Comparing Transit Agency Peer Groups using Cluster Analysis

Document Type

Journal Article

Publication Date


Subject Area

place - urban, place - north america, economics - operating costs, land use - urban density, organisation - performance, planning - service level, ridership - behaviour


urban, transit ridership, service level


Transit ridership has decreased steadily each year from 2014 to 2017 despite increasing urban populations and transit investment. Several analyses have examined trends in national transit ridership levels or within specific agencies or regions. However, the causes of transit ridership changes, and the changes themselves, are not likely to be the same everywhere. Thus, comparing the performance of transit agencies over time with their peers, which share similar characteristics, may yield more informative results than just examining national trends. This analysis groups metropolitan areas using relevant characteristics, which are correlated with transit ridership: total population, density, percentage of zero vehicle households, and transit agency operating expenditures. Data from the National Transit Database and the United States Census Bureau are used. Trends in transit ridership are then analyzed in relation to population and service levels within each cluster and by family of modes. Results suggest that although the change in population and service levels can partly explain the change in ridership for transit modes in dedicated right of way, the same is not true for modes in mixed right of way. In particular, transit ridership for mixed right of way modes in large metropolitan areas is not significantly correlated with the change in population or service levels.


Permission to publish the abstract has been given by SAGE, copyright remains with them.