Conventional travel survey data, whether household- or traveller-based, are scarce in developing countries, and it is suggested that household expenditure surveys, which are relatively common, be used instead. Although many of the traditional topics of travel surveys cannot be addressed with expenditure data, many others can, and there are compensating advantages of coverage and scope. This concept is illustrated with data from household surveys of Sri Lanka, Hong Kong, Thailand, and Tunisia. Broad regularities in travel expenditures are discussed, both within and across countries, and for the Tunisian survey, a more detailed regression analysis is presented that focuses on the interrelations between travel expenditures and vehicle ownership.
Abstract