The effect of land use and residential self-selection on automobile travel in the greater Sekondi-Takoradi area of Ghana using structural equation model

Despite the suggestions by numerous studies over the years that spatial land use pattern influences travel behaviour, uncertainty remains about the existence, form, and the extent of any causal relationship between spatial land use and urban travel behaviour. This study seeks to provide a new dimension from the perspective of a medium-sized city in Africa by simultaneously incorporating attitudinal and socioeconomic variables in examining the relationship using structural equation model (SEM). Primary data collected using questionnaires from 534 households of adult commuters with private cars and spatial data of the built environment for households in the Greater Sekondi-Takoradi Area (GSTA) in Ghana were used in this study. Principal component analysis (PCA) was used to reduce 25 statements attributed to travel attitude to four. Similarly, 22 statements attributed to neighbourhood preferences were also reduced to three. Latent constructs of four and 20 observed variables were analysed using SEM in LISREL 10.3 software. The fitness of the final model to the data was assessed using absolute, incremental, and parsimonious fit indices. The findings from the study indicated that socio-economic factors had a stronger influence on households’ automobile travel than self-selection, with land use having the least impact. The results are consistent with earlier studies that suggested self-selection has a greater impact than land use on households’ automobile travel. This suggests that land use policies can effectively be used in controlling households’ automobile travel when the effect of self-selection is simultaneously accounted for with the land use.