Journal of Transport Geography 9 (2001) 209-218

www.elsevier.com/locate/jtrangeo

Transport investment and the promotion of economic

growth
David Banister a,*, Yossi Berechman b
a
University College London, Gower Street, London WCJE 6BT, UK
b
Tel Aviv University, Tel Aviv, Israel

Abstract
One of the major unresolved research issues in transport is the question as to whether transport infrastructure investment
promotes economic growth at the regional and local levels. The concern is not with the transport benefits, principally measured as
travel-time savings, but whether there are additional development benefits from these investments. If they do exist, can they be
measured? In this paper, we have developed a new approach based on defining the set of necessary conditions for economic de-
velopment to take place - in addition to the economic conditions, there are the investment conditions and the political and
insti- tutional conditions. It is argued that it is only when all three sets of necessary conditions are operating at the same time will
measurable and additional economic development benefits be found. A conceptual approach will be presented to encompass these
conditions. The paper will also address some of the key questions that have haunted researchers over the last twenty years on this
subject area. These fairlydetailed findings on the relationship between transport investment and economic development are
followed by a series of more generic conclusions relating to other key concerns of analysis. Included here will be a discussion of the
di- mensions of analysis, a new proposal for project appraisal, decoupling transport from economic growth complexity and
causality, accessibility and proximity, and the role of policy design. © 2001 Elsevier Science Ltd. All rights reserved.

Keywords: Economic development; Infrastructure investment; Additionality; Causality; Decoupling; Accessibility; Project appraisal

1. lntroduction measureability of these assumed benefits that we are

concerned about. Failure to properly identify and
The key question addressed in this paper is that measure these alleged development benefits is bound to
whether transport infrastructure investments, in well- result in double counting of benefits, thereby running
developed economies, promote economic growth, the risk of implementing the wrong projects. The terms
mainly at the urban and regional level. Given the com- economic development and economic growth cover the
plexity of the issues, this paper summarises the issues long-term increase in economic activity which can be
and presents some of the implications for policy-making attributed to the specific investment, and which can be
and for further research. Our intention is to be contro- shown to be an addition to the direct transport
versial and thought provoking, as this question is one of benefits and not plainly their capitalised value.
the key determinants of decisions on transport infra- The design of this paper is as follows. Section 2 dis-
structure investment, and one of the principal unre- cusses the necessary conditions that underlie the rela-
solved challenges to transport researchers (see tionship between transport and economic development
SACTRA, 1999; Banister and Berechman, 2000). and this is followed by a summary table (Section 3) of
Before proceeding, it is necessary to emphasise cer- some of the main findings and questions that still remain
tain key conceptual issues that underlie this paper. The unresolved. Section 4 describes our conceptual frame-
potential ability of transport infrastructure investments work relative to key variables, the level of analysis and
to produce transport benefits, such as travel time re- suggests how benefit-cost analysis, which is the most
ductions, is not questioned. The debate is over whether commonly used method to evaluate transport projects,
there are additional benefits from these investments, should be enhanced to account for possible economic
generically referred to as economic development bene- development benefits. We then raise the questions of
fits, and how to measure them. It is the additionalityand how to reduce travel volumes of freight and people in
the face of economic growth (this is the decoupling of
*
transport and economic development - Section 5). We
Corresponding author.
further examine the complex relationship between
0966-6923/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights
reserved. PII: S 0 9 6 6 - 6 9 2 3 ( 0 1 ) 0 0 0 1 3 - 8
210 iste e ec m o l of s o t eo

transport investment and economic development rela-

tive to causality and modelling (Section 6). Finally, we of the investment and its location, the network
question the notion of accessibility improvement from effects (e.g., are there missing links in the network),
infrastructure investment and discuss the critical role of and the actual timing of the investment. Transport
policy design in influencing the relationship between infrastruc- ture investment decisions are not made in
transport investment and economic development. isolation, so the nature of the investment, including
its place in the network, is also one of the necessary
conditions that need to be considered. These factors
2. Transport and economic development on their own are again not sufficient, and this
particular focus has been a limitation of much of the
There are two basic premises underlying the argu- analysis found in the literature, where the spatial
ment of this paper: factors have been ex- amined in isolation
1. In developed countries, where there is already a well- (Vickerman, 2001).
connected transport infrastructure network of a high • The third set constitutes olitic l f cto s, which are
quality, further investment in that infrastructure will related to the broader policy environment within
not on its own result in economic growth. which transport decisions must be taken. To achieve
2. Transport infrastructure investment acts as a economic development, complementary decisions
comple- ment to other more important underlying and a facilitating environment must be in place, oth-
conditions, which must also be met if further erwise the impacts may be counterproductive. In-
economic develop- ment is to take place. Additional cluded in this group of factors are the sources of
transport invest- ment is not a necessary condition, finance, the level of investment (local, regional or na-
but acts in a supporting role when other factors are tional), the supporting legal, organisational and insti-
at work. tutional policies and processes, and any necessar
The argument is further progressed through a series of complementary policy actions (e.g., grants, tax
necessary conditions that must be present to allow breaks and training programmes). Again, on its
economic development to take place: own, even a favourable political environment will
• The first necessary condition is the presence of under- not result in economic growth, unless the other
lying positive eco omic exte lities, such as agglom- neces- sary conditions are also not present.
These three basic sets of necessary conditions are illus-
eration and labour market economies, the
trated in Fig. 1. These three necessary conditions indi-
availability of a good quality (well trained and highly
vidually will have little or no impact on development.
skilled) la- bour force, and underlying dynamics in
Even if they are combined on a pair-wise basis, their
the local economy. This is a fundamental condition,
effect will be limited. For example, in Fig. 1, if only the
as it is only when these factors are positive and the
investment and political conditions are present (box
local economy is buoyant that new transport
2 + 3), we can expect accessibility changes, but since the
investment will, in conjunction with other necessary
necessary economic conditions are not present, eco-
conditions, have an economic development impact.
nomic growth impacts will not transpire. In that case,
• Secondly, there are i vestme t f cto s which relate to
the relative attractiveness of particular locations may
the availability of funds for the investment, the scale
change, but this is merely a redistribution of existing

Fig. 1. Illustration of the necessary sets of conditions.

 Banister, Bere an o rnal o Trans ort Geogra y J J 21
1
Table 1
Questions and answers

Questions Answers

1. Is the growth impact of This is not true and its analysis is central to current the transport debate. The main lesson is that the
any new transport facility research agenda should focus on (1) critical thresholds where changes in accessibilitywould be expected to have
in developed countries an impact and (2) to establish the conditions necessary for development. In particular this is the case for sparse
likely to be significant? networks; poorly connected networks; critical interchanges, bottlenecks and bypasses; or complementary
infrastructure (e.g., technological and communications networks).
2. Are transport costs This is true and transport costs are likely to fall as a proportion of total production costs, as dematerialisation
a declining part of and service-information-based employment increase. Even with substantial increases in transport costs, there are
total production and means by which these costs can be lessened, through regional production, technological substitution and
labour costs? increases in productivity on the production side, and changes in lifestyle, residential location and travel
behaviour on the labour side. But transport costs at the margin can be important.
3. Are buoyant local eco- This is true. Other factors are more important and include the structure and composition of the labour markets,
nomic conditions more the skills of the labour force, availability of sites, availability of grants or incentives for investments, availability
important than transport of entrepreneurial skills, quality of the environment, knowledge and information. Transport infrastructure
infrastructure improve- investments can enhance these factors but cannot substitute for them.
ments in generating
growth?

4. Are the unique charac-

teristics of the area and Indeed they are. The modelling approaches examined in the review mainly focus on the conditions for change,
the spatial extent over but the empirical analysis needs control areas. All of the time-based analyses compare areas with and without
which the growth impacts the investment, but change often takes place outside the studied region. Moreover, mainly for political
to be felt important? reasons, economic development that occurs outside an area is irrelevant for local decision-making relative
to the transport and other complementary investments.
5. Can one generalise
about the results from Cautiously yes. It is mainlyat the micro-level that the true links must be established since this is the level at
specific case studies? which investments take place. At this level local conditions are quite crucial for development. Yet,
complementarity between studies at different scales and locations (countries) is important to build up a
database of case studies and models to allow general conclusions to be drawn from the local to the macro-
6. What is the role of level.
technological changes in
affecting the relationships The implications of technological change for transport investment impact on economic development are unclear.
between transport invest- They are likely to lead to greater complexity in the relationships, and they affect the means by which firms and
ment and economic individuals have reorganised themselves to reduce uncertainty, to increase productivity, and to improve their
growth? flexibility. This is related to the possibilities of reducing the amount of travel without impairing economic
growth.
7. How slow, long term
and complex are the ad-
justment processes within The absence of appropriate data and methods make assessment difficult. The treatment of time has presented
the regional economy fol- manyproblems, both in terms of the causalityarguments and the time over which the impacts are expected. Few
lowing a transport invest- transport investments open in their entirety at one point in time, and changes such as firms' location and labour
ment? market adjustments often take place before, during or after the investment.

8. Does a good transport

infrastructure raise the
image and the This seems to be the case, as the image of the area is important for developers to invest, but often the public
perceptions of an area, sector cannot recoup these benefits (e.g., through value capture). For urban rail systems, this image building is
thereby at- tracting key to the revitalisation of central areas and in maintaining strong city centres.
additional private
investment?

9. How can one assess the

course of economic de- The historical literature provides some retrospective lessons on this question but it is quite hard, if not
velopment in an area if impossible, to establish what would be the course of development without the transport investment as changes
the particular transport are being measured against a moveable situation. The counterfactual situation in the identification of
investment was not made? development effects is intriguing, but difficult to operationalise either analytical or empirically.

10. What is the role of

expectations, regarding It is not really known how to correctly assess expectations over the period of time that the impacts of the project
the impacts of the will be felt. However, expectations can significantly affect the consequences from a particular transport
invest- ment, in investment with respect to growth. This is rather important, as expectations are likely to influence patterns of
accomplishing growth? production, consumption and private investments before and during the infrastructure project.

Based on Banister and Berechman (2000, Table 12.1).

 2 Banister, Bere an o rnal o Trans ort Geogra y J
economic development rather than additional growth.
Similarly, if only the investment and economic condi- assess the investment and its impacts, and the time
tions are present (box 1 + 2), economic development duration of the effects (Banister and Berechman, 2000).
effects from the investment may not follow for the lack Table 2 shows the relationships between the types of
of supportive policies, or because of the presence of measured impacts and the level at which theyare likelyto
conflicting transport and land-use policies. It is only be felt. Although the national and regional levels of
when all three necessary sets of conditions are present impacts have been covered, our main focus is directed at
and operating together that economic growth will ensue. the local impacts of transport infrastructure investments.
The transport infrastructure investments are location Here, some examples of the main dimensions are given as
specific and have potential growth effects on local illustrations of the key variables and the necessary
economies. Hence, to actually identify and measure the conditions for economic development to take place.
economic growth resulting from such investment, anal-
ysis must take place at the local level. It is at this scale 4 J National level
that the impacts on local economic development, in-
come levels, accessibility and employment should be It is clear that all countries need a well-developed
assessed. On the practical level, as analysis becomes transport infrastructure to compete internationally in
more aggregate, many of the impacts are lost''. Sec- new global markets. As trade barriers are reduced and
ondly, if the economic growth effects are to be measured, as new markets are opened up, it is essential to have
then analysis has to move away from the primary con- high levels of accessibility. However, it is not only the
cern with user benefits (conventionally travel-time sav- quantity of the physical infrastructure that is important,
ings) to a much wider assessment of costs and benefits. but also the q ality o t e in rastr t re in a much wider
context (the political and institutional conditions in Fig.
1). The physical infrastructure needs to be extended, not
3. Key questions and answers just to the links in the network, but also the terminals
and interchanges. These nodes are often the points of
In an extensive review of the literature of transport congestion, which cause most delay. The control systems
and economic development, Banister and Berechman used on the network are crucial to ensure that maximum
(2000) developed ten key questions, and these are sum- efficien cyis gained from its use and decisions are based
marised here, together with their answers (Table 1). on high-quality real-time information. The manage-
Definitive answers have not been obtained for all (or ment, information and control systems are also impor-
any) of the questions, but the main debating points have tant, in many cases even more important than the
been identified. Inevitably, as progress is made in one physical infrastructure itself. In addition, the financial
direction, new directions emerge so that we may be left and organisational context within which the infrastruc-
with even more questions than answers. ture is placed can again influence decisions on con-
struction, on maintenance, on ownership, on charging
and on responsibility.
4. Dimensions of analysis

The methodological framework, within which this 4 Regional level

study has been structured, consists of three dimensions:
the scale of the analysis, the type of variables used to Here the concept of accessibility is central to the de-
bate. It has been argued that changes in accessibility

Table 2
Conceptual framework of analysis

Variables Scale of analysis

National Regional Local

Economic Productivity growth; social rate of Accessibility changes; changes in Employment level; time allocation;
development return the location of retail, housing labour productivity
measures and employment activities
Output measures of Annual GDP growth Spatial relocation; competitive Growth in jobs; welfare improve-
models advantage; industrial clustering ment; agglomeration economies
Scale and type of Total state capital; infrastructure High-speed regional rail and road Metropolitan roads and facilities;
transport investment stock networks; regional terminals; new rail links; underground rail;
airports; seaports, waterways; transport centres
pipelines
 Banister, Bere an o rnal o Trans ort Geogra y J J 21
3
resulting from transport infrastructure investment cau-
ses a redistribution of employment between regions. It is project, each new proposal needs to be analysed sepa-
unclear whether the changes in accessibility also create rately, so that the general rate of return (or rate of
new activities, which we have defined as economic GDP growth) from total public capital formation may at
growth. The conclusions reached here suggest that at the best serve as a kind of a guideline. A further problem
regional level redistribution will take place, often to the is that all planned infrastructure investment project
further advantage of the already accessible core parts of analyses are done ex ante, with little certain
the country. But, transport accessibility must be seen as information about future trends and development.
part of a much wider concept of accessibility that in- Hence specific issues that may affect the potential
cludes availability of skilled labour, good-quality loca- impact of this investment with economic growth need
tions, the necessary supporting infrastructure, and local to be considered.
road- and rail networks (Fig. 2). Two key factors need to be addressed. The first is the
In Fig. 2, movement along the accessibility axis will network onsideration, implying that each transport
not achieve economic growth on its own. It is only project should be considered within the framework of a
where there are open dynamic systems (the economic local, regional or even national network. For growth
and political sets of conditions are present, see Fig. 1) to occur, it must be the result of improvements at the
that it has a real impact, particularly where the existing network level and not at the single project level. The
infrastructure provides only poor levels of accessibility effect of a given investment on growth can be dramati-
(Quadrant 1). In dynamic systems with high levels of cally different when it links, for example, two disjoint
accessibility it will support growth (Quadrant 2), but in networks than being an additional link in an established
the lower part of the figure (Quadrants 3 and 4) im- network. The second factor is rioritising obje tives and
proving transport accessibility is not a sufficient condi- riteria. The majority of benefits need to be transport
tion for economic growth (for the lack of the economic, related, since otherwise why invest in transport facilities
investment and political sets of conditions). Movement in the first place. Double counting must be avoided in
has to first take place along the vertical axis to shift the measuring non-transport benefits. There needs to be a
political and economic factors from the closed static to functional linkage between primary transport benefits
the open dynamic systems. (e.g., accessibility improvements) and potential eco-
nomic growth effects.
4 .3 Lo al level If a transport investment is to take place, a twin ap-
proach to assessment is proposed, where conventional
A major criticism of the national level analysis is that cost benefit analysis is carried out on the project to
finding a rate of GDP growth from a certain increase de- termine the user benefits and costs of investment
of total capital accumulation actually tells us little about (Ta- ble 3). First, to achieve a given rate of return, this
the contribution of the next infrastructure project. analysis may account for some or all the necessary
That is, some projects mayhave a high and positive returns. If there is a shortfall, then a complementary
impact on GDP growth whereas others may not. All analysis needs to take place that takes a wider view of
that the na- tional level analysis can say is about the the investment proposal. This would include the
average con- tribution of capital accumulation. But contribution of the project to the transport network as a
since infrastructure investments are carried out whole through network analysis. Secondly, in addition
project by the value added of the project would be assessed
through its contribution to local employment, the
Table 3
A suggested twin approach to project appraisal

Project (common element: phasing and timing of investment)

Traditional cost benefit analysis

User transport benefits
Costs of investment
Complementary economic development analysis
Network effects - accessibility analysis
Value added from project
Changes in employment
Changes in factor productivity
Changes in environmental quality
Priority
Spatial impacts in terms of locations
Social impact in terms of
distribution
Fig. 2. Transport and economic development at the regional level. Note: The value added relates to the allocative externalities (Fig. 3).
 2 Banister, Bere an o rnal o Trans ort Geogra y J
potential for increases in productivity, and the
environmental impacts. Finally, it would also investigate Reduced use of physical materials (sometimes called
the distributional impacts in terms of the spatial effects de aterialisation) can make a more fundamental im-
on the regional and local distribution of services and pact. As production becomes more service-based, lighter
facilities, and the social impacts. products are moved about, and miniaturisation takes
This complementary analysis would use some of the place with a much greater emphasis on quality and de-
methods and models proposed here to measure the sign. The net effect is that less volume is moved around
contribution that the components outlined in Table 3 so less travel is required. It has been estimated that
would make to the overall project assessment. If addi- de- materialisation could result in a 15-20% reduction
tional benefits can be shown and are sufficient to raise in freight volumes between 1995 and 2020
the benefit-cost analysis above the crucial level for the (Schleicher- Tappeser et al., 1998). But as the material
rates of return, then investment could take place. This intensity of products declines, material consumption
more complex type of analysis seems to be increasingly may still in- crease as the economy is growing and
important, as the conventional benefits from any pro- demand is also expanding. Thus, this reduction due to
posed transport investment may be providing an ever- dematerialisation must be placed against the expected
decreasing proportion of the total returns (Banister and increase in freight traffic of 80% over this same
Berechman, 2000). period.
The second major change in organisation could be
the move from the global to the glo al. 1 Traditional
5. Decoupling transport from economic growth arguments strongly favour concentration of production
to take advantage of agglomeration economies. How-
ever, the development of flexible specialisation (Piore
Historically, there has been a close correlation be-
tween the growth in demand for freight and passenger and Sabel, 1984) has allowed a new complementarity
traffic and economic growth, as measured by GDP. The between global networks and regional production for
question raised in Section 1 (and Table 1) related to regional markets without the multinationals losing
the underlying rationale for this statistical relationship control. The multinationals still have the knowledge and
and whether that relationship should (or would) control of information, but produce for local markets in
continue into the future. There are also serious a dispersed manner through outsourcing. Production
questions re- garding the causality mechanism that units are downsized so that the new lean production
underlies this re- lationship, (e.g., does transport methods can be introduced. For example, in the state of
investment promote economic growth or does growth Baden-Wurttemberg, in south Germany, most suppliers
encourage more de- mand for transport, and thus for the car manufacturer Mercedes are based in and
further investment). In general, it seems that there is no around Stuttgart. Certain components are ordered to
reason why we should have transport growth in line the hour and must be available within a radius of 100
with economic growth (Stead, 2001). Indeed, there are km (Schleicher-Tappeser et al., 1998). Local
strong efficiency and environmental arguments for production networks provide the opportunity for short
breaking that link, and policy should be seeking to travel dis- tances and a reduction in freight traffic. This
reduce the transport inten- sity of activities whilst at same study estimated that the potential reduction in
the same time maintaining economic growth - this freight traffic through the use of regional markets,
is the de o ling arg ent. regional produc- tion, and the reduction in
Travel can be broken down into three components international flow of goods (that have been produced
- volume, distance and efficiency. The first two are for local markets) could amount to 20-30% (1995-
usually combined to give measures of performance 2020).
(i.e., passenger kilometres or tonne kilometres), but There is a substantial potential for decoupling in the
the third element is equally important as it relates to freight sector, with the new forms of production out-
modes, travel time and price, the use of resources, lined above. The scale economy arguments, together
technology and organisational factors. For example, in with economies related to specialisation and the com-
the freight sector efficiency can be increased through parative cost advantages of producing large quantities
the use of logistics, flat organisational structures, new for large markets, are now being questioned. Customer-
forms of handling, minimisation of warehouse re- driven requirements mean that products are now tai-
quirements, and spatial organisation to reduce distri- lored to individual specifications so that smaller-scale
bution costs. Given demand, all these measures can production units for regional markets become possible,
increase efficiency, while red ing the volumes and provided that the knowledge and skills are available.
distances travelled. However, efficiency can also pro-
mote increased demand so that these factors can also
result in increased efficiency and the volumes/distances 1
Glocal refers to a combination of global and local, where
travelled. production is still controlled by the large multinational companies,
but produced locally for local markets under franchising and other
arrangements.
 Banister, Bere an o rnal o Trans ort Geogra y J J 21
5
However, there is still a long way to go as global
cultures and internationalisation have been instrumental 6. Complexity and causality
in producing similar values, tastes and lifestyles, with
the consequent loss of community and locality. There may be a Catch 22 situation in the analysis of
Similarly, most of the changes in policy have tended to the links between transport infrastructure investment
encourage greater internationalisation through trade and economic development. To make analysis tractable
liberalisation, market-based strategies, subsidies to requires a series of simplifying assumptions to be made,
farmers and other groups, deregulation in transport, but this in turn reduced the usefulness of the analysis.
privatisation, and even markets for environmental and Even if more complex formulations were possible, then
consumer protec- tion. The most optimistic view is the limitations of the available data are quickly reached.
that in the freight sector, transport demand could This situation results in uncertainty and prevents clear
remain constant over the next 20 years (up to 2020), conclusions. This complexity is increasing, particularly
with the strong imple- mentation of decoupling when placed against new funding priorities and mecha-
strategies. nisms, the changing industrial base, the balancing of
In the assenger se tor, the opportunities for stabili- economic priorities with social and environmental con-
sation in demand seem harder to envisage, particularly cerns, etc. However, we must accommodate complexity
within the context of increased affluence and leisure in analysis and recognise the multi-dimensional nature
time (Banister et al., 2000). Decoupling must again be of the links between transport, location, development
seen as a combination of strategies to reduce the and the many other new factors relevant to our under-
volume of traffic, the distance travelled and measures to standing of these processes (Fig. 3).
increase efficiency, but at the same time maintaining The main idea underlying the relationship shown in
economic growth. The decoupling arguments follow this figure is that for transport-induced economic
the same structure as in the freight sector with growth to transpire, it is necessary that various econo-
dematerialisation of travel through less travel or travel mies be present in various markets. The principal ones
by more efficient modes, and through establishing local are firms' agglomeration, transport network, labour
travel patterns through the reorganisation of the market, land market and environmental quality en-
production and con- sumption patterns based on local hancement. Merely improving accessibility, which
and regional networks. Even if there has been a link in translates into reduced travel time and higher travel
the past between transport use and economic growth, volumes, would not be sufficient to generate growth.
there is no reason why this link should continue. This In a micro-economic analysis, Banister and Berech-
argument has both a strong economic and a strong man (2000) tried to include some of these factors and
sustainability rationale.

Fig. 3. Proposed scheme for the evaluation of economic growth benefits from a transport infrastructure investment.
 2 Banister, Bere an o rnal o Trans ort Geogra y J

Fig. 5. Equilibrium labour in the economy (l*) as a function of

transport infrastructure capacity (K).

lead to substantial increases in new employment, as

most of the potential savings will be realised through
increases in productivity with the existing labour force.
Changes in accessibility may induce relocation, but only
if those changes are above key thresholds and other
factors are working strongly in the same direction, these
relocations may further induce economic growth. In the
Fig. 4. Illustration of possible relationship between the production, micro-economic model, Banister and Berechman (2000)
household and transport sectors. have explored the impacts of infrastructure investment
on the choices that individuals make in how they choose
to use their time. Whether additional time (resulting
demonstrate the underlying dynamics of the from reduced travel times following an infrastructure
relationship (Fig. 4). But even here, the model
investment) is taken as leisure or additional work, is a
concentrates on work- related activities, single workers
major question. The proliferation of leisure-time activ-
households, standard notions of values of time, and only
ities in well-developed economies may mean that addi-
two firms. Perhaps the complexity issue needs to be
tional time may not be used to increase labour supply. If
approached through a combination of methods that link
that is the case, the causality linkage between accessi-
together more formal modelling approaches with
bility improvements and economic growth may be ra-
descriptive qualitative anal- ysis that can make links
ther weak. The analysis shows that additional growth,
between methods. Such an approach would begin to
mainly in terms of additional employment, from addi-
tackle the complexity in a robust and flexible manner
tional infrastructure is likely to abate quite rapidly, as
so that a clear picture can be achieved. 2
the network becomes fully developed (Fig. 5).
Complexity relates to one of the underlying problems The main conclusion reached here is that the cau-
investigated in this study, namely causality. One of the sality argument is weakening, as there seem to be de-
key elements in the basic economic argument is that creasing returns on public physical capital. Although
there must be a set of causal relations between transport this issue has not been explored here in detail, the ar-
investment and economic development, since otherwise, guments for increasing returns on knowledge and tech-
there would be no basis for the claims that transport nology may be seen as the new engines of economic
infrastructure investments also promote economic growth rather than the returns on transport
growth. Public infrastructure investment is regarded as infrastruc- ture (Romer, 1986).
the trigger mechanism to the increasing of private cap-
ital rates of return through increases in private capital
stock and labour productivity, which in turn results in
higher total output and economic growth. 7. Accessibility and proximity
Given this view, Banister and Berechman (2000) have
argued that this causality argument is one of many that As suggested by Figs. 3 and 4, new additions to the
can be used and that the empirical evidence is weak. infrastructure stock enhance accessibility, which pro-
It seems doubtful that public infrastructure investment vides the engine behind the potential increase in eco-
will nomic activity, mainly realised through employment
increases and factor productivity. However, the empir-
ical evidence is mixed on the role that accessibility
2
Further details of this approach, together with a formal changes have in generating economic development (see
represen- tation of the model, can be found in Banister and also SACTRA, 1999). It should be remembered that
Berechman (2000, Chapter 7).
 Banister, Bere an o rnal o Trans ort Geogra y J J 21
7
accessibility is a relative concept and infrastructure in-
vestment in one location may help that location, but at networks and accessibility are required to determine
the expense of a competing location. Within a metro- under what conditions the competitive position of one
politan area the net effect may be marginal (see Rietveld network will be changed as compared with another.
and Bruinsma, 1998). More important though is the This is a question of value added to influence
argument that in most advanced economies, levels of expectations, to facilitate co-ordination and to ensure
accessibility b y road (and rail) are already high. This compatibility.
means that most infrastructure investments will only Proximity is closely related to accessibility, but may
affect the accessibility in the system as a whole in a also be important in the context of development. There
marginal way. It is only where there is a major change in are two conflicting processes at work, one of which leads
the system-wide accessibility (e.g., a new link joining towards proximity through agglomeration factors and
two previously disjoint networks, or the opening up the other which suggests that in a well-connected soci-
of a previously inaccessible location) that major ety, non-proximity or diseconomies of agglomeration
relocation might take place. This suggests that may operate. Many services can (and are) provided re-
accessibility should not be investigated only as a relative motely. For example, in a telephone inquiry service, the
concept, but that we should be searching for minimum person actually answering your question need not be in
thresholds above which changes are likely to take place. close proximity to you or to others. Similarly, many
These thresholds are important if accessibility changes products can be outsourced to locations where labour
are seen as the means to improve the relative positions and other production factors are cheaper. Provided that
of regions, by increasing inward investment and the charge to the user is low (i.e., a local rate telephone
employment. call or cheap products), proximity is not important.
A second more subtle implication of changes in ac- High-quality communications and transport infrastruc-
cessibility is that it enhances existing trends rather than ture facilities have allowed a greater flexibility in the
creating new ones. If the conditions are advantageous location of many services and firms.
for firms to relocate or set up a new business in some Economies of scale and of agglomeration play an im-
areas (for example, where the labour force has appro- portant role in this process, and locations that possess
priate skills or where there are financial incentives), then specialisation tendencies may benefit more from a re-
improvements in the transport infrastructure may give duction in transport costs than other locations. Krugman
one location preference over another. On its own, (1991) shows an interesting example of the impact of
transport infrastructure is a second-order location a reduction in transport costs from infrastructure im-
variable where there is a well-developed network, but in provements, on manufacturing location in the presence
conjunction with other factors it may tip the balance'' of scale economies in production. With high transport
in favour for the (marginally) more accessible location costs, production will take place in dispersed locations
(see Fig. 2). (e.g., core and periphery). When transport costs fall
This debate relates to the complementarity found there will be a shift towards production in one location
within networks. Accessibility has tended to be viewed (either the core or the periphery), which can switch if
as the impact of one new link on the network as a whole. transport costs keep falling (see also Anas et al., 1998,
But many investments are complementary, so competi- for the incorpo- ration of similar ideas in the context
tion is really taking place between systems, not indi- of urban spatial structure). However, all these elegant
vidual links. Hence, accessibility should not only be and quite persua- sive ideas are largely based on
viewed as the changes taking place in one system (e.g., differences in transport costs, independent of network
rail), but the new competitive position of that system in structure and the organi- sation and efficiency of
relation to other systems (e.g., road). There is a strong transport delivery systems.
optionality value in accessibility to a particular system, In the situation where transport costs are decreasing
even if no use is actually made of it. 3 New concepts of (or are already low) and where there is a ubiquitous in-
frastructure, non-material flows may become more im-
portant (Burmeister and Colletis-Wahl, 1997). The
traditional framework, which treats transport as a cost,
3
We should note here the concept of positive network is becoming less relevant when trying to understand the
externalities developed mainly for communication and computer spatial dynamics of industry. Recent research proposes a
networks. It implies that the value of membership to one user is
richer, more complex set of relationships that are based
positively affected when another user joins and enlarges the network
(Katz and Schapiro, 1994). Hence the rationale for having more users on circulatory capacity. This concept includes the
joining and using the network. In transport this argument may be infra- structure, and the organisation and management
valid in developing economies where new transport investment open of flows of goods, information and people (i.e., the
up undeveloped areas. In the context of this paper which focuses on services that can be produced). The links between
transport networks in highly developed economies, this argument may
be invalid as there are strong congestion externalities present when
infrastructure and the utilisation of the production
another user joins the network. process are not deter- ministic. The role of transport
infrastructure in a network economy is no more than a
generic resource for circula- tion (Burmeister and
Colletis-Wahl, 1997, p. 239). Par-
 2 Banister, Bere an o rnal o Trans ort Geogra y J
ticular forms of use and the consequent impacts depend
on the strategies adopted by the actors in the production References
process. Proximity extends far beyond geographical ac-
Anas, A., Arnott, R., Small, K.A., 1998. Urban spatial structure.
cessibility to include the organisation and the degree of
J. Econ. Lit. 36 (September), 1426-1464.
control of the flows of goods, information and people. Banister, D., Berechman, J., 2000. Transport Investment and Eco-
Finally, policy design also has a crucial role in in- nomic Development. UCL Press, London.
fluencing and strengthening the potential impact of Banister, D., Stead, D., Steen, P., Aºkerman, J., Dreborg, K.,
transport infrastructure investment on local economic Nijkamp, P., Schleicher-Tappeser, R., 2000. European Transport
development. In Fig. 1 we have highlighted the notion Policy and Sustainable Mobility. Spon, London.
that for economic development to follow from transport Burmeister, A., Colletis-Wahl, K., 1997. Proximity in production
networks: the circulatory dimension. Eur. Urban Reg. Stud. 4 (3),
infrastructure investments, it is essential that three sets 231-241.
of necessary conditions be met. This representation, Katz, M.L., Schapiro, C., 1994. Systems competition and network
however, conveys the impression that all of these sets effects. J. Econ. Perspect. 23 (2), 177-200.
are equally important in affecting the economic develop- Krugman, P., 1991. Increasing returns and economic geography.
ment outcomes of infrastructure investments. Reality J. Political Econ. 99 (3), 483-499.
may be more complex as political and policy decision- Piore, M.J., Sabel, C., 1984. The Second Industrial Decline. Basic
Books, New York.
making actually affect (directly or indirectly) the other
Rietveld, P., Bruinsma, F., 1998. Is Transport Infrastructure Effective?
two sets of necessary conditions, namely the investment Transport Infrastructure and Accessibility: Impacts on the Space
type and the economic conditions. Economy. Springer, Berlin.
Policy-making is the crucial factor in realising eco- Romer, P.M., 1986. Increasing returns and long-run growth. J. Political
nomic growth benefits from a transport infrastructure Econ. 94 (5), 1002-1038.
investment. Although economic conditions are impor- Standing Advisory Committee on Trunk Road Appraisal (SACTRA),
tant, they are also embedded to their political context. 1999. Transport and the Economy. The Stationery Office, London.
In democracies, the ability of governments to co-ordi- Schleicher-Tappeser, R., Hey, C., Steen, P., 1998. Policy approaches
for decoupling freight transport from economic growth. Paper
nate their activities and design complementary policies
presented at the World Conference on Transport Research,
to gain maximum economic development effects from Antwerp, July.
their capital investments seems on many occasions to Stead, D., 2001. Transport intensity in Europe - indicators and trends.
have been rather limited. However, perhaps paradoxi- J. Transp. Policy 8 (1), 29-46.
cally, it is only in democratic societies that economic Vickerman, R., 2001. Transport and economic development. Paper
development reaches its maximum potential. to be presented at the 119th Round Table of the European
Confer- ence of Ministers of Transport, Paris, March.