9. Access and infrastructure –social models for extending the reach of the Internet

Not everyone has access to the internet but in many ways the digital divide is just a prism through which all other inequalities – whether of race, gender, class or whatever – are reflected. The gap in internet access between developed and developing countries is large and continues to grow. Low-income countries account for almost 60% of the world’s population but have just under 5% of the world’s internet users.

In the developed world, because internet access is fairly widely distributed throughout the population, the digital divide is less acute. Nonetheless governments spend significant sums of money on digital divide initiatives aimed at offering access to poorer people and those in rural areas. Although internet use has grown rapidly from a low base in developing countries (particularly in Africa), this progress is not keeping up with advances in the more developed world. And by their very nature, the governments of low-income countries find it hard to prioritise spending on internet access over more pressing demands like health care.

Because companies providing internet access in low-in-come countries need to make money (as indeed they do elsewhere), they will target customers who can afford to pay; either middle-class individuals or companies. Obviously there is a relationship between cost of access and the number of users.

Simple economics means that lower costs mean lower access costs and therefore more potential users. The example of the cost of a telephone line illustrates the impact of lowering costs quite clearly. As Dr Ashok Jhunjhunwala, pioneer in affordable telecom solutions, IIT-Madras in India puts it: “It currently costs (an investment of) Rs30,000 to install a single telephone line. To cover this investment, you need a revenue of at least Rs1000 per phone line per month. These rates are affordable to just 2-3% of the Indian population. But if you bring down the investment needed for a phone line to Rs10,000, then affordability of telephones could immediately go up to 30 per cent or more of our population”.

The same basic argument holds true for internet costs, which are inextricably linked with telecom costs. The cost of a telephone line is often a key component of overall access costs (see section 6.5). Where telecom companies are government-owned and/or have no competition, the cost of installing new lines is inevitably higher than the costs of lines installed by privately owned companies.

But however much you lower the costs, the market will only provide internet access to those who can afford it. Nevertheless the market can be ‘stretched’ to provide access to areas that might not otherwise be connected by governments or regulators, by offering continuous or one-off financial incentives. Again these ‘market-stretch-ing’ initiatives will take internet access out to another layer of people, often paying the capital costs of putting in connections and sometimes subsidising rural call costs.

Beyond the areas that might be served by these approaches lie vast swathes of poor people who do not have the means to buy internet access at all or who cannot afford market prices. For example in Africa, large numbers of its poorest people live widely dispersed across rural areas. Two problems converge: not only are these amongst the poorest people on the planet but they are also spread out and often in physically inaccessible villages. For these areas internet access can only really be delivered if it can be paid for by government or external donors as a social cost. The arguments advanced would be very similar to those advanced for access to telephones.

This has led to a rumbling debate in the development sector about whether the internet is actually a cost effective means of delivering communications benefits to poor people in low-income countries; those living on less than a dollar a day. These are perhaps best summarised in a World Bank paper by Charles Kenny called The Costs and Benefits of ICTs for Direct Poverty Alleviation.

In essence the argument is that the internet compares badly in cost terms as a means of reaching people. It is cheaper for an individual to own a radio, and radio broadcasting reaches people more cheaply in their own language. For example, a station in central Mali broadcasts to 92,500 people a year at a cost of just US$ 0.40 cents a person. Likewise a mobile or fixed phone – although more costly than radio – can serve more people at a lower cost than the internet.

The debate has been crystallised by international donors finding that providing internet access (through telecentres) is becoming one more demand on their already stretched resources. For as Kenny puts it, “there is a movement in the development community pushing for widespread rollout of community access points to the internet as a tool for direct poverty relief.”

Whether or not you accept this kind of argument, the underlying issue remains one of who pays for the social cost of rolling-out internet access to areas where people have very little money to pay for it. Again, what starts as an internet issue very quickly becomes part of broader discussions about the role of government. Until relatively recently, the telephone company would have been owned by the government and the social priorities of the government of the day would be reflected in what it did. For example, British Telecom was obliged to provide call boxes in a wide range of locations across Britain. Governments would usually cross-subsidise these social costs out of the monopoly profits made from the rest of the operation.

In the world of privatised telecom companies, the role of government shifts. It can no longer do everything, it has to become an ‘enabler’. What does this word (much beloved by consultants) actually mean? Well, in this context, it may mean that the government encourages the regulator to set up a universal access fund into which licensed companies pay a contribution, and that these contributions are used to fund rural roll-outs in poor areas.

For example Chile used its universal access scheme to run a reverse subsidy auction in which private companies bid to provide public telephones to un-served areas, the lowest subsidy bid being successful. In a similar fashion, Uganda’s regulator has franchised out rural telephony in a particular area and in this instance, the company providing the service will provide a telecentre in one part of the region.

By acting as a facilitator, government can seek to encourage internet take-up in a variety of different ways. E-rate is another example of an approach designed, in this case, to encourage schools to connect to the internet. Started in the USA, e-rate is at its simplest a nationally agreed discounted rate for internet access for schools: often this rate is enshrined in the relevant telecoms legislation at a national level, and therefore is the responsibility of the regulator. Brief descriptions of several schemes give a flavour of what it seeks to achieve:

USA: The US scheme is administered by a not-for-profit organisation that was established by the Schools and Libraries Universal Service Fund. The scheme has six different levels of discount in order to focus the maximum subsidy in poor and rural areas. The method used to measure poverty is the percentage of students eligible for the national school lunch programme that provides a free lunch to poor students. In its first two years, the e-rate programme connected one million classrooms.

Senegal: The Ministry of Education and Sonatel signed an agreement that establishes preferential terms for access to the internet to make it more affordable to learning institutions. Discounts vary depending on the type of connection but can go as high as 75%. Installation costs are also discounted. Sonatel is directly responsible for invoicing the schools. Sonatel and the Ministry of Education have appointed a co-ordinator for the programme.

South Africa: An amendment to the Telecom Act includes e-rate, which it wants to introduce “to stimulate and facilitate internet usage by public schools. The e-rate will allow public schools a 50% discount on calls to access the internet as well as internet access charges.” Although the provision has been made in the Act, implementation of a national scheme has been slow but things are now beginning to move.

The experience of telecentres offers another way providing wider internet access in developing countries. Like cyber cafes they offer the user a chance to connect to the internet without having to buy a PC and often at subsidised rates.

Based on a study carried out Peter Benjamin of the Link Centre in South Africa on telecentres in Africa¹
, it is possible to summarise what the experience has been and the costs of providing them. Of the centres surveyed, the average cost of providing a telecentre is up to US$250,000. These projects stress community participation and sustainability, but to date none have proven that they can be self-sustaining without external funding. Most of these centres are supported by foreign donors, though the national programmes for telecentres in South Africa and Egypt can be included in this category.

The Nakaseke Multipurpose Community Telecentres (MCT) in Uganda opened in March 1999. It aims to introduce and test new technologies and applications, and demonstrate the impact of such technologies on development of rural and remote areas. A baseline survey was conducted to establish the nature of information needs of the community and the services. Funding came from international donors (60%), and national government (40%).

The telecentre has eight computers, two printers, a scanner, a photocopier, VCR/TV, video camera and projector. However, frequent power cuts are a problem. As well as phone, fax and internet use, there is a paper and digital library, computer training, and an interesting indigenous knowledge programme where centre staff are building a resource of local health and crop experience.

The centre is just about covering operating costs (subsidised by the community) but there is no expectation that the centre could generate enough income to replace equipment in a few years (depreciation) let alone repay the major capital investment. This centre required great external support (financial and organisational) and so is unlikely to be a model that can become widespread.

The first of the major funded Multipurpose Community Telecentres in Africa was established in Timbuktu in Mali in May 1998. Sotelma, the national telecom was the main local implementer, with other main partners being the ITU, ORTF (TV Mali), UNESCO and IDRC. The majority of the funding of around US$200,000 has come from the external donors. The pilot telecentre is equipped with 11 computers. It offers copying, telephone, fax, and internet services. A major emphasis of the telecentre is to provide training to artisans to set up their web page to sell their handicrafts. The telecentre has proved to be particularly useful to the tour agents organising visits to Timbuktu. The services are subsidised and offered at fees decided by its steering committee.

In Mozambique, two pilot telecentres were established in 1999 in Manhiça and Namaacha (both in Maputo province), funded by the IDRC. They each have four computers, an internet hub and modem, two printers, backup equipment, a public phone, a fax/phone, an external cardphone, a photocopier, overhead projector, whiteboard, TV with video, radio and binder. Current operating costs are just being met by operating income, except for the phone bill. Initial conclusions are that longterm economic sustainability depends on the existence of a critical mass of users and the adoption of computerrelated services (over-reliance on phone and photocopy services for income means vulnerability to inevitable future competition and that the major telecentre investment is not justified); technical support, backup and continuous staff training are essential, especially for encouraging the developmental and information services; good communications channels with local authorities and community leaders, and maximum transparency and information regarding the project, are important to success.

In South Africa, 62 telecentres have been established by the Universal Service Agency. These cost around R200,000 each (US$30,000) and most have four computers, four phone lines, a printer, a copier and TV. Most are in rural areas. Only a few are economically sustainable, mostly through running computer training courses. There have been many technical, financial and managerial problems.

Peter Benjamin’s conclusions from the study help define the current limitations of the use of telecentres in low-income areas:

• Centres are managed better where the owners have a stake in them. In some projects, donated equipment is lying around unused. The entrepreneurial instinct is a strong force in making a centre effective.

• There is a great demand for telephony. ICT use can be built, but it takes time, training and local adaptation.

• Simple business models are more likely to be successful than complicated ones. The idea of a multi-purpose telecentre is ambitious. Without extensive training and support, many of the wider aims of telecentres are difficult to reach.

• Computers by themselves are not an information service. Few centres use IT systems to provide information for local use.

A study carried out by Samuel Kyabwe and Richard Kibombo on two villages in Uganda with telecentres showed that internet use among villagers was below 5% compared to 30% for telephone usage and 100% for radio usage.