Document(s) 9 of 12

South Africa and Uganda in the broader context

The conclusions, lessons, and suggestions in this final chapter are based largely on the empirical evidence from the two countries covered by the study, Uganda and South Africa. How widely applicable, then, are the lessons that can be drawn?

South Africa and Uganda are two specific examples of a process that is occurring, with different variations, across Africa and the world. Firstly, national governments and international agencies have since the 1990s, paid increasing attention to the importance of information and communication technologies, since these are drivers of the new emerging 'knowledge economy'. Secondly, 105 national governments made commitments to the WTO to privatize and liberalize their telecommunication markets under the GATS agreement beginning in 1997 (Bressie, Kende and Williams, 2004). Uganda and South Africa were two such governments. Since these agreements committed governments to similar conditions and processes of telecommunications reform, it is not surprising that some early results reveal similarities across countries. This reform has also influenced the penetration of newer telecom technologies such as the cellular phone. In this case, there were generally multiple licensed private service providers early on, in contrast to the situation for fixed-line telephone provision, which in most countries was rooted in public monopolies. The Research ICT Africa network has produced some interesting cross-Africa comparisons of telecom sectors in seven African countries (South Africa, Uganda, Cameroon, Ethiopia, Rwanda, Kenya and Zambia). These comparisons show that, while different countries have taken different routes to meet their commitments, common patterns exist. Most countries encounter barriers to achieving 'ideal' liberalization. Weak regulators and political lobbying by powerful private and state interests, for example, are fairly common. However, those countries in which the telecom sector has been more fully opened up generally appear to have better overall access at lower prices. Uganda, although far from being in a state of perfect competition, has gone further than most African countries in liberalizing its markets. In contrast, Ethiopia still has a state-run monopoly and telecommunication services are scarce, expensive and difficult to obtain (Adam, 2004). South Africa is in many ways a unique case, due to its higher GDP, much stronger existing telecom infrastructure, and high levels of inequality.

In short, applying lessons from South Africa and Uganda requires an understanding of the contexts in which the observations were made and in which they are to be applied. Referring back to Figures 1.1 and 1.2, we can reconsider Uganda and South Africa as two specific instances of a general pattern occurring within Africa and beyond.

Examining key differences in ICT prevalence and use between South Africa and Uganda also points to differences in causal factors. Some of these are noted in Figures 1.1 and 1.2, while other aspects, especially those related to local culture, have not been fully captured and explored by this analysis. This is true, for example, for the greater gender imbalance amongst ICT users in Uganda than in South Africa. Table 6.1 shows some of these differences in causal factors and resulting ICT accessibility, use or developmental impact.

Similarities between South Africa and Uganda in ICT access and use can equally be explained by corresponding similarities in policy intent, political and economic situations, and external factors. As already mentioned, the overall processes in each country were shaped by the WTO's GATS trade agreement. Both countries had relatively high costs of connectivity, related to still limited markets with powerful telecom operators which could be difficult to regulate. Equally, those managing access centres in both countries related stories of unresponsive service and incorrect billing from landline providers. In both countries, local political troubles had sometimes turned a subsidized access centre into a source of conflict, and some observers had concluded this was another argument for privately owned centres.

Lessons for policy makers

1. Market liberalization is more likely to lead to broader access than telecommunication monopolies with some obligations – especially where the regulator is weak.

Based on South Africa's experience, a monopoly telecommunication market results, not surprisingly, in high costs and a lack of response to consumer needs. The argument often made for telecommunication monopolies – that they are necessary for achieving universal access – does not bear up in the experience of either South Africa or Uganda. In both countries, the relatively more open and competitive cellular markets were responsible for the majority of new telephone users. This increase was never a major part of the original universal access strategies since the popularity of cellular services was not foreseen. Although this is generally a positive development, cellular phone use is still much more expensive than landline phone use, and cellular networks are not well-suited to large data transfers. While new technological developments may change these limitations, some close observers have argued that public investment in telecom infrastructure, including fibre optic cable, is necessary to move a nation towards universal access and to maximize the potential benefits of the information economy (Tusubira et al., 2003).

For Internet use too, the relatively more liberalized environment of Uganda has facilitated universal access. Although South Africa has the largest proportion of Internet users in Africa, it is also one of the most unequal countries in the world and patterns of access reflect this. Rates of Internet use in Bhamshela and Khayelitsha were lower than in many places in Uganda. In both places, dial-up had proven a poor option for Internet connectivity because it was expensive and of poor quality. Only in Uganda (and urban areas in South Africa not covered by this research) had people been able to use alternative methods of connectivity, including VSAT and wireless broadband.

2. Universal access and access centres need to be integrated into national development strategies, especially in rural areas.

Both South Africa and Uganda demonstrate that, left to market forces, certain populations are likely to be excluded from certain types of services and benefits associated with ICTs. Moreover, the ability to benefit from ICTs depends upon the provision of services that have traditionally been within the domain of public services – especially electricity, roads and education. At the same time, ICTs may play a complementary role in larger development strategies. In South Africa, the Multipurpose Community Centre programme focuses on creating structures for government service provision and community media in traditionally disadvantaged areas. In Uganda, the National Agricultural Advisory Service, which is integrated into the overarching national development strategy, focuses on creating demand and linking it with the supply of agricultural information. Both of these programmes can benefit from the presence of basic ICT infrastructure in rural and disadvantaged areas, and can in turn act as 'anchor clients' in creating effective demand for these services, thus contributing to their financial viability in poor areas.

In 2003, both South Africa's MPCC programme and Uganda's NAADS strategy had been operating for a few years, but it was too early to determine what the long-term viability and impact of these strategies would be. Both required the creation of new organizational structures and new working relationships, within government and between government and other stakeholders. This was a time-consuming process that committed to a longer-term vision. Such types of activities can be viewed as social investments in underdeveloped communities. If carefully co-ordinated and adequately resourced, they will create opportunities to ensure ICTs serve the public good.

Any attempt to integrate ICT access into development strategies will be strongly affected by the larger market and regulatory environment. Uganda, despite the fact that the national ICT policy had not yet been passed, seems to have faired better than South Africa, which had, from the beginning, an explicit policy and a special agency to focus on universal access. This was in part because much of this policy, while ambitious, was not clearly defined and planned out. Also, the South African regulator was generally perceived to be weak, especially because of regular interference from the Ministry of Communication. Any integrated ICT policy will be seriously hobbled unless it includes clear definitions and action plans, and operates in an environment where the national telecom sector is effectively regulated.

South Africa has experimented with direct government intervention in universal access service provision. As of 2003, these efforts had not been very successful. This cannot be said to prove that direct government intervention is always a bad idea – but rather that it is affected by larger market and infrastructure issues. The Universal Service Agency might have been more effective had it been set up as a branch of the regulator. As such, it would enjoy the same independence and power vis-à-vis the operators as the regulator, and universal access issues would always be framed within the larger context of the telecommunication market.

Besides being highly dependent on the wider national market, access centres operating through user fees are not adequate by themselves to build local demand. This is because of the multiple barriers to access that many populations face. Low levels of education, illiteracy, low awareness, lack of access to transportation, and lack of money all contribute to low levels of ICT use. They are reinforced by a popular conception of ICTs, and especially computers and the Internet, as the tools of the educated only. Most people are concerned with gaining immediate productive assets. Knowledge of computers is usually seen as a long-term investment for people who want to seek formal, professional jobs opportunities. While this understanding reflects current reality, it also overlooks the potential that these tools hold for other types of livelihood, and especially the potential value of information in productive activities. Unfortunately, this perception is widespread and such uses are therefore unlikely to be discovered spontaneously.

Because knowledge of computers is necessary in a growing proportion of formal-sector jobs, public schools are also important potential equalizers amongst the population. Creating a strategy aimed at giving all public school students a fair chance to use computers and learn related skills is therefore another important complement to universal access strategies. However, there are huge obstacles to achieving this goal in both South Africa and Uganda because many schools are under-resourced and have no electricity. Also, the presence of computers in schools by no means ensures that all students will receive effective instruction in their use or get the chance to apply them in other areas of the curriculum.

3. Large telecom operators seem to profit disproportionately from the telecom services offered through shared access centres, with little of the profit staying in local communities. Regulation can improve this situation by recognizing the value of the service provided by local intermediaries between big operators and the public.

Large telecom operators have much greater political clout than the fragmented groups of consumers and small businesses further downstream. Access centres in local communities are often small local businesses or not-for-profit organizations that actually help generate profit for the large operators by aggregating demand and creating use amongst people who would not be able to use the services by other means such as private ownership. In South Africa especially, the gap between rich and poor is one of the largest in the world, and it is therefore tempting for the large operators to ignore poorer markets. Again, it is the small entrepreneurs who often help to bridge this gap. Regulators can do much more to explicitly recognize the valuable contributions of these service providers both to the public and to the large operators. The latter do business in monopoly or near-monopoly situations, whilst the former face fierce local competition. For these reasons, regulators should seriously consider requiring large operators to give preferential ('wholesale') rates to any agent who either resells public access to telecom services or provides them free.

4. There is a clear practical distinction between phone and Internet access: the former has much greater local market demand than the latter.

The distinction between phone and Internet access may reflect a broader distinction between communication-related and information-related services, where the former commonly enjoy greater demand than the latter (Odlyzko, 2004). In terms of policy, it implies that it is worth thinking of telephone services as separate from, or at least as not necessarily connected with, the Internet. This is especially true as long as GSM (Global System for Mobile Communications) plays a major role in providing basic phone access, since the same technology is not suited to large-scale data transfer as required by most Internet applications. However, where and when VoIP is legalized, or GSM-related technologies become better suited for broadband data transmission, this distinction might once again cease to be important in terms of physical infrastructure. It will remain relevant, however, in terms of barriers to use. Telephones are simple to use and do not require literacy or knowledge of a major world language. Most computer-based applications, in contrast, require literacy and some training and time for mastery. And they are vastly more useful when the user has knowledge of a major world language. Because of this, any strategy to provide universal access to computers will need to be closely linked to public education and adult education programmes.

Lessons for those implementing access centres

1. Set a clear and simple initial goal that is based on a real local need and on which the centre can build.

Early telecentres were often conceived as extremely ambitious projects, offering a full range of services from virtual learning libraries to telehealth applications. But in South Africa and Uganda, they were largely unable to deliver on these high expectations. One of the biggest problems seems to be with the expectations themselves. For example, local stakeholders, including telecentre staff and community members, were often introduced to telecentres with great tales of what they could bring. Whatever the centre did manage, it almost always fell far short of these expectations. Local support for telecentres, essential to their continued survival, often decreased over time and staff became demoralized. Telecentres can reduce risk of big disappointment and lowered credibility by starting with modest expectations that they can in fact fulfil. This also allows local capacity to be built gradually. RANET, for example, began by focusing on the delivery of seasonal forecasts. As many of its sites flourish, it is natural for local stakeholders to experiment with other applications of the informational resources available to them. Any intervention by outside agencies also requires an honest assessment of risk and potential limitations.

2. Externally subsidized access strategies should be based on a full local market assessment that takes into account both current conditions and likely future developments.

Another hard-earned lesson from the early telecentres is that outside agents often did not assess the local market. Rather, they tended to assume that ICT access services were in short supply and that they should therefore go ahead and set up a centre to meet a need. This assumption was often true when the original centres were launched, but the local market often changed quickly –and not always in response to the telecentre itself. It also links to the sometimes arbitrary distinction often made between 'urban' and 'rural'. In the implementation of funded telecommunication projects in Uganda, 'rural' was sometimes used to refer to any place other than the capital city, Kampala. While Kampala is the largest, most important urban centre, Uganda also has large towns. 'Rural' telecentres situated on the outskirts of these towns are still competing with a small but active town market. In Uganda, many of the larger towns have hosted some kind of access centre since about 1998 – the same period in which the first telecentres began.

This implies that those implementing funded access centres should consider how these are likely to support local efforts and meet local needs, rather than usurping and undermining existing local efforts. Perhaps they are providing a particular service that the market cannot, or they are providing access for a particular segment of the community that market-based access centres cannot reach. They should further consider how they might work with and perhaps directly support local efforts, whether they are small businesses or NGOs. Supporting existing efforts is likely to be more cost-effective and to have greater development impact than unilaterally creating a new institution. It may add value to the work of the existing organization. Many local entrepreneurs also have an interest in their community that goes beyond a pure profit motive. Supporting them will in turn support the local economy and may create jobs.

3. Control for complexity and localize decision making as much as possible. Funding needs to be flexible and responsive to local market changes.

Local markets can change rapidly, especially with regard to ICTs. So access centres need to be able to respond quickly and flexibly. This requires a lot of locally based decision-making power, as well as capacity to make decisions. Good management has long been recognized as a key to telecentre success (eg, Fuchs, 1997). In both South Africa and Uganda, early experience also showed that placing scarce resources, in the form of telecentres, in resource-poor communities could lead to local political conflict, especially where the centre was under the control of local government authorities. This is another reason why entrepreneurial models sometimes seem to work better. For some of the people involved in the early Ugandan telecentres, this was the single most important lesson from their experiences. Certainly, a clear, simple decision-making and accountability structure is important. Telecentres that rely on boards filled with local politicians and funding decisions tied to a preplanned budget or a centralized authority are likely to run into serious operational difficulties, as these structures will not allow them to respond to community needs.

4. It is generally not realistic to expect a centre to be profitable immediately after start-up.

Financial viability over the short term is most likely where:

• there is already demonstrated need and willingness to pay;

• a centre is able to count on service of fair quality and reliability;

• running costs are not high and profit margins are fair;

• the access centre is either very simple, or is an addition to an already functioning organization;

• the access centre is in a highly visible place where people come regularly to shop and do business.

User fees will often play a role in sustaining access. This may not be a bad thing, insofar as willingness to pay is an indication of local relevance. However, the barriers that many people face in gaining access to ICTs and using them, especially computers and the Internet, are multiple and cumulative. User fees are an extra impediment for a large segment of the population. Thus, for any particular centre, depending on its aims and intended reach, it is important to consider the likely impact of user fees and the pricing structure on the services offered. For telephone services, simple pay-as-you go access offered at convenient hours has strong popular demand. Without a concerted marketing effort, fee-based access to computers and the Internet are likely to serve only a small segment of the population.

Lessons for those supporting access centres

1. Civil society organizations need to act as watchdogs and place consumer issues related to telecommunications into public debate.

Building telecommunications infrastructure requires huge amounts of capital. Because of this, national telecom services will never operate in an environment of 'true competition', where consumer choice amongst multiple providers promotes efficiency and good practice by competitors. Thus, governments create agencies to regulate the sector. The telecom operators, though, are large and powerful, and the regulators often do not have sufficient political might to manage them. Moreover, the government itself may have stakes in the major operator, as well as the power to override the regulator's decisions. This happened in South Africa on a number of occasions. Such behaviour can result in decisions that do not maximize public interest.

There is clearly a need for third parties which can serve as watchdogs: groups that are not directly engaged in telecom services, not profiting from them, and not directly tied to the government. While the regulator is supposed to play this role, this is demonstrably often not the case, and an outside agitator may help the regulator perform more effectively. In South Africa, for example, the LINK Centre at the University of the Witswatersrand has conducted public-domain research and provided expert commentary on telecommunication regulatory and policy issues. It has done this, for example, for the news media which has in some cases strongly criticized practices of government, the regulator and major operators.

There is an ongoing critical need for third parties to take on this role because of the large vested interests and amounts of money at stake. Major policy decisions are often made behind closed doors and the general public is not usually in much of a position to scrutinize carefully everything that happens. Yet the impacts of these decisions are felt all the way down to access-related activities at the grassroots, which may find themselves struggling simply because they cannot get basic phone service at an affordable price.

2. Small businesses and not-for-profit groups providing various forms of public ICT access often need support, especially in the form of capital, equipment, practical technical skills, and business management and marketing skills.

There are plenty of small enterprises struggling to make a success of providing ICT access and related services. Although the capacity of these enterprises in terms of capital, equipment, skills and knowledge varies considerably, most are weak in at least one of the major areas. The idea of applying SMME support services to access centres has a fairly successful history in Asia, mainly through Grameen. In Africa, IDRC supported entrepreneurial start-ups in Senegal and many of these thrived. In the areas covered by this study, however, there was relatively little support activity although some NGOs were beginning to think about it. In the summer of 2003, a project to replicate the success of Bangladesh's Grameen Village Phone Programme was just beginning in northern Uganda, but was too young to be captured in this research.¹ The Vodacom phone shop franchise was another example of an activity fitting this category, with great success. Other licensed national operators were considering how they could emulate the model to meet their own licensing obligations. Clearly, more could be done in this area, especially given the high demand in both countries for productive, financially viable activities. Third parties should help entrepreneurs assess both potential risks and benefits. Market research especially was an area that was often weak, resulting in many stillborn enterprises.

3. There have been many stand-alone pilot projects and innovative activities, but the knowledge gained from these is fragmented and poorly shared.

There are many examples of stand-alone initiatives – sometimes implemented by local agencies, sometimes by outsiders – that have shown some level of innovation but which were not covered by this research. Many of them address common problems. For example, Wizzy Digital Courier has developed an innovative solution to the high cost of Internet connections. While they have focused on schools in South Africa, the same idea could be applied to many other situations and places. Sharing these ideas and learning together have proven to be a challenge. Perhaps this is because there has been so much activity around ICTs for development in a reasonably short time period. Many of those who have taken independent action have also expressed some frustration with larger agencies and networks that they have found to be too bureaucratic and often ineffective. This has contributed to overall fragmentation.

For third parties trying to find their own role, there is a rich bank of on-the-ground experience in providing access to ICTs, often for a specific development-related goal. A number of agents have worked at documenting and publicizing these. For example, bridges.org is producing a number of ICT-for-development case studies. At this stage within both South Africa and Uganda, there is little reason for unilateral action. Interventions should be preceded by a concerted effort to understand existing efforts and then designed to complement those activities. The challenge is that the ICT scene is constantly changing and documentation becomes quickly outdated.

4. Partnerships between national or regional not-for-profit organizations and local access centres require careful attention to the capacity requirements of partners, and may take time to develop as this capacity grows.

Partnerships between access centres and other organizations were typically ad hoc, sometimes transient, and more common among not-for-profit centres. Such centres, however, often had difficulty maintaining partnerships due to ongoing crises. Some regional structures put in place to support access centres found that maintaining reliable communication with these centres could be difficult; whereas the centres themselves sometimes found such organizations hard to contact and unresponsive to their needs. For example, schools in the Western Cape were unimpressed with the activities of Schoolnet South Africa. So they began their own local network which they hoped to expand. Networking seemed to be more effective where it was rooted in local activities and partnerships expanded slowly, and where the participants had at least some basic potential. External agencies looking to co-ordinate such networks were sometimes far-reaching in their ambitions but had relatively few active linkages to on-the-ground projects.

What are the main lessons for third-party organizations seeking local partners? Firstly, in creating partnerships and networking opportunities, it is best to start small, with concrete activities. These may need to include capacity-building exercises. But if capacity building is not the intention – for example, an access centre is to be used to disseminate information on HIV/AIDS – then the capacity of the centre to effectively carry out such a task should be assessed.

Secondly, not-for-profit organizations are often considered to be prime candidates for partnership, but depending upon the task, small local businesses already providing ICT access or related services may also be valid partners. Partnering with locally owned business has the potential benefit of supporting the local economy while building on existing capacity. Sometimes small businesses exhibit more local ownership than NGOs which may have been opportunistically driven by external funding opportunities. Such businesses also face great challenges in accessing investment capital and sufficient management, market research and technical capacity, and may be equally valid targets for capacity-building efforts.

Research directions

The role of research

The issue of universal access is complex and subject to constant change as new technologies become available and the market changes. For this reason, ongoing research in this area is important. One role of research is to monitor what is happening across countries and to analyse emerging risks and opportunities. This can inform the process of policy revision.

As universal access is a public interest issue, there is a need for independent research and publicly available results. The telecom sector especially has a large body of private sector research often funded by and directed on behalf of major operators. Public research is necessary to inform regulators, smaller operators, consumers and consumer advocates, and to promote transparency. Independent research can also play an important role in monitoring the degree to which reality mirrors policy intent. For instance, in South Africa where there had been a clear discrepancy between policy intent and outcomes, independent research by organizations such as the LINK Centre at the University of Witswatersrand and bridges.org, as well as media coverage of the issue, put pressure on the government to revise its practices.

Much research has been done in this area at the project level and by independent academics, including Master's and doctoral students. While there is a wealth of detail in various case studies, little of this information has been analysed at a more general level. There is potentially much to be learned through meta-analysis. Better channels for sharing and disseminating research on universal access issues would also be useful. Some of the traditional academic methods of disseminating knowledge do not fit well with ICT issues because the information dates very quickly. For this reason, much master's and doctoral research remains unpublished.

The overall arena: actors and relationships

At the national level, ongoing independent research should monitor the impact of universal access policy, with a specific focus on the costs and benefits to various stakeholders. This type of monitoring requires clear definitions of key terms such as 'affordability' so that progress (relative to the definition) can be measured. Other potential areas of research are the effects of pricing structures on consumer behaviour; the impact of specific universal access policies, including the costs and benefits to various stakeholders; international examples and the potential applicability of other policies (for example, those from Latin America) to particular national contexts; and research into the issue of public versus private infrastructure development.

Start-up and sustainability

A key area that remains fuzzy and which is often overlooked is how to determine appropriate strategies for different types of areas. While different strategies are obviously necessary, implementing agencies are usually on their own when it comes to identifying which type of situation they are dealing with. It may be that the market can be left to itself to provide access centres. Or perhaps conditions are such that a subsidy or intervention in setting up a centre would help prime the market. Or it may be a situation where the cost of setting up an access centre outweighs the potential benefits until other conditions – for example, electricity, roads or the local economy – are improved. Research that provides general guidelines and issues to watch out for could help implementing agencies to avoid common errors.

Another area that could be better researched and documented is how best to support entrepreneurs who provide universal access. Some initiatives have already been well documented, including Vodacom's experiences and the Grameen phone programme in Asia. Some needs are already well understood, such as the development of skills in business management, marketing and basic technical maintenance. These particular skills are in high demand. As the importance of entrepreneurs in providing access becomes ever more apparent, these initial case studies and lessons need to be expanded and applied to potential policy in support of entrepreneurial activity. Research can also address the issue of how entrepreneurs make decisions about setting up a business and what types of services to offer. Understanding this could help those who support entrepreneurs to do so in a way that serves the public interest, reduces the risk to entrepreneurs, and maximizes productive potential. For example, a number of entrepreneurs in Lira put great emphasis on customer service and marketing. How could their counterparts in other places be encouraged to adopt such practices and, if they did, would the change be beneficial?

Reach and use

Many not-for-profit telecentres, especially those in or near towns with private competitors, struggle to establish an identity and role for themselves. To survive financially, they often end up competing directly with local businesses on services and prices and targeting the same customers. Further research could identify and document experiences that demonstrate the particular niches such telecentres might fill, ways to extend their reach and development impact, and strategies for survival. And perhaps it could identify when such centres are no longer necessary, in the face of a changing market.

Research on how different people decide on the cost they are willing to pay for a particular service has been rather limited. Largely anecdotal evidence in Africa suggests that ITU guidelines of affordability at 3 per cent of income spent on information and communication services have little connection with reality. In practice, people often pay a much larger proportion of their income on such services, particularly phone use. A related theme is the issue of how people, at the individual and household level, make decisions to start using ICTs and to continue to use them. A special focus on people in the 'subsistence plus' category could lead to greater insight as to when and how to build demand in rural areas. This approach could also be used to analyse differences as a function of languages spoken, sex, age and socioeconomic status.

Finally, one of the most important and potentially positive uses of ICTs is to maintain urban-rural and domestic-Diaspora links. Specifically, how much of an equalizing force do these links exert, given other factors like unequal access to education? What are the likely longer-term trends? Ideally, research could inform policies that favour these interactions and thus increase benefits.

Appropriating technologies

Most telecommunications R&D occurs in the industrialized North, targeted towards populations of the same. There is an ongoing need to focus research attention on determining which technologies might be usefully applied to universal access and national development in countries of the South. (GSM, VSAT and wireless are good examples of technologies that have already been of great value to the South.) Ideally, such research should be carried out in a way that supports and strengthens domestic technological capacity, especially so that domestic businesses can participate and gain.

Last words

Optimizing the benefits that new information and communication technologies present to all citizens of a nation is a goal that presents many challenges, constant change and much to be learned. Access centres are an important part of universal access strategies but are not sufficient by themselves. Other conditions are equally crucial, especially effective national telecommunication operators regulated in the public interest. The services that access centres can offer are not always the equivalent of private ownership of ICTs. Sometimes they cannot equal it; sometimes they accomplish more. For example, private cellular phones allow people to receive calls anywhere (within coverage) at any time. But the collaborative dimension of access centres can sometimes help to spark larger community development processes.

One side of universal access that cannot be overlooked is the potential for giving local people opportunities to provide ICT services and profit from that entrepreneurship. When this factor is included, universal access becomes a double-edged tool for community development. Otherwise, if the profit generated from ICT services is sucked out of poor communities and these services do not lead to local economic benefits, universal access is potentially a tool for further impoverishment. Universal access policy aims to maximize benefits and understand and minimize risks to public interest. Early policy, for example in South Africa, was sometimes overly optimistic in identifying potential benefits. Experience now reminds us that positive development outcomes do not spontaneously emerge from access, although the potential is there.

Document(s) 9 of 12