All ages and abilities cycling infrastructure, cycling activity, and perceived safety: Findings from a natural experiment study in three mid-sized Canadian cities

Abstract

As many cities aim to increase cycling ridership, they turn to research for evidence on the effectiveness of building cycling infrastructure. Our study team of researchers and practitioners designed a natural experiment study to assess the impacts of cycling infrastructure on two outcomes: self-reported cycling activity and perceived safety. The intervention was a 33 kilometre ‘all ages and abilities’ (AAA) network in Victoria, British Columbia, Canada. We included Kelowna, British Columbia and Halifax, Nova Scotia as control cites. In 2016, 2019, and 2021, we conducted population surveys on transportation behaviours with ~1000 respondents per wave, per city, using age and gender quotas. We measured exposure as the distance from each respondent’s postal code to the nearest AAA infrastructure. We completed a triple-difference analysis to estimate the effect of living closer to AAA infrastructure (≤ 500 metres) on cycling activity and perceived safety over time. Across study cities, cycling increased marginally during the study period, although there was no statistical difference in the change over time between respondents living closer and farther from AAA infrastructure. Perceived safety increased over time in Victoria, but for people living > 500 metres from AAA infrastructure (Odds Ratio: 1.59, 95% Confidence Interval: 1.01 – 2.50). Sensitivity analysis revealed a greater magnitude increase in perceived safety for women (Odds Ratio 2.08, 95% Confidence Interval: 1.07 – 4.04). The study results were influenced by a delay in the implementation of the AAA network in Victoria, substantial infrastructure investment in the control cities, and the COVID-19 pandemic. These unanticipated events likely affected cycling activity and the ability to assess impacts of cycling infrastructure interventions in this natural experiment study.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 10 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.