Research in Transportation Economics 40 (2013) 3e18

Contents lists available at SciVerse ScienceDirect

Research in Transportation Economics

journal homepage: www.elsevier.com/locate/retrec

Cities on the move e Ten years afterq

Kenneth Gwilliam
Institute for Transport Studies, University of Leeds, UK

a r t i c l e i n f o

a b s t r a c t

Article history:
Available online 25 August 2012

The World Bank urban transport strategy review, Cities on the Move analyzed urban transport problems in developing and transitional economies and articulated a proposed strategy framework for
national and city governments. This paper describes how the urban transport problems of the developing
world have changed in the last decade and assesses the extent to which the strategies recommended in
2002 have been successfully implemented. It shows that progress has been widespread in some areas e
particularly in mass transit analysis and investment and some environmental policies e and that there
have developed some good planning and public transport practices in a smaller number of model cities.
But more strategic institutional and policy issues, including the mobilization and regulation of private
sector initiative in meeting infrastructure and public transport supply deciencies, have tended to be
poorly developed. Above all, the growth of medium sized cities with weak institutions and nance
highlights the need for the international development institutions to put greater emphasis on helping
those cities by dissemination of best practice in strategic transport planning and trafc management.
2012 Published by Elsevier Ltd.

Keywords:
Urban transport
Developing countries
Institutional reform

1. Introduction
In 2002, the World Bank published its urban transport strategy
review, Cities on the Move (World Bank, 2002).1 The paper had
three objectives, (i) to develop a better understanding of urban
transport problems in developing and transitional economies; (ii)
to articulate an urban transport strategy framework for national
and city governments; and, (iii) to identify the role of the World
Bank in supporting governments in the development and implementation of urban transport strategies. The purpose of the present
paper is to take stock, a decade later, of how the urban transport
problems of the developing world have changed and the extent to
which, and success with which, the strategies recommended in
2002 have been implemented.
The report fell into four parts. The rst, identied and analyzed
the main strategic objectives with chapters on city economic
development, poverty reduction, the urban environment and
safety and security. The second part dealt with strategy for the

q This paper has benetted from input from former colleagues at the World Bank,
including Tony Bliss, Jean-Charles Crochet, Roger Gorham, Masami Kojima, Ajay
Kumar, Shomik Mehndiratta, Gerhard Menckhoff, Slobodan Mitric, Richard
Podolske, Jorge Rebelo, Tom Rickert, Chris Willoughby and Zhi Liu, They are not
responsible, however, for the opinions and judgments offered in the paper.
E-mail address: kgwilliam@yahoo.com.
1
A summarized version of the contents of the report is contained in Gwilliam
(2003).
0739-8859/$ e see front matter 2012 Published by Elsevier Ltd.
http://dx.doi.org/10.1016/j.retrec.2012.06.032

modes, comprising chapters on the road system, road passenger

transport, mass rapid transit, and non-motorized transport The
third dealt with the instruments of pricing and nancing and
urban transport institutions. The nal part considered how the
World Bank could assist countries to achieve their strategic
objectives. This review follows the same structure, with each
section briey describing the 2002 strategy and then discussing
subsequent developments.
2. Strategic objectives
2.1. City economic development
Cities on the Move argued that economies of agglomeration in trade
and industry generate the growth of cities where the advanced
sectors with the highest rate of growth of labor productivity are
located and that urban transport oils this engine of growth. However,
as cities grow and become richer, ownership and use of motorized
vehicles e including private vehicles (as in Eastern Europe), small
buses (as in much of Latin America and Africa) and 2e3 wheelers (as
in Asia) e grows more rapidly than the available road space. This
results in increased congestion and trafc-generated air pollution,
particularly in megacities with population of over 10 million. As the
sources of pollution differ so would the prescribed solutions.
The report identied several other factors contributing to urban
transport deterioration including inadequate quantity and structure
of road space and poorly developed institutional, scal and

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

regulatory arrangements at the municipal level. Excessive concentration of activity in the capital cities and megacities was to be
addressed by removal of the scal advantages which they enjoy in
many countries. Counterproductive policies on land use control
were to be improved by more reliance on market signals. And
inconsistencies between sector policies were to be addressed by the
incorporation of all sector policies within a coherent city development strategy.
2.1.1. Urbanization continues.
By 2008, more than half of the globes population, 3.3 billion
people, was living in towns and cities, and urban population is
projected to grow to 4.9 billion by 2030. At the global level, all
future population growth will thus be in towns and cities. Most of
this growth will be in developing countries. The urban population
of Africa and Asia is expected to double between 2000 and 2030. It
will also continue to expand, but more slowly, in Latin America and
the Caribbean (United Nations Population Fund, 2010). Poverty is
now growing faster in urban than in rural areas: a billion people
already live in urban slums.
2.1.2. .but is not conned to megacities
While the number of megacities with a population of over 10
million is expected to increase from the present 27, most of the
growth is expected to occur in smaller and medium sized cities.
The number of cities with more than one million people has
already grown from 74 in 1950 to 442 today (National
Geographic, 2011). The policy implication is that urban transport strategy must concentrate more than was suggested in Cities
on the Move on the medium sized cities, which are often less well
endowed in either nancial or human resources to deal with the
transport problems of urbanization. It is notable that in the
national transport strategy reviews currently taking place in
Russia and India, the focus of urban transport strategy has shifted
from the problems of megacities to the more general issues
which affect cities of all size.
2.1.3. In many countries the economic context is now more
favorable.
The rst decade of the new millennium was also a period of
growth for many of the developing countries. Economic growth
proceeded at about 8% per annum in India and 10% per annum in
both China, while sub-Saharan Africa, achieved overall growth in
excess of 5% per annum for a continuous period of over 5 years at
the end of the decade, despite the world recession and the
continued ravages of war in some of its countries. The improvement
of the economy in general is having signicant effects on urban
transport, including an actual or incipient explosion in private car
ownership in the wealthier countries or cities. The effect of that
rapid rate of income growth has been an unprecedentedly rapid
growth in car ownership which has had profound implications for
transport infrastructure, public transport, transport nance and
institutions.
2.1.4. .and car ownership is accelerating at an unprecedented rate
Recent research on car ownership trends in developing
countries throws some light on this. Dargay, Gately, and Sommer
(2007) use a conventional sigmoid curve to model the relationship between car ownership and income, but allow the saturation level to be determined endogenously to reect differences in
urbanization and population density. They conclude that saturation levels for countries presently in the earlier stages of
development are likely to be similar to those of already developed countries with similar density and urbanization, and that
the elasticity of car ownership with respect to income peaks at

a value over two somewhere in the range of average GDP per

capita of $5000e$10,000. The fact that many developing countries with GDP in this range also have relatively high rates of
growth of income over time thus explains the phenomenon of
rates of growth of car ownership which are consistently higher
than those previously experienced in developed countries. Taken
together with the relatively much lower income elasticity and
growth rate of road stock (Ingram & Liu, 2000) this also presages
a continuing deterioration in urban road conditions in developing countries.
2.1.5. Transitional economies are adjusting to a changed economic
paradigm.
Although by 2000 the Russian Federation had substantially
recovered from the post-liberalization economic depression, many
other countries of the former Soviet Union were still struggling to
overcome the traumatic economic changes associated with the
dismantling of the economic linkages of the FSU. Similarly, some of
the former COMECON member states had still to nd effective
realignment of their economic systems and trading patterns as they
lost their traditional markets or suppliers in the FSU. By 2010, the
necessary readjustments were largely achieved. The exploitation of
the oil and natural gas reserves of the region in a booming international market created wealth for a number of the states of the
region (including Azerbaijan, Kazakhstan, and Turkmenistan as
well as the Russian Federation), which had benecial spin-off
effects for their regional trading partners. Access of several of the
former COMECON states to the European Union also boosted their
economies.
2.1.6. .but an effective balance between planning and market
mechanisms remains elusive
In addition to the former COMECON countries many other
countries such as China, India and Vietnam have shifted in the
direction of market economics during the nineties to great effect.
Greater involvement of the private sector in manufacturing has
enabled high growth rates to be maintained during the last decade,
despite the world recession. But that shift has been less extensive or
effective in the urban transport sector. Two aspects have been
particularly disappointing. First, the private sector contribution to
transport infrastructure investment, of which so much was expected in the late nineties, has not ourished as hoped. Private
investment in infrastructure continues to be concentrated in Latin
America and East Asia. Second, private sector provision of urban
transport services, particularly bus services, has tended to be
accidental and chaotic rather than planned and well managed. In
both cases, as discussed later, the problem has arisen from a failure
to nd an appropriate procurement or regulatory mechanism to
reconcile the public need for efciently provided service at an
affordable price with the private sector requirement of an adequate
return on capital and management effort supplied.
2.1.7. .and freight transport efciency is an increasing concern
There is also a growing recognition of the importance of urban
freight transport to the efciency e and environmental sustainability of cities. For example in China, where road freight transport
accounts for over 50 percent of transport energy consumption, two
factors have hindered its efciency. The fuel efciency of Chinese
trucks is estimated to be 30 percent below that of the industrialized
countries, while logistics management is also comparatively backward, partly as a consequence of the fragmented nature of the
trucking industry. Donors are assisting efforts to develop comprehensive green freight programs at the national level in Brazil and
at the city level in Guangdong, China (Mehndiratta, Liu, & Fang,
2011). But those types of program are still rare.

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

2.2. Poverty reduction

Cities on the Move viewed poverty as a multi-dimensional
concept involving the lack of the economic social and cultural
means to procure a minimum level of nutrition and to participate
in the everyday life of society. Despite economic growth, it
observed that in many cities some sections of the community e
women, the elderly and the disabled as well as those with low
incomes e appeared to be getting worse rather than better off.
Transport contributed to the perverse distributional effects of
growth through a dynamic similar to that which limits the efciency of cities. As cities expand the price of more accessible land
increases. Poor people are forced to live on less-expensive land,
either in inner-city slums or on city peripheries. As average
incomes grow and car ownership and use increases, the patronage,
nancial viability, and eventually quality and quantity of traditional public transport diminishes, to be replaced by a proliferation of smaller, more congesting vehicles. Motorization, which is
permitted by the growth process, may thus also make some poor
people even poorer. In particular, in the absence of efcient
congestion pricing for road use, piecemeal investment to eliminate
bottlenecks will almost certainly benet the relatively wealthy at
the expense of the poor.
2.2.1. The affordability of transport is now better understood.
The concept of affordability of transport was introduced to the
World Bank by Armstrong-Wright (1986). In the developing
country context, much useful empirical data has been assembled
recently by Carruthers, Dick, and Saurkar (2005). Using a common
indicator reecting the costs of a specied number of public
transport work journeys, they show that the percentage of
average income spent on public transport varies from about 1
percent of city average income in Manila and Bangkok to 11
percent in Sao Paulo. For people in the bottom quintile the story is
different with some cities requiring 30% of income. The main
exception to this generality were in some Chinese cities where
low income workers were heavily dependent on walking and
cycling, and hence had low expenditures on journeys to work
(Peng, 2005) and in some Latin American and Indian cities where
the very poor often live in informal settlements near to sources of
employment.
2.2.2. Though it must be used cautiously in policy analysis
The affordability concept is still often used normatively. For
example, Litman (2008), while emphasizing the need to consider
housing and transportation expenditures together, argues that
transport is unaffordable if it amounts for more than 20% of the
household income. In contrast, Gmez-Lobo (2011) argues that
the required affordability benchmark to determine whether
transport costs are high or not is arbitrary and that any approach
that uses the absolute level of an affordability measures to assess
policies is meaningless. He argues that the overly simplistic
assumption that public transport should be subsidized because
public transport use is greater for the poor than the rich is
particularly to be avoided because it begs the question of why
transport rather than food, housing, education or health expenditures should be the subject of general subsidy. At the empirical
level there is evidence that, because they travel more, such
general subsidy may actually benet the rich more than the poor
(Carruthers et al., 2005).
On the basis of careful analysis some conclusions do nevertheless appear to be supportable. First, the theoretical objection to
subsidizing transport is weakened where there are spillover effects
to production which may arise if transport costs result in a less
effective job search or to higher levels of unemployment than

would be the case with lower transport costs.2 Carruthers et al.

(2005) quote some anecdotal evidence of such effects in Wuhan,
China, while Baker, Basu, Cropper, Lall, and Takeuchi (2004) document the differences of commuting distances between income
groups and the apparent exclusion of poorer suburban dwellers
from higher paid employment in the distant business district in
Mumbai, India. Second, where there is potential for service to be
provided economically at below standard fares (for example offpeak use of under-utilized public transport itineraries), economic
efciency requires that such price differentials should be offered.
Third, whatever conclusions are reached about public transport
subsidy as an instrument for the reduction of urban poverty, it is
clear that cheaper, more efcient, public transport is benecial to
the poor. Hence poverty considerations should enhance rather than
blur the emphasis which was put on efciency of public transport
supply arrangements, discussed in more detail later.
2.2.3. .and transport for the disabled is receiving increased
attention
Transport for the disabled has been receiving greatly increased
attention during the last decade. At the international level, the UN
adopted a Convention on the Rights of Persons with Disabilities in
2006, and in 2010 established a group to prepare practical guidelines for countries which have signed the Convention. The development banks have also incorporated provisions for the disabled in
their investment projects, for example the World Bank funding
accessibility improvements in sidewalks associated with the new
Lima Metro. At the national level some large developing countries
have now adopted laws on transport for the disabled (Brazil in 2004
and the Philippines, Vietnam and Mexico in 2010), though this has
not always had very immediate results. At the city level, accessibility
improvements are now being introduced particularly in the context
of new mass rapid transit facilities (elevators in metro systems and
extended platform boarding of BRT systems in Latin America) and
existing bus systems (for example low oor or wheelchair ramp
equipped buses and vans). In Shenzhen, China, a substantial
program of investments in curb ramps, tactile guide ways and
accessible taxi and metro services, has been developed on the basis
of a municipal regulation (Access Exchange International, 2011).
Not surprisingly poorer countries and cities tend to do least,
partly because the informal sector paratransit services on which
they depend are particularly difcult to inuence. But even in
middle income countries, if there is no legal obligation progress
tends to be concentrated in show cities, often the capital. For
example, while two thirds of buses in Buenos Aires are accessible
low-oor models, the proportion is only 5% in the rest of Argentina.
Retrotting to existing systems is more costly and therefore more
difcult to achieve than in new systems where good design rather
than huge extra investment is required. Co-ordination between
interacting agencies is also often poor. For example, it took protracted legal action to secure compliance in making the citys
subway stations fully accessible in Rio de Janeiro. So there is still
much to achieve.
2.3. The urban environment
In environmental policy Cities on the Move put the main
emphasis on the local air pollutants which were most immediately
damaging to health, particularly lead in gasoline, ne particulate
matter and ozone. As ozone is not directly generated by transport
this also meant attention to the volatile organic compounds and

2
This is similar to the merit good argument for subsidies to health and
education.

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

oxides of nitrogen which are its precursors. Reduction of global

warming gases was seen to be often synergistic with the reduction
of local air pollution (particularly through the reduction of fossil
fuel consumption) but also called for special approaches (such as
the global overlay of the Global Environment Fund) where that
synergy is weak. Noise and other disturbances were given little
attention. The recommended policies were partly technological and
partly managerial. In respect of fuels the emphasis was put on
eliminating lead from gasoline, reducing sulfur in gasoline and
diesel, with some caution expressed on most of the alternative fuels
e including biofuels. In respect of vehicles, the emphasis was put on
developing emission suppression technologies, reducing fuel
consumption rates and shifting from two stroke to four stroke
engine motorcycles. Improved inspection and maintenance
programs were considered to have high potential. In system
management, the paper called for better structured domestic
taxation and import duty policies, and better trafc management,
including priorities for public transport. Very little in this eclectic
agenda was new, but great emphasis was put on the adoption of
measures which were practical and well aligned with economic
incentives.
2.3.1. One major pollutant has been eradicated.
The most gratifying progress has been made in respect of lead.
In 2000 India and China moved to lead free gasoline, but by October
2005 there were still 81 countries selling leaded gasoline. By June
2011 only six countries (Afghanistan, Algeria, Iran, Democratic
Republic of Korea, Myanmar and Yemen) were reported to be still
selling leaded gasoline. The ght against lead in gasoline is therefore effectively over. While there have been some concerns about
the alternative octane enhancers MMT3 and MTBE,4 even these
have been diminished by the increasing use of ethanol.
2.3.2. .but others are increasing
The same is not true of most other local air pollutants in the
developing and transitional economies. For PM10, the World Health
Organization (WHO) advisory norm is a maximum annual average
of 20 micrograms per cubic meter (mg/m3) and 50 mg/m3 for the 24h mean. While most western European and North American cities
have ambient levels around or only a little above the norm, most
large Chinese and Indian cities had annual average values
exceeding 60 mg/m3, and some, including Tianjin, Chungking, Kolkata, Delhi, Kanpur and Lucknow, exceeded 100 (World Bank, 2008)
The contribution of the developing countries to global warming is
also increasing (IPCC, 2007). In aggregate, though not per capita,
China has already overtaken the United States as the chief emitter
of greenhouse gases (World Bank, 2010). Most of the policy
prescriptions proposed to address this includes some combination
of increasing vehicle efciency and/or fuel substitution with no
increase in vehicle kilometers of travel e which is proving very
difcult to achieve in developing countries with increasing car
ownership. Moreover, climate nance instruments, such as the
Clean Development Mechanism characterized by Cities on the Move
as potentially promising in the urban transport sector have generally not proved to offer a workable model for urban transport
nance.
2.3.3. Quality of traditional fossil fuels is increasingly critical.
Sulfur in fuel now appears to play a critical role in efforts to reduce
particulates because even relatively low levels of sulfur in fuels
reduce the efciency of catalytic converters and other exhaust control

3
4

MMT is short for Methylcyclopentadienyl manganese tricarbonyl.

MTBE is short for Methyl Tertiary Butyl Ether.

devices, and, for some such as diesel particulate lters, even render
them inoperative. The gasoline sulfur content standard in the United
States has been set at less than 30 ppm over 2004e2007, and diesel
sulfur at less than 15 ppm over 2006e2010, while EU Directive 2003/
17/EC introduced a new phase-in requirement for both gasoline and
diesel, restricting the maximum sulfur content to 10 ppm from
January 1, 2009. As late as 2006 few developing countries had
stringent standards for sulfur in diesel. In Argentina, Kenya, and
Bolivia, the maximum allowable limit for sulfur in fuels was 500 mg
per kg, which was quarter that in Pakistan, one third of that in
Guatemala, El Salvador, Honduras, Malaysia, and Tanzania, and half of
that in Bangladesh, the Philippines, Thailand, Columbia, Paraguay,
Nicaragua, and Panama (CONCAWE, 2006). However, national vehicle
emission standards are now changing rapidly. While there are still
countries in Latin America with standards for cars and light vehicles
equivalent to Euro 1 or less, many of the larger countries in the region
such as Argentina and Mexico are already adopting Euro 4 standards
for new vehicles. New member countries of the European Union
accept EU standards. And large countries in the rest of the world, such
as Russia, Thailand, Malaysia, Indonesia and the Philippines have
committed to introduce Euro 4 standards for new vehicles by 2012
(Delphi, 2011). Fuel standards will follow accordingly.
2.3.4. .alternative, cleaner fuels have not yet made a breakthrough
Alternative fuels, given some consideration in Cities on the Move,
have not made the expected breakthrough. Liqueed petroleum gas
(LPG), is the third most widely used motor fuel in the world with
14.6 million vehicles estimated to be fueled by propane gas
worldwide in 2008.5 It is supported by preferential taxation in
many European countries. It is relatively safe and easy to distribute
and has a signicant market share in some, particularly the
Republic of Korea, Turkey, Poland, and Japan.6
Compressed natural gas (CNG) powered 12.7 million vehicles
worldwide in 2010.7 It is most important in a few countries with
domestic reserves and a dense distribution market, such as
Pakistan, Iran, Argentina, and Brazil, and for high mileage public
transport eets (taxis and buses), particularly in locations where
the use of CNG is mandated (e.g. buses in Delhi, Ahmedabad and
Rangoon) or supported by signicant price advantage (for taxis in
Argentina and Brazil) or tax advantage (in China and Malaysia). It
maybe expected to increase its market share where there are large
local reserves (as in Peru), but elsewhere the costs of fuel distribution, the sparcity of distribution networks and the space of onvehicle storage, limits its use to a few heavily concentrated eet
operations.
2.3.5. .and the role of biofuels is increasingly controversial
Liquid biofuels have the advantage of being readily usable in
transport without signicant modication in existing vehicles or
infrastructure (Kojima, 2010). The International Energy Association
(IEA) estimates that liquid biofuels accounted for 1.5 percent of
global transportation fuel, and the volume of liquid biofuels
supplied could double by 2015 (IEA, 2009). With the aid of heavy
subsidy and protection in the United States, which with Brazil
accounted for about 90 percent of total supply, world fuel ethanol
production rose from 17 billion in 2000 to 66 billion liters in 2008.
Biodiesel production, 70% of which is from the European Union,
rose even faster, from less than 1 billion liters in 2000 to 12 billion
liters in 2008 (REN21, 2011, pp. 13e14).

5
See
http://www.worldlpgas.com/page_attachments/0000/2191/WLPGA_
ARTICLE.pdf.
6
See http://www.worldlpgas.com/gain/key-autogas-data/.
7
See http://www.iangv.org/tools-resources/statistics.html.

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

Most recently the contributions of biofuels to energy conservation and environmental protection have both been challenged
(Delucchi, 2010). In energy terms, estimates by Bourne and Clark
(2007, p. 41) point to modest results for corn ethanol produced in
the United States: one unit of fossil-fuel energy is required to create
1.3 energy units from the resulting ethanol compared with a ratio of
1:8 for Brazilian ethanol from sugar cane.8 In environmental terms,
while many biofuels directly generate one third less GHGs by volume
than gasoline, the advantage is reduced by the environmental effects
associated with their production, including depletion of natural
resources, razing of forests to open land for cultivation, and damage
to ecosystems (Phalan, 2009). Nearly half of all biofuels, including
U.S. maize ethanol, soy diesel, and Malaysian palm-oil dieseldmay
even have greater environmental costs than fossil fuels (Zah et al.,
2007).
The UN Food and Agriculture Organization (FAO) suggests that
diversion of maize and oilseeds to biofuel production and soaring oil
prices together were the major drivers for the record prices for basic
foods set in June 2008, pushing an additional 115 million people
into chronic hunger (FAO, 2009). This has already prompted some
governments to direct efforts for biofuel production away from
food-based feedstock. In 2007, the Chinese government stopped
approving new plants processing grains, including ethanol plants.
Indias biofuel strategy focuses on the use of nonfood feedstock:
molasses for ethanol and non-edible oilseeds for biodiesel. South
Africas biofuels industrial strategy, issued in December 2007,
excluded maizedan important staple among the poordfrom the
governments biofuel policy (DME, 2007). Mexico stopped
providing nancial support to maize-based biofuel projects in 2008.
Second-generation biofuels, derived from feedstock from non-food
crops, wastes and by-products and a possible third generation of
biofuels derived from algae may change the picture (Dragone,
Fernandes, Vicente, & Teixeira, 2010). But they are still under
development, and their costs are high and signicant technical
barriers remain (Brennan & Owende, 2010).
Despite these disadvantages there may still be an important role
for biofuels in developing countries. For example, Melillo et al.
(2009) consider Africa to be the best place to produce cellulosic
biofuels which will lead to most carbon capture in the long run,
though they also argue that the CO2 account might not benet until
mid-century. Modes of use may also improve. There is an emerging
consensus that biomass is better used in stationary applications
(UNEP 2009; WBGU, 2009); a recent analysis of full lifecycle GHG
emissions concluded that using biofuels in stationary applications
to produce energy for electric vehicles would reduce GHG emissions more than direct use to fuel conventional spark ignition
(Campbell, Lobell, & Field, 2009).
2.3.6. But the motorcycle problem now looks more tractable
Motorcycles were seen as a major problem in Asia both because
they were very polluting (especially the two stroke engines common
in South Asia) and because, even if clean, they were a precursor of an
unsustainable motorization as incomes increased (as in Vietnam
and China). The problem of trading up from motorcycles to cars still
exists. In Ahmedabad, India, two and three wheeler motorcycles
account for 30% of all trips. In Hanoi, Vietnam, though the average
ownership rate of motorcycles is already 0.5 per capita, and still
increasing at 14% per annum, increased car ownership has already
led to increased congestion and serious accident rates.
The environmental issue now looks rather different, however.
First, state of the art technology is a great improvement with

8
A more favorable estimate for the US is contained in nepis.epa.gov/Exe/ZyPURL.
cgi?DockeyP1006DXP.txt.

four-stroke engine motorcycles replacing two-strokes in Vietnam

and India. Second, the emergence of the electric motorcycle and
scooter offers and environmentally much cleaner alternative to the
gasoline vehicle, both at the point of use (where emissions are zero)
and overall (so long as the electricity is not generated from coal
red stations near to cities). China already produces over 20 million
electric motorcycles per year. Unfortunately, the current models of
electric motor scooters in the Chinese domestic market are relatively small and slow e concentrating on replacing bicycles for use
in the reserved bicycle paths. Thus, although cheaper to purchase
and operate, they have not yet made a breakthrough in the large
Indian and Vietnamese markets (ADB, 2009). So the environmental
objective would now seem to call for a shift to electric motorcycles,
more powerful than those currently spreading through the Chinese
market, supported by appropriate tax and car restraint policies to
maintain an advantage for the motorcycle even as incomes
increase.

2.4. Safety and security

Cities on the Move described the road safety situation in urban
areas in developing countries as a global pandemic with nearly
0.5 million deaths and up to 15 million people injured annually,
mostly poor pedestrians, cyclists and motorcyclists. It argued that
this could be reduced signicantly by improved road design and
trafc management policies, including the introduction of safety
audits on all road engineering schemes, as well as by better medical
response provisions. But it would also require increased safety
awareness to change driver and pedestrian behavior, and training
of staff for specic road-safety coordinating agencies or councils, at
both the national and municipal levels.
Personal security in public transport (particularly for women)
was also seen as a growing social problem in many countries,
calling for better data to enhance ofcial awareness of the problem,
and greater commitment of police authorities to arrest and the
courts to appropriately penalize offenders. Physical improvements
are also important, for example, street lightingddesigned to
improve pedestrian securitydcan be included in street improvement, and particularly in slum-upgrading projects. Franchise
conditions for public transport can give incentives for improved
attention to security by public transport operators.
2.4.1. .in the developed world the situation is improving
It is difcult to identify precisely what has happened to urban
road accident fatalities and injuries since 2000 because most data
sets, including those produced by WHO (2009), are at the national
level, and there is known to be very substantial under-reporting.
Despite those caveats, a number of characteristics are apparent.
At the national level, the high income countries of Western Europe
mostly have annual fatality rates between 3.5 and 6.0 per 100,000
population. In Australasia, the US and the highest income Asian
countries like Japan and Singapore, there has been continuing
improvement in performance. Rates in the large urban areas are
typically between one third and one half of the national rates. Both
at national and city levels they have also mostly been falling
steadily at 3 or 4% per annum through the decade. The transition
economies of Eastern Europe have fatality rates typically double
this and in many cases they are more or less stable.9

9
Interestingly, in a comparison between the countries which now form the EU
27 and the CIS countries showed them to have very similar fatality rates of around
13 per 100,000 in 1988, but by 2006 the CIS average rate had risen to about 17
while that of the EU 27 had fallen to about 8. Russia is particularly bad.

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

2.4.2. .but in the developing world the picture is still grim

Outside the industrialized countries experience is very different.
Latin American annual fatality rates fall mostly between 13 and 21
per 100,000 people, with some beginning to fall and some still
rising slightly. The two most populous countries of the world, China
and India, have rates between 16 and 17 as have two other very large
countries, Indonesia and Vietnam. The largest central Asian countries, Iran, Iraq and Kazakhstan all have eight times the Japanese
rates. African countries, almost without exception, have very high
rates, mostly between 28 and 35 per 100,000 people. In most of the
poor countries the rates are still increasing as car ownership is
increasing. As might be expected, the proportion of fatalities among
the vulnerable groups e pedestrians, cyclists and motorcyclists e is
roughly consonant with their proportion of trips by those modes.
Globally, this group accounts for 46% of all fatalities, but is less than
25% in the United States but up to 80% in Thailand. The long term
effects of injury accidents can also be catastrophic. It was estimated
that of the 95,000 people injured in trafc crashes in Turkey in 2005,
13% had a subsequent disability (Esiyok et al., 2005); in India it has
been estimated that there are 2 million people permanently
disabled by road crashes. During the last ten years the situation has
not improved in developing countries (Bliss & Breen, 2009).

systematized form of safety audit/inspection. An interesting

breakthrough with the evaluation methods has been the use of
a simple value of statistical life formula (70 times GDP per capita) to
assess the benets of safety improvements to improve the safety
rating of road sections which has helped to highlight ex ante the
scale of safety benets.
Capacity development is critical. The World Report highlights the
fundamental role of the lead agency in ensuring the effective and
efcient functioning of the road safety management system. It
argues that conduct of a safety management capacity review is
a vital rst step in the process of a country achieving this and that
the process of institutional strengthening must be carefully phased.
Initially, emphasis must be put on improving the focus on results
and related inter-agency coordination. As these institutional
management functions become more effective the remaining
management functions are in turn strengthened. Cities remain
a special case in this process as their safety management arrangements are often subordinated to trafc management functions and
remain ill-dened or absent.
3. The modal components
3.1. The urban road system

2.4.3. .though the situation is now changing

However, the stage is now set for action over the coming decade.
Safety has a higher prole, and more partners are engaged. In 2004
the World Bank partnered with the WHO to produce the World
Report on Road Trafc Injury Prevention, setting out a blueprint for
action. In 2009 the World Bank produced some extensive country
guidelines (Bliss & Breen, 2009) for implementing that action. In
March 2010 the United Nations General Assembly resolution proclaimed a Decade of Action for Road Safety 2011e2020 with a goal
of stabilizing and then reducing the forecast level of road trafc
fatalities around the world by increasing activities conducted at
national, regional and global levels. Achieving this goal would save
around 5 million lives and avoid 50 million serious injuries, for
a social benet of US$3 trillion. Nearly 60% of the lives saved and
serious injuries avoided would be in the East Asia Pacic and South
Asia regions alone, with another 18% in Sub-Saharan Africa. In 2011
the World Health Organization and the United Nations regional
commissions, in cooperation with other partners in the United
Nations Road Safety Collaboration and other stakeholders,
produced a Global Plan and also in 2011 the seven multilateral
development banks (MDBs) committed to a shared program of road
safety activities to support the U.N.
2.4.4. .and a common strategy appears to be emerging
In content, the various initiatives have much in common. All
emphasize the importance of institutional factors and similar lists
of main areas of action e road safety management, safer roads and
mobility, safer vehicles, safer road users, and post-crash response.
The main policy shift is to the adoption of the Safe System
approach under which the setting of speed limits is determined by
the intrinsic protective quality of the road sections and vehicles
concerned, rather than the speed behavior of road vehicle users.
The biomechanical tolerance of humans to crash impacts is the
limiting factor.
This approach represents a radical shift to making mobility
a function of safety, rather than vice versa. Its focus on safer and
reduced speeds harmonizes with other efforts to reduce local air
pollution, greenhouse gases and energy consumption. And its
priority to afford protection to all road users is inclusive of the most
vulnerable at-risk groups such as pedestrians, young and old,
cyclists and motorcyclists. In support of this is the emergence of
safety rating tools for infrastructure, which are a more

Cities on the Move recognized that the road system was the
fundamental core of all city transport systems, but observed that it
is usually managed in fragmented and uneconomic way. Decisions
on the management, maintenance and expansion of urban road
systems usually rest with separate public sector agencies, while
those concerning operations on the system are predominantly
private sector. The report therefore concluded that the most
important requirement of a strategy for urban roads was to link
those public and private sector decisions in a coherent way. The
framework of a metropolitan strategic transport plan was important for achieving this, but changes in the approach to the separate
functions were also needed. On road system investment the report
emphasized the need for more rigorous economic appraisal of
investments and the development of a logical hierarchical structure
with clear allocation of responsibility for provision and maintenance of each category. On road maintenance more stable funding
mechanism were usually needed, as were more scientic maintenance management systems and improved private sector contracting capabilities and arrangements. On trafc management it
was argued that improved technical capability (for example
computer linked signal systems), need to be complemented by
strategic decisions to use the system to improve the ow of people
rather than vehicles (which implies priority to public transport and
non-motorized movement). On demand management the report
emphasized the need to balance demand with capacity through
restraint measures such as parking control and congestion pricing.
3.1.1. Road infrastructure remains inadequate in many cities
In a study of 14 African cities Kumar and Barrett (2008) identied a range of deciencies in the urban road systems. In quantity
the networks were insufcient, accounting for less than 7 percent
of the land area in cities e only about one third of that in most
developed cities. Service lanes are absent, and street lighting is
minimal. Moreover the majority of the roads had one lane in each
direction; where the roads are wider, one lane is often taken up by
pedestrians and parked vehicles and outlying neighborhoods could
only be reached only by two-wheeled vehicles. Only one third of
these roads were paved, giving a paved road density is typically on
the order of 300 meters per thousand inhabitants (or close to two
kilometers per square kilometer), which is only about one third of
the average for developing country cities worldwide.

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

More generally, road systems in developing countries are poorly

structured, with little attention to ensuring that the system works
as a system. For example, the very large Chinese road program of
the last decade has concentrated almost entirely on primary roads.
Bangkok is also a classic example of the problems which arise when
there is inadequate local distribution capacity. Fortunately some of
the more sophisticated cities now recognize the importance of
having a clear functional classication of roads, so that road design
and management can be matched with trafc volumes, composition and safety priorities. For example, the recently adopted urban
transport strategy for St. Petersburg, Russia, identied the need for
a more appropriate classication of roads as the heart of its road
sector strategy.
Relatively modest expenditure on methodical assessment can
also improve the efciency of the road system considerably. For
example, in Bogota Colombia, despite having a problem of adequacy
of funds, the city keeps a relatively up-to-date asset register/road
management system which includes about 114,000 segments to
describe its road network. On average the condition of the primary
road network is surveyed every four years. Trafc counts are also
carried out every four years. With that data available it is possible to
determine the optimum road network maintenance strategy and
related budget, using computerized road maintenance management systems (RMMS) which takes into account vehicle operating
costs, unit cost of road works, and pavement strength and condition.
3.1.2. Road maintenance nancing arrangements remain poor in
many countries.
The condition of urban roads is still unsatisfactory in many
countries. While Sub-Saharan Africa is perhaps the extreme case,
some recent evidence from that region is instructive. The recent
AICD study (Gwilliam, 2011) included some estimates of the urban
road stock and its condition in 20 African countries, covering a total
urban road network of 268,490 km, of which 164,440 was in South
Africa. Excluding South Africa, only a little over 40% were in good or
fair condition (Table 1).
Table 1
Condition of urban roads in 20 AICD countries (percent of totals).
Including South Africa

Paved
Gravel
Earth

Excluding South Africa

Good

Fair

Poor

Good

Fair

Poor

26.1
26.7
0.5

11.2
12.7
0.5

8.2
13.3
0.6

25.5
15.5
1.4

14.4
15.0
1.3

6.7
18.5
1.6

Source: Gwilliam, 2011.

The reason for this poor state is partly institutional. Most African
countries have now established a second generation road fund, and
have moved toward a policy of funding road maintenance through
various kinds of road-user charges. But although most road use
occurs in urban areas, and hence most of the fuel levy revenue is
collected from urban road users, the allocation of the resulting
scal resources usually fails to reect that. While in Ghana and
Ethiopia 20e30 percent of the road fund is allocated for urban road
maintenance, of on an average only about 10 percent of road fund
revenues go to urban roads. A restructuring of road-fund allocations to more closely reect observed trafc patterns could help to
alleviate this problem. In addition, urban road funds could be
supplemented by charges on parking for private cars or on urban
developments that impose a measurable transport burden.
3.1.3. .and implementation of work is inefcient
Work implementation also needs to be efcient. While many
cities still undertake maintenance by force account, there is an

increasing trend to contract implementation out to private sector

contractors. In that case the contracting arrangements are critical.
Competitive tendering usually requires that there are a number of
separate packages to let and monitor. For efciency of this process,
the number of packages should not be too large. For example, the
city of Bogota recently reduced the number of packages from 330 to
six (organized on an area basis) in order to improve the efciency of
contract enforcement.
Other arrangements are possible. In particular longer term
performance based contracts are now being used in many countries. In 1995, Argentina introduced performance-based contracts,
which at present cover 44% of its national network. In the mid
nineties Uruguay also piloted PBC, rst on a small portion of its
national network and then on the main urban roads of Montevideo.
Shortly thereafter, other Latin American countries, such as Brazil,
Chile, Colombia, Ecuador, Guatemala, Mexico and Peru, also started
adopting a performance-based approach. Preparations for launching PBC programs are underway in Albania, Cape Verde, Chad,
Madagascar, Tanzania, Burkina Faso, India, Cambodia, Thailand,
Indonesia, Vietnam and Yemen.
3.1.4. Trafc management still receives inadequate attention
Poor trafc management is still common. On the basis of a study
of 14 cities in Sub-Saharan Africa, Kumar and Barrett (2008) reported that road intersections were usually too close together and
ill-designed for turning. Commercial activities (such as street
vendors) and vehicle parking force pedestrians off the sidewalks
into the roadway, reducing the capacity of the roadway and posing
safety hazards. Competition at bus stops causes localized congestion that can spillover into adjoining trafc lanes. Moreover, little
attention has been paid to facilitating the operation of public
transport systems; dedicated bus lanes are rare, bus stops and bus
shelters are scarce and in poor condition, and bus bays are narrow
and cannot accommodate multiple buses. In the few cases where
measures to favor bus travel have been introduced, they have not
been properly enforced. Nowhere have they been effective. Finally,
most cities have ignored the needs of pedestrians and cyclists.
While the African situation may be extreme, it does reect
a general lack of attention to trafc management. Even in countries
which have reached high levels of car ownership, city administrations often lack an adequately skilled and resources trafc
management unit. In both China and Russia, institutions are a key
element of this as trafc police, who have traditionally implemented
trafc control, have little coordination with construction bureaus.
There are signs of change however. Following the success of Moscows comprehensive reform and strengthening of its trafc function in 2000 the Russian federal government has issued instructions
that the police should concentrate on trafc law enforcement,
leaving the primary role in trafc management to civilian authorities. This is having some effect. For example, following the establishment of a Road Trafc Management Unit in the city of
Novosibirsk e a city with a population of 1.4 million e an automatic
road trafc management system was introduced. In the subsequent
four years, during which the total vehicle stock increased by 23% the
city reported that average trafc speeds had increased by 15e20%
and trafc fatalities and injuries decreased by 18e20%.
3.1.5. .and demand management is generally absent or ineffective
While the impossibility of providing enough road space for
unrestricted private car use is increasingly recognized, demand
management is usually weak. Direct charging for use of congested
road space has made little progress in the developing world.
Conscious restriction of the amount of road space provided e such
as the demolition of the Cheonggye and Samil elevated highway in
Seoul to re-establish an urban park e is rare.

10

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

The most commonly effective instrument e restraint through

parking control e is still little used in developing and transitional
countries. Parking policy is still usually concerned with satisfying
rather than managing trip demand. For example, Barter (2011)
shows that, while experience is very varied, most Asian cities still
approach off-street parking merely as a way of relaxing the pressure of on-street parking and hence have planning norms for new
developments specifying minimum rather than maximum parking
provisions. Economic pricing and time limits for parking are rare,
except in the developed cities of the region. Making things worse
there are often legal impediments e such as the prohibition on
charging for on-street parking in Russia, central government
imposed upper limits on parking fees in Poland or insurance laws
that discourage private provision of off-street parking in China. In
many countries the limitation is caused by an inability to enforce.
Driving restrictions are currently the most popular instrument
used to restrict congestion and environmental pollution in developing country cities. First introduced in Mexico City in 1988 (hoy no
circula), similar programs have been implemented in Bogota, Santiago, So Paolo, and, most recently, in Beijing in preparation for the
2008 Olympics. The peak period restrictions initially set in place in
the year 2000 in Bogot were extended in 2009 to the whole period
from 6 am to 8 pm on weekdays, in Bogot and other Colombian
cities. But the long term effectiveness of this instrument remains
questionable. In Mexico City pollution has continued to increase as
hoy no circula has led to an increase in the total number of
vehicles in circulation, that most of the additional cars are highemitting second hand cars imported from other parts of the
country, and that the use of taxis, which were also among the most
polluting vehicles in the city has increased.
3.2. Public road passenger transport
Cities on the Move observed that most urban bus services in
developing countries were now provided by the private sector,
many by fragmented informal sector operators, and usually in
a chaotic and often dangerous manner. Often this followed from the
decay of a badly regulated public or private sector monopoly. It was
argued, however, that, when appropriately regulated, competition
best guarantees efcient supply, mobilizing low cost operations to
provide the best possible service for any budget capability.
Competition for the market was generally preferred to free entry
to compete in the market, with gross cost tendering seen as the
most appropriate form of contract in complex urban situations. But
this was also seen to require a skilled regulatory and procurement
function and restructuring of any state enterprises as unsubsidized
commercial entities. Fare controls should be determined as part of
a comprehensive city transport nancing plan, and fare reductions
or exemptions nanced on the budget of the relevant line agency
(health, social sector, education, interior, and so on). The role of
paratransit in satisfying dispersed trip patterns and in exibly
addressing the demands of the poor was recognized. But it was
argued that cities should strive to nd ways to mobilize the
initiative potential of the informal sector through legalizing associations to participate in competitive processes, while meeting the
same environmental, safety, tax and insurance requirements as
formal operators.
3.2.1. The private sector continues to predominate in the road
passenger transport.
The private sector, and particularly the informal sector, has
continued to predominate in urban road passenger transport
provision in most parts of the world, though with signicant
regional and national differences. In Latin America supply
continues to be predominantly by the private sector with policy

development concentrated on trying to bring a fragmented private

sector into a more disciplined regulatory framework in cities such
as Santiago de Chile, Bogota, Buenos Aires and Lima In Eastern
Europe and Central Asia governments such as those of the Russian
Federation and Uzbekistan, while recognizing a need to accept the
supplementation of traditional public sector supply by the
emerging private sector, and using competitive tendering of franchising as a disciplined way of achieving this, have tended to
restrict it to commercial (as opposed to social) services. In SubSaharan Africa attempts to reconstitute the traditional public sector
operators in African cities such as Dakar and Ghana, have generally
had little impact. In south and east Asia the pattern is mixed. In
China, some cities, such as Shanghai and Guangzhou are now
attempting to use the private sector in a municipally regulated
regime, while others, such as Beijing, Taiyuan and Wuhan rely
predominantly on traditional municipal enterprises, albeit in corporatized form. In India the smaller cities rely on the private sector
while the larger cities such as Bombay and Delhi still have
municipal enterprises as the core of their systems. Only in some of
the North African countries do the traditional public enterprises
still dominate.
3.2.2. .and fragmentation of supply has proved difcult to handle
In most countries the private sector remains highly fragmented,
and authorities nd it difcult to enforce disciplined behavior.
Moreover, the private sector still concentrates on the small vehicle
sector in many countries. Vehicle size and operator size tend to be
closely correlated as the small operator is unable to nd the nance or
unwilling to take the risk of purchasing such an inexible xed asset
as a large bus. Institutional factors contribute signicantly to this.
Even where the municipalities attempt to manage the system
through contracting with private operators, they often fail to use
their contracting powers to stimulate consolidation. For example,
route franchising systems as diverse as those of Bogota, Santiago,
and many Russian cities have allowed the franchise holder to
subcontract operations to individual private suppliers. If the franchises are on a net cost basis the suppliers then have the incentive
to compete internally even within a route. This was the origin of the
penny wars common in Latin America. The tax system often
contributes to this as small operators are either taxed at a different
rate (as in Russian and many of the Central Asian countries), or
effectively avoid tax due to the difculties of enforcement, so there
is little incentive to consolidate.
Two reforms have been successful in addressing these issues.
First, requiring the franchisee to have effective control over all of
the vehicles operated under the franchise e imposed sufciently
tightly to prevent sub-contracting e has assisted the development
of companies or strong associations of operators in franchising in
Uzbekistan and Chile and now appears to be having some effect in
Colombia. Second, moving from net to gross cost contracting
managing the size of contracts, changes the operators incentives.
In Santiago Chile a reduction to a system of only 15 contracts has
revolutionized behavior.
3.2.2.1. .with the private sector poorly integrated into urban
transport systems. The origin of much private sector supply of
public transport was fragmented and opportunistic. As recommended by Cities of the Move private operators are increasingly
working under contracts with municipal authorities, often
competitively tendered. In the event that has not always resulted in
the private services becoming integrated into the urban transport
system in a number of circumstances.
The rst circumstance is where the regulation of the private
sector is quite separate from that for public sector services, as in Sri
Lanka. Similarly, in Bogota, Colombia, the Transmilenio system,

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

which was intended to be the core of a reformed trunk and feeder

system, was operated by an agency totally separate from the
Secretariat of Transportation which operated the traditional
permission system. And in that case there was antipathy between
the organizations. It is only in recent years that these two cases
have been addressed by the unication of the separate regulatory
systems. In neither case have the events yet totally played out.
A second common circumstance arises in the former socialist
economies where many traditional administrators see the growing
role of the private sector as an accidental abnormality outside the
traditional socially oriented supply arrangement. In those circumstances, existing in most Russian cities, the public sector still
operates the traditional social services, for which it receives
decit nancing, while the private sector provides commercial
services, sometimes at higher fares, not carrying concessionary fare
passengers but receiving no subsidy. These services may be operating under competitively tendered franchises. But there has
developed nevertheless a two-tier transport system often with
a declining social service sector. Anomalously, trams and trolleybuses, which should be the backbone of an integrated service
because of their higher xed costs are increasingly the residual
carriers of the lowest fare passengers.
3.2.3. .leaving the reform agenda incomplete
Some parts of the Cities on the Move agenda have been adopted
effectively. A more liberal approach to private sector entry has
increased supply; a move to contracting has improved the discipline of private sector supply; competitive tendering has reduced
costs and budget burden. But in two respects, experience of the last
decade demonstrates that the lessons of the previous decade have
not been learned.
First, the importance of a sufcient and predictable cash ow to
maintain satisfactory service has still not been recognized. Public
sector companies have been expected to continue to maintain
uneconomically low fares, or to provide extensive concessions on
the full fares, without adequate compensation. Even where, as in
some Russian10 and African cities, there is an occasional injection of
new capital equipment the revenue ow remains inadequate for the
continued maintenance and replacement of the assets, and the level
and quantity of service eventually regresses to its former low level.
Similarly the private sector will not invest in expensive equipment if
their revenues are constrained by uneconomically low fare control.
Second, available methods of mobilizing the efciency of
a competitively spurred supply within a publicly planned multimodal urban public transport sector have not been adopted.
Gross cost tendering as the basis of effective multi-modalism in
urban public transport is rare in the developing world. The full
potential of electronic ticketing has not been exploited. So the new
agenda must be one of improving the understanding of the ways in
which market processes (both tendering and integrated pricing)
can be mobilized to secure the social objectives as well as to
improve overall system efciency.
3.3. Rapid transit
Cities on the Move adopted a rather cautious tone on rapid
transit. While acknowledging that it could contribute to the

10
In Russia a common form of urban public transport enterprise is the unitary
municipal enterprise. This is a company solely owned by the municipality, operating, but not owning, the municipalitys buses. It relies solely on the municipal
budget for vehicle replacement. But the municipal budget is only 30% funded by
sources controlled by the municipality. The rest comes from transfers from the
regional government obtained annually on a request and negotiation basis. Bus
replacement, not surprisingly does not get consistently treated in this arrangement.

11

achievement of strategic objectives, the report observed that it did

not always do so, especially when it did not carry sufciently high
passenger volumes or when it imposed an insupportably high cost
burden. It therefore emphasized (a) the need for rapid transit to be
an integral part of a strategic plan including provisions for modal
co-ordination and for sustainable nance, and (b) the selection of
a rapid transit mode appropriate both to the travel volumes to be
served and to the nancial capabilities of the city. On planning it
was recommended that rapid transit development, even if privately
nanced, should be carefully integrated in the metropolitan
structure plan, with the public sector designing the network and
accepting responsibility for land acquisition, and that a range of
alternative modes should be considered. On nance it was recommended that the full nancial implications should be assessed ex
ante, and that implementation should not proceed until nancial
sustainability was assured. And on management it was recommended that rapid transit should be integrated with the rest of the
public transport system both physically and through an integrated
fares structure.
3.3.1. Rail rapid transit has grown explosively, particularly in East
Asia and Latin America.
Rail rapid transit has been long established in Europe, which has
75 of the worlds 175 systems, including 16 in the former COMECON
countries. Latin America has 19 operational systems, totaling 800
route kilometers, seven of which have opened since 2000. India,
despite its size, has only three systems (in Chennai, Kolkata and
New Delhi) with systems in construction in ve further cities. Cairo
has the only fully edged metro in Africa. excluding China and
Japan, most other large countries in East Asia only have MRT
systems in their capital cities (Bangkok, Manila, Kuala Lumpur). The
most explosive development has been in China, which already in
2011 has nearly 2000 km of route in 18 systems (16 cities), only
three of which existed before 1995. A further 18 are already under
construction, with plans to build 87 MTR lines totaling 2495 km in
25 cities between 2009 and 2015. Clearly this is a pace of development not envisaged by Cities on the Move.
Two issues e both ultimately nancial still limit this growth.
First, there is the problem of reconciling affordability to the user
with scal viability. While this has been overcome by central
government contributions to the capital nance in some cities (for
example in India), that tends to be a solution not available to the
poorer countries. Second, such long-lived and expensive undertakings are inherently complex and risky, prone to over optimistic
nancial forecasts and inviting corruption at the pre-investment
and procurement stages, especially with unsolicited proposals.
For those reasons at least it is sensible to consider other technological alternatives as a benchmark.
3.3.2. .and less expensive alternatives have emerged
Bus rapid transit (BRT) has now taken off as an alternative form
of mass transit. In its fullest form, in addition to larger vehicles and
segregated busways with stations in the middle of the road
(i.e. away from the kerb), BRT includes off-board fare collection
permitting bus entry through all doors, same-level (usually well
above the road level) boarding, control of bus access (the old busways in Bogot and Lima had no such control, resulting in
congestion caused by too many buses on the busway). Passing lanes
at stations permit express services, resulting in a quantum jump in
commercial speed for those services (>30 km/h in Bogot and
Lima) and line-haul capacity (44,000 passengers per hour in the
peak direction on Avenida Caracas in Bogot where many bus
services bypass otherwise overloaded stations). Of the 157 BRTs
now operating, only about 15 in Latin America, and a few elsewhere
(including Ahmedabad, Guangzhou, Johannesburg, Seoul and

12

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

Istanbul) meet these gold-standard conditions and can really

compete with rail systems in terms of commercial speed and
capacity.
There is an ever widening range of variants to be considered,
making it possible to choose a variant well suited to particular local
circumstances. For example, within the general class of BRT
systems, a distinction can be made between closed systems
meeting the full gold standard conditions and open systems
which do not use specialized vehicles, but operate as super buslanes, and hence can more easily provide a many to many route
structure instead of relying on a network of feeder services. The
relative merits of these two forms depends critically on the structure and density of the city and the extent to which congestion
impacts conventional bus services. The success of some of the
earlier (non gold standard) BRT systems, and the developments in
Light Rail Transit (LRT), has presented ample evidence on the
technical and economic possibilities of a range of technologies.
It is notable that alternative technologies are increasingly
explored. BRT can be introduced more quickly, more cheaply, and
sustained by lower trafc volumes than a rail mass transit system
but beyond some threshold trafc volume (about which there
remains some argument) it cannot match rail performance. It may
thus be seen less as a competitor with rail transit than as
a complement, allowing some of the advantages of MRT to be obtained at an appropriate timing in periods of rapid population and
income growth, and in parts of networks for which rail transport is
nancially unsustainable. Multi-technology MRT networks should
be, and are already becoming, the norm rather than the exception.
While there remains an underlying preference for the fully edged
underground or elevated rail metro, many cities, such as Bogota and
Curitiba have opted for the less expensive BRT in the rst phase of
a mass rapid transit development. Some of the largest cities in Asia
and Latin America (Beijing, Seoul, Mexico City, Santiago) have
invested in both, selecting the technology for specic lines based on
the expected trafc volumes as well as on the urgency with which
a development is expected.
Where cities have incorporated different technologies in their
mass rapid transit networks they have increasingly provided for
both physical interchange and for commercial integration in a citywide ticketing system. Some of the newer rail metros, such as
Guangzhou, have also followed the Bogota BRT example in planning
and providing for integration with cycling. In some cases, such as
Santiago, Chile, there has also been restructuring of bus networks to
integrate with a mass rapid transit trunk haul function. But more
generally lack of effective co-ordination with local bus services has
proved to be the greatest weakness of mass rapid transit development. Frequently this has its roots in the absence of a strategic
planning authority and the separation of mass rapid transit and bus
regulatory functions.
The benets of securing co-ordination of the widest possible
range of services has been amply demonstrated in London and
Paris, where not only bus and metro services, but also suburban
heavy rail services have been brought together in a commercially
and physically co-ordinated system. In the developing and transitional economies many such opportunities have been missed, often
because of the unwillingness of the national rail undertakings to
lose any of their prized autonomy. For example, it has proved very
difcult to secure co-ordination of the suburban rail services in
Mumbai with other modes. Similarly in St. Petersburg, Russia, the
potential to structure the expanding suburbs around existing
suburban rail lines are not currently being adequately exploited.
3.3.3. .but the strategic selection remains controversial
The selection of technology and timing of investment in MRT is
still somewhat controversial. Cities on the Move recommended that

these issues should be addressed through the traditional cost/

benet analysis. More recently, a more dynamic kind of analysis has
been suggested, reecting the Japanese tradition. Acharya and
Morichi (2007), argue that the relationship between motorization
and mass rapid transit is plagued by two vicious circles. In the
short run, increasing incomes permit increasing car ownership,
which in turn decreases public transport patronage, which
responds by reducing frequencies of service, which encourages
increased car ownership and use, and so on in a circular fashion. In
the longer term increased use of cars encourages low density
suburbanization which cannot be so easily served by public transport so the demand for and use of cars increases e also in a circular
fashion. Once set in motion these cycles are difcult to break. From
this they deduce that the critical problem is to identify the stage in
the development of income and population in a city before which
an attractive mass rapid transit option must be developed to forestall the vicious circles. They argue that historical data shows that
a subway system is warranted at a threshold population of 2.8
million and a threshold GDP per capita of 6202 PPP 1990 dollars,
and therefore they construct an Income-Population Normalized
Index (IPN) which can be applied to assess the appropriate timing
for MRT investment for any given city. This theory still requires
much further development if it is to become an operational tool.
But it does highlight the importance of taking a broader and more
dynamic view of the impacts of MRT than the approach in Cities on
the Move.
3.3.4. Private sector nance of new systems is developing slowly.
In So Paulo a privately concessioned Metro Line 4, inserted in
an otherwise publicly owned system, started operating in October
2011 and is already carrying 550,000 passengers per day. A
monorail is also being developed under PPP arrangements as are
new metro lines in Mumbai and Chennai, India. But most new
urban rail investments are publicly funded. In Delhi only the airport
link is concessioned. Even in Bangkok, where the rst line (BTS)
was nanced privately on a non-recourse basis, the extension of
that line, and the civil and mechanical engineering investments of
the second (blue) line, was publicly nanced. For the blue line only
the electrical and signals engineering and rolling stock investments
were nanced by the private sector concessionaire. In a rather
similar way the ostensibly privately nanced systems in Kuala
Lumpur have been taken into direct state control as a consequence
of their nancial problems. Overall, the role of private nance in the
new MRT systems has been limited.
3.3.5. .but the private sector role in existing systems has
broadened in Latin America
Existing systems are a different matter. Early concessions of the
existing suburban railways and metro in Buenos Aires were victims
of their own success, and had to be renegotiated in a politically
controversial process to adapt to the need for additional capital
investments to handle expanded demand (Guasch, Laffont, &
Straub, 2008). Subsequently the concessions of suburban rail and
Metro systems in Rio de Janeiro, which were originally for operations only, have been extended for another 25 years with obligations of the concessionaires to invest in rolling stock, signaling and
infrastructure. The So Paulo suburban railway organization (CPTM)
has also now entered a PPP for provision and maintenance of its
rolling stock in one of its lines and the So Paulo Government is
seeking further PPPs. Latin American experience clearly demonstrates that if private participation in urban rail development is to
be tted successfully in a comprehensive urban transport strategy,
in which the public sector retains a planning and regulatory role,
institutional and contractual arrangements need careful design,
with investment obligations and risks assumed by both parties.

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

3.4. Non-motorized transport

Cities on the Move argued that an explicit strategy for NMT is
necessary to redress a historic vicious policy circle that has biased
urban transport policy unduly in favor of sacricing the interests of
pedestrians and cyclists to those of motor vehicle users, with the
consequence that NMT becomes less safe, less convenient, and less
attractive, making the forecast decline of NMT a self-fullling
prophecy .It was concluded that there should be clear provision for
the rights of pedestrians and bicyclists in a national trafc law which
would be a facilitating framework for explicit formulation of local
NMT plans by municipal authorities. These plans should provide for
separate infrastructure where appropriate (for safe movement and
for secure parking of vehicles), while standards of provision for
bicyclists and pedestrians should be included in new road infrastructure design. Trafc management should be focused on
improving the movement of people rather than of motorized vehicles
and trafc police trained to enforce the rights of NMT in trafc
priorities. To ensure adequate nance for infrastructure responsibilities for provision for NMT should be included in Road Fund statutes and procedures, while nance of bicycles in poor countries
should be assisted by development of small-scale credit mechanisms.
3.4.1. The modal share of NMT is generally declining
In the last decade walking and cycling has been increasing
slightly in the industrialized countries e mainly for health and
recreation purposes. In Africa and Latin America little has changed in
aggregate, with 50% of trips still walking trips in Africa, though the
proportion of bicycle trips is very low e less than 1% of all trips in
South Africa. In contrast, there has been a steady decline in aggregate in the bicycle sector in Asia. In China, bicycle ownership fell
from 1.9 to 1.1 per household between 1995 and 2005 And the
bicycle share of trips has fallen e in Guangzhou, for example, from
34% in the early nineties to 16% in 2000 and 11% in 2010. At the
national level it is forecast that the bicycle proportion of trips will
fall from 50% at the beginning of the millennium to 17% in 2040. In
both Vietnam and India the bicycle is being rapidly replaced by the
motorcycle. In Hanoi, motorcycle ownership had reached 0.5 per
capita and they accounted for 65% of all vehicular trips in 2006
(World Bank, 2006). In Ahmedabad 30% of all trips (37.5% of vehicular trips) were undertaken by motorcycle in 2007 (ADB, 2009).
3.4.2. .partly as a result of unfriendly public policies
This has not been helped by public policy. The 2004 Road Trafc
Safety Law in China was predicated on protecting the weak from the
strong and afrmed the pedestrians right to life over the motorists
right of way. But that law took ten years in preparation and has
proved difcult to implement. In practice for most of the last decade
policy in China has actually been positively inimical to the bicycle. Up
to 2009 Guangzhou had not built a new bike lane for 12 years;
Shanghai prohibited the use of bicycles on trunk roads in 2007; and
Shenzhen had changed its urban road planning standards for roads,
removing bicycle lanes to improve capacity for cars. In Zhengzhou
the Bureau of Education prohibited primary school students cycling
to school. While many cities pay lip service to NMT by the creation of
cycle paths (for example, both Warsaw and St Petersburg have cycle
path plans), for the most part they are directed at leisure cycling (e.g.
in parks) or are not fully segregated from motorized trafc. Moreover,
when they have appeared to impinge at all to the disadvantage of
motorists they have been rejected, as in Nairobi (Earthtec, 2007).
3.4.3. .and cycle ownership is a continuing constraint
An early attempt to provide credit for cycle purchase in Lima,
Peru failed because of the problem of collateral with the traditional
municipal bank, as did an early lease program in Tanzania. But

13

group savings program in Kenya were more successful. Government support is critical, as in the South African shova kalula
program inaugurated in 2001, which now aims to roll out a million
bicycles by 2014, but had only introduced 72,000 bicycles by early
2011. The elimination of import duties helps, as in Kenya (Earthtec,
2007). An alternative approach is adopted in Hangzhou, China
where a short term bicycle rent program, has 2000 stations and
50,000 bicycles for hire.
3.4.4. But there are now signs of change in policies on cycling
Most signicantly Chinese policies are becoming more cycling
friendly. A number of recent World Bank projects have assisted
cities in providing a more cycle friendly environment, including
road design on a safe corridor philosophy, re-introduction and
strengthening of separation barriers, and a shift from investments in
primary to secondary and tertiary roads (Mehndiratta, 2011). Many
pedal cycles in China are being replaced by electric cycles which
have much of the same environmental advantage but offer greater
range, speed and comfort. Over 20 million electric cycles are being
produced annually. Moreover, China has chosen to limit the power
and speed of the electric cycle to allow them to continue to use the
cycle paths and to avoid the congested motor roads. Cycling is also
increasingly being addressed as part of a system. A pattern of
coherent, complete, segregated cycle networks has been introduced
with great success in Bogota, Colombia. Since the completion of the
300 km cicloruta network in 2001 bicycle use has been increased by
ve times, and now nearly 400,000 cycle trips are made daily in the
city. The synergy between the cycle (pedal or electric) as an access
mode to mass rapid transit in cities is increasingly recognized and
planned for. Bogotas Transmilenio and the new major BRT network
in Guangzhou are good examples of that integration with well
organized bicycle parking facilities provided.
Elsewhere, in 2007 the government of Botswana commissioned
a report on best practices in NMT planning implementation and
maintenance (Earthtec, 2007). In 2008 the South African government
published its draft proposals for NMT, which included the obligation
for NMT plans to be prepared at both the provincial and local levels,
and for an NMT fund to be established from part of the national Road
Fund revenues (South Africa Department of Transport, 2008).
3.4.5. .and the importance of walkability is increasingly
recognized
Most trips involve at least a short walk leg. And in many
developing country cities walking accounts for half of all total trips.
Yet it is notoriously uncomfortable and dangerous, having traditionally received little attention in the design of transport systems.
This limits accessibility, increases costs and lowers the quality of
life. A number of design factors contribute to this. Systems containing large block sizes, sparse secondary and tertiary networks
and wide roads with large set-backs reduce pedestrian accessibility
to buildings e less than one quarter of the numbers of jobs are
accessible within 10 min walk of major rail stations in Beijing than
in New York (Torres-Montoya, Yanan, Dubin, & Mehndiratta, 2010).
This is often accentuated by insufcient and poorly designed street
crossing places, lack of pedestrian phases in signal setting, poor
road condition and impediments such as cars parked on sidewalks.
These issues are now being addressed in a comprehensive way in
some of the World Bank projects in China (Mehndiratta, 2011).
4. The policy instruments
4.1. Pricing
Cities on the Move noted the complexities arising from the
multiplicity of objectives being pursued in urban transport and by

14

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

the institutional separation of road infrastructure from operations,

of infrastructure pricing from charging, and of roads from other
modes of urban transport. In the interests of both urban transport
integration and nancial sustainability it recommended that
developing countries should move toward prices reecting full
social costs (including congestion and environmental externalities)
for all modes; and to a targeted approach for subsidies reecting
strategic objectives.
4.1.1. Pricing policies are developing slowly
The emphasis on full cost pricing was driven by two objectives e
to generate the funds for asset maintenance and replacement and
to avoid distortions in mode choice e particularly where private car
users are not charged the full marginal social cost of congestion.
In respect of nancing maintenance the road sector has
improved considerably over the last decade. In Africa, where the
needs were greatest, not only have all countries now established
second generation road funds but they have progressively set fuel
tax surcharges at higher levels to fund maintenance (Gwilliam,
2011). The story is not so satisfactory in public transport. Many of
the remaining municipally owned operators in Russia and other
countries of the former Soviet Union continue to be decit
nanced, and to rely on a very uncertain ow of funds from the
regional budgets or from special central government targeted
programs to replace their vehicles. As a consequence the publicly
operated social services continue to decline despite a decade of
relative prosperity at the national level.
In respect of distortions, direct application of congestion charges
is still restricted to a few developed countries. But it is notable that
over the last decade fuel prices in China have risen from being 10%
less than those of the US to being nearly 30% greater, while India
continues to keep prices about halfway between US and European
levels (Mehndiratta et al., 2011).
4.1.2. And the emerging surrogate restraints on motorization
remain crude
In respect of modal distortions the main issue has been the
choice between public and private transport in congested areas.
The fact that congestion pricing has not been introduced does not
necessarily mean that the modal distortions go completely
unaddressed. In many industrialized country cities a combination
of parking restrictions and parking charges is used as a surrogate
for a congestion pricing measure. Unfortunately, few developing
countries exercise any effective control over parking. Even worse,
in some countries such as Russia it has not been legally possible to
restrict on-street parking or to charge for it. As a consequence
historic cities like St. Petersburg are being asphyxiated with
trafc.
Schemes to restrict car use have been implemented in Sao Paulo,
Athens, Beijing, Manila, Mexico City, Santiago, Seoul, and Shanghai
(ADB, 2008). For some of the above cities, these restrictions were
implemented only during days when air pollution reached critical
levels (such as in Santiago and Mexico City) or during periods of
extreme congestion (Seoul during the Olympic Games). These
schemes restricted plates to enter the target zone which was
either applied to the entire CBD or to a specic area of the city. In
certain cases, this measure has proved highly effective in curbing
congestion. In Seoul, the Sippujae program was put into operation
during periods of extreme congestion. The resulting outcome was
a reduction in total trafc by 7 percent, while vehicle speeds
increased by 14 percent. However, if employed for extended
periods, license plate control programs run the risk of increasing
the vehicle stock, as people purchase vehicles to avoid the restrictions of the system, as happened in Mexico City. Moreover, the
second vehicle was typically older and less energy-efcient.

The most important aspect of these restriction programs is their

initial design. If the goal of the program is to reduce congestion or
air pollution, vehicle exemptions cannot be allowed. For example,
in the original inception of Hoy No Circula, taxis and colectivos
were exempt from the restrictions which led to an increase in their
utilization, which were even less energy-efcient and created more
emissions than other passenger vehicles. As a result, in the
programs subsequent updates, taxis were added to the list of
vehicles that were prohibited. Another major problem concerning
plate restriction programs surrounds the issue of equity. In many
cities, lower income communities rely on passenger vehicles as
a means to commute to work, rather than for leisure purposes.
Restricting vehicular use can also unintentionally limit employment opportunities.
4.1.3. .but subsidies are still poorly targeted
Cities on the move ventured into the economics of the second
best by conceding (in a rather unusual way for a World Bank
document) that subsidies might have a role to play in achieving
efciency and environmental, as well as in social objectives. In the
rst two cases the mechanism would be through the subsidization
of less congesting or less polluting transport alternatives, while in
the latter the mechanism would be to transfer income in kind to
specic needy categories when the (generally preferable) direct
nancial transfers were not considered feasible. In the event,
experience has been very mixed in all three of these policy
dimensions.
For both economic efciency and environmental reasons a shift
from private to public transport would appear to be desirable. But
for lower income countries, where public transport still dominates,
there is little evidence that subsidies are designed for that end.
Even in middle income countries where emerging motorization is
a problem, public transport subsidy is not an important strategic
instrument. In Latin America and most of Asia public transport is in
any case commercial, while in Russia and central Asia subsidies are
usually socially rather than economically or environmentally
motivated. While electric transport (tram and trolleybuses) in the
former Soviet Union is more heavily subsidized than autobus
transport, this is mainly because the electric modes are publicly
owned and operated and carry a large share of concessionary fare
passengers, rather from any considered economic or environmental
strategy.
However, a number of developing countries are beginning to use
differential tax rates (a form of subsidy) to encourage less polluting
technologies. For example, in many major Chinese cities, local
governments provide nancial support to encourage the use of CNG
and Liquid Petroleum Gas in transport (Zhao, 2006). The most
generous incentives apply to biofuels, not only in Europe and the
United States but also in developing countries. Even governments
of countries that are low-cost producers of feedstocks have used tax
breaks and other scal incentives for biofuel manufacturers.
Argentina provides scal incentives including tax exemptions for
biofuels. In mid-December 2009, Thailand, a low-cost producer of
sugarcane, offered subsidies as high as $1.80 a liter of ethanol
blended for E10dmore than triple the ex-renery price of gasoline.
Other support policies include various scal incentives provided to
manufacturers and minimum price guarantees. Honduras provides
fundsdexempt from customs tariffs, income tax, and other related
taxes for 12 yearsdfor production of biodiesel, provided that at
least 51 percent of the feedstock is grown in the country while
China provides subsidies to ve fuel ethanol producers, amounting
to approximately $258 per tonne of ethanol in 2008, as well as
waiving the consumption tax and value-added tax entirely for fuel
ethanol (Kojima, 2010). Such scal incentives may help to support
domestic fuel production and possibly yield environmental

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

benets. But they also need to be seen in the wider context of their
adverse effect on scarce government revenues and on the distribution of income. In that context, the overall effect may be negative.
In respect of socially oriented subsidies two different
approaches are adopted. Controlling the general level of fares to
what is considered an affordable level is explicit policy in the FSU,
but in many cities has the perverse effect of progressively diminishing the level of service that the budget provision affords.
Personally targeted subsidies are usually targeted at students and
old people who are believed to have less ability to pay. But those
subsidies miss what may be the most vulnerable group, very low
income or unemployed adults. Even the work trip subsidy implemented through the Brazilian vale transporte fails to reach the
informal sector.
4.2. Finance
Cities on the Move said little about nancing of urban transport
except to emphasize the need for integrated urban transport
funding. Experience in the last decade suggests that this was
a serious omission as failures in the transport sector have often
been associated with a general nancial failure of the cities.
4.2.1. Municipal tax revenues are typically inadequate
In most countries municipal tax revenues are very limited. For
example, in Russia the municipalities only nance about 30% of
their expenditures from their own resources, with the rest coming
from intergovernmental transfers. The conventional explanation of
this is that only property tax meets the criteria for good local taxes
e ease of local administration, impact predominantly on local
residents, and not raising problems of harmonization or competition between sub-national local or regional governments or
between sub-national and national governments. In practice,
however, property tax is difcult to administer and property taxes
seldom account for more than 20 percent of local current revenues
or less than 1 percent of total public spending in developing
countries (Bahl & Martinez-Vasquez, 2007). For local property tax
to play a larger role local governments must be allowed to set their
own tax rates, the tax base must be maintained, property transfer
taxes reformed to prevent under-declaration of sales values, and
procedures reformed to improve collection efciency, valuation
accuracy, and the coverage of the potential tax base. Given these
limitations, Bahl and Bird (2008) conclude that in most developing
countries there are other potentially sound and productive taxes
that sub-national governments could use: PIT surcharges, taxes on
the use of motor vehicles, payroll taxes, and even sub-national VATs
and local business value taxes may all be viable options in
particular countries.
4.2.2. .and borrowing capabilities are weak
In principle, capital expenditures could be nanced by local
borrowing. While some large cities are at least as creditworthy as
their national governments, but that is not usually the case for
smaller cities, however, and borrowing may have to be by
government on their behalf. Sood (2010) argues that the most
critical avenue for sustained nancing of urban infrastructure will
be domestic credit markets. Several approaches have been tried.
Quasi-independent municipal credit institutionsdsuch as the
Municipal Development Fund in the Philippines, the Municipal
Fund for Infrastructure Finance in the Czech Republic, and the Tamil
Nadu Urban Development Fund of Indiadhave been established to
channel borrowed and grant funds to local governments. Credit
enhancement mechanisms have been developed e such as the
Local Government Units Guarantee Corporation in the Philippines
and the Infrastructure Credit Guarantee Fund in the Republic of

15

Korea. And some special purpose vehicles have been set up, such as
the Water and Sanitation Pooled Funds in Tamil Nadu and Karnataka in India to raise nance for small municipalities through
bonds, the Investment Fund for Urban Development in Vietnam,
and the Urban Development Investment Corporations in China to
manage specic sectors of urban infrastructure in each of the major
cities. However, ADB (2007) reports that the experience with these
initiatives has been mixed and that they have, at best, had limited
impact and, at worst, crowded out private nance.
4.2.3. .but new sources of funding have been developed
An alternative cost based approach, is to require developers to
provide their own common infrastructure and to recover their
costs through land sales. Developer exactions require developers
to install at their own expense the internal infrastructure needed
to meet development standards or to pay for infrastructure
elements provided by public authorities. Impact fees are designed
to cover the external infrastructure costs caused by demand for
system wide expansion in infrastructure capacity for roads and
other social services caused by a new development In Mumbai, it
has been estimated that a 10 percent development fee, imposed
on the cost of new construction, could nance as much as 40e50
percent of all regional infrastructure investment required over the
next two decades. However, there is often a considerable margin
of uncertainty about the amount of external infrastructure
expansion to associate with any particular new development so
that impact fees have to be negotiated and often cover less than
the full costs. For example, the government of Chile attempted to
nance the external development costs of the metropolitan
extension of Santiago in Chacabuca province by impact fees
negotiated with developers but were only able to recoup 39% of
the total estimated costs.
The establishment of new town corporations allows the developer to recoup the costs of infrastructure through land and property sales. These corporations can be entirely public, as in the UK or
public private participatory ventures, as in Denmark. Many new
towns in the developing world build directly on the experience
with new towns in Great Britain. In India the private sector has
been involved as developer in a national plan for 11 airports,
including Bangalore and Hyderabad. Some state agencies have also
acquired surrounding land, with the intention of nancing access
roads to the airports through prots from land sales (Kagonova and
Peterson, 2010).
Lease or sale of unused urban land can also contribute, in
Ethiopia, for example, 90 percent of the proceeds of municipal land
leasing should, by national law, be used to nance municipal
infrastructure investment. In other settings, the fact that publicly
owned land in cities is held by state or national development
agencies has made it more difcult to use the prots from land sales
to nance infrastructure investments that, by law, are the responsibility of municipal governments. In contrast in China, where all
urban land is owned by the municipal government, an entire
generation of urban highways and other urban public works has
been nanced throughout by sales of rural land rezoned for urban
development or by direct municipal land leasing combined with
borrowing against land collateral (Peterson, 2007). Of course, land
leasing cannot be the primary source of infrastructure nance
indenitely; the major cities of Chinas coastal region by now have
largely exhausted land-leasing revenue as a source of infrastructure
nance, after 15e20 years of reliance on it. Also the danger is that
the attractiveness of converting rural land for urban development
for a municipality containing a large amount of rural land (as many
municipalities do) leads to unnecessarily large conversions and
sprawling development with associated transport and environmental damages (Liu & Salzberg, 2012).

16

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

Sale of development rights is an alternative to the sale of land to

capture incremental value. In So Paulo, Brazil, the municipality
charges a preset fee for additional oor space beyond the normal
maximum density, in locations authorized for higher-density
development. This instrument has also been used in India. In
April 2007, the state of Maharashtra (India) approved a rise in the
maximum oor space index (FSI, the ratio of oor space to land area
on a lot) throughout two of Mumbais districts from 1.0 to 1.3, with
the extra FSI sold to developers.
4.2.4. .though integrated urban transport funding is rare
Integrated urban transport funding is still a rarity, though it can
be done. A striking example is that of the recently created Lagos
Area Metropolitan Transport Authority (see more on this institution
below). The law grants LAMATA several powers to facilitate the
discharge of its statutory functions, including the power to levy and
collect user charges in connection with the provision of its services
and to collect any other tariffs, fees and road taxes as may be
authorized by the Governor. A transport fund has now been set up
with the following sources: (a) Lagos State budget provision; (b)
license fees (hackney permit, road taxes, license plate registration,
auto registration); (c) concession fees; (d) other road user charges
(tolls). LAMATA has successfully made a case with the Joint Tax
Board to increase road user charges, to be shared between LAMATA
(50 percent), state treasury ofce (40 percent), Motor Vehicle
Authority (ve percent), and state MOT (ve percent). As a result,
LAMATA is able to meet 60 percent of the 2009 funding requirements from state support and the transport fund (from less than 20
percent in 2004), with the objective of becoming self nancing in
the coming years.
4.3. Institutions
Cities on the move attributed many of the observed failures in
urban transport to institutional weaknesses. National governments
were held responsible for failure to provide adequate nance for
nationally mandated policies, particularly in the elds of road
maintenance and transport, and to create an appropriate regulation
and competition policy. At the regional and local government level,
major failings were observed both in spatial and functional coordination. And at the operational level trafc management, road
maintenance and public transport operational agencies were found
to be often poorly staffed and extremely conservative in behavior,
with poor and inadequate mobilization of the possibilities of
commercially based nance and management.
The proposed strategy had four main components. The recommendations for functional capability and human resource development included the better identication and allocation of
responsibilities at the local level, the development of a national
strategy for training in urban transport, and greater collaboration
between authorities to create centers of excellence to share scarce
skills. For jurisdictional co-ordination it was recommended that
allocation of responsibility between levels of jurisdiction should be
clearly established by law with inter-governmental nancial
transfers consistent with the allocation of responsibilities, and that
formal arrangements for collaboration or integration should be
made at the metropolitan level, encouraged by national government. For functional co-ordination it was recommended that there
should be a strategic land use and transport strategy with which all
of the functional activities, including policing, should be aligned.
Finally, for effective involvement of the private sector it was recommended that planning and operating responsibility for public
transport should be separated as should technical and economic
regulation, while operating agencies should be fully commercialized or privatized.

4.3.1. The capability need is increasingly well recognized

Many countries already recognize weaknesses in their current
human skill endowments, particularly in secondary cities. In
Thailand the lack of skills at the city level was addressed by the
establishment of a centre of excellence in Bangkok which could
provide assistance to the cities. Similarly, in Russia, where
responsibility for urban transport is by the constitution located
with the regions the federal government is nevertheless planning
the development of a national urban transport advisory unit to
assist with training and the dissemination of good practice. But
more needs to be done, with the establishment of a respected,
portable professional qualication seen as the key to developing
and spreading skills.
4.3.2. .and some countries are approaching transport as
a metropolitan function
The need to approach urban transport on a metropolitan level is
increasingly recognized for the large conurbations. But the position
varies substantially from country to country. For example, the
municipal boundaries in China are usually so wide that most
metropolitan development takes place within the existing municipal structure. The ongoing reform of government structures in
India provided for the creation of a new layer of responsibility at the
metropolitan level. In contrast, the creation of metropolitan organizations in Russia is inhibited by the allocation of responsibilities
to federal or regional governments embodied in the constitution.
Though there are some constitutional provisions for local democratic self-government at the municipal level these only appear to
facilitate metropolitan organization through agreement between
jurisdictions (as in the developing Moscow hub).
A good deal of emphasis has been put on this need in Latin
America. In association with World Bank funding of urban rail
rehabilitation programs attempt were made to set up strategic
metropolitan transport planning bodies in most of the major cities.
Some e as in Sao Paulo e have developed into strong and sophisticated planning agencies.
A handful of African cities in the sample studied have established agencies with overarching responsibility for urban transport.
Where they do exist they usually lack the necessary executive
powers to implement their vision and must work through other
agencies of government, as in Abidjan (AGETU), Bamako (DRCTU),
and Dakar (CETUD). In Addis Ababa, a city-wide transport authority
does exist, but it is not autonomous with respect to the city
government. In Dar es Salaam a multi-sector regulatory agency, the
Surface and Marine Transport Regulatory Authority (SUMATRA),
was established in 2001 to regulate rail, road, and maritime
transport services.
Probably the most striking recent development has been in
Lagos, where Lagos Metropolitan Area Transport Authority
(LAMATA), launched in December 2003, is a corporate body with
a Board of Directors representing transport operators, transport
unions, the organized private sector, the general public, local
government areas, and transport related agencies. It has overall
responsibility for transport planning and coordination in the Lagos
metropolitan area. LAMATA has already been successful in (a)
preparing a strategic long-term plan for the transport sector in
Lagos; (b) coordinating activities of the multiple agencies involved
in the sector; (c) rationalizing motor vehicle tax administration,
resulting in a substantial increase in revenues; (d) maintaining,
upgrading, and rehabilitating the declared road network; (e)
implementing a pilot Bus Rapid Transit (BRT) system, and (f) most
important, changing the attitude among users toward bus transport
system. Two ingredients appear to have been critical in this success.
First, it had the support of a strong champion in the form of the
State Governor. Second, Staff remuneration and retention packages

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

are pitched at private sector levels so as to attract, retain, and

motivate high caliber staff.
4.3.3. .but a comprehensive strategy basis is often still lacking
A comprehensive strategy is very difcult to achieve without
a multi-functional authority able to exercise control over all of the
elements of the urban transport system. But the existence of
a comprehensive political control does not necessarily ensure that
a comprehensive strategy is developed and implemented. For
example, the Manila Metropolitan Development Authority has
done little real strategic planning. Even in the former planned
economies of the former Soviet Union, traditions of functional
separation of administration mean that, even now, planning is little
more than physical infrastructure planning through a general plan
(Genplan) containing little on such strategic issues as trafc
management, demand management, public transport procurement
and commercial strategy, etc. For example, only in 2011 has St.
Petersburg e the second largest city in Russia e developed
a genuine strategic plan.
There are important exceptions of course. Curitiba, Brazil, has
implemented an integrated approach to land use control, trafc and
public transport management for over thirty years, and more
recently other Latin American cities such as Sao Paulo and Santiago,
have developed both the technical skills and the political
commitment to adopt a more comprehensive strategic approach.
Seoul, Korea is following a similar policy. But there remains a great
need to develop that approach in most developing country cities.
4.3.4. .and the understanding of how best to mobilize the private
sector is weak
The private sector is viewed as increasingly important, even in
the former socialist countries, as unsubsidized operators of buses
and as investors in new rail systems. Most countries also recognize
that there is a residual public sector role as network planner and
performance monitor in the bus sector and as promoter and facilitator of right of way acquisition for rail systems. But in both subsectors the limitations of that understanding frequently detracts
from the value that could be obtained from the private sector.
Bus franchising is often seen primarily as a way of re-imposing
traditional supply procedures on suppliers of what are otherwise
purely commercial services. For example, in competitive tendering
in Russian cities much emphasis is put on traditional procedures
such as daily medical checks of staff, at-gate inspection procedures,
accession to public dispatching arrangements, etc., and little or no
emphasis on cost. What is typically missing is any appreciation of
the benets that can be obtained by a more pro-active use of
tendering to obtain lower cost supply of unremunerative services,
or to secure system integration through gross cost tendering
arrangements. Moreover, there is rarely any appreciation of the
potential benets of putting competitive pressure on high cost
public sector operators. Where this has been attempted, as in
Hanoi, Vietnam, competition has been ineffective because the
public operator has continued to receive decit nancing which
allows it to pervert the competitive process.
Rail concessioning has been undermined by similar conceptual
blind spots. While the rst round of concessions of urban services
in Buenos Aires were sophisticated in design and highly successful
in implementation, that very success raised the need for contract
adjustments and exibility which had not been provided for. In the
concessioning of African rail systems provisions for compensating
for the continuation of unremunerative passenger services were
included but not honored.
Nevertheless, some lessons have been learned. The franchising
system in Uzbekistan addressed the needs both for structural reorganization and vehicle lease nancing for the private sector.

17

After attempting a pure non-recourse nance approach with the

BTS Line system, the Thai government recognized that to achieve
their social and environmental objectives they would have to inject
public money, which they have done through taking the civil
engineering costs on to the public budget while concessioning
operations in the later Blue line investment and BTS line extensions. What is needed is a further codication of that experience in
a form which can be easily communicated to other administrations.
5. The future role of the development banks
Many developing countries have experienced a decade of
continuing urbanization and relatively fast economic growth since
the publication of Cities on the Move. In the urban transport sector
there have been some notable achievements. Transport generated
lead pollution has disappeared and price and tax differentials are
being used to encourage less polluting modes and fuels. Attention
to the transport problems of the disabled has increased, and there
are signs of a new international commitment to road safety. In
some cities urban road maintenance has been improved through
better contracting procedures. Failing public sector bus transport
supply has been supplemented by the private sector, which is
increasingly operating under competitively tendered contracts
which improve discipline and effectiveness. Mass rapid transport is
expanding rapidly, particularly in Asia, and new lower cost BRT and
LRT alternatives are being integrated into networks. Although nonmotorized transport is decreasing overall, some renewed efforts are
now being made to protect it by segregation of track. Novel sources
of land based nance are being tapped to support trunk road and
rail infrastructure development in the major. A few new metropolitan transport authorities have been developed, and several
countries such as India and Russia are considering their development on a systematic basis.
Despite those achievements there is a considerable downside. In
many middle income cities income growth has already stimulated
a rapid growth of car ownership so that the problems of congestion,
and the effects of motorization on the poor, have been accentuated.
Road trafc accidents and air pollution have also increased.
Improvements are still needed in most of the policy areas discussed
in Cities on the Move. Land use planning is poorly integrated with
transport in most cities. Municipal nances remain weak, and
require more secure revenue sources committed by tax reform.
Road systems are inadequate in quality, poorly nanced and
managed; safety remains a pressing priority; trafc management
institutions are particularly weak, and despite an increasing
recognition of the need to mange demand few cities have effective
instruments to achieve this. Parking policies are often perverse,
encouraging trips rather than restraining them in congested areas.
Private suppliers of public transport are frequently separated from
the traditional public transport networks rather than integrated
into them. Above all, even where there does appear to be an
appropriate institutional base for comprehensive urban transport
planning there is rarely a comprehensive planning system and
process. Major reforms are still needed.
The development banks e and particularly the World Bank e
have traditionally leveraged their efforts to assist reforms through
their investment lending. Given that many of the weaknesses
described above are regulatory and institutional it is tempting to
conclude that the recipe for the banks should be more of the same
thing. Certainly there is much left to achieve. But there are also
doubts about the sufciency of that formula in the changed
circumstances. Historically the successes in urban transport e for
example the institutional reforms achieved recently in Lagos e have
been concentrated on capital cities. But much of the growth of urban
population is now in medium and large cities rather than the

18

K. Gwilliam / Research in Transportation Economics 40 (2013) 3e18

megacities the problems are now impacting cities which are less well
equipped to handle them in either nancial or human resources. The
challenge for the next decade is to broaden that impact.
So the suggested agenda for the development banks is to
establish, with the support of national governments, sustainable
programs for urban transport in secondary cities. To do that, two
things are necessary. First, a method of rolling out the reforms to
secondary cities must be established. That has already been achieved in a number of cases. In Colombia, the Bogota BRT development has been extended, with central government support, to
secondary cities. In Uzbekistan the national government legislated
for the introduction of competitive tendering of bus services as
a response to a World Bank project limited to ve cities. And in
Parana state, Brazil, the Curitiba system is being replicated in
smaller cities. The common element in all these cases is the
commitment of national or state governments to support the
reforms. Second, the cities themselves must be convinced of the
benets of reform and capable of implementing them. This requires
both human resource development in the secondary cities as well
as scal reforms to make them self-sustaining. That is no mean task.

References
Access Exchange International (2011). website document at www.globalride-sf.org.
Acharya, S. R., & Morichi, S. (2007). Motorization and role of mass rapid transit in
East Asian megacities. IATSS Research, 31(2).
Armstrong-Wright, A. (1986). Urban transit systems: Guidelines for considering
options. World Bank Technical Paper 52. Washington, DC: World Bank.
Asian Development Bank. (2007). Infrastructure and private participation in Asia and
developing economies. Manila: Asian Development Bank.
Asian Development Bank. (2008). Urban transport energy use in the APEC region e
Trends and options. Manila: Asian Development Bank.
Asian Development Bank. (2009). Electric two wheeler vehicles in India and Vietnam.
Manila: ADB.
Bahl, R., & Bird, R. (2008). Subnational taxes in developing countries: the way
forward. Public Budgeting and Finance, (Winter).
Bahl, R., & Martinez-Vasquez, J. (2007). The property tax in developing countries:
current practice and prospects. In J. Riddell, & G. Cornia (Eds.), Toward a vision of
land in 2015 (pp. 23e46). Cambridge, MA: Lincoln Institute of Land Policy.
Baker, J., Basu, R., Cropper, M., Lall, S., & Takeuchi, A. (2004). Urban poverty and
transport: The case of Mumbai. World Bank Working Paper.
Barter, P. A. (2011). Parking policy in Asian cities. Manila: Asian Development Bank.
Bliss, T., & Breen, J. (2009). Implementing the recommendations of the world report on road
trafc injury prevention: Country guidelines for the conduct of road safety management
capacity reviews and the specication of lead agency reforms, investment strategies
and safe system projects. Washington DC: World Bank Global Road Safety Facility.
Bourne, J. K., Jr., & Clark, R. (October 2007). Green dreams. National Geographic
Magazine.
Brennan, L., & Owende, P. (2010). Biofuels from microalgaeea review of technologies for production, processing, and extractions of biofuels and co-products.
Renewable and Sustainable Energy Reviews, 14, 557e577.
Campbell, J. E., Lobell, D. B., & Field, C. B. (2009). Supporting online material for
greater transportation energy and GHG offsets from bioelectricity than ethanol.
Science Express. www.sciencemag.org/cgi/content/abstract/1168885v1.
Carruthers, R., Dick, M., & Saurkar, A. (2005). Affordability of public transport in
developing countries. Transport Paper No 3. Washington DC: World Bank.
CONCAWE. (2006). Motor vehicle emission regulations and fuel specications: Part 2
historic review (1996e2005). Brussels: CONCAWE.
Dargay, J., Gately, D., & Sommer, M. (2007). Vehicle ownership and income growth,
worldwide: 1960e2030. Washington DC: IMF.
Delphi. (2011). Worldwide emission standards; passenger cars and light duty vehicles.
Delphi Worldwide.
Delucchi, M. A. (2010). Impacts of biofuels on climate change, water use, and land
use. Annals of the New York Academy of Sciences, 1195, 28e45.
Department of Minerals and Energy. (2007). The biofuels industrial strategy of the
Republic of South Africa. http://www.dme.gov.za/pdfs/energy/renewable/
biofuels_indus_strat.pdf(2).pdf.
Dragone, G., Fernandes, B., Vicente, A. A., & Teixeira, J. A. (2010). Third generation
biofuels from microalgae. In A. Mndez-Vilas (Ed.), Current research, technology
and education topics in applied microbiology and microbial biotechnology.
Microbiology Series N 2, Vol. 2. Badajoz, Spain: Formatex.

Earthtec. (2007). Best practices in non-motorized transport planning, implementation

and maintenance. Report to the Government of Botswana and Gaberone City
Council, Gaberone.
Esiyok, B., Korkusuz, I., Canturk, G., Alkan, H. A., Karaman, A. G., & Hanci, I. H. (2005).
Road trafc accidents and disability: A cross-section study from Turkey.
Disability and Rehabilitation, 27, 1333e1338.
FAO. (2009). The Right to Food and the Impact of Biofuels (Agrofuels). Rome: FAO.
Gmez-Lobo, A. (2011). Affordability of public transport, a methodological clarication. Journal of Transport Economics and Policy (JTEP), 45(3).
Guasch, J. L., Laffont, J.-J., & Straub, S. (2008). Renegotiation of concession contracts
in Latin America: evidence from the water and transport sectors. International
Journal of Industrial Organization, 26(2), 421e442.
Gwilliam, K. M. (2003). Urban transport in developing countries. Transport Reviews,
23(2), 197e216.
Gwilliam, K. M. (2011). Africas transport infrastructure: Mainstreaming maintenance
and management. Washington DC: World Bank.
IEA (International Energy Agency). (2009). World energy outlook 2009. Paris: IEA.
Ingram, G. K., & Liu, Z. (2000). Determinants of motorization and road provision.
World Bank.
IPCC (International Panel on Climate Change). (2007). Fourth assessment report:
Climate change 2007.
Kagonova,O., & Peterson, G.E. (2010). Integrating Land Financing in Sub-National
Fiscal management. World Bank Policy Research Working Paper WPS 5409.
Kojima, M. (2010). Liquid biofuels: Background brief for the World Bank group energy
sector strategy. Washington DC: World Bank.
Kumar, A. J., & Barrett, F. (2008). Stuck in trafc: Urban transport in Africa. AICD
Background Paper No. 1. Washington DC: World Bank.
Litman, T. (2008). Transportation affordability, evaluation and improvement strategies.
Social Research in Transport (SORT) Clearinghouse Institute of Transport
Studies, Monash University.
Liu, Z., & Salzberg, A. (2012). China local government, municipal nance and land
use. In Low Carbon Development in China. World Bank.
Mehndiratta, S. (2011). Cycling and walking: Preserving a heritage, regaining lost
ground. World Bank.
Mehndiratta, S., Liu, Z., & Fang, K. (2011). Motorized vehicles: Demand management
and technology. World Bank.
Melillo, J. M., Reilly, J. M., Kicklighter, D. W., Gurgel, A. C., Cronin, T. W., Paltsev, S.,
et al. (2009). Indirect emissions from biofuels: how important? Science, 326,
1397e1399.
National Geographic. (December, 2011). National geographic magazine.
Peng, Z.-R. (2005). Urban transportation in Chinese cities and their impacts on the
urban poor. http://wwics.si.edu/index.cfm?topic_id1410&fuseactiontopics.
documents&group_id61692.
Peterson, G. E. (2007). Land leasing and land sale as an infrastructure nancing
option. In G. E. Peterson, & P. C. Annez (Eds.), Financing cities: Fiscal responsibility
and urban infrastructure in Brazil, China, India, Poland, and South Africa. London:
World Bank and Sage Publications.
Phalan, P. (2009). The social and environmental impacts of biofuels in Asia: an
overview. Applied Energy, 86, S21eS29.
REN21. (2011). Renewables 2011: Global status report. http://www.ren21.net/Portals/
97/documents/GSR/GSR2011_Master18.pdf.
Sood, P. (2010). Sub-sovereign nance for urban infrastructure. Manila: ADB.
South Africa Department of Transport. (2008). Draft national non-motorized transport policy. Pretoria.
Torres-Montoya, M., Yanan, L., Dubin, E., & Mehndiratta, S. (2010). Measuring
pedestrian accessibility: Comparing central business and commercial districts in
Beijing, London and New York city. Washington, DC: World Bank.
United Nations. (October 2009). Towards sustainable production and use of resources:
Assessing biofuels. Geneva: United Nations Environment Programme.
United Nations Population Fund. (2010). State of the world population. Geneva:
United Nations.
WBGU (Wissenschaftlicher Beirat Globale Umweltvernderungen). (2009). Future
bioenergy and sustainable land use. London: Earthscan. www.wbgu.de/wbgu_
jg2008_engl.html.
WHO. (2009). Global status report on road safety. Geneva: World Health
Organization.
World Bank. (2002). Cities on the move: A World Bank urban transport strategy
review. Washington, DC.
World Bank. (2006). Vietnam e Hanoi urban transport development project WB report
AB 2656 2006/11/22. Washington, DC.
World Bank. (2008). Biofuels: The promise and the risks in agriculture for development. World Development Report 2008. Washington, DC.
World Bank. (2010). World development report 2010: Development and climate
change. Washington, DC.
Zah, R., Gauch, M., Hischier, R., Lehmann, M., & Wger, P. (2007). Empa report. Life
cycle assessment of energy products: environmental assessment of biofuels.
Switzerland: Empa St Gallen.
Zhao, J. (2006). Whither the car? Chinas automobile industry and cleaner vehicle
technologies. Development and Change, 37(1), 121e144.