Meta-analysis of the built environment and transit use relationship in different countries

Many travel behaviour studies have demonstrated a relationship between public
transport ridership and built environment variables such as density, diversity, urban design and accessibility. However, results from past studies show large variability, which limits the transferability of findings outside the research setting. Context is important in travel behaviour research, encompassing many potential sources of this variability. Yet little is known about the link between the social and geographic context of the research setting and different impacts of the built environment on transit use.

In this paper, meta-analysis is used to synthesise evidence relating indicators of the built environment to transit use. Results are compared before and after grouping by the country in which research is conducted, to determine whether this explains variability.

As expected from previous research, results show indicators of transit-friendly urban design, density and accessibility are weakly correlated with increasing transit ridership. The average effect size for density showed significant variance between different countries. However, the majority of variance cannot be accounted for by differences associated with the country of the sample.

This study is limited by small sample sizes once data is grouped by country. This
reflects the lack of geographically diverse evidence in the field of land use and travel behaviour research. Furthermore, differences in study design were found to impact results. Future research that seeks to identify contextual sources of variability in the built environment and transit use relationship need to adopt a consistent study design across the locations being tested.