Distribution and Modal Split Models for Freight Transport in The Netherlands
G de Jong, Significance/ ITS Leeds/CTS Stockholm NL; A Burgess, R Versteegh, NEA, NL; L Tavasszy, TNO and Delft University of Technology, NL; M de Bok, Significance, NL; N Schmorak, Dutch Ministry of Infrastructure, NL
New models for distribution and modal split in Dutch freight transport were estimated. The paper will discus the model structure, estimation results for different specifications and elasticities.
Models play a crucial role in the complex realm of policy and decision making in The Netherlands. In an effort to improve the information provided to policy makers and to maximise the added value of the existing instruments and available knowledge and data, the Dutch Ministry of Infrastructure and Enviroment decided to work towards a new freight transport model. The first step in the building process is a basic model called BASGOED. This model is meant as fundamental corner stone for an incremental building process leading to a modular, transparent and flexible set of instruments for policy making in the area of freight transport.
For this purpose, BASGOED has been designed as a basic model, satisfying the basic needs of policy making, based on proven knowledge and data that will also be available in the future. BASGOED will therefore make use of those modules and data from the existing SMILE+ model, that are suitable for the new set-up. This is the case for the Economy Module which provide transports flows in tonnes by commodity type generated and attracted in each zone. New models have been estimated and implemented for distributing these flows to origin-destination pairs and for determining the modal split (road, rail and inland waterways). These two new models will be presented in the paper.
Both the distribution and the modal split model are based on the national Dutch freight transport database of 2004 ("Basisbestand Goederenvervoer"). Furthermore, transport distances and transport times from existing uni-modal road, rail and inland waterway transport network models were used as input (also taking account of road access to and egress from train stations and inland ports). The commodity classification used is NSTR1 (10 commodity groups) and the zoning system distinguishes 40 zones in The Netherlands and 29 zones abroad.
These models are aggregate models (e.g. the modal split model is an aggregate logit model). A recent revealed preference data base at the disaggregate level, such as a shippers survey or a commodity flow survey is not available in The Netherlands.
Various functional forms have been tested for the model specification in both models. This included the specification of transport cost and transport time, but also the influence of border resistance, zone-internal transport, containerisation and the residual influences of specific origin and destination zones or origin-destination combinations.
The link between the distribution model and the modal split model is a logsum variable that is calculated over the modes in the modal split model using the coefficients estimated for this model, and then converted to the transport resistance variable in the distribution model.
The estimated models were implemented in a computer code in a scientific programming language. Elasticities for changes in time and cost by mode have been calculated using both the estimation data and the implemented model (that also includes transport flows for which only one mode is available). In the model, changes in transport time and cost will not only affect the mode shares, but also the distribution.
The paper will discuss the structure of the overall model and of the submodels for distribution and modal split in particular. It will also contain a description of the model inputs and outputs and of the estimation results for different specifications. After this, the elasticities will be reported in the paper, and discussed on the basis of a comparison with the literature (using recent international overview studies of freight transport elasticities some of the authors were involved in).
Association for European Transport