The report explores a number of issues that have emerged with the publication of the most recent TRRL report on car ownership forecasting, reports on the car ownership element of the regional highway traffic model, and the report of the advisory committee on trunk road assessment. These issues include the form of model to be used, the numerical assumptions and definitions of variables used in calibrating the models, and the methods of calibration. One general conclusion is that, for a given model structure, the forecasts produced depend to an appreciable extent on the precise way in which such quantities as incomes and motoring costs are defined, as well as on the views taken on such matters as saturation levels and elasticities. Available data do not give clear indications of the most satisfactory choices. Better understanding of causal mechanisms is necessary to resolve the differences between the presuppositions made in current disaggregate cross-section models and current time-series-based aggregate models. (Author/publisher)
Samenvatting