This paper describes one aspect of a project that is being undertaken as a three year SERC CASE Award with The Rover Group. It describes the work carried out prior to and the calibration of what is believed to be the first behavioural model of GB car scrappage for 30 years. Since the model is to be used to forecast the number of cars scrapped in a year of inputting to a medium term model of new car purchases, evidence to support this modelling approach is briefly outlined which draws upon an understanding of the process of car scrappage in Great Britain. The theory behind this process assumes that owners scrap their cars when faced with an uneconomic repair; this could be initiated y either a mechanical breakdown, or the car failing its MOT. Likewise, the decision rule they are assumed to use in deciding whether or not it is economic to repair a car is that it should be scrapped if either its value once repaired, or the cost of buying a replacement car, is less than the sum of the repair cost of the car's value as "scrap". Both of these assumptions are investigated empirically to ensure that the model calibrated truly reflects both the events and decisions that take place in practice. From this background the paper then outlines the specification and calibration of the behavioural model. Wherever possible, the wider implications are drawn from this analysis for both new registrations and changes in car ownership, thereby making the paper of more general interest.
Abstract