| description abstract | The impacts of COVID-19 on for-hire vehicle (FHV) (e.g., Uber/Lyft, often referred to as transportation network companies in other locations) and taxi use have been relatively understudied compared with transit and personal vehicles. This study analyzed and estimated the changes in ridership for taxis and FHVs in New York City during the COVID-19 pandemic to determine whether it had disproportional impacts on these competing modes, how these impacts varied over time and space, and the associated factors. Data supporting the analyses came from the Taxi and Limousine Commission, the COVID-19 Data Repository, Google's Community Mobility Reports, the American Community Survey, and the Primary Land Use Tax Lot Output. Temporal change was measured by the daily taxi/FHV ridership deviation from a defined baseline, which showed that COVID-19 more negatively impacted taxis than FHVs. Temporal moving average models were then employed, which showed that COVID-19 had different temporal impacts on taxis and FHVs in relation to the parameters’ significance, magnitude, and temporal correlation patterns. In general, taxi/FHV ridership dropped when people spent more time at home and the number of confirmed COVID-19 cases was greater. The spatial variation in taxi/FHV ridership was measured by the coefficient of variation. Spatial regression models indicated that the land use of a zone affected taxi/FHV ridership during the pandemic. In addition, a zone with more carless/car-free households, older persons, or more children enrolled in school was more likely to experience a decrease in taxi/FHV ridership. A zone with more workers who commuted by walking or taking transit (excluding taxis) in pre-COVID times was more likely to see a decrease in taxi/FHV ridership. A zone with more people working from home pre-COVID, was more likely to see an increase in FHV ridership. The models showed that COVID-19 had greater spatial impacts on taxis than FHVs. Based on these results, this study provides insights as to what factors affected ridership of the two competing travel modes and suggests actions that transportation authorities could take to reduce temporal and spatial impact disparities. | |
