Ken Livingstone's plan to charge motorists £5 a day for driving into central London is a bold and necessary initiative that will be closely watched by other congested cities in the UK.
It bears all the marks of a political compromise, but that is a small price to pay to get the ball rolling.
A £5 charge is too expensive for poorer people to afford - excepting the really poor who do not have cars - while too low to deter the business user - the worst offender making one journey into town to a company parking lot and one journey home at peak times.
Ideally, the charge should rise exponentially with the size of engine and the extent to which it emits pollutants.
The revenue expected to be raised is £200m a year - not nearly adequate to build enough new buses and dedicated bus lanes to provide an efficient public alternative to the car.
But that is not the point. We have to start somewhere. Congestion charges have been bandied about for 40 years without being implemented in a big city. We now have the opportunity - and, hopefully, the technology - to start. If the experiment works, then it can be expanded in London and to other cities: if it does not work, then people will want to know why.
And the most likely reason will be the government's utterly myopic decision to go for a private finance solution for the tube that will give glossy new stations and improved signalling, but no new trains for 10 years.
If car usage in central London is to be cut, then we need far more than the 200 new buses promised by the mayor, plus draconian moves to enforce dedicated bus lanes.
If he was really serious, he would insist - as they do in other historic cities - that deliveries are made only at night to avoid the snarl-ups caused by a single van blocking a lane in the middle of town.
Mr Livingstone did not suggest these radical proposals because he is a politician. By aiming for the lowest common denominator, he kept the CBI and other bodies on board for the launch. Make haste slowly is the inevitable slogan for anything to do with transport in London.
There is the obvious danger that companies will pay the extra £5 a day, even though it is levied on individuals. An extra £5 may well be a comparatively small percentage of the real cost of what they are already paying for parking for employees.
But they should resist the temptation. If the charge is to have any chance of becoming a deterrent, it must be paid by motorists.
As Transport 2000 - which supports the initiative - points out, one of the interesting things about the Livingstone levy is that, for the first time, motorists could be faced with the choice of alternative costs between car and public transport.
Studies show that when drivers choose to travel by car, they do not take into account the depreciation on it nor the cost of the extra petrol they use. Under Mr Livingstone's scheme, they could face the alternatives of paying £5 to go by car or a flat fare of 70p each way to travel by bus.
But there is no getting away from the fact that there will still be an irresistible temptation to continue using the car if the buses or tube trains do not provide a fast and efficient alternative.
All routes lead back to the same conclusion. London, and other cities, must have massively more money spent on their public transport systems before road pricing can have any chance of success.
On present plans, that simply will not happen. Mr Livingstone must be betting that if this gloomy scenario turns out to be true, then commuters will blame parsimonious Brown rather than have-a-go Livingstone. On present evidence he will, sadly, be proved correct.
