Only deregulation and competition can help close the digital divide

Official efforts to get the developing world online are moving at Old Economy pace

Everyone must by now be aware of the "digital divide" - the gap been people and countries that are embracing the new communications technologies and those failing to do so. But policymakers have been relatively slow to acknowledge the seriousness of the situation and, more important, to do much about it.

Everyone must by now be aware of the "digital divide" - the gap been people and countries that are embracing the new communications technologies and those failing to do so. But policymakers have been relatively slow to acknowledge the seriousness of the situation and, more important, to do much about it.

True, there was a European summit in Lisbon last spring that fussed about the extent to which the European Union seemed to be slipping behind the United States. But the measures it proposed - things like getting all schools connected to the internet - were relatively minor.

Now the Organisation for Economic Co-operation and Development is tackling the subject. It has just held a four-day conference in Dubai on e-commerce, and as background it produced a report called "Understanding the Digital Divide".

Three big themes emerge from the report. The first and unsurprising one is that there are enormous differences between the developed and developing world in basic aspects of telecommunications and internet access. On the crude figures of telecoms access paths (that is, a telephone line or a mobile phone connection) the gap is nearly 10 to one. There are 72 access paths per 100 people in the developed world (defined as the OECD member states) and 7.8 in the developing and middle-income countries (see bar chart). There appears to have been some narrowing of this gap during the 1990s, but that is largely because of the explosion of telephone growth in China. Phone penetration in Africa has hardly risen at all.

On the internet the divide is even greater, nearly 100 to one, and is hardly narrowing at all. In October there were just over 94 million internet hosts in the world, of which 95.6 per cent were in the developed world - 82 hosts per 1,000 people. In the developing countries the figure was less than one host per 1,000 people. Africa has only 0.25 per cent of all internet hosts, mostly in South Africa, and its share is actually falling. The bright spot is Latin America, where internet penetration is booming in the richer countries such as Argentina, Brazil and Chile.

The second theme is that there are big cultural and language differences, particularly with the internet. For commercial applications, the Net is virtually a monolingual zone. English accounts for nearly 90 per cent of the links to secure servers, and German, French and Spanish each account for less than 1 per cent (see pie charts). I am sure that if you take into account other uses of the Net, such as chat, e-mail, information-gathering and so on, the non-English quotient would be a bit higher, maybe 20 per cent. But it is still small.

The language gap is associated with different levels of use in different countries. The highest levels of penetration, measured either by internet hosts or by proportion of people connected, is highest in the US and Scandinavia, followed by "old Commonwealth" countries, the Netherlands, and then the UK. It is unusually low, relative to the wealth of those countries, in France, Spain, Italy and Japan.

What seems to be happening is that Scandinavia and the Netherlands, where English is widely spoken, simply switch to English when using the Net, whereas other European nations tend to stick to their own languages.

The third element of the divide is socio-economic. Levels of income, education and physical location (town rather than country) also affect things, although in the more "wired" countries, levels of personal computer and internet access are so high that it is hard to argue that anyone is excluded who does not want to be. Nearly half of Dutch primary school children have access to a PC - and I suspect that statistic is rising fast too.

But then, countries such as the Netherlands are not the problem - nor indeed are countries such as France or Italy. Though they may be lagging behind the English-speaking world, that is really their choice. They are rich enough to adopt English, or to develop local language competence on the Net.

The Latin American countries are seeing explosive growth of the use of Spanish, which looks like becoming the second language of the Net quite soon. Ultimately the new communications technologies are language-neutral, as Japan has shown with its i-mode mobile phone technology. The largest commercial use of mobile telephony is in Japanese.

There are, however, two serious aspects to the digital divide, which will widen if not tackled. One is, with difficulty, fixable; the other is not.

The fixable problem is cost of access. The most important single reason why many developing countries are losing ground is the lack of competition in their telecoms sectors. Among OECD countries there are only five where there is still a monopoly of fixed-line services, and in another year there will be only two. In middle-income countries where liberalisation has been pushed hardest (such as Chile) there has been a collapse in access charges and explosive growth in services. But monopolies for fixed lines abound in the developing world.

The benefit of cheap telecommunications to the poorest parts of the world is potentially enormous. Simply having a communications system that works has been of great help to businesses in Africa, where e-mail has transformed the way companies operate. Mobile telephony carries similar potential for a revolution; in several African countries there are now more mobile connections than fixed.

But high costs matter little to businesses using e-mail: the time needed to send a burst of e-mail signals is tiny. Getting Africa fully online is tougher, but at least we know that the weapon of competition is an effective one.

The hardest task of all, surely, is to overcome the cultural barriers to greater internet use. It is very hard to see how you do fix this. Even if the cost of internet access were zero and PCs no more costly than a TV set - they seem to be heading that way - unless people can use the kit and want to do so, the divide will widen.

Following the OECD meeting, the Digital Opportunities Task Force is going to make a set of recommendations to the next economic summit. But wait a minute - that task force was set up last summer at the Okinawa summit. So a year has gone by between setting this thing up and a set of recommendations emerging. That is not internet time. Bureaucrats and governments, however well-meaning, operate at the pace of the Old Economy world.

Anyway, surely the internet is and will be a bottom-up culture? Maybe all governments can do is remove restrictions and let a thousand flowers bloom. If, for cultural reasons, they wither, then there ain't much the people at the top can do about it.