Investigating the Effectiveness of Innovation Policy Mix to Enhance the Adoption of EVs
Principal Investigator:
Joseph Cho
Abstract:
Despite efforts to deploy electric vehicles (EVs), the penetration rate is low, and the consequence of the slow diffusion is detrimental to climate change mitigation in the transportation sector. Various factors may affect consumer awareness of EVs, such as cost, charging time, charging infrastructure, vehicle performance, and warranty. Due to the data constraints, the evidence is limited. Significant factors leading to the low market share of EVs can be uncovered when using an agent-based model with an inductive-deductive approach. Thus, the study systematizes various determinants to capture consumer behavior toward EVs and the consumer awareness effect in our model to help illustrate the heterogeneity in response to multiple factors.
Description of Research:
Innovation policy has focused on promoting consumer acceptance of EVs to reach a critical mass of adopters. Currently, the federal government and most states offer various types of vehicle purchase-related financial incentives and charging-related rebates in the United States. Despite efforts to deploy EVs, the penetration rate is still low, and the consequence of the slow diffusion is detrimental to climate change mitigation in the transportation sector. In this regard, various factors affect the consumer’s decision-making: price, range anxiety, lack of infrastructure, lengthy charging time, technology resistance, uncertainty, and safety. Prior research measured how much subsidy would be required to promote the adoption of EVs (Helveston et al., 2015). However, the effect of policy incentives is still questionable in groups of people in different regions or countries since the effectiveness of incentives to aid EV adoption will vary based on demographics and network effects (Skerlos and Winebrake, 2010). In addition, most prior research efforts examined the impact of policy incentives on EV adoption in isolation without considering the pattern and number of charging stations, the change in electricity price, network effects, and the change in oil price. It is significant to characterize the effectiveness of the policy mix between market-pull or consumer-oriented EV policy instruments (e.g., tax incentives) and technology-push instruments (e.g., number of charging stations). Moreover, it is crucial to examine how various policy instrument mixes affect EV manufacturers' incentives to innovate since policy makers have dual policy objectives: the increase of EV adoptions and EV innovation associated with vehicle price and battey efficiency. Thus, this study estimates the effectiveness of policy incentives to promote EV adoption in the U.S. states. The research examines this through longitudinal analysis of EV adoption in the U.S. market employing panel data. This study provides an empirical investigation of the role played by two characteristics of the policy mix in increasing EV innovation and adoption. The econometric results will characterize the policy mix for more balanced use in demand-pull and technology-push instruments. We expect to present an important aspect to be considered in policy mix design.
The research objective is to investigate how EV tax incentives and other policy instruments affect EV adoption in the U.S. market. We collect a rich dataset of state-level EV registrations, EV prices, gas prices, electricity prices, number of charging infrastructures, number of patents, R&D expenditures, consumer spending, education level, annual wage, and policy instruments that can be assigned to a value from 2015 to 2021. This study addresses research questions by employing econometric techniques to panel data for the 50 U.S. states to estimate the effectiveness of various policy instruments on EV adoptions. We construct two measures of policy: demand-pull instruments and technology-push instruments. We specify a high-dimensional fixed-effect regression model and estimate it using an iterative procedure that allows the estimation of a parsimonious specification via ordinary least squares (OLS) (Correia, 2016). We use STATA procedure reghdfe uses the Frisch-Waugh-Lovell theorem to sweep away each of the fixed effects from all three-dimensional variables.