The report examines the suggestion that the total amount of time, money or generalised cost spent on personal travel may be constant over time, in cross-section or under changes of travel costs or policies. Invariant expenditure of either time or money on travel is inconsistent with rational economic behaviour and with conventional transport modelling. A theoretical analysis concludes that generalised expenditure on travel might be invariant. Data from the national travel survey indicate that generalised expenditure is much the same in urban as in rural areas, despite the wide variation in modes available and distances to destinations. In comparisons of households at different income levels, generalised expenditure per person is almost directly proportional to gross income per person; in other words, consumption of 'generalised time' is constant. Time series data from 1953 to 1976 are also examined and, though the data are poor, they indicate that generalised travel expenditure per person has increased over the years appreciably faster than has real income per person. Caution is therefore needed in using constant budget concepts in forecasting models. (Author/publisher)
Abstract