Introduction

In its ascent towards global economic preeminence, China has undergone transformative alterations in its provincial export trade architecture, metamorphosed by the intricate orchestration of economic vectors and technological advents within the globally interconnected milieu. Central to this paradigm shift is the synthesis of foreign direct investment, the burgeoning trajectory of e-commerce, the proper deployment of indigenous labor resources, and tactically channeled technological capital. An adept comprehension of these intricate dynamics becomes essential for informed forecasts pertaining to China’s export evolution and its symbiotic relationship with sustainable developmental objectives. The exponential proliferation of China’s export vertical can be attributed to its accurate incorporation of foreign direct investment, pivotal in catalyzing technological assimilations, fortifying workforce competencies, and forging novel market corridors. In tandem, the surge in e-commerce has revolutionized market penetration modalities, enabling Chinese offerings to seamlessly infiltrate global commerce arenas. Moreover, China’s abundant labor capital, juxtaposed with deliberate technological ventures, has accentuated its competitive foothold in global trade arenas. Yet, the velocity of this expansive trajectory beckons a meticulous assessment through a prism of sustainability, addressing facets of resource optimization, laboral integrity, and ecological prudence.

In the current academic landscape, a significant emphasis has been placed on dissecting the myriad ramifications of foreign direct investment on export enhancement, with studies underscoring its cardinal role in technological integration and amplifying operational efficacy. The integration of e-commerce facets, as delineated by Hao et al. (2023), offered a refined perspective, spotlighting the instrumental role of digital conduits in transcending conventional trade barriers. The interrelation of labor capital, as articulated by Zhang et al. (2016), in concert with technological advancements, as expounded by Autor et al. (2015), underscored the salience of indigenous assets and frontier innovations in the export dialog. However, despite the expansive literature, a comprehensive appraisal amalgamating these aspects, especially within the framework of China’s regional disparities, is palpably lacking. From a methodological standpoint, diverse econometric paradigms have been employed in antecedent research, yet province- and year-fixed effect models are increasingly lauded for their analytical precision. The eastern corridors, advantaged by their littoral proximity, have conventionally steered the export zeitgeist. Academic contributions, such as those by Duan et al. (2020), underscored this region’s proficiency in harnessing foreign direct investment and e-commerce potentialities. In juxtaposition, the central and western sectors, albeit resource-rich and labor-abundant, evince a marked lag in technological embrace and foreign direct investment influx. This regional polarization, as theorized by Zhong et al. (2022), accentuates the necessity for tailored policy interventions to promulgate balanced and sustainable growth vectors. In this context, our scholarly pursuit seeks to redress the prevailing knowledge chasm. The intricate interplay of foreign direct investment, e-commerce, labor dynamics, and technological innovation in molding China’s export tapestry is indubitable. Yet, exhaustive scrutiny, particularly one sensitive to regional grades, stands as an academic imperative. Grounded in methodological robustness and echoing sustainability principles, this study aims to demystify this intricate interconnection, catalyze informed policy deliberations, and buttress China’s odyssey towards a sustainable export paradigm.

Drawing upon the aforementioned analytical discourse, this research delves into the complex relationship between foreign direct investment, e-commerce, and the growth of exports in China from 2005 to 2022, situated within the context of a rapidly changing global economic landscape. Using advanced statistical methods, such as province- and year-fixed effects analysis along with fully modified ordinary least squares and dynamic ordinary least-squares methods, the study gives a more complete picture of how foreign direct investment and e-commerce work together to make China’s exports stronger. A key aspect of this study is its alignment with the global sustainable development agenda, examining how China’s export growth extends beyond basic international investment and digital commerce. It integrates sustainable practices, such as the effective use of local labor and a focus on technological advancement, offering insights into the role of knowledge capital and educational attainment in boosting export figures. Our analysis reveals a pronounced regional variation in the impact of foreign direct investment and e-commerce on export trade, with Eastern China showing greater responsiveness compared to the Western regions. This finding highlights the necessity for region-specific policies and sustainable strategies to ensure a balanced distribution of foreign direct investment and e-commerce benefits across the country. The study’s methodology stands out in the existing literature for its comprehensive approach, combining advanced econometric techniques to dissect the multifaceted influences on China’s export sector. It addresses a gap in previous research by providing a clearer picture of the interplay between foreign direct investment, e-commerce, and export growth within the unique context of China’s evolving economy. The research emphasizes the need for a nuanced understanding of China’s position in the global economy, exploring the relationship between foreign direct investment and e-commerce in a way that prior empirical studies have not fully captured. By doing so, the study offers valuable insights for policymakers and stakeholders, advocating for strategies that not only foster economic growth but also align with sustainable development objectives, ensuring the long-term prosperity and environmental sustainability of China’s economy.

This study presents several significant contributions to the current academic understanding of China’s export sector, particularly focusing on sustainable development. First, our analysis synthesizes the roles of foreign direct investment and e-commerce, offering fresh insights into their collective influence on China’s exports. This aspect builds upon the work of Fidrmuc and Korhonen (2010), who underscored the impacts of global capital and digital advancements on emerging economies. Our study extends this perspective by explicitly linking these factors to export growth in the Chinese context. Second, we introduce a nuanced approach by examining regional variations in export performance, moving beyond the limitations of previous studies that often treated China’s economy as a uniform entity. Grübler et al. (2007), who emphasized the value of regional analysis in producing more thorough economic insights than national overviews, served as an inspiration for this strategy. Third, our research highlights the role of local labor resources as a key component of sustainable export strategies. This aligns with Sun’s (2022) assertion of human capital as a critical driver of economic growth, positioning it as a sustainable asset in China’s export framework. Fourth, the study delves into the impact of technological investment on sustainable export growth, expanding upon Qian et al. (2021) thesis that technology is fundamental to achieving green growth. We explore how technological advancements contribute specifically to the sustainability of China’s export sector. Lastly, the research advocates for a balanced approach to economic growth and environmental sustainability, echoing Wright’s (2019) argument for the necessity of balancing economic development with ecological preservation. Our study furthers this dialog by illustrating how such a balance can be achieved within the context of China’s export dynamics. Together, these facets of our research offer new perspectives on the complex relationship between economic activities, technological innovation, and sustainable development in the context of China’s growing role in the global market. These insights are particularly relevant for policymakers and business leaders looking to navigate the challenges and opportunities presented by China’s evolving export landscape.

The subsequent sections of this article are structured as follows: section “Literature review” delves into a comprehensive review of extant literature, shedding light on prior research in this domain. In the section “Variables and model”, we elucidate the methodological approach, detailing the variables employed and the underlying model. Section “Empirical results” offers a synthesis of the empirical results, coupled with a discussion of the implications. Lastly, the section “Conclusions” culminates with conclusions, policy recommendations, and avenues for future research in this field.

Literature review

In today’s global trade environment, the interplay between foreign direct investment and e-commerce has become a critical factor influencing export trends. Current research highlights foreign direct investment’s pivotal role in driving technology transfer and expanding markets. Concurrently, e-commerce platforms have revolutionized trading patterns, facilitating instantaneous market connections and broadening international reach. Additionally, elements like labor resource allocation and technological progress intertwine with these primary factors, creating a multifaceted framework that reveals the complexities of modern export strategies.

In contemporary academic discourse, the impact of foreign direct investment on China’s export trade has received significant attention. Yet, the complex relationship with e-commerce remains insufficiently explored. The prevailing literature, as seen in the works of Li et al. (2019), Wang et al. (2020), and Jin and Huang (2023), mainly focuses on the direct effects of foreign direct investment on export efficiency through capital infusion and technological transfers. These studies, however, tend to overlook the burgeoning dimension of digital commerce. Addressing this gap, Fu et al. (2016), Chen et al. (2023), and Lei and Xie (2023) provide a more nuanced perspective by acknowledging the role of digital transformation in global trade. They underscore e-commerce’s potential to complement foreign direct investment, particularly in enhancing market access for Chinese exports. Expanding on this viewpoint, Qi et al. (2020), Klimenko and Qu (2023), and Yan et al. (2023) examine how e-commerce platforms democratize export opportunities, even for smaller entities, thus amplifying foreign direct investment’s impact. The insights of Zhang and Yang (2022), Mahalik et al. (2023), and Cordes and Marinova (2023) served as inspiration for this research’s more integrative approach. It goes beyond the traditional analysis of foreign direct investment’s influence on exports to include the transformative role of e-commerce. This methodological advancement builds upon and extends the analyses of Götz (2020), Auboin et al. (2021), and Ha (2022), who, despite their thoroughness, did not fully address the synergistic relationship between foreign direct investment and digital trade channels. Aligned with the analytical frameworks of Agarwal and Wu (2015), He et al. (2021), and Shanmugalingam et al. (2023), this study emphasizes a thorough understanding of trade dynamics in the digital era. By incorporating e-commerce as a key variable alongside foreign direct investment, it fills a critical gap in the literature. This approach resonates with the findings of Zhang and Zeng (2023), Xiao and Abula (2023), and Sun et al. (2024) on the growing influence of digital platforms on trade and extends their work by empirically quantifying this impact within the context of China’s export landscape. In conclusion, this research contributes significantly to the existing body of literature by integrating the crucial role of e-commerce. It provides a more comprehensive view of the dynamics shaping China’s export trade, thereby addressing a vital need in the ongoing academic conversation.

The existing literature recognizes the impact of e-commerce on China’s export trade but lacks a thorough exploration of its synergistic effects with foreign direct investment and traditional trade mechanisms. Previous studies, such as those by Giuffrida et al. (2017) and Li et al. (2019), have primarily focused on the direct impact of e-commerce on market expansion and customer engagement, emphasizing its role in broadening the global reach of Chinese products. However, these studies often treat e-commerce as an isolated factor, not integrating it with broader economic elements like foreign direct investment. A more nuanced perspective is emerging from research such as Blanchard, Jean-Marc (2019), Villegas-Mateos (2022), and Singh and Singh (2022), which begin to address the interaction between e-commerce and foreign direct investment but do not provide a comprehensive analysis. These studies show how e-commerce platforms can enhance export efficiency in conjunction with foreign direct investment, yet they stop short of examining how e-commerce is transforming traditional export models. This research addresses this gap by adopting an integrative methodology, drawing on the approaches of Wang et al. (2021) and Yin and Choi (2023). This methodology extends beyond evaluating the direct effects of e-commerce on exports to also consider its interplay with foreign direct investment. Such an approach expands upon the frameworks used in studies by Zhang (2019) and Phang et al. (2019), which, while insightful, did not fully capture e-commerce’s complex dynamics within China’s integrated market economy. Additionally, this study aligns with the emerging literature, such as the works of Gao (2018) and Li et al. (2020), advocating for a holistic view of digital trade’s role in economic growth. By incorporating a comprehensive array of variables, including technological advancement and digital infrastructure quality, this research provides a more robust analysis than previous studies like those by Katz and Callorda (2018), Sinha et al. (2020), and Wei and Ullah (2022). In conclusion, this study overcomes previous shortcomings in academic research by offering a detailed empirical examination of how e-commerce, in conjunction with foreign direct investment and traditional trade mechanisms, shapes China’s export landscape. It contributes significantly to academic discourse by presenting a more complete understanding of e-commerce’s role in the modern economy, thus fulfilling a critical need in the ongoing narrative on global trade and digital economics.

In analyzing China’s export sector, the influence of labor resource allocation, technological advancements, knowledge capital, and educational attainment, particularly in relation to e-commerce, warrants a deeper exploration. Initial research efforts, exemplified by Bhaumik et al. (2016), Song and Wang (2018), and Liu and Xie (2020), have individually evaluated the impacts of labor and technology on export performance, underscoring their roles in bolstering China’s position in international markets. Yet, these studies typically overlooked the integration of e-commerce into their analytical models. Recent scholarly works, including those by Kwak et al. (2019), Elia et al. (2021), and Tang and Li (2023), have started to recognize the combined effect of technological prowess and labor skills within the framework of e-commerce. However, these investigations fall short of comprehensively examining how knowledge capital and education intersect with e-commerce to affect export trends. The methodologies of Lin et al. (2020), Hanelt et al. (2021), and Abdul-Rahim et al. (2022) served as the foundation for this study’s holistic approach to closing this research gap. Our approach is comprehensive, assessing not just the direct impacts of labor, technology, education, and knowledge on exports but also situating these impacts in the context of the growing e-commerce domain. This method expands upon the analytical scope of previous studies like Wei et al. (2020) and Li et al. (2023), which, despite their thoroughness, did not fully delve into the complex relationship between e-commerce and China’s export dynamics. Furthermore, our study aligns with the evolving scholarly narrative, as seen in the works of Banalieva and Dhanaraj (2019) and Huang et al. (2023), advocating for an integrated view of digital commerce’s interaction with traditional economic variables. By including an extensive analysis of factors such as digital infrastructure and market development in e-commerce, this research offers a more detailed examination than earlier studies by Gorla et al. (2017) and Wang et al. (2024). In summary, this research fills existing gaps in the literature by thoroughly investigating how labor resources, technological investments, knowledge capital, and education, in conjunction with e-commerce, shape the export sector in China. It provides a comprehensive perspective on the synergy between traditional economic elements and digital trade, addressing a critical need in the ongoing discussion of global trade and economic progression.

Variables and model

Variables

Numerous studies have explored the significant impact of foreign direct investment on a nation’s export trends, highlighting foreign direct investment’s critical role in reshaping export strategies. Researchers like Choong (2012) and Otchere et al. (2016) have pointed out that foreign direct investment not only provides essential capital but also facilitates technological transfer, thereby boosting efficiency and productivity in host countries. Moreover, the aspect of sustainability is increasingly becoming interlinked with foreign direct investment, often bringing eco-friendly technologies and sustainable methodologies to the forefront, enhancing a nation’s prospects for long-term export stability, as noted by Perrini and Tencati (2006). Simultaneously, the influence of the digital revolution, particularly the rise of e-commerce, has significantly transformed the nature of exports. Studies by Wang (2010) and Teng et al. (2022) highlight that in China’s expanding digital landscape, e-commerce platforms have leveled the playing field, allowing even smaller businesses to access the global market. According to Rita and Ramos (2022), Amornkitvikai et al. (2022), and He et al. (2021), e-commerce is also in line with the global trend towards sustainable trading due to its traceable and transparent nature. Considering these complex interactions, export trade volume becomes an appropriate variable to study, representing the combined and sustainable effects of foreign direct investment and e-commerce. This research, therefore, focuses on the export trade volumes of China’s provinces, incorporating foreign direct investment inflows and e-commerce transaction data as independent variables. This approach aims to shed light on their hypothesized influence on provincial export patterns.

To fully grasp the complex factors affecting export trade, it’s crucial to look beyond conventional indicators like foreign direct investment and e-commerce. A deeper exploration into academic literature and fundamental economic theories uncovers the critical role of labor resource allocation and technological advancements in shaping export patterns. The foundational Heckscher-Ohlin theorem, supported by research from Castilho et al. (2012) and Antràs et al. (2017), underscores the vital impact of labor resources on global trade trends. Darku (2021) extends this perspective, emphasizing the sustainability aspects and suggesting that effectively managed labor resources can contribute to more equitable and environmentally responsible trading practices. Additionally, examining the role of technology provides insights into the nuances of export competitiveness. Rooted in Romer’s theory of endogenous growth and backed by findings from Jones (2019) and Anzoategui et al. (2019), there is a consensus that deliberate technology investments boost productivity and support sustainable growth through cleaner, more efficient production methods. Zhou et al. (2021) further clarify this idea, showing how sustainable technologies and competitive exports are interlinked. Recognizing the importance of these two factors, this study incorporates labor resource allocation and technological inputs as key control variables. To empirically anchor these theoretical concepts, we use urban employment data from various provinces as indicators of labor resource allocation and local government spending on technology as a reflection of technological investments. Building on the work of Mansion and Bausch (2020), Lyu et al. (2022), and Mohammad Shafiee et al. (2023), this paper also introduces knowledge capital quantified by the number of patent licenses. Following Atkin (2016), Ahmed et al. (2020), and Blanchard and Olney (2017), the paper incorporates education level, measured by the average number of schooling years.

Due to data availability, this paper selects balanced provincial-level data from 2005 to 2022. Since 2005, e-commerce across various Chinese provinces has seen rapid development, making this period particularly relevant to the study’s context. The unavailability of data from Tibet necessitates the inclusion of 30 other provinces and municipalities in China, namely Beijing, Tianjin, Shanxi, Inner Mongolia, Liaoning, Jilin, Heilongjiang, Shanghai, Jiangsu, Zhejiang, Anhui, Fujian, Jiangxi, Shandong, Henan, Hubei, Hunan, Guangdong, Guangxi, Hainan, Chongqing, Sichuan, Guizhou, Yunnan, Shaanxi, Gansu, Qinghai, Ningxia, and Xinjiang. The data used in this study is sourced from three official databases, each providing specific insights into our variables of interest. The Bureau of Statistics of China supplies data on export trade volume, knowledge capital, education level, labor resource endowment, and technological investment. Information on e-commerce is obtained from the China E-commerce Report, while data on foreign direct investment is sourced from the Statistical Bulletin of China’s Outward Foreign Direct Investment.

Model

In examining the interplay between foreign direct investment and e-commerce on export trade using China’s province data, it is imperative to adopt a robust econometric technique that effectively captures both time-invariant and entity-specific heterogeneities. The two-way fixed effects regression model, as elaborated upon by Wooldridge (2010) and advocated by Baltagi (2021), is particularly adept at mitigating potential omitted variable biases in panel data, making it especially suitable for our study’s empirical context. By incorporating both entity and time-fixed effects, this approach controls for unobserved province-specific factors that may influence export trade (such as local policies or geographical advantages) and time-specific shocks (like global economic trends or national regulatory shifts) that uniformly affect all provinces. By accounting for these dual dimensions of variability, the model ensures that the estimated effects of foreign direct investment and e-commerce are purged of confounding influences, thus bolstering the credibility of causal inferences drawn from the results. Given the dynamism of China’s economic landscape, combined with the evolving trajectories of foreign direct investment and e-commerce, leveraging the two-way fixed effects regression offers a rigorous and robust approach to discerning their impact on export trade. The model is shown as follows:

$$\begin{array}{ll}{{\rm{logex}}}_{{i},{t}}={{a}}_{0}+{{a}}_{1}^{+}{{\rm{logfdi}}}_{{i},{t}}+{{a}}_{2}^{+}{{\rm{logec}}}_{{i},{t}}\\\qquad\qquad+\,{{a}}_{3}^{+}{{\rm{loglab}}}_{{i},{t}}+{{a}}_{4}^{+}{{\rm{logtec}}}_{{i},{t}}+{{a}}_{5}^{+}{{\rm{logkn}}}_{{i},{t}}\\\qquad\qquad+\,{{a}}_{6}^{+}{{\rm{loged}}}_{{i},{t}}+{\eta }_{i}+{\delta }_{{t}}+{\in }_{{i},{t}}.\end{array}$$
(1)

In Eq. (1), the subscript ‘i’ represents individual provinces, while t delineates the temporal dimension, capturing the yearly variations. Within this model, ex symbolizes the export trade volume, serving as our dependent variable. On the explanatory side, ec corresponds to e-commerce metrics, ‘fdi’ quantifies foreign direct investment inflows, ‘lab’ encapsulates labor resource endowment, and ‘tec’ signifies the magnitude of technological investment. ‘kn’ indicates knowledge capital. ‘ed’ stands for education level. The term a0 denotes the intercept, providing a baseline measure for our regression. The vector [a1, a6] comprises the coefficients estimated for each explanatory variable, reflecting their respective strengths and directions of influence on export trade. To control for potential unobserved heterogeneities, η embodies province-specific fixed effects, while δ accounts for year-specific fixed effects, ensuring that time-invariant provincial attributes and common temporal shocks are appropriately adjusted for. The error term, ϵ, is presumed to follow a white noise process, indicating randomness and the absence of serial correlation. The empirical focal points of this study are the coefficients a1 and a2. In our analytical framework, the ‘+’ symbol is strategically used to represent the expected positive effect of various independent variables—including e-commerce, foreign direct investment, labor resource endowment, technological investment, knowledge capital, and education level—on the export volumes of Chinese provinces. This symbolism is central to our hypothesis, positing that these variables play a beneficial role in shaping export trends across different provinces. The underlying premise of this hypothesis is that variables like foreign direct investment, enhanced e-commerce capabilities, and other pertinent factors positively stimulate export activities. In essence, the '+' sign indicates a probable correlation where increases or improvements in these independent variables are likely to correspond with a rise in export volumes from the provinces. Such a correlation is instrumental in dissecting how various economic elements and technological progressions, specific to China’s varied regional landscapes, can bolster the nation’s export capacity. This exploration is particularly salient for understanding China’s export mechanics. It provides a nuanced view of how strategic investments in technology, human capital development, and leveraging local resources can collectively uplift the export sector, reinforcing China’s position in the global economy. The ‘+‘ sign, therefore, not only signifies a positive correlation but also serves as a gateway to understanding the multifaceted drivers that enhance export efficiency in the context of China’s evolving economic landscape.

Robustness test

Considering the relatively modest scale of our sample in this study, there exists a plausible risk of heteroskedasticity and autocorrelation in the outcomes estimated through the use of annual and provincial fixed effects models. These statistical issues could potentially introduce biases into our analytical results, thereby affecting the reliability of our conclusions. To mitigate this challenge, our research strategically implements two advanced econometric methodologies: fully modified ordinary least squares and dynamic ordinary least squares. The fully modified ordinary least-squares technique, a refined version of the ordinary least-squares methodology, is particularly adept at addressing complexities arising from heteroskedasticity and autocorrelation. The efficacy of this approach in handling such statistical nuances is well documented in the works of scholars such as Pedroni (2001), Christou and Pittis (2002), Trapani (2015), Li et al. (2020), Kripfganz and Sarafidis (2021), Norkutė et al. (2021), and Kheifets and Phillips (2023). These studies validate the use of fully modified ordinary least squares as a robust tool for enhancing the accuracy of econometric estimations, especially in scenarios similar to those in our study. Similarly, the dynamic ordinary least squares method offers a comprehensive solution for addressing the challenges of endogeneity and serial correlation, which are common in time-series data. Research by Chudik and Pesaran (2015), Moon and Weidner (2017), Liu et al. (2020), Ahn and Thomas (2023), Hartono et al. (2023), and Fingleton (2023) underscore the effectiveness of dynamic ordinary least squares in ensuring more precise and reliable results in econometric analysis. This technique, by adjusting for both the lead and lag dynamics of the variables, enhances the accuracy of regression coefficients, thereby providing a more nuanced understanding of the underlying data patterns. Both fully modified ordinary least squares and dynamic ordinary least squares are sophisticated enhancements of the traditional least-squares approach. These methods have been specifically adapted to address the intricate statistical issues inherent in panel data analysis, like the one employed in our study. By incorporating these advanced techniques, we aim to mitigate potential biases arising from heteroskedasticity, autocorrelation, and endogeneity, thereby enhancing the credibility and robustness of our findings. Equation (2) in our study meticulously outlines the application of these methods, demonstrating their integration into our analytical framework to yield more reliable and insightful results.

$${\rm{y}}_{\rm{t}}=\widetilde{{\Gamma }_{2{^\prime} 1}}{\rm{D}}_{1{\rm{t}}}+\widetilde{{\Gamma }_{2{^\prime} 1}}{\rm{D}}_{1{\rm{t}}}+\tilde{{\upepsilon }_{\rm{t}}}.$$
(2)

This outcome is directly derived from the differential regression, as shown in Eq. (3).

$$\Delta {\rm{y}}_{\rm{t}}=\widetilde{{\Gamma }_{2{^\prime} 1}}\Delta {\rm{D}}_{1{\rm{t}}}+\widetilde{{\Gamma }_{2{^\prime} 1}}\Delta {\rm{D}}_{1{\rm{t}}}+\widetilde{{\upepsilon }_{\rm{t}}}.$$
(3)

Let’s consider \(\tilde{\Theta }\) and \(\tilde{\Psi }\) to represent the long-term covariance matrix calculated using the residuals denoted by \([\tilde{{\uptau }_{{\rm{t}}}}=(\tilde{{\uptau }_{1{\rm{t}}}},\tilde{{\uptau }_{2{\rm{t}}{^\prime} }}){^\prime} ]\). Based on this assumption, we are able to represent the modified data as depicted in Eq. (4). This representation considers the complex interdependencies reflected in the covariance matrix, laying the groundwork for subsequent examination and understanding of the data within the framework of our chosen model.

$${\rm{y}}_{\rm{t}}^{\ast }={\rm{y}}_{\rm{t}}-\widetilde{{\upupsilon}_{12}}{\widetilde{{\Theta }_{22}}}^{-1}\widetilde{{\Psi }_{22}}.$$
(4)

In our study, the term for bias correction, crucial for refining our model, is detailed in Eq. (5). This component is essential for enhancing the accuracy and dependability of our results, as it compensates for possible biases encountered during the estimation phase.

$${\lambda }_{12}^{\ast }={\lambda }_{12}-\widetilde{{\upupsilon}_{12}}{\widetilde{{\Theta }_{22}}}^{-1}\widetilde{{\Psi }_{22}}.$$
(5)

Therefore, the formulation of the fully modified ordinary least-squares estimator, pivotal to our analysis, is encapsulated in Eq. (6). This estimator is integral to refining our estimations, as it addresses potential issues of serial correlation and endogeneity within our regression models. By employing the fully modified ordinary least-squares method, we gain a more accurate and insightful comprehension of the relationships present in our dataset.

$$\tilde{\upvarphi }=\left[\begin{array}{c}\widetilde{\upbeta }\\ \tilde{{\gamma }_{1}}\end{array}\right]={\left(\mathop{\sum }\limits_{\rm{t}=1}^{\rm{T}}\cdot \sum _{\rm{t}}\cdot \Sigma {\rm{t}}\right)}^{-1}\left(\mathop{\sum }\limits_{\rm{t}=1}^{\rm{T}}{\rm{Z}}_{\rm{t}}{\rm{y}}_{\rm{t}}^{\ast }-{\rm{T}}\left[\begin{array}{c}\widetilde{{\lambda }_{12}^{\ast }}{^\prime} \\ 0\end{array}\right]\right)$$
(6)

In Eq. (6), \({\rm{Z}}_{\rm{t}}=({\rm{y}}_{\rm{t}}^{{\prime} }{\rm{D}}_{\rm{t}}^{{\prime} })\). Developing estimators for the long-term covariance matrix, a critical component in the implementation of fully modified ordinary least squares, is highlighted in studies by Atil et al. (2023), Wagner (2023), Phillips and Kheifets (2024), and Pelagatti and Sbrana (2024). This process is essential for the precision and efficacy of the fully modified ordinary least-squares approach. It entails refining the OLS regression by including both preceding and subsequent factors, ensuring that the error component in Eq. (1) remains uncorrelated with the entire historical sequence of random regressor variations. This method, as detailed in the research by Mark and Sul (2003), Panopoulou and Pittis (2004), Bruns et al. (2021), and Wang et al. (2024), is efficiently captured in Eq. (7).

$${\rm{y}}_{\rm{t}}={\rm{x}}_{\rm{t}}^{{\prime} }\upbeta +{\rm{D}}_{1{\rm{t}}}^{{\prime} }{\gamma }_{\rm{t}}+\mathop{\sum }\nolimits_{\rm{j}=-{\rm{q}}}^{\rm{r}}\Delta {\rm{X}}_{\rm{t}+{\rm{j}}}^{{\prime} }\updelta +{\uptau }_{1{\rm{t}}}$$
(7)

By incorporating q lags and r leads of the differenced regressors, the persistent correlation between variables τ1t and τ2t is effectively neutralized. This adjustment allows the estimation of φ = (β′, γ′)’ through the least-squares estimator to align with the asymptotic distribution achieved via the fully modified ordinary least-squares method. These methods are notably effective, as emphasized by Bai et al. (2021), Chebrolu et al. (2021), De Menezes et al. (2021), Zhao et al. (2022), and Bollen et al. (2022), in overcoming challenges like endogeneity, serial correlation, and biases that are typically prevalent in studies with smaller sample sizes.

Empirical results

Descriptive statistical analysis

For the purpose of this study, data extraction was conducted, harnessing information from 31 distinct provincial datasets covering the temporal bracket of 2005–2022. This compilation was sourced directly from the authoritative National Bureau of Statistics of China, ensuring data authenticity and integrity. An initial stage of rigorous analytical procedures was executed, encompassing both qualitative descriptive statistical evaluations and quantitative correlation analyses. This served to provide a holistic view of the data landscape, enabling the identification of patterns and inter-variable relationships. The culminating findings from this analytical phase are methodically tabulated in Table 1. For clarity and comprehensive representation, the results are segmented into two distinct panels: Panel A elucidates the statistical analysis of variable description, while Panel B delineates the correlation matrices.

Table 1 Descriptive statistical analysis.
Full size table

Within Panel A of Table 2, an examination of the data yields insights into provincial economic dynamics. The export trade registers an average value of 2.226, complemented by a notably narrow standard deviation of 0.085. This suggests a trend of ascent in export trade across the majority of provinces. Conversely, the foreign direct investment landscape, with a mean of 0.241 and a slightly more dispersed standard deviation of 0.117, indicates a predominant trajectory of foreign direct investment enhancement among provinces, albeit with some variability. E-commerce, represented by a mean of 2.276, portrays a positive trend; however, its relatively expansive standard deviation of 0.572 implies a diverse range of advancements and perhaps volatility within this sector. This is emblematic of the rapidly evolving and heterogeneous landscape of e-commerce in China, a reflection that aligns with empirical observations on the nation’s digital commerce forefront. The labor resource endowment is quantified with a mean of 2.791 and a standard deviation of 0.315, providing insights into a generally favorable labor capital across provinces. The metrics for technological investment, with an average of 0.997 and a standard deviation of 0.441, underline the ongoing endeavors in technological innovation but also hint at disparities in the extent and pace of such investments across the provinces. Finally, the metric for knowledge capital is calculated with an average value of 4.012 and a standard deviation of 1.506. Meanwhile, the education level is measured, showing an average value of 0.907 and a standard deviation of 0.217.

Table 2 Results of the effects of foreign direct investment and e-commerce on export trade.
Full size table

In the wake of conducting a correlation analysis, the subsequent findings are articulated in Panel B of Table 1. An inaugural examination of the data reveals a discernible positive relationship between foreign direct investment and e-commerce relative to the scope of China’s provincial export trade. Parallel to this, a deeper analytical traverse into the data underscores a tangible connection between labor resource endowment and technological forays as pivotal determinants of export trajectories. This interrelationship accentuates the premise that provinces emphasizing sustainable labor methodologies and avant-garde technological endeavors are not solely shaping a resilient economic structure but are concurrently enhancing their export trade capacities. This synergy between sustainability-oriented strategies and burgeoning trade volumes fortifies the argument that sustainability stands as a potent stimulant, accentuating both foreign direct investment and e-commerce outcomes. Furthermore, the analysis essentially establishes a positive correlation between knowledge capital, education level, and the export trade of China’s provinces.

Results

In this investigation, a quintet of econometric techniques is deployed to discern the nuanced impacts of foreign direct investment and e-commerce on export trade. These methodologies encompass pooled ordinary least squares (Model 1), panel ordinary least squares (Model 2), province-specific fixed effects (Model 3), year-fixed effects (Model 4), and a provincial and year-fixed-effects approach integrating both provincial and annual dimensions (Model 5). The outcomes of these estimations are documented in Table 2. Upon evaluating the data through the prism of the Chow test, we discerned a clear rejection of the null hypothesis, indicating the inadequacy of pooled ordinary least squares for this dataset. Subsequent to this, the Hausman test was executed, which further rejected the null hypothesis, rendering the province-fixed effect model suboptimal. The decision to employ Model 5—integrating both province- and year-fixed effects—is grounded in several advanced econometric postulations. Kropko and Kubinec (2020), Hill et al. (2020), and Fernández-Val and Weidner (2018) posited that in the presence of unobserved heterogeneity—factors that remained constant over time but vary across entities or vice versa—implementing province- and year-fixed effects can yield unbiased and consistent estimators. This became particularly salient when considering phenomena such as global economic oscillations or overarching regulatory changes, which exerted a consistent impact across all provinces. By accounting for these twin axes of variability, Model 5 ensures the extrication of extraneous influences from the core relationship between foreign direct investment, e-commerce, and export trade. This approach enhances the robustness of the analysis, fortifying the validity of causal extrapolations drawn from the empirical results.

In Table 2, our primary focus is on the insights garnered from Model 5. However, it is crucial to recognize the crucial role that the outcomes of the additional four models played. These models act as a robustness check, lending further credibility to our main findings. Model 5’s empirical data highlights a robust and statistically significant link between the surge in foreign direct investment and the increase in export trade within Chinese provinces. Specifically, a 1% increase in foreign direct investment inflows is associated with a 0.209% rise in provincial export trade volume. Shifting our analysis to the impact of e-commerce on the export landscape of Chinese provinces, we observe a compelling dynamic. E-commerce is identified as a significant driver of export growth. Quantitatively speaking, a 1% growth in e-commerce activities results in a 0.405% increase in provincial export volumes. Moreover, our research identifies critical factors influencing export patterns in Chinese provinces, notably labor resources and technological investments. The study reveals that a 1% elevation in labor resource availability correlates with a 0.715% increment in export volumes at the provincial level. In the same vein, a 1% rise in technological investments is linked to a 0.304% boost in exports. Additionally, the study brings to light the constructive effects of knowledge capital and education levels on provincial export trade. An increase of 1% in these variables is found to enhance export volumes by 0.083% and 0.101%, respectively.

Discussion

The positive correlation between foreign direct investment inflows and increased export trade can be understood through various theoretical frameworks and empirical studies. Drawing on the research of Adikari et al. (2021), Rehman et al. (2023), and Zhang and Chen (2020), the eclectic paradigm suggests that foreign direct investment promotes export trade by transferring advanced technologies, managerial expertise, and marketing skills to the host country. These spillover effects enhance the competitiveness of domestic firms, boosting their export potential. Additionally, foreign direct investment helps to establish export-oriented industries within host economies, as seen in China’s Special Economic Zones, which act as production and export hubs (Chiang and De Micheaux, 2022; Ngoc et al., 2022; Huang et al., 2023; Vukmirović et al., 2021). This influx of capital, technology, and knowledge through foreign direct investment acts as a catalyst, creating a trade-friendly environment and aligning provinces with a more globally integrated economic path. Several factors support e-commerce’s positive impact on provincial export volumes. Firstly, e-commerce reduces informational disparities, fostering a transparent market conducive to robust exports. Additionally, as e-commerce platforms grow, their value proposition to users strengthens, encouraging an environment ripe for increasing transactions, including exports. Thirdly, e-commerce inherently reduces transactional friction, enabling businesses to engage more effectively in international trade. The theories and results of researchers like Onjewu et al. (2022), who contend that e-commerce lowers traditional trade barriers and enables even small businesses to participate in global markets, support this viewpoint. Lipton et al. (2018) and Fritz et al. (2004) show that online platforms allow businesses to overcome geographic limitations, thus expanding their export reach. Tolstoy et al. (2021) and Zhong et al. (2022) discuss how e-commerce’s digital footprint lessens the constraints of geographical distance, creating a more fluid international trade environment. Khan and Khan (2021) and Watson et al. (2018) illustrate how digital trade avenues boost export growth by adapting to market changes and consumer preferences. Additionally, Xi et al. (2023) and Deng et al. (2023) highlight the relationship between digital infrastructures and export portfolio diversification, with e-commerce spurring product innovation and differentiation. In conclusion, the integration of these theoretical insights and empirical evidence underlines the significant role of e-commerce as a key driver in enhancing the scale of export trade in Chinese provinces.

Labor resources and technological investments have been identified as key factors positively influencing the scale of export trade in Chinese provinces. This result is consistent with the Heckscher–Ohlin theorem, which states that regions typically export goods that effectively use their most abundant resources, according to research from Kunroo and Ahmad (2023) and Akther et al. (2022). Given China’s substantial labor force, provinces endowed with richer labor resources are naturally capable of higher production, thereby supporting larger export volumes. Conversely, the relationship between technological investments and the strength of exports is anchored in contemporary economic growth theories, particularly those emphasizing the role of technology in economic development. Aghion et al. (1998) reinforce this notion, demonstrating that technological investment in regions not only enhances productivity but also provides a competitive advantage in international markets, thus boosting export capacity. Moreover, the study finds that both knowledge capital and education levels positively impact the scale of export trade in Chinese provinces. This underscores the importance of intellectual resources and educational attainment as drivers of export dynamics in a rapidly evolving economy like China’s. The correlation with knowledge capital reflects China’s strategic emphasis on innovation and intellectual property. Liu et al. (2017) emphasize that investments in research and development, especially in technology and sciences, have significantly enhanced China’s export capabilities, leading to an increase in patents and technological breakthroughs. Due to these advancements, Chinese products now have a competitive advantage in the global market with higher value and higher quality. Similarly, the significance of education in boosting export trade is notable. Yang (2012) points out that China’s focus on higher education and vocational training has equipped its workforce with the necessary skills for export-oriented industries, facilitating the production of more sophisticated, high-value products. Chen et al. (2022) further discuss how the synergy between technological advancement and educational development contributes to a more dynamic and diversified export sector. This interplay is vital for China’s ability to adapt to global economic changes and more effectively participate in international trade. In conclusion, the increase in exports due to heightened knowledge capital and education levels signifies China’s strategic transition towards a knowledge-based economy. This shift is reshaping the structure of its domestic industries and redefining China’s role and competitiveness in the global market.

Robustness test

In this study, meticulous measures were taken to guarantee both the accuracy and reliability of the results, especially those obtained from the analysis using the province and year-fixed effect models. To ensure the dependability of our findings, an extensive robustness check was conducted on the outcomes of the province and year-fixed effect model. This involved the use of two econometric techniques: fully modified ordinary least squares and dynamic ordinary least squares. The implementation of fully modified ordinary least squares and dynamic ordinary least squares was critical in substantiating the integrity of the inferences drawn from the province and year-fixed effect model. The employment of these methods not only bolsters the solidity of our results but also reflects a commitment to the best standards of empirical rigor and methodological thoroughness. This approach to data verification underlies the credibility and trustworthiness of our conclusions. The specifics of these findings are systematically outlined in Table 3.

Table 3 Results of robustness test.
Full size table

Table 3 presents a detailed evaluation of the estimated parameters, focusing on both their magnitude and statistical significance. Remarkably, the findings obtained through the application of fully modified ordinary least squares and dynamic ordinary least squares align closely with those from the initial province and year-fixed effect model. This alignment between fully modified ordinary least squares and dynamic ordinary least squares, in comparison to the province and year-fixed effect model, robustly confirms the accuracy of the original model. The consistency observed across these varied econometric methods not only strengthens the trustworthiness of the province and year-fixed effect model but also substantiates the reliability of the study’s overall findings. The convergence of results across these methodologies indicates that the initial province and year-fixed effect model was meticulously crafted and successfully captured the essential dynamics of the variables under examination. The adoption of this comprehensive cross-validation process, which incorporates multiple analytical techniques, reinforces the solidity and validity of the research’s conclusions. This multi-faceted approach to analysis assures a high level of confidence in the integrity and reliability of the study’s results.

Regional heterogeneity analysis

Spanning a considerable geographical expanse, China is officially categorized into three distinct regional demarcations: eastern, central, and western. The eastern precinct is widely acknowledged as the epitome of China’s developmental zenith, encapsulating its most economically advanced locales. Conversely, the central sector is recognized for its intermediary developmental status, while the western swathes are often delineated by developmental lacunae. These territories, though unified under a single nationhood, manifest disparate attributes ranging from their economic growth trajectories, state-directed policy nuances, and infrastructural development gradients to their inherent geographical peculiarities. To delve into the multifaceted influence of foreign direct investment and e-commerce on regional export dynamics, our empirical approach disaggregated the core dataset, structuring it into three region-specific sub-samples. This strategic bifurcation aimed at discerning the variable intensities of foreign direct investment and e-commerce influences across these heterogeneous regions. The analytical outcomes derived from this region-centric examination are detailed in Table 4.

Table 4 Results of regional heterogeneity analysis.
Full size table

Reflected in Table 4, the repercussions of foreign direct investment on export trade reveal intricate regional gradations within China’s geographical tapestry. Concretely, a marginal ascent of 1% in foreign direct investment is associated with a 0.278% enhancement in the export dynamism of the eastern provinces. This increment tapers to 0.179% for central provinces and further diminishes to 0.161% for their western counterparts. The scholarly discourses of Contractor et al. (2020), Dang and Zhao (2020), and Batschauer da Cruz et al. (2022) elucidated that the synergy between foreign direct investment and export growth hinged upon a triad of factors: intrinsic firm capabilities, locational attributes, and the operational modus operandi. The eastern provinces, historically recognized as China’s economic epicenter, are imbued with a robust infrastructural matrix, streamlined trade corridors, and a business milieu that gravitates towards global market integration. These intrinsic locational advantages, complemented by the spatial competition theory proposed by Proost and Thisse (2019), Redding and Rossi-Hansberg (2017), and Goerzen et al. (2013), amplify the efficacy of foreign direct investment in spurring export trade. On the contrary, the central belt, despite its ascending economic trajectory, is intermittently stymied by transitional economic impediments, occasionally attenuating the foreign direct investment-export nexus. The western provinces, albeit burgeoning, still navigate developmental constraints, resonating with Wang and Zhao (2015) and Jiang et al. (2016)‘s backwash effects, wherein peripheral regions grapple to harness the complete spectrum of foreign direct investment benefits. From a sustainability lens, echoing the tenets of Milne and Gray’s (2013) Triple Bottom Line framework, the magnitude and mode of foreign direct investment’s assimilation should be judiciously balanced to ensure economic, social, and environmental equanimity. The immediate economic impetus observed, particularly in the eastern provinces, warrants an integrated approach wherein foreign direct investment infusion aligns with sustainable practices, ensuring that regional development dovetails with ecological stewardship and socio-cultural inclusivity. Such a harmonized trajectory ensures that the fruits of foreign direct investment are not ephemeral but perennial, fostering a resilient and sustainable export landscape across all regions.

Referring to the results presented in Table 4, an augmentation of 1% in e-commerce transaction volume is observed to lead to a differentiated impact on the export trade across China’s tripartite regional structure: specifically, a surge of 0.397% in the eastern provinces, an enhancement of 0.365% in the central provinces, and a growth of 0.325% in the western provinces. This regional heterogeneity in the influence of e-commerce on export trade can be supported by a confluence of academic perspectives and established theoretical underpinnings. Drawing insights from Porter, Michael’s (2011) Competitive advantage theory, the eastern provinces, having established themselves as economic powerhouses, have already harnessed advanced infrastructural frameworks and digital ecosystems. This enables them to efficiently leverage the capabilities of e-commerce, thereby reflecting a more pronounced augmentation in their export trade. The central provinces, as highlighted by North and Douglass’s (1989) theory of institutional change, are navigating through evolving institutional landscapes, mediating between traditional trade mechanisms and burgeoning digital frontiers. While they have made significant strides, the transformational gaps that exist temper the full realization of e-commerce benefits in the domain of exports. The western provinces, on the other hand, are still grappling with foundational challenges. Drawing from Sachs and Warner’s (2001) resource curse hypothesis, these provinces, abundant in natural resources, might have historically focused more on primary sectors, leading to a lag in the adoption and integration of e-commerce into their economic tapestry. This could partially elucidate the relatively muted growth in export trade from e-commerce advancements. Incorporating the sustainability ethos, as expounded in the triple bottom line approach by Elkington (1998), the expansion of e-commerce should not merely serve economic objectives. It should be orchestrated in a manner that respects ecological boundaries and promotes social inclusivity. Especially in regions like the western provinces, where development is paramount, it is critical to ensure that the surge in e-commerce-driven exports is not at the expense of environmental degradation or social disparities, thereby upholding a balanced, sustainable developmental trajectory.

Conclusions

Amidst the fast-paced evolution of the global economy, key factors such as foreign direct investment and e-commerce have become instrumental in reshaping China’s export sector between 2005 and 2022. Our analytical models, which utilize a combination of province- and year-fixed effects analysis along with fully modified ordinary least squares and dynamic ordinary least-squares methodologies, shed light on how foreign direct investment and e-commerce synergistically enhance China’s export capabilities. Significantly, this expansion in China’s exports aligns with the global agenda for sustainable development. It’s encouraging to see that China’s growth in exports extends beyond the realms of international investment and digital marketplaces, intertwining sustainable practices like optimizing local labor resources and prioritizing technological advancements. These approaches contribute to a more sustainable export environment. Our findings further reveal that variables such as knowledge capital and educational levels positively influence China’s export figures. Additionally, our analysis of regional disparities provides a deeper understanding. The eastern regions of China show greater responsiveness to foreign direct investment and e-commerce in driving export trade, whereas the western regions respond more modestly. This variation highlights the need for tailored policies and sustainability strategies to ensure a fair distribution of the benefits from foreign direct investment and e-commerce across all regions. In conclusion, while foreign direct investment and e-commerce are key drivers of China’s export growth, the broader story is one of sustainable development. Technological innovation and human capital development are pivotal to China’s continued success in exports. Moving forward, it is essential for policymakers to maintain a careful equilibrium between these economic drivers and sustainable development goals, fostering a balance between economic growth and environmental sustainability.

Drawing upon the insights derived from this study, we elucidate several policy recommendations along with practical solutions. First, for policymakers and business leaders, investing in technology and education is identified as a crucial strategy. The significant impact of technological innovation and a well-educated workforce on export growth underscores the necessity for ongoing investment in these domains. For academia, this opens avenues for further research into specific types of educational programs and technological innovations that most effectively enhance export capabilities. Businesses, especially in the export sector, should prioritize employee training and the adoption of cutting-edge technologies to maintain competitiveness. Second, considering the varied responsiveness to foreign direct investment and e-commerce between China’s eastern and western regions, it’s imperative for regional authorities and business managers to customize their policies and strategies to suit the unique needs and strengths of their regions. This could involve targeted investments in infrastructure and digital capabilities in the eastern regions while simultaneously focusing on cultivating other competitive advantages in the western regions. Academia can play a role by conducting region-specific research to identify the most effective strategies for each area. Third, the intersection of export growth with sustainable development goals necessitates a comprehensive approach to policymaking. Managers in the export sector are encouraged to integrate sustainable practices into their business models, such as the utilization of environmentally friendly technologies and adherence to fair labor practices. This area also presents an opportunity for academic research into the effective implementation of sustainable practices in the export sector, aiming to balance profitability and competitiveness. Finally, our findings suggest that although foreign direct investment and e-commerce are significant drivers of export growth, their benefits are not uniformly experienced across all regions. This indicates the need for balanced development strategies that ensure equitable benefits from foreign direct investment and e-commerce across various regions. Strategies might include enhancing e-commerce infrastructure in less-developed areas or offering incentives for foreign investment in regions currently less engaged with these investments. For academics, this highlights the necessity of researching ways to optimize the impact of foreign direct investment and e-commerce across diverse regions, promoting equitable economic growth. These policy implications offer a strategic roadmap for leveraging key drivers of export growth in China, highlighting the importance of regional customization, sustainable development, and balanced economic strategies.

In the course of this research, certain limitations emerged that warrant acknowledgment. Firstly, the study’s timeframe, spanning from 2005 to 2022, may not fully capture the evolving dynamics of foreign direct investment and e-commerce in the context of China’s longer-term economic history. A more expansive temporal analysis could provide deeper historical insights. Secondly, while the fully modified ordinary least squares and dynamic ordinary least-squares methodologies and fixed effect models offer robustness, they may not encompass all the nuanced intricacies of the interactions between the chosen variables. Future research could employ mixed-method approaches, blending quantitative and qualitative inquiries to attain a richer understanding. Thirdly, our focus on regional heterogeneity, while pivotal, may overlook intra-regional variances that can significantly influence export trends. Subsequent studies might delve deeper into micro-level analyses, probing district- or city-level data. Fourthly, the emphasis on sustainability, though aligned with global imperatives, is predominantly viewed through the lenses of labor and technology. Incorporating other sustainability metrics, such as environmental or social indicators, could render a holistic view. Lastly, the external validity of our findings, primarily centered on China, might be limited in their generalizability to other nations. Comparative studies juxtaposing China’s experiences with those of other global players could bridge this gap. Addressing these limitations would not only refine the existing body of knowledge but also ensure a more comprehensive alignment of economic strategies with sustainable development goals.