Wheres the Bus: Examining Factors Correlated with High and Low Public Transit On-Time Performance

TL;DR

This study analyzes various factors affecting bus on-time performance across 18 US transit agencies, identifying key influences like congestion, passenger load, and bus speed to inform improvements.

Contribution

It introduces a correlation-based approach tailored to individual agencies, highlighting specific factors that significantly impact bus OTP.

Findings

Higher city congestion correlates with better OTP.

Fewer passengers per hour negatively impact OTP.

Higher bus speeds and miles per vehicle improve OTP.

Abstract

An important aspect of public bus transit is its reliability of being on-time, which has a major impact on bus ridership. Currently, there is no unified national standard to measure bus on-time performance in the United States. This paper proposes a novel approach to improve the on-time performance (OTP) of buses at transit agencies throughout the country, based on factors unique to each agency. In this study, information on 18 public transit agencies data were obtained from the National Transit Database and were verified by the authors after contacting the respective agencies. A correlation analysis was conducted to evaluate the influence of different factors on bus OTP. The results show that a higher congestion ranking of the cities i.e. lesser traffic has a positive correlation with OTP while a lower number of passengers per hour has a negative correlation with OTP, both at a 95% confidence interval. Total miles per active vehicle and average speed of the bus were found to have a positive correlation with OTP at a 90% confidence interval. This potentially suggests that, if the average speed of the bus is higher i.e. if the bus operates at its scheduled speed, total miles per active vehicle will increase, which will in turn lead to a better on-time performance. A flow diagram interpreting the results of the correlation analysis was shown, where the main bus service process was considered and the predecessors and successors which could potentially affect the final OTP, were determined.