2. Introduction. 1. Summary

Transcription

1 Background paper on Universal Service and Universal Access issues Prepared for the Telia Telecommunications in Society 1999 Seminar, Sweden, June Peter Benjamin & Mona Dahms 1. Summary This is a background document on issues of Universal Access and Universal Service in Telecommunications for discussion at the Telecommunications in Society 1999' seminar organised by Telia in June This paper will focus on issues of relevance in the developing world and draws heavily on the ITU s World Telecommunication Development Report 1998 (ITU 1998). There is great investment in new telecommunications services throughout the world, with information systems fulfilling an increasingly important role in the world economy. Throughout the world there is a tendency towards a competitive liberalised telecommunications environment, and there is a real risk that the large numbers of people who cannot afford a telephone (as defined by current economics) will not have the benefits of telephony and will become increasingly marginalised in a wired world. This will hinder wider socio-economic development that requires a stable telecommunications infrastructure. For telecommunications to aid the wider development of poorer countries, special concern must be given to providing general access, issues which are quite different from the developed world where a majority have access to telephones. While universal service (a telephone in every home) will not be achievable in most countries soon, a realistic goal is to work for universal access, where a useable telephone is within reach of the whole population. There is a migration of aims and policy means as the telephone network of a country expands. For universal service, issues of Availability, Accessibility and Affordability need to be considered, especially regarding the economics of provision. To provide universal access, a range of public phone operators or telecentres can be used. There are a number of financial, regulator and policy measures that can be taken to promote universal service in a county. Developing a definition of universal service and universal access can be mechanisms for agreeing national targets with the many stakeholders involved, which must then be monitored to ensure compliance. Universal access and service definitions should be seen as moving targets when one level is achieved, a higher goal should be set. 2. Introduction As the Information Revolution has increasing impact around the globe, the issues of who has access to these technologies has great importance. Connection to these networks and services influences a people s access to jobs, education, health care and full participation as citizens. The convergence of telecommunications, computers, information production and broadcasting has a great impact on how societies are structured. There is a major question throughout the world: will these technologies and systems lead to increasing disparities between the connected information elite and a large majority of information poor (the digital divide ), or can these systems be used to support widespread development and a more just society? These communication and information networks are often now referred to as the Information Infrastructure, as important to the development of a country as the road and water networks. Should access to communications (especially a telephone service) be considered as a right, as the UN suggests (ITU 1997)? If it is accepted as a right then the cost benefit equation should become a secondary consideration but how is this possible in developing countries with very severe economic constraints? There now are technical solutions that would allow telephony throughout the world. The issue is one of political will, which often (but not always) means finding economic models that allow for provision to be paid for. Throughout the world there is a move from state owned monopoly telecommunications providers (PTOs - Public Telephony Operators) towards a more commerical and liberalised environment. In the former case, universal access action would come through national policy requiring certain action of the PTO. In the liberalised sector, other regulatory mechanisms are required. The Maitland Commission in 1984 stated that, We believe that by the early part of the next century virtually the whole of mankind should be brought within easy reach of a telephone. This is still the target most frequently referred to, and the International Telecommunications Union (ITU) believes it can be achieved by around 2010 with committed action from all involved in the sector especially if policy and regulatory frameworks are put in place to support this (ITU 1998). The goal of providing access to all is becoming increasingly important for promoting development in the Information Age. Page 1 Page 2

2 3. Overview 3.1 What are these terms The most common measure of telephones in a country is teledensity. This measures the number of telephones per 100 inhabitants. This is the easiest statistic to collect and so is most often used to compare different areas. However, it pays no regard to what phones are used for and who has access. For example, adding 500 phones to a banking office might not benefit other people living in the neighbourhood but would increase the teledensity. It normally refers to main telephone lines (connection to the local exchange) and normally does not include cellular phones. Universal Service means that everyone in an area has a telephone in their house (or the right to one if they can afford one). This is usually measured by household penetration (how many households have a telephone). A figure of 90% of households with a phone is normally considered to mean that full universal service has been achieved. The universal service percentage for an area is taken to be the percentage of households with a telephone service. Universal Access, by contrast, means that everyone can use a telephone (such as a payphone). What accessible means is open to debate, one example is saying that there must be a working affordable 24-hour public phone within 5 kilometres. What this definition entails is a major part of the discussion below. 3.2 Access throughout the world As in so many things, there is a great difference in access to telephones between the rich and the poorer countries. In High Income countries (with a GNP per capita US$ 8,956 or more), over 90% of households have a telephone, compared to only 16% of households in the rest of the world (ITU 1998, p 13). As significant is the difference within developing countries, especially between urban and rural areas. In developing countries 60% of the population live in rural areas. However in these countries 80% of the telephone lines are in urban areas (ibid, p 14). Income level Teledensity (lines / 100) High Income Countries 48.8 Upper Middle Income Countries 12.9 Lower Middle Income Countries 8.1 Low Income Countries 0.9 Source: Hudson (1997b) from ITU data Page 3 Urban Density Rest of country Telecommunications is an area of great growth, with a major expansion of networks happening in most countries. The developing world is certainly not homogenous, with some areas achieving explosive network expansion. The table below (taken from Mansell and Wehn 1998) illustrates the growth in telephone lines for regions of the world. It shows the Compound Annual Growth Rate in main lines between 1990 and 1995, with the teledensities in the two years. It shows that China has had tremendous growth while Sub-Saharan Africa and former soviet Central Asia has grown poorly. There has been low growth in developed countries due to their high starting level of teledensity. ( Asian first Newly Industrialised Economies include Korea, Singapore and Hong Kong; while the second group include Indonesia, Malaysia and Thailand). Region CAGR Main lines / 100 inhabitants Page 4 Main lines / 100 inhabitants 1990 Min lines / 100 inhabitants 1995 China 40.9% Asia 2 nd NIEs Other Asia 18.6% 15.3% Mahgreb West Asia 11.4% 10.0% Other N Africa South America Oceania Developing 8.0% 7.5% 7.5% Europe Developing Caribbean 6.5% 6.0% Asia 1 st NIEs 6.0% Eastern Europe Sub-Saharan Africa 4.9% 4.0% Developed 2.7% Central Asia 2.6% Source: Knowledge Societies. (Mansell and Wehn 1998) 3.3 It s not just about telephones Today telecommunications is not a luxury service for a few, but increasingly is an enabler for development in a wide range of areas. Without an adequate telecommunications network, it is impossible for a country to take advantage of the opportunities of the new information economy. The telecommunications infrastructure supports a range of other sectors and is a necessary support for many modern industries. There is strong evidence that investment in the telecommunications is closely related to increases in national productivity (Lamberton 1995). There is a close

3 relationship between the wealth of a country (as measured by GNP per capita) and national teledensity, with the graph of these variables revealing a fairly straight line known as the Jipp curve after the researcher who first observed this (Jipp 1963). While it can not be directly proven that either one of the variables causes the other, there clearly is a close inter-relation between the two. The ability to use telecommunications for instantaneous communications can support development by increasing: Efficiency: the ratio of output to cost; Effectiveness: improving the quality of services; Equity: better distribution of benefits throughout society. (Hudson 1998) Telecommunications forms the information infrastructure to support developments in many areas of life (Parker and Hudson 1995, ITU 1998). Many studies have shown the advantage to health care through training and organisational support for health workers, for education of communities on health issues, for consultations and in certain well-resourced areas for telehealth where a specialist at a central location can diagnose a patient in a rural area. There are several cases of news of the outbreak of a disease (such as cholera in the South Pacific or Ebola in central Africa) being made through telecommunications, allowing a more rapid response which saves lives. Education can similarly benefit greatly from telecommunications. While teleeducation and using the internet for teaching is of great advantage in wellresourced areas, basic telephony is very useful at all schools for general organisation and support purposes. One example of use of telecommunications in education is the Monterrey Institute of Technology in Mexico which uses satellite-delivered compressed video to reach 26 campuses nationally, allowing one lecturer to reach over 2,000 students (Hudson 1997a). Adequate telecommunications can also greatly reduce the need to travel to find information. responses. Depending on the current state of the network and what is feasible, different policies for universal service are appropriate at different stages. The longer term goal would be to provide telecommunication service to all households in developing countries, but realistically this will not happen in many countries for many years. In these countries, however, universal access is a realisable goal within a few years. A national telecommunications network evolves through different phases: network establishment, wide national reach, mass market expansion, network completion and complete individual service (Milne 1997). In each phase there are different pressures that require different policy objectives. See the diagram overleaf derived form Claire Milne s work (ITU 1998). The ITU s World Telecommunications Development report of 1998 suggests that developing countries should focus on Universal Access rather than Universal Service. This is because contemporary Universal Service is not a single concept but, rather, a composite, comprising nationwide coverage, non - discriminatory access and widespread affordability. Pursuing availability, accessibility and affordability can be in conflict to each other, e.g. extending the network geographically requires major investment, that the affordability criteria could work against coverage. Instead these 3 goals of availability, accessibility and affordability can be seen as different priorities as the telecoms infrastructure develops. Chosing the right balance of priorities and measures is the task of policy makers and regulators - a task that is both difficult and necessary for equitable access to a growing network. There are many examples of telecommunications supporting a range of rural economic activity such as in agriculture where farmeres can find current prices and levels of demand. In Peru, rural business users estimated that the average saving from making telephone calls compared to alternatives was US$7.30; and the manager of a chicken co-operative in China phrased it like this, When the telephone rings, business is coming (ibid). 3.4 From Access to Service Universal service and universal access measure different things, requiring different policy measures to promote and so they will be treated separately below. However, they should be seen as part of a continuum of goals and policy Page 5 Page 6

4 4. Universal Service 4.1 Development of the concept in High Income Countries The concept of Universal Service has gone through several stages of evolution over the past 90 years. In telecommunications, the term Universal Service was first brought into use in America in 1907 by Theodore Vail, President of AT&T. His slogan then was One System, One Policy, Universal Service. By this he meant that AT&T should have a monopoly, and other independent service providers should be excluded (Mueller 1994). After 1920 telecommunications regulations were focused on the development of a single integrated network and were not directly linked to a policy of promoting household telephone penetration. Universal service as a concept meaning a telephone in every house developed in the 1960s, especially after the Rural Electrification Authority worked to assist and subsidise telephone penetration in rural areas of the United States, largely through supporting the formation of telephone co-operatives. In Western Europe, Japan and Australia, universal service only became a stated policy goal after there already was widespread household penetration. With the move to privatisation and liberalisation in the 1980s, there was concern that the minority without service would be further excluded in the new system, and so policies to promote Universal Service were introduced. In several instances, existing monopoly PTOs used universal service as an argument in favour of their continued monopoly - though without having shown much concern previously. In the UK, the definition of universal service is the provision of affordable access to basic voice telephony or its equivalent for all those reasonably requesting it, regardless of where they live. Often in developed countries there are voices against intervention in favour of universal service, which argue that they will create economic distortions that will lead to inefficiency. Garnham (1997) argues against cross-subsidy or special network provision for universal service. Currently universal service exists in law in two European Countries (Austria and Spain) and in most other developed countries the arrangements for universal service exist in license conditions for operators. (Analysys 1997). Yet even in the developed world, there are many pockets without access. Schement and Mueller (1994) conducted a study looking at the economic reasons why 5.6 million households in the USA are unconnected, focusing on Camden, New Jersey. So, explicit concern for universal service in Developed countries only came when there was already high levels of household penetration (usually over 60%) and so the policy focussed on the minority. Many developed countries do now have over 90% of households with a telephone (the standard measure of achieving universal service). However, it has taken over 90 years in most of these Page 7 countries to reach this point, and universal service only became an issue when most people already had service, so policies only needed to deal with providing access to a disconnected few. 4.2 Universal Service in Developing C ountries The situation is very different in developing countries where a great majority of the population do not have a telephone at home. Very different economic considerations come into play, as there is not the option of marginal crosssubsidy from a majority to a minority. To consider what the possibilities are for universal service, we must look at economic factors of providing service. In many countries of the developing world, the telephone company was run as a state monopoly. Prices were often set at levels below costs for policy reasons, especially for local telephone services. These were often cross-subsidised by prices for international calls being charged for over cost. This system is starting to break down through greater competition reducing the possibility of cross-subsidy, and also the attack on the international accounting rate system from the United States Federal Communications Commission (FCC) which has reduced international call transfer costs to developing countries considerably. This trend for telephone prices to move closer to real costs is seen throughout the globe and is usually referred to as tariff rebalancing. This trend makes local and low cost services more expensive. Also the transparency requirements that often come with a liberalised telecommunications environment make it harder (and sometimes illegal) to have hidden cross-subsidy within pricing structures. Most developed countries which currently have over 90% household telephone service, achieved this while there were these cross-subsidies. This brings us to the issue of costs that could be afforded for telephone service. The ITU (1998) calculates this from considering what the average cost of running a telephone line would be for the operator, and then seeing what percentage of households in the population would be able to afford this (which assumes that telephony as a whole are not to be subsidised). The ITU estimates the median cost of providing a phone line as US$ 2,060 (and with a very efficient operator the cost could be brought down to US$ 1,340). The estimate of the percentage of households that could afford this price is arrived at by assuming that a household could spend up to 5% of their income on a phone service, and then seeing what percentage of households have an income allowing this (i.e. income 20 times the estimated telephone service cost). This leads to a level of household income under which it is not advisable to have a private telephone as the household is unlikely to be able to afford it. As mentioned above this level is usually set so that 5% would cover an average expected phone bill for the year. For example in South Africa the level is set at Rand 900 per month, which means that around 55% of the population are below Page 8

5 the level where they can afford a private telephone (again pointing to universal access rather than universal service as the short-term goal). This analysis provides the approximate figures for the 1.47 billion households in the world: 504 million (34%) households now have a telephone globally, 286 million (20%) households could afford one, 676 million (46%) households could not afford a telephone. (Adapted from ITU 1998). The above calculation does assume current market economics, and that there will be no additional investment for public service (for example from government). Each new phone attached to the network gives an advantage to the people already with a phone they have someone new they can call. This is different from most other forms of infrastructure (for example, you having a water tap does not help me, and it may be bad for me if you use up all the water). This peculiarity of the telephone system is referred to as network externalities and means there is a positive (economic and social) reason to increase the network the value of the network as a whole expands as more people are included by it. For economists, this can mean that the economic value of a person having a phone can be greater than that person s willingness or ability to pay, which can be used as an argument for subsidy. We now move to considering the policy and regulatory issues of providing universal service. The concept of universal service was extensively canvassed at the International Telecommunications Union's (ITU) second colloquium in 1993, with a consensus that there is no fixed and uniform definition of Universal Service (ITU 1993). This concept may mean different things in different countries and regions, and different things in different contexts within each country. It has been a changing concept over time as technology develops and expectations consequently change. This colloquium showed that Universal Service policy operated on three dimensions: Geographic Dimension - geographic availability of service (this is the policy emphasis in the majority of developing countries). Distribution Equity Dimension - accessibility and affordability of telecom service to low-income users. Disability Dimension - accessibility, usability and affordability of service to disabled people, and other specially disadvantaged groups. Universal service is now normally phrased as a requirement of the telecommunications service provider to meet criteria in the following three areas: Availability, Accessibility and Affordability (ITU 1998). Availability: There should be nationwide coverage of telephone service, wherever and whenever required. Page 9 Accessibility: Affordability: Users should be treated alike; there should be nondiscrimination in terms of price, service and quality, irrespective of location or race, sex or religion. Telephone service should be priced so that most users can afford it. These three aspects will be discussed in turn below, but it should be pointed out that seeking to achieve one can be opposed to achieving another - for example, subsidising some users could generate less revenue than would be needed in the expensive major rollout required to attain national availability. These three are not competing goals, but rather different priorites at different stages of the network growth (refer to the discussion of Claire Milne s analysis in section 3.4 above). 4.3 Availability Without a network presence all other issues of accessibility are irrelevant. While new satellite based systems (such as the Low Earth Orbit LEO networks and VSAT technology) mean that the whole globe can now be covered, current economic necessity means that terrestrial networks are the primary concern (though this may change). The key policy issue here is how to encourage network growth into previously uncovered areas. Many of the newer technologies are changing the means of doing this, making it much cheaper to provide service in previously isolated areas. These new systems include cellular radio (either for mobile telephony or fixed Wireless Local Loop systems), satellite based systems, bundled services (voice over video cable, electricity grids and the internet) and various hybrids. [These technical issues are dealt with elsewhere in this Telia workshop and are not considered in any detail in this paper.] As well as technical advances, availability issues can be dealt with by institutional innovations. The points made below refer to a liberalised environment in which a regulator or government body is attempting to promote universal service. In a country with a state owned monopoly, it may be easier to direct implementation, but financing must be dealt with differently usually from increasing artificially low prices through some form of rebalancing. Different organisational mechanisms have been found in various countries to quickly expand their networks. In a number of countries that have privatised their telephone company (such as Ghana, Argentina, South Africa and the Czech republic), high network expansion targets have been set as a condition of the sale. In general these targets have been met. Page 10

6 Particularly in South East Asia, a number of contracted network expansion projects have been successful. Companies are licensed to build the network, and then earn the profits from operation for a time before handing over the network to the State (BOT - Build Operate and Transfer; or BTO - Build Transfer and Operate). In various countries (especially in South Asia), markets have been opened to new entrants with the aim of increasing networks, especially for cellular operators. In Chile, the regulator opened an auction for the amount of subsidy operators would require to service 46 disadvantaged areas of the country. This became a competitive process with different operators bidding to request less subsidy than the others. This led to 16 areas receiving bids for zero subsidy, as the research for this process revealed that there was greater local demand than had been anticipated. This is a major process and should result in telephone service coming to around one third of the Chilean population. (ITU 1998). 4.4 Accessibility An accessible service is one where all users are treated alike, without discrimination. This is normally taken to mean there should be no difference in terms of the services, the price of service and the quality of service to all people. While there can be social, gender, ethnic or religious discrimination in a country, the issue that is most frequently of concern is geography. Often it costs more to the operator to provide a service in remote areas which usually are where the poorer people live. This definition of accessibility means that the operator should not charge more for service in these areas, especially in differentiating between urban and rural areas. A related issue is whether companies serving different areas of the same country (through some kind of regional monopoly) are allowed to charge differentiated rates, and in some countries this is regulated against. In many countries, specific groups of people are considered in need of special treatment. People living with certain disabilities can not use a telephone as others can, and special equipment (such as a text-telephone) is often too expensive for private purchase. In many countries there is special provision of equipment and lower usage charges whether provided by the operator or through some form of Universal Service Fund. In developed countries there frequently are packages or subsidies for other disadvantaged groups, such as the aged, unemployed or people living in poverty (such as the lifeline service in the USA). However in many developing countries the number of the people who would require such a subsidy could overwhelm the system, and other forms of provision can be provided as discussed below. Page 11 A major disadvantage of many forms of individual user subsidy is the high administrative costs. In South Africa it has been estimated that the administration of such a subsidy scheme for disadvantaged individuals could easily cost twice the level of subsidy given out (especially where prevention of fraud needs to be considered). 4.5 Affordability As mentioned above, it is estimated that globally 504 million (34%) of households now have a phone, 286 million (20%) households could afford one and 676 million (46%) households could not afford a telephone. This leaves a large number (in many developing countries a majority) of households unable to afford a private telephone in any mechanism with the current structure of telephone industry. The next expansion of the telecomm networks globally will be seeking to provide services to the 20% who can afford a service, who often are in the waiting lists in developing countries. Various technical innovations and efficiency savings could reduce the cost of telephones, and so increase the percentage who could afford service. However, for at least the medium term a very large number of households will be unable to afford a service. A number of regulatory and tariff issues that can promote wider usage are now discussed. INCOMING CALLS One issue is frequently overlooked in normal modelling of the economic viability of a network in an area. The amount of incoming calls a connection would generate is normally ignored, with the studies rather examining the outgoing calls an area could afford (on the usual caller pays principle). In many developing countries a phone in a rural area would generate many incoming calls from friends and relatives in the cities (especially where there has been a large urban drift) or migratory workers. The issue here is whether the price of the incoming call coming from the urban area will go to the company that invested in the network in the rural area, which is an interconnection and tariff issue. CHURN As Shement and Mueller (1994) point out in their study of phoneless people in Camden, access to a telephone is often the only form of unlimited credit that poor people have access to. The regular connection fees appeared to be less of a problem as they could be budgeted for. Once a phone is installed, a very large bill can be run up in a month (often by people other than the person who has to pay the bill). This can lead to quick default on payment, which leads to disconnection. Once the phone was disconnected, the cost became Page 12

7 insurmountable (especially because the subscriber now has a poor credit rating and so would be required to leave a deposit). This issue is usually referred to as churn, those who fall off the network through non-payment. This can be a considerable problem, for example in the network expansion in South Africa in the last year it has been estimated by Telkom SA that 75% of new rural connections leave the network withi n 6 months from economic factors. This can call into question the whole method of network expansion (though in South Africa there have been considerations to provide some rural connections to meet regulator s roll-out goals without adequate cost targeting of users). FLEXIBLE BILLING This covers a number of options such as different payment packages for high users and low users of the phone, user-settable caps on call charges and discounts for usage under a certain amount in a month. Also bills can be paid over a longer period than the usual one month. One method that has been successfully tried in a number of countries is using pre-payment (which overcomes the potential problem of unlimited credit). 3. Lifeline service: Emergency calls, operator services incoming calls. They can not be disconnected from this level. This system would provide a basic lifeline to all telephone users, though certainly would be fought against by most operators as the fear of total disconnection is often stated as the motivator for payment of bills. The author knows of no examples of this scheme being implemented and the real consequences. A system preventing total disconnection is supported by the network externalities argument mentioned above, where any new connection provides value to all other phones on the network (through providing another person to call). Revenue can be lost to the operator by preventing incoming calls to a disconnected phone. In South Africa, currently only 3.7 million (42%) households can afford to install and maintain a telephone without assistance. It is expected that the introduction of new technology and flexible billing system could increase the number by 1.6 million (19%) households. (Stavrou, 1997) LIFELINE TARIFFS The US Benton Foundation suggests a three-tier system of universal service whereby all subscribers, especially low-income users, are assured of continuous access to at least essential services (Blau et al 1996). 1. Full normal service. If the user is not able to pay for their normal telephone bill, rather than being disconnected they are moved to a reduced service: 2. Basic service limited to: Local outgoing calls Incoming calls Operator calls Access to emergency services Should they pay their outstanding bill they return to full normal service, or if they pay their current bill they stay at this level. Should they also default on this, they are moved to: Page 13 Page 14

8 5. Universal Access 5.1 Universal Access in Developing countries As discussed above, in many developing countries currently neither can many households afford a phone, nor can the operators supply them with one. This makes the developed counties goal of universal service unacheivable in the next few years, and so an alternative implementable goal of universal access has been adopted in numerous countries. This usually amounts to some form of community provision, either through a payphone owned by the operator, some other form of commercial paypoint, or a connection within some form of community centre (often referred to as a telecentre). As opposed to individual subsidy the costs of which are excessive (as mentioned above), costs of providing community access are more feasible. The debate is currently whether community access points should be run on purely commercial lines, or whether direct or indirect state funding is appropriate as they can offer such a great social service, especially in generally underserviced rural areas. The ITU has compiled a list of the targets set by different countries, which is given overleaf (ITU 1998). Targets can also be specified for different kinds of organisations. For example, South Africa has a list of priority customers that Telkom SA is obliged to provide with a telephone in its license conditions that includes schools, health clinics, government offices and community centres. Targets should include reliability standards (percentage of time useable, time to repair a fault, line quality for voice, fax or data communications) and the provision of emergency and operator services. Also the hours of access is relevant (e.g. 24-hours a gay or less). Targets for operator payphones can be included in the license conditions. For example, in many Latin American countries, the (newly privatised) telephone operators were set payphone targets. Often payphones prove to be more profitable than the operator had anticipated. Targets should be stated in terms of the services and standards expected, but should not specify the technology required to meet them. This changes too fast and can be very limiting on operators for no benefit. 5.2 Targets The key issue is what does accessible mean? Universal access has been defined in many different ways in different countries, such as a phone for every settlement with over X population (500 people in Ghana); a phone a certain distance of everyone (20 km in Burkina Faso); or a phone within a certain travelling time (such as 30 minutes). It is recommended that a time criteria is not used as it is open to too much interpretation (e.g. time travelled by who, using what means of transport?) Other countries focus on getting at least one line into all villages and localities, such as Mexico, Thailand and Poland. The national goals can be a mix of different elements, for example China has a policy of One family, one telephone in urban areas and a telephone in every village by the year A key point is that these targets change with time - as one target is met, universal access is redefined at a higher level. For example, in India the principle is, easier and quicker and increasing access to a public telephone for people who can not take a private subscription - and the definitions changes as the network increases, currently any village over 2,500 or a distance of 5 km qualify for a public telephone. Page Public Call Offices Small commercial telecom entrepreneurs have emerged in many developing countries. The number of such Public Call Offices has hugely increased in the last few years, usually with no external financing from governments or external donors. These companies usually offer one or a small number of telephones and sometimes a fax machine. A few also offer other services such as photocopying and even the Internet. Examples of this include the franchised call offices Wartels i n Indonesia and the 5,000 Telecentres prives in Senegal that have blossomed throughout Dakar and now also in rural areas. In Bangladesh, Grameen Telecom is working to establish 68,000 Village Pay Phones (VPP) using cellular technology that will provide access to over 100 million rural inhabitants. It will cost US$450 per VPP that is being financed by the Grameen Bank Small and Medium Enterprise (SME) loans, and is very successful. Grameen focuses on women operators, who they find are more likely to be successful entrepreneurs and responsible in using their money. 5.4 Multipurpose Community Telecentres Offering far greater services than a Public Call Office, in the last few years a movement in support of Multipurpose Community Telecentres (MCTs) has grown Page 16

9 up. A large number of projects are currently being established, usually by funding from international donors, to offer telecommunication in ways to support a range of developmental services. The International Telecommunications Union has been at the forefront of championing these, as has the International Development Research Centre (IDRC, Canadian donor), UNESCO and the World Bank. These MCTs provide telecommunications services (often including internet connectivity), and the majority of these centres offer computer services as well as telecommunications. MCTs are seen as a vehicle to support a range of development activities in a community, using the power of Information and Communications Technology (ICTs) as a tool. It is also hoped that the revenue that is generated by the telecommunications can be used to support the wider objectives of the centre. Many of the MCTs are linked for example to health, education, agriculture or government information services. The IDRC is running the Acacia programme supporting telecentres in Uganda, Senegal, Mozambique and South Africa, as well as other projects in Latin America and Asia. The ITU has supported telecentres in many countries (Benin, Bhutan, Honduras, India, Mali, Mozambique, Suriname, Tanzania, Uganda and Viet Nam) (Ernberg 1998). A key feature of most telecentres is a large degree of community involvement or ownership. Involvement of the local community in managing and running is stressed, and frequently techniques such as Participatory Rural Appraisal are used in developing the centres. However, telecentres have been much slower to get going than the public call offices mentioned above. Most of them have been established by external donor funding and the longer-term viability is as yet unproven. Social sustainability: Ensuring the services are used in the local area and are responsive to social needs. Economic sustainability: This is the key test to whether these centres can survive on a major scale. To promote the use of ICTs for development, developing skills, content and services is as important as providing access. The content required can only be determined by participation of the communities involved, and then seeing patterns of usage over time. There already have been a number of hard lessons that should be learnt for example centres where the equipment has been stolen, where community tensions have closed down centres, where incompetent or fraudulent management have made centres fail. It is clear that Multipurpose Community Centres must be developed through partnerships, between the public, private and community sectors (Benjamin 1998). 5.5 Virtual telephony In Brazil, Telebahia has instituted a service where people can subscribe to receive a telephone number attached to an automatic answering service (without a separate telephone line). Received calls can be retrieved by phoning the service s number and entering a security code, which is akin to the cellular phone voice mail service. This has proven very successful, and has provided an incoming call service, which other universal access systems do not do so well. This also is a relatively cheap way to increase the telephone network in a way that promotes the added growth from network externalities mentioned above. An international effort is now being made to evaluate these projects and develop guidelines for wider use. These pilot projects will only have a significant impact when mechanisms can be found for some form of national rollout to have a bearing on universal access. There are different levels to this goal of providing access to information and communication to communities that have been excluded: Infrastructure: Access to the telephone system (network roll-out) as well as computer equipment (many computers could be sourced from recycling of older computers) Education & training: Training the operators of these centres and the users in the systems, especially computerised and network systems. Information: Arranging links to local and national information sources, and developing local information creation (e.g. community directories) Page 17 Page 18

10 6. Universal service in what? Most of the discussion of universal service assumes that the service referred to is conventional voice telephony. However there are now many other telecommunications services available, and the definition of universal service needs to be specific. As it has been phrased, Plain Old Telephone Systems (POTS) are being replaced by Pretty Amazing New Services (PANS). Are only essential services included (such as voice telephony, including emergency services), or do you include all services in widespread use? Is the level of quality of service at the minimum necessary or at the best of the world standard? For example, Wireless Local Loop (WLL) technology such as DECT can greatly impact the cost effectiveness of rollout in rural areas, providing greater access and service. However, there are issues of whether WLL will be able to support broadband applications. Also, satellite technology, particularly Global Mobile Personal Communications via Satellite (GMPCS) could benefit remote areas though at a high cost. Full Internet access requires more bandwidth than standard telephony. And should include other services be included, such as tone dialling, broadband, fax/modem capacity, operator services, directory enquiries, call line identification, emergency services, itemised billing, call forwarding, multi-party lines and voice mail? Many voices are suggesting definitions of universal service should now include elements of the new information services (O Siochru 1996, ITU 1998). Otherwise there is the risk that this generation of technology to meet voice telephony targets will become rapidly obsolete as demand will soon be for newer broadband services. This also points to the fact that setting goals for universal service should be seen as a moving target. It is important to think ahead and not to choose a technological deadend that will become a liability soon. Some countries such as South Korea and South Africa are actively working to provide widescale computer literacy, including training and skills development in a programme for universal access to ICTs. In the next generation of telecommunications services such as the internet, a lack of skills is as strong a barrier to use as having no access. 7. Policy options This last section aims to bring together some of the issues for regulators and policy makers in promoting universal service. As the International Telecommunications Union states in their World Telecommunications Development Report, 1998: Technology that theoretically provides telecommunication access from anyplace on the surface of the Earth is already available. Universal access is now not so much an engineering or supply-side problem but rather a regulatory and policy challenge (ITU, 1998) Within the overall goal of increasing telephone access in developing countries, different policy options are required for each country. Differences in the state of development of the network, the existing market structure, technology and social goals require different national policies. If a suitable regulatory and policy framework is not in place, then increased investment in telecommunications can increase the access gap between the already advantaged minority and the increasingly isolated majority in a country. Also without appropriate institutions, investment can widen the technology gap between developed and developing countries, as in India where obsolete equipment was being bought rather than using investment to build local technological capacity (Pitroda 1993). This section outlines a few possible directions that policy makers could consider. 7.1 Funding mechanisms Funding for universal service projects can come from a number of sources: Cross-subsidy: Traditionally many operators have used money from profitable services such as international calls to keep the price down of other services, such as local calls. These benefit all users (they are not targeted on the disadvantaged) and in many countries it is only the rich who have phones, and are therefore the ones being subsidised. With increasing competition, crosssubsidy is less easy to maintain as competitors under-cut the prices on subsidygenerating services. Other funds can be raised in several ways, such as: Taxes on usage or revenues License or franchise fees Mandated profit reinvestment. Universal Service Fund: This amount of money usually comes from a levy on operators (most often as a percent of profits or turn-over). Either this is paid to the designated national carrier to compensate for maintaining services in all Page 19 Page 20

11 areas (as with France Telecom); or the fund pays money to specific projects in support of universal service (as in Peru and the South African Universal Service Agency). O Siochru (1996) describes the main instrument for universal service as follows: Sources of finance: Profit reinvestment Sector taxation License fees Interconnection fees Obligatory contributions Instruments: Government direct investment policy Rollout targets License twinning Multiple licenses BTO, BOT (Build Operate / Transfer) License subsidy Tariff controls: averaging, cross-subsidy Targeting certain users Innovation 7.2 Monitoring and evaluation As we have discussed, the traditional measures of telephone availability of teledensity and household usage do not fully capture the issue in many developing countries. Hudson (1997b) suggests that targets should be set at different levels: Level One: Household access Level Two: Community Access (such as through library, church, telecentres or other organisation central to a community) Level Three: institutional access (e.g. schools, clinics, agricultural centres, or government offices) A related role that the regulator can play is one of coordination and organised learning for the range of universal service projects that can develop. Some form of centralised database, preferably on a Geographical Information System, should track the different areas of the country with data on population and other demographics, telephone and other infrastructure services, organisations (such as schools, clinics, libraries and community centres) and related information. Quality of service: Waiting list, time for repair service, percentage of digital lines, percentage of phones with direct dial (national & international), percentage of multi -party lines. Price: Of installation, monthly charge, 3 minute local call, 3 minute national call (all as percentage of average per capita income). Internet access: Number of Internet gateways (per million population), percentage of universities and schools with internet connection. Mobile communications: percentage of land covered by mobile service, percentage of population in land covered, numbers using mobile service, cost of calls compared to fixed-line. (Adapted from Hudson 1997a) 7.3 The process for Universal Service Harmeet Sawhney (1994) describes the process of arriving at universal service as one of developing an overlapping consensus. Reviewing other things that have become universal services in many countries, such as education and the right to vote, he finds that progress came in small incremental steps and the grand theorising came later to justify the practice. The key to making progress is not theoretical discussion on the right definition, but rather practical coalition building to create a general agreement that universal service is in different people s interest. Broad goals should be agreed, rather than attempting detailed national plans that will lead to argument. The actual implementation will happen piecemeal at local level with many experiments, with the successful ones being taken up by others. The ITU suggests these steps in addressing universal service (ITU 1998) Definition: Develop an appropriate definition of Universal Access & Service Information: Collect the relevant statistics regularly Targets: Establish targets, preferably in license conditions Affordability: Ensure affordability, often with gradual tariff rebalancing Appropriate market structure: State or liberalised with effective regulation Funding mechanism: Especially relating to cross-subsidy and US Fund Technology: For example, Wireless-Local-Loop making access cheaper Monitoring: Important to monitor (and enforce) progress and targets The following statistics should be kept to identify trends in Universal Service: Availability of service: National teledensity, and in urban and rural areas. Universal Service: Percentage of households with a telephone, in urban and rural areas and by province. Universal Access: Percentage of households within defined distance of a working publicly accessible phone (e.g. 1 km or 10 km). Page 21 Page 22