The Survey of Motor Vehicle Use (SMVU) is the principal source of information on registered motor vehicle use in Australia. Several Australian transport analysts and organisations adjust the SMVU results with the aim of producing consistent time series estimates of vehicle use. These efforts sometimes produce conflicting estimates, presenting a problem for uninformed practitioners who wish to quote measures of motor vehicle use. This paper provides an overview of major changes to the survey methodology between 1971 and 2002 and describes the results of several Monte Carlo simulations designed to assess the impact of some of these methodological changes. The authors suggest that there are only a handful of factors that analysts need to adjust for in constructing consistent long-term time series estimates of vehicle use. The post-1998 survey results should be more reliable than the earlier results. Simulation results imply that the current survey design and other characteristics should produce reliable estimates of registered motor vehicle use. (Author/publisher) For the covering entry of this conference, please see ITRD abstract no. E211825.
Samenvatting