In outlining some of the many problems associated with numerical prediction of car ownership, it has been possible to concentrate on the debate surrounding the different development procedures. Although this book has concentrated on the detailed technical aspects of forecasting it should not be forgotten that there continues to be widespread scepticism regarding the basic philosophy of mathematical modelling. It is unlikely that lay and professional scepticism, often resulting from misapplication of forecasts at second hand, will be reduced in the 1980's. It is now clearly evident that causal models may appear more scientific and be less vulnerable to outside criticism, but they are heavily dependent upon accurate forecasting of at least one independent variable. Only recently Mogridge and others commented upon the high level of licence evasion in the UK. Average mileage statistics require careful interpretation and tend to conceal significant changes in the pattern of vehicle use.