Follow-up of Swedish Freight Transport Forecasts Since 1975
Inge Vierth, VTI, Magnus Landergren, VTI
This paper looks at the eight Swedish freight transport forecasts that have been created using different economic and network models.
The Swedish Transport Agency and its precursors makes forecasts of the long-term development of passenger and freight transports. These official forecasts are used to assess the need for infrastructure investments and to calculate the socio-economic consequences of different policy measures, among other things.
A research program (that will be finalized in the first half of 2016) follows up the in total 15 passenger and freight transport forecasts that have been published between 1975 and 2009. Main research questions are:
1) Have transport forecasts been systematically wrong when it comes to the overall development?
2) How much of the deviations can be attributed to false assumptions about input variables and how much to incorrect elasticities between the variables?
3) Which guidelines can be given to future forecasts?
We concentrate on evaluating the eight freight transport forecasts that have been created using different economic and network models. Tonne-km forecasts, total and by mode, are compared to transport statistics. This is done at the national level and as far as possible at the regional level. The results so far indicate that total tonne-km, road tonne-km and sea tonne-km conform for most of the forecasts. However, six out of eight forecasts for rail transports overestimated the development.
The official tonne-km forecasts (total and by mode) are compared to alternative forecasts that have been created using linear extrapolation. The results so far indicate a comparable or better fit for the alternative method.
Lack of reliable statistics for road transports (quality problems due low response rate for transports performed by Swedish trucks and limited information about the transports performed by foreign trucks) obstructs the possibilities to follow up the forecasts.
It is shown that the exploitation of economies of scale leads to different developments for tonne-km and vehicle-km. This is true for all modes. Ideally forecasts should be calculated both for tonne-km and vehicle-km.
Association for European Transport