IMPLICATIONS FOR FORECASTING: FUTURE RESEARCH

The development of Discrete Choice Modelling in the 1970s made it clear that the use of generalised cost in transport models was essentially equivalent to a (negatively) scaled estimate of the 'utility' of relevant alternatives. This utility formulation is not restricted to linear functions. This note considers briefly the implications of this departure from linearity for transport models. The literature of value-of-time (VOT) research has extensively discussed the possibilities for variation with the size and sign of the time change. It is generally agreed that it is reasonable to suppose that the unit valuation rate might be lower for gains than for losses ('sign' effect) and for small changes than for large changes ('size' effects). In the 'relativity' effect, the valuation of a given change may decrease as the base journey to which it applied becomes longer. The note first discusses these three effects, and considers that the size effect is probably the most controversial, for example, due to difficult questions about thresholds. It then discusses the methodological issues of 'willingness to pay' and stated preference in relation to 'inertia', and some key forecasting issues: 'path dependence', non-reversibility, and non-additivity. It finally makes some tentative proposals. For the covering abstract, see IRRD E100587.

Language

  • English

Media Info

  • Features: References;
  • Pagination: 9 p.

Subject/Index Terms

Filing Info

  • Accession Number: 00764334
  • Record Type: Publication
  • Source Agency: Transport Research Laboratory
  • Files: ITRD
  • Created Date: May 28 1999 12:00AM