The effect of COVID-19 slow streets on dockless scooter travel: a quasi-experiment

Source abstract

We used the implementation of slow streets during COVID-19 as a quasi-experiment to understand how street treatments that prioritize non-car travel influenced dockless scooter trips. We gathered data on slow streets implementation from Los Angeles, Oakland, Portland, and San Francisco and identified control group, non-implemented slow streets candidates in three cities (excluding Portland). We examined the impact of slow streets on dockless scooter trips using a before-after, experimental-control group design. We implemented differences-in-differences and panel regressions to analyze the effect of slow streets on scooter trips, finding a robust pattern of statistically significant, positive associations between slow streets implementation and dockless trips. The treatment effect is generally in the range of a 7% to 94% increase in dockless trips. The results indicate potential for infrastructure treatments that slow car travel and prioritize non-car modes to increase dockless scooter trip-making. We discuss the findings and suggest avenues for future research.