Speaker
Description
Electrification of vehicles is considered a promising strategy to mitigate the increasing environmental and energy challenges caused by the growing demand for vehicular travel. As these benefits of EVs are closely linked to their market size, comprehending the determinants of EV adoption is crucial.
Built environment (BE) characteristics, known to profoundly impact motor vehicle adoption, may exert a comparable influence on EV adoption. Some research has noted the role of BE in the adoption of EVs; however, such studies remain limited. The primary gaps in the existing research can be summarized into three aspects: Firstly, previous studies have mainly concentrated on low-density cities in North America. In contrast, studies focusing on China, a country characterized by its high-density urban areas and status as the world's largest EV market, are scarce. Previous work suggested that the strategies to promote EVs should be differentiated to suit the varying characteristics of high-density and low-density urban areas. Secondly, there is a lack of comparison regarding how different types of EVs (e.g., battery electric vehicles (BEVs) versus plug-in hybrid electric vehicles (PHEVs)) interact with BE characteristics. Given the variations in travel behavior and charging requirements among different types of EVs, these differences may result in different relationships between BE and EV adoption. Thirdly, most existing studies have relied on vehicle registration data, which faces challenges due to data incompleteness. To protect privacy, if a CBG has fewer than 100 vehicles, the vehicle registration location information will not be disclosed. For example, 68.7% of Tesla vehicles lack information in the vehicle registration data. This lack of complete data represents a significant obstacle to understanding the relationship between the adoption of different types of EVs and BE characteristics.
In this study, we empirically investigate the roles of the BE in EV adoption in China. We calibrate a set of Poisson regression models of EV ownership at the neighborhood (jiedao) level using a unique dataset comprised of a 5% random sample of all the EVs in Shanghai, China. In addition, our analysis considers the potential heterogeneity in this relationship.
Our findings demonstrate the critical role of the BE in the adoption of EVs in China. On average, the EV ownership rates have a positive correlation with land use mix, number of parking lots, accessibility to public charging facilities, and number of shopping malls, while exhibiting a negative correlation with population density, job availability, metro station density, and distance to CBD. In the meantime, the BE-EV relationship shows significant heterogeneity along multiple dimensions, such as EV type (BEV vs. PHEV), car classification (A-segment, B-segment, and C-segment) and residential location (urban vs. suburban). For instance, the anticipated decrease in EV ownership due to increased metro access does not apply to BEVs, possibly because choosing public transportation over private vehicles is often correlated with lower travel expenses. Interestingly, distance to CBD exhibits no significant effect on the ownership of low-end A-segment EVs, yet shows a strong negative influence on the ownership of more expensive B-segment and C-segment EVs, suggesting income as a mediating factor in this relationship. Another interesting observation is that direct current charging piles do not show a significant correlation with EV ownership in suburban areas while alternating current charging piles do not correlate significantly in urban settings.
These insights offer a foundational basis for developing nuanced land use and transportation planning policies aimed at achieving the goal of sustainable transportation in China.
| Keywords | Electric vehicle adoption; Built environment; Poisson regression; China |
|---|---|
| Best Congress Paper Award | Yes |
