The Impact of the Digital Economy on New Energy Vehicle Export Trade: Evidence from China

Abstract

In the digital economy era, artificial intelligence, big data, and 5G are widely applied across various industries. The deep integration of digitalization and traditional sectors has been facilitated by this trend, which has injected new momentum into industrial development. In this context, this paper employs panel data from 29 Chinese provinces that span the years 2017 to 2023. This paper transcends the constraints of current research by integrating the digital economy with the export of new energy vehicles. Furthermore, this paper provides a regional analysis of this impact, thereby contributing to the existing literature. The following are the conclusions: (1) The export of new energy vehicles is substantially stimulated by the development of the digital economy. (2) Exports are indirectly facilitated by the digital economy, which promotes technological innovation and financial services. (3) The digital economy shows a significantly greater impact on the export of new energy vehicles in the eastern and inland areas than in other regions. Based on these discoveries, the paper suggests four critical policy recommendations: expanded openness, technological innovation, intelligent digital marketing, and government support. The objective is to foster the sustainable growth of China’s new energy vehicle export trade. This paper offers theoretical support for the sustainability of Chinese enterprises’ competitiveness in the international market. It also provides policymakers and industry stakeholders with practical advice.

1. Introduction

The world is undergoing a substantial transformation characterized by digitization. Traditional sectors have significantly integrated emerging technology such as big data, artificial intelligence, and cloud computing. The digital economy is swiftly and profoundly altering the global industrial landscape. Moreover, it has become a crucial catalyst for global economic recovery and sustainability. Since the 2016 G20 Hangzhou Summit, China has ardently promoted the digital economy on a global scale. Furthermore, it initiated the G20 Digital Economy Development and Cooperation Initiative, indicating its elevation to a national strategic priority. In 2018, the National Conference on Cybersecurity and Informatization underscored the imperative to accelerate the development of the digital economy. Moreover, it promoted the progression of digital manufacturing, nascent industries, and inventive business paradigms. In 2021, the total value of China’s digital economy reached 45.5 trillion yuan, ranking it second globally, after the United States. The 14th 5-Year Plan for Digital Economy Development, released in 2022, emphasized the convergence of the digital economy with the real economy. It also set a target for related industries to constitute 10% of GDP, signifying China’s transition toward high-quality and sustainability-oriented digital advancement.

Simultaneously, the digital economy has expedited the advancement of green and low-carbon sectors, including new energy vehicles. Amid increasing global environmental consciousness and escalating energy challenges, the vehicle industry is experiencing a dramatic transition from conventional fuels to alternative energy sources. In this way, China assumes a prominent position in the exportation of new energy vehicles. It has additionally implemented digital technology throughout the industry chain to improve product quality and global competitiveness. According to the China Association of Automobile Manufacturers, the automotive sector is a major contributor to carbon emissions. Its emission levels are surpassed only by those of high-carbon industries, such as coal and heavy equipment. Nonetheless, pure electric vehicles produce nearly zero carbon dioxide, rendering their promotion crucial for energy conservation, emission reduction, and climate action. China commenced the development of the new energy vehicle sector in 2000 and currently possesses a robust technological basis. However, it persists in facing obstacles in international markets and rapid technical advancements. Consequently, it is imperative to leverage the digital economy to facilitate the industry’s global expansion. These endeavors are essential for facilitating industrial enhancement and fostering sustainability.

Digital technology persists in its advancement, augmenting the sustainability of economic and social institutions. Concurrently, the “dual carbon” plan has positioned new energy vehicles as a crucial green business endorsed by the government. Numerous studies concentrate on the impact of the digital economy on growth and manufacturing efficacy. Nonetheless, its effect on export performance in particular industries remains inadequately examined. The role of the internet economy in facilitating export growth within the new energy vehicle sector is notably underexamined. The techniques involved encompass technology empowerment, information alignment, and supply–demand optimization. However, scientific evidence about these pathways is limited. In this context, it is imperative to examine the influence of the internet economy on the exportation of new energy vehicles. It is essential to ascertain whether and how it improves product competitiveness and market expansion. This paper offers empirical validation for green trade policy and assists firms in attaining advantageous positions. These roles are essential throughout the reconfiguration of global value networks.

Existing literature generally agrees that the development of the digital economy could enhance industrial efficiency and improve market responsiveness. However, the specific mechanisms through which it promotes new energy vehicle exports—particularly through technological innovation—remain insufficiently explored. As a key driver of industrial upgrading, technological innovation may play a crucial mediating role in the impact of the digital economy on exports. The digital economy enhances information flow and improves resource allocation. These improvements encourage enterprises to increase R&D investment and accelerate technological progress. As a result, the quality and competitiveness of new energy vehicles are significantly improved. Specifically, the existing literature indicates that many scholars have undertaken pertinent research on the digital economy and the exportation of new energy vehicles. Prior research has extensively examined the definition, technological attributes, and sustainable development trajectories of the digital economy. Academics underscore the essential importance of developing digital infrastructure and allocating data resources to foster the high-quality advancement of the digital economy [1]. Additionally, many researchers have established multi-dimensional evaluation systems that integrate digital infrastructure, technological application, and industrial enhancement. Metrics such as internet user penetration and mobile device ownership are commonly employed to establish comprehensive quantitative frameworks [2]. Conversely, studies on new energy vehicle exports have predominantly concentrated on export volume, international competitiveness, and tariff-related matters. Most studies examine the subject through the lenses of industrial chain enhancement, brand impact, and policy assistance [3,4]. Nonetheless, numerous constraints persist in the current literature: (1) The correlation between the digital economy and exports of new energy vehicles has not been adequately explored. In addition, there is a deficiency of systematic theoretical frameworks and empirical evidence. (2) The indirect impact of the digital economy, especially via mechanisms like technological empowerment, on export performance has not been extensively examined. (3) Concerning regional heterogeneity, insufficient attention has been devoted to the impact of varying levels of digital economy development across regions in China on new energy vehicle exports.

This paper focuses on China’s new energy vehicle export trade and aims to systematically examine the impact of digital economy development on export performance. The specific research objectives are as follows:

(1)

Based on provincial panel data, this paper uses the entropy method to construct an index system for digital economy development. It measures the digital economy level in each province and establishes a regression model using export data of new energy vehicles to conduct empirical tests.

(2)

By analyzing the internal relationship between the digital economy and new energy vehicle exports, this paper constructs a mechanism model. It identifies the direct impact pathways, potential mediating mechanisms, and regional heterogeneity through which the digital economy affects exports.

(3)

According to the empirical results, this paper proposes targeted policy recommendations to promote China’s new energy vehicle exports. These suggestions aim to support green, high-quality, and export-oriented industrial upgrading.

This paper presents the subsequent marginal contributions: Initially, this paper presents a significant advancement in the research topic. The current literature predominantly concentrates on either the digital economy or the exportation of new energy vehicles. Nevertheless, limited research rigorously investigates the intrinsic connection between the two. This paper synthesizes the advancement of the digital economy with the new energy vehicle sector. It elucidates the mechanisms through which the digital economy influences exports from a theoretical standpoint and validates these mechanisms via empirical study. Secondly, this paper advances the utilization of research methodologies. This paper empirically analyzes the effect of the digital economy on new energy vehicle exports at the macro level through the development of a two-way fixed effects regression model. Additionally, it incorporates two mediating variables: financial services and technology innovation. It establishes a theoretical framework of “digital economy–mediating variables–new energy vehicle exports” to thoroughly study the indirect mechanisms. This offers a novel analytical viewpoint and methodological framework for pertinent research. Thirdly, this paper provides an innovative research perspective. This paper, having established that the digital economy facilitates both direct and indirect exports of new energy vehicles, further categorizes China into eastern, central, and western regions. This paper additionally analyzes heterogeneity by categorizing the 29 provinces into coastal and inland regions. It examines the geographical disparities in the impact of the digital economy on exports. This regional analysis provides policy ideas customized for local circumstances. At the same time, it also offers pragmatic suggestions for advancing the sustainability of China’s new energy vehicle sector development.

The structure of this paper is arranged as follows: (1) Literature Review. This section focuses on three aspects: the digital economy, the impact of digital economy development on export trade, and the impact of digital economy development on export trade. (2) Theoretical Analysis and Research Hypotheses. This part includes both the direct and indirect effects of the digital economy on new energy vehicle exports. (3) Research Design. This section covers sample selection, data sources, model specification, and variable definitions. Additionally, this part also includes descriptive statistical analysis and multicollinearity analysis. (4) Results. This part includes baseline regression results, robustness tests, heterogeneity analysis, and mechanism testing. (5) Discussion, Conclusions, Implications, and Research Limitations. This section presents the discussion, research conclusions, policy implications, limitations, and directions for future research.

2. Literature Review

2.1. The Digital Economy

The theoretical foundation of the digital economy could be traced back to the evolution of information economics. In the 1960s, systematic discussions on the knowledge industry recognized the independent role of information product manufacturing and service provision in economic activities. During this period, intellectual property also emerged as a core element. The preliminary research in information economics established the foundation for the theoretical framework of the digital economy and provided a robust basis for its further development. The digital economy initially referred to the integration of economic activity and information technology, using digital tools to build transaction systems and enhance productivity.

As digital adoption expands, it increasingly emphasizes sustainability, functional diversity, and structural complexity. Its structure now includes three key layers: technological implementation, e-commerce, and digital infrastructure. These layers are indicative of its multifaceted operational logic [5]. Additionally, as a novel economic paradigm, the digital economy exhibits unique structural characteristics. These could be understood in terms of industrial sectors, output structure, and production factors [6]. This method underscores the fundamental distinctions between conventional and digital economic models. As a result, the digital economy has developed into a model that is driven by innovation. Its core medium is the internet, and it is distinguished by its broad applicability, robust penetration, and high efficiency [7]. Its primary characteristic is the disruption and transformation of conventional production methods. Moreover, its operations are fueled by digital technologies, and its foundation is established by industries that offer digital goods and services [8]. Furthermore, aggregate output indicators may be implemented to evaluate their progression. As digital products and services become increasingly central, a fully digitalized value chain has emerged. It spans the entire process of production, exchange, and consumption [9].

Two interconnected development paths that jointly promote sustainability from within are followed by the digital economy. On the one hand, traditional industries are improved by information and communication technology, which optimizes processes and improves production factors. Conversely, industrial digitalization generates digital goods and services, which in turn alter business models and value chains [10]. In all, these two pathways foster the expansion of the digital economy from a technological system to a more comprehensive economic system. This transformation enhances the economy’s long-term sustainability and resilience. This evolution has also enhanced its theoretical significance. In technical terms, the digital economy encompasses fundamental digital services that are delivered through hardware and software. It encompasses sectors such as industrial control, healthcare, education, and publishing at the industrial level [11]. Consequently, it is a technological system and a fundamental component of the broader socioeconomic framework. In particular, digital trade has revolutionized the production and delivery of goods and services, thereby altering conventional business logic and operations [12]. It is widely recognized as a new form of economic structure driven by digital cognitive resources. Its development is further bolstered by sophisticated information technologies deployed worldwide. This perspective is also acknowledged in domestic research and serves as the theoretical foundation for the explanation of the digital economy’s development [13]. Thus, it is fundamentally a transformation model that is characterized by continuous industrial renewal, innovation, and adaptability.

Accurate measurement becomes indispensable as the digital economy expands in complexity and scope. Three primary methodologies are frequently implemented. Initially, the indicator system method establishes composite frameworks that encompass application levels, industrial development, and digital infrastructure. For instance, Wang et al. (2024) developed a multi-dimensional indicator system by utilizing data from Chinese cities [14]. Secondly, principal component analysis (PCA) is used to extract critical components in order to mitigate dimensionality and prevent multicollinearity. PCA was implemented by Zhang et al. (2023) to construct a digital economy index at the city level [15]. Third, in order to dynamically capture digital activity, big data and machine learning methods employ unstructured data, such as search indices or firm-level digital text [16]. The entropy method is particularly advantageous among these. It avoids subjective bias, considers both data variation and weight distribution, and objectively reflects the information weight of each indicator. Consequently, it is frequently implemented in empirical research to develop comprehensive evaluation systems.

2.2. Factors Influencing the Export of New Energy Vehicles

The export of new energy vehicles is affected by various factors. These factors include technological innovation, regulatory framework, industrial composition, market demand, and international collaboration. Technological innovation is widely acknowledged as a principal catalyst for export performance [17]. It enhances vehicle quality and value while fortifying the upstream industrial chain, thereby augmenting overall competitiveness. Recent research highlights a strong connection between the development of the new energy vehicle sector and industrial enhancement. Innovation drives this relationship by improving efficiency and quality, which are essential for export growth [18,19]. Furthermore, increased technological content alters export structures toward higher value-added segments, influencing global trade patterns. Innovation serves as a mediating factor in export promotion. Beyond acting as a direct catalyst, it indirectly enhances exports by refining product functionalities, strengthening brand recognition, and increasing international acceptance [20,21].

Government policies and the establishment of standardized systems constitute a crucial factor affecting the export of new energy vehicles. The U.S. government facilitates these exports via various policy supports and subsidy mechanisms, providing robust institutional backing for international expansion. Furthermore, standardized production processes and cohesive technical standards reduce production costs and enhance market access efficiency, thereby bolstering international competitiveness [22]. Simultaneously, variations in fuel prices, insurance frameworks, and regulatory environments among countries and regions directly influence export expenses and market receptivity [23]. The establishment of a robust export system relies not solely on technological advancements but also on efficient institutional collaboration.

Market factors are prominently addressed in the current body of research. The consumption capacity, human capital, and urban density of export destination countries significantly influence the import demand for new energy vehicles. In the EU market, per capita income is a significant determinant of demand, and the level of urbanization is also an important factor. The development of transportation infrastructure significantly influences consumers’ willingness to adopt new energy vehicles. Additionally, research results indicate that resource control and endorsements from influential opinion leaders could markedly improve consumers’ purchase intentions. Operational convenience is recognized as a significant practical factor affecting purchasing behavior. These factors collectively establish a linkage effect in export markets, influencing both demand and acceptance [24]. The export of new energy vehicles depends not only on supply-side factors such as technology and product quality. It also relies on the sustainability of socio-economic conditions and the development of supporting infrastructure in importing countries.

Regional and strategic cooperation pathways, including the Belt and Road Initiative and free trade agreements, significantly influence the development of export routes. China has formed partnerships with nations along the Belt and Road regarding lithium battery trade and infrastructure development. This synergy between resources and technology enhances export scale and optimizes the structural layout of export markets, thereby promoting sustainability [25]. Gravity model analysis indicates that free trade agreements and fuel bans contribute to the expansion of new energy vehicle exports, thereby supporting sustainable growth [26]. China’s new energy vehicle export data is sourced from the General Administration of Customs, as shown in Figure 1. From 2017 to 2023, the export value increased steadily, with a particularly rapid surge after 2020. In 2021, the export value exceeded USD 10 billion, representing a year-on-year growth of over 300%. This sharp increase is largely driven by China’s effective pandemic response, which ensured the smooth functioning of supply chains and logistics. Moreover, the expansion of the digital economy has significantly improved firms’ digital management and cross-border trade capabilities. These advancements have further supported the rapid growth of new energy vehicle exports. Leading Chinese manufacturers, including BYD, XPeng, and NIO, have accelerated their global expansion, particularly in the European market. By 2023, the export value had reached USD 41.812 billion, highlighting China’s rising competitiveness in the global new energy vehicle industry.

2.3. Impact of Digital Economy Development on Export Trade

The mechanism through which the digital economy influences export trade is marked by various levels and pathways. The effects are observable in the modification of trade structures at the macro-national level. Simultaneously, it facilitates the optimization of market behaviors at the micro-enterprise level, thereby supporting sustainability in both dimensions.

The extensive adoption of the internet and digital infrastructure is deemed essential for the expansion of service trade and goods exports at the macro level. Research conducted in 2002, which analyzed data from 14 service sectors, indicated that increased levels of network services significantly improve bilateral service trade [27]. Subsequent large-scale panel data studies encompassing over 160 countries further validated the dynamic interaction between internet penetration and economic growth. This interaction notably influences export trade. This effect is particularly evident in non-high-income countries [28,29]. These findings imply that digital infrastructure improves the efficiency of traditional economic systems. By enabling innovative service models, it strengthens export capacity and supports long-term sustainability. Furthermore, the advancement of the digital economy has been positively associated with export performance in the cultural and agricultural sectors. This suggests its growing influence across a wide range of industries. The positive relationship is particularly pronounced among trading partners exhibiting substantial economic, institutional, and cultural disparities [30]. This finding highlights the distinct function of digital technology in overcoming institutional barriers and cultural divides, thus promoting sustainability in international trade. Figure 2 uses a stacked area chart to show the changing shares of the digital economy and other sectors in China from 2016 to 2023. The digital economy’s share steadily increased, while that of other sectors declined. Compared with bar charts, area charts better illustrate cumulative contributions and structural shifts over time. In 2016, the digital economy accounted for 30.2% of GDP. By 2023, this figure had risen to 42.8%, representing an increase of 12.6 percentage points. This upward trend highlights the growing role of the digital economy in China’s economic structure. During the COVID-19 outbreak in 2019, digital trade models showed greater resilience than traditional trade. The rapid growth of online services helped cushion the impact on conventional service industries, underscoring the adaptability of the digital economy to external shocks.

At the micro-enterprise level, the digital economy profoundly influences corporate export behavior by reshaping transaction organization and market channels. With the rise of virtual organizations and e-commerce platforms, enterprises are increasingly able to offer customized products and services. As a result, they can bypass traditional export constraints and integrate more quickly into global markets [31]. Specifically, higher levels of internet access and e-commerce adoption are associated with greater export intensity among enterprises. It shows that digital platforms play a crucial role in driving corporate export performance [32]. Moreover, research highlights that e-commerce platforms offer particular benefits to agricultural and food enterprises. These platforms enable such firms to quickly enter global value chains and expand their economies of scale [33]. In addition, the application of robotics and automation technologies enhances production efficiency. It also increases product added value, thereby improving the competitiveness of export goods [34].

From a mechanistic perspective, the digital economy promotes export trade through six primary pathways. First, e-commerce reduces information search and transaction costs, thereby improving the matching efficiency between enterprises and consumers. Second, communication technology enhances information transparency, which reduces asymmetries and lowers international trade barriers [35,36]. Third, digitalization removes temporal and spatial constraints, enabling transactions to occur anytime and anywhere [37]. Fourth, digital technologies improve order management, supply chain coordination, and logistics efficiency, all of which are essential for cross-border trade [38]. Fifth, digital platforms generate network effects that expand international market coverage [39,40]. Sixth, highly digitalized enterprises tend to increase capital investment in innovation and R&D, which further optimizes the structure and value of export products [41,42].

The uneven development of regional digital economies results in disparities in export performance. The impact of the digital economy on export performance in China’s trade with ASEAN countries is stronger in markets that feature advanced digital infrastructure. This highlights the importance of digital readiness in enhancing export outcomes. Consequently, these countries demonstrate accelerated export growth. Digital platforms facilitate access to extensive market information for enterprises, thereby diminishing geographical and institutional barriers. As a result, companies could enter foreign markets more effectively, thus enhancing long-term sustainability in their global expansion efforts.

3. Theoretical Analysis

3.1. The Direct Impact of Digital Economy Development on China’s New Energy Vehicle Export Trade

In the context of the rapid development of the digital economy, China’s export trade in new energy vehicles has seen significant growth. This growth significantly enhances the sustainability of green industrial development. This trend is fundamentally driven by the deep integration of various digital elements. These encompass digital payment, data resources, digital marketing, and digital technologies, which function synergistically. The efficiency, scale, and competitiveness of new energy vehicle exports have been significantly enhanced by these factors.

Digital payment serves as a crucial facilitator of the digital economy by significantly lowering the costs and risks associated with cross-border transactions of new energy vehicles. Conventional payment systems frequently encounter issues, including intricate contract processes, convoluted remittance pathways, and elevated information security vulnerabilities. Conversely, digital payment employs advanced encryption technologies and secure protocols to facilitate instant and secure transactions. This enhances exporters’ expectations for stable payments, mitigates financial and fraud risks, and streamlines payment procedures. By streamlining intermediary steps in the cross-border payment chain, this approach improves payment efficiency. At the same time, it helps lower labor costs and reduces extra expenses resulting from logistics delays. Consequently, it offers enhanced efficiency and convenience in financial support for the export of new energy vehicles from China.

Data have emerged as a fundamental production factor in the digital economy era. It enhances the efficiency of resource allocation and production decisions in the export trade of new energy vehicles. Utilizing technologies like real-time data collection through Internet of Things devices and cloud-based processing enables corporations to effectively monitor essential information. This encompasses global market demand, supply chain stability, and policy trends. These insights facilitate improved decision-making and operational modifications. Utilizing these insights allows for adjustments in production plans, inventory management, and optimization of transportation routes [43]. Data-driven supply chain optimization enhances the efficiency of resource allocation. It enhances coordination efficiency between upstream and downstream firms via data-sharing mechanisms. This facilitates the stable and efficient integration of new energy vehicles into global markets. Simultaneously, big data analytics enables corporations to identify potential markets and assess risk factors. The approach enhances decision-making foresight and precision, while partially mitigating resource misallocation resulting from information asymmetry in conventional trade.

Digital marketing influences the demand side, consistently increasing the export scale of new energy vehicles. In contrast to conventional marketing techniques, digital marketing utilizes mobile internet, social media, and big data technologies. It attains precision, interactivity, and globalization in marketing. New energy vehicle companies could utilize user behavior analysis to precisely determine consumer preferences and purchasing power across various countries and regions. This enables the development of more localized and differentiated product and pricing strategies. Furthermore, digital marketing overcomes the constraints of unidirectional information distribution, thereby enabling immediate engagement with consumers via social media platforms. Such interaction improves brand recognition and strengthens user engagement. It also fosters consumer trust and raises their willingness to purchase new energy vehicles. In addition, the development of online sales platforms allows new energy vehicle companies to transcend traditional geographical market boundaries. As a result, they achieve greater global brand influence and more efficient market expansion.

Digital technology is essential for optimizing and intelligently upgrading the structure of the new energy vehicle industry. It facilitates collaboration within the industry chain and enhances the efficiency of supply chain management. It further promotes technological advancement in the domestic replacement of essential components. In the motor, battery, and electronic control system, domestic firms have progressively reduced the technological disparity with foreign brands via digital innovation. This has increased the technological content and value of exported products [44]. In addition, the incorporation of intelligent technologies, including artificial intelligence and big data algorithms, enhances the intelligence of new energy vehicles. This encompasses driving assistance, energy management, and remote-control functionalities. Additionally, these technologies enhance vehicle longevity and improve market acceptance internationally. Furthermore, digital platforms facilitate the rapid diffusion of technology and the sharing of knowledge. They facilitate the development of industrial clusters. The Yangtze River Delta region has effectively established a “4-h new energy vehicle industry circle.” This greatly improves the region’s capacity for export responsiveness and coordination.

The digital economy facilitates the high-quality development of China’s new energy vehicle exports through various dimensions. This approach improves sustainability by optimizing payment systems and reallocating resources. It also enhances marketing strategies and supports technological upgrades [45]. Based on the above analysis, the following research hypothesis is proposed:

H1. 

The digital economy shows a positive effect on China’s new energy vehicle export trade.

3.2. The Indirect Impact of Digital Economy Development on China’s New Energy Vehicle Export Trade

3.2.1. Digital Economy, Technological Innovation, and New Energy Vehicle Export

The advancement of the digital economy facilitates technological innovation in the export of new energy vehicles. The information diffusion path is experiencing a digital transformation. This transitions the information distribution mechanism from a centralized model to a decentralized one. Consequently, there is an expansion in consumer demand. Users’ understanding of new energy vehicle exports extends beyond that of conventional vehicles. Cars are anticipated to develop into terminals featuring autonomous driving, zero-emission travel, and intelligent interaction. Corporations must implement technological innovation to address diverse consumer demands. The digital economy promotes collaboration in research and development between corporations and research institutions. This enhances the innovation capacity of corporations. Simultaneously, the integration of the new energy vehicle export sector with digital technology facilitates the advancement of innovative technologies. The innovation spillover effect of digital technology is also realized in the new energy vehicle export sector.

Technological innovation enhances the export of new energy vehicles by improving product quality and production efficiency. It further facilitates firms’ upgrading within the global value chain. Moreover, as a key production factor, it could redefine a country’s comparative advantage, thus reinforcing its position in international trade. Technology serves as the primary driver in the development of new energy vehicle exports. Advancements in three primary domains—battery technology, motor efficiency, and electronic control systems—could significantly improve performance, safety, and endurance. This enhances export stability and competitiveness. Secondly, corporations exhibit considerable variation in production efficiency, organizational structure, and scale. Only firms exhibiting elevated productivity levels could adequately cover the fixed costs associated with exportation [46]. Technological innovation enhances production efficiency. It assists corporations in addressing diminishing marginal returns and enhancing the marginal benefit of input factors. The export of new energy vehicles is characterized by its reliance on advanced technology. Technological innovation represents the most effective method for enhancing production efficiency. It increases the propensity of corporations to engage in export activities. Furthermore, technological innovation facilitates the advancement of developing countries within the global value chain. Their focus transitions from processing and assembly to research and development. The improvement in global value chain position increases export competitiveness [47]. Based on the above analysis, the following research hypothesis is proposed:

H2. 

The development of the digital economy could positively impact new energy vehicle exports by promoting technological innovation, which acts as a mediator.

3.2.2. Digital Economy, Financial Services Development, and New Energy Vehicle Export

The development of the digital economy has fostered the growth of digital finance. It encompasses a range of innovative models, including mobile payments, online lending, and blockchain-based solutions. These models reduce barriers to financial services, enhance service efficiency, and broaden coverage. The application of digital technology significantly improves risk assessment and management in financial institutions. It leverages big data analysis, artificial intelligence, and a range of advanced analytical techniques to achieve greater precision. This facilitates enhanced personalization and efficiency in financial services [48]. By enabling the rapid global flow of capital, the digital economy expands financing channels for new energy vehicle exporters. As a result, it contributes to lower financing costs and improved access to international markets.

With advancements in financial development, new energy vehicle export companies are increasingly able to secure funding. This funding facilitates technology enhancement, production growth, and market expansion. Utilizing digital finance platforms enables corporations to efficiently obtain loans and issue bonds. Innovations in financial instruments enable corporations to more effectively manage risks associated with exchange rates and market fluctuations. This enhances their competitiveness in the global market. Digital finance development fosters innovations in supply chain finance. It assists corporations in optimizing supply chain cash flow management and minimizing operating expenses. Chinese new energy vehicle export corporations exemplify how the digital economy offers substantial financial backing and facilitates market expansion. Online digital finance platforms provide corporations with access to cost-effective financing options for technology adoption and market expansion. Corporations utilize big data analysis to effectively target the European and Southeast Asian markets. Cross-border e-commerce platforms facilitate rapid export growth.

The advancement in financial services development offers renewed impetus for the export of new energy vehicles. Financial innovation and accelerated capital flow enhance corporate access to funding and improve risk management capabilities. The utilization of digital technology facilitates the gathering of market information, enhances e-commerce growth, and improves brand development. The expansion of the digital economy would enhance the development potential of new energy vehicle exports. Based on the above analysis, the following research hypothesis is proposed:

H3. 

The development of the digital economy could positively affect new energy vehicle exports by promoting financial services development, which plays a mediating role.

As shown in Figure 3, this paper summarizes the above theoretical analysis. The paper first analyzes the effect of the development level of the digital economy on the export of new energy vehicles. Furthermore, it analyzes the mediating roles of technological innovation and financial services development. In conclusion, based on these analyses, Hypotheses 1–3 are proposed. This paper will next analyze and test these hypotheses, further validating the effectiveness of the theoretical model.

4. Research Design

4.1. Sample Selection and Data Sources

To ensure data consistency and completeness, this paper focuses on 29 Chinese provinces from 2017 to 2023. Qinghai, Tibet, Hong Kong, Macao, and Taiwan are excluded due to data limitations. Stata 18.0 is used to examine how the development of the digital economy affects new energy vehicles in China. The data used in this paper are primarily obtained from authoritative official sources to ensure reliability and accuracy. The dependent variable, the logarithm of new energy vehicle export value (LnED), is derived from the General Administration of Customs of China. The core explanatory variable, the level of the digital economy (Digital), is calculated using the entropy weight method based on indicators from multiple dimensions. Control variables are sourced from several official statistical publications. Provincial GDP (Gdp), trade openness (Open), urbanization rate (Urb), and foreign direct investment (Fdi) are collected from the China Statistical Yearbook and the statistical yearbooks of individual provinces and autonomous regions. The share of the secondary industry (industry) is obtained from the National Bureau of Statistics of China. In addition, two mediating variables are included. The level of technological innovation (Lnip) and the development of financial services (Fin) are extracted from the provincial statistical yearbooks and the China Financial Yearbook, respectively. The variables and related descriptions are as follows.

4.1.1. Explanatory Variable

In the baseline regression, this paper employs the entropy method to construct a composite index of the digital economy. This method allows for objective weighting and effectively captures the relative contribution of each indicator, making it suitable for multi-dimensional index construction. In addition, principal component analysis (PCA) is used in the robustness test to remeasure the digital economy. PCA reduces dimensionality by extracting key information, minimizing redundancy, and avoiding multicollinearity, thereby helping to verify the robustness of the regression results.

Specifically, this paper applies the entropy method to calculate a composite score for the digital economy. An evaluation system of digital economy development level is constructed. Based on this system, the information entropy method is used to determine the weight of each indicator. First, all selected indicators are standardized. Absolute values are converted into relative values. Since all indicators of digital economy development are expected to positively influence industrial export trade, they are uniformly treated as positive indicators. Accordingly, a positive transformation is applied to all variables to ensure consistency in the direction of their effects. Standardized data are then obtained.

First, positive and negative indicators are processed using the following formulas:

$$x_{ij}={\displaystyle \frac{x_{ij}-min\left\{x_j\right\}}{max\left\{x_j\right\}-min\left\{x_j\right\}}}$$

(1)

$$x_{ij}={\displaystyle \frac{max\left\{x_j\right\}-x_{ij}}{max\left\{x_j\right\}-min\left\{x_j\right\}}}$$

(2)

Specifically, max{x_j} denotes the maximum value of the indicator across all years, and min{x_j} denotes the minimum value. The normalized value is denoted as x_ij.

Subsequently, the weight of the indicator is calculated and represented as ω_ij.

$${\omega }_{ij}={\displaystyle \frac{x_{ij}}{{{\displaystyle \sum }}_{i=1}^m{x}_{ij}}}$$

(3)

The information entropy of each indicator is calculated as follows:

$$e_j=-{\displaystyle \frac{1}{lnm}}{\displaystyle \sum }_{i=1}^m{\omega }_{ij}\times ln{w}_{ij}$$

(4)

The redundancy of information entropy is calculated as follows:

$$d_j=1-e_j$$

(5)

The indicator weights are calculated based on information entropy redundancy as follows, where m represents the number of evaluation years:

$${\phi }_j={\displaystyle \frac{d_j}{{{\displaystyle \sum }}_{i=1}^m{d}_j}}$$

(6)

This paper designs an evaluation system for the digital economy development level based on its essential characteristics and development patterns. The design follows the principles of scientific rigor, objectivity, systematization, and comprehensiveness. It draws on the indicator framework proposed by Guo Feng (2020) [49]. The system includes three secondary indicators: digital infrastructure, digital industry, and digital finance. It further consists of 13 tertiary indicators. The detailed calculation results are presented in

Table 1. Indicators and weight of digital economy development levels of China.

Primary Indicator	Secondary Indicator	Tertiary Indicator	Weight	Data Source
Digital Economy Development Level	Digital Infrastructure	Number of Domain Names	0.0647	China Statistical Yearbook China City Statistical Yearbook
		Number of IPv4 Addresses	0.0609
		Number of Internet Broadband Access Ports	0.0889
		Mobile Phone Penetration Rate	0.0965
		Optical Cable Length per Unit Area	0.0686
	Digital Industry Development	Number of Information Technology Enterprises	0.3131	China Statistical Yearbook
		Number of Enterprises with Websites per 100 Enterprises	0.1027
		Proportion of Enterprises Engaged in E-commerce Transactions	0.0614
		E-commerce Sales Volume	0.0879
		Software Business Revenue	0.0450
	Digital Inclusive Finance Index	Coverage Breadth Index	0.0934	Measuring the Development of Digital Inclusive Finance in China: Index Compilation and Spatial Characteristics
		Usage Depth Index	0.0964
		Digitalization Level Index	0.1017

.

4.1.2. Explained Variable

This paper follows the approach of Jing Shouwu (2023), using the natural logarithm of the total new energy vehicle export value of each province as the explained variable [50]. The data are sourced from the General Administration of Customs. A higher value indicates a greater export volume of new energy vehicles.

4.1.3. Control Variables

(1)

Economic development level (Gdp). This paper follows Fan Xin (2020) [51] and uses per capita GDP to measure economic development level. Higher per capita GDP usually indicates stronger technological innovation capacity. It helps improve production efficiency and reduce production costs, thereby promoting new energy vehicle export.

(2)

Trade openness (Open). Based on the method by Niu Xiaoyu (2021), this paper uses the ratio of total annual import and export trade value to regional GDP for each province as the core indicator of trade openness [52]. This indicator effectively reflects the degree of regional integration into the global value chain and international economic cooperation. The data are sourced from provincial statistical yearbooks.

(3)

Urbanization level (Urb). This paper uses the proportion of the urban population in each region to measure the urbanization level. The data come from provincial statistical yearbooks. Yan (2019) [53] found a significant correlation between regional industrial agglomeration and local urbanization level. Higher urbanization rates lead to industrial concentration and scale effects, which promote export trade growth.

(4)

Foreign direct investment (Fdi). This paper analyzes the intensity of foreign capital inflow based on the percentage of foreign direct investment in regional GDP. The data are from provincial statistical yearbooks. Chen (2019) argues that technology spillover effects brought by foreign direct investment could enhance corporate innovation and management, thus promoting industrial structure optimization and export trade development [54].

(5)

Industrialization level (Industry). The level of industrialization is measured by the proportion of industrial added value to regional GDP, following the World Bank’s method. Data come from the China Statistical Yearbook. Zhu (2017) argues that improving industrialization could optimize the new energy vehicle supply chain, reduce production costs, and enhance export competitiveness [55].

4.1.4. Mediating Variables

(1)

Technological innovation (Lnip). Technology level is a key factor in enhancing export competitiveness. Especially with the rise of the digital economy, its value continues to emerge. The development of the digital economy could accelerate technological progress and drive innovation. Zhao (2021) [56] points out that technological innovation could promote product quality upgrading, thereby improving export stability and competitiveness. Therefore, technological innovation is used as a mediating variable, measured by the logarithm of the number of patent grants.

(2)

Financial services development (Fin). Digital financial services, such as supply chain finance and insurtech, reduce trade risks and improve transaction efficiency. Thus, financial services development is chosen as a mediating variable. Following Zhou (2004), it is measured using the ratio of the sum of deposits and loans of financial institutions to GDP in each province [57]. The data come from the China Financial Yearbook. The specific variables and their descriptions are presented in

Table 2. Variable symbol and source.

Variable	Code	Symbol	Source
Explained Variable	LnED	New Energy Vehicle Export Trade Volume	General Administration of Customs of China
Explanatory Variable	Digital	Digital Economy Development Level	Entropy Weight Method Calculation
Control Variables	Gdp	Economic Development Level	China Statistical Yearbook
	Open	Trade Openness Degree	Statistical Yearbooks of Provinces and Autonomous Regions
	Urb	Urbanization Level	Statistical Yearbooks of Provinces and Autonomous Regions
	Industry	Industrialization Level	National Bureau of Statistics of China
	Fdi	Foreign Direct Investment	Statistical Yearbooks of Provinces and Autonomous Regions
Mediating Variables	Lnip	Technological Innovation	Statistical Yearbooks of Provinces and Autonomous Regions
	Fin	Financial Services Development Level	China Financial Yearbook

.

4.2. Research Model

4.2.1. Baseline Regression Model

This paper uses a panel data model for empirical analysis. The dependent variable is the natural logarithm of the new energy vehicle export trade value. The core independent variable is the digital economy development index measured by the entropy method. To ensure the robustness and reliability of the model, a series of control variables is added. These include the level of regional economic development, the degree of trade openness, the level of urbanization, the scale of foreign direct investment, and the level of industrialization. The model is specified as Equation (7).

LnED_it = â0 + â1Digital_it + â2Gdp_it + â3Open_it + â4Urb_it + â5Industry_it + â6Fdi_it + ì_it

(7)

In the model, i represents different provinces, and t represents different years. ì_it is the random error term. LnED is the dependent variable. Digital is the independent variable. The control variables include five indicators: economic development (Gdp), openness (Open), urbanization (Urb), industrialization (Industry), and foreign direct investment (Fdi).

4.2.2. Mediating Effect Model

The development of digital economy in China may promote technological innovation and financial services. These factors then play a mediating role in boosting new energy vehicle exports. Therefore, this paper adopts the three-step method proposed by Wen Zhonglin (2014) to test the mediating mechanism [58]. The mediating variables are technological innovation (Lnip) and financial services (Fin). Technological innovation is measured by the logarithm of patent applications granted. Financial services level is measured by the ratio of total deposits and loans of financial institutions to GDP in each province. The mediating effect model is as follows:

LnED_it = â0 + â1Digital_it +â2Industry_it + â3Open_it + â4Urb_it + â5Gdp_it + â6 Fdi_it + ì_it

(8)

M_it = á0 + á1Digital_it + á2Industry_it + á3Open_it + á4Urb_it + á5Gdp_it + á6Fdi_it + ì_it

(9)

LnED_it = è + è0M_it + è1Digital_it+è2Industry_it + è3Open_it + è4Urb_it + è5Gdp_it + è6Fdi_it + ì_it

(10)

Here, M_it represents the mediating variable. Other indicators are consistent with the baseline regression model.

4.3. Descriptive Statistics

Table 3. Descriptive statistics.

Variable	Obs.	Mean	Median	SD.	Min	Max
LnED	203	16.23	16.71	3.516	7.467	0.747
Digital	203	0.188	0.140	0.130	0.0580	0.747
Industry	203	0.304	0.308	0.0730	0.101	0.498
Open	203	0.254	0.147	0.230	0.0270	0.977
Gdp	203	13,317	10,016	8980	5771	49,352
Urb	203	0.634	0.622	0.112	0.334	1.166
Fdi	203	0.0170	0.0140	0.0180	0	0.101
Lnip	203	10.11	10.09	1.195	7.394	12.40
Fin	203	3.608	3.387	1.177	1.279	8.164

presents the descriptive statistics of the key variables. The average value of new energy vehicle exports is 16.23, with a median of 16.71. This indicates a relatively high and concentrated export level across regions. However, the standard deviation reaches 3.516, reflecting significant regional disparities. Some regions export far more than others, showing clear spatial inequality. In comparison, the digital economy index has a mean of 0.188 and a median of 0.140. The standard deviation is 0.130, with values ranging from 0.058 to 0.747. This suggests uneven digital development across regions. A few regions are highly digitized, while most remain underdeveloped. The distribution is also right-skewed, further confirming digital concentration in advanced areas. Notably, both variables show substantial spatial variation. Regions with high digital economy scores often correspond to higher export volumes. This similarity suggests a potential link between digital capacity and export performance.

Furthermore, a comparison of the distributional characteristics of the two variables reveals considerable regional heterogeneity. This heterogeneity exists in both new energy vehicle export performance and digital economy development. Their similar patterns in terms of standard deviations and extreme values suggest a potential structural linkage. Specifically, the much higher maximum value of the digital economy index relative to the mean reflects a concentration of digital advantages. These advantages are evident in areas such as infrastructure, data application, and information system development. Likewise, the wide range in new energy vehicle export values shows that some regions are also leading in terms of export performance. Therefore, whether such leading positions are related to the level of digital economy development warrants further analysis.

In terms of the mediating variables, there are significant differences in technological innovation and financial services development across provinces in China. On the one hand, these differences result from variations in talent resources and levels of economic development. On the other hand, the development of the financial services sector shows a clear trend of concentration. Regions with abundant production factors and better infrastructure tend to have higher levels of development.

4.4. Correlation Analysis

Correlation analysis helps to preliminarily assess the rationality of selected variables in the empirical study by examining whether correlations exist among them. The correlation coefficient measures the linear relationship between variables, ranging from −1 to 1. A larger absolute value indicates a stronger correlation, while the sign reflects the direction.

Table 4. Correlation test results.

LnED	Digital	Industry	Open	Gdp	Fdi	Urb
LnED	1
Digital	0.538 ***	1
Industry	0.133 *	0.0910	1
Open	0.461 ***	0.704 ***	−0.166 **	1
Gdp	0.351 ***	0.589 ***	−0.280 ***	0.883 ***	1
Fdi	0.258 ***	0.606 ***	−0.0290	0.542 ***	0.374 ***	1
Urb	0.389 ***	0.442 ***	−0.0520	0.717 ***	0.645 ***	0.275 ***

Note: ***, **, and * represent significance at the 1%, 5%, and 10% levels, respectively.

presents the correlation test results and further confirms the relationship between the core variables. The correlation coefficient between new energy vehicle export and digital economy development is 0.538, significant at the 1% level. This indicates a strong and positive association. Combined with the descriptive statistics in Table 3, this finding suggests that regions with a higher digital economy index tend to achieve better export performance. Among all variables, digital economy development shows the strongest correlation with new energy vehicle export. The coefficients for openness, urbanization, economic development, and foreign direct investment are 0.461, 0.389, 0.351, and 0.258, respectively. All are positive and statistically significant, yet notably weaker than that of the digital economy. This pattern highlights the central role of digital transformation in enhancing export capacity. The strong correlation may reflect how digital infrastructure improves production efficiency, reduces information and transaction costs, and supports integration into global value chains. These mechanisms help explain why provinces with stronger digital economies tend to export more new energy vehicles.

4.5. Multicollinearity Test

Among the methods for diagnosing multicollinearity, the variance inflation factor (VIF) is one of the most commonly used indicators. The VIF value is positively related to the severity of multicollinearity. This paper uses the VIF test to diagnose potential multicollinearity in the model. The detailed results are shown in

Table 5. Results of multicollinearity test.

Variable	VIF	1/VIF
Open	8.180	0.122
Gdp	5.330	0.188
Digital	2.570	0.389
Urb	2.190	0.456
Fdi	1.820	0.550
Industry	1.260	0.794
Mean	VIF	3.560

. The test results show that the VIF value is 3.560, which is below the critical threshold of 5. Therefore, it can be concluded that there is no significant multicollinearity among the selected explanatory variables.

5. Results

5.1. Regression Results

As shown in

Table 6. Baseline regression results.

Variable	(1)	(2)	(3)	(4)	(5)	(6)
	LnED	LnED	LnED	LnED	LnED	LnED
Digital	28.541 ***	32.332 ***	32.890 ***	28.089 ***	27.961 ***	24.914 ***
	(6.527)	(7.331)	(7.833)	(6.997)	(6.986)	(5.704)
Industry		22.112 ***	25.571 ***	28.652 ***	26.954 ***	25.394 ***
		(3.271)	(3.945)	(4.727)	(4.379)	(4.103)
Open			22.595 ***	15.666 ***	14.365 ***	12.082 **
			(4.352)	(3.126)	(2.831)	(2.313)
Gdp				0.001 ***	0.001 ***	0.001 ***
				(5.231)	(5.174)	(4.413)
Fdi					−22.605	−19.200
					(−1.451)	(−1.229)
Urb						7.458 *
						(1.695)
_cons	10.848 ***	3.404	−3.494	−12.506 ***	−11.113 ***	−13.134 ***
	(12.922)	(1.408)	(−1.251)	(−4.009)	(−3.415)	(−3.808)
N	203	203	203	203	203	203
R2	0.198	0.245	0.320	0.414	0.421	0.431
F	42.597	27.843	26.811	30.049	24.617	21.221

Note: ***, **, and * indicate significance at the 1%, 5%, and 10% confidence levels, respectively. Values in parentheses are t-statistics based on robust standard errors.

, the coefficient of digital economy is positive in all models and significant at the 1% level. This indicates that the development of the digital economy has a significant positive impact on new energy vehicle exports. As more control variables are added, the coefficient of digital economy decreases from 28.541 to 24.914. Although there is a slight decline, the coefficient remains at a high level. This indicates that the positive impact of the digital economy on new energy vehicle exports is significant.

In terms of other control variables, the level of industrialization has a substantial positive impact on the export of new energy vehicles. A stronger promotion effect of industrial development on exports is suggested by the gradual increase in its coefficient as the model improves. Additionally, exports are significantly bolstered by trade openness. The reduction of manufacturing costs and the promotion of export expansion are facilitated by enhanced resource allocation efficiency. Furthermore, exports are positively influenced by the level of economic development. Although the coefficient is small, it is statistically significant, indicating that sustained economic growth consistently supports export performance. The coefficient of foreign direct investment is negative and insignificant, indicating that it has a limited impact on the export of new energy vehicles. Additionally, the level of urbanization has a positive but relatively limited impact on exports. Industrial agglomeration effects gradually emerge as regional integration progresses, thereby improving export competitiveness. In general, the regression results are highly explanatory, and the models are well fitted.

5.2. Mechanism Test

5.2.1. The Role of Technological Innovation

The results of the mechanism test for technological innovation (Lnip) are shown in

Table 7. Mediation effect test of technological innovation.

Variable	(1)	(2)	(3)
	LnED	Lnip	LnED
Digital	12.399 ***	5.533 ***	2.633 *
	(4.948)	(9.570)	(0.947)
Industry	4.309	2.957 ***	−0.910
	(1.372)	(4.082)	(−0.304)
Open	4.862 *	2.709 ***	0.081
	(1.930)	(4.660)	(0.033)
Gdp	−0.000	−0.000	−0.000
	(−1.561)	(−0.990)	(−1.261)
Fdi	−28.736 *	−4.264	−21.210
	(−1.921)	(−1.236)	(−1.543)
Urb	4.339	−2.307 ***	8.410 ***
	(1.617)	(−3.726)	(3.308)
Lnip			1.765 ***
			(6.226)
_cons	10.163 ***	9.168 ***	−6.019 **
	(5.841)	(22.839)	(−1.974)
N	203	203	203
R2	0.347	0.699	0.456
F	17.384	76.019	23.309

Note: ***, **, and * indicate significance at the 1%, 5%, and 10% confidence levels, respectively. Values in parentheses are t-statistics based on robust standard errors.

. In column (1), digital economy has a significantly positive impact on new energy vehicle exports. Column (2) shows that the coefficient of digital economy is 5.533, indicating that digital economy significantly promotes technological innovation. The progress of technological innovation in China is closely related to the development of digital economy. From column (3), the coefficient of technological innovation is 1.765, which confirms its positive effect on new energy vehicle export. Compared with column (1), the coefficient of digital economy drops significantly from 12.399 to 2.633 in column (3). This result confirms that technological innovation plays a mediating role in the relationship between digital economy and new energy vehicle export, thereby supporting hypothesis H2.

5.2.2. The Role of Financial Services

The mechanism test results for financial services (Fin) are presented in

Table 8. Mediation effect test of financial services.

Variable	(1)	(2)	(3)
	LnED	Fin	LnED
Digital	24.914 ***	5.290 ***	14.512 ***
	(5.704)	(7.856)	(2.973)
Industry	25.394 ***	2.958 ***	19.578 ***
	(4.103)	(3.100)	(3.219)
Open	12.082 **	0.746	10.615 **
	(2.313)	(0.926)	(2.121)
Gdp	0.001 ***	−0.000 ***	0.001 ***
	(4.413)	(−5.421)	(5.846)
Fdi	−19.200	−6.071 **	−7.262
	(−1.229)	(−2.521)	(−0.477)
Urb	7.458 *	2.283 ***	2.969
	(1.695)	(3.366)	(0.684)
Fin			1.966 ***
			(4.111)
_cons	−13.134 ***	1.970 ***	−17.007 ***
	(−3.808)	(3.705)	(−4.961)
N	203	203	203
R2	0.431	0.446	0.483
F	21.221	22.525	22.325

Note: ***, **, and * indicate significance at the 1%, 5%, and 10% confidence levels, respectively. Values in parentheses are t-statistics based on robust standard errors.

. Column (1) shows that the growth of digital economy significantly promotes new energy vehicle export. In column (2), the coefficient of digital economy is 5.290, indicating that digital economy strongly improves the level of financial services. The development of financial services is closely linked to the digital economy. In column (3), the coefficient of financial services is 1.966 and is significant at the 1% level. This confirms that financial services effectively promote the growth of new energy vehicle exports. Compared with column (1), the coefficient of digital economy decreases from 24.914 to 14.512 in column (3). This finding supports that financial services act as a key mediating variable through which digital economy influences new energy vehicle export, thus confirming hypothesis H3.

5.3. Robustness Tests

5.3.1. Alternative Measurement of Digital Economy Development Level

In the previous regression analysis, the level of digital economy development was measured using the entropy method. To improve the robustness of the findings, this paper follows the research paradigm by Chao Xiaojing and Ren Baoping (2011) [59]. It then applies the principal component analysis (PCA) method to remeasure the digital economy development level. A new panel regression is then conducted based on the recalculated index. The results are presented in

Table 9. Robustness test by changing the calculation method.

Variable	(1)	(2)
	LnED	Fin
Digital	28.797 ***	27.743 ***
	(17.467)	(12.819)
Industry		7.901 *
		(1.693)
Open		8.170 **
		(2.049)
Gdp		0.000
		(1.593)
Fdi		6.229
		(0.508)
Urb		−3.134
		(−0.891)
_cons	6.880 ***	1.817
	(12.607)	(0.634)
N	203	203
R2	0.638	0.657
F	305.105	53.567

Note: Standard deviations are shown in parentheses. ***, **, and * represent significance at the 1%, 5%, and 10% levels, respectively.

.

The regression results using the alternative measurement of digital economy development are reported in Table 9. Model (1) does not incorporate any control variables. The digital economy and the export of new energy vehicles exhibit a positive correlation. The regression coefficient is 28.797. All control variables are incorporated into Model 2. The positive impact of the digital economy on exports is slightly diminished after these variables are incorporated. Nevertheless, the regression serves as confirmation that the digital economy fosters the growth of exports of new energy vehicles. The coefficient reaches 27.743. The conclusion remains resolute in light of the empirical findings of both models. Exports of new energy vehicles are substantially stimulated by the digital economy, regardless of the utilization of an alternative measurement methodology. This further validates the baseline regression in Table 9 and strengthens the support for Hypothesis H1.

5.3.2. Bootstrap Test

According to the previous three-step mediation test, both financial services and technological innovation act as mediating channels between the digital economy and new energy vehicle exports. This paper employs the Bootstrap sampling technique for robustness testing in order to confirm the reliability of the mediation effect. Confidence intervals of the mediation effect are generated by Bootstrap through repeated sampling. The distribution characteristics of the mediation effect are more accurately captured by this method. Additionally, it mitigates the issue of multiple testing and enhances the reliability of the results.

Consequently, the Bootstrap method is implemented in this paper to evaluate the mediation effect. The null hypothesis (H0) presupposes that ab = 0, which implies that there is no mediation effect. In order to enhance the precision of estimations, 1000 repetitions are implemented. The null hypothesis is rejected if the 95% confidence interval does not include zero. This finding substantiates the statistical significance of the mediation effect.

The detailed Bootstrap test results are presented in

Table 10. Robustness test by bootstrap.

Variable	Coefficient	Z	p	95%Confidence Interval
Mediation Effect	9.766106	4.39	0.000	(6.094558, 14.88182) (p)
				(6.101089, 14.99698) (BC)
Direct Effect	2.633364	0.99	0.321	(−2.807706, 7.542252) (p)
				(−2.825381, 7.40504) (BC)
Replications			1000

Note: (p) represents uncorrected bias, and (BC) represents bias correction.

. The mediation effect coefficient is 9.767, as indicated by the regression. The p-value is significantly lower than the 0.1 threshold. The null hypothesis is rejected at the 1% significance level. Zero is not included in the bias-corrected 95% confidence interval. The mediation effect is further verified following the application of the Bootstrap method. This suggests that the export of new energy vehicles is indirectly influenced by the digital economy. The empirical findings are consistent and dependable.

5.3.3. Changing the Regression Model

To test the robustness of the regression results, this paper employs a two-way fixed effects panel model that systematically controls for both time and provincial fixed effects. This approach fully accounts for the dynamic evolution of the digital economy over time and addresses regional heterogeneity in new energy vehicle exports across provinces. On the one hand, time fixed effects capture potential external shocks and policy changes that may impact new energy vehicle exports, thereby reducing omitted variable bias. On the other hand, provincial fixed effects control for unobserved regional heterogeneity.

According to the results of the two-way fixed effects regression shown in

Table 11. Two-way fixed effects regression results.

Variable	(1)
	LnED
Digital	24.9143 ***
	(4.367)
Industry	25.3942 ***
	(6.189)
Open	12.0818 **
	(5.223)
Gdp	0.0007 ***
	(0.000)
Fdi	−19.2000
	(15.620)
Urb	7.4581 *
	(4.399)
_cons	−13.1337 ***
	(3.449)
year	Yes
province	Yes
N	203
R2	0.4311

Note: Standard deviations are shown in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

, the digital economy has a positive impact on China’s new energy vehicle exports. The robustness test results do not differ significantly from the previous regression outcomes. This indicates that the panel regression results in this paper are robust.

5.4. Heterogeneity Test

5.4.1. Analysis Based on Eastern, Central, and Western Regions

The regression results in

Table 12. Analysis based on eastern, central, and western regions.

Variable	Eastern Region	Central Region	Western Region
Digital	20.971 ***	−23.466 *	−1.754
	(4.384)	(13.136)	(14.362)
Industry	43.952 ***	−4.918	18.800
	(13.375)	(6.880)	(12.214)
Open	6.655	−25.648 **	38.715 **
	(5.250)	(12.640)	(14.860)
Gdp	0.001 ***	0.000	0.000
	(0.000)	(0.000)	(0.001)
Fdi	5.764	6.910	103.803 *
	(20.076)	(19.702)	(56.517)
Urb	0.788	132.142 ***	78.559 ***
	(4.142)	(19.900)	(18.724)
_cons	−19.812 **	−57.991 ***	−43.868 ***
	(7.456)	(8.687)	(7.072)
N	77.000	56.000	70.000
R2	0.471	0.817	0.652
R2_a	0.330	0.760	0.556

Note: Standard deviations are shown in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

reveal significant regional heterogeneity in the relationship between new energy vehicle exports and the digital economy. In the eastern region, the estimated coefficient is 20.971 and is statistically significant at the 1% level. This indicates that the digital economy significantly promotes new energy vehicle exports, likely due to the region’s advanced technologies and well-developed infrastructure. In the central region, the negative effect may reflect a conflict between digital development and industrial restructuring. The expansion of the digital economy may crowd out resources from traditional manufacturing sectors. In the western region, the impact is not significant. This may be attributed to the relatively low level of digital development, which has not yet formed effective support for new energy vehicle exports.

The effects of control variables also vary across regions. Industrial development plays a significant role in the eastern region, possibly due to its strong industrial foundation. However, the effect is not significant in the central and western regions. This is likely due to the fact that their industrial structures are less aligned with the requirements of the new energy vehicle industry. The level of openness has a positive effect in the western region, likely driven by inflows of foreign capital and technology. In the central region, the effect is negative, which may result from unequal resource allocation or intensified market competition. Economic development significantly promotes new energy vehicle exports in the eastern region, where growth is more dependent on high-tech industries.

In contrast, the effect is not significant in the central and western regions, possibly because their development models have not fully shifted toward high-tech sectors. Foreign direct investment has a strong positive effect on the western region. This may be due to its role in addressing local capital shortages and supporting the growth of the new energy vehicle industry. In the eastern and central regions, where capital is more abundant, the effect of foreign investment is limited. Urbanization promotes new energy vehicle exports in the central and western regions. This may be because urbanization drives market demand and infrastructure improvements. In the eastern region, the effect is marginal, as urbanization has already reached a high level. It is worth noting that foreign direct investment does not exhibit a statistically significant impact on new energy vehicle exports in the full-sample regressions. However, the results from the regional heterogeneity analysis indicate that fdi plays a more influential role in the western provinces. This divergence may stem from the relatively lower baseline of industrial and infrastructural development in these regions. In such areas, foreign direct investment plays a crucial role as a source of capital and technological spillovers. These findings suggest that the impact of fdi is context dependent and more pronounced in underdeveloped regions.

5.4.2. Heterogeneity Analysis According to Coastal and Inland Geographic Features

To further analyze the regional differences in the impact of the digital economy on new energy vehicles, this paper follows the coastal–inland classification. It divides 29 provincial-level regions into two groups: coastal and inland areas. Specifically, the coastal areas include 11 provinces such as Shandong, Liaoning, Zhejiang, and Hainan. In contrast, the inland areas include 18 provinces such as Jilin, Anhui, and Henan. The regression results in

Table 13. Analysis based on coastal and inland regions.

Variable	Coastal Region	Inland Region
Digital	−10.819 **	38.099 **
	(−2.256)	(2.461)
Industry	16.194	13.874 *
	(1.365)	(1.912)
Open	0.383	3.056
	(0.085)	(0.491)
Gdp	8.358 *	3.214
	(1.840)	(1.333)
Fdi	−22.126	56.286 **
	(−1.165)	(2.160)
Urb	18.481	15.835
	(1.502)	(0.783)
_cons	−53.817	−58.124 **
	(−1.122)	(−2.214)
N	77	126
R2	0.859	0.722
R2_a	0.802	0.637

Note: Standard deviations are shown in parentheses. **, and * indicate significance at the 5%, and 10% levels, respectively.

show significant differences in the impact of the digital economy on new energy vehicle exports between coastal and inland regions. Specifically, the digital economy has a negative and significant effect on coastal regions, with a coefficient of −10.819 at the 5% level. In contrast, it exerts a strong positive effect on inland regions, with a coefficient of 38.099, also significant at the 5% level. The divergence may be attributed to the advanced stage of the new energy vehicle industry in coastal areas. Here, rapid digital development exerts structural adjustment pressures that could temporarily hinder export growth. Conversely, inland regions are still developing their digital economy. This development facilitates technological progress and information flow, thereby promoting industrial growth and exports.

Regarding control variables, industrial development positively influences inland exports at the 10% significance level but shows no significant effect in coastal areas. Openness does not significantly affect exports in either region, suggesting a limited short-term impact. Economic development positively affects coastal exports at the 10% level but is insignificant inland, likely due to weaker economic foundations. Foreign direct investment strongly promotes inland exports, with a coefficient of 56.286, while its negative and insignificant impact in coastal regions may reflect capital saturation. Urbanization shows no significant effect in both regions, indicating marginal contributions to export growth.

Overall, inland regions depend more on the digital economy, industrial development, and foreign investment, reflecting a phase of industrial growth and upgrading. Coastal regions face challenges of industrial restructuring and optimization, where digital economy growth may temporarily hinder exports. Therefore, policies should be tailored to promote industrial transformation in coastal areas while strengthening digital infrastructure in inland regions. Such targeted measures are essential for sustaining the growth of new energy vehicle exports.

6. Discussion, Conclusions, Implications, and Research Limitations

6.1. Discussion

This paper offers empirical evidence on how the digital economy contributes to the development of China’s new energy vehicle exports, providing several implications when aligned with practical realities:

(1)

The results indicate that the digital economy significantly enhances new energy vehicle export performance. This effect is achieved through reductions in transaction costs, improved logistics efficiency, and expanded access to international markets. These findings suggest that government investment in digital infrastructure—such as broadband networks, data platforms, and cross-border e-commerce systems—remains crucial. In practice, export-oriented new energy vehicle enterprises are especially sensitive to improvements in digital connectivity. These upgrades, in turn, directly enhance their competitiveness in global markets.

(2)

The empirical results also verify the existence of dual mediating effects: technological innovation and financial services. These channels indicate that the digital economy facilitates direct improvements in export outcomes. Moreover, it enhances the enabling environment in which green industries operate. Therefore, policymakers should strengthen support systems by increasing subsidies for digital R&D and providing tax incentives for digital adoption. At the enterprise level, firms should integrate digital technologies such as blockchain for supply chain transparency and artificial intelligence for international market forecasting. Financial institutions should also develop customized services that cater specifically to the financing needs of green exporters.

(3)

Significant regional differences emerge from the analysis. In eastern regions—where digital infrastructure is more developed—the digital economy enhances exports by improving service quality and driving innovation. In contrast, the inland regions, though traditionally disadvantaged, experience a stronger marginal benefit from digital development. This suggests that strengthening basic digital capabilities and institutional support in these regions can yield disproportionately large gains. As such, tailored regional strategies are required: developed regions should move toward advanced digital applications, while developing regions should focus on foundational capacity-building. Interprovincial collaboration and regional digital integration platforms may also help narrow the development gap and promote balanced export growth.

(4)

This paper adopts a robust empirical strategy, incorporating the entropy method, mediation models, and regional heterogeneity analysis. This analytical framework validates the core hypotheses. In addition, it provides a model that can be applied to other green sectors. Industries such as smart transportation, battery manufacturing, and renewable energy equipment may benefit from similar digital transformation strategies. The approach is particularly valuable for emerging economies seeking to enhance export competitiveness through digital means.

6.2. Conclusions

This paper investigates the impact of the digital economy on the export performance of China’s new energy vehicle sector through a comprehensive empirical framework. The findings demonstrate that the digital economy significantly promotes new energy vehicle exports through both direct and indirect pathways.

First, the digital economy directly improves export performance by reducing transaction costs, streamlining logistics, and expanding access to international markets. These benefits are especially important for new energy vehicle enterprises, which rely on fast, data-driven supply chains and efficient cross-border operations to compete globally.

Second, it indirectly drives export growth by enhancing technological innovation and improving the availability and quality of financial services. These two channels help enterprises upgrade their products and lower their capital constraints. As digital tools become more integrated into production and financing processes, firms gain greater operational flexibility. As a result, they are better positioned to invest in innovation and respond to international market demands.

Third, the effect of the digital economy on exports varies significantly across regions. Inland regions benefit more due to their relatively underdeveloped digital infrastructure, which leaves more room for marginal gains. By contrast, eastern regions show improvements primarily through advanced innovation systems and service-oriented digital applications. This regional variation underscores the need for differentiated digital development strategies that match local conditions.

Fourth, this paper extends existing research by combining dual transmission mechanisms with regional heterogeneity in one empirical framework. This approach deepens the understanding of how digital transformation interacts with export dynamics and provides a flexible model for application in other green industries. Sectors such as battery manufacturing, renewable energy equipment, and smart transportation may similarly enhance their international competitiveness through targeted digital upgrades.

In conclusion, the digital economy serves as a crucial force in strengthening the global position of China’s new energy vehicle industry. It supports both industrial transformation and sustainable trade expansion by improving efficiency, fostering innovation, and narrowing regional disparities. These insights provide valuable guidance for governments, businesses, and researchers aiming to integrate digital tools into green development strategies.

6.3. Implications

To further illustrate the practical significance of this paper’s conclusions, a series of feasible recommendations is proposed for enterprises based on the above findings. The details are as follows:

(1)

The regression results reveal that the digital economy significantly promotes new energy vehicle exports, with technological innovation serving as a key mediating channel. Promoting digital transformation is thus essential for enhancing export competitiveness. The integration of IoT, big data, and AI improves production efficiency and product quality while reducing costs. Smart factories enable automated manufacturing and real-time monitoring. Digital tools also enhance supply chain management by making material flows visible and improving response speed. AI forecasts supply and demand, helping reduce risk and increase efficiency. In R&D, technologies such as CAD, simulation software, and machine learning support faster, smarter design. AI can optimize batteries and motors, improving energy efficiency and power density. Cloud platforms facilitate data sharing across departments, boosting development efficiency. Big data helps firms analyze usage behavior and adapt designs to user needs. The development of intelligent connected vehicles is also critical. Vehicle-to-everything networks support real-time data sharing, safer driving, and personalized in-car experiences. Remote management and customized settings enhance convenience and user satisfaction. Overall, digital technologies enhance the entire value chain, which in turn drives export growth.

(2)

The empirical results show that the digital economy significantly boosts China’s new energy vehicle exports. This effect is stronger in regions with a higher level of openness. This suggests that outward-oriented cooperation and digital infrastructure work together to amplify export performance. Chinese automakers should expand globally by building service networks, strengthening R&D, and sharing resources. Joint efforts in technology and talent development can reduce costs and increase export scale. The government should offer policy support to guide firms toward innovation, branding, and market diversification. Improvements in cross-border finance, logistics, and after-sales systems are also essential for raising service quality and responsiveness. Moreover, China should increase its role in setting international trade rules. New energy vehicles, as a strategic global industry, should target developing countries and Belt and Road markets. Active participation in free trade zones and bilateral agreements can ensure institutional support. Companies should also utilize WTO mechanisms to safeguard their interests. International cooperation and standardization are key. Firms should engage in joint innovation in batteries, smart driving, and connectivity. They must take part in international standard-setting bodies such as ISO, IEC, and 5GAA. The government should back these efforts through policy and funding, promoting global recognition of Chinese technology and supporting cross-border R&D centers.

(3)

The mechanism analysis reveals that financial services and digital infrastructure are key mediating factors through which the digital economy promotes new energy vehicle exports. Therefore, developing a data-driven intelligent marketing system could further amplify the positive impact of the digital economy on export performance. Enterprises need to establish comprehensive big data platforms covering R&D, production, sales, and service, enabling real-time data collection and unified storage. Leveraging big data and AI enables firms to analyze consumer behavior and market demand. This capability helps optimize product design and enhance supply chain efficiency, thereby driving continuous technological innovation. Data sharing and collaboration among firms, research institutions, and government agencies further support industrial upgrading. Financial services facilitate digital marketing by providing funding, risk management, and secure cross-border payments, enhancing system efficiency. Data security and privacy must be ensured through encryption and access controls. Government support is essential to promote big data and digital finance infrastructure. Furthermore, expanding digital marketing channels is key to accessing global markets. Firms should utilize cross-border e-commerce platforms and social media to increase brand visibility and engage customers, especially in Europe, North America, and Southeast Asia. Technological innovation enhances product differentiation, providing compelling marketing content, while financial services ensure smooth international transactions. The synergy among digital marketing, innovation, and finance fosters sustainable export growth.

(4)

Heterogeneity analysis shows that China’s eastern region has advantages in digital finance, talent, and supportive policies. This results in a stronger impact of the digital economy on new energy vehicle exports there. Thus, institutional efforts should focus on digital financial infrastructure, talent development, and regional coordination. First, digital financial infrastructure is key to improving export finance services. Companies should adopt global payment innovations to shorten settlement times to T + 0.5, enhancing overseas payment efficiency. A dynamic credit evaluation system integrating customs certification, sales, and social media data can create digital credit passports for international recognition. Satellite remote sensing and machine learning can reduce bad debt risks and speed up financing approvals. Second, talent cultivation supports the integration of the digital economy with the vehicle industry. Education authorities should promote interdisciplinary courses in intelligent vehicles, batteries, and AI. School-enterprise partnerships can build practical training bases. Companies should run regular training to boost digital skills and innovation. Finally, balanced regional development requires local adaptation. The central government should fund digital infrastructure upgrades in central and western regions, including broadband and cloud platforms. Digital literacy programs via community training should be expanded. Regional digital platforms should offer technical support. These measures will integrate digital technology deeply into production and daily life, fostering high-quality economic growth.

6.4. Research Limitations

This paper still has certain limitations, mainly reflected in two aspects:

(1)

The Ministry of Industry and Information Technology of China did not officially define new energy vehicles until 2017, despite the fact that the concept was first proposed. This was accomplished through the “Administrative Measures for the Access of New Energy Vehicle Production Enterprises and Products.” Consequently, it is challenging to acquire systematic and standardized export data of new energy vehicles from the General Administration of Customs of China prior to that year. This restricts the duration of the study.

(2)

The theoretical framework of the digital economy is still in its infancy, as it is a rapidly developing field. Current research on the specific impact of the digital economy on the export trade of new energy vehicles in China is still in the early stages. This paper continues to be devoid of a systematic and comprehensive theoretical framework.

It would be valuable for future studies to focus on the following aspects: Initially, the study period could be extended forward or backward to capture the long-term dynamic changes in new energy vehicle exports as additional data accumulates. Subsequently, broaden the connotation and measurement dimensions of the digital economy. This entails the incorporation of micro mechanisms, including intelligent manufacturing, platform economy, and big data applications, in order to establish a more explanatory analytical framework. Third, the examination of regional disparities and industry heterogeneity could be further enhanced. This paper examined the distinct effects of the digital economy on new energy vehicle exports across different segments of the industry chain. It also considers variations at different levels of regional development. The breadth and depth of research will be improved by such endeavors.