This report presents the results of a range of tests of organised car-sharing schemes. The performance of the schemes is predicted using a sophisticated microsimulation model. A brief resume of the model is followed by a description of the tests and analysis of their results. A comparison is made between the model predictions and empirical evidence of the performance of car-sharing schemes, and the implications of organised car sharing within broader transport policy are discussed. Significant results include the following: a typical scheme would draw applications from 8 per cent of trip makers; a matching system could provide match lists for up to 90 per cent of applicants provided that the pool of applicants was large (>1000) or that they shared broadly the same origin and destination; unless aimed at specific groups of commuters, small schemes are likely to be extremely ineffective; of people who actually become car sharers, almost half will previously have been public transport users; most car-sharing would involve simple lift-giving rather than true car pooling; a large municipal scheme might result in 1 1/2 per cent of trip makers becoming car sharers; this would result in a reduction in peak period public transport patronage of about 1 1/2 per cent and a reduction in private vehicle mileage of about 0.3 per cent; a scheme based on a group of long-distance commuters might result in 5 per cent of them becoming car sharers; a doubling of both fuel prices and public transport fares would increase the impacts of the scheme by a factor of two, while the provision of free reserved parking for car sharers would increase considerably the number of participants but would have very little effect on private vehicle kilometres travelled. In summary, car-sharing schemes are unlikely to have more than a marginal effect on congestion, parking requirements or energy use, but they can confer considerable benefits on their participants and they may have a significant effect on public transport. (Author/publisher)
Abstract