A 3-Step, Open-Data, Ride-Hailing Ridership Model with Pricing Applications

Author

Richard A. Mucci, University of KentuckyFollow

Author ORCID Identifier

https://orcid.org/0000-0002-7217-6836

Date Available

7-24-2025

Year of Publication

2024

Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation

College

Engineering

Department/School/Program

Civil Engineering

First Advisor

Dr. Gregory Erhardt

Abstract

Researchers and practitioners studied the effects ride-hailing had in cities before the covid-19 pandemic. Previous research found ride-hailing to produce negative externalities, such as reducing transit ridership and increasing congestion in various cities. Since the pandemic, ride-hailing ridership has nearly recovered to pre-pandemic levels in Chicago. Ride-hailing ridership has grown steadily since the pandemic while a rider’s willingness to share their trip stagnated. Ride-hailing ridership nearly recovering to pre-covid levels in Chicago suggests that transportation planners, and policy makers, will need to continue assessing the impacts ride-hailing trips have in their cities.

Pickup and drop off locations in the Chicago ride-hailing dataset are suppressed to community areas when there are less than 3 trips to/from a census tract within a 15-minute period. Suppressed and unsuppressed trips are compared to find that dropping suppressed trips could bias results because most of the ride-hailing trips in low-income areas of Chicago are suppressed. Data suppression typically occurs in areas and during times of infrequent ride-hailing use. Data suppression is necessary to protect a user’s privacy, but dropping suppressed trips will exaggerate the differences between frequent and infrequent trips. Dropping suppressed trips could underestimate trip lengths, underestimate overnight trips, and underestimate trips in low-income areas. The effect of dropping suppressed trips correlating with variables of interest would lead to biased model estimation results.

During 2019, approximately 1 in 5 ride-hailing users in Chicago were willing to share their trip. Understanding a user’s willingness to share a ride-hailing trip and the process of matching shared trips together will help transportation planners, and policy makers, to promote sustainable transportation. Travel time, fare, and choice models determine which factors influence a user’s willingness to share a ride-hailing trip and whether a shared trip is matched with another shared trip. Trips to/from airports are less likely to be shared. Trips to/from low-income areas are more likely to be shared. Longer shared trips are more likely to be matched, shared trips to/from dense areas are more likely to be matched, and shared trips between areas with a high number of shared trips are more likely to be matched. Matching with another shared trip adds approximately 4 minutes to a trip. The value of time of ride-hailing users is found to be $34.82 per hour. Understanding travel behavior of ride-hailing users is important for transportation planners aiming to fill mobility gaps that transit does not fill and aiming to reduce congestion. The results of the willingness to share choice model can guide the development of a ride-hailing tax that promotes ride-hailing users to share their trips. The results of the matching choice model can help develop a ride-hailing tax that increases the likelihood of shared trips being matched together.

Predicting transportation demand is an important step in making informed decisions about transportation policies and investments. Predicting demand for emerging modes allows new public policy scenarios to be evaluated. Many cities do not have ride-hailing ridership data, which makes it difficult to predict ride-hailing ridership. Previous research has focused on predicting ride-hailing pickups, determining why ride-hailing users are willing to share their trips, and/or determining whether ride-hailing trips produces negative externalities within cities. This paper showcases a 3-step, open-data, ride-hailing ridership model that predicts ride-hailing pickup totals within zones, assigns a destination zone for each pickup, determines whether a rider is willing to share each trip, and determines whether each shared trip is successfully matched. An in-sample validation is performed to confirm the accuracy of the model, which predicts monthly ride-hailing ridership during the study period within 5%, on average.

Chicago implemented an innovative ride-hailing tax change on January 6^th, 2020. A counter-factual scenario is simulated by applying the 3-step, open-data, ride-hailing ridership model without the tax change. The simulated ridership totals are compared to the modeled ridership totals, with the tax change, to evaluate the effect of the ride-hailing tax. The simulation finds that the tax reduced private trips and had a negligible impact on shared ridership. Private trips to/from/within downtown decreased by 22%, or approximately 33,000 average weekday trips. Shared trips throughout Chicago increased by 1%, or approximately 250 average weekday trips. Private trips throughout Chicago decreased by 13%, or approximately 35,000 average weekday trips. Ride-hailing ridership is found to be inelastic with the elasticity of ride-hailing use in response to the tax change being -0.83. The results suggest that pricing policies are an effective way to control private ride-hailing ridership while also generating revenue that can be used to promote more sustainable transportation modes. Simulating the effect of a ride-hailing tax change is one of many applications the 3-step, open-data, ride-hailing ridership model in this paper is capable of performing.

Digital Object Identifier (DOI)

https://doi.org/10.13023/etd.2024.317

Recommended Citation

Mucci, Richard A., "A 3-Step, Open-Data, Ride-Hailing Ridership Model with Pricing Applications" (2024). Theses and Dissertations--Civil Engineering. 147.
https://uknowledge.uky.edu/ce_etds/147