The latest Wolfson Prize seems deeply flawed to me because it assumes the premise of transport policy is to build roads and the problems of transport policy is how we pay for roads.
How can we pay for better, safer, more reliable roads in a way that is fair to road users and good for the economy and the environment?”
The assumption appearing to be a Chris Graying like one that road users are car users.
No the economic purpose of roads is to provide a ‘Transport Service’to get from A to B.
However the amount a society has to pay for this transport service is not an economic good, it is a deadweight loss which measures the loss to the economy of having to transport goods or people from A to B because the co-location of A to B is geographically inefficient – we are forced to travel. Of course the fact that all land uses cannot be stacked like angels on the head of the most accessible point means we have to travel, and the arrangement of land uses and transport networks is the art and science of minimising this loss.
Given that we can minimise the loss by providing a transport service from A to B to minimise the total time society takes between A and B. That includes all users of the total transport system between A and B including the opportunity costs of those deterred from travelling.
We know from the work of Dupuit that the market price of say a toll will be inefficient because it is not the marginal cost of travel – but includes an extractive ricardian rent caused by the monopoly of road ownership.
Charging a road price based on congestion and reinvesting it in public transport helps, but congestion is often a poor measure of the transport service forgone. A better measure is the gap between an efficient transport network, which moves people the most efficiently for least cost, and the current inefficient one. Congestion as Mogridge argued in Jam Yesterday Jam Today is inevitable in successful cities and a raw cost of congestion’ assumes that the ideal situation is one where anyone can travel as fast as possible from A to B at no cost – an impossibility.
We know from the Downs-Thompson Paradox that the only way to increase road speeds is to increase the equilibrium speed of public transport, and from the Lewis-Mogridge position that building roads simply leads to them clogging up and increasing welfare loss. We also know from Braess’s theorem that adding roadspace to congested networks increases journey time and taking them away reduces it. I recently tested this in a complex traffic model for a new city, and contrary to the prior warnings of the traffic engineers the only way to valence the network was to take away car links and add public transport only links.
Given all of this the best way to charge for road space is based on the inverse of the non public transport service that the road offers and to spend proceeds on improving those services on that link. That is if road has plentiful cycleways, pedestrian facilities and bus routes on it it should be free of a charge, however if it is a pure car and truck based road then it will be able to move far fewer people and should be subject to a much higher charge. For example the Dubai metro manages to move as many people as the parallel 16 lane highway. The paradox here is that drivers would be attracted to congested slow moving roads – so what congestion per-se is not the problem but a lack of none car capacity. It is much better that traffic moves slowly to enable a mixing of modes – Copenhagen style – the mixing that improves transport services, it is also better to raise most funds on arterial roads designed only for cars to fund facilities like bus rapid transit – South American Style -that can most cost effectively improve transport services.
Spending on complete streets approaches would gain support because drivers would be charged less on such roads, and spending on simply expanding roads without modal choice would be less popular as drivers would spend more. This leaves a final question, how to set the charge. Theoretically the ideal level of charge is the funding cost calculated to present value of the cost of upgrading the public transport/cycling/walking on a link so that equilibrium speeds are maintained across a network before that link is added