Individuals in households with higher incomes spend more money on bus travel than those from lower income households, but it does not follow that as people get richer over time they will travel more by bus. This study looks at some cross-sectional data on bus travel as a function of household income, and the way in which it changes over time, with the intention of seeing whether a demand elasticity with respect to earnings (after adjustment for inflation) can be estimated. The data used have come from two sources, the annual family expenditure surveys (FES) and the national travel surveys (nts) and have considered mainly non-car-owning households. No consistent pattern emerged from the analysis to enable a demand elasticity with respect to earnings to be fixed with precision. The only statistically significant results using the FES data suggest that the earnings elasticity was negative, which is counter-intuitive. Using the nts data, non-significant results were obtained except when the elasticity of bus travel with respect to the level of service (as measured by the number of bus-kilometres provided) was assumed to be perhaps implausibly large at 0.9, or when the earnings elasticity was assumed to vary with household income, but in the latter case the variation between low and high income households appeared to be implausibly large. Excluding the elderly from the sample of households, because of their concessionary fares, did not improve the level of significance of the earnings elasticity. Similarly, analysing households by household structure (for example 2-person households) or by population density did not alter the picture. Overall, consideration of these different analyses, and of changes in trip rates over time as found in the nts data, suggests that the effect of earnings on public transport use can be represented by an elasticity which is probably very small, and with an upper limit of 0.4. (Author/publisher)
Abstract