Rising Gas Price and Transit Ridership: Case Study of Philadelphia, Pennsylvania

Document Type

Journal Article

Publication Date

2009

Subject Area

ridership - elasticity, ridership - demand, mode - rail, mode - mass transit

Keywords

Travel models (Travel demand), Travel demand, Travel behavior, Transit, Ridership, Regression analysis, Regression, Regional railroads, Public transit, Prices, Philadelphia (Pennsylvania), Petrol, Patronage (Transit ridership), Multivariate analysis, Mathematical models, Mass transit, Local transit, Gasoline, Elasticity (Economics), Case studies

Abstract

In July 2008, gas prices peaked at unprecedented levels in both nominal and real dollars. Americans also took more transit trips in 2008 than in any year since 1956. Past research has demonstrated a correlation between increases in gas price and increases in transit ridership. Using the Philadelphia, Pennsylvania, metropolitan area as a case study, this research confirms and provides new insight into this relationship. Multivariate linear regressions were developed both to demonstrate and to measure the correlation while accounting for seasonal differences and to provide insight into what future gas price scenarios could mean for a transit system in a city like Philadelphia. One set of models was developed for the Regional Rail system and another was developed for the City Transit system. Both models demonstrated correlations of statistical significance. Comparison of the models showed that the price of gas had both a more significant correlation with and a higher impact on Regional Rail ridership than on City Transit ridership. The models were also used to explore the cross elasticity for transit demand. The results suggest various cross elasticities of between 0.15 and 0.23 for City Transit services and between 0.27 and 0.38 for Regional Rail services. The problems with attempting to isolate a value for cross elasticity are discussed. This analysis also explores the possibilities of a nonlinear relationship that explains past trends and discusses the difficulties in using these models to predict future transit demand.

Share

COinS