Transportation Network Companies (TNCs) like Uber and Lyft are rapidly changing urban transportation. Their widespread adoption, fueled by smartphone accessibility and convenience, has led to significant increases in ridership (Schaller, 2018). While proponents highlight increased travel options and potential reductions in vehicle ownership, concerns remain regarding induced travel demand and negative impacts on public transit ridership. The uncertainty surrounding TNCs' system-level effects on energy use, the environment, and traffic congestion (Conway et al., 2018) necessitates a deeper understanding of individual travel decisions. This study addresses this gap by analyzing survey data from Strategic Vision Incorporated to investigate the preferences of TNC travelers and drivers in the United States. The research focuses on understanding traveler preferences for being a rider, driver, or non-user of TNC services, as well as the choice of pooled ridesourcing. For drivers, the study examines their inclination to switch to fuel-efficient vehicles and their vehicle purchase decisions influenced by TNC driving. The analysis utilizes multinomial and binary logistic regressions to model these preferences and provide insights into the complex interplay of socio-demographic factors and TNC usage.

The literature review examines the evolution and impacts of TNC services on vehicle ownership and transit ridership. While some studies suggest TNCs reduce vehicle ownership (Rayle et al., 2016; Clewlow and Mishra, 2017; Conway et al., 2018), the extent of this reduction remains unclear. The impact on transit ridership is equally debated, with some studies indicating negative impacts (Sadowsky and Nelson, 2017; Babar and Burtch, 2017), while others suggest complementarity (Dias et al., 2019; Hall et al., 2018) or a potential for synergy through first/last mile connectivity. Studies have explored the socio-demographic characteristics of TNC users, identifying younger, highly educated, and urban-dwelling individuals as frequent users (Clewlow and Mishra, 2017; Kooti et al., 2017; Alemi et al., 2018). Research on TNC drivers has focused on safety, wages, and demographics (Feeney, 2015; Berger et al., 2018; Kooti et al., 2017; Hall and Krueger, 2018; Berliner and Tal, 2018), but lacks exploration of their vehicle preferences. This study fills this gap by examining driver preferences related to fuel efficiency and vehicle purchase decisions.

The study utilizes data from a 2017 Strategic Vision Inc. survey of 11,902 U.S. consumers residing in TNC-served areas. The survey collected information on household characteristics (age, gender, marital status, education, income, ethnicity, residence location, commuting mode, household size), TNC usage, preferences, and vehicle ownership. To address potential sampling bias, the researchers employed iterative proportional fitting (IPF) to weight the sample based on demographic data from the 2016 American Community Survey and 2010 U.S. Census Bureau data. The weighted sample was then used to analyze various preferences. Four logistic regression models were developed: 1. **Model 1 (Multinomial Logistic Regression):** This model examines the preference for being a TNC rider, driver, or non-user, using socio-demographic characteristics as explanatory variables. 2. **Model 2 (Binary Logistic Regression):** This model analyzes the preference for pooled ridesourcing among TNC users. 3. **Model 3 (Binary Logistic Regression):** This model investigates TNC drivers' preference for switching to fuel-efficient vehicles. 4. **Model 4 (Binary Logistic Regression):** This model explores whether driving for TNCs influenced drivers' decisions regarding buying, renting, or leasing a new vehicle. The models incorporate various socio-demographic variables and account for interaction effects to capture non-linear relationships. Predicted choice probabilities were plotted to visualize the effects of continuous variables.

The key findings are summarized below: **TNC User Preferences (Model 1):** * Likelihood of being a TNC user increases with age until 44, then decreases. * Higher-income individuals are more likely to be TNC riders but less likely to be drivers. * Households with more vehicles are less likely to use TNCs. * Early adopters of technology and metropolitan residents are more likely to use TNCs. **Pooled Ridesourcing Preferences (Model 2):** * Preference for pooled ridesourcing decreases with age. * Younger women are more likely to pool rides than younger men. * Households with more vehicles are less likely to use pooled rides. * Metropolitan residents are more likely to use pooled rides. **Fuel-Efficient Vehicle Preferences (Model 3):** * Younger TNC drivers are more likely to prefer fuel-efficient vehicles. * Postgraduate drivers in metropolitan areas are more likely to prefer fuel-efficient vehicles (particularly those under 48). * Married drivers are more likely to prefer fuel-efficient vehicles (especially those under 60). * Early adopters of TNC services are more likely to prefer fuel-efficient vehicles. * Drivers who drive daily are more inclined towards fuel-efficient vehicles. * Drivers with higher vehicle ownership have a higher tendency to switch to fuel-efficient vehicles. **Vehicle Purchase Decisions (Model 4):** * Tendency to buy a new vehicle (influenced by TNC driving) decreases with age. * Postgraduate and married drivers (under 55) are more likely to consider TNC driving when buying vehicles. * Higher income drivers are less inclined to buy a new vehicle for TNC driving, with different patterns observed across single and married drivers. * Male drivers, metropolitan residents, early adopters, and those with higher vehicle ownership who drive daily are more likely to consider TNC driving in vehicle purchases.

The findings provide valuable insights into the preferences driving TNC usage and vehicle choices. The non-linear relationship between age and TNC usage highlights the need for targeted marketing strategies. The impact of income on TNC usage and vehicle purchases suggests potential policy implications regarding affordability and access. The findings on pooled ridesourcing and fuel-efficient vehicle preferences offer opportunities for policy interventions to promote sustainable transportation. The study’s focus on driver preferences helps to understand the broader implications of TNCs on the automotive market. This study contributes to a richer understanding of micro-level decisions that influence the system-level impacts of TNCs. The results can inform stakeholders in designing effective policies and strategies.

This study offers novel insights into TNC user and driver preferences, particularly concerning pooled ridesourcing and fuel-efficient vehicles. The findings reveal non-linear relationships and highlight the influence of socio-demographic characteristics. Future research could explore causal relationships through randomized controlled trials and investigate the long-term impacts of TNCs on vehicle markets and the environment.

The study relies on survey data, which may be subject to recall bias and self-reporting errors. The cross-sectional nature of the data prevents the establishment of causal relationships. The sample, while large, might not perfectly represent the entire U.S. population. Future research using longitudinal data and experimental designs could strengthen the findings.