Investigating heterogeneous preferences for plug-in electric vehicles: Policy implications from different choice models

Plug-in electric vehicles (PEVs) adoption has been a promising strategy to address climate change and improve energy security. Understanding consumer preferences for PEV attributes and policies is critical to accelerate mass PEV penetration. This study examines consumer preference het-erogeneity in adopting PEVs with mixed logit (MXL), latent class (LC), and latent class-mixed logit (LC-MXL) models based on a stated preference survey in Virginia, U.S. Consistently, all three models indicate that providing a monetary incentive is most effective in increasing the overall PEV market share, followed by deploying more charging infrastructure, while improvement in battery range shows a weak impact. Furthermore, the choice models that capture preference heterogeneity provide more nuanced results on the effectiveness of policies on a specific con-sumer segment and for a specific vehicle powertrain type. Lastly, when considering various model evaluation measures (e.g., model fit, prediction accuracy, and behavioral interpretation), results show that no model is unanimously superior to the other models. Rather, altogether they uncover a more comprehensive picture of EV preference structure. Findings provide insights into the usefulness of each choice modeling framework for future PEV adoption research. Also, it informs policymakers to be aware of alternative models which can provide different perspectives on policy implications.