Revolution that will banish the city slog

We are in the age of the giant city. Hardly a day passes without some new paper being published, or conference organised on the future of big cities, for this is a subject upon which everyone feels he or she has something to say. Thus the World Bank is concerned about the development of giant agglomerations of 20 million people and more in the richer developing countries: places like Mexico City or Sao Paulo. The British architect Richard Rogers will devote his forthcoming Reith Lectures to the future of cities. There is a string of plans for spending the proceeds of the Millennium Fund on new grands projets in London and elsewhere. Even this week's World Economic Forum meeting in Davos had a special session on the future of thecity.

And if you fancy yourself as a budding Richard Rogers (or even an embryonic Ken Livingstone) and want to do a bit of DIY design on cities, the new hot ticket in the computer games market is called SimCity. Plug it into your PC and you not only create your ideal city, complete with homes, offices, factories, public services, and taxpayers to foot the bill, but you run its entire finances. If you end up with a bad case of urban blight, you have only yourself to blame.

Why this sudden outburst? The millennium is concentrating people's minds on the mess we've made of our cities in the latter half of this century. And underlying this are two deep forces which are making us more aware of the social and economic costs of urbanisation.

In the developing world there is the population explosion: adding 90 million people to the world each year.Even if two-thirds of these people remained in the country or small towns, we would still be adding 30 million city-dwellers a year. The world's largest urban agglomeration, the Tokyo plain, is a bit less than 30 millions. So the mathematical task is to build the equivalent of a new city larger than anything on earth every year simply to keep pace with rising population.

That process will continue for another generation at least, more likely two, until the growth of world population starts to level off. How these cities manage themselves will depend, in some measure, on the lessons they can learn from their giant counterparts in the developed world. And here the forces are more complex, for they concern the changing economic function of the city: in particular the influence of technology in changing both the sorts of jobs which are being done, and the need for physical proximity in carrying out those jobs.

Cities do not grow large simply by being administrative centres: Ottawa, Brasilia and even Washington are middle-ranking in size despite being capital cities. Cities grow large because they are successful producers of goods and services. Thus London and New York are sustained by a large financial and business services industry; Tokyo by this plus a sizeable manufacturing base; Los Angeles by being the capital of the world entertainments industry. Physically, all large cities exhibit much the same characteristics. There will be a scattering of public buildings, but the highest rents (and the highest buildings) will be commercial office blocks - the factories in the sky that produce these services.

Looked at rationally it is pretty odd: enormously expensive buildings, used by pretty expensive people for perhaps ten hours out of the 24, who have to spend on average two hours a day in considerable discomfort travelling to and from work on overloaded transport systems which only operate at full throttle for four hours out of the 24.

Before the railways the practical limit was not much above a million, the size of ancient Rome, 18th-century Edo (as Tokyo was called), or pre-railway London. Railways, particularly underground systems, enabled the size to be pushed up to 10 millions andmore, but at the cost of considerable diseconomies of scale. Tokyo, the most extreme example, subjects its people to extremely long commuting times - typically one-and-half hours each way - which inevitably make companies in Tokyo less efficient producers. An extra hour commuting is an hour not available for working. One of the ironies of the progress of workers' rights during the past 100 years is that most of the additional leisure time won by establishing shorter working hours has been absorbed in longer commuting times.

But people still commute in giant cities - because that is where the jobs are located. True, more and more of these jobs may be in fringe locations, with the result that a string of mini-cities have grown up on the edge of conurbations: a process brilliantly catalogued by Joel Garreau in his 1991 book Edge City - Life on the New Frontier.

There is now more office space in London's edge city locations (Slough, Croydon, etc.) than in the West End and the City. This has resulted in a change in commuter patterns: from travel in and out of the centre on mass transit systems, to driving round the edges of the conurbation. But it still takes up a lot of time. Commuting is actually a relatively recent invention, made possible only because we have the technology to do it. For much of the industrial revolution people lived within walking distance of their place of work.

But now telecommunications create the possibility of eliminating the commute. I suspect we are in the early stages of a revolution as dramatic as the invention of the railways and the motor car: telecommunications which will be immensely capable and virtually free. We talk about the implications for jobs (will white-collar clerical jobs move abroad to countries like India?). We wonder whether we will all become teleworkers and move to Galloway (or Norfolk, or Provence)? But while we think about the possible consequences for us as people, and businesses ponder whether they could manage with smaller head offices by "hot desking", we think much less about the implications for our cities.

Let's suppose that some, not all, of the white-collar jobs that drive the economies of places such as New York, London or Paris do move out, what will be the glue that holds cities together? The answer is fun. Cities can generate, in a way that no low-density semi-rural community can, cultural excitement: the shops, restaurants, theatres and nightclubs that make them the magnets they are. City centres work badly as factories of white-collar services, but wonderfully as factories of entertainment. The more that the commuting load can be reduced the more the urban transport networks can be directed to moving people about for fun, rather than for work.

But since our cities are already too big to be efficient, how can we remodel them for this purpose? The answer is, we have to take them apart. Not physically, of course, for we have to live with much of the structures that we inherit. No, take it apart conceptually: create new business centres on the fringes, but buttress these by creating new centres of fun, too. In London, the Millennium Commission would support a new opera house at the Brent Cross shopping centre, or a museum near Heathrow. Do not think of a city of 12 million: imagine instead that there are four cities of 3 million, which just happen not to have any countryside in between.