A number of studies of the interaction between car ownership and the availability and use of public transport in Great Britain and the united states are presented, based on time series, household level cross-sections and area-based cross-sections. The household level cross-sections show that public transport use increases with increasing income, at a fixed level of car ownership, but decreases with increasing car ownership. When car ownership levels in different areas are compared with public transport levels in the same areas, strong negative correlations are usually obtained. Usually these remain strong even when allowance is made for a number of non-transport variables. One study, however, shows a slight positive correlation. Some of the studies were designed to show particularly the effect of public transport level on car ownership, and these suggest that when for extraneous reasons public transport levels vary over a wide range, ownership may as a result vary by about 0.05 cars per person, mainly associated with second cars. Causality in the reverse direction is shown most clearly by time series data and by household level cross-sections. The effects in the two directions combine to produce the area-based correlations. (Author/publisher)
Abstract