University of California Transportation Center

In this study using data from Sylhet, Bangladesh, transportation expenditures are examined in detail for different income groups living in different natural and built environments. We employ Tobit models and a dataset of more than 22,000 persons in 4162 households to correlate transportation expenditures by separate income groups to land use, transportation and other infrastructure, as well as household and worker within household characteristics. Among other findings, urban high income households are not negatively affected by their environments. They have the ability to increase their transportation expenditures to reach most destinations and have access to almost all opportunities. This is vastly different from the urban poor; they are restricted by distance because most of their trips take place on foot. Moreover, rural workers are less affected by environmental indicators, presumably because they are restricted by their location and the lack of road networks near their residence. The lack of modal choices may be the determining factor on transportation expenditures. Lack of basic services increase expenditures, showing a link between underdeveloped or temporary housing and lack of near work opportunities.