This chapter has summarized theoretical and empirical developments in automobile demand and type choice modelling that have emerged over the past 20 years in response to the needs of transport modellers who provide analytical tools to support policy analysis decision-making at all levels of government. Models of auto-ownership level and auto-type choice are major components in integrated forecasting and microsimulation systems, thus becoming part of the ongoing debate concerning the tactical or strategic role such tools should play in making policy decisions. The major emphasis has been on developing approaches that are theoretically consistent with individual/household-level behaviour, and researchers have striven toward ever increasing "behavioural realism" as a means of advancing the state of the art. In this regard, advances in auto-choice modelling have been intimately linked with the development of discrete-choice models within a RUM framework. The move toward more behavioural realism is likely to continue, but increased data requirements and other costs will be major issues. More generally, the trend towards modelling all aspects of transportation behaviour as part of an integrated analysis of household activity patterns requires much more detailed datasets. Moreover, the explosion of information technology as we move into the new millennium portends a complex future filled with alternative "new mobility options" (e.g., telecommuting, "virtual mobility" through e-commerce, car sharing) that will all interact with automobile ownership and usage decisions in a fundamental way. Gaining understanding of quickly changing markets and new behaviour patterns will provide additional challenges for, e.g., models using stated preference data. Finally, the vast majority of modelling efforts have focused on the demand side, but more recently there has been increased attention on modelling the supply side as well, and taking into account market equilibrium effects. This, and related issues of vehicle scrappage and used-vehicle price equilibration have not been addressed in any detail here (see, e.g., Berkovec, 1985). Addressing all these issues could be costly. However, current rapid advances in information technology may hold the key to solving these problems.
Abstract