AN ECONOMETRIC ANALYSIS OF AUTOMOBILE TRANSACTIONS

Most models conceal the decision making process of households; the author describes the use of a model based on household intertemporal choice activities, providing reliable short-term forecasts which when periodically updated lead to improved long-term predictions. A random utility minimisation framework is used to characterise a household's transactions in its automobile market. Multinomial logit analysis is adopted to eliminate the model and the results indicate that household behaviour depends upon the base period level of attributes, including income, number of licensed drivers and the number of newer vehicles in the household stock, and also short-term variations in factors relevant to the choice environment including prior state occupancy and brand loyalty. A simplified model, requiring less data is found to approximate the more general model quite well. Forecasting characteristics of the simpler model are examined showing that short-term predictions are significantly improved if prior household decisions on ownership levels are considered. (TRRL)

  • Availability:
  • Corporate Authors:

    IJTE

    8 VIA G A GUATTANI
    Roma, 00161, ZZ Italie   
  • Authors:
    • McCarthy, P S
  • Publication Date: 1985-2

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00451055
  • Record Type: Publication
  • Source Agency: Transport Research Laboratory
  • Files: ITRD, TRIS
  • Created Date: Aug 27 2004 9:57PM