A longitudinal comparison of shared and private e-scooter crashes, near-miss rates, attributions and characteristics

Abstract

As e-scooters have become increasingly common in urban areas, safety concerns have grown. However, most safety research does not distinguish between shared and private e-scooter use. Moreover, no studies have examined within-person changes in safety outcomes over time. This study followed Australian e-scooter riders (n = 285) across two survey waves conducted 12 months apart (follow-up and final) and comprised two analyses. First, we compared riders’ most recent near misses and crashes on shared and private e-scooters in terms of their characteristics and riders’ attributions. At follow-up, 32.7% of participants were private-only riders, 51.2% shared-only, and 16.1% mixed-use. Most near misses involved another road user (78.1% at follow-up, 70.1% at final), while a substantial share of crashes did not (55.6%, 59.1%). Private e-scooters accounted for the majority of near misses involving another road user at both waves (64.0%, 76.6%). Near misses occurring on shared e-scooters were consistently attributed to riders’ own lack of experience. Second, we modelled within-person changes in near-miss rates. A Poisson generalised linear mixed model, with a random intercept for participant and riding exposure as an offset, tested whether near-miss rate changed between waves, adjusting for sociodemographic characteristics. Near-miss rate did not differ significantly between waves. Riders outside Queensland and Victoria had a lower rate than Queensland riders (IRR = 0.27). Riders with primary/secondary or bachelor’s degree education were associated with lower rates than those with post-school qualifications. Riders’ attribution of shared e-scooter incidents to inexperience suggests onboarding and practical training warrant further evaluation.