If we had no railways, we'd never invent them

No one would even think of using rail for rural transport Anyone who wants rail preserved should be shouting for privatisation

Do we really need railways? One of the underlying assumptions of the Great British Rail Debate is that the railway network is a public good, a service the community needs. The country as a whole, therefore, has an interest in maintaining the netwo rk through taxes. The question that is rarely asked is whether we still need this particular technology at all. Just occasionally, when there is a rail strike and the economy continues without much apparent damage, a few brave souls are allowed, tentativ ely, to pose the question. But most of the time the assumption that we need railways - and, therefore, railway subsidies - goes unchallenged.

But suppose, just suppose, that we had invented cars, buses and planes before we invented railways; that the railway was a new technology that had just been developed by some aerospace company, and we were wondering whether it would become an effective competitor to existing transport systems. What would we be saying? Would railways seem as significant an invention as the car and the plane, or would they be an interesting disappointment like the hovercraft ... or a successful specialist transport system, like the "eggs" at ski resorts?

I think the most likely reaction would run like this. There are two areas where the technology is potentially attractive. The first is urban transport. Our cities would have grown somewhat differently had there been no railways, but whatever form they had taken, the availability of a transport system that can move large numbers of people in acceptable discomfort would seem an ideal solution to the problems of congestion in the high-density parts of cities.

There would be a range of experiments to determine the appropriate level of investment: should, for example, development be mainly low-cost, lightweight surface systems like trams, or would it be economically sensible to install higher-capacity underground railways? The majority of the systems would probably be lightweight, for the risk-reward equation appears more attractive than it would for the very expensive systems to which we are now accustomed. But in very busy corridors (airports to city centres, perhaps) there would be a case for high-capacity systems.

The other likely area for railways would be between large population centres, where they have a potential time advantage over air travel: distances up to about 400 miles. A few experimental lines would be built, but the airlines would continue to drive down their costs, undercut the railways, and the experiment would at best be in trouble, at worst abandoned.

Other applications? Hard to see any. There might be some attempts to use railways to carry goods, but the cost and inflexibility of the systems would make them unattractive. And no one would even think of using railways for rural transport. The densitiesof population would never justify it.

Besides, the new technology would run into severe criticism. If they needed subsidies, the urban networks would be attacked on ethical grounds, in that money would be taken from people in the country to support (probably richer) people in towns. The railways would be attacked by the environmental lobbies, who would argue that they would simply encourage more people to travel who did not need to. They would be attacked by groups living near planned lines, who would argue that a plane 30,000 feet overhead caused far less nuisance than a high-speed train running at ground level.

But the most telling objections would come from urban planners. They would find fault with, first, rail's inflexibility: its very high fixed costs relative to other forms of transport. If forecasts of use proved wrong, the financial catastrophe would be far greater than, say, for air transport. In a world of uncertainties, the train would be at a serious disadvantage.

Second, while the train's urban applications might appear a seductive way of coping with peak-hour travel, they come at a time when people may not need to travel so much. As telecommunications improve, it may be possible to substitute electronic communication for physical movement. The high-capacity urban transport system is the answer for cities of the past, not of the future.

If these arguments are right, the likely success of a newly invented railway would seem an invention more akin to the ski-lift "eggs" than the jet plane: useful for specialist applications, particularly in city centres, but without sufficient economic advantages over cars, buses and planes to justify more general use. And the idea of subsidising this new form of transport would seem absurd.

But railways are not newly invented. There are enormous sunk costs and enormous sunk subsidies. The networks exist, the environmental damage largely done. Even if one would not build them now, they can still have a useful role.

Moreover, they have distorted development. We have built cities that need these very expensive urban transport networks, used at full capacity for only four hours of the 24, to keep performing as economic entities. We will need these systems for at leastanother generation, probably longer, though the loading on them will tend to decline. And the sunk costs will enable inter-city services to continue for another generation, maybe more.

But in both cases, particularly the second, it will be a case of managing a declining share of the transport market rather than meeting an expanded demand. How can this best be done?

To rely on subsidies is a desperately dangerous strategy.Sooner or later the ethical objection to transport subsidy - why should one take money from people who do not want to travel and give it to those who do? - will dominate political attitudes.

There are two practical responses for the rail industry. One is to grind down costs, in the way that the most efficient US regional airlines are finding ways of running short-haul routes more cheaply. The other is to make rail travel nicer, in the way Richard Branson is reinvigorating long-haul routes.

Ask whether either of these responses is likely to come from the public sector and one is drawn to a novel conclusion: anyone who wants to see as much as possible of our rail network preserved ought to be shouting for it to be privatised (though perhaps not in the way this government has in mind). And anyone who secretly hates railways ought to be quietly praying that it remains in the public sector till it suffers the death of a thousand cuts.