Modelling the Factors Which Influence New Car Purchasing



Modelling the Factors Which Influence New Car Purchasing

Authors

PAGE M, WHELAN G and DALY A, ITS, University of Leeds, UK

Description

Models of new car pumhasing behaviour are important in order to predict what types of new car might be purchased in the future in response to changing circumstances. This forms an important input into the modelling of the evolution of the car fleet.

Abstract

Models of new car pumhasing behaviour are important in order to predict what types of new car might be purchased in the future in response to changing circumstances. This forms an important input into the modelling of the evolution of the car fleet.

The UK Department Of the Environment, Transport and the Regions (DETR) have a transport model which was developed to improve the modelling and forecasting of fuel consumption and CO2 emissions. The model is called the Vehicle Market Model (VMM) and it incorporates a model of the GB vehicle fleet disaggregated by vehicle type, engine size, fuel type and ownership type. The most recent version of this model was developed in 1998 (Kirby et al, 2000). The VMM is used by the DETR to model the effects of various policy levers such as fuel price increases, changes in vehicle taxation and changes in technology.

The DETR commissioned the Institute for Transport Studies (ITS) to develop a model which could be used to forecast the future distribution of new car sales. It was anticipated that the new model would provide an improved input to the existing VMM and therefore had to be consistent with the disaggregations of the vehicle fleet used in the VMM. The new model therefore had to split new car sales by:

* engine size (<700cc, 700-1000cc, 1001-1200cc, 1201-1500cc, 1800cc, 1801-2000cc, 2001-2500cc, 2501-3000cc, >3000cc)

* fuel type (petrol/diesel); and

* ownership type (retail/private or fleet/company)

The survival rate submodel within the VMM provides an estimate of the overall number of new cars sales so the model developed in this project only had to provide an estimate of the proportion of this figure in each of the 36 different disaggregations (9 engine sizes by 2 fuel types by 2 ownership types).

The objectives of the project were therefore:

* To improve knowledge of the factors that influence people's decisions when they buy new cars

* To develop a computer model to forecast the future distribution of new car sales

Publisher

Association for European Transport