A comparative assessment of synthetic indices to measure multimodality behaviours

The study of how people jointly use different travel means is one of the key issues in contemporary transport research. However, measuring multimodality behaviours presents some intricacies that deserve more attention in order to come up with an instrument that is effective both on a modelling and on a policy viewpoint. The present work considers some methods that have been proposed in different disciplinary ambits to measure diversity and assesses to what extent they are useful to measure multimodality. A broad set of indices is then analysed, ranging from welfare economics (Gini, Dalton and Atkinson indices) to information theory and ecology (entropy, Herfindahl index). Theoretical investigations and empirical experiments on the properties of such indices show that there is not a measure of multimodality that consistently outperforms all the others in any circumstance. On the other hand, it emerged that some methods are clearly preferable for specific problem instances, as discussed in the conclusions.