Analysis of the influences of built environment measures on household car and motorcycle ownership decisions in Hubli-Dharwad cities

This paper investigates the impacts of objective and subjective built environment measures on household car and two-wheeler ownership levels in Hubli and Dharwad cities in India. The empirical analysis bases on primary household vehicle ownership data and secondary land use information. The study accommodates self-selection due to travel attitudes. An important finding of the study is that households’ perceptions about neighborhood attractiveness and cleanliness are significant predictors of car and two-wheeler ownership decisions in both cities. The correlation of this latent variable is strong in Dharwad than in Hubli, probably due to the less urbanized context of Dharwad when compared to Hubli. The land-use diversity indicator improved its statistical significance in motorcycle ownership models after controlling for travel attitudes in Hubli, while in Dharwad, the coefficient's magnitude worsened. In Dharwad, road network density and distance measures revealed ‘causal’ impacts on motorcycle ownership levels, even after controlling for travel attitudes. Distance to hospitals and bus stops are highly important for a Dharwad resident's motorcycle ownership decisions. The models show that self-selection is not at work in car ownership decisions in Hubli, and built environment variables have independent effects. Among the built environment variables, land use diversity and population density can decrease car ownership levels in Hubli city. Besides, with an improvement in the women travel friendliness of neighborhoods in Hubli, households are less likely to increase car ownership levels. In Dharwad, none of the objective built environment attributes remained significant after controlling for travel attitudes, suggesting self-selection effects. The statistical relevance ‘children-friendliness’ factor was unaltered due to the addition of attitudinal variables.