Monday 18 February 2013

10 years of London's congestion charge - a success or disappointment?

I, along with thousands of other policy, economics, consultancy and transport planning professionals watched the news from London closely today 10 years ago.  That was because, although Singapore had pioneered fully electronic congestion pricing five years beforehand, this was London.

Singapore's stunning success was seen, perhaps a little unfairly by some, as reflecting more a culture of obedience towards officialdom, and a democracy that has been unafraid of being authoritarian when it was thought of as being in the national interest.   So London was seen as the acid test.  If that old-world major city could introduce a congestion charge, it would mean it would be possible for others.  

What happened, of course, is now well known.  It worked.  It cost a lot of money, it reduced the numbers of charged vehicles entering the charged zone by around 20%, and raised a little money.  It's become a political non-issue, with there being little dispute that retention of the charge makes sense.

Since then it was extended to the west for four years, and that extension was abolished, because of local opposition (including genuine concern about effects on businesses in more suburban areas).  The price has doubled in nominal terms, but the system has evolved from being entirely declaration based (all users having to pay, in advance, whether or not they drove in the charging area during charging times) to offering a detection based account (whereby for an annual charge, you get an account billed automatically for driving within the charging zone).   The system costs have dropped significantly, partly due to technology (a shift from using live video to digitised images, and improved accuracy of ANPR) and partly due to a change in contractor (as the initial contractor Capita understandably charged a lot to ensure that the system would be working in a tight time frame).  

However, after ten years some key questions should be asked.  

Has the congestion charge reduced congestion, sustainably?
Has the congestion charge made a lot of money for Transport for London?
Has the congestion charge positively or negatively affected business in London?
Did the success of London make it easier for other cities to introduce congestion charging?
What next?

Has the congestion charge done the job that was intended?

Critics of the congestion charge say that while it had an initial impact, congestion today is about the same as it was in 2003.    I wrote about the success of the charge in August 2012, and made the wider point that London has developed around a dense network of rail and underground public transport corridors, so has not had the challenges of many other cities.

By 2007 there was a 18% reduction in charged vehicle trips (12% reduction in total vehicle trips given buses and taxis).  However TfL admits that this congestion reduction has been significantly reversed by "road space reallocation to improve conditions for pedestrians, cyclists, public transport and the urban realm".  In other words, lanes were surrendered for buses and bikes, footpaths widened at intersections, reducing road capacity.   The charge was used to reduce demand, and enable the network to be altered to promote modal shift.  Whether this has further supported reducing congestion through modeshift or exacerbated conditions for those with little choice (e.g. freight) is debatable.   Bear in mind that only 30% of vehicles in central London are even liable for the full charge, as you can remove buses, taxis, disabled people, residents (with a 90% discount) and "environmentally friendly" vehicles. 

Bear in mind also that the modeshare for car trips into central London in the AM peak was only 9.8% in 2002 and was 6.6% in 2007.  In other words, the affected number of vehicles is remarkably small.   The most important mode for commuting into central London is rail (not underground) with 44.1% mode share in 2007 (a rise of 42.2% in 2002).

The latest report on "Travel in London" states that car driver trips in London dropped by 13% between 2001 and 2011, whilst the population increased by 13%, and total trips increased by 13%.  By contrast, the number of cycling trips increased 66.6%, bus by 59.7% and rail (not underground) by 41.9% (and car passenger trips by 3.9% suggesting an increase in vehicle occupancy).  That parallels a significant increase in cycle lanes and promotion for cycling, a major increase in bus subsidies after 2000 (recently reduced by a third).   In other words, total trips increased, while car driver trips decreased.  London has achieved a significant modal shift in the past decade.   Indeed, when looking across all of London it looks the same, as the mode share for cars have dropped (as main mode in a trip) from 46% to 38% from 2002-2011.  Public transport grew from 29% to 36%.   Bus trips increased 54% between 2001 and 2011, attributable in part due to massive increased in bus services, but also new concessions for free travel by those under 16 and those over 60 (so that as much as one third of bus trips are "free").  

Yet how much of this can be attributed to the congestion charge?  

A clue may be seen in looking at the figures more closely.  By 2011, the overall vehicle kms in central London (wider than the charging zone) was 21.1% lower than in 2000, whereas in outer London the figure had dropped by 8.4%, yet TfL admits that the reduction in outer London largely occurred from 2007, suggesting the impacts of both higher fuel prices and the recession are significant.  The congestion charge impacts on the central zone are clear, but had much less impact on the traffic levels in greater London overall.  It is also notable by comparing UK wide kms driven to London, that UK wide road traffic declined between 2007 and 2011 by around 3%, whereas in greater London it was nearly 7%. (13% in central zone).   It seems to suggest that there was already a trend of reducing traffic, and a trend of modeshift, that the charge supported and increased, but which in itself did not contribute much outside the area that was serviced by the charge.

So the congestion charge has supported reductions in overall road traffic and increases in use of other modes, yet the evidence on congestion per se is far less clear.  The figure I choose to use to compare this is average traffic speeds, as this will reflect both the charge effects (and other impacts) on demand, and the changes in road capacity.  Note in the figure below, offpeak means the interpeak weekday period, during which the congestion charge operates.

London traffic speeds since congestion charge introduction
The central zone saw speeds increase immediately after the charge was introduced, but speeds are now lower than they were before the charge was introduced.  Given overall traffic figures are down, it suggests that road capacity has been reduced significantly, although much of that has been temporary, as much work has been undertaken on London's Victorian water and drainage network, and more recently on the vast tunnelling project Crossrail - which will see a quantum increase in east-west public transport capacity.  Another factor is the price of parking, which is hard to average, but has meant a reduction in on street parking (coinciding with reallocation of roadspace for pedestrians and cycles) and local authorities increasing parking charges, both of which have simply made it more inconvenient to drive.

Speeds outside the central zone have been more stable, in part because there have been far smaller losses of highway capacity due to reallocation of road space and less roadworks.  It is notable that speed increased after 2007 in the inner and outer areas, but central London speeds decreased, indicating that the recession and higher fuel prices were not enough to offset reductions in network capacity in central London during this time.  Bear in mind the real price of the charge has increased 60% since 2003.  This suggests that the elasticity of demand for the remaining users is low, probably because those who can afford to park in central London during weekdays and who do not tolerate public transport despite its intensity of service, are relatively wealthy (except freight users who simply absorb it as a cost of doing business).

The abolition of the Western extension has had little effect on congestion in that area.  Traffic speeds have dropped by about 3%, indicating that the charge was effective in reducing traffic, but not by much.

Air quality impact conclusions are mixed.  One report in 2008 said the impacts were negligible.  25% of the benefits would have occurred anyway as the vehicle fleet got cleaner, according to one study, which noted a 16% reduction in overall emissions.    The Londonist notes the Mayor's recent talk of an ultra low emission zone in central London from 2020, effectively making the congestion charging zone banned to cars that are not electric or plug-in hybrids.

On emissions the overall impact is very small, and dwarfed by the larger trend towards less overall road traffic, but even more dwarfed by changes in the vehicle fleet.  On its own it hasn't had much effect on air quality, but more of an effect on CO2 emissions.

So my conclusion is two-fold.

Yes, the congestion charge reduced congestion in central London, but this congestion reduction has been overriden by the removal of road capacity, both permanently and temporarily, meaning that congestion is no better than it was before in central London.   Of course had the road capacity been removed without the charge, the situation would now be worse.

The charge, combined with the improvements to other modes, has helped reduce overall car trips in greater London.  However, the impact outside central London has been small.  Far more significant has been the recession and increases in fuel prices (which was, until 2010, supported by annual above inflation increases in fuel taxes).

So for motorists, paying the congestion charge (particularly the freight sector), they would be right to claim that they don't get much for what they pay for, as service standards are no better today than they were before, and because 90% of the net revenues earned have not been spent on road improvements.

Yet, more broadly it is fair to say that the charge contributed towards a broader shift in mode in London.   However, I'd argue that this would not have happened without large improvements in public transport, and promotion of cycling, and a London transport policy to support this.  It's worth noting that bus fares in London are amongst the lowest in the UK (with single flat fares regardless of journey length).  

More interesting will be to see what the result of changes in policy since 2008 will mean, particularly as Mayor Boris Johnson has been changing priorities towards improving traffic management (through more intelligent traffic signals), modest investments in upgrading intersections and focusing more on cycling and less on buses.   The other dimension is that it has barely been two years since the Western extension of the congestion charge was closed.  It is too soon to fully assess the impacts of this.

Has the congestion charge made a lot of money for London?

The media has reported £1.2 billion of revenue since it was introduced (with gross revenue of £2.6 billion). This suggests costs of 54% of revenue. The 2012 annual report noted £226.7 million in gross revenue from the congestion charge, with £81.2 million in "toll facilities and traffic management" costs and another £8.7 million in "administration, support services and depreciation", leaving net revenues of £136.8 million.  A cost factor of around 40%.   This suggests that had London wanted a new source of revenue, this wasn't an efficient way of delivering it.  Certainly costs were significantly reduced from when it was introduced, but it suggests there is a long way to go.  Bear in mind also that the nominal charge has increased from an average £5 to £9.50 (given half of current users pay £9 and the others £10).  Yet, the "costs" line in the accounts suggests that more than the mere costs of running the scheme are included, given "traffic management" is included.  

Total gross revenues of the congestion charge are small, representing only slightly more than 5% of total revenues for Transport for London.  So talk about it being about money would seem to be misleading, but there would appear to be some scope to get costs down further.

Has the congestion charge been positive or negative for business?

The last full impact assessment of the congestion charge was published in 2008, and it indicated that there was a neutral impact.  Businesses that are transport dependent saw a small cost increase, offset by small improvements in travel times, but with such a small proportion of car commuters being by car it is not surprising.  There was a bigger impact with the Western extension, because some small retail premises in west London had a catchment of customers from outside the charging area, which could choose alternative shops.  However, with that zone gone, the net effect on business is now largely seen as neutral.   It has had little effect on businesses, in terms of impacts on commuters.  It has had little effect on retail, as most retail trips are undertaken using public transport.  It has had a deadweight impact on freight, as it costs lorries and delivery vans to drive into central London, but the £10 cost is low.  Perhaps the most disappointing factor is the lack of improvement of congestion overall, which would mean a more positive result.  However, given the overall increase of trips in London, albeit not by car, it is fair to say that business, overall has barely noticed the presence of the charge.

Did the success of London make it easier for other cities to introduce a congestion charge?

Yes, but not by much.   Multiple UK cities looked at a charge, but only Edinburgh and Manchester went as far as developing complete projects, and foolishly asked the public what they thought.  In both cases, the public by enormous margins said "No".   Quite simply, voters didn't think  London's example was applicable.  In part because neither city has public transport networks or service frequencies remotely akin to London, in part because car commuting is far more important in those cities than in London, and finally because they didn't trust politicians to spend the net revenues wisely.   

Outside London, Stockholm has been a notable success.  London no doubt helped pave the way.  Yet, no other major European city has followed.  Talk of congestion charges for Amsterdam, Copenhagen, Dublin, Budapest and Helsinki, have all come to nothing so far.  Italian and German cities have introduced inner city access controls for environmental reasons.  The two major cities that have introduced types of congestion charge schemes similar to London are Dubai and Tehran.  Across the Atlantic, New York was inspired, but the State legislature thwarted the desires of the City.  In China, it is being considered

Of course had London been a disaster technically or in terms of results, it would have put back congestion pricing by some years.  So no doubt London's success has been helpful, but hardly decisive. A step forward, but London's rather unique geography and intensive rail based public transport network has meant few cities have seen it as being their parallel.  The obvious other city with some parallels - Paris - has barely even entertained the idea.

What next?

I wrote before that the congestion charge has gone from being a decisive issue in the 2000 and 2004 Mayoral elections to being an issue of the Western extension alone in 2008, and in 2012 no issue at all

Yet there is little immediate likelihood of any major changes.  Recent proposals have focused on reducing operating costs and reducing the scope of the environmental discount.

Parallel to the congestion charge has been the London wide Low Emission Zone for commercial vehicles, which has been largely uncontroversial.

The Western extension reduced traffic in the zone that it covered, but was politically unacceptable and the way it was introduced meant that it increased traffic in the central zone (as residents of the Western zone got the 90% residents' discount for trips into the central zone) and particularly affected some small businesses (because it applied from 7am-6pm rather than just the peaks when congestion was at its worst in the zone).  In short, it was too blunt.

So congestion charging in London will have to be far more clever.

For me, there are two main options.  One is to divide London into zones, like Italian cities, and charge for crossing between the zones, but only at peak times.  Yet this will create an inevitable problem, as London does not have natural boundaries that easily divide the suburbs, meaning a lot of uncongested roads would be charged, and whilst it would charge more by distance, it would mean some very short trips would cost a lot.  The decisions as to where to draw such lines would inevitably mean winners and losers.  The further you go out of inner London, the more likely it is that this will seem unfair, because whilst London's public transport is great for radial trips, and a few orbital trips (i.e if your origin and destination happens to match stations on the Overground), the car is dominant for outer London to outer London trips.  

The other option is the rosetta stone of congestion charge - distance, time and place based charging.  Given that London is covered in narrow streets that are frequently used by taxis and drivers in the know "rat-running" to avoid jams or simply to find direct routes in the absence of decent major highways, it would make sense to charge by distance.  It obviously makes sense to target congested roads at congested times.  However, there are several big issues:

1.  The need for most vehicles to be equipped to charge by distance, place and time.
2.  The need to have a secondary system for out of town vehicles to pay, which means having an expensive flat access charge.
3.   The sheer costs and logistics of implementation.
4.   What to do with the money?  

The Green Party has proposed a similar idea, but apparently wants to price cars off the road punitively                 and spend all of the net revenue on more public transport subsidies.  An alternative approach would be to price roads at levels to optimise throughput (which makes it cheap at off peak times) and use the money to offset other taxes, or on road improvements.

However, that is all besides the point.  The politics in London for expanding the congestion charge are, in my view, negative.  Such a large expansion would be perceived by many as being punitive and intrusive, and without it offering a sufficiently good deal in say reducing council tax or serious improvements in roads, it simply couldn't be sold.   Baroness Jo Valentine, chief executive of London First (a business lobby group aiming to improve London's competitiveness) in City AM has suggested the sort of transformation I envisage.

If the congestion charge was expanded, Singapore like (but using GPS based technologies) it could seriously address congestion, but the revenue raised could be spent in ways that improve traffic flow by making more traffic signals dynamic and responsive to traffic conditions, by upgrading intersections and (this is serious money) being bold about new road tunnels to bypass bottlenecks.  Talk of new road capacity, in the context of road pricing may seem counterintuitive, but it is what much of the Stockholm congestion tax revenue is being spent on, and the congestion charge can ensure new capacity is sustainable and can support the movement of freight and buses where rail based options simply aren't viable.

There may be discussions about going further in about ten years, but not in the near future.  London's congestion charge will evolve a little in the near future, but is unlikely to expand.   


London's congestion charge was a bold step forward and then Mayor Ken Livingstone deserves credit for the initiative.  It delivered a reduction in traffic, although he chose to reallocate the freed up capacity away from those paying the charge to other modes.  So the effect has been that congestion is now really no better than it was before it was introduced, despite a 60% increase in price in real terms.   

It works, it costs a lot of money to run, and raises a little money, but it has not resolved London's congestion problems, and is only widely accepted because it affects so few road users.  Most London car trips don't go near the charging zone during charging periods.  Most trips to central London during charging periods are not by car.  

It is an achievement for it to be so widely accepted, but the Western Extension was a mistake, as it was bluntly introduced in a way that did not target congestion.  If it is to have a future beyond the current geography of operation, it will have to be radically different, it will need to target congestion by time and place, and at least some of the money collected recycled into reductions in other taxes or noticeable improvements to roads.

Charles Komanoff on Streetsblog has a great article about the charge.  He makes a key point about costs, which he attributes to technology and the number of charging points.  The latter is also a function of the former, as it is an area charge, so there are charging points at many points within the charging zone as well as on the boundary.  The former had been considered and ruled out in London, although I think there could be merits in using DSRC technology for exempt/discounted vehicles.   However, while I share his positive outlook about the charge, I think that the key policy purpose - congestion relief- has been negatively effected by reallocating road space - so that, unfortunately, those who pay it don't really notice that congestion is improved, because, quite simply, it isn't.

A success?  Yes, but also a disappointment.  It will take much more than this small charging zone to make a major change to traffic congestion in London.  The good news is that the increase in public transport and cycling network capacities are such, that any further steps to modestly improve road capacity are likely to have more success in reducing congestion, rather than inducing mode shift back.

A more radical expansion on a grand scale could transform London, but the politics of doing so make this unlikely for some years.  Perhaps the only way it could be done would be if it involved a countervailing reduction in other taxes (council tax, vehicle excise duty or fuel duty), and that money raised was directed more substantially towards improving roads (and also importantly, privacy concerns are appropriately addressed).

It's not the technology stopping this, but the political will and the technocratic expertise to structure a radical reform of the governance and pricing of roads in London.  Maybe peak charging on the Blackwall Tunnel, helping to pay for the new Silvertown Tunnel, will be the toe in the water for such a change.


Finally some simple mistakes people make about the congestion charge:

1.  London was not the first city in the UK to introduce a congestion charge.  That was Durham (although that system is a tiny historic central city area).

2.  London's congestion charge is not just a cordon charge, but an area charge.  A vehicle already within the charging area will be liable if it drive within it, not just crossing the boundary.

3.  Most London residents have never used the congestion charge.  Only 60% of London households have private cars, and only 7% of commuter trips into central London are by car drivers and passengers.

Oh and if TfL (and Westminster Council) wonder where congestion charge money can be spent to improve conditions for pedestrians and to stop motorists frustratingly near missing them, how about Great Cumberland Place having a signalled pedestrian crossing?   This busy intersection sees pedestrians unable to ever cross this street in one go without having to give way to traffic on green traffic lights coming either out of the street itself or turning into it.  It is metres away from the Marble Arch underground station.  It is utterly atrocious that pedestrians in this area should have to dodge traffic to make this simple crossing across a side street.

Great Cumberland Place, where pedestrians (green) have no signal or formal crossing against traffic signals that alternate between giving traffic turning into or from the street priority.

No comments:

Post a Comment