Top Story

Interest is using ICT for development is probably well over 20 years old. But interest in the digital divide seemed to come into sharp focus with the G8 Dot Force three years ago. The attention span of world politicians is notoriously fickle. Other important issues - Afghanistan, HIV/AIDS and Iraq - to name but a few always seem to come along. Democratic politicians are often gone in five years. So as we’re about halfway through the five year cycle it seemed like a good idea to take stock of what’s happening on the digital divide front in Africa. In this off-the-record briefing, Russell Southwood tries to make sense of the successes and failures up to this point.

Take these random examples from the frontline of Africa’s digital divide initiatives:

- There is a European fund that seeks to invest in African ICT projects. Over 4 years it has only invested in 8 projects. It has a total of 2 million euros to invest. Like many organisations formed to address the digital divide, it has sought to invest for a (lower) commercial return in companies and organisations doing socially useful things. The wet dream for this type of investment would be as follows: an application that involved women in a rural village doing something that affects health or mortality rates.

There has been much talk of new models of social investment but the actual practice has been a confused mix of traditional NGO funding with an overlay of upbeat venture capital rhetoric. Not surprisingly, it has proved extremely difficult to get the crosshairs on the target when the social needs are great and the markets (with users who can pay) are tiny. Doing good and doing business can overlap but they are not always the same thing.

- In a Northern rural province of an African country there is an NGO cyber-centre. The huge sign announcing it has been pushed over but people still seem to know where it is. It has not had internet access for over a year. The cost of connectivity went up and it could no longer afford it. A lone child plays games on one of the computers. A grant application is pending but the process has been slow, probably at both ends. In the same town there are two busy cyber-cafes with 15 to 20 computers respectively. Connectivity’s not great but people are queuing to use them,

A telecentre in another country is run by people coming out of the public sector opens between 9-6pm. Asked when users would most like to use the Centre, they replied:"6-7 pm". A school in yet another country has solar-powered computers but no money to connect to the internet. A high-profile telecentre with a great deal of support from the private sector and the "great and the good" in yet another country (which we covered extensively) has collapsed.Its private sector backers said it would become self-sustaining and be the start of a chain of cyber-centres.

Only a small number of people are acknowledging any of these problems and beginning to talk about the need for more entrepreneurial attitudes and a hard-headed discussion about financial sustainability.

- A European NGO launched a portal to encourage trade with the South using e-commerce. But they had few contacts in the South with which to start trading and they ran out of money and closed down before they could correct this rather fundamental oversight. From the discussion list they ran on the topic it was clear that they had little or no detailed knowledge of business ­ e-business or otherwise ­ or of the mechanics of international electronic transactions.

The African regional group of the UN ICT Task Force has met several times and spawned a number of poorly used discussion list topics. These are not insincere or dishonest people. But it has simply proved very difficult to carve out an agenda for action ­ whether policy or project specific ­ on which it can focus. It is also not entirely clear what its parent body ­ the UN ICT Task Force ­ has achieved.

Many of those working on digital divide issues share the same basic assumptions and analysis. In headlines, it might read: Africa is lagging behind in ICT terms. ICT can overcome various development problems. Create a facilitating policy environment and ICT will start to overcome these problems and in so doing, close the digital divide.

That started being said at high volume three years ago and the mantra is now in danger of sounding pretty thin. With a number of exceptions, all involved (NGOs, Government and the private sector) have difficulty converting generalised propositions like the one summarised above into actual working examples of how ICT will make a difference. It has all too often been a case of solutions in search of problems. As a speaker at a Digital Partners conference in Accra observed:"ICT experts are not tuned to creating solutions (in this context) and NGOs have few skill sets to find solutions (using ICT)."

All too often applications have been created for a user (or in commercial terms a client) group who are often not driving the discussion, whether for example farmers or teachers. ICT development rhetoric insists on the primary virtues of working with users. But nearly always the "gatekeepers" are Government or other NGOs. It ought to be the teachers of farmers defining the service or software but somehow the process is strained through secondary "clients" (ministries, software developers themselves, other NGOs) who become the organisation to whom the grants are made.

Each of these random examples (and there are many more which could have been included) contain some or all of the fundamental assumptions that launched a wide range of digital divide initiatives. It is often said in Hollywood that there are only a few stories in the world: for example, boy meets girl, revenge and so on. So it is with digital divide initiatives: there are perhaps five basic assumptions and the time has come to look at whether they are working or not.

Assumption one: Business can do it

The assumption here is that a mixture of less regulation, pump priming funding for initiatives, entrepreneurship and skills transfer will all make things happen that didn’t before. The reality is that African technology markets are amongst the toughest in the world. They are nearly all relatively small; carry high levels of political risk and in many cases an overhead of corrupt payments ("marketing and facilitation costs"). There are only really three large markets: South Africa, Egypt and Nigeria.

The process of finding a good idea, piloting it and "rolling it out" stumbles on the sheer scale and number of obstacles to doing business continentally: tariff and licensing barriers; the high cost of air travel and the small size of the management cadre to name but three. Few companies can get the savings from scale by operating across the continent that you might get elsewhere: a handful of multinational companies, one or two regional players (for example, Econet) and the brave but struggling Africa Online. If we take the internet it is not hard to see why. It is small, almost entirely price-driven business with low returns except from corporate customers. Until recently, there has been little or no competition for its primary commodity, international bandwidth.

With little or no significant intercontinental infrastructure between countries Africa has had to pay large sums to carry local and regional traffic via the US and Europe. Local IXP initiatives and the proposed African Internet Exchange are beginning to address these issues but they have often taken place in a less than sympathetic regulatory climate.

So how does this assumption look? Well there have been considerable successes brought about by the private sector. The vertiginous rise in mobile use and the spread of cyber-cafes has brought both jobs and significant social changes. It’s a guess but Africa’s next political generation will probably come from its cyber-café users and they will be younger by several decades than many of the current incumbents. But there are relatively small numbers of internet users as the costs of devices that can deliver it are high relative to incomes and there are few sustainable local content models.

With success comes the question of what will happen when the growth curve in ICT flattens? Since there is little or no manufacturing on the continent, it is a service industry and its growth will rise or fall along with the production of other commodities or services. So what will African countries do that will earn them hard currency in the global economy? As India and others have shown, outsourcing is one route. It is not without its problems but only a handful of countries are currently involved in any significant way: Ghana and South Africa being the most notable examples.

The successful growth of ICT has focused attention on other shortcomings in African countries. The success story of the Nigerian mobile market is rather undercut by the failings of its power supply industry. You cannot deliver cost-effective ICT if you need to run dual-power systems. Shiny new computers (or even refurbished ones) in schools chime badly with the generally low standard of educational assets in the continent. Viewed optimistically, these contradictions can perhaps sharpen the imperative for change. Pessimistically, they are the complex "layers of the onion" that unfold as ICT initiatives seek to make an impact.

The private sector has had considerable success in enlarging access to connectivity despite the still considerable regulatory and governmental barriers in many markets. But although "private sector time" moves faster than "African government time", it’s been slow compared to what might happen if the remaining obstacles were removed.

Assumption 2: Technological innovation and the internet will make it happen

This assumption was that technology (particularly the internet) delivered it for us in the developed world, so why not elsewhere?

Several of the North American companies that signed up to digital divide initiatives were motivated by a practical mix of social responsibility and self-interest. They wanted to know of it would be possible to create say a low cost device or new forms of local content. Maybe there were undiscovered streams of outsourcing activity. They wanted to leverage their "can do" skills to make it happen and in so doing help change the world and open up new markets.

Let’s take the example of the low cost delivery device. Africa’s mobile users have acquired a device in their millions at a price that shows how large the market might be. If you could make a robust, largely voice driven device with wireless connectivity below the USD50 point, there’s just a chance that the world might be your oyster. But the hard realities of current production costs seem to make this a shimmering mirage.

In the pre-production phase, the Indian produced Simputer was talked about as a sub-USD200 device. Now it’s been delivered it retails for between USD3-400 and it’s yet to set African markets alight. A secondhand, refurbished PC will cost around USD200 in many African countries. MediaSolv talked of producing a wireless-enabled PDA for around USD25-30 but this probably translates into a retail price of USD75-100. It’s not for want of trying but it’s difficult to get the device below the threshold needed.

So how does this assumption look? There have been few revelatory, paradigm-shifting technological moments out of which new businesses have emerged. There have been almost no Africa-specific, technological developments in either the hardware or software fields. A lot of seriously bright people looked at it very hard and it simply didn’t happen.

In the field of delivery devices, there are a number of interesting projects using PDAs to gather data but these are largely donor-funded. Hybrid systems ­ getting e-mail delivered to your mobile ­ might yet be successful.

Assumption 3: E-commerce will cut out the middleman

This assumption was that if sellers in Africa (particularly small-scale sellers) could contact their markets directly, they would find larger markets and get a bigger proportion of the final sale price.

The reality check on this one is quite stark. 99% of all electronic transactions in sub-Saharan Africa take place in South Africa. This is both an argument for and against why it won’t happen in the rest of the continent.

Outside of South Africa, the main large-scale opportunities are in tourism and small-scale things like gift sites. (Budget airfares delivered by selected cyber-cafes might also be an attractive possibility). The volume of goods sold from gift sites is currently tiny. They have been forced to cobble together payment systems using overseas bank accounts, something that is much harder to do after 9/11. They have discovered that simply having a website doesn’t deliver customers. You need money and know-how to market.

The commercial barriers are almost insurmountable. If you have a Visa merchant account in a developed country, they charge you 2-2.5% of each transaction. Paypal, a system which you must join to use, charges 3.2%. Banks in discussions outside South Africa have talked of charging 7%. This may be unfair but it reflects the perception of financial risk on the continent.

In one of the continent’s larger markets, a big company was turned down by its bank when it asked for a merchant account. It had a long history of taking (and repaying) major loan facilities. What chance an enterprising small company if the big players get turned down? A massive shake-up is required in many parts of Africa’s financial services sector. Too many markets are either de-facto monopolies or duopolies: there is not enough competition to drive innovation and cost-cutting.

As Barry Coetzee of iVeri insists tourism must be the "killer app" and there is no reason why there should not be a massive increase in e-transactions in Africa’s tourist economies. Why should this matter? Because it will open up e-transaction opportunities in other parts of the economy in its wake.

So how does this assumption look? It was fundamentally wrong. No-one looked at or even understood the basis of electronic transactions. Even now (at both national and international levels) you will be hard pushed to meet anyone who can "do the detail" of the arguments and is actually working on bringing about changes.

It’s a shame because there is tremendous potential. Hybrid electronic payment systems ­ like making payments and confirming them via mobiles ­ makes good sense for vulnerable delivery drivers who are forced to carry large amounts of cash.

Assumption 4: Information/knowledge is power

The assumption here is that better information ­ whether in the market or for the governance of citizens ­ will better, more efficient or effective choices to be made.

This assumption has been powered by evangelising anecdotes that are repeated endlessly amongst the faithful. One which must be apocryphal involves a farmer using a phone to ring the local market to discover the price his crop sells at. In this way he discovers how much the "middleman" is taking from him. Enraged, this leads him to gather his fellow farmers together to start a transport co-operative to take his crops to market. Over the years, the crop and the country seem to have changed and the teller can never quite put his or her finger on the source of the story. Maybe the farmer exists, maybe he doesn’t but it’s an uplifting tale to the teller.

We tell each other ICT success stories almost as if they were parables for a better life. Often these isolated examples do not measure up to the hype we give them on a day-today basis.

However real life examples do exist. Fishermen on Uganda’s Lake Victoria can get an SMS message on their MTN mobile that allows them to know which shoreside market is achieving the best price. Senegal’s Manobi is offering a similar service for both fish and fruit prices. These are functioning value-added services delivered by mobiles.

But equivalent progress from Government has been painfully slow. India’s Uttar Pradhesh made its land sale registration process electronic to cut out corruption and aid transparency. This happened four years ago. Kenya announced it would start work on the process in May of last year. However the continent does not always appreciate the difference between announcement and implementation. Although Senegal seems to have made progress in implementing an e-version of its customs clearance procedures.

Perhaps as effective but less visible than either the private sector or government has been the proliferation of e-letters, web-sites and mail-lists that have offered for the first time a working continental grapevine, often connecting a country’s citizens with the diaspora living elsewhere.

So how well has this assumption worked? Well the most effective has been the informal and often unplanned growth of e-mail networks. In our own field, the speedy mobilisation of African ISPs in protest when Jambonet cut off Kenya’s ISPs was both fast and impressive. E-mail used like this can actually speed things up and get things done in ways that challenge "African government time". The more formal projects may work over time but the jury’s still out. As one of those told us:"We’ve built castles in the air. Now we’ve got to build the foundations."

The key is always does the information have use-value" and will it make a difference to the user. We listened to a presentation of a project to collect health statistics electronically in a country which barely has the connectivity to make this possible. Even if by some miracle it had been possible to do it by tomorrow, how would the information produce different decisions? No-one seemed terribly clear. One might guess that the local doctor makes intolerable choices about allocating resources between many overwhelming health demands. But how would a statistical description of his or her dilemma help change the shape of the problem? Perhaps the fallacy is that better information will always drive better decision-making. The project seems to have difficulty passing the "will it make a difference?" test.

Assumption 5: Open source will free Africa from unnecessary cost burdens

The assumption here is a compellingly simple one: Africans have no money to pay for Microsoft’s software. Linux is free. It’s a marriage made in heaven. A sub-variant of this argument is that important information should be free to Africa on the same basis.

We tread softly here and with great caution as the worldwide linux movement has some of the same distinguishing features as a religious jihad. Any obstacle is merely something to be overcome through true belief and any suggestion that there might be two sides to this argument often produces more heat than light. To be fair, most of the proponents on the continent itself are as sweet-tempered as you might wish for. But there are nearly always more people arguing for it than actually using it.

The reality on the continent is not wholly promising for open source. Linux is widely used for server operations by ISPs and as elsewhere, by those with server-based applications. There are interesting thin client applications like DireqLearn’s Open Lab, But recent improvements in desktop applications has yet to translate into anything resembling a sizeable user base.

Push and pull comes into play here. Without a substantial user base, the continent will not grow the critical mass of engineers it needs. Without a substantial number of linux engineers, how will the applications get written (and tested and updated) that will create the user base?

One person closely involved guessed that there were probably 300 linux engineers in the whole of sub-Saharan Africa. One country-based linux mail list has participants who can be counted on the fingers of two hands. Even the committed find it hard to survive: one dedicated linux programmer has recently put Windows back on his desktop.

For any organisation running a linux-based application that is not already in existence, it would require them to have a technologically confident member of staff and a peer group user base that it could turn to when problems arose. There would be no manuals or hotlines, however unhelpful.

That said, the argument remains compelling but there are precious few detailed and costed examples. The recent study is helpful on the arguments that might inform a decision but we must await later work for costed examples.

Open source could work well in Africa but it would require a massive commitment from the linux community worldwide. It would require the newly created FFOSSA to act more as an effective, self-help skills transfer group between the developed world and Africa. The African Network Operators Group (AFNOG) perhaps provides a model that might be followed.

As one close person closely involved in digital divide initiatives told us:"Casual observers have heard so much nonsense about the magical pixie dust of ICT and the coming age of digital bliss it augers in the developing world - it’s necessary for those in the field to counteract the hype with some hard-nosed scepticism. It’s almost like a penance - we need to rebuild credibility for the movement by showing we appreciate the folly of all the misadventures that have gone on around us."

Perhaps it’s time for this kind of wake-up call if interest in closing the digital divide is to show better results in the next three years...