In the era of the digital economy, data elements have emerged as critical drivers of industrial transformation, profoundly influencing sectors worldwide. The electric vehicle industry, particularly in China, exemplifies this shift, where data-driven insights are reshaping product development and market strategies. As a researcher focused on the China EV market, I investigate how commuting distance affects the range selection behavior of electric vehicle users, leveraging large-scale data to uncover patterns that can inform industry practices. This study is situated within the context of China’s rapid adoption of electric vehicles, supported by policies like the “14th Five-Year Plan for Digital Economy Development” and the “Data Twenty Articles,” which aim to establish robust data governance frameworks. Understanding user preferences in the China EV landscape is essential for optimizing product offerings and enhancing market segmentation.
The transition to electric vehicles is accelerating globally, with China leading in adoption rates due to government incentives and technological advancements. However, range anxiety remains a significant barrier for potential buyers, especially for battery electric vehicle (BEV) users who rely solely on battery power. Unlike plug-in hybrid electric vehicle (PHEV) users, BEV users must carefully consider their daily commuting needs when selecting a vehicle’s range. In this study, I explore the correlation between commuting distance and range demand, hypothesizing that longer commutes drive the selection of higher-range electric vehicles. By analyzing data from Tianjin in 2024, I employ econometric models to quantify this relationship, providing insights that can help manufacturers tailor their offerings to diverse user profiles in the China EV market.
Data for this analysis were sourced from the human-vehicle big data platform jointly developed by Automotive Data of China and China Mobile, which integrates anonymized user data with vehicle specifications. The dataset includes 11,219 observations from Tianjin in 2024, covering both BEV and PHEV users. Key variables include range mileage (in kilometers), commuting distance (categorized into short: 0–10 km, medium: 10–25 km, and long: >25 km), vehicle type (BEV or PHEV), price segments (e.g., below 200,000 CNY vs. above), and vehicle class (e.g., A00, A0, A, B, C). Control variables encompass demographic attributes such as gender, age, presence of children, and household structure. To ensure privacy, data were processed using privacy-preserving techniques that prevent raw data exposure, aligning with ethical standards for big data research in the China EV sector.
The primary methodological approach involves a dummy variable linear regression model to assess the impact of commuting distance on range selection. The model is specified as follows:
$$ Range_i = \beta_0 + \beta_1 Commute\_medium_i + \beta_2 Commute\_high_i + \alpha X_i + \epsilon_i $$
Here, \( Range_i \) represents the CLTC range mileage for user \( i \), \( Commute\_medium_i \) and \( Commute\_high_i \) are dummy variables indicating medium and long commuting distances, respectively, with short distance as the baseline. \( X_i \) denotes a vector of control variables, and \( \epsilon_i \) is the error term. Parameters \( \beta_1 \) and \( \beta_2 \) capture the marginal effects of commuting distance on range choice. To ensure robustness, I estimate two versions of the model: one without control variables and one with them, reporting results from the latter for comprehensive analysis. This approach allows me to isolate the influence of commuting distance while accounting for confounding factors in the China EV context.
The regression results, summarized in Table 1, reveal a statistically significant relationship between commuting distance and range selection. Users with medium and long commutes tend to choose electric vehicles with higher range mileage compared to those with short commutes. Specifically, the coefficients indicate that medium commuters select ranges approximately 17.823 km higher, while long commuters opt for ranges about 39.944 km higher, after controlling for demographic variables. These findings underscore the importance of commuting patterns in the decision-making process for electric vehicle purchases in China.
| Variable | Coefficient (Without Controls) | Coefficient (With Controls) |
|---|---|---|
| Commute_medium | 18.983*** (p<0.001) | 17.823*** (p=0.001) |
| Commute_high | 40.064*** (p<0.001) | 39.944*** (p<0.001) |
| Constant | 250.607*** (p<0.001) | 243.276*** (p<0.001) |
| Controls Included | No | Yes |
| Observations | 10,677 | 10,624 |
To delve deeper into the heterogeneity of these effects, I extend the analysis by incorporating interaction terms between commuting distance and vehicle type (BEV vs. PHEV). This is motivated by the premise that BEV users, due to their reliance on battery power alone, may exhibit greater sensitivity to commuting distance. The extended model is formulated as:
$$ Range_i = \beta_0 + \beta_1 (Commute\_medium_i \times BEV_i) + \beta_2 (Commute\_high_i \times BEV_i) + \beta_3 Commute\_medium_i + \beta_4 Commute\_high_i + \beta_5 BEV_i + \alpha X_i + \epsilon_i $$
In this equation, \( BEV_i \) is a dummy variable indicating pure electric vehicle users. The coefficients \( \beta_1 \) and \( \beta_2 \) represent the differential effects of commuting distance on range selection for BEV users relative to PHEV users, which I term the “BEV-commuting-range elasticity.” Results from this model, presented in Table 2, highlight that BEV users demonstrate significantly higher range sensitivity. For instance, in medium commuting scenarios, BEV users choose ranges 12.311 km higher than PHEV users, and this gap widens to 30.032 km for long commutes. In contrast, commuting distance does not significantly influence range choices for PHEV users, as indicated by the non-significant coefficients for \( \beta_3 \) and \( \beta_4 \). This divergence underscores the unique challenges faced by BEV users in the China EV market, where range anxiety is more pronounced.
| Variable | Coefficient (Without Controls) | Coefficient (With Controls) |
|---|---|---|
| Commute_medium × BEV | 12.506** (p=0.010) | 12.311** (p=0.011) |
| Commute_high × BEV | 30.405*** (p<0.001) | 30.032*** (p<0.001) |
| Commute_medium | -0.572 (p=0.721) | -1.451 (p=0.371) |
| Commute_high | 1.598 (p=0.554) | 0.801 (p=0.766) |
| BEV | 338.857*** (p<0.001) | 339.780*** (p<0.001) |
| Constant | 102.498*** (p<0.001) | 243.276*** (p<0.001) |
| Controls Included | No | Yes |
| Observations | 10,677 | 10,624 |
Further heterogeneity analyses examine how price segments and vehicle classes moderate the relationship between commuting distance and range selection among BEV users. I introduce dummy variables for price (Price_i = 1 if vehicle price < 200,000 CNY) and vehicle class (Mini_i = 1 for A00, A0, or A classes). The models are specified as follows for price and class interactions, respectively:
$$ Range_i = \beta_0 + \beta_1 (Commute\_medium_i \times Price_i) + \beta_2 (Commute\_high_i \times Price_i) + \beta_3 Commute\_medium_i + \beta_4 Commute\_high_i + \beta_5 Price_i + \alpha X_i + \epsilon_i $$
$$ Range_i = \beta_0 + \beta_1 (Commute\_medium_i \times Mini_i) + \beta_2 (Commute\_high_i \times Mini_i) + \beta_3 Commute\_medium_i + \beta_4 Commute\_high_i + \beta_5 Mini_i + \alpha X_i + \epsilon_i $$
The coefficients \( \beta_1 \) and \( \beta_2 \) in these models capture the “price-commuting-range elasticity” and “class-commuting-range elasticity,” measuring how commuting distance effects vary by affordability and vehicle size. Results in Tables 3 and 4 show that for BEV users, long commuters in lower price brackets select ranges 28.867 km higher than short commuters, while those in mini-class vehicles choose ranges 21.248 km higher. These effects are statistically significant, indicating that economic constraints and vehicle purpose amplify range sensitivity. Notably, for high-price or large-class vehicles, commuting distance has no significant impact, suggesting that premium segments prioritize other features over range. The negative coefficient for \( \beta_5 \) in both models reflects that in short-commute scenarios, higher-priced or larger-class vehicles tend to have longer ranges, consistent with market trends where cost and size correlate with advanced capabilities in the China EV industry.
| Variable | Coefficient (Without Controls) | Coefficient (With Controls) |
|---|---|---|
| Commute_medium × Price | 5.809 (p=0.363) | 5.502 (p=0.392) |
| Commute_high × Price | 27.234*** (p=0.006) | 28.867*** (p=0.004) |
| Commute_medium | 1.109 (p=0.819) | 0.790 (p=0.872) |
| Commute_high | 0.628 (p=0.929) | -1.368 (p=0.847) |
| Price | -187.712*** (p<0.001) | -186.955*** (p<0.001) |
| Constant | 577.486*** (p<0.001) | 573.775*** (p<0.001) |
| Controls Included | No | Yes |
| Observations | 4,775 | 4,752 |
| Variable | Coefficient (Without Controls) | Coefficient (With Controls) |
|---|---|---|
| Commute_medium × Mini | 9.757 (p=0.118) | 9.757 (p=0.120) |
| Commute_high × Mini | 19.915** (p=0.026) | 21.248** (p=0.018) |
| Commute_medium | -0.457 (p=0.924) | -1.303 (p=0.787) |
| Commute_high | 5.201 (p=0.401) | 3.217 (p=0.606) |
| Mini | -195.200*** (p<0.001) | -194.785*** (p<0.001) |
| Constant | 580.144*** (p<0.001) | 575.632*** (p<0.001) |
| Controls Included | No | Yes |
| Observations | 4,775 | 4,752 |
The findings from this study have profound implications for the electric vehicle industry, particularly in China, where market segmentation is crucial for sustained growth. The strong correlation between commuting distance and range selection emphasizes the need for manufacturers to integrate commuting data into product design and marketing strategies. For instance, electric vehicle companies can use these insights to develop targeted campaigns for BEV users with long commutes, highlighting vehicles with extended ranges. Additionally, the heterogeneity results suggest that affordable and compact electric vehicles should prioritize range enhancements to meet the demands of cost-conscious consumers who face lengthy daily travels. This approach can alleviate range anxiety and boost adoption rates in the competitive China EV market.
From a policy perspective, these results support initiatives that promote data-driven decision-making in the automotive sector. The Chinese government’s focus on data infrastructure, as outlined in the “Data Twenty Articles,” aligns with the need for granular analyses like this one. By fostering collaborations between data providers and industry stakeholders, policies can facilitate the development of electric vehicles that better match user needs. For example, urban planning could incorporate commuting patterns to optimize charging infrastructure, reducing barriers for BEV users in high-commute areas. Moreover, subsidies or incentives for long-range electric vehicles in regions with extensive commuting networks could accelerate the transition to sustainable transportation in China.
Despite these contributions, this study has limitations that warrant attention. First, the use of categorical commuting distance data, rather than continuous values, may obscure nuanced relationships. Future research could employ precise GPS-based metrics to enhance accuracy. Second, the straight-line distance calculation for commuting may not reflect actual road travel, potentially introducing measurement error. Incorporating navigation data or real-time traffic conditions would provide a more realistic assessment. Third, unobserved factors such as regional climate, terrain, and charging infrastructure availability could influence range selection but are not accounted for here. Expanding the model to include these variables would offer a holistic view of the China EV ecosystem. Lastly, the focus on Tianjin may limit generalizability; replicating this analysis across diverse Chinese cities could validate and refine these findings.
In conclusion, this research underscores the pivotal role of commuting distance in shaping electric vehicle range preferences, with BEV users exhibiting heightened sensitivity. The integration of econometric models with large-scale data has yielded actionable insights for manufacturers and policymakers in the China EV market. As the industry evolves, continued emphasis on user-centric design and data analytics will be essential for driving innovation and meeting the diverse needs of electric vehicle adopters. By addressing the identified limitations and exploring additional moderating factors, future studies can further illuminate the dynamics of electric vehicle adoption, contributing to a more sustainable and efficient transportation landscape in China and beyond.