Tuesday, 9 June 2015

Call for much wider road pricing in London

London's congestion charging scheme is world famous, in the esoteric world of road pricing, because it was the first major Western city, to adopt charging of existing roads.  Its success is such that it ceased to become a political issue after its implementation, (notwithstanding the poorly targeted Western extension of the original central charging zone, which was scrapped by the current Mayor Boris Johnson because of local unpopularity and modest traffic impacts). 

Discussion about expanding the scheme further has been largely limited to the Green Party, which as a vocal minority has keenly supported an ambitious concept of charging cars and trucks by distance largely to penalise such traffic to reduce congestion, and to raise revenue for its own preferred projects to favour cycling and public transport.  Whilst this would make a significant impact on the environmental impacts of road transport (and congestion), it would appear to reflect more of an ideological opposition to motorised road transport that involves private cars or lorries, rather than an interest in optimising the use of the network or an efficient level of pricing.   

Yet the merits of road pricing are widely acknowledged not only by some environmentalists and opponents of growth in motorised road transport on the political left, but also laissez-faire free-market proponents who are more neutral about growth in road transport on the political right, who believe in more efficient allocation of road space.

So it is interesting to see Baroness Jo Valentine, Chief Executive of lobby group London First, in the Guardian advocating a vastly expanded road pricing scheme for London

London First is described as:

representing the capital’s biggest employers in financial services, property, transport, hospitality and retail, along with its universities. Its stated aim is “to make London the best city in the world in which to do business.”

The article cuts across a number of major issues for London, such as housing and governance, but transport is always a big issue.  

Baroness Valentine proposes a radical expansion:

“You need pan-London road pricing,” she says. “Probably not right out to the M25, but to the north and south circular. The population’s growing, the roads are never going to keep up with the natural growth in demand, so you’ve got to ration it in some way. I would do more sophisticated road pricing than we have at present more widely. You’ll get a version of it with the new river crossings, if those are ever built.” Again, she thinks the sums would soon add up: “If you relieve congestion in London, that produces economic benefits and the Treasury benefits too.” 

How could you expand pricing to the A406/A205?

There could be a few different ways of doing this.  The existing zone is relatively tiny in the context of greater London, as seen below

Existing London congestion charge zone

Expanding out to the North Circular (A406) and South Circular (A205) roads would be a significant expansion of the charged area, as can be seen below. 

London congestion charge if expanded to the "circular" roads.

The most economically efficient way of doing this would be with distance based charging, that had a time and location element to it, but to do this would require the use of either dedicated on board equipment or the realisation of the concept of using a mobile phone app, securely linked to the vehicle, to enable such charging.  The potential to cleverly target congestion, discouraging "rat running" on local roads and maximising utilisation of the network is considerable.

The bigger problem is dealing with occasional drivers into the zone.  Having a single flat charge, as exists with the current congestion charge, to be effective would need to be high, and so excessively blunt for those crossing the outer cordon (or taking a single trip within the area).  An alternative would be to adopt an Italian style multiple zone scheme, splitting the area within the ring roads into multiple cordons, so that motorists pay to cross multiple zones.  Of course, any multiple zone system creates distortions at the boundaries of zones, so design would have to be careful to minimise these.

However, any options to create a new congestion charge based on existing ring roads have their own difficulties, because in all cases they involve compromises.

The first point to note is that at the eastern end, the two roads don't meet up over the Thames, but are connected by a free car ferry which is wholly unsatisfactory as a major arterial crossing in a major city.  However, one option could be to adapt the route to be bounded by the A12/Blackwall Tunnel or to build the proposed Thamesmead Crossing to bridge the gap.  Another point is that parts of the North Circular and all of the South Circular roads are far from being dual-carriageway grade separated main highway standard, but are actually residential streets in many cases indistinguishable from neighbouring ones.  Quite simply, if these roads are meant to carry traffic around a charging zone they are severely inadequate in some locations, as can be seen below.  The blue lines are where the roads are 2 or 3 lanes each way, grade separated, the red lines are where the roads are either not grade separated (and have significant bottlenecks) and between 1 and 3 lanes each way.  Note also the gap to the east.  

Gaps in London North and South Circular roads

A simple approach would be to create a second charging zone outside the existing one, but that would only penalise movements from outside the zone to inside it, not around it.  Of course the lack of any real differences between outside and inside the boundary in some locations would make such a charge quite arbitrary.  Look here at the suburban commercial district Forest Hill, which would be divided by a road that is indistinguishable.

Forest Hill, red line is the south circular road
Assuming that the billions of pounds needed to build serious orbital highways to fix this aren't going to come soon, if at all (given it would involve heroic levels of tunnelling), then it is difficult to envisage a congestion charge being introduced to the South Circular road without it causing serious disruption along that route.


None of this is a reason not to consider various options, one floated is a "vignette" whereby motorists from outside London pay to cross the greater London boundary (which could mean most journeys within the M25 ring motorway), although the congestion reduction impact would not be significant beyond that point.  However, once again, it could be a start, charging both to use a vehicle within London and for entering London, although what is the value gained from such a blunt move?

Better deal for motorists?

What will be essential is to link any charge to delivering a better standard of service to motorists in London.  A lot of money is being spent on upgrading intersections in London, by and large to accommodate cyclists.  For safety reasons (and because of significant increases in cycling at peak times), some of these projects are justified, but in some cases they are increasing motorised traffic congestion.

Jo Valentine is a cyclist, but says that the current programme to reallocate road space on many routes to dedicated (and in some cases segregated) cycling lanes also imposes costs on other road traffic, in the form of congestion.  From 1996 to 2009, central London has seen approximately 25% of its road network capacity transferred to walkways, cycleways, bus lanes or public space (Source: Travel in London Report 4, 2011, Transport for London, Figure 4.12), although over than time demand for that road space has declined by about 12% (in significant part because of the congestion charge) (Source: Travel in London Report 4, 2011, Transport for London, Figure 4.13).  In effect, it means the congestion charge has meant reallocation of road space has been more tolerable than it would have been otherwise.  

I couldn't source readily the most up to date data on central London traffic figures, but extrapolating from 2007 data which indicated that 42% of motorised vehicle trips are chargeable, it appears that around 21% of vehicle trips into central London are by cars that are not exempt or 100% discounted from the congestion charge (because of disability or being ultra low emission vehicles).  So the scope for modal shift in central London appears to be low.  This is hardly surprising, as driving in central London is notably slower than using the Underground or cycling until after around 10pm and before 6am.

So if there are going to be more charges for motorised road users, there needs to be a consideration of using much of the new revenue either to offset other charges or to improve roads, either by addressing bottlenecks (the Bounds Green bottleneck on the A406 seems obvious), or by more tunnelled highways to take traffic away from pedestrians, cyclists and town centres.

The potential is there for traffic congestion to be significantly improved in London through charging, but the quid pro quo needs to be for the revenue from charging to be recycled either by reducing other taxes (e.g. the council tax contribution to road maintenance) or addressing the major shortfalls in the network.  The Greens understandably want to use congestion charging in London as a stick to relentlessly contain road traffic, but I believe most of their objectives can be achieved by taking a more consumer led approach.

It's time to talk about roads as a utility.


  1. A cordon based charge was always going to be a blunt instrument. Better a blunt instrument than none at all, but the bigger the cordon, the blunter the instrument. Rather than try to find ways of making a bigger cordon work, effort needs to go into a more sophisticated mechanism. Smartphone type technology seems a possible approach but it will be necessary to find a way to link the car to the phone. The basis for the charge should be the travel time. If we assume the fundamental diagram is consistent with a BPR function with power 4, then the time-cost imposed by a car on all other motorists is 4 x (actual time - free flow time). Not only is this theoretically easy to measure, it also is something that motorists should be able to estimate in advance (given a fixed charge per minute) and so their responses can be rational.

  2. @david You wrote your comment in 2015. Have you had any additional ideas or seen other articles that tackle the issue of 'linking the car to the phone'? I think there are several ways of doing this, including a tag on the engine block that includes temperature, pressure and accelerometers. What I am looking for is a review of different combinations of sensors and devices that might together create a robust, secure road charging system that's infrastructure light and inherently tamper proof.