Measuring travel behaviour change

Increasing interest in 'soft' policy approaches to demand management, initially in personal travel and now in water and power usage, poses the question of how to measure the effectiveness of interventions. Much of the focus has been on the statistical reliability of measured change, where sample surveys are the primary means of estimating the change. Sample surveys also pose methodological issues that can introduce non-sampling errors, especially when the 'measure' is the difference between two estimates ('before' and 'after'). In some cases, direct measurement of usage appears less difficult - both electricity and water consumption are metered in Australia. The challenge in these cases is to establish the extent to which changes in consumption are real and to what extent they are the result of factors such as seasonal differences (either between seasons or from year to year) or demographic change in the household. This paper outlines the principles and pitfalls in measuring behaviour change. It draws on experience with voluntary travel behaviour change, using a number of approaches, including but not limited to Individualised Marketing. A key issue is the extent to which repeated experience can validate the effectiveness of voluntary behaviour change interventions in general, despite statistical errors of individual measurements. Experience demonstrates that the scale of the intervention is important. Interventions with more than 5000 households are consistently more successful than small ones, even allowing for the greater statistical variability of measurement for smaller projects. Large-scale also offers opportunities for intervention design to benefit from the potential for diffusion beyond those directly involved in the project.