Who Uses Ride-Hailing? Using Cluster Analysis to Identify Traveler Markets in the Greater Toronto & Hamilton Area

No Free Rides: Winners and Losers of the TTC U Pass
January 20, 2020
What Makes the Gears Go Round? An Analysis of Factors Influencing Bicycling to Suburban Rail Stations
January 7, 2020
by
Hong Yun (Eva) SHI & Matthias SWEET

Background
Ride-hailing has become popular, but how is this technology used? And how does it align with long-term planning and policymaking?

Research Design
Using data from an online survey of 3,200 participants in the Greater Toronto and Hamilton Area conducted in December 2018, this study uses descriptive statistics and cluster analysis results.
  1. Descriptive statistics are used to gather a general understanding of correlations between individual, household, and travel characteristics, on the one hand, and ride-hailing use.
  2. Descriptive statistics are used to gather a general understanding of correlations between individual, household, and travel characteristics, on the one hand, and ride-hailing use.
  3. Traveler sub-markets identified through cluster analysis are compared with ride-hailing use.
Cluster Analysis Findings
Cluster A: Multi-Modalists
  • Biggest users of shared-mobility tools including bike-share, car-share, and ride-hailing, mirroring the characteristics of “Supersharers”, as coined by Clewlow and Mishra (2017).
  • Highly mobile with access to many mobility tools: vehicles (1.6 vehicles per household) , bikes, transit passes.
  • Already engages in the most multi-modal travel (notably public transit and regional transit commuting).
Cluster B: Low-Mobility Travelers
  • High ride-hailing use rate.
  • Low levels of mobility and access to fewest mobility tools.
  • 0.8 vehicles per household.
  • 61.3% rely on transit for commuting.
  • Low smartphone ownership.
  • Have the highest proportion of lower income households (9.4% of households earn under $14,999, 21.5% earn between $15,000 and $39,999).
  • Highest proportion of unemployment (14.8%), with 5.6% not in the workforce and 10.1% being full-time students.
Cluster C: Auto + Private Mobility Traveler
  • Low users of ride-hailing (70.5% have never used ride-hailing).
  • Heavily reliant on their cars but also have a high bicycle ownership rate.
  • 1.7 vehicles per household.
  • Highest proportion of retired individuals at 30.8%, just below half (45.0%) of respondents are full-time employed
Cluster D: Car-Dependent Travelers
  • Lowest ride-hailing adoption rate (76.1% have never used ride-hailing).
  • Reliant solely on their personal vehicles.
  • 57.1% of respondents are employed full time.
  • 91.0% of respondents drive, either alone or with others, to work.
  • Typical commute time is the lowest among the groups (30.8 minutes).
Conclusion
There are three main findings in this study:
  1. Car dependent lifestyles are not associated with ride-hailing.
    Auto + Private Mobility Travelers and Car-Dependent Travelers are infrequent users and least likely to ride-hail. This finding suggests that policymakers should be cautious to conclude that ride-hailing is likely to significantly reduce auto-dependence among those auto-oriented residents.
  2. Ride-hailing is filling a transportation gap for the Low Mobility Travelers.
    Low Mobility Travelers have lower household incomes, lower levels of educational attainment, and lower levels of mobility. Although there are many equity-based critiques of ride-hailing - e.g. not being available for the “unbanked” – ride-hailing appears to be advancing needs of economically vulnerable transportation system users.
  3. Multi-Modalists are the single largest ride-hailing user group.
    This group is highly multi-modal, owns and uses cars, owns transit passes and uses transit, and uses other forms of shared mobility, such as car-sharing and bike-sharing. This cluster is significantly younger and more highly-educated compared to other traveler sub-markets. Overall, ride-hailing appears to be used as one additional form of mobility to augment many diverse other sources of mobility.

Acknowledgement

This research was funded by the City of Toronto's Transportation Services Division. Thank you to each of their staff for their support

Presented at the 99th Annual Meeting of the Transportation Research Board, January 2020