Lessons From London After 10 Years of the Congestion Charge

A Republican member of Congress told me last week that he recently was in London for the first time in a long while. “Traveling was so much better,” he said. “You can actually get around. That traffic-charging system they’ve got seems to be doing a lot of good.”

London’s system — known formally as congestion charging — started up 10 years ago this Sunday, on February 17, 2003. In the decade since then it has been meticulously monitored, analyzed and debated — perhaps more than any traffic-managing scheme since Moses parted the Red Sea. It has also spawned a raft of charging programs elsewhere, most notably in Stockholm, and, starting last month, in Gothenburg, Sweden’s second city. Of course, an all-out effort to enact a comparable system here, the proposal to toll motor vehicles entering Manhattan south of 60th Street, died in the state legislature five years ago.

Ten years on is a good time to take stock. Let’s have a look.

What It Is: Cars and trucks pay £10 (roughly $15.60) to drive into or within the charging zone between 7 am and 6 pm on weekdays. The zone is London’s commercial and financial hub and, at 8 square miles, rivals Manhattan’s 8.5-square mile Central Business District. Taxis are exempt, as are qualifying low-emission vehicles. Cars registered to zone residents, who account for 2 percent of Greater London’s 7 million people, pay one-tenth the standard charge.

How Drivers Pay: London’s system deploys 1,360 closed-circuit cameras at 348 sites within the charging zone and on its boundaries to record the license plates of vehicles entering and moving within the zone. The plates are continuously matched against a database of monthly accounts, and “spot” payments are made via Internet or at kiosks, drawing down accounts or billing license-plate holders. This cumbersome system arose not only from the absence in the U.K. of electronic toll collection systems such as E-ZPass when the system was launched a decade ago, but also from the decision to charge for car trips entirely within the zone in addition to vehicle entries. A byproduct is the relatively meager net revenue available for transport improvements.

Traffic Outcomes: In its first few years, the London charging scheme was heralded as a solid traffic-buster, with 15-20 percent boosts in auto and bus speeds and 30 percent reductions in congestion delays. Most of those gains appear to have disappeared in recent years, however. Transport for London (TfL), which combines the functions of our NYCDOT and MTA and which created and operates the charging system, attributes the fallback in speeds to other changes in the streetscape and traffic management:

TfL established that the primary reason for the continued reductions to traffic speed, which would otherwise have been unexpected given falling traffic levels, was a substantial increase in interventions that reduced the effective capacity of the road network for general traffic. These interventions ranged widely, including policies to increase road safety, improve the urban realm, and prioritise public transport, pedestrian and cycle traffic, as well as a large-scale increase in road works by utilities and general development activity. (Travel in London, Report 5 [pdf], p. 77)

Travel Outcomes: Without those “interventions,” which include extensive Olympics-related works, travel speeds likely would be a good deal higher than before the congestion charge. In any event, London is enjoying improved travel choice, access, dependability and safety, as my Congressmember friend observed. According to TfL‘s Travel in London, Report 5:

  • The London Underground runs 5 percent more train-miles on the Tube, and traveler delays are down around one-third, versus a decade ago (Table 4.3).
  • Bus usage reached a 50-year high in 2011, with 30 percent more service and 20 percent less waiting compared to 2000-01 (Table 4.4).
  • Bike trips increased 79 percent from 2001 to 2011, after having stagnated between 1993 and 2001 (Table 2.2).
  • Travel fatalities and serious injuries were the lowest on record in 2011 (Figure 5.1), although cycling casualties have risen in recent years, perhaps owing to increased cycling (p. 85).

The bus service gains stem from bus fleet expansion and provision of bus lanes, the former financed by congestion charge revenues, the latter enabled by the initial reductions in auto traffic attributable to the charge itself. Here’s how an FHWA team of experts summed up London’s congestion revenues and expenditures for one recent year (2008):

Revenues from the congestion charge were £268 million (US$435 million) in 2008. When accounting for expenses (about 50 percent), the congestion charge generated about £137 million (US$222 million) in the same year, which by law must be spent on transportation in greater London. Of the 2008 net revenues, 82 percent went for bus improvements, 9 percent for roads and bridges, and the remaining 9 percent for road safety, pedestrian and cycling facilities, borough plans, and environmental improvements.

Lessons for NYC: We’ll compare London’s charging scheme with proposals for NYC, including Sam Schwartz’s developing toll plan, in a future post. For now, these differences are worth noting:

  • The cost of toll administration, which soaks up roughly 50 percent of congestion charge revenue in London, would consume only around 10 percent here. The difference arises from both E-ZPass and the vast difference in the quantity of tolling points: 348 in London vs. a few dozen to cross into the Manhattan CBD.
  • Gains in CBD auto speeds account for just a fraction (less than a quarter) of traveler time savings projected from tolling in New York. The majority of time savings will be on the road and highway network “approaching” or leaving the CBD, and on the city and region’s transit network through improvements financed by the far greater net revenues. Thus, any creep-back of CBD congestion, as London has experienced as it re-allocates road space, should not cancel improved travel times.
  • Travel-speed gains from eliminating “toll-shopping,” as the Schwartz plan would do by equalizing tolls on the East River bridges and tunnels (as well as across 60th Street), should also survive any bounce-back in traffic levels.

As for London, ten years on, no one is proposing scrapping or diluting the congestion charge. Motorists pay to drive in that city’s congested center, the sky hasn’t fallen, and life on the ground is demonstrably safer, healthier, and better.

  • Andy

    A variable charge would be much better, and could apply in many more places. Instead of choosing between $15 or nothing, the system should attempt to calculate density (based on how many others paid in the last 10 minutes) and charge based on a calculation. Few people driving at 2am means a very low charge. When everyone wants to drive in at 9am though, it can charge a lot. People will quickly learn about how much to expect and make the decision if they want to continue driving at those times or use transit next time.

    This can even be used for multilane highways – If you want to move more swiftly, pay the going rate for the fast lane. The price would depend on how many others are in the lane to assure that it continues moving at the proper speed. Once the capacity is too high, entering the lane would cost a higher rate, and a choice will need to be made. Either an in-car box could rely these prices, or signs on the road could do it.

  • Ian Turner

    @ggAndy:disqus : Under such a system, you would have no way to know how much the charge would be in advance, which would defeat the behavior-modifying intentions of the program.

    I’m not saying the charges shouldn’t be variable, but they should be predictable.

  • Kara Kockelman

    Thanks for the summary, Charles K.!  I have long been disappointed in London’s heavy admin costs, and the constant, cordon nature of the toll, but excited about ETC applications (especially by link/corridor). 
    Ian & Andy: Variable toll rates are typically pre-determined & thus “certain” (rather than truly dynamic), by tracking typical (past) traffic loads. Semi-dynamic rates can reflect real-time conditions, while being moderated to ensure no big swings in rate & maxing out as needed (e.g., when an unexpected incident takes away a lane of capacity, resulting in very high [and publicly unacceptable] penalties, if true marginal cost pricing is applied).

  • Charles_Siegel

    Great summary.  Thanks for writing this. 

  • I moved in August from London to New after nine years when I lived in London and wrote about transport, including congestion charging systems. I wrote in a blogpost after leaving London that the congestion charge was one of the two big improvements to cycling conditions between 1997, when I first lived in London, and last year, when I left: 
    http://invisiblevisibleman.blogspot.com/2012/09/farewell-to-london-where-cyclings.html The other big improvement (which I recently wrote that parts of New York badly needed to emulate: 
    http://invisiblevisibleman.blogspot.com/2013/02/staten-island-and-why-invisible-visible.html) was greater traffic calming and junction redesigns.

    It sounds like an excuse to say that the scheme is still working, even though congestion’s got worse again. There are certainly people who will claim it doesn’t work. But the figures show a continuing fall in the numbers of vehicles entering the charging zone, so the argument about road space changes holds up, I think. It’s all the more convincing if one’s spent much time in London in recent years. There are a series of huge infrastructure programs – including wholesale renovation of the gas mains and water systems – under way at present. They are causing dreadful disruption to traffic. The truth is that, while congestion has returned, it would be far, far worse given the work that’s under way without the congestion-charging scheme.

    One big lesson for New York, incidentally, lies in what has produced the reductions in traffic.Many of them come from a fall in the number of vehicles choosing to drive across Central London when there’s another route available. I think about the potential for such a change in New York every morning as I cycle through TriBeCa battling with vehicles that are in the area only because they want to drive free between Long Island and New Jersey or Staten Island, instead of paying the Verrazano Bridge tolls.

    It’s true, meanwhile, that the London system isn’t sophisticated enough and costs too much to run. One of the most impressive congestion charging systems I’ve seen is the one in Singapore, where every vehicle has an electronic token and has to pay a pre-set fee for driving past certain toll gates around the city center, depending on the time of day. It seems to work very well – and to tackle the problem mentioned below that variable systems don’t send clear price signals. The payment rates in Singapore are fixed for three months at a time – but every three months the speed of traffic on each road is measured and the toll adjusted to keep traffic flowing at the optimum speed.

  • Bronxite

    What are the chances Bloomberg will give this another go before he’s out the door?

  • JK

    I’d like to see a high resolution map like this projected traffic reductions from the Sam Schwartz fair tolling plan. The traffic reductions in neighborhood like Long Island City and Downtown Bklyn, Qns Blvd and BQE would look pretty dramatic

  • Arbolioto

    The jury is still out there about the benefits of creating such a huge surveillance & bureaucratic system of control in London. It helped it to become the city with the biggest array of CCTVs in the world. All we need now is an array of megaphones to truly become a sophisticated version of Orwell’s 1984

  • Anonymous

    First, I think it is outrageous to have a system like that eating up almost half of its revenues just to keep the collection system going. It reminds me of certain paid parking on  smaller cities where the wages or traffic wardens amount to the better part of revenue raised with paid parking.

    Now, in relate to those UK graphs, I find particularly wrong to make inferences on bike or bus services based on how much more or less are out there. Bus traffic is heavily dependent on planning decision, bike traffic is much a factor of bicycle-specific infrastructure. 

  • Markmorris69

    I think it is great that a Republican member of Congress thinks:

    “That traffic-charging system they’ve got seems to be doing a lot of good.”

    Well, perhaps the US embassy should stop insulting its host city and start paying the Congestion Charge – it is shameful the US leads the list of countries that dodge paying this charge that everyone else has to pay if they wish to drive into central London at certain tmes of the day Monday to Friday.   The US record causes great resentment amongst Londoners.


  • Anonymous

    Uhuh, so the goal is to make it easier for visiting Republican congressmen to get around in your city. Roll out the red carpet for the elite, peasants off the road!

  • Am wondering if air quality in this area was monitored over time–I expect it would have improved?  Also wondering if walking mode share went up.

    The amount of money the system consumes concerns me as well. To do a congestion system for downtown San Francisco would require monitoring close to a hundred intersections, and this is with the advantage of the Bay making a natural border on the east and north sides! If a city has some sort of natural choke points entering and exiting (i.e. bridges) it seems easier to do a congestion charge. 

    Other cities have successfully dealt with congestion in other ways. Vienna took their inner ring (their historic core), divided it into five pie shaped wedges, and made it impossible to drive from one wedge to another. (You have to leave your wedge, go the outer ring, and go into the wedge you want from there.) They did this through a series of one-way streets and forced right-hand turns. Of course you can walk and bike between wedges.  It doesn’t prevent anyone who lives or works in a wedge from driving there, but it pretty much eliminated cut through traffic and seriously discouraged any tourist from even considering driving in the city due to how confusing the streets are. They have also made large stretches of the historic center car-free entirely. As a result, the streets in the inner ring are amazingly calm and pleasant and provide a very fine walking experience.  Of course Vienna has an excellent transit system (but so does London!) and their biking numbers are higher than any US city and increasing.

    People drive because it is cheaper/and or more convenient than the alternatives. Traffic volume is caused either by people going to a destination (which requires parking) or they are cutting through on their way to somewhere else. I wonder if congestion can be dealt with by managing traffic flow to make cut-throughs unattractive or impossible (like Vienna), reducing the availability and/or increasing the cost of parking, and/or making the densest areas car-free entirely. Private parking in congested areas could be levied a congestion inducement fee. Employers and businesses that offer parking to employees or customers could be taxed if they charge under a certain amount per month. Street parking could be gradually reduced like Paris has done. With car-free zones and/or street designs, cut through traffic in some areas could be reduced or eliminated entirely.  I’ve noticed in San Francisco when there is a huge event drawing tens of thousands of people from outside the city, when the media makes a big deal about a parking apocalypse, everyone freaks out, takes transit, and congestion and parking aren’t so bad. When there is no parking apocalypse moaning and everyone brings their car, things are horrible.

  • In reply to the point about bike-specific infrastructure, the big increase in cycling in London at the time of the congestion charge’s introduction was not accompanied by a significant improvement in cycling infrastructure. I speak as someone who cycled in London daily from 1997 to 1999 and from July 2003 until August last year.

  • Anonymous

    @KarenLynnAllen:disqus : Air quality on European cities improved dramatically last 15 years as a result of Euro-standards for engine emissions with much tighter emission standards for new vehicles (with the progressive phase-out of older cars). The reduction has been massive even in cities where car traffic and fleet size actually increased a lot, such as Lisbon, Rome, Madrid, Belfast etc.

  • @andrelot:disqus  That’s very true about emissions standards, so I suppose it’s hard to separate out what air quality improvements might be due to lower congestion. I remember when I visited Budapest in 1989 the cars and motorcycles spewed out exhaust so awful that when I washed my face at the end of the day, the residue on my skin literally turned the rinse water black. (Okay, not entirely black, more like dark grey, but appalling nonetheless.)

  • Decongestion charge. It’s not a congestion charge, it’s a decongestion charge!

    Via Jarrett Walker.



Pricing Advocates Call for Impact Study and New Parking Policies

Congestion pricing advocate Carolyn Konheim and consulting partner Brian Ketcham are advising the Bloomberg administration to drop its resistance to a congestion pricing Environmental Impact Study. The two say a study is needed to head off "likely 11th hour litigation" aimed at stopping the three-year pilot program from taking effect, a possibility Streetsblog alluded to […]

Congestion Charging Rumor Mill

Three congestion charging rumors, all from excellent, though, un-named sources: 1. Bob Kiley, the Transport for London Commissioner who oversaw the design and implementation of that city’s successful congestion charging system is joining Parson’s Brinckerhoff in New York City. Parsons is the global engineering firm that has been leading the Partnership for New York City’s […]

Three Concrete Proposals for New York City Traffic Relief

This Morning’s Forum: Road Pricing Worked in London. Can It Work in New York? Three specific proposals to reduce New York City’s ever-increasing traffic congestion emerged from a highly anticipated Manhattan Institute forum this morning. One seeks variable prices on cars driving in to central Manhattan, with express toll lanes and higher parking fees to keep things […]