Assessment of paratransit system performance in mixed traffic using a two-stage bootstrap-DEA and ordinary Least Square approach

There has been a surge in the use of paratransit services including minibuses, two and three-wheelers, and other variants in Sub-Saharan African cities. Despite being characterized as unsafe; they continue to address a greater percentage of urban travel demand. Amidst the ongoing debate to integrate formalized public transport (FPT) and paratransit services, there is a need to understand the current performance of the existing modes operating in mixed traffic. The present study attempts to evaluate the performance measures of paratransit services in Freetown, Republic of Sierra Leone using a two-stage approach. In the first stage, efficiency and effectiveness scores for paratransit services along two corridors were estimated and corrected for bias by bootstrapping using data envelopment analysis (DEA). In the second stage, the bias-corrected DEA efficiency scores were regressed against a set of service and operational indicators. Data collection included videography, onboard vehicle, user perception, and operator surveys. A total of 1119 drivers and 1010 passengers were intercepted, and 300 onboard trips were made. Data were analyzed in STATA and R package (deaR) software programs. Findings revealed that as compared to low-capacity modes, relatively high-capacity modes are more efficient and effective in terms of revenue-generating vehicle kilometers travelled and daily ridership. Efficiency was found to be positively influenced by passenger density, speed density, and vehicle capacity. Stop density, capacity utilization ratio, in-vehicle travel time, trip delay and fuel consumption had a significant negative impact on paratransit services. These findings provide an opportunity to develop more precise planning strategies aimed at improving paratransit service performance especially for cities moving toward public transport system integration.