Impact Assessment and Evaluation in Transportation Planning

In the past years the context and substance of transportation planning in many countries have exhibited dramatic changes. Transportation planning is no longer a ‘fixed route’ planning, but is increasingly characterized by the need for flexible and visionary policy strategies and decision processes in an uncertain environment.

The actual impact of the research community on public decision making — both in general and in the transportation sector — is not always impressive. Institutional bottlenecks, administrative procedures, opportunistic behaviour or unanticipated events are often obstacles on the way toward clear and straightforward choice behaviour of policy makers (cf. Leman and Nelson, 1981). In a noteworthy article Verdier (1984) points to the need to bridge the gap between experts and policy-makers by suggesting inter alia the following guidelines:

Rational and consistent policy analysis presupposes a reliable assessment and balanced evaluation of all foreseeable consequences and choice possibilities in relation to policy initiatives. The aim of generating and judging alternative frameworks of policy measures is a far from easy task for mainly two reasons. The process of generating meaningful choice options in the context of policy analysis is extremely complicated in an open, multi-actor social system with diverging interests, while also the assessment of expected impacts of policy measures — especially in a dynamic spatial system — is fraught with many difficulties inherent in the uncertainty context of decision-making.

In the past, various methods and models have been designed in order to analyze the allocative efficiency of (urban, regional or transportation) policies, their equity consequences and their impacts on a broader set of socio-economic and environmental development objectives (e.g., energy use, environmental quality). Some good illustrations of spatial impact analyses can be found in Moore and Rhodes (1973, 1976). Clearly, the impacts of policies can be measured by means of a multiplicity of indicators, such as employment, income, investments, amenities, and so forth. In general, policy objectives (and hence the achievement of these objectives) have to be represented by a multidimensional profile, so that in principle the appraisal of policies has to be based on multiple indicators (see Nijkamp, 1979).

In the previous chapter it has been asserted that models are one possible structured method — and certainly not the only one — for assessing the impacts of a certain policy measure. Clearly, impact assessment requires a careful estimation of all expected consequences of a stimulus (e.g., a rise in gasoline tax, a new underground system) on the behaviour of individuals, groups or society as a whole. In order to generate testable results, very often researchers have to resort to statistical and econometric techniques.

In this chapter we will discuss the results of a case study which was undertaken for assessing the impact of infrastructure endowments on three different regions in the Netherlands. This case study was part of a study on infrastructure and employment effects with the aim to provide a more solid empirical basis for the question whether (public) infrastructure investments are significantly contributing to an improvement of the employment situation in the Netherlands (see Bruinsma, 1990). For the case study a micro-oriented approach — related to the method of quasi-experimentation (see Section 4.2) — was used to assess the programme effects of infrastructure endowments. Programme effects refer to the supply of infrastructure promoting long-term structural employment effects based on maintenance and management and spin-off effects caused by changes in the relative locational attractiveness of places or regions for new enterprises. In the study both the direct effects (related to the design, construction and building aspects of infrastructure provisions) and indirect employment effects of infrastructure investments (related to derived (second-order) consequences of the creation of infrastructure) were assessed also for the Netherlands as a whole. Our main interest however, is in the regional programme effects.

Infrastructure policy is one of the popular components of regional development policy. Infrastructure has in general a long range impact on the structure of regions or nations. Since socio-economic disparities are usually not the result of short-term economic fluctuations but of structural differences in space, it is conceivable that regional policy often assigns a critical role to infrastructural provisions. Moreover, in most cases, transport infrastructure appears to have the highest financial share among all public infrastructure endowment expenditures.

The improvement of infrastructure is one of the major policy instruments of the European Regional Development Fund (ERDF), created in 1975 in order to support and develop backward regions in the European Community. Infrastructure is regarded here as (material and immaterial) public capital which forms the foundation of — and is a critical success factor for — all other productive or socio-economic activities in a country or region. Especially since 1978 — when it was decided that a wide variety of public goods could be considered as infrastructure to which the Regional Fund (and other financial policies of the Community) might financially contribute — subsidies for infrastructure projects gradually rose to a significant part of the Regional Fund’s annual budget.

In this Chapter we will present a simple explanatory model for assessing the regional impact of the ERDF (European Regional Development Fund) expenditures. This model was developed as an extension of the frequency analysis as an impact assessment tool for ERDF Commitments, described in the previous chapter. The frequency table analysis is mainly meant to be an exploratory tool. The explanatory model however, to be developed in the present chapter serves to test the existence of a causal quantitative relationship between the economic development of a region and various explanatory background variables for this development. The ERDF Commitments in this model are only one of such explanatory variables. In a way analogous to the frequency table analysis, this model is tested for two empirical studies, viz., Italy and The Netherlands.

Modern societies demonstrate an increasing complexity. The impact patterns of decisions and actions of individuals and groups are often intricate, far-reaching and conflicting. In the recent past, this has provoked the need for many kinds of impact assessment analyses, for instance, environmental assessment, socio-economic assessment, technology assessment, etc. Impact analysis however is a necessary but not yet sufficient stage in (transportation) planning (see Chapter 3–5); impacts also have to be judged. Especially the field of public policy-making is often encountering difficult evaluation problems regarding the impacts of choice alternatives. In this context, various types of conflicts may be distinguished:

After the discussion of multiple criteria evaluation as a tool in planning (see Chapter 10), we will now present a concise overview of some technical and substantive aspects of multiple criteria analysis. Multiple criteria analysis has essentially its roots in the history of economics. A first major contribution to the development of multiple criteria decision analysis was given by Pareto in 1896 in his trade-off analysis which generated the concept of Pareto-optimality. Pareto was looking for a criterion to judge a certain distribution of goods among people. In his opinion, interpersonal utility comparisons were hardly possible, so that an unambiguous social welfare function approach was not very meaningful (see Tarascio, 1986). Thus, he ended up with the much weaker concept of Pareto-optimality: a distribution of goods among people is Pareto-optimal, when it is impossible to improve a given person’s utility performance without making other people worse off. Clearly, the Pareto-optimality concept is closely related to the efficiency concept which plays a major role in multiple criteria decision analysis.

In the previous chapter it has been shown that in the past decade a wide variety of multiple criteria evaluation methods has been designed, which aimed at structuring, systematizing and judging complex decision problems marked by multiple appraisal criteria. In this period, the general principle for rationalizing such complex choice and trade off problems was based on a straightforward approach: given (i) a certain well-defined evaluation problem and (ii) a certain specific evaluation technique, the aim is to identify the most plausible outcome for the decision problem concerned.

Cities are a source of economic activity, and house at the same time a wealth of historico-cultural heritage which needs to be protected. In many cities a conflict has arisen between the aim of reorienting the urban structure towards a more modern and efficient spatial lay-out and the need to preserve the historico-cultural heritage (see e.g. Nijkamp, 1988). Various aspects of the historico-cultural heritage are hardly measurable in cardinal units, so that ‘soft’ evaluation tools have to be used. In the present chapter a numerical illustration of conflict analysis in this area will be given, based on a qualitative sign analysis (see also Nijkamp, 1980). The aim is to present a simple method in case of qualitative information on impacts of new urban infrastructure plans.

In the practice of transportation planning, cardinal information (measured on a ratio scale) is often not available, so that then conventional evaluation methods — either monetary-based methods (such as cost-benefit analysis, cost-effectiveness analysis or shadow project methods) or cardinal multiple criteria methods (such as goals-achievement analysis or weighted summation analysis) — cannot be applied as methodologically sound tools for decision analysis. In the past years a wide variety of qualitative multiple criteria choice methods has been developed (see also Chapter 11). Examples of classes of adjusted methods are: survey table methods (for instance, score card methods and computer-graphic methods); interactive computer methods (for instance, based on an interplay between expert and decision-maker); weighted methods (based on a set of weights reflecting the relative importance attached to the successive value criteria).

In recent years, airline policies have become a focal point of transportation research, mainly because in this field deregulation principles have been experimented and applied in many countries. The liberalization of airline policy has had tremendous impacts on air fares, competitive positions, level of services and/or capacity of carriers between different airports (cf. Banister and Button, 1991; Banister et al., 1992; Barrett, 1990).

Transportation planning is not only concerned with moving vehicles and network infrastructure (such as roads or railways), but also with point infrastructure (such as parking lots). In the present section an empirical application will be dealt with, which addresses the issue of an optimal location of an urban parking lot. The case study concerns the city of Enkhuizen (in the province of North-Holland in the Netherlands). The city houses an interesting museum on the history of the interior lake of the Netherlands (currently named the IJsselmeer, but in former times the Zuiderzee). The construction and opening of a new exterior part of this Zuiderzee-museum — located near the border of the IJsselmeer — required sufficient parking facilities for private cars. The city itself is an extremely interesting old place characterized by an impressive architectural and historico-cultural heritage which deserves strict protection, so that parking policy in this city does not only have a transport aspect but also a conservation aspect. The city is also a centre of tourism, with a strong orientation towards water sports (e.g., sailing).