Telecommunications has made it possible to develop global markets for goods, services, money and information. The rise of the global information economy in turn is transforming human life, nationally, regionally, locally, and within the family. Today, everything is changing because of telecommunications -- the nature of work, relationships with people, media, messages, and patterns of political life. The talk is of information superhighways and global information infrastructures, with the potential to improve the `human capital' -- the health, education and skills of everyone.

• Dr. Pekka Tarjanne, Secretary-General, International Telecommunication Union (Tarjanne 1995)

ABSTRACT:

People's ability to communicate will be a major determining factor in the shape of future trade, economics and development. Developments in communication technology and capacity are creating exciting new opportunities both for trade and for participation in civil society. The shape of these new forms of trade, and the scope of civil participation, will be determined by who has the power to communicate: who is taking part.

The key questions are: Who will get "wired"? When? What determines this?

What may be the outcomes of different development policies? Is there such a social category as "information poverty" and, if there is, what policies will exacerbate or reduce it?

SUMMARY & INTRODUCTION:

Both development economists, and economists who try to understand the implications of a world where information is a primary item of trade, increasingly appeal to the concept of "human capital".

The UN Human Development Report 1996 (UN 1996: 64), for example, quotes estimates that 80% of all economic activity in countries such as Germany and Japan is founded on "human capital" -- skills and knowledge.

It seems obvious that communication is to "human capital" as "the velocity of money" is to money supply: in the end, it is not how much one knows (has) that is important; but rather how much one can make pass through one's hands. Yet that Report does not mention communication as a factor in economic development; nor is communication capacity a factor in its Human Development (quality of life) Index.

In an increasingly knowledge-based economy, information is becoming at least as important as land and physical capital. In the future, the distinction between developed and non-developed countries will be joined by distinctions between fast countries and slow countries, networked nations and isolated ones.

• (Baranshamaje 1995: 2)

Any attempt to assess the impacts of economic policies, and particularly of development policies, must take into account the impact of new communication technologies.

The key feature of the most interesting of these technologies is the possibility they offer of low-cost, many-to-many communication through computer networks.

It is a cliché among authors dealing with this technology to compare the possible impact of this to the undoubtedly large impact on the course of history of the development of printing by Gutenberg and others. It is a cliché which is difficult to avoid while doing justice to the subject.

Gutenberg et al made possible exact communication by the few (writers with publishers) to the many (those who were literate). Without this, it is absolutely clear that the modern form of academic endeavour could not exist; it is probable, in a wider sense, that modern forms of economic activity could not exist either.

The fundamental difference with the new communication technologies is that they are inherently "many-to-many".

This fact lies behind the novelty of the exciting opportunities which the new technologies offer. Some of these will be reviewed at the end of Section 1. Many of them are yet to be discovered. The author concludes that the major risk facing the world, in the face of these opportunities, is an increase in relative information poverty, as inhabitants of the industrialised North move away from the rest of the world at an accelerating rate.

Remember that a significant proportion of the world's people have yet to become full participants in the "Gutenberg revolution" -- because they cannot read. They certainly experience relative poverty as a result of this. The question of whether they suffer increased absolute poverty is an extremely difficult one.

It is possible to construct scenarios in which mistaken or simply absent policies for the development of communication lead to an increase in absolute poverty. Future historians will doubtless never be able to agree whether this in fact happened.

For brevity and relative familiarity, this paper will refer to these new communication technologies as "the internet"; but this is to be understood as including the immediate descendants of the internet proper. (see Section 1)

The possibility of high-speed global communication raises interesting practical questions, too, about future trends in trade and economics.

There is no room for technological determinism here (see Section 2 ). The critical questions are about who controls the technological infrastucture and who has access to it.

However, it is worth examining some of the possible economic changes (see Section 3) which the new technology may facilitate. If nothing else, this will serve as a pointer to the scope of possible changes. A certain amount of specifically internet-based trade and commerce is emerging now. The situation will almost certainly change completely between the revision and the publication of this paper.

The term "internet years" gained some currency while this paper was being written, as a tongue-in-cheek analogy to pet-lovers' "cat years", reckoned at seven for each human year. An "internet year" is reckoned to be about two or three human months. Forecasting more than one internet year ahead is hazardous.

The shape of internet communication in Africa, for example, changed radically (see Section 4) in the first few months of 1996. Entrepreneurs began providing internet access service in those cities where a business market appeared to exist.

It is apparent that the goals and priorities of these pioneer entrepreneurs differ radically from previous internet service provision in Africa, which was focused on Non-Governmental Organisations (see Section 5) and on providing communication as a good in itself. The emergence of commercial enterprise, in a field which was previously the preserve of development organisations, raises questions over how those organisations can promote their goals in the new circumstances.

Substantial development funds have already been allocated to improving telecommunications infrastructure in the South, particularly in Africa (see Section 6).

As such work continues, however, more sophisticated measures of its progress will be required. The measures used in the first 25 years of the internet's expansion may lead to complacency (see Section 7).

What will be the effects of the differing priorities of the commercial and the NGO-based internet service providers? It appears highly likely that the commercial providers' need for rapid return on investments which are large in local terms will lead to a further concentration of communication capability in the cities (see Section 8) and in the existing business elites.

More is at stake than the formal economy. A major motivation for those who invested their effort in quasi-voluntary projects in support of NGOs was to encourage communication to build "civil society" (see Section 9). As is common in development matters, the delivery of aid to and participation by women is problematic when goals are set with reference to external models of the formal economy.

What, then, is to be done?

It is probably not too late to re-direct a portion of these funds from national infrastructure to local and regional infrastructure, and to training, education and community access provision (see Section 10). We -- you -- probably have about five internet years -- perhaps two human years -- to make a difference.

1: A breathless tour of the technologies

The internet is more of a concept than a thing. It is a set of rules which allow rapid transfer of information between computers, more than it is a physical machine that one can touch. It is also best thought of as a new means of transport for information -- the "tracks" over which actual information services "run". In the same way that the building of railways made regional and national newspapers possible, the arrival of the internet (and its successors -- see below) makes new information services possible. It could also be pointed out that railways were essential to the emergence of such diverse economic phenomena as commuter suburbs and summer holidays, and that the people laying the first wooden rails in coal-mines could scarcely have imagined either.

The first physical manifestation of the internet was in September 1969, (Hobbes 1994) when four computers on the West Coast of the United Stated were connected together as part of an Advanced Research Projects Agency (ARPA) experiment.

Notoriously, the purpose of the ARPA project was to develop a communications system which would keep the various parts of the US military's Command and Control hierarchy in communication after a nuclear attack. Military planners in the US (and the other nuclear powers) were deeply worried about the prospect of what they called "decapitation" -- a nuclear attack on a central command post which would leave their forces "headless".

The ARPA researchers -- perhaps the most unconventional group of people ever to get top-level security clearances -- determined that the way to make a communication system proof against decapitation was for it to be totally decentralised -- for there to be no head to cut off. So their design for ARPAnet -- the basis for the internet -- had each computer connected by high-speed data cables to a number of neighbours. When computer A wants to send a message to computer B, it divides it into "packets". Each packet is sent to the neighbouring C with a note of the "address" of B. Computer C looks up the best available route in the general direction of B, and forwards the message. If computer C disappears, A tries its other neighbours. In this sense, each packet is thrown into the network and left to "swim" to its destination. This detail is included as explanation of the oft-quoted maxim "the internet interprets censorship as damage, and routes around it".

The internet spread rapidly through universities, starting in the USA. At a time when only a dozen or so university and defence sites had hard-wired internet connections, thus forming a "backbone", many universities got their first connection through "UUCP". Ancestrally, UUCP is the Unix-to-Unix Copy Program -- a utility to allow two computers running the Unix operating system to exchange files over an ordinary dial-up telephone line. This was adapted to "fake" an internet connection. The maximum speed of data transport achievable over a high-quality normal phone line with modern digital switches (telephone exchange/office) is about 50 kilobits per second. A Southern rural telephone line, under attack by pests over a long distance between the subscriber and an antiquated switch, is more likely to achieve 2400 bits per second.

Following the spread of relatively affordable personal computers from the early 1980s, computer hobbyists developed FIDOnet starting in 1983. (Hardy 1993) The computers which form FIDOnet store messages, usually until night-time, and then dial a neighbouring computer to exchange "what's new" through their modems. Selected computers dial "hosts" with full internet connections to exchange messages between FIDOnet and the internet.

Another network, using different technology, is BITNET, a proprietary protocol linking IBM mainframe computers. Many internet services are made available to BITNET users, and vice versa, through "gateway" computers which translate streams of information between the two formats.

A "real" internet connection (not FIDO or UUCP) may use almost any technology which permits higher-speed data transport than a telephone line. The ultimate is fibre-optic cable: in laboratory tests, over 100 gigabits per second have been transmitted over a single fibre (notoriously, thinner than a human hair). The next is a "coaxial" land-line -- a home television's aerial cable is an example, routinely carrying 100 megabits per second. Telecommunication satellites in geostationary orbit, such as those in the Intelsat and Inmarsat networks, can offer leased 1 megabit-per-second channels, over most of the earth's surface -- at a price.

In order to make use of the internet, an individual needs at the minimum:

• a computer (a 1990 "antique" will do nicely for sending messages);
• a modem (to convert computer files to and from a code of squawks which can be sent over the phone);
• a functioning phone line (of relatively high quality by Southern standards);
• a reliable supply of electricity; and
• an "account" for their modem to dial up (that is, permission to use facilities on a computer which itself has a link to the rest of the internet).

Table 2 shows the availability of telephones and PCs in the top 40 world economies. It demonstrates the wide disparities even within this severely limited range of countries.

Until late 1995, for example, people in most of Africa who had electronic mail used this equipment and only this equipment. Most were associated with Non-Governmental Organisations (see Section 5). Most used FIDO networking. In late 1995, therefore, the actual internet connections for much of Africa, outside Egypt and South Africa, were routed through GreenNet in London. GreenNet is a charitable organisation depending on subscription income from Northern individuals, activists and NGOs, with minor amounts of funding from development organisations. It is affiliated to APC, the Association for Progressive Communications (APC 1996).

Most of the remainder passed through the French government development research organisation ORSTOM (Orstom 1996) in Montpellier. (See Table 1).

FIDO users can exchange messages -- delayed by up to 24 hours -- with anyone on the internet. They can extract information from some databases and parts of the World-Wide Web by sending carefully formatted email to "servers", which interpret their requests and return the results by email. They can participate in some international discussion fora -- but they cannot do anything which requires a continuous, "real-time" connection.

The attention of development agencies and telecommunications corporations alike has naturally focused on the infrastructure question. As Cable and Wireless chair Lord Young told shareholders in 1992: "There are only a billion and a half people with adequate telecommunications, another billion with poor services and more than three billion people in the world who can today only dream of a telephone." (Holderness 1993a) These people have no immediate hope of using even FIDOnet.

Possible interim solutions include, in urban areas, fixed-location cellular phones, which are often cheaper than putting a lot of copper into the ground. However, the amount of digital information which these systems can transport (the "bandwidth", measured in "bits per second"), is not likely to be sufficient for more demanding applications than electronic mail. Installing cellular base stations in rural areas with few potential subsctibers is not a profitable activity, though in some cases it may offer a cheaper means of installing subsidised telephones than copper wire.

There are a number of proposals for Low-Earth Orbit (LEO) satellites which would provide telephone service to a hand-held receiver anywhere on the surface of the planet. The Iridium scheme, (Lockheed 1993) led by Motorola, (Motorola 1995) proposes hand-held mobile phones with 9600-bit- per-second data connections for $2000 (Holderness 1993b).

The similar TeleDesic scheme (Teledesic 1995), proposed by Bill Gates of Microsoft fame and another billionaire, Craig McCaw, promises higher bandwidth. As with Iridium, TeleDesic's coverage for the South is a side-effect of Newton's laws of gravity: it is simply not possible, as a result of Newton's laws, to put satellites into orbits which cover only the US or only the industrialised North. To cover the plum markets using low-orbit satellites, in other words, entails covering the entire world.

Iridium was in 1993 promising service in 1998, and TeleDesic in 1995 was promising to go into service in 2000. It is not unreasonable to guess that the recent lack of announcements from either party is related to difficult investment decisions: the projected cost of TeleDesic is US$9 billion. It is rumoured that Iridium managers have declared their target market to be the "one million people who pay first-class air-fare from their own pockets".

All these systems, based on radio transmissions, suffer from the problem that "bandwidth" in the radio spectrum is a limited resource. (One can no more create new spectrum than one can insert a brand-new colour between green and blue.) Massive use in urban areas would rapidly exhaust it. "Bandwidth" in fibre-optic cables, however, is limited only by the number of cables one can lay and connect; and a single fibre can theoretically provide as much bandwidth as the entire radio spectrum.

Africa One, (AT&T 1996) proposed by AT&T, is a grand-scale solution to Africa's backwardness in telecommunications. The proposal is to spend US$2.65 billion laying fibre-optic cables around the shore of the continent. The total bandwidth would be 40 giga-bits per second. The financial decisions on the scheme were due to be made in late 1996. No news on implementation was, however, forthcoming by March 1997. If the telecommunication authorities of (initially) the coastal African countries pre-buy enough capacity, it may still go ahead. Connecting land-locked countries presents interesting questions of political trust and stability.

Meanwhile, internet technology is not sufficient for the multimedia entertainment and "infotainment" applications envisaged for relatively wealthy consumers in the North. As some Southern countries are expanding their entire internet connectivity from an intermittent 9600 or 14,400 bits per second to a whole 1 megabit per second... "virtual reality" applications which require more than 1 megabit per second per household are being dreamed up. (Holderness 1996b) Revised internet "protocols" to deal with greater need for security, and a massively expanded demand for "addresses", are under development (Castineyra 1992) This is already being transformed into a subset (Malis 1995) of ATM (Asynchronous Transfer Mode). (ATM Forum 1997)

It seems, however, that technical solutions exist at the level of providing workable connectivity (rather than full-motion virtual reality) to countries and territories, at a price. What might be the effects of this happening?

Albert Gore, Vice President of the United States, was quoted by told the World Bank in 1995 that:

The liberating effects of these technologies have been clear around the world. Satellite stations brought medical advice to those tending to the suffering in Rwanda. Radio and TV broadcasts in South Africa promoted the role of voting in a democracy. Wireless technologies are allowing emerging nations to leap-frog the expensive stages of wiring a communication network -- for example, in Thailand, where the ratio of cellular telephone users to the population is twice that of the US.

• (Gore 1995)

A report from the Rand Corporation -- hardly a hotbed of aid lobbying -- noted that:

With greater than 99.9 percent certainty... one can reject the null hypothesis that there is no relationship between democracy and interconnectivity. Furthermore, the coefficient on interconnectivity is large. A single point increase on the interconnectivity scale corresponds to an increase of 5 points in democracy rating....

Governments that try to squelch the new information technologies to protect their monopoly on power do so essentially at the peril of economic growth. This is... precisely what leading analysts have been predicting: "For nations to be economically competitive, they must allow individual citizens access to information networks and computer technology. In doing so, they cede significant control over economic, cultural, and eventually political events in their countries" (Builder, 1993: 160).

Almost everyone who comes to use the new communication technologies can see liberatory potentials -- often, different potentials according to their fields of interest and activity.

If one focuses on women achieving equal participation in society and the economy, then it is easy to see that these technologies for communication-at-a-distance are relatively gender-neutral. (But development like video-conferencing may, within 15 years in the North, challenge this assumption of gender-neutrality.)

If one focuses on universal access to education, then probably no better technology than the internet can be imagined. Anyone, anywhere, can already explore for themselves a huge body of knowledge, and stands some chance of being able to consult leading experts in many fields. (For "anywhere", of course, read "anywhere with phone lines and computers".)

If one focuses on political self-organisation, then enormous potential is apparent for people who have mobilised in geographical communities to build global networks of mutual support. Fishing villages on India's West Coast could make their concerns heard in European Union debates on fishing policy; and communities in Vietnam could offer their experience and support directly to mine-clearance programmes in Bosnia. This does presuppose a common language; and at present the internet handles Serbo-Croat poorly and Vietnamese not at all.

For those of us who are deeply immersed in the technologies it is, however, often tempting to take widespread understanding of these (various) liberatory potentials for granted. In any case, it is easier to write with rigour about problems than potentials.

For those involved in development projects, it is tempting to concentrate on the relative risks of the potentials not being widely realised:

Africa needs to seize this opportunity, quickly. If African countries cannot take advantage of the information revolution and surf this great wave of technological change, they may be crushed by it. In that case, they are likely to be even more marginalized and economically stagnant in the future than they are today. Catching the wave will require visionary leadership in Africa. The World Bank, other international agencies, bilaterals, and NGOs can all help.

• (Baranshamaje 1995: 2)

2: Some philosophy: how revolutionary is all this?

This paper will not deal in detail with the question of whether more information or more communication is, in fact, better than less. Interesting questions can be raised about real choice and freedom. A Northern consumer in a supermarket, as an analogy, may be faced with so many different brands of soap powder that their freedom (or ability) to choose even among the offerings of capitalism is swamped. Only the product which "shouts loudest" gets through. Similar observations apply to newspapers -- and to TV news, especially in a multi-channel environment. (Wangler 1996) Heavy internet users frequently describe themselves as suffering "information overload".

There are two observations to be made on this. Firstly: of those people who have the capability to access vast quantities of information, some will acquire the skills to filter and make use of it. They will become their own journalists and their own librarians. Secondly: in the two years since the World-Wide Web became widely known in the North, it has been confused with the internet as a whole and thus the internet has been popularly defined as a publishing medium. This is incorrect: the fundamental property of the internet is that it is a many-to-many communication medium.

This paper will deal with the possible results of some people -- and not others -- gaining extended power to communicate.

The new communications technologies do open up a range of economic and social possibilities. Particularly when we consider trade in information itself, the quantitative increase in speed and ease of transport and replication is sufficient to generate qualitative changes in the kinds of transaction which are possible.

The tired, but necessary, analogy is with the application of movable type to printing, by Gutenberg and others. This "merely" led to a quantitative increase in the speed and affordability of the generation of identical texts. One outcome of this quantitative change, however, was the development of academic publishing and hence the whole modern concept of science. The internet is already, within a decade of its widespread adoption in US academia, radically changed the practise of academic research and publication in a growing number of disciplines. (Holderness 1992) This will be an additional brake on economic development in places which lack connectivity.

We can no more know the impact of the new communication technology than Gutenberg could have foreseen his role in causing modern science or, arguably, Protestantism and hence, even more arguably, capitalism.

We should be wary of over-stretching the Gutenberg analogy -- journalist Jim McClellan refers to it as an "historical fairy-tale". (McLellan 1995) Experience also shows that many of the social issues allegedly raised by the new technology are in fact very old issues, usefully thrown into sharper focus by a new context.

One comparison is fairly safe. In Northern societies the car has altered geography, removing (for those who have access to it) spatial meaning from the concept "proximity". (In Los Angeles, people measure distance in quarter-hours.) The internet will do the same, but more so, for those who have access to it.

Everyone who is on the internet is in the same place.

Access to cars has, until relatively recently in Northern Europe, been restricted largely to men. Town-planning which was centred on the needs of those with cars decreased the mobility of those without.

There are, however, two difference with the internet.

From the individual point of view, to get connected is expensive: a reasonable computer costs about one year's unemployment benefit in the UK, or about the annual income of three schoolteachers in Calcutta. It is not, however, as expensive as getting and running a car. There is, double-however, as yet no widespread effective equivalent to the bus or taxi -- though the "multi-purpose community centres" proposed in South Africa and briefly discussed in Section 9, and the "internet cafes" spreading in Northern metropolises, provide respective examples.

The other difference is that highways have almost universally been installed as public works, with an explicit goal of universal access. The Information Highway is everywhere being installed by private enterprise, with the goal of meeting the most profitable demands.

To consider what the effects of this might be, it may be helpful to contemplate an extreme scenario.

3: A thought experiment

Imagine, for a moment, a future in which the only multinationals are AT&T and UPS.

In a fully wired world, there is no reason why all economic transactions cannot be negotiated directly between those who have the goods and those who have the needs.

If you want mangoes, and you want them now, why not browse the producers' offers, select an exquisite variety from a farmers' co-operative in East Timor, place and pay for an order, and have them delivered to your door?

AT&T conveys the "bits" of contract and funds; United Parcels Service delivers the "atoms" of physical trade. The bits/atoms distinction was, of course, popularised by Nicholas Negroponte (Negroponte passim) in his Being Digital (Negroponte 1994)

In principle, considering trade in the abstract, all other goods and services -- from crude oil supply and tanker-leases to cigarette papers and horoscopes -- could be traded in a continuous global "farm-gate" auction.

This is no more than some current management shibboleths -- downsizing, just-in-time delivery, delayering and autonomous divisions -- taken to a logical conclusion.

The implications are quite startling. For the past two millennia, for example, the terms of trade between Africa and the rest of the world have been set by the middle-men -- whether they be Phoenicians, or the Swahili-speakers who traded with the Romans, or British imperialists. Eliminating the trader, the wholesaler and the ship-owner holds out the prospect of, at long last, achieving fair terms of trade.

Ultimately, if someone can sew shirts at home and make $4 each, selling them to shops, or even directly to consumers, for maybe $15 including shipping -- why go to a factory and sew them for $0.05 each when they sell for $40?

The real world is more complicated.

Ranil Senanayake, a founder of internet access in Kenya, points out concrete reasons why he does not believe the "mango picture" will happen: "the movement of pathogenic organisms and diseases are a very real threat in trans-boundary movement of goods." (Senanayake 1996) So grain brokers Booker McConnell, United Fruit and so on may be saved by the volume of paperwork required to meet foodstuff export/import regulations. Unless, that is, food irradiation becomes widely acceptable...

"Further, it seems unlikely that that the producer will be able in many instances to capitalize to a level that processes to export standards. As I have been involved in the production, import and export of primary good I know what a zoo this is." It must be asked, however, whether this is to define the problem circularly.

Consider a different future in which a chain of stores like Britain's Marks and Spencer will place orders only electronically, and only through an "all-bells-and-whistles" system of its choosing. The cost of connecting to this system is $2000 down and $1500 a year -- and it can only be done in the capital city. The smaller shirt factories, and all those in the countryside, are driven to seek poorer markets, or go out of business. Wages are driven down and economic power is concentrated in the hands of the wired few.

This is more than speculation. "Air freight is now so cheap to make the shipment of humble vegetables economic and computers enable Hilbre and its customers to make, track and record orders all the way from field to dining room table.... If you are buying baby corn from Marks and Spencer tomorrow, it may have been picked in Zimbabwe yesterday." Hilbre is a farm an hour's drive west of Harare: owner Ian Gordon also exports to mainland Europe, Australasia and the Far East. (Prest and Bowen 1996)

In any case, an increasing proportion of trade in and with the developed world is trade in pure "bits" of information. The imminent advent of electronic cash (Chaum 1990-) is likely to create a parallel economy, largely divorced from the "real world". (Holderness 1996a) Initially the items of trade will be information, and advice. Notably, the most visible item of trade on the internet today are news and pornography (in that order).

This "parallel economy" may not immediately be an issue for people who suffer a severe shortage of actual cash, let alone the virtual kind. In 1994, however, the European Union's Directorate-General XIII projected that nearly half the new jobs created in the EU before 2010 would be information-related (Wilkinson 1995) The major area of uncertainty in that projection is whether those jobs will be created in the EU -- or in the few high-literacy, highly-connected, relatively-low-wage areas of the South. The low-literacy, poorly-connected areas can be projected to slide to an even lower rung on the wage economy ladder. (Holderness 1995)

The future world where all trade with all is not a thought-experiment confined to these pages. People are busily trying to build it.

4: Entrepreneurs wire Africa

World-wide, there has been considerable expansion in the number of people with access to telephones.

Over the past ten years about 200 million new main telephone lines have been installed world-wide, bringing the total to some 575 million. The proportion of lines in lower income countries increased from 20 per cent to 30 per cent over the same period. Progress in the right direction certainly. But the lower income countries contain 85 per cent of the world's population -- there is a long way to go.

The grossly uneven distribution between population and access to telephone lines is frequently illustrated in terms of teledensity -- the number of main telephone lines per 100 inhabitants. Average teledensity in high income countries is now 49, increasing from 38 ten years ago. Over the same period the average teledensity in the rest of the world rose from just 2.0 to 3.5, some countries even experiencing a drop in teledensity as the installation of main lines failed to keep up with population growth....

In most high income countries about 10 per cent of the population live in the capital city, which is served by about 10 per cent of the telephone lines in the country. The situation is dramatically different in many low income countries with a predominantly rural economy. Here only 4 per cent of the population is concentrated in the capital city, but they benefit from nearly a third of the telephone lines. Rural areas suffer from really severe telecommunications blight.

• (Sharrock 1995)

In Ghana, for example, in Spring 1995 the country's total network connectivity was ten 14,400 bit-per-second leased lines to Cambridge, England, costing $7500 a month each, reports Dutch journalist Michiel Hegener. (Hegener 1995) One was used by the SWIFT inter-bank clearance system, and another by the air-traffic control network SITA -- which reaches parts of the world which no other network touches and plans to offer internet services wherever governments allow competition with the state telephone company.

By early 1996, the private Network Computer Systems internet host in Ghana's capital had 140 subscribers paying US$1300 a year each. Martin Mulligan of the Financial Times points out that this is the entire annual income of a Ghanaian journalist. (Mulligan 1995) "Our customers are expats, large companies, and a few Ghanaian researchers," Network Computer Systems Deputy Director William Tevie told Hegener. (Hegener 1995)

UUnet-Pipex is active in South Africa and half a dozen other countries in Sub-Saharan Africa, and is looking for partner companies in more. (Pipex1996) Business Development Manager Ashley Oliver thinks aloud about the company's motivation: "Opportunism would be the wrong word... 'accident', no: We followed up on African contacts when some other companies regarded the continent as wholly impracticable. Africa is not significant in terms of our total business, but we'd hate there to be a part of the world where we didn't have a presence." (Oliver 1996)

Mr Oliver lists the requirements for commercial investment in an internet service provider as, in this order:

• Political stability
• Availability of bandwidth -- though the possibility of starting up with a satellite link makes this nearly universal, at a price
• Availability of technical expertise
• Telecommunications deregulation

And what does Mr Oliver believe the effect of wiring Africa will be? "It's a genuine stimulant for communication at all levels. The internet is an adaptive technology and does tend to produce what communities require, rather than meeting the goals for which it was put in in the first place. I guess it's about 'empowerment'."

Ashley Oliver estimates that it costs about US$500,000 to set up a full-service internet service provider (ISP). "If you're the smallest phone company in the world, that's not a lot, but if you're a community-based enterprise it's a great deal of hard-to-get hard currency."

To compile an exhaustive list of commercial projects in Africa would be exhausting. It seems reasonable to predict, however, that within the next three years, most capital cities which are not in the thick of armed conflict are likely to have a satellite connection.

5: The internet and Non-Governmental Organisations: actors and facilitators

The first steps towards wiring Africa were taken by intrepid activists with a mission to communicate, fanning out across the continent with rucksacks full of modems. (The French did their bit in a rather more elegant way.)

By January 1996, it was possible to use electronic mail at least in the capital cities of most countries. The breakdown of technologies was as follows (excluding South Africa):

Table 1: Email hosts in Africa, January 1996

 internet & BitNet:	 7* 

 GreenNet/gnfido	21

 other FIDO		 3

 ORSTOM UUCP     	14** 

 other UUCP		 8

 Total:			53

 From an internet information posting 
 maintained by Randy Bush 

 *	(Egypt, Tunisia, Algeria, Zambia) 

 **	(Dakar has IP-over-X.25 too)

 Note that the version 15 of Landweber's
 map  (Section 8) shows an increase in
 full internet access since this 
 table was issued.

Between mid-1993 and the end of 1995 GreenNet forwarded 2.2 million messages between the developed and less-developed worlds, at a total cost of US$80,000: an average of US$0.036 per message. (Banks 1996) It did this mostly using the lowest technology applicable to the internet, direct-dial FIDOnet connections.

Michiel Hegener calculates that to move a 2000-word message between the Netherlands and Ghana costs US$34 if you make a voice call from Amsterdam and read it out; $7.00 if you send it as a closely-typed fax; and $0.20 to $0.40 if you send it as email. If you put it in the post, it may or may not arrive. (Hegener 1995)

This kind of low-tech electronic mail is thus the only kind of inter-continental communication that many Southern groups can afford.

Karen Banks of GreenNet says that, when she started six years ago, "no-one else was working with local indigenous initiatives [in the South], and there was no technical or moral support. One reason why GreenNet's partners in Africa have been so successful," she says, is "that there is a personal, co-operative working relationship -- so they felt neither isolated nor dependent." (Banks 1996)

Though the FIDO networks thus established are slow and cumbersome by the standards of Northern techno-enthusiasts, they meant that by the beginning of 1996 NGOs in Africa had better connectivity than many or most governments.

Karen Banks identifies one reason for the success of these projects in their own terms as the establishment of an equal co-operative relationship between users and initiators in the South and facilitators in the North. (Banks 1996) There are two further reasons:

• the users and initiators in the South are motivated by a pre-existing perceived need to communicate regularly and rapidly with counterparts elsewhere in the world; and
• the guiding principle of the facilitators in the North is simply that communication is A Good Thing.

This is not to say that activists such as Banks proceeded from a theoretical assessment of the importance of communication to development. Rather, is seems likely that many found themselves simultaneously possessing interests in the politics of development and in electronic communications, and created for themselves opportunities to combine the two. Attempts in interview to get Karen Banks to say why she had chosen this field of work produced:

"I was very interested in the politics of the continent, but am not and will never be an academic of any kind nor a politician."

Why, the author asked, communications and not something else?

"It was something that used my mathematical bent... it had to be good for the world in general -- it was something that I absolutely knew I had to do, and that was it." (Banks 1996)

6: Development programmes

The Gondwana project, based in Belgium and Zimbabwe and dedicated to improving connectivity worldwide, has identified at least 40 agencies and NGOs which have internet-related projects in Africa. (See Appendix 1.) (Gondwana 1996) One, for example, the US-based Leland initiative, has $15 million to spend in 20 countries. It was launched in January 1996 after US Vice-President Al Gore had invited the USAID agency to lead a U.S. effort to "bring the Global Information Infrastructure to Africa". By the end of January 1997 it had signed Memorandums of Understanding with the governments of Mali, Madagascar, Rwanda, Mozambique, Ghana, Cote d'Ivoire, Guinea, Benin and Guinea-Bissau. In February 1997 it commenced a programme of technical training for African instutions and individuals. (Leland 1997)

In March 1995 a World Bank report noted that, using Very Small Aperture Terminal satellite dishes, "full Internet nodes could be established in the 47 sub-Saharan African countries currently lacking this level of service for approximately $30 million, including the first year's operating cost and exclusive of local labor costs." (Baranshamaje 1995: 14)

Compare the situation in India, where in late 1996 the author was able to identify two internet service providers and a total of 23 FIDO "bulletin boards" in the entire country (See list, Appendix 2). These bulletin boards' description of their content gives a strong computer-hobbyist flavour. Subscriptions started at US$8 a month -- payable in hard, not local, currency.

The Indian state telecommunications enterprise has been criticised for pricing independent operators out of the market by charging such high prices for the relatively high-speed connections they would need that they cannot compete with the prices it offers for 9600-bits-per-second dial-up access in Bombay and Delhi. In India, a high-speed link is 64 kilobits per second; the academic ERNET has lines of this speed from Bombay to Bangalore, Juhu and the USA. The university sites in Madras and Pune must make do with 9600 bits per second each -- whereas the author is struggling to meet his information needs with three times that capacity for himself alone.

Internet connectivity in India was thus "thinner" in 1996 than that in the African continent. However, the opening of the Indian economy to direct inward investment is likely to produce pressure for a very rapid change in this situation.

7: How much of the world is wired?

From the beginnings of the international expansion of the internet, statistics have been maintained on the numbers of states and territories in which the network is accessible. These record an impressive growth, especially when presented in graphical form. Larry Landweber's maps, distributed by the Internet Society, categorise territories as having full internet access, Bitnet but not the internet proper, electronic mail only through UUCP or FIDOnet links (see Section 1) or no connectivity.

Landweber's latest map (Landweber 1996) suggests an almost-complete project.

Almost the whole world, it seems from a casual inspection of this map, has turned internet-coloured. The sun never sets on the internet; it appears to reach everywhere except some war-torn corners of the world.

Consider, however, that only in Europe, Australia, North America and Japan is there a genuine national communications infrastructure. There may be a full internet connection at the University in Ulan Bator, but ten kilometers away there are no telephones.

As a first approximation at a realistic map, the author has faded the colours of non-metropolitan areas outside the OECD countries. In addition, in an attempt to correct the perceptual distortion inherent in the map projection used, the author has "greyed out" major tundras, icefields and deserts:

The picture is rather different; and it is not dissimilar to a map of per-capita income.

This map, it should be noted, is an approximate freehand adjustment for illustrative purposes. The author is cuurrently seeking funding for a research project to develop accurate metrics of the availability of the internet to people rather than to territories.

8: Who is getting wired? The implications of exclusion

We should be aware that the geography of communication capacity is much more complicated than a North-South divide. The author lives in a largely Bangladeshi-populated area of London, ten minutes' brisk walk from the Bank of England. That walk crosses the route of a British Telecom tunnel carrying huge quantities of fibre-optic bandwidth for the City financial institutions. But cable television -- which can be used as a low-cost, high-speed internet connection -- will not be available in this pocket of the city for several years. (It must be understood that the author wants cable television only for the side-effect of fast internet access, not for any frivolous reason.)

Nevertheless, the disparities in the South are enormously greater. The disparities between urban and rural areas in the South are great and some current development policies will increase them. The Action Plan from the first International Telecommunications Union World Telecommunications Development Conference, held in Buenos Aires in March 1994, set these goals:

Special assistance to the least developed countries is given top priority. And for the LDCs specific targets were set for the year 2000:

Fully meeting the demand for telecommunication services in urban areas. This means virtually eliminating the waiting list for services by that year, which translates into an average urban main line density of five per 100 population

Achieving a rural main line density of one main line per 10,000 inhabitants.

• "The Missing Link" (Sharrock 1995)

The FIDO systems and other low-tech solutions cannot be expected to meet the entire long-term communications needs of the populations they serve. But they do have a mission to extend communication to as wide a range of their societies as possible, and to offer support and training.

There seems to be a risk that large-scale projects such as the Leland initiative will, in many African countries, be left behind as commercial enterprises move to meet business needs.

Many of the small-scale projects are concerned that the income upon which they depend for their training and outreach work is jeopardised by the arrival of commercial ISPs, creaming off their heaviest users. Those users cannot be blamed for preferring a fully-interactive connection to the delays of store-and-forward technology.

Some of Karen Banks's colleagues at GreenNet are outraged by this development, seeing it as commerce threatening their strategy. "Mukla in Uganda is an indigenous organisation," says Jill Small, "which has been working in the country for the past 10 years... and now..." In January StarCom arrived in Kampala, with the backing of US Sprint and the Norwegian phone company Telenor. It provides 1 Mbit/second of full internet capacity in Kampala through a "VSAT" satellite dish. (Banks 1996)

But Mike Jensen, who was in at the beginning of WorkNet, and now manages to combine street-level assistance to community groups with high-level conferences, stresses that the people on the ground are more pragmatic. (Jensen 1996) Charles Musisi in Uganda has negotiated with StarCom for Mukla to use its bandwidth -- part of that set aside, under the agreement with Norway, for free use by universities.

In Dakkar as in London, commercial ISPs start by wiring the richest sectors of the population. In other words, bandwidth is distributed by ability to pay. Since increased bandwidth will increase ability to pay, there is a serious risk of starting another vicious circle of increasing economic inequality.

High-quality connectivity is also at present being made available only in the cities. It can be argued that a major reason why people migrate from the countryside to towns, and from small towns to cities, is "bandwidth": the higher communication capacity of a crowded place, of a factory compared to a rural workshop. It is not entirely fanciful to suggest that current trends in telecommunications will exacerbate the problems of urban migration -- when, in theory and leaving aside the economics -- they ought to do the opposite by dissociating communication from physical proximity.

"Information poverty" clearly exists as a relative condition. The privileged -- overwhelmingly concentrated in the North -- are moving rapidly away from the rest of the world's population in terms of their (our!) unprecedented access to information and communications.

This is enough cause for concern. The same communication technology, used in broadcast mode, ensures that almost all citizens of the planet are informed of the conditions of life of the most privileged.

Watching street children in Calcutta watching "Dallas" and MTV on satellite television raises a number of interesting philosophical questions about the nature of poverty. It must raise, too, immediate political questions. What will be the effects on the aspirations of this generation?

It is widely accepted by ex-citizens of the ex-German Democratic Republic that the widespread availability of Western television played a major role in motivating people to seek political change and "openness". (All the ex-East-Germans with whom the author has discussed the matter speak bitterly of the experience of discovering what the West was actually like, after the changes had been achieved...)

One immediate and clearly apparent result of the massively increased visibility of relative poverty in the past decade is massive resentment among the younger generation. Some of this may be channelled into backlash and fundamentalist rejection: the author suspects that the burning down of a Kentucky Fried Chicken franchise in Bangalore in February 1996 may turn out to be a kind of iconic event -- but whether its historical equivalent turns out to be a church in Dresden in 1968, or a beer-cellar in Munich in 1923, remains to be seen.

One risk which development organisations can and must address is that inequality of access to communications may promote increases in absolute poverty. The scenario above, in which control of (for example) textile export markets is concentrated and wage levels in the remaining home-market sector are depressed, is a risk to be taken seriously.

It is not clear how any of this will affect the very poorest -- who have only heard gossip about telephones, perhaps -- but it is all too often unclear what any kind of infrastructural development has to do with these people.

The consequences go beyond the economic sphere. Political power in these days rests to a significant extent on access to information and the means to disseminate information. This, too, is being concentrated in the hands of urban elites.

9: Connectivity and Civil Society

In the sense popularised by Czech poet-president Vaclav Havel, "civil society" is all the clubs, associations, pressure groups and play-groups without which there is, indeed, no such thing as society: only investors, hired hands and consumers.

At a workshop on women and technology in Delhi, participants were asked by an aid worker "but how does email improve the quality of women's lives?" Their answer should be sufficient: "It allows us to communicate!" (Banks 1996)

The participants also recounted how email had enabled them to communicate the actual results of an election count to colleagues in Delhi, before "revised" results could be published.

In South Africa, networking was seeded by people who can now admit to being associated with COSATU, the Communist federation of trade unions, and with Church-based human rights groups.

Their goals were precisely to help "civil society" to communicate.

WorkNet started in 1987 as a FidoNet operation. Andriette Esterhuysen, now its director, says the motive was simple: "I was working on human rights information, and wanted to feed information about South Africa to other countries directly, not through the North." From the beginning, WorkNet workers travelled the country with rucksacks of modems, connecting community organisations in small towns and villages. (Esterhuysen 1996)

Now rechristened SangoNet, the project is one of many full-internet service providers in the country -- and the low-tech part is still essential to communication with small towns and villages.

Andriette Esterhuysen reports "a lot of discussion about community access in South Africa -- for example through multi-purpose community centres. But there's not a lot of clear direction or commitment from the centre." At present, the onus for providing that sense of direction almost certainly rests with SangoNet.

Provision of connectivity through community centres of various kinds does seem to be the main route to providing access for civil society. In particular, in most of the world connectivity provided to commercial enterprises will reach only locally-rich men. Collective or communal access points seem to be the only way to extend access to women, children and male employees.

It might be objected that large areas of the world have no libraries and poor schools: why worry about modems for people who have no books? But is there a reason why aid agencies should not assist these areas to leap directly to remote-learning capability -- in other words, for development aid financed on the back of interest in communications to create "virtual" libraries and schools where there are now no "real" ones?

10: Conclusions

Communications policy must be a part of any aid and development policy. The exact consequences of growing information inequity are hard to predict precisely, but the chances of them being beneficial to human welfare as a whole are vanishingly small.

In those countries in the South with reasonable political stability, commercial enterprise is likely to provide connectivity, to the capital city at least, in the very near future. Initially this may happen through entrepreneurs -- with overseas backing, which may include logistical support by aid agencies -- installing satellite dishes.

In the longer term, projects of the scale of Africa One -- the proposed high-capacity fibre-optic communication ring around Africa, described in Section 1 -- are likely to mature to meet growing demand. It is also possible that projects to provide global satellite connectivity, such as Iridium and/or TeleDesic and/or one of their competitors will render physical location practically irrelevant, for those who can afford the receiver equipment.

It is the unstable countries -- those which US aid agencies are rumoured to be avoiding -- which will need serious assistance to acquire publicly accessible internet connectivity. As discussed in Section 1, direct satellite links such as those promised by Iridium and TeleDesic offer a solution only for an absolutely wealthy minority.

In this sense, the problem for development agencies and economists is not now one of connecting countries, but of connecting people. To be provocative, the author proposes that country-level infrastructural support be provided only to those countries which in the throes of civil war.

Development organisations should form relationships with these commercial ISPs where they exist, as with the Norwegian support for StarCom mentioned above. They should use these relationships, and their influence with governments, to promote telecommunications regulation regimes which promote community access.

One possibility is to lobby countries to impose "Universal Service Requirements", with timetables for compliance, as a condition of licenses to new ISPs. A telephone company operating under a Universal Service Requirement is legally bound to provide service to anyone who asks for it, wherever they are in the relevant territory; the extra charges which the company may make for remote connections are regulated. This, however, runs counter to a very powerful "de-regulatory" current in world-wide telecommunications policy. It may be practicable, however, to require operators to provide low-cost connections to schools and community facilities as capacity is installed in their areas.

Commercial enterprises may well resist even such mild regulatory initiatives. There appears to be an illogical prejudice that all regulation should be resisted. It is illogical because efforts of any kind to promote widespread access to communication, beyond the very small elite markets which exist now, are building a potentially huge market for such services in the near future. Commercial enterprises should be reminded at every opportunity that the large market for private internet connectivity in the United States was built entirely on university students being given access free at the point of use -- and then looking around for ways to stay connected when they finished their courses.

The primary focus of financial aid for communications should be shifted immediately from the country-level infrastructure programmes to the village level -- with the exception of International Telecommunications Union-level support for programmes such as Africa One. This reverses the priorities set out by the Buenos Aires Conference and listed in Section 8.

The greatest effects have over the last decade been achieved by the programmes with the lowest budgets, such as GreenNet's. There is scope -- and funding can in principle be redirected to -- for hundreds or thousands of such outreach projects to work at local level throughout the South. The major restriction on the emergence of such projects may be the supply of individuals having enough persomal committment to the concept to go through the organisational hoops required to obtain funding.

Such projects stand the highest chance of avoiding the "rusting tractor" outcome of failed development projects. Maintenance of computers and internet connections in the South is difficult; it is likely to be achieved only where there is a close connection between indigenous organisations with the need, and outside facilitators with the means.

Such organisations also need to shift their focus from the country level to the village level. They may find that their work becomes less "glamorous": eventually it will integrate completely with health and education aid work. Eventually, in other words, telecommunications should disappear as a separate concern, and become an integral part of human capacity-building work.

home

This preprint is © copyright 1997 Mike Holderness; moral rights are asserted.

---

This paper was presented at a United Nations University INTECH seminar in Maastricht, Netherlands on 25 October 1996.
Comments are very welcome.

---

The sections:

1: A breathless tour of the technologies

2: Some philosophy: how revolutionary is all this?

3: A thought experiment

4: Entrepreneurs wire Africa

5: The internet and Non-Governmental Organisations actors and facilitators

6: Development programmes

7: How much of the world is wired?

8: Who is getting wired? The implications of exclusion

9: Connectivity and Civil Society

10: Conclusions

Table 1: Connectivity in Africa

Table 2: Multimedia access: top 40 territories compared

Appendix 1: Projects Related to Internet in Africa

Appendix 2: Access in India

Biblography and links

Glossary

---