2.1 Multicriteria analysis
Usually, a decision-making problem has more than one goal to reach, and there is always a trade-off between the different goals, advocated by different interest groups or stakeholders. Within this context, the MCA is seen by some authors as the most appropriate tool to adopt [
11‐
15]. The MCA is a tool for selecting alternative projects, which have significant social, economic, environmental impact, that allows to take into account several criteria and the stakeholders’ opinions. Therefore, the inclusion of multiple stakeholders in the decision making process is widely acknowledged and it is often a crucial factor for the successful implementation of the measure or project under consideration in the transport sector [
15].
Within the MCA, the objectives to reach must be specified and corresponding attributes or indicators must be identified. The actual measurement of indicators need not to be in monetary terms, but it is often based on scoring, ranking and weighting of a wide range of qualitative impact categories and criteria.
The MCA is mainly organised into the following phases:
-
Phase 1. Definition of the projects or actions to be judged.
-
Phase 2. Definition of judgment criteria.
-
Phase 3. Analysis of the impacts of the actions.
-
Phase 4. Judgment of the effects of the actions in terms of each of the selected criteria.
-
Phase 5. Aggregation of judgments.
Multi-criteria methods can be classified according to the nature of the decision problem restrictions (implicit or explicit), and the nature of the results (deterministic or random) [
14].
There are several methods that might be applied to transport evaluation, the most suitable are: (i) Analytic Hierarchy Process (AHP), (ii) Analytic Network Process (ANP), (iii) REGIME, (iv) ELECTRE family, (v) Multi Attribute Utility approach, (vi) ADAM type (for a detailed review, see [
16,
17]). In the present work particular attention is placed on the AHP, ANP and REGIME, being the AHP and the ANP two of the most used and well-known multi-criteria techniques, and the REGIME adopted to assess sustainability at the neighbourhood level.
The AHP, developed by Saaty [
18,
19], is a three-stage method: (i) building the hierarchy; (ii) weighting the indicators by a pair-wise comparison, and (iii) calculating the final value for the alternatives. The AHP, indeed, consists of decomposing a complex decision making process into a hierarchical structure, with the ultimate goal at the top of the hierarchy, the primary criteria in the second level, and the subcriteria in the third level. In the following level there might be additional subcriteria, while at the bottom level of this “probability tree” [
14] there are the discrete options under consideration (for a review see, among the others, [
20]). The elements of the hierarchy can relate to any aspect of the decision problem—tangible or intangible, carefully measured or roughly estimated, well- or poorly-understood—anything at all that applies to the decision at hand.
The stakeholders, public and private, might take part to the construction of this hierarchy, and “especial care must be taken when building up the hierarchy such that pernicious double counting of attributes is avoided” [
14]. Once the criteria and sub-criteria have been settled, a set of weights is required. These weights represent the relative importance of the criteria, subcriteria and attributes belonging to a specific nest in the hierarchy. According to the original procedure developed by Saaty, these weights are obtained from pairwise comparison matrices, for each nest in the hierarchy. Once weights are available, the hierarchical structure is collapsed, following a folding back procedure. For each option under study, there will be a final weight. These final weights are used to rank the options.
Another procedure that is widely adopted is the ANP. The ANP provides a general framework to deal with decisions without making assumptions about the independence of higher level elements from lower level elements and about the independence of the elements within a level. In fact the ANP uses a network without the need to specify levels as in a hierarchy. Influence is a central concept in the ANP. The ANP is a useful tool for prediction and for representing a variety of competitors with their surmised interactions and their relative strengths to wield influence in making a decision (for a review see, among the others, [
21]).
The ANP is a coupling of two parts. The first consists of a control hierarchy or network of criteria and subcriteria that check the interactions. The second is a network of influences among the elements and clusters. The network varies from criterion to criterion and a different supermatrix of limiting influence is computed for each control criterion. Finally, each of these supermatrices is weighted by the priority of its control criterion and the results are synthesized through addition for all the control criteria.
With the ANP a problem is often studied through a control hierarchy or control system of benefits, a second for costs, a third for opportunities, and a fourth for risks, and each is represented in the controlling system. The synthesized results of the four control systems are combined by taking the quotient of the benefits times the opportunities, to the costs times the risks for each alternative, then normalizing the results over all the alternatives to determine the best outcome.
The REGIME method uses pairwise comparison (for a detailed review see [
22,
23]) on the basis of which a synthetic index is calculated. The index defines a ranking among alternative options: the higher is the index, the most preferable is the option. In this case, the synthetic index expresses the level of performance of the different selected indicators as respect to the criterion referring to each analyzed experiences, and make them explicit in the ‘information matrix’. In the REGIME analysis the main focus is the sign of differences between impacts of alternatives. In general terms, an evaluation table is given and composed by scores of a number ‘n’ of alternative options with respect to ‘m’ criteria. In the case of ordinal information, the weight can be represented by means of rank orders in a weight vector: the higher the value of the weight, the better the correspondent criterion. The alternative options will be compared pairwise for all criteria and for two alternative choice options, the difference of the criterion scores is assessed.
1
2.2 Cost benefit analysis
Cost Benefit Analysis is the most used evaluation technique for assessing infrastructural investments. In the transport field, it is the basic tool in the majority of countries in Europe ([
24]; OECD, ECMT [
25]) and in the rest of the world ([
26]; EVA TREN [
27]), and it is also widely adopted by all the international bodies ([
28,
29]).
A number of official guidelines exist.
2 They, despite some differences in how a CBA must be actually performed, always refer to one single common theoretical framework. It is not the aim of the present paper to go in deep with the well known CBA theory, however, some key aspects can be pointed out.
Firstly, the CBA is based on monetisation and inter-temporal discount. Money is the measure unit used as a common numerary to translate all costs and benefits associated to an investment or a policy. Apart direct monetary costs in perfect markets (e.g. untaxed cost of energy) whose monetisation is trivial, also non-market goods and goods traded in an imperfect market are quantified. The first ones (e.g. time or environmental costs) are translated into the common numerary by means of the willingness to pay or by deriving prices from substitute markets (hedonic prices method). The second ones are translated into their opportunity cost by subtracting taxes (e.g. fuel prices) and by looking at the direct effect only (e.g. shadow price of labour cost).
Once all relevant effects of an investment are quantified, the concept of inter-temporal discount is used to translate future costs and benefits to present day by means of a social discount rate. In this way, the future can be compared with present.
The core of CBA technique is the social surplus, sum of users’ surplus, producers’ surplus, and, if the case, non-users’ and Government surplus. Surplus is the difference between the willingness to pay/sell of users/producers for a good (which is the combined effect of perceived utility and income distribution), and the effort needed to obtain such good (the monetary cost or any other kind of effort). A scheme generates a variation of surplus, between the situation “with” and “without” it. Following this concept, CBA essentially compares among trade-offs: total benefits must exceed the total opportunity cost of consumed resources (labour, time, monetary costs, etc.) to make a project feasible. Otherwise, social cost exceeds social benefits and the scheme should be rejected.
In order to have a significant result, two hypotheses must be fulfilled. The first is represented by the Kaldor-Hicks criterion [
30], which states that a resources allocation change is efficient if the surplus obtained by some actors exceeds the surplus losses paid by other actors, i.e. if the benefit for a person can be compared with the cost of another one. The second hypothesis assumes that a scheme is marginal, i.e. does not change upstream and downstream markets.
From a practical viewpoint, transport CBA usually quantifies the investment plus running cost of a scheme and compare it with direct benefits, that usually are represented by time, running costs and environmental cost savings. Recently, CBA can also include wider benefits, i.e. macroeconomic benefits that are not subsumed in the direct benefits (agglomeration effects, labour pooling, efficiency; see
3.2).
Moving to the topic of SM policies, several quantitative studies do exist. For example, Farrell et al
. [
31] and Beria [
32] assess quantitatively the effect of different mobility policies in terms of avoided emissions, but only in terms of policies’ effectiveness. They, however, do not evaluate the costs associated to such policies, at least in strict microeconomic terms, and thus they do not perform a socio-economic assessment.
In contrast with the importance of CBA for infrastructural investments, in fact, the use of such tool for softer measures, typically those associated with sustainable mobility, is less common. The above cited examples of CBA guidelines actually refer to infrastructural investments. Motivations that lay behind to the less common application of CBA to policies assessment are clear and will be discussed in a while. However, some examples of CBA applied to policy analysis exist in scientific literature, together with some applications.
From a general perspective, Farrow and Toman [
33] state that CBA can be used to improve environmental regulation. Its limits must be known, but it is a necessary tool to cope with scarcity of resources. They describe an evaluation process that reflects the same flow used in physical investments: the definition of a base scenario in which the state of the world goes on with already decided actions, the definition of a complete set of policy alternatives, the identification of the changes to the costs and benefits due to the policies, the assessment and finally a sensitivity/risk analysis in order to—to evaluate robustness.
A warm debate is open on the limits the use of CBA in the environmental field policies. Among many, Heinzerling and Ackerman [
34], Hahn [
35], and Turner [
36] analyse, from radically different viewpoint, pros and cons of the approach. While the first heavily criticise the approach, claiming that it is trying to price priceless things, the other two conclude that, despite the well-known limitations of the technique, it still plays a role also in the environmental policy appraisal. In particular, Turner [
36] revises the relevant literature on the topic, underlining that CBA usefulness is particularly true if moving from a prescriptive role to that of information and decision support and when scarcity of resources exists. Only in the policies or actions involving “values” (poverty, cultural aspects, beauty, etc.) the role of CBA must be only partial. In general, CBA is seen as part of a multiple-criteria policy analysis process.
Another specific field of policy assessment in which the debate on CBA is particularly active is that of safety in transport. For example, Elvik [
37] analyses the conditions under which applying CBA in safety policies is justified or not, showing that it is appropriate if used as a tool to find “the most cost-effective measures to reduce the number of accidents and injuries”. Wijnen et al
. [
38] go more into the practice and present the method to assess safety effects (by the estimation of the value of a statistical life).
However, apart from specific ethical and theoretical reflexions on the applicability of CBA to policy analysis, fewer contributions exist explaining “how” to perform a CBA of a generic policy instead of an infrastructure. In fact, all manuals and theories usually refer to the “physical” projects, involving an investment cost and some future benefits. Policies, instead, represent a much broader world, where infrastructures are only one of the possible options, aside to technological investments, education, pricing, etc. to obtain a chosen goal.
One of the policies more studied in literature is road user charging. Rich and Nielsen [
39], Transport for London [
40], Eliasson [
41], and Rotaris et al
. [
42] perform a CBA to road user charging schemes in Copenhagen, London, Stockholm and Milan, respectively. A similar measure, but applied at a country scale is described in Glaister and Graham [
43]. In this kind of policy, however, main benefits and costs are similar to those related to infrastructures: investment and fixed running costs, time savings or costs, revenues, reduction of congestion and possibly of pollution.
Another typical sustainable mobility policy is car sharing. Fellows and Pitfield [
44] use the standard British methodology (the so called “COBA”) to analyse a soft measure like car sharing. Methodologically, however, they perform it “exactly in the same way as new road schemes”, quantifying benefits from reduced vehicle kilometres, increased average speeds and savings in fuel, accidents and emissions.
Moving to measures less and less dominated by an investment cost, Sælensminde [
45] performs a CBA of walking and cycling networks, taking into account all the relevant aspects, some of which are usually not considered: travel time, insecurity, accidents, savings for school buses, health effects, parking costs, environmental external costs. On the same topic - cycling - Börjesson and Eliasson [
46] demonstrate that, in a CBA of cycling support policies, health effects on cyclists, which are usually considered as external benefits, are actually internalized. In fact, the generalized cost of a cyclist seems to include also the health benefit of cycling: he/she accepts a slower mode vs. car or public transport also because perceiving a personal benefit in terms of health. For this reason, considering health benefits as an external benefit to be added in case of an improvement to cycling network, introduces a double counting if consumers’ surplus is correctly accounted.
Finally, a practical application for environmental policies is offered by Massé [
47]. The author is calculating the costs and benefits associated to the compulsory introduction of anti-particulate matter filters on cars and trucks in France. He obtains a very good B/C ratio, thanks to the large benefits in terms of saved human lives. On the same topic of pollution, Bollen et al
. [
48] perform a CBA to evaluate policies that reduce jointly or separately local and global pollutants.
2.3 MCA and CBA: integrated approaches
The combination of MCA and CBA has been applied in several studies. For instance, CBA has been combined with AHP or ANP in order to cope with the CBA’s weakness in reflecting stakeholders’ knowledge in the evaluation process of projects [
49]; while a financial CBA plus a MCA matrix was created for an economic evaluation of urban transformations [
50].
As concerns specifically transport issues, in the EU member states (for a detailed review see [
51]), different evaluation techniques are adopted with a great prevalence of CBA. In most of the cases (i.e. in Sweden, Netherlands, United Kingdom), CBA is supplemented by a specific appraisal for impacts that are difficult to be monetized; in some others (i.e. Belgium, Austria and Greece), MCA is used, but it includes CBA as one of the criteria. Finally, in France CBA has recently been considered weak in stimulate stakeholders’ interactions, thus, in order to create a larger public debate, MCA tools have also been adopted [
51]. Even at the urban scale of transport investment evaluation, as the outcome of CBA did not match with the one selected by MCA, a combination of the two methods was suggested [
14].
As regards sustainable development, for instance, in the ExternE-Pol project [
52] it was possibile to integrate multicriteria methodology into the wider structure of CBA by using a specific framework. By doing so, all the stakeholders’ preferences could be exploited, thus, deriving indirect monetary values for environmental goods and impacts .
Besides, as for the specific environmental impacts assessment from transportation projects, either for small-scale (local) or large-scale (regional/national), a combination of CBA and MCA has been developed in the Evaluation Framework of Environmental impacts and Costs of Transport (EFECT) [
53]. EFECT is a methodological framework, which aims to cover all kinds of transport environmental initiatives, namely policies, plans and projects, by using an additive function.
Finally, as concerns the evaluation of SM at neighbourhood level, as it will be described in details in the following sections, a first attempt to use MCA has been done [
54], while no specific evidence has been provided for the application of CBA. This is due to the fact that CBA well fits in the assessment of specific infrastructures or policies where monetary or monetizable costs and benefits prevail (investment, time, environmental benefits, etc.) (see
3.2 for a discussion), therefore when it is applied to assess soft policies at the urban scale, it can be well supplemented by a broader evaluation approach like MCA. The MCA, indeed, effectively evaluates effects like social inclusion, change in behavior of citizens, change in the use of city, quality of life, etc.