Identifying the wider benefits of transport investments

Roger VickermanCentre for European, Regional and Transport Economics

University of Kent, Canterbury, UK

XREAP Symposium, Barcelona, 26 November 2010

2

Introduction and motivation• Transport as a determinant of land use and

economic development (wider economic benefits -WEB) the subject of much controversy

• Formal appraisal techniques tend either– to exclude the possibility of wider economic impacts

because of the fear of double counting – or simply include an arbitrary add on

• Recent work has improved our understanding of the way in which accessibility– affects the performance of firms, – labour markets.

3

Introduction and motivation• However, the empirical evidence remains

problematic– endogeneity and causality questions– conflicts between macro-and micro-based estimates– the interrelationship and spillovers between different

areas• This has policy implications:

– underinvestment in transport infrastructure could lead to

• lower growth• congestion

– overinvestment could lead to • problems for public budgets• negative externalities associated with over expansion.

4

Transport and the local economy

• The multiple nature of transport– Transport as a derived demand– Transport as a substitutable input– Transport as an engine of growth

• Transport infrastructure and accessibility– External accessibility and the ‘two-way’ road– Internal accessibility and efficiency

• Accessibility, the cost of transport and economic efficiency

• User benefits and the wider economic benefits of transport

5

The agglomeration issue• ‘New Economic Geography’ provides the

necessary linkages– Transport costs as determinant of the price of an urban

location– And hence of the real wage – Thus going beyond the simple value of time savings as

a transport benefit• The theoretical basis of agglomeration

– Increasing returns, transport costs and market size – Linkages in the local economy– The role of real wages in cumulative causation– Labour market impacts

• Is agglomeration universal and inevitable?

6

Labour market impacts • Removal of barriers - integration of labour

markets• Agglomeration benefits in labour markets

– Changing participation rates– Increased working hours– Moves to more productive jobs - increased size of

commuting area has impacts on productivity and wage differentials

• Dangers of transferring parameters – Between modes– Between cities– Does analysis of effects on individual city regions

work for inter-city connections?

7

Costs/Benefits

Labour market sizeL0

B0

C0

α

β

γ

δW (= wage gap)

City size and agglomeration benefits

After Venables (2007)

No agglomeration economies: Wage gap unchanged at B0

LM size increases to L1’ Net Benefits = α+β

β′

γ′

Transport costs fall C0 to C1

Agglomeration economies: Wage gap increases to B1

LM size increases to L1

Net Benefits = α+(β+β′)+δ

B1

L1′ L1

C1

8

Looking for the evidence• Theoretical explanations and numerical simulations

demonstrate relevance, but useful application requires empirical evidence based on real data

• But such evidence is not straightforward and depends on:– The geographical scale of the empirical study– The unit of analysis– The ability to control for other factors which determine urban

development

• Look at three levels and types of study– Those which only look at macro aggregates– Those which examine the working of individual markets– Those which look in detail at the behavioural responses of

individual agents.

9

Evidence – macro studies• The Aschauer legacy

– The productivity of public infrastructure– Public infrastructure and the enhancement of the productivity of

private infrastructure– Debunking crowding out – Excessive optimism– Geographic scale

• Econometric problems– Causality– Spatial autocorrelation and spatial spillovers

• What should we measure?– Output– Employment– Productivity

10

Evidence – market studies• Exploring detail of agglomeration models• Competition effects

– Ambiguity– Pro-competitive effects from lower transport costs– Limited by existence of imperfect competition and rent seeking

• Agglomeration effects– Localisation economies– Urbanisation economies– Productivity effects – elasticities typically 0.01 to 0.4 for industry

but 0.2 or higher for services – Spatial scale variations

• Linkage effects – Labour markets

11

From Graham (2007)

12

Estimating elasticities• Agglomeration measured as ‘effective density’. Total effective

density (U) of employment that is accessible to any firm located in area i is

where Ei is total employment in area i, Ai is its area, Ej is total employment in area j, and dij is the distance between i and j.

• Effective density measure captures the scale and proximity of economic activity that is available in particular locations.

• Agglomeration economies are treated as a technology component that shifts a firm's production or cost function:Y = g(U)f(X)where Y is the output level of the firm, X is a vector of factor inputs, and g(U) is a vector of influences on production that arise from agglomeration economies.

• Firm level data used to estimate a translog production function

Estimated elasticities of productivity with respect to agglomeration, evidence for London

13From Graham (2007)

14

Evidence – micro studies• Why micro studies – changes in behaviour and

organisation– London Congestion Charge impacts

• Labour market effects– Accessibility and property prices– Jubilee Line impact (Gibbons and Machin)

• Increase in values +9.3% in areas with new stations• 1km reduction in access led to 1.5% increase in values

• Business organisation – HSR effects and internal restructuring – Concentration to access to network rather than along

network

CBA: the standard approach

15

C

C’

Q Q’

GeneralisedCost

Volume

D

CBA: the standard approach• But what are the assumptions lying behind this?

– Perfect competition so that p=mc– No externalities so that mc=smc– No returns to scale so mc constant– Demand is only responsive to a change in price, not a

change in supply (i.e. a fixed trip matrix)• Suppose we change these assumptions

– mc is upward sloping and smc>mc– But with increasing returns mc could slope down– p≠mc– And D could shift outwards in response to changing

opportunities– But suppose that agglomeration also caused mc to shift

downwards– Is the outcome now so determinate?

16

CBA and transport appraisal• The pure transport CBA:

– perfect competition in transport-using sectors– transport cost and benefits an acceptable approximation of final

costs and benefits • The best practice CBA:

– takes account of all indirect and direct responses by economic agents under conditions of perfect competition in the economy as a whole

– includes all dimensions of travel choice, repercussions on land use and economic activity, all externalities including environmental impacts

• The theoretically optimal CBA: – all direct and indirect responses by economic agents included – assumption of perfect competition relaxed (price ≠ msc)– imperfections or failures in both goods and factor markets: imperfect

competition in product markets, wages exceeding the opportunity cost of labour, taxation effects, external costs and benefits.

18

Two UK examples• Crossrail

– Urban rail project in London– Cost GBP16bn– Direct user benefits insufficient– But could have significant agglomeration benefits?

• HS2– High-speed rail line London-Birmingham (with

possible extensions northwards) – Cost GBP25.5bn– Direct user benefits estimated sufficient– But wider benefits add (although relatively smaller

than urban situation)– Have all the impacts been captured?

19

20

Source: Department for Transport (2005)

21

HS2 Proposed Route, Dec 2009

HS2 Benefits

22

23

24

Transport policy objectives• Is transport over-burdened?

– Transport issues subordinate to wider policy interests? – Over-reliance on transport to solve all a region’s problems is

unwise. – Where this policy involves significant capital expenditure it may

also be inefficient – lead to over-investment. – But if wider benefits of transport are significant then if appraisal

techniques ignore these leads to underinvestment.• Transport infrastructure a major part of any economy. • Transport has more complex impacts on output and

growth– If transport costs are reduced industries become more competitive – Improved transport contributes to productivity growth. – Changes in the location of activities– Employment growth

25

Implications for policy

• Simple rules are dangerous– Investment in transport can damage your health– Failure to invest in transport can damage it too

• Appraisal rules need to be comprehensive but transparent – Decisions have to be robust– But clearly understood by all stakeholders

• Levels of decision making– Spillovers– Policy refraction in multi-level governments– Jurisdictional competition and over- or under-

investment

26

Concluding remarks• Full circle on wider benefits

– From “transport is critical”– To “beware double counting”– To “wider benefits are the key”

• But beware all simple rules in transport appraisal• There remains much on the research agenda

– Imperfect competition and the productivity gains from transport

– Micro-behavioural evidence– Link versus network effects– Spillovers and jurisdictional competition– More ex post studies, does transport investment

really make the difference claimed?