New Article: ‘An alternative to slow transit, drunk driving, and walking in bad weather: An exploratory study of ridesourcing mode choice and demand’

Based on the number of vehicles with Uber and Lyft stickers on them I’ve seen around Ann Arbor, it’s obvious that ridesourcing (or ridesharing) has become a significant transportation mode even in small cities. But given the difficulty of obtaining data from these companies, how can we study this important new phenomenon? I decided to try a novel approach, and conducted a survey of ridesourcing users in Washtenaw County (which contains Ann Arbor and Ypsilanti) by recruiting respondents via geographically targeted online ads, as well as ads in local businesses and buses. The resulting data, summarized in a paper published today, provides a detailed look into the nature of these trips: why riders are choosing ridesourcing, and the neighborhoods they’re traveling to and from. Here is the abstract:

Companies providing ridesourcing, or the use of mobile phone apps to request rides from drivers of privately-owned vehicles, have expanded rapidly in many cities in recent years. To shed light on this phenomenon, this paper reports an exploratory study of ridesourcing trip patterns and mode choice in Washtenaw County, Michigan, USA, which obtained a convenience sample of 167 respondents (reporting 192 trips) via geographically targeted online and offline ads. Consistent with previous empirical studies, ridesharing users are younger and a greater percentage are female than the the general public, and most trips occur in a small number of high density block groups. When asked what other options were available for ridesourcing trips, respondents reported transit (63%), private vehicles (32%), walking (32%) and bicycling (18%). Specific reasons for choosing ridesourcing instead of these options included the frequency of transit, alcohol use for driving, and weather and distance for walking and biking. A multivariate analysis found variables related to greater ridesourcing use for a block group included job density, jobs-housing balance, bar and restaurant density, and presence of households without vehicles. The paper demonstrates the potential of survey data to generate greater geographic insights into ridesourcing use, as well as the potential for extending established travel-behavior research approaches to ridesourcing.

Goodspeed, Robert, Tian Xie, Tawanna R. Dillahunt, and Josh Lustig. 2019. “An alternative to slow transit, drunk driving, and walking in bad weather: An exploratory study of ridesourcing mode choice and demand.” Journal of Transport Geography 79:102481. doi: (open access)

Author: Rob Goodspeed