Understanding Factors Influencing Shared e-Scooter Usage and its Impact on Auto Mode Substitution 

Abstract
Background: Since the debut of e-scooter sharing (ESS) in 2017, hundreds of cities worldwide embraced this transportation mode to provide more options to road users, hoping to address ‘last mile’ needs and alleviate traffic congestion, etc. According to existing studies, shared moped scooter has some complementarity with the pedestrian, while it seems to substitute private transport (either car or moto) and, up to date, has no clear effects on public transport in Madrid, Spain (Aguilera-García, Gomez, & Sobrino, 2020) or a considerable share of trips that ESS could potentially replace are currently undertaken with more environmentally friendly modes (walk, bicycle, public transport) in Taiwan (Eccarius & Lu, 2018). In the City of Tampa, ESS mainly replaces walking, ride-hailing and private vehicle driving according to our survey results. Thus, the mode shift caused by ESS varies from city to city and whether and how the socio-demographic and other attributes affect the mode shift is still an open question.
Research objectives: In this study, taking the ESS program in the City of Tampa as an example, we designed a survey questionnaire to collect data from both users and non-users. By analyzing the survey data, we attempt to understand user behaviors, including the main factors leading to the use of the program and its potential as an auto-trip substitute (including private vehicle and ride-hailing). We will compare the results with literatures of bike sharing to gain deeper insights.
Data: A survey was disseminated to the general public after 5 months of the program launch. In this research, we mainly focus on shared e-scooter users. Users are asked to answer questions regarding user experience, safety concerns, mode shift, collision experience, opinions on the program and sociodemographic information. To address our research interests, we specifically asked about the frequency of using shared e-scooters, and for a recent particular trip, what other transportation modes would the respondent had used if shared e-scooters were unavailable.
Methods: Descriptive statistics are presented to provide an overview of users’ riding behaviors and mode substitution. To gain a deeper understanding, a random-parameter ordered probit model is estimated for the usage of e-scooters and a mixed logit model is estimated for auto-mode substitution (private vehicle and car-hailing).
Results: From modeling results, we find users who ride e-scooters on bike lanes are more likely to be frequent users and users who prefer a lower speed limit are more likely to ride e-scooters once a while. Thus, service providers offering training programs for new users and the environment with more protected bike lanes will encourage the use of ESS and leverage the advantages of this emerging transportation mode. In terms of mode substitution. We find users who ride e-scooters more frequently are more likely to replace ride-hailing, indicating the competition between the two modes. While users who live in downtown Tampa, and difficult parking increase the probability of replacing private vehicles with e-scooters, indicating ESS helps the City relieve the shortage of parking issues in downtown.