American Economic Association

The London Congestion Charge

Author(s): Jonathan Leape
Source: The Journal of Economic Perspectives, Vol. 20, No. 4 (Fall, 2006), pp. 157-176
Published by: American Economic Association
Stable URL: http://www.jstor.org/stable/30033688
Accessed: 29-08-2017 13:08 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

American Economic Association is collaborating with JSTOR to digitize, preserve and extend
access to The Journal of Economic Perspectives

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
 Journal of Economic Perspectives-Volume 20, Number 4-Fall 2006-Pages 157-176

The London Congestion Charge

Jonathan Leape

By the 1990s, the average speed of trips across London was below that at the
By beginning of the twentieth century-before the car was introduced (New-
bery, 1990, p. 35). Traffic speeds in central London had fallen more than
20 percent since the 1960s, from an average 12.7 mph for the morning peak period
in 1968 (and a high of 14.2 mph in 1975) to 10 mph in 1998. Even in the larger area
of inner London, drivers in 1998 spent almost 30 percent of their time stationary
during peak periods and more than half their time traveling at speeds of less than
10 mph (Department of the Environment, Transport and the Regions, 1998). By
2002, the all-day average travel speed in central London was just 8.6 mph (14.3
km/hour), compared to an uncongested (night-time or "free flow") average speed
of around 20 mph (32 km/hour). Congestion, measured in terms of minutes of
delay per mile compared to uncongested conditions, averaged 3.7 minutes/mile
(2.3 min/km) (Transport for London, 2003a, p. 11). With more than one million
people entering central London between 7:00 and 10:00 a.m. on an average
workday, and more than one-quarter of those by road, the cost of congestion was
clearly considerable.
Public concern over levels of traffic congestion was high. An independent
survey in 1999 identified public transport and congestion as the two most "impor-
tant problems requiring action"-selected by 46 and 33 percent of London resi-
dents, respectively, compared to 20 percent for crime or law and order. Ninety
percent of London residents said "there is too much traffic in London" (ROCOL,
2000, chap. 2, p. 5).
Proposals for congestion charging in London have been made since the early
1960s. New car registrations in the United Kingdom doubled from 500,000 in 1958

* Jonathan Leape is Senior Lecturer in Economics, London School of Economics, London,

United Kingdom. His e-mail address is ([email protected]).

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
158 Journal of Economic Perspectives

to over one million in 1963, and concerns over congestion led the Ministry of
Transport to commission what became known as the Smeed Report in 1964
(Ministry of Transport, 1964). The report concluded that existing vehicle taxes had
little impact on congestion and recommended the introduction of direct road user
charges that would take into account the very different congestion costs of different
journeys. At about this time, Vickrey (1959, 1963) and Walters (1961) formalized
Pigou's (1912, 1920) ideas on the application of marginal social cost pricing to the
case of congested roads.
However, urban congestion pricing schemes were generally thought to be
unworkable (for example, Ministry of Transport, 1967, Better Towns with Less Traf-
fic). In theoretical terms, the difficulty was that while identifying optimal congestion
charges for a single road or bridge is relatively straightforward, calculating optimal
charges for a network of roads is hard. Congestion at intersections is a critical
feature of network congestion, and optimal charges therefore vary by road and by
intersection as well as by the time of day (Newbery, 2005; Santos and Newbery,
2001; Newbery, 1990). Moreover, simulation exercises show that the benefits of
congestion charging schemes depend critically on the location of the cordon that
defines the charging area, but the theory of optimal cordon design has proved
largely intractable (Santos and Newbery, 2001; Verhoef, 2002; Shepherd and
Sumalee, 2004; Sumalee, 2004). In practical terms, the cost of collecting tolls in an
urban area was expected to be high. Moreover, the complicated nature of any
system of variable charges, by time of day and even perhaps by road, would make
enforcement difficult and undermine the desired effects on driver behavior. The

perceived difficulties of determining the appropriate cordon and level of charge,

along with the feared costs of enforcement and compliance meant that, for many
years, the idea of congestion charging had little practical appeal.
But in early 2003, London imposed a daily charge for driving or parking a
vehicle on public roads within central London between the hours of 7:00 a.m. and
6:30 p.m. on workdays. Traffic congestion has declined substantially, and the
program is largely popular. This article describes the origins of the London
congestion charge, how it overcame practical and theoretical difficulties, and what
effects it has had. The introduction of the London congestion charge is, in
important respects, a triumph of economics. It represents a high-profile public and
political recognition of congestion as a distorting externality and of road pricing as
an appropriate policy response.

The Birth of the Congestion Charge

Concerns about London's traffic congestion in the 1990s led the national
Department of Transport to create the London Congestion Charging Research
Program. Its 1995 report (MVA Consultancy, 1995) concluded that a congestion
charge would reduce congestion, offer rapid payback of the initial setup costs, and
generate net revenues as well as broader net economic benefits. The report

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
 Jonathan Leape 159

proposed a o4.00 toll (one British pound currently equals approximately $1.85 in
U.S. dollars) on vehicles entering a central London area that is effectively identical
to the charging zone ultimately chosen.
This proposal led to a working group for a Review of Charging Options for
London (ROCOL), which issued a 2000 report, Road Charging Options for London.
The 1995 report had suggested a charge for each trip entering the zone (equivalent
to a "fare"). In contrast, the ROCOL report focused on an "area license" that would
allow the payee to travel into, around, and out of the charging zone as frequently
as desired. The report concluded that, to have a noticeable impact on congestion
in London and to produce significant revenues, the initial focus of any scheme
should be central London, rather than the larger "inner London" area. A paper
license scheme was rejected on the grounds that effective enforcement would be
prohibitively expensive. The report recommended two alternatives: an area licens-
ing scheme for central London based on video camera enforcement and a work-
place parking levy.
The final impetus to the London congestion charge came with the election of
Ken Livingstone as the first Mayor of London in May 2000, following a decision by
the new Labour government to grant limited autonomy to new municipal govern-
ments. Livingstone is a high-profile London political figure who, as leader of the
Greater London Council until its abolition in 1986, was the focal point of London
protests against the policies of Margaret Thatcher. After narrowly losing the Labour
nomination, Livingstone had run for Mayor as an independent, with overwhelming
popular support. Prominent in his platform was a commitment to introduce
congestion charging.
After an 18-month period of extensive public consultation-which some ob-
servers have argued was a critical factor in making the scheme publicly acceptable-
Livingstone decided to propose an area licensing scheme applied to central Lon-
don. Livingstone's decision to choose area licenses, rather than parking levies, was
supported by evidence that it would be more effective in reducing congestion and
was consistent with his campaign promise to introduce congestion charging.

How the Congestion Charge Works

The London experience demonstrates how introducing real-world congestion

charges requires not only careful modeling and analysis of traffic patterns and
commuting behavior, but also a thorough assessment of compliance and enforce-
ment issues, which present a range of practical, political, and technological chal-
lenges.

Amount of the Charge

On February 17, 2003, London imposed a o5.00 daily charge (increased to
o8.00 in July 2005) for driving or parking a vehicle on public roads within the
congestion charging zone between the hours of 7:00 a.m. and 6:30 p.m., Monday

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
160 Journal of Economic Perspectives

to Friday, excluding public holidays. Economic analysis played a decisive role in

determining the appropriate level of congestion charge. For the reasons noted
above, it was neither feasible nor appropriate to use a measure of the marginal cost
of congestion as the basis for the charge.' Mayor Livingstone relied, instead, on the
extensive modeling of household behavior and the resulting traffic patterns un-
dertaken by the London Congestion Charging Research Program and, especially,
the ROCOL report (2000). The predicted impact on congestion was then used, in
the ROCOL report, to estimate the net economic benefits of alternative levels of
charge. Both the predicted impact of the scheme and the costs and benefits are
discussed further in a separate section below.
In adopting the initial o5.00 congestion charge, the Mayor was endorsing the
recommendations of the ROCOL report. The report evaluated the impact of
setting the charge at o2.50, o5.00, and o10.00 under various scenarios. Assessing the
scenario in which the congestion charge was assumed to be accompanied by
"associated transport measures" (principally, a o100 million package to improve bus
transport), the estimates presented in the report indicated that while the impact on
traffic levels and net revenues increased proportionally with the level of the charge,
estimated net benefits did not. Although there was considerable uncertainty re-
garding the level of net benefits, the point estimates suggested that a o5.00 charge
might yield net benefits almost 40 percent higher than a charge of o2.50, with little
additional benefit from a o10.00 charge (ROCOL, 2000, pp. 89-90). These con-
siderations in favor of a o5.00 charge were undoubtedly reinforced by political
concerns regarding the public reaction should the charge be set as high as o10.00.
In making the case for the recent increase in the charge from o5.00 to o8.00,
Mayor Livingstone continued to stress economic considerations. An increase to
o8.00 would, he argued, result in a significant further reduction in congestion
(from a reduction of 30 percent with the current o5.00 charge to an estimated
reduction of 34-38 percent); an increase in net revenues (from o80 million/year
to o115-125 million); and a rise in net benefits (from o50 million/year to an
estimated o70-100 million). Moreover, since, by agreement with central govern-
ment, all congestion charge revenues would remain earmarked for public transport
for a further ten years, the net revenues could be expected to increase the impact
still further. He argued against a smaller, incremental increase in the charge (say,
to o6.00) on the grounds that it might have little impact for behavioral reasons,
because "the difference is sufficiently small for people to carry on and say, 'I will pay
it.' " He also rejected a more substantial increase to o10.00 on the grounds that the
additional benefits were too small (net benefits would only rise to an estimated
o75-110 million) to justify the additional burden on charge payers and the resis-
tance to the change that might arise (London Assembly, 2005).

1 It is nevertheless interesting that a recent attempt to estimate the marginal cost of congestion in
London (ignoring the network complications discussed earlier) concluded that the o5.00 charge would
correspond to the optimal congestion charge if, on average, charge payers undertake journeys of about
two miles, or just over half the width of the charging zone (Santos and Shaffer, 2004).

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
 The London Congestion Charge 161

A much discussed feature of the London scheme is the flat-rate nature of the

charge. One of the historical objections to congestion charging in London and

elsewhere has been the infeasibility of implementing a system with the degree of
time variation in the charge that was assumed necessary for such a scheme to be
effective.

The decision to opt for a single rate charge is supported by London travel
speed statistics, published by the Department of Transport. Although off-peak
daytime travel speeds in outer London are, as one might expect, markedly higher
than those during the morning and afternoon peaks (Transport for London,
2003b, p. 63), the same is not true in central London. Over the period 1968 to 1998,
off-peak daytime travel speeds in central London averaged 11.6 mph, compared to
11.9 mph in the morning peak and 11.6 mph in the evening peak period. Average
speeds may even have converged over time; the off-peak daytime travel speed
recorded in the 1998 survey was 10.0 mph, compared to 10.0 mph and 10.2 mph in
the morning and evening peak periods, respectively (Department of the Environ-
ment, Transport and the Regions, 1998). A similar flat pattern was evident in
central London traffic levels, as measured in vehicle-miles driven within the charg-
ing zone in 2002 (Transport for London, 2003b, p. 77). If congestion is similar
throughout the day, then a flat charge throughout the day makes sense.

The Geography of the Charging Zone

The congestion charge is an area license, or "day pass," that allows unlimited
travel into and around the charging zone. The congestion charging zone, illus-
trated in Figure 1, covers the area known as "Central London," which includes the
financial center (the "City of London") as well as Parliament and the principal
government offices, the major tourist sites, and the main centers of business, law,
and entertainment (the "West End").2 Although the zone covers an area of only
eight square miles, little more than 1 percent of the total area of Greater London,
it includes the main areas of worst congestion. Central London is defined as the
area inside the Inner Ring Road, which is made up of the roads shown on the map.
Since around one-quarter of the central London traffic was, before the charge was
introduced, simply passing through, an important consideration in the choice of
the Inner Ring Road as the boundary of the zone was its usefulness as a peripheral
route to accommodate through traffic diverted by the charge. The charge is
payable only by vehicles traveling inside the zone and not those on the ring road.
The standard charge applies to private cars and commercial vehicles entering
the zone during the charging hours. As mentioned earlier, individuals who live
within the zone are entitled to a 90 percent discount on the charge, and do not
have to pay the charge if their car remains parked off-street or in a residents'
parking bay throughout the charging period. Some vehicles are exempt from the
charge, including motorcycles and bicycles; buses and taxis registered in London;

2 In September 2005, the Mayor announced that the charging zone would be extended westward in early
2007 to include the boroughs of Kensington and Chelsea.

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
 162 Journal of Economic Perspectives

Figure 1
The Congestion Charging Zone

l :::ing::::~s Grossoo:::
'::':
An*l"": o ::'

:::t P'::::::acn:'::'::s:
':::: I :~ '
niiii:iiir
::::::::::j:::-:;:::-:::::::: ::o::
:::::::::::::::::::::i::::::::::i:::::::

~i~aAb
':::
:::::::::::ii::::::::i:::::::::::::::::: ::::::::::::::::::::::::::::::::::::::::
i::j::::::::::i::::i:::::::::::o:::::::: ::::::::::::::::::::::::::::::::ji:
::::::::::::::::::::j:::::::::::::1:::::
:::::::::::::::::::::j::::j:::::::::::::

~ ::::::::::::::::::::::i
::::::::::::,- :':':::

Old Stre
o:o:o::Io:
:o:o:o:--
:o:
::::o:-::

~ao~'!
Mia:::::::: ::rylebo'ne:
r.
:i:::::::::::::::::::::

pO
:::::::::::::::::::j:::::::::::::::::
::--:-:-:::::::::o::o:o:

::::::::::

Ijverpool5reetA
Oxford City
::::::::::::::::::::::::::::::j*r ~:::::::

~Riu1iwJ,
::::-o::o:o::o:
:o:o::o::::-::::::-o::::o:o
o-
:i:::::::::::::::o-::::::::::::::::::::7 eii8sink
Circus
:::::::::::::::::::::::::::::i:::::::::: :::::'::':~:

-iottorn
iiiiiia L"-~LLI
iiAiiiiiii!iiiiliidgate
::::::::i::::j::::::::::i::i::::::i::i::

~iiIii.i~~
0?~4~~
iiiii:~
;Z%% Blackfriars -each
i Piccaditty
~ Circus streetl Tower
Bridge
Trafalgar
Loedon
Square ha5ri5g
rosr~s
Bridge
Westminster Waterlooa
~::i::j::::::::::::::::::::lil:::::::li'
';::"::::::--::'::::::'::':: '';: ::': :':::':':

" :-:-::::::o:o:-::::::-:-:::-::::-:-:::::

Glpin asse ::i::::::::::ii::::::::1::i::i::::::::::

:-::::o:-.:-:::::o::o-:-:o:o:o:o::::o:::
~:~:~:~:~:_:~:~::~: :... ..-..
-::::i::i-::ii:: :l:~:-:-lil :ii:::::::::::::::::::::::jj:j:B:-

Central London
::-:-:oo:o::o:o:o::o:o:o:o:-:o:-:o:o:o:o

::-:::'o'::-":::':':;:':::::'::;:

ooI
congestion charging zone
::::::i:::i:::::::ur~r;:li::i:::::::::::
o:;:;:o:;:o:-:;-: ::;::r rr 1~:::o::::o:o::::: ::::::::::::-:
:::::::;:::-:;::-:;:-:-::ov, rs~i-:o:o:::::::::::::::-:::o::::

:::::::j::::::j::j:::
ilijoii:joi''''''''':::: .
o:o:o::-:-:-:-:-::-:::-:-:o::-:: ::::::

Vau
::::i::::::::::::::::::::j:::.:::.:.
:j:::::::::::::::::::::::::::::::::i::::
:i::::i::::::::::i::::::::::::::::::::::

o:::o:::::;:::::o:o:::::::::;:::::::::::

Source: Transport for London, 2006

vehicles used by disabled individuals; alternative fuel v

and vehicles used by some medical staff and fire fig
exempt from the charge. It is the driver's responsibili
day of travel within the charging zone.

Logistics of Payment
The charge may be paid in advance, on a daily, week
the day of travel. To discourage people from paying a
made between 10 p.m. and midnight are subject to a s
charge to o10.00 (unchanged after July 2005). There a
charging zone. Instead, a variety of payment mechani
retail outlets, kiosks, by telephone, over the Internet, an
phones. While use of the text message and call cen
remained relatively stable since the scheme was introd
22 and 14 percent of all payments, respectively), the
changed over time. Payment through retail outlets has de
percent in the first year to less than 30 percent in 20
of payment by Internet, which has increased from aro

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
 Jonathan Leape 163

total number of charge payments has remained relatively stable at around 110,000
per day (Transport for London, 2005b, p. 142).
Individuals submit their vehicle registration number when paying the conges-
tion charge, and this number is then entered into a database for trips within the
charging zone (during the hours of operation) for a particular day or period.

Enforcement

Video cameras at every entry point and in mobile units within the zone capture
images of vehicles entering, leaving, or driving within the zone.3 Automatic number
plate recognition technology is then used to identify the vehicle registration
number. The current automatic number plate recognition system has a 70-80
percent success rate for a single pass; the average detection rate is estimated to be
85-90 percent, as almost all vehicles in the charging zone pass multiple camera
sites. A recent report on detection technologies concluded that while the current
level of accuracy would not support an account-based charging system (for which a
vehicle would have to be detected in order to be charged), it is sufficient for the
enforcement of the current pre-pay system; individuals must pay the charge without
knowing whether they have been detected, and even the reduced penalty charge (if
paid within 14 days) is more than six times the daily charge (Transport for London,
2005c, pp. 19, 52).
At the end of each day, the numbers captured are compared to the database
of paid or exempt vehicle registrations. If a match is made, the images and vehicle
details are removed from the database. If no match is made, the numbers are sent
to the Driver and Vehicle Licensing Agency (the national agency responsible for
issuing all drivers licenses and collecting the annual car tax), which supplies the
name and address of the vehicle's registered owner as well as the vehicle make and
model. The final step of the process is a manual check of the image to confirm that
the registration number captured by the plate recognition system is correct and
that the vehicle type corresponds with the information supplied by the Driver and
Vehicle Licensing Agency.
The registered owners of vehicles that have not paid the charge as required are
then sent penalty notices for o100 (o80 before July 2004). As with parking tickets in
London, the penalty is reduced to o50 if paid within two weeks and increases to
o150 if not paid after a month.
Problems of compliance and enforcement affected the scheme at the outset.
One source of confusion was that both the automatic number plate recognition
system and the individual payers of the charge when submitting their number plate
details, had trouble distinguishing the digit zero from the letter O, and the digit 1
from the letter I. In the early weeks of the scheme, Transport for London received
appeals or complaints on almost two-thirds of the penalty notices issued, most of
which were due to errors in compliance or enforcement.

The potential infringement of civil liberties has not been a major issue, perhaps because of the
increasingly widespread use of closed-circuit television in public areas of London for crime prevention.

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
164 Journal of Economic Perspectives

Transport for London took a series of steps to improve enforcement, including

increased staff training and improved information for charge payers-notably, the
inclusion of a picture of the vehicle on the penalty notice from July 2004 (Trans-
port for London, 2005b, p. 146). By the first quarter of 2004, enforcement errors
were down to 2 percent of appeals and the majority were associated with vehicles
that had recently been sold, with the penalty notice sent in error to the previous
owner. The number of penalty notices also rose steeply in 2003, from 15,000 per
week in the early weeks to over 40,000 per week towards the end of the year. Actions
to improve compliance together with the increase in the penalty charge from o80
to o100 in July 2004 were associated with a fall in the number of penalty notices to
less than 30,000 per week by February 2005.
Looking to the future, Transport for London has recently undertaken a review
of alternative detection technologies. "Tag and beacon" systems based on infrared
(or microwave) detection were found to have 99 percent accuracy, but involve
substantially higher compliance costs as each car must have a tag or beacon
installed, as well as having readers or tollbooths for the beacons at all entrances to
the zone. Systems based on the Global Positioning System (GPS), which works by
installing beacons in cars that can be identified by satellite, were found to require
a buffer zone of 60 to 250 meters in order to achieve 99 percent accuracy, while cell
phone systems (based on the Global System for Mobile Communication or GSM)
not only require that all drivers have a phone, but need a buffer of 800 to 2,400
meters-clearly inadequate to support congestion charging (Transport for Lon-
don, 2005c).

Impact of the Congestion Charge

Predictions of the Impact on Traffic Levels and Congestion

Before the congestion charge was implemented, extensive modeling was un-
dertaken to estimate its likely impact on traffic and congestion (ROCOL, 2000).
The results of surveys were used to model household demand, by type of household
and type of trip, for the area license. These results were fed into a separate model
to assess the impact on traffic patterns (p. 67). One of the schemes assessed was a
o5.00 charge for central London, which very closely resembled the scheme that was
originally adopted. This congestion charging scheme was predicted to reduce car
miles in central London by 20-25 percent during the charging period and to
reduce total vehicle miles traveled in the zone by 10-15 percent. Most of this
reduction was due to a 20 percent expected reduction in car trips, but it was partly
accounted for by diversions around the charging area (pp. 69-70).
The reductions in vehicle flows were predicted to increase average traffic
speeds in central London from 9.3 to 11.2 mph (15 to 18 km/h) during the
morning peak period (and from 9.9 to 11.2 mph on average over the day), the
improved speeds being comparable to those prevailing in the early 1980s (ROCOL,
2000, chap. 5, p. 76).

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
 The London Congestion Charge 165

Table 1

Impact of the Congestion Charge on Traffic in the Congestion Charging Zone

(in thousands of vehicle-kilometers and percent)

Percentage
2002 2003 change

Cars 771 (47%) 507 (35%) -34%

Vans 287 (18%) 273 (19%) -5%
Trucks 73 (4%) 68 (5%) -7%
Taxis 256 (16%) 312 (21%) 22%
Buses 54 (3%) 65 (5%) 21%
Motorcycles 129 (8%) 137 (9%) 6%
Bicycles 69 (4%) 89 (6%) 28%
All vehicles 1,640 (100%) 1,451 (100%) -12%

Source: Transport for London. Data provided to the author, May

Notes: The first column for each year shows the number of vehicle
type of vehicle, within the congestion charging zone. The secon
proportion of total vehicles within the congestion charging zone

Actual Reductions in Traffic and Congestion

The first impact studies showed that the reductio
met or exceeded predictions. Moreover, the initial re
in subsequent periods (Transport for London, 2005b
The number of private cars, vans, and trucks com
dropped 27 percent between 2002 (before the charg
drop in inbound car traffic represents some 65,000-7
made. Of these, Transport for London estimates on
over half have transferred to some form of public tran
divert around the charging zone; about 10 percent h
private transport, predominantly taxis and bicycle
either stopped traveling or shifted their trips to out
for London, 2005b). These data suggest that, as exp
has influenced the decisions of road users on variou
particular trip, which mode of transport to use, and
The net result has been a significant change in t
traffic, as measured by vehicle miles driven within t
shown in Table 1, private cars, which accounted for
traffic before the charge was introduced, now represen
34 percent. The level of commercial traffic (vans and tr
there has been a sharp rise in taxis (up 22 percent),
bicycles (up 28 percent).4

4 A degree of caution is appropriate in interpreting these figure

of trend decreases in inbound traffic and in central London traf
Historical data suggest that overall traffic levels in central Lon

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
166 Journal of Economic Perspectives

Evidence on average travel speeds on roads inside the charging zone indicates
that the impact of the congestion charge has been considerable. All-day average
network travel speeds increased from a precharging average of 8.9 mph (14.3
km/h) to 10.4 mph (16.7 km/h) in May/June 2003, a rise of almost 17 percent
(Transport for London, 2003a).
The congestion charge has had an even more significant impact on congestion
levels. One way to measure congestion is in terms of the minutes of delay experi-
enced compared to an uncongested (night-time) travel rate. On this measure,
congestion has dropped an average 30 percent from the start of the charge in
February 2003 to mid-2005, at the top end of expectations (Transport for London,
2005b, p. 14).
The drop in congestion levels, and increase in average speeds, reflects mainly
a decrease in queuing time at junctions. Figure 2 shows time spent traveling at
different speeds in the charging zone during charging hours. For each speed
category, the first six shaded bars show each of the two-month periods from
March/April 2002 through January/February 2003. The second six bars show the
same two-month periods one year later (March/April 2003 through January/
February 2004). While the figure shows small increases in the amount of time spent
traveling at higher speeds, between 30 and 40 km/h (19 and 25 mph), the principal
change is the drop in the time spent in traffic jams. The time spent traveling at
speeds below 10km/h (6 mph) has been reduced by one-third compared to
precharging levels (Transport for London, 2004b, p. 17).
Decreased congestion within the zone has also, as expected, contributed to
reduced congestion and travel times on routes into the central zone. One source of
data comes from what are called "moving car observer" (or "floating car") surveys,
in which cars are fitted with instruments to measure the speed and travel times and
the drivers are instructed to match average traffic speeds while driving a predefined
schedule of routes around the network. This survey data indicates that conges-
tion on the main radial approach roads dropped from a precharging level of
1.5 min/km to 1.2 min/km, a drop of 20 percent (Transport for London, 2004b,
p. 16); for comparison, the night-time "uncongested" travel rate is about
1.5 min/km. These findings were corroborated by evidence from surveys of regular
drivers between November 2002 and April 2003, which showed that journey times
during the morning peak decreased an average of 14 percent following the intro-

percent between 1986 and 2002, with sharper declines in the years since 1994. An even more pro-
nounced drop is evident in car traffic, which decreased by somewhere between a quarter and a third
since 1986, again more sharply since the mid-1990s. On the other hand, the significant drop in inbound
and central London traffic between 2001 and 2002 was due, at least in part, to an unusually high level
of road works in preparation for the congestion charge, and might have been reversed in 2003 had the
charge not been introduced (Transport for London, 2003b). For this reason, comparisons between 2002
and 2003 seem unlikely to overstate the impact of the charge. Note that the first caveat, based on
historical data, does not apply to estimates of the impact of the congestion charge on travel speeds and
congestion, both of which had steadily worsened over time, as discussed in the introduction and shown
in Transport for London (2003a, p. 12).

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
 Jonathan Leape 167

Figure 2
Impact of the Congestion Charge on Time Spent Traveling at Different Speeds in
the Charging Zone during Charging Hours

180,000
1st U Mar 1st
AprU Mar 2002
Apr 2003
160,000 2nd U May2ndJune
O May June 2003
2002
3rd U July3rdAug
O July Aug
20022003
140,000 4th U Sep 4th
Oct
O Sep 2002
Oct 2003
5th U Nov Dec 2002 5th O Nov Dec 2003
120,000 6th O Jan Feb 2004
6th U Jan Feb 2003

100,000

80,000

60,000

-e

40,000

20,000

0-
0-10 10-20 20-30 30-40 40-50 60+

Speed bands (km/h)

Source: Transport for London. Data provided to the author, May 2006.

duction of charging. For an average round trip of 80 minutes, this implies a time
savings of 10 minutes (Transport for London, 2004b, pp. 19-20).
There were concerns that the diverting impact of the congestion charge could
lead to higher levels of congestion on the inner ring road that borders the zone, but
is not included in it, as well as in areas just outside the zone. These higher traffic
levels did not materialize, at least partly as a result of improved traffic management
systems that adjust traffic lights to manage the flow of traffic on and approaching
the ring road (Transport for London, 2003b, 2004b). Traffic changes in the
bordering areas have been variable, with some experiencing small net increases in
traffic (between 2 and 6 percent) in the first year of the scheme and others
registering net reductions. In the second year of the scheme, traffic levels fell in
almost all areas.

It was expected that reduced congestion in central London as a result of the

charge would not only shorten travel times but also increasejourney time reliability.
This, too, was borne out. Results of the survey of "regular" drivers found that the
standard deviation of travel times decreased 27 percent during the morning peak
and 34 percent for return journeys (Transport for London, 2004b, pp. 19-20).
As noted earlier, before the congestion charge was levied, the level of congestion
in London was remarkably constant throughout the day. This pattern was apparent
after the charge was introduced, as well. In particular, there is no reduction in
congestion during the midday "interpeak" period, contrary to what one might expect.

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
168 Journal of Economic Perspectives

Impact on Public Transport

The congestion charge sought to reallocate road space from private cars to
public transportation. The ROCOL (2000, chap. 5, p. 70) report predicted that the
charge would result in a 3 percent increase in public transport trips into London
(mainly more rail trips) and a 4 percent increase in public transport trips within the
charging zone (mainly more bus and underground trips). However, projecting
changes in public transport use was complicated by the fact that a number of
measures aimed at improving bus services were implemented in the lead up to the
congestion charge.
The expected increase in rail trips did not materialize following the intro-
duction of the congestion charge, and the number of underground trips
fell-although the latter was widely attributed to other factors including the pro-
longed closure of the Central Line, the downturn in the local economy, the war in
Iraq, and the drop off in tourism. By contrast, the rise in the number of individuals
entering central London by bus exceeded predictions by almost 50 percent. From
autumn 2002 to autumn 2003, bus passengers entering the charging zone in the
morning peak period by bus rose by 29,000, an increase of 38 percent. Transport
for London (2004b, p. 40-41) estimates that roughly half the increase is due to the
improved bus service and half to the congestion charge.
The reason that the rise in bus ridership exceeded expectations appears to lie
in a "virtuous circle" for bus transport that can result from congestion charging
(Small, 2005, p. 12). The higher price of rush-hour car travel induces some to
switch to public transport, increasing revenues to transport providers. At the same
time, reduced congestion leads to increased travel speeds for buses which, in turn,
further encourage patronage while also reducing average costs per passenger to
transport providers. Increased passenger numbers and reduced average costs en-
able providers to offer some combination of improved service levels (more routes,
higher frequencies) and lower fares. Improved services and reduced fares stimulate
further shifts from car travel to public transport, resulting in additional reductions
in congestion and gains to public transport.
Small (2005) simulates these interactions. The initial 6 percent rise in patron-
age, as a direct effect of the rise in the cost of car travel resulting from the London
congestion charge, is, in his model, magnified to 16 percent as a result of the
virtuous circle (p. 14). Small also simulates these effects for a typical American city.
Since the share of weekday trips that are made by public transport is much lower
than the 85 percent in London (even before the charge), and fares are more
heavily subsidized, the shifts to public transport resulting from road pricing are
likely to be much larger in the context of U.S. cities, resulting in more dramatic
reductions in fares, user costs, and costs to transport providers.
These estimates may underestimate the beneficial feedback effects from con-
gestion charging, since they do not incorporate the effects of increased reliability
in bus services. In the first full year after the introduction of charging, excess
waiting times for buses (caused by service irregularity or missing buses) dropped 30
percent in the charging zone. Excess waiting times dropped further in the second

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
 The London Congestion Charge 169

year, with an additional reduction of 18 percent in and around the charging zone
in the period from March to December 2004, compared to the same period in 2003
(Transport for London, 2005b, p. 50). The resulting reduction in user costs will
have further reinforced the virtuous circle.

Impact on Local Business

A congestion charge is likely to have different effects across businesses. A
decrease in car trips could have an adverse impact on some retail businesses, while
a reduction in congestion and in travel times could benefit others.
A department store chain called John Lewis commissioned a study on how the
congestion charge affected the sales of its central London store. Using sales data
provided by the store, Bell, Quddus, Schmoecker, and Fonzone (2004) found a
statistically significant negative impact of the charge on sales. The authors revisited
the question in a broader study published in 2005, and found a continuing negative
impact of the charge at the John Lewis store, but no significant effect for total
central London retail sales (Quddus, Carmel, and Bell, 2005).
A survey of 500 firms carried out in early 2004 found that 72 percent of
businesses felt that the road charging experiment was working (with 14 percent
saying it was a failure) and 58 percent felt it improved London's image (with 15
percent saying it gave London a bad image to outsiders). Overall, a plurality of
firms felt the impact on London's economy was neutral (32 percent) with equal
numbers identifying positive and negative effects (26 percent) (Clark, 2004).

Costs and Revenues

The unexpected success of the congestion charge in reducing traffic had a

negative impact on revenues from the charge. At the same time, the set-up an
operational costs of the scheme were considerably higher than expected. As
result, net annual revenues have fallen far short of expected levels. The ROCOL
(2000) report estimated that the annual operating and enforcement costs of the
scheme would be o30-50 million (on top of one-off set-up costs of the sam
magnitude). The annual revenues from a o5.00 charge were estimated at o230-28
million and the annual revenues from penalty payments at o30-40 million. Thu
net annual revenues were expected to be in the range o230-270 million. Even th
revised, more conservative estimates from Transport for London (2002, p. 14-3
put expected net revenues in the range of o130 million to o150 million pounds.
Actual net revenues have been far lower, at an estimated o68 million in 2003-4
and o97 million for 2004-5 (Transport for London, 2003a, p. 30; 2005b, p. 138).
A large part of the explanation for the lower-than-expected net revenues is that the
congestion charge payments have averaged o115 million/year, or just half the leve
originally predicted. This shortfall in revenues is primarily due to the scheme'
greater-than-expected impact on car traffic-a drop of 30 percent in potentially
chargeable vehicles entering the zone compared to the mid-point prediction of 20

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
170 Journal of Economic Perspectives

percent used to project revenues. Revenues have also been affected by higher-than-
expected numbers of exempt or discounted vehicles, the former reflecting, in part,
the success of the scheme's incentives for using low-emission vehicles.
Another contributing factor has been the level of noncompliance, which
exceeded expectations. As a result of steps taken to improve compliance and
enforcement (discussed above), income from penalty notices reached o70 million
in 2004-2005 (Transport for London, 2005b, p. 138), twice the level expected in
the ROCOL (2000) report, somewhat narrowing the shortfall in scheme revenues.
The shortfall in net revenues from the scheme has also been driven by
higher-than-expected costs: implementation costs averaged o95 million in the first
two years-more than twice the level expected. The higher costs are, in part,
attributable to the increased enforcement expenditures resulting from measures
taken to address noncompliance and to raise income from penalty notices. Never-
theless, the high resource costs of running the scheme are an important feature of
the London congestion charge.
This revenue shortfall has political implications, since one selling point of the
scheme was that the revenues from the congestion charging scheme would be
earmarked for spending on public transport in London. Earmarking served various
purposes. Mayor Livingston had promised "radical improvements" in bus services,
and the revenues from congestion charge revenues were seen as part of achieving
that objective. It was hoped that earmarking would reinforce the effect of conges-
tion charging by encouraging further shifts to public transport through service
enhancements. The earmarking of revenues to public transport was also intended
to ensure that the scheme facilitated access to central London for all households,
rather than simply increasing mobility for drivers. Earmarking was seen as impor-
tant in offsetting the potentially regressive effects of the congestion charge which,
as a flat-rate charge, imposes a heavier burden on low-income drivers. Thus, for a
variety of reasons, earmarking played a crucial role in securing political support for
the congestion charge-as predicted in the congestion pricing literature (for
example, Small, 1992, 1993).
As expected, some 80 percent of scheme revenues have been spent on bus
network improvements, amounting to approximately o80 million in 2003-2004. A
further 11 percent was spent on road safety and the remainder on other initiatives
like supporting walking and cycling). The central government announced in 2005
that the earmarking, originally scheduled for the first 10 years, would be extended
for a further ten years.

Social Costs and Benefits

Estimating the social benefits and costs of congestion charging in London was
a key focus of the London Congestion Charge Research Program and the subse-
quent ROCOL report. This work continued following the implementation of the
scheme and, as noted above, was used by Mayor Livingstone in making the case for
the July 2005 increase in the charge from o5.00 to o8.00.

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
 Jonathan Leape 171

For the purposes of social cost-benefit analysis, the costs of the scheme fall
under five headings: the initial set-up costs associated with the installation of the
necessary infrastructure and services; the scheme operation costs, which is the
largest single cost (and consists primarily in the payments made to Capita, the
private company contracted to collect the charge); the supervisory costs incurred by
Transport for London, which are associated with management of the scheme and
administration of contracts; the traffic management costs directly associated with
the scheme (primarily associated with changes to the network of bus services); and
charge-payer compliance costs. The charge payments made by drivers are not
included, as they represent a transfer rather than a resource cost.
The measurable benefits of the scheme can be grouped in six categories. The
first and most important of these is the time savings to drivers and passengers of
vehicles that continue to use the road system after charging is introduced-
including cars, taxis, buses, and commercial vehicles within and outside the charg-
ing zone. All of the London analyses have followed the standard approach of using
separate values for time savings in business and nonworking time (including
commuting). The inconvenience of those who switch to public transport ("deterred
drivers") as a result of the charge enters negatively, valued at half the level of the
avoided charge. Improved journey time reliability is another source of benefits, in
addition to direct time savings, as are reduced accidents and lower carbon dioxide
emissions. The net proceeds from the scheme (charge and penalty payments) are
not included since they, like the charge payments, are transfers. However, it is
worth noting that, given the UK government requirement that all charge revenues
be spent on transport improvements, the charge payments are likely to be gener-
ating significant additional benefits in reduced travel times and accidents and in
other savings that contribute to popular perceptions of the scheme's impact.
Table 2 shows the most recent detailed estimates of the social benefits and
costs, which are drawn from Transport for London (2006, pp. 171-72). The table
also includes the author's estimates of the opportunity and depreciation costs
associated with investment in the precharge period (2000-2003), which, as
Prud'homme and Bocarejo (2005) and Mackie (2005) note, are omitted from the
Transport for London estimates. The total estimated annual costs of the congestion
charging scheme are o163 million (o140 million if implementation costs are
excluded), while the total annual benefits are o230 million.
The cost-benefit estimates produced by Transport for London have been the
subject of some controversy. The robustness of their initial estimates (Transport for
London, 2003a) was questioned by Prud'homme and Bocarejo (2005), who provide
alternative estimates suggesting that the economic benefits represent less than 60
percent of the scheme's resource costs. Mackie (2005) argues, however, that
Prud'homme and Bocarejo significantly underestimate the benefits of the scheme,
by omitting the benefits to road users in Greater London outside the city center
(which may account for as much as two-thirds of total benefits as a result of the
large numbers involved); by using an inappropriately low value of time; and by
underestimating the benefits to bus users. In addition, as Mackie (2005) notes, the

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
172 Journal of Economic Perspectives

Table 2

Estimates of the Social Benefits and Costs of the London Congestion

Charging Scheme
(in o millions, 2005 prices and value, for a o5 charge)

Annual costs

Transport for London administrative costs 5

Scheme operation costs 85
Setup costs (opportunity costs and depreciation) 23
Traffic management costs (primarily increased bus services) 20
Charge-payer compliance costs 30
Total 163

Annual benefits

Time savings and reliability benefits:

Car users
business 65

commuting, other 45
Vans, trucks
business 35
Taxis
business 30

commuting, other 10
Buses
business 2

commuting, other 40
Deterred drivers
business -5

commuting, other -20

Reduced accidents 15
Reduced CO2 emissions 3
Other resource savings 10
Total 230

Source: Transport for

Note: Author's estima
o170 million in the
estimated depreciatio
Book, 2003).

authors' approach
London fails to ta
heterogeneous val
road pricing sch
2004).
All of the above estimates focus on direct costs and benefits and make no
attempt to incorporate second-best considerations, which may be significant. Parry
and Bento (2002), for example, show that congestion charges on commuting traffic
can stimulate or discourage labor force participation at the margin, depending on
how the revenues are used.

While the net benefits of the congestion charge scheme do appear to be clearly

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
 The London Congestion Charge 173

positive, it is evident that the total resource costs of the scheme are a substantial fraction
of the benefits. These high costs have raised questions about the operational efficiency
of the London scheme; for example, Mackie (2005) asks whether the scheme contains
inessential elements that are not cost effective and whether the high costs reflect an
element of rent-seeking on the part of the private contractor operating the scheme.
The high costs also suggest the need for careful analysis before extrapolating the
benefits of the scheme to other cities in the United Kingdom and elsewhere.

Conclusion

Attempts to introduce road pricing in Britain have historically encountered fierce

political resistance. The famous Rebecca's Rioters of 1843 were farmers who took to th
streets dressed in women's clothing and dismantled toll booths in rural Wales. Mor
recently, increased gasoline taxes and rising oil prices triggered a fuel price rebellion
in 2000, in which truck drivers in Britain blocked refineries and major roads, precip-
itating a serious political crisis for the government. Against this background, man
politicians regarded attempts to introduce congestion charging as foolhardy.
But the London congestion charge has been both a practical success in reducing
congestion and a popular success. Traffic delays inside the zone have decreased by
around 30 percent, with a reduction of 15 percent in traffic circulating within the zone
and 18 percent in traffic entering the zone during charging hours. Journey tim
reliability has improved by an average of 30 percent. Political opposition to the scheme
has been minimal and popular support is now widespread.
A similar, but somewhat more muted, picture emerges from an analysis of the
scheme's social costs and benefits. The benefits of the congestion charge, primaril
associated with time savings and increased reliability of car and bus transport,
appear to be largely in line with expectations. The resource costs of running th
scheme-a factor which, as Mackie (2005) notes, is omitted altogether from the
conventional diagrammatic analysis of road pricing-have been twice as high as
expected, representing more than two-thirds of the scheme benefits. Net benefit
of congestion prices seem to be positive, but less than commonly anticipated.
For this and other reasons, a degree of caution is appropriate before gener-
alizing from the London experience. In the United Kingdom, the success of the
London congestion charge has given rise to a new government initiative to intro-
duce national road pricing-a major policy shift in a country where the first and s
far the only toll highway is a 27-mile stretch of the M6 outside Birmingham, opened
in 2003. Various studies have illustrated the potential efficiency gains from a shif
to environmental and congestion-based charges, especially where accompanied b
a reduction in fuel taxes (for example, Glaister and Graham, 2005). Ironically,
however, the successful introduction of the London charge means that the poten
tial additional benefits of an ambitious national scheme may be limited. Th
estimates produced by the Department for Transport (2004) show that the impac
of congestion charging in London is responsible for more than 80 percent of the

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
174 Journal of Economic Perspectives

reduction in congestion that would result from a national road pricing scheme
applied to all urban and interurban roads. In fact, Newbery (2005) argues that a
national distance-based road pricing scheme for interurban roads in the United
Kingdom would be uneconomic, as existing fuel taxes generate much of the
benefits at a small fraction of the cost; not surprisingly, the principal efficiency
gains from congestion-based charging arise in urban areas. Moreover, Edinburgh
voters rejected in February 2005 plans for a citywide congestion charge, which
suggests that popular support for congestion charging derives from the perception
of concrete benefits, not abstract principle.
London has some particular characteristics that helped to make its congestion
charge a success. Traffic congestion in London was severe, even by the standards of
other large cities in high-income countries. London has a comprehensive and
well-functioning public transport system-including rail, subway, and bus systems-
which offers good alternatives to road users. The geography and roads of London
suggested the possibility of using the "ring road" around inner London as a suitable
boundary for the congestion charge. In cities with a milder traffic congestion
problem that face obstacles to developing their public transport system or have a
geography that makes drawing a boundary difficult, congestion charges might not
function as well. Moreover, any city considering congestion charges must face the
practical problems of how to set and enforce charges and collect payments and
then assess the net benefits of any scheme given their particular circumstances.
But the London experience also illustrates that the practical problems of
congestion charges are not insurmountable. Cities around the world display high
levels of traffic congestion, which impose large costs; the annual costs of congestion
(including wasted time and fuel) in the 85 largest U.S. cities, for example, are
estimated at $63 billion or $794 per traveler (Schrank and Lomax, 2005). More-
over, numerous studies demonstrate that the values placed by individuals on travel
time and reliability are high and heterogeneous (for example, Small, Winston, and
Yan, 2005). These factors suggest that congestion charging can yield significant
economic benefits by inducing a more efficient use of scarce road space. A
simulation exercise in Small (2005) suggests that the benefits of the London
scheme should be replicable in U.S. cities. The benefits of carefully designed
congestion charges in large urban areas appear, on the basis of the successful
London scheme, to be both significant and within reach.

* The author is grateful to Transport for London, and to Charles Buckingham in particular,
for providing the data used in this paper; to Yen Mooi and Prakarsh Singh for excellent
research assistance; and to Elisabetta Bertero, Margaret Bray, James Hines, Timothy Taylor,
and Michael Waldman for their helpful comments.

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
 Jonathan Leape 175

References

Bell, Michael G. H., Mohammed A. Quddus,

Georgina Santos, ed. Research in Transporta-
Jan-Dirk Schmoecker, and Achille Fonzone.tion Economics, vol. 9. Elsevier Science, pp, 61-
2004. "The Impact of the Congestion Charge on85.

the Retail Sector." Report submitted to JohnMay, Anthony D., R. Liu, Simon P. Shepherd,
Lewis Partnership, UK. and Agachai Sumalee. 2002. "The Impact of Cor-
Blow, Laura, Andrew Leicester, and Zoc Smith.
don Design on the Performance of Road Pricing
Schemes." Transport Policy. July, 9:3, pp. 209-
2003. "London's Congestion Charge." Briefing Note
No. 31. London: The Institute for Fiscal Studies. 220.

(http://www.ifs.org.uk/bns/bn31.pdf). Ministry of Transport. 1964. Road Pricing: The

Clark, Andrew. 2004. "London Companies
Economic and Technical Possibilities. London:
HMSO.
Learn to Love Congestion Charge." The Guard-
ian. London, February 16. Ministry of Transport. 1967. Better Towns with
Department of the Environment, Transport Less Traffic. London: HMSO.
and the Regions. 1998. "Traffic Speeds in Inner
MVA Consultancy. 1995. The London Conges-
London: 1998" DETR Statistics Bulletin. 98:22. tion Charging Research Programme: Final Report.
London: HMSO.
Transport Statistics Division. (http://www.dft.gov
.uk/stellent/groups/dft_transstats/documents/ Newbery, David M. 1990. "Pricing and Con-
downloadable/dft_transstats_021875.pdf). gestion: Economic Principles Relevant to Pricing
Department for Transport. 2004. "FeasibilityRoads." Oxford Review of Economic Policy. Summer,
6:2, pp. 22-38.
Study of Road Pricing in the UK: A Report to the
Secretary of State." London. (http://www.dft.Newbery, David M. 2005. "Road User Charg-
gov.uk/stellent/groups/dft_roads/documents/ ing and Finance." Taxation Seminar, London
page/dft_roads_029788.hcsp ). [The Department School of Economics, December.
for Transport also provides a useful annotatedNewbery, David M. and Georgina Santos.
bibliography of documents on road pricing: 1999 "Road Taxes, Road User Charges and Ear-
marking." Fiscal Studies. 20:2, pp, 103-132.
(http://www.dft.gov.uk/stellent/groups/dft_roads/
documents/page/dft_roads_029710.hcsp).] Parry, Ian W. H. and Antonio Bento. 2002.
Glaister, Stephen and Daniel. J. Graham.
"Estimating the Welfare Effect of Congestion
2005. "An Evaluation of National Road User
Taxes: The Critical Importance of Other Distor-
Charging in England." Transportation Research
tions within the Transport System." Journal of
Part A. 39:7-9, pp, 632-50. Urban Economics. 51:2, pp. 339-65.
HM Treasury. 2003. The Green Book: Appraisal
Pigou, Arthur C. 1912. Wealth and Welfare. Lon-
and Evaluation in Central Government. London. don: Macmillan.
(http://greenbook.treasury.gov.uk/). Pigou, Arthur C. 1920. The Economics of Welfare.
Livingstone, Ken. 2001 "The Mayor's Trans- London: Macmillan.

port Strategy, Draft for Public Consultation." See Prud'homme, Remy andJuan Pablo Bocarejo.
esp. Annex 4: "The Congestion Charging2005. "The London Congestion Charge: A Ten-
Scheme for London." London: Greater London tative Economic Appraisal." Transport Policy.
Authority. (http://www.london.gov.uk/mayor/ 12:3, pp, 279-87.
strategies/transport/pdf/annex4.pdf). Quddus, Mohammed A., Alon Carmel, and
London Assembly. 2005. "Proposed Conges- Michael G. H. Bell. 2005. "The Congestion
tion Charge Increase: The Mayor Answers Charge's to Impact on Retail: The London Experi-
the London Assembly." London Assembly ence." GLA Economics Working Paper No. 12.
Transport Committee, February. (http://www Greater London Authority.
.london.gov.uk/assembly/reports/transport/ROCOL [Review of Charging Options for
cong_charge_feb05.pdf). London] Working Group. 2000. Road Charging
Mackie, Peter. 2005. "The London Conges- Options for London: A Technical Assessment. London:
tion Charge: A Tentative Economic Appraisal.HMSO. A (http://www.gos.gov.uk/gol/transport/
Comment on the Paper by Prud'homme and 161558/228862/228869/?view= Standard).
Bocajero." Transport Policy. 12:3, pp, 288-90. Santos, Georgina and David Newbery. 2001.
May, Anthony D. and David S. Milne. 2004. "Urban Congestion Charging: Theory, Practice
"The Impact on Network Performance of Driv- and Environmental Consequences." CESifo
ers' Response to Alternative Road Pricing Working Paper No. 568.
Schemes," in Road pricing: Theory and evidence. Santos, Georgina and Blake Shaffer. 2004.

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms
176 Journal of Economic Perspectives

"Preliminary Results of the London Congestion Charging: Update on Scheme Impacts and Opera-
Charging Scheme." Public Works Management and tions. London, February. (http://www.
Policy. 9:2, pp. 164-81. tfl.gov.uk/tfl/downloads/pdf/congestion-
Schrank, David and Tim Lomax. 2005. "The charging/cc-12monthson.pdf).
2005 Urban Mobility Report." Texas Transpor- Transport for London. 2004b. Impacts Monitor-
tation Institute. (http://mobility.tamu.edu). ing Programme: Second Annual Report. London,
Shepherd, Simon, and Agachai Sumalee. 2004. April. (http://www.tfl.gov.uk/tfl/cclondon/cc
"A Genetic Algorithm Based Approach to Opti- monitoring-2nd-report.shtml).
mal Toll Level and Location Problems." Network Transport for London. 2005a. Central London
and Spatial Economics. 4:2, pp, 161-79. Congestion Charging Scheme Impacts Monitoring:
Small, Kenneth A. 1992. "Using the Revenues Summary Review. London, January. (http://
from Congestion Pricing." Transportation 19:4, www.tfl.gov.uk/tfl/downloads/pdf/cclondon/
pp. 359-81. impacts-monitoring-report-january-2005.pdf ).
Small, Kenneth A. 1993. "Urban Traffic Con- Transport for London. 2005b. Impacts Monitor-
gestion: A New Approach to the Gordian Knot." ing Programme: Third Annual Report. London,
The Brookings Review. 11:2, pp. 6-11. April. (http://www.tfl.gov.uk/tfl/cclondon/pdfs/
Small, Kenneth A. 2005. "Road Pricing and ThirdAnnualReportFinal.pdf).
Public Transit: Unnoticed Lessons from Lon- Transport for London. 2005c. London Conges-
don." Access. Spring, no. 26. tion Charging Technology Trials: Stage 1 Report.
Small, Kenneth A., Clifford Winston, and London,
Jia February. (http://www.tfl.gov.uk/tfl/
Yan. 2005. "Uncovering the Distribution of Mo- downloads/pdf/congestion-charging/technology-
torists' Preferences for Travel Time and Reliabil- trials.pdf ).
ity." Econometrica. July, 73:4, pp, 1367-82. Transport for London. 2006. Impacts Monitor-
Small, Kenneth A. and Jia Yan. 2001. "The ing Programme: Fourth Annual Report. London,
Value of 'Value Pricing' of Roads: Second-BestJune. (http://www.tfl.gov.uk/tfl/cclondon/
Pricing and Product Differentiation." Journal ofpdfs/FourthAnnualReportFinal.pdf).
Urban Economics. 49:2, pp, 310-36. Verhoef, Erik T. 2002. "Second-best Conges-
Sumalee, Agachai. 2004. "Optimal Road User tion Pricing in General Networks: Heuristic Al-
Charging Cordon Design: A Heuristic Optimiza- gorithms for Finding Second-Best Optimal Toll
tion Approach." Computer-Aided Civil and Infra- Levels and Toll Points." Transportation Research B.
structure Engineering. 19:5, pp. 377-92. 36:8, pp, 707-729.
Transport for London. 2002. The Greater Lon- Verhoef, Erik T. and Kenneth A. Small. 2004.
don (Central Zone) Charging Order 2001: Report to"Product Differentiation on Roads: Constrained
the Mayor, London, February. (http://www.tfl. Congestion Pricing with Heterogeneous Users."
gov.uk/tfl/cclondon/cc_report_mayor.shtml). Journal of Transport Economics and Policy. 38:1, pp,
Transport for London. 2003a. Congestion 127-56.
Charging: Six Months On. London, October. Vickrey, William. 1959. "Statement on the
(http://www.tfl.gov.uk/tfl/downloads/pdf/ Pricing of Urban Street Use," in Hearings: U.S.
congestion-charging/cc-6monthson.pdf). Congress, Joint Committee on Metropolitan Washing-
Transport for London. 2003b. Congestion ton Problems. November, 11, pp, 466-77.
Charging: Three Months On. London, June. Vickrey, William. 1963. "Pricing in Urban and
Transport for London. 2003c. Impacts Moni- Suburban Transport." American Economic Review.
toring Programme: First Annual Report. London, May, 52:2, pp. 452-65.
June. (http://www.tfl.gov.uk/tfl/cclondon/ Waiters, Alan A. 1961. "The Theory and Mea-
cc_monitoring-lst-report.shtml). surement of Private and Social Cost of Highway
Transport for London. 2004a. Congestion Congestion." Econometrica. 29:4, pp. 676-99.

This content downloaded from 146.155.32.57 on Tue, 29 Aug 2017 13:08:33 UTC
All use subject to http://about.jstor.org/terms