From New to Second-Hand: Consumer Trade-offs Between Price, Range and Vehicle Condition for BEVs and Hybrids in Australia
Access status:
Open Access
Type
Working PaperAbstract
Most empirical work on vehicle choice has focused on new-vehicle purchase decisions, even though households often acquire vehicles through the second-hand market and face a different set of constraints and information conditions. This paper addresses that gap by estimating a single ...
See moreMost empirical work on vehicle choice has focused on new-vehicle purchase decisions, even though households often acquire vehicles through the second-hand market and face a different set of constraints and information conditions. This paper addresses that gap by estimating a single choice framework that spans new and second-hand passenger vehicle markets, allowing direct comparison of how consumers value vehicle attributes and how substitution patterns differ across the two segments. We use a discrete choice experiment administered to a New South Wales sample, with alternatives that include new and second-hand vehicles and an opt-out, and attributes that reflect both markets, including purchase price, body type, vehicle size, powertrain, range, delivery availability for new vehicles, and odometer and condition for second-hand vehicles. Preferences are estimated using a two-class latent class model with error components to capture correlation in unobserved utility and segment-specific substitution within new and second-hand markets. The results show substantial heterogeneity in valuations and clear differences between new and second-hand decision processes, with second-hand quality signals exerting economically meaningful effects and price sensitivity being stronger for second-hand choices. Powertrain attributes matter, but their implications vary by market segment and by class, indicating that technology preferences interact with the institutional and informational features of the market in which the vehicle is acquired. Applying the estimated class-membership model to population microdata, we generate conditional parameter and willingness-to-pay distributions for a large synthetic population and predict market shares for simulated vehicle profiles. The simulation results underscore that secondary-market dynamics materially shape predicted demand patterns, which has implications for policies and market designs that aim to influence fleet composition through interventions that operate in, or propagate through, the second-hand market.
See less
See moreMost empirical work on vehicle choice has focused on new-vehicle purchase decisions, even though households often acquire vehicles through the second-hand market and face a different set of constraints and information conditions. This paper addresses that gap by estimating a single choice framework that spans new and second-hand passenger vehicle markets, allowing direct comparison of how consumers value vehicle attributes and how substitution patterns differ across the two segments. We use a discrete choice experiment administered to a New South Wales sample, with alternatives that include new and second-hand vehicles and an opt-out, and attributes that reflect both markets, including purchase price, body type, vehicle size, powertrain, range, delivery availability for new vehicles, and odometer and condition for second-hand vehicles. Preferences are estimated using a two-class latent class model with error components to capture correlation in unobserved utility and segment-specific substitution within new and second-hand markets. The results show substantial heterogeneity in valuations and clear differences between new and second-hand decision processes, with second-hand quality signals exerting economically meaningful effects and price sensitivity being stronger for second-hand choices. Powertrain attributes matter, but their implications vary by market segment and by class, indicating that technology preferences interact with the institutional and informational features of the market in which the vehicle is acquired. Applying the estimated class-membership model to population microdata, we generate conditional parameter and willingness-to-pay distributions for a large synthetic population and predict market shares for simulated vehicle profiles. The simulation results underscore that secondary-market dynamics materially shape predicted demand patterns, which has implications for policies and market designs that aim to influence fleet composition through interventions that operate in, or propagate through, the second-hand market.
See less
Date
2026-02-10Licence
Copyright All Rights ReservedFaculty/School
The University of Sydney Business School, Institute of Transport and Logistics Studies (ITLS)Share