Atanaz Dorrani-Arab

This paper explores the relationship between public transit mode share and population density. It critically reviews the long-held belief that an increase in population density (compact built form) will result in an increase in public transit ridership. The research developed a longitudinal data set of travel behavior, transit supply, and proxies of built form for 1996 and 2016 for the City of Toronto. The data set is spatially disaggregated at the Traffic Analysis Zone (TAZ) level such that the TAZs that divide the City into 480 mutually exclusive and collectively exhaustive zones.
The paper found that a cross-sectional analysis of population density and transit mode share captures mostly the contemporaneous relationship between the two and does not, by default, lend credence to the argument that if the density increases over time at a place, it will subsequently result in higher public transit ridership. Such a question will require a longitudinal analysis where the impact of a change in public density over time is examined to determine its impact, if any, on transit ridership. Using Linear Mixed Models for longitudinal data, the paper found that the contemporaneous relation between density and transit mode share holds, but the change in population density over time does not automatically correlate with an increase in transit ridership.