Evaluating Regional Economic Impacts from Heavy Vehicle Fees - a Comparison of Different Approaches
O Kveiborg, Danish Transport Research Institute, DK
The paper describes three model approaches to evaluate economic impacts from heavy vehicle fees; System dynamics modelling, Input-output base models and Spatial CGE models. Three practical approaches using IO and SCGE models are then discussed.
Many countries have introduced different types of distance dependent heavy vehicle fees (HVF). The primary objective for this introduction has been to raise revenues for infrastructure financing (e.g. Germany and Switzerland) or to regulate the number of vehicles on the main network (e.g. Austria). The interest in infrastructure financing and traffic regulation has let analyses of the different schemes to focus on the traffic impacts from which revenues can be calculated. The fees collected are in all cases substantial. This raises two additional questions: who is paying the fees and what are the impacts for them, and what are the impacts of the use of the revenues.
These questions cannot be answered easily due to the following:
? In the first instance it is the carrier that has to pay the charge, but he will pass the bill to the transport buyers. The transport cost change induced by the introduced HVF is thus spread to all different industries either directly because they use freight transport to convey their goods to and from the market, or because they purchase goods from other industries that have paid the fees. The impact from the vehicle fees is therefore transmitted to the entire economy.
? The introduced fees further raise substantial revenue (even after the transaction costs are deducted). The use of this revenue can have impacts of similar magnitude on the regional economies depending on how it is spent.
A way of addressing the evaluation of the economic impacts across industries and regions is by using a spatial economic model. Many such models exist, but they have only to a very limited degree been applied to analyses of the relation between HVF and the regional economies. This is interesting because the same models have to a very large extent been applied to analyses of the regional economic impacts of large infrastructure investments. The impacts of infrastructure investments and HVF are very likely to be of similar magnitude even though HVF may have a wider regional distributed impact. It is thus important to evaluate the HVF in a similar way.
The paper describes three overall model approaches that can be used to address the question of regional economic and industry impacts. These are regional input-output (based) models (IO), systems dynamics models (SDM) and spatial general equilibrium models (SCGE).
The models are described with focus on understanding why and how the models can be used to evaluate regional economic impacts of HVF. The IO and SCGE models are described independent of actual model examples because both approaches are seen as more general theoretical approaches with many different practical examples. This is in contrast to the SDM models, which best can be described as a complex of independent models. To illustrate the properties of a SDM the Transeuropean ASTRA model is used. The ASTRA model contains a number of submodels including a traffic model, a macro economic model, a regional economic model and an environmental model. For the specific purpose in this paper it is the interrelation between the regional economic model and the traffic model that is of main interest.
The important properties of the three different model approaches are discussed and compared with respect to the regional economic evaluation task. A further evaluation of the approaches is completed through practical cases. The practical cases include a German IO model approach, a Norwegian SCGE approach and a Danish IO based SCGE model. The latter model does not contain some of the complex non-linear features often used in SCGE models, but is contains a high industry and regional differentiation. It has not been possible to find a SDM approach to this specific question so no practical illustration is possible here.
It turns out that the practical implementation of the different models shows rather similar types and magnitudes of impacts. It is not surprising that all the models are capable of showing similar types of impacts as these are directly modelled in all cases. However, it is quite surprising to see the similarities in the magnitude of the impacts. One would suspect that the highly linear and non-flexible IO approach would overstate the behavioural responses of the industries, but this does not seem to be a very big problem. Some variation can be found though. Some of this can be related to the model type and some of it to the specific case being studied.
Generally it is concluded that the SCGE approach is the best suited for the evaluation because it contains more of the complex relations involved (non-linear relations, a theoretical basis, specific focus on the supply and demand relations).
Association for European Transport