Analysis of Chicago Trip Data Shows Neighborhood and Transit Service Factors Associated with Ride-Hailing Usage Patterns.
Modern on-demand transportation solutions, such as ride-hailing apps, have been credited with providing new modes of mobility for travelers who live in "transit deserts," or geographic areas with little or no conventional transit services. However, research remains mixed on whether these ride-hailing services are a complement or competition to other transit services. A report focusing on trips taken in the City of Chicago sought to learn more about the ways in which on-demand ride-hailing and conventional transit use interacted.
Researchers used ride-hail trip data from the City of Chicago Transportation Network Providers trip database, which collected required trip-level data from providers. From the database, trips where the census-tract origin and destination locations were both identified, consisting of more than 86 million data points, were analyzed. Transit deserts were identified through a combination of transit stop density, route density, median daily transit service headway, and the number of overnight stops; the four variables were standardized and averaged to create an overall transit supply score. Neighborhoods were then analyzed for demographic factors to provide context for the other data. Bivariate analyses examined the spatial relationship between ride-hail trip ends and socio-demographic factors and transit supply. Multivariate regression models were developed to relate transit availability and ride-hailing at the census tract level.
- The authors reported that there was little evidence that ride-hailing trips were associated with a low level of overall transit service. In almost all instances with statistical significance, the relationship between transit accessibility and ride-hail usage was positive. Clusters with the most transit service in central Chicago also had the highest number of ride-hailing trips. Areas with low transit service tended to have low ride-hailing trips. Only a few census tracts, in a mixed-use neighborhood, were found to have high ride-hailing and low transit service.
- When transit supply was analyzed in more depth, as a suite of four individual factors rather than as an overall measure, characteristics within the census tract were found to be significantly related to ride-hailing trips.
- Higher transit stop density (dominated by presence of bus stops) was associated with lower ride-hailing trips. A 10 percent increase in stop density was associated with 2 percent fewer ride-hail trips. However, overnight transit service stops were found to have a positive relationship with ride-hailing trip frequency.
- The largest effect was whether a rail station was located nearby; some travelers may be using on-demand transit to fill first-mile / last-mile gaps. Census tracts within 400 meters of rail stations had 18 percent more ride-hail pickups and 22 percent more drop-offs than others.
- The greatest socio-economic effect on ride-hailing trip frequency was found to be associated with income. Census tracts with a 10 percent higher median household income were estimated to have a 4.1 percent higher number of ride-hail pickups and a 3.7 percent higher number of drop-offs.