The case of ADSL, opportunities for operators and challenges for regulators,

Broadband technologies and services in sub Saharan Africa: the case of ADSL, opportunities for operators and challenges for regulators,

Michel Rogy

Abstract: Sub Saharan African (SSA) countries generally suffer from a lack of fully rolled out fixed infrastructure to enable the spread of ICT use among the population, especially in remote and/or rural areas. Nevertheless, the rapid development of dial-up internet access out of the existing footprint of the fixed network provides a clear indication of latent demand for internet access. Average annual growth of internet dial-up traffic, approximately around 40%-60%, is currently mainly driven by increases in the number of customers, through both private (residential) and public (cyber cafe) access. Various broadband initiatives have therefore been launched and the number of African countries offering commercial ADSL services tripled to 15 in 2004 from just 4 (Tunisia, South Africa, Nigeria, Senegal) in March 2003. Achievable penetration rates (typically 3% of fixed lines after 2 years, 5% after 5 years, 10% after 10 years) provide a sound basis for companies, which are fairly concentrated geographically, a basis that has been reinforced by the significant fall DSL equipment costs. In SSA countries, confronted both with the development of broadband internet usage and the liberalisation of fixed telecommunications, the major challenge for regulators is to ensure an appropriate set of options for “DSL Make or Buy” with respect to ALL the various players in the market (non infrastructure-based ISPs, infrastructure based-ISPs, infrastructure-based telcos (voice, data, internet, retail and wholesale etc.) whilst preventing incumbents from abusing their dominant position.

Key words: ADSL, Sub Saharan Africa, business models, regulation.

**********

Faced with the diffusion of new technologies and services such as VoIP, WIFI and WIMAX that bring both services and high-speed internet access to urban and rural areas, as well as the launch of ADSL offerings in urban areas with/out TV or VOD, regulators and policy makers in Sub Saharan African (SSA) countries are encountering major challenges in the design and implementation of a new set of models and tools to reform and regulate broadband.

Broadband value chain(s)

Broadband technologies enable triple play services, comprising of internet access, voice and video services. The economic logic behind triple play is quite straightforward, due to the importance of:

* Economies of scale, at a subscriber level: joint fixed costs between services are high, so once the first two services (broadband internet + VoIP, for example) have been sold, a third service (such as TV) can be offered at a very low marginal cost.

* Economies of scope, at a network level: when new services are added, the potential for attracting new clients increases as a higher number of new customers will be interested in the offering; for example, by TV programmes only (sports events etc.) or unlimited calls only (VoIP). The broadband value chain comprises of 4 main components: content, backhaul, delivery and end user terminal (see figure 1 below). When considering SSA countries, the following specificities as compared to Europe have to be taken into account:

* Content: if content provision by African stakeholders does not increase, most of the traffic flows are likely to be from North to South, which constitutes a major difference between the broadband market and voice telephony.

* Backhaul: issues of international internet bandwidth capacity and pricing, as well as an appropriate number of internet exchange points, are of paramount importance to the development of broadband in Africa. Africa’s combined international internet bandwidth capacity is projected to increase tenfold to over 6 GB/s through 2006, although significant differences in availability exist depending on access to deep-sea cable capacity and regional backhauls. Furthermore, insufficient bandwidth availability sometimes leads existing broadband customers (using internet leased lines) of incumbents to switch to a VSAT solution, thus bypassing the incumbent.

* Delivery: most incumbent operators have succeeded in upgrading and modernizing their backbone and access network on the basis of the existing footprint. Investments in new local loops, on the other hand, have been very scarce, and fixed networks basically only cover major urban areas. Although the world’s most rapidly growing market for mobile telephony and home to its fastest growing fixed telephony markets, Africa still has the world’s lowest penetration rates and basic telephony provision remains a major need in many parts of the continent, particularly in rural areas.

* End user terminal: the availability of computers is still insufficient.

[FIGURE 1 OMITTED]

Single play “broadband internet”: DSL offers an interesting solution for incumbent operators in SSA countries

SSA countries generally suffer from a lack of fully rolled out fixed infrastructure to enable the diffusion of ICT use among the population, especially in remote and/or rural areas. The fixed revenues of an incumbent operator in an SSA country can typically be broken down as follows: 75% in the capital city, 20% in major provincial cities and only 5% in remote and/or rural areas.

Nevertheless, the rapid development of the internet dial-up market out of the existing footprint of the fixed network provides a clear indication of a latent demand for internet access. The average annual growth of internet dial-up traffic of approximately 40%-60% is currently mainly driven by increases in the number of customers, through both private and public (cyber cafe) connections. Not so long ago launching ADSL in SSA countries would have seemed like a crazy commercial strategy, but the following supply- and demand-related factors have improved the business case for such a move in recent years.

Major supply-side related factors include:

* Prices for DSL equipment (DSLAMs in particular) have dropped sharply in recent years (DSLAMs manufactured by Chinese suppliers currently retail at under EUR 10.000 per unit).

* Incumbents’ networks are geographically highly concentrated (see figure 2 below). The number of main digital frames to be equipped is relatively small, at least to cover the main areas of demand in the existing footprint, thus making the relevant investment compatible with a typical 5 year business plan of an incumbent telephone operator in SSA countries.

[FIGURE 2 OMITTED]

Major demand-side related factors include:

* Narrowband usage of internet via fixed lines is fairly high, so an appropriate pricing policy would definitely encourage migration to ADSL lines.

* Flat rate use of ADSL (as opposed to per minute pricing for narrowband internet access) enables far better control over communication expenditure from the customer’s point of view and is therefore more. appropriate for public programmes aimed at supporting the diffusion of ICTs (in schools and universities, for example).

Various broadband initiatives have therefore been launched and the number of African countries offering commercial ADSL services tripled to 15 during 2004 from just 4 (Tunisia, South Africa, Nigeria, and Senegal) in March 2003. The SSA countries that have introduced ADSL include Ghana (in June 2003), Benin (in April 2004) and the Ivory Coast (in August 2005). An analysis of past and future market developments indicates that, given appropriate pricing, ADSL penetration in typical SSA countries could reach 2.5% to 3% of fixed lines within 2 years, around 5% within 5 years and 10% within 10 years, taking into account the existing segmentation of demand (high ARPU), growth in public access (cyber-cafes) and ICT governmental projects.

So even if widespread deployment of ADSL within SSA countries is clearly impracticable (which advocates the simultaneous usage of alternative technologies such as WiFi and WiMAX), deployment where copper infrastructure exists could enable the rapid roll out of broadband services in major areas of demand.

As illustrated by Senegal, demand in the initial phase of market development is predominantly for low bit rates, typically 256 kbit/s or 128 kbits/s (see figure 3 below). Appropriate pricing for the first level of bit rates should not be more than around twice the price of a narrowband internet connection in the SSA country.

Viability of business models for competitors is nevertheless highly dependent on the regulatory environment

In SSA countries, which face both growth in internet usage and the continued liberalisation of fixed telecommunications, the major challenge for regulators is to ensure an appropriate set of options for “DSL Make or Buy” with respect to ALL of the various players in the market (non infrastructure-based ISPs, infrastructure-based ISPs, infrastructure-based telcos (voice, data, internet, retail and wholesale etc.) whilst preventing incumbents from abusing their dominant position.

A review of the current situation in SSA countries that have introduced DSL offers reveals potential regulatory concerns both in terms of operational issues (different levels of quality of service between an incumbent and its competitors’ clients, refusal to let competitors offer one-stop shopping for ADSL connections etc.) as well as in terms of pricing issues.

When devising the new regulatory framework for facility-based competition in fixed telecommunications, regulators should therefore:

–introduce an appropriate set of options for competitors, including resale, bit stream access and unbundled access

–adopt a clear stance against price squeezes by incumbents at the retail and/or wholesale level

Introduce an appropriate set of options for competitors, including resale, bit stream access and unbundled access

ISP subsidiaries of incumbent telephone operators in SSA countries generally face competition from rival ISPs, despite their retention of a very significant market share. It is therefore important for regulators to prevent incumbents from launching retail DSL products in cases where a wholesale product, enabling rival ISPs to offer their own DSL retail products and compete against the incumbent, does not exist.

Different wholesale DSL products could be made available by incumbents to ensure that rival ISPs are not foreclosed from the ADSL market (see figure 4 below):

[FIGURE 4 OMITTED]

Resale

As regulations in most African countries do not allow ISPs to obtain their own international bandwidth capacity, incumbents have to develop a resale offer to rival ISPs, as illustrated by Senegal.

In Senegal, the launch of retail DSL offerings by the incumbent Sonatel / Sonatel Multimedia in March 2003 was accompanied by the launch of a resale offer for ISPs named “Net ISP”. Upon intervention by the regulatory authorities, one-stop shopping was imposed, enabling the rival ISP to order DSL access on behalf of its customers (rival ISPs had claimed that the purchase of DSL connections in the incumbent’s outlets gave the latter an opportunity to win customers for its own internet subsidiary). However, the issue of “one-stop billing” for rival ISPs is still under discussion.

Bitstream access

Two main technical forms of bitstream access that could be provided with different geographical granularities are possible depending on the nature of the network (technology, footprint) operated by the new entrant. Based on experiences in Europe, the following bitstream offers could be envisaged in SSA countries:

* An IP bitstream at national level. This option would benefit rival ISPs as soon as competitive provisioning of international bandwidth capacity is allowed.

* An IP bitstream or an ATM bistream at a regional level, if rival ISPs are facilities-based and operate a long-distance network linking the capital and some of the country’s major cities. Compared to IP bitstream, ATM bitstream would offer more scope for differentiation compared to the incumbent, and especially suits those new entrants that operate an ATM network as they offer services to residential and business customers.

By using bitstream access, new entrants will be put in a position to follow a smoother path between non facilities-based competition (simple resale of DSL services offered by the incumbent: small infrastructure investment, small potential for differentiation) and facilities-based competition (Shared Access/Local Loop Unbundling for DSL services by new entrant: high infrastructure investment, high potential for differentiation).

Shared access/Full unbundling of the local loop

These options are associated with a higher amount of investment, as new entrants have to install their DSLAMs in the incumbent’s premises in all of the geographical areas that they intend to address. In SSA countries, however, fewer areas are “structurally” interesting for competitors with shared access or full unbundling of the local loop (5 to 15 areas in most of the cases to serve most potential clients). In these areas, the length of lines is generally compatible with DSL, but quality can be an issue (see Ghana). Nevertheless the civil engineering costs of accessing those areas could prove high: compared to shared fiber MAN, costs can multiply be a factor of 3 with leased lines and 5 with proprietary MAN. As a result, shared access or full unbundling of the local loop have more relevance in the context of the licensing of long distance or local operators for voice and ADSL services (second national operator), as these operators will search for local interconnection for voice and ADSL services and can thus share the associated costs.

Regulators need to adopt a clear stance against price squeezes by incumbents at a retail and/or wholesale level

Whenever confronted with a competitive ADSL market, regardless of whether the incumbent and its ISP subsidiary hold a very dominant position, regulators need to monitor the relative pricing of retail and wholesale ADSL products closely. This monitoring may be fairly complex depending on the diversity of products on the market, but it is crucial to ensure that there is appropriate economic space for all players in the ADSL market, as shown by experiences in the UK and France (see figure 5 below).

[FIGURE 5 OMITTED]

Clear potential for dual or triple play in SSA countries

Traditional voice telephony (PATS) has been subject to very strong substitution by mobile telephony for almost a decade in Africa. VoIP, which is currently growing rapidly throughout Europe and expanding in SSA countries, poses the threat of large-scale substitution for traditional voice telephony.

* DSL VoIP (also called voice over broadband or managed telephony), offered by telcos or facilities-based ISPs to clients with a DSL connection, targets “heavy” voice telephony and/or internet users. DSL VoIP is fairly easy to use and offers a service close to carrier-call reliability. When priced aggressively, DSL VoIP is a direct substitute for traditional voice telephony, so that only users with low telephone traffic and no desire to use the internet may eventually remain clients of the incumbent if it does not react to the DSL VoIP offerings of its competitors.

* Internet VoIP is also taking off, although it addresses a nearby market segment: low fare / low cost and “nomads” with no fixed line. Today internet VoIP is still limited to users equipped with PCs, but QoS and reliability are not guaranteed.

As SSA countries are characterized both by the importance of public access (revenues in an SSA country can typically be broken down as follows: 15% public service, 35% business, 20% residential via private access, 30% residential via public access) and by relatively high prices for international calls. Very rapid growth in internet VoIP is to be expected for international telephony, with both competitors and incumbents offering VoIP services (in Algeria, for example, international telephony prices have decreased by almost 40% due to competitive pressure from VoIP services).

Furthermore, incumbent operators in SSA countries will seek to leverage their DSL equipped local loop and enter the media market. This process is well underway in North African countries, offering yet another example of the latent demand for such services, while experiences in Senegal with ADSL2+ technology with both streaming (movie, sport etc.) and broadcasting (6 channels of Canal Horizon etc.) have proven successful to-date. This evolution towards triple play services will very probably confront regulators and competition authorities in SSA countries with the same new challenges as their EU counterparts, with ownership of diffusion rights becoming a major competitive advantage for those operators that have secured them.

Conclusion

Although SSA countries are considered the most under-served region in telecommunications infrastructure and network access, and the success of mobile telephony casts even more doubts over the further development of fixed telecommunications, the successful launch of DSL shows a clear potential for this broadband technology to supply internet access, but also voice and TV services.

Even if only limited to the existing footprints of fixed infrastructures, DSL offers to bring customers in SSA countries higher quality internet access and better value for money, and thus to contribute to lasting ICT diffusion. Furthermore, DSL is particularly suitable for cyber cafes, whose successful development plays a vital role in increasing the demand for broadband access and bandwidth capacity. Failure to deploy ADSL technologies may therefore deny SSA countries an opportunity to participate fully in the knowledge economy of the 21st century.

Michel ROGY

Tera Consultants, Paris

Figure 3: Mix in ADSL bit rates in Senegal

Bandwidth

Year 256 Kbits/s 512 Kbits/s 1024 Kbits/s

March 2156 711 133

2003

December 5507 1816 340

2004

# DSL # Fixed % DSL lines

Year lines lines over fixed lines

March 3000 229,000 1.3%

2003

December 7663 229,000 3.3%

2004

Source: Tera consultants

COPYRIGHT 2005 IDATE

COPYRIGHT 2008 Gale, Cengage Learning