The increasing adoption of electric vehicles (EVs) necessitates the development of robust charging infrastructure. The Infrastructure Investment and Jobs Act (IIJA) in the US underscores this need, allocating significant funding for nationwide EV charging network expansion. However, the impact of public EV charging stations (EVCS) extends beyond simply providing charging capabilities. EV drivers often spend time at nearby businesses while their vehicles charge, potentially increasing foot traffic and revenue for local establishments. While research exists on the environmental benefits of EVCS, a significant gap remains in understanding their economic impact on local businesses. Previous studies by EV charging companies have shown increased customer visits and spending, but these were often limited in scale and lacked robust causal analyses. This study aims to fill this gap by quantifying the impact of EVCS on customer counts and spending at nearby businesses in California, a state at the forefront of EV adoption and charging infrastructure deployment. The study employs a Difference-in-Differences (DID) methodology along with propensity score matching to address endogeneity concerns and analyze data from two distinct periods (2019 and January 2021 to June 2023) to account for the COVID-19 pandemic.

Existing literature extensively documents the environmental benefits of EVs and EVCS in reducing greenhouse gas emissions and promoting sustainable transportation. Studies have shown the positive effects of various transportation infrastructures, such as new rail and subway stations, bus stops, and improved street accessibility on retail sales. However, research specifically addressing the impact of EVCS on local businesses and foot traffic is lacking. While some pilot studies and surveys by EV charging companies have reported increased customer visits and spending at nearby businesses, these studies suffer from limitations in scale and lack of rigorous causal analysis, making generalization challenging. This research directly addresses this gap by employing a robust methodology and a large-scale dataset to determine the causal impact of EVCS on local business activity.

The study leverages data from the US Department of Energy's Alternative Fuels Data Center (AFDC) for EVCS information, including location, installation dates, and number of ports. Safegraph's Places and Spending data provided information on points of interest (POIs), monthly customer counts, spending amounts, and customer income distribution. Advan's Pattern data offered customer metrics such as median distance from home and median dwell time. Socio-demographic data came from the American Community Survey (ACS). The study period encompassed 2019 and January 2021 to June 2023, excluding 2020 due to the COVID-19 pandemic and data integration anomalies. The analysis focused on public EVCS and three POI categories: accommodation and food services, retail trade, and arts, entertainment, and recreation. To address endogeneity concerns (the possibility that EVCS are strategically placed in locations that already benefit from high foot traffic), the researchers used a three-pronged approach: propensity score matching (PSM) to create balanced comparison groups, difference-in-differences (DID) specifications with POI and county-by-month fixed effects, and event study analysis to assess parallel trends before EVCS installation. PSM was conducted in two steps: exact matching on POI categories and nearest-neighbor matching on various covariates, including built environment factors, socio-demographic variables, EV sales, and pre-treatment POI customer counts and spending. The DID analysis used a regression model with the natural log of customer counts or spending as the dependent variable and a binary treatment indicator along with the number of charging ports as independent variables. The model controlled for POI fixed effects and county-by-month fixed effects. Distance-varying treatment effects were explored by dividing the distances between POIs and EVCS into five bins (0-500m). Event study analysis examined the temporal dynamics of treatment effects. Additional analyses were conducted to explore heterogeneity in treatment effects based on EV charger type (Level 2 vs. DC fast chargers), POI type, and customer income groups.

The study's key findings consistently demonstrate a positive and statistically significant impact of EVCS installations on customer counts and spending at nearby POIs. Installing one EVCS boosted annual spending by 1.4% ($1478) in 2019 and 0.8% ($404) from January 2021 to June 2023. The impact was more significant for POIs within 100 meters of an EVCS (2.7% in 2019 and 3.2% in 2021-2023). The effects were also statistically significant in underprivileged communities, emphasizing the potential for EVCS to stimulate economic activity in these areas. In 2019, DC fast chargers had a more prominent effect on spending, potentially due to early EV adoption patterns among higher-income individuals. However, from 2021 to 2023, Level 2 chargers had a stronger impact, likely due to their increased prevalence and longer charging times encouraging more spending at nearby businesses. The analysis also revealed that EVCS installations, particularly Level 2 chargers, increased spending at restaurants, grocery/clothing stores and hotels from 2021-2023. In 2019, EVCS attracted customers from all income groups, but in 2021-2023, the effect was strongest for higher income groups. Finally, the study calculated the cumulative monetary impact of EVCS, showing that the addition of one EVCS resulted in an average annual spending increase of $22,813 in 2019 and $3412 in 2021-2023 across all nearby POIs. Considering all EVCS in California during the study periods, the total estimated spending increase was $6.7 million in 2019 and $19.5 million from January 2021 to June 2023.

The findings address the research question by demonstrating a clear positive causal relationship between EVCS installations and increased spending at nearby businesses. This effect is not only statistically significant but also economically meaningful, especially when considering the cumulative impact across all surrounding POIs. The results highlight the potential of EVCS to stimulate local economies, particularly in underprivileged areas where such economic boosts are particularly valuable. The shift from DC fast chargers being more influential in 2019 to Level 2 chargers in 2021-2023 reflects changing market dynamics and consumer behavior. The results have implications for policymakers, suggesting the importance of supporting EVCS deployment, especially in underserved communities, and for EVCS providers, who can leverage the positive economic externalities to create more profitable business models. The study's findings contribute to a deeper understanding of the complex interplay between transportation infrastructure, consumer behavior, and local economic development, enriching both the sustainable transportation and urban planning literature.

This study demonstrates the substantial positive economic impact of EVCS installations on nearby businesses in California. The findings highlight the economic benefits of EVCS deployment, particularly for businesses within 100 meters and in underprivileged areas. The shift in the relative importance of Level 2 and DC fast chargers over time reflects the evolving EV market. Future research should explore alternative identification strategies, investigate the specific mechanisms driving the observed spending increases, and examine the long-term effects of EVCS installations in the post-pandemic era.

The study acknowledges limitations related to the potential for unobserved confounding factors despite employing robust identification strategies. Data limitations, such as the exclusion of cash transactions from Safegraph's Spending data, might underestimate the true economic impact of EVCS. The analysis focused on public EVCS in California, limiting the generalizability of the findings to other regions or private EVCS. The COVID-19 pandemic also created unique conditions, possibly affecting the study's findings. Future research could explore alternative identification techniques and delve deeper into the mechanisms driving the observed effects.