BP10_ICT_maintxt

BACKGROUND PAPER 10 (PHASE II)

Information and Communications Technology in Sub-Saharan Africa:

A Sector Review

Mavis Ampah, Daniel Camos, Cecilia Briceño-Garmendia, Michael Minges, Maria Shkratan, and Mark Williams

JANUARY 2009

© 2009 The International Bank for Reconstruction and Development / The World Bank

1818 H Street, NW

Washington, DC 20433 USA

Telephone: 202-473-1000

Internet: www.worldbank.org

E-mail: [email protected]

All rights reserved

A publication of the World Bank.

The World Bank

1818 H Street, NW

Washington, DC 20433 USA

The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily

reflect the views of the Executive Directors of the International Bank for Reconstruction and Development / The World Bank or the governments they represent.

The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors,

denominations, and other information shown on any map in this work do not imply any judgment on the part of The

World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries.

Rights and permissions

The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without

permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The

World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the

work promptly.

For permission to photocopy or reprint any part of this work, please send a request with complete information to the

Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923 USA; telephone: 978-750-8400; fax:

978-750-4470; Internet: www.copyright.com.

All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the

Publisher, The World Bank, 1818 H Street, NW, Washington, DC 20433 USA; fax: 202-522-2422; e-mail:

[email protected].

About AICD

This study is a product of the Africa Infrastructure Country Diagnostic (AICD), a project designed to expand the

world’s knowledge of physical infrastructure in Africa.

AICD will provide a baseline against which future improvements in infrastructure services can be measured,

making it possible to monitor the results achieved from

donor support. It should also provide a better empirical foundation for prioritizing investments and designing

policy reforms in Africa’s infrastructure sectors.

AICD is based on an unprecedented effort to collect

detailed economic and technical data on African infrastructure. The project has produced a series of reports

(such as this one) on public expenditure, spending needs,

and sector performance in each of the main infrastructure sectors—energy, information and communication

technologies, irrigation, transport, and water and sanitation.

Africa’s Infrastructure—A Time for Transformation, published by the World Bank in November 2009,

synthesizes the most significant findings of those reports.

AICD was commissioned by the Infrastructure Consortium

for Africa after the 2005 G-8 summit at Gleneagles, which recognized the importance of scaling up donor finance for

infrastructure in support of Africa’s development.

The first phase of AICD focused on 24 countries that together account for 85 percent of the gross domestic

product, population, and infrastructure aid flows of Sub-

Saharan Africa. The countries are: Benin, Burkina Faso, Cape Verde, Cameroon, Chad, Côte d'Ivoire, the

Democratic Republic of Congo, Ethiopia, Ghana, Kenya,

Lesotho, Madagascar, Malawi, Mozambique, Namibia,

Niger, Nigeria, Rwanda, Senegal, South Africa, Sudan, Tanzania, Uganda, and Zambia. Under a second phase of

the project, coverage is expanding to include as many other

African countries as possible.

Consistent with the genesis of the project, the main focus is

on the 48 countries south of the Sahara that face the most

severe infrastructure challenges. Some components of the

study also cover North African countries so as to provide a broader point of reference. Unless otherwise stated,

therefore, the term “Africa” will be used throughout this

report as a shorthand for “Sub-Saharan Africa.”

The World Bank is implementing AICD with the guidance

of a steering committee that represents the African Union,

the New Partnership for Africa’s Development (NEPAD),

Africa’s regional economic communities, the African Development Bank, the Development Bank of Southern

Africa, and major infrastructure donors.

Financing for AICD is provided by a multidonor trust fund to which the main contributors are the U.K.’s Department

for International Development, the Public Private

Infrastructure Advisory Facility, Agence Française de Développement, the European Commission, and Germany’s

KfW Entwicklungsbank. The Sub-Saharan Africa Transport

Policy Program and the Water and Sanitation Program

provided technical support on data collection and analysis pertaining to their respective sectors. A group of

distinguished peer reviewers from policy-making and

academic circles in Africa and beyond reviewed all of the major outputs of the study to ensure the technical quality of

the work.

The data underlying AICD’s reports, as well as the reports themselves, are available to the public through an

interactive Web site, www.infrastructureafrica.org, that

allows users to download customized data reports and

perform various simulations. Inquiries concerning the availability of data sets should be directed to the editors at

the World Bank in Washington, DC.

iii

Contents

Summary iv

What price access? v

From monopoly to competition vi

Competition and performance vii

Reform and regulation viii

The persistence of state-owned enterprises x

The path to wider access to telecommunications services xi

1 Sector developments 1

Access 1

Prices 8

Quality 19

2 Institutional framework 23

Sector reform 24

Regulation 43

Governance of state-owned enterprises 46

The anatomy and impact of institutions: emerging patterns 48

3 The path to wider access to telecommunications services 50

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

iv

Summary

frica is undergoing a revolution in information and communications and technology (ICT) that is

bringing telecom services within the reach of hundreds of millions of people. The revolution is

based on wireless technologies, which are bypassing the fixed-line networks on which the

telecom markets of developed countries were built.

In the 53 countries surveyed, the rolls of fixed-line subscribers grew by 11 million, from 19.5 million

lines in 2000 to 30.6 million in 2007. That is an improvement, but it pales against the growth in mobile

networks, which added 252 million subscribers over the same time period. The number of mobile

subscribers in the 53 countries jumped more than 15 times—from 15 million in 2000 to 267 million in

2007. By 2006, 57 percent of Africans were living under the footprint of the mobile networks (figure A).

These averages mask wide variations among countries. Of the 53 countries studied, the small middle-

income group had seven times more fixed lines per 100 inhabitants than the low-income countries. The

range in mobile penetration rates is equally great. The average rate for the region is 28.1 mobile

subscriptions per 100 inhabitants, but this ranges from just over 1 in Eritrea to 98 in Seychelles.

Similarly wide variations in

access can be found within countries,

between rich and poor households,

and between rural and urban areas.

Less than 3 percent of rural African

households have access to a fixed

telephone lines, whereas 20 percent of

urban households have them. Rural-

urban differentials in access to cellular

services are less marked, as networks

have extended into remoter areas, with

42 percent of rural dwellers versus 91

percent of urbanites living under the

mobile footprint.

Access to the Internet is much less

widespread than access to basic voice

services. There were less than four

million subscribers in 2007. Of these, more than three-quarters were living north of the Sahara or in the

Republic of South Africa. The most common form of access to the Internet is through shared facilities

such as Internet cafes and telecenters so, in practice, we estimate that there were about 44 million Internet

users in 2007—around 5 percent of the population, less than half the rate in Pakistan.

A

Figure A GSM mobile telephone population coverage in AICD countries, 1998–2006

Source: AICD.


v

What price access?

As mobile network coverage increases across the region, the primary determinant of popular access to

services is price, which is high by international standards and in relation to household incomes in the

region. In 2007 the average monthly prepaid package for mobile service in the countries studied was

priced at around US$12, almost the same as the average package for fixed lines. There is great variation

across countries, with pre-paid mobile baskets ranging from a high of US$19 to a low of US$4. The price

of a fixed-line package covers a similar range, from $2 to $25 per month.

Broadband Internet is not always available, but where it is, services are usually charged on a flat-rate

basis. Here, too, prices tend to be very high. The average price of an entry level monthly DSL

subscription in Sub-Saharan Africa was over US$100. In comparison, the average monthly price in

OECD countries for a broadband connection was US$34.

But although mobile tariffs are still high given the low incomes in the region, they are falling steadily

and, as the networks expand, are reaching lower-income customers. One indication of the drop in prices is

the steady fall in the average revenue generated by network subscribers. The monthly average revenue per

mobile user (ARPU) stood at US$16 in 2007, less than half the level of US$40 in 2000. There is plenty of

room, however, for prices to come down even further. Over the same time period, ARPUs in Bangladesh,

India, and Pakistan fell by almost 90 percent—to US$4 per month. In the fixed-line realm, some prices

have increased and others fallen as competition has forced the fixed operators to rebalance tariffs. The

average price of a call to the United States from the region, for example, was cut in half between 2000

and 2007.

Despite high service prices, the mobile networks have been able to provide access to low-income

users through flexible retail packages. Over 90 percent of the region’s consumers are on prepaid plans,

which allow them to purchase services in tiny increments and control their spending precisely. High

connection charges are rare, so the minimum cost of access to mobile services is generally lower than for

fixed networks, which traditionally operate on a post-paid subscription basis. For the operators,

prepayment dramatically reduces credit risk and the cost of revenue collection. Moreover, the absence of

credit checks, proof-of-address requirements, and other “know your customer” measures has reduced the

cost of customer acquisition in the region and increased the flexibility of markets.

Other factors—notably taxes and energy shortages—keep prices of ICT services in Africa higher than

they would be if market dynamics alone were at work. These taxes include import duties on mobile

handsets, taxes on services, and, in some countries, particularly in East Africa, excise charges on calls.

Value-added taxes range from 5 percent to 23 percent in the countries studied. The combined effect of

these taxes and duties is to add significantly to the cost of mobile ownership, putting ICTs outside the

reach of many consumers who otherwise might be able to afford them. Uganda ranks second in the world

in taxes as a percentage of total mobile revenues, while Kenya and Tanzania are above the world average.

Shortages of electricity contribute to higher costs, as service providers must rely on their own generators

to power mobile base stations and other telecommunications equipment. Scarce and unreliable electricity

also affects operators’ earnings because mobile subscribers, particularly in rural areas, have difficulty

recharging the batteries in their mobile phones.


vi

From monopoly to competition

Competition is the quickest route to lower prices and wider access to services. Perspectives on

telecom governance have changed radically over the past few decades. The most important implication

has been the shift from monopoly to competition. Greater competition has brought expanded networks,

lower prices, and new efforts to reach previously underserved groups of customers.

Since 1993, most of

the countries studied

have introduced some

degree of competition in

their telecommunications

markets (figure B). Less

than ten countries have a

monopoly mobile market

and the majority of

African countries have

more than two mobile

operators.

More than two dozen

of the countries allow

some degree of

competition in fixed-line

and international

markets, but only a few

have more than two operators in these segments. Few outright bans on competition remain in national

legislation.

The popularity of the Internet has resulted in growing demand, and most countries have issued

licenses to several Internet service providers (ISPs). Some countries have relaxed their authorization

regimes, requiring low or no license fees, so there are a number of licensed or registered ISPs; in some

countries, only a fraction of these are operational.

Despite the large number of ISPs licenses that have been issued, many of the countries have imposed

restrictions on what ISPs can do. For example, ISPs are often not allowed to provide their own

infrastructure unless they obtain other licenses. Restrictions on entry into the fixed-line and international

gateway markets have meant that ISPs have often had to lease infrastructure from incumbent operators

sometimes at prices that are not cost-based. ISPs are sometimes not allowed to directly obtain

international bandwidth, which is particularly onerous for landlocked countries that rely on satellite. Even

in countries with other options for international connectivity (such as undersea fiber optic networks),

incumbent operators often have a stranglehold on landing stations and belong to consortiums that own the

networks. This means that ISPs have no choice but to go through the incumbent operator for fiber-based

international connections.

Figure B Status of mobile competition

Source: AICD.


vii

Of the 14 AICD countries for which the International Telecommunications Union has data, most do

not allow full competition in international gateways. Six reported that international gateways were a

monopoly of the incumbent, four reported partial competition; only four responded that full competition

existed in this market segment.

Despite widespread de jure liberalization, the process of de facto liberalization has been moving more

slowly. Barely half of the countries studied have more than three active operators. Constraints on

competition often appear in the licensing process. In some countries there is no clear procedure for

obtaining a license, perpetuating the de facto monopoly. In other countries, the complexity of the

licensing process can discourage new entrants. For example, some countries divide the market into many

segments and require a license for each. Sometimes it is not clear which licenses are needed to provide

which service, or whether the scope of a license allows the licensee to provide the services it wishes to

provide.

Competition and performance

Countries that pursued early market liberalization for mobile telephony had an average penetration

level that was 2.2 points higher in 2005 than would be expected from their average income (figure C).

Liberalized countries are those that have established an independent regulatory agency, partly privatized

government-owned operators, and maintained competition for at least five years. Those that did not had

average mobile penetration rates 2.1 points below the level suggested by their income. As competitive

markets develop, the gap between countries that are reforming and those that are not is getting wider. The

performance gap in the fixed-line sector followed the same pattern but was less pronounced.

The effects of competition

deepen as reform proceeds. Mobile

penetration speeds up, for example,

as the number of operators in the

market increases. For the sample

countries, mobile subscriptions

increased by more than 3 percentage

points annually after the entry of the

fourth operator. The growth occurs

earlier in higher-income countries

(those where annual per capita

income is greater than $1,000),

where subscriptions jumped 11

percentage points after the entry of

the second operator.

Competition in mobile services

also forces fixed operators to

rebalance their tariff structure

Figure C Difference between expected and actual mobile penetration, 2000–05

Source: AICD.


viii

(figure D). This has resulted in higher average fixed tariffs, but significantly lower prices for international

calls.

Competition is the primary driver of reductions in the price of long-distance and international

services. However, access to submarine fiber-optic connectivity also plays an important role in

determining the price of international services, particularly Internet access. When access to the submarine

cables remains in the hands of the incumbent operator, the incumbent is likely—in the absence of

adequate regulatory controls—to prevent the full cost advantage of this technology from being passed on

to consumers. In countries with multiple international gateways, some competitive pressure is exerted,

and service prices are significantly lower than in countries where the submarine cable provides the only

international gateway. In the case of broadband and Internet access in general, access to fiber has

undoubtedly had a significant impact on prices.

The SAT-3 undersea fiber-optic

cable has helped to alleviate the

shortage of bandwidth for a number

of countries on Africa’s west coast.

In addition, Cape Verde and Sudan

have been able to connect to other

fiber-optic submarine cable systems.

Although landlocked, Ethiopia is

sending a overland fiber-optic cable

to Sudan to tap into that country’s

fiber link to Saudi Arabia. Some

nations on the west coast, which lack

their own international fiber outlet,

are also using terrestrial links to

connect to neighbors with a SAT-3

landing station. For example,

Namibia has a fiber link to South

Africa. East Africa has been

particularly affected by a shortage of fiber-based international Internet connectivity and as a result faces

high retail prices. Most East African countries are collaborating to create the East African Submarine

Cable System (EASSy), which would provide high-speed fiber optic connectivity at lower costs. Progress

toward EASSy, however, has been hampered by governments keen to ensure that the system will provide

access to those outside the consortium.

Reform and regulation

How telecommunications markets are structured and regulated affects competition—for example, by

limiting entry to the market—and therefore affects access, by keeping prices higher than they otherwise

would be.

Figure D Tariff rebalancing following reform of the telecommunications sector, 2000–05

Index: 2000 = 100

Source: AICD.


ix

All the countries studied have laws and regulations covering the telecommunications sector; over 20

adopted their present legislation after 2000. A typical bill establishes a national regulatory agency (NRA)

to supervise the telecommunications sector and contains general provisions governing competition,

licensing, interconnection, allocation of scarce resources (such as numbering and spectrum), pricing, and

market entry. Wide variations are found between countries in the extent of reform.

Compared with other infrastructure sectors, there has been intensive institutional reform. These

changes have been predominantly driven by market reforms that fostered competition and facilitated

various forms of private participation. Policy oversight evolved accordingly.

Progress is also evident on the regulatory front, although the picture is more mixed. By 2007, 45

AICD countries had NRAs in operation, but governments continue to interfere with their decisions.

Effective NRAs depend on legal frameworks that make them accountable to the public, encourage them

to operate transparently, give them the enforcement powers and other tools they need to do their job, and

grant them autonomy and freedom from political interference. Regulatory quality differs from country to

country (figure E), but autonomy is a particularly scarce commodity.

Figure E Telecom regulatory score

Source: AICD

On the other hand, progress on accountability has been encouraging, particularly if compared with

other variables of regulation and other infrastructure sectors. Achievements in transparency have been

mixed and generally incomplete, with much ground yet to be covered. Almost all the NRAs have Web

sites. Yet availability and quality of online information remain uneven.


x

Failure to reform has a direct

fiscal cost because competitive

markets generate higher taxable

revenues. Telecom revenue as a

percentage of GDP in the liberalized

countries is 5.6 percent, compared

with 3.5 percent in the nonliberalized

countries (figure F). In the

liberalized countries telecom

revenues increased by 2.5 percent of

GDP between 2000 and 2005,

compared with the 1.2 percent rise

observed in the nonliberalized

countries.

The persistence of state-

owned enterprises

Given the private sector’s success in delivering ICT services, it is striking that half of the fixed-line

operators in Africa remain in public hands, despite low productivity and poor quality of service. Only

those in South Africa and Sudan put into effect even half of international best practices for governance

(figure G).

In countries with

state-owned fixed-line

incumbents, public

spending on telephone

service averaged 2

percent of GDP, an

extraordinarily large

amount to be spent by

the public sector in a

market that is

increasingly

competitive. The reason

for the high spending is

clear enough. It is not

uncommon for public

utilities to be used as

social buffers,

redistributing wealth via excessive employment. However, this practice carries with it substantial hidden

costs of redundancy and inefficiency that are as much as 0.3 percent of GDP in Tanzania or US$200 per

subscriber in Chad.

Figure F Telecom revenue as share of GDP, selected African countries, 2000–05

Note: Liberalized countries are those that have established an independent regulatory agency, partly privatized government-owned operators, and maintained competition for at least five years.

Figure G Telecom governance score

Source: AICD


xi

The path to wider access to telecommunications services

Telecommunications reforms have led to more competitive markets in many of the countries studied.

The result has been impressive growth during the first half of the 2000s, particularly in mobile telephony.

The challenge will be to sustain this growth in the face of significant barriers.

The key to extending communications access in Africa is to leverage the unprecedented success of

mobile technology on the continent. Given that mobile operators tend to have the largest

telecommunications networks in most countries, the incremental costs of extending mobile coverage into

underserved areas is probably less than that of other solutions, such as extending fixed-line networks or

promoting the voice-over-Internet protocol. A companion project to this study has estimated the cost of

extending mobile coverage to areas that currently do not have a mobile signal.

A number of key policy recommendations, if followed, would sustain growth and deepen access to

telecommunications in the region.

There is ample scope for further sector reform in most countries. According to a 2006 report from

the GSM Association, poor regulation has reduced telecommunications investment in Africa by

US$4.6 billion. Countries that have not yet privatized incumbent operators should do so in order

to reduce direct state intervention in operations, encourage a more level playing field, and attract

investment and innovation. Additional competition should be introduced by not limiting the

number of operating licenses available. Regulatory agencies should be strengthened and allowed

to operate independently.

Countries should pursue liberalization by simplifying licensing regimes, lifting remaining bars to

market entry, and examining the feasibility of introducing mobile number portability and mobile

virtual network operators.

Efforts should be increased to lower prices for telecommunications services. Average per capita

income in Sub-Saharan Africa was just US$970 in 2007—less than US$3 a day. Any incremental

efforts to lower prices would have a tremendous impact on affordability and hence access. A few

ways to push prices down are to lower taxes and termination fees, and, where competition is

limited, through regulatory action.

Mobile telephone access should be incorporated into established goals for universal access so as

to leverage the successful spread of mobile communications. Mobile telephony has probably done

more to increase access through a competitive environment than any other policy, yet, for the

most part mobile operators have not been involved in formal universal access programs. Adapted

universal access policies might require mobile operators to expand coverage as a condition of

licensing or allow mobile operators that expanded coverage to receive money from universal

service funds.

High-speed connectivity over fiber optic cable is a prerequisite for e-government and other

socioeconomically beneficial applications. Private-public partnerships can play a useful role in

developing and expanding national, regional, and international fiber optic links throughout the

region, allowing Sub-Saharan Africa to join fully in the global information society. Although


xii

governments should play an active role in encouraging the deployment of fiber networks, their

participation should not delay the badly needed implementation of fiber backbones.


1

1 Sector developments

The AICD project here reviews the telecommunications sector in African countries. It analyzes

trends, scrutinizes benchmark performances, and identifies barriers to sustainable sector development.

The review begins by outlining the existing legal and regulatory framework for the sector and

determining the degree of competition. It then analyzes the status of telecom privatization in the region.

Pricing, international Internet bandwidth, and universal access policies are also studied. After establishing

the levels of liberalization in the countries, their performances are compared. The review concludes with

key policy recommendations.

Access

Africa is not exempt from the transformation of the world’s telecommunications services industry,

dating back to the 1980s. Still, the region has the lowest penetration rate in the world for fixed phone lines

and Internet services and its penetration rate for mobile services is the lowest along with South Asia

(figure 1.1). This mixed picture nevertheless suggests that Africa is poised for an information and

communication technology (ICT) explosion and might be positioned for a technological leapfrog because

its booming mobile market is essentially bypassing the capital constraints associated with fixed-line

telephony.

Global averages, of course, mask significant variations from country to country. In fact, among the

countries studied, fixed and mobile penetration rates are markedly different. The region’s high and upper-

income countries (HUICs) are the best performers in fixed-line penetration. HUICs have ten times more

main lines per 100 habitants than low-income countries (LICs) do. Average fixed-line penetration was 3.2

in 2007 and just over half a dozen Sub-Saharan African countries had more than five fixed lines per 100

inhabitants (figure 1.2). The country variation in mobile subscribers per 100 inhabitants is also enormous.

Although the average was 28, country values ranged from 98 in Seychelles to just over 1 in Eritrea.

Similar to fixed-line patterns, mobile penetration for HUICs is significantly higher than for LICs

(figure 1.3). Income level is, in fact, a second-order predictor for penetration rates. But the study found no

consistent pattern of other environmental variables—like geography, urbanization, or country size in

terms of population—that explains the variation in phone penetration.


2

Figure 1.1 ICT penetration rates per 100 inhabitants, 2007

Figure 1.2 Fixed telephone lines per 100 inhabitants, 2007

Source: AICD.


3

Figure 1.3 Mobile subscribers per 100 inhabitants, 2007

Source: AICD.

These penetration rates disproportionately affect poor and rural households (figures 1.4 and 1.5). In

fact, less than 3 percent of rural African households have access to a fixed telephone line, compared with

a figure of about 20 percent for urban residents. Similarly, access disparities are enormous: the richest

quintile has close to 30 percent household access for fixed and cellular services. The poorest quintile has

less than 2 percent household access for the same. These disparities, together with government-led efforts

to address them, underpin attempts to provide shared, or public, services. Providing ICT services at public

facilities such as pay phones, Internet cafes, and so forth is emerging as an attractive solution in efforts to

provide universal access.1

There are a number of ways to measure universal access. Public telephone facilities provide one

measure, and the available data suggest that a number of countries have committed to providing public

telephones. Namibia, Seychelles, South Africa and Togo have more than two public telephones per 1,000

inhabitants—above the regional average (figure 1.6). Public telephones account for a significant portion

of main lines in Togo and Uganda. Uganda liberalized its pay phone market several years ago, doubling

its number of pay phones between 2004 and 2005. The statistics probably understate the provision of

public calling availability because they often do not include the region’s large informal market that sells

mobile airtime. Nevertheless, overall pay phone access is low, particularly when compared with other

1 This means universal access as opposed to universal service, in which every household has a private connection.

For more information on universal service and access, see ITU (1998).


4

low-income regions. For example, in Bangladesh, India, and Pakistan, pay phone penetration is more than

2 per 1,000 inhabitants, almost eight times the average for Sub-Saharan low-income countries and above

the average for Sub-Saharan middle- and upper-income countries.

Figure 1.4 Percentage of households with a telephone, latest available data, selected African countries—urban/rural split

Figure 1.5 Percentage of households with a telephone, latest available data, selected African countries by quintile

Source: AICD adapted from national household surveys, UNDP. Note: Simple average of countries with data.

Figure 1.6 Public pay phones, per 1,000 people, 2007

Source: AICD.


5

Another way to measure universal access is the number of take-up addresses, or the number of

households in an area where the service is provided. In the case of mobile services, the equivalent

measure is the percentage of the population within range of a mobile signal (regardless of whether they

own or even use a mobile phone). On average, only about 15 percent of African households have a fixed

line (figure 1.7). For mobile services in 2006, 57 percent of the population lived within signal range

(figure 1.8). Comoros, Mauritius, Seychelles, South Africa and Uganda have already achieved full or

nearly full signal coverage for their populations.

In terms of market potential, the key measure is infrastructure footprint—the signal coverage of the

global system for mobile communication (GSM) or the availability of wires for fixed lines. What share of

the population has access to network facilities but chooses not to connect, either because of affordability

or consumer preference? Are the networks and signals reaching customers? Services are only one step

removed from reaching households and customers directly.

Figure 1.7 Percentage of households with a fixed-line telephone, latest year available

Source: AICD Household Survey Database.


6

Figure 1.8 Mobile population coverage, 2007

Source: Adapted from Winrock International, operator reports.

But presenting this static view of ICT leaves out a key characteristic of the sector: its technological

dynamism. Among Sub-Saharan African countries, fixed-line subscribers saw a net increase of 2.7

million between 2000 and 2007 (rising in 2007 to 12.0 million from 9.3 million in 2000). However this

increase did not keep pace with population growth and penetration barely rose from 1.4 in 2000 to 1.5

main lines per 100 inhabitants in 2007. This growth is dwarfed by statistics for the mobile sector, which

added 172 million subscriptions over the same time period. The number of mobile subscriptions in Sub-

Saharan Africa jumped from 11 million in 2000 to more than 183 million in 2007. This has boosted

mobile penetration from 1.7 in 2000 to 23.1 in 2007. The market has also become more evenly

distributed. While South Africa accounted for 73 percent of subscribers in 2000, this had dropped to 24

percent in 2007.

But the growth rate for mobile penetration may be slowing. The annual growth rate fell through the

year 2003 but then picked up in 2004. However the rate of growth has been declining since 2005

(figure 1.9).

GSM coverage, meanwhile, continues a steady expansion, and by 2006 reached more than 91 percent

for urban areas (figures 1.10 and 1.11). This points to what is undoubtedly one of the primary challenges

in the sector: providers are not intent on providing services to marginal users who have limited resources

and live in remote areas. Another challenge is the provision of information/data services through the

network to a larger segment of the population.


7

Figure 1.9 Growth in new mobile subscribers Figure 1.10 GSM population coverage in AICD countries

Source: AICD. Source: AICD.

Figure 1.11 GSM footprint January 1999 and September 2006

Source: AICD.

Internet access is perhaps the most important secondary service delivered over the telephone

infrastructure (although most Internet access in Africa is obtained via wireless). It is estimated that there

are less than two and half million Internet subscribers in Sub-Saharan Africa (as distinguished from

users), and over half are in Nigeria and South Africa alone. In addition, most users have only shared

access at offices, schools, and Internet cafes. It is estimated there were some 23 million Internet users in

Sub-Saharan Africa in 2007—or some ten users to one subscription. Overall penetration was just 3

percent of the population (figure 1.12).


8

There are no official surveys on the number of Internet users in Africa, so estimates are typically

based on multipliers of subscriptions or international bandwidth. Given the unreliability of Internet data

for the region, there may, in fact, be many more users in the region than believed. But these assumptions

would apply equally to other developing-country regions. Sub-Saharan Africa therefore ranks low in

terms of international comparisons of Internet access.

For example, the average Internet penetration in low-income nations Bangladesh, India, and Pakistan

was estimated at almost 6 per 100 people in 2007. This is almost two times the Sub-Saharan average.

Figure 1.12 Internet users per 100 inhabitants, 2007

Source: AICD.

Prices

With promising take-up ratios, pricing helps to determine access to telephony. For fixed-line

networks, the pricing structure is characterized by high connection charges, which impede the addition of

new subscribers. By way of contrast, the level and flexibility in pricing underpins the success of mobile

telephony in Sub-Saharan Africa.

Monthly packages for subscription-based, conventional fixed-line telephone service averaged US$12

in 2007; prices varied a great deal across countries (figure 1.13). Countries such as Cape Verde and

Ethiopia that have monopoly operators tend to have relatively low tariffs because they have not yet fully

rebalanced them (i.e., charges for monthly line rental and local calls are set below cost while charges for

national and international calls are above cost). In countries that privatized their fixed-line operator (Côte

d’Ivoire, Senegal, South Africa, Tanzania, and Uganda), operators rebalanced tariffs, a move that has led

to above-average monthly packages. Kenya, Mozambique, and Namibia use state-owned operators who

rebalanced fixed-line pricing with above-average monthly packages.


9

Connection charges for fixed lines tend to be higher than those for mobile networks. In addition,

fixed-line networks have typically charged monthly rental fees for copper line access. Also, fixed-line

calls to mobile networks—where most subscribers are—tend to be more expensive than on-net mobile

calls. With the new fixed wireless networks, however, fixed-line operators are moving toward pricing

structures that were once unique to mobile services. These rely on prepayment and waive monthly

subscription charges; usage fees, however, are higher.

In the countries studied, the average monthly prepaid mobile package in 2007 was US$12; packages

ranged from a high of US$20 to a low of US$4 (figure 1.14).2 Although not significantly less costly, on

average, than a fixed line, mobile pricing is flexible and offers users an adaptable menu of options.

Mobile prepaid tariffs are fairly complex given the range of prices (which are keyed to networks used)

and the time of the call (peak or off-peak). Complexity aside, the pricing arrangements have allowed more

people to access mobile service. Connection charges are rare and credit checks unnecessary. Over 90

percent of African mobile subscribers are using prepaid plans.

Figure 1.13 Fixed-line monthly package, US$, 2007

Source: AICD.

Note: The package is based on one-fifth the connection charge, the monthly subscription charge, and 15 three-minute peak and off-peak calls each.

2 The OECD low-user mobile package is used to compare prepaid tariffs across the countries studied. The monthly

package includes 25 calls distributed over different destinations and calling periods. It also includes 30 text

messages.


10

Figure 1.14 Mobile prepaid monthly tariff package, US$, 2007

Source: AICD.

Note: Based on OECD low-user package methodology.

Furthermore, mobile prices have dropped as networks have expanded and less-affluent customers sign

on. The monthly mobile average revenue per user (ARPU) stood at US$16 in 2007, almost half the figure

for 2000 of US$40 (figure 1.15). Moreover, prices could easily fall even further. For example, the ARPU

in the three South Asian countries of Bangladesh, India, and Pakistan stood at US$40 in 2000. By 2007,

average ARPU in these South Asian nations fell to around US$4—almost a fourth of that of the African

region.


11

Figure 1.15 Mobile ARPU, US$

Source: AICD.

Note: BIP = Bangladesh, India, and Pakistan.

Although mobile prices have been falling, further decreases are needed, especially given the high

connection rates and high taxes in some countries. Mobile connectivity has been an ongoing problem in

Africa just as in other regions of the world. Mobile termination rates (MTRs) are an issue because

operators often fail to agree on fees for terminating calls on mobile networks. More often than not, the

rates are unrelated to costs and therefore discourage competition through different pricing structures.3

Despite retail competition among mobile operators, regulators in a number of countries have determined

that they have a monopoly over the termination of calls on their networks and have begun to regulate

them. Given the lower incomes in Africa, its MTRs are relatively high (the region has wide variations—

see figure 1.16) and termination rates need to be cost based.

3 This is the situation in Kenya, where the smaller mobile operator complained to the NRA about the practice of the

dominant mobile operator charging much cheaper for calls made within its network: “Early in the year, Celtel wrote

to the CCK complaining of alleged monopolistic practices by Safaricom, including the locking in of subscribers

through high charges to other networks. Safaricom currently charges its subscribers up to Ksh50 (US$0.71) per

minute to access the Celtel network, and Ksh45 (US$0.64) a minute for calls to Telkom. In contrast, calls

terminating within the network are charged as little as Ksh8 (US$0.11) per minute. On its part, Celtel charges as low

as Ksh16 (US$0.22) per minute to call other networks.” “CCK Caps Interconnection Charges for Warring Mobile

Firms.” The East African (Nairobi). February 25, 2007


12

Figure 1.16 Mobile termination rates, US$ per minute, 2006

Source: AICD database.

Note: * Monopoly mobile market owned by fixed-line operator, therefore mobile termination rate is not applicable.

In Africa, mobile interconnection has generally been left to operators to resolve. The national

regulatory agencies (NRAs) get involved only when the operators cannot agree. Some NRAs are taking a

more active approach, however, through the establishment of MTR ceilings. In Tanzania, the NRA

introduced a glide path calling for annual MTR reductions. Nigeria’s NRA has intervened several times

by establishing MTR targets.4 In Kenya, the NRA recently established a ceiling MTR in addition to a cap

on retail fees for off-net calls.

These developments are in line with global trends whereby dominant operators (or those with

significant market power [SMP]) are given extra obligations—including the obligation to provide

wholesale, cost-based connection to their networks. Dominant operators sometimes are asked to publish a

so-called Reference Interconnection Offer that supplies technical and economic details on

interconnection. The European Union (EU) declares dominance and SMP and the determination often

includes an analysis of market share. The EU has identified mobile call termination as an individual

market, and operators have been declared dominant in these markets in many countries. This has led to

termination rate controls by most regulators.

The Senegalese NRA conducted a market-dominance analysis for 2006. Rather than looking at

individual markets, it looked at the entire telecommunication sector and determined the incumbent was

4www.ncc.gov.ng/interconnection/Interconnect%20Rate%20Determination%202006.pdf.


13

dominant on the basis of its 89 percent market share of overall sector revenues.5 South Africa’s NRA is

undertaking a market dominance investigation that the newest mobile operator hopes will level the

playing field.6 In Niger and Rwanda, operators have also been found to be dominant, but measures to

control their influence have been limited. For example, the Niger NRA found the leading mobile operator

dominant and required it to publish an interconnection catalog.7

Two additional factors might be deterring the expansion of mobile services: (a) taxes/duties on

equipment and (b) energy shortages. Taxes on telecommunication equipment and services raise prices,

which put ICTs outside the reach of some consumers. Import duties on mobile handsets add to the cost of

what is often perceived as one of the biggest barriers to increased penetration: the price of the cellular

telephone. Taxes on services increase prices and discourage usage. Value-added taxes (VAT) on

communication services range from 5 to 23 percent in the countries studied (figure 1.17). In addition,

5 www.artp-senegal.org/telecharger/Decision%20%20fixant%20la%20liste%20des%20operateurs.pdf 6 “Cell C welcomes ICASA inquiry into introduction of cost-based interconnection rates.” Press Release. February

6, 2006. www.cellc.co.za/common/includes/news_headlines_detail.asp?cl_pkiArticleNo=119 7 Autorité de Régulation Multisectorielle. “Décision No 13 du 2 Aout 2005 Conseil National de Régulation (CNR)

portant liste des opérateurs dominants.” http://niger.arm-niger.org/decision013.pdf

Figure 1.17 Value-added and excise taxes, 2007

Source: GSM Association.


14

some countries, particularly in East Africa, charge an excise tax on calls. Combining import duties on top

of VAT and excise taxes can add significantly to the cost of mobile ownership. For example, Uganda has

the second highest taxes in the world as a percentage of total mobile ownership; Kenya and Tanzania are

above the world average.

Energy shortages also affect pricing. Shortages of electricity contribute to higher costs because

operators must operate their own generators for mobile base stations and other telecommunication

equipment.8 Shortages also dampen operator earnings because mobile subscribers, particularly in the

countryside, have difficulty recharging their mobile phone batteries.

Another segment of the ICT market with an impressive trend is the long-distance market. Prices for

overseas calls have declined as operators move to rebalance tariffs and compete with Internet-based

calling solutions. On average, long-distance prices were halved between 2000 and 2007 (figure 1.18).

Nevertheless, prices remain high by global standards.

Figure 1.18 Average price of a one-minute peak-rate call from African countries to the United States, US$

Source: AICD.

Long-distance prices for calls within Africa are slightly higher than calls made to the United States

(figures 1.19 and 1.20). A one-minute call within Africa is US$0.99, while a one-minute call to the

8 “Mainly in Africa, with the exception of Mauritius, the electricity supply is insufficient due to the growth

experienced in most of the countries where we operate. We therefore have to rely on diesel-powered generators that

we source, install, maintain and refuel. In Chad and Sierra Leone, at March 31, 2007, close to 100 percent of our

radio sites were powered by diesel-powered generators, and in the Democratic Republic of Congo it was the case for

about 75 percent of our sites. This increases our costs and impacts the profitability of our African operations.”

Millicom International Cellular SA. Form 20-F. 2007.


15

United States would cost about US$0.87. Significant variations exist within the region. Intra-Africa call

prices range from $0.31 to $3.98 per minute.

Figure 1.19 Price of one-minute peak-rate call to the United States, US$, 2007

Figure 1.20 Price of one-minute peak-rate call within Africa, US$, 2007.

Source: AICD.


16

A pattern of lower long-distance tariffs is evident for countries that are cosignatories to specific trade

and economic agreements in comparison with countries outside the agreements (figure 1.21)

Figure 1.21 Price of one-minute peak-rate call within and outside AFR trade agreements, USD per minute, 2006

Source: AICD.

Note: Excluding Chad and Cape Verde.

For services over the phone, Internet access prices are high. The high prices are the result of

restrictions on Internet service providers (ISPs), limited international connectivity options, small markets,

and the practice of charging telephone usage fees for dial-up access.

The average price for 20 hours of Internet use was US$46 per month (figure 1.23). One factor is

telephone usage fees. A few countries offer lower fees for Internet access, but most charge Internet dial-

up at conventional voice-calling rates. The move toward tariff rebalancing, which has raised local calling

fees, exacerbates the high cost of Internet access in the region.


17

Figure 1.23 Price structure of Internet package, US$ per month, 2007

Source: AICD. * Broadband price is cheaper than dial-up.

The high cost of Internet dial-up should motivate users to move to broadband such as Asymmetric

Digital Subscriber Lines (DSL), which is charged on a flat-rate basis. But DSL is not always available (by

2007, around a dozen countries in Africa had not launched DSL), and where it is available, DSL prices

tend to be quite high. The average price of a monthly DSL subscription was over US$100 in Africa

(figure 1.24). In comparison average prices in the OECD was US$34. There are about a dozen countries

where an always-on DSL package was cheaper than 20 hours of dial-up although the speed is sometimes

less than 256 kbps. All of the North African countries except Algeria had monthly DSL prices of less than

US$20 (for a 256 kbps download connection). In contrast, most countries in Eastern Africa charge high

rates for DSL access, primarily because of a lack of cheap, fiber-optic-based international Internet

connectivity. Most of the countries that charge less than US$100 have access to international fiber

networks. Internet market development is linked to broadband pricing. In general, countries with the

highest Internet penetration have the cheapest DSL prices. One exception is Kenya, which, as noted,

suffers from a lack of fiber-based international connectivity. Another is Benin, which, despite a landing

point for the SAT-3 fiber submarine cable, has above-average DSL prices. One reason is limited

competition in the ISP market, including the need to gain access to SAT-3 connectivity through the

incumbent.


18

Figure 1.22 Broadband Internet prices, US$ per month, 2007

Source: ITU, AICD.


19

Quality

Fixed-line quality of service is measured by the number of faults per 100 main lines per year. Recent

data are lacking in most of the countries, but for those that do report this figure, the average value was 69

in 2005, or some 7 out of 10 fixed lines being out of service at some point during the year. In contrast, the

average figure for 2003 for 14 Organisation for Economic Co-operation and Development (OECD)

countries was only 1 in 10; Canada and the Republic of Korea reported fault rates of 1 in 100 (OECD

2005).

Mobile quality of service is rarely monitored (let alone published) by the region’s regulatory

authorities. One exception is Senegal’s Telecommunications and Post Regulatory Authority, which

contracted a company to conduct a quality survey in October–November 2006 across four applications:

voice, short message service (or texting), connecting the two operators, and data (general packet radio

service).9 For the voice tests, the survey assessed the ease of establishing a call, maintaining a call for two

minutes, and audio quality. The results were aggregated into a single indicator ranging from 0 percent to

100 percent; higher values represented better quality. The two mobile networks were found to have

perfect to acceptable quality with ratings of between 80 percent and 94 percent.

9 Directique. Enquête Qualité de Service des Réseaux GSM au Sénégal–Octobre-Novembre 2006.

Figure 1.24 DSL price, USD per month, 2007

Source: AICD.


20

Productivity is an important issue for the region because it affects costs. The more efficient a firm, the

lower its costs. This is particularly critical in Africa, where incomes are low and any measure to reduce

costs will tend to reduce prices and thus increase access. The traditional industry measure of productivity

is the number of subscribers per employee. There are wide variations in the region both between fixed and

mobile networks and within similar networks.

Productivity for fixed-line networks is much lower than for mobile networks. Among operators that

publish this information, the number of fixed-line subscribers per employee ranges from 35 to 186,

whereas the minimum figure is 724 for mobile productivity. Within the fixed-line segment, gains in

productivity are low because the market is stagnant. Productivity increases tend to proceed from staff

reductions rather than from any increase in the number of lines. For example, although Telkom South

Africa saw fixed lines increase per employee (from 121 in March 2002 to 184 in March 2006), the per-

employee increase in lines was actually caused by the loss of nearly 14,000 staff over the same period. In

fact, the number of fixed lines actually fell. A number of state-owned operators face the problem of

excessive staff. This affects not only productivity, but also to the ability to privatize.10 From the limited

data, it appears that partly privatized firms are more productive than fully state-owned operators. For

example, Sonatel of Senegal and Telkom South Africa, both partly private, have higher productivity

levels than TDM of Mozambique or Telecom Namibia, which are fully state owned.

As mentioned, productivity levels in the mobile sector are far higher than for fixed lines.

Furthermore, mobile productivity continues to grow because new subscribers far exceed the need for

additional staff (figure 1.25). There are significant variations in mobile productivity throughout the region

including differences within subsidiaries of the same company. For example, Vodacom Congo (D.R.) had

4,760 mobile subscribers per employee in 2008, whereas Vodacom Mozambique had more than one and

one half times this—or 8,166 subscribers per employee. One surprising statistic is that productivity in

African mobile networks tends to be higher than in developed nations. In 2008, Vodacom’s five mobile

networks in Africa had an average of 5,628 subscribers per employee, whereas at Vodafone in the UK the

average was only 1,788. This is partly explained by the many prepaid subscribers in Africa who tend to

create a significant amount of downstream employment in airtime card sales,11 allowing operator staff to

focus on core activities.

10 Kenya was involved in a staff reduction for Telekom Kenya prior to privatization: “Kenya’s state-owned fixed-

line telephone firm Telkom Kenya plans to lay off more than 12,000 workers this year as it prepares for privatization.” See “Kenya’s state-owned Telkom to retrench 12,000 workers.” People’s Daily Online, March 17,

2006. http://english.people.com.cn/200603/17/eng20060317_251527.html 11 East African employment from airtime sales in December 2006 was estimated at 90,000 in Kenya, 14,250 in

Rwanda, 60,000 in Tanzania and 30,450 in Uganda. See: GSM Association. 2007. Taxation and the growth of

mobile in East Africa.


21

The quality of the services provided over the network varies significantly and is highly dependent on

connectivity. Overall, the level of international Internet connectivity is quite low. As a whole, the

countries had some 56 Gbps of international bandwidth in 2007, of which almost three quarters was just

in two countries: Egypt and Morocco. Per-capita bandwidth is also low compared with other low-income

countries; the median bits per capita of the 24 Sub-Saharan African countries is only one-third that of the

average for Bangladesh, India, and Pakistan. Countries with access to undersea fiber-optic cable networks

have a significantly higher per-capita bandwidth than those without. Apart from South Africa, Senegal is

notable for its relatively ample international bandwidth (figure 1.26).

Figure 1.25 Mobile subscribers per employee, Vodacom networks

Source: Vodacom, Vodafone.


22

Figure 1.26 International Internet bandwidth (bits per person), 2007

Source: AICD.

The limitation in international Internet bandwidth is widely acknowledged as a critical impediment

for Africa: “Bandwidth is the life-blood of the world’s knowledge economy, but it is scarcest where it is

most needed—in the developing nations of Africa which require low-cost communications to accelerate

their socioeconomic development.”12

12 Association for Progressive Communications, Open Access: Lowering the costs of International Bandwidth in

Africa. APC Issue Papers Series. October 2006. http://rights.apc.org/documents/open_access_EN.pdf


23

2 Institutional framework

Institutional frameworks are crucial for achieving a sustainable, improved performance in the ICT sector.

This section describes the relationship between institutions in the ICT sector with indicators of

performance. The focus is on three major building blocks of the institutional framework: market reform,

regulation, and governance.13

A scorecard was developed to allow cross-country comparisons of institutional reform. The scorecard

assigned a score to the three different dimensions of reform, with a cumulative 1 being the top score.14

Figure 2.1 shows the scores for each of the 24 countries studied.

Figure 2.1 Telecom institutional score for 24 African countries

Source: AICD.

Compared with other infrastructure sectors, there has been intensive institutional reform. These

changes have been predominantly driven by market reforms that fostered competition and facilitated

various forms of private participation. Policy oversight evolved accordingly.

Progress is also evident on the regulatory front, although the picture is more mixed. The majority of

countries have set up independent regulators, although in some cases, governments continue to interfere

with their decisions. There is significant room to improve transparency and accountability.

Governance has received less attention from policy makers, which is why the scores are generally

lower in this area. Although management and board autonomy (and perhaps disclosure of standards) have

13 Many of the examples in this section use analysis carried out on the original 24 countries of AICD’s first phase.

The doubling of AICD’s coverage in phase 2 may affect the analysis in some respects. 14 Admittedly, this has an ex-ante judgment value of every aspect of institutions.


24

improved, many incumbents remain in public hands. This has severely restrained labor-related decisions

and created indiscipline in the capital markets.

Sector reform

Numerous studies have found a link between reform of the telecom sector and performance.15 Reform

requires new sector legislation, market restructuring, policy oversight, and private-sector participation

(figure 2.2).

Figure 2.2 Telecom reform score for 24 African countries

Source: AICD.

Most of the African countries in the study have enacted laws and regulations affecting the telecom

sector. This has typically included establishing a NRA while making general provisions for competition,

licensing, interconnection, scarce resources (such as numbering and spectrum), and pricing. This basic

law is usually supplemented by decrees, resolutions, or other legal documents dealing in more detail with

specific issues. Over 20 countries have issued new laws affecting the telecom sector since 2000.

Some regional organizations—for example, the Common Market for Eastern and Southern Africa

(COMESA 2007), Economic Community Of West African States (ECOWAS),16 and Southern African

15 For example, Fink, Mattoo, and Rathindran examine the impact of privatization and competition. See www-

wds.worldbank.org/servlet/WDSContentServer/WDSP/IB/2002/11/11/000094946_02102904035023/Rendered/PDF/multi0page.pdf.

Wellenius looks at privatization, the creation of a regulator, and competition, arguing that all three consolidate sector

reform. See http://rru.worldbank.org/Documents/PublicPolicyJournal/130welle.pdf. Wallsten argues that the sequencing of the

reforms is important. See www.inomics.com/cgi/repec?handle=RePEc:wbk:wbrwps:2817. 16 See ITU. West African Common Market Project: Harmonization of Policies Governing the ICT Market in the

UEMOA-ECOWAS Space. Model ICT Policy and Legislation. Available from www.itu.int/ITU-D/treg/projects/itu-

ec/Ghana/modules/FinalDocuments/Model_ICT_Law_Policy.pdf


25

Development Community (SADC)—have model laws that countries can use in writing their legislation.

These model laws have been instrumental in encouraging governments to revise outdated sector

legislation, even if not all countries have adopted the model laws in a comprehensive way.

Over the past few decades, perspectives on how to govern the telecommunications sector have

changed radically. The most important aspect of this is a change in the perception that the ICT market is

one that should be run as a monopoly to one that should be competitive. The successful introduction of

competition has also had implications for state-owned incumbent operators that have lost market share

and have increasingly become a drain on public finances.

Communications technology has also evolved rapidly, which has itself driven policy reforms. For

example, the dominance of wireless-based networks and the convergence of services made possible by

digitization all have implications for radio-spectrum, numbering, and universal service policy.

Insufficient reforms and ongoing government restrictions on private companies are emerging as the

main hurdles for increasing the availability of ICT services. A recent study concludes that private

investment in the GSM market has the potential to expand the GSM footprint to cover the majority of the

population of Africa, despite low income levels. A key reason for cross-country variation in GSM

network coverage is in the extent to which countries have reformed. Countries that have been slow to

reform have a larger “efficiency gap” than countries that have not (figure 2.3).

Figure 2.3 Inverse relation between untapped GSM coverage and advances in ICT reform in 24 African countries

Source: AICD.

This finding has clear policy implications: attention should be paid to increasing private investment

and improving the conditions for competition among operators. Public resources should be targeted at the

small percentage of the population lying outside the areas of commercial viability for the GSM networks.


26

Reform and performance: A first broad approach

A simple country typology is used to see how reforms have affected performance. According to this

typology, the 24 countries are classed as liberalized if they established a NRA, privatized incumbents, and

opened the mobile market to competition—reforms that have to have been in place for more than five

years. The length of time from the introduction of reforms is seen to be important because reforms take

time to have an effect. According to this definition, only six countries were deemed liberalized: Côte

d’Ivoire, Ghana, Senegal, South Africa, Tanzania, and Uganda.17

This classification is used to analyze sector performance by comparing liberalized and non-liberalized

countries from 2000 to 2005.18 The most significant determinant of sector outcomes is gross domestic

product per capita so the exercise compares the difference between the actual performance of the country

with the performance that we would expect, given its average per capita income.

This analysis (Figure ) shows that countries that liberalized had significantly higher rates of mobile

penetration than those that did not. Countries that pursued early liberalization had an average penetration

level 2.2 points above their income levels in 2005. Those that did not had average access levels 2.1 points

below their income level. A closer inspection of the data reveals some interesting trends. Among

countries that had not introduced mobile competition, all were performing below expectations considering

their per capita incomes. Indeed, their performance worsens each year, and their penetration was 9.2

points below their expected performance.

Figure 2.4 Difference between expected and actual fixed penetration

Figure 2.5 Difference between expected and actual mobile penetration

Source: AICD.

Note: Liberalized countries refer to those that have established an independent regulatory agency, partly privatized government-owned operators, and introduced competition for at least five years.

17 Of those six, five have made the additional reform of a World Trade Organization telecommunication

commitment, enshrining liberalization through binding international obligations. A list of countries making commitments is available at www.wto.org/english/tratop_e/serv_e/telecom_e/telecom_commit_exempt_list_e.htm. 18 The performance measures are based on the indicators used by in: Bressie, K., M. Kende and H. Williams. (2004)

Telecommunications trade liberalization and the WTO. Paper presented to the 15th ITS Biennial Conference,

'Changing peoples, societies and companies: Telecoms in the 21st Century', Berlin, 5–7 September 2004.

http://www.harriswiltshire.com/Telecommunications%20Trade-%20Liberalization%20and%20the%20WTO.pdf.


27

The result of a similar analysis for the fixed sector shows the opposite result (figure 2.4). Countries

that liberalized their fixed markets had slightly lower penetration rates than countries that did not. The

underlying reason for this is that when the fixed market is liberalized, average tariffs rise as tariff

structure is rebalanced (figure 2.6), reducing the number of fixed subscribers. A second effect that

happens during liberalization is that as the mobile networks expand, customers substitute from fixed

networks to mobile networks, which offer the same and sometimes better functionality. This is evident

when comparing fixed and mobile tariffs as a percentage of GDP per capita between the two groups of

countries (figure 2.7).

Figure 2.6 Tariff rebalancing in African countries with a liberalized telecom sector

Figure 2.7 Fixed and mobile price packages as percentage of GDP per capita, 2005

Source: AICD.


Sector liberalization has a positive effect on the total size of the ICT industry. The revenue generated

by the telecommunications sector in the liberalized countries is 5.6 percent of GDP, compared with 3.5

percent in the non-liberalized countries (figure 2.8). The industry is also growing faster in countries that

have liberalized than in countries that have not. Telecom revenues as a percentage of GDP rose by 2.5

percentage points between 2000 and 2005 in the liberalized countries, compared with just 1.2 percent

points in the non-liberalized countries.


28

It is worth noting that

telecom revenue as a

percentage of GDP did not

increase significantly in the

first three years post

liberalization. This was

because competition drove

down prices as demand

was expanding, so the net

effect on total revenue was

small. However, in

subsequent years, the

growth of the market

outweighed the declines in

prices, so total revenue

rose.

Competition (market restructuring) and performance

Understanding the links between competition and performance starts with distinguishing nuances in

the legal reforms designed to encourage competition. There are few outright bans on competition in the

telecom sector in the countries that were studied. In fact, most national legislation in the countries under

discussion introduced some degree of competition in their telecom markets (table 2.1). Half the countries

allow competition in the fixed-line and international markets and only one has yet to open mobile markets

to competition. The provision of Internet access has also been liberalized in most countries. However,

there is often a difference between de jure and de facto liberalization. Only a few countries in the sample

have more than two operators in the fixed market segment, and barely half the sample has more than three

active mobile operators. In the Internet segment of the market, many ISPs are not allowed to obtain

international bandwidth directly so there is competition on paper, but this is severely constrained.

Table 2.1 Status of telecommunication market competition, African countries, 2007

Legal status of competition Number of operators

M P C NA 1 2 >2 NA

Mobile 3 18 27 5 7 18 27 1

International 21 15 16 1 ... ... ... ...

Internet 6 6 34 1 6 0 32 15

Local fixed 24 11 14 4 37 10 5 1

International gateway 9 11 12 21 ... ... ... ...

Source: Adapted from ITU and regulator Web sites.

Note: The table shows the number of countries in each category (i.e., in two countries the legal status of mobile is a monopoly, and in three countries there is only one mobile operator). M=Monopoly, P = Partial competition, C= Competition, NA=Not available.

The gap between the number of countries that have liberalized their telecommunications markets on

paper but not in practice arises primarily through the licensing process. Namibia and Zambia are both

Figure 2.8 Telecom revenue as percentage of GDP, selected African countries

Source: AICD.



29

examples of countries that have legislation that provides for competition in the fixed-line and

international gateway facilities markets, but the failure to issue licenses for new entrants has prevented

competition from taking off.

In other countries, licenses may be available, but the complexity of the process can discourage market

entry. For example, many countries have license frameworks that segment their telecom markets,

requiring operators to hold multiple licenses. Sometimes it is not clear which licenses are necessary to

provide which service, or whether the scope of a license allows the licensee to provide the services it

wants to. The convergence of the ICT sector leads to the ability to provide multiple services over a single

network. Ideally, this should be reflected in the regulatory environment through a simplification of the

licensing process such that one license allows any type of service to be provided. This is the situation in

Tanzania with the introduction of the Converged Licensing Framework (CLF) in 2005. There are now

four license categories: network facilities, network services, applications services, and content services.19

As of December 2007, eight network facilities licenses had been issued under the new regime, which

allows licensees to offer any facilities-based telecom service. One outcome of the new license framework

is that both the incumbent operator (TTCL) and a new operator (Benson Informatics) were allowed to

launch mobile services in 2007 without having to obtain additional licenses.

License fees can also present a significant barrier to potential entrants, particularly domestic

companies interested in entering the telecom sector. Some countries charge relatively large sums for some

licenses while they require operators to have a number of licenses for different market segments, which

adds to the cost of doing business. When license fees are combined with other regulatory charges, such as

universal service contributions and spectrum usage fees, the total can be a significant barrier to market

entry. For example, an international voice license in Zambia costs US$12 million, with the result that the

incumbent remains the sole facilities-based provider for international telephone calls. According to the

United Nations Conference on Trade and Development (UNCTAD):

International call costs in Zambia are among the highest in the region, not all connections (incoming and

outgoing) are successful and calls are often of poor quality. This has been frequently cited by investors as

contributing to the high cost of doing business in the country. All international calls are currently routed

through an international gateway operated by ZAMTEL. However, this gateway is unable to provide for the

required traffic because of a lack of investment in equipment and the fact that ZAMTEL has no competition which could provide the incentive to do so.20

Revisions in license fees affect existing operators by creating market uncertainty. The government in

Benin ordered mobile operators to pay an additional retroactive license fee of around US$50 million. If

the fee is not paid, the licenses can be revoked.21

Competition in fixed-line markets

Fixed-line retail telephone service is one of the region’s least competitive segments of the ICT sector.

This is partly because of the exclusivity periods granted to incumbent operators, but it is also the result of

19 See “Licensing Information” on the TCRA Web site at www.tcra.go.tz/licensing/licensing.php. 20 UNCTAD. 2007. Blue Book on Best Practice in Investment Promotion and Facilitation: Zambia. Available from

www.unctad.org/Templates/webflyer.asp?docid=8183&intItemID=1397&lang=1&mode=downloads. 21 “Benin: Beninois Regulator Withdraws Operating Licenses from MTN and Moov.” Global Insight Perspective.

July 17, 2007. www.globalinsight.com/SDA/SDADetail9935.htm


30

a perception that incumbents have de facto control over fixed-line markets. Most of the countries have

ended both the exclusivity arrangements and the de jure restrictions, but effective competition has not

developed.

Where exclusivity arrangements have ended for at least two years, only about half of these countries

have seen new operators or a licensing regime that supports an open market. For example, Telkom South

Africa’s exclusivity ended in May 2002, but a second national operator’s license was not issued until

December 2005 and did not commercially launch until August 2006. In Namibia and Zambia, no

exclusive arrangements exist, but the incumbent operators enjoy a de facto monopoly. These are examples

of countries that have liberalized their markets in law but that retain implicit and explicit regulatory

barriers to the development of competition.

Elsewhere, exclusivity is written into licenses and governments are not able to introduce competition

into the fixed-line market. Sometimes these exclusivity arrangements are not publicly available, leaving

government policy on sector structure unclear. Other countries have no licensing regimes, or their regimes

are so complex and costly that they have discouraged new market entrants. 22

Only sixteen countries had more than one fixed-line operator providing service in 2007 (table 2.2),

and new fixed-line operators have gained significant market share only in Morocco and Nigeria.

22 In Kenya, the three main reasons put forth for the delay in the launch of local loop operators are regulatory delays,

interconnection obstacles, and a lack of investment capital. See www.cck.go.ke/llo_consultative.pdf.


31

Table 2.2 Status of fixed-line competition, 2007

Country / Operator Subscribers Date Market share HHI Note

Algeria

Algérie Télécom Incumbent

Lacom

Joint venture of Telecom Egypt and Orascom awared a license in 2006. Was to be divested in 2007 until regulator adapted license conditions.

Angola

Angola Telecom Incumbent

Mundo Startel

44% owned by Telecom Namibia. Construction of an NGN IP network with commercial operations slated for 2008.

Congo D.R. 10,579 Dec-05

OCPT No data provided to regulator

Congo Korea Telecom 884 Includes ISP subscribers and fixed wireless users

Sogetel 9,695 Includes ISP subscribers and fixed wireless users

Cote d’Ivoire

CI Telecom Incumbent

Arobase Telecom

Ghana 376,509 Dec-07 100 9,857

In April 2006, invitations to bid on providing regional fixed telephone services but currently on hold.

GT 373,798 99 9,857 Incumbent

Westel 2,711 1 1

Kenya 339,199 Jun-07 100 9,437

Second National Operator announced in October 2006 but later cancelled

Telkom Kenya 329,358 97 9,428 Incumbent

Others 9,841 3 8 19 licensed "local loop operators" with limited scale and scope

Madagascar

Telma Incumbent

Gulfsat

Mauritania Dec-07 100 9,418

Mauritel 97 9,409 Incumbent

Chinguitel 3 9 Since August 2007 (owned by Sudan Telecom)

Mauritius 361,300 100 8,417

Mauritius Telecom 330,000 91 8,342 Incumbent

MTNL 31,300 9 75

Morocco 2,393,767 Dec-07 100 4,945

Maroc Telecom 1,273,675 53 2,831 Incumbent

Meditel 19,790 1 1

Wana 1,100,302 46 2,113 Limited mobility wireless

Nigeria 1,545,984 Mar- 100


32

Country / Operator Subscribers Date Market share HHI Note

08

NITEL 59,850 4 Incumbent

Others 1,486,134 96 17 fixed including fixed wireless

Rwanda 22,291 Sep-07 100 9,570 MTN awarded fixed line license in 2006 but has not entered market.

Rwandatel 21,801 98 9,565 Incumbent

ARTEL 490 2 5 Rural infrastructure provider

South Africa

Telkom Incumbent

Neotel Granted license in 2005, started providing business services in 2007 and consumer market in 2008.

Underserviced Areas Licensees (USALs)

Licenses for underserviced areas with a teledensity of less than 5%. 27 areas identified with 7 licenses awarded.

Sudan 636,905 Dec-06 100 6,410

Sudatel 487,584 77 5,861 Incumbent

Canartel 149,321 23 550 License awarded in 2004

Tanzania 146,419 Mar-07

Fully liberalized but no entry to the fixed line market due to the cost effectiveness and convenience of mobile services

TTCL 146,419 Incumbent

Zanzibar Telecom

Although awarded "fixed" license, using GSM-fixed wireless and subscribers counted as mobile. Also, until 2006, only operated on island of Zanzibar.

Uganda 100,777 Dec-05 100 7,195

UTL 83,777 83 6,911 Incumbent

MTN 17,000 17 285 Second National Operator

Source: AICD.

One of the most competitive fixed-line markets in the region is Nigeria, where growth has been

strong. One reason is that Nigeria allows fixed wireless operators to offer limited mobility services.

Elsewhere in the region, fixed lines have seen steady if unimpressive growth. After Sudan introduced

mobile competition, there was a rapid shift to mobile, and in 2005 the number of fixed lines fell

dramatically after massive disconnections. Similarly, in South Africa, despite the introduction of a

prepaid fixed-line pricing platform, fixed lines fell by some 250,000 since 2000 (figure 2.9).


33

Despite the poor performance of the fixed-line market in to date, the prospects have been improved as

a result of some newly developed technologies. For instance, wireless local loop (WLL) products,

typically based on CDMA2000 1x technology, have led to renewed interest in fixed markets. One feature

is that when used in lower frequencies (for example, at 450 MHz), WLL systems have wide transmission

abilities suitable for rural and remote areas. According to the CDMA Development Group, over 30

African countries studied had commercially deployed a CDMA2000 1x wireless network by mid-2007.

Some operators have added features like limited mobility and free inter-network calling in an effort to

attract customers. There are regulatory implications for the mobility aspects of fixed wireless that will

likely intensify if these platforms succeed.23 The WLL networks often come with billing platforms that

support different price plans and prepaid packages.

WiMAX (Worldwide interoperability for microwave access) is another promising technology being

investigated by both incumbents and new fixed-line market entrants including mobile operators with

fixed-line licenses. WiMAX is a wireless broadband technology over which telephony can also be

provided using VoIP. Over a dozen African countries have launched WiMAX networks.

The de jure and de facto of mobile markets

More and more mobile networks are being deployed in the region, a process that is deepening

competition. As noted, most countries have allowed de jure entry and licensing of mobile operators. Not

23 For example, in Lesotho, the incumbent fixed-line operator was ordered to limit the range of its fixed wireless

offering to one cell site. See www.lta.org.ls/Consultations/Orders/Order4_LekomoFlexi.pdf. In Namibia, the two GSM operators

have complained to the regulatory authority about the mobility features of the incumbent’s fixed wireless service.

See www.telecom.na/index.php?go=news&sel=view&nid=52.

Figure 2.9 Net change in fixed lines, Sub-Saharan Africa

Source: AICD.


34

surprisingly, by 2007 all the countries possessed a mobile network, 46 countries had more than one active

operator, and more than half had three or more active operators. In 1993, by way of stark contrast, two

thirds of the countries in the study had no mobile network, and those with mobile networks functioned as

monopolies (figure 2.10).

The de facto competition of the mobile segment cannot be established by counting the active

operators. Market concentration is often measured with the Herfindahl-Hirschman Index (HHI).24 A HHI

of 10,000 indicates a monopoly; the lower the HHI score, the more diluted the market power as exerted

by one company/agent.

In a perfectly competitive market, one would expect a strong negative correlation between the HHI

and number of operators. But cases like Burundi and Zambia are not uncommon. Burundi has the same

number of active operators as Nigeria, Benin, and Ghana, but its HHI scores are higher. The same is true

of Zambia vis-à-vis Uganda, Burkina Faso, and South Africa (figure 2.11): the same number of operators

but higher HHI scores.

24 The HHI is “a commonly accepted measure of market concentration. It is calculated by squaring the market share

of each firm competing in the market and then summing the resulting numbers. For example, for a market consisting

of four firms with shares of 30, 30, 20, and 20 percent, the HHI is 2600 (302 + 302 + 202 + 202 = 2600).” See

www.usdoj.gov/atr/public/testimony/hhi.htm.

Figure 2.10 Status of mobile competition, 1993–2006

Source: AICD.


35

With four mobile operators and a HHI of 3,246, Nigeria has the least concentrated market, partly

because it awarded three digital GSM licenses at the same time to level the playing field. Although the

incumbent was awarded one of the licenses, its market position was hampered by an antiquated analog

mobile system that had few subscribers and insufficient capacity. A fourth license was issued several

years later.

Figure 2.11 Number of mobile operators and market concentration, 2007

Source: AICD.

Kenya, Mozambique, Senegal, and Sudan gave their incumbent operators several years’ lead time

before introducing competition. New entrants in their markets all struggled to gain market share.

Similarly, incumbents in Ghana and Uganda struggled to gain market share after being awarded mobile

licenses. Late entrants have been able to successfully gain market share by entering markets at a time of

technological transformation (from analog to digital) and when they are part of a multinational group.

Examples include the former Areeba (now MTN) in Benin and Ghana; Celtel (now Zain) in Burkina

Faso, Ghana, and Zambia; and Vodacom in DRC and Tanzania.

The number of operators active in the market as well as market concentration levels have important

effects on access and pricing behaviors. The year-to-year gains, measured in new subscribers, explode


36

after a second operator enters the market, particularly in higher-income countries that have fewer

affordability issues (figure 2.12).

Figure 2.12 Subscription increments, 2006

Source: AICD.

Active operators and mobile price levels are showing a positive trend. Prices decrease faster after the

entry of the second operator, but after an initial decrease, prices seem to occupy the narrow range from $8

to $15 for the monthly package. The exceptions among the 24 countries of AICD’s first phase are

Ethiopia and Madagascar at the low end ($3.4 and $4, respectively), where low income combines with

low penetration levels, and Cape Verde at the high end ($20), because of high demand. Given the income

and penetration levels, this should go down soon.


37

Figure 2.13 Mobile price changes after second operator enters the market

Source: AICD.

The causality between competition and performance is highlighted by measuring market

concentration. High competition (defined as HHI below 5,000) is clearly related to higher penetration and

GSM coverage as well as to lower prices (figure 2.14). Note, however, the very strong relation between

country income level and performance (figure 2.15).

Figure 2.14 Mobile access and competition levels, 2006 Figure 2.15 Mobile access and country income levels, 2006

Source: AICD.

The introduction of competition causes prices to rise temporarily and then to fall as HHI scores fall

below a certain threshold (table 2.3). In the initial stages of mobile competition (indicated by high HHI

scores), the market dynamics appear to push operators to focus on coverage instead of prices. In

monopolies, mobile prices seem low, but access is problematic, leading to an unsustainable equilibrium.

In fact, in cases where the incumbent becomes the first mobile provider, tariff rebalancing brings

competition to the fore. Mobile companies also tend to become profitable very quickly. The failure of the

fixed-line operators to provide adequate service tends to push prices up for mobile, with its much higher


38

quality, particularly where few operators are able to capture the consumer surplus. As the market matures,

however, it shifts to pricing issues as operators seek to attract consumers.

Table 2.3 Average mobile monthly price package and Herfindahl-Hirschman Index

HHI

Data in US$ 10,000 7,500–9,999 6,000–7,500 4,500–6,000 3,000–4,500*

Mean 12.96 11.17 16.20 15.13 12.08

Mean for low-income countries 10.82 11.17 16.20 15.56 11.97

Source: AICD.

Note: Minimum HHI for the sample equals 3,022

For all its recent activity, however, African mobile markets—with the exception of Egypt, Morocco

and South Africa—have not adopted a number of advanced features like mobile virtual network operators

(MVNOs), mobile number portability (MNP), and regulatory oversight of market dominance—

particularly of interconnection rates.

In South Africa, Virgin Mobile launched as an MVNO in 2005, using the infrastructure of Cell C, one

of South Africa’s licensed mobile operators. MNP allows users to retain their telephone numbers when

switching operators and leads to more competitive markets because users are less hesitant to switch

operators when they can keep their number. After several delays, the first MNP was finally launched in

South Africa in November 2006. By March 2007, there were 49,794 portings, and the newest operator,

Cell C, obtained the most new subscribers. Both Egypt and Morocco have subsequently introduced MNP.

Spectrum liberalization is another emerging wireless market trend. The need for spectrum is growing

along with the expansion of second- and third-generation mobile products and the emergence of WiFi,

WiMAX, and WLL. Countries are finding it increasingly difficult to price spectrum. Although auctions

can help, they can result in high prices, which are passed on to consumers. Auctions also take time to

organize, inhibiting operators’ needs to react quickly to market developments. In order to provide greater

flexibility, some countries are moving toward liberalized spectrum regimes whereby licensees can trade

spectrum bands. A number of regulatory challenges are involved, including ensuring that spectrum does

not get monopolized (enforceable by spectrum limits) and adopting policies for industrial, scientific, and

medical (ISM) band spectrum, which is typically unlicensed and used for applications like WiFi. The

countries under study in this paper have yet to reform their spectrum policies. A number of them,

however, have devised liberal policies for ISM band frequency—for example, by simply requiring an

authorization or issuing a license on demand.

Competition in long-distance and international gateways

Access to international submarine fiber-optic cables and competition are fundamental to driving down

international voice prices. In the absence of adequate regulatory controls, when an incumbent operator

controls access to the submarine cables, the full cost advantage of this technology is not passed on to

consumers. Multiple international gateways exert competitive pressure and push service prices much

lower than in countries where the submarine cable provides the only international gateway. In the case of


39

broadband and Internet access in general, access to fiber has a marked impact on prices. In contrast,

competition in international gateways has only a relatively modest impact on prices (table 4.4).

Table 4.4 Benefits associated with access to submarine cable

US$

Percentage of countries

Price per minute call within SSA

Price per minute call

to US

Price 20-hour per month dial-up

Internet access

Price ADSL broadband

Internet access

No access to submarine cable 67 1.34 0.86 67.95 282.97

Access to submarine cable 33 0.57 0.48 37.04 110.71

Monopoly on international gateway 16 0.70 0.72 37.36 119.88

Competitive international gateways 16 0.48 0.23 36.62 98.49

Source: World Bank.

The SAT-3 undersea fiber-optic cable has helped to alleviate the shortage of bandwidth for a number

of countries on Africa’s west coast. In addition, North African countries along with Cape Verde and

Sudan have been able to connect to other fiber-optic submarine cable systems. Although landlocked,

Ethiopia is sending an overland fiber-optic cable to Sudan to tap into that country’s fiber link to Saudi

Arabia. Some nations on the west coast that lack their own international fiber outlet are also using

terrestrial links to connect to neighbors with a SAT-3 landing station. For example, Namibia has a fiber

link to South Africa. East Africa has been particularly affected by a shortage of fiber-based international

Internet connectivity and, as a result, faces high retail prices. Most East African countries are

collaborating to create the East African Submarine Cable System (EASSy),25 which would provide high-

speed fiber-optic connectivity at lower costs. Progress toward EASSy, however, has been hampered by

governments keen to ensure that the system will provide access to those outside the consortium.

Competition in Internet provision

The popularity of the Internet has resulted in growing demand, and most countries have issued

multiple ISP licenses. As a result, the Internet market segment has attracted the most entrants. Some

countries have relaxed their authorization regimes, requiring low or no license fees, so there are a number

of licensed or registered ISPs (figure 2.16).

25 For additional information, see the EASSy Web site, www.eassy.org/. Since this study was written, other

submarine cable initiatives have begun. In East Africa, SEACOM is already up and running. In West Africa, Glo is

now functioning.


40

Figure 2.16 Number of ISPs, 2007

Source: AICD.

Despite the large number of ISP licenses that have been issued, many of the countries have imposed

restrictions on what ISPs can do. For example, ISPs are often not allowed to provide their own

infrastructure unless they obtain other licenses. Restrictions on entry into the fixed-line and international

gateway markets have meant that ISPs have often had to lease infrastructure from incumbent operators

sometimes at prices that are not cost based. This is particularly onerous for international bandwidth,

especially in landlocked countries that rely on satellite. Even in countries with other options for

international connectivity (such as undersea fiber-optic networks), incumbent operators often have a

stranglehold on landing stations and belong to consortiums that own the networks.26 This means that ISPs

have no choice but to go through the incumbent operator for fiber-based international connections. Of the

14 AICD countries for which the International Telecommunication Union (ITU) had data, most do not

26 Commenting on the SAT-3 fiber-optic cable that runs up the west coast of Africa, one reporter notes: “As the

gatekeepers to the international connection, the SAT-3 consortium has maintained high surcharges for any other

telcos or organizations that want to use the connection, crippling competitors and keeping customers hostage.” http://newsroom.cisco.com/dlls/2006/ts_053106.html


41

allow full competition in international gateways.27 Six reported that international gateways were a

monopoly of the incumbent, four reported partial competition; only four responded that full competition

existed in this market segment.

Besides the obvious virtues of the Internet for education and information development, the

proposition of making telephone calls via Internet is attractive. From the viewpoint of cost, voice over the

Internet compares very favorably to a conventional public switched telephone network. But the status of

Voice-over Internet Protocol (VoIP) telephony is far from clear in the countries studied.

Reforming by allowing private participation

With the exception of Comoros, Djibouti, Eritrea and Ethiopia, all of the countries studied allow

foreign investment in their telecommunication sector. And most allow foreigners to have at least 51

percent ownership. A number of countries have gone further, allowing foreign investors to have complete

ownership of subsidiaries. The vast majority of investments in ICT are greenfield projects (i.e.,

investments in new businesses, rather than investments made as part of privatizations), particularly new

operations in mobile communications. New asset investment was in the order of US$20 billion, while

incumbent divestitures—which have been the most visible and controversial form of PPI—accounted for

only US$3.3 billion.28

By the end of 2007, just over half of the African countries had sold shares in their incumbent

telecommunications operator to the private sector. Most privatizations have been to strategic investors,

but there are some notable exceptions. In Sudan, the government has released its holdings on the

Khartoum and regional stock markets in several sales since 1993.29 The government of Kenya recently

sold 25 percent of its 60 percent stake in Safaricom, the leading mobile operator, through an initial public

offering on the Nairobi Stock Exchange. The total value of incumbent privatization transactions between

1993 and 2008 was just short of US$13 billion, of which northern Africa nations and South Africa

accounted for almost three quarters.

After purchasing stakes in a number of incumbent operators in the late 1990s and early 2000s,

developed-country investors largely withdrew from telecom privatizations in Africa. Recent privatizations

have either been public offerings (e.g., Egypt, South Africa and Sudan), sales to developing-country

investors (e.g., ZTE of China in Niger and Maroc Telecom in Burkina Faso, Mauritania and Gabon), or

sales to domestic investors (e.g., Malawi and Nigeria). One barrier to incumbent privatization has been

the high asking prices. Resistance to foreign ownership of key strategic assets is also a factor. Indeed,

some governments such as Ghana, Guinea and Rwanda have renationalized by repurchasing shares in

incumbent operators. In Tanzania, the government signed an agreement for a Canadian company to

manage Tanzania Telecom. A couple of recent high profile transactions may signal the return of

27 According to the “Level of Competition” information extracted from the ITU ICT Eye available from

www.itu.int/ITU-D/icteye/Regulators/Regulators.aspx# [Accessed August 2, 2007] 28 This is based on the World Bank’s PPI (Private Participation in Infrastructure) database http://ppi.worldbank.org/. The

PPI database lists 82 transactions between 1992 and 2005 for the countries studied. 29 Although not strictly a strategic investor, ETISALAT, the incumbent operator in the UAE, owned 5 percent of

SUDATEL’s shares at the end of 2005.


42

developed country investors. These include the sale of 51 percent of Telkom Kenya to France Telecom in

December 2007 and the sale of 70 percent Ghana Telecom to Vodafone of the UK in August 2008.

But, as mentioned before, the largest volume of private investment transactions have been for new

greenfield operations. Of these, most have involved the sale of licenses to private investors for the

creation of new mobile operators. Most mobile operators are controlled by one of the multinational firms

operating in the region (table 2.5). As result, almost all of the countries have at least one strategic foreign

investor present in their mobile sector and these strategic investors account for over 80 percent of mobile

subscribers in the region.

The increasing influence of large

pan-African mobile groups is also

powered by their participation in

privatizations and acquisitions of

existing incumbent platforms. For

example, some of France Telecom’s

properties were the result of buying

incumbent fixed-line operators that

had mobile operations (for example,

Côte d’Ivoire Telecom and Sonatel of

Senegal). MTC’s ownership is the

latest in a line of acquisitions

involving its African subsidiaries.

MTN doubled its holdings when it

acquired the mobile operations of

Lebanon’s Investcom in 2006.

The fact that large groups now

dominate the market is a sign of the

desirability of the African mobile sector and other regions where large mobile groups own a number of

subsidiaries (for example, Telefonica and America Movil in Latin America and Singapore Telecom in

Asia, to name two). The benefits of strategic groups owning mobile operators were highlighted in 2005

by the Zambian Competition Commission (ZCC), which approved the sale of local mobile operators to

MTC and MTN.30 As noted by the ZCC, strategic mobile investors offer many benefits, including access

30 “But the board noted that mere acquisition of a dominant market position was not anti-competitive per se if

acquired through efficiencies such as better technology, low operational costs, high turnover due to better innovative

marketing techniques, superior branding and highly trained technical staff. ‘With the proposed acquisition of the two

leading mobile telephone operators in Zambia, there are likely significant synergies to accrue that would be used to

develop the telecommunications industry,’ Mr Lipimile said. He also said that benefits such as increased investment

as well as technical and other economies of scale were likely to accrue to the two operators with other envisaged

consumer benefits. Among the consumer benefits to accrue include the efficient and real time internetwork short

message system at marginal rates and lower tariffs as a result of lower costs of operations and interconnection fees. Also envisaged are more widespread and reliable international roaming possibilities where subscribers would not

need to buy a simcard in each country they visited but could still use their Zambian simcard to communicate.”

“Zambia Competition Commission (ZCC) approves Celtel, Telecel takeovers.” Times of Zambia. July 27, 2005. http://rights.apc.org/africa/index.shtml?apc=s21819e_1&x=554587

Table 2.5 Strategic mobile investors, 2007

Source: AICD adapted from company reports.


43

to capital,31 know-how, and group-wide purchasing strategies to lower costs and increase roaming

opportunities.

Given the high penetration of mobile communications in the region, roaming agreements—which

allow users in one country to use their mobile phones in another—are to be encouraged. One welcome

development is therefore the launch of Celtel’s One Network, which allows mobile users in the East

African countries of Kenya, Tanzania, and Uganda to use their mobile phones in the other countries and

pay local rates, subject to taxes (Celtel International 2006). MTN Uganda, Safaricom in Kenya, and

Vodacom subsequently cooperated to launch a similar service, called “Kama Kawaida.” More such

roaming arrangements are bound to appear. Pan-African mobile operators also sometimes offer cheaper

overseas calling rates to their home subscribers to group subscribers in other countries. For example, in

July 2007, MTN Rwanda charged Frw 354 per minute for calls to MTN operations in other countries,

compared with a rate of Frw 413 per minute to non-MTN destinations in Africa (excluding East Africa).

Divestiture and privatization have changed the fixed-line industry structure and performance. Two

patterns are emerging. In terms of access, fixed-line subscription doubled or more than doubled when the

public incumbent was privatized. For fixed-line prices, horizontal integration of the incumbent toward a

competitive mobile market results in tariff rebalancing and, consequently, significant tariff increases in

the short run. When the incumbent integrates toward a noncompetitive mobile market, the dominant

government response has been to reduce prices through cross-subsidies that are mostly aimed at predatory

practices (table 2.6).

Table 2.6 Fixed-line subscriptions and price by privatization status and mobile HHI, 2006

Fully or partially privatized incumbent

fixed-line operator

State-owned incumbent fixed-line

operator

High mobile competition 2.64 0.68 Fixed-line subscribers, number per 100 people Low mobile competition 2.70 1.61

High mobile competition 20.24 8.78 Price of fixed-line monthly package, USD Low mobile competition 11.75 14.88

Regulation

Practically every country studied has created a national regulatory authority (NRA) responsible for

the telecommunications sector. The first were created in 1994. By 2007, 44 AICD countries had NRAs in

operation. But effective NRAs depend on a regulatory framework that is accountable, transparent, and

autonomous. Progress on regulation differs significantly from country to country, but overall the level of

autonomy of the NRA is quite low across the board (figure 2.17)

Autonomy is measured—among other things—by financial independence and evidence of

government noninterference. Most NRAs in the region are financed through levies on telecom operators

(such as license and spectrum fees, special taxes on revenues, fines, and penalties). This generally assures

31 For example, MTC recently obtained the largest local loan ever raised in Tanzania for the expansion plans of its

mobile subsidiary (US$70 million). www.celtel.com/mobile/en/news/press-release39/index.html


44

a reliable source of income given the rapid development of telecommunication networks in the region. In

not a few instances, however, the funds the regulators collect must be remitted to the government, with

the NRA financing met by central government allocations. This increases the unpredictability of funding

while making NRAs subject to political interference.

Figure 2.17 Telecom regulatory score

Source: AICD.

Some observers believe that multisectoral NRAs are less susceptible to interference. Whether or not

this is true, most NRAs in the region are primarily responsible for telecommunications and occasionally

postal services.32 The Gambia, Mauritania, Niger and Rwanda have multisector regulators responsible for

other utilities, such as electricity and transport. One global trend is the evolution toward converged ICT

sector regulators covering telecommunications, broadcasting, and information technology services.

Converged regulators have been created in Australia, Brazil, Finland, Italy, Malaysia, Singapore, and the

United Kingdom.33 These regulators are better suited for environments where telephone, broadcast, and

data networks each offer a variety of electronic services. Among the countries studied, several have

32 Four regulatory institutional models have been identified: “…single-sector regulator whose sole function is to

oversee the telecommunications sector…The second design is known as the “converged” regulator, meaning those

regulatory entities that oversee a broader range of services which, in addition to telecommunications, also include

information and communications technologies, including broadcasting…The multi-sector regulatory

authority…usually encompasses various industry sectors that are considered public utilities, e.g.,

telecommunications, water, electricity, and transportation. The fourth category is not a regulatory authority per se, but an approach in which general competition policy is the main method of overseeing the telecommunications

sector…” See The ICT Regulation Toolkit, “Overview and Comparison of Different Institutional Designs.”

Available from www.ictregulationtoolkit.org/en/Section.2033.html. 33 Victoria Shannon. “Communications regulators ‘converge’ with the times.” International Herald Tribune.

December 10, 2006.


45

moved in this direction. The first was the Independent Communications Authority of South Africa

(ICASA), regulator of telecommunications and broadcasting, established in July 2000. Tanzania went this

route in 2003 with the Tanzania Communications Regulatory Authority (TCRA), which merged the

Tanzania Communications Commission and the Tanzania Broadcasting Commission. The Nigerian

government is also exploring the creation of a converged regulator.

Accountability in the telecom sector is making promising progress, particularly if compared with

other aspects of regulation and other infrastructure sectors. But transparency is a mixed and incomplete

story. Almost all the NRAs have Web sites with the potential to increase transparency by publishing

information about the sector. Yet the availability and quality of information is uneven. A review of

telecommunications regulatory Web sites in Southern Africa (Mahan and Melody 2003) found that,

“Apart from providing information relating to legislation, there is no other category in which all of the

surveyed NRAs fulfill [the criterion] on their Web site. By looking at the content, functionality, usability,

and design of the … surveyed Web sites, substantial differences are noted.”34 The study notes that a best-

practice Web site:

Creates a reliable, comprehensive and up-to-date repository of regulatory information. This

reduces the burden of stakeholders having to search for new regulatory information from different

sources, and makes reliable national level information accessible to potential investors and

stakeholders;

Helps users and stakeholders to understand regulations, rules and regulatory processes; raises

awareness about regulatory compliance, rights and responsibilities—not only through mere

posting of regulatory acts and laws, but also in providing additional information which explains

the regulatory instruments, such as responses to frequently asked questions (FAQs); provides

access to information about further means of assistance and intervention such as public hearings,

contact information, regulatory structure and process information;

Makes forms for different regulatory processes accessible, helps channel official communication

to the proper departments or recipients, and overall facilitates the anticipated (and desired!)

compliance and reporting on the part of different stakeholders.

Voice-over Internet Protocol telephony is an important regulatory matter. The status of VoIP is far

from clear. Some countries ban it, others allow it, while its status is murky elsewhere: VoIP might be

legal for licensed telecom operators, and even if illegal for others, it is tolerated. This situation has a

negative incidence in the regulatory score overall. Part of these regulatory difficulties stem from the

various modalities of Internet telephony, such as computer-computer and computer-phone, and whether

licensed telephone operators and/or ISPs can provide them.

Universal service and access is an important regulatory aim. In practice, these have been difficult to

implement. This is particularly true in Africa, where incomes are low, infrastructure is limited, and the

regulatory framework imperfect.

Although many of the countries studied have defined universal service and access, they have varying

methods for achieving this goal. To keep service charges low, several countries allow explicit cross-

34 “Benchmark Indicators for African NRA Websites” (2004), http://lirne.net/resources/nra/NRAchapter5.pdf


46

subsidies for incumbent operators. This approach is losing ground, however, as operators become private

and move to reorient prices. To finance universal access, some countries have created a fund for operators

to give to. In some cases, the funds have yet to disburse payments. The funds provide direct support to

incumbent operators or to low-subsidy bids by any operator interested in providing service in an

underserved area. Nineteen countries reported the existence of a fund to expand telecommunications in

rural and other underserved areas.

A few countries have not been able to incorporate mobile operators into formal universal access

programs. Although mobile operators are generally required to contribute to universal service funds, they

are typically not obligated to roll out the network. One exception has been South Africa, where mobile

operators have universal access obligations. So South Africa has been at the forefront of linking mobile

licenses and spectrum needs to universal access obligations. When the initial GSM licenses were issued,

mobile operators were obligated to install a certain number of so-called community service telephones.

Additional spectrum was provided to the operators in exchange for the offer of 5 million free subscriber

identity module (SIM) cards to new customers. The operators were then required to provide Internet

access to 5,000 public schools in exchange for the provision of third-generation mobile spectrum.

Mobile operators in other countries are contributing to universal access with public phone schemes.

MTN in Rwanda and Uganda has launched village pay phone projects, modeled after the successful

Bangladesh program35 where microfinance is extended to rural inhabitants to buy mobile phones in order

to sell airtime to the public. MTN Rwanda’s community program had 3,300 subscribers in 2005. In

Uganda, MTN had installed 2,000 “VillagePhones” by August 2005; some require booster antennas and

solar or car-battery power where electricity is not available. Vodacom offered community service

telephones—95,000 were installed in South Africa by March 2007; 28,000 in the Democratic Republic of

Congo; 4,000 in Lesotho; and 10,000 in Tanzania.

Governance of state-owned enterprises

Governance is not progressing as well. In the midst of reform and technological innovation, half the

fixed-line operators in Africa remain in public hands. Although data are scanty, labor market discipline

appears poor, suggesting that public telecom utilities are being used as social buffers (figure 2.18).

Outsourcing bill collection and metering is not even in the radar screen of the countries. Only South

Africa and Sudan satisfied, on average, as many as half of the best-practice requirements.

35 The VillagePhone scheme in Bangladesh was spearheaded by Muhammad Yunus, founder of Grameen Bank and

winner of the 2006 Nobel Peace Prize. The Grameen Foundation is involved in both the Rwanda and Uganda

operations. See Telenor. “Deep admiration for Mohammad Yunus.” Press Release. 12 October 2006. Available at:

http://presse.telenor.no/PR/200610/1081182_5.html.


47

Figure 2.18 Telecom governance score

Source: AICD.

For countries with state-owned fixed-

line incumbents, public expenditures

reached on average 2 percent of GDP (table

2.7), an extraordinary figure in an

increasingly competitive market on a

dynamic technological path. In the rest of

world, telecom services are primarily

provided by the private sector, which also

bears the risks of cost recovery.

For those African countries with public

incumbents, cost recovery and productive

spending rely mostly on the governance

and operational efficiency of state-owned

enterprises (SOEs). More than 90 percent

of public expenditure is channeled through

the SOEs, and from that amount only 25

percent goes to capital. Capital investment is the most common proxy for productive spending, and the

fact that only a minor share of spending is productive raises stark questions about efficient spending,

particularly in a capital-intensive sector like telecom. This adds to the large labor bills of ICT SOEs.

Public utilities appear to be used as social buffers, redistributing wealth via excessive employment.

This is a sign of labor market indiscipline. The dollar value of labor redundancies—or the hidden costs of

excessive employment—is estimated to add up to 0.3 percent of GDP (Tanzania) or cost in excess of

Table 2.7 Public expenditure on telecom, annual average, 2001–06

Total expenditure (%

of GDP)

SOE expenditures on

total (%)

Share of SOE expenditure on

capital (%)

Benin 2.87 99.16 50.99

Cameroon 1.48 100.00 56.21

Chad 0.58 96.92 -

Ethiopia 1.55 77.55 41.14

Ghana 2.99 98.38 25.84

Kenya 2.28 100.00 14.99

Mozambique 1.70 98.70 —

Namibia 3.32 86.96 13.62

South Africa 2.11 96.46 29.47

Tanzania 1.07 95.91 —

Source: World Bank.

Note: Countries with fully or partly publicly owned telecom incumbent.


48

US$200 per subscriber (Chad) (table 2.8). These estimates take as a norm or point of reference the

number of subscribers per employees seen in OECD fixed-line operators.

Table 2.8 Dollar value of labor redundancies, annual average, 2001–06

Total telephone subscribers (fixed+mobile) per employee, last

year available Hidden costs, percentage of

GDP, OECD benchmark Hidden costs, US$ per

subscriber, OECD benchmark

Benin 621 0.01 2

Cameroon 730 0.00 0

Chad 127 0.12 202

Ethiopia 104 0.07 9

Ghana 563 0.11 20

Kenya 220 0.00 0

Mozambique 605 0.02 1

Namibia 470 0.12 14

South Africa 1,145 0.00 0

Tanzania 219 0.24 32

OECD benchmark 634

Source: World Bank.

These striking labor inefficiencies underscore the importance of external governance mechanisms.

The relation between increased governance and lower labor costs attributed to inefficiencies is a no-

brainer (figure 2.19)

The anatomy and impact

of institutions: emerging

patterns

Rapid technological change

has, as we have seen, spurred

greater competition among

providers. In many cases, cellular

telephones have become

substitutes for fixed-line services;

consequently, demand for fixed-

line services is in decline.

Massive restructuring of the

sector to allow for competition

and private participation has also made price regulation less relevant but has caused a temporary tariff

rebalancing. Competition and free mobile licensing have also shifted focus in the short run from price

competition to competition for expanding access. Private participation in the fixed-line monopolistic

segment of the market was very intense in the 1990s but stalled after 2001. In fact, a surprising number of

fixed-line incumbents remain in public hands. Thus, governance of SOEs remains a thorny issue.

Figure 2.19 Link between governance and cost of labor redundancy

Source: AICD.


49

Table 2.9 Succinct relation between institutional scores and key performance indicators

Penetration (subscribers per 100 people) Price service package (index) Efficiency

Correlation Mobile Fixed line Mobile Fixed line Untapped GSM

potential (% pop.)

Cost of labor redundancy

($/subscriber)

Reform 29.0 -7.1 24.6 59.9 -48.1 6.1

Regulation 35.0 -3.9 -7.1 34.1 -34.1 -23.8

Governance 58.5 47.5 -13.9 37.4 -38.8 -56.2

Source: World Bank.

Well-functioning institutions and efficient markets are rare on both the consumer and the producer

sides of the market. From a fiscal perspective, the more profitable and competitive the provision of

service, the larger the revenue collection (figures 2.20 and 2.22). This benefit is not well documented in

the literature. Still, it is very important in resource-strapped countries.

Figure 2.20 Link between reform and fiscal revenues Figure 2.21 Link between regulation and fiscal revenues

Source: World Bank.


50

3 The path to wider access to telecommunications

services

Telecommunications reforms have led to more competitive markets in many of the countries studied.

The result has been impressive growth during the first half of the 2000s, particularly in mobile telephony.

The challenge will be to sustain this growth in the face of significant barriers.

The key to extending communications access in Africa is to leverage the unprecedented success of

mobile technology on the continent. Given that mobile operators tend to have the largest

telecommunications networks in most countries, the incremental costs of extending mobile coverage into

underserved areas is probably less than that of other solutions, such as extending fixed-line networks or

promoting the voice-over-Internet protocol. A companion project to this study has estimated the cost of

extending mobile coverage to areas that currently do not have a mobile signal.

A number of key policy recommendations, if followed, would sustain growth and deepen access to

telecommunications in the region.

There is ample scope for further sector reform in most countries. According to a 2006 report from

the GSM Association, poor regulation has reduced telecommunications investment in Africa by

US$4.6 billion. Countries that have not yet privatized incumbent operators should do so in order

to reduce direct state intervention in operations, encourage a more level playing field, and attract

investment and innovation. Additional competition should be introduced by not limiting the

number of operating licenses available. Regulatory agencies should be strengthened and allowed

to operate independently.

Countries should pursue liberalization by simplifying licensing regimes, lifting remaining bars to

market entry, and examining the feasibility of introducing mobile number portability and mobile

virtual network operators.

Efforts should be increased to lower prices for telecommunications services. Average per capita

income in Sub-Saharan Africa was just US$970 in 2007—less than US$3 a day. Any incremental

efforts to lower prices would have a tremendous impact on affordability and hence access. A few

ways to push prices down are to lower taxes and termination fees, and, where competition is

limited, through regulatory action.

Mobile telephone access should be incorporated into established goals for universal access so as

to leverage the successful spread of mobile communications. Mobile telephony has probably done

more to increase access through a competitive environment than any other policy, yet, for the

most part mobile operators have not been involved in formal universal access programs. Adapted

universal access policies might require mobile operators to expand coverage as a condition of

licensing or allow mobile operators that expanded coverage to receive money from universal

service funds.

High-speed connectivity over fiber optic cable is a prerequisite for e-government and other

socioeconomically beneficial applications. Private-public partnerships can play a useful role in


51

developing and expanding national, regional, and international fiber optic links throughout the

region, allowing Sub-Saharan Africa to join fully in the global information society. Although

governments should play an active role in encouraging the deployment of fiber networks, their

participation should not delay the badly needed implementation of fiber backbones.