BP10_ICT_maintxt by Michael Minges

BACKGROUND PAPER 10 (PHASE II)

Information and Communications Technology in Sub-Saharan Africa: A Sector Review

Mavis Ampah, Daniel Camos, Cecilia Brice単o-Garmendia, Michael Minges, Maria Shkratan, and Mark Williams

JANUARY 2009

ÂŠ 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: feedback@worldbank.org 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: pubrights@worldbank.org.

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 SubSaharan 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.

Contents iv

Summary v vi vii viii x xi

What price access? From monopoly to competition Competition and performance Reform and regulation The persistence of state-owned enterprises The path to wider access to telecommunications services

1 Sector developments 1 8 19

Access Prices Quality

2 Institutional framework 24 43 46 48

Sector reform Regulation Governance of state-owned enterprises The anatomy and impact of institutions: emerging patterns

3 The path to wider access to telecommunications services

iii

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

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 middleincome 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. Ruralurban 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.

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

Access to the Internet is much less widespread than access to basic voice Source: AICD. 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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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.

Figure B Status of mobile competition

More than two dozen of the countries allow some degree of competition in fixed-line and international Source: AICD. 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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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, 2000â&#x20AC;&#x201C;05

Source: AICD.

vii

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA (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 Figure D Tariff rebalancing following reform of the telecommunications cable has helped to alleviate the sector, 2000–05 shortage of bandwidth for a number Index: 2000 = 100 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 Source: AICD. 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.

viii

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 stateowned enterprises

Figure F Telecom revenue as share of GDP, selected African countries, 2000â&#x20AC;&#x201C;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.

Given the private sectorâ&#x20AC;&#x2122;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 Figure G Telecom governance score 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 Source: AICD 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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 2007â&#x20AC;&#x201D;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

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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.

xii

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

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 upperincome 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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

Figure 1.1

ICT penetration rates per 100 inhabitants, 2007

Figure 1.2

Fixed telephone lines per 100 inhabitants, 2007

Source: AICD. 2

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

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 inhabitantsâ&#x20AC;&#x201D;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 regionâ&#x20AC;&#x2122;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). 3

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 countriesâ&#x20AC;&#x201D;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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

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 footprintâ&#x20AC;&#x201D;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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 SubSaharan 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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

Figure 1.9 Growth in new mobile subscribers

Figure 1.10 GSM population coverage in AICD countries

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 2007â&#x20AC;&#x201D;or some ten users to one subscription. Overall penetration was just 3 percent of the population (figure 1.12). 7

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 networksâ&#x20AC;&#x201D;where most subscribers areâ&#x20AC;&#x201D;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. 9

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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$4â&#x20AC;&#x201D;almost a fourth of that of the African region.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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.

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 11

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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. Nigeriaâ&#x20AC;&#x2122;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 obligationsâ&#x20AC;&#x201D;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

www.ncc.gov.ng/interconnection/Interconnect%20Rate%20Determination%202006.pdf. 12

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 Figure 1.17 Value-added and excise taxes, 2007

Source: GSM Association.

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

“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 13

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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

â&#x20AC;&#x153;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.â&#x20AC;? Millicom International Cellular SA. Form 20-F. 2007. 14

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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. 15

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

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 dialup 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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA Figure 1.22 Broadband Internet prices, US$ per month, 2007

Source: ITU, AICD.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

Figure 1.24

DSL price, USD per month, 2007

Source: AICD.

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.

Directique. Enquête Qualité de Service des Réseaux GSM au Sénégal–Octobre-Novembre 2006. 19

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 peremployee 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.

Kenya was involved in a staff reduction for Telekom Kenya prior to privatization: “Kenya’s state-owned fixedline 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. 20

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

Figure 1.25 Mobile subscribers per employee, Vodacom networks

Source: Vodacom, Vodafone.

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).

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

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

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 22

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

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

Many of the examples in this section use analysis carried out on the original 24 countries of AICDâ&#x20AC;&#x2122;s first phase. The doubling of AICDâ&#x20AC;&#x2122;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. 23

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 organizationsâ&#x20AC;&#x201D;for example, the Common Market for Eastern and Southern Africa (COMESA 2007), Economic Community Of West African States (ECOWAS),16 and Southern African

For example, Fink, Mattoo, and Rathindran examine the impact of privatization and competition. See wwwwds.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/ituec/Ghana/modules/FinalDocuments/Model_ICT_Law_Policy.pdf

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 countries

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

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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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. 26

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

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

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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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.

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

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

Mobile

International

...

Internet

Local fixed

International gateway

...

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 28

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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. 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 20

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 Africaâ&#x20AC;&#x2122;s exclusivity ended in May 2002, but a second national operatorâ&#x20AC;&#x2122;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.

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. 30

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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

Dec05

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 100

9,857

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

373,798

9,857

Incumbent

Westel

2,711

Kenya

339,199

100

9,437

Second National Operator announced in October 2006 but later cancelled

Telkom Kenya

329,358

9,428

Incumbent

Others

9,841

19 licensed "local loop operators" with limited scale and scope

Ghana

376,509

Dec07

Jun07

Madagascar Telma

Incumbent

Gulfsat Dec07

Mauritania

100

9,418

Mauritel

9,409

Incumbent

Chinguitel

Since August 2007 (owned by Sudan Telecom)

Mauritius

361,300

100

8,417

Mauritius Telecom

330,000

8,342

MTNL

31,300

Morocco

2,393,767

100

4,945

Maroc Telecom

1,273,675

2,831

Meditel

19,790

Wana

1,100,302

2,113

Nigeria

1,545,984

Dec07

Mar-

Incumbent

Incumbent Limited mobility wireless

100

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

Country / Operator

Subscribers

Date

Market share

HHI

Note

08 NITEL

59,850

Incumbent

Others

1,486,134

17 fixed including fixed wireless

Rwanda

22,291

Rwandatel ARTEL

Sep07

100

9,570

MTN awarded fixed line license in 2006 but has not entered market.

21,801

9,565

Incumbent

490

Rural infrastructure provider

South Africa Telkom

Incumbent

Neotel

Granted license in 2005, started providing business services in 2007 and consumer market in 2008. Licenses for underserviced areas with a teledensity of less than 5%. 27 areas identified with 7 licenses awarded.

Underserviced Areas Licensees (USALs) Sudan

636,905

Sudatel

487,584

Canartel

149,321

Tanzania

146,419

TTCL

146,419

Dec06

100

6,410

5,861

Incumbent

550

License awarded in 2004

Mar07

Fully liberalized but no entry to the fixed line market due to the cost effectiveness and convenience of mobile services Incumbent Although awarded "fixed" license, using GSM-fixed wireless and subscribers counted as mobile. Also, until 2006, only operated on island of Zanzibar.

Zanzibar Telecom Uganda

100,777

UTL MTN

Dec05

100

7,195

83,777

6,911

Incumbent

17,000

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).

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

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

Source: AICD.

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 incumbentâ&#x20AC;&#x2122;s fixed wireless service. See www.telecom.na/index.php?go=news&sel=view&nid=52. 33

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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). Figure 2.10 Status of mobile competition, 1993–2006

Source: AICD.

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.

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. 34

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 yearsâ&#x20AC;&#x2122; 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 35

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 AICDâ&#x20AC;&#x2122;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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 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

Data in US$

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

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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

1.34

0.86

67.95

282.97

Access to submarine cable

0.57

0.48

37.04

110.71

Monopoly on international gateway

0.70

0.72

37.36

119.88

Competitive international gateways

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 Africaâ&#x20AC;&#x2122;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 countryâ&#x20AC;&#x2122;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 highspeed 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).

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. 39

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

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

Commenting on the SAT-3 fiber-optic cable that runs up the west coast of Africa, one reporter notes: â&#x20AC;&#x153;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.â&#x20AC;?

http://newsroom.cisco.com/dlls/2006/ts_053106.html

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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

According to the “Level of Competition” information extracted from the ITU ICT Eye available from [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. www.itu.int/ITU-D/icteye/Regulators/Regulators.aspx#

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 Telecomâ&#x20AC;&#x2122;s properties were the result of buying incumbent fixed-line operators that had mobile operations (for example, CĂ´te dâ&#x20AC;&#x2122;Ivoire Telecom and Sonatel of Senegal). MTCâ&#x20AC;&#x2122;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 Lebanonâ&#x20AC;&#x2122;s Investcom in 2006.

Table 2.5 Strategic mobile investors, 2007 1!/".(!$./ /

1*!$. ,% ",1+0.($/

,10' %.("

,# ",* ,10' %.("

. /",* &5-0

(+ 13 (0

,# %,+$

(2$+#( .," $)$",*

. +"$ $)$",*

,.01& ) $)$",*

())(",* 14

,,2

Source: AICD adapted from company reports.

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

â&#x20AC;&#x153;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. â&#x20AC;&#x2DC;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,â&#x20AC;&#x2122; 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.â&#x20AC;? â&#x20AC;&#x153;Zambia Competition Commission (ZCC) approves Celtel, Telecel takeovers.â&#x20AC;? Times of Zambia. July 27, 2005.

http://rights.apc.org/africa/index.shtml?apc=s21819e_1&x=554587

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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

Fixed-line subscribers, number per 100 people

High mobile competition

2.64

0.68

Low mobile competition

2.70

1.61

Price of fixed-line monthly package, USD

High mobile competition

20.24

8.78

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

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 43

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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. 44

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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 bestpractice 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 cross34

“Benchmark Indicators for African NRA Websites” (2004), http://lirne.net/resources/nra/NRAchapter5.pdf 45

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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.

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. 46

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

Figure 2.18 Telecom governance score

Source: AICD.

For countries with state-owned fixedline 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.

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

For those African countries with public Mozambique 1.70 98.70 — incumbents, cost recovery and productive Namibia 3.32 86.96 13.62 spending rely mostly on the governance South Africa 2.11 96.46 29.47 and operational efficiency of state-owned Tanzania 1.07 95.91 — enterprises (SOEs). More than 90 percent Source: World Bank. of public expenditure is channeled through Note: Countries with fully or partly publicly owned telecom incumbent. 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 47

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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, 2001â&#x20AC;&#x201C;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

Cameroon

730

0.00

Chad

127

0.12

202

Ethiopia

104

0.07

Ghana

563

0.11

Kenya

220

0.00

Mozambique

605

0.02

Namibia

470

0.12

1,145

0.00

Tanzania

219

0.24

OECD benchmark

634

South Africa

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 nobrainer (figure 2.19) Figure 2.19 Link between governance and cost of labor redundancy

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 fixedline services is in decline. Massive restructuring of the Source: AICD. 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.

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

Table 2.9 Succinct relation between institutional scores and key performance indicators Penetration (subscribers per 100 people)

Correlation

Mobile

Fixed line

Price service package (index)

Mobile

Fixed line

Efficiency 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

Source: World Bank.

Link between regulation and fiscal revenues

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA

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.

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 50

INFORMATION AND COMMUNICATIONS TECHNOLOGY IN SUB-SAHARAN AFRICA 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.