A new type of transport model UMoT (unified mechanism of travel) has been developed by Zahavi for the World Bank (and subsequently for the US Department of Transport and the Federal Republic of Germany Ministry of Transport) as an aid to urban transport planning. The feature which distinguishes UMoT from practically all other transport models is the way it takes into account known regularities in the amount of time and money which people spend daily on travel (these are commonly referred to as travel budgets). the relationships which express the travel budgets in terms of demographic variables are expected to be transferable from town to town and so little, if any, calibration of the model is necessary. The aim of this special programme of study is to see how the UMoT model responds to changes in the major parameters and to compare predicted travel characteristics with values observed in reading, UK. Both the original version of the model (set up for two modes - car and bus) and its extension to three modes (car,bus and walk) gave a fairly satisfactory level of fit when the model outputs were compared with the observed daily travel distances (discrepancies varying from 1 to 33 per cent) and car ownership values (1 to 4 per cent). sensitivity analysis, however, showed that some of the elasticity values had signs which were counter-intuitive on current evidence and experience. For example, higher bus fares produced more bus travel and less car travel - as did faster car speeds. (car speed elasticity was based on a change in access time, not network speed.) it is understood from private correspondence with Dr. Zahavi during the preparation of this report that the latest version of the model gives elasticities with the expected signs except for the bus fare elasticity, which is still counter-intuitive unless one accepts that higher bus fares over the long term will results in more, not less, bus travel. (Author/publisher)
Abstract