In North American cities, the growing use of ride-hailing services, such as Uber and Lyft, has occurred at a time when public transit ridership has either stagnated or declined. Previous studies on the topic have found that the relationship between ride-hailing and public transit services tends to be context-specific in nature. In some cases, ride-hailing has been shown to complement public transit, while, in others, it can be a substitute. This paper takes an integrated approach to modeling the generation of demand for ride-hailing and public transit services, using trip-level ride-hailing data and data from a regional household travel survey. In addition, the relationship between the demand for ride-hailing and transit services is also explored. This paper uses the bivariate ordered probit model and the recursive regression model to study the role that built environment and socio-demographic attributes play in the generation of transit and ride-hailing demand. The model results reveal that there are several built environment attributes, such as transit accessibility and the density of commercial and recreational establishments, that influence the generation of both ride-hailing and transit demand. The results also indicate that the relationship between ride-hailing and transit services tends to be more complementary than substitutive in nature. This does not, however, mean that ride-hailing services are never used as a substitute for public transit in Toronto. The results of this study aim to provide transit agencies with a means of identifying locations where ridership may be threatened by ride-hailing services.

在北美城市，Uber和Lyft等乘车服务的使用日渐增多，而公共交通的乘车率则停滞不前或下降。先前对该主题的研究发现，乘车服务与公共交通服务之间的关系在本质上倾向于特定于上下文。在某些情况下，打车已被证明可以补充公共交通，而在另一些情况下，它可以替代。本文采用整合的方法，使用出行级别的打车数据和来自区域性家庭旅行调查的数据，对打车和公共交通服务的需求生成进行建模。此外，还探讨了乘车需求与公交服务之间的关系。本文使用二元有序概率模型和递归回归模型来研究环境和社会人口属性在交通和乘车需求的产生中的作用。模型结果表明，存在一些建成的环境属性，例如交通可及性以及商业和娱乐场所的密度，这些都会影响乘车和交通需求的产生。结果还表明，乘车服务与公交服务之间的关系在本质上往往比替代服务更互补。但是，这并不意味着在多伦多，乘车服务从未被用来代替公共交通。