In recent years, the global shift toward sustainable transportation has accelerated the adoption of electric vehicles (EVs), with the Regional Comprehensive Economic Partnership (RCEP) agreement playing a pivotal role in shaping trade dynamics. As a researcher focused on international trade networks, I aim to explore the structural evolution and influencing factors of electric vehicle trade among RCEP member countries from 2018 to 2022. The electric vehicle sector, particularly China EV developments, has become a cornerstone of green industrial policies, driving economic integration and environmental goals. This study employs social network analysis (SNA) and quadratic assignment procedure (QAP) regression to dissect the trade relationships, emphasizing the roles of economic scale, energy structures, logistics, and environmental regulations. By leveraging data from UN Comtrade, the World Bank, and other sources, I construct a dynamic trade network to identify key nodes and clusters, such as China EV hubs, and assess how differences in GDP, population, lithium battery trade, oil consumption, and carbon emissions influence trade ties. The findings reveal a densifying network centered on China, Japan, and South Korea, with core-periphery structures stabilizing over time. This analysis not only highlights the growing importance of electric vehicle trade in regional cooperation but also offers actionable insights for policymakers to foster synergy and sustainable growth within the RCEP framework.
The electric vehicle industry represents a transformative force in global markets, with China EV production leading innovation and export volumes. Under the RCEP agreement, which facilitates tariff reductions and streamlined customs procedures, member countries have witnessed enhanced trade flows in high-tech sectors like electric vehicles. However, existing literature often overlooks the network-based interdependencies and multi-factorial drivers specific to electric vehicle trade. My research addresses this gap by applying SNA to visualize and quantify the trade linkages, using metrics such as network density, centrality, and core-periphery analysis. For instance, the network density $D$ is calculated as:
$$D = \frac{2m}{n(n-1)}$$
where $m$ is the number of actual connections, and $n$ is the number of nodes (countries). This formula helps gauge the closeness of trade relationships, with values approaching 1 indicating tighter integration. Similarly, the average clustering coefficient $\bar{C}$ measures the tendency of nodes to form clusters:
$$\bar{C} = \frac{1}{n} \sum_{i=1}^{n} C_i, \quad C_i = \frac{e_i}{K_i(K_i-1)}$$
where $e_i$ is the number of edges between neighbors of node $i$, and $K_i$ is its degree. These SNA tools, combined with QAP regression, allow me to model the electric vehicle trade network as a function of economic, energy, and environmental variables, providing a holistic view of the RCEP landscape.
Data for this study spans 2018–2022, covering 15 RCEP member countries, and includes electric vehicle trade data classified under HS codes 870220 to 870380, as outlined in Table 1. These codes encompass various EV types, from hybrid to fully electric vehicles, ensuring a comprehensive scope. The variables selected for analysis, such as GDP, population, lithium battery trade, oil consumption, logistics performance index, and CO2 emissions, are derived from authoritative sources like the World Bank and BP Statistical Review. To construct the trade network, I represent countries as nodes and bilateral trade volumes as weighted edges, forming an adjacency matrix $M$ for each year. The overall network $G$ is defined as:
$$G = (N, M), \quad N = \{1, 2, \dots, n\}, \quad M = (E, A, W, T)$$
where $E$ denotes trade relationships, $A$ the number of connections, $W$ the trade value, and $T$ the time dimension. This setup enables dynamic tracking of electric vehicle trade evolution, with a focus on China EV exports and imports, which dominate the network.
| HS Code | Product Definition |
|---|---|
| 870220 | Vehicles with compression-ignition internal combustion piston engine and electric motor |
| 870230 | Vehicles with spark-ignition internal combustion piston engine and electric motor |
| 870240 | Vehicles solely with electric motor, no internal combustion engine |
| 870340 | Other vehicles with spark-ignition engine and electric motor, not externally chargeable |
| 870350 | Other vehicles with compression-ignition engine and electric motor, not externally chargeable |
| 870360 | Other vehicles with spark-ignition engine and electric motor, externally chargeable |
| 870370 | Other vehicles with compression-ignition engine and electric motor, externally chargeable |
| 870380 | Other vehicles solely with electric motor, no internal combustion engine |
The evolution of the electric vehicle trade network from 2018 to 2022 shows a consistent increase in connectivity, as summarized in Table 2. Network density rose from 0.438 to 0.500, indicating stronger trade ties, while the average clustering coefficient grew from 0.653 to 0.726, reflecting heightened clustering among countries like China EV producers and their partners. The average path length decreased from 1.613 to 1.399, suggesting improved efficiency in trade routes. These trends underscore the rapid integration of electric vehicle markets within RCEP, driven by initiatives such as China EV subsidies and regional green policies. Despite this progress, the overall density remains moderate, highlighting disparities where developed nations like Japan and South Korea lead, while developing countries like Myanmar and Cambodia lag due to infrastructural gaps.
| Year | Number of Relations | Network Density | Average Clustering Coefficient | Average Path Length |
|---|---|---|---|---|
| 2018 | 291 | 0.438 | 0.653 | 1.613 |
| 2019 | 340 | 0.486 | 0.685 | 1.610 |
| 2020 | 305 | 0.419 | 0.691 | 1.468 |
| 2021 | 342 | 0.471 | 0.689 | 1.516 |
| 2022 | 382 | 0.500 | 0.726 | 1.399 |
Centrality analysis reveals that China, Japan, and South Korea consistently exhibit high degree, closeness, and betweenness centrality, positioning them as hubs in the electric vehicle trade network. For example, in 2022, these countries achieved a degree centrality of 1, indicating direct connections to all other nodes, and a betweenness centrality of approximately 7.811, highlighting their role as intermediaries. This centrality can be expressed mathematically for node $i$ as:
$$C_B(i) = \sum_{j \neq k \neq i} \frac{\sigma_{jk}(i)}{\sigma_{jk}}$$
where $\sigma_{jk}$ is the total number of shortest paths between nodes $j$ and $k$, and $\sigma_{jk}(i)$ is the number of those paths passing through $i$. Such metrics emphasize how China EV exports facilitate trade flows between peripheral nations, reinforcing the network’s resilience. Countries like Thailand and Indonesia also show rising centrality, contributing to a multipolar structure that supports regional electric vehicle diffusion.
Core-periphery analysis further delineates the hierarchical structure of the electric vehicle trade network. Using Ucinet’s Core&Periphery algorithm, I classify countries based on coreness values: core (above 0.30), semi-periphery (0.20–0.30), and periphery (below 0.20). As shown in Table 3, China, Japan, South Korea, and Thailand maintained core status throughout 2018–2022, with China EV initiatives driving their dominance. In 2022, the core expanded to five countries, including Indonesia, while periphery nations like Laos and Cambodia saw minimal changes. This “layered固化” pattern suggests that while electric vehicle trade is growing, structural inequalities persist, necessitating targeted interventions to uplift marginalized economies.
| Country | 2018 Coreness | 2022 Coreness |
|---|---|---|
| China | 0.207 | 0.337 |
| Japan | 0.344 | 0.337 |
| South Korea | 0.226 | 0.337 |
| Thailand | 0.256 | 0.337 |
| Indonesia | 0.228 | 0.305 |
| Australia | 0.417 | 0.269 |
| Singapore | 0.256 | 0.265 |
| Malaysia | 0.225 | 0.252 |
| New Zealand | 0.304 | 0.252 |
| Vietnam | 0.207 | 0.226 |
| Philippines | 0.225 | 0.191 |
| Myanmar | 0.352 | 0.176 |
| Cambodia | 0.151 | 0.164 |
| Brunei | 0.156 | 0.150 |
| Laos | 0.151 | 0.121 |
To identify the factors influencing electric vehicle trade relationships, I conduct QAP regression analysis, which accounts for the relational nature of network data. The model specifies the trade network matrix $G$ as a function of multiple variable matrices:
$$G = f(\text{GDP}, \text{POP}, \text{Li}, \text{Oil}, \text{CO2}, \text{Log})$$
where GDP represents economic scale differences, POP population disparities, Li lithium battery trade variations, Oil oil consumption gaps, CO2 carbon emission differentials, and Log logistics performance distinctions. Variables like electrification rate and common language were excluded due to insignificance in preliminary correlations. The QAP results, based on 4,000 random permutations, are summarized in Table 4. For instance, GDP differences show positive coefficients (e.g., 0.252 in 2022), indicating that larger economic disparities foster electric vehicle trade, aligning with comparative advantage theory. Conversely, POP differences exhibit negative coefficients (e.g., -0.0339 in 2022), suggesting that similar population sizes enhance trade, possibly due to aligned market demands for China EV products.
| Variable | 2018 Coefficient | 2019 Coefficient | 2020 Coefficient | 2021 Coefficient | 2022 Coefficient |
|---|---|---|---|---|---|
| GDP | 0.0245* | 0.074* | 0.048* | 0.0611** | 0.0059** |
| POP | 0.0981*** | -0.0051** | -0.0101* | -0.0063* | -0.0339* |
| Li | 1.145* | 1.292** | 0.26* | 0.229* | 0.848* |
| Oil | 0.147*** | 0.153*** | 0.026*** | 0.028*** | 0.294*** |
| CO2 | 0.348* | 0.0293 | 0.0086** | 0.0198** | 0.028** |
| Log | 0.008** | 0.006* | 0.019* | 0.014* | 0.009* |
Energy-related factors, such as lithium battery trade (Li) and oil consumption (Oil), consistently positively influence electric vehicle trade, with coefficients like 0.848 for Li in 2022. This underscores the importance of battery supply chains in enabling China EV exports and regional integration. Similarly, oil consumption differences promote trade, as high-oil-use nations seek electric alternatives to meet environmental targets. Carbon emission (CO2) disparities show positive effects in most years (e.g., 0.028 in 2022), reflecting how emission pressures drive electric vehicle adoption. Logistics performance (Log) also positively correlates with trade strength, emphasizing that efficient transport infrastructure, crucial for electric vehicle distribution, enhances network connectivity. These findings collectively highlight the multi-dimensional drivers of electric vehicle trade, where economic, energy, and environmental factors intertwine to shape RCEP dynamics.
Building on these insights, I propose several policy recommendations to optimize the electric vehicle trade network within RCEP. First, establishing a graded division of labor in the electric vehicle industry can leverage strengths: China EV sectors should focus on battery technology and smart driving systems, Japan on solid-state batteries, and South Korea on semiconductors. This approach fosters technological complementarity and avoids duplication. Second, a “core-linked, periphery-nurtured” strategy can address disparities; for example, Australia and New Zealand could harmonize charging standards, while Vietnam and Cambodia benefit from EV industrial parks to boost local participation. Third, implementing a carbon footprint compensation mechanism, such as quantifying battery recycling rates, can align trade with sustainability goals, particularly for fossil-dependent nations like Brunei. Lastly, developing a digital logistics platform using blockchain technology can streamline supply chains between key hubs like China EV ports and ASEAN nodes, reducing delays and costs. These measures aim to create a synergistic electric vehicle ecosystem that promotes inclusive growth and green transformation across RCEP.
In conclusion, this study demonstrates the evolving structure of electric vehicle trade among RCEP members, characterized by increasing density, centralization around China EV actors, and stable core-periphery hierarchies. The application of SNA and QAP regression reveals that economic disparities, energy trade variations, and environmental regulations positively drive trade, while population differences hinder it. As the electric vehicle market expands, these findings offer a roadmap for enhancing regional cooperation, emphasizing innovation, equity, and sustainability. Future research could explore temporal shifts post-2022 or integrate machine learning to predict trade patterns, further enriching our understanding of this dynamic field.