Exploring travel behavior and multimodal accessibility in Dhaka

Abstract

A global paradigm shift from mobility-geared to accessibility-oriented planning is observed in recent years due to the ability of accessibility-oriented planning to integrate land use aspect with transportation planning and consider the 'derived demand' nature of travel. Since accessibility, defined here as the ease of reaching desired destinations, is a measure of potential travel, a large body of studies have been conducted to explore whether it influences actual travel behavior. However, such studies are extremely limited in the context of developing countries and almost non-existent in the context of Bangladesh. The multimodal nature of public transport travel has also received little attention. In this backdrop, the aim of this thesis is to explore and quantify multimodal accessibility and its effects on travel behavior in Dhaka city and synthesize its research and policy implications. Accessibility has been considered both spatially and temporally in this study. In doing so, the study has considered the data-deprived context of cities of developing countries like Dhaka where traditional methods of incorporating temporal dimension in accessibility analysis is hardly possible. To develop an alternative approach to integrate temporal dimension in accessibility studies, the study has used the simple observation that trip purpose, time-of-day and trip destinations are always interlinked in an urban area. Using spatial autocorrelation approach, the statistically significant destination clusters of Dhaka by peak and off-peak hour were identified. Calculating Modal Accessibility Gap (MAG) showed that the major destinations are highly accessible by car compared to public bus irrespective of any time of the day. There is a general consensus that socio-economic and demographic, trip, and built environment characteristics influence travel behavior although this influence is far from homogenous. Especially developing countries do not necessarily correspond to evidence found from developed countries particularly when it comes to the effects of built environment characteristics. Binary logistic regression models revealed statistically significant effects of several built environment attributes including accessibility to major destinations on mode choice and household car ownership in Dhaka. This indicates the potential for enhancing public transport accessibility, and future infrastructure investments oriented towards those locations. Alongside that, indications were also found that built environment alone will not be able to bring any drastic change in travel behavior unless mobility-limiting strategies complement accessibility-enhancing strategies and land use planning. The study findings confirm that accessibility could be a powerful tool to impact travel behavior and there exists a solid basis to evaluate the extent to which accessibility and travel behavior have been meaningfully integrated in the policy and planning documents in Dhaka’s context. Reviewing the five most relevant and widely cited policy and planning documents revealed that there exists significant lack of clarity in defining concepts like mobility, accessibility and connectivity. The documents do not explicitly incorporate accessibility and accessibility-related goals and objectives have not been translated to quantifiable measures or performance indicators. In a nutshell, this research suggests that accessibility is crucial in travel behavior decisions in Dhaka and therefore, can be a fundamental tool in our planning process. A clear conceptualization of accessibility is of huge importance to integrate it in the policymaking and planning process and effectively translate the policies into practice.