Impact of Covid-19 on mode choice behavior: A case study for Dhaka, Bangladesh

• Main Authors: Chakraborty, Rohit, Paul, Tonmoy, Ratri, Salma Afia, Debnath, Mithun
• Format: Conference Proceeding
• Language: English
• Subjects: Accuracy; Coronaviruses; COVID-19; Developing countries; Independent variables; LDCs; Modal choice; Pandemics; Parameter estimation; Parameter identification; Travel; Utility functions
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0130123

Since the beginning of the COVID-19 pandemic, major changes have been observed in travel behavior. With the occurrence of multiple waves of this pandemic, Bangladesh as a least-developed country needs to deal with these changes, particularly changes in mode choice behavior. As Dhaka has been marked as a hotspot for the virus contagion, this paper has focused on the changes in mode choice behavior of Dhaka people while they are on the road. A web-based questionnaire survey was conducted for more than ten weeks to capture the information on mode preferences and perspectives on travel characteristics for commute and discretionary trips before and during COVID-19. After a careful, manual screening of the data, 571 responses were produced. Multinomial Logit (MNL) model based on a utility function has been used to investigate the significance of the socio-demographic attributes and travel characteristics of the trips on the mode choice behavior. The utility function contains gender, age, net monthly income (NMI), occupation, vehicle ownership, reliability, availability, and cost-effectiveness of the trip to calculate the maximum utility of the mode choice. The parameter estimates of the models identify the independent variables which are significant for mode choice behavior for this study. The MNL model of this study maintained overall prediction accuracies of 67.3% and 55% for commute trips and with an accuracy of 68.7% and 61.6% for discretionary trips before and during COVID-19 respectively. In addition, temporal comparison of trip characteristics perspective using stacked graphs show that public transports are the least reliable modes during the pandemic and NMVs are the most cost-effective modes in commute trips. Furthermore, this paper suggested some viable policy-making implications to cope with the current pandemic and relatable future national and global crises. Finally, the paper concludes by suggesting some future research insights.