Fowkes, A.S., Nash, C.A., Toner, J.P. and Tweddle, G (1993) Disaggregated Approaches to Freight Analysis: A Feasibility Study. Working Paper. Institute of Transport Studies, University of Leeds , Leeds, UK.
Available under licence : See the attached licence file.
Forecasting the demand for freight transport is notoriously difficult. Although ever more advanced modelling techniques are becoming available, there is little data available for calibration. Compared to passenger travel, there are many fewer decision makers in freight, especially for the main bulk commodities, so the decisions of a relatively small number of principal players greatly influence the outcome. Moreover, freight comes in various shapes, sizes and physical states, which require different handling methods and suit the various modes (and sub-modes) of transport differently.
In the face of these difficulties, present DTp practice is to forecast Britain's freight traffic using a very simple aggregate approach which assumes that tonne kilometres will rise in proportion to GDP. Although this simple model fits historical data quite well, there is a clear danger that this relationship will not hold good in the future. The relationship between tonne kilometres and GDP depends on the mix of products produced, their value to weight ratios, number of times lifted and lengths of haul. In the past, a declining ratio of tonnes to GDP has been offset by increasing lengths of haul. This has come about through a complicated set of changes in product mix, industrial structure and distribution systems. A more disaggregate approach which studies changes in all these factors by industrial sector seems likely to provide a better understanding of the relationship between tonne kilometres and GDP.
However, there are also problems with disaggregation. As we disaggregate we get more understanding of what might change in the future, but are less able to project trends forward. This can be seen if we consider the future amounts of coal movements. Theoretically there is clearly scope for better forecasting by allowing for past trends to be overturned by a movement towards gas powered electricity generation and more imports of coal direct to coastal power stations. However, making such a sectoral forecast is extremely difficult, and inaccuracy here may more than offset the theoretical gain referred to earlier. This is because it is usually easier to forecast to a given percentage accuracy an aggregate rather than its components. For example, the percentage error on sales forecasts of Hotpoint washing machines will be greater than that for the sales of all washing machines taken together. This occurs because different makes of washing machines are substitutes for each other, so forecasts for Hotpoint washing machines must take into account uncertainty over Hotpoint's market share as well as uncertainty over the future total sales of washing machines. Nevertheless, a disaggregate investigation of the market could spot trends which were `buried' in the aggregate figures. For example, rapidly declining sales for one manufacturer might indicate their leaving the market, which with less competition would then price up and so reduce the total future sales.
We have assumed above that the use of the term disaggregate in the brief refers to disaggregation by industrial sector. An alternative usage of the word disaggregate in this context is when referring to modelling at the level of the individual decision making unit. Disaggregate freight modelling in this sense would involve analysing decisions in order to determine the utility weight attached to different attributes of available transport options. Because data on suitable decisions is not readily available in this country, due to commercial confidentiality, we have recently undertaken research in which we have presented decision makers with hypothetical choices, and obtained the necessary utility weights from their responses. Whilst initial scepticism is understandable, this method has produced results acceptable for use in major projects. ITS itself has provided algorithms (known as Leeds Adaptive Stated Preference) which have been used to derive utility weights for use by British Rail in forecasting cross-channel freight, by DTp in evaluating the reaction of commercial vehicles to toll roads, and by the Dutch Ministry of Transport in modelling freight in the Netherlands. In the light of the above, the following objectives were set for the feasibility study:
(1)To determine if a forecasting approach disaggregated by industrial sectors, as under the first definition above, can be used to explain recent trends in freight transport;
(2)To test the feasibility of the disaggregated approach for improving the understanding of likely future developments in freight markets, this being informed by current best understanding of the disaggregate decision-making process as under the second definition above.
|Item Type:||Monograph (Working Paper)|
|Copyright, Publisher and Additional Information:||Copyright of the Institute of Transport Studies, University Of Leeds|
|Institution:||The University of Leeds|
|Academic Units:||The University of Leeds > Faculty of Environment (Leeds) > Institute for Transport Studies (Leeds)|
|Depositing User:||Adrian May|
|Date Deposited:||24 May 2007|
|Last Modified:||20 Jul 2014 05:10|
|Publisher:||Institute of Transport Studies, University of Leeds|
|Identification Number:||Working Paper 399|