Ride-Hailing Data Suppression and Exclusion Strategies Can Lead to Biased Outcomes

The Chicago ride-hailing data set is one of the few data sets in the United States containing details of individual ride-hail trips. To protect privacy, locations and times are aggregated, and locations are further suppressed when the frequency of trips is low. Most researchers using this data remove the trips with suppressed locations or external destinations from their analysis. This research finds that when suppressed and external trips are excluded, the trip length, cost, and distance are all underestimated, as are trips in low-income neighborhoods. Future research should consider including these trips at a more aggregate spatial resolution.