The Impact of Digital Trade on the Export Competitiveness of Enterprises—An Empirical Analysis Based on Listed Companies in the Yangtze River Economic Belt

Abstract

With the in-depth development of globalization and informatization, digital trade, as an emerging form of trade, is gradually reshaping the global economic landscape and becoming a new engine for driving economic growth. Among them, the impact of digital trade on the export competitiveness of enterprises in developing countries has become a common concern in academia. To reveal the causal relationship between the development of digital trade and the enhancement of export competitiveness in developing countries, this paper first constructs a theoretical model based on product quality heterogeneity and analyzes the impact of digital trade on the export competitiveness of enterprises on the basis of achieving supply and demand equilibrium; then, this paper constructs a comprehensive index system for measuring digital trade and enterprise export competitiveness, and establishes an empirical analysis model; on this basis, this paper uses the data of listed companies in the A-share market in the Yangtze River Economic Belt area from 2011 to 2021 for empirical analysis. The results of the empirical analysis show that for every one-unit increase in the level of digital trade development in the region, there will be a positive impact of 0.9041 units on the export competitiveness of enterprises. After a series of robustness tests and endogeneity analyses, the above empirical results are confirmed to be robust and reliable. Furthermore, this paper conducts a heterogeneity analysis and finally puts forward corresponding policy recommendations based on the above theoretical and empirical research results.

1. Introduction

Under the impetus of new generations of information and communication technology, digital trade is becoming a new engine for global economic growth [1,2]. As the world’s largest developing country, China’s digital trade has developed rapidly in recent years, with the scale of digital delivery service trade reaching 2.5 trillion yuan in 2022 and the import and export scale of cross-border e-commerce reaching 2.1 trillion yuan, accounting for 63.3% of the global cross-border digital service trade [3].

To reveal the causal relationship between the development of digital trade in developing countries and the enhancement of their enterprises’ export competitiveness, the Yangtze River Economic Belt in China is selected in this study as the research object. The Yangtze River Economic Belt encompasses 11 provinces and municipalities: Shanghai, Jiangsu, Zhejiang, Anhui, Jiangxi, Hubei, Hunan, Chongqing, Sichuan, Yunnan, and Guizhou. Covering an area of approximately 2.0523 million square km, it accounts for 21.4% of China’s total land area, with both its population and GDP exceeding 40% of the national total [4]. This choice is justified for two reasons: firstly, the region holds an irreplaceable significance in China’s national strategy as the world’s largest developing country [5,6]. It encompasses several economically advanced provinces (comprising three city groups: the Yangtze River Delta, the Middle Yangtze River, and the Sichuan-Chongqing. Among them, the Yangtze River Delta includes cities under Shanghai, Jiangsu, Zhejiang, and Anhui; the Middle Yangtze River includes cities under the jurisdiction of Hubei, Hunan, and Jiangxi; Sichuan−Chongqing includes Chongqing and cities under the administration of Sichuan), and has actively pursued the goal of comprehensive green transformation by vigorously developing digital trade in recent years [7,8,9]. Secondly, the Yangtze River Economic Belt plays a crucial role in China’s opening up and international cooperation [10,11], serving as a hub that intersects multiple dimensions, including domestic and foreign interactions, as well as land and maritime exchanges [12,13,14]. This provides a typical case for studying the role of digital trade in the core economic zones of developing countries within the international economy.

This paper aims to explore whether and how digital trade can enhance export competitiveness. Based on this exploration, it will propose corresponding development strategies and recommendations to support the sustained and healthy development of digital trade for Chinese enterprises, empower high-quality regional economic growth, and provide evidence for the formulation and implementation of relevant national policies in the future. Additionally, this research will enrich the theoretical content regarding the impact of digital trade on export competitiveness and offer insights for future studies in terms of data analysis, indicator selection, and variable setting.

Although relevant research has achieved fruitful results, there are still few discussions on the causal relationship between firms’ digital trade and export competitiveness, especially regarding the lack of analysis of the heterogeneity of this causal relationship among different urban agglomerations and enterprises of different sizes. In addition, existing measurement methods for digital trade have certain limitations and cannot fully reflect the multidimensional characteristics of digital trade.

In view of this, this paper first constructs a theoretical model based on product quality heterogeneity analyzes the theoretical mechanism by which digital trade affects the export competitiveness of enterprises on the basis of achieving supply and demand equilibrium, and puts forward corresponding research hypotheses; then, this paper constructs a comprehensive index system for measuring digital trade and enterprise export competitiveness, and establishes an empirical analysis model; on this basis, this paper uses the data of listed companies in the A-share market in the Yangtze River Economic Belt from 2011 to 2021 for empirical analysis, and carries out a series of robustness tests and endogeneity analyses on the benchmark regression results, and then discusses the heterogeneous situation of the above causal relationship between different city clusters and enterprises of different sizes. Through the above research, this paper strives to achieve the following innovations on the basis of the existing literature:

This study demonstrates innovations in several aspects:

• A theoretical model is constructed that discusses the impact of digital trade on the export competitiveness of enterprises from both the demand and supply dimensions.
• A more comprehensive index system for measuring digital trade and enterprise export competitiveness is constructed, further enriching the quantitative calculation methods for digital trade and enterprise export competitiveness.
• By analyzing the core economic region of the world’s largest developing country, this paper not only enriches the existing international trade and e-commerce research literature but also provides decision-making references for the vigorous development of digital trade in developing countries.

The subsequent content of this paper is arranged as follows: the second part constructs a theoretical model of the impact of digital trade on the export competitiveness of enterprises; the third part establishes a comprehensive index system for measuring digital trade and enterprise export competitiveness; the fourth part constructs an econometric analysis model, obtains the benchmark regression results, and carries out related tests and heterogeneity analysis; and the fifth part summarizes the whole paper and puts forward corresponding policy recommendations.

2. Literature Review

2.1. Digital Trade

In the field of digital trade, current studies primarily focus on the following main directions:

Conceptual Boundaries and Definition of Digital Trade: Research on the definition and conceptual understanding of digital trade is one of the foundational directions in this field. Since there is no unified consensus internationally on the definition of digital trade, significant differences exist among countries and international organizations in terms of its scope and interpretation. A widely accepted resource in academia is the 2023 second edition of the Handbook on Measuring Digital Trade, jointly published by the World Trade Organization (WTO), the Organization for Economic Cooperation and Development (OECD), the International Monetary Fund (IMF), and the United Nations Conference on Trade and Development (UNCTAD) [15]. This handbook adopts a transaction-based perspective, highlighting “digital ordering” and/or “digital delivery” as the core characteristics of digital trade, independent of product type or participant characteristics. This definition, based on the nature of transactions rather than product features, provides a foundation for international convergence in the concept of digital trade and underscores the essential role of digital channels in their operation.

Measurement and Data Collection Methods in Digital Trade: The methods for measuring digital trade have been a central focus in this field. Early academic efforts often used data from the information and communication technology (ICT) sector as proxy indicators for digital trade. However, as the complexity of digital trade has grown, research has gradually shifted to the utilization of statistical data defined by governments or international organizations. For instance, statistical data on digitally delivered trade, jointly released by the OECD [16,17], WTO [18], and IMF [19], are now widely used as a primary data source in numerous academic studies. As research deepens, scholars continue to refine data and classification methods to enhance the cross-national comparability and accuracy of digital trade statistics, striving to depict the true scale and dynamic characteristics of digital trade more precisely from a multidimensional perspective [20,21,22].

Analysis of the Economic Impact of Digital Trade: This research direction primarily examines the profound impact of digital trade on economic structures and traditional trade models, particularly in terms of how digital services reduce transaction costs and expand market access [23]. Studies in this area analyze how digital trade reshapes global value chains [24,25] and explore how digital platforms empower smaller economies and small and medium enterprises with new competitive advantages [26,27], enabling them to enter international markets without the need for traditional trade infrastructure [28]. Moreover, scholars are concerned with the challenges that digital trade poses to existing regulatory frameworks [29,30] and the potential of digital innovation to foster inclusive economic growth [31].

2.2. Export Competitiveness

In the field of export competitiveness, current studies primarily focus on the following main directions:

National-level Export Competitiveness: Studies at the national level examine export competitiveness by evaluating a country’s ability to gain a favorable position in the global market [32]. This direction often considers factors such as currency exchange rates [33], labor costs [34], technological advancement [35], trade regulations [36], political stability [37], and domestic market size [38]. These factors influence a country’s relative advantage and are pivotal in establishing its products’ appeal and accessibility on an international scale [39,40]. National-level competitiveness research aims to understand how macroeconomic and policy environments shape overall export performance, providing insights into economic policy’s impact on international trade [41].

Industry-level Export Competitiveness: At the industry level, research focuses on factors that contribute to specific sectors’ performance in global markets [42]. Key aspects include cost control [43], technological innovation [44], research and development (R&D) investment [45], brand influence [46], and supply chain management [47]. Studies in this area highlight how industry-specific strategies and innovations contribute to sustaining and enhancing competitiveness [48]. This direction underscores the role of sectoral strengths and challenges in adapting to global market demands and maintaining resilience against international competition [49,50].

Firm-level Export Competitiveness: At the firm level, export competitiveness is typically assessed based on product and service quality [51], pricing strategies [52], and innovation capabilities [53]. Research in this area investigates how individual firms leverage these factors to succeed in international markets, often emphasizing the importance of quality management [54], adaptive pricing [55], and continuous innovation [56]. Firm-level studies provide a granular perspective on the dynamics of export competitiveness, illustrating how company-specific strategies can differentiate products in the global market and create sustainable competitive advantages [57,58].

2.3. The Impact of Digital Trade on Export Competitiveness

In the field of the impact of digital trade on export competitiveness, current studies primarily focus on the following main directions:

Enhancement of Market Expansion Capability through Digital Trade: Research indicates that the development of digital trade significantly strengthens firms’ competitiveness in overseas markets, particularly in market expansion and customer acquisition [59,60]. Through digital channels, firms can enter international markets more quickly and at a lower cost, achieving targeted marketing and thereby expanding market coverage [61,62].

Digital Transformation and Increased Export Competitiveness: At the firm level, digital transformation has become a critical element in enhancing export competitiveness [63,64]. Relevant studies point out that companies can significantly reduce costs and improve production efficiency by adopting digital production methods and intelligent supply chain management, thereby strengthening their international competitiveness [65,66]. The application of digital technology in production and operations enables firms to respond to market demands, customize production, and ensure rapid delivery [67,68,69].

Promotion of Industrial Integration through Digital Trade: Digital trade not only alters the operational models of export-oriented firms but also drives deep integration between different industries, creating new points of export growth [70,71]. Academic research shows that the application of digital technology fosters the integration of manufacturing and services [72], promoting the development of servitization in manufacturing [73,74] thereby enhancing the overall value chain level of exports [75,76].

3. Theoretical Model

When examining the impact of digital trade on the export competitiveness of enterprises, this paper constructs a model based on product quality heterogeneity. Inspired by the research of Hallak and Sivadasan [77], the model aims to explore the determinants of product quality and analyze how the development of digital trade affects the quality of enterprises’ export products.

The reason for selecting Hallak and Sivadasan’s study is that while constructing the model, we also referred to other literature related to digital trade. However, these studies primarily focus on the impact of digitalization on aspects such as operational efficiency, cross-border costs, and market entry barriers and offer relatively limited explanatory power regarding the relationship between quality heterogeneity and competitiveness in the export trade indicators selected in this paper. Therefore, they were not used as the primary theoretical basis for our model.

3.1. Market Demand

The paper assumes a Constant Elasticity of Substitution (CES) utility function for consumers, which is as follows (refer to Equation (1)):

$$U_g = {\left(\sum _g{\left({\lambda }_g{q}_g\right)}^{{\displaystyle \frac{\sigma -1}{\sigma }}}\right)}^{{\displaystyle \frac{\sigma }{\sigma -1}}}, \sigma > 1$$

(1)

Here, ${\mathsf{\lambda }}_g$ represents the quality of product g, $q_g$ is the demand for product $g$, $\sigma$ is the coefficient of substitution elasticity, and $\sigma >1$. The consumer expenditure is denoted by $E$, and the product price index is $P$. The demand function is derived from the utility function with the constraint (refer to Equation (2)):

$$E=\sum _g{q}_g{p}_g$$

(2)

The equilibrium condition is (refer to Equation (3)):

$${\displaystyle \frac{{MU}_n}{p_n}}={\displaystyle \frac{{MU}_m}{p_m}}=\dots ={\displaystyle \frac{{MU}_g}{p_g}}$$

(3)

This leads to (refer to Equation (4)):

$$q_g={\displaystyle \frac{{\mathsf{\lambda }}_g}{{\left(p_g\right)}^{\mathsf{\sigma }}}}\cdot {\displaystyle \frac{E}{P}}$$

(4)

Here, $p_g$ is the price of product g, and $P$ is the product price index, which can be defined as (refer to Equation (5)):

$$P=\sum _g(({\displaystyle \frac{{\lambda }_g}{p_g}}{)}^{\sigma -1})$$

(5)

Here, $P$ is determined by the prices and quality of all products. According to the demand function, this paper shows that the product price is inversely proportional to demand, while product quality is directly proportional to demand, meaning that lower prices or higher quality lead to greater demand and competitive advantage in the market.

3.2. Market Supply

In terms of market supply, digital trade accelerates the flow and sharing of information, enabling enterprises to obtain market trend data, consumer preferences, and competitor information more efficiently. This affects production decisions, allowing enterprises to respond quickly to market changes, reduce inventory backlogs, improve the flexibility of resource allocation, and effectively reduce operational risks and marginal costs. Therefore, we can extend the marginal cost function to include the variable of digital trade to reflect the relationship between information advantage and cost (refer to Equation (6)):

$$MC={\displaystyle \frac{h}{\phi D_t}}{\lambda }^{\alpha }$$

(6)

The improvement in a firm’s production capacity contributes to a reduction in its fixed costs; thus, the fixed cost function is set as follows (refer to Equation (7)):

$$FC=F{C}_0+{\displaystyle \frac{f}{\xi S_d}}{\lambda }^{\beta }$$

(7)

where $MC$ is the marginal cost, $FC$ is the fixed cost, $\phi$ represents production efficiency, $\alpha$ denotes the effect of production efficiency on product quality $\left(\alpha >1\right), F{C}_0$ is the required operating cost of the firm, $\xi$ represents the firm’s production capacity, $\beta$ is the impact of production capacity on product quality $\left(\beta >1\right)$, and $h$ and $f$ are constant coefficients.

3.3. Market Equilibrium

In the context of demand and cost functions, a firm aiming to maximize profit selects the optimal product quality as follows (refer to Equation (8)):

$$\lambda =\left[{\displaystyle \frac{(1-\alpha \beta )}{\sigma -1}}\left({\displaystyle \frac{\sigma }{\phi h(\sigma -1)}}\right){\left({\displaystyle \frac{\zeta f}{EP}}\right)}^{{\displaystyle \frac{1}{\beta -1-\alpha (\sigma -1)}}}\right]$$

(8)

The quality of an enterprise’s export products is related to its production efficiency and production capacity. By taking partial derivatives, we can conclude that the improvement in production efficiency $\phi$ and production capacity $\xi$ is conducive to improving the quality of the enterprise’s export products $\lambda$. Let $\gamma =\beta -(1-\alpha )(\sigma -1)$, and take the first-order partial derivative of this formula with respect to production efficiency $\phi$ and production capacity $\xi$ to obtain (refer to Equations (9) and (10)):

$${\displaystyle \frac{\partial \lambda }{\partial \phi }}={\left[{\displaystyle \frac{1-\alpha }{\beta }}{\left({\displaystyle \frac{\sigma -1}{\sigma }}\right)}^{\sigma }{\left({\displaystyle \frac{\phi }{h}}\right)}^{\sigma -1}{\displaystyle \frac{\xi }{f}}{\displaystyle \frac{E}{P}}\right]}^{{\displaystyle \frac{1-\gamma }{\gamma }}}{\left({\displaystyle \frac{\sigma -1}{\sigma }}\right)}^{\sigma }{\displaystyle \frac{E}{P}}{\left({\displaystyle \frac{\phi }{h}}\right)}^{\sigma -2}{\displaystyle \frac{\xi \left(\sigma -1\right)\left(1-\alpha \right)}{\gamma \beta hf}}>0$$

(9)

$${\displaystyle \frac{\partial \lambda }{\partial \xi }}={\left[{\displaystyle \frac{1-\alpha }{\beta }}{\left({\displaystyle \frac{\sigma -1}{\sigma }}\right)}^{\sigma }{\left({\displaystyle \frac{\phi }{h}}\right)}^{\sigma -1}{\displaystyle \frac{\xi }{f}}{\displaystyle \frac{E}{P}}\right]}^{{\displaystyle \frac{1-\gamma }{\gamma }}}{\left({\displaystyle \frac{\sigma -1}{\sigma }}\right)}^{\sigma }{\displaystyle \frac{E}{P}}{\left({\displaystyle \frac{\phi }{h}}\right)}^{\sigma -1}{\displaystyle \frac{1-\alpha }{\gamma \beta f}}>0$$

(10)

Improving production efficiency and production capacity has a positive effect on improving the quality of enterprises’ export products. With the development of digital trade, enterprises are turning to service-oriented, digital, and intelligent, which helps to improve their production efficiency and production capacity. Therefore, these changes help improve product quality and technical level, and increase the technical complexity of enterprises’ export products.

4. Indicator System and Measurement Method

4.1. Digital Trade Level Indicator System and Measurement Method

This paper constructs an indicator system to measure the development level of China’s digital trade based on the principles of data availability and reliability, replicability, scientificality, and comparability of the indicator system, referring to the existing research method [78]. It starts from the perspective of the digital trade industry chain (

Table 1. Comprehensive indicator system for digital trade.

Level	Dimension	Specific Measurement Indicator
Digital Trade Development Environment	Internet Broadband	Number of Internet broadband access ports per ten thousand people (+)
	Domain Names	Number of domain names per ten thousand people (+)
	Optical Cable Network	Length of optical cable per square kilometer (+)
	Mobile Internet	Number of mobile Internet users per ten thousand people (+)
Supporting Industries for Digital Trade	Software and Information Technology Services Industry	Fixed assets of the information transmission, software, and information technology services industry (+)
		Main business revenue of the electronic and communication equipment manufacturing industry (+)
		Revenue of software and information technology services (+)
		Number of cross-border payment companies (+)
	E-commerce	E-commerce sales (+)
	Logistics Network	Revenue from express delivery services (+)
Potential of Digital Trade	Diffusion Level	Proportion of enterprises with e-commerce transaction activities (+)
	R&D Investment	Fixed asset investment in the manufacturing of computers, communications, and other electronic equipment (+)
	Regional Economy	Per capita regional gross domestic product (+)

Source: Own elaboration.

). All indicators in this system are positive indicators.

The comprehensive indicator system mentioned above is designed to assess and measure the development level and potential of digital trade. This system includes the following three main levels:

• Digital Trade Development Environment: This level focuses on the infrastructure and technical level of digital trade, which forms the foundation of its progress. At this level, the paper considers indicators such as Internet broadband, mobile Internet, the number of domain names, and the optical cable network. For example, the number of Internet broadband access ports and mobile Internet users per ten thousand people can reflect the degree of network popularity and user access capabilities in a region, which is crucial for the conduct of digital trade. The length of optical cables per square kilometer reflects the level of network infrastructure construction in the area, which directly affects the speed and stability of data transmission.
• Supporting Industries for Digital Trade: This level is concerned with industries that directly support digital trade, such as the software and information technology services industry and e-commerce. At this level, the paper selects indicators such as fixed asset investment, main business revenue, information technology service revenue, and software business revenue to measure the development level of the software and information technology services industry. These indicators can reflect the economic scale and growth potential of the industry. At the same time, indicators such as e-commerce sales, express business revenue, and the volume of express delivery are directly related to the activity level of e-commerce and the efficiency of the logistics network, which are key factors for the smooth progress of digital trade.
• Potential of Digital Trade: This level examines the future development potential of digital trade, including its popularity of digital trade, R&D investment in the manufacturing industry, and the overall level of the regional economy. The proportion of enterprises participating in e-commerce transactions can reflect their activity levels in digital trade. At the same time, the situation of fixed asset investment in the manufacturing of computers, communications, and other electronic equipment shows the strength and determination of investment in the digital transformation of the manufacturing industry. Per capita regional GDP (Gross Domestic Product), as a key indicator, reflects the strength of a region’s economic power. It indirectly affects the development environment and the potential of digital trade.

Based on the above indicator system, this paper uses the entropy weight method to measure the development level of digital trade in China’s Yangtze River Economic Belt. This method assigns weights to each evaluation indicator based on the entropy value, thereby ensuring the objectivity and scientific nature of the evaluation by comprehensively considering the role of each indicator in the system.

The initial step involves standardizing the raw data to eliminate the impact of differences in dimensions and data scales among various regions. This standardization process adjusts the values of each indicator in each region to relative values, ensuring comparability between indicators during the weight calculation process. The calculation method for the standardization process is as follows:

$$x_{ij}^{\prime }={\displaystyle \frac{X_{ij}-X_{min}}{X_{max}-X_{min}}}$$

(11)

Here, $\left(i=\mathrm{1,2},\dots ,n\right)$ represents regions, and $(j=\mathrm{1,2},\dots ,m)$ represents sub-indicators. $X_{ij}$ and $x_{ij}^{\prime }$ represent the original value and the standardized value of the j-th indicator for the i-th province, respectively, and $X_{max}$ and $X_{min}$ are the maximum and minimum values for each indicator, respectively.

Next, calculate the proportion of each indicator in the overall evaluation system. The following formula is used for calculation:

$$P\left(x_{ij}\right)={\displaystyle \frac{x_{ij}^{″}}{{\sum }_{i=1}^m{x}_{ij}^{″}}}$$

(12)

And calculate the entropy value for each sub-indicator:

$$e_j=-\left({\displaystyle \frac{1}{\ln m}}\right)\sum _{i=1}^{m}P\left(x_{ij}\right)\ln P\left(x_{ij}\right)$$

(13)

Then, determine the weight of each indicator by calculating the indicator’s difference coefficient $(d_j=1-e_j)$:

$$W_j={\displaystyle \frac{d_j}{{\sum }_{j=1}^n{d}_j}}$$

(14)

The implementation of the above steps will form a weight matrix, providing a basis for the weights for the subsequent evaluation of the level of digital trade development.

Finally, based on the weights of each indicator, the sample data are weighted and summarized to calculate the evaluation score of the degree of digital trade development of each province in the Yangtze River Economic Belt for the corresponding year:

$$F=\sum _{j=1}^n{x}_{ij}{W}_j$$

(15)

The larger this value, the higher the level of digital trade in that region.

4.2. Corporate Export Competitiveness Indicator System

In the research on measuring corporate export competitiveness, the Export Sophistication Index (ESI) is an important indicator. It reflects the international competitiveness of products by calculating the Revealed Comparative Advantage (RCA) and adjusting for quality based on per capita GDP [79]. In addition, export product quality, product variety diversity, market share, and the Revealed Comparative Advantage Index (RCA) are also commonly used indicators to measure export competitiveness. In empirical research, researchers may combine these indicators to comprehensively analyze the export performance of enterprises. This provides a new perspective for understanding the competitiveness of enterprises in the international market.

Therefore, we select the Export Sophistication Index as a proxy variable for corporate export competitiveness. The Export Sophistication Index is obtained by calculating the Revealed Comparative Advantage (RCA) of products in the international market and adjusting for quality based on per capita GDP. The specific calculation method is as follows:

First, calculate the technical sophistication of product k, ${PRODY}_k:$

$${PRODY}_k=\sum _c{\displaystyle \frac{\left(x_{ck}/X_c\right)}{\sum \left(x_{ck}/X_c\right)}}\times pcgd{p}_c$$

(16)

Here, ${PRODY}_k$ represents the technical sophistication of product k indicated by the HS6 code, $x_{ck}$ represents the export value of country c, product k, $X_c$ represents the total export value of all products of country c, and $pcgd{p}_c$ represents the per capita GDP of country c. In fact, the first term on the right side of the equation ${\displaystyle \frac{\left(x_{ck}/X_c\right)}{\sum \left(x_{ck}/X_c\right)}}$ represents the Revealed Comparative Advantage of country c in product k.

Secondly, adjust the quality of the technical sophistication of the product:

$$QUALIT{Y}_{ck}={\displaystyle \frac{Pric{e}_{ck}}{{\sum }_n\left({\mu }_{nk}\times Pric{e}_{nk}\right)}}$$

(17)

Here, $Pric{e}_{ck}$ represents the export price of country c, product k. ${\mu }_{nk}$ represents the proportion of the export value of country n, product k in the total export value of the world’s product k, so the quality and competitiveness of the export product can be measured by calculating the relative price of the product.

Furthermore, based on the quality adjustment results, process the technical sophistication of the product:

$$PRO{DY}_k^{adj}={\left(QUALIT{Y}_{ck}\right)}^{\lambda }\times PROD{Y}_k$$

(18)

Here, $\lambda$ is the adjustment factor used to adjust $QUALIT{Y}_{ck}$ obtained from Equation (16) to reflect the impact of the quantity of product k on technical sophistication. In the above formula, $PRO{DY}_k^{adj}$ is the quality-adjusted version of $PROD{Y}_k$, taking into account the contribution of product quality to technical sophistication.

Finally, sum up the adjusted technical sophistication of the products according to the weight to calculate the adjusted Export Sophistication Index of the enterprise, ${ESI}_i^{adj}$:

$${ESI}_i^{adj}=\sum _k\left({\displaystyle \frac{x_{ik}}{X_i}}\right)\times PRO{DY}_k^{adj}$$

(19)

In the regression model of the empirical part of the following text, ${ESI}_i^{adj}$ will be log-transformed.

5. Empirical Model

Based on the theoretical model and indicator system described above, this paper will further explore the impact of digital trade on the export competitiveness of A-share listed companies in the Yangtze River Economic Belt of China through empirical analysis to verify the causal relationship between digital trade and export competitiveness, and further explore the mechanism by which digital trade affects export competitiveness.

5.1. Econometric Model

To analyze the impact of digital trade on export competitiveness, we construct an econometric model as follows:

$$\mathit{ln}\left(ES{I}_{it}\right)={\beta }_0+{\beta }_1\cdot {Digital}_{it}+\gamma \cdot Control{s}_{it}+{\gamma }_i+{\lambda }_t+{ϵ}_{it}$$

(20)

Here, $ES{I}_{it}$ is the dependent variable, representing the export competitiveness of firm i in year t, which is calculated as described in Section 3.2 of the previous text, and we take the natural logarithm of $ES{I}_{it}$ to help stabilize the variance and linearize potential nonlinear relationships; ${Digital}_{it}$ is the core independent variable, which quantifies the level of digital trade of firm i in year t, and this variable is designed to capture the degree of firm engagement and implementation in the process of digital trade and how these activities are transformed into enhancing its competitiveness in the international market [80], which is calculated as described in Section 3.1 of the previous text; the vector of control variables $Control{s}_{it}$ includes the firm’s sales expenses ($Market$), return on total assets ($ROA$), the proportion of shares held by the top five shareholders ($Top5Share$), financial leverage ($Leverage$), and the Herfindahl-Hirschman Index ($HHI$); ${\gamma }_i$ and ${\lambda }_t$ are firm fixed effects and year fixed effects, respectively, the former can eliminate all factors common to all firms that do not change over time, while the latter is used to control for macroeconomic conditions, policy changes, and other factors that do not change with firms; ${ϵ}_{it}$ is the random disturbance term.

5.2. Indicator Selection and Data Sources

Given that the dependent variable and core independent variable have been discussed in the previous text, this section introduces the control variables:

(1)

Sales Expenses ($Market$): The level of sales expenses is closely related to the export competitiveness of a firm. A firm can increase its sales expenses to strengthen market share and brand awareness, thereby enhancing the premium ability of its products in the international market; however, if the sales expenses are too high, the pricing and profitability of the products will be under pressure, thereby reducing their competitiveness in the international market. In the empirical analysis, we have log-transformed this indicator.

(2)

Return on Total Assets ($ROA$): Return on total assets is an important indicator for measuring the profitability and asset use efficiency of a firm. The higher the value of this indicator, the more effectively a firm can use its assets to generate profits, which will affect its export competitiveness.

(3)

The Proportion of Shares Held by the Top Five Shareholders ($Top5Share$): The shareholder structure of a listed company significantly affects the firm’s decision-making and strategic direction; therefore, this indicator can reflect the concentration of ownership in a firm. Highly concentrated ownership may lead to more aggressive strategies, such as increasing R&D investment and market expansion, thereby enhancing export competitiveness; on the contrary, dispersed ownership may lead to more conservative strategies, affecting the firm’s performance in the international market.

(4)

Financial Leverage ($Leverage$): Financial leverage reflects the extent to which a firm relies on debt financing. Firms with high leverage may face greater financial risks in the face of market fluctuations, which may limit their competitiveness in the international market; at the same time, moderate financial leverage can bring financial leverage effects to the firm, improving its return on investment, thereby helping to enhance its competitiveness in the international market.

(5)

Herfindahl-Hirschman Index ($HHI$): This indicator measures market concentration, reflecting the market share of dominant firms in the market. In a highly concentrated market, firms may face intense competition, which may affect their export strategies and competitiveness.

In summary, the variables involved in the empirical analysis are shown in

Table 2. Variables and explanations in the empirical analysis.

Variable Type	Variable Name and Symbol	Variable Meaning
Dependent Variable	Export Sophistication Index ($ESI$)	Reflects the competitiveness and influence of a region’s products in the international market, taking into account the diversity and technical complexity of export products
Independent Variable	Digital Trade Development Level ($Digital$)	Measures the degree of development of enterprises in digital trade, affecting their position in the global value chain and export competitiveness
Control Variables	Log of Sales Expenses ($Market$)	Affects the total cost of enterprises, related to market positioning and brand image, which may affect export competitiveness
	Return on Total Assets ($ROA$)	Reflects the advantages of cost control, product pricing, and market positioning of enterprises, affecting export competitiveness
	Top Five Shareholders’ Shareholding Ratio ($Top5Share$)	Affects the decision-making process, risk-bearing capacity, and long-term investment of enterprises, thereby affecting export competitiveness
	Financial Leverage ($Leverage$)	Affects the competitiveness of enterprises in the international market, excluding the interference of financial risks and capital structure
	Herfindahl-Hirschman Index ($HHI$)	Reflects the market share of dominant enterprises in the market, affecting the export strategies and competitiveness of enterprises

Source: Own elaboration.

.

5.3. Data Sources and Processing

The research period for the empirical analysis in this paper spans from 2011 to 2021, and the analysis focuses on companies listed on the A-share market in the Yangtze River Economic Belt region of China. The following criteria were applied during the sample selection process: First, ST stocks, which can affect the health of the securities market and are considered high-risk, were excluded from the sample selection; second, samples with incomplete data over the study period were also excluded; finally, to address extreme values that might affect the results of the empirical analysis, a winsorizing process was applied at the 1% upper and lower tails.

Specifically, we selected listed companies in the Yangtze River Economic Belt that are of a certain size and representativeness within their industries to ensure that the sample covers major sectors, thereby enhancing the generalizability of our conclusions. Due to missing data for some companies in certain years, we excluded those with incomplete information to maintain data integrity and accuracy in our analysis. Following the initial data collection, we conducted multiple rounds of screening, excluding companies with financial irregularities, regulatory penalties, and other factors that could introduce bias, ultimately resulting in a sample of 5902 data entries.

After the above screening steps, 5902 sample data points from 1199 companies were ultimately determined for empirical analysis. The original data for the enterprise digital trade index and the financial information required for the study were primarily obtained from the CSMAR database [81]; data on enterprise R&D investment were sourced from the Wind Economic Database [82]; regional-level indicator data were derived from the annual “China City Statistical Yearbook” [83]; and data related to enterprise export trade were obtained from the statistical data of the General Administration of Customs of China [84].

5.4. Variable Descriptive Statistics

Based on the aforementioned variable settings and data sources, this paper presents the descriptive statistics for each variable, as shown in

Table 3. Descriptive statistics of variables.

Variable	Observations	Mean	SD	Min	Median	Max
$ESI$	5902	10.77	2.12	0.00	11.13	13.89
$Digital$	5902	0.22	0.12	0.02	0.21	0.42
$Market$	5902	21.28	1.27	8.43	21.19	27.51
$ROA$	5902	0.05	0.07	−1.40	0.05	0.59
$Top5Share$	5902	54.14	14.83	9.50	54.46	95.55
$Leverage$	5902	1.37	1.88	−14.32	1.05	79.09
$HHI$	5902	0.16	0.12	0.04	0.13	1.00

Source: Own elaboration.

:

The descriptive statistics reveal key insights into the study’s variables. The Economic Sentiment Indicator ($ESI$) has a mean of 10.77, indicating moderate sentiment, while the Digital Trade Indicator ($Digital$) shows a mean of 0.22, reflecting low digital trade development. The Market Size Indicator ($Market$) has a mean of 21.28, suggesting consistency in market sizes. The Return on Assets ($ROA$) averages 0.05, indicating low profitability. The Top 5 Market Share ($Top5Share$) averages 54.14, highlighting significant market concentration, and the Financial Leverage Ratio ($Leverage$) has a mean of 1.37, indicating varied debt levels. The Herfindahl-Hirschman Index ($HHI$), with a mean of 0.16, suggests a competitive market structure.

6. Empirical Results and Recommendations

6.1. Benchmark Regression Analysis

Before conducting the benchmark regression, we first carried out the Hausman test, as this method compares fixed effects and random effects models and provides reliable results for identifying endogeneity. The test results are shown in

Table 4. Results of the Hausman test.

	Coef.
Chi-square test value	128.313
p-value	0.000

Source: Own elaboration.

.

Based on the Hausman test results, it is significantly noticeable at the 1% level, indicating that the fixed effects model is consistent at the 1% significance level. Thus, compared to the random effects model, the fixed effects model is more appropriate.

According to these results, the results of baseline regression are shown in

Table 5. Results of the baseline regression.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	$\mathit{ln}\left(ESI\right)$	$\mathit{ln}\left(ESI\right)$	$\mathit{ln}\left(ESI\right)$	$\mathit{ln}\left(ESI\right)$	$\mathit{ln}\left(ESI\right)$	$\mathit{ln}\left(ESI\right)$	$\mathit{ln}\left(ESI\right)$
$digital$	0.8978 *	0.7252 *	0.7115 *	0.7245 **	0.7245 **	0.8966 ***	0.9041 ***
	(1.9020)	(1.9497)	(1.9194)	(2.3824)	(2.3824)	(3.5047)	(3.5361)
$Market$		−0.1884 ***	−0.1612 ***	−0.1555 ***	−0.1555 ***	−0.1058 ***	−0.1057 ***
		(−7.1615)	(−6.0571)	(−7.1491)	(−7.1491)	(−5.3967)	(−5.3934)
$ROA$			1.0034 ***	1.0732 ***	1.0732 ***	0.4753 **	0.4979 **
			(5.8814)	(7.6614)	(7.6614)	(2.2297)	(2.3346)
$Top5Share$				−0.0038 ***	−0.0038 ***	−0.0027 **	−0.0028 **
				(−2.9244)	(−2.9244)	(−2.5196)	(−2.5499)
$Leverage$						−0.0019	−0.0021
						(−0.5097)	(−0.5551)
$HHI$							−0.3401 **
							(−2.4265)
Time FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes
IndividulFE	Yes	Yes	Yes	Yes	Yes	Yes	Yes
$N$	5902	5902	5902	5902	5902	5902	5902
$R^2$	0.6330	0.6612	0.6597	0.6812	0.6825	0.6833	0.6838

Source: Own elaboration. Note: ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively, and the $t$ value is in parentheses.

as follows:

From Table 5, the core explanatory variable “digital” in columns (1) to (7) are all positive and significant, at least at the 10% level, particularly in Column (7), which includes all control variables; the regression coefficient of the core explanatory variable is 0.9041 and significant at the 1% level. This finding indicates that digital trade has a positive impact on export competitiveness. With the inclusion of all control variables, for every unit increase in the level of digital trade development, export competitiveness will increase by 0.9041 units. These results demonstrate the positive role of digital trade development in enhancing enterprises’ export competitiveness.

Regarding the control variables, the regression coefficient of sales expenses ($Market$) is negative across all columns and significant at the 1% level. This finding indicates a negative correlation between a company’s sales expenses and export competitiveness. If sales expenses are too high, company profits will be negatively affected, thus reducing competitiveness in the international market. Moreover, the coefficient of financial leverage ($Leverage$) in the model is negative but not significant. This may be due to the multifaceted impact of financial leverage on a company’s export competitiveness: (1) Financial leverage can enhance a firm’s available capital through increased debt financing, supporting production expansion and technological upgrades, thereby boosting export competitiveness. However, excessive leverage may elevate financial risk, leading to higher interest expenses and reduced profitability, which may undermine competitiveness. (2) The impact of financial leverage on export competitiveness may vary significantly across industries due to differences in capital structures and market characteristics. (3) Factors such as economic cycles, interest rates, and exchange rate fluctuations influence the costs and benefits of financial leverage, thereby affecting a firm’s export competitiveness in diverse ways.

6.2. Robustness Test

To ensure the robustness of the above benchmark regression results, this paper employs the following three methods for testing:

• Replacement of the core explanatory variable. To more accurately measure the actual performance of enterprises in digital trade, this paper selects cross-border e-commerce data as a replacement for the core explanatory variable. Cross-border e-commerce data directly reflect the digital trade activities of enterprises in the global market. As a key component of digital trade, cross-border e-commerce has rapidly developed in global trade and has become a major channel for driving exports. Through cross-border e-commerce platforms, enterprises can bypass traditional trade barriers, directly access overseas markets, improve transaction efficiency, and reduce costs. Therefore, cross-border e-commerce data can more precisely capture the breadth and depth of an enterprise’s participation in international digital trade. This paper replaces the original core explanatory variable representing the level of digital trade with cross-border e-commerce transaction data (**Cross-border_eCom**). The regression results after this substitution are shown in Column (1) of

  Table 6. Results of the robustness test.

  	$\left(1\right) \mathit{E}\mathit{S}\mathit{I}$	$\left(2\right) \mathit{C}\mathit{o}\mathit{s}\mathit{t}\_\mathit{r}\mathit{a}\mathit{t}\mathit{i}\mathit{o}$	$\left(3\right) \mathit{E}\mathit{S}\mathit{I}$
  	(Replacement of the core explanatory variable)	(Replacement of the dependent variable)	(Reduction of the sample space)
  $eCom$	0.001 **
  	(7.4752)
  $digital$		–0.0006 **	0.001 *
  		(–2.2183)	(5.7957)
  Control variable	Yes	Yes	Yes
  Time FE	Yes	Yes	Yes
  Individual FE	Yes	Yes	Yes
  $N$	5252	4758	4257
  $R^2$	0.8356	0.8653	0.8529

  Source: Own elaboration. Note: **, and * indicate significance at the 5% and 10% levels, respectively, and the $t$ value is in parentheses.

  , demonstrating the significant role of cross-border e-commerce in enhancing the export competitiveness of enterprises.
• Replacement of the dependent variable. In the context of increasingly intense global competition, operating costs are a core element in measuring an enterprise’s international competitiveness. Therefore, by calculating the ratio of an enterprise’s operating costs to its operational income, we can obtain the “Cost_ratio” indicator representing their cost competitiveness. This will be used as a proxy variable to measure the firm’s export competitiveness, thereby replacing the dependent variable “ESI” in the benchmark regression. The regression results after replacing the dependent variable are shown in Column (2) of Table 6.
• Reduction of sample space. As China officially issued 4G licenses and launched their respective commercial services in 2013 [85], we further shortened the research period to 2013–2021 to test the robustness of the benchmark regression results. At the same time, we have also excluded companies that were penalized by the China Securities Regulatory Commission (CSRC) for information disclosure issues during the research period to further eliminate potential outliers and biases. Because companies penalized by the CSRC often exhibit improper or non-compliant information disclosure practices, which can result in their financial data, operational information, and other key metrics being inaccurate, incomplete, or untimely [86]. This distortion of data may affect these companies’ performance on indicators of the digital economy and export competitiveness, leading to significant discrepancies when compared to companies in full compliance. Such discrepancies could skew the accuracy of the regression analysis. Moreover, companies penalized for information disclosure violations during the study period may also face other management or financial risks, such as weak internal controls, financial manipulation, or liquidity issues [87]. These risk factors can have a unique impact on their performance in digital trade and export competitiveness, rendering them different from most compliant companies and potentially introducing systematic bias.

The regression results after reducing the sample space are shown in Column (3) of Table 6.

From the table above, we can see that

• After replacing the core explanatory variable, the digital transformation of enterprises still significantly and positively impacts export competitiveness at the 1% level. This indicates that the digital transformation of an enterprise will significantly enhance its export competitiveness.
• After replacing the dependent variable, the level of digital trade of a company has a significant negative influence on its cost advantage at the 5% level. This demonstrates that enhancing the level of digital trade will significantly reduce the ratio of an enterprise’s operating costs to its operational income, and thus acquire more cost competitiveness in the international market.
• After reducing the sample space, the digital transformation of an enterprise still significantly positively impacts its export competitiveness, which validates the robustness of the benchmark regression results from before.

6.3. Endogeneity Test

Considering potential endogeneity issues such as measurement errors, omitted variables, or bidirectional causality in the econometric model mentioned above, we selected the terrestrial distance between the company’s city and Hangzhou to serve as an instrumental variable for endogeneity testing of the benchmark regression results. Hangzhou is a pioneer in digital trade in China [88,89]. The rise of digital payment platforms such as Alipay has set a benchmark in the field of domestic digital trade and provided invaluable experience and insights for the global development of digital trade [90].

On the one hand, concerning the relevance of the instrumental variable, the closer a company is to Hangzhou, the richer the digital trade resources it can access. Such resources include, but are not limited to, efficient logistics networks, convenient cross-border payment systems, and market access and trade facilitation services provided by government and industry organizations. These resources are crucial for enterprises to enhance their digital trade capabilities, optimize supply chain management, improve market response speed, and increase customer satisfaction. Additionally, Hangzhou’s open trade environment and innovative culture provide enterprises with a good platform for learning and communication, which is conducive to achieving innovation and breakthroughs in digital trade applications.

On the other hand, considering the exogeneity of the instrumental variable, there is no direct causal relationship between the geographical index, the distance between the company and Hangzhou, and the error term of the model. An enterprise’s export competitiveness is affected by various factors, such as product quality, brand influence, market positioning, and international trade policies, and these factors have no direct link with geographic distance.

Based on the above analysis, we choose the terrestrial distance between the company’s city and Hangzhou as an instrumental variable ($distance$) to carry out endogeneity testing. The results are shown in

Table 7. Results of endogeneity test.

	(1)	(2)
	First Stage	Second Stage
$distance$	−0.002 ***
	(−4.84)
$Market$	−0.641 ***	−0.114 ***
	(−3.13)	(−11.17)
$ROA$	−6.275	0.334
	(−1.40)	(1.57)
$Top5Share$	−0.041 ***	−0.004 ***
	(−2.88)	(−5.00)
$Leverage$	−0.222 **	−0.009 **
	(−2.37)	(−2.06)
$HHI$	5.548 ***	−0.303 ***
	(2.65)	(−3.14)
$digital$		0.000 ***
		(3.03)
$\_cons$	−6.173	8.904 ***
	(−1.46)	(54.76)
$N$	2601	2601
$R^2$	0.015	0.121

Source: Own elaboration. Note: *** and ** indicate significance at the 1% and 5% levels, respectively, and the $t$ value is in parentheses.

.

In the first stage, the relationship between the terrestrial distance between the company’s city and Hangzhou and the digital trade level of the enterprise was examined to ensure that it satisfied the relevance criteria. In the second stage, the results obtained from the first stage were utilized to analyze the specific impact of digital trade on an enterprise’s export competitiveness. From the table above, it can be seen that the first-stage results indicate a significant negative correlation between distance and level of digital trade. The second-stage results demonstrate the significant positive influence of digital trade levels on the export competitiveness of enterprises. Hence, the endogeneity concerns of the benchmark regression results mentioned earlier have been dismissed.

6.4. Heterogeneity Analysis

• Analysis of City Group Heterogeneity

As mentioned in the Introduction, China’s Yangtze River Economic Belt principally comprises the Yangtze River Delta, the Middle Yangtze River, and the Sichuan-Chongqing. Among them, the Yangtze River Delta cities have the most robust overall economic strength and significant advantages in digital infrastructure construction, e-commerce development, and financial services. Although the Yangtze River Midstream cities are late starters in digital trade, they have solid foundations in manufacturing and agriculture. Meanwhile, the Sichuan-Chongqing cities serve as an important economic hub in southwestern China.

According to the classifications of the three city clusters mentioned above, we categorize all listed companies. The results of the heterogeneity analysis are shown in

Table 8. Results of the heterogeneous analysis by city group.

	(1)	(2)	(3)
	The Yangtze River Delta City Group	The Middle Yangtze River City Group	The Sichuan-Chongqing City Group
$digital$	2.798 ***	8.866 ***	2.414 ***
	(17.38)	(9.16)	(3.41)
Control variable	YES	YES	YES
Time FE	YES	YES	YES
Individual FE	YES	YES	YES
$\_cons$	10.07 ***	8.788 ***	9.005 ***
	(34.38)	(14.39)	(21.01)
$N$	3904	665	683
$R^2$	0.01	0.05	0.07

Source: Own elaboration. Note: *** indicates significance at the 1% level, and the $t$ value is in parentheses.

.

The results in the table indicate that the level of digital trade has a significant positive impact on enterprise export competitiveness across all three city groups. Among them, the influence is the greatest in the Middle Yangtze River city group, followed by the Yangtze River Delta city group, and the impact is the least in the Chuan-Yu city group. These results may be associated with the industrial structure, policy environment, infrastructure construction, enterprises’ acceptance level, and application capability of digital technologies within the three city groups.

Since enterprises in the Yangtze River Delta city group inherently have stronger export competitiveness, the impact of digital trade levels is not leading among the three city groups. Although the Middle Yangtze River city group has the largest digital trade level influence coefficient, it still has significant developmental potential that is waiting to be further unleashed. The situation is similar for the Chuan-Yu city group. However, while harnessing potential, enterprises need to focus on expanding their exports using digital technology and enhancing competitiveness.

2.

Heterogeneity Analysis Based on Enterprise Scale

According to the “Methods for Categorizing Large, Medium, Small, and Micro Enterprises for Statistical Purposes (2017)” published by the National Bureau of Statistics of China [91], the listed company samples analyzed in this article are categorized into large enterprises and medium-small enterprises. A heterogeneity analysis was then conducted, and the results are shown in

Table 9. Results of the heterogeneous analysis by enterprise scale.

	(1)	(2)
	Large Enterprises	Medium-Small Enterprises
$digital$	1.891 ***	1.724 ***
	(13.91)	(6.12)
Control variable	YES	YES
Time FE	YES	YES
Individual FE	YES	YES
$\_cons$	9.758 ***	8.371 ***
	(34.93)	(6.71)
$N$	4769	483
$R^2$	0.02	0.03

Source: Own elaboration. Note: *** indicates significance at the 1% level, and the $t$ value is in parentheses.

.

The results from the table above indicate that the level of digital trade of both large enterprises and medium-small enterprises significantly positively impacts their own export competitiveness. However, the impact is greater in large enterprises (1.891) than in medium-small enterprises (1.724). The reason for this result is that large enterprises often have a clear advantage in terms of digital technology reserves and applications. They can invest heavily in advanced digital technologies, such as big data analytics, cloud computing, and automated systems, to optimize production processes and improve supply chain efficiency. Furthermore, compared to medium-small enterprises, they would be better at precisely grasping international market demands. These advantages make large enterprises internationally competitive in the global market, allowing them to effectively handle the intense international competitive environment.

3.

Industry-Specific Heterogeneity Analysis

Furthermore, we analyzed the heterogeneous impact of digital trade on different industries. The results are shown in

Table 10. Results of the heterogeneous analysis by different industries.

	(1)	(2)	(3)
	Labor-Intensive Industries	Knowledge-Intensive Industries	Capital-Intensive Industries
$digital$	1.741 **	1.829 ***	2.054 ***
	(3.25)	(12.07)	(8.21)
Control variable	YES	YES	YES
Province FE	YES	YES	YES
Year FE	YES	YES	YES
$\_cons$	10.49 ***	10.01 ***	9.224 ***
	(7.18)	(31.77)	(28.44)
$N$	307	3548	1397
$R^2$	0.22	0.02	0.03

Source: Own elaboration. Note: *** and ** indicate significance at the 1% and 5% levels, respectively, and the $t$ value is in parentheses.

.

The results from the table above indicate that in labor-intensive manufacturing, the coefficient for digital trade is 1.741, and while it is significant (t-value = 3.25), the coefficient is relatively small. This may indicate that digital trade has a weaker influence on the export technological complexity of labor-intensive manufacturing, which aligns with the fact that such industries typically rely on manual labor rather than high-tech equipment.

In contrast, for knowledge-intensive manufacturing, the coefficient for digital trade is 1.829, with a significance level of *** (t-value = 12.07). This finding suggests that digital trade has a significantly positive impact on the technological complexity of knowledge-intensive manufacturing. Considering that this type of industry often involves highly technological production processes, the support provided by digital trade in areas such as information exchange, innovation, and R&D appears to be more critical for knowledge-intensive manufacturing.

As for capital-intensive manufacturing, the coefficient for digital trade is 2.054, with a significance level of *** (t-value = 8.21). This indicates that digital trade has a substantial impact on the technological complexity of capital-intensive manufacturing, which may be related to the fact that such industries are more likely to benefit from digital technologies. Capital-intensive manufacturing typically requires substantial financial investment, and digital trade may offer more efficient ways of facilitating capital flows and resource allocation, thus enhancing its technological complexity.

These results align with the theoretical framework outlined earlier in this paper. The factors driving the impact of digital trade vary across different types of manufacturing industries, and this has been confirmed by empirical findings. The significant effects of digital trade observed in knowledge-intensive and capital-intensive manufacturing highlight the importance of digital technologies in high-tech and capital-intensive sectors.

6.5. Mechanism Analysis

1.

Mediation Model Setup and Variable Selection

According to the theoretical analysis, the mediating variables—innovation efficiency ($Innoeff$), corporate governance ($Tobinq$), and supply chain efficiency ($Supply$)—have direct and clear impacts on export competitiveness. Therefore, the mediation effect model in this paper is set as follows:

$$Me{d}_{it}={\beta }_0+{\beta }_1{digital}_{it}+{\beta }_{j}Controle+{\gamma }_i+{\lambda }_t+{ϵ}_{it}$$

(21)

where $i$ represents firms, $t$ represents time, and the meanings of ${ESI}_{it}$, ${digital}_{it}$, ${Controls}_{it}$, ${\gamma }_i$, ${\mathsf{\lambda }}_t$ and ${\mathsf{ϵ}}_{it}$ are consistent with the baseline regression model. $Me{d}_{it}$ refers to the selected mediating variables: innovation efficiency ($Innoeff$), corporate governance ($Tobinq$), and supply chain efficiency ($Supply$). The selection of these mediating variables is explained below.

First, innovation efficiency ($Innoeff$) is used as a comprehensive indicator to measure the efficiency of corporate innovation activities, calculated by the logarithm of the number of patent applications plus 1 for listed companies. This indicator effectively reflects the input-output ratio in R&D and innovation. The paper expects that the application of digital technology will enhance firms’ ability to acquire, process, and utilize information, thereby accelerating the innovation process and improving innovation efficiency.

Second, corporate governance ($Tobinq$) is measured by Tobin’s Q, which is the ratio of a company’s market value to the replacement cost of its assets. This indicator reflects market expectations for a firm’s future profitability and represents the efficiency of its governance structure. The paper hypothesizes that good corporate governance can promote efficient resource allocation and improve decision-making, further enhancing firms’ export competitiveness with the aid of digital technology. This study explores the relationship between digital technology adoption and corporate governance, as well as how this relationship varies under different governance structures and market environments.

Third, supply chain efficiency ($Supply$) is measured by inventory turnover days, an indicator reflecting the flexibility and responsiveness of supply chain management. The paper argues that digital technology can optimize inventory management, reduce stockpiling, and improve overall supply chain efficiency. By analyzing the relationship between inventory turnover days and digital technology, the paper will verify the role of digital technology in enhancing supply chain efficiency and how this improvement translates into export competitiveness.

2.

Mediation Effect Test Results and Analysis

The results of the mediation effect test are reported in

Table 11. Results of the mediation effect test.

	(1)	(2)	(3)
	$\mathit{I}\mathit{n}\mathit{n}\mathit{o}\mathit{e}\mathit{f}\mathit{f}$	$\mathit{S}\mathit{u}\mathit{p}\mathit{p}\mathit{l}\mathit{y}$	$\mathit{T}\mathit{o}\mathit{b}\mathit{i}\mathit{n}\mathit{q}$
$digital$	0.913 **	375.2 *	428.6 *
	(3.31)	(2.10)	(2.46)
Control variable	YES	YES	YES
Province FE	YES	YES	YES
Year FE	YES	YES	YES
$\_cons$	−106.3 ***	12702.4 ***	11714.9 ***
	(−17.04)	(34.10)	(32.13)
$N$	5252	5252	5252
$R^2$	0.08	0.13	0.17

Source: Own elaboration. Note: ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively, and the $t$ value is in parentheses.

.

In the table above, Column (1) shows that there is a significant positive relationship between digital trade and innovation efficiency. This finding indicates that firms experience significant improvements in their innovation capabilities after adopting digital technologies. These improvements may stem from enhanced information-processing capacity, optimized R&D processes, and faster market response. However, it is surprising that while the improvement in innovation efficiency positively influences export competitiveness, the impact is not statistically significant. This may suggest that certain factors, such as market acceptance, the commercialization path of product innovation, and the protection of innovation outcomes, have not been fully considered in transforming innovation efficiency into market competitiveness. Therefore, while pursuing innovation efficiency, firms should also focus on how to translate innovation achievements into market competitiveness.

Column (2) shows that digital trade has a significantly positive impact on supply chain efficiency. These findings reveal that improvements in supply chain efficiency significantly enhance export competitiveness. This demonstrates the important role of digital technology in optimizing supply chain management, improving logistics efficiency, reducing costs, and increasing responsiveness to market demand. These improvements directly boost a firm’s competitiveness in international markets, as an efficient supply chain ensures that products are delivered to customers promptly, meeting market demand, and increasing customer satisfaction and loyalty. Firms should continue leveraging digital technologies to improve supply chain efficiency while maintaining flexibility and sustainability to adapt to the changing global market environment.

Column (3) shows that digital trade has a positive effect on corporate governance. Digital technology improves information transparency, optimizes decision-making processes, and enhances risk management, significantly boosting corporate governance. Improved corporate governance not only enhances internal management efficiency but also helps build more stable and reliable business partnerships, which is crucial for establishing a strong corporate image and reputation in international markets. These findings emphasize the importance of optimizing governance structures during digital transformation to ensure that technological advantages translate into long-term competitive advantages.

6.6. Policy Suggestions

Based on the above research conclusions, this paper proposes the following policy suggestions:

• Strengthen the construction of regional digital infrastructure. In order to further enhance the export competitiveness of companies in the region through the development of digital trade, it is suggested that the government establish a high-speed network corridor covering the whole area of the Yangtze River Economic Belt, ensuring that high-quality network services can be enjoyed equally from the upstream Sichuan, Chongqing to the downstream Shanghai, Jiangsu, etc. At the same time, it is also suggested that governments at all levels in the Yangtze River Economic Belt region strive to build intelligent logistic platforms, integrating the information resources of all ports and logistics parks along the river and achieving digitalization and intelligence in goods tracking, transport scheduling, warehousing management, and other aspects. By using the Internet of Things technology to improve logistics efficiency, reduce logistics costs, and enhance the logistics competitiveness of the Yangtze River Economic Belt.
• Promote cooperation between enterprises to form synergies in the industrial chain. The Yangtze River Economic Belt in China spans several provinces, with different regions each having advantages in resources, industries, and technology. It is suggested that cross-regional enterprise cooperation should be encouraged to promote resource sharing and complementary advantages through policy guidance and financial support. At the same time, strong support should be given to technology R&D cooperation between enterprises, especially for cross-industry technology integration. The establishment of joint R&D centers, innovation laboratories, etc., is recommended to promote cooperation in new materials, intelligent manufacturing, environmental protection technology, and other fields, as well as to jointly solve common industry problems and promote industrial upgrading.
• Improve the level of enterprise digital governance. Businesses should be incentivized to establish decision-making mechanisms based on data and market analysis to ensure the science and foresight of decisions. Advanced decision support systems, such as big data analysis and AI-assisted decision-making, should be introduced through policy guidance and financial support. At the same time, businesses should be guided to establish a comprehensive digital risk management system, helping them identify potential risks in international competition through regular risk assessments and audits, as well as flexible strategy development.

7. Conclusions

By constructing theoretical models, comprehensive indicator systems, and econometric models, this paper explores the causal relationship between digital trade development and the enhancement of enterprise export competitiveness. From a theoretical perspective, this paper constructs a theoretical model based on product quality heterogeneity, analyzing the impact of digital trade on enterprise export competitiveness on the basis of supply-demand balance. From an empirical perspective, this paper builds a comprehensive indicator system for measuring digital trade and enterprise export competitiveness, establishes an econometric analysis model, and uses data from listed companies in the Yangtze River Economic Belt region in China from 2011 to 2021 for empirical analysis. The research results show that for every 1 unit increase in the level of digital trade development in this region, a positive effect of 0.9041 units will be generated on enterprise export competitiveness. After a series of robustness tests and endogeneity analyses, these empirical results have been confirmed as robust and reliable.

The findings of this research, focusing on the Yangtze River Economic Belt, present insights that can be adapted and applied to other geographical areas abroad. The core mechanisms identified in this study—digital trade development, regional digital infrastructure, inter-enterprise cooperation, and digital governance—offer a universally applicable framework for enhancing export competitiveness through digital advancements.

In conclusion, while tailored to the Yangtze River Economic Belt, the policies and mechanisms proposed in this study offer a strategic roadmap that can be adapted to other regions seeking to improve export competitiveness through digital trade and innovation. By adopting these strategies, international regions can achieve similar advancements and foster sustainable growth in their respective markets.

The limitations of this paper are, on the one hand, its focus mainly on the Yangtze River Economic Belt. While representative, these findings may not be universally applicable to other regions. On the other hand, the data in this study covers the period from 2011 to 2021. Given the rapid development of the digital economy, however, influencing factors may change over time.

To further deepen the findings of this paper and expand research on digital trade and enterprise export competitiveness to other regions and international contexts, future research can focus on the following areas: (1) Future studies could include data from companies in more countries and regions and extend the time frame to validate the applicability of the impact of digital trade on export competitiveness across different geographical contexts. (2) Considering that external factors, such as geopolitical conditions and international policies, may influence the development of digital trade and its role in exports, future studies could incorporate these external variables into models to enhance explanatory power and predictive accuracy. (3) Future studies could investigate the potential of digital trade to support corporate green transformation and the export of environmental technologies. Analyzing how digitalization promotes resource conservation, reduces carbon emissions, and enhances firms’ green competitiveness in global markets would contribute theoretical support to the development of green trade in the digital economy context.