A partnership with academia

Building knowledge for trade and development

Vi Digital Library - Text Preview

Information Economy Report 2013: The Cloud Economy and Developing Countries

Report by UNCTAD, 2013

Download original document (English)

The Information Economy Report 2013 marks the first time the United Nations is examining the economic potential of cloud computing for low- and middle-income countries, where rates of adoption are currently low. With governments, businesses and other organizations in the developing world considering whether to migrate some or all of their data and activities to the cloud, this publication is especially timely.

In
fo


r
m


a
tIo


n
Ec


o
n


o
m


y
r


Ep
o


r
t 2


0
1


3
UN


CTA
D


UNITED NATIONS


The Cloud Economy and Developing Countries


InformatIon
Economy


rEport 2013


U n i t e d n at i o n s C o n f e r e n C e o n t r a d e a n d d e v e l o p m e n t


EMBARGO
The contents of this Report must not be


quoted or summarized in the print,
broadcast or electronic media before


3 December 2013, 17:00 hours GMT


13-51665
ISBN 978-92-1-112869-7




The Cloud Economy and Developing Countries


INFORMATION
ECONOMY


REPORT 2013


U N I T E D N AT I O N S C O N F E R E N C E O N T R A D E A N D D E V E L O P M E N T


New York and Geneva 2013




ii INFORMATION ECONOMY REPORT 2013


NOTE


Within the UNCTAD Division on Technology and Logistics, the ICT Analysis Section carries out policy-oriented
analytical work on the development implications of information and communication technologies (ICTs). It is
responsible for the preparation of the Information Economy Report. The ICT Analysis Section promotes international
dialogue on issues related to ICTs for development, and contributes to building developing countries’ capacities
to measure the information economy and to design and implement relevant policies and legal frameworks.


In this Report, the terms country/economy refer, as appropriate, to territories or areas. The designations
employed and the presentation of the material do not imply the expression of any opinion whatsoever on the part
of the Secretariat of the United Nations concerning the legal status of any country, territory, city or area or of its
authorities, or concerning the delimitation of its frontiers or boundaries. In addition, the designations of country
groups are intended solely for statistical or analytical convenience and do not necessarily express a judgement
about the stage of development reached by a particular country or area in the development process. The major
country groupings used in this Report follow the classification of the United Nations Statistical Office. These are:


Developed countries: the member countries of the Organization for Economic Cooperation and Development
(OECD) (other than Chile, Mexico, the Republic of Korea and Turkey), plus the new European Union member
countries that are not OECD members (Bulgaria, Cyprus, Latvia, Lithuania, Malta and Romania), plus Andorra,
Liechtenstein, Monaco and San Marino. Countries with economies in transition: South-East Europe and the
Commonwealth of Independent States. Developing economies: in general, all the economies that are not specified
above. For statistical purposes, the data for China do not include those for Hong Kong Special Administrative
Region (Hong Kong, China), Macao Special Administrative Region (Macao, China), or Taiwan Province of China.


Reference to companies and their activities should not be construed as an endorsement by UNCTAD of those
companies or their activities.


The following symbols have been used in the tables:


Two dots (..) indicate that data are not available or are not separately reported. Rows in tables have been
omitted in those cases where no data are available for any of the elements in the row;


A dash (-) indicates that the item is equal to zero or its value is negligible;


A blank in a table indicates that the item is not applicable, unless otherwise indicated;


A slash (/) between dates representing years, for example, 1994/95, indicates a financial year;


Use of an en dash (–) between dates representing years, for example, 1994–1995, signifies the full period
involved, including the beginning and end years;


Reference to “dollars” ($) means United States of America dollars, unless otherwise indicated;


Annual rates of growth or change, unless otherwise stated, refer to annual compound rates;


Details and percentages in tables do not necessarily add up to the totals because of rounding.


The material contained in this study may be freely quoted with appropriate acknowledgement.


UNITED NATIONS PUBLICATION


UNCTAD/IER/2013


Sales No. E.13.II.D.6


ISSN 2075-4396


ISBN 978-92-1-112869-7


e-ISBN 978-92-1-054154-1


Copyright © United Nations, 2013


All rights reserved. Printed in Switzerland




iiiPREFACE


PREFACE


Innovation in the realm of information technology continues its rapid pace, with cloud computing
representing one of the latest advances. Significant improvements in the capacity to process, transmit
and store data are making cloud computing increasingly important in the delivery of public and private
services. This has considerable potential for economic and social development, in particular our efforts to
achieve the Millennium Development Goals and define a bold agenda for a prosperous, sustainable and
equitable future.


The Information Economy Report 2013 marks the first time the United Nations is examining the economic
potential of cloud computing for low- and middle-income countries, where rates of adoption are currently
low. With governments, businesses and other organizations in the developing world considering whether
to migrate some or all of their data and activities to the cloud, this publication is especially timely. I
commend its information and analysis to all those interested in learning more about the benefits and risks
of the cloud economy.


BAN Ki-moon
Secretary-General


United Nations




iv INFORMATION ECONOMY REPORT 2013


ACKNOWLEDGEMENTS


The Information Economy Report 2013 was prepared by a team comprising Torbjörn Fredriksson (team leader),
Cécile Barayre, Shubhangi Denblyden, Scarlett Fondeur Gil, Suwan Jang, Diana Korka, Smita Lakhe and
Marie Sicat under the direction of Anne Miroux, Director of the Division on Technology and Logistics.


The report benefited from major substantive inputs provided by Michael Minges, David Souter, Ian Walden and
Shazna Zuhyle. Research ICT Africa provided original research for five country case studies. Additional inputs
were contributed by Tiziana Bonapace, Axel Daiber, Nir Kshetri, Rémi Lang and Howard Williams.


Comments on the initial outline of the report were provided by experts attending a brainstorming meeting
organized in Geneva in February 2013, including Jamil Chawki, Alison Gillwald, Abi Jagun, Martin Labbé, Juuso
Moisander, Jason Munyan, Jorge Navarro, Thao Nguyen, Marta Pérez Cusó and Lucas von Zallinger. Valuable
feedback on various parts of the text was also given by experts attending a peer review meeting organized in
Geneva in July 2013, including Chris Connolly, Bernd Friedrich, Alison Gillwald, Angel González-Sanz, Nir Kshetri,
Matthias Langenegger, Mpho Moyo, Tansuğ Ok, Daniel Ramos and Carlos Razo.


Additional comments were received at various stages of the production of the report from Dimo Calovski,
Padmashree Gehl Sampath, Esperanza Magpantay, Markie Muryawan and Marco Obiso. Ngozi Onodugo
provided helpful assistance and inputs during her internship with UNCTAD.


UNCTAD is grateful for the sharing of data by national statistical offices and responses received to UNCTAD’s
annual survey questionnaire on ICT usage by enterprises and on the ICT sector. The sharing of data for this report
by the International Telecommunication Union, LIRNEasia, Research ICT Africa and TeleGeography is highly
appreciated.


The cover was done by Sophie Combette. Desktop publishing was done by Nathalie Loriot, graphics were
carried out by Stephane Porzi and Christian Rosé and the Information Economy Report 2013 was edited by
Maritza Ascencios, Lucy Annette Deleze-Black and John Rogers.


Financial support from the Government of Finland and the Republic of Korea is gratefully acknowledged.




vCONTENTS


CONTENTS


Note ............................................................................................................................................................ ii


Preface ...................................................................................................................................................... iii


Acknowledgements .................................................................................................................................. iv


List of abbreviations ................................................................................................................................. ix


Overview .................................................................................................................................................... xi


CHAPTER I THE CLOUD ECONOMY ECOSYSTEM ............................................. 1


A. Evolution to the cloud ...........................................................................................................2


B. Definitionsofkeytermsconcernedwithcloudcomputing ..............................................4
1. Cloud computing and cloud services ..........................................................................................4


2. Cloud service categories ............................................................................................................5


3. Cloud deployment models ..........................................................................................................6


C.Driversandbarriersrelatedtocloudcomputing ...............................................................8


D.Thecloudeconomyecosystem ..................................................................................................10


E. Implicationsfordevelopingcountries ...............................................................................12


CHAPTER II TRENDS IN THE CLOUD ECONOMY AND RELATED
INFRASTRUCTURE ..................................................................... 15


A. Trendsinthecloudeconomy .............................................................................................16
1. Cloud markets and traffic .........................................................................................................16


2. Leading corporate players in the cloud economy ......................................................................17


3. Can cloud computing trends be seen in trade statistics? ..........................................................20


B. Trends in cloud-related infrastructure ..............................................................................23
1. Factors determining the cloud readiness of countries ...............................................................23


2. International broadband connectivity ........................................................................................25


3. National backbone, Internet exchange points and data centres ................................................25


4. Broadband infrastructure trends ...............................................................................................27


5. Quality of service of broadband networks .................................................................................30


6. Affordability of broadband services ...........................................................................................34


C. Conclusions .........................................................................................................................36


CHAPTER III IMPLICATIONS FOR DEVELOPING COUNTRIES ............................. 43


A. Driversandbarriersofcloudadoption .............................................................................44
1. Drivers of cloud adoption..........................................................................................................44


2. Barriers to cloud adoption ........................................................................................................45


3. Assessing drivers and barriers to cloud adoption in developing countries .................................49


B. Cloudadoptioninselectedcountries:experiencesandopportunities .........................51
1. Supply-side cloud opportunities in developing countries ...........................................................51


2. Cloud use by different stakeholders in developing countries .....................................................57


C. Conclusions .........................................................................................................................63




vi INFORMATION ECONOMY REPORT 2013


CHAPTER IV GOVERNANCE,LAW AND REGULATION OF CLOUD SERVICES
IN DEVELOPING COUNTRIES ...................................................... 67


A. Internationallegalandregulatorytrends ..........................................................................68


B. Public law and cloud services ............................................................................................69
1. Cloud as a regulated activity .....................................................................................................69


2. Regulating in a multi-jurisdictional environment .........................................................................70


C. Private law and cloud services ..........................................................................................72


D. Implicationsfordevelopingcountries ...............................................................................74
1. Provision of cloud services .......................................................................................................74


2. Conditions of supply .................................................................................................................74


3. Treatment of data ....................................................................................................................75


E. Concludingremarksandpolicyrecommendations .........................................................76


CHAPTER V POLICY RECOMMENDATIONS .................................................... 81


A. Recognizingthediversityofcloudadoptionstrategies ..................................................82


B. Recommendations to Governments .................................................................................83
1. Designing strategic policy approaches to the cloud ..................................................................83


2. Addressing the infrastructure challenge ....................................................................................85


3. Strengthening the legal framework ...........................................................................................88


4. Map opportunities on the supply side .......................................................................................89


5. The human resource challenge .................................................................................................90


6. Government use of the cloud ..................................................................................................91


7. Implementation and follow-up of the strategy ...........................................................................91


8. Raising awareness ...................................................................................................................91


C.Recommendationstodevelopmentpartners ...................................................................92


References ...............................................................................................................................................95


STATISTICAL ANNEX ..................................................................................... 101


Glossary .................................................................................................................................................111


Listofselectedpublicationsintheareaofscience,technologyandICTfordevelopment ..........113


Readershipsurvey .................................................................................................................................116




viiCONTENTS


Boxes


I.1. The expansion of computing and communication power ................................................................2
I.2. Economies of scale and the cloud ..................................................................................................6
II.1. Cloud computing services in international classifications ...............................................................20
II.2. Cloud services over satellite ..........................................................................................................29
III.1. Barriers to cloud adoption in Ghana ..............................................................................................47
III.2. Government data centres in developing countries .........................................................................53
III.3. Examples of cloud service provision in developing countries .........................................................55
III.4. Selected cloud aggregators in developing countries .....................................................................56
III.5. Government cloud use in the Republic of Korea ...........................................................................58
III.6. Towards a government cloud in India ............................................................................................60
III.7. Benefiting from the cloud’s elasticity – the case of Zenga Media ...................................................63
IV.1. Protecting personal data in the cloud – different options and challenges .......................................69
IV.2. Legislating for cloud: the case of Mexico ......................................................................................71
IV.3. UNCTAD support to e-commerce legislation harmonization ..........................................................77
V.1. Promoting the cloud economy in Kenya ........................................................................................84
V.2. Approaches to measuring the quality of broadband services.........................................................86
V.3. The African Internet Exchange System Project ..............................................................................88


Tables


I.1. From mainframes to the cloud ........................................................................................................3
I.2. Different computing categories .......................................................................................................5
II.1. Estimates and forecasts of cloud revenue, 2010 and 2015 (Billions of dollars)...............................16
II.2. Top 10 companies by estimated number of servers, 2012 ............................................................18
II.3. Top 10 generators of SaaS revenue among the global 100 software firms, 2011 ..........................19
II.4. Top importing and exporting economies of computer servers of the type described under HS


847150, 2008–2012 .....................................................................................................................22
II.5. Infrastructure-related indicators used in different cloud readiness indices ......................................25
II.6. Theoretical highest mobile broadband speeds, W-CDMA family ...................................................30
II.7. Comparison of 3G and LTE speeds and latency in Finland ............................................................30
II.8. Relevance of broadband characteristics by cloud service .............................................................31
II.9. Quality of service performance, by groups of economies ..............................................................33
III.1. Selected cloud service providers in Ghana, 2013 ..........................................................................54
V.1. Potential advantages and risks/disadvantages of cloud adoption ..................................................82


Box table and figure


I.1. Cost implications of scale economies in data centre operations ......................................................6
III.1. Perceived challenges of migrating to cloud service .......................................................................47




viii INFORMATION ECONOMY REPORT 2013


Figures


I.1. Characteristics and types of cloud computing ................................................................................7
I.2. Main stakeholders and market relationships in the cloud economy ...............................................10
II.1. Global cloud data centre traffic (2011–2016) and distribution by region (2012) ..............................17
II.2. Getting to the cloud ......................................................................................................................24
II.3. Incremental lit capacity of submarine cables on major routes (Tbps) .............................................24
II.4. Distribution of IXPs, by region, June 2013 (Percentage) ................................................................26
II.5. Distribution of co-location data centres, by group, 2013 and secure Internet servers per


1 million people, 2012 ...................................................................................................................26
II.6. Global broadband subscriptions and distribution by fixed or mobile technology ............................27
II.7. Global fixed broadband subscriptions and per 100 people by region, 2007–2012 ........................28
II.8. Fixed broadband market share by technology and theoretical maximum download speeds ..........28
II.9. Economies with more than 10 per cent household penetration of fibre to the home/building plus


local area networks, 2012 (Percentage) ........................................................................................29
II.10. Global active mobile broadband subscriptions per 100 people, 2010–2012..................................30
II.11. Levels of cloud sophistication and related quality of service requirements .....................................31
II.12. Fixed broadband price basket (left) and annual change (right), 2008–2011 ...................................35
II.13. Mobile broadband prices as a percentage of gross national income per capita, by region and


level of development, 2011 ..........................................................................................................35
II.14. Pre- and post-paid ADSL and mobile broadband prices per month, 2013 (United States dollars) .....36
III.1. Government expectations of cloud impact, 2011 (Percentage share of replies) .............................45
III.2. Drivers and barriers to cloud adoption ..........................................................................................50




ixLIST OF ABREVIATIONS


LIST OF ABBREVIATIONS


3G third generation (refers to mobile phones)


ACP The African, Caribbean and Pacific Group of States


ADSL asymmetric digital subscriber line


API application programming interface


BPaaS business process as a service


BPO business process outsourcing


bps bits per second


BRICS Brazil, the Russian Federation, India, China and South Africa


CaaS communication as a service


CERT computer emergency response team


CIO chief information officer


CPC Central Product Classification


CPU central processing unit


CRM client customer relationship management


ERP enterprise resource planning


GATS General Agreement on Trade in Services


GB gigabyte


Gbit/s, Gbps gigabits per second


GDP gross domestic product


IaaS infrastructure as a service


ICT information and communication technology


IDC International Data Corporation


IP Internet protocol


ISO International Organization for Standardization


ISP Internet service provider


IT information technology


ITU International Telecommunication Union


ITU-T ITU Telecommunication Standardization Sector


IXP Internet exchange point


kbit/s, kbps kilobits per second


LDC least developed country


LTE long-term evolution


m2m mothers-2-mothers organization


Mbit/s, Mbps megabits per second


ms millisecond


NCIA National Computing and Information Agency (Republic of Korea)


NDC national data centre


NGO non-governmental organization


NIST National Institute of Standards and Technology




x INFORMATION ECONOMY REPORT 2013


NTT Nippon Telegraph and Telephone Corporation


OECD Organization for Economic Cooperation and Development


PaaS platform as a service


PC personal computer


PPP public–private partnership


PUE power usage effectiveness


QoS quality of service


RTT round-trip time


SaaS software as a service


SLA service level agreement


SME small and medium-sized enterprise


SMS short message service


Tbps terabits per second


TDF transborder data flow


TNC transnational corporation


UNCTAD United Nations Conference on Trade and Development


WTO World Trade Organization


XaaS x as a service




xiOVERVIEW


Cloud computing accentuates the quality dimension
of the digital divide.


The differential between countries in access to and
use of information and communication technologies
(ICTs) – the digital divide – has long been a significant
concern of Governments and the international
community. Over time, its nature has changed. The
gap in access to basic telephone services, once
very substantial, is now significantly diminished and
expected to shrink further in the next few years. In
its place has come a gap in access to the Internet
and, particularly, in access to broadband services. The
digital divide in broadband capacity and quality leads
in turn to a divide between countries and regions in the
extent to which individuals, businesses, economies
and societies are able to take advantage of new ICT
innovations and applications.


Cloud computing is a recent manifestation of this
evolving ICT landscape. Given its potential, it is
becoming increasingly important for Governments and
enterprises. In simple terms, cloud computing enables
users, through the Internet or other digital networks,
to access a scalable and elastic pool of data storage
and computing resources, as and when required.
Some predict that cloud technology will be among the
most significant disruptive technologies over the next
two decades, with major implications for markets,
economies and societies. Against this background,
the Information Economy Report 2013 provides an
objective analysis of the possible implications for
developing countries of the evolving cloud economy.


Massive improvements in storage, processing and
transmission capacity have paved the way for the
cloud economy.


The metaphor of the “cloud” can be misleading. Rather
than representing an amorphous phenomenon in the
sky, cloud computing is well anchored on the ground
by the combination of physical hardware, networks,
storage, services and interfaces that are needed to
deliver computing as a service. A key feature of cloud
computing is that it often involves transferring data to
a server controlled by a third party.


The shift that is taking place towards the cloud
represents a step change in the relationship between
telecommunications, business and society, and has


been enabled by massively enhanced processing
power, data storage and higher transmission speeds.
For example, Intel’s 22-nanometre central processing
unit is 4,000 times faster than that which the same
company introduced in 1971, and between 1986
and 2007 the world’s “technological memory” roughly
doubled every three years. Meanwhile, the fastest
theoretical speed of a dial-up connection in 1993,
the year the Internet browser was introduced, was
56 kilobits per second (kbit/s); as of 2013, consumer
broadband packages of 2 gigabits per second (Gbit/s)
are available, almost 36,000 times faster than dial-up.
Major cloud service providers today have hundreds of
thousands of servers located in massive data centres
in different parts of the world.


According to definitions proposed in April 2013 by the
International Telecommunication Union (ITU) and the
International Organization for Standardization (ISO),
cloud computing is a paradigm for enabling network
access to a scalable and elastic pool of shareable
physical or virtual resources with on-demand self-
service provisioning and administration. Cloud
services are defined as services that are provided and
used by clients on demand at any time, through any
access network, using any connected devices that
use cloud computing technologies. The implications
of cloud computing and cloud services on wider
economic development are discussed in this report in
the context of the cloud economy.


The cloud economy comprises various cloud service
categories and deployments.


However, the “cloud” and “cloud services” are not
homogenous products but come in different shapes
and configurations. Three categories of cloud services
– infrastructure as a service (IaaS), platform as a
service (PaaS) and software as a service (SaaS) – are
commonly used to encompass the whole range of
cloud service categories that are currently available.
The defining characteristic of each of these variations
of the cloud is the type of computing or information
technology (IT) facilities that is made available remotely
to a cloud service customer, on a rental or subscription
basis, by a cloud service provider:


In the case of IaaS, the cloud provider’s processing,
storage, networks and other fundamental computing
resources allow the cloud customer to deploy and run


OVERVIEW




xii INFORMATION ECONOMY REPORT 2013


software. The elasticity of IaaS allows an organization
or enterprise to access computing infrastructure in a
flexible and timely manner.


In the case of PaaS, the cloud customer deploys its
own applications and data on platform tools, including
programming tools, belonging to and managed by the
cloud provider.


With SaaS, the cloud customer takes advantage of
software running on the cloud-provider’s infrastructure
rather than on the customer’s own hardware. The
applications required are accessible from various client
devices through either a thin client interface, such as
a web browser (for example, web-based email), or a
program interface.


Cloud services can also be deployed to users in
a variety of ways, the most significant of which are
summarized below:


• Publicclouds: open resources that offer services
over a network that is open for public use. Many
mass market services widely used by individuals,
such as webmail, online storage and social media
are public cloud services.


• Privateclouds: proprietary resources provided for
a single organization (for example, a Government
or large enterprise), managed and hosted internally
or by a third-party.


• Communityclouds: resources/services provided
for and shared between a limited range of clients/
users, managed and hosted internally or by a
third-party.


• Hybridclouds: a mix of the deployment models
described above, for example, public and private
cloud provision.


Different cloud configurations offer both
opportunities and risks for potential cloud service
customers.


As a basis for the analysis, the Information Economy
Report 2013 uses the concept of the cloud economy
ecosystem, which highlights the deployment and
impacts of cloud computing and cloud services
within the wider information economy and, thereby,
their relevance to national economic development.
The cloud economy ecosystem includes a complex
set of relationships between technology and
business, governance and innovation, production and
consumption. It is how this ecosystem evolves, rather
than the potential of the technology alone, that will
determine the outcomes for developing countries.


As Governments, enterprises and other organizations
in the developing world consider whether to migrate
some or all of their data and activities to the cloud,
they need to assess the potential advantages and
risks of such a move.


Potential advantages include:


• Reduced costs for rented IT hardware and
software compared to in-house equipment and IT
management;


• Enhanced elasticity of storage/processing capacity
as required by demand;


• Greater flexibility and mobility of access to data
and services;


• Immediate and cost-free upgrading of software;
• Enhanced reliability/security of data management


and services.


Potential risks or disadvantages include:


• Increased costs of communications (to
telecommunication operators/Internet service
providers (ISPs));


• Increased costs for migration and integration;
• Reduced control over data and applications;
• Data security and privacy concerns;
• Risk of services being inaccessible, for example,


due to inadequate ICT or power infrastructure;


• Risk of lock-in (limited interoperability and data
portability) with providers in uncompetitive cloud
markets.


The cloud’s potential to improve efficiency is a
strong incentive for organizations in the private and
public sectors to transfer activities to the cloud. At
the same time, there are important trade-offs to be
made, for example, between cost savings on the one
hand and considerations related to data security and
privacy on the other. Various cloud customers will
assess the opportunities and risks associated with
the cloud differently, therefore opting for different
solutions. Some businesses, Governments and other
organizations are better positioned to reap the benefits
of a shift to the cloud, or can gain greater advantage
than others because of the nature of their activities
or business model. This is the case, for example, for
those that have high fixed costs in maintaining in-
house IT departments, recurrently need IT software
and hardware, face large or unpredictable variations
in demand for IT resources or can gain substantial
added value from more efficient exploitation of data
and market opportunities.




xiiiOVERVIEW


The cloud economy is expanding fast but is still
small.


There are various estimates of the size of the market
for cloud services. Fee-generated revenues from
public provision of IaaS, PaaS and SaaS have been
forecast to reach somewhere between $43 billion and
$94 billion by 2015. To this can be added the revenue
generated through advertising on cloud-enabled
web applications that are available at no cost to the
user. Such revenue is currently considerably larger
than the fees generated from public cloud provision.
Estimates of the value of private cloud services also
vary greatly – from about $5 billion to about $50
billion. Discrepancies in projections reflect different
methodologies, but most forecasts agree that cloud
adoption will continue to expand rapidly over the next
few years.


This is still very small compared with the revenue
of the global ICT sector, which was estimated at
about $4 trillion in 2011. However, most segments
of the ICT sector are in some way affected by cloud
computing. The demand for bandwidth will drive
telecommunication services revenue, although
revenues from voice services could be affected
as more people switch to cloud-based voice over
Internet protocol applications. Demand for equipment
and computer hardware, particularly data servers and
network equipment, will rise as more services move
to the cloud.


The shift to the cloud is generating considerable growth
in data traffic. During an average minute in 2012, Google
received two million search requests, Facebook users
shared around 700,000 content items and Twitter sent
out 100,000 tweets. In 2012, 60 per cent of such cloud
traffic on the Internet emanated from Europe and North
America. Asia–Pacific was responsible for another third
while Latin America and the Middle East and Africa
together accounted for only 5 per cent. However, the
highest growth rates in the next few years are expected
in the Middle East and Africa.


On the supply side, the cloud economy is currently
dominated by a few very large cloud service providers,
almost all headquartered in the United States. Their
early entry into cloud computing gave them first-
mover advantages, not least in terms of building large
networks of users and massive data storage and
processing capacity. The absolute levels of investment
required for major cloud-computing estates are very
high; it can cost more than half a billion dollars for a
cluster of data centres.


While the cloud service provider market is likely to
continue to be dominated by a small number of global
IT businesses, some factors may favour national or
regional players. Some Governments and enterprises
are required (by law or corporate policy) to locate their
data within national jurisdictions, or prefer to do so for
security or geopolitical reasons. Large corporations
and Governments have hitherto shown a preference
for private over public clouds, eschewing some
cost saving to ensure a greater sense of security
and control over their data and services. Recent
international publicity concerning data surveillance
may have reinforced such a preference.


Cloud adoption in developing countries has
potential implications for both the supply and the
user side of the cloud economy.


The most significant activities and potential supply
opportunities for enterprises in developing countries
are concerned with: (a) data-centre and related
cloud provision; (b) the development and provision
of local cloud services for groups of customers,
including local businesses and individual citizens; (c)
cloud aggregation, system integration, brokerage
and related services. In addition to these explicitly
cloud-based areas of activity, opportunities
exist for national communications businesses
(telecommunications operators and ISPs) which
can gain from increased data traffic using their
networks. Despite the advantages of global cloud
service providers, there are some factors that offer
scope for local or regional data centres to expand
in developing countries, such as growing demand
for private cloud solutions, national data-protection
laws or corporate policies requiring data to be kept
within national jurisdictions, and high costs of or
unreliable international broadband connectivity.


There has been extensive adoption by individuals in
developing countries of free cloud services such as
webmail and online social networks. This is true in
almost all countries, in particular those with higher
levels of Internet use and cloud readiness. The most
popular cloud-based applications are generally
those provided at a global level. In low-income
countries at a nascent stage of cloud readiness,
IaaS is often the first category of cloud services to
emerge. As the infrastructure situation improves
and if the SME sector expands, the market for SaaS
in developing countries will become more important
and eventually dominant as it already is in developed
countries.




xiv INFORMATION ECONOMY REPORT 2013


Foreign affiliates in developing countries make
extensive use of the cloud as part of their parent
companies’ global networks. With some wariness,
Governments in developing countries are also moving
towards the cloud. Some are developing systematic
cloud strategies, as part of broader ICT strategies
or sometimes alongside these. Where government
departments and larger corporations are concerned,
there is so far a general preference for private over
public cloud approaches. There is planned adoption
of the cloud in domestic enterprises, although less
extensive than anticipated by cloud advocates.


Experience of cloud computing in developing countries
is too recent for there to be a strongly established
evidence base on which to assess impacts. Businesses,
Governments and other organizations should carefully
examine the potential for cloud services to improve their
management and service delivery. They should only
migrate data and services to the cloud when they are
confident that the cloud offers significant benefits and
that attendant risks can be appropriately mitigated. Both
public and private cloud solutions should be considered
in this context, taking into account implications for data
security and privacy.


Infrastructure deficiencies seriously hamper the
uptake of and benefits from cloud computing in
many developing countries.


For several reasons, the options for cloud adoption
available in low- and middle-income countries look
different from those in more advanced economies.
Critical factors relate, among other things, to the
availability and quality of cloud-related infrastructure,
cost considerations and inadequate legal and
regulatory frameworks to address data protection and
privacy concerns.


As regards access to and availability of cloud-related
infrastructure, and despite significant improvements in
broadband connectivity in many developing economies,
the gap between developed and developing countries
keeps widening. Average fixed broadband penetration
is now more than 28 subscriptions per 100 people in
developed economies, 6 in developing countries and
only 0.2 in the least developed countries (LDCs). In the
case of mobile broadband, the gap is also significant.
The average number of subscriptions in 2012 was
about 67 per 100 people in developed countries, 14
in developing countries and below 2 in the LDCs.


In addition, in most low-income countries, mobile
broadband networks are characterized by low speed


and high latency and are therefore currently not
ideal for cloud service provision, especially of the
more advanced kinds. The net value of cloud-based
solutions will be lower in countries with a heightened
risk of communication- and power-network outages.
The lack of supporting infrastructure, such as Internet
exchange points (IXPs), reliable and inexpensive
electricity and robust fibre-optic backbones also affect
the deployment of national data centres. Indeed, as
much as 85 per cent of data centres offering co-
location services are in developed economies. This
“data centre divide” is reflected in the availability of
servers; whereas there were in 2011 more than 1,000
secure data servers per million inhabitants in high-
income economies, there was only one such server
per million inhabitants in LDCs.


The cost of communication remains another critical
obstacle for adoption of cloud services in many
developing countries. The fees paid to cloud service
providers and for broadband access and usage,
charges by the ISP and the hardware and software
costs incurred are likely to form a much higher
proportion of the total costs of cloud provisioning
than in advanced economies. The combination of few
national data centres and high costs of international
broadband communications further weighs on the net
value of relying on cloud solutions.


The cloud raises legal and regulatory challenges,
especially concerning data protection and privacy.


The rapid emergence of cloud computing has raised
concerns about its legal and regulatory implications.
Issues of data protection and security are among the
concerns most frequently mentioned by potential cloud
customers in both developed and developing countries.
Such concerns have intensified following the disclosure
in 2013 of national surveillance programmes and
reports on access by law-enforcement agencies to data
hosted by global cloud service providers. Governments
need to protect national interests and their citizens;
service providers require a stable framework to facilitate
innovation and investment; and users require assurance
and trust to encourage the take-up of such services.
Policy responses may range from a do-nothing attitude
to the adoption of cloud-specific laws.


Public law is essential to secure the basic rights of
end users. While there is no imperative to develop
specific laws or regulations on cloud computing,
areas requiring reform are relatively clear: privacy, data
protection, information security and cybercrime. For
Governments of developing countries, it is essential




xvOVERVIEW


that appropriate laws and regulations are adopted and
enforced in these areas. As of 2013, there were 99
countries with data-privacy laws. As far as is known,
Mexico is the only country which has adopted cloud-
specific provisions in relation to data protection. There
is no international harmonized privacy framework
regulating data transfers across borders, but
developing countries could benefit from implementing
strong domestic-privacy regimes.


In addition to public law, contractual agreements
between cloud service providers and cloud service
customers also greatly impact on the operation and
effects of the cloud economy. In some circumstances,
regulatory intervention in the freedom to contract
may be necessary to protect the public interest. The
placement of data in the cloud may require regulatory
intervention to address concerns related to personal
privacy, commercial secrecy or national security.
For example, within data protection laws, imposing
minimum responsibilities on the cloud service providers
– to ensure the security of customer data and to notify
its customers if there is a security breach – could help
to provide greater transparency about vulnerabilities
and to enable mitigation in a timely manner.


Where there are apprehensions on relying heavily on
cloud services offered by providers based in a foreign
jurisdiction, it may be difficult to address this market
reality through regulatory intervention. An alternative
policy response may be to encourage the establishment
of domestic cloud services, either by offering foreign
investors a favorable environment to invest in the
building of local infrastructure (such as data centres) or
encouraging domestic enterprises to enter the supply
side of the cloud economy. Whereas such measures
may involve regulatory components, such as imposing
“localization” requirements, they would be designed to
facilitate the provision of cloud services rather than to
constrain them. Several Governments of developing
countries are building government clouds to serve the
needs of the Government itself and sometimes others.
In Europe, there have been calls for the development
of a secure European cloud and some national cloud
initiatives have been launched to offer an alternative
source of cloud service provision.


Governments should facilitate benefits from the
cloud economy but be aware of pitfalls.


Although cloud adoption is still at a nascent stage in
developing countries, policymakers should waste no
time in enhancing their understanding of how it may
affect their economies and societies, in order to be


able to make informed policy decisions. Government
policies should be based on an assessment of the
pros and cons of cloud solutions and be rooted in
a thorough understanding of existing ICT and cloud
use within countries. Governments need to recognize
the diversity of business models and services within
the cloud, the multiplicity of customers of cloud
services, and the complexity of the cloud economy
ecosystem. In view of its relevance for both public
service delivery and business competitiveness, it
is important to integrate any cloud strategy in the
overall national development plan, and to plan for
its execution, monitoring and evaluation. Policy
approaches should be tailored to the circumstances
of individual economies, and be consistent with the
overall strategic framework for national economic
development and for leveraging ICTs.


On the whole, Governments should broadly welcome
and support the development of a cloud economy
and the adoption of cloud services. In principle,
there is no general case for government policy and
regulation to discourage migration towards the cloud.
Rather, policies and regulatory approaches should
seek to create an enabling framework that supports
firms and organizations that wish to migrate data and
services to the cloud so that they may do so easily
and safely. However, this does not mean that cloud-
based solutions are always preferable to alternative
approaches. In addition, there are multiple ways
of making use of cloud technology – using public,
private or hybrid clouds at national, regional or global
levels. Governments should seek to facilitate those
approaches that seem most likely to deliver wider
economic benefits in their particular context.


A number of steps could be considered by
Governments that wish to translate the potential of
the cloud into tangible development gains. In terms
of scope, at the national level policymaking would be
advised to consider measures related to the following
areas:
• Assess the cloud readiness of the country.


Governments should start by carefully assessing
the current situation in their countries, to identify
bottlenecks and weaknesses that need to be
addressed if the cloud is to be effectively exploited,
and clarify what kind of cloud solutions are most
propitious.


• Develop a national cloud strategy. Based
on the readiness assessment, a national cloud
strategy could be drafted either as a stand-alone




xvi INFORMATION ECONOMY REPORT 2013


policy document or as an integral part of the
national ICT strategy.


• Address the infrastructure challenge. This
would involve measures to improve the provision
of reliable and affordable broadband infrastructure
and to monitor regularly the quality of broadband
services. Effective communications regulations are
here of the essence. Attention should also be given
to the role of IXPs and the provision of electricity.


• Addressrelevantlegalandregulatoryissues
related to cloud adoption to ensure that
cloud service users’ interests are properly
protected. Key areas include the location of data,
e-transactions and cybercrime. Efforts should be
made to reflect international best practice in the
development of new legislation.


• Map opportunities in the supply side of
the cloud economy. Three key areas deserve
particular attention: the development of national
data centres, the potential for cloud aggregation
services, and the development of new cloud
services.


• Address the need for human resources.
Skill areas that are likely to become increasingly
important include those related to the IT and
software skills needed to manage the migration
and integration of cloud services; management and
organizational skills to handle the reorganization
and re-engineering of business processes; and
legal and procurement skills.


• Government use of cloud services. Given their
important role in the information economy in many
developing countries, the role of Governments
should be explored with regard to the establishment
of national data centres, e-government systems
and related public procurement.


Development partners should work with
Governments in responding to the cloud economy.


Addressing the many challenges that developing
countries face in seeking to benefit from the evolving
cloud economy will require both expertise in various
fields and financial resources. Development partners
could help in that respect, by ensuring that cloud-
related development challenges are incorporated in
their agendas to reduce the risk that the move towards
the cloud economy may result in a widening of the
digital divide. They may also provide support at the
country level in contributing to financing broadband
infrastructure, establishing appropriate legal and
regulatory frameworks, and building capacity in
relevant areas.


International agencies could facilitate this assistance
through some of their existing activities. UNCTAD and
other international organizations can, for example,
facilitate an exchange of experiences with regard to
the policy challenges that developing countries face
to derive benefits from the cloud economy and avoid
pitfalls.


Another key area in which development partners can
play a role concerns international standards for cloud
services, which are essential to facilitate interoperability
and to help customers understand what they are
purchasing. Standardization forums should consider
how to engage developing countries and their users to
ensure that their specific needs and requirements are
addressed. More research is also needed in a number
of areas to allow for a more comprehensive assessment
of the impact of different forms of cloud adoption. As
the evidence base expands, it will become feasible
to assess macroeconomic implications for economic
growth, employment, productivity and trade.


As with other ICT areas, the pace of change
in cloud technology and markets is rapid. The
experiences described in this report relate to present
circumstances. The nature of cloud services and of
the cloud economy will continue to develop fast, and
may be very different in five years’ time. Governments,
businesses and development partners need to bear
these changes in mind, and to re-evaluate their
policies and strategies concerning the cloud regularly
to ensure that they continue to maximize potential
benefits and minimize potential risks to their citizens,
businesses and customers.


Mukhisa Kituyi
Secretary-General, UNCTAD


24


Dr. Mukhisa Kituyi


Secretary-General of UNCTAD




THE CLOUD
ECONOMY ECOSYSTEM1


Improvements in processing power, storage capabilities and communication
transmission speeds have together facilitated the emergence of what has
come to be known as cloud computing. This phenomenon is likely to influence,
in various ways, both the production and use of services delivered through
ICTs over coming years. As with other new technological developments, it
is difficult to predict the full implications of cloud computing. While it offers
significant potential benefits to those who are able to leverage them, it
also raises concerns for users, not least with regard to data protection and
privacy. Moreover, the spread of cloud computing may widen the digital
divide between those – countries, businesses and individuals – that are and
those that are not well placed to benefit from it.


Although cloud adoption is still at a nascent stage in developing countries,
policymakers should waste no time in improving their understanding of how it
may affect their economies and societies, in order to be able to make informed
policy decisions. The Information Economy Report 2013 seeks to assist in this
context, by stimulating a nuanced discussion of the potential role and value
of the “cloud economy“. This first chapter explains the evolution towards the
cloud, defines key terms, outlines the most important drivers and barriers
of cloud adoption, and presents an overall framework for analysing the
cloud economy ecosystem. It thereby provides a road map for the following
chapters of this report.




2 INFORMATION ECONOMY REPORT 2013


A. EVOLUTION TO THE
CLOUD


Cloud computing and the emergence of a cloud
economy based upon it are becoming increasingly
important considerations for Governments and
enterprises. A recent study has described cloud
technology as one of the most significant disruptive
technologies which will develop over the next two
decades, with major implications for markets,
economies and societies (Manyika et al., 2013). It
predicts that, by 2025, most information technology
(IT) and web applications and services could be cloud
delivered or cloud enabled, and that most enterprises
might be using cloud facilities and services. While it is
always difficult, even risky, to predict the development
of IT so far ahead, policymakers and business leaders
should start examining the opportunities and potential
risks of this growing phenomenon.


In simple terms, cloud computing enables users,
through the Internet or another digital network, to
access a scalable and elastic pool of data storage and


computing resources, as and when they are required
(see also section I.B). The metaphor of the “cloud“
can be misleading here. Rather than representing
something amorphous or in the sky, cloud computing
is enabled by the combination of the physical
hardware, networks, storage, services and interfaces
that are needed to deliver computing as a service. A
key feature of cloud computing is that it often involves
transferring data and computing to a server controlled
by a third party.


The shift towards cloud computing that is currently
taking place can be seen as a step change in the
relationship between telecommunications, business
and society, enabled by greatly enhanced processing
power, data storage and higher transmission speeds,
accompanied by sharp price reductions (box I.1).


In one sense, the cloud has taken the relationship
between user and computer back towards an earlier
era (table I.1). Computing began in the 1950s and
1960s with access to large “mainframe“ computers
through “dumb terminals“, typewriter-like devices with


Box I.1. The expansion of computing and communication power


A few examples can help to illustrate the dramatic evolution of processing, storage and transmission capacity.


• Processing power. The first transistor, precursor to the central processing unit (CPU) that forms the “brain“ of a
computer, was built by hand in 1947. Gordon Moore, one of the founders of microprocessor manufacturer Intel,
predicted in 1965 that the number of transistors on a chip will double roughly every year (“Moore’s Law“). Intel
introduced its first CPU (the 4004) in 1971; Intel’s current 22 nanometre CPU is 4,000 times faster, uses only 0.02 per
cent of the energy and costs 50,000 times less (Intel, 2011).


• Digital storage capabilities. The world’s “technological memory“ roughly doubled every three years between 1986 and
2007, about the time that cloud computing began to take off. It grew in this period from 2.5 optimally compressed
exabytesa (1 per cent digitized) to around 300 (94 per cent digitized) (Hilbert and López, 2011). The first IBM personal
computor (PC), which was introduced in 1981, cost $3,000 and could only accept diskettes of 160 kilobytes storage.b
By 2010, for $600 it was possible to buy a hard disk that could store all the world’s recorded music.c


• Transmission speeds. At both user and backbone levels these have increased dramatically. The fastest theoretical
speed of a dial-up connection in 1993, the year the Internet browser was introduced, was 56 kbps. As of 2013,
consumer broadband packages of 2 Giga bit per second (Gbps) are available, almost 36,000 times faster than dial-up
(see also section II.B). The first transatlantic fibre optic cable, TAT-8, was introduced in 1988 with a speed of 280
Mbps.d In 2011 the Hibernia Atlantic Submarine Cable System reached speeds of 100 Gbps, some 350 times faster.e


Source: UNCTAD.
a The byte (8 bits) is the unit used for expressing digital information. One exabyte is equal to one quintillion bytes.
b See http://www-03.ibm.com/ibm/history/exhibits/pc25/pc25_press.html (accessed 2 October 2013) and


http://www-03.ibm.com/ibm/history/exhibits/pc25/pc25_fact.html (accessed 2 October 2013).
c Kevin Kelly, Web 2.0 Expo and Conference, March 29, 2011. Video available at www.web2expo.com/


webexsf2011/public/schedule/proceedings (accessed 2 October 2013).
d See http://atlantic-cable.com/Cables/speed.htm (accessed 2 October 2013).
e See http://www.submarinenetworks.com/systems/trans-atlantic/hibernia-atlantic/hibernia-atlantic-trials-100g-


transatlantic (accessed 2 October 2013).




3CHAPTER I : THE CLOUD ECONOMY ECOSYSTEM


no intelligence of their own. The “personal“ computer
appeared in the 1980s, providing users with their
own processing power and storage independent of
these earlier mainframe devices. When there was
a need to obtain remotely stored information in this
new paradigm the user‘s client computer accessed
a server (client/server computing). The introduction
of the Internet initially allowed only slow telephone
dial-up access, first for email and file downloads, later
to access websites. As telecommunications networks
have developed they have enabled ever faster access
to more sophisticated applications.


While the evolution of computing towards the cloud
has been gradual, it is only in the past few years that
the economics of cloud computing have triggered
a move away from processing and storing data on
client hardware to using cloud-based data centres. As
stated in a recent review of cloud computing (Crémer
et al., 2012):


The seeds of cloud computing were deposited as
soon as the Internet began developing and…recent
evolutions have been questions of degrees, not
of nature. But when there are enough degrees of
difference, the nature of the beast becomes totally
different.


Although there are similarities between the mainframe
model and today’s cloud paradigm, there are also
crucial differences. The mainframe model was designed
to manage scarce and expensive data storage and
processing resources, rationing their availability to
a limited set of users within an organization. By
contrast, cloud computing is being deployed in an era
of abundant data storage and processing resources.
Cloud providers aim to ensure that the much wider
availability, greater diversity of applications and lower
costs that result from this abundance are made widely
available to the growing numbers of Internet and cloud
users. Cloud computing’s success depends explicitly
on the conjunction between great computing capacity


(data storage and processing power) and reliable high-
speed communications (which enable users to access
that computing capacity in real or near-to-real time).


As will be discussed below, this new paradigm has
the capacity to reduce capital expenditure on IT,
improve operational efficiency (by introducing new
administrative and service-delivery models), enable
new applications, and improve customer service.
A general shift in business and government data
and data handling to the cloud, it has been argued,
could thereby stimulate efficiency, productivity and
economic growth. At the same time, there is nothing
automatic about these potential gains, which are
likely to be unevenly distributed, geographically,
within societies, and over time. Cloud adoption
in developing countries faces several significant
constraints which are discussed later in this report,
including inadequacies in infrastructure and legal and
regulatory frameworks. As in other countries there
are also concerns about security, data protection,
privacy and reliability. Moreover, since the early stages
in the development of the cloud, some experts have
cautioned against the risk of surrendering control over
the software used by relying on third-party servers and
without being able to study the source code used.
These are complex aspects of the cloud economy
with important implications not just for individuals,
businesses and Governments, but for economic and
social development in general.


Cloud computing is already having an impact on the
ways in which Governments, businesses and citizens
interact online. Consumer-oriented services such
as webmail, online social networks and file-sharing
applications are among the most popular applications
currently used on the Internet. All are fundamentally
cloud based. Enterprises and Governments,
particularly in developed countries, are increasingly
transferring data to the cloud and making use of cloud
services rather than relying on their own hardware and


Table I.1. From mainframes to the cloud


Source: UNCTAD, adapted from (Microsoft, 2010).


Technology Centralized computation & storage; thin clients
High up -front costs for
hardware and software


Economic
PCs and servers for distributed computation, storage,


etc.
Optimized for agility because of the low cost


Perpetual licence for
operating system and
application software


Business model
Large data centres, ability to scale, commodity


hardware, devices
Efficiency and agility an order of magnitude


better
Ability to pay as you go, and


only for what you use


Optimized for efficiency because of high cost


Characteristics


1960 1970 1980 1990 2000 2010
Mainframe
computing


Client/server
computing


Cloud
computing>>>> >>>>


>>>>>>




4 INFORMATION ECONOMY REPORT 2013


software. For some, the use of the cloud has already
become the norm, at least for data storage, with new
services set to follow. There have been warnings of a
hype cycle, reminiscent of the earlier dot.com boom,
with regard to cloud computing, with overly optimistic
expectations over the next few years (Renda, 2012:
25). For example, high prices are being paid by
major cloud providers for the acquisition of smaller
businesses that have developed new cloud services
and applications.2


The remainder of this chapter is structured as follows.
Section B provides definitions of cloud computing,
cloud services and other relevant terms. Section
C briefly summarizes the main advantages and
concerns that are attributed to cloud adoption from
the perspectives of different users. Section D identifies
the principal stakeholders in the cloud economy and
draws these together into a descriptive model of the
cloud economy ecosystem. Finally, section E outlines
the main research questions that are addressed in
the report and provides a roadmap to subsequent
chapters.


B. DEFINITIONS OF KEY
TERMS CONCERNED
WITH CLOUD
COMPUTING


In order to understand the implications of the
emerging cloud economy – the interactions between
cloud service providers and cloud service customers
in different parts of society – it is important first to
clarify exactly what is meant by cloud computing,
to describe what it does and to summarize how it
works. This section of the chapter provides core
definitions which underpin the understanding of
cloud computing and the cloud economy in the
remainder of the report.


1. Cloud computing and cloud services


The concept of cloud computing is still evolving. What
cloud computing offers today differs from what it
offered five years ago and from what it will offer five
years hence. Different definitions therefore emphasize
different aspects of the cloud: its technology, its
economics or how it is experienced.


Cloud computing is a way of delivering applications,
services or content remotely to end users, rather


than requiring them to hold data, software or appli-
cations on their own devices, a process known as
“virtualization”. The automation of infrastructure and
application processes (including configuration, pro-
visioning, auto-scaling and failure recovery) enables
cloud providers to hold data and run applications/
services on behalf of cloud customers. Data and
applications are managed through application pro-
gramming interfaces (APIs) between cloud software
and cloud customers’ own IT systems. Virtualization
also allows resources such as servers and storage
devices to be shared between many different users,
an arrangement known as “multi-tenancy” that maxi-
mizes the operational efficiency of data centres and
enables sharing of overhead costs among numerous
cloud customers, so further reducing the total costs
of usage.


Cloud computing can be described as a rental or
pay-as-you-go model for the exploitation of IT assets
and resources, by contrast with the ownership model
(for both hardware and software) which has become
prevalent during the client/server era of computing.
It exploits the storage capacity of very large-scale
data centres and the communications capabilities
of broadband infrastructure, where these are both
available, to achieve economies of scale and scope
that can be passed on as lower costs for end users.
The legal and economic terms of such arrangements
are contractually agreed between cloud providers
and cloud customers, and governed by service-
level agreements (SLAs) reached between them
(European Telecommunications Network Operators’
Association, 2011).


One widely used definition of cloud computing, which
draws on these descriptions, is that put forward by
the National Institute of Standards and Technology
(NIST) of the United States of America, as follows
(National Institute of Standards and Technology,
2011):


Cloud computing is a model for enabling ubiquitous,
convenient, on-demand network access to a
shared pool of configurable computing resources
(e.g., networks, servers, storage, applications, and
services) that can be rapidly provisioned and released
with minimal management effort or service provider
interaction.


An alternative definition has been proposed
more recently (April 2013) by the International
Telecommunication Union (ITU) and the International
Organization for Standardization (ISO):3




5CHAPTER I : THE CLOUD ECONOMY ECOSYSTEM


Cloud computing is a paradigm for enabling network
access to a scalable and elastic pool of shareable
physical or virtual resources with on-demand self-
service provisioning and administration.


Both of these definitions establish the fundamental
nature of the concept. Following on from them, “cloud
services” have been defined as services that are provided
and used by clients “on demand at any time, through
any access network, using any connected devices [that
use] cloud computing technologies” (ITU, 2012a). A
further distinction can be made between cloud services
that can be accessed via a web browser without having
to install any software on the user’s device, and “cloud-
based services” which require software installation in
order to make use of cloud resources. For the sake of
simplicity, the term “cloud services“ will be used in this
report to represent both kinds of services.


2. Cloud service categories


A number of different cloud service categories are
consistent with these characteristics. These are
generally described using the formula “X as a service”
(XaaS), where X represents the kind of facility made
available, on the terms described above, by the cloud
service provider to the cloud service customer (see
ecosystem diagram in figure I.2). A number of different
formulations of XaaS are described in the literature,
depending on the granularity with which they seek
to describe the services that are being offered. Three
categories of cloud services – infrastructure, platform
and software as a service (IaaS, PaaS and SaaS) –
are commonly used to encompass the whole range
of cloud service categories that are currently available.


These are compared in table I.2 with one another and
with traditional IT systems.
• IaaS. In this category, the cloud provider’s


processing, storage, networks and other
fundamental computing resources allow the cloud
customer to deploy and run software, which
can include operating systems and applications.
The cloud customer does not manage or control
the underlying infrastructure but has control
over operating systems, storage and deployed
applications, and may have limited control of select
networking components (for example, host firewalls).
Making use of the elasticity of IaaS for data storage
and processing capacity allows an organization
or enterprise to access computing infrastructure
in a flexible and timely manner, for example when
demand is uncertain or unpredictable.


• PaaS. In this category, the cloud customers
deploy their own applications and data on platform
tools, including programming tools, belonging to
and managed by the cloud provider. Application
developers working on mobile applications
commonly use cloud-based platforms to develop
and launch their services. The cloud customer
does not manage or control the underlying
cloud infrastructure such as network, servers,
operating systems, or storage, but has control
over the deployed applications and perhaps over
configuration settings for the application-hosting
environment. For example, a developer working on
Android applications can use PaaS to ensure that
an application can automatically take advantage
of changes implemented in, and follow the look
and feel of, new releases of the Android operating
system as they appear.


Table I.2. Different computing categories


Source: Microsoft, 2010.


Applications Applications Applications Applications


Data Data Data Data


Runtime Runtime Runtime Runtime


Middleware Middleware Middleware Middleware


Operating system Operating system Operating system Operating system


Virtualization Virtualization Virtualization Virtualization


Servers Servers Servers Servers


Storage Storage Storage Storage


Networking Networking Networking Networking


Traditional IT IaaS PaaS SaaS


Us
er


m
an


ag
es


De
liv


er
ed



as


a
s


er
vi


ce
Us


er
m


an
ag


es Us
er



m


an
ag


es
De


liv
er


ed
a


s
a


se
rv


ic
e


De
liv


er
ed


a
s


a
se


rv
ic


e




6 INFORMATION ECONOMY REPORT 2013


• SaaS. In this category, the cloud customer takes
advantage of software running on the cloud provider’s
infrastructure rather than on the customer’s own
hardware. The applications required are accessible
from various client devices through either a thin client
interface, such as a web browser (for example, web-
based email), or a program interface. In SaaS services,
the customer has no control over the underlying cloud
infrastructure, accessing applications through a web
browser or separate programme interface.4 Another
formulation of XaaS that might broadly be included in
SaaS is CaaS (communications as a service), which
includes cloud services for messaging and voice-
over Internet protocol (IP).5


The defining characteristic of each of these variations
of the cloud is the type of computing or IT facilities
that are made available remotely to a cloud service
customer, on a rental or subscription basis, by a cloud
service provider. Globally, the most widespread of
these service models today is SaaS, followed by IaaS
and PaaS (Chapter II). However, as will be illustrated
in subsequent chapters, the degrees of prevalence
differ considerably between countries, depending
on their level of development in terms of broadband
infrastructure and regulatory framework.


3. Cloud deployment models


As well as different service categories, cloud services
can be deployed to users in a variety of ways, the most
significant of which can be summarized as follows:


• Public clouds are an open resource that offers
services over a network that is open for public use.
Examples of public clouds include Amazon Elastic
Compute Cloud (EC2), IBM’s Blue Cloud, Sun
Cloud, Google AppEngine and Windows Azure
Services Platform. Many mass market services
widely used by individuals, such as webmail and
online storage, are public cloud services.


• Private clouds, by contrast, are a proprietary
resource provided for a single organization
(for example, a Government or large business
customer). It can be managed internally or by a
third party, and hosted internally or externally.


• Community clouds are a resource/service
provided for and shared between a limited range
of clients/users. It can be managed internally or by
a third party and hosted internally or externally. It
might be considered halfway between public and
private cloud provisioning.


Box I.2. Economies of scale and the cloud


The scale of computing power involved in the very large data centres that are at the heart of major cloud provider businesses
has markedly different economic impacts from earlier models of data centre provisioning. A traditional data centre comprises
a collection of corporate IT resources. It typically houses hardware that can support a variety of applications (and software
architectures), and therefore has to manage complex workloads. This requires multiple management tools and incurs high
maintenance costs. As a result, traditional data centres have typically maintained substantial surplus capacity and been
relatively expensive to run. The differences in costs between traditional data centres and the very large-scale data centres
characteristic of cloud computing are illustrated in box table I.1. While costs vary in absolute terms and in relation to one
another, this underscores the significant potential for cost reduction that becomes available through the cloud.


Box table I.1. Cost implications of scale economies in data centre operations


Cost in medium-sized data centre Cost in very large data centre Ratio
Network $95 per Mbps/month $13 Mbps/month 7.1
Storage $2.2 per Gbyte/month $0.40 per Gbyte/month 5.7


Administration Approx. 140 servers/administrator <1000 servers/administrator 7.1


Source: Armbrust et al., 2009.


Economies of scale grow even more important through the agglomerations of data centres that form the cloud. Major cloud
providers do not necessarily warehouse data from an individual customer in any particular data centre but may distribute
the data across several large data centres, often in diverse jurisdictions. Indeed, the ability to do this forms a significant part
of their competitive advantage. By creating interconnected agglomerations, cloud providers can offer much more powerful
computational outcomes. The ability to reap optimal benefits from such accumulation of computing power usually requires
cross-border data flows – the source of some policy and regulatory challenges associated with the cloud.


Source: UNCTAD.




7CHAPTER I : THE CLOUD ECONOMY ECOSYSTEM


• Hybrid clouds expand deployment options for
cloud services by mixing the deployment models
described above, for example, public and private
cloud provision. This may be valuable, for example,
where the data and applications involved require
different levels of security or are subject to different
regulatory requirements.


In practice, major cloud providers often use the
same data centres when offering both public and
private services. Governments and individual
businesses may make use of more than one
cloud and more than one cloud service provider
for different purposes. Interoperability between
clouds and service providers is therefore important
in providing flexibility and competition in service
provision, not least in enabling customers to switch
between providers.


Different cloud users have different needs and are
therefore likely to prefer different deployment models.
At present, individual users and small and medium-
sized enterprises (SMEs) make widespread use
of public clouds, either consciously (for example,


through cloud storage services), or in many cases
without realizing that they are doing so (such as
when using social networks or webmail services).
Many public cloud services with mass markets are
either offered free of charge to customers and funded
by advertising revenue, or are available at low cost.
Such business models are made possible by the
massive economies of scale that can be achieved
by cloud computing data centres handling very
large quantities of information (box  I.2). In spite of
such economies of scale, however, major enterprise
and government customers have so far generally
preferred to procure private cloud provision, which
can be more easily tailored to individual requirements
and which is perceived to have security and data
protection advantages.


The diverse aspects of cloud computing defined above
are drawn together in figure I.1, which also highlights
certain characteristics that give cloud computing a
comparative advantage over other models of handling
data (National Institute of Standards and Technology,
2011). These can be summarized as:


Broad network
access


IaaS


Infrastructure as a
service


PaaS


Platform as a
service


SaaS


Software as a
service


SERVICE CATEGORIES DEPLOYMENT MODELS


Rapid elasticity


Measured
service


On-demand
self-service


Resource pooling


Multi-tenancy


CORE CHARACTERISTICS


Public cloud


Community cloud


Private cloud


Figure I.1. Characteristics and types of cloud computing


Source: UNCTAD, adapted from National Institute of Standards and Technology, 2011.




8 INFORMATION ECONOMY REPORT 2013


• Broad network access. Capabilities can be
accessed over the network using standard
terminal devices such as PCs, laptops, tablets and
smartphones;


• Rapidelasticity.The capability for rapid scaling,
up or down, of the access or services provided in
accordance with user requirements;


• Measured service. The business model that
data access or service provided is monitored and
charged by usage (on a pay-as-you-go basis);


• On-demand self-service. Based on the principle
that users can provision relevant resources as and
when required, on their own initiative, without having
to negotiate access terms at the time of need;


• Resource pooling. The resources supplied by
a cloud provider serve multiple users rather than
being dedicated to a single user, and they are
assigned as required by user demand, resulting in
them being less costly per unit than they would be
if provided for a single user;


• Multi-tenancy.Physical and virtual resources are
allocated in such a way that multiple tenants and
their computations and data are isolated from, and
inaccessible to one another.


C. DRIVERS AND
BARRIERS
RELATED TO CLOUD
COMPUTING


A transition to the cloud involves both potential
advantages and potential risks and costs. This makes
it important for Governments and organizations in the
public and private sectors to analyse carefully how
they might be affected by such a transition and how
to respond.


The cost-effectiveness of the cloud model with
its large economies of scale would suggest that it
could become the norm, rather than the exception,
for data storage and service provisioning. This is
already the case for globally provided mass market
applications, such as webmail and online social
networks. In developed countries, almost all listed
companies and other large private companies make
significant use of the cloud, though cloud adoption
among SMEs is more limited. While developing
countries do not yet exhibit such high adoption rates,
enterprises are starting to explore cloud provisioning,
and their economies will be impacted by its growing
prevalence (chapter III).


There is a burgeoning literature on the potential
advantages of cloud computing. Much of this consists
of studies conducted or sponsored by the cloud
industry rather than reflecting independent empirical
research. Experience of cloud computing, particularly
in developing countries, is too recent for there to be
a strongly established evidence base on which to
assess impacts on individual firms, organizations or
national economies. There is, however, a consensus
on the principal potential benefits that can be derived.
These can be summarized in the following three points,
which will be discussed in greater detail in chapter III:


1. Cost savings. Hardware and software in con-
ventional business environments are rarely used to,
or approaching, full capacity. Aggregation of de-
mand in the cloud means that cloud service provid-
ers’ equipment is used more efficiently than their
customers’ equipment would be, enabling benefits
from economies of scale to be passed on to users.
These factors may result in cost savings in three
main areas:


(a) Hardware requirements for devices that
primarily use the cloud can be much lower, with
lower specifications and less frequent need for
hardware upgrades;


(b) Latest versions of software can be accessed
as and when required, rather than purchased
(or pirated) and uploaded to all hardware on
which they might be required;


(c) There is less need for in-house IT maintenance
and support staff (though staff dealing with IT
management, procurement, and the like, need
to be as highly skilled in cloud as in conventional
environments).


Overall, transition to the cloud can permit firms
to divert significant resources from capital to
operational expenditure, enabling earlier return on
investment. The potential for reaping benefits from
these economies of scale is greater in public and
community clouds, where there is more aggregation
of demand. Nonetheless, many Governments and
enterprises prefer private cloud solutions, balancing
these advantages against other issues such as
control over data management and security.


2. Flexible access to facilities on demand,
with a high degree of elasticity. Cloud
procurement enables businesses to implement
new administrative or customer service approaches
more quickly than would be possible through the




9CHAPTER I : THE CLOUD ECONOMY ECOSYSTEM


acquisition, development and implementation of
bespoke applications. Cloud service customers
can also more or less instantly increase or decrease
the services/resources they require from the cloud
in response to fluctuations in demand. This makes
it easier for cloud service customers to manage
peak workloads, and to downsize when necessary.


3. Improved system management, reliability
and IT security. Cloud providers claim that
their services enable businesses to access new
administrative and service delivery models which are
more agile and responsive to changes in consumer
behaviour and demand, as well as enabling smooth
migration of data and services between software
generations. For organizations that lack the
necessary in-house IT staff to protect and secure
their systems, cloud-based solutions may also offer
more effective and systematic backup of data, and
access to the specialized IT expertise of the cloud
service provider.


However, as stressed above, such potential gains
cannot be guaranteed merely by migrating data or
applications to the cloud. Businesses with highly
variable data-management requirements, or with
complex customer relationships requiring frequent
interaction, can be expected to benefit more than
others from the inherent flexibility of cloud provisioning.
SMEs in manufacturing and non-ICT services, which
are generally too small to maintain skilled IT personnel,
may gain by reducing their direct IT costs and relying
on services rented from the cloud.


At the same time, the cost of additional services and
facilities procured from cloud providers may undercut
some of the cost savings that would otherwise be
achieved. In most cases, there will be some initial
costs in cloud migration which must be factored into
cost–benefit assessments, and which may inhibit
smaller firms from taking the plunge.


As is generally the case with new technology, potential
benefits such as those described above can only be
fully realized if and when management and operational
systems are re-engineered and staff retrained to take
advantage of them. Legacy administrative processes
need to be overhauled and realigned in ways that
optimize the capabilities of cloud applications and
communications costs. Legacy hardware and
software may need to be written off. In-house IT
departments will need to be restructured – a matter
of ensuring the continued availability of essential skills
as well as shedding those that are no longer needed.


The need for reorganization and re-engineering of
business processes can involve significant migration
costs that businesses and other organizations need
to juxtapose against financial benefits when assessing
the net benefits.


Another important aspect in this context is the
development of industry standards. Cloud computing
is relatively new and changing rapidly, with new
services and modalities of provisioning continuously
becoming available. Different cloud service providers
offer different approaches, based on different business
models, different capabilities and different customer
profiles. Each seeks to secure advantage in a growing
marketplace by differentiating its service from its
competitors, often using proprietary systems. From
the perspective of cloud service customers, however,
standards are critical to facilitate more competition
between different cloud service providers. Customers
can only make effective use of multiple cloud providers if
those providers’ services are interoperable, and they can
only migrate data and applications located in the cloud
operated by one provider to another, to take advantage
of lower prices or better provisioning, if they are not
locked into proprietary standards. The limited degree of
standardization at present contributes to the reservations
that some businesses have about migration.6


In addition to cost implications, two particularly
important concerns for potential cloud service
customers are related to data security/privacy and
to management control of data transferred to the
cloud. A shift towards cloud computing can imply a
loss of control over applications and data when the
computing is conducting on a server that is not owned
by the user. It may not be possible, for example, for
cloud service customers to know where data are
processed and stored, to ensure that data are not
shared or manipulated in ways that lie outside their
cloud agreements, or to see the software source
code related to the services that they have bought.
Governments and enterprises are wary of the risk of
becoming dependent on large cloud service providers,
especially if their headquarters and facilities lie in another
jurisdiction. Concerns related to data protection and
privacy are compounded by the possibility that third
parties (including foreign Governments) could gain
access to sensitive national, business or personal
data. Such aspects have received increased attention
as a result of revelations that some Governments may
have with cloud providers to access data passing
through their servers.7 There is furthermore the risk
that cloud service providers may exercise market




10 INFORMATION ECONOMY REPORT 2013


power to lock clients into relationships which might
become progressively less satisfactory and more
costly to them.


The feasibility of cloud computing depends to a great
extent on the availability of reliable, affordable, high-
quality communications networks. In practice, this
means broadband networks that link all parties in the
supply chains described in the following section on
the cloud economy ecosystem. Variations in access
to such networks, and in their quality, in different
countries fundamentally affect the ability of firms and
other organizations to participate successfully in the
cloud economy. Many low-income countries suffer
from limited or inadequate access to high-quality and
affordable broadband infrastructure, as well as from a
shortage of relevant IT skills (chapter II). Against this
background, it is essential to place any discussion of
the cloud alongside broader considerations of how
to strengthen ICT readiness. Existing bottlenecks in


traditional areas of ICT deployment can act as an
effective barrier to cloud adoption.


From a supply-side perspective, developing countries
will also take into account to what extent their local
enterprises can play an active role in the cloud
economy, either by offering their own related services
or by adding value to existing cloud services. The
advent of the cloud may influence the growth and
trajectory of local IT sectors, enabling some enterprises
(including new start-ups) to flourish while undercutting
the business models of others.


D. THE CLOUD ECONOMY
ECOSYSTEM


The “cloud economy ecosystem” encompasses the
deployment and impacts of cloud computing and
cloud services within the wider information economy
and, thereby, their relevance to national economic


Cloud service customers


Citizens/
consumers SMEs


Large
enterprises


Cloud service providers


Inter -cloud service
providers


Cloud aggregators,
integrators and


brokers


Governments


Cloud service providersL
eg


al
a


nd
re


gu
la


to
ry


fr
am


ew
or


k


Advertisers


Communications network providers
and power suppliers


Source: UNCTAD.


Figure I.2. Main stakeholders and market relationships in the cloud economy




11CHAPTER I : THE CLOUD ECONOMY ECOSYSTEM


development. The cloud economy is more than cloud
computing, just as the information economy is more
than ICTs. Its ecosystem includes a complex set of
relationships, synergies and interactions between
technology and business, governance and innovation,
and production and consumption, which involve
different stakeholders and which contribute in different
ways to economic and social development. It is how
this ecosystem evolves, rather than the potential of
the technology alone, that will determine outcomes
for developing countries.


At the heart of the cloud economy lies the interface
between a rapidly evolving technology and those –
in business, Government and other organizations
– that seek to use it to achieve administrative and
commercial goals with greater efficiency and at lower
cost. A number of attempts have been made to
develop models of the cloud economy ecosystem,
which are valuable for different explorations of the
cloud and its impact on society and economy.8 The
model proposed in this report is primarily concerned
with the relationships between leading stakeholders in
the cloud economy and how those relationships can
influence the cloud’s potential in developing countries.


As with all market ecosystems, the principal
relationships revolve around the supply and demand.
Figure  I.2 illustrates schematically the various
relationships that connect cloud service providers
with cloud service customers and other entities in the
cloud economy.


Cloud service providers in this ecosystem are those
businesses that own the cloud computing centres
and other infrastructure that form the cloud, and make
services, platforms and/or infrastructure available
through them to cloud service customers and users.
Some providers specialize in one or other provisioning
relationship – IaaS, PaaS or SaaS – while many of
the larger providers offer customers all three. As will
be elaborated in chapter II, cloud service provision is
currently dominated by a few large IT corporations,
mostly based in the United States, companies which
can offer cloud service customers the benefits of
lower costs arising from economies of scope and
scale which are derived from global operations in the
cloud and other IT sectors.


Inter-cloud service providers act as intermediaries and
rely on one or more other cloud service providers to
offer IaaS, PaaS or SaaS services to cloud service
customers. Where this is the case, a cloud service


customer will have a contractual relationship with an
inter-cloud service provider, which in turn will have a
contractual relationship with another cloud service
provider. An example of an inter-cloud service provider
of this kind is Salesforce.com, a provider of customer
relations management and other cloud applications
with over 100,000 customers worldwide, and which
runs its products from data centres managed by co-
location companies, such as Equinox.


Cloud service customers as defined here, include all
those that acquire cloud services directly from cloud
service providers and inter-cloud service providers.
Cloud service customers are diverse – including
citizens, consumers, enterprises and Governments
– and have varying requirements and objectives
in making use of the cloud. A variety of different
relationships results between cloud providers, cloud
service customers and end users of cloud services,
depending on the type of customer involved and the
characteristics of the applications.


Cloud service customers obtain cloud services in
diverse ways:


(a) Customers – not just individual citizens but
also SMEs, larger businesses, Governments
and other organizations – may access some
services free of charge directly from cloud
and inter-cloud service providers. This applies
particularly to consumer-oriented, mass-
market services, such as search facilities,
webmail and online social networks that are
often financed by advertising revenue (indicated
to the left of figure I.2). The user interface for
these services provides an ideal platform for
advertising placement, while data mining of the
accumulated information that service providers
hold about many millions of end users enables
much more precise targeting of advertising
than is possible in print and broadcast media.
Advertisers are, therefore, also significant
stakeholders in the cloud economy.


(b) Other cloud services (IaaS, PaaS or SaaS) are
provided on a fee-paying basis, whereby end
users pay either a flat-rate subscription or a
variable fee dependent on the level or extent
of service use. Such services may be acquired
directly from different service providers. Larger
businesses, government agencies and other
organizations, however, also procure a wider
range of services that are specifically adapted
to their commercial or administrative needs.




12 INFORMATION ECONOMY REPORT 2013


These more tailored services may also come
from cloud service or inter-cloud service
providers.


As cloud markets become more diverse and
competitive, cloud service customers – especially
enterprises and Governments – are likely to make
use of aggregators, systems integrators and brokers
(sometimes referred to as cloud service partners)
that can help them to identify the best solutions and
integrate services from different cloud and inter-
cloud service providers. This presents a significant
opportunity for new business development at a
national level (chapter III).


The relationships between the various actors in this
cloud economy ecosystem depend at a fundamental
level on some key infrastructure providers. The most
significant of these are the communications network
operators that provide the infrastructure of the Internet.
These businesses – principally telecommunications
operators and Internet Service Providers (ISPs) – gain
from the increased volume of data traffic that results
from cloud adoption. They may also see opportunities
themselves for diversifying into cloud provisioning
and become a cloud service provider (or inter-cloud
service provider) in their own right. Another important
set of stakeholders are the power suppliers. Access
to reliable and affordable electricity is particularly
essential for running large data centres, as well as
being necessary for users to enjoy reliable cloud
services.


The development of the cloud economy is influenced
by stakeholders that fall outside these core
commercial relationships. Governments are important
actors in the cloud economy in several ways. Some
Governments not only buy services from the market
(acting as cloud service customers), but also act as
cloud service providers to their own private clouds or
even to third parties (see chapter III). More generally,
through their policies and actions, they shape the legal
and regulatory frameworks that can foster or constrain
the migration to the cloud by businesses and other
organizations within their jurisdictions. For example,
data protection and other legislation set limits around
the extent to which data can be shared or stored
outside national boundaries.


Government departments can set benchmarks through
their own adoption of the cloud, their confidence or
otherwise in it encouraging or discouraging that of
others. Governments which develop their own cloud
strategies across the whole of government show


that they are taking the opportunities and challenges
of the cloud seriously. Development partners may
also play a part in encouraging them to do so. The
roles of government and development partners in
relation to the cloud economy are discussed, with
recommendations, in chapters IV and V.


Finally, it is worth noting that the impact of the cloud
on end users does not stop at those who make direct
use of cloud services. Many people and organizations
that do not directly use cloud services are still affected
indirectly by the use of them by Governments and
businesses to manage public and commercial services
that are important to them.


E. IMPLICATIONS
FOR DEVELOPING
COUNTRIES


There is currently a general debate on the significance,
value and risks of cloud computing. This is particularly
the case for developing countries. On the one hand,
it can be seen as a major opportunity, representing
the next major step in the Internet’s evolution into the
most important driver of the world economy. This view
rests upon a belief in the power of technology as a
principal driver of social and economic outcomes.
On the other hand, there are important factors that
suggest caution with regard to the cloud’s potential
value for developing countries, including privacy-
related concerns. Processes of adoption and diffusion
are shaped by many complex human as well as
technical factors, and past predictions of rapid growth
in IT and services – in areas such as e-commerce –
have not always fulfilled their expectations.


Nevertheless, cloud computing is becoming
increasingly important in the delivery of public and
private services around the world. Even where the
cloud has not yet been adopted substantially by
Governments and businesses, widespread use
of mass-market cloud services by citizens has
established the significance of a cloud economy that
will become more important over the next few years.


Governments and businesses in middle-income
and low-income countries should, therefore, actively
seek to assess the relevance and potential of the
cloud for their particular context. At a practical level,
they need to develop a realistic understanding of
the opportunities and risks associated with cloud
provisioning of internal administration and service
delivery before making commitments that might be




13CHAPTER I : THE CLOUD ECONOMY ECOSYSTEM


beneficial but also costly and difficult to reverse. At a
policy level, they need to understand the implications
of cloud computing for national economic outcomes
in order to consider appropriate policy and regulatory
changes. This Report offers a frame of reference to
help them in considering these issues.


These are still early days for the cloud economy in
developing countries. Most of the literature about cloud
computing is concerned with its perceived potential
rather than with what actually has happened or is
happening. Moreover, most of the literature that has
considered cloud computing in developing countries
has focused on emerging markets, particularly larger
countries with growing economies and relatively large
installed IT. It is important in assessing the potential
of cloud computing for developing countries, not to
consider only those that are best equipped to take
advantage of the cloud economy, but also the wider
diversity of developing countries, their economies
and capabilities, including least developed countries
(LDCs) and other low-income countries.


The analysis in this report draws on the limited
evidence that has been published about cloud
adoption in developing countries and on selected
individual case studies. There is, however, a need
for more research into current experience on the
ground, including changing attitudes and perceptions,
and into new thinking about the cloud’s potential in
developmental contexts. This should include the
needs of developing-country SMEs, the potential
value and impact of the cloud for local ICT sectors,
and the ability of Governments to provide e-services.


Potential applications of the cloud differ according
to customer needs. It is equally important, therefore,
to pay attention to the diversity of businesses within
developing countries, from transnational corporations
(TNCs) to SMEs, and to both users and producers
of cloud services. While proponents of the cloud
argue that its most efficient use requires full freedom
of location and free movement of information – that
the economies of scale achievable make global
corporations the most cost-effective and efficient
cloud providers – other approaches, extending beyond
mere efficiency, may be appropriate or better suited to
particular circumstances. Migration to the cloud may
incur new costs and there is little empirical evidence to


date on the total cost of paying as you go for software
and other applications. In addition, the cost dimension
needs to be weighed against other considerations, for
example, those related to data protection and privacy.


Three sets of research questions emerge from this
context, and form the basis for the investigation in this
Report:


(a) Thecurrentstatusof thecloudeconomy
indevelopingcountries. How far are cloud-
based approaches being adopted, which
service and deployment models are proving
attractive and/or effective, and what specific
drivers and barriers stimulate or constrain
development and adoption? This is addressed
in chapters II and III.


(b) The potential impacts of the cloud
economy for different stakeholders in
developingcountries,intheshort,medium
and longer term. Which developing countries
and developing-country businesses are best
equipped to take advantage of the opportunities
presented by the cloud economy? What
opportunities and challenges do cloud models
pose for Governments, enterprises of different
size and for local ICT sectors? Chapter III
examines the available evidence in this context.


(c) The policy and regulatory approaches
which will facilitate the adoption and
effective utilization of the cloud, to
maximize its value and minimize negative
impacts. How should Governments in
developing countries assess the cloud
economy in their specific national contexts?
What investments and policy interventions
should they undertake, and how might these be
supported by international agencies, including
UNCTAD? Chapters IV and V are devoted to
this set of questions.


The Information Economy Report 2013 seeks to
address the challenges of evidence and interpretation
described above and to respond, at least tentatively,
to these research questions. While it is too early to
draw many firm conclusions, it hopes to contribute
to a nuanced discussion of the potential role and
value of the cloud economy among policymakers and
business leaders across developing countries.




14 INFORMATION ECONOMY REPORT 2013


NOTES


1 See, for example, “Cloud computing is a trap, warns GNU founder Richard Stallman“, The Guardian, 29 September
2008, available at http://www.theguardian.com/technology/2008/sep/29/cloud.computing.richard.stallman
(accessed 2 October 2013).


2 See, for example, “The cloud builds up steam“, The Financial Times, 6 June 2013, available at http://www.ft.com/
cms/s/0/e2b826a2-ce20-11e2-8313-00144feab7de.html#axzz2chNG8oKc (accessed 2 October 2013).


3 This definition was agreed by the Collaborative Team between ITU Telecommunication Standardization Sector
(ITU-T) and ISO on cloud computing vocabulary. It was expected to be finally approved in September 2013.


4 For example, by providing software applications as a service, users can access only those that they need as and
when required, rather than buying a full package of software applications and installing all on a particular computer.


5 Special reference is sometimes also made to Business Process as a Service (BPaaS), which builds on the other
three categories of cloud services. It can be seen as a cloud-based variation of more traditional business process
outsourcing (BPO) and is expected to grow in the future.


6 At present, the challenges of standardization are being addressed by a large number of organizations, including
industry associations, ISO, ITU and the Distributed Management Task Force. See, for example, http://cloud-
standards.org/wiki/index.php?title=Main_Page (accessed 3 October 2013).


7 See also chapter IV.


8 One such model is built around three layers of “technical architecture” – provision of cloud services based on
data centres, provision of networks based on communications infrastructure, and the manufacture and supply of
terminal devices (Kushida et al., 2012). This is particularly useful for exploring the relationship of the cloud to the
global ICT industry, and for analysing the strategies of major cloud providers.




TRENDS
IN THE CLOUD
ECONOMY
AND RELATED
INFRASTRUCTURE2


The market for cloud services is growing rapidly, but is still small in developing
countries. While the supply side of the cloud economy is dominated by global
cloud service providers headquartered in the United States, various regional
and local players are also emerging in different parts of the world. The ability
to leverage the opportunities created by cloud computing – on the supply as
well as demand side – is greatly influenced by the quality of ICT infrastructure.
As the shift towards the cloud continues, the digital divide becomes less a
question of basic access and more about the quality of use.


This chapter examines recent developments in the cloud economy and
related trends in the infrastructure needed for cloud services. Leading cloud
providers and other stakeholders in the cloud economy are identified. The
performance of different countries with regard to their readiness to benefit
from the cloud economy is discussed, with special emphasis on broadband
infrastructure for effective usage of cloud services of varying degrees of
sophistication. The analysis furthermore considers the cost of broadband
access and the spread of data centres and Internet exchange points (IXPs).
The chapter concludes with a summary of the main findings.




16 INFORMATION ECONOMY REPORT 2013


A. TRENDS IN THE
CLOUD ECONOMY


This section reviews various estimates of the size
and growth of the cloud computing market, noting
the importance of advertising revenues generated in
connection with the provision of public cloud services
in addition to fees charged for the direct use of cloud
services. It finds that the markets for both public and
private clouds are expected to grow considerably in
the next few years. This is followed by a discussion on
the leading global cloud providers and other relevant
actors in the cloud economy ecosystem, and on the
link between cloud computing and international trade
statistics.


1. Cloudmarketsandtraffic


Cloud market classifications are highly fluid with varying
scope, depending on whether analyst definitions
are applied or the wider ecosystem is considered.
As noted in chapter I, there is general agreement
regarding three standard categories of cloud services:
IaaS, PaaS and SaaS. There are basically two main
models of creating revenue from the provision of these
cloud services to end users: service provision against
a flat-rate or variable subscription fee dependent on
the level or extent of service use, and advertising.


It is estimated that sales of worldwide public cloud
services were worth $111 billion in 2012.1 According
to this source, the largest revenue segment by far
was advertising, which accounted for more than
$53 billion. With regard to the fee-generated revenue
for the three main cloud service categories, SaaS was
the largest category (about $17  billion), followed by
IaaS (about $6  billion) and PaaS (about $1  billion).2
Forecasts of the future growth of cloud computing
differ significantly. For example, while one consultancy
firm anticipates that the total market for public cloud
deployments of IaaS, PaaS and SaaS will grow from
$14  billion to $43  billion between 2010 and 2015,


another expects that the market will be twice that size
by 2015 (table II.1). Discrepancies in projections may
reflect different methodologies as well as difficulties in
predicting the speed at which the cloud computing
phenomenon will evolve. Nevertheless, most forecasts
suggest that cloud adoption will continue to expand
rapidly over the next few years.


Estimates of the value of private cloud services also
vary. According to one private source, private cloud
revenue was estimated at around only $5  billion in
2012, and expected to grow to about $24 billion by
2016.3 According to another private source, however,
private cloud provisioning generated more than
€40 billion ($53 billion) in 2012, and it was expected
to rise to about €75 billion ($99 billion) by 2016 (Pierre
Audoin Consultants, 2013). Whereas SaaS dominates
the market for public cloud services, IaaS is the
dominant feature for private cloud services. There
also appear to be different preferences among cloud
service customers in different regions. For example,
public cloud provisioning, and SaaS in particular,
is most frequently used in North America and the
United Kingdom, whereas French enterprises invest
mainly in private clouds, preferably hosted by French
companies (Pierre Audoin Consultants, 2013).


As stressed in chapter I, the cloud economy has wider
implications. A broader market perspective would
encompass activities such as access, support and
usage of web-enabled services. This includes, for
example, IT equipment purchased by cloud providers in
order to offer their services and connectivity purchased
by businesses, Governments and households to
access cloud services and user purchases of services
delivered from the cloud.4 Such a broader view would
generate a considerably higher estimate of the value
and potential of the cloud economy.


The value of the global ICT sector in 2011 has been
estimated at €2.9 trillion ($4 trillion), a figure forecast
to rise to €3.2  trillion ($4.5  trillion) by 2014 (IDATE
Foundation, 2012).5 Telecommunication services


2010 2015


SaaS Paas Iaas Total SaaS Paas Iaas Total
Gartner 10


(70.9%)
1.3


(9.2%)
2.8


(19.9%)
14.1


(100%)
21.3


(49.2%)
2.4


(5.5%)
19.6


(45.3%)
43.3


(100%)
Forrester 13.4


(91.1%)
0.3


(2.2%)
1


(6.7%)
14.7


(100%)
78.4


(83.5%)
9.8


(10.4%)
5.8


(6.1%)
94.1


(100%)


Table II.1. Estimates and forecasts of cloud revenue, 2010 and 2015 (Billions of dollars)


Source: Berry and Reisman, 2012.
Note: This table only includes revenue from the services shown.




17CHAPTER II : TRENDS IN THE CLOUD ECONOMY AND RELATED INFRASTRUCTURE


accounted for 36  per cent of this, making it the
largest segment, followed by software and computer
services. All segments are in some way affected by
cloud computing. For example, the demand for higher
bandwidth will drive telecommunication services
revenue, although revenues from voice services could
be affected as more people switch to cloud-based
Voice over Internet Protocol (VoIP) applications.
Demand for equipment and computer hardware,
particularly data servers and network equipment, will
rise as more services move to the cloud. Television
services will be affected by increasing demand for
video streaming.


The shift to the cloud is generating considerable
growth in data traffic, requiring further investment in
broadband capacity. Consider that during an average
minute in 2012, Google received two  million search
requests, Facebook users shared around 700,000
content items and Twitter sent out 100,000 tweets.6


According to some estimates, cloud traffic crossed
the zettabyte7 milestone in 2012 (figure II.1, left). The
same source predicts an annual growth rate of over
40 per cent between 2011 and 2016 (Cisco Analysis,
2012). Some 40 per cent of all traffic reflects storage
requests (retrieving or sending data from/to the cloud).
By 2016, cloud traffic is expected to represent as
much as 64 per cent of all data centre traffic.8


In 2012, the majority of such traffic emanated from the
Europe and North America, which together accounted
for 60 per cent of the total (figure II.1, right). Asia-Pacific
was responsible for another third, while Latin America


and the Middle East and Africa together accounted
for only 5 per cent. Asia-Pacific is forecast to overtake
North America as the top cloud traffic region by 2014.
Meanwhile, the highest growth rates are expected in
the Middle East and Africa between 2011 and 2016,
albeit from a low base.


2. Leading corporate players in the
cloud economy


Cloud service providers include businesses that
own cloud computing centres and other cloud
infrastructure, and those that provide cloud services,
platforms and/or storage capacity available to cloud
service customers. Identifying the leading cloud
service providers is far from straightforward. In terms
of financial size, some companies do not separate
their cloud activities from other revenues;9 and some
providers might be significant in a particular niche
but small in terms of overall market revenue. There
is furthermore no consensus about whether cloud
revenues should be related only to the cloud service
model categories or to the wider cloud ecosystem. In
terms of the number or size of clients, there is scarce
information. Against this background, this subsection
uses various alternative metrics, such as number of
data servers or geographical client reach, as well as
information from qualitative rankings to identify the top
corporate actors in the cloud economy.10 The pattern
that emerges is one of a concentrated industry with
most of the major providers coming from the United
States.


Figure II.1. Global cloud data centre traffic (2011–2016) and distribution by region (2012)


Source: Cisco Analysis, 2012.
Note: Data for 2012–2016 are estimates.


683


1 181


1 694


2 324


3 166 3 129


0


500


1 000


1 500


2 000


2 500


3 000


3 500


2011 2012 2013 2014 2015 2016


Cloud traffic (exabytes) Distribution of cloud traffic, 2012











Middle East
and Africa


2%










Central and
Eastern
Europe


3%


Asia-Pacific
32%


Latin America
3%


Western
Europe
23%


North
America


37%




18 INFORMATION ECONOMY REPORT 2013


Company
Estimated
number of


servers


Estimated
cloud revenue
(billion dollars)


Comments


Google
(United States) 900 000 50


The Google Cloud Platform includes software applications and data
storage. It has six data centres in the United States and three in
Europe; four more are under construction (three in Asia and one in
South America). Over 90 per cent of revenue stems from advertising
on the web.


Microsoft
(United States) 300 000 n.d.


Offers both commercial and consumer cloud services. Cloud revenue
is earned primarily from usage fees and advertising. Products include
cloud-based versions of Office (365), Skype, Xbox and Azure. Data
centres throughout the Americas, Europe and Asia.


Amazon Web Services
(United States) 250 000 2.1


Launched by Amazon in 2006, providing infrastructure services to
hundreds of thousands of enterprises, Governments and start-ups
in 190 countries around the world. AWS offers over 30 different
services. Data centre locations in the United States, Brazil, Europe,
Japan, Singapore and Australia.


Facebook
(United States) 180 000 5.1


Social network company data centres in the United States and
Sweden. Generates most revenue from advertising and fees from
purchase of virtual and digital goods from platform developers. One
billion users (at least once a month) at December 2012.


Akamai
(United States) 127 000 1.4


Data servers located in 81 countries. Provides content delivery
and cloud infrastructure services, including delivery of conven-
tional content on websites, tools supporting delivery and operation
of cloud-based applications, and live and on-demand streaming
video.


OVH (France) 120 000 n.d. Has eight data centres. Is the leading webhosting company in Europe with more than 400,000 customers.


Softlayer
(United States) 100 000 n.d.


Employs a unified global infrastructure platform to provide on-
demand, hour-to-hour or month-to-month billing through one
portal and one API. Has 13 data centres in the United States, Asia, and
Europe. Has 100,000 devices under management, claiming it is the
largest privately held IaaS provider in the world.


Rackspace
(United States) 79 805 (2011) 1


Offers traditional hosting as well as PaaS. Served 172,510 customers
in 2012.Has nine data centres located in the United States, the United
Kingdom, Hong Kong, China and Australia.


Intel
(United States) 75 000 n.d.


The microprocessor giant has 69 worldwide data centres. Offers
on-demand applications for staff, partners and private cloud
clients, and online netbook applications such as services to help
validate applications and software.


1&1 (United States) 70 000 n.d. Web-hosting provider with around 12  million customer contracts served through five data centres in Europe and the United States.


Table II.2. Top 10 companies by estimated number of servers, 2012


Source: UNCTAD, based on information from company reports and other sources.
Note: n.d. = cloud revenue not disclosed and no reliable estimate exists.
a http://huanliu.wordpress.com/2012/03/13/amazon-data-center-size/ (accessed 2 October 2013).
b http://online.wsj.com/article/SB10001424127887324442304578236000008569908.html (accessed 2 October 2013).
c http://perspectives.mvdirona.com/2012/08/13/FunWithEnergyConsumptionData.aspx (accessed 2 October 2013).
d http://www.akamai.com/html/about/facts_figures.html (accessed 2 October 2013).
e http://www.ovh.co.uk/aboutus/ (accessed 2 October 2013).
f http://cdn.softlayer.com/PS_DataCenterOverview.pdf (accessed 2 October 2013).
g http://www.intel.com/content/www/us/en/it-management/intel-it/intel-it-data-center-solutions.html


(accessed 2 October 2013).
h http://www.1and1.com/Facts?__lf=Order-Tariff (accessed 2 October 2013).




19CHAPTER II : TRENDS IN THE CLOUD ECONOMY AND RELATED INFRASTRUCTURE


Given that the provision of cloud services requires
storage of applications and data in dispersed data
centres, a practical comparable metric is the number
of data servers. Although such information is often
considered commercially confidential, estimates
made by market analysts can serve as a basis. Table
II.2 identifies the top 10 companies ranked according
to the number of worldwide data servers.11 Almost
all operate globally and their headquarters are largely
concentrated in the United States. Some are giants,
such as Amazon, Microsoft and Intel, for which cloud
computing comprises only a relatively small portion
of total revenue. Some (e.g. Google or Facebook)
are large Internet-based companies that derive a
significant share of their revenue from cloud advertising.
Others specialize in specific cloud services, mainly
hosting but also software and application integration
for clients. In terms of revenue, Google is by far the
largest, generating most of its some $50  billion of
revenues in 2012 from its overall web-based services,
almost all from advertising.


At the same time, several inter-cloud service providers
have few or no servers and instead lease capacity
to offer their services to cloud service customers.
For example, the top 10 SaaS providers among
the world’s 100 largest software firms, based on
2011 revenue, are shown in table II.3. Some of the
listed SaaS companies are also active in other cloud
services, such as PaaS (Salesforce.com, Microsoft) or
have a significant number of data servers (Microsoft,
Google). Again, apart from DATEV (Germany), all firms
included in the list are based in the United States.


To broaden the perspective to the cloud economy,
additional companies should be considered. This
includes, for example, major vendors of IT equipment
and software used for the cloud, such as Cisco,
EMC, IBM, VMware and HP (all from the United
States) and major telecom operators that provide the
communication networks, such as AT&T, Level 3 and
Verizon (all three from the United States), BT (United
Kingdom) and Nippon Telegraph and Telephone
Corporation (NTT) (Japan).


Though global cloud service providers operate
across different regions, there is scope for regional
players to compete by providing more localized and
personalized niche services. Regional cloud provision
trends have been marked by alliances between
telecommunication operators and IT companies. This
match unites the operator’s direct contact with the
customer with companies that have specific cloud


expertise. Outright purchases of cloud service firms by
telecommunication operators is one model, but this
has mainly been limited to the North American market
(e.g. Verizon’s acquisition of cloud company Terremark
for $1.4  billion). In other regions, links between
telecommunication operators and cloud providers
have with some exceptions been established primarily
through partnerships.13
• In Africa, Vodacom, a South African telecommu-


nication operator, has partnered with Novell, an IT
provider, to offer cloud services.14 Vodacom oper-
ates across a number of African countries. MTN,
another South African-based operator with subsidi-
aries in several African countries, provides e-mail
and other cloud services to customers in Cam-
eroon, Ghana, Nigeria and South Africa. Seacom,
the company behind the undersea cable of the
same name, has launched a new company, called
Pamoja, to offer SMEs the ability to provide cloud
computing-based services to their customers with-
out the capital outlay such services usually require.
France Télécom has appointed a regional director
for cloud computing in Africa and the Middle East.


• In Latin America, Telefonica (Spain) has partnered
with NEC Corporation (Japan) to offer cloud


Table II.3. Top 10 generators of SaaS revenue among
the global 100 software firms, 2011


Company Country headquarters


2011
SaaS


revenue
(million
dollars)


2011
software
revenue
(million
dollars)


SaaS
revenue
as share


of software
revenue


(percentage)


Salesforce.
com United States 1 848 2 008.7 92


Intuit United States 950 2 456.5 38.7


Cisco United States 831 1 796.9 46.3


Microsoft United States 788 57 668.4 1.4


Symantec United States 572 6 330.3 9


Google Inc. United States 462 575.6 80.3


Oracle United States 446 26 175.9 1.7


Adobe United States 410 4 154.1 9.9


Blackboard United States 396 411.7 96.2


DATEV Germany 395 974.2 40.5


Total 7 098 102 552.2 6.9


Source: PWC. See http://www.pwc.com/us/en/technology/
publications/global-100-software-us-rankings.jhtml
(accessed 3 October 2013).




20 INFORMATION ECONOMY REPORT 2013


services.15 Products include an SaaS called
Aplicateca featuring an app store aimed at
businesses.16 According to Pyramid Research,
the market for cloud services in Latin America
was $4.8  billion in 2012 (Ramos, 2012). The
same company reckons that cloud revenues from
serving SMEs may reach $12.7 billion by 2017.


• In Asia, Telkom (Indonesia) offers its cloud offer-
ing through the IT company TelkomSigma, while
Indosat has partnered with Dimension Data to
launch an enterprise-class public cloud service.17


India’s Tata Communications is another telecom-
munication operator that is expanding into cloud
services. It has developed its own pay-as-you-go
IaaS called InstaCompute.18 Around 1,000  com-
panies are trialling InstaCompute, and some 300
are commercial clients.19


Global cloud services can generally be accessed
directly online by users from any place in the world.
At first glance, it may appear difficult for purely
local providers to compete with regional or global
suppliers of cloud services. However, that presumes
that potential domestic customers are aware of
these global or regional sources or how to use them.
Furthermore, factors such as security, localization and
latency may create a role for local software companies
to create domestic cloud solutions and for brokers and
aggregators to interface between the local market and
regional and global cloud providers (see also chapter
III). In China, for example, Microsoft has entered the
cloud market by partnering with 21Vianet, a Chinese
carrier-neutral provider of data centre services in
33 cities.20 This allows Microsoft to offer its services
with customers using a Chinese company that
guarantees that the data used will remain in China.


Local telecommunication operators have an
important strength through their direct relationship
to the customer. Cloud services need to get to
the customer over the networks operated by the


telecommunications firms. As noted above, many
network operators are forging partnerships with cloud
service companies and this is flowing down to the local
level in some countries through subsidiaries. Individual
domestic telecommunication operators could be at a
disadvantage in terms of cloud expertise. Moreover,
cloud adoption may be slower in countries without a
certain level of competition, where operators are not
part of a global consortium and may lack expertise.


3. Can cloud computing trends be seen
in trade statistics?


The expansion of cloud computing has various
implications for the development of international
trade in both ICT services and ICT goods. The cloud
business model rests in large part on the possibility to
locate large data centres (hardware) in the most cost-
effective location and then to send and receive data
(services), often across borders, between the storage
location and the cloud customers as well as between
different data centres.


However, measuring international trade related to
cloud computing is challenging. Official data only
give a partial impression of related developments.
International classifications do not allow for separate
identification of trade in cloud services (see box II.1),
and it is difficult to determine how much of the trade
in data servers and other communication equipment
that is linked specifically to the provision of cloud
computing. Most analysis on the topic thus also relies
on estimates from specialized market research firms
and industry associations.


(a) Trade in services


In the United States, attempts have been made to
measure the value of trade in cloud services. A first
conservative estimate of the value of that country’s
exports of public cloud computing services in 2010,


Box II.1. Cloud computing services in international classifications


Even though it does not have a subclass or group of products corresponding directly to cloud computing services,
Central Product Classification (CPC) Version 2 is the most amenable international classification for identifying computer-
related services linked to cloud computing. The subclasses of CPC Version 221 mentioned below can be identified as
most relevant. However, international trade in services data are not yet available at this level of disaggregation.


83151 Website hosting services


83152 Application service provisioning


83159 Other hosting and IT infrastructure provisioning services


Source: UNCTAD, based on information provided by the United Nations Statistics Division.




21CHAPTER II : TRENDS IN THE CLOUD ECONOMY AND RELATED INFRASTRUCTURE


arrived at $1.5  billion for cross-border transactions
(or 3.4 per cent of United States exports of services
likely to involve cloud computing)22 and $1.4 billion for
sales of majority-owned foreign affiliates (0.5 per cent
of their total sales in 2009) (Berry and Reisman, 2012).
The estimates are conservative in that they do not
take into account private cloud computing services,
typically developed in-house by individual companies.
Neither do they take into account sales of businesses
whose main activity falls outside the computer and
data processing services and royalties and license
fees sectors but that may nevertheless offer cloud
services.23 In India, for example, many firms use cloud
computing platforms located in Singapore due to lower
bandwidth costs and more robust infrastructure.24


Moreover, these estimates do not fully capture the
value of trade in cloud services financed through
advertising revenue or that are based on other
secondary-content-supported business models.
In these cases, the value created by Internet data
exchange is not closely associated with the financial
transaction and is therefore difficult to seize in trade
statistics.25


Examining the case of the United States is relevant
in this context, as it is the home of most of the
main cloud service providers. At the same time, the
estimation method applied in the case of the United
Sates cannot be easily extrapolated to other countries
and economies that are at a less advanced stage in
terms of both cloud service provision and use.


(b) Trade in goods


The impact of cloud computing on trade in ICT
goods is similarly difficult to gauge. Still, preliminary
information suggests that the value of trade in goods
related to cloud computing may be at least of a similar
magnitude with that of trade in services.


Relevant trade flows include those involving the ICT
devices needed for using cloud services such as
computers and smartphones. In 2012, the value of
world imports of portable computers amounted to
$136 billion. In the same year, world imports of cellular
telephones of all types amounted to $186  billion.26
A certain portion of these devices (computers and
smartphones) is likely to be used for cloud services,
but it is difficult determine its magnitude. Estimates
from market consultancy firms suggest that the
number of smartphones – which are more suited for
accessing cloud services – sold globally is now higher
than that of feature phones.27


From a cloud service production perspective, relevant
trade flows involve ICT goods needed to set up and
run cloud computing services, notably computer
servers and related communication equipment.
Common components of a cloud computing hardware
architecture include the following:28


• Server computers, such as application servers,
chat servers, database servers, e-mail servers, file
servers, gateway servers, load balancers, media
servers, office servers, presentation servers, print
servers, security servers, telephone servers, web
accelerators, web servers and wireless gateway
servers.


• Storage devices, such as disk drives, disk library,
tape drives, and tape library.


• Support hardware, such as air conditioners,
fire suppression, physical security devices, and
uninterruptable power supply.


• Network components such as extranets, grid
networks, the Internet, intranets, local area
networks, mesh networks, metropolitan area
networks, peer-to-peer networks, ring networks,
star networks, storage area networks and wide
area networks.


• Network connectivity devices, such as access
adapters, bridges, cables, caches, connectors,
firewalls, gateways, hubs, modems, multiplexers,
network interface cards, optical amplifiers,
receivers, repeaters, routers, switches and
transmitters.


• Network connections, such as coaxial cables, fibre
optic cables, infrared radiation and twisted wire
cables.


Consider the example of computer servers. This
broad term denotes many types of automated data
processing machines that can be sold in a variety of
configurations. Depending on the product specifics,
computer servers can be imported under any of the
following three HS 2007 classification subheadings:
847141, 847149 and 847150.29 For simplicity this
section uses the example of data reported only
under category HS 847150, which covers computer
servers without keyboard and monitor. This category
includes servers that are more likely to be used in
data centres, as well as some server computers used
for other purposes.30 World exports of this product
category reached $42  billion in 2012, reflecting a
modest 4  percentage point increase from 2008 to
2012 (table  II.4). China was responsible for 31  per
cent of the global export value, developed countries
for 46 per cent and the remaining 23 per cent came




22 INFORMATION ECONOMY REPORT 2013


from other developing and transition economies.
Large exporters with high growth rates over recent
years include China, Hong Kong, China,31 Malaysia,
Mexico and Singapore.32


In terms of import market shares, the United States
accounted for more than one third (36  per cent) of
the global value in 2012, which is considerably higher
than that country’s share of total ICT goods imports
(15 per cent). The United States also displayed some
of the highest import growth rates during the past five
years. This may reflect its prominent role in the cloud
economy. Other developed countries accounted for
another 40 per cent, leaving the remaining 24 per cent
for developing and transition economies.


Market estimates of global sales of computer
servers indicate faster growth for servers that are the
most relevant for the production of cloud services.


According to International Data Corporation (IDC), the
worldwide server systems factory, revenue reached
$51.3  billion33 in 2012, up a modest 6  per cent per
year from 2009.34 This overall number masks diverging
trends for different kinds of servers. Those that are
most used in cloud infrastructure deployments appear
to have grown faster. For example, the share of Linux
servers rose from 17 per cent of all server revenue in
2010 to 20 per cent in the first quarter of 2013.35 In
addition, two types of modular data servers that are
used extensively by cloud computing providers – blade
servers, used mostly in private cloud data centres, and
density optimized servers, used in large public cloud
data centres – saw rapid growth. The market share of
blade servers increased from 13 per cent in 2010 to
almost 18 per cent in the first quarter of 2013. Density-
optimized servers, which represent about 7  per cent
of the server market, registered strong growth in 2013.


Top 15 importers Top 15 exporters


Economy
2012


(million
dollars)


2012 market
share


(percentage)


Annual
growth rate
2008–2012


(percentage)


Economy
2012


(million
dollars)


2012
market
share


(percentage)


Annual
growth rate,
2008–2012


(percentage)


United States 15 199 36.4 22 China 11 471 31.0 13


Japan 2 927 7.0 -4 United States 6 302 17.0 0


United Kingdom 2 223 5.3 -9 Mexico 4 328 11.7 17


Canada 1 983 4.7 0 Czech Republic 3 603 9.7 2


Germany 1 977 4.7 -11 Germany 1 862 5.0 -9


China 1 919 4.6 18 Singapore 1 771 4.8 35


France 1 565 3.7 -10 Hong Kong, China 1 574 4.3 95


Netherlands 1 496 3.6 2 Netherlands 1 445 3.9 -13


Italy 1 045 2.5 -0 Hungary 731 2.0 -11


Hong Kong, China 993 2.4 33 France 624 1.7 -14


Republic of Korea 903 2.2 6 Ireland 540 1.5 -22


Mexico 875 2.1 -12 United Kingdom 396 1.1 -18


Australia 796 1.9 11 Canada 338 0.9 -4


India 688 1.6 26 Japan 337 0.9 2


Switzerland 676 1.6 -5 Malaysia 177 0.5 31


Other economies 6 532 15.6 3 Other economies 1 478 4.0 -1


World 41 798 100.0 1 World 36 978 100.0 -1


Table II.4. Top importing and exporting economies of computer servers of the type described under HS 847150,
2008–2012


Source: United Nations Commodity Trade Statistics Database.
Note: Discrepancies between world exports and imports can be due to a variety of factors, such as different valuation, differences


in the time of recording, misclassification on importer or exporter side, exchange rates and markups during consignment.




23CHAPTER II : TRENDS IN THE CLOUD ECONOMY AND RELATED INFRASTRUCTURE


Computer servers are only one of the identified
components of the typical cloud computing hardware
architecture. To get a more complete picture, it would
be necessary to also analyse trade in goods such as
storage devices, network connectivity devices and
network components of the type closely associated
with providing cloud computing.


(c) Final observations


It is difficult to identify cloud computing trends in
international trade statistics. Conservative estimates
for the United States suggest that cloud services may
today represent about 3–4 per cent of all exports of
computer and data processing services. Given the
dominance of United States cloud service providers
and the relatively high level of cloud adoption in
that country, the share is likely to be lower in most
other countries. Compared with worldwide sales
of public cloud services, estimated at $111  billion
(section  II.A.1), the value of cloud service exports
remains modest.


Meanwhile, trade in computer servers, storage
devices and network components, which are needed
for setting up and running cloud services, represents
a non-negligible part of ICT goods trade, probably at
least similar in magnitude with the value of trade in
cloud services. To determine the exact value, however,
it is necessary to single out those ICT goods that are
used mainly for cloud computing. Aggregated data
often mask diverging trends for different kinds of ICT
goods within the same generic classification.


For both goods and services, estimating the cloud-
related trade value is also made difficult by the evolving
nature of cloud computing business models, for some
of which the dollar value of transactions is decoupled
from the Internet data exchange itself. More research is
clearly needed to better understand the implications of
cloud computing for international trade, both in goods
and services, and to identify potential opportunities
and niche markets for developing countries.


B. TRENDS IN
CLOUD-RELATED
INFRASTRUCTURE


This section reviews ICT infrastructure trends with a
special focus on aspects of the broadband landscape
that may affect the ability of developing countries to
adopt and benefit from cloud-based applications. It


begins by reviewing the concept of “cloud readiness”,
highlighting the ICT network components and
services considered important for the use of cloud
services. This is followed by a review of broadband
infrastructure, quality and pricing that impact the
utilization of cloud services.


1. Factors determining the cloud
readiness of countries


In order to identify factors that influence the ability of
countries to benefit from the cloud economy, it is useful
to consider various attempts that have been made to
assess the cloud readiness of different countries and
economies. Several indices available use different
methodologies. Some focus on infrastructure, while
others rely on a broader set of both qualitative and
quantitative indicators. The geographical coverage is
currently limited in available indices. Moreover, there
is as yet no indicator that shows actual level of cloud
adoption:
• The Enterprise Cloud Readiness Index, designed


by Pyramid Research, defines enterprise cloud
readiness “as the degree to which service providers
in a given country can potentially leverage cloud
services for the enterprise segment” (Pyramid
Research, 2012). The index covers 49 countries
and is purely quantitative, based on nine economic,
demographic and ICT infrastructure indicators
(annex table 1).


• The Global Cloud Computing Scorecard of the
Business Software Alliance ranks 24 developed
and developing countries in seven categories that
measure “preparedness to support the growth of
cloud computing” (Business Software Alliance and
Galexia, 2012). The scorecard uses both qualitative
and quantitative data (annex table 2).


• The Asia Cloud Computing Association Readiness
Index ranks 14 economies in the Asian region
across 10 categories critical to “successful
deployment and use of cloud computing
technology” (Asia Cloud Computing Association,
2012). It includes both qualitative and quantitative
factors (annex table 3).


• The Cisco Global Cloud Readiness tool is based
on only three indicators, all related to broadband:
download and upload speeds and latency. These
indicators are seen as reflecting a country’s ability
to optimally support different levels of cloud
services (basic, intermediate and advanced) for
both fixed or mobile broadband (Cisco Analysis,




24 INFORMATION ECONOMY REPORT 2013


Figure II.2. Getting to the cloud


Figure II.3. Incremental lit capacity of submarine cables on major routes (Tbps)


Download speed Upload speed


Bandwidth


Latency


Servers


International
Cloud


National
Cloud


Domestic
data centre


National
backbone


International
connectivity


User
connection


IXP


Source: UNCTAD.


0


2


4


6


8


10


12


14


1997–2002 2002–2007 2007–2012


Tb
ps


Transatlantic


Trans-Pacific


United States–Latin
America
Intra-Asia


Europe–Asia via Egypt


Source: TeleGeography. Available at www.telegeography.com (accessed 3 October 2013).
Note: Lit capacity is the actual traffic-carrying capability of the system today, based on what has been equipped to date.




25CHAPTER II : TRENDS IN THE CLOUD ECONOMY AND RELATED INFRASTRUCTURE


2012). Cisco does not compute an actual index
value but publishes the top 10 countries for
both fixed and mobile performance based on an
unspecified combination of the three indicators
(annex table 4).


Given the novelty of the cloud computing phenomenon,
more time is needed to refine assumptions about
what drives cloud computing and to establish the best
indicators for representing it. As none of the indices
include statistics related to actual cloud adoption,
it is difficult to develop a model to show objectively
the impact and thus relevance of specific factors.36
Considering infrastructure-related factors, broadband
subscriptions are used across the first three but are
not considered relevant by Cisco (annex table 4). On
the other hand, broadband speed and latency are
included in the Asian index and by Cisco (table II.5).


Infrastructure-related factors influencing the
development of cloud computing include international
and national backbones that aggregate user data
for transmission back and forth to cloud services
as well as end-user broadband access. These
components are interrelated; if cloud services are
hosted domestically, international connectivity may
not be as critical (figure II.2). In the next subsections,
special attention is given to international broadband
connectivity, national backbone and Internet
exchanges, broadband penetration, quality of service
(QoS) and the affordability of broadband.


2. International broadband connectivity


International Internet bandwidth, which is critical for
accessing data servers located abroad, has grown
significantly in recent years. According to the telecom
market research firm TeleGeography, it increased
annually by 53  per cent between 2007 and 2012.
Expanding demand has been met by the construction
of new fibre optic networks and the upgrading of
existing ones. As much as 54 terabits per second
(Tbps) of capacity were added during this period
(figure  II.3). Demand in developing countries is rising
the fastest. A key driver has been cloud services
and new traffic to data centres over international
connections between users and overseas locations
where applications and data are located. However, this
is mitigated somewhat by caching widely accessed
content on national networks in efforts to improve
performance and reduce international bandwidth
costs (TeleGeography, 2013).38


ACCA BSA Pyramid Cisco


Download speed ✔ ✔


Upload speed ✔ ✔


Latency ✔ ✔


Fiber optic subscrip-
tions




International band-
width


✔ ✔


Wired broadband
subscriptions


✔ ✔ ✔


Mobile broadband
subscriptions


✔ ✔ ✔


Internet users ✔


Computers ✔ ✔


Smartphones ✔


Table II.5. Infrastructure-related indicators used in
different cloud readiness indices


Source:ACCA = Asia Cloud Computing Association Readi-
ness Index; BSA = Business Software Alliance’s
Global Cloud Computing Scorecard; Pyramid
= Pyramid Research Enterprise Cloud Readi-
ness Index; Cisco = GCI Cloud Readiness tool.
Excluding compound indices and subjective indicators
(e.g. opinion surveys).


Note: UNCTAD analysis.


The availability and affordability of international
broadband bandwidth in Africa, which used to lag
far behind in this area, have increased enormously
as a result of the landing of competing submarine
fibre cables on the continent’s eastern and western
seaboards. The principal barriers to affordable
connectivity now lie in regional backbone and domestic
backhaul networks, many of which are seeing
upgrade investments. Significant improvements to
terrestrial infrastructure, comparable to those that
have occurred with submarine connectivity, could
enable a step change in the adoption of cloud services
by businesses, especially in Africa.


3. National backbone, Internet
exchange points and data centres


National backbones, Internet exchanges and data
centres are essential facilities for transmitting and
processing cloud data streams. National backbones
are vital for getting data to their destination. If the
backbone is not robust, the users will not achieve the
full speed of their local connection. There is extensive
anecdotal evidence that the lack of connectivity




26 INFORMATION ECONOMY REPORT 2013


to high-speed national backbones in rural areas
is a concern in many countries. However, unlike
international bandwidth, globally agreed metrics and
country data for quantifying and comparing national
backbone capacity are lacking.39


As the name implies, IXPs are facilities where ISPs
co-locate equipment to exchange national traffic. This
reduces the cost of international bandwidth, since
domestically destined traffic does not need to transit
overseas. It also improves performance through lower
latency (see below). The establishment of IXPs can
help build national human capacity in networking
skills. Once traffic reaches a certain level, IXPs can


allow content providers to cache data in the country.
With growing expertise and traffic, a next step may
be to locate data servers at IXPs and offer domestic
cloud services.


There were in June 2013 some 397 IXPs in the
world (figure  II.4).40 More than 60 per cent of them
were located in Europe or North America. In Africa,
which was home to only 6  per cent of the world’s
IXPs, such exchanges would dramatically improve
the web experience through better performance. For
example, a study of IXPs in Kenya and Nigeria found
that they sharply improved latency, reducing it from
over 200  milliseconds (ms) to less than 10 ms on
average.41


Like public IXPs, private peering facilities similarly offer
exchange of traffic but generally charge some or all
participants for this privilege. Unlike most public IXPs,
however, private peering facilities are profit-oriented
and typically operated by a single company. They also
far outnumber public IXPs; according to PeeringDB,
there were some 1,221 private peering facilities around
the world in July 2013.42 Another distinction between
IXPs and private peering facilities is that the latter
often include centres where companies can lease
or co-locate data servers. Such data centres are the
concrete manifestation of the cloud, as they host the
servers that store and process cloud data. This critical
piece of cloud infrastructure requires special operating
conditions. The most important condition is a stable
energy supply to keep the servers available non-stop
and to power the cooling systems so the servers do


Figure II.5. Distribution of co-location data centres, by group, 2013 and secure Internet servers per 1 million people,
2012


Source: See annex table 5.
Note: Regions in right-hand chart refer to developing economies only.





Developing
economies


13%


Economies in
transition


2%




Least developed
countries


0%






Developed
economies


85%


Co-location data centres, 2013


1


3


4


6


36


48


184
1 067


0 200 400 600 800 1 000 1 200


Least developed countries


Middle East and North Africa


East Asia and Pacific


Sub-Saharan Africa


Latin America and Caribbean


Europe and Central Asia


World


High-income countries


Secure Internet servers per 1 million people, 2012


Africa


6%


Europe


38%


North


America


23%


Latin
America


12%


Asia-Pacific
20%


Figure II.4. Distribution of IXPs, by region, June 2013
(Percentage)


Source: Packet Clearing House, see https://prefix.pch.net/
applications/ixpdir/summary/growth-region/?sort1=
bandw id th&so r t2=_pe rcen t_change&o rde r=
desc (accessed 10 October 2013).




27CHAPTER II : TRENDS IN THE CLOUD ECONOMY AND RELATED INFRASTRUCTURE


not overheat.43 Other considerations include the price
and the security of energy supply.


Several demanding conditions need to be met for data
centres to be financially viable in a location. Risk is a
factor influencing decisions by companies on where
to locate their servers or build data centres, and many
developing countries are perceived to pose a higher
risk level.44 The cost of energy and ensuring supply on a
reliable basis can easily outweigh the demand for data
centres. According to one source, as much as 85 per
cent of data centres offering co-location services are
in developed economies (figure II.5, left).45 This digital
divide in data centres is also reflected in the availability
of servers. There were over 1,000 secure data servers
per one million inhabitants in high-income economies
in 2011, compared with just one per million inhabitants
in LDCs (figure II.5, right). Countries without local data
centres need to access overseas servers for cloud
services, increasing international bandwidth costs
and impacting latency-related performance. Further,
some cloud service customers such as Governments
require secure cloud environments that are likely best
provided through national data centres.


4. Broadband infrastructure trends


There were an estimated 2.1  billion broadband
subscriptions in 2012, with mobile broadband
accounting for over 70 per cent of the total (figure II.6).
In terms of relative importance for cloud access, the
statistics should be interpreted carefully. As discussed
below, fixed broadband is often of a higher and more


consistent speed, with less latency and typically shared
in a household or enterprise, thus having many more
users than suggested by the number of subscriptions.
By contrast, mobile broadband is typically used by a
single user, and so-called “active” subscriptions may
sometimes include subscriptions that are “broadband
capable” but may not actually be used to access data.46


In 2012, fixed broadband subscriptions around the
world stood at 632 million (figure II.6, left). Growth has
been fairly flat, with global fixed broadband penetration
rising only about one per 100 people for the past
few years. Although the number of subscriptions in
developing economies surpassed those in developed
economies in 2012, the gap in penetration keeps
widening. Fixed broadband penetration in developed
economies was almost five times higher than in
developing regions in 2012. In LDCs, the total
number of fixed broadband subscriptions was only
1.3 million – about the same as in Singapore – and
average penetration was just 0.2 per 100 people.
Though penetration in transition economies remained
significantly below developed economies, it was
higher than the world average.


The broadband user experience is greatly influenced
by the kind of connection. The three main fixed
broadband technologies in use around the world
are digital subscriber line (DSL), cable modem and
fibre optic. Although these technologies account for
the vast majority of subscriptions, satellite remains
important for isolated rural areas where terrestrial-
based broadband is not available (box  II.2). DSL
delivered over copper telephone lines accounts for


Figure II.6. Global broadband subscriptions and distribution by fixed or mobile technology


Source: ITU, World Telecommunication/ICT Indicators Database.























0


500


1 000


1 500


2 000


2 500


2007 2008 2009 2010 2011 2012


Fixed (wired) broadband subscriptions


Active mobile broadband subscriptions


M
ill


io
ns


Active mobile
broadband


subscriptions
70%


Fixed (wired)
broadband


subscriptions
30%


Broadband subscriptions by technology, 2012




28 INFORMATION ECONOMY REPORT 2013


the largest share of fixed broadband technology
(figure  II.8, left). These technologies offer different
theoretical speeds depending on various factors, such
as hardware versions used and distance of user from
exchange. Maximum theoretical speeds vary between
technologies, with fibre supporting the highest velocity
(figure II.8, right). Further, fibre is generally available with
symmetrical upload and download speeds, whereas
the other technologies are predominantly asymmetric,
with higher download speeds. Average speeds in most
countries are far below these theoretical maximums
(see annex table 5).


Fibre is the most cloud-friendly of the fixed broadband
technologies since the higher the speed, the better
cloud services perform.47 Gigabit (1,000 megabits) per-


second offerings are available in several economies,
primarily in East Asia. The Hong Kong Broadband
Network was one of the first in the world to launch
such a service in 2010.48 Sony recently launched a
2  Gbps service to several parts of Japan.49 In the
United States, Google offers gigabit service in several
cities, hoping to encourage traditional broadband
operators to follow.50 Some question whether such
speeds are actually needed today, as few cloud
services exploit 1 Gbps.51 Further, even though these
speeds are theoretically available to the end user,
bottlenecks emerge due to the inability of backbone
networks to handle such large amounts of data.
However, as applications have a history of evolving to
exploit available speeds, fibre offers the most future-
proof fixed broadband technology.


Figure II.7. Global fixed broadband subscriptions and per 100 people by region, 2007–2012


Figure II.8. Fixed broadband market share by technology and theoretical maximum download speeds


Source: ITU.


Source: Point Topic and Broadband Strategies Toolkit, available at http://broadbandtoolkit.org/5.7 (accessed 3 October 2013).
Note: In the left-hand chart, market share information refers to the second quarter of 2012. In the right-hand chart, the cable modem


refers to the EuroDOCSIS 3 standard and asymmetric digital subcriber line (ADSL) refers to VDSL2, ITU-T G.993.2 recommendation.
Speeds are symmetrical.


0


100


200


300


400


500


600


700


2007 2008 2009 2010 2011 2012


M
ill


io
ns


Fixed broadband subscriptions


Least developed countries
Economies in transition


Developing economies
Developed economies


DSL
58%





Cable modem
19%





FTTx
19%


Other
4%


Fixed broadband market share by technology


0


5


10


15


20


25


30


2007 2008 2009 2010 2011 2012


Fixed broadband subscriptions per 100 people


Developed economies
Developing economies


Economies in transition
Least developed countries World


100


200


1 000


0 200 400 600 800 1 000 1 200


ADSL


Cable
modem


FTTH


Speed (Mbps)




29CHAPTER II : TRENDS IN THE CLOUD ECONOMY AND RELATED INFRASTRUCTURE


Asia leads in fibre subscriptions, accounting for
three quarters of the global total.52 All of the six
highest-ranked fibre economies (figure  II.9) are in
Asia. Most economies with a high proportion of
fibre have high incomes and are densely populated,
making them conducive to investments in fibre
to the premises.53 Elsewhere, fibre penetration is
generally low. According to one source, less than
three dozen economies had a household fibre
penetration of more than 1  per cent in 2012.54 In
developing countries with limited purchasing power,
private operators have little incentive to invest in
fibre investment beyond central business district
areas. In such cases, public involvement is typically
required to achieve more widespread fibre coverage.
Azerbaijan offers an example of public investment in
fibre optic networks. The Government of Azerbaijan
recently launched a three-year fibre optic project that
will increase end-user data transmission speeds to
100 Mbps in the capital Baku, to 30 Mbps in other
cities and to 10 Mbps in towns and villages.55


In contrast to fixed broadband, mobile broadband
has been growing rapidly in developing countries as
well. The number of global subscriptions stood at an


Box II.2. Cloud services over satellite


In many parts of the developing world that lack adequate terrestrial infrastructure or mobile broadband coverage, the only
way to access the Internet or other communication networks is through satellite. Running cloud services over satellite
communication, however, is difficult for two main reasons.


The first is latency. Geostationary orbit satellites are 35,786 km above the earth, and communication requires data to
cover that distance twice. Therefore, there is always a latency between 500 ms and 800 ms – far above the benchmarks
noted for optimal user experience of even basic cloud services. The second reason is that satellite communication is
expensive. Depending on the technology used, the market price for 1 megabit per second (Mbps) and about 1 megahertz
is about $3,500 plus service and one-time fees for hardware and installation. Some providers offer services at lower cost
by overbooking the available capacity, but with a negative impact on end-user experience as a result.


Recent innovations may be making satellite a somewhat more attractive option. The latency problem can be partly
addressed by reducing the amount of communications data that has to be transmitted between the remote site and the
central data centre. This can be done using a combination of caching and intelligent data management. With regard to
the cost of using satellites, the price can be reduced by sharing the bandwidth between sites and customers. A positive
side effect of a reduction in data traffic that is not related to communication (handshake and communications traffic) is that
less bandwidth is consumed for the transmission. This can lower the price for bandwidth by up to 70 per cent.


In South Africa, the data centre operator Business Connexion in 2011 implemented a solution to provide customers with
cloud services such as central data storage and centralized software management over satellite. One of its customers
is using this service to enable several subsidiaries in Zambia, Zimbabwe and the Democratic Republic of the Congo to
connect with headquarters in Johannesburg, and to use applications such as customer relationship management (CRM).


In the medium term, it is likely that in many parts of developing countries it will still not be financially viable to develop
terrestrial infrastructure. Making satellite communication solutions more affordable represents the main option for remote
areas to obtain access to cloud services. The potential may be further enhanced by of mobile telephony operators using
the global system for mobile communication connecting rural areas over satellite to their network.


Source: UNCTAD, based on information provided by the company named “meanswhat”.


62


59


55


39


34


34


25


21


18


18


17


16


15


15


14


12


11


10


0 10 20 30 40 50 60 70


Republic of Korea
United Arab Emirates


Hong Kong, China
Japan
Qatar


Taiwan Province of China
Singapore


Sweden
Iceland
Norway


Bulgaria
Denmark


Estonia
Slovakia


Russian Federation
Slovenia


Lithuania
Latvia


Figure II.9. Economies with more than 10 per cent
household penetration of fibre to the
home/building plus local area networks,
2012 (Percentage)


Source: ictDATA.org adapted from national regulatory authorities,
the European Union and Fibre to the Home Council Europe.


Note: Data for the United Arab Emirates, Japan, Bulgaria,
Estonia and Latvia refer to 2011.




30 INFORMATION ECONOMY REPORT 2013


estimated 1.5 billion in 2012, up almost 350 million or
30 per cent from the previous year.56 The world average
was estimated at 21 subscriptions per 100 people in
2012, with a large gap between developed countries at
67 and developing at 14 (figure II.10).57 With just some
15  million active mobile broadband subscriptions in
LDCs, penetration was below 2 per 100  inhabitants
in 2012.


As with fixed broadband, a variety of speeds is available,
depending on the technology used for the mobile
broadband network (table II.6). Mobile broadband
is optimally used as a single-user technology with
drops in performance if the connection is shared. It
generally has higher latency than fixed broadband.
Furthermore, coverage is still far from ubiquitous
in developing nations. Long-term evolution (LTE),
fourth-generation mobile technology, offers far higher
speeds than third generation (3G) networks – with a
theoretical 100 Mbps for mobile use and 1 Gbps for


stationary use – and less latency. Tests carried out
in Finland confirm this observation and suggest that
LTE performance is comparable to some fixed line
offerings (table II.7).


Figure II.10. Global active mobile broadband
subscriptions per 100 people, 2010–2012


0


10


20


30


40


50


60


70


80


2010 2011 2012


Pe
rc


en
ta


ge


Developed economies
Developing economies


Economies in transition
Least developed countries World


Source: ITU, World Telecommunication/ICT Indicators Database.


Technology Download speed Upload speed


W-CDMA 384 kbps 384 kbps


HSDPA 14.4 Mbps 384 kbps


HSUPA Specification for upload and not download 5.7 Mbps


HSPA 42 Mbps 11 Mbps


Table II.6. Theoretical highest mobile broadband
speeds, W-CDMA family


Source: World Bank, Broadband Strategies Toolkit.
Abbreviations: HSPA – high-speed packet access, HSDPA


– high-speed downlink packet access, HSUPA – high-
speed uplink packet access, kbps – kilobit per second,
W-CDMA – wideband code division multiple access.


3G LTE


Download
speed 4.1 Mbps 36.1 Mbps


Latency 117ms 23ms


Table II.7. Comparison of 3G and LTE speeds and
latency in Finland


Source: Epitiro, 2011, LTE ‘Real World’ Performance Study, http://
www.epitiro.com/assets/files/20-103-1006.001%20LTE-
Report-World-Report.pdf (accessed 3 October 2010).


5. Quality of service of broadband
networks


As connectivity in the cloud is achieved primarily via
the Internet, reliable broadband access is imperative.
The cloud is capable of an array of tasks, ranging
from simple webmail and word processing to
complicated reporting, visualization, data processing
and management information systems – all of which
are expected to have a quick turn-around time. As the
need for reliable and fast Internet access becomes
evident, so does need for higher QoS experienced by
the user.


Several dimensions can be considered when
assessing broadband connectivity QoS, (Gonsalves
and Bharadwaj, 2009):
• Download speed is the most often cited


characteristic when describing the quality of
broadband services. It refers to the time taken to
transfer data packets from a server to an end-user
device. Usually measured in kilobits per second
(kbps) or Mbps.


• Upload speed refers to the time taken to transmit
data packets from an end user device to a server.
Usually measured in kbps or Mbps.


• Latency or round trip time (RTT) is the time taken
for a packet to reach the destination server and
return to the client (the end-user device). Usually
measured in ms.


• Jitter is the variation of latency or the variation in time
of data packets arriving. Usually measured in ms.


• Packet loss refers to the share of packets that fail to
arrive at the destination server. Usually measured
as a percentage of the total number of packets
transferred.




31CHAPTER II : TRENDS IN THE CLOUD ECONOMY AND RELATED INFRASTRUCTURE


Various applications demand different performance
measures in order to function within acceptable
standards (table II.8). Upload and download speeds
and latency affect various levels of cloud sophistication
differently (figure  II.11). For example, throughput is
vital for streaming media, while RTT or latency is more
important for networked games.58 As a result of the
nature and design of cloud services, upload speed
and latency are gaining prominence for the use of
broadband.


The Cisco Global Cloud Readiness tool proposes
a set of thresholds with regard to QoS requirements
that should be met for cloud services of different levels
of sophistication (figure II.11). Low speeds and high
latency do not preclude the use of cloud computing but
rather the use of different cloud-based applications. For
example, QoS requirements are considerably lower for
basic cloud services such as webmail, web browsing
and VoIP than for more advanced applications such
as high-definition (HD) video streaming. A key usability
factor is performance, or how immediate the interaction
with the application seems.


(a) Download speed


Download speed has traditionally been a key metric
of broadband performance. Prior to Web 2.0, few
users generated content and most were passive
consumers of downloaded information. Though social
networking, blogs, wikis and video sharing have
grown, users still download more than they upload.
User-generated content generates more information


Service


Down-
load


speed
(kbps)


Upload
speed
(kbps)


Latency
(ms)


Jitter
(ms)


Packet
loss
(per-


centage)
Browsing ++ – ++ – –
Web-based
e-mail +++ ++ ++ – +


Streaming
media (con-
sumption)


+++ – ++ ++ ++


Streaming
media (pro-
duction)


++ +++ ++ ++ ++


Data storage – +++ – – –
VoIP + + +++ +++ +++
Gaming ++ + +++ ++ ++
Data storage
and analysis
(real time)


+++ +++ ++ – +


Web services + ++ +++ ++ +++


Management
information
systems
(ERP/CRM)


++ ++ +++ + ++


Software
development
services


– ++ ++ – +++


Table II.8. Relevance of broadband characteristics by
cloud service


Source: UNCTAD, based on LIRNEasia research; Gonsalves
and Badarwaj, 2009.


Key: +++ Highly relevant; ++ Very relevant;
+ Somewhat relevant; – Irrelevant.


Basic
Download: 750 kbps


Upload: 250 kbps
Latency: 160 ms


Single player gaming


Text communications
(E-mail, instant messaging)


Stream basic video/music


Web conferencing


Web browsing


VoIP (Internet telephony)


Intermediate
Download: 751–2 500 kbps


Upload: 251–1 000 kbps
Latency: 159–100 ms


ERP/CRM


HD video streaming


Multi-player gaming


Online shopping


Social networking
(multimedia/interactivity)


Video conferencing


Advanced
Download: >2 500 kbps


Upload: >1 500 kbps
Latency: <100 ms


3D video streaming


HD video conferencing


Stream super HD video


Connected education/
medicine


Group video calling


Virtual office


Figure II.11. Levels of cloud sophistication and related quality of service requirements


Source: Adapted from http://www.cisco.com/en/US/netsol/index.html (accessed 11 October 2013).
Note: Concurrent and multiples instances of applications will require a faster network.
ERP: enterprise resource planning.




32 INFORMATION ECONOMY REPORT 2013


to be downloaded. The growth of broadband has
also triggered more feature-rich content, especially
video, with higher quality resulting in many more bits
coming down to users. This is amplified by cloud
services offering feature films in HD format. It has been
estimated that some 1.3 billion Internet users around
the world watched around 162 online videos a month
in 2012.59 This has increased the ratio of downloaded
to uploaded information. A 2012 study of 5,000 users
in Japan found that, on average, each one downloaded
20 times more the volume of information than they
uploaded, a ratio that has doubled since 2005.60


(b) Upload speed


Increased use of offline storage, file sharing and
back-up services require more consideration to
upload speeds. Meanwhile, network operators
generally promote broadband plans primarily based
on download speeds with little or no mention of the
upload features. As noted previously, most fixed and
mobile broadband technologies other than fibre are
designed so that downlinks have greater bandwidth
and are faster than upstream links.61 The concept of
storage in the cloud requires users to upload data
to a remotely located server. Therefore, upload
speeds and data caps are critical elements when
assessing the ability of a network to support cloud
applications. Lack of bandwidth on the uplink can be
more of an issue for home or small enterprise users
than for large enterprises if the latter have access to
dedicated links.


(c) Latency


The performance of applications and services is
a function of the speed of which the data can be
accessed and processed. For instance, data on a local
hard disk can usually be accessed much faster than
data located at the end of a large network. The cloud
consists of multiple servers physically located around
the world. The collection of these servers creates a
virtual data centre. Within such a virtual environment
there can be varying degrees of latency. This is in
addition to the latency that exists between the end
user and the server that has been queried. Therefore,
unlike with traditional Internet services, different types
of latencies adversely affect a user’s overall experience.
This aspect is sometimes overlooked as the focus
tends to be on the performance and reliability of the
cloud as opposed to the connectivity to the cloud, a
necessity in order for it to offer a viable solution.


Less RTT means that a web page or document loads
faster. Latency can be a key bottleneck for the adoption
of cloud services. While a number of countries may
achieve acceptable levels of download and upload
speed, many do not meet the latency requirement for
intermediate and advanced cloud services. In terms
of fixed networks, only two economies (Hong Kong,
China and Lithuania) can reportedly fully support
concurrent advanced cloud services in an efficient
way.62 In the case of mobile broadband, no country
can efficiently support intermediate and advanced
services, due to high latency. However, as noted in
table II.7, LTE is comparable to many fixed broadband
offerings. There can be significant differences in terms
of network latency when accessing a server within an
individual ISP domain and in the international domain,
respectively.


Comparing latency across countries is not
straightforward. Since bits cannot travel faster than
light, there will always be some element of latency.63
As these delays increase with distance, it would be
appropriate to compare countries in terms of the
latency citizens experience over the same distance.
Such data are not universally available for most
countries.64 At the same time, not all cloud computing
servers may be accessible from within some countries.
Users in these countries would then need to access
overseas sites.


(d) Country comparison of broadband quality
of service


With a view to presenting a crude assessment of
countries’ performance in this area, table II.9 uses
data concerning 138 economies to illustrate whether
they are able to meet the QoS benchmarks proposed
by Cisco (figure II.11) for making use of basic and
advanced cloud services. The table further shows
in which areas (download speed, upload speed or
latency) countries encounter bottlenecks.


Of the 43 economies that meet all the minimum
benchmarks for advanced cloud services, 7 are from
developing Asia and Oceania and 6 are transition
economies. By contrast, no African or Latin American
countries are represented in this group.


The largest group consists of the 61 economies that
meet at least all the proposed minimum requirements
for basic cloud services. It includes 9 developed
economies and 7 transition economies, in addition to
6 economies from Africa, 18 from Asia and Oceania
and 20 from Latin America and the Caribbean. The




33CHAPTER II : TRENDS IN THE CLOUD ECONOMY AND RELATED INFRASTRUCTURE


Economies that
meet minimum


requirements for
advanced cloud


services


Economies
that meet minimum


requirements for
basic cloud


services


Bottleneck Economies that
do not yet meet
requirements


for basic cloud
services


Bottleneck


Download
speed


Upload
speed Latency


Download
speed


Upload
speed Latency


Armenia Albania X X Afghanistan X X


Austria Argentina X Algeria X


Belgium Aruba X X Angola X


Bulgaria Australia X Antigua and Barbuda X


Canada Azerbaijan X X Bangladesh X


China Bahrain X X X Belize X X


Hong Kong Barbados X X
Bolivia
(Plurinational
State of)


X X


Taiwan Province Belarus X X Cayman Islands X


Czech Republic Bermuda X Côte d'Ivoire X


Denmark Bosnia and Herzegovina X Haiti X


Estonia Brazil X Iraq X


Finland Brunei Darussalam X X X Lebanon X X


France Cambodia X Maldives X


Georgia Chile X Mauritius X


Germany Colombia X X Mozambique X


Hungary Costa Rica X X X Myanmar X


Iceland Croatia X Namibia X


Japan Cyprus X New Caledonia X


Republic of Korea Dominican Republic X X X Nigeria X


Latvia Ecuador X X Paraguay X


Lithuania Egypt X X X Peru X


Luxembourg El Salvador X X X Samoa X


Malaysia Ghana X X Senegal X


Moldova Greece X Sudan X


Mongolia Guatemala X X X Suriname X X


Netherlands Honduras X X Syrian Arab Republic X X


Norway India X X X Tajikistan X


Poland Indonesia X X X Turkmenistan X X X


Portugal Iran (Islamic Republic of) X X X Uganda X


Romania Ireland X United Republic of Tanzania X


Russian Federation Israel X Uzbekistan X


Singapore Italy X Yemen X


Slovakia Jamaica X X Zambia X


Slovenia Jordan X X X Zimbabwe X


Spain Kazakhstan X


Sweden Kenya X


Switzerland Kuwait X X X


Table II.9. Quality of service performance, by groups of economies




34 INFORMATION ECONOMY REPORT 2013


Economies that
meet minimum


requirements for
advanced cloud


services


Economies
that meet minimum


requirements for
basic cloud


services


Bottleneck Economies that
do not yet meet
requirements


for basic cloud
services


Bottleneck


Download
speed


Upload
speed Latency


Download
speed


Upload
speed Latency


The former
Yugoslav Republic
of Macedonia


Malta X


Ukraine Mexico X


United Arab Emirates Montenegro X


United Kingdom Morocco X X X


United States Nepal X X X


Viet Nam New Zealand X


Nicaragua X X


Oman X


Pakistan X X X


Panama X X X


Philippines X X X


Puerto Rico X


Qatar X


Saudi Arabia X


Serbia X


Solomon Islands X


South Africa X X


Sri Lanka X X X


Thailand X


Trinidad and Tobago X X


Tunisia X X X


Turkey X


Uruguay X X


Venezuela (Bolivarian
Republic of) X X X


Source: Adapted from Cisco Analysis, 2012. See annex table 5.
Note: See figure II.11 for QoS requirements for advanced and basic cloud services.


Table II.9. Quality of service performance, by groups of economies (continued)


main bottleneck impeding these economies from
graduating to the next level was the benchmark for
upload speeds – which more than 90 per cent of the
economies failed to meet – followed by latency.


The final group comprises 34 economies that failed to
meet at least one of the QoS benchmarks proposed
for basic cloud services. In this case, Africa dominates
with 13 entries, followed by Asia and Oceania (10),
Latin America and the Caribbean (8) and transition
economies (3). Latency is the most critical bottleneck
for this group, faced by all but three of the economies
listed.


6. Affordability of broadband services


The price of Internet access in some countries
presents a key barrier to cloud adoption, potentially
offsetting the cost savings from remote storage and
applications. Fixed broadband prices have declined
sharply in the past few years but are still high in many
developing countries. While a monthly fixed broadband
basket was, on average, equivalent to 40 per cent of
per capita income in developing countries in 2011,
it was less than 2  per cent in developed countries
(figure II.12). In the case of mobile broadband, prices




35CHAPTER II : TRENDS IN THE CLOUD ECONOMY AND RELATED INFRASTRUCTURE


are lower – the world average post-paid dongle mobile
broadband basket was 14  per cent of per capita
income in 2011. Nevertheless, the cost of mobile
broadband also remains high for developing countries
(20 per cent of income for a post-paid package and
31  per cent for pre-paid) (figure  II.13). Furthermore,
mobile broadband plans tend to have lower usage
ceilings than fixed broadband.


In Africa in particular, most broadband users rely
on either ADSL or mobile broadband for last-mile
connectivity. In addition to raising concerns of QoS
and limited deployment, ADSL access remains
relatively expensive. Recent studies confirm that at
low usage, mobile broadband tends to be cheaper


than ADSL (figure II.14) (Stork et al., 2013). However,
at higher usage, ADSL becomes a more competitive
option. Fixed broadband tends to be packaged
with higher volume limits than mobile, and this
advantage increases with greater usage. This implies
that in regions where mobile vastly outnumbers
fixed subscriptions, cloud access will tend to be
constrained by the usage limits in mobile broadband
pricing structures. The data for Africa also show
sharp price differences between countries that shift
over the usage profile. While ADSL tends to have
much bigger price divergence between countries at
low usage, mobile broadband prices diverge more
sharply at higher usage levels.


Figure II.13. Mobile broadband prices as a percentage of gross national income per capita, by region and level of
development, 2011


Figure II.12. Fixed broadband price basket (left) and annual change (right), 2008–2011


Source: ITU.
Note: Based on simple averages. For definition of regions, see http://www.itu.int/ITU-D/ict/definitions/regions/ (accessed 3 October 2013).


Source: ITU.
Note: Based on simple averages. For definition of regions, see http://www.itu.int/ITU-D/ict/definitions/regions/ (accessed 3 October 2013).


13.7


1.2


19.5


1.0


4.6


4.9


12.1


17.5


53.9


20.4


1.6


30.7


1.6


4.9


10.9


17.6


53.1


63.6


0 10 20 30 40 50 60 70


World


Developed economies


Developing economies


Europe


Arab States


The Americas


Asia and the Pacific


Commonwealth of
Independent States


Africa


Post-paid computer-based prices
Pre-paid handset-based prices


164.7


85.1


59.4
40.3


115.1


59.7


41.8 28.5
2.5



2.0



1.7


1.7 0
20
40
60
80


100
120
140
160
180


2008 2009 2010 2011




Developing


World


Developed


Fixed broadband basket as percentage
of gross national income per capita


1.4


-17.0


-19.1


-31.8


-30.0


- 48.1


-32.2


-30.2


-48.3


-60.0 -50.0 -40.0 -30.0 -20.0 -10.0 0.0 10.0


2010/11


2009/10


2008/09


Fixed broadband basket annual change


Developing


World


Developed




36 INFORMATION ECONOMY REPORT 2013


C. CONCLUSIONS
At the global level, revenues from formal cloud service
segments are small compared with the ancillary revenues
from the cloud economy (equipment sales, broadband
access and advertising-supported free cloud services).
Various private sources give different estimates of the
size of public and private cloud revenues, but all seem
to agree that the cloud phenomenon is set to expand
quickly in the next few years.


Metrics such as the number of data servers,
revenue and estimated market shares indicate that
practically all the major corporate players in the
cloud economy hail from the United States. This
is not a surprise, given the country’s dominance
of many aspects of the Internet ecosystem and its
position as an early adopter of cloud computing.
Their early entry into cloud computing has given
them first-mover advantages, not least in terms of
building large networks of users and massive data
storage and processing capacity. The absolute
levels of investment required for major cloud
computing estates are very high; it can cost more
than half a billion dollars for a cluster of data centres.
Investment requirements on this scale constitute
significant barriers to entry, which helps explain why
the global market is dominated by a small number
of providers and may raise concerns about the
extent of competition in the market.


In addition, the global cloud providers are also buying
up new cloud businesses, often at high valuations,
before they reach the point of being able to challenge


the dominant providers. Some cloud customers are
concerned about the risks of limited competition in
the cloud provision market (Renda, 2012). If cloud
providers can lock customers into contracts because
of the way in which they store their users’ data and
the services they provide them with, then customers
may not benefit as much as they should from the cost
savings that the cloud makes possible. It may also
raise questions about the scope for local cloud service
providers to emerge and grow, including in developing
countries (see chapter III).


At the same time, as cloud computing expands
geographically, it is possible that regional and local
companies will assume a greater role due to the need
for localization, system integration and aggregation,
combined with the need to host cloud servers close
by for better performance. Telecommunication service
providers are increasingly entering the cloud market
through acquisitions or partnerships and in the future
can be expected to play an active role due to their
direct link to the end user.


The second part of this chapter focused on the
infrastructure-related determinants of the ability of
countries to use and benefit from cloud computing.
Compared with other types of Internet use, some
factors are more relevant when considering the user
experience of cloud applications. Upon analysis, five
main observations can be made:
• Firstly, QoS is key. In particular, both upload speed


and latency become critical to ensure optimal use
of cloud applications. As some degree of latency


Figure II.14. Pre- and post-paid ADSL and mobile broadband prices per month, 2013 (United States dollars)


Source: Stork et al., 2013.
Note: ADSL lines were used for the fixed broadband baskets, developed in accordance with ITU broadband definitions: (a) monthly


cost of 1 gigabyte (GB) use per month with at least 256 kbps connection for a period of 24 months, (b) monthly cost of 5 GB use
per month with at least 256 kbps connection for a period of 24 months and (c) monthly cost of uncapped use per month with at
least 256 kbps connection for a period of 24 months. For comparative purposes, the mobile broadband baskets were designed
to match the fixed broadband baskets. The baskets were calculated for post- and pre-paid products.


0
10
20
30
40
50
60
70
80


1GB


ADSL


Post-paid


Pre-paid


Ca
m


er
oo


n


M
oz


am
bi


qu
e


Bo
ts


w
an


a


So
ut


h
A


fr
ic


a


N
am


ib
ia


Ke
ny


a


G
ha


na


Et
hi


op
ia


U
ni


te
d


Re
pu


bl
ic


of
T


an
za


ni
a


ADSL


Post-paid


Pre-paid


0


50


100


150


200


250


5GB


Ca
m


er
oo


n


M
oz


am
bi


qu
e


Bo
ts


w
an


a


So
ut


h
A


fr
ic


a


N
am


ib
ia


Ke
ny


a


G
ha


na


Et
hi


op
ia


U
ni


te
d


Re
pu


bl
ic


of
T


an
za


ni
a




37CHAPTER II : TRENDS IN THE CLOUD ECONOMY AND RELATED INFRASTRUCTURE


always exists, data centres should be located
within countries or fairly nearby to obtain optimal
performance using advanced cloud services.
This has implications for the configuration of the
underlying infrastructure. As documented above,
there is currently a significant digital divide in
terms of data centre and server availability across
countries. Moreover, few operators advertise their
upload speeds, making it difficult for users to
compare different offerings.


• Secondly, cloud services are not homogenous. It
is important to distinguish between different kinds
of applications when considering the ability of
countries to use and potentially benefit from the
cloud economy. Most Internet users make use of
some basic cloud services, such as webmail and
VoIP. These types of applications require far less
speed and can tolerate more latency than more
advanced cloud services relevant to the business
world.


• Thirdly, most mobile broadband networks are
capable of supporting only basic cloud services.
Although the speeds of mobile networks
continue to increase, they still fall short of those
achievable with fibre optic connectivity. Mobile
broadband typically involves higher latency than
fixed technologies. It is uncertain to what extent
mobile broadband below LTE standards can be
a substitute for fixed technology with respect to


efficient access to advanced cloud services. This
has implications for countries that are currently
relying primarily on wireless broadband, as in most
low-income countries.


• Fourthly, there is a need for better, more relevant
and international comparable data and definitions
for measuring the potential of cloud computing.
This includes official data on broadband speeds
and latency as well as primary source indicators on
cloud uptake such as the number of cloud servers,
data centres and cloud services available.


• Lastly, in the immediate future, many developing
countries, particularly in Africa, will face challenges
in order to fully benefit from the cloud, especially
with regard to intermediate and advanced cloud
services. Although most countries have now
deployed mobile broadband networks as a
substitute for fixed ones, they are characterized by
low speed and high latency and are therefore not
ideal for cloud provision. The cost of broadband
access in particular remains an obstacle for the
adoption of cloud services in many developing
countries. Long distances to data servers located
in foreign locations increase latency, and the
deployment of national data centres is affected by
the lack of supporting infrastructure (IXPs, reliable
and inexpensive electricity, robust fibre optic
backbones) and the lack of adequate regulations
(see chapter IV).




38 INFORMATION ECONOMY REPORT 2013


NOTES
1 That amount was forecast to would rise to $131 billion in 2013. See press release, “Gartner says worldwide public


cloud services market to total $131 billion”, 28 February 2013. Available at http://www.gartner.com/newsroom/
id/2352816 (accessed 4 October 2013).


2 Gartner also includes a value of BPaaS, which is attributed with generated revenues of about $31 billion in 2012.


3 This figure was derived based on the forecast of $24 billion in 2016, growing at a 50 per cent annual rate from
2012. See “IDC: Private cloud-based services to see rapid growth”, The Wall Street Journal, 28 February 2013.
Available at http://online.wsj.com/article/BT-CO-20130228-708787.html (accessed 4 October 2013).


4 See for example Kushida et al., 2012, p. 82 .


5 It should be noted that IDATE’s statistics may include an element of double counting in overlapping subsectors.
Furthermore, the data are based on consumption. For certain categories, disparities with production data can
be significant in cases of a high volumes of international trade (IDATE Foundation, 2012).


6 See http://www.iacpsocialmedia.org/Resources/FunFacts.aspx (accessed 4 October 2013).


7 One zettabyte is equal to 1,000 exabytes.


8 According to Cisco, most Internet traffic has originated or terminated in a data centre since 2008. Factors
contributing to intra-data centre traffic include functional separation of application servers, storage and databases,
which generates replication, backup and read/write traffic traversing the data centre. The ratio of traffic exiting
the data centre to traffic remaining within the data centre could be expected to increase over time, mainly due to
increased use of are bandwidth-heavy video files. Meanwhile, the increasing use of applications such as desktop
virtualization may offset this trend (Cisco Analysis, 2012).


9 See GigaOM, “IBM says its cloud biz is growing like crazy but provides no real numbers”, 23 January 2013.
Available at http://gigaom.com/2013/01/23/ibm-says-ibms-cloud-biz-is-growing-like-wild-fire-but-provides-no-
real-numbers/ (accessed 4 October 2013).


10 For example, one analyst firm compares 10 public cloud storage providers by various capabilities, such as
accessibility, manageability, pricing, security and resiliency. See http://www.gartner.com/technology/reprints.do?id=1-
1D9C6ZM&ct=121216&st=sg (accessed 4 October 2013). Another ranks the top 10 cloud providers by various features:
see http://ichitect.com/best-cloud/(accessed 4 October 2013). There is also a list of the top 100 cloud services providers
but no details are given on the methodology used: see http://talkincloud.com/tc100 (accessed 4 October 2013).


11 It should be noted that all servers of a particular company are not necessarily used for cloud service provision.


12 See “Verizon buys Terremark”, The Wall Street Journal, 27 January 2011. Available at http://online.wsj.com/article/
SB10001424052748703399204576108641018258046.html (accessed 4 October 2013).


13 One exception is NTT’s (Japan) acquisition of Data Dimension (South Africa); see http://www.telegraph.co.uk/
finance/newsbysector/mediatechnologyandtelecoms/telecoms/7893094/NTT-in-2.1bn-deal-to-buy-Dimension-
Data.html (accessed 4 October 2013).


14 See http://www.novell.com/news/press/2010/5/vodacom-business-and-novell-partner-to-securely-manage-and-
optimize-cloud-services.html (accessed 4 October 2013).


15 See http://www.nec.co.jp/press/en/1002/1803.html (accessed 4 October 2013).


16 See https://www.aplicateca.es (accessed 4 October 2013).


17 See http://www.digitalnewsasia.com/mobile-telco/special-report-telco-cloud-strategies-in-asean?page=0%2C1.


18 See http://instacompute.com (accessed 4 October 2013).


19 See http://itknowledgeexchange.techtarget.com/cloud-computing/tata-approaches-us-cloud-market-with-caution/
(accessed 4 October 2013).




39CHAPTER II : TRENDS IN THE CLOUD ECONOMY AND RELATED INFRASTRUCTURE


20 See “Microsoft launches office, Azure services in China”, Information Week, 5 November 2012. Available at http://
www.informationweek.com/cloud-computing/infrastructure/microsoft-launches-office-azure-services/240044374
(accessed 4 October 2013).


21 For further information, see explanatory notes available at http://unstats.un.org/unsd/cr/registry/docs/CPCv2_
explanatory_notes.pdf (accessed 4 October 2013).


22 The services sectors included were computer and data processing services and royalties and license fees for
general use of computer software, amounting to a total value of $43.8 billion cross-border transactions in 2010.


23 Examples of such industries include telecommunication businesses, retail trade businesses and computer and
electronic product manufacturers.


24 See “Made outside India”, The Economist, 10 August 2013. Available at http://www.economist.com/news/
international/21583285-growth-slows-and-reforms-falter-economic-activity-shifting-out-india-made-outside
(accessed 4 October 2013).


25 Communication from the Computer and Communications Industry Association to the United States International
Trade Commission regarding Digital Trade Hearing of 7 March (Investigation No. 332-531). Available at http://www.
ccianet.com/libraryfiles/ccLibraryFiles/Filename/000000000764/CCIA%20Digital%20Trade%20Follow-up%20
Letter.pdf (accessed 10 October 2013).


26 As reported in the UNCOMTRADE database for HS category 851712.


27 See, for example, Gartner press release, “Gartner says smartphone sales grew 46.5 percent in second quarter
of 2013 and exceeded feature phone sales for first time”, 14 August 2013. Available at http://www.gartner.com/
newsroom/id/2573415 (accessed 10 October 2013).


28 Information provided by e-mail by TechAmerica.


29 Information provided by e-mail by the World Customs Organization.


30 The HS 847150 category is described as covering the traded value of “processing units other than those of
sub-heading 8471.41/8471.49, whether/not containing in the same housing one/two of the following types
of unit: storage units, input units, output units”.


31 Hong Kong, China appears as a top importer and exporter in table II.4. This is partly due to high activities of
re-exports through that economy to and from China.


32 Although not featured in table II.4, Taiwan Province of China also witnessed rapid growth of computer server
production and exports.


33 See “Worldwide server market rebounds sharply in fourth quarter as demand for x86 servers and high-end systems
leads the way, according to IDC”, IDC, press release, 27 February 2013. Available at http://www.idc.com/getdoc.
jsp?containerId=prUS23974913 (accessed 10 October 2013).


34 See “IDC: Q4 server revenue declines again, but shows improvement”, CRN, 25 February 2010. Available at http://
www.crn.com/news/components-peripherals/223100777/idc-q4-server-revenue-declines-again-but-shows-
improvement.htm (accessed 10 October 2013).


35 See “Worldwide server market accelerates sharply in fourth quarter as demand for heterogeneous platforms
leads the way, according to IDC”, IDC, press release, 28 February 2010. Available at: http://www.idc.com/about/
viewpressrelease.jsp?containerId=prUS22716111 (accessed 10 October 2013).


36 Frameworks for analysing cloud computing in developing countries have been proposed, but performance is
defined in a broad sense, such as availability of cloud services, cloud awareness and diffusion of cloud for which no
comparable statistics exist (Kshetri, 2010).


37 See http://www.telegeography.com/products/commsupdate/articles/2013/04/17/international-bandwidth-demand-
is-decentralising/ (accessed 11 October 2013).




40 INFORMATION ECONOMY REPORT 2013


38 Caching refers to the process of storing previously requested Internet information (such as a web page) on a server
near the user so that it can be re-used. This minimizes data traffic back to the source of the information, thereby
reducing bandwidth requirements.


39 Indicators are beginning to emerge with regard to length, coverage and capacity of terrestrial networks.
According to one study on terrestrial fibre backbone networks in sub-Saharan Africa, there was twice as
much fibre under deployment across Africa in 2011, compared with 2009; 313 million people (36 per cent)
were within reach of an operational fibre node; and cross-border networks more than doubled to 20 Gbps
by the end 2010 (Hamilton, 2011).


40 Some studies suggest that a significant share of these IXPs are non-operational (Ryan and Gerson, 2012).


41 See http://www.internetsociety.org/news/new-study-reveals-how-internet-exchange-points-ixps-spur-internet-
growth-emerging-markets (accessed 14 October 2013).


42 See https://www.peeringdb.com/help/stats.php (accessed 14 October 2013).


43 The recommended temperature and relative humidity range for data centres is 18°C–27°C and between 40 per cent
relative humidity (5.5°C dew point) and 60 per cent (15°C). See http://www.cisco.com/en/US/solutions/collateral/
ns340/ns517/ns224/ns944/white_paper_c11-680202.pdf (accessed 14 October 2013).


44 A data centre risk index has been designed to assist companies in making strategic investment and operational
decisions about where to locate their data, whether it be server rack deployments or the creation of new facilities,
see (Source8 et al., 2013).


45 See http://www.datacentermap.com/datacenters.html (accessed 14 October 2013).


46 A study of mobile Internet in Europe found that the number of reported mobile broadband subscriptions was almost
twice as high as the number of people who reported in surveys that they use the Internet from mobile devices over
mobile networks; see http://www.ictdata.org/2013/01/mobile-internet-in-europe.html (accessed 14 October 2013).


47 See http://www.ftthcouncil.org/p/cm/ld/fid=50 (accessed 14 October 2103).


48 See “HKBN launches 1 Gbps broadband for US$26/month”, press release, 14 April 2010. Available at
http://uk.reuters.com/article/2010/04/14/idUS20842+14-Apr-2010+GNW20100414 (accessed 14 October 2013).


49 See “Sony ISP launches world’s fastest home Internet, 2Gbps”, Computer World, 15 April 2013. Available at
http://www.computerworld.com/s/article/9238392/Sony_ISP_launches_world_39_s_fastest_home_
Internet_2Gbps (accessed 14 October 2013).


50 See “5 reasons you want Google Fibre in your city”, CNN, 12 April 2013. Available at http://edition.cnn.
com/2013/04/11/tech/innovation/google-fiber-austin-cities/index.html (accessed 14 October 2013).


51 See “Fibre to the home: A highway too super”, The Economist, 6 December 2012. Available at
http://www.economist.com/blogs/babbage/2012/12/fibre-home (accessed 14 October 2013).


52 See http://www.ftthcouncilap.org/index.php?option=com_content&view=article&catid=6:media-releases&id=221:apac-
represents-75-of-ftthb-subscribers-worldwide&Itemid=36 (accessed 14 October 2013).


53 It is cheaper to lay fibre in areas where many people live in close quarters, such as apartment buildings, and where
incomes can afford the multiple services (e.g. broadband and video) fibre providers market to recoup investment.


54 See http://www.ftthcouncil.eu/documents/Presentations/20130220PressConfLondon_Online.pdf (accessed
14 October 2013).


55 See http://regionplus.az/en/articles/view/1983 (accessed 14 October 2013).


56 From ITU World Telecommunication/ICT Indicators Database.


57 As noted earlier, the data should be interpreted carefully for possible discrepancies between the reported and actual
levels of mobile broadband usage.




41CHAPTER II : TRENDS IN THE CLOUD ECONOMY AND RELATED INFRASTRUCTURE


58 One study found that a one-second delay reduces customer conversation by 7 per cent, customer satisfaction by
16 per cent and page views by 11 per cent (Aberdeen Group, 2008).


59 See http://www.comscore.com/Insights/Presentations_and_Whitepapers/2013/The_Past_Present_and_Future_
of_Online_Video (accessed 13 October 2013).


60 See http://www.iij.ad.jp/en/company/development/iir/pdf/iir_vol16_report_EN.pdf (accessed 13 October 2013).


61 Empirical assessments confirm weaker performance with significantly slower upload than to download speeds
(see LIRNEasia, 2011).


62 See http://www.cisco.com/en/US/netsol/ns1208/networking_solutions_sub_sub_solution.html (accessed 14 October
2013).


63 It takes around 20 ms for a packet travelling at the speed of light to cover around 3,000 km. Current fibre optic
technology adds a delay, increasing latency. Recent experimental research suggests that new fibre optic design
could come close to reaching the speed of light (Poletti et al., 2013).


64 Statistics that do not use a universal latency distance thus produce misleading results. See “Use of latency in
broadband ranking is silly”, Digital Society, 2 October 2009. Available at http://www.digitalsociety.org/2009/10/
use-of-latency-in-broadband-ranking-is-silly/ (accessed 14 October 2013).


65 See “The billion dollar data centers”, Data Center Knowledge, 29 April 2013. Available at http://www.
datacenterknowledge.com/archives/2013/04/29/the-billion-dollar-data-centers/ (accessed 14 October 2013).


66 See “The cloud builds up steam”, Financial Times, 6 June 2013. Available at http://www.ft.com/intl/cms/s/0/
e2b826a2-ce20-11e2-8313-00144feab7de.html#axzz2hiA0qwBY (accessed 14 October 2013) and “Cisco
acquires enterprise Wi-FI startup Meraki for $1.2 billion in cash”, Techcrunch.com, 18 November 2012.
Available at http://techcrunch.com/2012/11/18/cisco-acquires-enterprise-wi-fi-startup-meraki-for-1-2-billion-
in-cash/ (accessed 14 October 2013).






IMPLICATIONS
FOR DEVELOPING
COUNTRIES 3


The level of cloud adoption in most developing countries is still low.
Nevertheless, this is an appropriate moment to consider what opportunities
and challenges the evolving cloud economy may bring to developing
countries. The picture differs between countries as well as between types of
cloud service customer. The ability to seize opportunities presented by cloud
computing and to avoid the pitfalls associated with it will depend substantially
on the level of cloud readiness, particularly in terms of broadband connectivity
(chapter II) and appropriate legal and regulatory frameworks (chapter IV).


Using the cloud economy ecosystem presented in chapter I, this chapter
reviews critical drivers and barriers for cloud adoption from a developing
country perspective. It draws on the still limited available evidence, which
mainly consists of various surveys of prospective or existing users of cloud
services. It describes opportunities for developing-country enterprises on
both the supply and the demand sides of the cloud economy. The chapter
recognizes that the extent to which countries, and stakeholders within
countries, will be able to take advantage of the cloud economy can be
influenced by proactive policies, which are considered in chapter V.




44 INFORMATION ECONOMY REPORT 2013


A. DRIVERS AND
BARRIERS OF CLOUD
ADOPTION


1. Drivers of cloud adoption


The main advantages attributed to cloud computing
can be summarized in three main categories (see also
chapter I):
• Cost savings in hardware, software and personnel,


derived from the economies of scale that are
available from the cloud;


• Flexible access to processing and storage capacity
on demand, with a high degree of elasticity;


• Improved system management, reliability and IT
security.


Cost savings are most frequently cited in the literature
among these drivers. These are expected to result
from cloud service customers paying for the use of
data storage capacity and application software, rather
than having to purchase otherwise the necessary
hardware and software. This allows them to gain
from the economies of scope and scale that are
available through cloud service providers, and from
their specialized management expertise. This model
of provisioning effectively transfers resources from
capital to operational expenditure, making it possible
for companies with limited financial resources – a
situation that is common in developing countries – to
make use of applications that they could not otherwise
afford. Savings at the cloud service customer level
can also arise from reductions in IT staff costs,
though companies and other organizations that make
extensive use of cloud facilities will need to retain
skilled personnel to manage continuing in-house
IT functions, procure cloud services, oversee their
relationship with cloud providers and manage cloud
data and applications internally.


Flexible access to processing and storage capacity
is another important driver, particularly for companies
whose business activities are seasonal, whose
computing requirements vary substantially at different
stages of a business cycle, or which operate in
sectors where demand is unpredictable. Start-up
ventures often have difficulty scaling up resources if
they are met with unexpected levels of demand. Cloud
provisioning can enable them to respond without
having to seek capital, at short notice, to invest in
new equipment. Likewise, companies experiencing a
business downturn benefit from the ability to reduce


software and hardware requirements without having
to write off expensive IT assets. In addition, cloud
provisioning enables users to apply the latest software
versions as and when these become available, at
limited cost, mostly associated with less need to buy
and install software upgrades or replace the hardware
required to run them.


Smaller businesses in developing countries often find
it difficult to hire competent personnel in IT and other
professional disciplines (see, for example, UNCTAD,
2011a). Cloud provisioning enables them to outsource
some of the IT skills that they would otherwise have to
resource internally, benefiting from the expertise that
cloud service providers can offer in areas such as IT
management and security. Taken to the national level,
in countries with a shortage of IT skills, this could allow
for a more efficient use of scarce IT expertise. From a
security standpoint, the relative value of this aspect
may be higher for SMEs in developing countries that
may be less likely to be in a position to take measures
against cyberthreats than larger enterprises or SMEs
in more advanced economies (Kshetri, 2010).


Drivers such as those mentioned above apply in
principle to both large and small firms, but play out
differently. Larger companies, with more complex
needs, are more likely to search out cloud solutions
that are tailored to their specific needs. Smaller firms,
with less capacity to assess different IT options, may
rather rely primarily on generic, public applications.
The size of an enterprise may also influence its access
to the Internet, affecting the degree of sophistication
of the cloud services that it can use.


Some evidence concerning the drivers of cloud
adoption in non-profit sectors, where decision-
makers have different priorities, can be set alongside
this analysis. Although a 2011 survey of over
400 government executives in 10 developed and
developing countries confirmed that cost savings
were the most significant expected benefit from cloud
adoption, the next most significant expected benefits
were concerned with the nature of government
activity (figure III.1) (KPMG, 2012). About 39 per cent
of the respondents stated that they expected the use
of cloud applications to change how Government
interacted with citizens and 37  per cent that cloud
would increase government transparency.


In a global survey of non-governmental organizations
(NGOs), responses indicated that administrative
advantages (such as easier software access, rapid
deployment and reduced system administration)




45CHAPTER III : IMPLICATIONS FOR DEVELOPING COUNTRIES


were considered more important than cost reductions
(Tech Soup Global, 2012).1 Nonetheless, more than
60 per cent of the respondents reported that they still
considered cost reductions important. These surveys are
useful reminders that different cloud users have different
objectives that are related to the nature of their activities.


2. Barriers to cloud adoption


Barriers to cloud adoption in developing countries fall
into two main categories: those that are internal to the
potential cloud customer’s business/organization, and
those external barriers that relate to the wider economic
and communications environment. Internal barriers
to adoption include attitudes, concerns and anxieties
among managers about data security in the cloud,
the location of data and reliability of service. Internal
barriers also relate to the customer organization’s ability
to adapt its own systems to use the cloud or to benefit
from cloud provisioning. External barriers result from the
business, legislative and communications environment
in which a firm or organization operates.


(a) Internal barriers


Many of the concerns expressed by potential cloud
service customers stem from reservations about the
reliability and quality of contractual terms offered by


cloud service providers (see also chapter  IV). Six
anxieties in particular emerge from survey evidence
and discussions with cloud users in developing
countries, as follows:
• Concernsrelatedtothesecurityandprivacy


of data. In a 2010 survey of Indian enterprises,
72 per cent of respondents said that privacy and
data security issues were extremely significant
concerns for them (Ernst and Young, 2010).
Security concerns cited by both Government and
business customers of cloud services relate, among
other things, to the confidentiality of company and
customer data, identity management and the risk
of identity theft, the risk of data being compromised
or altered intentionally or inadvertently in the
cloud, and to the fate of data at the end of time-
limited contracts.2 It has been argued that the
security arrangements of cloud service providers
are often more sophisticated and comprehensive
than those that potential customers can deploy
themselves (see, for example, Capgemini, 2012).
Nevertheless, potential cloud service customers
are likely to be reluctant to leave security in
the hands of a third party, particularly if security
arrangements are unclear to them. Such concerns
are compounded by the possibility that third parties
(including foreign Governments) could gain access
to sensitive national, business or personal data.


Figure III.1. Government expectations of cloud impact, 2011 (Percentage share of replies)


Note: Survey was conducted from February to May 2011 and canvassed 429 government executives in Australia, Canada,
Denmark, Italy, the Netherlands, Singapore, South Africa, Spain, the United Kingdom and the United States.


Source: Adapted from (KPMG, 2012)


3


18


24


28


37


39


50


0 10 20 30 40 50 60


Other


No signicant impact


Accelerate time to market


Fundamentally change the business model


Provide management with greater
transparency on transactions


Change the interaction with
constituents/citizens and suppliers


Reduce costs




46 INFORMATION ECONOMY REPORT 2013


Worryingly, data stored in the cloud can be a potential
goldmine for cybercriminals. In 2009, Google
reported on an attack on its infrastructure that it had
found was part of a larger operation that infiltrated
the infrastructures of at least 20 major companies
(Information Warfare Monitor and Shadowserver
Foundation, 2010). In January 2012, the
Government of Kenya suffered extensive hacking,
involving more than 100 official websites, as a result
of attacks by an individual hacker in Indonesia.3
Security breaches and other cybercrimes may also
go unreported because of fears that information
about them will negatively influence the valuation
or reputation of the companies involved (Kshetri,
2010). Online payment and trading systems will
become more attractive targets for cybercrime as
traffic volumes and revenues on them grow, and
businesses will need to pay increasing attention to
security protocols throughout their IT operations as
the threat and sophistication of attacks expand. The
need for coherent interfaces between local and cloud
security norms requires consistent application of
stronger security standards throughout customers’
own systems, which they may lack the expertise
to deploy. Even the best cloud security can be
compromised by inadequate security measures in
customers’ own (terrestrial) systems.


• Concerns over the geographical location of
data(includingbackups).The movement of data
into and out of a cloud service will often result in its
falling under the rules of different jurisdictions. The
cross-border transfer of data is generally opaque
to the user, raising issues of control. Subsequently,
questions may arise on who is responsible for the
data at any given point in its cross-border movement.
A cloud service customer may, for example, obtain
services from an inter-cloud service provider, which
in turn obtains additional cloud services from another
cloud provider. Some potential users are concerned
that data that are located in foreign jurisdictions could
be misused or exploited by third parties. In some
cases, corporate policies require data to be held
within national territorial jurisdictions, barring the use
of internationally provided cloud services and facilities.


In addition to legal and other concerns about
security, there are other factors that may influence
potential cloud customers in favour of more local
provisioning. Half of the informants in a survey of
3,000 SMEs saw value in obtaining the services
from a locally based provider, and for 31 per cent
this was seen as critical (Microsoft, 2012). This


preference for local expertise illustrates that factors
other than cost savings influence cloud adoption
decisions, potentially creating incentives for the
expansion of local provision of data centre and
other cloud services as the cloud economy evolves.


• Concernsrelatedtothereliabilityofservice.
Potential cloud customers are concerned that
services will be disrupted by system failures in the
cloud itself, in the communications networks that
connect them with the cloud, or through power
cuts that make it impossible for them to access
data and services when required. Such problems
can result in management inefficiencies and/or
poor service to their customers. In practice, there
is little evidence of system failures occurring in the
systems of major cloud service providers. The
greater risks of unreliability are likely to be external
to cloud contracts, in national communications
and power networks (see external barriers, below)
or in the international broadband access when
data are stored abroad. Such systemic failures are
much more likely to occur in countries with less
complex communications networks, lower levels
of redundancy and poorer power infrastructures.


• Concerns related to the non-availability
of suitable terminal devices. Although
computerization is becoming more entrenched in
developing-country business environments, SMEs
there tend to be less well equipped with terminal
devices than their peers in other countries, and
many smaller firms remain reliant on mobile
phones (UNCTAD, 2011b). Although some cloud
systems can work effectively with less specified
in-house computer systems (or with smartphones
and other recent mobile terminals), others require
access to devices with sufficient computing power
and capability to use them effectively.


• Concerns related to the migration of data and
upgradability. An important consideration for
users is to avoid becoming locked into contractual
agreements with a particular cloud provider, thus
making it difficult or too expensive to migrate to
an alternative provider. Another consideration is
that the services of various cloud providers will be
insufficiently interoperable for the user to build a
portfolio of services. Cloud proponents argue that
the development of competitive markets in cloud
provision should obviate these risks. At the same
time, as noted in chapter II, economies of scale
and first-mover advantages may lead to further
concentration of the industry. The development




47CHAPTER III : IMPLICATIONS FOR DEVELOPING COUNTRIES


Box III.1. Barriers to cloud adoption in Ghana


Some of the perceived challenges to cloud computing were identified in a 2013 survey of 72 public sector managers,
administrators and IT managers in Ghana (box figure III.1).


The cost of migration was the principal challenge. Respondents were concerned not only about charges made by cloud
service providers, but also about surrendering the ownership and management of data to a third party. They referred to
the cost of being locked-in to a particular cloud service vendor. Various costs may be incurred if the organization wanted
to change a cloud service in terms of data reconfiguration to meet the requirements of the new operator or purchase new
software. The risk of such future costs may deter the organization from changing its cloud service provider.


Box figure III.1. Perceived challenges of migrating to cloud service


Source: Tweneboah-Koduah, 2013.


Data security and privacy was the second most important challenge identified in the survey. Institution/enterprise readiness
to migrate to cloud services was in third place.


Service availability and Internet connectivity were mentioned by more than half of the respondents. The main concern
here was related to the need to find appropriate last mile connections (ADSL and mobile broadband) to access services.
Dependence on ADSL raised concerns over quality, limited deployment and cost.


Source: UNCTAD, based on Tweneboah-Koduah, 2013.


0 10 20 30 40 50 60 70 80 90


Security and
privacy


Migration cost


Internet connectivity


Service
availability


Standardization


Institution/enterprise
cloud readiness


of common cloud standards should facilitate
interoperability, although that in itself is insufficient
to ensure the operation of a competitive market.
The need for users to secure interoperability
when considering different solutions and different
providers creates opportunities for cloud brokers
and aggregators (see section III.C).


• Institutional collaboration and coordination.
Effective use of cloud computing requires a
high level of organizational collaboration and
coordination internally (between government
departments or between business divisions).
Interdepartmental coordination remains a challenge
for many developing country Governments, for
both institutional and technical reasons. There are


often demarcation disputes between ministries and
other government agencies, which may or may
not be resolved by the establishment of special
bodies concerned with ICT sector development.
At a technical level, in order to leverage cloud
services more effectively, there may be a need to
strengthen various government data networks and
make them more interoperable.


The implementation of new IT systems, such as
those enabled by the cloud, requires substantial
organizational change – the re-engineering of
management structures and business models – if the
potential benefits of the cloud are to be fully realized.For
a firm of any size, this requires significant investment in
business restructuring and staff training, as well as the




48 INFORMATION ECONOMY REPORT 2013


writing off of existing hardware and software assets.
However, failure to implement necessary changes –
for example, maintaining legacy systems which should
preferably be replaced – can be even more costly
as problems of integration are then likely to persist.
In some developing countries, such as Ghana, the
evidence suggests the cost of migration may be the
most important barrier to cloud adoption, followed by
security and privacy concerns (box III.1).


Developing-country businesses furthermore have to
address challenges in implementing cloud services,
as revealed in a global survey of managers in large
corporations undertaken for Oracle in 2013 (Dynamic
Markets Ltd., 2013).4 This indicated that more than
half had experienced problems with integration of
cloud services, leading to staff downtime and non-
use of cloud applications by parts of their businesses,
and to the abandonment of some cloud applications.
Poor integration, for example, with other software, had
hindered innovation more in the Asia-Pacific region
than elsewhere, with figures as high as 70  per cent
and above among respondents in India and Singapore.
Asian respondents also reported higher than average
problems with customization of cloud services to
company needs. Over 40  per cent had experienced
usability challenges and/or security breaches with
cloud applications, with a third of the respondents
noting the need to call in support from their companies’
IT departments in order to resolve these problems.
Other problems raised by managers included poor
implementation of cloud services on mobile devices.
The experience reported by these managers illustrates
problems at various points within the cloud economy
ecosystem, in the relationship between cloud providers
and corporate cloud service customers as well as
within the latter.


The previously cited survey of NGO perceptions of
the cloud found that lack of knowledge and skills to
manage cloud resources was the most significant
concern, followed by issues of cost (for example,
for migration to the cloud, set-up and recurrent
expenditure on both cloud providers’ charges and
communications) (Tech Soup Global, 2012). The
latter is likely to be particularly relevant in countries
with high costs for broadband connectivity. NGOs in
countries with relatively low gross domestic product
(GDP) per capita were more likely to identify barriers to
adoption, citing foreign currency problems alongside
lack of reliable Internet and electrical connectivity and
government regulations concerning the location of
data storage. Some expressed concerns about data


security, while others rather expected that the cloud
would help to improve data security.


Lack of awareness of what cloud computing actually
involves and its implications is also holding back cloud
adoption. Nevertheless, organizations in developing
countries with relatively advanced communications
networks (for example, Egypt, Mexico and South
Africa) anticipate faster deployment of the cloud within
their IT systems in the next two years.5


In view of the above, a number of critical decisions face
potential users of cloud services. Potential cloud service
customers need to make judgments about whether
migration of data and services to the cloud will lead,
for them, to better or worse outcomes in efficiency
and service delivery; about how long it will take for
improvements to materialize; and about the degree
of risk there is that factors related to data security,
privacy, reliability and internal management and costs
of the migration will negate the advantages which
may otherwise arise. As well as inhibiting migration to
the cloud, such reservations will likely affect choices
about the kind of cloud provisioning that cloud service
customers may adopt. Concerns about control over
data, for example, have so far led to a preference by both
Governments and businesses for private rather than
public cloud solutions, in spite of the greater cost savings
that could be afforded by the latter. Concerns about
interoperability may also make them more cautious in
their choice of cloud providers. Against this background,
contractual issues will remain of central importance to
the decision-making of potential cloud users (chapter IV).


(b) External barriers


External barriers fall into three main categories, as
follows:


(i) Inadequate infrastructure


As stressed in chapter II, several aspects of
communications infrastructure affect the extent to
which a company’s migration to the cloud will yield
the benefits that are generally attributed to cloud
computing. Infrastructure-related deficiencies are
especially critical external barriers to cloud uptake in
rural areas of low-income countries. Even in Kenya,
for example, where there has been a relatively good
progress in ICT development, the lack of reliable power
and broadband connectivity remains a major concern.
Optic fibre cable vandalism occasionally results in fibre
cuts that can run for a week at a time (Research ICT
Africa, 2013). The lack of reliable power is a particular
constraint on the building of local data centres.




49CHAPTER III : IMPLICATIONS FOR DEVELOPING COUNTRIES


(ii) Legal and regulatory barriers


The second set of external barriers inhibiting adoption
and provision of cloud services in developing countries
concerns the legal and regulatory frameworks. For some
sectors, cloud-related transfers and storage outside the
jurisdiction of the regulated entity may imply a breach
of national rules by failing to provide national authorities
with “effective access” to the data (chapter IV).


Many cloud services depend on the ability of users
to make secure electronic transactions. This is only
possible where legislation and financial regulation
give electronic transactions equivalent status to
their physical counterparts, and where banks and
other commercial interests are willing to treat them
as equivalent. Many developing countries still lack
adequate legal and regulatory frameworks for
electronic commerce (chapter IV). Relevant legislation
needs to cover issues such as the recognition of
digital signatures and electronic transactions, data
protection, evidential standards and transborder data
flows. Although model legislation has been available in
these areas for some years, it has been adopted only
in a limited number of developing countries.6


Similar barriers result from weak or uncertain
legislation and regulation concerning cybersecurity.
The term cybersecurity is wide-ranging, including the
prevention of disruption to networks and services
from accidental or criminal activity, the prevention of
fraud, and efforts to prevent spam and other malware
from corrupting online services. As with legislation for
e-commerce, many countries still need to introduce
and implement necessary legislation, together with the
institutions required to support cybersecurity regimes
(see also chapter IV). In 2012, for example, a regional
assessment by the ITU for the African, Caribbean and
Pacific Group of States (ACP) indicated that only five of
fifteen ACP countries in the Caribbean had legislation
governing e-transactions (HIPCAR, 2012).


The regulatory framework for cybersecurity can also
influence the choice which cloud service customers
make between different types of cloud configurations.
For example, the cyber-control measures put in place
by the Government of China have contributed to
greater reliance on nationally based cloud services
because of higher latency experienced when using
servers located outside the country. According to one
study, the Chinese firewall – the Golden Shield Project,
which has been used since November 2003 – has led
to an increase in the loading time by 450 ms or more
for an object hosted on a server outside of China.7


(iii) Weaknesses in the wider business
environment


The third set of external barriers concerns the wider
business environment and the extent to which a
‘digital culture’ has developed. This is partly a matter of
business performance, including performance in trade
in goods and services, and partly of the availability of
underlying skills and resources within society.


The shortage of IT skills is a significant concern
especially in low-income countries, where IT education
(at school and university levels) is underdeveloped and
where those with IT skills have opportunities to earn
more and build more attractive careers if they take
their skills elsewhere. The migration of skills outside
the country is a particularly serious problem for smaller
countries, such as island states in the Caribbean and
the Pacific. Even with a shift to the cloud, expertise
is still required at the cloud service customer level in
both use of the technology and its legal and regulatory
requirements. IT skills are even more important where
countries and businesses are seeking to establish new
business opportunities through the cloud, whether
managing data centres, acting as cloud aggregators or
establishing new cloud services targeted at export or
domestic markets. The lack of training infrastructure,
and of skilled personnel benefiting from it, can be a
significant barrier to cloud adoption and successful
leveraging of cloud computing in developing countries.


The drivers and barriers to cloud adoption described
above are summarized diagrammatically in figure III.2.
While the drivers illustrated in the figure are broadly
similar for businesses and other potential users in all
countries, the barriers differ significantly depending
on a country’s level of development and business and
communications environments.


3. Assessing drivers and barriers
to cloud adoption in developing
countries


Potential cloud service customers, be they in business
or Government, should weigh potential gains (such
as lower operational and transaction costs) against
additional costs which they may incur by using the
cloud (notably for communications and migration).
It is possible to overestimate the internal savings
that may result from migration to the cloud as well
as underestimate the costs of the migration itself.
Like other new technologies, cloud computing can
prove significantly more successful if accompanied




50 INFORMATION ECONOMY REPORT 2013


by organizational change that enables a business
or organization to make proper use of it, a point
emphasized above. Making the necessary changes
may take time and require new investment.


Factors of particular significance in developing
countries that may lead to different decisions from
those that would be made by decision-makers in a
more developed economy include the following:
• The lesser prevalence of ICT use in Government


and business in developing countries;
• Cloud costs: the cost of fees paid to cloud


service providers and for communications
access and usage, ISP charges and the
hardware and software costs incurred when
accessing the cloud are likely to be significantly
higher in developing-country environments, and
so form a higher proportion of the total costs of
cloud provisioning;


• The presence or absence of Internet exchange
points (IXPs) and local data centres that can


reduce costs and address challenges caused by
international connectivity;


• The risks and potential costs associated with
unreliable communications networks and high
latency. Cloud-based systems may experience
more downtime in developing countries and are
likely to suffer slower connectivity;


• Inadequacies in the legal and regulatory
framework to address concerns related to data
protection and privacy;


• The costs of automating processes that can be
done efficiently by human staff. Labour costs
are likely to be significantly lower in developing
countries, making it more cost-effective to
use personnel rather than to automate some
functions;


• The availability of content and applications that
are specifically addressed to the requirements
of developing country or local enterprises and
organizations;


Broad network
access


IaaS


Infrastructure
as a service


PaaS


Platform as a
service


SaaS


Software as a
service


SERVICE CATEGORIES DEPLOYMENT MODELS


Rapid elasticity


Measured
service


On-demand
self-service


Resource pooling


Multi-tenancy


CORE CHARACTERISTICS


Public cloud


Community cloud


Private cloud


DRIVERS


Economies
of scale and
cost savings:


• Hardware
• Software
• IT staff


Privacy
and security


Reliability
of service


Migration and risk
of lock-in


Location of data,
including back-ups


BARRIERS – INTERNAL Availability of
terminal devices


Costs of transition


DRIVERS


Rapid
deployment


Flexible
access


Reliability and
security


Accessibility
and mobility


BARRIERS – EXTERNAL


Infrastructure &
affordability


Legislative
framework


Institutional
weaknesses


Availability
of ICT skills


Lack of
awareness


Poor business
integration


Figure III.2. Drivers and barriers to cloud adoption


Source: UNCTAD.




51CHAPTER III : IMPLICATIONS FOR DEVELOPING COUNTRIES


• The ability to source equipment and software in a
timely manner for service provision, whether using
conventional computing or the cloud. Established
sources of conventional supply may be preferable
if potential gains from cloud provisioning are
relatively small;


• The extent to which potential users are likely
to take up cloud services that may be offered
to them. This is likely to be lower in developing
countries because communications networks are
less widespread and less affordable.


The outcome of calculations of the expected net
benefit will vary between countries according to their
level of cloud readiness. It will also vary between
organizations, according to the contexts, data and
services involved. And it will evolve over time as the
level of cloud readiness in national communications
environments improves.


It should be recognized, too, that if firms and
Governments adopt cloud provisioning they may not
always do so systematically. While some large firms
and some Governments are developing overarching
cloud strategies, in others, decision-making is devolved
to individual divisions or departments. Some of these
may adopt cloud services while others eschew them,
and different departments may contract with different
providers. SMEs, conscious of the risk to their
businesses of getting things wrong, may move more
cautiously towards the cloud, adding cloud applications
one by one, rather than adopting a comprehensive
approach. In some cases, cloud services may already
be used extensively, without any management decision
in favour of such use having been taken. Staff in
government and the private sector often prefer to
make use of cloud services (such as webmail) for work
purposes rather than using in-house email and other
facilities if they are deemed inconvenient or less reliable.


An organization-wide strategic approach aimed at
integrating changes in management and operational
systems with a coordinated move towards the cloud is
more likely than ad hoc approaches to leverage cloud
benefits successfully. An organization-wide approach
can enable more efficient procurement of cloud
services, whether from a single cloud provider or
through an aggregator, avoiding potential problems of
incompatibility between cloud services and enabling
systematic integration of legacy and cloud services
throughout the business. As indicated by the high
incidence of integration problems (Dynamic Markets
Ltd., 2013), companies at all levels and Governments


would benefit from more coherent consideration of
cloud options by senior management.


It is by no means certain that cloud adoption and
the cloud economy will evolve in similar ways in
developing countries as in developed countries. As
suggested above, for example, some potential cloud
service customers value a local presence of the cloud
provider. Different situations in terms of the affordability
and QoS of broadband networks may tilt the balance
in favour of more national or regional rather than global
cloud provision. Meanwhile, as in developed countries,
some cloud service customers will prefer to forgo
some potential cost savings and opt for private rather
than public cloud facilities, in the belief that these can
provide better security, greater control of data, lower
risk of becoming dependent on a cloud provider and
more certainty about communications costs.


B. CLOUD ADOPTION
IN SELECTED
COUNTRIES:
EXPERIENCES AND
OPPORTUNITIES


The cloud economy ecosystem described in chapter
I and the discussion of drivers and barriers outlined
above provide a framework for considering levels of
adoption and potential opportunities for the cloud
economy. The evidence base concerning cloud
adoption in developing countries is very limited.
Substantial differences are nevertheless evident
in the extent to which the cloud economy has
emerged in countries with diverse characteristics.
Much of the variation that can be identified can be
attributed to differences in the national business and
communications environments (chapter II).


This section draws on available evidence of cloud
development in selected countries to explore opportunities
on both the supply and the demand sides of the cloud
economy ecosystem. The discussion leverages the
experience of countries with a high level of readiness as
well as those with nascent cloud economies.


1. Supply-side cloud opportunities in
developing countries


The most significant activities and potential
opportunities for enterprises on the supply side of the
cloud economy in developing countries are concerned




52 INFORMATION ECONOMY REPORT 2013


with the following: (a) data centre and related cloud
provision; (b) the development and provision of local
cloud services for groups of customers, including
local businesses and individual citizens; (c) cloud
aggregation, system integration, brokerage and
related services. In addition to these explicitly cloud-
based areas of activity, opportunities exist for national
communications businesses (telecommunications
operators and ISPs) which can gain from increased
data traffic using their networks.


(a) Provision of data-centre services


Major cloud service providers own and manage the
computing infrastructure (the interconnected data
centres and related facilities that store data) and enable
services on behalf of cloud service customers. As
stressed earlier (chapter I), the cloud business model is
highly dependent on economies of scale and the ability
to manage networks of data centres or server farms with
huge computational and storage capacity. Very high
levels of capital investment are needed to challenge their
predominance and few potential competitors may have
the resources to do so. However, the dominance of the
incumbent providers may be somewhat challenged by
growing concerns among cloud customers about data
privacy and security.8


Only a few larger and more technologically advanced
emerging economies such as the BRICS countries
(Brazil, the Russian Federation, India, China and South
Africa), Malaysia, Singapore and Thailand meet the
scale and connectivity requirements for global data
centre operations.9 However, there is potential for a
wider range of developing countries and developing
country businesses to establish data centre operations
which could attract customers from a regional level.
This potential is likely to be greatest in countries
with large and dynamic ICT sectors, with strongly
established IT businesses and with an established role
as regional business or communications hubs.


Despite the advantages of leading global cloud
service providers, there are various other factors that
may offer scope for local or regional data centres to
expand in developing countries, such as the following:


• Demand forprivatecloudsolutions. The fact
that large corporations as well as Governments
have hitherto preferred private over public clouds
suggests that they have preferred to eschew
significant economies of scale to ensure a greater
sense of security and control over their data and
services;


• Non-cost factors which may require local
presence (such as national data protection
laws). Some Governments and businesses are
required (by law or corporate policy) to locate their
data within national jurisdictions, or prefer to do so
for security or geopolitical reasons. Such policies
may increase demand for local cloud service
provision;


• High degree of alternative added value. More
effective customer service or familiarity which may
be enhanced by a local or regional presence are
examples;


• High costs of or unreliable international
broadband connectivity, making it more
advantageous to rely on local data centres.
While these hamper and reduce potential benefits
from cloud adoption in general, inadequate
infrastructure may also undermine the cost
advantage of global cloud provision in developing
countries.


Reducing the physical distance between cloud
computing resources and end users would reduce
the costs of broadband communication and reduce
latency. As noted in a recent report on cloud computing
in Africa (ITU-D, 2012):


Despite the development of international data
transmission links between Africa and the rest of
the world, the costs associated with the bandwidth
necessary for transferring “African data” to and from
cloud computing resources located outside Africa…
are so high that it is more advantageous to construct
data storage centres in Africa than to pay for accessing
centres located tens of thousands of kilometres away
from the continent.


An increasing number of Governments are establishing
their own data centres to manage government data
and services. These can form the basis for what are in
effect private clouds (box III.2). In other cases, new data
centres are being established by the communications
providers. In Nigeria, for example, the state-owned
provider of connectivity, ICT services and applications,
Galaxy Backbone, is in the process of implementing a
national cloud infrastructure that will deliver IaaS, PaaS
and SaaS (Research ICT Africa, 2013). Meanwhile, the
Nigerian telecommunications operator Globacom has
set up a large Internet Data Centre in Lagos to provide
co-location, disaster recovery and dedicated hosting
services to its clients all over the country.10 In Kenya,
Safaricom began to host its MPESA mobile money
service at a local data centre in 2011. This was in




53CHAPTER III : IMPLICATIONS FOR DEVELOPING COUNTRIES


response to downtimes caused by problems related
to international broadband connectivity. However, it
has been difficult to convince other local businesses
to adopt cloud services in Kenya due to concerns
over issues associated with connectivity and power
availability.11


An alternative opportunity for local cloud service
providers and data centres is to develop partnerships
with global providers with the objective of offering
customers more control over their data while still
leveraging economies of scale. A hybrid cloud
arrangement utilizing local data centres may allow
sensitive government or corporate data to be retained
in-country while less sensitive data can be distributed
across a global public cloud. There are both cost and
strategic advantages for global cloud providers in
locating data centres in different world regions, whether
operating these within their own estates or making use
of data centres established by locally based partners.


There has been significant investment by TNCs in data
centres in developing countries since the final years
of the last decade. IBM, for example, has opened a
number of cloud centres in China and India since 2008,
providing services to local businesses, universities
and software developers. Other international firms
that have established cloud operations in India
include the software provider, Parallels, the business
applications provider Salesforce and the virtualization
specialist VMware (Kshetri, 2010). At the same time,


local companies in developing countries have also
been developing their own cloud facilities, including
significant international players such as the Indian
telecommunications company, Airtel, and the South
African firms MTN and Dimension Data.


One factor to keep in mind for countries aiming to
encourage the creation of new large-scale data
centres is the need for low cost, sustainable power.
Data centres have enormous electricity demands,
much of it for cooling servers. Estimates of their
total power consumption are hard to clarify, though
it has been asserted that they consume between
1.1 and 1.5 per cent of total electricity consumption
worldwide, equivalent to the output of thirty nuclear
power stations.12 Countries with fragile power
supplies or regular shortages of power simply cannot
accommodate large data centres without placing
significant additional burdens on existing electricity
users.


(b) Cloud service provision


The second, and perhaps most significant, area of
supply-side potential within the cloud economy lies in
the development and marketing of new cloud services.
Because of the versatility of cloud provisioning, the
range of potential services is wide, as is the range
of potential target clients (chapter I). Countries with
established communities of IT professionals and
entrepreneurs will be best placed to develop more


Box III.2. Government data centres in developing countries


Different approaches have been used by Governments to set up national data centres, sometimes involving the public–
private partnership (PPP) sector. A few examples are provided below.


• The Government of Kenya has developed data centre capacity for its own use as well as for public access in
order to reduce costs for businesses and organizations that need to host data in-country. Such intervention is not
uncontroversial, since government-owned data centres may compete with commercial businesses operated in the
private sector (Research ICT Africa, 2013).


• The Government of Ghana is constructing three data centres that will host data from all government ministries,
departments and agencies (Research ICT Africa, 2013).


• In India’s roadmap to develop the “GI Cloud” (see also chapter V), the Government opens the possibility of inviting
private cloud service providers to set up dedicated government clouds or to manage them according to the policy,
standards and guidelines adopted for the government cloud (India, Department of Electronics and Information
Technology, 2013).


• Ecuador has included the implementation of a national data centre in its Digital Ecuador Strategy. The centre is to
be established and operated under a PPP arrangement led by the Ministry of Telecommunications and Information
Society.a


Source: UNCTAD.
a See eLac2015 newsletter, number 19, July 2012, available at http://www.cepal.org/socinfo/noticias/paginas/3/44983/


newsletter19ENG.pdf (accessed 7 October 2013).




54 INFORMATION ECONOMY REPORT 2013


innovative, sophisticated cloud applications and
services.


For low-income countries at a nascent stage of cloud
readiness, IaaS is likely to be the first category of cloud
services to be demanded. Businesses and Governments
in developing countries are recognizing this opportunity.
In Ghana, for example, a number of companies already
provide various types of cloud services (table III.1). By
contrast with the global pattern observed in chapter II,
most of the cloud services in Ghana are in the areas of
IaaS and PaaS, which together accounted for 69 per
cent of the market, exceeding SaaS deployments,
with IT training and consulting services making up
the rest (Yeboah-Boateng and Cudjoe-Seshie, 2013).
The majority of the companies listed in the table act
as local representatives of global cloud providers. In
Nigeria, similarly, most cloud adoption appears so far


Cloud service provider Cloud delivery model(s) Target market segment Comment


Locally owned providers


CIS Ghana PaaS, IaaS SMEs, ASPs


DreamOval SaaS SMEs, individuals


Gesatech Solutions IaaS SMEs Linked to Kaspersky Security software


Ghana Dot Com IaaS SMEs, individuals


Maafo-Visions IaaS SMEs, individuals Targets regional markets


National Information Tech-
nology Agency BPaaS, PaaS, IaaS Government Government agency


NetSolutions Ghana PaaS, IaaS SMEs


Radius Consulting PaaS, IaaS SMEs
Targeting West African markets. Part of
several vendor accreditation schemes (Cisco,
Dell, Microsoft, Symantec Gold, and the like)


Rancard Mobility PaaS
Mobile operators, content
providers, mobile marketers,
developers


Established in 2001, grew regional in 2010
(office in Nigeria)


Red Mango SaaS, IaaS SMEs Has a network of global partners including Cisco, HP, Microsoft


Sylversys Consulting SaaS, PaaS, IaaS SMEs Offices in Ghana and Mali and targets West African markets


Foreign affiliates


ACT ICT SaaS Government, SMEs Founded by ACT ICT (Israel) in partnership with Lifeforms (Ghana)


Huawei IaaS Enterprises, Government Huawei (China)


IIHT Ghana SaaS, PaaS SMEs, individuals
Owned by Indian Institute of Hardware
Technology, India’s leading IT services
training company


Internet Solutions IaaS SMEs, individuals


Pan-African division of Dimension Data
(South Africa), which is owned by NTT
Group (Japan). Partner of Cisco, Microsoft,
Symantec


MTN SaaS, PaaS, IaaS SMEs Affiliate of MTN Group (South Africa)


Table III.1. Selected cloud service providers in Ghana, 2013


Source: UNCTAD, adapted from Research ICT Africa (2013).


to be focusing on IaaS (Research ICT Africa, 2013).
The emphasis on storage and infrastructure services by
virtually all cloud providers in that country can be seen
as a natural evolution following on from the significant
investments made in recent years by various TNCs,
banks and other large enterprises in their own network
infrastructure, particularly data centres and hosting. As
of early 2013, however, there was very limited PaaS
and SaaS activity in the country.


As the market for IaaS matures, the scope for PaaS
and SaaS tends to expand. Some of the services
which are being made available in developing countries
replicate the success of global applications (such as
social networks) at a national or regional level, offering
a more targeted experience. Other cloud services
have been developed in response to more specific,
localized requirements (box III.3).




55CHAPTER III : IMPLICATIONS FOR DEVELOPING COUNTRIES


Box III.3. Examples of cloud service provision in developing countries


The following list includes examples of the provision of cloud services that either provide a local version of globally suc-
cessful applications or that are targeted to specific local needs:


• Sonico.com is a social networking website headquartered in Argentina, with over 48 million users, the majority in Latin
America. To compete with global giants such as Facebook and LinkedIn it needed to keep costs down. Sonico uses
Amazon Web Services applications for file storage and sharing of more than 1 billion images uploaded by its members.
The company says that this has led to a 70 per cent cost saving over its previous file management architecture.a


• A host of domestic cloud service providers has emerged in China, offering services for the local market which are sig-
nificantly protected from global competitors by language and regulatory constraints. Examples include e-commerce
platforms (such as Alibaba and Taobao) and indigenous web platforms for micro blogging and social networking (such
as Sina Weibo and Renren) and local search engines (such as Baidu) (UNCTAD, 2012a).


• Viet Nam Technology and Telecommunication offers server storage capacity and system capability for its clients,
mainly SMEs. It has the intention to launch applications that are designed specifically for construction and for real-
estate companies.b


• India-based AdventNet’s Zoho division operates a suite of web-based applications. As of September 2009, Zoho had
over 2 million users, the bulk of which were from North America and Europe and some 20 per cent from India and China.
Zoho’s applications are used by hospitals and banks in India to develop new products and services and by some insur-
ance companies to develop innovative services such as a personalized insurance for diabetics (Kshetri, 2010).


• Cloud services can be focused on individual or neighbouring countries rather than international markets. Ping.sg is a
blog aggregator and blogging community in Singapore, whose site brings together more than 100,000 blogs. As its
business grew, it chose to migrate its data and services to the cloud. It has derived cost savings and other benefits
from the ability gradually to increase computing capacity rather than having to procure new equipment as the busi-
ness expands.c


• A growing range of cloud services are targeted at enterprises. The South Africa-based media and telecommunications
company MTN launched a suite of cloud services for SMEs in Ghana and Nigeria in December 2012, following pilot proj-
ects in six countries. It offers packages to support small business accounting, human resource and customer-relations
management as well as webmail and videoconferencing, storage and backup. It also offers an SaaS application that
provides microfinance institutions with a platform on which to run their banking operations. According to MTN, this pack-
age has been adopted by SMEs in the manufacturing, hospitality, microfinance and advertising industries (Research ICT
Africa, 2013). Pamoja Cloud Services, which is owned by the South African-led submarine cable company SEACOM, is
also targeting growing demand for IT-as-a-Service from SMEs in Africa (Research ICT Africa, 2013).


Source: UNCTAD.
a See http://aws.amazon.com/solutions/case-studies/sonico/ (accessed 8 October 2013).
b See “VNTT rides on cloud to deliver new services”, search SMB Asia, 31 August 2009, available at http://www.


searchsmbasia.com/en/content/vntt-rides-cloud-deliver-new-services?page¼0%2C0 (accessed 8 October 2013).
c See http://aws.amazon.com/solutions/case-studies/ping-sg/ (accessed 8 October 2013).


In countries where national broadband coverage is
inadequate, but where mobile connectivity is widespread,
there may be opportunities to leverage a combination
of cloud and mobile services. The mothers-2-mothers
(m2m) organization, a South African NGO, combines
the cloud with database technology and mobile services
to reduce the incidence of HIV/AIDS transmission from
mothers to children.13 The m2m NGO digitizes patient
records and shares them with counsellors across its
networks of more than 700 sites in Africa. The records
contain information on treatment plans, and advanced
reporting tools that allow quick response. Women in


African villages authenticate children’s medication with
text messages via short message service (SMS).14 As of
2011, m2m served more than 1.5 million women in nine
sub-Saharan African countries. In a similar way, the core
customer interface of MPESA in Kenya relies on SMS,
but its underlying data and applications management
reside in the cloud.


Although PaaS applications of cloud services are less
widespread than IaaS and SaaS deployments, they
can offer supply-side opportunities for small scale IT
businesses in developing countries, which may be




56 INFORMATION ECONOMY REPORT 2013


expected to grow in coming years. The Information
Economy Report 2012 emphasized in that respect
the important role that local IT sectors, including local
software developers, have in building a culture of digital
innovation in developing countries. Innovation centres
like the iHub in Nairobi and Campus Tecnológico in
Guatemala (UNCTAD, 2012a) have brought together
vibrant groups of developers and entrepreneurs
producing software, developing applications and
supporting the adoption of ICTs by enterprises.


Most developing countries now have at least a
small number of individuals with the skills to develop
software applications, but their business development
is often hampered by limited access to development
tools, by the cost of necessary hardware, and by
small domestic markets for their services. A critical
advantage of cloud computing for innovative firms,
particularly such entrepreneurial start-ups, is that they
can access the ICT capabilities they need to innovate
without significant capital expenditure.


The cloud can support open source software
development, which is preferred by many application
developers. The software company, Canonical, for
example, has built cloud capabilities into its Ubuntu
Server platform,15 and offers a developer store platform
on top of its application store, Ubuntu Software
Center.16 The Ubuntu App Developer allows developers
anywhere in the world to create Linux applications for
the Ubuntu platform and make them available either
for a fee or free of cost. These applications can be


directly downloaded to the Ubuntu Linux Desktop
and Server platforms (UNCTAD, 2012a). There are
other platforms that allow such entrepreneurs to
develop new ideas, such as mobile applications, at
low cost, using the latest available tools, and then
offer them to global rather than national markets,
where they may generate enough revenue to become
sustainable and finance the next phase of enterprise
development. Global or national PaaS providers could
partner with these enterprises, sharing the risk, cost
and profit associated with the development of new
products. OrangeScape is one of several examples of
Indian companies that offer a PaaS application. This
enterprise claims to be “the world’s only cross-cloud
PaaS platform”.17


(c) Cloud aggregation, integration and related
services


A third area of opportunity for the supply of cloud
services in developing countries is in the provision
of aggregation, consultancy and brokerage services
that facilitate the relationship between cloud
providers and their government or corporate clients.
Cloud aggregators work on behalf of government
or corporate customers to assemble packages
of services from different cloud providers that suit
their clients’ overall requirements, rather like system
integrators in more traditional IT environments.
Aggregators assume the role of identifying the most
suitable cloud services, integrating these with one


Box III.4. Selected cloud aggregators in developing countries


• Aggregation is one of the services offered by the IT company Clogeny, based in Pune, India. It works with businesses
and organizations on a range of issues across the spectrum of cloud management, offering services in areas such as
cloud assessment and planning, application development and cloud testing to end-to-end cloud support. It acts as
an intermediary between clients and major cloud computing providers.a


• A number of indigenous and foreign-owned cloud aggregators and system integrators are active in the Nigerian market
(Research ICT Africa, 2013). Nigerian companies providing aggregation services include Computer Warehouse,
Resourcery, City Business Computers and Computer Information System Nigeria. The companies generally offer
aggregation alongside other IT services, helping their clients to integrate cloud provisioning more effectively within
their overall IT strategy.


• Smaller businesses are less likely to need the services of an aggregator, but may benefit from consultancy and
procurement services offered by a specialist broker with the expertise and connections to identify and negotiate
the most appropriate cloud deals. Descasio, for example, describes itself as “the leading Google Apps and cloud
computing solutions provider in Nigeria and the sub-region”, and as offering “the bridge between cloud-based
solutions and legacy on-premise solutions”.b


Source: UNCTAD.
a See http://www.clogeny.com/.
b See http://www.descasio.com/ (accessed 8 October 2013).




57CHAPTER III : IMPLICATIONS FOR DEVELOPING COUNTRIES


another, managing relationships with cloud providers,
and thereby offer clients a simpler interface for their IT
and cloud activities. Box III.4 provides some examples
of such businesses in developing countries.


The aggregator/broker market is a promising area for
developing-country IT firms to explore. While larger,
transnational firms may prefer to obtain aggregation
services at a multinational level, national businesses
are more likely to see value in working with local
partners. They can leverage their experience of their
national business and communications environments,
including the particularities of national legislation in
areas such as data protection, the preferences of
national client groups, and the availability of national
or regional cloud providers.


The discussion above illustrates opportunities on the
supply side of the cloud for potential developing-
country businesses. A number of factors will influence
the extent to which these are likely to emerge in the next
few years. Some developing countries are clearly better
placed to take advantage of supply-side opportunities
than others – particularly larger countries with more
developed economies, large domestic IT businesses
and more sophisticated communications networks.
Countries in which connectivity is reliable and cheap
will be at a considerable advantage over others. So
will countries that already have the skills to develop IT
services for domestic and export purposes. Political
and other factors also play a part. When looking for
potential providers, cloud service customers in both
government and the private sector are likely to give
significant weight to the quality of security protection
afforded by cloud providers and to the risk that data
will be accessed by third parties.


2. Cloud use by different stakeholders
in developing countries


This section reviews the available evidence concerning
cloud adoption in developing countries within four
groups of potential cloud service customers – citizens
and others using mass market cloud services;
Governments and government agencies; TNCs and
large national businesses; and SMEs.18


(a) Mass market use of free cloud services


There has been extensive adoption by individuals in
developing countries of free-to-use cloud services
such as webmail and online social networks. This
is true in almost all countries, though particularly


so in those with higher levels of Internet use and
cloud readiness. The most popular cloud-based
applications are generally those provided at a global
level. Facebook, for example, by the end of 2012
had some 835  million registered users worldwide,
more than 10  per cent of the global population.19 It
is estimated that more than 40  per cent of Internet
users on any given day access the Facebook site.20


The video file-sharing site YouTube, owned by
Google, is estimated to reach around 35 per cent of
global Internet users.21 Both appear among the top
four websites in all the 17 African countries for which
statistics are available.22 In the few national markets
where these global services are not available, such
as China, local equivalent cloud services are equally
predominant (see box III.3). Social media platforms are
increasingly accessed on mobile phones rather than
computers, and through apps rather than through the
conventional Internet.23


There is widespread anecdotal evidence that cloud-
based webmail services are extensively used by
Government and business professionals when
they consider the in-house email systems to be
less efficient or reliable (which is the case in many
developing countries). Data-storage and file-sharing
services that are available on the public cloud are also
experiencing increased use in both developed and
developing countries. Many users of these services
may be unaware that they are cloud-enabled, but their
widespread use illustrates the growing pervasiveness
of the cloud in core aspects of personal and
professional use of computing and the Internet.


Publicly-available mass market cloud services are
of value not just to individuals but throughout the
supply chain. As has already been noted, they are free
because data mining enables their providers to target
advertising much more precisely on likely customers
than conventional media, and so to attract advertising
revenue. They are also potentially useful, however, for
small business users. Social media platforms allow
SMEs to build customer relationships by sending
occasional text messages or emails, or by building
their profiles through Facebook (and similar) pages
rather than conventional websites. Social media,
in this sense, add a new dimension to the “word of
mouth” on which many small businesses have always
depended to extend their market reach. Following their
customers’ interests online can even offer SMEs an
opportunity to target their own services more precisely
on consumer trends. Cloud-based information
services also offer opportunities for marketing. Public




58 INFORMATION ECONOMY REPORT 2013


cloud services can therefore add value to SMEs,
particularly those that deal directly with individual
consumers, though care needs to be taken to protect
the privacy and security of both SMEs themselves and
of their customers.


(b) Government use of the cloud


In many developing countries, Governments are the
largest purchasers of IT, spending substantial amounts
annually on hardware, software and ancillary goods and
services, including in-house IT management. Much of
this procurement – for example, of software licences
– exceeds requirements as the majority of authorized
users rarely need more than basic applications. The
availability of in-house IT management s kills is also
often a challenge in Governments of developing
countries.


Government agencies are increasingly considering
cloud solutions to manage large data sets of the kind
on which modern administration and service delivery
depend, in some cases building their own estates
of data centres. Coordination among government
departments to maximize the value of cloud solutions
requires proper managerial structures as well as


national strategies. In the Republic of Korea, the
decision to build a government cloud has generated
significant benefits in terms of cost reductions,
improved security and greater client satisfaction (box
III.5). Several developing and transition economies are
moving towards greater government use of the cloud.
A few examples are given in the following paragraphs.
• Thailand’s Electronic Government Agency


launched its government cloud in 2012, bringing
together 200 different IT systems from more than
100 government agencies. A PPP approach
was adopted, with local software companies
developing SaaS products that enable legacy
government systems to be redeployed in the
cloud and reduce duplication between different
agencies.24 The Electronic Government Agency
has signed a memorandum of understanding
with the Cloud Security Alliance – a not-for-
profit organization which promotes the use of
best practices for providing security assurance
within cloud computing – and plans to make
Cloud Security Alliance certification frameworks
mandatory for cloud service providers.25


• The Government of Moldova launched its Moldova
Cloud, developed in partnership with the private


Box III.5. Government cloud use in the Republic of Korea


The Government of the Republic of Korea has been a leading exponent of network connectivity and data centre
management. It enjoys very high broadband connectivity, has among the highest average download and upload
speeds in the world, and enjoys low levels of latency (annex table 5). An important part of the Government’s overall
strategy to boost the cloud economy has been the holistic development of its own cloud use.a


When Korean agencies were responsible for building and operating their own e-government systems, a number of
problems frequently arose – includingoverlapping infrastructure investment, lack of relevant expertise, inadequate
provisioning of facilities, weak security systems and inferior computing environments. The National Computing and
Information Agency (NCIA) was established in 2005 under the Ministry of Security and Public Administration to serve
as a government integrated data centre and build the foundation for more reliable and sustainable e-government.


In order to provide government cloud services, it was decided that the co-location of ICT resources was needed,
followed by hardware integration through hardware pooling and the organization of software platforms (Kang et
al., 2011). As a first step, a standardized and automated government integrated data centre management system
– the National Total Operation Platform System – was developed. To reduce the risk of serious government-wide
impacts caused by natural disasters or terror attacks, a disaster recovery system was built, including two inter-
copied centres in Daejeon and Gwangju to ensure business continuity. K-net (Korea net), an intranet separated
from commercial networks, was also established to provide safe, secure and scalable communication services to
central and local government agencies and public organizations.


Government-wide integration of hardware resources began in 2008, involving the conversion of multiple servers
into one unified hardware platform by applying virtualization technologies. Since 2011, the Government of the
Republic of Korea has started the process of embedding service platforms in integrated hardware to enable cloud
computing. The cloud computing platform, G-Cloud, was deployed in 2011. In the same year, a comprehensive




59CHAPTER III : IMPLICATIONS FOR DEVELOPING COUNTRIES


Box III.5. Government cloud use in the Republic of Korea (continued)


defense system, the Electronic Advanced National Security Infrastructure, which implemented physical and
cyber security and access control, was also put in place.


NCIA today has the ICT resources and operational management needed for government agencies to conduct
their business electronically. It also provides a reliable e-government platform, which offers the general public
some 1,200 e-government services at all times of the day and night. By 2015, half of all the ICT resources of
central agencies and a smaller percentage of those of local agencies and their affiliates will have been moved
into cloud environments.


NCIA provides various forms of PaaS and a validated standard environment:


• TheG-cloudplatformserviceprovides government agencies with a server, storage and a network in a
virtualized environment on demand along with the eGovFrame software environment;b


• The G-mobile platform service offers common systems and test environments for the development
of mobile government services, including basic systems (link, security, authentication, mobile device
management, push) needed for developing mobile or web applications;


• TheSmartofficeplatformservice provides an environment for flexible working by mobile workers; it also
designs and provides a cloud PC infrastructure architecture.


Finally, NCIA provides a business continuity guarantee service, which offers ICT-based government services
without interruptions even in the event of a disaster.


More efficient management of national ICT resources has generated significant cost savings, estimated at
about $300 million per year. Government agencies have been able to reduce their budgets for the purchase
and operation of ICT resources by 30 per cent through better procurement, shared resources and integrated
development of common functions and services. The total investment involved was recouped seven years
after the establishment of NCIA. Other returns on investment continue to be generated in the areas of security,
operation, resource integration and maintenance. In addition, on the power usage effectiveness (PUE) scale,
energy efficiency improved from 2.0 in 2006 to 1.71 as of June 2013, that is, the second highest level.c


One of the reasons for entrusting NCIA with the task of operating cloud services in-house, rather than outsourcing
these to the private sector, was to ensure that data owned by the Government would be protected. NCIA
has strengthened its cyber response system to detect and block external attacks automatically.d In addition,
using dual-power supply equipment and generators, the e-government system equipment downtime has been
dramatically reduced; the monthly average downtime per item of equipment fell from 67 minutes in 2004 to only
3.04 seconds in June 2013.


Public satisfaction with e-government services has increased. Government agencies are now able to provide
such services anytime and anywhere. By ensuring the quality of services through the application of service level
agreements between each agency and NCIA, the level of satisfaction among government agencies has also
increased. Each agency is now able to improve its business competencies and productivity by using NCIA’s
various ICT resources as well as managerial and technical support.


Source: Ministry of Security and Public Administration, Republic of Korea.
a In 2009, the “Government-wide Cloud Computing Promotion Master Plan” was announced which was followed


in 2011 by the “Strategy for Promoting the Expansion and Competitiveness of Cloud Computing”.
b eGovFrame (e-Government Standard Framework) is a standardized set of software tools for developing and


running e-government applications. It was developed by the Government of the Republic of Korea.
c The PUE indicates the annual power consumption of a data centre. The closer the number gets to 1, the more


power efficient it is. The average PUE for data centres around the world is 1.8.
d Over 99 per cent of invasion attempts are cut off automatically. Known “distributed denial of service” attacks can be


blocked in real time, whereas unknown new attacks can be blocked within ten minutes by applying the rule-sets (about
18,500 as of June 2013).




60 INFORMATION ECONOMY REPORT 2013


sector, in February 2013 as part of the country’s
e-Transformation Agenda. The cloud makes
standardized IaaS, SaaS and PaaS services
available to government agencies to reduce costs,
improve information management and share data.
It also consolidates capacity in government data
centres to improve resource allocation and lower
maintenance costs. The Government now applies
a cloud first policy, requiring every agency to
consider using the Moldova Cloud before procuring
hardware or upgrading legacy IT systems.


• Galaxy Backbone, a state-owned company which
provides infrastructure for government entities in
Nigeria, is in the process of developing a national
cloud infrastructure which will offer its public
sector clients backup and disaster recovery, IP


telephony and unified communications, and IT
project management, in addition to other IaaS,
SaaS and PaaS services (Research ICT Africa,
2013). One of Nigeria’s state governments – in
Rivers State – has launched its own private cloud
(known as RivCloud) to provide public agencies
with storage and application hosting, likely to
include an application for tax filing (Research ICT
Africa, 2013).


• The Government of India is taking steps to develop
a government cloud, which will serve both the
national and state government levels (box III.6).


These national government cloud initiatives are new,
and it remains to be seen how they will work out in
practice. As well as large-scale coordinated cloud


Box III.6. Towards a government cloud in India


The Government of India has taken a proactive stance in fostering government cloud use. The Department of Electronics
and IT (DeitY) unveiled its national “GI Cloud” initiative in 2013.a The GI Cloud is a private cloud computing environment
that will be used by government departments and agencies at the national as well as state levels. It is intended to
support the implementation of the National e-Governance Plan of India. Ultimately, the GI Cloud will provide services to
government departments, citizens and businesses over both the Internet and mobile phones.


To provide strategic direction and guidance to DeitY on matters pertaining to the functioning of the GI Cloud, an
“Empowered Committee” has been proposed under the chairmanship of the Secretary of DeitY, with representation from
Central/State line ministries and other government entities. In addition, a task force has been set up to give direction to the
creation of a detailed cloud strategy, including cloud architecture, cloud implementation and a roadmap for future activity.
This task force comprises representatives not only from national and state governments but also from the private sector
(National Association of Software and Service Companies, Gartner, Tata Consultancy Services, Cisco, Microsoft and HP).


The DeitY in April 2013 released a roadmap to GI Cloud.This assesses the ICT infrastructure currently in place and sets
out an implementation plan.b The roadmap underlines that cloud computing environments will be established at the
national and state levels using new and existing data centres. There are already large national data centres (NDCs) run by
the National Informatics Centre in Delhi, Hyderabad and Pune. Another NDC is being set up at Bhubaneswar. The largest
NDC, in Delhi, is at an advanced stage of virtualization. State data centres in 21 states have been made operational and in
four states are in an advanced stage of implementation. As of 2013, in about 10 of these centres, infrastructure utilization
had reached more than 50 per cent.


The roadmap also identifies potential challenges, such as those related to security and interoperability, shortage of
technical competencies and the need for changes in government procurement norms. In terms of security-related
aspects, DeitY will prescribe standards around interoperability, integration, data security, portability, operational aspects
and contract management for the cloud. A dedicated unit will be responsible for defining guidelines on security and for
prescribing appropriate steps to mitigate risks.


Ultimately, the GI Cloud will provide IaaS and PaaS as well as SaaS. It is anticipated that the national cloud will be able to
provide computing, storage and network infrastructure, backup and recovery, and application development as a service,
supported by the state clouds. Different business models will be considered, such as pay-per-use, subscription or offering
services free of charge. The national cloud will be owned and run by a central government agency.


Source: India, Department of Electronics and Information Technology (2013).
a See http://deity.gov.in/sites/upload_files/dit/files/GI-Cloud%20Strategic%20Direction%20Report%281%29.pdf (accessed


9 October 2013).
b See http://deity.gov.in/sites/upload_files/dit/files/GI-Cloud%20Adoption%20and%20Implementation%20Roadmap


%281%29.pdf (accessed 9 October 2013).




61CHAPTER III : IMPLICATIONS FOR DEVELOPING COUNTRIES


deployments, individual government agencies also
have experience of using cloud applications to deliver
specific public services, for example in the education
and health areas.26 In education, there is potential to
extend the range of content available to students at all
levels – both through access to published material and
greater accessibility for teaching resources. Health
records managed through the cloud may be much
more accessible at point of need, and more easily
shared between health professionals, enabling better
coordination of patient care. Cloud applications can
also be used to deliver other government services,
including transactions such as bill payments and
individual certification (driving licences, pension
entitlements, land records, and the like). Data
protection and privacy considerations, however, are
obviously important in these contexts.


There are differences in the ways government
departments and enterprises make use of cloud
provisioning.


First, Governments are likely to have many more data
sets available. Aggregating these data sets is likely to
add value for both administration and service delivery
(though it may also have privacy implications). For
these benefits to be achieved, Governments need to
ensure that different Government departments have
the necessary resources to share data effectively
and that the cloud services which they commission
are interoperable. A government-wide approach
to cloud procurement can help to maximize
administrative benefits from cloud provisioning, while
joint procurement should also reduce costs. The
experience of the Republic of Korea is illustrative in
this context (UNCTAD, 2012a).


Secondly, Governments may have particular security
concerns, and may be more constrained by legal
restrictions on where national data can be held,
inhibiting the use of data centres that are distributed
around the world. Given the volume of government
data, and the potential scale of cloud provisioning,
these constraints offer opportunities for local data
centres and cloud providers. Partnerships between
local data centres and global cloud providers, locating
sensitive and non-sensitive data in separate locations,
could enable Governments to maximize cost benefits
while fulfilling requirements for security and control.
Successful local data centres and cloud providers in
one country may subsequently be able to provide a
service for neighbouring countries, particularly within
regional economic communities.


A shift to the cloud may generate some additional
opportunities for Governments. Some Governments
use cloud applications to enhance e-government
initiatives which are explicitly targeted to businesses,
including transaction interfaces concerned with trade
facilitation, licensing and taxes. The introduction of
single-window processes in trade facilitation is an
example, which, like open data, can be implemented
through the cloud or through managed proprietary
systems (Adam et al., 2011). Cloud provisioning can
make it easier for different government agencies
to make official data available, and to do so flexibly,
enabling businesses, citizens and NGOs to target and
make innovative use of information about their own
localities. The cloud is also well suited to support
open data and other transparency initiatives aimed at
making public information more widely accessible.


(c) Large business customers


At the global level, almost all major TNCs are now
using cloud computing to some degree. For example,
it has been estimated that in 2012, 80  per cent of
Fortune 1000 enterprises were paying for some cloud
services.27 The cloud strategies of large international
businesses with operations in developing countries are
determined at a global rather than national level. Like
Governments, such companies often rely on private
cloud arrangements with global providers that can
meet their needs in every country and maximize the
value of economies of scale. Sometimes, businesses
deploy their own resources in this way. However, where
global corporations (such as banks and large retail
businesses) reach extensively into domestic markets,
they may make use of local cloud providers and cloud
services to extend beyond their own marketing and
customer relations operations.


Large national businesses are also adopting cloud and
cloud services where these offer competitive advantage
in managing their businesses. This is clearly the case
in developed countries.In the United States, the
share of businesses adopting cloud services doubled
between 2010 and 2011, from 22 per cent to 45 per
cent of large companies, from 17 per cent to 36 per
cent of medium-sized businesses and from 7 per cent
to 13 per cent of small ones (Gentzoglanis, 2012). It is
also increasingly the case in developing countries. In
Kenya, for example, all large companies in the Nairobi
Stock Exchange are in the cloud in one way or the
other (Kituku, 2012). To varying degrees they have
moved messaging, collaboration, human resources,




62 INFORMATION ECONOMY REPORT 2013


payroll, CRM/sales management, accounting, finance,
project management and application development to
the cloud.


Anxieties about reliability, security and control influence
decisions by major corporations as well, as confirmed
in a study of companies listed on the Nairobi Stock
Exchange (Kituku, 2012). Similar concerns also
featured prominently during a 2013 meeting of chief
information officers (CIOs) in South Africa. They
stressed the importance of looking at the relative
merits of different solutions, including the competing
benefits of traditional and cloud-based approaches
to business management, and of integrated and non-
integrated approaches to cloud deployment.28 Many
businesses have experienced difficulties in integrating
cloud applications with their existing software and
business practices (Dynamic Markets Ltd., 2013):
“68  per cent of cloud adopters have attempted
integration, but 86  per cent of these encountered
negatives along the way – in fact, 55 per cent have
tried and failed”.Many businesses in the same survey
had abandoned at least one cloud application in
frustration.29 This further underlines the importance of
making thorough and realistic assessments of likely
outcomes of cloud adoption rather than relying on
assumptions that gains will be made.


(d) Small and medium-sized enterprises


The SME category of enterprises is highly diverse.
Depending on the nature of their size, industry, nature
of activities and geographical location, they may be
more less intensive users of ICTs.30 Consequently,
the adoption of cloud services by SMEs is unevenly
distributed. Those with very few employees have
significantly lower adoption rates. For example, a
survey of SMEs commissioned in 2012, covering
13 countries (including Brazil, China, Turkey and the
Russian Federation) confirmed that cloud use is greater
in medium-sized than in small firms (Microsoft, 2012).
However, almost all respondents expected their use
of the cloud to grow in the next few years. This would
suggest that cloud provisioning will displace legacy
systems in the majority of the SMEs surveyed as and
when the latter come to need upgrading.


Many SMEs may find webmail services preferable
to proprietary alternatives, particularly when they are
able to cloak these in their own commercial identities.
Some, as discussed earlier in this chapter, can gain
considerably from the use of other cloud-based
social media services in marketing and customer


relations. In some cases, specialist cloud services
will be particularly valuable – for example, for some
professional freelancers or for SMEs in sectors such
as tourism that relate to large groups of remote
potential customers.


On the other hand, as mentioned above, the challenges
of integrating cloud and legacy software and the
loss of control over data and applications can make
migration undesirable for some SMEs. Some small
businesses report having moved away from the cloud
altogether in light of such experience. Some start-ups,
for example, have benefitted from the cloud initially
but then found that they were better off investing
in their own equipment as their businesses mature
and stabilized; others may experience a different or
opposite trajectory.31


Among the SMEs that have adopted cloud services,
the number of applications that have been taken up is
low. On average, the 3,000 respondents that took part
in the survey cited earlier were using four paid services
(in addition to free services such as webmail) (Microsoft,
2012). They expected this to increase to an average of
six applications over a two to three year period. Based
on the results from that survey, the cloud experience
for SMEs to date, even in developed countries, seems
to relate mostly to standard, consumer-oriented cloud
services which are available cost-free or at low cost,
rather than to specialized services oriented explicitly
at businesses. The most commonly used services
included:


• Webmail (40 per cent of current users);
• Voice communications (23 per cent);
• Instant messaging (23 per cent);
• Data storage and back up (22 per cent);
• File-sharing (21 per cent).


In the same study, informants reported using a range
of sources to obtain information about cloud options,
the most commonly cited being the websites of cloud
providers and other software companies, industry
analysts and blogs, IT consultants with whom they
were familiar and business associates. Thus, even
where they have access to a wide range of potential
sources of advice, SMEs may not be assessing
options systematically and may benefit from the kind
of services offered by cloud brokers and aggregators.


For many small firms, mass-market free-to-use
services can provide effective resources for marketing
and customer relations management. Medium-sized
firms, however, and those in more complex markets,




63CHAPTER III : IMPLICATIONS FOR DEVELOPING COUNTRIES


could gain advantage from using more sophisticated
cloud services where possible, as illustrated by the
case of Zenga Media (box  III.7). At the other end of
the scale, independent professionals can also benefit
from cloud services. Global online marketplaces allow
independent entrepreneurs to enter product and
service markets, including North–South and South–
South markets, which were previously unavailable to
them because there were no suitable procurement
channels for low-value service exports.


These services give small developing-country
enterprises with suitable skills the opportunity to
market services beyond their local communities. The
economic value which individuals derive from them
comes not from cost savings but from improved
access to markets. Where this access is international,
it contributes to national export performance as well
as revenue generation for individuals. However, much
more analysis is needed to assess the significance of
these job-search services in developing countries.


Developing-country SMEs need to consider the
pros and cons of using cloud-based applications
for their specific businesses and business models.
This consideration includes benefits and potential
disadvantages for their customers, the costs and
benefits for the SME itself, the risks that may be
incurred by relying on services that run in the cloud,
and the risks of not using the cloud when competitors
do so. The calculation of these costs and benefits will
depend partly on the circumstances of the firm itself,
and partly on the wider business and communications
environment in which it operates. Much will depend
on how easily the gains attributable to the cloud can
be secured in practice. Migrating data and services to
the cloud may make sense for a company offering a
particular range of consumer services in one location,
but not for a company offering identical services
elsewhere, if the infrastructure is less adequate or


reliable. The value of cloud migration will also change
over time as national communications environments
evolve.


C. CONCLUSIONS
Cloud computing is only in its early stages in developing
countries, but its reach is rapidly growing and is likely
to continue to do so. There are several potential
advantages for businesses and other organizations
in adopting cloud services related to cost reduction,
lower capital expenditures and improvements in the
quality of administration and services to end users.
But risks also need to be accounted for, related to
data security and privacy, integration challenges and
reliability of services. It is difficult to assess the overall
impact on the national economy of different countries,
for example, in terms of employment, productivity and
growth, which depends on the extent and nature of
cloud uptake, and on the degree to which domestic
enterprises will be able to expand and develop on the
supply side of the cloud economy.


There is already significant cloud usage within countries
by citizens making individual use of available public
cloud services, including in their work places. Foreign
affiliates of TNCs make extensive use of the cloud as
part of their parent companies’ global networks. With
some wariness, Governments in developing countries
are also moving towards the cloud. Some Governments
are developing systematic cloud strategies, as part
of broader ICT strategies or sometimes in parallel
(chapter V). There is increasingly significant planned
adoption of the cloud in domestic enterprises, though
this appears to be less extensive than anticipated
by cloud advocates. Important barriers need to be
overcome within companies. Nevertheless, the trend
is towards more adoption, across a growing range of
services. Where government departments and larger


Box III.7. Benefiting from the cloud’s elasticity – the case of Zenga Media


Zenga Media is an Indian company that provides real-time access to video and television content for mobile phones.
Zenga saw the demand for its services rise from 50,000 to 7 million users over a period of about six years. It now
delivers more than 50 million items of video content to its users every month. Cloud computing enabled this growth
to take place smoothly, without recurrent capital expenditure peaks to upgrade server capacity. It also allowed it
to manage large surges in demand at particular times, such as during the Indian Premier League cricket season.
Transition to cloud-based delivery has taken place gradually, but almost all content is now delivered using cloud
resources. The company has at the same time reduced the number of IT professionals managing its services from 35
to 4 while benefiting from increased turnover.


Source: See http://www-07.ibm.com/in/city/pdf/cloudburst_for_smes.pdf (accessed 10 October 2013).




64 INFORMATION ECONOMY REPORT 2013


corporations are concerned, there is so far a general
preference for private over public cloud approaches.


The potential for cloud adoption is inhibited by
underlying infrastructure and legislative/regulatory
barriers, which raise the costs and reduce the benefits
of adoption. Security concerns are also holding back
movement towards the cloud. Government initiatives
to promote improvements in the availability, reliability
and affordability of infrastructure, and to revise legal
and regulatory frameworks for e-commerce and data
protection would be necessary to overcome such
barriers (see chapter IV).


Businesses, Governments and other organizations
should carefully examine the potential for cloud
services to improve their management and service
delivery. They should migrate data and services to
the cloud only when this can offer significant benefits
and few attendant risks. Both public and private cloud
solutions should be considered in this context, taking
into account implications for data security and privacy.


When doing so, internal systems need be adapted to
reap maximum gains from the change. Cloud providers
can tailor their offerings more clearly for developing-
country markets, recognizing the infrastructure
constraints within those markets and the security and
other concerns of potential customers.


Although the cloud provisioning market is dominated
by a relatively small number of very large providers,
there are opportunities for developing-country firms to
participate on the supply side of the cloud economy.
These include data-centre provision and management,
both independently and in conjunction with global
cloud providers; cloud aggregation and integration
services; and the development and provision of cloud
services to different groups, including local businesses
and individuals. It is important that, when they are
designing national ICT strategies, or dedicated cloud
strategies, Governments take the supply side as well
as the demand side of the cloud fully into account (see
chapter V).




65CHAPTER III : IMPLICATIONS FOR DEVELOPING COUNTRIES


NOTES
1 The survey aimed to gain a better understanding of the state of the technical infrastructure of NGOs, nonprofits


and charities and their future plans for adopting cloud technologies. Answers were received from more than 10,500
respondents in 88 countries (Tech Soup Global, 2012).


2 For example, in a survey of 3,000 SMEs, where four out of ten informants felt that cloud services remained unproven
and therefore risky, security and control were the prime concerns (Microsoft, 2012). About 40 per cent of the SMEs
responding were in a developing or transition economy. More than two thirds of the SMEs were anxious to know
where their data would be located, and more than half expressed that concerns over data privacy would deter them
from moving some functions to the cloud.


3 See CIO East Africa “103 Government of Kenya websites hacked overnight”, available at http://www.cio.co.ke/
news/main-stories/103-Government-of-Kenya-websites-hacked-overnight (accessed 4 October 2013).


4 The survey collected responses from 1,355 companies with revenues of £50 million (about $80 million) or more.
The companies were based in 17 countries, including Brazil, China, India, the Russian Federation, Singapore, South
Africa, Turkey and the United Arab Emirates. At least 50 responses were collected from each country.


5 See, for example, “Why more SMEs are embracing cloud computing”, Times of India, 2 July 2013; Research ICT
Africa (2013) and Tech Soup Global (2012).


6 For example, the UNCITRAL Model Law on Electronic Signature from 2001 has so far been adopted by
26 developing economies. See http://www.uncitral.org/uncitral/en/uncitral_texts/electronic_commerce/2001Model_
status.html (accessed 7 October 2012).


7 For a typical website hosted in Asian cities such as Hong Kong (China), Singapore or Tokyo, the firewall allegedly
added 10 to 15 seconds to load in China. For a website hosted in the United States, it would take 20 to 40 seconds to
load in China. See “How to do online business with China”, TechWeek Europe, 22 February 2013, available at http://
www.techweekeurope.co.uk/comment/how-to-do-online-business-with-china-108291 (accessed 7 October 2013).


8 The Information Technology and Innovation Foundation estimates that United States cloud service providers might
lose between 10 per cent and 20 per cent of the non-United States market for cloud provision as a result of recent
revelations concerning government surveillance of data communications, (see Castro, 2013).


9 See also Source8 et al. (2013).


10 See “Glo to build Nigeria’s biggest data centre”, BiztechAfrica, 27 June 2013, available at http://www.biztechafrica.
com/article/glo-build-nigerias-biggest-data-centre/6327/ (accessed 7 October 2013).


11 See “Safaricom offers locally hosted cloud service: Plans to run its Mpesa pan-African mobile money offering on
new platform”, Computerworld Kenya, 1 November 2011, available at http://www.pcadvisor.co.uk/news/mobile-
phone/3314899/safaricom-offers-locally-hosted-cloud service/ (accessed 7 October 2013).


12 See “Power, pollution and the Internet”, The New York Times, 22 September 2012, available at http://www.
nytimes.com/2012/09/23/technology/data-centers-waste-vast-amounts-of-energy-belying-industry-image.
html?pagewanted=all&_r=2& (accessed 8 October 2013). See also http://www.analyticspress.com/datacenters.
html (accessed 8 October 2013).


13 See “African NGO taps IT to help prevent HIV transmission”, Computer World, 28 January 2011, available at http://
news.idg.no/cw/art.cfm?id=E669CBCD-1A64-67EA-E47FFECA76A12C2A (accessed 8 October 2013).


14 See Leo Apotheker (2011), “Opinion: Connectivity is no longer a tool but a global way of life”, San Jose Mercury
News. 3 December. Available at http://www.mercurynews.com/opinion/ci_17596575?nclick_check%C2%BC1
(accessed 8 October 2013).


15 Canonical was founded by a South African entrepreneur, Mark Shuttleworth. It is registered in the United Kingdom.
See also Ubuntu Linux Cloud, http://www.ubuntu.com/download/cloud (accessed 8 October 2013).


16 See Ubuntu Software Center, https://wiki.ubuntu.com/SoftwareCenter (accessed 8 October 2013).




66 INFORMATION ECONOMY REPORT 2013


17 See http://www.orangescape.com/paas/platform-as-a-service/ (accessed 8 October 2013).


18 NGOs are similarly moving to cloud services, notably standardized public cloud services. The 2012 survey of
ICT decision-makers in more than 10,000 NGOs cited earlier in this chapter found that the most frequently
used applications were webmail, social networking and web 2.0 applications, file storage and file sharing, web
conferencing and office productivity (Tech Soup Global, 2012).


19 See http://www.internetworldstats.com/facebook.htm (accessed 8 October 2013).


20 See http://www.alexa.com/siteinfo/facebook.com (accessed 8 October 2013).


21 As estimated by the web information provider Alexa, see http://www.alexa.com/topsites/countries (accessed
8 October 2013).


22 See www.alexa.com (accessed 8 October 2013).


23 For example, Facebook’s revenue from mobile advertising increased from virtually nothing a year ago to 41 per cent of
its total advertising revenue of $1.6 billion in the second quarter of 2013. See, for example, http://bits.blogs.nytimes.
com/2013/07/25/daily-report-facebooks-mobile-ad-revenue-cheers-investors/ (accessed 8 October 2013).


24 See “Govt pushing its agencies into the cloud”, The Nation, 28 May 2013, available at http://www.nationmultimedia.
com/technology/Govt-pushing-its-agencies-into-the-cloud-30206986.html (accessed 9 October 2013).


25 See “Cloud Security Alliance and Electronic Government Agency (EGA) of Thailand partner to drive cloud
computing adoption in the Association of Southeast Asian Nations.” Cloud Security Alliance press release.
Available at https://cloudsecurityalliance.org/csa-news/csa-electronic-government-agency-ega-of-thailand-
partner/ (accessed 9 October 2013).


26 See, for example, Cowhey and Kleeman (2012) and http://www.grameenfoundation.org/sites/grameenfoundation.
org/files/MOTECH_Suite_Overview_Nov2012.pdf (accessed 10 October 2013).


27 See “Cloud computing – An enterprise perspective”, presentation by Raghavan Subramanian, Infosys Technologies,
available at http://research.microsoft.com/en-us/people/sriram/raghu-cloudcomputing.pdf (accessed 9 October 2013).


28 See http://www.brainstormmag.co.za/index.php?option=com_content&view=article&id=4756%3Acio-roundtable-
castles-in-the-clouds&Itemid=124 (accessed 9 October 2013).


29 For an assessment of cloud experience, see Trip Advisor, available at http://highscalability.com/blog/2012/10/2/
an-epic-tripadvisor-update-why-not-run-on-the-cloud-the-gran.html (accessed 9 October 2013).


30 The different impacts of ICTs on these different types of SME, and the different policy approaches appropriate to
them, were explored in the Information Economy Report 2010 (UNCTAD, 2010).


31 See “Why some startups say the cloud is a waste of money”, Wired, 15 August 2013, available at http://www.wired.
com/wiredenterprise/2013/08/memsql-and-amazon/ (accessed 9 October 2013).


32 See http://www.freelancer.com/?utm_expid=294858-31.GWNoRHpfQwCRcsmjo9a25Q.0 (accessed 10 October
2013) and https://www.odesk.com/ (accessed 10 October 2013).




GOVERNANCE,LAW
AND REGULATION
OF CLOUD
SERVICES
IN DEVELOPING
COUNTRIES 4


The rapid emergence of cloud computing has inevitably raised concerns
about the legal and regulatory implications of such developments. From the
regulatory characterization of cloud services to the privacy implications of
processing personal data remotely, policymakers, legislators, regulators,
service providers and users all have an interest in the governance of cloud
services. Governments will want to protect national interests, including the
protection of their citizens; service providers require a stable framework to
facilitate innovation and investment; users require assurance and trust to
encourage the take-up of such services. This chapter examines some legal
developments relating to the cloud, identifies emerging regulatory responses,
and considers certain issues of particular concern for developing countries.




68 INFORMATION ECONOMY REPORT 2013


A. INTERNATIONAL
LEGAL AND
REGULATORY
TRENDS


As discussed in preceding chapters, cloud computing
can be seen both as a return to an earlier model of
computing, as well as an innovation in the exploitation
of ICTs, through the exploitation of shared and
location-independent processing infrastructure.
This duality of perspective is also reflected in the
governance debates surrounding cloud computing.
Concerns about data sovereignty and privacy1 first
appeared during the 1970s, as the emergence of
more sophisticated telecommunication systems
made it feasible for large TNCs to move bulky data
sets between countries to take advantage of more
advanced technological capabilities available in
certain locations, primarily the United States, as
well as economies of scale. Such transborder data
flows (TDFs) were viewed by some with concern
(Seidman, 1986). For some Governments, the transfer
of critical commercial data to another country was
seen as generating an unwelcome dependency
and vulnerability for the State, with implications for
national security, analogous to recent concerns
about the provision of energy. In response, countries,
including Brazil and Canada, introduced legislation
requiring that certain data remain within the national
territory. National security was not the only concern;
the economic impact in terms of a perceived loss of
domestic economic activity from TDFs also lay behind
some such policy responses. A parallel and related
concern arose from the fact that TDFs could enable
companies to circumvent emerging national privacy
laws, especially in Europe, which were designed to
protect the privacy of citizens in an age of computing
(Walden and Savage, 1990).


The economic and data-sovereignty concerns of
the 1970s generally dissipated in the face of the
liberalization of trade in services, manifest in the
General Agreement on Trade in Services (GATS)
in 1994, under the auspices of the World Trade
Organization (WTO). The TDFs that underpin the
cross-border provision of services were seen
as offering greater benefits to nations than any
concurrent enhanced vulnerability. Developing
countries, in particular, have reaped comparative
economic advantages in certain sectors, with the
emergence of services such as the offshoring of


business processes. However, data-sovereignty
concerns have re-emerged as a governance issue
within cloud computing for a number of reasons (CIO,
2012). First, there are concerns that data placed with
global cloud service providers may be accessible by
law-enforcement agencies of their home country.
Second, as Governments increasingly look to
improve their efficiency and effectiveness through
e-government initiatives, there are concerns that
data fundamental to the functioning of Governments
may move offshore into the cloud (see, for example,
United Kingdom, Cabinet Office, 2011). Third,
the threat of cybercrime, and even cyberwarfare,
and the associated vulnerability of critical national
infrastructure and data, is another manifestation of
the national security concern (Kshetri, 2010).


In response to the privacy concerns of the 1970s, legal
instruments, particularly European data-protection
laws, expressly addressed the possibility of restricting
TDFs. Some harmonization measures in the field were
driven primarily by the desire to reduce the possibility
that privacy rules would operate as a barrier to
international trade, the leading example being the
1980 OECD Guidelines (OECD, 1980). Others, such
as the 1981 Council of Europe Convention, included
default rules stating that TDFs should only take place
between jurisdictions with “equivalent” levels of
privacy protection (Council of Europe, 1981).2 Indeed,
under the GATS, there is specific provision permitting
member States to provide for a general exception from
the liberalized and non-discrimination provisions for
the purpose of protecting privacy and confidentiality.3


Over the intervening years, while the nature of
TDFs has evolved way beyond that imagined some
40 years ago, concerns about the potential erosion
or circumvention of national privacy protections
remain at the forefront of the policy response to cloud
computing. There is no harmonized international
privacy framework regulating data transfers across
borders (see box IV.1), but developing countries could
benefit from implementing strong domestic privacy
regimes.


The governance framework for cloud computing can be
divided broadly into two parts. On the one hand, there
are laws and regulations adopted by Governments,
public administrations and independent regulatory
authorities – collectively referred to as “public law” for
the purpose of this chapter – that are either directly
targeted at the provision or use of cloud services, or
impact such provision and usage by virtue of being




69CHAPTER IV : GOVERNANCE, LAW AND REGULATION OF CLOUD SERVICES IN DEVELOPING COUNTRIES


applicable across a range of similar activities or
behaviour (such as data-protection laws). To date, the
public law response to cloud computing has primarily
been of the latter kind, that is, applying existing rules
in a cloud-based environment. A second governance
framework comprises the contractual agreements
entered into between the various providers of the
cloud ecosystem and the end users of the services.
While such “private law” arrangements are given
legal recognition and enforceability by the public-law
framework of a State, including the incorporation
of national mandatory requirements, the provisions
contained within such agreements more directly
impact on the operation of the cloud economy, for
providers and users.


These public and private law aspects to the
governance of cloud should be viewed as
complementary layers of protection, rather than
as substitutes. Public law approaches are more
comprehensive at meeting public policy objectives
(for example, cloud promotion) and addressing public
policy concerns (for example, consumer protection),
although they can be time-consuming to adopt
and require adequate resources to ensure effective
implementation and enforcement.


Box IV.1. Protecting personal data in the cloud – different options and challenges


While the importance of protecting personal data in a cloud environment is broadly accepted, jurisdictions diverge
significantly over the most appropriate regulatory mechanisms to achieve this, reflecting differing cultural and social
attitudes towards the concept of privacy and its protection. As of mid-2013, there were some 99 countries with data-
privacy laws of varying types (Greenleaf, 2013). European Union law imposes a general prohibition on transfers to
jurisdictions without “adequate” protection, to ensure that individual rights are not subverted by moving data outside the
jurisdiction, and requires the oversight of an “independent supervisory authority”. Examples of such national authorities
are the Office of the Information Commissioner in the United Kingdom and La Commission nationale de l’Informatique et
des Libertés in France.


By contrast, the Asia-Pacific Economic Cooperation Privacy Framework (2005), to which the United States is party,
subjects data transfers to an “accountability” principle,a but no general prohibition on transfers to countries with no data
protection laws. There is no requirement for an independent enforcement authority.


These very real differences in approach present developing countries with potentially difficult choices when adopting
domestic data-protection measures. Such adoption is often driven more by international trade concerns than by domestic
policy concerns. Yet data protection regimes – as with all legislation – also represent a bureaucratic and compliance cost
to developing-country Governments and domestic businesses. The current absence of an international consensus on
what constitutes an “adequate” data-protection regime, in terms of core content principles, as well as procedural and
enforcement mechanisms, can itself represent a barrier to cloud computing (Kuner, 2013).


Source: UNCTAD.
a According to article 26 of the Asia-Pacific Economic Cooperation Privacy Framework “A personal information


controller should be accountable for complying with measures that give effect to the Principles stated above. When
personal information is to be transferred to another person or organization, whether domestically or internationally,
the personal information controller should obtain the consent of the individual or exercise due diligence and take
reasonable steps to ensure that the recipient person or organization will protect the information consistently with
these Principles.”


B. PUBLIC LAW AND
CLOUD SERVICES


1. Cloud as a regulated activity


A fundamental distinction exists between activities
subject to the general law and those that are subject
to specific legal and regulatory obligations by virtue
of the type of activities being undertaken (such
as banking and finance, and health services and
telecommunications). Therefore, one question raised
by cloud computing is whether it falls within existing
regulatory spheres and, if not, whether it should be
made subject to some form of cloud-specific regime.


The cloud-economy ecosystem includes the provision
of transmission services to enable data to be
transferred to and from remote data centres (chapter
I). The provision of such transmission services is
generally subject to sectoral regulation as a form of
telecommunications. Progressive global liberalization
of the telecommunications sector since the 1980s
has resulted in a broadly harmonized but complex
regulatory environment, which can be represented by
the WTO “Telecommunications services: Reference




70 INFORMATION ECONOMY REPORT 2013


paper” for telecommunications.4 National regulatory
authorities exercise varying degrees of control and
oversight over market participants, imposing specific
obligations on the provision of services.


Companies that provide PaaS and SaaS are generally
viewed as falling outside the regulatory concept of a
telecommunication service. However, certain SaaS
applications that provide call-handling functionality
for enterprises could be regarded as a regulated
telecommunication service, depending on a country’s
particular regulatory regime.5 While such cloud services
are dependent on telecommunication networks and
services for communicating with their customers,
they are not per se generally characterized as being a
telecommunication facility, network or service. Certain
IaaS, however, may provide processing, storage
and connectivity services that may be viewed as a
telecommunication service depending on the primacy
of the connectivity component (ITU, 2012a). In 2012,
the Korean Communications Commission published a
proposed “Bill for the development of cloud computing
and protection of users”, which would have classified
cloud computing as a telecommunications service.
The proposal met with strong opposition from cloud
providers and was being reconsidered at the time of
preparing this report.6


In some jurisdictions, primarily in developing countries,
the regulatory net is cast wider than the provision of
telecommunication services to include the provision
of data processing services.7 In such cases, cloud
services – be they SaaS, PaaS or IaaS – are likely to fall
squarely within the regulatory sphere. Such regimes
enable Governments to impose conditions on the
supply of services and might help address concerns
including data sovereignty and privacy. In Indonesia,
for example, recent regulations require those providing
“electronic systems” for the provision of public services
to locate the applicable data and disaster-recovery
centres for public services in Indonesian territory
“for the purpose of law enforcement, protection and
enforcement of national sovereignty to the data of its
citizens”.8


In other countries, while cloud services are not subject
to comprehensive sectoral regulatory regimes, they
may fall within regulatory concepts designed to shield
certain service providers from liability for the content
that they make available on behalf of others. In South
Africa, for example, the provision of “information
system services”, which would extend to most cloud
services, is subject to special provisions limiting the


liability of service providers in respect of unlawful
content made available by others.9


Uncertainties over the regulatory characterization
of cloud computing may deter its acceptance until
legislators or regulators clarify the situation. Issues
of administrative competence are obviously linked to
such uncertainty. If viewed as a telecommunications
service, then the telecommunications regulator can
exercise jurisdiction. If a cloud service is seen as
a data-processing service, then competence may
lie with the national ICT regulator, if there is one, or
alternatively the media regulator if it is viewed as a
content-like service. Such sectoral regulators may
also have to operate in conjunction and cooperation
with any horizontal national regulators, such as data-
protection or consumer-protection authorities, in
respect of certain issues.


To date, few jurisdictions have attempted to draft
regulations expressly designed to regulate the
provision of cloud services. This probably reflects both
the broad range of services that fall within the concept
of cloud, as well as the flexibility of scope within
existing regulatory concepts. As far as is known,
Mexico is the only country which has adopted cloud-
specific provisions in relation to data protection (box
IV.2). Some of these rules are merely restatements of
generally applicable obligations, while others are more
“cloud appropriate”, designed to address concerns
that have arisen in a cloud context, specifically
regarding transparency about the layered nature
of the cloud supply chain (see also section IV.C),
regarding the treatment of user data following service
termination, and regarding law enforcement access
(Bradshaw et al., 2011). The Mexican approach
intends to encourage the domestic take-up of cloud
solutions.


2. Regulating in a multi-jurisdictional
environment


If regulatory characterization is one key concern for
policymakers, another is the perceived loss of oversight
and control over applications and data resulting
from the remote nature of cloud service provision.
Location independence is a key characteristic of cloud
computing, which exploits the “death of distance”
made possible where adequate communication
networks are in place (Cairncross, 1997).10 This results
in a multiplicity of jurisdictions potentially competing
to govern the various parts of a cloud service. The




71CHAPTER IV : GOVERNANCE, LAW AND REGULATION OF CLOUD SERVICES IN DEVELOPING COUNTRIES


movement of data into and out of a cloud service will
often result in the data becoming subject to the rules
of both the cloud user’s jurisdiction and those of the
providers of the cloud and inter-cloud services. The
transfer of data out of the user’s jurisdiction is often
opaque to the user, raising issues of control and, for
the relevant national regulator, effective oversight and
audit. For some regulated sectors, such as financial
services, cloud-related transfers and storage outside
the jurisdiction of the regulated entity may itself breach
national rules by failing to provide national authorities
with “effective access” to the data.11


Due to concerns about sovereignty, national
regulators may be unwilling to surrender jurisdiction
to a foreign authority unless adequate mutual
recognition arrangements are in place. For some
types of processing, such as those relating to national
security, no foreign involvement may be acceptable


and a national solution will be required. Effective
mutual recognition requires greater transparency,
more dialogue and closer cooperation between
national regulators to resolve conflicts of law and
regulation in a cloud environment, and to facilitate the
free flow of data. Significant examples of cooperative
networks of national regulators and law enforcement
agencies already exist in the areas of consumer
protection, cybercrime and data protection,12 while
legislative proposals have been put forward to
facilitate international cooperation and interoperability
with respect to cloud.13 The European Commission,
in its cloud strategy, noted that “being born global”,
cloud computing called for “a reinforced international
dialogue on safe and seamless cross-border use”,
including the need for “legal adjustments to promote
efficient and effective cloud roll-out” (European
Commission, 2012).


Box IV.2. Legislating for cloud: the case of Mexico


Policies on ICT and the regulatory framework constitute a basis for developing cloud computing services in Mexico. In
relation to privacy and data protection, a federal law on the “Protection of Personal Data Held by Private Parties” was
adopted in 2010. Regulations made under this law were adopted in December 2011 and include specific provision for
cloud computing.


Article 52 imposes certain obligations on cloud service providers offering “services, applications, and infrastructure” to
data controllers “by general contractual conditions or clauses”. The provider must ensure that it (the provider):


(a) Complies with at least the following:


(i) Policies to protect personal data similar to the applicable principles and duties set out in the Law and its
regulations;


(ii) Makes transparent any subcontracting that involves information about the service which is provided;


(iii) Abstains from including conditions, when providing the service, that authorize or permit the provider to
assume the ownership of the information related to the service provided;


(iv) Maintains confidentiality with respect to the personal data associated with the service provided.


(b) Ensures that the contract contains mechanisms at least for:


(i) Disclosing changes in its privacy policies or conditions of the service it provides;


(ii) Permitting the data controller to limit the type of processing of personal data associated with the service
provided;


(iii) Establishing and maintaining adequate security measures to protect the personal data associated with the
service provided;


(iv) Ensuring the suppression of personal data once the service has been provided to the data controller and
ensuring that the latter may recover the data;


(v) Impeding access to personal data by those who do not have proper access or in the event of a request duly
made by a competent authority, to inform the data controller.


The user of the cloud services, as “data controller”, is obliged not to use services that “do not ensure the proper protection
of personal data”.


Source: Mexico, Federal Institute for Access to Public Information, see http://inicio.ifai.org.mx/_catalogs/masterpage/
ifai.aspx (accessed 22 October 2013).




72 INFORMATION ECONOMY REPORT 2013


Another dimension of the multi-jurisdictional nature of
cloud is concern about the potential implications of the
regulatory regime of the country of origin of the cloud
provider. This has been highlighted, for example, in the
context of data placed with global cloud service providers
and the possibility for law-enforcement agencies to
access such data even if stored outside of the jurisdiction
of their home country. As noted in a recent report for the
European Parliament (Bigo et al., 2012):


Cloud providers are transnational companies subject
to conflicts of international public law. Which law they
choose to obey will be governed by the penalties
applicable and exigencies of the situation, and in
practice the predominant allegiances of the company
management.


The launch of Microsoft’s Office 365 in June 2011,
for example, was accompanied by expressions of
concern that Microsoft would not guarantee that
data of European customers could not be accessed
by agencies acting under United States jurisdiction.14


Similar concerns were behind statements by the
Government of the Netherlands, suggesting that cloud
providers from the United States in government bids
and contracts could be excluded.15 In Australia, the
Office of the Victorian Privacy Commissioner in 2011
warned government agencies about the use of cloud
computing, suggesting that it could be “impossible”
to protect personal information held about Australians
when located offshore or even outside Victoria.16


In 2013, media reports brought the issues of law
enforcement access to cloud-based services to the
forefront of the policy debate on cloud, indicating
that the United States National Security Agency
was obtaining large volumes of data from leading
cloud service providers, such as Apple, Facebook,
Google, Microsoft, Yahoo and Skype.17 The service
providers have denied suggestions that their services
permit some form of “back-door” access to national
agencies, but have accepted that they regularly
disclose customer data in response to requests from
law-enforcement agencies in a number of countries.18


It has been stressed that such disclosures, whether
carried out under legal mandate or voluntarily, might
expose the cloud service provider to liability under the
laws of the customer’s jurisdiction (Bigo et al., 2012;
Walden, 2011).


Related concerns may widen the scope for new
competitors to enter and expand in the market.19


Some European providers have called for certification
schemes that would indicate where data is protected


from such access.20 At the same time, almost all
countries have legislation that gives their authorities
the right for reasons of national security to request
data on cloud services stored in their jurisdiction.21


The issue is ensuring that these powers are used only
when necessary.


To the extent that cloud computing services are
located outside the domestic jurisdiction, policy and
legal issues arise as to whether large-scale data
transfers are appropriate or permissible and whether
regulatory control measures can be imposed or
maintained once the data have left the jurisdiction.
Harmonized statutory regimes would significantly
mitigate such concerns, but at present, contractual
agreements between providers and customers are the
main mechanism available to address them, at least to
some extent.


C. PRIVATE LAW AND
CLOUD SERVICES


As noted earlier, the cloud ecosystem is to a high
degree built on private law agreements between
service providers and between service providers
and cloud service customers. Such contracts offer
service providers and customers a self-regulatory
mechanism for generating a framework of legal
certainty and security in cloud computing. Cloud
contractual arrangements come in varying shapes
and sizes, but will generally comprise four distinct
components, whether in a single agreement or a set of
linked documents (generically referred to as the “cloud
contract”):
• Terms of service, detailing the key features of


the relationship, both cloud-specific and general
boilerplate provisions (for example, choice of law);


• Service-level agreement, detailing the service
features being provided, the standards that they
should meet (for example, service uptime) and any
compensation mechanism when the standards
are not met;


• Acceptable-use policies, detailing permitted
or impermissible conduct by users (for example,
copyright infringement);


• Privacypolicy, detailing the approach taken to the
storage and processing of user data, particularly
consumer information.


The terms of a cloud contract can be distinguished
into standard terms and cloud-specific provisions.
The standard terms will include such matters as




73CHAPTER IV : GOVERNANCE, LAW AND REGULATION OF CLOUD SERVICES IN DEVELOPING COUNTRIES


provider liabilities, dispute resolution and applicable
law. Both are of equal importance in terms of defining
the provider–user relationship. The cloud-specific
provisions generally focus on two key aspects: first,
the treatment of the data submitted by the cloud user
into the cloud service; second, the specifications of
the service being offered to the cloud user.


With respect to the treatment of data held in the cloud,
“cloud appropriate” terms would include provisions
ensuring that the users’ rights of ownership over the
data are not compromised, that the data are secure
against inadvertent or unauthorized disclosure, and
that they will reside on infrastructure located in a
specified jurisdiction or region. On the service side,
users will want cloud-appropriate service levels that
address concerns about dependency on the cloud
service (see for example, European Network and
Information Security Agency, 2012).22 Conversely,
providers will generally argue that the utility and
commodity nature of a “public” cloud service must
limit their responsibilities concerning issues such
as the integrity and backup of user data (Hon et
al., 2012).


From a public policy perspective, self-regulation
through contractual agreements can raise concerns
when market practice facilitates a situation where
contracts do not result in a fair balance of liabilities
and responsibilities between cloud providers and
customers, especially when the latter are SMEs and
individuals. In addition, contractual terms can have
implications for third parties that not a part of the
agreement, particularly regarding privacy and data
protection issues. In these circumstances, regulatory
intervention in the freedom to contract may be
considered necessary to rebalance the relationship
with a view to protecting third parties or the wider
public interest. Such intervention may occur through
the imposition of obligations that cloud providers are
required to address in any cloud contractual process
or agreement (see box IV.2), through certain terms and
conditions being statutorily implied in any agreement,
or the development of model cloud industry terms and
conditions or service-level agreements.23 For example,
the European Commission recently stated that model
terms may be needed “to create transparent and
fair cloud services contracts” following calls from
respondents to a survey.24 For most developing
countries, however, the challenge of intervening
in cloud contracts is obviously much greater, with
domestic consumer laws often being either inadequate
or unenforceable.


A cloud service customer will only contract directly
with the cloud service provider and, usually, a
communication service provider, and this customer
may be unclear in their understanding of the entire
chain of suppliers providing elements of the cloud
service, such as IaaS. This may represent a risk,
since the customer may not be aware of the layers
of contracts that underpin the provision of the
service and, significantly, whether commitments
entered into by the contracting service provider are
adequately reflected at each stage of the supply
chain. One proposed regulatory response is to impose
transparency obligations on the cloud service provider
(ITU, 2012b).


In the consumer market, cloud service providers
will generally dictate the terms on which the service
is offered. Such standard terms and conditions are
often biased in favour of the provider, even though
they may vary considerably according to the markets
from which the cloud provider originates: providing
hardware, software, outsourcing, communications
services or retail products. At the enterprise level,
however, service providers are increasingly being
forced to negotiate agreements that have provisions
more favourable to the customer in order to win the
business (Hon et al., 2012). It has also been stressed
that certain market entrants, such as cloud integrators
and aggregators,25 seem to be willing to offer more
user-orientated terms (Hon et al.,  2012). The issues
on which most negotiation takes place with respect
to the terms of service are: provider liability, service-
level agreements, data protection and security, and
intellectual property rights. In terms of the mechanism
of the agreement itself, the right of service providers
to unilaterally amend service features and termination
rights are key areas of dispute. While enterprise cloud
contracts will remain distinct from the consumer
segment, some of the concessions achieved in
enterprise negotiations may later trickle down into the
provider’s standard terms of business, especially if
regulators become more involved (Hon et al., 2012).


Another obvious influence on the contractual
environment for cloud services is the procurement
practices of government administrations, as they are
usually among the largest customers in the emerging
cloud market. As public authorities embrace cloud
services for the provision of e-government-related
services, they are – similarly to large enterprise users
– in a potentially strong position to negotiate more
favourable terms and conditions with cloud service
providers, including measures to protect the privacy




74 INFORMATION ECONOMY REPORT 2013


interests of data subjects/citizens. In addition, they
can utilize their “procurement weight”, expressed in
national public procurement rules, to promote other
policy objectives, including “open technologies and
secure platforms” (European Commission, 2012: 3.2)
that benefit the wider user community. In the United
States, for example, the Chief Information Officer
within the Office of Management and Budget has
issued best practice guidance for the acquisition of
cloud services (United States, Federal CIO Council
and Chief Acquisition Officers Council, 2012). This
guidance addresses the selection of the service
provider, the service-level agreement, end-user
agreements, e-discovery and record-keeping issues.
Inevitably, a key concern for the public sector is that of
security in the cloud.


While contractual arrangements are not a substitute
for public law measures, especially when protecting
third party interests, they comprise an important layer
of cloud governance that should be acknowledged
and can be influenced to help address certain policy
concerns. However, the ability to negotiate and
enforce contractual obligations can vary considerably,
especially in a developing country context.


D. IMPLICATIONS
FOR DEVELOPING
COUNTRIES


Despite calls for new cloud laws from certain sectors
in both developed and developing countries,26 what
exactly would such regulatory and governance
measures address for developing countries? This
chapter has identified three overlapping areas: (a) the
provision of cloud services; (b) the conditions under
which such services are provided to end users; (c) the
treatment of data placed in the cloud service. These
areas are generic and relevant for all jurisdictions
– whether in developed or developing countries –
but achieving effective resolution through law and
regulation is generally more difficult in developing
countries.


1. Provision of cloud services


Should the provision of cloud-related services be
a distinct regulated activity? Governing the right to
establish and/or supply a cloud service in the domestic
jurisdiction, through some form of prior licensing or
authorization requirement, is about enabling regulatory


control over market entry. In developed countries,
there are few, if any, examples of such cloud-
specific regulations. Some components of the cloud
ecosystem, particularly the communication network
component, already fall within existing regulatory
regimes, but those regimes are typically designed to
facilitate market entry and competition, rather than to
constrain market participants.


For developing countries, while the dominance of
foreign companies offering cross-border services may
be a concern, it is both practically and legally difficult
to address this market reality through regulatory
intervention. From a practical perspective, preventing
access to foreign cloud services would likely require
a radical intervention in the global connectivity of the
State, especially in respect of the Internet. From a legal
perspective, a majority of developing countries are
members of the WTO and signatories to the GATS,
committing themselves, at least in some sectors, to
liberalized trade in services, including through cross-
border supply.27 As such, regulatory intervention is
permissible only under limited exceptions, including
the protection of the privacy and confidentiality of
data.


An alternative policy response is to encourage the
establishment of domestic cloud services, either
by offering foreign service providers a favourable
environment to invest in the building of local
infrastructure (such as data centres) or encourage
domestic enterprises to enter the supply side of
the cloud economy. However, the former approach
would not necessarily overcome the problem of
the application of foreign laws, as outlined above,
especially if the local entity remains under the
control of the foreign parent. Such measures may
involve regulatory components, such as imposing
“localization” requirements, but would be designed
to facilitate the provision of cloud services rather than
to constrain them. As noted in chapter III, several
Governments of developing countries are also taking
steps to build government clouds to serve the needs
of the Government itself and sometimes also other
cloud service customers.


2. Conditions of supply


Irrespective of whether the establishment and
provision of cloud services is regulated, Governments
may choose to intervene to impose certain conditions
on the supply of cloud services to address specified
public policy concerns, such as data security and




75CHAPTER IV : GOVERNANCE, LAW AND REGULATION OF CLOUD SERVICES IN DEVELOPING COUNTRIES


privacy. Such conditions may be applicable only to
certain types of cloud services (for example, SaaS),
certain types of end users (for example, consumers
or public administrations) or be imposed upon all
cloud services. The conditions may directly impose
behavioral obligations on cloud service providers,
such as requirements to retain some types of data
within certain geographical locations or for certain
periods of time, or they may indirectly govern the
contractual arrangements between the provider and
end user, such as rendering certain “unfair” terms
unenforceable. Such public law intervention in the
conditions of supply may also have an impact on a
user’s cost–benefit analysis for cloud adoption, if
it results in a higher cost service, which should be
factored into any initiative.


Among developed countries, there has been little
adoption of cloud-specific obligations to date
(Business Software Alliance and Galexia, 2012).
Instead, relevant regulatory authorities have tended to
issue opinions as to how existing obligations, whether
for particular sectors (for example, financial services)
or of a general nature (for example, data protection)
should be applied within a cloud context (United
Kingdom, Cabinet Office, 2011).28


In Mexico, the Government has gone one step further,
imposing an express obligation on all regulatory
authorities to draft guidelines in relation to the cloud.29
For most other developing countries, relevant laws
and regulations – where they exist – are more likely to
be considered an obstacle to cloud adoption rather
than a facilitator, as was the case with the emergence
of e-commerce during the first decade of the century.
Indeed, many e-commerce and cyberlaw reform
initiatives are directly relevant to cloud computing, and
the emergence of cloud computing can act as a further
spur towards the comprehensive implementation of
these reforms.30


The ability to enforce any regulatory obligations upon
a cloud service provider will depend, in reality, on
whether the provider is either in some way established
within the jurisdiction (such as having a local office,
against which any enforcement action can be
pursued), or if the provider is prepared to voluntarily
submit itself to local rules, or at least comply with
them, even though these may be unenforceable. A key
example of the latter case are foreign service providers
responding to domestic law enforcement requests in
relation to content considered illegal, which is held
and made available via the provider’s remote servers.


Companies such as Google and Microsoft have made
public their handling of such requests in respect of
specific jurisdictions, whilst noting that they are often
not legally required to cooperate.31


Another strategy for regulators in developing
countries is to make reference, in laws or regulations,
to international technical and business standards
that specify certain good practices that are either
generally applicable to,32 or specifically designed for
cloud computing.33 Requiring compliance with such
standards, particularly when they are implemented
through an external and independent audit and
certification procedure, carried out by an accredited
certification body,34 can offer end users, third parties
and regulators some assurance with respect to the
provision of trustworthy and quality cloud services.


3. Treatment of data


In addition to regulating the conditions under
which cloud services are supplied to cloud service
customers, there are also public policy concerns that
extend beyond the relationship between provider
and customer. Data placed in the cloud can engage
third-party interests that may require regulatory
intervention, whether concerning personal privacy,
commercial secrecy or national security. Within data
protection laws, for example, imposing security-
breach notification obligations on cloud service
providers provides transparency about vulnerabilities
and enables mitigating measures to be taken in a
timely manner.


Governments will have particular data security
concerns where the end users are public
administrations, involving data about citizens or data
considered relevant to national security, as is the
case with Indonesia. In February 2013, for example,
the European Commission published a proposal for
a Directive on information security that would require
member States to designate a competent authority
to provide regulatory oversight, establish a national
computer emergency response team (CERT) to
handle security incidents and risks, and require certain
“market operators” that enable the provision of online
services, including cloud service providers, to notify the
competent authority of security breaches and undergo
security audits when required.35 These proposals are
equally valid recommendations for consideration in
developing countries, where the vulnerabilities and
risks associated with cloud computing are shared.




76 INFORMATION ECONOMY REPORT 2013


A growing number of developing countries are also
establishing CERTs and are at different stages of
development in their operations, procedures and
related activities. Many of these countries are also
faced with challenges, particularly in the areas of
capacity building, funding, the legislative framework
and other soft and hard resources. Africa has the
smallest number of CERTs.36


E. CONCLUDING
REMARKS
AND POLICY
RECOMMENDATIONS


Cloud computing is another manifestation of a
rapidly evolving ICT environment that confronts both
developed and developing countries. How to respond
appropriately to cloud computing from a legal and
regulatory perspective is not obvious and can involve
different approaches, ranging from a do-nothing
attitude to the adoption of cloud-specific laws. The
nature of the response will often depend on whether
the implications of cloud computing are viewed
overall to be positive, and therefore to be facilitated,
or negative, and therefore to be constrained. Given
the immature state of the cloud market, especially in
developing countries, considerable caution and care
should be taken by both policymakers and regulators
in terms of rushing to legislate for cloud.


While there is no imperative to develop specific
laws or regulations on cloud computing, some of
the areas requiring law reform are relatively clear:
privacy, data protection, information security and
cybercrime measures.37 As noted earlier, these are
issues of equal importance to e-commerce and other
aspects of the networked economy. This, in itself,
indicates that approaching cloud computing issues in
isolation would not be advisable. For Governments of
developing countries, it is essential that appropriate
laws and regulations are adopted in these areas and
enforced.


Governments and public administrations, as potential
users of the cloud, can offer leadership to the national
market in terms of addressing areas of concern for
users. Regulators can also assist, by issuing guidance
on the applicability to cloud solutions of the regimes
for which they are responsible. This could include, for


example, explicit recognition that certified compliance
with international technical and business standards,
especially concerning information security, will be
accepted as prima facie evidence of good market
practice. For many developing countries, such
initiatives will require the provision of external assistance
to users, policymakers and regulators, on the technical,
managerial and legal issues raised by cloud computing.


Several additional steps can be taken by policymakers
in developing countries when considering the
appropriate policy response to cloud computing:
• Consider the need to draft and ensure a coherent


and appropriate national cloud strategy, of which
legal and regulatory issues are one component.
The drafting process should be inclusive and
transparent to stakeholders (see chapter V);


• Review existing legal frameworks and enact
comprehensive laws relating to, as a minimum,
privacy and data protection, information security
and cybercrime. Aligning such laws with leading
international legal instruments in these areas is
highly recommended to achieve a harmonized set
of laws and regulations. Such instruments include
the United Nations Committee on International
Trade Law texts on e-commerce,38 in particular the
United Nations Convention on the Use of Electronic
Communications in International Contracts,39 the
Council of Europe Convention for the Protection
of Individuals with regard to Automatic Processing
of Personal Data,40 and the Council of Europe
Convention on Cybercrime;41


• Build capacity to ensure enactment and
enforcement of laws and regulations, including
establishing CERTs, training regulatory agencies
and the judiciary, and making use of support
from international organizations such as UNCTAD
(box IV.3);


• Consider the use of public procurement rules as a
vehicle for promoting particular cloud-related policy
objectives, such as enhanced privacy protection;


• Promote awareness-building in government
institutions and public administrations on the
potential that cloud computing technologies offer;


• Monitor regulatory developments in the field of cloud
computing. For such an evolving area it is important
to stay abreast of regulatory developments, in
particular in cross-border cooperation between
regulatory and law-enforcement authorities.




77CHAPTER IV : GOVERNANCE, LAW AND REGULATION OF CLOUD SERVICES IN DEVELOPING COUNTRIES


Box IV.3. UNCTAD support to e-commerce legislation harmonization


UNCTAD’s E-commerce and Law Reform Programme assists member States in preparing legal and regulatory frameworks
to facilitate e-commerce and e-government. It has become the leading capacity-building programme within the United
Nations system in support of the harmonization of legal frameworks governing the use of electronic services in developing
countries. In addition to training workshops and briefing sessions for policy and law makers, including parliamentarians,
UNCTAD reviews existing legislation and helps prepare domestic laws and regional legal frameworks to facilitate this
process. Key legal issues addressed include electronic transactions, electronic signatures and authentication, data
protection and privacy, consumer protection, computer crime, intellectual property, competition, taxation and information
security at large.


To date, UNCTAD has assisted more than 35 countries, notably in the Association of Southeast Asian Nations (UNCTAD,
2013), Central America (UNCTAD, 2009a), Latin America (UNCTAD, 2009b), the East African Community (UNCTAD,
2012b) and the Economic Community of West African States. This has successfully contributed to raising the profile
of cyberlaw issues at the national and regional level and has resulted in the adoption of law reforms. For example, in
August 2013, the second phase of the East African Community Framework for Harmonized Cyberlaws was approved for
implementation by the tenth meeting of the Sectoral Council on Transport, Communications, and Meteorology. This and
the first phase of the Framework were developed with UNCTAD assistance. Partner States are currently advancing in their
implementation of domestic laws in the relevant areas.


Source: UNCTAD.




78 INFORMATION ECONOMY REPORT 2013


NOTES
1 See, for example, chapter 6, “Protecting privacy rights in an online world” in UNCTAD (2004).


2 What constitutes “equivalent” protection is not further defined, but is considered to require a higher standard of
harmonization than the term “adequate” adopted under European Union data protection law (see box IV.1).


3 See GATS, article XIV(c)(ii).


4 Available at http://www.wto.org/english/tratop_e/serv_e/telecom_e/tel23_e.htm (accessed 22 October 2013).


5 For example, CaaS solutions are available from companies such as Level 3, Cisco and Skype.


6 See BSA case study on the Republic of Korea, available at http://cloudscorecard.bsa.org/2013/assets/PDFs/case_
studies/2013Scorecardcs_Korea.pdf (accessed 22 October 2013). See also ETNews “Cloud Computing Act came
to a deadlock due to opposition by global cloud service providers”, 16 November 2012, available at http://english.
etnews.com/computing/2677775_1301.html (accessed 22 October 2013).


7 See, for example, Indonesia, Law No. 11 on Electronic Information and Transactions 2008, which at article 1(6)
regulates the provision of “electronic systems”. A more recent regulation, No. 82 of 2012 on the Operation of
Electronic Systems and Transactions, furthermore requires, inter alia, that data centres and disaster-recovery
centres for public services have to be located in Indonesia, and that Indonesian citizens are employed in certain
cases. Moreover, under the GATS, “data processing services” is a category of business service distinct from
“communication services”, such as telecommunications. For further information, see WTO, “Services sectoral
classification list”, MTN.GNS/W/120, 10 July 1991. See also WTO, “Communication from the United States”,
S/C/W/339, 20 September 2011: “notwithstanding convergence, the category of computer and related services
remains the fundamental locus for cloud computing under the GATS”, available at http://www.wto.org/english/
thewto_e/minist_e/min11_e/brief_ecom_e.htm (accessed 22 October 2013).


8 Regulation No. 82 of 2012, at article 17(2).


9 South African Electronic Communications and Transactions Act 2002, chapter XI.


10 However, it is worth noting that the “tyranny of distance” is not overcome for all services; for example, content
delivery networks, or applications (for example, real-time financial trading systems), where minimizing latency is
critical.


11 See European Commission Directive 2006/73/EC implementing Directive 204/39/EC “as regards operational
requirements and operating conditions for investment firms and defined terms for the purposes of that Directive”
(OJ L 241/26, 2.9.2006), at article 14(2)(i).


12 For example, the International Consumer Protection and Enforcement Network (icpen.org) and the Global Privacy
Enforcement Network (www.privacyenforcement.net).


13 See, for example, United States Senator Amy Klobuchar’s Cloud Computing Act of 2012 (S. 3569), available at
http://beta.congress.gov/bill/112th/senate-bill/3569/text (accessed 22 October 2013).


14 See, for example, “Patriot Act and privacy laws take a bite out of US cloud business”, Arstechnica, 8 December
2011, available at http://arstechnica.com/tech-policy/2011/12/patriot-act-and-privacy-laws-take-a-bite-out-of-us-
cloud-business/ (accessed 22 October 2013) and “BAe Systems: Office365 doesn’t fly”, Computer Weekly, 5
December 2011, available at http://www.computerweekly.com/blogs/it-fud-blog/2011/12/bae-systems-office365.
html (accessed 22 October 2013).


15 See the statement (in Dutch) by Dutch Minister for Security and Justice, Ivo Opstelten, available at https://zoek.
officielebekendmakingen.nl/ah-tk-20102011-3516.html (accessed 22 October 2013).


16 See “Offshore cloud privacy may be ‘impossible’, says commissioner”, Delimiter, 4 May 2011, available at http://
delimiter.com.au/2011/05/04/offshore-cloud-privacy-may-be-impossible-says-commissioner/ (accessed 22 October
2013), and Office of the Victoria Privacy Commissioner (2011).




79CHAPTER IV : GOVERNANCE, LAW AND REGULATION OF CLOUD SERVICES IN DEVELOPING COUNTRIES


17 See “NSA Prism program taps in to user data of Apple, Google and others”, The Guardian, 7 June 2013, available
at http://www.theguardian.com/world/2013/jun/06/us-tech-giants-nsa-data (accessed 22 October 2013), and
the statement by the President of the United States on 7 June 2013, available at http://www.whitehouse.gov/
the-press-office/2013/06/07/statement-president (accessed 22 October 2013).


18 See, for example, “PRISM scandal: tech giants flatly deny allowing NSA direct access to servers”, The Guardian,
7 June 2013, available at http://www.theguardian.com/world/2013/jun/07/prism-tech-giants-shock-nsa-data-
mining (accessed 22 October 2013).


19 It has been estimated that United States cloud service providers might lose between 10 per cent and 20 per
cent of the non-United States market for cloud provision as a result of recent revelations concerning government
surveillance of data communications (Castro, 2013). See also http://www.cio.com/article/734919/European_and_
US_cloud_providers_go_head_to_head_after_NSA_revelations (accessed 22 October 2013).


20 See “Deutsche Telekom wants ‘German Cloud’ to shield data from US”, Business Week, 14 September 2011.


21 See, for example, “Patriot Act not a cloud computing threat: IDC”, ContinuityCentral, 24 October 2012, available
at http://www.continuitycentral.com/news06514.html (accessed 22 October 2013); “GCHQ taps fibre-optic
cables for secret access to world’s communications”, The Guardian, 21 June 2013, available at http://www.
guardian.co.uk/uk/2013/jun/21/gchq-cables-secret-world-communications-nsa (accessed 22 October 2013);
“En France, la DGSE au cœur d’un programme de surveillance d’Internet”, Le Monde, 11 June 2013, available
at http://www.lemonde.fr/international/article/2013/06/11/en-france-la-dgse-est-au-c-ur-d-un-vaste-programme-
de-surveillance-d-internet_3427837_3210.html (accessed 22 October 2013).


22 For example, European Network and Information Security Agency (2012).


23 For example, Cloud Industry Forum, “Code of practice for cloud service providers”, available at http://www.
cloudindustryforum.org/code-of-practice/code-of-practice (accessed 22 October 2013).


24 See European Commission (2012: 3.4) and IDC (2012).


25 That is, entities that do not themselves own much of the cloud “stack”, but offer users integrated ICT solutions that
will increasingly comprise cloud components.


26 “New cloud laws needed for business data”, TechWeek Europe, 16 August 2012, available at http://www.
techweekeurope.co.uk/comment/cloud-laws-update-eu-89567 (accessed 23 October 2013).See also ITU-D
(2012).


27 For example, Indonesia’s Schedule of Specific Commitments under the GATS: GATS/SC/43, 15 April 1994.


28 For example, United Kingdom, Information Commissioner’s Office, “Guidance on the use of cloud computing”,
September 2012, available at http://www.ico.org.uk/news/latest_news/2012/~/media/documents/library/Data_
Protection/Practical_application/cloud_computing_guidance_for_organisations.ashx (accessed 25 October 2013).


29 Regulations to the Federal Law on the Protection of Personal Data Held by Private Parties (2011), at article 52, (see
further box IV.2).


30 For example, the East African Community Cyberlaws Framework, see UNCTAD (2012b).


31 See http://www.google.com/transparencyreport/ (accessed 23 October 2013) and http://www.microsoft.com/
about/corporatecitizenship/en-us/reporting/transparency/ (accessed 23 October 2013).


32 For example, ISO/IEC 27001 for information security management systems.


33 For example, recommendation ITU-T Y.3510, “Cloud computing infrastructure requirements”, which has recently
reached first-stage approval. ISO/IEC WD 27018 is a proposed “Code of practice for data protection controls for
public cloud computing services” being developed under the ISO 27001 family. See also the work of the Cloud
Security Alliance, which is developing mechanisms such as the CloudTrust protocol (for further details see https://
cloudsecurityalliance.org/research/ctp/, accessed 23 October 2013).


34 See the International Accreditation Forum, available at http://www.iaf.nu/ (accessed 23 October 2013).




80 INFORMATION ECONOMY REPORT 2013


35 European Commission, proposal for a Directive “Concerning measures to ensure a high common level of network
and information security across the Union”, COM(2013) 48 final, 7 February 2013.


36 With the support of the ITU, the region has recently established four new such teams and another four are underway;
see http://www.itu.int/en/ITU-D/Cybersecurity/Pages/Projects.aspx (accessed 23 October 2013).


37 Other identified areas include protecting intellectual property, ensuring data portability and liberalizing trade rules.
See, for example, Business Software Alliance and Galexia (2012).


38 See http://www.uncitral.org/uncitral/uncitral_texts/electronic_commerce.html (accessed 23 October 2013).


39 Ibid.


40 See http://conventions.coe.int/Treaty/Commun/QueVoulezVous.asp?NT=108&CL=ENG (accessed 23  October
2013).


41 In signatories to the Budapest Convention on Cybercrime, domestic law-enforcement agencies can ask foreign
ISPs for disclosure of data, but only on a voluntary basis. See http://conventions.coe.int/Treaty/Commun/
QueVoulezVous.asp?NT=185&CL=ENG (accessed 23 October 2013).




POLICY
RECOMMENDATIONS5


The emergence of the cloud economy has implications for countries at all
levels of development. Whereas a general shift to the cloud can stimulate
major gains in efficiency, productivity and economic growth, these are
not automatic. Where they are realized, they will be unevenly distributed,
geographically, within societies and over time. There are also potential
pitfalls, some of the most important of which are related to data security and
privacy. It is important, therefore, for both individual Governments and the
international community as a whole to consider policy responses that can
help bring about development outcomes from the cloud economy which are
as favourable as possible.


This final chapter builds on the analysis in earlier chapters of the report and
makes recommendations to Governments and their development partners.
Given the early stage of development of the cloud economy in developing
countries, and an evidence base which is still thin, the policy advice remains
at a fairly general level. This chapter is intended to provide a basis for
policymakers in developing countries and their partners as they seek to
enable the development of a cloud economy and translate its potential into
development gains. It identifies specific areas and aspects that should be
considered when designing a strategic response.




82 INFORMATION ECONOMY REPORT 2013


A. RECOGNIZING
THE DIVERSITY OF
CLOUD ADOPTION
STRATEGIES


Cloud computing has the potential to offer users
in developing countries access to unprecedented
resources of computing power and storage. This is a
welcome development which is likely to expand further
in the future. At present, however, cloud adoption in
developing countries is low and most countries face
significant barriers to participating effectively in the
cloud economy.


This report has sought to demystify the cloud
phenomenon and to analyse its implications. While
the metaphor of the “cloud” suggests something
rather new, cloud computing can also be seen as the
latest stage in the long-term evolution of computing
and communications services, enabled by greatly
expanded computing power, storage capacity and
reliable broadband communications networks. As
a consequence, however, the lack of adequate ICT
infrastructure and power supply in many developing
countries holds back their ability to make productive
use of the cloud. Cloud policy needs to be formulated
and implemented in this wider context of ICT markets
and usage.


While the adoption of cloud services in developing
countries is motivated by more or less the same basic
drivers that attract users in developed countries,
there are significant barriers that are more acute in
developing economies. Depending on the country,
these include insufficient broadband connectivity, high
costs for broadband access and use, unreliable power
supply, long distances to data centres (with higher
latency as a result), a lack of skills to make effective
use of ICTs and inadequate legal protection of data.
The level of cloud adoption is greatly affected by
such contextual factors. As a result, notwithstanding
the significant potential benefits that can be derived
from cloud computing, cloud-based solutions are
not always preferable to alternative approaches.
Moreover, different cloud configurations will be more
or less appropriate in different contexts.


The first question for Governments, enterprises (large
and small) and other organizations is whether to
migrate services to the cloud or not. Their response
will be influenced by their assessment of potential
advantages and risks/disadvantages (table V.1)


Potential adopters will be influenced by the nature of
their operations as well as by the national context.
Some businesses and organizations are better
positioned to reap the benefits of a shift to the cloud
or can gain greater advantage than others because
of the nature of their activities or business model.
This is the case, for example, for those that have high
fixed costs in maintaining in-house IT departments,
recurrently need IT software and hardware, face
large or unpredictable variations in demand for
IT resources or can gain substantial added value
from more efficient exploitation of data and market
opportunities. The outcome of an individual business
or organization’s assessment will also depend on
the economic and communications environment in
which it operates.


As a potential cloud service customer, a business or
organization needs to think through a number of issues
if it decides to migrate significant services to the cloud.
The business or organization must explore the range
of cloud solutions available to find that which best
meets its needs. While it is beyond the scope of this
report to make technical recommendations on what
solutions to adopt, businesses and organizations may
consider the following:


Table V.1. Potential advantages and risks/
disadvantages of cloud adoption


Potential advantages Potential risks/disadvantages


Reduced costs for rented
IT hardware and software
compared to costs for in-house
equipment (cost advantages
can be reaped through the
economies of scope and scale
of cloud solutions and from the
shift from capital to operational
expenditure)


Increased costs of
communications
(to telecommunication
operators/ISPs)


Reduced cost of in-house IT
management


Increased costs for data or
service migration and
integration


Enhanced elasticity of storage/
processing capacity as required
by fluctuation in demand


Reduced control over data and
applications


Greater flexibility and mobility
of access to data and services


Data security and privacy
concerns


Immediate and cost-free
upgrading of software


Unreliable services, e.g. due
to inadequate ICT or power
infrastructure


Enhanced reliability/security of
data and services


Risk of vendor lock-in
(limited interoperability and
data portability) with providers
in uncompetitive cloud markets


Source: UNCTAD.




83CHAPTER V : POLICY RECOMMENDATIONS


• Whether to opt for a public, private, community or
hybrid cloud solution


• Whether to opt for a national, regional or global
cloud solution


• What activities and data to migrate (taking data
protection and privacy concerns into account)


• What changes are required within the organization
and what business processes need to be re-
engineered in order to benefit fully from migration


• How the need for IT, procurement and
organizational skills may be affected


• Possible legal implications of choosing different
types of cloud provisioning.


The efficiency-enhancing potential of cloud computing
is a strong incentive for organizations in the private
and public sectors to move increasingly to the cloud.
However, there are important trade-offs to be made,
for example, between cost savings on the one hand
and considerations related to data protection and
privacy on the other. Different cloud customers will
assess the opportunities and risks associated with
the cloud differently, therefore opting for different
solutions. As noted in chapter III, there are visible
differences between cloud adoption strategies even
among large enterprises in the United States and in
continental Europe, respectively, reflecting different
preferences and risk assessments. Similarly, the
options for cloud adoption available to enterprises in
low-income countries look very different from those in
developed countries. The future shape of the cloud
economy at the global level, as well as its regional
and national configurations, will also be influenced by
government policies.


B. RECOMMENDATIONS
TO GOVERNMENTS


There is good reason for Governments to take
a proactive role in the development of the cloud
economy ecosystem. Although there is, as yet, no
statistical analysis of the relationship between the
cloud economy and economic growth, it seems
likely that by generating cost savings, leveraging new
opportunities and participating in learning processes,
enterprises that are early adopters of cloud computing
can gain a competitive advantage.


There is no case for government policy and regulation
to generally discourage migration towards the cloud.
Policies and regulatory approaches should seek to
create an enabling framework that supports firms


and organizations that wish to migrate data and
services to the cloud in doing so easily and safely and
that enables the introduction of new cloud services
at the national level where there is the necessary
infrastructure to support them. At the same time,
cloud-based solutions are not always preferable to
alternative approaches, and there are multiple ways
of using cloud technology. Governments should focus
on those which seem most likely to deliver wider
economic benefits to their countries.


The remainder of this section offers a number of
policy recommendations that can be used as a
foundation for Governments in developing countries
and for development partners that wish to translate
the potential of the cloud into tangible development
gains. These recommendations relate to both the
demand and the supply sides of the cloud economy
ecosystem.


1. Designing strategic policy
approaches to the cloud


Government policies towards the cloud need to
be rooted in a thorough understanding of existing
ICT and cloud use within countries. This requires
research and analysis. Policies should recognize the
diversity of business models and services within the
cloud, the diversity of customers of cloud services
and the complexity of the cloud economy ecosystem
described in this report (chapter I). These various
understandings need to be brought together to tailor
policy approaches to the particular circumstances of
individual countries and economies, consistent with
each country’s overall strategic framework for national
economic development and for the use of ICTs in
development.


(a) Assess the cloud readiness of the country


There is no one-size-fits-all strategy that Governments
can apply to benefit from the cloud. Governments
should therefore start by carefully assessing the
current situation in their countries, to identify how
best to make countries “ready” to leverage the
opportunities offered by the cloud and to address
concerns associated with increased cloud adoption.
This report has pointed to a number of tools that
can be used in this context (chapter  II). Moreover, it
may be useful to establish a national task group with
representatives of different stakeholder communities
(including relevant government ministries/agencies,




84 INFORMATION ECONOMY REPORT 2013


private sector businesses on supply and demand
sides of the cloud economy, civil society and other
stakeholders). The assessment undertaken by this
task group should help to identify bottlenecks and
weaknesses that need to be addressed if the cloud
is to be effectively exploited and clarify what kind of
cloud solutions are most propitious.1


Reliable evidence about a country’s cloud readiness
and economy should be collected as a critical
starting point in assessing policy objectives. In most
developing countries, there is as yet little reliable
quantitative information about the extent of cloud
adoption or about business attitudes towards the
cloud. Empirical evidence, including data from studies
of both perceptions and practice on the demand side
of the cloud, can help policymakers better understand
the needs of different categories of cloud service
customers. A baseline of knowledge derived from
such sources will give both government departments
and local enterprises a better platform on which to
take decisions about what is feasible within specific
contexts.


(b) Develop a national cloud strategy


Based on the assessment described above, a national
cloud strategy could be drafted. This could either be a
stand-alone policy document or an integral part of the
national ICT strategy. In either case it is crucial that the
cloud strategy be consistent with the ICT strategy and
that appropriate synergies be drawn. It should also be
linked to a national broadband strategy where one has
been adopted.


A separate national cloud strategy may be most
appropriate in developing countries which are planning
extensive government use of cloud computing, which
have markets of sufficient size to support local cloud
provision and/or which already have highly developed
sectors engaged in the outsourcing of IT services
and business processes. Elsewhere, it may be better
to integrate the cloud dimension into broader ICT
for development strategies, as in the case of Kenya
(box V.1). Many of the measures required for a positive
enabling environment for cloud computing – such
as improvements in infrastructure and in legal and


Box V.1. Promoting the cloud economy in Kenya


The implementation of cloud services in Kenya mirrors global trends in terms of service and deployment models,
but its success is dependent on local conditions. Presently, most of the cloud services offered are data backup,
archiving and disaster recovery, i.e. mainly IaaS. While Kenyan SMEs are at an incipient stage of cloud adoption,
all large companies listed on the Nairobi Stock Exchange have started migrating some services to the cloud,
including messaging, payroll, accounting, human resources and customer relationship management. For large
companies, the main concerns with regard to cloud adoption are related to data security, privacy and service
reliability. Cloud adoption by SMEs is hampered by the lack of awareness and trust, high cost of broadband
services and limited access to electricity.


The Government of Kenya recognizes the potential benefit of cloud computing for the economy at large and
has designed a strategy to address relevant constraints. The basis for the Government’s activities in this area is
the national ICT policy that was launched in 2006 and which mainstreamed ICT into governance and social and
economic applications.


The cloud economy is promoted through several actions. A key policy priority is to enable a large-scale, nation-
wide expansion of broadband infrastructure. The Government has invested in international submarine optical
fibre projects and in a national fibre optic network. Furthermore, the Communications Commission of Kenya has
promoted a competitive environment to facilitate network expansion while reducing costs and improving the
quality of services. In addition, the elimination of certain taxes on ICT goods and services has led to lower prices
for computers, software and cellular phones, which has helped to raise ICT use in the country.


As noted in chapter III, the Government has also established national data centres aimed at meeting the needs
of the Government itself and to serve as a neutral provider of SaaS to citizens and private sector customers.
To raise awareness among citizens and consumers, some e-government services have been migrated to the
cloud, and the Government has become one of the most significant cloud adopters in Kenya. The next step is
the adoption of data protection legislation, which is to be presented shortly to Parliament.


Source: Research ICT Africa (2013).




85CHAPTER V : POLICY RECOMMENDATIONS


regulatory frameworks, and enhancements to the ICT
skill base – are also relevant to the broader ICT for
development agenda.


A review by the Economic and Social Commission
for Asia and the Pacific of its 58 member economies
in Asia and the Pacific in 2013 concluded that 14
had already adopted a specific cloud computing
strategy or featured cloud computing prominently
in their overall ICT for development strategies.2
This group includes both OECD countries and
LDCs. Moreover, the Broadband Commission has
stated that as of mid-2013 there were 134 national
broadband plans in place, though these were
of varying type and quality, and there was still a
need to assess to what extent they address cloud
computing as such (Broadband Commission for
Digital Development, 2013).3


Whether concerned with ICTs in general, broadband
or the cloud economy in particular, such strategies
should be informed by inputs from a wide variety of
government departments, not just those concerned
with communications. Ministries for trade and
economic development are important in this context,
as are those concerned with education and research.
Ministries of finance need to be involved, because of
the implications for taxation, job creation and economic
growth, while ministries of justice need to be engaged
to develop an adequate legal and regulatory system.
Strategies should also be informed by close interaction
with other stakeholders, for example, through the task
group proposed above.


As mentioned before (chapter III), however, migration
to the cloud can also have less desirable impacts.
Examples may include job displacement within cloud
service customer businesses and reduced demand
for the services offered by local SMEs in the IT sector
when users opt for global cloud providers. These
potential negative aspects of the cloud economy need
to be considered in a strategy alongside potential
benefits.


In terms of scope, a cloud strategy should include
policy measures to address at least the following
main areas: infrastructure, legal and regulatory issues,
the supply side of the cloud economy ecosystem,
human resources, government cloud use and financial
implications. It should address the role of development
partners, set realistic targets and allocate responsibility
for implementation and follow-up with a view to
leveraging macroeconomic gains – such as improved
productivity, job creation, innovation and economic


growth – from the cloud economy. These areas are
discussed below.


2. Addressing the infrastructure
challenge


Fulfilling the potential of the cloud depends on action
to address a range of infrastructure challenges set out
below.


(a) Improve the provision of reliable broadband
infrastructure


Without reliable broadband networks, the cloud
economy can have only a limited impact in a
country. Governments in developing countries are
increasingly facilitating the deployment of broadband
networks, often through public–private partnerships
and/or with finance derived from government
revenues or loans from international financial
institutions. International bandwidth has improved
in most countries, especially through improved
access to submarine cables, and more attention is
now being devoted to national broadband networks
and to regional interconnection (e.g., for landlocked
countries in Africa). Making access to data and
services in the cloud a seamless experience requires
attention to backbone and backhaul infrastructure,
and also to local access networks, both wireline
(where available) and wireless (for the majority of
developing country users in the short to medium
term).


At the same time, cloud services do not all have
the same communications needs. Some require
high traffic volumes but are not time-sensitive, while
others – such as more sophisticated services that
may be used for customer relations management
or in the financial sector – need reliable, high-quality
infrastructure capable of implementing complex
applications in real time. Wireless broadband
access, which is the predominant form available in
many developing countries, does not offer the same
capabilities at present as fixed alternatives. Global
and local public and private clouds are likely to have
different infrastructure needs. Where international
communications are unreliable or very expensive,
this will favour the use of local data centres over
international ones (ITU-D, 2012). However, in
many developing countries, local data centres and
cloud providers are absent. As one entrepreneur
described the situation in Kenya:4




86 INFORMATION ECONOMY REPORT 2013


For now we have to make do with international hosts...
Their products are miles ahead in terms of offerings
and they are at the bleeding edge of innovation, set
on making those offerings even better. Their prices
are pretty reasonable… There’s just one problem.
They are continents away! They do not have data
centres anywhere near here. When I host my service
with them, any new data has to travel across oceans
to get to my user. The undersea fibre-optic cables
suffer cuts every so often and those of us who rely
on their data capacity are routinely left waiting for
days till things return to normal. This is no life for web-
based innovation – which whether or not we like it is
something we will need to make everything else thrive.
The sooner we acknowledge this as an industry, the
better off we will all be. We need to invest in support
structures for the ecosystem we keep talking about.


(b) Pay attention to quality


Cloud applications put higher demands on the quality
of service of broadband networks than many other
forms of ICT use. In addition to download speeds, the
user experience is greatly affected by upload speeds
and latency. Low levels of latency are particularly
important where a service has to be provided in real
time. Network redundancy is also crucial to ensure
that access to data and services is not disabled by
natural disaster, accident, equipment failure or criminal
activity. As noted in chapter II, many developing
countries currently suffer from inadequate quality of
service in their networks, with high latency and poor
upload speeds representing critical bottlenecks.


In order to be able to make informed decisions,
Governments should improve the measurement of quality
of service of broadband networks. Such measurement is
in the interests of all stakeholders. Regulators mandated
to act in the interest of consumers need information
in order to set standards, define policies and resolve
disputes based on evidence. Regulated businesses need
information on their performance, not least in comparison


Box V.2. Approaches to measuring the quality of broadband services


There are advantages and disadvantages to the measurement of quality of service being conducted by
different stakeholders (Wattegama and Kapugama, 2009).


• Independentconsumermeasurements typically lack uniformity in the methodology adopted, making
comparability and representativeness a challenge. A number of factors (e.g. operating systems, browsers
and virus-infected machines) can reduce the accuracy of the results. Nonetheless, consumer-based
initiatives are valuable for holding networks operators accountable.


• Networkoperators usually produce diagnostics for internal quality monitoring purposes. However, they
often consider those network segments with the best connectivity rather than the average performance of
the network.a Diagnostic results may therefore differ from what consumers actually experience.


with other operators, in order to maximize revenue.
Consumers need data to make informed decisions when
selecting service providers and broadband plans, as well
as for lodging complaints (Epitiro, 2011). In this context,
several approaches can be considered to ensure the
most accurate data (box V.2).


(c) Make connectivity more affordable


Connectivity needs to be competitively priced to avoid
having the cost advantages from holding data and
using applications in multi-tenancy data centres be
cancelled out by high communication costs. While
the costs of broadband communications in developed
countries are now relatively low, this is not the case
in many developing countries (chapter  II). In order to
address the affordability challenge, attention needs to
be given to the roles of network operators, Internet
exchange points and related regulation.5


(d) Implement effective communications
regulations


The infrastructure improvements referred to above are
fundamental requirements for the effective use of all
ICTs in developing country contexts, not only cloud
computing. They also need to be accompanied by
competent, effective and independent regulation of
the telecommunications markets that underpin them.
This includes liberalization of markets and regulatory
intervention to promote competition. Competitive and
interoperable ICT markets, facilitated by independent
regulators, should deliver services that are higher
quality, more reliable and affordable. As broadband
usage grows in developing country markets
dependent on wireless networks, policymakers and
regulators also need to make radio spectrum available
to communications operators to ensure that cloud
services can be reliably accessed through the wireless
networks which are currently the norm for local access
in developing country markets.




87CHAPTER V : POLICY RECOMMENDATIONS


Box V.2. Approaches to measuring the quality of broadband services (continued)


• A number of nationalregulatoryauthorities measure broadband performance. In addition to securing
the necessary resources, the challenges for them include the adoption of a sound methodology and ways
of presenting data transparently, in a manner that can be easily understood by the general public.


Several approaches are currently used to measure broadband quality. “SamKnows” (http://www.samknows.
com) is a hardware-based approach that has been adopted, for example, by the United Kingdom, the
European Commission, the United States and Singapore. In this case, a box directly linked to the consumer’s
network utilizes network idle time to run a series of tests. Although it is less intrusive on the network and
offers high levels of accuracy, this method may be too expensive for widespread adoption. Some regulatory
authorities carry out software-based tests, while others set guidelines and request operators to report results
periodically, with possible financial penalties for non-compliance. Examples of software include Ookla, Speed
Test, AT Tester and in-house solutions.


Irrespective of the approach taken, certain principles should be considered.


• Scope of measurement. The approach should be able to test multiple metrics. Download speeds,
upload speeds and latency are particularly relevant for cloud services.


• Timedependentvariance. Tests should be run over a period of time, on multiple days (both weekdays
and weekends) and at multiple time slots each day.


• File size. When testing download and upload speeds, the file should ideally consume the whole bandwidth
to increase the accuracy of measurement.


• User willingness to engage in tests. Desktop applications that run speed tests offer high levels of
accuracy. However, users may see downloading a test application as a potential security threat and the
uptake of such applications may be reduced.


• Network features. Quality of service experience is among other things a function of the physical medium
used in the access network (e.g., copper, fibre) and the distance from the exchange/base transceiver
station. In the case of wireless networks, the number of users of a base transceiver station also varies.
Therefore, a country’s broadband performance derived from results received in only one location can be
misleading as an indicator of quality at the country level. For transparency reasons, it may be better to
report results for a city or locality rather than the whole country.


In South Asia, a testing methodology designed by LIRNEasia and the Indian Institute of Technology (Madras)
has been applied to measure broadband quality of service. The software measures quality along various
dimensions (download and upload speeds, RTT, jitter, packet loss and availability) at six time slots a day, on
multiple days per week including weekends (Galpaya and Zuhyle, 2011). It can carry out the tests in different
domains – within the ISP network, nationally (within the country but through a server located on the network
of a different ISP) and internationally. This distinction can help detect problematic segments of the network.


At a meeting in 2012, the ITU Expert Group on Telecommunication/ICT Indicatorsb agreed on two indicators
on fixed broadband quality of service (service activation time and complaints per 100 fixed (wired)-broadband
subscriptions) and one on mobile broadband QoS (complaints per 100 mobile broadband subscriptions).
These three indicators will be included in the statistical data collection of the ITU. The indicator on data
transmission speed achieved should be collected at the national level (upload and download speed), for both
fixed and mobile broadband. Other indicators, such as on jitter, throughput, packet loss and latency, should
be collected at the national level using a representative sample and a verified methodology, as that developed
by LIRNEasia.


Source: UNCTAD.


a The access network is only a part of the entire end-to-end connection. While easy to measure, good QoS
in the access network does not necessarily imply good QoS experience for the end user.


b The ITU Expert Group on Telecommunication/ICT Indicators is open to all ITU members and experts in
the field of ICT statistics and data collection (see http://www.itu.int/en/ITU-D/Statistics/Pages/definitions/
default.aspx; accessed 14 October 2013).




88 INFORMATION ECONOMY REPORT 2013


(e) Promote the establishment of Internet
exchange points


IXPs should also be established to reduce the cost
of Internet access and bring down latency. There
are different kinds of IXPs (chapter II). Rather than
prescribing the model to apply, public policy may focus
on supporting an environment of fair competition and
establishing a licensing structure that can enable IXPs
to succeed (Ryan and Gerson, 2012). A key challenge
is to ensure effective cooperation with the ISPs in
each market. As was noted in chapter II, there is a
particular need for more IXPs in Africa. The African
Union has obtained financial support to help remedy
this situation (box V.3).


(f) Ensure access to reliable power


The power infrastructure is another critical bottleneck
to cloud adoption in many countries. Regular power
outages which disrupt communications between
cloud customers and providers – as well as between
different cloud service providers – are a serious
problem. Power outages are common in developing
countries, and smaller countries are often dependent
on their neighbours for a substantial part of their power
supply. For countries aspiring to host international
data centres, which consume considerable amounts
of electricity, access to a reliable power supply is even
more essential (see below).


3. Strengthening the legal framework


A national cloud strategy needs to address a
number of legal and regulatory issues related to
cloud adoption, ensuring that the interests of cloud
service users are properly protected. Concerns over
data protection and security are among the most
frequently mentioned. Most developing countries still
lack much of the legislation required to address such
concerns. While there is not necessarily a need to


Box V.3. The African Internet Exchange System Project


The African Union Commission signed an agreement in 2012 with the Luxembourg Development Cooperation to support
the implementation of the African Internet Exchange System project funded by the European Union–Africa Infrastructure
Trust Fund and the Government of Luxembourg. This project seeks to secure the establishment of IXPs in Member
States of the African Union. The project also seeks to make real-time and historical traffic data accessible via web-based
visualization system and to develop a certificate curriculum on Internet exchange technologies. A first steering committee
meeting was held in May 2012. Capacity-building and technical assistance to support establishment of regional Internet
exchange points was to begin in 2013.


Source: African Union. Available at http://pages.au.int/axis/about (accessed 14 October 2013).


develop laws or regulations which are specific to cloud
computing, legal reforms are clearly required in areas
including privacy, data protection, information security
and cybercrime. It is essential that Governments of
developing countries adopt and enforce appropriate
laws and regulations in these areas. Putting such
legislation in place will give local businesses, including
local entrepreneurs, new opportunities to innovate
in services and applications, seeking to address
export as well as domestic markets. While public law
is essential to secure the basic rights of end users,
private law contractual agreements between cloud
service providers and cloud service customers also
directly impact on the operation of the cloud economy.
Governments may wish to consider the following
recommendations in this context.


(a) Assess the adequacy of current legislation


An important starting point is to review current
legislation to see if it adequately addresses relevant
areas for cloud adoption and to begin to consider
ways of further improving the legal framework. This is
one area in which international organizations, including
UNCTAD, can offer assistance.


(b) Clarify situation concerning the location of
data


Many Governments and other cloud service customers
have serious concerns about data security. These are
most pronounced when data are kept on servers
located in other jurisdictions. In particular, there are
concerns that holding national data overseas may
make them more accessible to third parties (including
the Governments of other countries). At the same
time, it is important that regulations do not restrict the
ability to take advantage of cost savings where there
are not, in practice, real security threats. Assessing
this balance is challenging, and there are no simple
solutions. As part of their response, Governments




89CHAPTER V : POLICY RECOMMENDATIONS


may consider the value of data centres situated locally,
including the potential of these centres to interact
with private-sector cloud providers in hybrid (public/
private, local/global) clouds. Such ideas are being
explored in both developed and developing countries.
In Europe, for example, there have been calls from the
highest political level for the development of a secure
European cloud,6 and the Government of France
has launched a sovereign cloud project to offer an
alternative source of cloud provision to French and
other European enterprises.7


(c) Facilitate electronic transactions and
payments


The business model based on “pay as you go” that
is often applied to monetize many cloud services
requires a legal and regulatory framework which
accommodates electronic transactions and payments.
The ability to make secure micropayments is an
essential part of many services offered over the cloud,
including those that could provide new business
opportunities for SMEs in developing countries.


(d) Address the cybersecurity challenge


Growing use of the cloud is driving the demand for
adequate security solutions. The concentration of data
in a few locations represents an attractive target for
cybercriminals. In addition, the interaction of different
laws and regulatory schemes in various jurisdictions
increases the complexity of data security in the cloud.
All relevant stakeholders need to take action to limit
the vulnerabilities of their systems. Governments need
to design cybercrime and data protection laws and
regulations that take cloud vulnerabilities adequately
into consideration (chapter IV). Cloud providers need
to implement robust security enforcement tools. End
users also have a responsibility to be prudent when
making use of cloud services (and the Internet more
generally).8 Countries should also continue moving
towards the establishment of CERTs to handle security
incidents and risks and to ensure that the responsible
authorities are notified of security breaches.


(e) Take international legal frameworks into
account


Aligning relevant national laws with international legal
instruments and models is highly recommended to
achieve an internationally harmonized set of laws
and regulations (chapter IV). Governments should
monitor regulatory developments in the field of cloud
computing to be able to adjust as appropriate to


changes in markets and technologies. In view of the
limited resources available to many Governments
of developing countries, there is a potential role
here for regional economic communities to advise
Governments, possibly with financial support from
international development partners.


4. Map opportunities on the supply side


A national cloud strategy should not only consider
the implications for potential cloud customers in the
public and private sectors. Three areas of supply-side
opportunity in particular were highlighted in chapter III:
the development of national data centres and related
cloud provision, the potential for cloud aggregation
and the scope for new cloud services to be developed
for both domestic and export markets.


(a) Explore the market for national data centres
and cloud provision


The market for cloud provision, at a global level, is
currently dominated by a small number of very large
corporations which are predominantly headquartered
in the United States (chapter II). Moreover, it has
been estimated that as much as 85 per cent of data
centres offering co-location services are in developed
economies (chapter II). For many Governments
and businesses in developing countries, this raises
questions about the security and control of their data
and of critical applications. It also raises questions
concerning the dependence of developing countries
on developed countries to manage data resources that
are critical to their social and economic development,
something which can be considered a new dimension
of the digital divide.


Global cloud providers, notably in the United States,
have the advantages of being first movers and having
achieved economies of scale. Whereas these same
advantages can represent formidable entry barriers
for new competitors, there are opportunities for local
IT companies in the cloud economy in developing
countries, either independently or in collaboration with
international cloud service providers. Some of these
arise from a desire of cloud customers to exercise
greater control over their data and applications.
As noted above, the high cost of international
communications may also favour local data centre
provision.


Several developing countries are actively promoting
the development of national data centres, either by




90 INFORMATION ECONOMY REPORT 2013


companies in the ICT sector or by the Government
itself (chapter III). Data centres which offer commercial
storage and application may be oriented purely towards
domestic markets and/or towards export markets.
National data centre markets are most likely to arise
in developing countries with large domestic markets
such as Brazil, China, India, Nigeria and South Africa.
Governments that wish to foster the development of
data centre markets oriented towards exports need
to pay particular attention to the adequacy of power
and communications infrastructure, cybersecurity
arrangements and the cost of international bandwidth.


(b) Consider the inclusion of cloud-related
activities in foreign direct investment
promotion strategies


In countries that offer the right conditions for cloud
services, it may be appropriate to include cloud-related
activities as a new focus in national strategies for the
attraction of foreign direct investment. Several developed
countries are actively promoting themselves as strategic
locations for the establishment of international data
centres. Some developing countries may offer attractive
conditions in this area. Foreign cloud providers may
also be interested in supporting the development of
domestic capacity to store and process data.


(c) Explore the market for the aggregation and
brokerage of cloud services


Cloud service customers often need to rely on
several different cloud providers to meet their varying
needs. A market has emerged for cloud aggregators
packaging services from different providers on
behalf of customers and brokers providing advice
and guidance on the best cloud options for different
businesses. This activity offers opportunities for
IT companies in developing countries with local
expertise. Policymakers can engage in dialogue with
the local industry on the main areas of opportunity as
well as current bottlenecks.


(d) Support the development of cloud
applications for local needs


A growing number of IT companies are developing
and deploying cloud services targeted at national or
regional communities or at niche groups within them.
In low-income countries at a nascent stage of cloud
readiness, IaaS is likely to be the first category of
cloud services to be demanded. As the infrastructure


situation improves in low-income countries and in
countries with an expanding SME market with more
purchasing power, the market for SaaS will become
more important and eventually dominant as it is
today in developed countries. Cloud services cover a
wide range of activities, from the distribution of retail
goods to providing access to major content libraries.
Opportunities for local offerings may be particularly
relevant in countries where international connectivity
is expensive and latency is high.


5. The human resource challenge


Many of the underlying factors that influence propensity
to benefit from the cloud economy apply not just to
cloud services but to the ICT sector as a whole. This is
clearly the case for the national skill base, particularly
for the ICT sector itself.


The relationship between education/human resource
development, enterprise and the ICT sector was
discussed in the Information Economy Report 2010
(UNCTAD, 2010). Many developing countries have
struggled to incorporate IT skills development within
their school and education systems, not least because
of the shortage of IT-skilled teachers. Meanwhile,
without adequate skills on both the supply and user
side of the cloud economy, it is difficult to achieve
proper integration of new cloud applications. It is
therefore important for Governments, together with
the private sector and academia, to consider how
skill shortages within the ICT sector and elsewhere
may affect their capacity to take advantage of cloud
computing, and what priority interventions they should
make in order to address these shortfalls.


A national strategy could map skill requirements.
Likely areas to consider include:
• IT and software skills, for managing the migration


and integration
• Management and organizational skills, for


addressing the need for reorganization and re-
engineering of business processes


• Legal and procurement skills, to ensure that
contracts with providers or cloud services respond
well to the needs of the user.


Government and the private sector should work
together to address these human resource challenges,
from the revision of curricula at different stages in
the education system, to in-work training and the
development of specialized professional skills. Among
the many challenges involved, developing countries face




91CHAPTER V : POLICY RECOMMENDATIONS


significant loss of skilled IT personnel through migration
to more lucrative job markets in developed countries.
A continual flow of IT-skilled recruits into employment
is therefore required. The most successful countries in
leveraging the potential of the cloud in the longer term
may be those that are able to build up a significant
group of skilled IT entrepreneurs with the capability of
developing innovative cloud-based businesses.


6. Government use of the cloud


One way in which government action can affect the
development of the cloud economy is through its
own cloud use for administration and for the provision
of public services. Governments are among the
most important – in smaller developing countries
often the largest – purchasers of IT equipment and
services. Computer equipment and software licences
represent significant expenditure for Governments,
and the possible cost savings are important drivers
for government cloud adoption. Government policies
vis-à-vis cloud adoption are hence important.


(a) Consider the role of the Government in the
establishment of national data centres


Governments in developing countries should consider
investing directly in data centre capacity, taking into
account the experience of other countries.


(b) Improve e-government services through
government cloud use


There are many instances in which individual
government services are being provided through, or
supported by, the cloud. In education, cloud services
can make available larger libraries of content. They
can also support the delivery of mass-market services
that rely on personal data, such as health, tax and
benefit/welfare systems. Similarly, processes such as
business registration, customs administration and the
payment of taxes and licence fees can be facilitated
through online provision or through the cloud. Some
Governments have adopted a “cloud first” approach
(e.g., the Republic of Moldova and the United States),
requiring government agencies to consider cloud
provisioning before more conventional alternatives.


Although some cloud-related initiatives form part
of comprehensive government strategies, in many
countries, decisions to introduce particular cloud
services are being taken on an ad hoc basis by individual
government departments. One consequence is that


cloud provisioning can become unsystematic, risking
limited interoperability between systems that deal with
the same citizens and losing some of the economies of
scope and scale that might otherwise have been realized.


(c) Use public procurement as a tool to support
local IT sector development


There is a strong case for more coordination in government
procurement of cloud services (as indeed there is for
other IT services).9 In this context, Governments can seek
to ensure that their procurement practices are conducive
to greater involvement of domestic suppliers, especially
SMEs. This involves attention to transparency, openness
and clarity of specifications. There are also various ways
in which the structuring of the procurement process can
be adapted to fit the skills and capabilities of software
SMEs. At the same time, local IT firms will only be able to
compete effectively if they have the necessary skills and
capabilities. Thus, it is important to connect the public
procurement strategy with efforts aimed at strengthening
the capabilities of the local IT industry (UNCTAD and
Germany, Federal Ministry for Economic Cooperation
and Development, 2013; UNCTAD, 2012a).


7. Implementation and follow-up of the
strategy


Given the cross-cutting and multipurpose nature of
ICTs and cloud computing, with consequences for both
public service delivery and business competitiveness,
it is important to integrate any cloud strategy into
the overall national development plan. Moreover,
a detailed approach should be crafted for strategy
execution, monitoring and evaluation.10 To allow for
effective follow-up, realistic and measurable goals
in each area should be set for use when assessing
performance. The choice of indicators should reflect
discussions with relevant stakeholders to ensure that
targets are shared across society.


8. Raising awareness


Once a strategy has been formulated in collaboration with
relevant stakeholders, it is important to raise awareness
concerning its content. There is still much confusion and
uncertainty about the meaning and implications of the
cloud economy. This extends to government officials
and business leaders and affects important issues in
determining the potential for cloud computing.


Efforts to demystify the cloud, in order to build awareness
of its potential and limitations for businesses and




92 INFORMATION ECONOMY REPORT 2013


national economies, should be made by Governments
and business associations alike. Regulators can assist
by issuing guidance on the applicability to cloud
solutions of the regimes for which they are responsible.
Governments may need to build capacity in various
departments to ensure enactment and enforcement
of laws and regulations and to promote awareness in
government institutions and public administrations of
their implications. Awareness-raising activities should
seek to help potential users (including SMEs) better
understand the potential gains from using cloud-
based solutions as well as the potential costs and risks
associated with such migration. There may, for example,
be a need to educate users on available approaches to
protect data that are transmitted to and stored in the
cloud (e.g., through encryption techniques), contracting
with cloud providers and government policies vis-à-vis
the cloud.


C. RECOMMENDATIONS
TO DEVELOPMENT
PARTNERS


The differential between developed and developing
countries in access to and use of ICTs – the digital
divide – has been a significant concern of Governments
and international development agencies since “The
missing link” (ITU, 1985), known as the Maitland
Report, was published by the ITU almost 30 years
ago. Since that time, the nature of the digital divide
has changed. The gap in access to basic telephone
services, which was once very substantial, is now
greatly diminished and expected almost to disappear
in the next few years. In its place has come a gap
in access to the Internet and particularly in access to
broadband services, i.e. to services which have the
capacity to deliver a full Internet experience and enable
access to the full range of data communications
opportunities which can now be made available. A
digital divide in broadband capacity and quality leads
in turn to a divide between countries in the extent to
which they and businesses located in them are able to
take advantage of ICT innovations.


Many developing countries face considerable
challenges in seeking to benefit fully from the evolving
cloud economy. Addressing these challenges will
require both expertise in various fields and substantial
financial resources. In order to reduce the risk that
the move towards the cloud economy at the global
level results in a further widening of the digital divide,


development partners should ensure that cloud-
related development challenges are included on their
agendas, particularly where low-income countries
and those with limited IT-management capacity are
concerned. Development partners may explore the
need for bilateral support in specific areas in different
countries. In addition, there are several horizontal
issues for which support from bilateral and multilateral
donors will be needed.


At a country level, assistance may be relevant in some
or all of the following areas:


• Empirical analysis and research is needed
to map the current state of the cloud economy,
assess potential implications of greater cloud
adoption and make policy recommendations. As
the evidence base expands, it will be possible to
make better assessments of the implications of the
cloud economy for economic growth, employment,
productivity and trade. Special attention may also
be given to the link between cloud computing and
climate change, taking into account the intensive
use of energy by huge data centres.


• Financing of broadband infrastructure,
including public–private partnerships. Different
infrastructure layers need to be addressed
separately from a policy and financing perspective,
rather than using a one-size-fits-all approach
(Broadband Commission for Digital Development,
2012). Adequate attention should also be given
to power infrastructure, whose deficiencies
undermine the capabilities of communications
networks.


• Supportfortheestablishmentofappropriate
legal and regulatory frameworks. Many
developing countries still lack appropriate
legislation and ancillary institutions for electronic
commerce, data protection and cybersecurity.
Development partners can support Governments
of developing country in transposing existing
legislative models into national legal and regulatory
frameworks.


• Capacity-building activities. Development
partners can provide support through technical
training in the deployment, procurement and
integration of cloud services as well as by advising
on law reforms and undertaking policy reviews.11


International agencies could facilitate this assistance
through some of their existing activities. For
example, the ITU, UNCTAD and United Nations




93CHAPTER V : POLICY RECOMMENDATIONS


regional commissions could facilitate an exchange of
experiences with regard to the policy challenges that
developing countries face in deriving benefits from
the cloud economy and avoiding pitfalls. This could
be based on evidence-gathering from Governments,
cloud providers and cloud customers in developing
countries, with the aim of building a nuanced analytical
framework to help policymakers in government
and business address the challenges and seize the
opportunities of the cloud economy as it develops.


Another area of significance for developing countries
concerns international standards for cloud services,
which are essential to facilitate interoperability and to
help customers understand what they are purchasing.
The number of interfaces presently used by cloud
service providers can make it difficult for customers
to move from one system to another. The various
standard-setting activities that are currently undertaken
in this area (Sakai, 2011) are overwhelmingly in
and by developed countries, whose interests in
cloud services may differ from those of developing
countries. Standardization forums should consider
how to engage developing countries and their users
in standard-setting activities so as to ensure that their
specific needs and requirements are addressed.


This report has summarized the experience to date
with the cloud economy in developing countries and
raised a number of issues which need to be addressed
by Governments and businesses. The cloud economy
is, however, a recent phenomenon, and the evidence
base available for analyses of the kind in this report is
still limited. It is important to build a stronger evidence
base for decision-makers that relies more on empirical
evidence and independent analysis than on marketing
and other advocacy literature. More research is
needed in a number of areas, including:


• The extent of cloud adoption in developing
countries. In addition to national reviews required
for national policymaking purposes, developing
countries would benefit from a systematic and
comprehensive overview of the extent and manner
in which cloud services are being adopted in the
developing world as a whole and perceptions
of the cloud by developing country businesses.
National case studies in countries with diverse
experience would also be valuable.


• The macroeconomic implications of cloud
computing. Currently, the evidence base does
not allow for empirical assessments of the cloud’s
impact on various macroeconomic variables.


It is important to deepen the analysis to better
understand the implications for countries at
different levels of development and for the global
economy as a whole.


• Implications for data security and privacy.
Research into perceptions of cloud security and
data protection in determining attitudes towards
the cloud would be valuable in understanding how
related concerns affect cloud adoption. Research
might include assessment of the potential for
moving towards a common global approach
to data protection, rather than the diverse
approaches currently found in different jurisdictions
(see chapter IV).


• Competition in the cloud provider market.
Research into the extent to which cloud markets
are currently competitive and the potential impact
on competitiveness, and viability, of new cloud
businesses based in developing countries would
be valuable.


• Theimpactofthecloudontaxation. Many cloud
services, including mass market services, depend
on cross-border connectivity and transactions.
Some displace markets for physical goods or
services traded in more conventional ways in
either domestic or export markets. The impact of
this changing business model on taxation could be
significant. Research into the cloud’s impact on tax
revenue and into appropriate means of avoiding
double taxation when cloud services are supplied
across borders would be valuable.


• Theimpactofthecloudontradeinservices,
particularlyIT-enabledservices. Research into
the impact of the cloud on trade in services would
be valuable to Governments concerned about
maximizing export potential and would indicate
where domestic sectors may be vulnerable to
service competition from developed countries.
Research into the impact of the cloud on IT-enabled
services, such as BPO, and into its impact on
intraregional trade would be particularly valuable.


• The scope for South–South collaboration.
The ICT sector has seen significant South–South
collaboration, including public–private partnerships
in infrastructure investment and the growth in
South-based TNCs providing communications
services. Research would be valuable into the
experience and particularly the potential of South–
South collaboration in the cloud, especially at the
regional level.




94 INFORMATION ECONOMY REPORT 2013


As with other ICT areas, the pace of change
in cloud technology and markets is rapid. The
experiences described in this report relate to present
circumstances. The nature of cloud services and of the
cloud economy will continue to develop quickly and
may be very different in five years’ time. Governments,


businesses and development partners need to bear
these changes in mind and re-evaluate their policies
and strategies concerning the cloud regularly in order
to ensure that they continue to maximize potential
benefits and minimize potential risks to their citizens,
businesses and customers.


NOTES
1 In India, for example, the Government has established an Empowered Committee with representatives from different


government departments, and a special task force with private sector participation, to develop a strategy towards
the development of a government cloud (box III.6).


2 Australia, Bangladesh, Bhutan, China, India, Japan, Kazakhstan, Malaysia, New Zealand, the Philippines, the
Republic of Korea, Singapore, Sri Lanka and Thailand (Economic and Social Commission for Asia and the Pacific,
2013).


3 The Broadband Commission was established in May 2010 to boost the importance of broadband on the international
policy agenda. It is co-chaired by President Paul Kagame of Rwanda and Mr. Carlos Slim Hélu, Honorary Lifetime
Chair of Grupo Carso. Mr. Hamadoun Touré, Secretary-General of ITU, and Ms. Irina Bokova, Director-General
of the United Nations Educational, Scientific and Cultural Organization (UNESCO), serve as joint Vice-Chairs. See
http://www.broadbandcommission.org.


4 Hapa Kenya (2013). Why Kenyan telcos need to start thinking like web companies. Available at http://www.
hapakenya.com/why-kenyan-telcos-need-to-start-thinking-like-web-companies/ (accessed 10 October 2013).


5 See, for example, http://broadbandtoolkit.org/en/home and http://www.broadbandcommission.org/work/
documents.aspx.


6 Charlemagne (2013). Reaching for the clouds. The Economist. 20 July.


7 See, for example, Reuters (2013), Analysis: European cloud computing firms see silver lining in PRISM
scandal, 17 June. Available at http://www.reuters.com/article/2013/06/17/us-cloud-europe-spying-analysis-
idUSBRE95G0FK20130617 (accessed 11 October 2013).


8 For links to more information on the adoption of regionally and internationally harmonized, appropriate legislation
against the misuse of ICTs for criminal or other purposes, see, for example, http://www.itu.int/en/ITU-D/
Cybersecurity/Pages/Legal-Measures.aspx.


9 The Info-Communications Development Authority of Singapore facilitates public sector procurement of public cloud
services through several “bulk tenders” (Singapore, Info-Communications Development Authority, 2013). In India,
the roadmap to the GI Cloud highlights the importance of procurement and recognizes the importance of adjusting
current practices (India, Department of Electronics and Information Technology, 2013; p. 36).


10 A useful tool for governments to consider in this context are the IT sector manual and toolbox developed by the
Federal Ministry for Economic Cooperation and Development of Germany (Germany, Federal Ministry for Economic
Cooperation and Development, 2011).


11 An assessment of cloud computing in Africa identified training needs in several areas, including understanding
cloud computing, technical considerations, data centres, broadband connectivity and QoS (ITU-D, 2012).




95REFERENCES


REFERENCES


Aberdeen Group (2008). Application performance management: The lifecycle approach brings IT and busi-
ness together. Aberdeen Group. June. Available at http://www.riverbed.com/docs/WhitePaper-Riverbed-
ApplicationPerformanceManagement_LifeCycle.pdf (accessed 11 October 2013).


Adam L, Souter D, Jagun A and Tusubira FF (2011). Transformation-ready: The strategic application of informa-
tion and communication technologies in Africa. ict Development Associates Ltd.. Report prepared for the
African Development Bank, the World Bank and the African Union. Available at http://siteresources.worldbank.
org/EXTINFORMATIONANDCOMMUNICATIONANDTECHNOLOGIES/Resources/282822-1346223280837/
RegionalTradeandIntegration_Fullreport.pdf (accessed 9 October 2013).


Armbrust M, Fox A, Griffith R, Joseph AD, Katz RH, Konwinski A, Lee G, Patterson DA, Rabkin A, Stoica I and Zaharia
M (2009). Above the clouds: A Berkeley view of cloud computing. No. UCB/EECS-2009-28. Electrical Engineering
and Computer Sciences University of California at Berkeley. Berkeley. Available at http://www.softwareresearch.
net/fileadmin/src/docs/teaching/SS09/VS/Above_the_Clouds.pdf (accessed 1 October 2013).


Asia Cloud Computing Association (2012). Cloud Readiness Index 2012. Available at http://www.asiacloud.org/
index.php/2012-07-17-08-34-39/2012-11-12-10-08-55/index-2012 (accessed 11 October 2013).


Berry R and Reisman M (2012). Policy challenges of cross-border cloud computing. Journal of International
Commerce and Economics. May.


Bradshaw S, Millard C and Walden I (2011). Contracts for clouds: A comparative analysis of terms and condi-
tions for cloud computing services. International Journal of Law and Information Technology. 19(3): 187–223.


Bigo D, Boulet G, Bowden C, Carrera S, Jeandesboz J and Scherrer A (2012). Fighting cyber crime and pro-
tecting privacy in the cloud. Directorate General for Internal Policies. Policy Department C: Citizens’ Rights
and Constitutional Affairs. Study commissioned by the European Parliament’s Committee on Civil Liberties,
Justice and Home Affairs. Available at http://www.europarl.europa.eu/studies (accessed 25 October 2013).


Broadband Commission for Digital Development (2011). Broadband: A Platform for Progress. ITU and UNESCO.
Geneva and Paris. Available at http://www.broadbandcommission.org/Reports/Report_2.pdf (accessed
11 October 2013).


Broadband Commission for Digital Development (2012). The State of Broadband 2012: Achieving Digital
Inclusion for All. ITU and UNESCO. Geneva and Paris.


Broadband Commission for Digital Development (2013). Planning for Progress: Why National Broadband Plans
Matter. ITU and UNESCO. Geneva and Paris.


Business Software Alliance and Galexia (2012). BSA Global Cloud Computing Scorecard: A Blueprint for
Economic Opportunity. Washington D.C.. Available at http://cloudscorecard.bsa.org/2012/assets/PDFs/
BSA_GlobalCloudScorecard.pdf (accessed 16 October 2013).


Cairncross F (1997). The Death of Distance: How the Communications Revolution Will Change Our Lives.
Harvard Business Review Press. Boston.


Capgemini (2012). Trends in cloud computing: Secure journey to the cloud – a matter of control. Available at
http://www.capgemini.com/resources/trends-in-cloud-computing-secure-journey-to-the-cloud--a-matter-
of-control (accessed 2 October 2013).


Castro D (2013). How much will PRISM cost the U.S. cloud computing industry? The Information Technology and
Innovation Foundation. Washington D.C..


CIO (2012). Strategy guide: Data sovereignty and security. Available at http://www.cio.com.au/
whitepaper/371067/strategy-guide-data-sovereignty-and-security/download/?type=other&arg=0&location=
tag_detail_page (accessed 21 October 2013).


Cisco Analysis (2012). Cisco Global Cloud Index: Forecast and Methodology, 2011–2016. Cisco. Available at
http://www.cisco.com/en/US/solutions/collateral/ns341/ns525/ns537/ns705/ns1175/Cloud_Index_White_
Paper.pdf (accessed 2 August 2013).




96 INFORMATION ECONOMY REPORT 2013


Council of Europe (1981). Convention for the Protection of Individuals with regard to Automatic Processing of
Personal Data. http://www.conventions.coe.int/Treaty/en/Treaties/Html/108.htm (accessed 21 October
2013).


Cowhey P and Kleeman M (2012). Unlocking the benefits of cloud computing for emerging economies – a policy
overview. Univesity of California, San Diego.


Crémer J, Gassot Y, Lanvin B and Pupillo LM, eds. (2012). Introduction. Communications & Strategies. 85
(1st Quarter): 13–21.


Dynamic Markets Ltd. (2013). Cloud for business managers: the good, the bad and the ugly. Commissioned by
Oracle. Dynamic Markets Ltd.. Abergavenny, United Kingdom. Available at https://emeapressoffice.oracle.
com/imagelibrary/downloadmedia.ashx?MediaDetailsID=2905&SizeId=-1 (accessed 10 October 2013).


Economic and Social Commission for Asia and the Pacific (2013). A review of cloud adoption by Asia–Pacific
Governments. United Nations. Unpublished.


Epitiro (2011). Regulatory challenges for measuring national broadband performance. Epitiro. Cardiff, United
Kingdom. Available at http://www.epitiro.com/assets/files/20-102-1008.001%20Challengers%20for%20
Regulators.pdf (accessed 7 November 2013).


Ernst and Young (2010). Cloud adoption in India. Available at http://twelvedot.com/blog/wp-content/up-
loads/2011/07/Cloud_computing_adoption_in_India.pdf (accessed 2 October 2013).


European Commission (2012). Unleashing the potential of cloud computing in Europe. No. COM(2012) 529. Euro-
pean Commission. Luxembourg. Available at http://eur-lex.europa.eu/smartapi/cgi/sga_doc?smartapi!celexpl
us!prod!DocNumber&lg=EN&type_doc=COMfinal&an_doc=2012&nu_doc=529 (accessed 22 October 2013).


European Network and Information Security Agency (2012). Procure Secure: A guide to Monitoring of Security
Service Levels in Cloud Contracts. ENISA. Heraklion, Greece. Available at http://www.enisa.europa.eu/activi-
ties/Resilience-and-CIIP/cloud-computing/procure-secure-a-guide-to-monitoring-of-security-service-levels-
in-cloud-contracts (accessed 22 October 2013).


European Telecommunications Network Operators’ Association (2011). ETNO reflection document replying to
the public consultation on cloud computing. ETNO. Brussels. Available at http://www.etno.be/home/posi-
tions-papers/2011/32 (accessed 7 November 2013).


Galpaya H and Zuhyle S (2011). South Asian broadband service quality: Diagnosing the bottlenecks. Social Sci-
ence Research Network, No. 1979244. Rochester, United States. April.


Gentzoglanis A (2012). Evolving cloud ecosystems: Risk, competition and regulation. Communications & Strate-
gies. 1(85): 87–108.


Germany, Federal Ministry for Economic Cooperation and Development (2011). IT Sector Promotion in Develop-
ing and Emerging Countries: Manual. Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH.
Bonn and Eschborn.


Gonsalves TA and Bharadwaj A (2009). Comparison of AT-tester with other popular testers for quality of service
experience (QoSE) of an Internet connection. TeNet Group, Department of Computer Science and Engineer-
ing, Indian Institute of Technology Madras.


Greenleaf G (2013). Global tables of data privacy laws and bills (third edition). University of New South Wales Law
Research Paper No. 2013-39. Available at http://ssrn.com/abstract=2280875 (accessed 21 October 2013).


Hamilton P (2011). Measuring backbone transmission networks. Document C/10-E. Paper presented at the ninth
World Telecommunication/ICT Indicators Meeting, Mauritius, 7 December. Available at http://www.itu.int/
ITU-D/ict/wtim11/documents/cont/010-E.pdf 8 (accessed 16 October 2013).


Hilbert M and López P (2011). The world’s technological capacity to store, communicate and compute informa-
tion. Science. 332: 60–65.


HIPCAR (2012). Electronic transactions: Assessment report. ITU. Geneva.


Hon WK, Millard C and Walden I (2012). Negotiating cloud contracts: Looking at clouds from both sides now.
Stanford Technology Law Review. 16(1): 79–128.




97REFERENCES


IDATE Foundation (2012). Digiworld Yearbook 2012: The Challenges of the Digital World. Idate. Montpellier,
France.


IDC (2012). Quantitative estimates of the demand for cloud computing in Europe and the likely barriers to
up-take. No. 2011/0045. D4 final report. Available at http://ec.europa.eu/information_society/activities/
cloudcomputing/docs/quantitative_estimates.pdf (accessed 22 October 2013).


India, Department of Electronics and Information Technology, Ministry of Communications and Information
Technology (2013). GI cloud (Meghraj) adoption and implementation roadmap. New Delhi.


Information Warfare Monitor and Shadowserver Foundation (2010). Shadows in the cloud: Investigating cyber
espionage 2.0. Available at http://www.scribd.com/doc/29435784/SHADOWS-IN-THE-CLOUD-Investigating-
Cyber-Espionage-2-0 (accessed 5 November 2013).


Intel (2011). Over 6 decades of continued transistor shrinkage, innovation. Available at http://download.intel.
com/newsroom/kits/22nm/pdfs/Intel_Transistor_Backgrounder.pdf (accessed 5 November 2013).


ITU (1985). The missing link: Report of the Independent Commission for World Wide Telecommunications Devel-
opment. Telecommunication Journal. 52(2):67–71.


ITU (2012a). Focus Group on Cloud Computing technical report part 1: Introduction to the cloud ecosystem:
definitions, taxonomies, use cases and high-level requirements. ITU. Geneva.


ITU (2012b). GSR12 Best Practice Guidelines on regulatory approaches to foster access to digital opportuni-
ties through cloud services. ITU. Geneva. Available at https://www.itu.int/ITU-D/treg/Events/Seminars/GSR/
GSR12/consultation/GSR12_BestPractices_v3_E.pdf (accessed 24 October 2013).


ITU-D (2012). Cloud computing in Africa: Situation and perspectives. ITU Regulatory and Market Environment
Division. Geneva. April. Available at http://www.itu.int/ITU-D/treg/publications/Cloud_Computing_Afrique-e.
pdf (accessed 23 October 2013).


Kang HY, Lee HH, An HS and Yang HD (2011). Cloud computing: Its implication on industries and strategic direc-
tion. No. 11-08. Korea Information Society Development Institute (DISDI). Gwacheon.


Kituku KM (2012). Adoption of cloud computing in Kenya by firms listed in the Nairobi Stock Exchange. MBA
thesis. School of Business, University of Nairobi. Available at http://erepository.uonbi.ac.ke:8080/xmlui/
handle/123456789/13578 (accessed 9 October 2013).


KPMG (2012). Exploring the cloud: A global study of government’s adoption of cloud. Available at http://
www.kpmg.com/BS/en/IssuesAndInsights/ArticlesPublications/Documents/exploring-cloud.pdf (accessed
4 October 2013).


Kshetri N (2010). Cloud computing in developing economies. IEEE Computer. 43(10): 47–55.


Kuner C (2013). Transborder Data Flow Regulation and Data Privacy Law. ISBN 9780199674619. Oxford Uni-
versity Press.


Kushida KE, Murray J and Zysman J (2012). The gathering storm: Analysing the cloud computing ecosystem and
implications for public policy. Communications & Strategies. 1(85)(first quarter): 63–85.


LIRNEasia (2011). Broadband quality of services experience (QOSE) indicators. No. Q3 2011. LIRNEasia.
Colombo. Available at http://lirneasia.net/wp-content/uploads/2010/10/Broadband-QoSE-report-OCT-2011_
V5.pdf (accessed 16 October 2013).


Manyika J, Chui M, Bughin J, Dobbs R, Bisson P and Marrs S (2013). Disruptive technologies: Advances that will
transform life, business and the global economy. McKinsey Global Institute. Available at http://www.mckinsey.
com/insights/business_technology/disruptive_technologies (accessed 5 November 2013).


Microsoft (2010). The economics of the cloud. Available at http://www.microsoft.com/en-us/news/presskits/
cloud/docs/the-economics-of-the-cloud.pdf (accessed 1 October 2013).


Microsoft (2012). Drivers & inhibitors to cloud adoption for small and midsize businesses. Available at http://www.
microsoft.com/en-us/news/presskits/telecom/docs/SMBCloud.pdf (accessed 2 October 2013).


National Institute of Standards and Technology (2011). The NIST definition of cloud computing. Available at
http://csrc.nist.gov/publications/nistpubs/800-145/SP800-145.pdf (accessed 1 October 2013).




98 INFORMATION ECONOMY REPORT 2013


OECD (1980). OECD Guidelines on the Protection of Privacy and Transborder Flows of Personal Data. OECD
Publishing. Paris. Available at http://www.oecd.org/internet/ieconomy/oecdguidelinesontheprotection-
ofprivacyandtransborderflowsofpersonaldata.htm (accessed 21 October 2013).


Pierre Audoin Consultants (2013). PAC webinar: Cloud services – six key opportunities for software and IT
services providers. 21 March.


Poletti F, Wheeler NV, Petrovich MN, Baddela N, Numkam Fokoua E, Hayes JR, Gray DR, Li Z, Slavic R, Rich-
ardson DJ (2013). Towards high-capacity fibre-optic communications at the speed of light in vacuum. Nature
Photonics. 7(4): 279–284.


Pyramid Research (2012). Enterprise cloud readiness index excerpt. Available at http://www.pyramidresearch.
com/enterprise-cloud-readiness-index-excerpt.htm?sc=PRN121012_PPCRI (accessed 16 October 2013).


Ramos D (2012). How Latin American Telcos Are Tackling the SME Cloud Opportunity. Pyramid Research, Inc.


Renda A (2012). Competition, neutrality and diversity in the cloud. Communications & Strategies. 85 (1st quarter):
23–44.


Research ICT Africa (2013). Cloud computing country reports prepared for the UNCTAD Information Economy
Report 2013. Unpublished.


Ryan PS and Gerson J (2012). A primer on Internet exchange points for policymakers and non-engineers. Social
Science Research Network, No. 2128103. 12 August.


Sakai H (2011). Standardization activities for cloud computing. NTT Technical Review. 9(6): 1–6.


Seidman DR (1986). Transborder data flow: Regulation of international information flow and the Brazilian
example. Journal of Law and Technology. 31 (Spring).


Singapore, Info-Communications Development Authority (2013). Cloud computing in Singapore: Empowering
the next generation businesses. Singapore.


Source8, hurleypalmerflatt and Cushman & Wakefield (2013). Data Centre Risk Index: Informing global invest-
ment decisions. Available at http://www.business-sweden.se/PageFiles/9118/DCRI%202013.pdf (accessed
7 October 2013).


Stork C, Calandro E and Gamage R (2013). Future of broad band in Africa. 13 May. Available at: http://www.
researchictafrica.net/presentations/Presentations/2013_Stork_Calandro_Gamage_-_The_Future_of_Broad-
band_in_Africa.pdf (accessed 16 October 2013).


Tech Soup Global (2012). 2012 global cloud computing survey results. Available at http://www.techsoupglobal.
org/2012-global-cloud-computing-survey (accessed 4 October 2013).


TeleGeography (2013). Global Bandwidth Research Service: Executive summary. Available at http://www.
telegeography.com/page_attachments/products/website/research-services/global-bandwidth-research-
service/0003/8368/gb13-exec-sum.pdf (accessed 16 October 2013).


Tweneboah-Koduah S (2013). Unleashing the potential of cloud computing: what is it and what does it mean for
public organizations in Ghana. Paper presented at the GhanaTechnology University Centre for Information
and Media of University of Aalborg conference. Accra. 16–17 May.


UNCTAD (2004). E-Commerce and Development Report 2004. United Nations publication. UNCTAD/SDTE/
ECB/20. New York and Geneva.


UNCTAD (2009a). Estudio sobre las Perspectivas de la Armonización de la Ciberlegislación en Centroamérica y
el Caribe. United Nations publication. UNCTAD/DTL/STICT/2009/3. New York and Geneva.


UNCTAD (2009b). Study on Prospects for Harmonizing Cyberlegislation in Latin America. United Nations publica-
tion. UNCTAD/DTL/STICT/2009/1. New York and Geneva.


UNCTAD (2010). Information Economy Report 2010: ICTs, Enterprises and Poverty Alleviation. United Nations
Publication. UNCTAD/IER/2010. Sales No. E.10.II.D.17. New York and Geneva.


UNCTAD (2011a). ICT Policy Review Egypt 2011. United Nations publication. UNCTAD/DTL/STICT/2011/6. New
York and Geneva.




99REFERENCES


UNCTAD (2011b). Information Economy Report 2011: ICTs as an enabler to private sector development. United
Nations publication. UNCTAD/IER/2011. Sales No. E.11.II.D.6. New York and Geneva.


UNCTAD (2012a). Information Economy Report 2012: The Software Industry and Developing Countries. United
Nations publication. Sales No. E.12.IID.14. UNCTAD/IER/2012. New York and Geneva.


UNCTAD (2012b). Harmonizing Cyberlaws and Regulations: The Experience of the East African Community.
United Nations publication. UNCTAD/DTL/STICT/2012/4. New York and Geneva.


UNCTAD (2013). Review of E-commerce Legislation Harmonization in the Association of Southeast Asian
Nations. United Nations publication. UNCTAD/DTL/STICT/2013/1. New York and Geneva.


UNCTAD and Germany, Federal Ministry for Economic Cooperation and Development (2013). Promoting local IT
sector development through public procurement. United Nations publication. UNCTAD/DTL/STICT/2012/5.
New York and Geneva. Available at http://unctad.org/en/PublicationsLibrary/dtlstict2012d5_en.pdf
(accessed 21 October 2013).


United Kingdom, Cabinet Office (2011). Government ICT offshoring (international sourcing) guidance. Available
at https://www.gov.uk/government/publications/government-ict-offshoring-international-sourcing-guidance
(accessed 21 October 2013).


United States, Federal CIO Council and Chief Acquisition Officers Council (2012). Creating Effective Cloud
Computing Contracts for the Federal Government: Best Practices for Acquiring IT as a Service. February.
Available at https://cio.gov/wp-content/uploads/downloads/2012/09/cloudbestpractices.pdf (accessed
22 October 2013).


Walden I (2011). Accessing data in the cloud: The long arm of the law enforcement agent. Queen Mary School
of Law Legal Studies Research Paper No. 74/2011. Queen Mary College, University of London. London.
November.


Walden I and Savage N (1990). Transborder data flows. In: Walden I, ed., Information Technology and the Law.
Macmillan. London: 121.


Wattegama C and Kapugama N (2009). Prospects of volunteer computing model in performance data gathering
for broadband policy formulation: A case study from South Asia. Available at http://lirneasia.net/wp-content/
uploads/2009/10/Broadband-Quality-of-Service-Experience_LIRNEasia.pdf (accessed 14 October 2013).


Yeboah-Boateng E and Cudjoe-Seshie S (2013). Cloud computing: The emerging of application service provid-
ers (ASP) in developing economies. International Journal of Emerging Technology and Advanced Engineering.
3(5): 703–712.






STATISTICAL
ANNEX


Annex table 1. Enterprise Cloud Services Readiness Index, 2012 ................................................... 102


Annex table 2. 2013 BSA Global Cloud Computing Scorecard ....................................................... 103


Annex table 3. Asia Cloud Computing Association Readiness Index, 2012 ..................................... 104


Annex table 4. Cisco global cloud readiness, 2012 ......................................................................... 105


Annex table 5. Key cloud service infrastructure indicators ............................................................... 106




102 INFORMATION ECONOMY REPORT 2013


Annex table 1. Enterprise Cloud Services Readiness Index, 2012


Americas Europe Asia-Pacific Africa and Middle East


2. United States 5.54 3. Germany 5.33 1. Japan 5.74 9. United Arab Emirates 4.4


5. Canada 5.03 6. France 4.87 4. Republic of Korea 5.17 16. Saudi Arabia 3.71


13. Argentina 4.11 7. United Kingdom 4.64 8. Taiwan Province of
China


4.56 19. Turkey 3.63


18. Mexico 3.64 11. Spain 4.14 10. China 4.24 21. Israel 3.47


20. Brazil 3.61 12. Italy 4.13 14. Hong Kong, China 3.94 30. South Africa 2.71


24. Colombia 3.31 17. Romania 3.7 15. Philippines 3.88 34. Morocco 2.37


27. Puerto Rico 2.85 19. Turkey 3.63 22. Malaysia 3.47 36. Egypt 2.07


28. Venezuela (Bolivar-
ian Republic of)


2.75 23. Russian Federation 3.45 26. India 3.01 42. Nigeria 1.65


29. Costa Rica 2.74 25. Poland 3.02 33. Viet Nam 2.43


31. Chile 2.61 39. Czech Republic 1.91 35. Singapore 2.31


32. Panama 2.47 38. Thailand 1.95


37. El Salvador 2.02 40. Indonesia 1.87


41. Guatemala 1.75


43. Peru 1.59


44. Ecuador 1.46


45. Uruguay 1.38


46. Bolivia (Plurinational
State of)


0.94


47. Honduras 0.76


48. Paraguay 0.67


49. Nicaragua 0.44


Source: Pyramid Research.




103STATISTICAL ANNEX
Ra


nk


Ec
on


om
y


Ov
er


al
l s


co
re


Da
ta


p
riv


ac
y


Se
cu


rit
y


Cy
be


rc
rim


e


In
te


lle
ct


ua
l p


ro
pe


rt
y


Su
pp


or
t f


or
in


du
st


ry
-le


d
st


an
da


rd
s


an
d


in
te


rn
at


io
na


l h
ar


m
on


iz
at


io
n


of
ru


le
s


Pr
om


ot
in


g
fr


ee
tr


ad
e


IC
T


re
ad


in
es


s
an


d
br


oa
db


an
d


de
pl


oy
m


en
t


1 Japan 83.3 8.8 8.4 10 17.2 8.8 9.2 20.9


2 Australia 79.2 7.9 6 9.4 17.6 10 7 21.3


3 Germany 79 6.6 6.4 10 16.8 9.8 9.2 20.2


4 United States 78.6 6.5 7.6 8.8 16.6 9.4 8 21.7


5 France 78.4 6.5 7.6 10 16.4 9.6 8.8 19.5


6 Italy 76.6 6.2 7.6 9.6 17.4 9.8 8.8 17.2


7 United Kingdom 76.6 6.9 8 6.8 17.4 9.2 6.8 21.5


8 Republic of Korea 76 9.3 6 4.8 17.6 9.6 7 21.7


9 Spain 73.9 6.5 6.4 8.8 15.2 9.8 9.4 17.8


10 Singapore 72.2 3.2 3.6 9 17.2 8.8 8.6 21.8


11 Poland 70.7 6.4 5.6 8.8 16.8 9.8 8.4 14.9


12 Canada 70.4 8.1 6.8 6.2 10.8 10 9.6 18.9


13 Malaysia 59.2 7.1 5.6 5.4 11.4 10 3.8 15.9


14 Mexico 56.4 7.5 4.8 8.6 12.4 9.2 3 10.9


15 Argentina 55.1 5 6 8.2 12.4 4.6 5.8 13.1


16 Russian Federation 52.3 5.4 6.4 6.8 8.4 6.6 5.2 13.5


17 Turkey 52.1 3.5 4 6.4 14 8.6 2.8 12.8


18 South Africa 50.4 2.8 3.2 9.8 13.6 9.8 1.8 9.4


19 India 50 4.1 4.4 7.4 9.2 10 6.4 8.5


20 Indonesia 49.7 4.6 3.2 7 11.2 8.2 5.2 10.3


21 China 47.5 3.5 2 4.6 13.6 7.8 4.8 11.2


22 Thailand 42.6 3.5 1.6 7.4 7.2 8.8 3 11.1


23 Viet Nam 39.5 4.1 2.8 5 9.2 7 1.4 10


24 Brazil 35.1 4.7 3.6 1.6 7.2 3.4 2.2 12.4


Annex table 2. 2013 BSA Global Cloud Computing Scorecard


Source: Business Software Alliance and Galexia, 2012.




104 INFORMATION ECONOMY REPORT 2013


Ec
on


om
y


Da
ta


p
riv


ac
y


In
te


rn
at


io
na


l c
on


ne
ct


iv
ity


Da
ta


s
ov


er
ei


gn
ty


Br
oa


db
an


d
qu


al
ity


Go
ve


rn
m


en
t o


nl
in


e
se


rv
ic


es
a


nd


IC
T


pr
io


rit
iz


at
io


n


Po
w


er
g


rid
a


nd
g


re
en


p
ol


ic
y


In
te


lle
ct


ua
l p


ro
pe


rt
y


pr
ot


ec
tio


n


Bu
si


ne
ss


s
op


hi
st


ic
at


io
n


Da
ta


c
en


tr
e


ris
k


Fr
ee


do
m


o
f i


nf
or


m
at


io
n


a
cc


es
s


Cl
ou


d
Re


ad
in


es
s


In
de


x


Ra
nk


Japan 9.0 10.0 5.6 7.6 7.9 7.8 7.6 8.4 6.0 8.9 78.8 1


Republic of Korea 9.0 8.0 6.2 9.0 9.1 7.1 5.9 6.9 7.4 7.7 76.3 2


Hong Kong, China 7.5 7.4 7.6 7.6 8.4 5.7 7.9 7.1 8.0 8.7 75.9 3


Singapore 4.5 9.2 8.1 6.3 9.5 5.7 8.7 7.3 6.4 7.1 72.8 4


Taiwan Province of
China


7.0 7.5 5.9 6.1 8.8 7.1 7.1 7.5 6.5 8.9 72.4 5


New Zealand 9.0 1.3 8.1 5.4 7.8 8.3 8.3 6.6 7.1 8.9 70.8 6


Australia 7.5 2.7 7.3 6.0 8.2 7.5 7.6 6.7 5.6 8.6 67.7 7


Malaysia 7.5 4.6 5.6 3.7 8.2 6.2 7.0 7.1 6.2 6.9 63.0 8


India 6.0 8.4 4.7 2.4 6.3 3.3 5.0 6.1 3.1 7.6 52.7 9


China 4.0 5.0 3.5 3.5 6.6 4.5 5.7 6.2 5.1 7.1 51.2 10


Indonesia 6.0 4.8 2.1 2.2 5.7 4.9 5.1 6.0 3.1 7.2 47.1 11


Philippines 2.5 4.6 4.3 2.3 5.5 5.8 4.0 5.9 3.6 7.5 46.0 12


Thailand 3.0 2.8 1.5 5.9 5.5 4.8 4.4 6.0 3.6 7.4 44.9 13


Viet Nam 5.0 3.2 3.9 2.2 5.9 3.8 3.6 5.3 5.4 6.6 44.9 13


Annex table 3. Asia Cloud Computing Association Readiness Index, 2012


Source: Asia Cloud Computing Association.




105STATISTICAL ANNEX


Top 10 fixed network performing economies Top 10 mobile network performing economies


Do
w


nl
oa


d
(k


bp
s)


Up
lo


ad
(k


bp
s)


La
te


nc
y


(m
s)


Do
w


nl
oa


d
(k


bp
s)


Up
lo


ad
(k


bp
s)


La
te


nc
y


(m
s)


Bulgaria 18 973 12 256 35 Austria 3 671 1 864 107


Hong Kong, China 27 710 22 570 32 Canada 5 824 2 980 128


Japan 20 335 17 326 39 Denmark 3 445 1 316 121


Republic of Korea 23 222 22 682 38 Finland 3 439 1 791 136


Latvia 19 240 14 146 48 Hungary 3 607 1 278 99


Lithuania 26 810 21 308 39 Poland 3 080 1 207 135


Netherlands 22 495 5 822 29 Portugal 2 875 1 561 105


Romania 22 937 11 834 39 Romania 3 133 1 178 115


Singapore 19 399 12 255 38 Sweden 3 377 1 446 110


Sweden 20 835 9 657 53 United Arab Emirates 3 133 1 353 120


Annex table 4. Cisco global cloud readiness, 2012


Source: http://www.cisco.com/en/US/solutions/collateral/ns341/ns525/ns537/ns705/ns1175/CloudIndex_Supplement.html (accessed
10 October 2013).


Note: Countries are listed in alphabetical order.




106 INFORMATION ECONOMY REPORT 2013


Ec
on


om
y


Fi
xe


d
do


w
nl


oa
d


(k
bp


s)


Fi
xe


d
up


lo
ad


(k
bp


s)


Fi
xe


d
la


te
nc


y
(m


s)


M
ob


ile
d


ow
nl


oa
d


(k
bp


s)


M
ob


ile
u


pl
oa


d
(k


bp
s)


M
ob


ile
la


te
nc


y
(m


s)


Co
-lo


ca
tio


n
da


ta
c


en
tr


es
, 2


01
3


IX
Ps


, 2
01


3


In
te


rn
at


io
na


l I
nt


er
ne


t b
an


dw
id


th
,


20
11


(b
ps


p
er


In
te


rn
et


u
se


r)


Se
cu


re
In


te
rn


et
s


er
ve


rs
, 2


01
2


(p
er


1
m


ill
io


n
pe


op
le


)


Afghanistan 626 413 561 1.0


Albania 2 568 1 051 102 2 843 1 015 197 1 19 038 18.3


Algeria 683 306 144 695 340 253 8 933 1.4


Angola 1 044 498 317 1 3.6


Antigua and Barbuda 2 007 1 462 193 56 545 1 071.2


Argentina 2 959 757 74 2 649 763 205 6 10 25 712 42.3


Armenia 3 059 2 358 88 4 056 2 531 158 1 26.4


Aruba 2 972 554 133 396.0


Australia 7 876 1 261 54 2 894 950 161 73 14 50 396 1 777.5


Austria 9 103 2 141 56 3 671 1 864 107 10 3 81 919 1 147.0


Azerbaijan 2 196 732 96 2 480 1 053 171 1 19 102 6.3


Bahrain 2 229 803 114 1 870 736 132 1 1 14 719 137.6


Bangladesh 739 501 153 642 542 346 2 0.8


Barbados 2 506 756 157 38 177 397.0


Belarus 2 468 1 259 85 2 794 1 187 184 2 1 52 833 18.4


Belgium 16 045 2 024 39 8 484 1 730 94 28 1 131 137 697.3


Belize 594 451 315 286.8


Bermuda 3 083 1 979 128 5 030.1


Bolivia (Plurinational State of) 592 302 641 1 4 162 10.0


Bosnia and Herzegovina 2 879 587 69 3 377 732 106 1 17 767 27.5


Brazil 4 891 975 67 3 187 725 220 22 23 29 041 55.5


Brunei Darussalam 1 699 646 136 1 451 678 211 21 995 116.3


Bulgaria 18 973 12 256 35 8 115 5 869 105 17 2 65 832 167.3


Cambodia 2 194 1 804 78 1 840 1 289 181 3 1 13 530 3.0


Canada 10 215 1 564 59 2 972 1 293 308 86 7 70 150 1 277.2


Cayman Islands 2 187 557 276 2 603.8


Chile 6 080 1 448 66 3 563 999 243 4 1 20 414 82.3


China 5 075 2 960 87 1 499 709 655 17 4 2 692 3.1


Hong Kong 27 710 22 570 32 6 751 5 335 112 26 2 964 616 643.2


Taiwan Province 17 106 2 901 42 2 213 327 180 4 3 ..


Colombia 2 631 994 105 2 021 791 218 2 1 16 796 28.0


Costa Rica 1 934 644 145 684 310 134 2 36 216 98.7


Côte d'Ivoire 871 611 415 1 18 044 1.5


Croatia 4 228 707 62 3 038 1 063 145 2 1 19 948 244.8


Cyprus 3 871 733 68 7 1 53 569 969.7


Czech Republic 11 788 5 836 35 6 420 3 499 88 16 3 91 064 510.6


Annex table 5. Key cloud service infrastructure indicators




107STATISTICAL ANNEX


Annex table 5. Key cloud service infrastructure indicators (continued)
Ec


on
om


y


Fi
xe


d
do


w
nl


oa
d


(k
bp


s)


Fi
xe


d
up


lo
ad


(k
bp


s)


Fi
xe


d
la


te
nc


y
(m


s)


M
ob


ile
d


ow
nl


oa
d


(k
bp


s)


M
ob


ile
u


pl
oa


d
(k


bp
s)


M
ob


ile
la


te
nc


y
(m


s)


Co
-lo


ca
tio


n
da


ta
c


en
tr


es
, 2


01
3


IX
Ps


, 2
01


3


In
te


rn
at


io
na


l I
nt


er
ne


t b
an


dw
id


th
,


20
11


(b
ps


p
er


In
te


rn
et


u
se


r)


Se
cu


re
In


te
rn


et
s


er
ve


rs
, 2


01
2


(p
er


1
m


ill
io


n
pe


op
le


)


Denmark 14 268 7 218 44 3 288 1 492 88 28 2 159 511 2 243.0


Dominican Republic 1 561 634 111 1 308 833 198 1 11 205 23.9


Ecuador 2 053 1 903 109 1 834 1 045 286 2 27 742 23.8


Egypt 1 029 330 132 9 2 6 754 3.7


El Salvador 2 128 1 174 113 2 036 774 270 4 176 20.9


Estonia 9 282 3 716 54 4 423 2 187 134 7 3 24 378 653.8


Finland 10 848 3 435 60 3 508 1 453 149 11 4 118 445 1 621.6


France 9 214 1 974 66 2 196 1 050 133 123 19 78 590 424.9


Georgia 4 076 2 784 69 4 891 3 718 136 15 796 27.2


Germany 13 680 2 536 54 2 847 1 144 150 144 14 74 786 1 102.4


Ghana 2 618 984 122 1 592 749 234 1 225 3.1


Greece 5 438 677 69 4 781 810 110 9 1 26 008 169.6


Guatemala 1 916 814 130 1 553 780 287 14.5


Haiti 1 623 918 246 1 1.7


Honduras 2 210 1 351 94 1 679 804 195 4 866 9.5


Hungary 12 325 4 031 43 6 659 2 641 92 8 1 12 245 252.0


Iceland 17 611 11 404 32 7 399 5 373 90 3 1 287 139 3 064.0


India 1 468 907 121 1 246 748 323 69 7 5 423 3.5


Indonesia 1 009 440 139 839 450 336 22 8 7 196 4.0


Iran (Islamic Republic of) 1 031 504 138 206 191 673 5 3 540 1.3


Iraq 1 877 1 464 174 2 046 1 600 190 0.1


Ireland 5 693 1 409 79 15 3 69 031 1 056.9


Israel 8 460 1 060 52 4 032 830 135 5 1 11 335 429.9


Italy 4 393 765 81 37 7 60 820 210.7


Jamaica 2 748 706 122 3 355 975 265 23 077 51.4


Japan 20 335 17 326 39 6 141 3 352 96 36 16 23 111 774.3


Jordan 2 172 660 131 2 338 656 235 3 6 337 29.7


Kazakhstan 4 049 2 736 101 3 331 2 068 224 1 23 590 7.8


Kenya 2 856 1 740 114 2 354 1 118 311 2 2 4 544 4.2


Kuwait 2 467 1 081 123 1 854 761 177 1 202.3


Latvia 19 240 14 146 48 7 487 5 991 110 18 1 44 779 245.2


Lebanon 920 230 263 1 300 344 166 1 1 2 257 51.3


Lithuania 26 810 21 308 39 7 691 6 052 111 6 2 57 571 262.4


Luxembourg 16 706 8 295 66 6 561 2 552 84 12 2 89 564 2 050.0


Malaysia 2 864 2 324 82 26 1 10 651 66.1


Maldives 1 453 807 379 1 30 659 104.8




108 INFORMATION ECONOMY REPORT 2013


Ec
on


om
y


Fi
xe


d
do


w
nl


oa
d


(k
bp


s)


Fi
xe


d
up


lo
ad


(k
bp


s)


Fi
xe


d
la


te
nc


y
(m


s)


M
ob


ile
d


ow
nl


oa
d


(k
bp


s)


M
ob


ile
u


pl
oa


d
(k


bp
s)


M
ob


ile
la


te
nc


y
(m


s)


Co
-lo


ca
tio


n
da


ta
c


en
tr


es
, 2


01
3


IX
Ps


, 2
01


3


In
te


rn
at


io
na


l I
nt


er
ne


t b
an


dw
id


th
,


20
11


(b
ps


p
er


In
te


rn
et


u
se


r)


Se
cu


re
In


te
rn


et
s


er
ve


rs
, 2


01
2


(p
er


1
m


ill
io


n
pe


op
le


)


Malta 12 811 1 345 35 6 853 1 153 104 7 1 47 850 1 662.6


Mauritius 1 186 243 97 1 060 278 309 7 1 12 714 132.4


Mexico 3 688 1 204 90 9 8 743 29.7


Moldova 13 567 9 659 56 6 573 5 330 127 1 91 118 27.6


Mongolia 3 263 2 043 92 1 53 576 17.6


Montenegro 3 534 701 75 2 608 608 170 28.4


Morocco 2 139 398 133 1 898 444 201 2 7 558 3.8


Mozambique 797 436 334 1 1 244 1.4


Myanmar 798 715 334 8 180 0.1


Namibia 1 277 445 334 2 349 19.0


Nepal 919 738 127 985 673 293 1 1 1 531 2.2


Netherlands 22 495 5 822 29 2 209 638 215 71 5 162 532 2 880.0


New Caledonia 3 418 1 072 177 208.7


New Zealand 7 024 1 348 65 5 537 1 282 108 10 5 23 706 1 505.2


Nicaragua 1 996 866 99 1 12 857 10.2


Nigeria 1 299 1 005 239 1 106 763 428 1 1 368 1.8


Norway 15 771 7 665 52 6 874 3 846 140 13 7 151 257 1 924.8


Oman 2 674 626 98 11 648 60.6


Pakistan 1 245 511 131 9 1 4 752 1.3


Panama 2 442 968 158 1 637 614 324 3 1 44 121 144.8


Paraguay 1 276 806 184 1 616 687 292 1 9 482 11.4


Peru 1 174 343 615 1 9 319 22.1


Philippines 1 361 470 102 844 237 346 2 12 360 8.8


Poland 8 006 2 297 57 27 5 40 244 309.7


Portugal 17 267 3 047 43 8 379 2 578 87 22 1 135 332 238.1


Puerto Rico 2 685 733 77 1 1 113.0


Qatar 3 001 1 430 85 3 22 333 149.1


Republic of Korea 23 222 22 682 38 8 065 7 852 106 2 4 17 170 2 733.4


Romania 22 937 11 834 39 8 235 4 365 96 32 3 114 451 67.2


Russian Federation 9 172 9 161 62 35 16 31 911 38.6


Samoa 1 522 406 400 27.1


Saudi Arabia 4 030 755 92 2 696 897 159 10 1 32 985 29.9


Senegal 1 216 319 395 2 909 2.2


Serbia 4 167 861 59 3 648 1 013 127 2 1 76 761 27.2


Singapore 19 399 12 255 38 2 132 289 153 14 3 547 064 651.6


Slovakia 10 356 4 298 46 5 088 2 300 113 13 3 12 276 221.0


Annex table 5. Key cloud service infrastructure indicators (continued)




109STATISTICAL ANNEX


Ec
on


om
y


Fi
xe


d
do


w
nl


oa
d


(k
bp


s)


Fi
xe


d
up


lo
ad


(k
bp


s)


Fi
xe


d
la


te
nc


y
(m


s)


M
ob


ile
d


ow
nl


oa
d


(k
bp


s)


M
ob


ile
u


pl
oa


d
(k


bp
s)


M
ob


ile
la


te
nc


y
(m


s)


Co
-lo


ca
tio


n
da


ta
c


en
tr


es
, 2


01
3


IX
Ps


, 2
01


3


In
te


rn
at


io
na


l I
nt


er
ne


t b
an


dw
id


th
,


20
11


(b
ps


p
er


In
te


rn
et


u
se


r)


Se
cu


re
In


te
rn


et
s


er
ve


rs
, 2


01
2


(p
er


1
m


ill
io


n
pe


op
le


)


Slovenia 7 199 4 198 50 5 975 3 301 85 7 1 68 250 567.1


Solomon Islands 6 361 1 996 156 3 893 8.8


South Africa 2 045 764 95 2 069 865 151 17 5 18 874 84.6


Spain 10 203 1 642 69 42 5 64 069 295.9


Sri Lanka 2 193 513 125 1 839 597 278 1 5 224 7.3


Sudan 821 368 633 0.0


Suriname 738 312 183 43.1


Sweden 20 835 9 657 53 2 462 890 150 29 12 244 440 1 535.1


Switzerland 16 864 3 532 41 49 3 167 636 2 379.0


Syrian Arab Republic 725 275 215 3 489 0.3


Tajikistan 1 619 615 221 1.0


Thailand 5 200 965 71 1 626 542 165 8 1 10 622 19.1


The former Yugoslav
Republic of Macedonia


5 609 2 139 60 4 699 1 859 114 1 1 17 945 41.1


Trinidad and Tobago 6 184 1 039 109 4 305 1 000 303 19 753 97.7


Tunisia 1 433 626 147 1 565 628 249 1 14 832 13.0


Turkey 4 508 1 121 59 27 1 33 938 125.8


Turkmenistan 472 202 478 1 567 0.2


Annex table 5. Key cloud service infrastructure indicators (continued)




110 INFORMATION ECONOMY REPORT 2013


Ec
on


om
y


Fi
xe


d
do


w
nl


oa
d


(k
bp


s)


Fi
xe


d
up


lo
ad


(k
bp


s)


Fi
xe


d
la


te
nc


y
(m


s)


M
ob


ile
d


ow
nl


oa
d


(k
bp


s)


M
ob


ile
u


pl
oa


d
(k


bp
s)


M
ob


ile
la


te
nc


y
(m


s)


Co
-lo


ca
tio


n
da


ta
c


en
tr


es
, 2


01
3


IX
Ps


, 2
01


3


In
te


rn
at


io
na


l I
nt


er
ne


t b
an


dw
id


th
,


20
11


(b
ps


p
er


In
te


rn
et


u
se


r)


Se
cu


re
In


te
rn


et
s


er
ve


rs
, 2


01
2


(p
er


1
m


ill
io


n
pe


op
le


)


Uganda 912 422 586 1 1 752 1.5


Ukraine 11 300 8 710 70 6 875 5 334 165 21 8 9 835 23.9


United Arab Emirates 8 002 2 952 51 4 1 27 609 206.8


United Kingdom 11 460 2 173 54 2 480 948 335 191 9 166 073 1 534.1


United Republic of Tanzania 837 519 623 2 902 0.8


United States 10 332 2 956 64 2 366 923 241 1 144 86 47 174 1 501.0


Uruguay 2 392 530 75 2 083 540 158 3 32 078 84.3


Uzbekistan 1 428 659 256 579 0.8


Venezuela (Bolivarian Republic of) 1 153 386 139 836 207 369 8 108 10.9


Viet Nam 5 851 4 133 70 2 009 593 251 5 3 9 998 6.5


Yemen 939 410 348 1 082 0.4


Zambia 786 463 397 1 452 2.4


Zimbabwe 981 502 355 1 1 748 3.2


Sources and notes: Download and upload speeds and latency: Data are sourced from Cisco (based on analysis dated 2012; see
http://www.cisco.com/en/US/solutions/collateral/ns341/ns525/ns537/ns705/ns1175/CloudIndex_Supplement.
html (accessed 10 October 2013). The performance data are gathered from network tests and represent the av-
erage for the economy. Cisco breaks down the performance data by fixed and mobile networks, and business
and consumer users. The fixed network data refer to overall consumer and business usage. The mobile network
data generally refer to business usage that had the largest data availability. If this data was not available, overall
consumer and business usage is used. If there are insufficient test results, data for the country are not included.
Co-location data centres: Data are sourced from the database of Data Centre Map (see http://www.datacentermap.com/
datacenters.html (accessed 10 October 2013)). They were extracted in July 2013. Data refer to the number of co-location
data centres in the country. A co-location data centre is a facility that provides space for customers to place their data servers.
IXPs: Data sourced from Packet Clearing House (see https://prefix.pch.net/applications/ixpdir/summary/ (accessed
10 October 2013)). They were extracted in July 2013. Data refer to the number of IXPs in the economy. An IXP is a facility
ISPs exchange Internet traffic. International Internet bandwidth (bps per Internet user): Data are sourced from ITU. See
http://www.itu.int/ITU-D/ict/publications/idi/ (accessed 10 October 2013). They refer to 2011. International Internet
bandwidth refers to the total used capacity of international Internet bandwidth. It is measured as the sum of used capac-
ity of all Internet exchanges (locations where Internet traffic is exchanged) offering international bandwidth. If capacity
is asymmetric (i.e. more incoming (downlink) than outgoing (uplink) capacity), then the incoming (downlink) capacity is used.
Secure Internet servers (per 1 million people): Data are sourced from the World Bank (see http://data.worldbank.
org/indicator/IT.NET.SECR.P6 (accessed 10 October 2013). They refer to 2012. Secure servers are servers using
encryption technology in Internet transactions.


Abbreviations: bps – bit per second, ms – millisecond


Annex table 5. Key cloud service infrastructure indicators (continued)




111GLOSSARY


cloud aggregators, systems
integrators and brokers


Sometimes referred to as cloud service partners, they help customers identify
the best solutions and integrate services from different cloud service providers
(section I.D). They identify the most suitable cloud services, integrate them, manage
relationships with cloud providers and, as a result, offer clients a simpler interface
for their IT activities (section III.B.1(c)).


cloud computing A way of delivering applications, services or content remotely to end users, rather
than requiring them to hold data, software or applications on their own devices
(section I.B.1).


cloud economy ecosystem The complex set of relationships, synergies and interactions between technology
and business, governance and innovation, production and consumption, different
businesses and different stakeholders, that contributes to economic and social
development. Encompasses the deployment and impacts of cloud computing
and cloud services within the wider information economy and thereby for national
economic development (section I.D).


cloud service providers Businesses that own the cloud computing centres and other infrastructure, that
form the cloud, and make services, platforms and/or infrastructure available through
them to inter-cloud service providers and cloud service customers (section I.D).


cloud services Services that are provided and used by clients “on demand at any time, through
any access network, using any connected devices [that use] cloud computing
technologies” (see section I.B.1).


cloud-based services Cloud services that require software installation to make use of the cloud’s resources
(section I.B.1).


cloud service customers Citizens, consumers, enterprises and Governments that procure various kinds of
cloud services directly from cloud service providers or inter-cloud service providers
(section I.D.).


co-location data centre A facility that provides space for multiple customers to place their data servers
(section II.B.3).


community cloud A resource/service provided for and shared between a limited range of clients/users
from a specific community with common concerns. It can be managed internally or
by a third party and hosted internally or externally. It might be considered halfway
between public and private cloud provisioning (section I.B.3).


download speed The time taken to transfer data packets from a server to an end user device. Usually
measured in kilobits per second (kbps) or megabits per second (Mbps) (section
II.B.5).


hybrid cloud Expands deployment options for cloud services by mixing, for example, public and
private cloud provision (section I.B.3).


infrastructure as a service
(IaaS)


Cloud service category. In this mode, the provider’s processing, storage, networks
and other fundamental computing resources allow the customer to deploy and run
software, which can include operating systems and applications. The customer
does not manage or control the underlying cloud infrastructure but has control over
operating systems, storage, and deployed applications; and possibly limited control
of select networking components (for example, host firewalls) (section I.B.1).


inter-cloud service providers Providers of cloud services that rely on one or more other cloud service provider(s)
(section I.D).


GLOSSARY




112 INFORMATION ECONOMY REPORT 2013


Internet exchange point A facility where Internet service providers (ISPs) exchange Internet traffic (section
II.B.3).


jitter The variation of latency or the variation in time of data packets arriving. Usually
measured in milliseconds (ms) (section II.B.5).


latency or round trip time
(RTT)


Time taken for a packet to reach the destination server and return to the client (the
end-user device). Usually measured in milliseconds (ms) (section II.B.5).


multi-tenancy Physical and virtual resources are allocated in such a way that multiple tenants
and their computations and data are isolated from and inaccessible to one another
(section I.B).


packet loss The share of packets that fail to arrive at the destination server. Usually measured
as a percentage of the total number of packets transferred (section II.B.5).


platform as a service (PaaS) Cloud service category. In this case, the customer deploys its own applications and
data on platform tools, including programming tools, belonging to and managed
by the cloud provider. The consumer does not manage or control the underlying
cloud infrastructure including network, servers, operating systems, or storage, but
has control over the deployed applications and possibly configuration settings for
the application-hosting environment (section I.B.2).


private cloud A proprietary resource provided for a single organization (for example a Government
or large enterprise). It can be managed internally or by a third party and hosted
internally or externally (section I.B.3).


public cloud An open resource that offers services over a network that is open for public use.
Many mass market services widely used by individuals, such as webmail, storage
and social media are examples of public cloud services (section I.B.3).


secure Internet servers Servers using encryption technology in Internet transactions (section II.B.3).


software as a service (SaaS) Cloud service category. In this case, the customer takes advantage of software
running on the provider’s cloud infrastructure rather than on its own hardware.
The applications required are accessible from various client devices through either
a thin client interface, such as a web browser (for example, web-based email),
or a program interface. In SaaS services, the customer has no control over the
underlying cloud infrastructure, accessing applications through a web browser or
separate program interface (section I.B.2).


upload speed Time taken to transmit data packets from an end user device to a server. Usually
measured in kilobits per second (kbps) or Mbps (section II.B.5).


virtualization Implies the creation of a “virtual version” of a device or resource, such as a server,
storage device, network or an operating system where the framework divides the
resource into one or more different execution environments (section I.B.1).




113SELECTED UNCTAD PUBLICATIONS


LIST OF SELECTED PUBLICATIONS IN
THE AREA OF SCIENCE, TECHNOLOGY


AND ICT FOR DEVELOPMENT


A. Flagship reports
Information Economy Report 2013: Cloud Computing and Developing Countries. United Nations publication.


Sales no. E.13.II.D.6. New York and Geneva.
Information Economy Report 2012: The Software Industry and Developing Countries. United Nations publication.


Sales no. E.12.II.D.14. New York and Geneva.
Information Economy Report 2011: ICTs as an Enabler for Private Sector Development. United Nations publication.


Sales no. E.11.II.D.6. New York and Geneva.
Technology and Innovation Report 2012: Innovation, Technology and South-South Collaboration. United Nations


publication. UNCTAD/TIR/2012. New York and Geneva.
Technology and Innovation Report 2011: Powering Development with Renewable Energy Technologies. United


Nations publication. UNCTAD/TIR/2011. New York and Geneva.
Technology and Innovation Report 2010: Enhancing Food Security in Africa through Science, Technology and


Innovation. United Nations publication. UNCTAD/TIR/2009. New York and Geneva.
Information Economy Report 2010: ICTs, Enterprises and Poverty Alleviation. United Nations publication. Sales


no. E.10.II.D.17. New York and Geneva. October.
Information Economy Report 2009: Trends and Outlook in Turbulent Times. United Nations publication. Sales


no. E.09.II.D.18. New York and Geneva. October.
Information Economy Report 2007–2008: Science and Technology for Development – The New Paradigm of ICT.


United Nations publication. Sales no. E.07.II.D.13. New York and Geneva.
Information Economy Report 2006: The Development Perspective. United Nations publication. Sales no. E.06.


II.D.8. New York and Geneva.
Information Economy Report 2005: E-commerce and Development. United Nations publication. Sales no. E.05.


II.D.19. New York and Geneva.
E-Commerce and Development Report 2004. United Nations publication. New York and Geneva.
E-Commerce and Development Report 2003. United Nations publication. Sales no. E.03.II.D.30. New York and


Geneva.
E-Commerce and Development Report 2002. United Nations publication. New York and Geneva.
E-Commerce and Development Report 2001. United Nations publication. Sales no. E.01.II.D.30. New York and


Geneva.


B. ICT Policy Reviews
ICT Policy Review of Egypt. United Nations publication (2011). New York and Geneva.


C. Science, Technology and Innovation Policy Reviews
Science, Technology & Innovation Policy Review of the Dominican Republic. United Nations publication. UNCTAD/


DTL/STICT/2012/1. New York and Geneva.
A Framework for Science, Technology and Innovation Policy Reviews. United Nations publication. UNCTAD/DTL/


STICT/2011/7. New York and Geneva.
Science, Technology & Innovation Policy Review of El Salvador. United Nations publication. UNCTAD/DTL/


STICT/2011/4. New York and Geneva.
Science, Technology and Innovation Policy Review of Peru. United Nations publication. UNCTAD/DTL/


STICT/2010/2. New York and Geneva.




114 INFORMATION ECONOMY REPORT 2013


Science, Technology and Innovation Policy Review of Ghana. United Nations publication.
UNCTAD/DTL/STICT/2009/8. New York and Geneva.
Science, Technology and Innovation Policy Review of Lesotho. United Nations publication. UNCTAD/DTL/


STICT/2009/7. New York and Geneva.
Science, Technology and Innovation Policy Review of Mauritania. United Nations publication. UNCTAD/DTL/


STICT/2009/6. New York and Geneva.
Science, Technology and Innovation Policy Review of Angola. United Nations publication. UNCTAD/SDTE/


STICT/2008/1. New York and Geneva.
Science, Technology and Innovation Policy Review: the Islamic Republic of Iran. United Nations publication.


UNCTAD/ITE/IPC/2005/7. New York and Geneva.
Investment and Innovation Policy Review of Ethiopia. United Nations publication. UNCTAD/ITE/IPC/Misc.4. New


York and Geneva.
Science, Technology and Innovation Policy Review: Colombia. United Nations publication. Sales no. E.99.II.D.13.


New York and Geneva.
Science, Technology and Innovation Policy Review: Jamaica. United Nations publication. Sales no. E.98.II.D.7.


New York and Geneva.


D. Other publications
Review of E-commerce Legislation Harmonization in the Association of Southeast Asian Nations. UNCTAD/DTL/


STICT/2013/1. United Nations publication. New York and Geneva.
Mobile Money for Business Development in the East African Community: A Comparative Study of Existing


Platforms and Regulations. UNCTAD/DTL/STICT/2012/2. United Nations publication. New York and Geneva.
UNCTAD Current Studies on Geospatial Science and Technology for Development. United Nations Commission


on Science and Technology for Development. United Nations publication. UNCTAD/DTL/STICT/2012/3.
New York and Geneva.


Promoting Local IT Sector Development through Public Procurement. UNCTAD/DTL/STICT/2012/5. United
Nations Publication. New York and Geneva.


UNCTAD Current Studies on Applying a Gender Lens to Science, Technology and Innovation. United Nations
Commission on Science and Technology for Development. United Nations publication. UNCTAD/DTL/
STICT/2011/5. New York and Geneva.


UNCTAD Current Studies on Implementing WSIS Outcomes: Experience to Date and Prospects for the Future.
United Nations Commission on Science and Technology for Development. United Nations publication.
UNCTAD/DTL/STICT/2011/3. New York and Geneva.


UNCTAD Current Studies on Water for Food: Innovative Water Management Technologies for Food Security
and Poverty Alleviation. UNCTAD Current Studies on Science, Technology and Innovation. United Nations
publication. UNCTAD/DTL/STICT/2011/2. New York and Geneva.


UNCTAD Current Studies on Measuring the Impacts of Information and Communication Technology for
Development. UNCTAD Current Studies on Science, Technology and Innovation. United Nations publication.
UNCTAD/DTL/STICT/2011/1. New York and Geneva.


UNCTAD Current Studies on Financing Mechanisms for Information and Communication Technologies for
Development. UNCTAD Current Studies on Science, Technology and Innovation. United Nations publication.
UNCTAD/DTL/STICT/2009/5. New York and Geneva.


UNCTAD Current Studies on Green and Renewable Energy Technologies for Rural Development. UNCTAD Current
Studies on Science, Technology and Innovation. United Nations publication. UNCTAD/DTL/STICT/2009/4.
New York and Geneva.


Study on Prospects for Harmonizing Cyberlegislation in Central America and the Caribbean. UNCTAD/DTL/
STICT/2009/3. New York and Geneva. (In English and Spanish)


Study on Prospects for Harmonizing Cyberlegislation in Latin America. UNCTAD publication. UNCTAD/DTL/
STICT/2009/1. New York and Geneva. (In English and Spanish.)




115SELECTED UNCTAD PUBLICATIONS


Manual for the Production of Statistics on the Information Economy 2009 Revised Edition. United Nations
publication. UNCTAD/SDTE/ECB/2007/2/REV.1. New York and Geneva.


WSIS Follow-up Report 2008. United Nations publication. UNCTAD/DTL/STICT/2008/1. New York and Geneva.
Measuring the Impact of ICT Use in Business: the Case of Manufacturing in Thailand. United Nations publication.


Sales no. E.08.II.D.13. New York and Geneva.
World Information Society Report 2007: Beyond WSIS. United Nations and ITU publication. Geneva.
World Information Society Report 2006. United Nations and ITU publication. Geneva.
The Digital Divide: ICT Diffusion Index 2005. United Nations publication. New York and Geneva.
The Digital Divide: ICT Development Indices 2004. United Nations publication. New York and Geneva.
Africa’s Technology Gap: Case Studies on Kenya, Ghana, Tanzania and Uganda. United Nations publication.


UNCTAD/ITE/IPC/Misc.13. New York and Geneva.
The Biotechnology Promise: Capacity-Building for Participation of Developing Countries
in the Bioeconomy. United Nations publication. UNCTAD/ITE/IPC/2004/2. New York and Geneva.
Information and Communication Technology Development Indices. United Nations publication. Sales no. E.03.


II.D.14. New York and Geneva.
Investment and Technology Policies for Competitiveness: Review of Successful Country Experiences. United


Nations publication. UNCTAD/ITE/IPC/2003/2. New York and Geneva.
Electronic Commerce and Music Business Development in Jamaica: A Portal to the New Economy? United


Nations publication. Sales no. E.02.II.D.17. New York and Geneva.
Changing Dynamics of Global Computer Software and Services Industry: Implications for Developing Countries.


United Nations publication. Sales no. E.02.II.D.3. New York and Geneva.
Partnerships and Networking in Science and Technology for Development. United Nations publication. Sales


no. E.02.II.D.5. New York and Geneva.
Transfer of Technology for Successful Integration into the Global Economy: A Case Study of Embraer in Brazil.


United Nations publication. UNCTAD/ITE/IPC/Misc.20. New York and Geneva.
Transfer of Technology for Successful Integration into the Global Economy: A Case Study of the South African


Automotive Industry. United Nations publication. UNCTAD/ITE/IPC/Misc.21. New York and Geneva.
Transfer of Technology for the Successful Integration into the Global Economy: A Case Study of the Pharmaceutical


Industry in India. United Nations publication. UNCTAD/ITE/IPC/Misc.22. New York and Geneva.
Coalition of Resources for Information and Communication Technologies. United Nations publication. UNCTAD/


ITE/TEB/13. New York and Geneva.
Key Issues in Biotechnology. United Nations publication. UNCTAD/ITE/TEB/10. New York and Geneva.
An Assault on Poverty: Basic Human Needs, Science and Technology. Joint publication with IDRC. ISBN


0-88936-800-7.
Compendium of International Arrangements on Transfer of Technology: Selected Instruments. United Nations


publication. Sales no. E.01.II.D.28. New York and Geneva.


E. Publications by the Partnership on Measuring ICT for Development
Measuring the WSIS Targets - A statistical framework. ITU. Geneva.
Core ICT Indicators 2010. ITU. Geneva.
The Global Information Society: A Statistical View 2008. United Nations publication. Santiago.
Measuring ICT: The Global Status of ICT Indicators. Partnership on Measuring ICT for Development. United


Nations ICT Task Force. New York.






117READERSHIP SURVEY


READERSHIP SURVEY


Information Economy Report 2013: The Cloud Economy and Developing Countries


In order to improve the quality of this report and other publications of the Science, Technology and ICT Branch of
UNCTAD, we welcome the views of our readers on this publication. It would be greatly appreciated if you would
complete the following questionnaire and return it to:


ICT Analysis Section, Office E-7075


Science, Technology and ICT Branch


Division on Technology and Logistics


United Nations


Palais des Nations,


CH-1211, Geneva, Switzerland


Fax: 41 22 917 00 50


ICT4D@unctad.org


1. Name and address of respondent (optional)


...........................................................................................................................................................................


...........................................................................................................................................................................


...........................................................................................................................................................................


2. Which of the following best describes your area of work?


Government ministry
(please specify) .......................................... .


❏ Not-for-profit organization ❏


National statistics office ❏ Public enterprise ❑


Telecommunication regulatory authority ❑ Academic or research institution ❑


Private enterprise ❑ Media ❑


International organization ❑ Other (please specify) ..................................... ❑


3. In which country do you work? ......................................................................................................................


4. What is your assessment of the contents of this publication?


Excellent


Good


Adequate


Poor




118 INFORMATION ECONOMY REPORT 2013


5. How useful is this publication to your work?


Very useful


Somewhat useful


Irrelevant


6. Please indicate the three things you liked best about this publication.


a) .......................................................................................................................................................................


b) .......................................................................................................................................................................


c) .......................................................................................................................................................................


7. Please indicate the three things you liked least about this publication.


a) .......................................................................................................................................................................


b) .......................................................................................................................................................................


c) .......................................................................................................................................................................


8. What additional aspects would you like future editions of this report to cover:


...........................................................................................................................................................................


...........................................................................................................................................................................


...........................................................................................................................................................................


9. Other comments:


...........................................................................................................................................................................


...........................................................................................................................................................................


...........................................................................................................................................................................




In
fo


r
m


a
tIo


n
Ec


o
n


o
m


y
r


Ep
o


r
t 2


0
1


3
UN


CTA
D


UNITED NATIONS


The Cloud Economy and Developing Countries


InformatIon
Economy


rEport 2013


U n i t e d n at i o n s C o n f e r e n C e o n t r a d e a n d d e v e l o p m e n t


13-51665
ISBN 978-92-1-112869-7




Login