Cycling in a Crisis: Employing Quasi-Experimental Designs to Estimate the Effects of Provisional Bicycle Infrastructure.

This article examines the semi-causal impact of provisional cycling infrastructure implemented during the COVID-19 pandemic on bikeshare trip durations in Philadelphia, focusing on the closure of Martin Luther King Drive (MLK) to automobiles. Using a differences-in-differences (DID) quasi-experimental model and longitudinal trip data from the Indego docked bikeshare system, the study compares trip durations originating from stations within 500 meters of MLK (treatment group) to those 500–1,000 meters away (control group) before and during the closure period (March 2020–August 2021). Results indicate that while trip durations increased citywide during the pandemic, stations near the MLK closure experienced an additional statistically significant increase of approximately 2.4 minutes—about an 11% rise above baseline—suggesting that supportive, separated cycling infrastructure can meaningfully extend bikeshare trip durations. The study acknowledges limitations including lack of route-level data, potential recreational trip bias, and the single-city case study design, and recommends further research incorporating geolocated trip data and diverse urban contexts to better understand infrastructure impacts on cycling behavior.