Scholarly article on topic 'CBA § Darwin: The Case of Transport Infrastructure in France'

CBA § Darwin: The Case of Transport Infrastructure in France Academic research paper on "Social and economic geography"

CC BY-NC-ND
0
0
Share paper
OECD Field of science
Keywords
{"Cost-benefit analysis" / evolution / Darwin / "project assessment" / "sustainable development" / "transport infrastructure"}

Abstract of research paper on Social and economic geography, author of scientific article — Emile Quinet, David Meunier

Abstract CBA has a long history in the field of transport projects. Still, its methodologies have been developed in a rather simple context which is now much more complex: is CBA still adapted to this context? CBA is now asked by decision makers not only to represent monetary flows and time savings, but also economic effects, social and distributive concerns, land use, climate change, and wider effects. It has to be implemented in a framework where private sector is involved and numerous entities are involved in project financing. Another strong evolution is that CBA has to interact more with public communication processes, and to take into account new types of governance. Adopting a Darwinian point of view, we consider the evolutions of CBA's environment, its resources and past adaptations. Then we pay attention to the potential competitors of CBA in this new environment. This analytical part is illustrated using several recent examples of new assessment questions and answers considered in France, at different scales: a great transport infrastructure scheme for the Parisian region, and the French national transportation infrastructure plan. They present more concretely how the new evaluative questions take form, and the kind of answers that have been found (or still have to be found). Building partly on the on-going evolution of the evaluation system in France, some ideas on the axes of adaptation of CBA will be presented in the last part of the paper, covering not only CBA's technical performance but also CBA's governance and integration in a more comprehensive assessment system.

Academic research paper on topic "CBA § Darwin: The Case of Transport Infrastructure in France"

Available online at www.sciencedirect.com

SciVerse ScienceDirect

Procedia - Social and Behavioral Sciences 48 (2012) 3051 - 3064

Transport Research Arena- Europe 2012

CBA § Darwin: the case oftransport infrastructure in France

Emile Quineta David Meunier

aEcole des Ponts ParisTech, Paris School of Economies bUniversité Paris-Est, LVMT, UMR T9403 Ecole des Ponts ParisTech INRETS UPEMLV

Abstract

CBA has a long history in the field of transport projects. Still, its methodologies have been developed in a rather simple context which is now much more complex: is CBA still adapted to this context?

CBA is now asked by decision makers not only to represent monetary flows and time savings, but also economic effects, social and distributive concerns, land use, climate change, and wider effects. It has to be implemented in a framework where private sector is involved and numerous entities are involved in project financing. Another strong evolution is that CBA has to interact more with public communication processes, and to take into account new types of governance.

Adopting a Darwinian point of view, we consider the evolutions of CBA's environment, its resources and past adaptations. Then we pay attention to the potential competitors of CBA in this new environment. This analytical part is illustrated using several recent examples of new assessment questions and answers considered in France, at different scales: a great transport infrastructure scheme for the Parisian region, and the French national transportation infrastructure plan. They present more concretely how the new evaluative questions take form, and the kind of answers that have been found (or still have to be found).

Building partly on the on-going evolution of the evaluation system in France, some ideas on the axes of adaptation of CBA will be presented in the last part of the paper, covering not only CBA's technical performance but also CBA's governance and integration in a more comprehensive assessment system.

©2012 Published by Elsevier Ltd. Selection and/or peer review under responsibility of the Programme Committee of the Transport Research Arena 2012

Cost-benefit analysis; evolution; Darwin; project assessment; sustainable development; transport infrastructure

1 Corresponding author. E-mail address: david.meunier@enpc.fr

1877-0428 © 2012 Published by Elsevier Ltd. Selection and/or peer review under responsibility of the Programme Committee of the Transport Research Arena 2012

doi:10.1016/j.sbspro.2012.06.1272

1. Introduction

Cost-benefit analysis (CBA) has a long history in the field of transport projects, in France and in many other countries. It is widely used and considered as a very useful tool.

But CBA methodologies have been developed in a rather simple context which is now much more complex: is CBA still adapted to this context?

This paper intends thus to adopt a Darwinian point of view, considering the evolutions of CBA's environment, analysing CBA's resources and past adaptations. This analysis is developed in the next section on the case of CBA in France. In section 3, we pay attention to the potential competitors of CBA in this new environment and develop the idea that instead of competitors the diverse types of assessment tools may be complements.

Section 4 illustrates and develops this analytical part using several recent examples of the new assessment questions and answers considered in France, at different scales: a great transport infrastructure scheme for the Parisian region, and the French national transportation infrastructure plan. They present more concretely how the new evaluative questions take form, and the kind of answers that have been found (or still have to be found).

Building partly on the on-going evolution of the evaluation system in France, some ideas on the axes of adaptation of CBA are presented in the conclusive section, covering not only CBA's technical performance but also CBA's governance and integration in a more comprehensive assessment system.

2. Evolution of CBA in France

The history of CBA in France gives a striking example of Darwinian adaptative transformations in response to changes occurring in the environment. We will first describe these transformations, that were punctuated in the last decades by official methodological reference guides (called « circulates », the latest ones being Ministry for Transport Infrastructure 1998, 2004, 2005). These circulates, approximately every 5 to 10 years, were issued for updating methodological principles and reference values, and they introduced inflexions in the general process of CBA implementation.

We will see that, most of the time, these transformations resulted from the upsurge of new practical questions (such as urban congestion) or of new concerns (for instance environmental concerns). We will also see that the answers given in response were often engineering approaches, such as management of urban traffic flows, or legal tools as for environment. It sometimes needed many years before new economic methodologies appeared that allowed CBA to take into account the new situations and concerns.

2.1. The origins

The theoretical heart of CBA, Dupuit's surplus, was put to light more than 150 years ago using the case of a transport investment (Dupuit, 1849), but it remained without any noticeable real implementation during one century. The fifties saw the move from theory to practice, when it came clear that the public budget devoted to the improvement of interurban road infrastructure was far too low in front of the growing needs linked to the rapid development of car traffic, and that the best use had to be made of this budget.

In this context, Dupuit's purely theoretical recommendations had to experience a first evolution in order to be adapted to the specific issues addressed. Indeed, the main impacts of road investments did not translate into a decrease in transport prices, but into improved quality of service, under the shape of time and security benefits. The invention of the notions of "value of time" and "value of life" then allowed to issue in I960 a reference guide indicating how to compute benefits and costs, with official values to be used for implementing these two notions in the calculations.

2.2. Evolution observed up to now

In the mid seventies, environmental concerns appeared and developed, at first through the rejection of motorway projects that created deep cuts into important forests or damaged valuable natural sites2. Other environmental concerns such as acid rains linked to air pollution caused by the traffic led to project modifications or even project disappearance. The first response came from the engineering field (technical project modification) and for many years no economic response was given, I.e. CBA did not get modified.

About at the same time, the need appeared to improve traffic conditions in urban areas and to fight against increasing congestion. Evaluation in urban areas stumbles on new problems. First, environmental problems are more acute (at least those related to human environment such as noise emission). Second, managing urban traffic means managing flows on a meshed network, a novel and more complex problem than the archetypal interurban problem of flow affectation on parallel links. New types of models had to be developed, able to represent more numerous possibilities of substitute or complementary network links, for which surplus computation became more delicate. Last, improvements in urban transport have impacts on spatial organization and on urbanization that are more visible and clearer than similar interurban investments. The methods for helping in decision-making with these new issues took time to be produced. At first, they consisted in engineering methods for traffic modelling (the well-known four step model, and French "inventions" of the so-called Abraham model and the "price - time" model, or the affectation software Davis, ancestor of many present modelling tools). Then CBA was progressively enriched, introducing for instance comfort parameters valuation in public transport (P. Merlin, 1991 estimated monetary values for comfort in the metro, measured by the density of occupation of metro cars). But it appeared soon that the impacts on spatial organization were not correctly taken into account by conventional CBA and by generalized cost computations. Therefore, methods for assisting decisionmaking did less and less rely on only one criterion, the variation of social surplus, and moved progressively towards multi-criteria methods.

This movement was accentuated by changes in the decision-making process, which became more and more collective, associating an increasing number of stakeholders. At first, in the sixties and seventies, investment decisions were taken within the Ministry in charge of transports, with a major role of relevant modal directorates. External interventions were few and of low influence. Progressively, local public authorities gained importance with decentralization reforms in France. European authorities also acquired intervention powers, notably due to the subventions they distributed. Last, citizens took an increasing share in the decision-making process, through non-profit organizations, especially environmental associations.

The decision process relative to big transport infrastructures have included public enquiries that became more numerous and more complex. We are now in a situation where decision is shared and can

One example is the A86 motorway in the western Parisian region: the first project was rejected under the pressure of the local population, opposed to an alignment going through the forest.

occur only after a process of dialogue reaches a certain degree of consensus. This evolution reduced the influence of CBA, perceived as very technocratic and whose conceptions were not well understood by the stakeholders, as opposed to more simple methods for presenting a project's impacts. In these methods, CBA was merely an information source among other and more numerous ones, among which the impacts of traffic on the environment, and tentative estimations of the project's impact on local economic development.

Still, these evolutions were a driving force for economic analysis and for CBA promoters. Important efforts were devoted to translating into monetised terms the new stakes and situations CBA was confronted with. At the end of the nineties, there was a movement back to economic orthodoxy and to "mono-criterion analysis", with the intention to fight against the perceived dangers of subjectivity and arbitrary introduced in the decision process by MCA. The following reference guides introduced monetised values for environmental impacts such as noise emission, air pollution and global warming. They also codified how to take care of the macro-economic impacts on employment, with a somewhat questionable method3.

Another transformation was linked to the changes in the financing schemes of transport infrastructures. At first, about sixty years ago, the main source of financing was the public budgets, national or local. Private financing appeared in France in the mid seventies for motorways, took an increasing share and from the nineties have become a major source of financing for great transport infrastructures. Private intervention led to give pay more attention to the financial returns of the investments and not only to the social returns. Thus, all recent reference guides include methods for estimating financial returns. The private sector is more sensitive to risk, and risk adverse than public actors. Contract design also gives a great importance to risk allocation; therefore risk analyses were given more attention. These analyses were very approximate at first, using scenarios which nature and parameters were chosen rather arbitrarily by the analyst: they were more like sensitivity analyses. More recently, the question was addressed using economic analysis' tools. Recently, a working group examined this topic and recommended to make use of tools coming from financial analysis, notably analysis of volatility and correlation between project risk and market risk.

CBA is now more and more confronted with questions concerning the impacts of investments on economic development. This is not a new issue; it had already appeared in the seventies with the urban projects. But it took recently more importance with the great urban projects such as Crossrail in London; examples in France are EOLE4 and "Grand Paris"5.

This type of project may intuitively have a major influence on the urban agglomeration's development. This hypothesis and its consequences cannot be tested in the present conditions of project assessment. Indeed, only the impact on affectation of traffic Hows on transport networks and environmental externalities are estimated. Induced traffic is unsatisfactorily treated, and its origins are not explained (increased number of trips with unchanged location, or impact on location choices). No information is given on the impact of transport investment on the population's trip generation, nor on the

3 The method for estimating the number of jobs created by infrastructure construction and operation did not specify if it consisted in newjobs orjustrelocalization of existingjobs, without any analysis of the impacts onjobs at the national level

EOLE is a rail project in the Parisian region, which would double the long distance "RER" transit line that goes East-West. RER infrastructures are highly saturated and notoriously insufficient. It costs 3 billion Euros for 40 kilometers including some new sections and improvement of existing lines.

The core of the project is an automatic metro consisting in a circular line about 60 kilometers long around Paris, for an approximate cost of 30 billion Euros.

relative competitiveness of great agglomerations. Although recent research works show their importance and estimates their order of magnitude (Venables 2007; Vickerman 2007), agglomeration effects on productivity levels is not yet taken care of by CBA in France. These themes are currently under work in order to design a methodology based on the most recent developments of economic analysis and to integrate the corresponding mechanisms in CBA.

3. Evolution and competition

Since we adopt here a Darwinian approach, we have not only to analyse the evolution of the environment of CBA, but also its « competitors » within this environment.

Although CBA has remained central among project studies other than technical conception studies, new analyses have appeared and have grown in importance.

3.1. CBA's competitors

Environmental studies are a first such group: besides focussed environmental studies, such as fauna and flora inventories, the central study in this group is the impact study. In European Union, the impact study is compulsory and has to follow a strict process set by the directive n° 85/337/CEE and its national implementation rules.

Clearly, the basic objective of this impact studies is orthogonal to economic analysis: it deals with environmental impacts indeed, but it also has other goals that relate it closely to the decision process: help in designing a better project as regards environmental concerns (design optimization tool) justify project options that have been decided

give information to the public and make the public participate into the process.

Thus, diffusion of impact studies is a key item, together with public participation.

Besides environmental analytic tools and data, impact studies need output from traffic studies (environmental consequences being closely related to traffic levels, as regards noise, greenhouse gas emissions or air / water quality, for instance), but also from CBA, since thejustification objective implies looking at the relative stakes in both economic and environmental dimensions. Besides, CBA needs output from environmental studies, at least for quantifying environmental external costs.

Finally, environmental studies are far from being purely technical studies: they are designed and already interact with other types of studies including CBA. Also, the great importance taken by environmental concerns in public debates and in public decision has led to a growing role of environmental studies, sometimes at odds with the result of CBA studies.

Another group of studies has developed with the use of specific public infrastructure managers or private financing: finance analysis has become a key issue, in the same time as the part of public infrastructure financing shrank. These rather technical studies are needed, not only for assessing the financial sustainability of a concession, for instance, but also for quantifying expected profit levels and helping in customizing risk allocation between public parties and infrastructure managers, whatever the organizational form. Thus, these studies apply to key dimensions in public decision, and intervene not only for feasibility or opportunity questions6, but also for contract optimization, at a more operational stage.

These studies need the output of traffic studies, including sensitivity tests to external factors or to infrastructure pricing. They give elements that are inputs for CBA but do not necessarily need input from CBA.

For instance: what level of public financing would be needed for a concession to be profitable? Is the rate of return asked by the private partner reasonable when risks are taken into account?

Moreover, with the increasing limitations in public financing, their role becomes more important, and may reduce the weight of CBA in the public decision process, would the public decision maker pay less importance to long term collective output than to short term action and budgetary limits.

Confronted with highly diversified lobbying groups expressing various concerns, public decision makers tend to use Multi-Criteria Analysis methods. This trend benefits also from the increasing number of financing partners in infrastructure projects: higher costs with smaller budgets, together with the intrinsic interest of risk sharing, lead to a strong development of multi-financing schemes7. MCA methods are also applied by some local authorities, and the MCA approach is gaining audience for big projects. Among many reasons, it may help to understand stakeholders' attitudes and to discuss with them, and it may be presented sometimes as an external and « objective » way of analysing diverse dimensions. Still, the core misunderstanding with MCA methods remains: do we speak here of MCA as a tool for giving a multidimensional view of a complex issue, or, as very often is the case, as a tool for classifying projects or project options ? Two basic limits interfere with the growth of the second version of MCA's role in public debates and decisions: making a projection on one composite dimension does imply high loss of information, and lack of confidence for the stakeholders concerned. The technical process has to fix a composite projection axis 8 which is very easily subject to the « black box » objection arguments.

MCA analysis needs, by nature, every other possible analytic study, whether environmental of financial studies, traffic studies or CBA. The reverse is not true, and MCA does not produce additional information on the nature or the magnitude of the project's impact: it simply gives a general view of the diverse impacts, selects or highlights some of them, or builds from them a composite view. CBA and MCA do compete in the common goal of giving a « simple and rational» view of the impacts of a project, including as many natures of stakes as possible, and both need basic information on the impacts. But MCA cannot generate by itself this information, which lies on the positive side of economics, not on the normative one. As regards public decision process, the multi-dimensional MCA approach may develop in the future, as grows the association of the public to the decision.

3.2. Competition versuscooperation?

We have seen that other kinds of studies have developed besides CBA, which may be competitors as regards the influence in the public decision or in communication on the project. But we have also seen that all these studies are not« self-supporting » and are linked by input/output relations.

If we stick to the Darwinian observation of nature, we may ask another question: cooperative species do exist, could it be the case for CBA and its competitors? Besides the clear necessity of cooperation due to input / output relations, another strong concern could lead us to give a positive answer to the cooperative question. Indeed, the philosophy behind sustainable development is, besides developing a multi-dimensional and general view, consists also in a progressive development of knowledge and aggregation of new information, in order to progressively optimize the public decision. Taking this point of view, it becomes clear that developing structured interactions between the different kinds of studies would be keen to be highly profitable. For instance, CBA and financing studies could interact so as to test alternative pricing policies (which unfortunately is very rarely done); or CBA, environmental study and traffic / technical design could more effectively build together an iterative process of project optimization.

The high speed line Tours-Bordeaux in France associates more than fifty public entities in the financial scheme,

whether it is purely quantitative, semi-qualitative, with « flag » methods or not, under certainty assumption or with probability distributions, or even with fuzzy logic.

This is clearly a situation where a collective work would largely over-perform the sum of individual contributions.

The respective time scales of these studies, the present organisation of project design processes9 and cultural differences between people that make these analyses are examples of the big obstacles in the way towards better consideration.

But would it not be an interesting and useful evolutionary process to develop these coordinations?

4. Criticism of CBA and illustrative examples in France

We will now review some criticisms addressed to CBA in France, and get into more detail for some real cases in order to illustrate better how the new environment of CBA may have practical incidences implying adaptations of conventional CBA:

• the great project of « Grand Paris », relying mainly on a new automatic metro network, deals with the scale of a huge project for very big conurbations

• the national transportation infrastructure plan (SNIT) will illustrate the national scale.

The second case represents a new kind of assessment for France, but the other type of assessment topic was until now assessed using conventional rules of surplus estimation. These rules rely on the results of the traffic studies, valuing time savings with a standardised unit value, and adding to the user surplus a monetised value of environmental impacts and the variation of transport actors' profits. Traffic models usually estimate induced traffic through a generation law using a gravity model, more or less rigorously calibrated on past experience.

What will be of interest in these two real cases is not the projects' description nor their assessment results, but the issues addressed by their assessment system and the specific studies involved, in order to bring some recent material for the analysis of CBA evolution.

4.1. Criticism of CBA

Present studies are subject to questions and dissatisfaction on several topics:

• first, they give a very small weight to environmental concerns ; the question is not about weak identification of environmental problems: they correctly emerge from environmental studies. It bears on the way CBA integrates them into its logics and on the final weight they are awarded (generally very low in CBA outputs, but on the opposite quite high in public debates, which do reflect some kind of stated preference value). As an illustration of these affirmations, consider for instance SETRA (2008) which shows that the share of environmental effects rarely reaches more than 15% of the total effects of motorway projects; Chapulut and Taroux (2010) present the same conclusion for HST; yet environmental concerns hold a major place in the public opinion which is very sensitive to pollution and local nuisance issues, and also in more in depth debates where problems linked to global warming are a major concern. It could be inferred that public opinion and knowledgeable debates are in the wrong, but instead of jumping to such a conclusion, it is good to consider the alternative hypothesis, i.e. that CBA has gaps in this domain.

• the rates of return obtained by present studies are quite low, just a few percent. And including the effect of productivity increases due to accessibility increases does not greatly modify their order of magnitude. The WEB (wider economic benefits) computed according to the official CBA method in Great Britain does add at most 15 to 30% to conventional impacts (for instance Worsley 2011), as is the case for British projects. This low rates contrast a lot with the opinion of urbanism experts and of

Common process do give room to such or such type of study only at precise moments in the long life of process design and implementation, and these studies are produced, in practice, within time periods that very often do not leave room nor time for « external» coordination.

local politicians, which think the such a project may have a great impact on the town's future -including high uncertainty on the magnitude of the change, and perhaps on its sign, too -. This practical knowledge leads to think that the consequences of a great investment are subject to debates, a question that should induce CBA to introduce uncertainties about diversified consequences. But this is not the case; up to now CBA generally passes by this question without even mentioning it.

• generally speaking, they do not give answers to the decision maker's concerns. They are for instance interested in transport performances, but also, for great projects, in questions that relate mainly to economy and space, for instance: will the project lead to changes in urbanization? Will it induce or reduce urban spread? Will the project favour high density or low density urbanization? Which parts of the agglomeration will get the most positive impacts of the project? How the agglomeration's competitiveness will be modified relatively to neighbour or competing agglomerations? How will migratory flows change, and with what outside territories will these exchanges take place? How will the revenue distribution be modified, and the growth rates?

• CBA does not answer this type of question10 (ITF 2011, Quinet 2010); it gives only one or very few figures, such as net present value or internal rate of return. For a long time, this figure has been considered to represent the social surplus stemming from the project, even if the detail allocation of this social surplus was not given according to the above components. But this equivalence relies on several hypotheses among which absence of externalities, of increasing returns and of firms' power market. From the new economic geography, we know that these hypotheses do not capture the mechanisms of urban economies, which are based on increasing returns, agglomeration externalities, and taste for diversity. Statistical analyses have even showed that it is possible to measure the consequences of these theoretical mechanisms and that they are not negligible, a typical order of magnitude being for instance an elasticity of productivity to density of about 3% add ref(see for instance Prager § Thisse 2008 or Vickerman 2007) . In the same time, spatial modelling has made some progress, and many land use transport integration (LUTI) models have appeared which allow simulating consequences of a transport improvement on spatial organization (see for instance Wegener 2011, Echenique 2004 or Brocker and Mercenier 2011 for reviews of these models).

Thus, tools are becoming available that allow to tell something on decision makers' concerns. The results are not enough advanced for a complete definition of methodology that would open CBA to these new mechanisms and issues. But they allow to identify research axes and to maintain the hope to be able to get some results, would they be mere qualitative results at first, and thus, in a relatively short term, to be able to enrich CBA.

4.2. The "GrandParis"project

Following this line of thought, many initiatives are currently being launched in several directions for great transport projects in the Parisian area, which should in the coming years produce material for building a new CBA for urban projects.

Estimation of the impact of the project on the Parisian region's attractiveness for foreign direct investment:

Announcing a great transport network for Grand Paris could increase international investors' interest for the Parisian region, comparatively to other very big agglomerations which are competing with it at the

10 The point here is to consider the items on which CBA could be improved. As concerns the previous series of questions, they have a lot to do with the technical ability to evaluate indirect and long term impacts. Surely enough, MCA is prone to the same flaws; nevertheless it is better considered than CBA as regards the attention paid to environmental or local impacts - as long as MCA does not drown these issues into a kind of global indicator -.

world's scale. This transport network may appear as an amplifier of the efficiency of public authorities' local strategies, which are intended to strengthen research productivity and the links between research and firms. It appears necessary to try to estimate this effect in terms of employment, based on the rare international studies available on this topic.

Effects on firm localization choices and employment:

The analysis of job localization constitutes a key element in the assessment of a big urban transport infrastructure. This analysis may begin with the study of the main factors of creation, destruction, relocalization and variations of the number of jobs in the regions' firms, and with the design of a localization model forjobs and firms inside the major cities.

Estimates of agglomeration effects resulting from the project:

This item would be expressed in terms of firm productivity and additional GDP. The literature on agglomeration effects is now important, but gives quite variable numerical results, probably due to the econometric difficulties encountered in their measure. An in-depth review of past and current literature on this topic could help to estimate the relative magnitude of impacts to be expected from the polarisation induced by the transport network, as compared to other expected impacts.

Computation of the overall impact of the project on job localization and population localization, together with land values:

The idea here is to use one (or rather several) land use models based on household behaviour and firm localization. Many tools have been developed in France and all other the world, which have not yet been fully satisfactory. Parallel implementation of several of these models should give a more precise and less uncertain view of the potential spatial consequences of the project.

Estimation of transport improvement's impact on the whole growth dynamics of the region:

This would suppose developing a macro-economic model for the whole Parisian region. Transport improvement has an impact on the economic agents' anticipations. Consumption and investment levels of the region's economic agents, households or firms, partly depends on their anticipations of growth. It can be imagined that a shock on anticipations could result in a modification in the regions' growth rate. The economic literature is abundant on characterization and estimation of this effect on a national level, but works produced on this topic at a regional level are rarer.

Prospective study:

In order to evaluate a mid or long term project such as are most great projects, it is not satisfactory to extrapolate present trends without considering future transformations or inflexions that could occur on the long term. This is particularly true for the estimation of long term impacts of the transport network on household and firm behaviour and on urbanization types, considering the most significant technological trends observed now in trip organization's evolution (potential development of new services driven by technological progress, especially with NICT) and for the possible energy futures.

4.3. The national transportation infrastructure plan in France (SNIT)

The national plan for transport infrastructure (SNIT) is under work in France (MEDDTL, 2011), and the assessment studies produced for the moment, which relate to a project version of the SNIT, deserve some comments in this paper. It is a good example of how complex the objects submitted to assessment may become due to the adjunction of new objectives inspired by sustainable development concerns.

As a planning exercise, the SNIT has to comply with directive 2001/42/CE (European Union, 2001), i.e. an environmental assessment must be produced and made public. The French government decided to produce also a more comprehensive assessment, taking into account the other dimensions of sustainable development, within the limits of available data and degree of precision of the decisions included into the SNIT. Indeed, the SNIT is not only an infrastructure long term planning document; it outlines also the national policy as regards maintenance, renewal and operation of national networks, and national contribution to urban transit development.

It covers all modes including intermodal supply such as motorways of the sea. More than fifty innovative actions are specified besides the strategic orientations given in the SNIT. Another key feature is that the existence of the SNIT has been decided by law. More precisely, it appeared within a wide range law that covered many topics as a result from the environmental summit « Grenelle de l'environnement». These debates held a great importance in France and led to numerous decisions intended to make the whole society move towards more environment friendly ways of thinking and acting, and more generally towards sustainable development. The SNIT has therefore been given by law several objectives that go well beyond the usual functional objectives of infrastructure networks, such as contributing to greenhouse gases emission reduction,

Thus, the SNIT is a complex planning object with no previous example to base its assessment on. We will focus here on the role CBA held and on the interactions between the diverse assessment studies produced for the SNIT.

CBA has mainly been used at the scale of the individual infrastructure projects that were reviewed for deciding on their inclusion (or exclusion) in the SNIT. Its results provided some of the numerous criteria that were used for this review, organized through a kind of non-weighted MCA. One of the technical issues in this process was to use homogeneous methodologies and reference values in order to ensure correct comparability between projects. Among the difficulties encountered, the projects were at highly diverse stages of maturity, some where highly advanced and known in high detail, others were at very preliminary stages. Nevertheless, this problem was less acute for economic analysis than for environmental assessment, which requires localized information and had therefore to use less precise indicators such as « surface of valuable natural area potentially concerned » since the localization of projects was not precisely known11.

At the national scale, the economic impact of the SNIT could be analysed from several angles, we will select two of them. The most classical one was trying to get aggregated indicators such as net present value of the infrastructure development scheme. This was done in practice with the data and tools that were available: output from individual CBAs was used, after corrections for ensuring better homogeneity, to obtain a gross order of magnitude of the whole development scheme. On theoretical grounds, this is not correct since additivity conditions were not strictly met. Nevertheless, since road projects were not very numerous, concerned mostly short links, and were scattered on the present network, the approximation of additivity did not introduce a great distortion. The high speed rail network caused more concern, since the SNIT plans to double and more the present network. In order to take account of network effects and of possible effects in the intermodal competition, a multimodal traffic model was used. This allowed checking that the order of magnitude obtained for the long-distance traffic effects of the scheme was similar to the results obtained by aggregation of individual projects. Still, this model could not go further in overall CBA analysis and a countercheck of aggregated economic indicators' order of magnitude could not be done. Quantitative indicators have been produced that were directly derived from the detailed traffic simulation results of this multimodal model, such as yearly С02 emissions avoided (2 to 3 million tonnes) or reduction of time spent in interurban transportation (about 4%).

For some projects of long rail links, localization could vary within a band which was 10 kilometers large and much more.

Another overall economic question focussed on indirect macroeconomic effects of the SNIT. This disputed issue could be analysed from a wider economic benefits (WEB) point of view: schematically, better accessibility translates into higher productivity and improved economic performance. Still, the recent debates on WEB tend to be much more cautious for interurban infrastructures than for (sub)urban ones on this methodology. Besides, an existing macroeconomic model was tested for the SNIT case12, and concluded that, if the short term effect would indeed be beneficial, the final result could be negative, due to complex dynamic interactions (through, for example, delayed effects on prices).

5. Conclusion

The previous analysis gives indications on the desirable evolutions of CBA. These evolutions relate to the underlying economic theory, and concern branches of this theory that CBA should invest in. They also have to do with the highly diverse information that needs to be gathered. Last, they relate to new modes of integrating CBA to the decision process and to project governance.

As for economic theory, CBA makes a greater use, as we have seen, of mechanisms and specializations outside the usual limits of the transport sector. Successively, new needs of CBA appeared for environmental economics, with concerns about externalities, pollution and global warming; for spatial economics, with the consequences of investments on economic activities' organization; for macroeconomics, with the attention given to unemployment problems and to the impacts on economic competitiveness of territories; for financial economics, with the increase in private financing and with the growing concern on risks; last, for the issues of equity and acceptability which are of utmost importance in a world where redistribution mechanisms and identification of winners and losers become a key item in investment projects.

The needs for statistical data are oriented towards increased disaggregation and diversification. This trend results primarily from the progress of knowledge: thus, in addition to monetary costs and time costs, CBA now generally considers costs related to reliability or to unwished shifts from desired transport schedules. This trend is driven by the increasing use of models, that are more and more complex and data demanding.

Even for the basic case of traffic modelling, the conventional four steps models, based on rough zoning and approximate definition of networks' and demand's characteristics, have evolved towards an ever more detailed network description; the progress of modelling techniques allow - and incite - to use additional data, for example data on trip chaining and timing. The LUTI models encounter a good success now: they need much more data since they have to be fed with economic information on production processes, firms and their exchanges, and also on households. Data processing cannot do without information systems that have GIS functionalities, thus displacing model implementation from handcraft practice to industrial approach. Similarly, unit values used for conventional reference valuation in CBA, which were once very simple to express, become more and more complex: starting from one time value, CBA has now to deal with a whole collection of time values, depending on revenue levels, transport modes, trip purposes, and soon(Hensher 2011).

The third direction of change has to do with the integration of CBA in project governance.

12 Both the methodology and the assumptions were source of controversial debates, but we mention it to show that other kinds of economic analysis may now address some of the issues on which CBA needs to evolve.

Formerly, the public decision was taken, once for all, by only one person or a very small group. Now decision consists in several decisions taken during the project's lifetime, each of them being potentially able to modify deeply the project, as concerns its functions or its alignment. CBA has to adapt to each step of the project, to make the best use of available data at that time, and to deal with more numerous decision makers, whose concerns may change along the project's life, and imply that CBA develop new angles. This two fold diversification means to make use of knowledge in directions that arejust emerging. Thus, our knowledge on the role of transport infrastructures on location choices is still not very robust, but it is progressing and CBA methods have to adapt continuously to this progress.

Finally, the analyst in charge of CBA has to adjust closely to modifications linked to the evolution of knowledge and of decision makers' concerns. This means that the analyst has to be watching and reactive to changes in the project's environment. The consequences of this evolution may be questioned from an ethical point of view. Surely, enlarging the set of decision makers and opening the public dialogue process constitutes a democratic progress. Nevertheless, the objectivity - real or supposed - that was attached to pure conventional surplus calculation moves away. But this is another debate that goes far beyond the scope of this paper.

Acknowledgment

The authors are indebted to two anonymous referees for helpful comments and suggestions.

References

Beuthe, M., Bouffioux, Ch., Krier, С. § Mouchart, M. (2008). A Comparison of Conjoint, Multi-Criteria, Conditional Logit and Neural Network Analyses for Rank-Ordered Preference Data. In M. Ben-Akiva, H. Meersman and E. Van de Voorde, RecentDevelopments in TransportModelling, Emerald Group Publishing Ltd., Bingley, UK. (Ch. 9, 157-178).

Boiteux, M. § Baumstark, L. (2001). Transports: choix de investissements et coût des nuisances. Paris : Commissariat General du Plan.

Brôcker, J. § Mercenier, J. (2011). Generalequilibrium models for transportation economics . In Handbook of Transport Economics by de Palma A., Lindsey R., Quinet E., and Vickerman R., Edward Elgar

Calthrop, E., De Borger, B. § Proost, S. (2009). Cost-Benefit Analysis of Transport Investments in Distorted Economies. Transportation Research Part B, forthcoming.

CGDD (General Commission for Sustainable Development) (2009). Annual Report to the Parliament on ImplementingFrance's Environment Round Table Commitments. Paris: MEDDTL.

CGDD (General Commission for Sustainable Development) (2011). Rapport d'évaluation globale de l'avant-projet consolidé de SNIT (Integrated assessment of the draft French national transportation infrastructure plan ). Paris: MEDDTL.

Chapulut, J.N. § Taroux, J.P. (2010). Trente ans de LGV; comparaison des prévisions et réalisations.

Revue Transport, 462, 229-239,-

Department for Transport (2009). NATA Refresh: Appraisal for a Sustainable Transport System. London: Department for Transport.

Department for Transport (2011). TAG unit 2.5: The Appraisal Process. London: Department for Transport.

Dupuit, J. (1849). De l'influence des péages sur l'utilité des voies de communication. Annales des Ponts et Chaussées, 17, MémoiresetDocuments, 207.

Echenique, M. (2004). Econometric Models of Land Use and Transportation. In Hensher, D.A., Button, K.J. (Hg.), Transport Geography and Spatial Systems, Volume 5 of Handbook in Transport, Kidlington, UK: Pergamon/Elsevier Science, 185-202.

European Union, 2001. Directive 2001/42/EC of the European Parliament and of the Council on Strategic Environmental Assessment (assessment of the effects of certain plans and programmes on the environment). Official Journal of the European Communities L 197/30, 30-37.

Florio, M. (editor) (2007). Cost Benefit Analysis andlncentives in Evaluation. Edward Elgar, 2007.

Flyvberg, B. (2009). Survival of the Unfittest: Why the Worst Infrastructure Gets Built. Oxford Economic Review, 25/3, 344-367.

Graham, D.J. (2007). Agglomeration, Productivity and Transport Investment. Journal of Transport Economics and Policy, 41, 317-343.

HEATCO (2005). Current Practice in Project Appraisals in Europe: Analysis of Country Reports. Deliverable 1. http://heatco.ier.uni-stuttgart.de/.

Hensher, D. (2011). Valuationof Travel Time Savings. In Handbookof Transport Economics by de Palma A., Lindsey R., Quinet E., and Vickerman R., Edward Elgar.

International Transport Forum ( 2011). Improving the practice of cost-benefit analysis in transport. Summary and Conclusions of the Roundtable on Improving the Practice of Cost Benefit Analysis in Transport. Discussionpaper 2011-1. Paris: OECD/ITF.

Mackie, P. (2010). Cost-Benefit Analysis in Transport: A UK Perspective. Discussion Paper 2010-16. Paris: OECD/ITF.

MEDDTL (Ministry for Ecology, Sustainable Development, Transport and Housing) (2011). Preliminary Project for the National Transportation Infrastructure Plan submitted for discussion. Paris: MEDDTL.

Merlin, P. (1991). Géographie, économie et planification des transports. Paris: PUF.

Meunier, D. § Quinet, E. (2010). Application of Transport Economics and Imperfect Competition. In

Research in Transportation Economics (forthcoming).

MEEDDAT (Ministry forEcology, Energy, Sustainable Development and Land Use Planning) (2008 a).

Circular of 9 December 2008 on the Reference Frameworkfor Evaluation at the Ministryfor Ecology, Energy, Sustainable Development and Land Use Planning. Paris: MEEDDAT.

MEEDDAT (2008 b). Circular of 9 December 2008 on the Establishment ofa Quality Charter for Evaluation. Paris: MEEDDAT.

MEEDDAT (2009). Impact Study of the 2008 Grenelle Environment Bill, Mimeograph. Paris: MEEDDAT.

Ministry for Transport Infrastructure (1998). Circular No. 98-99 of 20 October 1998, Methods for the Economic Evaluation oflnterurban Road Investment. Paris: METL.

Ministry for Transport Infrastructure (2004). Framework Instruction of 25 March 2004 on the Harmonisation of Methods for Evaluating Major Transport Infrastructure Projects. Paris: METLTM.

Ministry for Transport Infrastructure (2005). Circular of 27 May: Update of the Framework Instruction of 25 March 2004. Paris: MTETM.

Ottaviano, G.I.P., Tabuchi, T. § Thisse, J.-F. (2002). Agglomeration and Trade Revisited. International Economic Review 43, 409-436.

Prager, J.C. § Thisse, J. F. (2008). Economie Géographique du Développement. Repères. Paris: La Découverte.

Quinet, A., Baumstark, L. § Celestin-Urbain, J. (2008). la valeur tutélaire du carbone. Paris : Centre d'Analyse Stratégique.

Quinet, E. (2007). Cost Benefit Analysis of Transport Projects in France. In M. Florio (Ed.) Cost Benefit Analysis and Incentives in Evaluation. Edward Elgar.

Quinet, E. (2010). The Practice of Cost-Benefit Analysis in Transport: The Case of France,. Discussion Paper 2010-17. Paris: OECD/ITF.

SETRA (2008) (Service d'études des transports, des routes et de leur aménagement). Analyse

transversale de bilans IOTI deprojets routiers : volets « Transport» et « Effets socio-économiques ». Paris: MEEDDAT.

Venables, A.J. (2007). Evaluating Urban Transport Improvements: Cost-Benefit Analysis in the Presence of Agglomeration and Income Taxation. Journal of Transport Economics and Policy, 41, 173-188.

Vickerman, R. (2007). Recent Evolution of Research into the Wider Economic Benefits of Transport Infrastructure Investment. Discussionpaper No. 2007-9, ITF.

Wegener, M. (2011). Transport in spatial models of economic development. In A Handbook of Transport Economics by de Palma A., Lindsey R., Quinet E., and Vickerman R., Edward Elgar.

Worsley, T. (2011). The evolution of the London's Crossrail Scheme. Discussionpaper n°2011-27, prepared for the Roundtable on Major transport infrastructure projects and regional economic developments: Assessment and implementation. Paris: ITF.