-adopted list of indicators is the core list of ICT indicators developed by the Partnership

on Measuring ICT for Development, a multi -stakeholder initiative launched in 2004.24 This list includes, among others, the key-infrastructure

access and individual-use indicators that ITU collects and disseminates. Some of these indicators are amenable for augmentation using

big data analytics. 25 The core indicators on ICT infrastructure and access include indicators on mobile-cellular and

mobile-broadband subscriptions, which remain some of the most widely used and internationally comparable telecommunication indicators

produced for tracking the information society One of the main issues with mobile-cellular and

mobile-broadband subscription data is that they do not refer to unique subscriptions, or mobile

users. Since one person can have multiple subscriptions, or share a subscription with another person, it is not possible to determine

how many individuals subscribe to, or use, the mobile service. It is often the case that countries

with large numbers of prepaid subscriptions display relatively high penetration rates since prepaid cards can often be purchased at no initial

to monitor the time during which a SIM CARD remains inactive. In June 2014, for example GSMA estimated that, globally, the number of

unique mobile subscriptions was just over 50 percent, whereas the number of connections per 100 population far exceeded 100 per cent. 26

Survey-based data, for example on Internet users and mobile-phone users, do not entail the same issues as subscription data.

They are collected through household surveys, directly from citizens, and their level of reliability is relatively

high. The advantage of surveys is that they can go into more depth on the use of ICTS.

example, one of the core indicators reflects the types of online activity pursued by Internet users

and includes response categories such as seeking health information, obtaining information from government entities or participating in social

Survey-based data can also be broken down by individual characteristics, including gender, age, educational level and occupation

these data is that they are not widely available in particular, many developing countries do not yet collect data on individual use of

or household access to, ICTS), are relatively expensive to produce, and are much less timely

than subscription data (often with a time lag of one year. Consequently, data on users of

the information society and the types of online service they consume are limited, and in many

networks and mobile big data could be used to identify alternative, less costly and faster ways of

administrative data from operators and survey data collected by NSOS, it is particularly interesting to assess some of the ways in

which big data can be used to overcome the shortcomings of existing key ICT indicators and to provide additional insights into ICT access

and use, user behaviour, activities and also the individual user. Big data could help in obtaining

more granular information in several areas, and big data techniques could be applied to existing data to produce new insights.

In particular operators†big data could produce information in the following areas Individual subscriber characteristics: Additional categorization across both time and space are

possible for subscription indicators, and big data could provide additional information on gender, socioeconomic status and user location

Information on gender or age, for example, could be derived from customer registration information notwithstanding a number of challenges and

privacy issues, as discussed later in Section 5. 5 The socioeconomic status of the person linked

to a subscription could be derived from big data techniques applied to users†consumption information, as well as other data contained in

customer registration information. In addition, the analysis of customers†mobility patterns will often allow for an understanding of important locations

work and home being the two most important and of the use of mobile services in rural versus

urban areas. It would thus be possible to gain a more reliable and more granular understanding

of service penetration across space on the basis of actual behaviour/activity, rather than of what

All subscription data could provide information as to location. In the case of fixed-telephone and fixed-broadband

subscriptions, which are linked to an address through the billing information, it is possible to obtain information on the administrative division

to other (administrative) databases in order to Box 5. 9: Using mobile big data and mobile networks for implementing surveys

An important measurement for assessing the development of the information society is the extent to which households

have access to ICTS. Given the need for continued recourse to surveys for collecting the corresponding data, and the declining

response rates where traditional surveys are concerned (Groves 2011), mobile operators could develop platforms to facilitate the

collection of survey data. This could include targeting a wide variety of respondents covering the full spectrum of appropriate

using big data analytics. In 2011, for example, UN Global Pulse partnered with Jana, 27 a mobile-technology company,

the feasibility of using mobile phones for the deployment of rapid global surveys on well-being. 28 This requires, however, that the

mobile users targeted for the survey match the requisite survey profile. For instance, if one of the requirements was for the survey

-phone user. To that end, the World bank has experimented with the use of mobile phones to conduct statistically representative

monthly household surveys in Latin america and the Caribbean. 29 Source: ITU Chapter 5. The role of big data for ICT monitoring and for development

194 create new information. Consumption patterns could also deliver additional information on the socioeconomic status of the person/household

Particularly rich possibilities exist where data from mobile-cellular and mobile-broadband subscriptions are concerned, since they are

linked to mobility profiles. The indicators for such subscriptions could be broken further down to understand the utilization of services †including

voice, data and VAS †over time, and the intensity of use. Mobile operators are able to provide

3g, LTE-Advanced, etc. but also on the types of service that subscribers are using, and the

therefore, potentially identify Internet and VAS usage patterns between rural and urban areas and identify the kinds of application or webpage

that mobile-Internet users access. Combined with individual subscriber characteristics, this information could provide new and rich insights

into the digital divide and help understand usage patterns, including intensity of use, by gender socioeconomic status and also location

Internet activities carried out by individuals This information is collected currently only by countries that carry out household ICT surveys

In addition, mobile-operator data could be combined with customer information from popular online services, such as Facebook

Google or other, local (financial, social etc services to provide additional insights. This could be done by using probabilistic analyses to match

the profiles developed using data from online services with customer profiles generated from analyses of mobile-operator data.

This would require telecommunication operators, OTT providers and other Internet content providers to work together and share information

This technique is, currently, probably the least developed one, also because of the lack of a

good ontology and of privacy issues. However, as operators seek to gain a better understanding of

their customers in terms of the type of content they consume (as revealed through clickstream analyses), DPI may provide greater insights for

websites could be classified individually in terms of the information they provide, then Internet -user activities, including their frequency and

intensity, could be understood much better By applying big data techniques to survey data and administrative data from operators, new

insights could be derived, in particular, in respect of the following Subscriptions versus subscribers: Big data techniques could help extrapolate the actual

number of unique mobile subscribers or users rather than just subscriptions, by comparing subscription numbers to user numbers derived

from household surveys, and by taking into account usage patterns or data from popular Internet companies such as Google or Facebook

By linking data collected from different sources and combining subscription data and usage patterns, a correlation algorithm could be

developed to reverse engineer approximate values for these indicators, in order to estimate user numbers in between surveys, and possibly

in real time. This could be pursued in a similar way to the work done by Frias-Martinez and

Virseda (2012) on estimating socioeconomic variables using mobile-phone usage data, as described in greater detail at the beginning of

this section. It is important to note here that depending on such correlation techniques, big data methods only complement existing surveys

rather than replacing them completely (see Section 5. 5 for a further discussion of this

In sum, relatively simple big data techniques can help analyse and provide complementary information on existing ICT data, and provide

new insights into the measurement of the information society. This includes information on the use of different services and applications

intensity, frequency, and the geographic 195 Measuring the Information Society Report 2014 locations from which subscribers access ICT

services and applications. All of these insights on subscribers could potentially be further disaggregated by different demographic

and socioeconomic profiles. However, all of them relate to subscriptions. Given multiple SIM usage and the fact that users will in many

cases be using ICT services from more than one operator or device, additional techniques need to be leveraged if the insights articulated

for subscribers are to be extended to unique individuals. Such techniques will often include combining data from surveys with big data to

build new correlation and predictive analytic techniques Finally, it should be noted that the methods that

could help improve the indicators on individual and household access and use could also be used

and for other big data for development projects, big data analysis cannot replace survey data, which is needed to build and

test correlations and to validate big data results While the opportunities discussed above present what is analytically possible, data access

and privacy considerations are complex and nuanced, and therefore place constraints on what is practically feasible or advisable.

Section 5. 5 discusses these challenges in greater detail 5. 5 Challenges and the way

forward Attempting to extract value from an exponentially growing data deluge of varying structure and variety comes with its share of

challenges. The most pressing concerns are those associated with the standardization and interoperability of big data analytics, as

well as with privacy and security. Addressing such privacy and other concerns with respect to data sharing and use is critical,

and it is important for big data producers and users to collaborate closely in that regard. This includes

raising awareness about the importance and potential of producing new insights, and the establishment of public-private partnerships

to exploit fully the potential of big data for development Data curation, standardization and continuity Data curation and data preparation help to

structure, archive, document and preserve data in a framework that will facilitate human understanding and decision-making.

Traditional curation approaches do not scale with big data and require automation, especially since 85 per cent of big data are estimated to be

unstructured (Techamerica Foundation, 2012 Dealing with large heterogeneous data sets calls for algorithms that can understand the data

shape while also providing analysts with some understanding of what the curation is doing to

the data (Weber, Palmer and Chao, 2012 Telecom network operators themselves have to contend with interoperability issues arising

from the different systems (often from different vendors) they employ. It is not uncommon for

operators to write customized mediation software to overcome potential inter-comparability issues among data from different systems.

The problems are compounded when one has to take account of secondary third-party users that may seek to

leverage the data. The framework used by an NSO to organize data would be different from that

used by a network engineer or a marketing or business intelligence specialist. Naturally, telecom network operators have curated their data

based on their needs. To be able to use telecom big data for development and monitoring and to guarantee its continuity, the creation of

a semantic framework would require greater consensus among the many diverse stakeholders involved (telecom operators, network equipment

manufacturers, system developers, developmental practitioners and researchers, NSOS, etc Chapter 5. The role of big data for ICT monitoring and for development

196 Accessing and storing data, and data philanthropy Big data for development is still in its nascent

stages and, as such, comes with its share of challenges, not least of which is obtaining access to what is essentially private data.

Private corporations would hesitate to share information on their clients and their business processes in case such sharing is illegal, precipitates a

loss of user confidence and/or accidentally reveals competitive business processes. More importantly, companies will not share until there

are incentives to do so. Until holders of big data become more comfortable about their release it is going to be difficult for third-party research

entities to gain access Researchers (mainly from developed countries with some exceptions such as LIRNEASIA) have

recently succeeded in obtaining mobile network big data, but it has taken them considerable time to build and leverage the necessary

relationships with operators. Such privileged access is conditioned for the most part by lengthy legal agreements whose preparation requires major investments of time.

All the parties to such agreements have to address the necessary parameters as to how data are to

be used, including the manner in which they are to be extracted anonymized and, and with regard to time periods for access, etc.

Even once agreements are in place, both researchers and operators face costs arising from the technical

challenges associated with extraction of the data on account of different curation approaches and problems relating to the interoperability of

different systems Some mobile operators are taking tepid steps towards sharing data more publicly. Orange

for example, hosted a â€oedata for Development Challenge, †releasing an aggregated anonymized mobile dataset from CÃ'te d†Ivoire to researchers

and convening a conference at MIT in early 2013 where 84 papers from different researchers were

presented. A follow-up conference, this time using Orange data from Senegal, is planned for 2015.30 In 2014, Telecom italia initiated a similar

challenge, making data from the territories of Milan and the Autonomous Province of Trento available to researchers for analysis. 31 It has

however, gone one step further: in addition to releasing some of its own telecom datasets, it partnered with other data providers to curate

and release additional big datasets containing weather, public and private transport, energy event and social network data.

In both the Orange and Telecom italia cases, researchers had to go through an approval process in order to

gain access Organizations such as UN Global Pulse are seeking to popularize the concept of â€oedata

philanthropyâ€, aimed at systematizing the regular and safe sharing of data by building on the precedents being created by the ad hoc

activities outlined earlier. Such efforts by UN Global Pulse, as well as by other organizations such as LIRNEASIA, that seek to bring different

stakeholders to the same table, remain critical to the efforts being made to open up private-data

stores in order to obtain actionable development insights There is a gap that needs to be addressed if large-scale pooling and sharing of such data

are to become a reality. Cross-sector and cross -domain collaboration would benefit greatly from facilitators or intermediaries capable of

addressing issues related to standardization and data-curation practices when pooling data from multiple sources. This facilitatory role may

even be played by a third-party organization able to subsume regulatory and privacy burdens faced by data providers, effectively acting as

a gatekeeper to ensure that data are used transparently and in a way that contributes to overall scientific knowledge generation, while

ensuring that any safeguards that may be applicable in respect of private information are applied. Such an approach was taken recently

by the pharmaceutical company Johnson and Johnson, which decided to share all of its clinical trial data.

To facilitate the process, they hired Yale University†s Open Data Access (YODA) Project 197

Measuring the Information Society Report 2014 to act as gatekeepers (Krumholz, 2014. YODA undertakes the necessary scientific review of any

make use of the data and ensures that necessary privacy and data usage guidelines are followed. 32

The question remains as to who is placed best to act as gatekeepers and standard-bearers when it comes to telecom network big data

Some have argued that NSOS are placed well to ensure that best practices are followed in the collection and representation of big data

and to provide a stamp of trust for potential third-party data seekers. Telecom operators for their part, are regulated mostly by sector

-specific regulators who can also have purview and dictate terms governing the privacy and data-reporting responsibilities of operators

Ultimately, however, the decision as to who takes on the gatekeeper and standardization function requires the confluence of multiple

actors. It is here that organizations such as ITU UN Global Pulse and others have a greater role

to play in building an institutional model for data sharing and collaboration, in consultation with all

protocols) of privately held data such as mobile -phone records can be mutually beneficial to both government and private sector.

For example mobile network operators monitor and forecast their revenue at the cell-tower level. Emerging research in Africa shows how reductions in

revenues, including airtime top-ups, could presage declines in income in specific regions This could allow for targeted and timely policy

-action system shows how data sharing could be considered a business risk mitigation strategy for operators in emerging markets.

big data is linked closely to advances in the ICT sphere, including the falling cost of data storage

Depending on the data volume, storage can still be costly, especially where privacy considerations preempt the use of specialized third-party

cloud-based services. But as storage prices continue to fall, they are expected to be less of

As social scientists look towards private data sources, privacy and security concerns become paramount. To mitigate the potential risks, all

such data sharing. These stakeholders include not just the public and private sectors, but also significantly, the general public, who in many

cases are the primary producers of such data through their activities. It is also the public that

must ultimately decide on how the data they produce may be used. The World Economic Forum†s â€oerethinking Personal Data†project

personal data economy, and hosts consultations to deepen understanding of what type of trust frameworks are needed between individuals

new data ecosystem. 33 Discussions must address the individual†s privacy expectations, as well as those of private-sector stakeholders looking

of personal data that needs to be protected OECD, for example, defines personal data as â€oeany information relating to an identified or

identifiable individual (data subject) †(OECD 2013). ) The result of such an approach has been the policy of â€oeinform and consent†practised by

most companies to inform users of what data are Chapter 5. The role of big data for ICT monitoring and for development

198 being collected and how they will be used. It has been argued, however, that in a big data world

the â€oeinform and consent†approach is woefully inadequate and impractical, and that a new approach is needed (Mayer-Schã nberger and

Cukier, 2013; WEF, 2013 Firstly, user-privacy policies have morphed into long documents written in †legalese†that most

users can hardly comprehend and have little patience for reading in full. Secondly, in the big

data paradigm, the greatest potential often lies in secondary uses, which may well manifest long after the data was collected originally.

It is thus impractical for companies to have a priori knowledge of all the potential uses and to seek

permission from the user every time such a new use is found. Given the volumes of data

that individuals are now generating, companies would find themselves struggling to maintain meaningful control Of greater concern is how to articulate the

privacy issues that may arise when data from one source is combined with data from other sources

to reveal/infer new data and insights. This blurs the lines between personal and nonpersonal

information, allowing seemingly nonpersonal data to be linked to an actual individual (Ohm 2010). ) Digitized behavioural data crumbs may in

fact greatly diminish personal privacy. The use of DPI, for example can technically reveal all of a user†s online activity.

Going one step further it is possible to understand a person†s needs behaviours and preferences by using data

-mining techniques on the digital breadcrumbs For instance, a recent study showed how Facebook â€oelikes†could accurately predict a range

of behavioural attributes such as, inter alia, sexual orientation, ethnicity, religious and political views, and use of addictive substances (Kosinski

Stillwell and Graepel, 2013 Data anonymization35 (i e. methods designed to strip data of personal information), employed

by computational social scientists, has been called into question (Narayanan and Shmatikov 2008). ) A recent study of mobile CDRS for 1. 5

million anonymized users covering a 15 month period showed how the authors were able to

identify 90 per cent of the users with just four data points, and 50 per cent with just two points

de Montjoye, Hidalgo, Verleysen and Blondel 2013). ) Although the actual real-world identities of the users were unknown, the authors point

out that the data could in fact be de-anonymized completely by cross-referencing them with other

data sources. The attendant privacy concerns about such cross-referencing are clear, and have to be taken seriously and addressed

However such de-anonymization concerns remain, for the time being, somewhat premature for developing countries, mainly because the

connections, the registered user and the actual user may not be one and the same. Depending on the country, SIM resellers may pre-register

the SIMS they sell under their own name, and SIMS that are registered by one family member

may be used by other members of the same family. Sri lankan operators, for example, see a great mismatch between the person registering

a subscription and the person using it. The same may also be the case in many other developing

countries. 36 Irrespective, there is a consensus that there have to be safeguards in place, be they technological

data are kept secure. Data breaches undermine consumer confidence and hinder efforts to exploit big data for the greater social good.

Encryption virtual private networks (VPNS), firewalls, threat monitoring and auditing are some potential technical solutions that are employed currently

199 Measuring the Information Society Report 2014 but they need to be mainstreamed (Adolph 2013).

) The paradigmatic shift required to address privacy has started, but it will be some time before a consensus is achieved on the most appropriate

of big data, the World Economic Forum (WEF initiated a global multi-stakeholder dialogue on personal data that advocated a principle-based

approach, with the principles arising from a new approach that shifts governance from the data

per se to its use; acknowledges the importance of context rather that treating privacy as a binary

tools to actively engage users, enabling them to make clear choices based on an actual value

privacy and data protection in a big data world the danger is that these questions may take too

use of big data for broader development. Hence a balanced risk-based approach may be required in the context of what is under discussion here

i e. the use of telecom big data for monitoring and development. This does still require the

of big data for development can be â€oesandboxed†with appropriate privacy protections imposed on researchers, while still ensuring that the broader

Veracity in data analysis and results â€oegarbage in, garbage outâ€, or GIGO for short is a computer science concept that refers to

the fact that the veracity of the output of any logical process depends on the veracity of the

input data. In the big data paradigm, it is easy to overlook that concept, given the expectation

that when dealing with vast volumes of (often unstructured) data from a multitude of sources

â€oemessiness†is to be expected. As Mayer -Schã nberger and Cukier (2013) note, â€oewhat we lose in accuracy at the micro level we gain

Data quality and their provenance do matter, and the question is important in establishing the

generalizability of the big data findings Data provenance and data cleaning Understanding data provenance involves tracing the pathways taken by data from the

originating source through all the processes that may have mutated, replicated or combined the data that feed into the big data analyses.

This is no simple feat. Nor, given the varied sources of data that are utilized, is it always as feasible as

the scientific community would wish. However at the very least it is important to understand

some aspects of the origin of data. For example the fact that some mobile network operators choose to include the complete routing of a call

that has been forwarded means that there may be multiple records in the CDRS for the same

call. If that is not taken into consideration, the subsequent social network analysis could contain errors (overstating or understating tie strength

for example. While it may not be possible to establish data provenance as envisaged by scientists, it is at the very least important to

understand the underlying processes that may have created the data Data cleaning remains a key part of the process

to ensure data quality. It is important to verify that the quantitative and qualitative (i e

categorical) variables have been recorded as expected. In a subsequent step, outliers must be removed, using decision-tree algorithms or

other techniques. However, data cleaning itself is a subjective process (for example, one has to decide which variables to consider) and not a

truly agnostic one as would be desirable, and is thus open to philosophical debate (Bollier, 2010

Are the data representative Related to the question of data provenance is the issue of understanding the underlying

Chapter 5. The role of big data for ICT monitoring and for development 200 population whose behaviour has been captured

The large data sizes may make the sampling rate irrelevant, but they do not necessarily

make it representative. Not everyone uses Twitter, Facebook or Google. For example ITU estimates suggest that 40 per cent of the

world†s population uses the Internet. In other words, more than four billion people globally are not yet using the Internet,

and 90 per cent of them are from the developing world. Of the world†s three billion Internet users, two-thirds

are from the developing countries. Even though mobile-cellular penetration is close to 100 per cent, this does not mean that every person in

the world is using a mobile phone. This issue of representativeness is of high relevance when considering how telecommunication data

may be used for monitoring and development While the potential benefits to be gained from leveraging mobile network operator data for

monitoring and development purposes hinges on the large coverage, close to the actual population size, it is nevertheless not the whole

population. Questions such as the extent of coverage of the poor, or the levels of gender

representation among telecom users, are all valid considerations. While the registration information might provide answers, the reality

is that the demographic information on telecom subscribers, for example, is not always accurate. With prepaid subscriptions being

the norm in most of the developing world, the demographic information contained in mobile -operator records is practically useless, even with

mandatory registration as discussed above The issue of sampling bias is illustrated best by the case of Street Bump, a mobile app developed

accelerometer to detect potholes while users of the app are driving around Boston and notifies City hall.

demographics of app users, who often hail from affluent areas with greater smartphone ownership (Harford, 2014.

Hence, the â€oebig†in big data does not automatically mean that issues such as measurement bias and methodology

internal and external data validity and data interdependencies can be ignored. These are fundamental issues not just for â€oesmall data†but

also for â€oebig data†(Boyd and Crawford, 2012 Behavioural change Digitized online behaviour can be subject to

personas, so studying people†s data exhaust may not always give us insights into real-world dynamics.

where in essence the data artefact is itself a by-product of another activity. Telecom network big data, which mostly fall under this category

may be less susceptible to self-censorship and persona development, but the possibility of these phenomena cannot be ruled out.

Nor is it inconceivable that users may stop using their mobiles, or even turn them off, in areas

where they do not wish their digital footprint to be left behind. In a way, big data analyses

of behavioural data are subject to a form of the Heisenberg uncertainty principle, whereby as soon as the basic process of an analysis is

known, there may be concerted efforts to exhibit different behaviour and/or actions to change the outcomes (Bollier, 2010.

For example the famous Google pagerank algorithm has spawned an entire industry of organizations that claim to enhance website page rankings, and

search-engine optimization (SEO) 38 is now an established part of website development Changes in behaviour could also partially explain

the declining veracity of Google Flu Trends GFT), researchers having found influenza-like illness rates as reflected by Google searches to

be no longer necessarily correlating with actual influenza virus infections (Ortiz et al. 2011 Recent research has shown that since 2009

when GFT failed to reflect the nonseasonal influenza outbreak), infrequent updates have not improved the results

and GFT has in fact persistently overestimated flu prevalence (Lazer Kennedy, King and Vespignani, 2014. GFT does

actions of the population that turned to Google with its health queries, and which contributed

the robustness of the original algorithm. For example, the enthusiasm surrounding GFT may well have created rebound effects, with more

and more people turning to Google with their broader health questions, thereby introducing additional search terms (due to different

considering big data analyses for monitoring purposes. Dr Nathan Eagle, a pioneer in the use of cellphone records to understand phenomena

related to social development and public health stresses the importance of weeding out false assumptions by conducting an a priori survey

when CDR data from Rwanda showed low mobility in the wake of flooding he theorized that this was due to an outbreak

it comes to the generalizability of telecom -data analyses based on big data. For example prior research had established a power-law

distribution between the frequency of airtime recharges and average recharge amount. 39 It was further found that the poor tended to top

researchers working with Sri lankan mobile datasets attempted to use these findings to help them segregate their analyses for different socio

the big data paradigm, leading to the discovery of misleading patterns. As Google†s Chief Economist, Hal Varian, notes, â€oethere are often

) Big data draws many of its techniques from machine learning, which is primarily about correlation and predictions. 40 Big data are by

their very nature observational and can measure only correlation and not causality. Supporters of big data have predicted the end of theory and

hypothesis-testing, with correlation trumping causality as the most relevant method (Anderson 2008; Mayer-Schã nberger and Cukier, 2013

However, such predictions may be premature The behavioural economist Sendhil Mullainathan notes that inductive science (i e. the algorithmic

mining of big data sources) will not drown out traditional deductive science (i e. hypothesis testing), even in a big data paradigm.

Among the three Vs in the traditional big data definition volume and variety produce countervailing forces.

More volume makes big data induction techniques easier and more effective, while more variety makes them harder and less effective

It is this variety issue that will ensure the need for explaining behaviour (i e. deductive science

Chapter 5. The role of big data for ICT monitoring and for development 202 rather than merely predicting it (Mullainathan

Causal modelling is possible in a big data paradigm by conducting experiments. Telecom network operators themselves use such

techniques when rolling out new services or, for that matter, for pricing purposes. The question

with â€oesmall data, †in this case the statistics collected by the Centers for Disease Control and

note, when combined with small data, â€oegreater value can be obtained by combining GFT with other near†real time health data. †Where data

from mobile network operators are used for syndromic surveillance, as in the case of malaria in Kenya (Wesolowski et al.

2012a), big data are most useful as a basis for encouraging timely investigation, rather than as a replacement

for existing measures of disease activity. Even when engaging with the broader question of how telecommunication network data could be

used for monitoring, surveys and supplemental datasets will remain important to sharpen the analyses and especially to verify the underlying

prior to data anonymization to build a training dataset. This enabled them, for example, to understand variations in mobility, social networks

and consumption among men and women and between different socioeconomic groups which would not have been possible using only

Virseda (2012) needed census data to build their algorithms and provide training data for their

algorithms to reverse engineer approximate survey maps. Official statistics will thus continue to be important to building the big data models

and for periodic benchmarking so that the models can be tuned fine to reflect ground realities Transparency and replicability

The issues with GFT also illustrate transparency and replicability problems with big data research The fact that the original private data may

in many cases not be available to everyone underscores the importance of opening up such

private-data sources (in a manner that addresses potential privacy concerns) or of peer reviews that can hone

and improve the analyses. Instead consumers of such research have no option but to take the analysis and the results on faith.

data calls for a combination of specialized skills in the areas of data mining, statistics and

domain expertise, as well as data preparation cleaning and visualization. NSOS may have deep statistical skills in house, but this is not

enough when it comes to working with large volumes of big data calling for computer science and decision-analysis skills that are not

emphasized in traditional statistical courses Mcafee and Brynjolfsson, 2012. NSOS recognize this shortcoming. In a recent global survey of

NSOS from 200 economies, conducted by UNSC respondents identified the development and 203 Measuring the Information Society Report 2014

data sources (UNSC, 2013. Currently, there is a mismatch between the supply of and demand for talented individuals with the requisite

i e. data scientists. Mckinsey predicts that by 2018 the demand for data-savvy managers and analysts in the United states will

amount to 450 000, whereas the supply will fall far short of this, at only 160 000 (Manyika et al

to leverage big data for development will face competition from the private sector when seeking to attract the right talent.

the most from the use of telecommunication big data to complement official statistics, have a shortage of advanced analytical skills by

comparison with developed economies. Until such time as systematic capacity development yields proper rewards, it will remain essential

Current research suggests that new big data sources have great potential to complement official statistics and produce insightful

of big data will have to overcome a number of barriers. This includes the development of models which protect user privacy while still

allowing for the extraction of insights that can serve development purposes, in particular where developing countries are concerned.

for using big data to complement official ICT statistics. Although this report highlights some of the big data sources and techniques that

could be used, further research is needed to understand and confirm the usefulness of big data sources for monitoring the information

society As with other official statistics, it is paramount for big data producers and big data users

to collaborate and to initiate a dialogue to identify opportunities and understand needs and constraints.

Since many of the big data sources lie within the private sector, close cooperation between NSOS, on the one hand

and telecommunication operators and Internet companies, including search engines and social networks, on the other, is necessary and could

be institutionalized through public-private partnerships Operators and Internet companies Business interests will naturally provide operators

and Internet companies with the incentive to talk to commercial vendors of big data analytics In addition, operators and Internet companies

can benefit greatly from engagement with academia and researchers to understand how to leverage big data for different purposes

Such engagement will also broaden their understanding of the limitations and assist them in the development of new methodologies

algorithms and software techniques that can be repurposed for business-use cases. Indeed, where the applications of data use for development

are concerned, operators also have an interest in maximizing the economic well-being of their customer base

Operators and Internet companies need to take advantage of their existing customer relationships to elicit a greater understanding

of consumer concerns and needs in relation to privacy. They are placed well to develop a

Chapter 5. The role of big data for ICT monitoring and for development 204 privacy framework, in consultation with other

stakeholders Given their business concerns, operators and Internet companies may hesitate to pool and share their data with those from other sources

including from competitors), but this is something that is worth exploring. Combining big data sources has great potential to increase

added value and produce new insights. There is scope for exploring established models for such pooling †for example, the sharing by banks of

some of their customer data with credit bureaux Governments Governments have different opportunities and different roles to play in the exploitation of big

data for monitoring and development. They can use big data to identify areas where rapid intervention may be necessary,

to track progress and make sure their decisions are based evidence and to strengthen accountability. More and more governments are recognizing the importance of

big data and have set up communities of practice and working groups to study their use and potential impact (UNSC 2013

Governments should also facilitate the legislative changes that are required and take a lead in setting big data standards.

To this end, national regulatory authorities (NRAS) and NSOS, in consultation with other national stakeholders are placed best to lead the corresponding

big data clearing houses that promote analytical best practices in relation to the use of big data for complementing official statistics and for

development. Those standards, which NSOS are in the best position to enforce, would also have

to encompass best practices in relation to data curation and metadata standards. To this end NSOS must also prioritize the upgrading of the

handle big data, while at the same time investing in the necessary computational infrastructure As the main regulatory interface to the telecom

sector, NRAS are placed well to co-champion the national discussion on how telecommunication big data may be leveraged for social good

Regulators have a role to play in facilitating the introduction of legislation that addresses privacy

concerns while encouraging data sharing in a secure manner. The following recommendations were made in a recently published ITU draft

paper (ITU, 2014 •Establishing mechanisms to protect privacy: Regulators could develop a regulatory mechanism that would

collecting personal data to accountable and responsible uses of personal data This mechanism would foresee a well

-resourced privacy regulator with the expertise and power to enforce such a use-based privacy protection mechanism

In return, data users would be permitted to reuse personal data for novel purposes where a privacy assessment indicates

minimal privacy risks •Restricting the use of probabilistic predictions: While the use of big data

can help better decision-making through probabilistic predictions, this information should not be used against citizens Regulators should restrict the ways in

can utilize big data predictions •Fostering big data competition and openness: Regulators could foster big data competition in increasingly

concentrated big data markets, including by ensuring that data holders allow others to access their data under fair and

reasonable terms 205 Measuring the Information Society Report 2014 International stakeholders International stakeholders †including UN

agencies and initiatives (such as ITU and UN Global Pulse), the Partnership on Measuring ICT for Development, ICT industry associations

and producers of big data (Google, Facebook etc.)) †have an important role globally. More work is needed to understand fully the potential

of big data and examine the challenges and opportunities related to big data in the ICT sector. To this end, the key international

stakeholders have to work together to facilitate the global discussion on the use of big data

UN Global Pulse, as one of the main UN initiatives exploring the use of big data,

can do much to inform and motivate the discussion on global best practices and the use of big data for

development Where using big data for monitoring the information society is concerned, new partnerships, including public-private

partnerships between data providers and the ICT statistical community, including ITU, could be formed to explore new opportunities and

address challenges, including in the area of international data comparability and standards As one of the main international bodies working

on issues related to the telecommunication sector, ITU could leverage its position to facilitate global discussion on the use of telecom big data

for monitoring the information society Together, ITU and UN Global Pulse could facilitate the work that needs to be done by

NRAS and NSOS, through awareness raising and engagement on privacy frameworks, data sharing, and analytical global best practices

ITU could help reduce the transaction costs associated with obtaining telecommunication big data, for example by facilitating the

standards-setting process. Standardized contracts for obtaining data access as well as standards on how the data are stored,

collated and curated can collectively reduce the overall transaction costs of accessing and leveraging telecommunication big data for social good

Academia, research institutes and develop -ment practitioners The research into how telecom data may be used to aid broader development is being done

mainly by academia, public and private research institutes and, to a lesser degree, development practitioners. This makes them important

stakeholders in defining the state of the art with respect to leveraging big data for development They, more than others, have been the first to

engage with telecommunication operators with a view to using their data for development They therefore understand the potential and

challenges from multiple perspectives. Their collective experiences will be valuable as big data for development becomes mainstreamed

207 Measuring the Information Society Report 2014 Chapter 5 Annex The mobile-telecommunication data that

operators possess can be classified into different types, depending on the nature of the information they produce.

They include traffic data, service access detail records, location and movement data, device characteristics, customer

details and tariff data Traffic data Operators use a range of metrics to understand and manage the traffic flowing through their

networks. These include •Data volume: both uplink and downlink volumes for Internet traffic can be captured at various levels of

disaggregation down to the individual subscriber, or even to the level of a base station (in the case of a mobile operator

or local switch (in PSTN networks. These can be analysed to understand subscriber demand for data at both an individual

level and at aggregate levels, and the understanding thus gained can be used for billing purposes and for network

management •Erlang: a dimensionless metric used by mobile network operators to understand the offered and utilized network load. 41

Erlang data are used to understand the load on a base station at any given time

•Call, SMS and MMS volumes are used for a variety of purposes from billing to

customer relationship management, as well for network planning •Deep packet inspection42 (DPI) is used to scan the information that goes over a

network. Operators employ DPI to varying degrees, and it is not always feasible for the entire data stream to be captured and

stored, owing to the storage requirements that would be needed and also to privacy concerns. Often only the header

information, which includes originating and recipient Internet protocols (IPS is captured for a variety of purposes

including to manage the network and understand the demand for particular applications and websites Service access detail records

Whenever a user utilizes a telecommunication service, each access is recorded not only for infrastructure management but also for billing

purposes. Depending on the type of service, the resulting records may be referred to as call detail records (CDRS), SMS/MMS detail records, Internet

access detail records, etc. and may include the following information •A timestamp of when the service was

accessed •The duration of use of the service (for example, duration of a call

•The technology used (2g, 3g, etc The most common use of such data is for basic

billing purposes, in addition to which they can be used to build a rich profile of customers, as

Chapter 5. The role of big data for ICT monitoring and for development 208 Location and movement data

Mobile networks can, depending on their sophistication, capture a range of movement and location variables, which can be broadly

classified into two different types: passive and active positioning data (see Annex Box 1 Annex Box 1:

Active versus passive positioning data Passive positioning transaction generated data (TGD) is automatically generated by the network

and captured in the operator†s logs for billing and network management purposes to understand network load

and to keep track of the handset in relation to its network elements. Active positioning data (which is

of relevance only to mobile networks) is location and movement data that is captured in response to a specially initiated network

query to locate a handset using either network or handset-based positioning methods. GPS location data can also be considered

as active positioning data Active positioning data Active positioning data can be generated using either device

-centric or network-centric methods, as well as via satellite i e. GPS. The use of these methods has developed either in

response to national regulations requiring operators to capture higher-precision location data, and/or to a business case for

providing location-based services. The large-scale capture of such higher-resolution data is undertaken mainly by operators

in developed economies. Operators in developing economies use some of these methods, but often on a case-by-case basis

and not for their entire subscriber base. 43 However, this trend is currently changing, and an increasing number of regulators are

considering mandating operators to collect higher-resolution location information. 44 The table below lists some of the active

Passive positioning data Passive location data are contained in the subscriber†s registration information, which includes some form of mailing address

billing address in the case of postpaid. While fixed-telephone network operators have access only to static location data

mobile networks have much richer and dynamic location data CDRS, SMS detail records and Internet access records are the

main sources of passive positioning data for mobile operators and reside in their data warehouses. These records include the

ID of the antenna (cell ID), which in turn has a geolocation an azimuth (i e. antenna orientation information) and an

angular tilt. 45 It is also possible to obtain such data in real time through data-mediation services,

but these are not universally implemented Passive location data from the billing records are obviously

sparse and generated only when the phone is used and when the network knows which cell a particular handset is currently

connected to. However, many operators choose not to archive these data if they do not have a business case to justify the

additional storage costs. Where they are archived, such cell -handoff data provide a time-stamped sequence of cells that the

phone was attached to, and provides for a rich mobility profile as compared to the event-based billing records

Passive positioning data based on cell IDS is inexpensive when compared to active positioning data, but the tradeoff

lies in their lower precision, usually at the level of network cells. This lower-resolution location estimate can range from

a few hundred metres in urban areas with a higher-density base station coverage, to a few kilometres, especially in rural

areas with sparse coverage. Furthermore, handsets are not always served by the nearest antenna, for a variety of reasons

associated with signal strength, topography and saturation loads at the nearest antenna during peak times.

chosen by the user on the handset. For example, if the closest cell supports only 2g but the user has chosen to connect only

to 3g networks, the handset will always connect to an antenna that supports 3g, even if it is further away.

Despite these location errors or limitations that can occur in analyses using such passive location data, at an aggregate level (temporal

and/or spatial) these data remain very valuable Source: ITU Device characteristics All mobile user devices used to access mobile

telecommunication services come with an international mobile station equipment identity IMEI) number. This 15 or 16 digit number is

captured whenever a device is used to access 209 Measuring the Information Society Report 2014 Table Annex Box 1:

Active positioning methods Method Description Cellular triangulation using angle of arrival Aoa The Aoa method uses data from base stations that have been augmented with arrays of smart antennas.

It is then possible to determine the location of a handset by triangulating known signal angles from at least two base

stations. The location estimate varies from around 50 to 150m and is often towards the latter end, especially when

handsets are far from the base stations Cellular triangulation using time of arrival Toa This method uses the geo-coordinates of the cell

GSM-GPS and assisted GPS (A-GPS Both of these utilize the network (mainly via triangulation from multiple base stations) to augment the satellite

Such location data have high spatial resolution, but are costly for operators to implement Enhanced observed time

telecommunication services. In addition to serving as a unique serial number for the handset, parts of it can reveal information

technology (e g. 2g, 3g, LTE), and it can be used for the collective categorization of handsets

Furthermore, devices used to access the Internet mobile handsets, routers, modems) also have a unique identifier known as a media access

control46 (MAC) address. Such identifiers can provide details of the device used to access the network.

Mobile network operators can use the IMEI number to identify the specific mobile handset being used by a subscriber, and

Telecom network operators capture various items of demographic data during the customer registration process. These can include the

customer†s age, gender, billing address and national identity card number (where available In addition, operators store the order history of

a rich customer/user and usage profile that the operator can leverage for a variety of purposes

Tariff data Operators maintain the complete tariff sheet and billing records for their current and past services

Mobile operators can associate such data with traffic data to understand the revenue that is being generated by specific network elements

e g. base stations), not only retrospectively, but also, possibly, in real time Chapter 5. The role of big data for ICT monitoring and for development

210 1 The report of the UN Secretary-General†s High-level Panel of Eminent Persons on the post-2015 Development Agenda

calls for a data revolution that â€oewould draw on existing and new sources of data to fully integrate statistics into decision-making

promote open access to, and use of, data and ensure increased support for statistical systems. †This suggests that efforts to

improve the availability of, and complement, official statistics have turned to the search for new data sources,

including big data. In addition, the European Statistical System Committee (ESSC) in 2013 adopted the Scheveningen Memorandum

on  Big data and Official Statisticsâ, which acknowledges that Big data represents new opportunities and challenges for

Official Statistics, and which encourages the European Statistical System and its partners to effectively examine the potential

of Big data sources, see: http://epp. eurostat. ec. europa. eu/portal/page/portal/pgp ess/0 docs/estat/SCHEVENINGEN

MEMORANDUM%20final%20version. pdf 2 This term was discussed first in 1991, although the term then used was generated â€oetransaction information†(Mcmanus, 1990

3 For more information, see http://www. cityofboston. gov/doit/apps/streetbump. asp 4 See http://www. donotpay. treas. gov/About. htm

5 In Europe, the collection and processing of personal data or information is regulated currently by Directive 95/46/EC

on the protection of individuals with regard to the processing of personal data and on the free movement of such data

Data protection Directive) 1 and Directive 2002/58/EC, as amended by Directive 2009/136/EC, on privacy and electronic

which focuses more specifically on the processing of personal data in the electronic communications sector. Article 7 of the Data protection directive establishes the principle of opt-in, according

to which personal data cannot legitimately be processed without the consent of the data subject, except if necessary to

preserve public order or morality, as well as to further the general interest of society or individuals. Building on this principle

Article 5 of the eprivacy Directive further provides that the processing of personal data can be effected only with the consent

of the data subject, who should be given clear and comprehensive information as to the manner and purpose of such

http://policyreview. info/articles/analysis/big data-big-responsibilities 6 See https://www. google. org/denguetrends

/7 A good example of this is the Conference Board Help Wanted Online (HWOL) data series that measures the number of new

first time online jobs and jobs reposted from the previous month for over 16 000 Internet job boards, corporate boards and

smaller job sites in the United states. More information can be found at http://www. conference-board. org/data/helpwantedonline. cfm

8 See http://bpp. mit. edu /9 According to Peerreach. com, 20 per cent of Indonesia†s online population uses Twitter, the second highest ratio in the world

See http://www. ibtimes. com/twitter-usage-statistics-which-country-has-most-active-twitter-population-1474852

10 See http://www. broadband. gov/qualitytest/about /11 ITU World Telecommunication/ICT Indicators database, 17th edition, 2014, available at

http://www. itu. int/en/ITU-D/Statistics/Pages/publications/wtid. aspx 12 See http://blog. stephenwolfram. com/2013/04/data science-of-the-facebook-world

/13 Mobile phone records have been used to infer locations of economic activity within, and patterns of migration to, slum areas

in Kenya (Wesolowski and Eagle, 2010) and internal migration in Rwanda (Joshua E. Blumenstock, 2012.

Other work has sought to understand international societal ties in Rwanda (Joshua E. Blumenstock, 2011) and the effects of migration on

mobile phone records to forecast the socioeconomic levels of localities, thereby yielding approximate census maps (Frias

mobile phone usage and regional economic development in CÃ'te d†Ivoire 14 The term â€oemetadata†is used also quite extensively to refer to TGD from telecommunication operators

15 Deep packet inspection (DPI) is a process that utilizes specialized software to scan all of the data packets traversing a

particular IP network. It can be employed by network operators (especially ISPS) to filter for malicious content (e g. spam) or

http://en. wikipedia. org/wiki/Deep packet inspection 16 It should be noted that there is not a harmonized methodology for the allocation of revenues from bundled packages to

Taking into account the increasing trend towards the bundling of telecommunication services (e g double-and triple-play offers), revenue figures disaggregated per service are in most cases not comparable across operators

in order to understand which sites were accessed, in what order and how much time was spent at each Endnotes

18 Comments by SK TELECOM CEO Jinwu So to Mobile Asia Expo attendees -http://www. lightreading. com/document. asp?

19 For more information on Cignifi, see their website at http://www. cignifi. com /20 The full report can be accessed at http://unstats. un. org/unsd/statcom/doc14/2014-11-Bigdata-E. pdf

21 Syndromic surveillance refers to the collection and analysis of health data about a clinical syndrome that has a significant

impact on public health, with the data in question being used to drive decisions about health policy and health education

22 A vector-borne disease is a disease that is transmitted through an agent (person, animal or microorganism

23 A geographic information framework is a representation framework that codes the components of geospatial data in a

standardized manner so as to facilitate analyses and data exchange 24 See http://www. itu. int/en/ITU-D/Statistics/Pages/intlcoop/partnership/default. aspx for more information regarding the

25 For the latest list of core ICT indicators refer to 2014 edition of the â€oemanual for Measuring ICT Access and Use by Households

and Individuals available at http://www. itu. int/dms pub/itu-d/opb/ind/D-IND-ITCMEAS-2014-PDF-E. pdf

27 Jana has integrated its systems with 237 mobile operators worldwide, giving them a reach of almost 2 billion subscribers

29 For more information regarding this project, see http://web. worldbank. org/WBSITE/EXTERNAL/EXTABOUTUS

514426,00. html 30 More information about the Data for Development (D4d) challenges using Orange data can be found at

http://www. d4d. orange. com/home 31 See http://www. telecomitalia. com/tit/en/bigdatachallenge/contest. html

32 More information about the Yale university Open Data (YODA) project can be found at http://medicine. yale. edu/core/projects/yodap/index. aspx

33 For more information, and to view the outcome reports of this project, see http://www. weforum. org/issues/rethinking-personal data

34 It should be noted that there is no single ITU definition of privacy, that the above definition is accepted not universally and

that there are divergent views on the exact scope of the right 35 Anonymization and security techniques are very rich.

37 See, for example, http://www. unglobalpulse. org/privacy-and-data protection for an understanding of the privacy

whereby a website†s structure and content are optimized to make the site more visible to the webpage-indexing process of one or more search engines,

thereby ensuring that the website and/or webpage appears higher up in the results of a search query

39 In a power-law distribution, the functional relationships between two variables is such that the value of one variable varies as

42 DPI is a process that utilizes specialized software to scan all of the data packets traversing a particular IP network.

http://en. wikipedia. org/wiki/Deep packet inspection 43 Based on interviews between LIRNEASIA and operators in South and Southeast asia

regulation on the Indian Department of Telecommunications website http://dot. gov. in/sites/default/files/Unified%20licence 0. pdf

Chapter 5. The role of big data for ICT monitoring and for development 212 45 Most network operators use multiple sectorized antennas on a single base station.

When a sectorized antenna is placed on the base station it needs to be positioned properly. This includes setting the direction of the antenna (assuming a horizontal

For more information, see http://en. wikipedia. org/wiki/Sector antenna 46 A media access control (MAC) address is a unique identifier that is assigned to network interfaces mostly by a hardware

manufacturer. For example, the telecom operator captures the MAC address from a modem or router or handset that

accesses its network and maintains the mapping of this network interface to a particular customer.

http://en. wikipedia. org/wiki/MAC ADDRESS 47 An international mobile subscriber identity (IMSI) number is a 15-digit number unique to the particular SIM in a subscriber†s

handset. The mobile operator†s system retains a mapping between an IMSI number and the particular mobile number

assigned to a user. The IMSI conforms to Recommendation ITU-T E. 212. For more information, see

http://en. wikipedia. org/wiki/International mobile subscriber identity 213 Measuring the Information Society Report 2014 List of references

the data deluge makes the scientific method obsolete. Wired Magazine, 16 (7 Askitas, N. and Zimmermann, K. F. 2009), Google econometrics and unemployment forecasting.

Retrieved from http://ftp. iza. org/dp4201. pdf Azman, A s.,Urquhart, E. A.,Zaitchik, B. and Lessler, J. 2013), Using Mobile phone Data to Supercharge Epidemic Models of

Cholera Transmission in Africa: A Case study of CÃ'te d†Ivoire. Data for Development: Net Mobi 2013

Bengtsson, L.,Lu, X.,Thorson, A.,Garfield, R. and von Schreeb, J. 2011), Improved response to disasters and outbreaks by

tracking population movements with mobile phone network data: a post-earthquake geospatial study in Haiti. PLOS

Medicine, 8 (8), e1001083. doi: 10.1371/journal. pmed. 1001083 Berlingerio, M.,Calabrese, F.,Lorenzo, G. Di, Nair, R.,Pinelli, F. and Sbodio, M. L. 2013), Allaboard:

mobility and optimizing public transport using cellphone data. Data for Development: Net Mobi 2013 Blumenstock, J. E. 2011), Using mobile phone data to measure the ties between nations.

Proceedings of the 2011 iconference pp. 195†202. New york, USA: ACM Press. doi: 10.1145/1940761.1940788

Blumenstock, J. E. 2012), Inferring patterns of internal migration from mobile phone call records: evidence from Rwanda

Information technology for Development, 18 (2), 107†125. doi: 10.1080/02681102.2011.643209 Blumenstock, J. E. and Eagle, N. 2012), Divided we call:

Disparities in Access and Use of Mobile phones in Rwanda. Information Technologies & International Development, 8 (2), 1†16

Bollier, D. 2010), The Promise and Peril of Big data. WASHINGTON DC Bouckaert, J.,Van dijk, T. and Verboven, F. 2010), Access regulation, competition, and broadband penetration:

An international study. Telecommunications Policy 34.11 (2010): 661-671 Boyd, D. and Crawford, K. 2012), Critical questions for big data.

Information, Communication & Society, 15 (5), 662†679. doi: 10 1080/1369118x.2012.678878 Broadband Commission (2013a), The State of Broadband 2013:

Universalizing Broadband. Available at http://www. broadbandcommission. org/Documents/bb-annualreport2013. pdf Broadband Commission (2013b), Planning for Progress:

Why National Broadband Plans matter. Available at http://www. broadbandcommission. org/documents/reportnbp2013. pdf References

214 Broadband Commission for Digital Development (2011), Broadband Targets for 2015. Available at http://www. broadbandcommission. org/Documents/Broadband targets. pdf

Brynjolfsson, E.,Hitt, L. M. and Kim, H. H. 2011), Strength in Numbers: How Does driven Data Decisionmaking Affect Firm

Performance? Retrieved from: http://papers. ssrn. com/sol3/papers. cfm? abstract id=1819486 Byrne, C. 2013), How The UN€ s New Data Lab in Indonesia Uses Twitter to Preempt Disaster, Fast Company.

Retrieved 19 march 2013 from: http://www. fastcolabs. com/3007178/open-company/how-uns-new-data-lab-indonesia-uses-twitter

-preempt-disaster Caceres, N.,Wideberg, J. P. and Benitez, F. G. 2007), Deriving origin†destination data from a mobile phone network.

IET Intelligent Transport Systems, 1 (1), 15. doi: 10.1049/iet-its: 20060020 Calabrese, F.,Di Lorenzo, G.,Liu, L. and Ratti, C. 2011), Estimating origin-destination flows using mobile phone location data

IEEE Pervasive Computing, 10 (4), 36†44. doi: 10.1109/MPRV. 2011.41 CGALIES (2002), Report on implementation issues related to access to location information by emergency services (E112) in the

European union, Coordination group on access to Location Information for Emergency Services. European commission DG information society

http://www. cetic. br/usuarios/tic/2012/apresentacao-tic-domicilios-2012. pdf Choi, H. and Varian, H. 2009), Predicting initial claims for unemployment benefits.

http://static. googleusercontent. com/media/research. google. com/en//archive/papers/initialclaimsus. pdf Choi, H. and Varian, H. 2012), Predicting the present with Google trends.

Economic Record, 88, 2†9. doi: 10.1111/j. 1475 -4932.2012.00809. x Cisco (2013), Cisco Visual Networking Index:

paper c11-481360. pdf Cisco (2014), Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update, 2013-2018, White paper, February 2014

Available at: http://www. cisco. com/c/en/us/solutions/collateral/service-provider/visual-networking-index-vni/white

paper c11-520862. pdf CCK (2010), Communications Commission of Kenya: Interconnection Determination No. 2 of 2010, August 2010.

http://www. cck. go. ke/regulations/downloads/interconnection determination no2 2010. pdf CCK (2013), Communications Commission of Kenya:

http://www. cck. go. ke/resc/downloads/Sector statistics report q2 201314. pdf David, T. 2013), Big data from Cheap Phones.

Technology Review, 116 (3), 50†54 De Montjoye, Y.-A.,Hidalgo, C a, Verleysen, M. and Blondel, V. D. 2013), Unique in the Crowd:

Department of Census and Statistics Sri lanka (2009), Computer literacy Survey †2009, December 2009. Available at

http://www. statistics. gov. lk/CLS/Buletincomputerliteracy 2009. pdf 215 Measuring the Information Society Report 2014

D. 2009), Inferring friendship network structure by using mobile phone data. Proceedings of the National Academy of Sciences of the United states of america, 106 (36), 15274†8. doi:

European commission (2014), E-Communications and Telecom Single Market Household Survey. Special Eurobarometer 414 March 2014.

http://ec. europa. eu/public opinion/archives/ebs/ebs 414 en. pdf European commission (2014a), Broadband access in the EU:

Fink, C.,Mattoo, A. and Rathindran, R. 2002), An assessment of telecommunications reform in developing countries.

Information technology for Development, 18 (2), 91†106. doi: 10.1080/02681102.2011.630312 Frias-Martinez, V.,Virseda-Jerez, J. and Frias-Martinez, E. 2012b), On the relation between socioeconomic status and physical

Information technology for Development, 18 (2), 91†106. doi: 10.1080/02681102.2011.630312 Giles, J. 2012), Computational social science:

http://www. udesa. edu. ar/files/Admtecysociedad/12 eng. pdf Ginsberg, J.,Mohebbi, M. H.,Patel, R. S.,Brammer, L.,Smolinski, M. S. and Brilliant, L. 2009), Detecting influenza epidemics

using search engine query data. Nature, 457 (7232), 1012†4. doi: 10.1038/nature07634 Groves, R. M. 2011), A possible data future for the observational social sciences.

COSSA 30th Anniversary meeting Washington, D c Gruber, H. and Koutroumpis, P. 2013), Competition enhancing regulation and diffusion of innovation:

Gruber, H. and Verboven, F. 2001), The diffusion of mobile telecommunication services in the European union.

GSMA and Deloitte (2013), Arab States Mobile Observatory 2013. Available at http://www. gsma. com/publicpolicy/wpcontent/uploads/2012/03/GSMA MOBILEOBSERVATORY ARABSTATES2013. pdf

GSMA (2013a), Country overview: Sri lanka. Retrieved from: https://wiki. gsmaintelligence. com/gsma kb/images/e/e0

/Mobile for development country overview sri lanka. pdf GSMA (2013b), The Mandatory Registration of Prepaid SIM CARD Users, White paper. Retrieved from: http://www. gsma

com/publicpolicy/wp-content/uploads/2013/11/GSMA WHITE-Paper mandatory-Registration-of-Prepaid-SIM -Users 32pgwebv3. pdf References

216 Gual, J. and Trillas, F. 2004), Telecommunications policies: Determinants and impacts. CEPR Discussion Paper No. 4578

Gutierrez, L. 2003), Regulatory governance in the Latin american telecommunications sector. Utilities Policy, 11, 225†240

Gutierrez, L. and Berg, S. 2000), Telecommunications liberalization and regulatory governance: Lessons from Latin america Telecommunications Policy, 24 (10†11), 865†884

Gutierrez, T.,Krings, G. and Blondel, V. D. 2013), Evaluating socioeconomic state of a country analyzing airtime credit and

mobile phone datasets, 1†6. Retrieved from: http://arxiv. org/abs/1309.4496 Harford, T. 2014), Big data:

are we making a big mistake? Financial times. pp. 7†11. Retrieved 28 march 2014 from http://www. ft com/intl/cms/s/2/21a6e7d8-b479-11e3-a09a-00144feabdc0. html#axzz2xkbwrrlj

IBM (2012), Analytics: The real-world use of big data. How innovative enterprises extract value from uncertain data.

Retrieved from: http://public. dhe. ibm. com/common/ssi/ecm/en/gbe03519usen/GBE03519USEN. PDF IBM (2013), The Four V†s of Big data.

Retrieved from http://www. ibmbigdatahub. com/sites/default/files/infographic file/4-Vs-of-big data. jpg

ICT Qatar (2014), Qatar†s ICT landscape 2014 households and individuals. Available at: http://www. ictqatar. qa/en/documents

/download/Qatar†s%20ict%20landscape%20report%202014-Household%20and%20individuals 12. pdf ITU (2006), Security in Telecommunications and Information technology:

An overview of issues and the deployment of existing ITU-T Recommendations for secure telecommunications.

Retrieved from http://www. itu. int/dms pub/itu-t/opb/hdb/T-HDB-SEC. 03-2006-PDF-E. pdf

ITU (2009), Measuring the Information Society 2009. Available at http://www. itu. int/ITU-D/ict/publications/idi/material/2009/MIS2009 W5. pdf

ITU (2010), Monitoring the WSIS Targets a mid-term review. Available at http://www. itu. int/ITU-D/ict/publications/wtdr 10/material/WTDR2010 EXECSUM-en. pdf

ITU (2011), Handbook for the collection of administrative data on Telecommunications/ICT. Available at http://www. itu. int/en/ITU-D/Statistics/Pages/publications/handbook. aspx

ITU (2011), Measuring the Information Society 2011. Available at http://www. itu. int/ITU-D/ict/publications/idi/material/2011/MIS 2011 WITHOUT ANNEX 5. pdf

ITU (2012a), ICT adoption and prospects in the Arab region. Available at http://www. itu. int/ITU-D/ict/publications/arab/2012/index. html

ITU (2012b), Measuring the Information Society 2012. Available at http://www. itu. int/en/ITU-D/Statistics/Documents/publications/mis2012/MIS2012 WITHOUT ANNEX 4. pdf

ITU (2013), Big data: Big today, normal tomorrow. Retrieved from: http://www. itu. int/oth/T2301000022/en

ITU (2013a), Measuring the Information Society 2013. Available at http://www. itu. int/en/ITU-D/Statistics/Documents/publications/mis2013/MIS2013 WITHOUT ANNEX 4. pdf

ITU (2013b), Trends in Telecommunication Reform 2013: Transnational aspects of regulation in a networked society.

Available at http://www. itu. int/pub/D-REG-TTR. 14-2013 ITU (2014), The world in 2014:

ICT facts and figures. Available at http://www. itu. int/en/ITU-D/Statistics/Documents/facts/ICTFACTSFIGURES2014-e. pdf

ITU (2014), ITU manual for Measuring ICT access and Use by Households and Individuals. Available at

http://www. itu. int/dms pub/itu-d/opb/ind/D-IND-ITCMEAS-2014-PDF-E. pdf

217 Measuring the Information Society Report 2014 ITU GSR discussion paper (2014), Big data-Opportunity or Threat.

Retrieved from: http://www. itu. int/en/ITU-D/Conferences /GSR/Documents/GSR2014/Discussion%20papers%20and%20presentations%20-%20gsr14/Session3 gsr14

-Discussionpaper-Bigdata. pdf Jones, S. 2012), Why â€oebig Data†is the fourth factor of production,

Financial times. Retrieved 27 december 2012 from www. ft com/intl/cms/s/0/5086d700-504a-11e2-9b66-00144feab49a. html

Kosinski, M.,Stillwell, D. and Graepel, T. 2013), Private traits and attributes are predictable from digital records of human

behavior, 2†5. doi: 10.1073/pnas. 1218772110 Koski, H. and Kretschmer, T. 2005), Entry, standards and competition:

firm strategies and the diffusion of mobile telephony Review of Industrial Organization 26.1 (2005): 89-113

Krumholz, H. M. 2014), Give the data to the People, New york times. Retrieved 2 february 2014 from

http://www. nytimes. com/2014/02/03/opinion/give-the-data-to-the-people. html? hp&rref=opinion& r=1

Laney, D. 2001), 3d data management: Controlling data volume, velocity and variety. Gartner, 6 Lazer, D.,Kennedy, R.,King, G. and Vespignani, A. 2014), Big data.

The parable of Google Flu: traps in big data analysis. Science New york, N y.),343 (6176), 1203†5. doi:

10.1126/science. 1248506 Lazer, D.,Pentland, A.,Adamic, L.,Aral, S.,Barabasi, A l.,Brewer, D.,Van alstyne, M. 2009), Computational social science

Science (New york, N y.),323 (5915), 721†3. doi: 10.1126/science. 1167742 Lenormand, M, Picornell, M.,Cantu-Ros, O.,Tugores, A.,Louail, T.,Herranz, R.,Barthelemy, M.,Frias-Martinez, E.,Ramasco, J

Cross-checking different sources of mobility information, 2014 Levy, B. and Spiller, P. 1994), The institutional foundations of regulatory commitment:

telecommunications regulation. Journal of Law, Economics and Organization, 10 (2), 201†246 Levy, B. and Spiller, P. 1996), Regulations, institutions and commitment:

Comparative studies of telecommunications Cambridge, UK: Cambridge university Press Lokanathan, S.,Silva, N. de, Kreindler, G.,Miyauchi, Y. and Dhananjaya, D. 2014), Using mobile network big data for informing

transportation and urban planning in Colombo Lu, X.,Bengtsson, L. and Holme, P. 2012a), Predictability of population displacement after the 2010 Haiti earthquake

Manyika, J.,Chui, M.,Brown, B.,Bughin, J.,Dobbs, R.,Roxburgh, C. and Byers, A. H. 2011), Big data:

Mao, H.,Shuai, X.,Ahn, Y.-Y. and Bollen, J. 2013), Mobile Communications Reveal the Regional Economy in Cote d†Ivoire.

Data for Development: Net Mobi 2013 Mayer-Schã nberger, V. and Cukier, K. 2013), Big data:

A revolution that will transform how we live, work, and think. Houghton Mifflin Harcourt Mcafee, A. and Brynjolfsson, E. 2012), Big data:

the management revolution. Harvard Business Review, 90 (10), 60†6, 68,128 Retrieved from: http://www. ncbi. nlm. nih. gov/pubmed/23074865

Mcmanus, T. E. 1990), Telephone transaction generated information: Rights and restrictions. Cambridge, MA. Retrieved from

http://www. pirp. harvard. edu/pubs pdf/mcmanus/mcmanus-p90-5. pdf Mullainathan, S. 2013), What big data means for social science.

Headcon †13 Part I. Retrieved from http://edge. org/panel/headcon-13-part-i

Naef, E.,Muelbert, P.,Raza, S.,Frederick, R.,Kendall, J. and Gupta, N. 2014), Using mobile data for development.

Gates Foundation Whitepaper Narayanan, A. and Shmatikov, V. 2008), Robust de-anonymization of large sparse datasets. 2008 IEEE Symposium on Security

http://www. oecd. org/dataoecd/9/43/27125982. pdf OECD (2013), The OECD Privacy Framework.

http://www. oecd. org/sti/ieconomy/oecd privacy framework. pdf Ofcom (2011), Measuring Mobile Broadband in the UK:

performance delivered to PCS via dongles/datacards, September to December 2010, May 2011. Available at:

http://stakeholders. ofcom. org. uk/binaries/research/telecoms-research /bbspeeds2010/Mobile bb performance. pdf Ofcom (2013), Ofcom Communications Market Report 2013, August 2013.

Available at http://stakeholders. ofcom. org. uk/binaries/research/cmr/cmr13/2013 uk cmr. pdf Ohm, P. 2010), Broken promises of privacy:

Responding to the surprising failure of anonymization. UCLA Law Review, 57 (6 Ortiz, J. R.,Zhou, H.,Shay, D. K.,Neuzil, K. M.,Fowlkes, A l. and Goss, C. H. 2011), Monitoring influenza activity

a comparison of traditional surveillance systems with Google Flu Trends. Plos One, 6 (4), e18687. doi:

http://unstats. un. org/unsd/statcom/doc12/2012-12-ICT-E. pdf Phithakkitnukoon, S.,Calabrese, F.,Smoreda, Z. and Ratti, C. 2011), Out of sight out of Mind-How Our Mobile Social network

Changes during Migration. 2011 IEEE Third Int†l Conference on Privacy, Security, Risk and Trust and 2011 IEEE Third Int†l

Conference on Social Computing (pp. 515†520. IEEE. doi: 10.1109/PASSAT/Socialcom. 2011.11 Pindolia, D. K.,Garcia, A j.,Wesolowski, A.,Smith, D. L.,Buckee, C. O.,Noor, A m.,Tatem, A j. 2012), Human movement data

for malaria control and elimination strategic planning. Malaria Journal, 11 (1), 205. doi: 10.1186/1475-2875-11-205

The effects of telecommunications reform on network expansion and efficiency. Journal of Regulatory Economics, 1 (15), 65†92

Ros, A. 2003), The impact of the regulatory process and price cap regulation in Latin american telecommunications markets

Web search query volume as a measure of pharmaceutical utilization and changes in prescribing patterns.

2012), Demystifying Big data: A Practical Guide To Transforming The Business of Government Retrieved from: http://www. techamerica. org/Docs/filemanager. cfm?

f=techamerica-bigdatareport-final. pdf The White house (2014), Big data: Seizing Opportunities, Preserving values. Executive Office of the President.

Retrieved from http://www. whitehouse. gov/sites/default/files/docs/big data privacy report 5. 1. 14 final print. pdf UN Global Pulse (2012), Taking the Global Pulse:

Using New Data to Understand Emerging Vulnerability in Real-time UN High-level Panel (2013), A new global Partnership:

Eradicate Poverty and Transform Economies through Sustainable Development. The Report of the High-level Panel of Eminent Persons on the Post-2015 Development Agenda.

Available at: http://www. post2015hlp. org/wp-content/uploads/2013/05/UN-Report. pdf UN-OHRLLS (2013), State of the Least Developed Countries 2013.

Available at http://unohrlls. org/custom-content/uploads/2013/10/State-of-the-LDCS-2013. pdf

UNSC (2013), Report of the Secretary general on Big data and modernization of statistical systems UNSC (2014), Big data and modernization of statistical systems, December 2013.

Available at http://unstats. un. org/unsd/statcom/doc14/2014-11-Bigdata-E. pdf Van damme, E. 2002), The European UMTS-auctions, European Economic Review, Volume 46, Issues 4-5 may 2002, pp. 846

-858. doi: 10.1016/S0014-2921 (01) 00219-7 Varian, H r. 2013a), Beyond Big data. NABE Annual Meeting.

San francisco, CA Varian, H r. 2013b), Big data: New Tricks for Econometrics Wallsten, S. 2001), An econometric analysis of telecom competition, privatization,

and regulation in Africa and Latin america Journal of Industrial Economics, 49 (1), 1†19 Wang, H.,Calabrese, F.,Di Lorenzo, G. and Ratti, C. 2010), Transportation mode inference from anonymized and aggregated

mobile phone call detail records. 13th International IEEE Conference on Intelligent Transportation systems (pp. 318†323 IEEE. doi:

10.1109/ITSC. 2010.5625188 Ward, J. S. and Barker, A. 2013), Undefined By Data: A Survey of Big data Definitions, 2. Databases.

Retrieved from http://arxiv. org/abs/1309.5821 Waverman, L. and Koutroumpis, P. 2011), Benchmarking telecoms regulation:

The Telecommunications Regulatory Governance Index (TRGI. Telecommunications Policy, 35,450-468 Weber, N m.,Palmer, C. L. and Chao, T. C. 2012), Current Trends and Future Directions in Data Curation Research and Education

Journal of Web Librarianship, 6 (4), 305†320. doi: 10.1080/19322909.2012.730358 WEF (2013), Unlocking the Value of Personal data:

From Collection to Usage. Geneva, Switzerland. Retrieved from http://www3. weforum. org/docs/WEF IT UNLOCKINGVALUEPERSONALDATA COLLECTIONUSAGE REPORT 2013. pdf

WEF (2014), The Global Information technology Report 2014: Rewards and Risks of Big data (B. Bilbao-Osorio, S. Dutta and B

Lanvin, Eds References 220 Wesolowski, A. and Eagle, N. 2010), Parameterizing the Dynamics of Slums.

AAAI Symposium on Aritiï cial Intelligence and Developmen (pp. 103†108 Wesolowski, A.,Eagle, N.,Tatem, A j.,Smith, D. L.,Noor, A m.,Snow, R. W. and Buckee, C. O. 2012a), Quantifying the impact of

human mobility on malaria. Science (New york, N y.),338 (6104), 267†70. Doi: 10.1126/science. 1223467

Wesolowski, A.,Eagle, N.,Tatem, A j.,Smith, D. L.,Noor, A m.,Snow, R. W. and Buckee, C. O. 2012b), Quantifying the impact of

human mobility on malaria. Science (New york, N y.),338 (6104), 267†70. doi: 10.1126/science. 1223467

World bank (2014), Ending poverty requires more than growth, Press release No. 2 014/434/DEC. Available at

http://www. worldbank. org/en/news/press-release/2014/04/10/ending-poverty-requires-more than-growth-says-wbg

Wu, W.,Cheu, E. Y.,Feng, Y.,Le, D. N.,Yap, G. E. and Li, X. 2013), Studying Intercity Travels and Traffic

Using Cellular network Data Data for Development: Net Mobi 2013 Xia, J. 2011), The third-generation-mobile (3g) policy and deployment in China:

Current status, challenges, and prospects Telecommunications Policy, Volume 35, Issue 1 february 2011, pp. 51-63. doi:

10.1016/j. telpol. 2010.12.008 221 Measuring the Information Society Report 2014 Annex 1. ICT Development Index (IDI

methodology This annex outlines the methodology used to compute the IDI, and provides more details on

various steps involved, such as the indicators included in the index and their definition, the imputation of missing values, the normalization

objectives, data availability and the results of various statistical analyses such as the principal component analysis (PCA.

Data for all of these indicators are collected by ITU. 2 1. Fixed-telephone subscriptions per 100

inhabitants Fixed-telephone subscriptions refers to the sum of active analogue fixed-telephone lines, voice -over-IP (Voip) subscriptions, fixed wireless local

loop (WLL) subscriptions, ISDN voice-channel equivalents and fixed public payphones. It includes all accesses over fixed infrastructure

wire, voice services using Internet Protocol IP) delivered over fixed (wired)- broadband infrastructure (e g. DSL, fibre optic), and voice

services provided over coaxial-cable television networks (cable modem. It also includes fixed wireless local loop (WLL) connections, which

are defined as services provided by licensed fixed-line telephone operators that provide last-mile access to the subscriber using radio

fixed-line telephone network (and not a mobile -cellular network. In the case of Voip, it refers to

subscriptions that offer the ability to place and receive calls at any time and do not require a

computer. Voip is also known as voice-over -broadband (Vob), and includes subscriptions through fixed-wireless, DSL, cable, fibre-optic and

other fixed-broadband platforms that provide fixed telephony using IP 2. Mobile-cellular telephone subscriptions per

100 inhabitants Mobile-cellular telephone subscriptions refers to the number of subscriptions to a public mobile

-telephone service which provides access to the public switched telephone network (PSTN) using cellular technology. It includes both the number

of postpaid subscriptions and the number of active prepaid accounts (i e. that have been Annex 1. ICT Development Index (IDI) methodology

all mobile-cellular subscriptions that offer voice communications. It excludes subscriptions via data cards or USB modems, subscriptions to

public mobile data services, private trunked mobile radio, telepoint, radio paging and telemetry services 3. International Internet bandwidth (bit/s) per

Internet user International Internet bandwidth refers to the total used capacity of international Internet bandwidth, in megabits per second (Mbit/s

It is measured as the sum of used capacity of all Internet exchanges offering international bandwidth.

If capacity is asymmetric, then the incoming capacity is used. International Internet bandwidth (bit/s) per Internet user is calculated by

converting to bits per second and dividing by the total number of Internet users 4. Percentage of households with a computer

A computer refers to a desktop computer a laptop (portable computer or a tablet or similar handheld computer.

It does not include equipment with some embedded computing abilities, such as smart TV SETS, and devices with

telephony as a main function, such as mobile phones or smartphones. Household with a computer means that the computer is available

for use by all members of the household at any time. The computer may or may not be owned

by the household, but should be considered a household asset. 3 Data are obtained by countries through national

household surveys and are provided either directly to ITU by national statistical offices NSO), or ITU carries out the necessary research

to obtain them, for example from NSO websites There are certain data limits to this indicator insofar as estimates have to be calculated for

many developing countries which do not yet collect ICT household statistics. Over time, as more data become available, the quality of the

indicator will improve 5. Percentage of households with Internet access The Internet is a worldwide public computer

network. It provides access to a number of communication services, including the World wide web, and carries e-mail, news

entertainment and data files, irrespective of the device used (not assumed to be only a computer

†it may also be a mobile telephone, tablet, PDA games machine, DIGITAL TV, etc..Access can be

via a fixed or mobile network. Household with Internet access means that the Internet is available for use by all members of the household at any

time Data are obtained by countries through national household surveys and are provided either directly to ITU by national statistical offices

NSO), or ITU carries out the necessary research to obtain them, for example from NSO websites

There are certain data limits to this indicator insofar as estimates have to be calculated for many developing countries which do not yet

collect ICT household statistics. Over time, as more data become available, the quality of the

indicator will improve b) ICT use indicators The indicators included in this group capture ICT intensity and usage.

Data for all of these indicators are collected by ITU. 4 1. Percentage of individuals using the Internet

Individuals using the Internet refers to people who used the Internet from any location and

for any purpose, irrespective of the device and network used, in the last three months.

It can be via a computer (i e. desktop computer laptop computer or tablet or similar handheld 223

Measuring the Information Society Report 2014 computer), mobile phone, games machine DIGITAL TV, etc..Access can be fixed via a or

mobile network Data are obtained by countries through national household surveys and are provided either directly to ITU by national statistical offices

NSO), or ITU carries out the necessary research to obtain them, for example from NSO websites

There are certain data limits to this indicator insofar as estimates have to be calculated for many developing countries which do not yet

collect ICT household statistics. Over time, as more data become available, the quality of the

indicator will improve 2. Fixed (wired)- broadband subscriptions per 100 inhabitants Fixed (wired)- broadband subscriptions refers to the

number of subscriptions for high-speed access to the public Internet (a TCP IP connection. High -speed access is defined as downstream speeds

equal to, or greater than, 256 kbit/s. Fixed (wired broadband includes cable modem, DSL, fibre

and other fixed (wired)- broadband technologies such as Ethernet LAN, and broadband-over -powerline (BPL) communications.

Subscriptions with access to data communications (including the Internet) via mobile-cellular networks are excluded 3. Wireless-broadband subscriptions per 100

inhabitants Wireless-broadband subscriptions refers to the sum of satellite broadband, terrestrial fixed wireless broadband and active mobile-broadband

subscriptions to the public Internet •Satellite broadband subscriptions refers to the number of satellite Internet

subscriptions with an advertised download speed of at least 256 kbit/s. It refers to the retail subscription technology

and not the backbone technology •Terrestrial fixed wireless broadband subscriptions refers to the number

of terrestrial fixed Wireless internet subscriptions with an advertised download speed of at least 256 kbit/s. This

includes fixed Wimax and fixed wireless subscriptions, but excludes occasional users at hotspots and Wi-fi hotspot

subscribers. It also excludes mobile -broadband subscriptions where users can access a service throughout the country

wherever coverage is available •Active mobile-broadband subscriptions refers to the sum of standard mobile

-broadband subscriptions and dedicated mobile-broadband data subscriptions to the public Internet. It covers actual

subscribers, not potential subscribers even though the latter may have broadband-enabled handsets. Standard mobile-broadband subscriptions refers

to active mobile-cellular subscriptions with advertised data speeds of 256 kbit/s or greater that allow access to the

greater Internet via HTTP and which have been used to set up an Internet data connection using Internet Protocol (IP) in

the past three months. Standard SMS and MMS messaging do not count as an active

Internet data connection, even if the messages are delivered via IP. Dedicated mobile-broadband data subscriptions refers to subscriptions to dedicated data

services (over a mobile network) that allow access to the greater Internet and which are purchased separately from

voice services, either as a standalone service (e g. using a data card such as a USB modem/dongle) or as an add-on data

package to voice services which requires an additional subscription. All dedicated mobile-broadband subscriptions with

recurring subscription fees are included regardless of actual use. Prepaid mobile -broadband plans require use if there is no

monthly subscription. This indicator could also include mobile Wimax subscriptions Annex 1. ICT Development Index (IDI) methodology

224 c) ICT skills indicators Data on adult literacy rates and gross secondary and tertiary enrolment ratios are collected by the

UNESCO Institute for Statistics (UIS 1. Adult literacy rate According to UIS, the Adult literacy rate is defined

as â€oethe percentage of population aged 15 years and over who can both read and write with

understanding a short simple statement on his/her everyday life. Generally, †literacy†also encompasses †numeracyâ€,

the ability to make simple arithmetic calculations. †The main purpose of this indicator is â€oeto show the accumulated

achievement of primary education and literacy programmes in imparting basic literacy skills to the population, thereby enabling them to apply

such skills in daily life and to continue learning and communicating using the written word. Literacy

represents a potential for further intellectual growth and contribution to economic-socio -cultural development of society. †5

2. Gross enrolment ratio (secondary and tertiary level According to UIS, the Gross enrolment ratio is the

â€oetotal enrolment in a specific level of education regardless of age, expressed as a percentage of the eligible official school age population

corresponding to the same level of education in a given school-year. †2. Imputation of missing data

A critical step in the construction of the index is to create a complete data set, without missing values

There are several imputation techniques that can be applied to estimate missing data. 6 Each of the

imputation techniques, like any other method employed in the process, has its own strengths and

weaknesses. The most important consideration is to ensure that the imputed data will reflect a country†s

actual level of ICT access, usage and skills Given that ICT access and usage are correlated both

with national income, hot-deck imputation was chosen as the method for imputing the missing data, where previous year data are not

available to calculate the growth rates. Hot -deck imputation uses data from countries with â€oesimilar†characteristics,

such as GNI per capita and geographic location. For example, missing data for country A were estimated for a certain indicator by

first identifying the countries that have similar levels of GNI per capita and that are from the same region

and an indicator that has known a relationship to the indicator to be estimated. For instance, Internet use data of country A was estimated by using

Internet use data of country B from the same region with similar level of GNI per capita and similar

level of fixed Internet and wireless-broadband subscriptions. The same logic was applied to estimate missing data for all indicators included in

the index 3. Normalization of data Normalization of the data is necessary before any aggregation can be made

in order to ensure that the data set uses the same unit of measurement For the indicators selected for the construction

of the IDI, it is important to transform the values to the same unit of measurement, since

some values are expressed as a percentage of the population/of households, whereby the maximum value is 100, while other indicators

can have values exceeding 100, such as mobile -cellular and wireless-broadband penetration or international Internet bandwidth (expressed as

bit/s per user There are certain particularities that need to be taken into consideration when selecting the

normalization method for the IDI. For example, in 225 Measuring the Information Society Report 2014 order to identify the digital divide,

it is important to measure the relative performance of countries i e. the divide among countries.

Second, the normalization procedure should produce index results that allow countries to track progress of their

evolution towards an information society over time A further important criterion for the selection of

the normalization method was to choose one that can be replicated by countries. Indeed, some countries have shown a strong interest in applying

the index methodology at the national or regional level. Therefore, certain methods cannot be applied for example those that rely on the values of other

countries, which might not be available to users For the IDI, the distance to a reference measure was

used as the normalization method. The reference measure is the ideal value that could be reached

for each variable (similar to a goalpost. In all of the indicators chosen, this will be 100,

except for four indicators •International Internet bandwidth per Internet user, which in 2013 ranges from

136 (bits/s/user) to almost 6 445 759 Values for this indicator vary significantly

between countries. To diminish the effect of the huge dispersion of values, the data were transformed first to a logarithmic

log) scale. Outliers were identified then using a cut off value calculated by adding two standard deviations to the mean

of the rescaled values, resulting in a log value of 5. 90 •Mobile-cellular subscriptions, which

in 2013 range from 5. 6 to 304 per 100 inhabitants. The reference value for mobile-cellular subscriptions was

reviewed and lowered to 120. This value 120) was derived by examining the distribution of countries based on their

mobile-cellular subscriptions per 100 inhabitants value in 2013. For countries where postpaid is the predominant

mode of subscriptions, 120 is the maximum value achieved, while in countries where prepaid is dominant

•Fixed-telephone subscriptions per 100 inhabitants, which in 2013 range from zero to 124.

In line with fixed-telephone subscriptions the ideal value was defined at 60 per 100 inhabitants After normalizing the data, the individual series

were rescaled all to identical ranges, from 1 to 10. This was necessary in order to compare the

Fixed-telephone subscriptions per 100 inhabitants 0. 20 Mobile-cellular telephone subscriptions per 100 inhabitants 0. 20

International Internet bandwidth per Internet user 0. 20 Percentage of households with a computer 0. 20

Percentage of households with Internet access 0. 20 ICT use 0. 40 Percentage of individuals using the Internet 0. 33

Fixed (wired)- broadband Internet subscriptions per 100 inhabitants 0. 33 Wireless-broadband subscriptions per 100 inhabitants 0. 33

ICT skills 0. 20 Adult literacy rate 0. 33 Secondary gross enrolment ratio 0. 33

•ICT access is measured by fixed-telephone subscriptions per 100 inhabitants, mobile -cellular subscriptions per 100 inhabitants

international Internet bandwidth per Internet user, percentage of households with a computer and percentage of households with Internet access

•ICT use is measured by percentage of individuals using the Internet, fixed wired)- broadband Internet subscriptions

per 100 inhabitants and wireless -broadband subscriptions per 100 inhabitants •ICT skills are approximated by adult

literacy rate, secondary gross enrolment ratio and tertiary gross enrolment ratio The values of the sub-indices were calculated

first by normalizing the indicators included in each sub-index in order to obtain the same unit

of measurement. The reference values applied in the normalization were discussed above. The sub-index value was calculated by taking the

For computation of the final index, the ICT access and ICT use sub-indices were given 40 per cent

the data were transformed first to a logarithmic (log) scale The ideal value of 787†260 bit/s per Internet user is equivalent to 5. 90 if transformed to a log scale

Source: ITU DENMARK Indicators 2013 ICT access Ideal value *a Fixed-telephone subscriptions per 100 inhabitants 60 37.4

b Mobile-cellular telephone subscriptions per 100 inhabitants 120 127.5 c International Internet bandwidth per Internet user**787'260 261'221

d Percentage of households with a computer 100 93.1 e Percentage of households with Internet access 100 92.7

ICT use f Percentage of individuals using the Internet 100 94.6 g Fixed (wired)- broadband Internet subscriptions per 100 inhabitants 60 40.2

h Wireless-broadband subscriptions per 100 inhabitants 100 107.5 ICT skills i Adult literary rate 100 99.0

j Secondary gross enrolment ratio 100 124.7 k Tertiary gross enrolment ratio 100 79.6 Normalized values Formula Weight

ICT access z1 Fixed-telephone subscriptions per 100 inhabitants a/60 0. 20 0. 62

z2 Mobile-cellular telephone subscriptions per 100 inhabitants b/120 0. 20 1. 00 z3 International Internet bandwidth per Internet user log (c)/ 5. 90 0. 20 0. 92

z4 Percentage of households with a computer d/100 0. 20 0. 93 z5 Percentage of households with Internet access e/100 0. 20 0. 93

ICT use z6 Percentage of individuals using the Internet f/100 0. 33 0. 95

z7 Fixed (wired)- broadband Internet subscriptions per 100 inhabitants g/60 0. 33 0. 67

z8 Wireless-broadband subscriptions per 100 inhabitants h/100 0. 33 1. 00 ICT skills

z9 Adult literary rate i/100 0. 33 1. 00 z10 Secondary gross enrolment ratio j/100 0. 33 0. 80

z11 Tertiary gross enrolment ratio k/100 0. 33 0. 99 Sub-indices Formula Weight

ICT access sub-index (L) y1+y2+y3+y4+y5 0. 40 0. 88 y1 Fixed-telephone subsriptions per 100 inhabitants z1*.

*20 0. 12 y2 Mobile-cellular telephone subscriptions per 100 inhabitants z2*.*20 0. 20 y3 International Internet bandwidth per Internet user z3*.

*20 0. 18 y4 Percentage of households with a computer z4*.*20 0. 19 y5 Percentage of households with Internet access z5*.

*20 0. 19 ICT use sub-index (M) y6+y7+y8 0. 40 0. 87

y6 Percentage of individuals using the Internet z6*.*33 0. 32 y7 Fixed (wired)- broadband Internet subscriptions per 100 inhabitants z7*.

*33 0. 22 y8 Wireless-broadband subscriptions per 100 inhabitants z8*.*33 0. 33 ICT skills sub-index (N) y9+y10+y11 0. 20 0. 93

y9 Adult literary rate z9*.*33 0. 33 y10 Secondary gross enrolment ratio z10*.*33 0. 27

y11 Tertiary gross enrolment ratio z11*.*33 0. 33 IDI ICT Development Index((L*.40)+(M*.40)+(N*.20))*10 8. 86

in the computation of the index, including the selection of individual indicators, the imputation of missing values and the normalization

weighting and aggregation of the data Each of the processes or combination of processes affects the IDI value.

computations using different methodologies This confirms the results conveyed by the IDI 229 Measuring the Information Society Report 2014

1 Principal component analysis was used to examine the underlying nature of the data. A more detailed description of the

2 More information about the indicators is available in the ITU †Handbook for the collection of administrative data on

telecommunications/ICT€ 2011, see ITU 2011 and the ITU â€oemanual for Measuring ICT Access and Use by Households and

EGH. pdf). As some of the data used in the calculation of the IDI were collected before that meeting,

however, the data may not necessarily reflect these revisions 4 More information about the indicators is available in the ITU â€oehandbook for the collection of administrative data on

telecommunications/ICT€ †2011, see ITU 2011b and the ITU â€oemanual for Measuring ICT Access and Use by Households and

Individualsâ€, see ITU 2014 5 UIS †Education Indicators: Technical Guidelinesâ€, see http://www. uis. unesco. org/ev. php?

ID=5202 201&id2=DO TOPIC 6 See OECD and European commission (2008 7 For more details, see Annex 1 to ITU (2009

Annex 2. ICT price data methodology 1. Price data collection and sources The price data presented in this report were

collected in the fourth quarter of 2013. The data were collected through the ITU ICT Price Basket

questionnaire, which was sent to the administrations and statistical contacts of all 193 ITU Member States in October 2013.

Through the questionnaire contacts were requested to provide 2013 data for fixed-telephone, mobile-cellular, fixed-broadband

and mobile-broadband prices; the 2011 and 2012 prices were included for reference, where available. For those countries that did not reply

prices were collected directly from operators†websites and/or through direct correspondence Prices were collected from the operator with the

largest market share, as measured by the number of subscriptions. Insofar as, for many countries, it

telecommunication operator. In some cases especially when prices were advertised not clearly or were described only in the local language

the fixed-telephone, mobile-cellular and fixed-broadband sub-baskets. The IPB is the value calculated from the sum of the price

The collection of price data from ITU Member States and the methodology applied for the IPB was agreed upon by the ITU Expert

Group on Telecommunication/ICT Indicators EGTI) 1 and endorsed by the eighth World Telecommunication/ICT Indicators meeting held

in November 2010 in Geneva, Switzerland The fixed-telephone sub-basket The fixed-telephone sub-basket refers to the

monthly price charged for subscribing to the public switched telephone network (PSTN plus the cost of 30 three-minute local calls to

the same (fixed) network (15 peak and 15 off -peak calls. It is calculated as a percentage of a

country†s average monthly GNI per capita, and also presented in USD and PPP$ Annex 2. ICT price data methodology

232 The fixed-telephone sub-basket does not take into consideration the onetime connection charge.

This choice has been made in order to improve comparability with the other sub -baskets, which include only recurring monthly

of the fixed-telephone sub-basket The cost of a three-minute local call refers to

equipment (i e. not from a public telephone It thus refers to the amount the subscriber

take into consideration the price of a telephone set (see Annex Box 2. 1 The ICT Price Basket includes a sub-basket for

fixed telephony because fixed-telephone access remains an important access technology in its own right in a large number of countries

Additionally, the conventional fixed-telephone line is used not only for dial-up Internet access, but also as a basis for upgrading to

DSL broadband technology, which in 2013 still accounted for the majority of all fixed -broadband subscriptions.

Internet access still remains the only Internet access available to some people in developing countries. Since the IPB does not include dial

-up (but only broadband) Internet prices, and since dial-up Internet access requires users to subscribe to a fixed-telephone line, the fixed

-telephone sub-basket can be considered as an indication for the price of dial-up Internet

access The mobile-cellular sub-basket The mobile-cellular sub-basket refers to the price of a standard basket of mobile monthly

usage for 30 outgoing calls per month on-net, off-net to a fixed line and for peak

and off-peak times) in predetermined ratios plus 100 SMS messages. It is calculated as a percentage of a country†s average monthly

GNI per capita, and also presented in USD and PPP$. The mobile-cellular sub-basket is

based on prepaid prices, although postpaid prices are used for countries where prepaid subscriptions make up less than 2 per cent of

all mobile-cellular subscriptions The mobile-cellular sub-basket is based largely on, but does not entirely follow, the 2009

methodology of the OECD low-user basket which is the entry-level basket with the smallest

number of calls included (OECD, 2010b. Unlike the 2009 OECD methodology, which is based on the prices of the two largest mobile operators

account calls to voicemail (which in the OECD basket represent 4 per cent of all calls), nor non

for a SIM CARD. The basket gives the price of a standard basket of mobile monthly usage in USD

Rules applied in collecting fixed-telephone prices 1. The prices of the operator with the largest market share (measured by the number of fixed-telephone subscriptions) should be

used 2. Prices should be collected in national currency, including taxes. 3 3. Only residential, single user prices should be collected.

If prices vary between different regions of the country, the prices applying to the largest city (in terms of the population) should be provided.

4. From all fixed-telephone plans meeting the above-mentioned criteria, the cheapest postpaid plan on the basis of 30 local calls (15

6. The same price plan should be used for collecting all the data specified. For example, if a given Plan A is selected for the fixed

television reception, over their networks. They often bundle these offers into a single subscription. This can present a challenge

for data collection, since it may not be possible to isolate the prices for one service. It is preferable to use prices for a specific

to low-income users, who might not have a regular income and will thus not qualify for a

the cheapest option available, the mobile -cellular sub-basket therefore corresponds to a basic, representative (low-usage) package

Annex 2. ICT price data methodology 234 Annex Table 2. 1: OECD mobile-cellular low-user call distribution (2009 methodology

Note: N/A: Not applicable Source: ITU, based on OECD (2010b To fixed On-net Off-net TOTAL

data usage of (a minimum of) 1 Gigabyte GB). ) For plans that limit the monthly amount

of data transferred by including data volume caps below 1 GB, the cost for the additional

bytes is added to the sub-basket. The minimum speed of a broadband connection is 256 kbit/s

will be taken into consideration for the calculation of the mobile-cellular sub-basket 4. If per-minute prices are advertised only in internal units rather than in national currency,

then this is taken into consideration in the formula for the mobile-cellular sub-basket, based

the mobile-cellular sub-basket formula will be calculated on the basis of 30 calls or 50.9 minutes.

Annex 2. ICT price data methodology 236 Annex Box 2. 3: Rules applied in collecting fixed-broadband Internet prices

1. The prices of the operator with the largest market share (measured by the number of subscriptions) should be used

3. Only residential, single-user prices are collected. If prices vary between different regions of the country,

and an advertised download speed of at least 256 kbit/s should be selected. If there is a price distinction between residential and

7. The same price plan should be used for collecting all the data specified. For example, if a given Plan A is selected for the fixed

the volume of data that can be downloaded, etc 8. Prices should be collected for regular (non-promotional) plan

television reception over their networks. They often bundle these offers into a single subscription. This can present a challenge

for price data collection, since it may not be possible to isolate the prices for one service.

GB of data volume. If providers set a limit of less than 1 GB on the amount of data that can be

transferred within a month, then the price per additional byte is added to the monthly price

so as to calculate the cost of 1 GB of data per month. Preference should be given to the most

widely used fixed (wired)- broadband technology DSL, cable, etc..The sub-basket does not include installation charges, modem prices or

3. Mobile-broadband prices In 2012, for the first time, ITU collected mobile -broadband prices through its annual ICT Price

-broadband price data from ITU Member States was agreed upon by the ITU Expert Group on

Telecommunication/ICT Indicators (EGTI) 9 in 2012, and revised in 2013 by EGTI in view of the

lessons learned from the first data collection exercise. The revised methodology was endorsed by the eleventh World Telecommunication/ICT

Indicators Symposium held in December 2013 in Mexico city, Mexico To capture the price of different data packages

covering prepaid and postpaid services, and supported by different devices (handset and computer), mobile-broadband prices were

collected for two different data thresholds, based on a set of rules (see Annex Box 2. 4

For plans that were limited in terms of validity less than 30 days), the price of the additional days was calculated

of data allowance (or validity. The customer i) continues to use the service and pays an

offers that include the minimum amount of data for each respective mobile-broadband plan The guiding idea is to base each plan on what

customers would and could purchase given the data allowance and validity of each respective plan

Annex 2. ICT price data methodology 238 Annex Box 2. 4: Rules applied in collecting mobile-broadband prices11

1. Prices should be collected based on one of the following technologies: UMTS, HSDPA+/HSDPA, CDMA2000, and IEEE 802. 16e

Prices applying to Wifi or hotspots should be excluded 2. Prices should be collected in national currency,

including taxes 3. Only residential, single-user prices should be collected. If prices vary between different regions of the country,

the prices applying to the largest city (in terms of population) or to the capital city should be provided

and b) computer-based mobile-broadband subscriptions 5. Mobile-broadband prices should be collected from the operator with the largest market share in the country, measured by the

number of mobile-broadband subscriptions. If this information is not available, mobile-broadband prices should be collected

from the mobile-cellular operator with the largest market share (measured by the number of mobile-cellular subscriptions) in

the country 6. Different operators can be chosen, for a different mobile-broadband service, if: a) there are differing market leaders for specific

segments (postpaid, prepaid, computer-based, handset-based; b) there is no offer available for a specific sub-basket

7. Prices should be collected for prepaid and postpaid services, for both handset and computer-based plans.

If there are several plans, the plan satisfying the indicated data volume requirement should be used 8. Where operators propose different commitment periods for postpaid mobile-broadband plans,

the 12-month plan (or the closest to this commitment period) should be selected. A note should be added in case only longer commitment periods are offered

9. Price data should be collected for the cheapest plan with a data volume allowance of a minimum of

i. 1 GB for USB/dongle (computer-based) subscriptions ii. 500 MB for handset-based subscriptions

The selected plan should not be the one with the cap closest to 500 MB or 1 GB,

Data volumes should refer to both upload and download data volumes. If prices are linked to †hours of use†and not to data volumes, this information should be added in a separate note

Note: ITU will most likely not be able to include these cases in a comparison

10. A validity period of 30 days should be chosen. If this is not available, 15 days should be used.

Preference should be given to packages (including a certain data volume. Pay-as-you-go offers should be used

since most often there are limits in the data volumes, either applied by throttling (limiting the speed)

services besides mobile-broadband access, these should be specified in a note 15. Prices refer to a regular (non-promotional) plan

and exclude promotional offers and limited discounts or special user groups for example, existing clients. Special prices that apply to a certain type of phone (iphone/Blackberry,

ipad) should be excluded Allowances during the night are included not Source: ITU 239 Measuring the Information Society Report 2014

1 The Expert Group on Telecommunication/ICT Indicators (EGTI) was created in May 2009 with the mandate to revise the list

of ITU supply-side indicators (i e. data collected from operators), as well as to discuss outstanding methodological issues and

new indicators. EGTI is open to all ITU members and experts in the field of ICT statistics and data collection.

It works through an online discussion forum (http://www. itu. int/ITU-D/ict/Expertgroup/default. asp) and face-to-face meetings.

to the World Telecommunication/ICT Indicators Symposium (WTIS 2 In some cases, it is not clear

5 On-net refers to a call made to the same mobile network, while off-net and fixed-line refer to calls made to other

competing) mobile networks and to a fixed-telephone line, respectively 6 In some cases, it is not clear

8 Data for fixed-telephone, mobile-cellular and fixed-broadband have been collected since 2008 through the ITU ICT Price

10 Some operators throttle speeds after the data allowance included in the base package has been reached.

11 These rules were presented to the Expert Group on Telecommunication/ICT Indicators (EGTI) in September 2012.

to measure mobile-broadband prices were endorsed by the eleventh World Telecommunication/ICT Indicators Symposium held in December 2013 in Mexico city, Mexico

Fixed-telephone subscriptions per 100 inhabitants Mobile-cellular subscriptions per 100 inhabitants International Internet bandwidth

Bit/s per Internet user Percentage of households with computer Percentage of households with Internet access Economy 2012 2013 2012 2013 2012 2013 2012 2013 2012 2013

1 Afghanistan 0. 3 0. 3 65.5 70.0 1†229 2†774 2. 3 2. 5 1. 9 2. 1

2 Albania 9. 9 8. 9 110.7 116.2 17'358 20'974 20.0 21.7 20.5 24.5

Fixed-telephone subscriptions per 100 inhabitants Mobile-cellular subscriptions per 100 inhabitants International Internet bandwidth

Bit/s per Internet user Percentage of households with computer Percentage of households with Internet access Economy 2012 2013 2012 2013 2012 2013 2012 2013 2012 2013

84 Lao P. D. R. 6. 8 10 10.0 13 64.7 15 66.2 9'397 9 10'636 8. 7

9. 6 5. 1 5. 1 85 Latvia 24.3 23.4 112.1 136.6 59'738 68'069 69.5 71.7 68.7 71.6

Data in italics refer to ITU estimates. For further notes, see p. 248 Source: ITU World Telecommunication/ICT Indicators database

Annex 3. Statistical tables of indicators used to compute de IDI 244 Percentage of individuals using the Internet

Fixed (wired)- broadband subscriptions per 100 inhabitants Wireless-broadband subscriptions per 100 inhabitants Economy 2012 2013 2012 2013 2012 2013

using the Internet Fixed (wired)- broadband subscriptions per 100 inhabitants Wireless-broadband subscriptions per 100 inhabitants

Data in italics refer to ITU estimates. For further notes, see p. 248 Source: ITU World Telecommunication/ICT Indicators database

Annex 3. Statistical tables of indicators used to compute de IDI 246 Skills indicators Gross enrolment ratio Adult

Data in italics refer to ITU estimates Source: ITU World Telecommunication/ICT Indicators database Annex 3. Statistical tables of indicators used to compute de IDI

248 Access indicators Fixed-telephone subscriptions per 100 inhabitants, 2012 1) Incl. 524 958 WLL subscriptions. 2) Incl. payphone, excl.

VOIP. 3) Incl. ISDN channels measured in ISDNB channels equivalents 4) Incl. Voip. 5) Bhutan Telecom is the only service provider as of now in Bhutan. 6) By December 7) Fixed and WLL. 8) Total access

lines. 9) First trimester 2012.5 431 registered subscriptions. 10) Estimate. 11) Incl. public payphones. 12) Decrease is caused by

change in tariff policy of the biggest WLL operator. 13) Data excluding own (NRA) consumption. 14) Excl. voice-over-IP (Voip

Telecom italia access lines, ULL, Virtual ULL, Naked DSL Wholesale line Rental, Fiber, Public Telephony. 17) The number of fixed public payphones is as of March 2012.18) Fixed Wireless

Local Loop. 19) Including digital lines. Without ISDN channels. 20) Excl. ISDN channels and fixed wireless subscriptions. 21

Incl. inactive subscriptions. 22) Preliminary. 23) Refers to active Fixed Wired/Wireless lines. 24) POTS, ISDN BRA & ISDN PRA. 25

Decrease due to cleaning out of inactive accounts. 26) Excluding fixed wireless. 27) Excl. internal lines and WLR of incumbent

Fixed-telephone subscriptions per 100 inhabitants, 2013 1) Incl. 420 000 WLL subscriptions. 2) Incl. payphone, excl.

VOIP. 3) Incl. ISDN channels measured in ISDNB channels equivalents. 4) Incl. Voip. 5) Estimate. 6) Bhutan Telecom is the only service provider as of now in Bhutan. 7) By December

2013.8) Excl. voice-over-IP (Voip) subscriptions, fixed wireless local loop (WLL) subscriptions, ISDN voice-channel equivalents

Telecom italia access lines, ULL, Virtual ULL, Naked DSL, Wholesale line Rental, Fiber, Public Telephony 12) The number of fixed public payphones is as of March 2013.13) Preliminary. 14) Incl. digital lines.

separate ISDN channels (abonnements au tã lã phone fixe. 15) Excl. ISDN channels and fixed wireless subscriptions. 16

Break in comparability. Only active subscriptions. Inactive subscriptions are: 45 609.17) February 2013 NTA MIS. 18) Based

on 2013q3 data. 19) Refers to active Fixed Wired/Wireless lines. 20) Per June 2013.21) Operators†data. 22) Residential:

Definitive data (annual report) may change because quarterly reports use a smaller sample of operators than annual report. 24) Fixed and fixed-wireless subscriptions. 25) Excl. internal lines and WLR of

incumbent. Data for the third quarter of 2013 Mobile-cellular subscriptions per 100 inhabitants, 2012

1) Numbers are down due to data cleanse. 2) ACMA Communications Report 2011-12.2) incl. payphone, excl.

Voip. 3) Active subscriptions. 4) Bhutan Telecom and Tashi Cell are the only two service providers in Bhutan. 5) Activity criteria:

voice or data communication in the last month. 6 december 2012.7) Total number of subscriptions (including non-active:

2 082 589.8 Excl. 2 720 698 prepaid cards that are used to provide Travel SIM/World Mobile service. 9) Excl. data-only SIM CARDS and M2m

cards. 10) Decrease due to merge of second and third operators of the mobile market. 11) Incl. fixed wireless local loop (WLL

Including PHS and data cards undividable. 15) Decrease was due to registration of SIMS. 16) Figure obtained from all five mobile (GSM & CDMA) operators

currently providing service in the country. 17) Incl. inactive subscriptions. 18) Preliminary. 19) Active subscriptions (87.85

Incl. inactive. 22) Incl. data-only subscriptions not possible to disaggregate the information at this point.

excl. 226 827 M2m subscriptions 24) Incl. active (in the last 6 months) prepaid accounts. 25) Registered SIM CARDS (incl. inactive:

261 887 751.26) Break in comparability: from this year only counting prepaid subscriptions used in the last 90 days. 27) No differentiation between

active and non-active subscriptions. 28) Incl. data dedicated subscriptions. 29) Decrease due to the closing of MTS

1) Preliminary. 2) Active subscriptions. 3) Bhutan Telecom and Tashi Cell are the two service providers in Bhutan. 4) Activity

voice or data communication in the last month. 5) By December 2013.6) Incl. all mobile cellular subscriptions that

but excl. mobile data subscriptions (via data cards, USB modems and M2m cards). 7) Estimate

8) Incl. data-only subscriptions. 9) Data based on NRA estimates. Excl. data-only SIM CARDS and M2m cards. 10) Incl. fixed

wireless local loop (WLL) subscriptions. 11) Incl. PHS and data cards, undividable. 12) Active subscriptions (after clearing

simcards that became inactive. 13) Figure obtained from all four mobile (GSM & CDMA) operators currently providing

service in the country. 14) Break in comparability. Incl. only active subscriptions. 15) Active subscriptions (83.44%total

16 february 2013 NTA MIS. 17) Based on 2013q3 data. 18) Measured using subscriptions active in the last 90 days. 19) Per

June 2013.20) Excl. 463 646 M2m subscriptions. 21) Incl. active (in the last 6 months) prepaid accounts. 22) Registered SIM

Definitive data (annual report) may change because quarterly reports use a smaller sample of operators than annual report. 26) Data for the fourth quarter of 2013.27) Incl. data dedicated subscriptions

28) Reduction is due to implementation of sim cards registration International Internet bandwidth Bit/s per Internet user, 2012

1) Refers to a survey conducted with the following companies: Global crossing, TIWS, Embratel e Globenet. 2) This is from

one ISP only. No response received from other ISPS. 3) Symmetric. 4) Total installed capacity. 5) Purchased capacity. 6) By

MOT. 11) Data obtained from nine service operators. 12 may 2012 purchased capacity. Lit capacity: 43 096 471 Mbit/s. 13) Incoming capacity;

weekly incoming capacity, averaged over 4 weeks in December 14) Preliminary. 15) SLT Data. 16) Refers to the total capacity

International Internet bandwidth Bit/s per Internet user, 2013 1) Purchased capacity. 2) As at December 2013.3) Total installed capacity.

4 may be revised with comprehensive data from mobile broadbnad providers. 5) Estimate. 6 june. 7) Sum of incoming capacity of all ISPS in the country. 8) Activated external

capacity. 9) By September 2013.10) Data obtained from eight service operators. 11) 1st april 2013 purchased capacity.

Percentage of households with computer, 2012 1) Estimated based on 2011 proportion of households with a computer and using annual growth rate of 3%.2) Preliminary

3) Refers to PC or laptop. 4) Data correspond to dwellings (not households). 5) Ghana Living Standards Survey 2012/2013

The estimate is based on households who own and/or have access to a desktop, laptop or TABLET PCS.

Sample weights have been applied. 6) Personal computer included desktop computer, laptop/notebook/netbook/tablet and palm top /Personal digital assistant (PDA),

but excluded digital diary and electronic dictionary. 7) Estimate. 8) From Household Socio -Economic survey-2012.9) Census data. 10) Computer includes the number of personal computer, Notebook, and PDA

11) U s. Census bureau, table 4: http://www. census. gov/hhes/computer/publications/2012. html Percentage of households with computer, 2013

1) Labour force Survey 2013.2) Cambodia Inter-censal Population Survey. 3) Refers to PC, laptop or a tablet. 4) Data

correspond to dwellings (not households. 5) Ghana Living Standards Survey 2012/2013. The estimate is based on

households who own and/or have access to a desktop, laptop or TABLET PCS. Sample weights have been applied

6) Preliminary. 7) Estimated Percentage of households with Internet access, 2012 1) Estimated based on 2011 proportion of households with internet and using estimated annual growth rate of 2. 8

%2) Preliminary. 3) Data correspond to dwellings (not households. 4) Ghana Living Standards Survey 2012/2013.

The estimate is based on households who own and/or have access to internet. Sample weights have been applied.

Not restricted to access at home. 5) Incl. desktop computer, laptop/notebook/netbook/tablet, but excluded palm top/Personal Digital

Assistant (PDA) and other devices for Internet connection (e g. smartphone, game console and e-book reader.

6) Accessing from personal computers. 7) Estimate based on 2011 Census Population Household Projection Estimates. 8) From Household

Socioeconomic survey-2012.9) Break in comparability: Refers to access at home, on cell phone or other mobile device and

via mobile modem. 10) Census data. 11) Excl. households which didn†t know type of internet access 172 346 households

12) U s. Census bureau, table 3: http://www. census. gov/hhes/computer/publications/2012. htm Percentage of households with Internet access, 2013

1) Labour force Survey 2013.2) Corresponds to all type of internet connections. 3) Data correspond to dwellings (not

households). ) 4) Ghana Living Standards Survey 2012/2013. The estimate is based on households who own

and/or have access to internet. Sample weights have been applied. Not restricted to access at home. 5) Included desktop computer

laptop/notebook/netbook/tablet, but excluded palm top/Personal digital assistant (PDA) and other devices for Internet connection (e g. smartphone, game console and e-book reader

Use indicators Percentage of individuals using the Internet, 2012 1) 15 years and older. Last 12 months. 2) Users in the last 3 months. 3) Estimated based on 2011 Residential consumer survey

result and TRA analysis of the growth. 4) Individuals aged 16 and over. 5) Preliminary. 6) In the last 3 months.

Population 10 +7) Residents of Canada 16 years of age or older excluding: Residents of the Yukon, Northwest territories and Nunavut, Inmates

of Institutions, Persons living on Indian Reserves, and Full time members of the Canadian Forces. 8) Estimated based on surveysâ€

results. Population age 5+.9) In the last 3 months. Population 5+.10) 12+years. 11) Population 5+.Direct response from

individuals 15 years and above. 12) The methodology depends basically on the number of internet users using hard indicators

instead of data survey. 13) Ghana Living Standards Survey 2012/2013. The estimate is based on weighting households who use

internet by the household size over the total estimated population. Sample weights have also been applied.

The question was asked at household level. 14) All persons aged 10 and over. 15) Age 20+.

computer and Internet usage is 3 months only. 28) U s. Census bureau, Table 2: http://www. census. gov/hhes/computer

/publications/2012. htm Percentage of individuals using the Internet, 2013 1) Individuals aged 15 years and over. 2) Population age 16-74.3) Labour force Survey 2013.4) Individuals aged 6 and over

5) Cambodia Inter-censal Population Survey 2013.6) Permanent residents at the age of 6 or above.

The estimate is based on weighting households who use internet by the household size over the total estimated population

population living in workers†camps. 18) Population age 10+using internet in the last 3 months. 19) Individuals aged 15 to 72

1) Internet Activity Survey, June 2) Incl. fixed wireless broadband. 3) Fixed broadband in Bhutan is provided via ADSL/DSL

based on the data provided by 89.1%of operators 6) The figure is corrected. The previous figure was 1†636†700.7) Only ADSL, excl. cable modem. 8) Speeds greater than, or

equal to, 512 Kbps. 9) By December 2012.10) Only ETL and LTC. 11) Preliminary. 12) Full VDSL. 13) Speeds equal to or greater

than 144 kbit/s. 14) Operators data/ictqatar estimate. 15) Incl. subscriptions at downstream speeds equal to, or greater than

16) Q3. 17) Excl. 3203 Wimax subscriptions. 18) Excl. corporate connections. 19) Data reflect subscriptions with associated transfer rates exceeding 200

kbps in at least one direction, consistent with the reporting threshold the FCC adopted in 2000.20) Incl.

ADSL and FTTH Fixed (wired)- broadband subscriptions per 100 inhabitants, 2013 1 november 2013.2) Preliminary. 3) Internet Activity Survey, June 2013.4) Fixed broadband provided through ADSL/DSL

and Fiber links. 5) Estimate, no specific data collected for â ¥ 256 kbit/s. 6) CRC estimation as of 31.12.2013.7) Estimate. 8) Data

based on NRA estimates. 9) Only ADSL, excl. cable modem. 10) Speeds greater than, or equal to, 512 Kbps. 11 december

2013. These are the subscriptions with the minimum download speed of 512 kbps. This is as per the revised definition of

Broadband in India with effect from 18th july 2013.12) December 13) ADSL and Leased lines. 14) Based on 2013q3 data

15) Per June 2013.16) Operators†data. 17) Incl. subscriptions at downstream speeds equal to, or greater than, 144 kbit/s (the

number of subscriptions that are included in the 144-256 range is insignificant. 18) Q4 report.

Definitive data (annual report may change because quarterly reports use a smaller sample of operators than annual report. 19) Estimate.

2013.20) Excl. 3175 Wimax subscriptions. 21) Excl. corporate connections. 22) 2013 data is an estimate as of June 30, 2013

Data reflect subscriptions with associated transfer rates exceeding 200 kbps in at least one direction, consistent with the

1) Only fixed Wimax subscriptions. 2) Internet Activity Survey, June 3) Break in comparability: including all categories of

mobile broadband. 4) Total number of EDGE/GPRS subscribers: 97 520.5) Break in comparability: from this year incl.

USB modems and dongles, mobile broadband(>256kbps at least in one direction up to HSPA+),Wimax, Pre Wimax, SID and

satellite. 6) Change in definition, break in comparability. 7) High use of mobile phones to access the internet. 8) Incl.

Home Box and RLANS. 9) Break in comparability, from this year incl. prepaid mobile-broadband subscriptions. 10) Incl. subscriptions

to Wifi hotspots. 11) Methodology changed from ability to have mobile broadband to actual mobile broadband usage

12) Satellite, BWA and active mobile subscriptions. 13) Estimate based on partial SIT data and ITU estimates. 14) Speeds

greater than, or equal to, 512 Kbps. 15) Rightel (Tamin Telecom) has been given license to operate 3g services and started

services from February 2011 (http://www. rightel. ir/./Data refer to the sum of fixed wireless broadband and active mobile

-broadband subscriptions. 16) Incl. mobile broadband and Wimax. 17) Estimate. 18) ETL and LTC. 19) Incl. narrowband

connections. 20) Drop in mobile-broadband subscriptions because in 2011 the operator offered free Internet access for a

limited amount of time so that many people used the free service. 21) Preliminary. 22) Mobile broadband only.

Fixed wireless and satellite exist but data are not available. 23) Operators data/ictqatar estimate. 24) Refers to active mobile-broadband

subscriptions only. 25.dec 26) Incl. 4125165 active mobile-broadband subscriptions plus 3203 Wimax subscriptions. Excl satellite subscriptions. 27) Excl. satellite and fixed wireless. 28) Incl. mobile subscriptions with potential access

Wireless-broadband subscriptions per 100 inhabitants, 2013 1) Only fixed Wimax subscriptions. 2) Preliminary. 3) Internet Activity Survey, June 2013.4) Total number of EDGE/GPRS

subscribers is 112 898.5) Incl. LTE subscriptions from ENTEL. 6) Change in definition, break in comparability. 7) 2012 figures

Still auditing the 2013 figures. 8) Incl. WCDMA, LTE, dedicated mobile-broadband and fixed wireless. 9) CRC estimation as of

31.12.2013. speeds equal to or greater than 144 kbit/s/.10) Estimate. 11) Estimate. Incl. subscriptions to Wifi hotspots

12) Satellite, BWA and active mobile subscriptions. 13) Incl. VSAT. 14) Speeds greater than, or equal to, 512 Kbps

15) subscriptions with minimum download speed of 512 kbps. This is as per the revised definition of Broadband in India with

effect from 18th july 2013.16) Data refer to the sum of fixed wireless broadband and active mobile-broadband subscriptions

17) 2013 data is an estimate as of June 30, 2013.18) Incl. mobile broadband and Wimax. 19) Estimate based on 1. Standard

mobile subscriptions using data services 2. Dedicated data subscriptions 3. Add on data packages. 20) Based on 2013q3

data. 21) Per June 2013.22) Mobile broadband only. Fixed wireless and satellite exist but data are not available. 23) Operatorsâ€

data. 24) As at Dec 2013.25) Q4 report. Definitive data (annual report) may change because quarterly reports use a smaller

sample of operators than annual report. 26) Wireless Broadband services are not being offered in St vincent as yet.

We anticipate that Mobile broadband and terrestrial fixed broadband services would be in place by the end of 2014.27) OFCOM

estimate. 28) Includes: active mobile-broadband subscriptions plus 3175 Wimax International Telecommunication Union Telecommunication Development Bureau

Place des Nations CH-1211 Geneva 20 Switzerland www. itu. int ISBN 978-92-61-14661-0 SAP id

9 7 8 9 2 6 1 1 5 2 9 1 8 3 9 4 6 4

Price: 86 CHF Printed in Switzerland Geneva, 2014 Photo credits: Shutterstock Measuring the Information Society Report

2014 M ea su rin g th e In fo rm at io n So

ci et y Re po rt 2 01 4 Measuringthe Information Society Report2014 Foreword Acknowledgements

Table of contents Chapter 1. Recent information society developments 1. 1 Introduction 1. 2 The voice market

1. 3 The broadband market and Internet access 1. 4 Revenue and investment in the telecommunication sector

1. 5 Use of ICTS 1. 6 Emerging ICT measurement issues Chapter 2. The ICT Development Index

2. 3 Monitoring the digital divide: Developed, developing and least connected countries 2. 4 Geography, population size, economic development and the IDI

4. 2 Fixed-telephone and mobile-cellular prices 4. 3 Broadband prices 4. 4 Income inequality and broadband prices

4. 5 The impact of competition and regulation on telecommunication prices Chapter 5. The role of big data for ICT monitoring and for development

5. 1 Introduction 5. 2 Big data sources, trends and analytics 5. 3 Telecommunication data and their potential for big data analytics

5. 4 Big data from mobile telecommunications for development and for better monitoring 5. 5 Challenges and the way forward

List of references Annex 1. ICT Development Index (IDI) methodology Annex 2. ICT price data methodology

Annex 3. Statistical tables of indicators used to compute the IDI

< Back - Next >

Overtext Web Module V3.0 Alpha
Copyright Semantic-Knowledge, 1994-2011