About LINK Centre Policy Research Papers

Abstract

1 Background and Procedure
2 Models Used in Other Parts of the World to Address Rural Access

During the 1990s some Latin American countries facing the global challenge of cost-effectively meeting the pent up demand for telephones decided in the absence of successful roll out into rural areas despite liberalisation of their markets to provide governments subsidies to stimulate telephone access in remote rural areas.

In Chile, licences were granted to provide a minimum number of payphones to the highly mountainous regions of the country. Despite considerable gains in the number of connections following the liberalisation of the sector in the early nineties, the rural areas remained underserved due to the high cost of reaching remote areas, the low income levels of people in the rural areas and the failure of operating companies to consider innovative ways of servicing these areas with new technologies or payment packages. In 1994 the Government established a fund to create incentives for private investment in unserviced rural areas. The telecommunications development fund was set up for five years only with the express purpose of achieving public access for Chileans living in areas without public access to telephony. The fund was financed through the national budget and was administered by a council, chaired by the telecommunications minister, who awarded subsidies against an annual programme of projects eligible for funding on the basis of competitive bidding. The sector regulator served as the Council's secretariat. It identified the locations requiring public access and assessed their commercial viability. Projects were assessed in terms of their social net present value (NPV). Where projects appeared commercially viable they were drawn to the attention of the private sector to apply for licences. Where they were found not to be commercially viable they were 'ranked by NPV per unit of subsidy need to make them viable and by other factors'. Forty-six projects covering 1285 locations throughout the country with subsidies totalling US $4,3million, were launched in 1995 and ultimately provided service to about a third of the population without access, just below half a million people. Rollout of services was encouraged by the subsidy for projects, which averaged around US$3 500, being paid for only once the facilities had been built. This programme resulted in the proportion of Chiles's population living in places without access to basic voice communications to decrease from 15% to 1% in 2002. According the telecom analyst Bjorn Wellenius: "Success was due largely to extensive reliance on market forces to determine and allocate subsidies, minimal regulatory intervention, simple and relatively expeditious processing and effective government leadership."

Although Peru began a similar process much earlier in 1992 with the establishment of a dedicated telecommunication fund, Fondo de Inversión en Telecommunicaciones (Fitel) and the regulator, Organismo Supervisor de la Inversión Privade and Telecommicaciones (Osiptel) was ready to call for a tender by 1996, the process only got underway in 2001.

The goal of Fitel was to ensure the provision of pay phone service in 5000 rural towns and public access to the Internet in all 554 district capitals within ten years. Fitel was funded through a 1% levy on the telecom operators' gross operating revenues. While legally distinct from the regulator, Osiptel provides technical and administrative services to Fitel and approves polices and projects. Osiptel was responsible for identifying the locations requiring services, tendering and monitoring performance.

The decision was taken to divide the country into six regions, each with more than 700 towns, and then hold two tenders. The winning bid was to receive a 20-year non-exclusive concession, which would require them to install one public payphone in each targeted rural location and one point of public access to Internet in each district capital. Licensees would be required to provide public service over the entire 20 years despite only receiving subsidy payment for the first five years. They would however be permitted to use their facilities to provide additional services to individual subscribers, though Osiptel would regulate retail prices for rural services and interconnection.

Payment of subsidies is tied to project implementation and service quality with partial payment at the beginning of the project, another quarter once facilities are installed and the remaining amount in instalments over the five year period. This allows the regulator to penalise operators for failure to meet operational deadlines or for payphone or system outages.

Lack of political will together will institutional infighting stalled a very promising process but by March 2001 three competitive tenders had been conducted for six projects covering all 5000 rural towns to be connected by 2003. Despite these delays the first pilot project revealed some interesting outcomes. Firstly, "the winning bid requested a subsidy 41% lower than Osiptel had estimated and 74% lower than a previous offer from the incumbent operator". Traffic also exceeded Osiptels' forecasts by 32% in the first year. The pilot project mobilised private investment at a rate of US$22 per inhabitant while requiring only a US$11 subsidy per inhabitant.

The project was not without operational problems however. Despite meeting deadlines for initiating services in all the nearly 200 localities, the operator failed to meet service reliability targets resulting in fines equivalent to over one and half months revenue. Failed grade of service targets also resulted in delays by Osiptel in the first annual subsidy payment until problems were rectified, which were equivalent to nearly two months revenue.

Telecommunications co-operatives have been another mechanism of extending access to rural communities in several parts of the world. In the United States and Scandanavia co-ops were major contributors to the high levels of universal services enjoyed in those countries. More recently in Bolivia, the fifteen telecom co-operatives owned by local municipalities with a monopoly on local services in their geographic territories, provide the majority of local service in both urban and rural areas. Each subscriber pays an installation fee to receive the service which translates into equity in the company and entitles the subscriber to one vote, irrespective of the number of lines they have installed.

Co-operatives have also contributed to the improved teledensity figures in Poland, which sat at about 3% in rural areas in the 1990s. Following the passing of legislation that allowed independent telephone providers to compete with the state-owned monopoly in the early nineties, a number of co-operatives were successfully established to provide services to rural villages. Financing was secured from municipal governments, Polish banks and international loan agencies. In some instances loan guarantees were secured from equipment suppliers.

"Revenues from tolls and special services enable the cooperatives to fulfil loan payments, hire office and maintenance staff and begin network build-out and subsequent upgrades…Today, these systems service some 40 000 home and businesses with modern, state-of-the-art digital technology including Internet. The resultant economic growth in these communities has been impressive. For example, local incomes rose some 30 % in the first four years of some of the cooperatives' service territory and some 300 cottages industries reported opening in on cooperative's service area…. Perhaps most importantly, the success of building these locally owned and operated telecommunications systems energised the communities to undertake other self-help projects, such as financing and building sewage treatments plants and natural gas pipelines."

The extensive experience of co-operatives all indicates however that a level of financial resources to support the co-op system, and some existing or potential economic base, are necessary conditions for the sustainability of co-operatives.

The United States based National Telecommunications Co-operatives Association have identified the following as necessary for an enabling environment for co-operatives:

• National policy in support of locally-owned telecommunications operators;
• A regulatory framework that supports interconnect tariff structures between local and national operators;
• Access to funding, and the establishment of institutional structures that can provide favourable loans for small operators;
• Affordability of service provision to consumers; and
• Institutional support in areas such as technical support, lobbying, training and financial management."

The result of such conditions in the United States was the delivery to large parts of rural America, telecommunication services not provided by the large private telephone monopolies. Critical to the success of the co-operatives was the loans and technical assistance they received from the Rural Electrification Administration (REA) as part of a supportive national policy.

Another successful model used to leverage private investment into public infrastructure, used in Ghana, the Philippines, Thailand and Indonesia is that of the Build Operate Transfer (BOT). In this arrangement an investor builds out the network and accrues the revenues for a fixed period of time before transferring ownership back to the government or local operator. In some instances foreign strategic investors partner with local operators. Very often the terms of the licence require that installation of a certain number of phones within a particular period or to serve a targeted area within a specific period.

Where these have been successful, emphasis has been placed, as in the other models, on the need for well-structured licences and regulations, transparent processes. Several commentators believe that few communities or SMME will have the resources to successfully operate or manage even small networks. An experienced investor may be vital in providing wider access to capital markets, better management skills and access to latest technology. A BOT model would allow local communities or SMME consortia to gain ownership of the licences in the longer term without the having the necessary capital or skill to operate the network currently.

The lessons arising from these experiences are that:

• whatever the model used to access under-serviced areas, the models needs to be supported by a clear funding model whether reverse subsidies or low interest loans or government guarantees;
• small amounts of state funding can be leveraged to secure local and international investment in support of access programmes;
• competition can be used to allocate limited resources effectively within a simple but clear regulatory framework;
• areas regarded as unprofitable for traditional large-scale telecos can be efficiently and profitably serviced with the deployment of more cost-effective new technologies and lower charges;
• there are nevertheless always likely to be areas requiring state support for public access to communications service, however this need for support can be reduced by mainstreaming many areas traditionally not believed to be viable markets;
• while a simple and flexible regulatory environment seems to be the key to successful remote access, it would appear that there are critical areas requiring regulatory intervention;
• financial assistance seems to be most effective when it is linked to the timely completion of roll-out targets or to penalise non-performance through non-payment; and
• in so far as possible such regulation should be geared at alleviating market failure and challenging traditionally non-cost based pricing, both in tariffs and interconnection.

Despite these high costs it appears that there is no clear funding framework attached to the policy to licence under-serviced areas. While the final ITA has dropped the financial bid from the evaluation criteria the licensee will face several potential constraints, such as ownership restrictions and rollout undertakings in order to secure the licence without any state funding arrangements to leverage in support of a business plan.

Indications from across the globe are that such initiatives are only sustainable on the basis of direct or indirect state support. In some countries where such licences or concession have been granted, they have been granted on an exclusive basis. In South Africa the USALs will be competing with the public fixed line incumbent and the highly successful mobile operators. Where licences have been granted for remote, low- income areas, many countries have provided subsidies to operators to provide relief. Some of the traditional inefficiencies associated with subsidies have been removed through competitive allocation, such as in the reverse bids used in Latin America. Evidence from there indicates that substantial private capital can be leveraged from such subsidies. This is true of co-operative models in the United States, Latin America and Central Europe where government guarantees and low interest loans have been used.

This use of the Universal Service Fund in South Africa for this purpose - now that the R20m ceiling on the universal service levy has been lifted, and at least R100million should be available - would seem obvious. It would benefit all industry players and indeed the economy, by extending the reach of existing networks and creating a multitude of new calling opportunities. In addition the multiplier effects of improved communication in previously under-serviced areas have been well documented.

While the Universal Service Agency has recently indicated that it would be able to fund communities wishing to apply for USALs, there has been no indication of how much or how such funding will take place. This lack of certainty has inhibited the potential of state support to spur interest and leverage private capital.

With no loan guarantees or capital grant subsidies, one must assume that the initiative will be private sector funded through venture capital. Besides the Development Bank of South Africa which has an obligation to consider loans for infrastructural development there appears to be little private investment interest. Without the introduction of the state funding mechanisms mentioned above, how can the investment environment be made more conducive?

The risk profile of these licences could be reduced by demonstrating a clear regulatory framework that is not burdensome on the new entrant with regard to licence fees, reporting requirements and tariff fillings but which will ensure rapid entry into a market that can demonstrate likely revenue streams. With the lack of certainty of this market, cross-ownership rules need to be not too restrictive and reviewable. This is critical for investors, who need to have a clear indication of their options should the operator fold.

In addition a substantively differential termination charge for terminating calls on USAL networks will become pivotal in demonstrating a business case for the USALs.

Regulatory issues critical to the building of a feasible business plan for USAL operators can be summarised as:

• The sharing and leasing of existing facilities of major carriers;
• The sharing and aggregation of USAL operator facilities and services;
• Cost-based asymmetrical termination charges between the USAL and PSTN on national long-distance.

Facilities sharing will be critical to developing a viable USAL business case both with regard to sharing existing network operators' facilities at reasonable rates but also the sharing of a common platform to provide certain services. As indicated earlier the high entry costs associated with operating telecommunications network, even on the basis of new cost effective technologies, makes network capital costs a major barrier to entry. As the DBSA/IDRC Financial Viability of the Under-serviced Area Licences Report indicates, the shared platform "converts a major capital cost into an incremental capital cost and incremental lease cost". A shared platform further provides a number of efficiencies resulting from increased scale of operations. Not only does it provide the facilities but the trained personnel to provide all customer activities for the USAL operators.

Interconnect is critical both as a direct cost for outgoing calls and as a revenue stream for incoming calls. This will require a firm regulatory regime that compels major carriers to provide facilities and interconnection speedily and at fair cost and that USAL operators are recognised as public operators with the associated benefits of wholesale pricing.

It will also require an interconnection regime that recognises the asymmetrical cost of terminating calls in high density, low cost urban areas and low density, high cost rural areas. Traditionally, largely due to monopoly provision and the send and receiver being on the same network costs have failed to recognise both origination costs and termination costs. However, as markets have been opened up and calling and pricing charges have become more cost based, the reality of the differential costs in the provision of services has become apparent. In the US, recognition of this has long been evident in the asymmetrical charges between major centres and more remote areas. In Rhode Island for example termination of traffic in rural areas is 38% higher than urban terminations in peak time and as high as 315% higher in off-peak time. In Chile the cost of terminating a call in rural areas is more than 18 times higher in peak time than terminating in a city and nearly 10 times as high off-peak. Even internationally, with the dispute over settlement rates, the ITU through protracted discussions in Study Group 3 has recommended asymmetrical costs associated with terminating international calls in developing countries.

Without cost based asymmetrical interconnection prices a sustainable business case cannot be made for USALs. A financial viability study of the Under Serviced Area Licences in South Africa concluded that the 30% proposed in the draft regulations proposed by the regulator would compromise the viability of the USALs and that the differential would need to be at least 70% for the USAL to be viable. There are limited revenue streams and without these there is no sustainable rollout. If it is forced to rely on the revenues from symmetrical origination and termination charges the only way the USAL will survive will be to control its costs by minimising rollout. At best it will survive as a niche player, but even this is unlikely considering that it will be competing with five operators in the designated areas and it will presumably have some roll-out undertakings in order to have won the licence. Asymmetrical interconnection rates together with a degree of regulatory flexibility have the potential to dramatically improve the viability of USAL operators. Increased revenues from termination fees should also permit the new entrant to reduce origination prices in order to compete with the incumbents, which should create greater demand for their services, permitting greater rollout and creating more revenues, which provides a virtuous cycle of sustainability. Price competition results in faster rollout. Origination price competition in Morocco's mobile duopoly led to an 1800% a year growth with over 3 million new customers (over which over 2 million were new incumbent subscribers).

This could also result in innovation of new services and business opportunities generating further revenue in the designated under-serviced area. For example, with even a small operating profit on each call resulting from termination in the under-serviced area, USAL operators might induce ISPs to relocate to underserved areas for marginal incentive payments for incoming traffic terminated by the USAL operator.

The matter is of such critical importance that the regulator should prepare a default interconnection agreement in case of any delay in the negotiation process with incumbents.

Although not permitted by the legislation, direct access to the international gateways, could have bolstered the viability of the USAL, especially with an asymmetrical termination charge for international calls terminating on their networks.

Ownership restrictions, while understandable again from a public interest point of view and facilitating the entry of an increased number of players, again has the potential to undermine the viability of operators. While technological gains have made possible the provision of telephone services at much lower cost than ever before, network economies still require a critical mass of people to justify the relatively high cost of network development. Many aspirant licensees were hoping to have at least significant ownership in a number of licences in order to benefit from economies of scale. Some aspirant licensees have indicated that as much as 50% of network costs would be duplicated in each licence and could not be justified in terms of the potential market in each area. If ownership rules are going to restrict ownership and control to one or two licences with minority ownership in one or two more, it may still be necessary for the regulator to adopt a very flexible approach to the sharing of facilities and services among operators, to encourage viable enterprises.

While the viability of these licences might be regarded questionable by industry pundits, for many SMMEs and other interests in rural areas these licences present once in a life time opportunities and with people likely to be putting their personal wealth on the line, the licensing process needs to be as simple, predictable and incontestable as possible.

An excellent case study in this regard is provided by the GSM licence awarded in Morocco in 1999. US$1.1 billion was paid for the fifteen-year licence to operate under 'relatively unfettered competition'. One of the highest fees per capita ever paid for a mobile cellular licence has been attributed to the leadership, clarity, transparency and lack of subjectivity in the process.

To ensure that the licence award process is as objective as possible and that the terms of the licence are clear and unequivocal, the rights and obligations of the licensee, together with the undertakings made against the evaluation criteria, should become the terms of the licence.

Drawing on the Moroccan experience, but accounting for local policy, the licence could consist of three areas to include the special requirements of South African policy – the licence description, the rights attached to that and the obligations.